(PDF) Theory and practice in Late Antiquity archeology - DOCUMENT.TIPS (2023)

THEORY AND PRACTICE IN LATE ANCIENT ARCHEOLOGY

ARCHEOLOGICAL SENANTIC

Editorial Board

MANAGING EDITOR

LUKE LAVAN

EDITORS

Averil CameronJames CronSimon Ellis

Jean-Pierre SodiniBryan Ward-Perkins

BIND 1 (2003)

THEORY AND PRACTICE IN LATE ANTIQUITY

ARCHEOLOGY

EDITED BY

LUKE LAVAN and WILLIAM BOWDEN

OF GLASSES • BOSTON

2003

This book is printed on acid-free paper.

The German Library - CIP unit recording

Theory and practice in late antique archeology / edited by Luke Lavan-Leiden ; Boston: Brill, 2003

(Late Antiquity Archaeology; Volume 1)

ISBN 90-04-12567-1

Library of Congress Cataloging-in-Publication Data

Library of Congress Cataloging-in-Publication Data is also available

ISSN 1570-6893 ISBN 90 04 12567 1

© Copyright 2003 from Koninklijke Brill NV, Leiden, Holland

All rights reserved. No part of this publication may be reproduced, translated, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior written permission of

permission from the publisher.

Permission to photocopy items for internal or personal use is granted by Brill provided that

the applicable fees are payable directly to The CopyrightClearance Center, 222 Rosewood Drive, Suite 910

Danvers MA 01923, USA. Fees are subject to change.

printed in the Netherlands

content v

CONTENTS

Late Antiquity Archaeology: An Introduction .................................... vii

Luke Lavan

Part one: Ideologies and agendas

Ideologies and Agendas in Late Antique Studies ........................... 3Averil Cameron

Second part: The social life

Archeology and late antique social structures .......................... 25Jean-Pierre Sodini

The Construction of Identities in Post-Roman Albania ................ 57William Bowden

Some aspects of the transformation of the Roman domus between Late Antiquity and the Early Middle Ages ...................... 79

Barbara Polci

Third part: The economy

Late Antiquities Trade: Research Methods and Field Practice 113Sean A. Kingsley

Coinage and the Late Roman Economy ......................................... 139 Richard Reece

Fourth part: Topography

Late Antiquity Urban Topography: From Architecture to Human Space ........................................ .. ................................................... ............ 171 Luke Lavan

The Urban Ideal and Town Planning in Byzantine New Towns in the Sixth Century AD. ................................... ............... ....... 196 Enrico Zanini

Christian Topography in the Late Antique City: Recent Findings and Open Questions ................................... .. .............. 224Gisella Cantino Wataghin

Topographies of production in North African cities during the Vandal and Byzantine periods ................................ ...... .. 257Anna Leone

content vi

Archaeological Survey and Visualization: The View from Byzantium ........................................... ........................................ 288Richard Bayliss

The political topography of the late antique city: activity spaces in practice ................................... ..... .............................. 314Lukas Lavan

Fifth part: Change

Attitudes to Spolia in Some Late Antique Texts .............................. 341Robert Coates-Stephens

The Value of Re-Employment: The Case of Ephesus ......................................... 359 Ida Leggio

Part Six: Decline and Fall? Studying long-term change

Studying Long-Term Change in the West, AD 400-800................ 385 Chris Wickham

Fall and fall? Studying Long-Term Change in the East .......... 404Mark Whittow

Index ................................................ .. ............................................ 425

content vii

LATE ANCIENT ARCHEOLOGY: AN INTRODUCTION

LUKE LAVAN

The justification for a specialism

Late Antiquity archeology has been slow to emerge. Although Late Antiquity has been a legitimate period of study for the past thirty years, archaeologists have generally preferred to remain as Roman, Early Medieval, Byzantine or 'Christian'. This partly reflects the long-term perspectives common in archaeology. It also reflects the fact that Late Antiquity was a period of profound change that led some to look in one direction, others in another. That said, there are strong reasons to distinguish between a late antique archaeology. The most important of these is the now wide acceptance that in territories that later constituted the Middle Byzantine Empire, the 7th c. then breakdown in economic life, urban development and rural development. 1

According to the same criteria, the Mediterranean was in the 3rd-6th century. AD looks very Roman; in contrast, the wretched world of Middle Byzantium offers only forms of continuity, not continuity of scale. It is the overall scale of society that seems to many archaeologists to be the most straightforward way of organizing history. In these terms Late Antiquity, both in East and West, essentially belongs to the ancient world; only in Egypt and the Levant is it difficult to identify sweeping changes in scale in the later 7th century. Late Antiquity archeology sits most comfortably as a sub-period within Roman archaeology, not as part of Western or Byzantine medieval studies.2

Nevertheless, there are some arguments for distinguishing Late Antiquity from Roman archeology as a whole. This cannot be done in connection with rural settlement, production, trade or technology, except

1 For a recent view of the 'rupture' within later Byzantine territories, see M.Whittow, The Making of Orthodox Byzantium (606-1025) (London 1996) 53-68, 89-95.

2 This is not to deny that ruptures occurred in different regions at different times during the 5th to 7th centuries. but it is intended to describe broad patterns of change in the period as they relate to the Mediterranean.

introduktionviii

in outlying areas where the change comes early. However, it seems possible in terms of urban, religious or military archaeology. Here, long-term changes in Roman society and external pressures create a recognizably distinct Late Antiquity situation. For urban history this means dealing not only with new political and religious structures, but also with new problems in the types of evidence we have, with the reuse of old architectural structures, a decline in unclear ideological planning and the use of less well-defined buildings. ing types. Some of this ground has of course been covered by Christian archaeology, but the specificity of this sub-discipline does not satisfy archaeologists who wish to study society holistically in terms of broad structures; tracing specific cultural traits across radically different periods has little appeal. There therefore seems to be some justification for an archeology from Late Antiquity. The idea of ​​this series of books is essentially to support this development by bringing together emerging and established scholars to address general themes such as rural settlement, social structure and economics. As such, it does not intend to exclude historians: indeed, historians often show themselves. better at synthesizing archaeological information than their archaeological counterparts.

Archaeological theory in late antique archaeology

This volume is based on two conferences held in 2001 under the banner of "Late Antique Archaeology", at the Ashmolean Museum, Oxford and the Swedish Institute in Rome: survey of "New Research, Method and Practice" and "Topographical Studies in Late" Antiquity respectively ”. Both aimed to form a strong foundation for the series by examining a range of topics from theoretical and methodological points of view. This was something of a gamble Late antique archeology tends to be extremely empirical; we didn't even dare put the word "theory" in the title of the conference. The resulting volume will not be readily recognizable to anyone familiar with theoretical archeology as part of that genre. This is the result of a conscious choice. Theoretical archeology is a highly developed field of thought, especially in Scandinavia, Great Britain and the United States. Nevertheless, it is widely recognized that there is something of a gulf between those who can read in theory and the rest of the archaeological community. This is less strong in prehistory, but too much

introduction ix

of historical archaeology, especially classical archaeology, it is striking; one often gets the impression that many classical archaeologists have never heard of theoretical archaeology. There has been a heroic effort by the Theoretical Roman Archeology Conference, based in Great Britain, to bridge this gap through a series of annual conferences. However, this work inevitably tends to reflect the predominantly British interests of the scholars active in the group and their attachment to theoretical ideas developed by prehistorians. Some subjects deal with the Mediterranean, but these tend to relate to the early Roman period. Those written about Late Antiquity are unfortunately not widely read in the mainstream of empirical science in the Mediterranean.3

The lack of penetration of archaeological theory into mainstream Roman and late antique archeology is intellectually deplorable. A healthy awareness of the shape of our ideas and the problems with our evidence cannot but bring benefits. Nevertheless, it is fair to say that the widespread sense of caution in applying theoretical archeology to Mediterranean historical archeology is probably justified. It is very difficult to translate the tools of interpretation that were primarily developed in prehistory into a cultural zone that is rich in texts. Here, interpretation is largely limited by information from written sources, and many approaches widely accepted elsewhere in archeology may seem inappropriate in their terminology and assumptions. Before the wider scholarly community that is LateAntique Studies, one cannot speak of 'the elite' without speaking of curiales and senators, one cannot speak of 'social life' without slaves or humiliores, or cannot speak of economy without theannona, or identity without reference to Romanitas or Christianity. Work that doesn't do this just isn't taken seriously.

Thus, if archaeological theory is to make a strong contribution to Late Antiquity archaeology, it will be necessary to find out the utility of different approaches to very specific problems – to try out new ideas and intellectual frameworks while fully engaging with rich opportunities. Mediterranean evidence, both archaeological and textual. The rationale behind this volume was to encourage work of this nature by examining theory and methodology regarding specific topics currently being addressed by researchers. Con-

3 For a review of the volumes from this series see R. Laurence, "TheoreticalRoman Archaeology", in Britannia 30 (1999) 387-90.

introductionx

as a result, these articles reflect the current theoretical development of the discipline rather than how anyone might wish it: authors consider problems using concepts familiar to their late antique colleagues, rather than addressing a wider theoretical audience in archaeology. All these papers were produced by invitation, many for a 'brief'; Contributors were often asked to write in ways they had not done before. The result of this has sometimes been closer to theory, sometimes closer to pure practice. Although this volume contains articles that do not always resemble 'theory', they represent the thinking of a group of scholars who are not only open to reflection but also fully concerned with the problems and possibilities of evidence.

Traditions of scholarship in late antique archaeology

The integration of theoretical archeology into Mediterranean research is actually a relatively unimportant topic in late antique archaeology. Much more serious fundamental disagreement exists between scholars about basic aspects of methodology. These do not deserve to be omitted from this volume, lest it create a wholly unrepresentative idea of ​​the intellectual climate of the subject. The archaeological community is deeply divided in its approaches and interests; so deep that various factions neither respect nor learn from the research done by their colleagues working in other intellectual traditions. Originally we hoped to have an essay for this volume on "Ideologies and Agendas in Late Antiquity Archaeology", but this proved too large a task to be carried out rigorously in the time available. However, to outline the nature of the problem, I provide here a brief sketch of some of the most important divisions between various researchers and some reflections on their implications. This should give readers an idea of ​​the nature of the theoretical and methodological difficulties that Late Antiquity archeology currently faces. My description, intended as a struggle between two opposing traditions, is simplistic and somewhat stereotypical, but somehow comes close to reality.

The first tradition is centered on Paris, although it has adherents at Oxford and in classics and religion departments around Europe and America. It is probably best labeled as 'Continental'. Scholars in this tradition tend to have had an education in ancient languages

introduction xi

or art history. They are used to the idea that the role of archeology is rather small: the intellectual landscape from which they are separated is dominated by literary studies. Philology is king and serious history is made from a detailed knowledge of Greek, Syriac and Armenian, not potsherds. In some quarters there seems to be an ingrained belief that the proper study of "Byzantium" is limited to literary production and art. The view of many Anglo-American scholars that textual history is a collection of half-truths waiting to be broken down by the "wrecking-ball" of archeology has no place here. Archeology is usually treated as an extension of art history and architecture. Fieldwork reflects this, being mainly aimed at the recovery of monumental architecture and, where found, art. Excavation is preferred, although not universal. The choice of sites to be examined is most revealing: they are monumental urban centers or rural monasteries and villas, not the dwellings of the poor.

The work in this first tradition is above all extremely empirical, not prone to far-reaching intellectual speculation: objects and architecture are studied, not artefacts in use or human spaces. This keeps research primarily focused on 'elite' concerns which have produced the most visible record; it can also mean that we concentrate archeology on what the literary sources will tell us about: political and religious life. The broad question-based approach of the Annales school, which has inspired so many northern European archaeologists, does not really interest this group. Little time is also given to theoretical considerations, such as philosophers and critical theorists working in Paris itself. Most researchers still do 'scientific' work and see their data as straightforward, not complex. To do a good job, gather as much information as possible in a catalog with attached literary sources. This can result in some very solid research and produce useful reference works. However, buildings and objects often seem to be studied for their own sake, as 'architecture' and 'art', rather than as 'artifact', that is, a source of historical information about the people who used them. The method is also very conservative. Established approaches are generally respected, and the patronage of a past master may be reflected in a grateful desire to continue the practices he laid down. Indeed, this tradition as a whole tends to attract people with more conservative instincts, in politics and in other areas of life. Again, it must be said that this first 'continental' tra-

introductionxii

dition does not dominate all work done in Paris, and certainly not all that is done in France. It is found elsewhere on the continent of Europe, in some German and Italian work, and in Greece, and one or two places in Great Britain. Archaeological work carried out in collaboration with Dumbarton Oaks also broadly fits into this group.

Another great tradition is represented by archaeologists mainly from Great Britain and America and their collaborators in Spain, Italy and France. It can be quite uncomfortably labeled 'Atlantic'. It is part of an archaeological tradition that has developed primarily in Northern Europe, particularly in Britain and Scandinavia, and in North America, which places great faith in modern fieldwork and has derived much of its research agenda from anthropology. This tradition is also now particularly strong among early medieval archaeologists in France and Italy, especially those based in archaeological units. Scholars in this second tradition are known for the high quality of their fieldwork, whether excavation or survey. They are also known to publish their earlier caves more fully with serious artifact studies. This tradition has produced enormous amounts of new information for the study of Late Antiquity. Indeed, the location of high-quality excavations can have the curious effect of reversing the geography of the Roman and early medieval world in the minds of Atlantic scholars. For many archaeologists in the second tradition, serious study of the Roman Empire, from well-excavated sites, is limited to Great Britain, Germany, the Netherlands, parts of France and northern Italy. The Mediterranean becomes a 'backwater' in a bizarre way.

Even Turkey is considered something of an archaeological desert, as almost no farms or villages have been excavated there, and urban excavations have been of poor quality. Richard Reece has recently discussed this issue in relation to coin reports - while perhaps two hundred significant sites in Britain have full coin lists, there are only about 20 sites in the Mediterranean with lists of over 300 coins, of which only two sites so far include details of context .4 The intellectual effects of geographical biases in the quality of evidence can be strong; Hodges and Whitehouse's claim that early medieval trade in the North Sea overpowered that in the Mediterranean is perhaps the best example; other researchers have suggested that their

4 This remark was made in the spoken version of the paper presented in this volume, at Oxford in March 2001.

introduction xiii

conclusions simply reflect the underdevelopment of Mediterranean archeology in the same period.5

Another characteristic of the second tradition is that it is generally more question-based and feels that empirical research should be subordinated to the study of issues; data should not dominate. Having strong concepts and methods is more important than having read all previous bibliography. Scholars in this tradition are also often familiar with mainstream theoretical archaeology, although this is far from universal. Many of them share a broad interest in studying the overall structure of society from below. These concerns may attract scholars with leftist sympathies, even if scholars with other political views share the same interests. Nevertheless, research that concentrates on elites (emperors, palaces, icons and silver) is sometimes perceived as a bit morally suspect. British archaeologists and field archaeologists working in other countries often show great skepticism about the use of textual sources. These archaeologists have often had a purely archaeological training rather than a training in ancient languages. They are also iconoclastic rather than conservative. They are prepared and indeed expected to attack the works of the previous generation, even if it is based on very high quality scholarship. They are particularly happy to disprove, through archaeological evidence, the textual claims of Moses Finley or A. H. M. Jones about late-century economy, urbanism, or rurality. These subjects represent the areas of ancient life in which such scholars are most interested. However, they also want to start rewriting social, political and religious history based on archaeological evidence. At best, this looks like a much more difficult direction to pursue. In their eyes, 'good' historians are those, like Chris Wickham and Mark Whit-tow, who are prepared to accept defeat by seriously incorporating archeology into their source base.

Tensions between the two traditions

How do these two traditions interact? In some areas they more or less coexist: Christian archeology contains elements of both. Some Mediterranean archaeologists in the Atlantic tradition are trying.

5 R. Hodges and D. Whitehouse Mohammed, Charlemagne and the Origins of Europe: Archeology and the Pirenne Thesis (London 1983). S. J. B. Barnish "The Transformation of Classical Cities and the Pirenne Debate" in JRA 2 (1987) 385–402.

introduktionxiv

seriously considering texts. However, there is also a large area where the two traditions simply do not interact; yes, they seem to be in competition without any significant dialogue. It's not because there aren't strong opinions. Scholars in the continental tradition can be very free with their criticism in book reviews, as in the "Bulletin Critique" of Antiquité Tardive. Scholars in the Atlantean tradition tend to keep their views private or reserve them for specialized publications that will never enter the libraries of the former. The main ruptures on the fault line between the two groups are as follows: first, and most seriously, work done by Atlantic archaeologists is often ignored by their continental counterparts, and textual work by the latter is not fully taken into account by the former: this is partly because the bibliography is not coincidental: the Continentals want to study architecture, religion and politics, the Atlantic economy, technology and rural development. Where it coincides, the continental attitude may be that the archaeological work of the Atlantic is impressive but perhaps unscholarly in its bibliography and use (or non-use) of literary sources: it is just 'bad'. Much traditional continental work, on the other side of the Atlantic, is perceived as 'boring' and without unhistorical significance, because it does not have a strong enough conceptual framework. Atlantics like to laugh privately at the stylistic dates offered for buildings by architectural historians in the continental tradition, or at the claims of their philological colleagues trying to write economic history. Continental scholars, meanwhile, are supremely confident in their mastery of the literary sources; nor are they impressed with the intensity that Atlantean archaeologists pour into individual sites to uncover facts they consider relatively trivial and unrelated to what they believe are areas of human cultural achievement: "I don't care with whether people eat shish kebab at Amorium”, as a professor put it to me recently.

Change?

Much of the last twenty years has seen the gradual erosion of the continental tradition in favor of the Atlantic. The biggest roads have been done in Spain, Italy and French field archaeology. For iconoclastic Atlantean archaeologists, of course, this development is inevitable: it is written in a Darwinian script whereby the new vital organism of rigorous Late Antiquity field archeology must

introduction xv

victory over the decaying mass of architecture, Christian archeology and literary scholarship. However, this vision of the future is more than a little naive. Indeed, both groups are currently experiencing periods of vitality, and it is far from clear who will make the greatest contribution to the subject over the next few decades.

Above all, there is no sign of revolutionary change in Paris. Philology and textual history remain firmly enthroned, with architecture and 'objects as art' present. This is largely because these topics remain popular in their traditional format. Most importantly, the vast amount of excavations carried out over the last twenty years in Israel, and to a lesser extent in Jordan, have largely operated within the first tradition of searching for architecture through clearance excavation. By contrast, the methodical puritans of northern Europe and their allies in urban archeology elsewhere have seen their budgets slashed and now undertake far fewer research excavations than back in the 1970s and 1980s. In the most dynamic region for fieldwork, the Near East, the Atlantic now faces the violent spectacle of clearance archeology triumphing over stratigraphic excavation, rendering irrelevant a hundred years of methodological refinement in their own work. From the pages of recent preliminary reports on excavations in Israel and Jordan, it sometimes seems that the future of Late Antiquity archeology is in the clearing excavation of mechanical graves rather than in simple context recording. Such encounters are not limited to excavation: at a recent conference on Roman mosaics it was noted that several contributions by researchers from the continental tradition made absolutely no use of the typological classifications developed by other researchers, in this case mainly French since the 1970s . There are likewise examples of serious textual work by scholars working in the Mediterranean being disregarded by fierce opponents of literary sources in Britain.

Conclusion

These somewhat stereotypical comments give an idea of ​​the context in which the articles presented in this volume are being written. It is an image that different intellectual traditions do not learn from each other and especially are not prepared to conduct a serious dialogue about theory and method. I don't imagine this book will really provoke this; it is more likely to be read by one group

introductionxvi

and ignored by the other. It is possible that only open polemics in major journals can build bridges. However, the beginnings of the debate may be developing. We are getting closer to each other, partly as a result of the growth of the EU. Many of the younger scholars in this volume have not only spent time at foreign universities, but have taken entire degrees abroad. Some of them have been so deeply affected that they can only write in the language in which their university course was conducted, even though they have chosen to return home and start working there. In such a situation, it seems that the time is ripe to start a single methodological debate about our differences, to be somewhat less tolerant than we have been, to say what we really mean and to challenge each other overexcavation methodology, historical theory, the legitimate use of texts and bibliographical rigour. Although this will be a difficult debate, it is certainly in everyone's long-term interest to build a late antique archeology that is as strong as it can be, both in theory and practice.

Recognitions

I would like to thank all the contributors to both this volume and the two conferences, as well as all those who helped organize and support both events. Bryan Ward-Perkins was co-organizer at Oxford and consulting editor for this volume. Barbara Polci acted as partner for the session in Rome. The Craven Fund Committee, the Oxford Byzantine Studies Committee and the British Academy made generous grants towards the costs of the Oxford meeting, which was also supported by the Association for Late Antiquity. The Swedish Institute hosted our meeting in Rome, where Arja Karivierian and Olof Brandt were particularly helpful. Olof's evening tour in the excavations under St. Lorenzo in Lucina was particularly memorable. The British School in Rome facilitated an important presentation by James Crow and Richard Bayliss through the loan of a data projector. Teresa Shawcross, Susanne Bangert, Daniel Farrell and Anna Leone also provided practical assistance during the conferences. Carlos Machado worked hard to compile an index. I am also grateful to the Leverhulme Trust for funding my academic year 2001-2002, which made the editing of these papers possible. Finally, I would especially like to thank Julian Deahl for agreeing to publish these papers and to him, Marcella Mulder and others at Brill for their work on this manuscript.

ideologies and agendas in late antiquity studies 1

SHARE ONEIDEOLOGIES AND THE AGENDA

this page is intentionally blank

ideologies and agendas in late antiquity studies 3

IDEOLOGIES AND THE AGENDA IN LATE ANTIQUE STUDIES

AVERIL CAMERON

Abstract

This paper sets a framework by discussing the trends and approaches that can be observed in the study of Late Antiquity during the last few decades. It takes the points made in a recent article by A. Giardina and considers the models of continuity and change adopted in several recent collective publications. It questions whether the current enthusiasm for the 'Long Late Antiquity' and the privileging of cultural over social and economic history are likely to continue in their current form. It draws attention to differences in emphasis between historians and archaeologists and between analyzes of the eastern and western parts of the empire, and emphasizes the complementarity between historical and archaeological approaches.

The purpose of this article is to set a framework for the essays that follow, which address the current work of late anti-archaeologists. It considers the discipline of Late Antiquity studies as a whole, rather than specifically the discipline of Late Antiquity archaeology, and therefore poses a question for archaeologists themselves to answer as to whether, or to what extent, their own discipline follows the same overall trends. The title itself was suggested by the organizer of the conference where this paper was originally presented, and it immediately raises the question of how an ideology can differ from an agenda. To this I am inclined to reply that, in relation to current and recent scholarship on Late Antiquity, there are plenty of agendas, which I will set out below. By this I mean programs or characteristics – topical issues and questions where we see a striking degree of coherence between different researchers. How far these agendas rest on serious ideologies is perhaps a further question. It has recently been suggested that there is indeed a recognizable type of current writing on the history of Late Antiquity which rests on a number of identifiable ideological assumptions, although these are not

L. Lavan og W. Bowden (red.) Theory and Practice in Late Antique Archaeology (Late Antique Archaeology 1 2003) (Leiden), s. 3-21

averil cameron4

expressly stated by the author.1 I have dealt with this claim elsewhere2 and return to it below. But the first part of what I have to say is more about agendas.

One cannot help but be struck by the number of recent publications which, in different ways, set agendas for Late Antiquity studies. EuropeanScience Foundation project of the 1990s on The Transformation of the Roman World, or volumes XIII and XIV of the new edition of Cambridge Ancient History (the latter are essentially additions to the original publication, which ended with volume XII and did not even lead the reader to that the end of Constantine's reign).3 Peter Brown's work has created its own industry of discussion and commemoration.4 But we also have such reflective collections as Shifting Frontiers in Late Antiquity, together with a volume edited by Stephen Mitchell and Geoffrey Greatrex on the problem of ethnicity.5

There are also numerous publications on late antique urbanism, while several recent sourcebooks for student use also impose agendas through their selection and presentation of material.6 So in another way, it could be argued, the Translated Texts for Historians series published by Liverpool University Press with a ever-growing number of volumes from late antiquity as well as the early medieval period. Late Antiquity is not only firmly established; it generates its own methodological discourse and it is very self-aware.

1 See Giardina (1999); Liebeschuetz (2001a).2 Cameron, Av. (2002, with longer discussion of some of the points here;

but on the notion of 'decline' see the debate in Lavan (2001).3 Bowersock, Brown and Grabar (eds.) (1999); CAH XIII and XIV: Cameron

and Garnsey (eds.) (1998); Cameron, Ward-Perkins and Whitby (eds.) (2000). Discussion of vol XIII with reflections on the changing perceptions of the period: Marcone (2000). The impact of the Journal of Roman Archeology and the many subsidiary volumes already published has been great, although not directly considered here.

4 Especially the debate, "The world of late antiquity revisited", SymbOslo 72 (1997)5-90; the collection of articles in Journal of Early Christian Studies 6 (1998) and Howard-Johnston and Heywood (eds.) (1999). See now Brown (2002) especially74-112.

5 Mathisen and Sivan (eds.) (1996); Mitchell and Greatrex (eds.) (2000); See also Miles (1999) on identity, another popular topic at the moment.

6 Maas (2000) and Lee (2000), both from the same publisher, Routledge (publishers also have their agendas or are quick to spot others' agendas); see also Valantasis (2000).

ideologies and agendas in late antiquity studies 5

It may be convenient to start with a look at the Harvard Guide, edited by Bowersock, Brown, and Grabar, and published in 1999. It has eleven introductory chapters and then nearly five hundred pages of entries on late antique subjects and individuals. It has pictures, but perhaps not very many in relation to the text or in relation to the importance now placed on visual evidence. Its subtitle is "A Guide to the Postclassical World". As far as I am aware as a contributor, the titles of the introductory essays were planned by the editors and, as is appropriate in a volume where one of the editors is Oleg Grabar, they include a chapter on Islam. Here is the first agenda, or rather a pair of agendas: we have here the model of the "long late antiquity", according to which continuity is assumed up to and into the Abbasid period, and Islam is considered to belong to the late world. Antiquity, not to mark a decisive break. This scenario is also evident in Garth Fowden's book Empire to Commonwealth, in Glen Bowersock's Hellenism in Late Antiquity and in Peter Brown's The Rise of Western Christendom, although not entirely in the latter's The World of Late Antiquity, published in 1971.7 In the workshops, organized by the Late Antiquity and Early Islam project, the tendency has likewise been to minimize the degree of disruption caused by the Arab invasions, and indeed to minimize the number and organization of the invaders.8 In response to these tendencies, however, a sense of discomfort is detected, and it is remarkable how often researchers now choose to discuss the question of long-term change as a methodological issue. Chris Wickham has warned that the continuity model overlooks the profound impact caused by the disintegration of the financial system of the late Roman state, and Wolfgang Liebeschuetz has recently argued strongly for a recurrent torupup and decline coming earlier in the West than in the East , but it is clear. also in the East before the end of the 6th c.9. Nevertheless, he does so against a strikingly widespread model of continuity. But the chronological scope of the guide is not its only striking feature: further expansion of the scholarly field can be seen in the geographical and cultural realm. Almost all the essays in the HarvardGuide, with the exception of those deliberately limited by their subject matter, span the full panorama of Christianity, pagan-

7 Fowden (1993); Bowersock (1990); Brown (1996), jf. Brown (1971).8 Se essayene i Cameron (1995).9 Liebeschuetz (2001b).

averil cameron6

ism, Judaism and Islam, moving geographically from the barbarian West to the Eastern Empire and its borders with Iran. Religious and cultural diversity on the one hand, and the Eastern Mediterranean world on the other, can be said to have come into their own as central subjects of historical attention.

Another index to current agendas in late antique studies consists in the titles of books now being published. To judge from current book catalogues, gesture, behavior and emotion are becoming the new fashionable subjects that follow in the wake of gender and masculinity (or more often 'masculinities'). Also in the Harvard Guide, time, memory and identity are themes as important as invasion and political narrative. In contrast, despite a noticeable return to the historiography of more modern periods, the narrative is out of favor when it comes to Late Antiquity. Indeed, the narrative chapters that appear in CAH XIII have recently been described by a reviewer as 'old-fashioned'.10 The very inclusion of the narrative has become an occasion for comment. Whatever one may think of this as a historian of Late Antiquity, it has the unfortunate effect of leaving much of the vast body of contemporary work in the field unread by a wider audience.

As for the selection of entries in the second section of the HarvardGuide, it would be interesting to compare their scope and emphasis with the late antique parts of the Oxford Dictionary of Byzantium11

and with the latest source books I have mentioned. A. D. Lee's book is explicitly called Pagans and Christians in Late Antiquity, and Michael Maas also has significant chapters on Christianity, polytheism, Judaism and Islam. Both reflect the emphasis on religion as part of cultural history that we have been familiar with since Peter Brown. But Maas's book also has broader coverage, including, for example, chapters on women, philosophy and medicine, as well as on Roman law, education and literature. The OxfordDictionary of Byzantium, which begins its chronological coverage with Late Antiquity, stands apart from all these works in that it deliberately minimizes the traditional concentration on spirituality, asceticism, or theology in extant Byzantine scholarship and favors material culture. This is very strikingly a reflection of the approach and preferences of the work's editor-in-chief, Alexander Kazhdan.

10 So Fergus Millar in JRA 13 (2000) 754-55.11 Kazhdan (1991).

ideologies and agendas in late antiquity studies 7

In the company of the Harvard guide, Michael Maas also gives space to war, the army and military affairs. In fact, I detect in the latter a certain reversal from the earlier Brownian emphasis on cultural history against law on the one hand and barbarians and warfare on the other. Thus Brent Shaw's chapter in the Harvard guide is entitled "War and violence" and it is prefaced by Patrick Geary on "Barbarians and ethnicity".12 However, I would argue that both themes are approached in ways that differ from the traditional.

New emphases can be seen. We have certainly seen tremendous interest in asceticism, and some very subtle and powerful work has been done in this area. However, there are other weights: take, for example, Judaism, now a strong player in the field. Leonard Rutgersha has recently written about Jews in late antique Rome; they are prominent in Bowersock, Hellenism in Late Antiquity, and the theme of Christian/Jewish disputation and polemic has been a fruitful subject.13

There are several reasons for this: one is perhaps the impact on late antiquity of the important recent work that has been done on the Jewish diaspora in the early empire, which shows the important role Jews played in the cities of the empire. Another reason is surely the growing body of archaeological evidence for synagogues and synagogue decoration, especially from Israel, dating up to the 7th century. A third stems from a more sophisticated reading of the Christian texts about Jews and Judaism.14 But finally and fundamentally, regardless of the title of Richard Lim's essay in HarvardGuide,15 the history of Late Antiquity is not today written in terms of a 'triumph of Christianity,' or primarily by Christian or confessional historians. Late Antiquity religious history, once seen in terms of a 'victory' of Christianity over paganism, is now more often studied in its own right. There is therefore more room for the non-Christian groups, both Gentiles and Jews.

Another theme, or rather another tone, that emerges from most of these works is optimism. While it was previously taken as a given that antiquity gave way to the Dark Ages and that the Islamic invasions brought rupture, the Harvard guide and other cur-

12 Shaw (1999); Geary (1999), with useful bibliography.13 For some of the extensive bibliography, see Cameron (1996), with Stroumsa

and Limor (eds.) (1996); cf. also Rutgers (1995).14 Basic work in this area has been done by Gilbert Dagron, Vincent

Déroche and Bernard Flusin, especially for the 7th century; see the articles in TravMém11 (1991), with Flusin (1993).

15 Lim (1999).

averil cameron8

rental works operate in a post-classical world with an identity of their own, not inferior, but different. In a sweeping Braudelian study that dissolves chronological divisions into a seamless synthesis, PeregrineHordern and Nicholas Purcell have argued that the Mediterranean itself is given over many centuries, rather than the various ruptures that may have occurred over that long span of time, which they cover.16 Fracture spelled decline and fall: continuity signals, if not quite optimism, at least authorial neutrality.

If we return to the barbarians, what do we see? Borders 'shift'; barbarians, once a non-problematic category, are now seen as undergoing, in their various ways, a process of ethnogenesis, becoming rather than being. The latter process is the subject of distinguished work in European Science Foundation project publications, notably by Walter Pohl and others.17 The broader topics of ethnicity and acculturation are topics for e.g. the recent and very interesting collection edited by Mitchell and Greatrex. Although naturally variable in its content, as collections are, it collectively addresses, through both textual and archaeological evidence and indifferent contexts, the very definition of ethnicity. Although we may have become unsure of our categories, the overall effect has been to provide a stimulus. Historians and archaeologists are fruitfully rethinking their old assumptions.

Serious work in the style of A. H. M. Jones on late Roman administration has been less prominent in recent decades, perhaps not least because it has seemed to offer less excitement than cultural and religious history. It may be suggested that in recent years it has been advanced through the continuous publication of the epigraphy of particular cities, such as Aphrodisias and Ephesus, or through publications on fiscal and economic matters, but Andrea Giardina takes English-speaking scholars to task with some justice to their neglect of institutional and administrative history.18 Peter Brown comments on what he sees as a changed emphasis from the old administrative

16 Hordern and Purcell (2000).17 From many examples, see Pohl and Reinitz (ed.) (1998) and cf. Geary (1999).18 However, there is a continuing French tradition in this area, which is also late

ancient epigraphy and economic history: cf. eg. the papers in Les gouverneurs deprovince dans l’antiquité tardive, Ant Tard 6 (1998); Durliat (1990). A French Late Antiquity would be represented on the one hand by the journal Antiquité Tardive, with its legacy from L. Robert, and the active promotion of N. Duval, and on the other by the work of Gilbert Dagron, with much greater emphasis on symbolic forms, about the imperial power and about the power of Christianity and Orthodoxy.

ideologies and agendas in late antiquity studies 9

history against a new 'symbiotic' model of power.19 I myself would place Ramsay MacMullen's book on Corruption and the Decline of Rome,20

which returns to the traditional emphasis on the supposed corruption of the later Roman ruling class, in the category of morality rather than administrative history. Yet definitional problems also arise in this area. Christopher Kelly examines, for example, fruitfully the meaning of the term 'corruption' as it is commonly used by modern writers on the late Roman bureaucracy (and indeed the word 'bureaucracy' itself carries a heavy load of negative connotations for us). In his contributions to both CAH XIII and HarvardGuide, Kelly points out the delicate balancing act on which the practice of the late Roman state rested. Patronage and pecuniary benefits are hardly unknown in modern states, but the simple transference of modern judgments to late ancient situations must be resisted. Interpreters of Late Antiquity must, in addition to looking towards various markers, for example the visuality and display of court ceremonies.21 We live in a media age, and it can be expected that we must now also be conscious in our subjects of performance, competition and visibility . Likewise, we are now more likely to be able to understand the workings of imperial and official power in terms of complex negotiations and transactions rather than in old-fashioned simple black-and-white terms.

When I taught the Roman Empire in London in the 1970s and 1980s, the ancient historians placed great emphasis on the Roman economy, and consequently on the size of the army, as the main object of state expenditure. A prevailing model of money circulation and the role of production, reminiscent of Moses Finley's classic book, The Ancient Economy,22 downplayed trade, presented cities as centers of consumption rather than production, and saw the Roman army through its purchasing power, mainly around the borders of the empire, as the main agent for the distribution of coins , which was then withdrawn through taxation. Such a model was extremely powerful. It permeates Michael Hendy's massive book on the Byzantine-Tinian monetary economy, which covers the late antique period and treats

19 Brown (2002) 81-85.20 MacMullen (1988).21 Kelly (1998), (1999).22 Finley (1973, revised edition 1985, 1999); the model was particularly assocd

along with the names of Keith Hopkins and (in an earlier period) Michael Crawford.

averil cameron10

the 4th to the 15th centuries as in some sense a unit.23 This model has since been subject to revision and the importance of trade has subsequently been re-emphasized. In relation to our present topic, the amount of production and type of exchange in Late Antiquity has therefore become a major topic of discussion, particularly in relation to the effects of the Vandal conquest of North Africa on long-distance trade.24 But the model in its pure form provided a tentative dichotomy between the High Empire, when it is claimed to have prevailed, and the 3rd century and later when it seemed in many ways to begin to break down. We were struggling hard at that time with '3. crisis of the century', only to discover that even there the questions were about pause versus continuity. Diocletian too, once the undisputed bringer of change and a decisive shift towards late Roman bureaucracy and oppression, has been dissolved into a process rather than an individual. Not so long ago was '3. century crisis' a major topic for both students and teachers; now it has become something of a mirage. Instead of worrying so much about the meaning of what is essentially equivocal evidence, historians focus directly on Late Antiquity itself. This may be because now historians often do not come to Late Antiquity from an Early Empire background and simply lack that point of comparison. The 3rd century is of vital interest if it is a question of understanding the transition from 'High Empire' to 'Late Antiquity', but much less if one starts from the medieval west or Byzantium. And if I feel the lack of a similarly simple model for the late empire, it's probably just as well that we don't have one, in that we are free to reflect the actual complexity and diversity of the data.

Many archaeologists and many historians have written about cities in Late Antiquity, not least recently Gianpietro Brogiolo and BryanWard-Perkins.25 One can say that urbanism has become the main question for Late Antiquity archaeologists, just as the crisis in the 3rd century was a main question. a generation ago for Roman historians. There was no simple answer then, nor is there now. It's a matter of changing priorities: to quote from Marlia Mango, “Period

23 Hendy (1985). 24 See Garnsey, Hopkins and Whittaker (eds.) (1983). To the discussion of his critics,

see Finley's new chapter, "Further thoughts", in Finley (1985) 182, and see the foreword by Ian Morris to Finley (1999) xxviii for 'the new consensus' which allows for a much greater amount of production and trade. For trade in the 6th century, see Hodges and Bowden (1998).

25 Brogiolo and Ward-Perkins (1999); Brogiolo, Gauthier and Christie (2000).

ideologies and agendas in late antique studies 11

425 to 600 is traditionally only seen as active and innovative in connection with the building of churches. But although this was much more marked in the East than in the West (reflecting the greater political stability and prosperity of the former), the 5th and 6th centuries. so also a significant continuity in the ideal and reality of the ancient city.”26 Both regionalism and the priorities of the scientific investigator play their part. Differences between East and West have become well known, in terms of the prosperity of the East as opposed to the signs of decline in the West.27 Textual evidence is often difficult to interpret, not least in the case of the Buildings of Procopius, one of the central texts for late Roman archaeologists . After reading the papers from a recent colloquium on this work, I concluded that there was still little real agreement, either on the text or on the topic.28 Although now central as a topic, late antique urbanism also involves difficulties. cult judgments about the use of spolia in new construction or restoration, or about change of use, churches being built, peristyle houses being divided and so on. Towns in Transition, the careful title of the collection edited by Neil Christie and Simon Loseby,29 suggests a change but leaves the reasons open. This change worked on many fronts, not purely economic, through patterns of thought, social practices, religious developments – the calendar, for example burial customs, development of Christian charity, family patterns. In the East (though not until the latter part of our period) alleys reddened by medieval markets began to superimpose the Roman street plans. Yet the evidence is far from consistent. Perhaps it is indeed the variety of data itself that gives rise to such a large amount of publication. But it can also all too easily lead to overgeneralization, which must be corrected by the most careful attention to local peculiarities.

If I were to point to a recent development in late antique studies that I find striking, it is this greater degree of sensitivity to the importance of evidence. Probably the most obvious field where this

26 Mango (2000) 971.27 Whitby (2000) 727, writes for example of "rapid and catastrophic collapse"

in the Balkans from the 6th century; similarly, Hodges (1998) 11 writes of "catastrophic decline" in the West from the 6th century.

28 Cameron (2000b); on the issue of texts versus material evidence cf. Brandes(1999) 36-41; Ward-Perkins (2000) 315, contrasting current scholarship with that of a generation or more ago: "Today, one is more likely to find accounts that bypass the contemporary written texts or dismiss them as hopelessly unreliable."

29 Christie og Loseby (red.) (1996).

averil cameron12

sensitivity has been shown is the hagiography, where there are many examples of new and much more nuanced readings. This is undoubtedly a product (though not acknowledged) of the deconstructive turn. One such read that I can think of immediately is Derek Krueger's recent book on Narsen's Life of Symeon, which he sees as a way of understanding eastern cities in the 6th century. The barbarian plain takes a single place of pilgrimage, the one for St. Sergius of Rusafa, weaving together historical, textual and archaeological evidence to bring it to life.31

There are countless other examples of this imaginative reading of texts. Indeed, late antique historians have applied techniques and approaches to their subject that have become widespread in wider classical and medieval scholarship. The new reading (of visual as well as textual material) is evident in the secondary literature of all periods of the empire, not just Late Antiquity. JásElsner titles a section of a recent article dealing with the Constantinian period "Rethinking the Roman Empire".32 We are encouraged to ask why it needs to be rethought, and then experience how contemporary scholars are re-thinking it.

It is not entirely unreasonable to ask in what ways the Late Antiquity of historians differs from the Late Antiquity of archaeologists. The simple answer is that it shouldn't: everyone's goal should be to integrate all kinds of evidence as closely as possible. Without denying a significant number of striking exceptions, it is fair to say that there is indeed a gap. The late antiquity of many historians, especially of the Brown school, is characterized by themes of competitiveness, ascetic rivalries, imaginative leaps, and by the visibility of saints and the lives of saints. In contrast, most archaeologists' Late Antiquity rests on trying to make sense either of city remains or of scattered evidence of settlement and survey archaeology. In both cases he must try to place this within a historical framework provided by other forms of evidence and secondary literature. It is useful to remember how difficult it is to match the archaeological evidence with the texts, even when texts are available. There is rarely a good fit, and not infrequently there is no textual evidence for an archaeologist to turn to.

30 Krueger (1996).31 Fowden, E. Key (1999).32 Elsner (2000) 194-95.

ideologies and agendas in late antiquity studies 13

Conversely, the case of Procopius's buildings cautions against the too-ready attempt to interpret physical evidence in ways that fit a superficially credible ancient text.

One field in which the use of written evidence has been transformed is the reading of the late Roman legal texts. A lot of work has recently concentrated on Theodosian law.33 Attention has again been drawn to the fact that late Roman law functioned in practice rather differently than it appears in the textbooks, not least in the areas in which it overlapped with barbarian law. legal codes. A generation ago, it was still possible to see the later Roman Empire as a society characterized by a rigid and legalistic bureaucracy, where the main evidence lay in the law codes. RamsayMacMullen questioned this in the 1960s, making the simple observation that repetition of the content of a law probably does not mean that there was heavy-handed repression, but rather that the law was flouted or not enforced.34 Thirty or more years later, the Late Roman government has proven formidable, to be sure, but also much weaker than we once thought, relying on rhetoric where it lacked real enforcement power. An increased awareness of regionalism and recognition of the great diversity of forms of late Roman urbanism make broad generalizations of the older type less plausible or appealing. At the same time, the public role of the bishops in relation to their own cities and to the imperial and provincial government has emerged as a central theme, which also makes the study of official buildings in late antique cities and their relationship to the bishops' houses and churches. buildings a very current topic.35

To return to the legal texts, it is no longer possible to read the late Roman law codes as anything other than very complex texts whose interpretation requires great sophistication and whose practical impact on individual provinces can be extremely varied.

However, there are other areas where the pendulum swings back. I notice that questions are being raised about what is seen in some circles as too much emphasis on cultural history, as conveying, one might say, too much transformation and not enough rupture. Take Peter

33 Harries and Wood (eds.) (1993); Harry (1999).34 MacMullen (1964), a then classic article; he returned to the problems in MacMullen

(1988). On these models of late Roman government see the revisionism of Brown (2002) 81.

35 Bishops; see Kelly (1999) 186, with bibliography at n. 103; also Brown (2002)67-70 and passim.

averil cameron14

Heather in CAH XIV, charting the old problem of "the fall of the Western Empire": "The nature of the crisis has been much debated and in some ways does not now appear as catastrophic as it did to the tool of generations of historians".36 He reviews (in a very considered and fair way) the now well-known arguments which are used to soften the sense of rupture and which have in fact led almost to the disappearance of the problem of the fall of the Western Empire from recent Anglo-Saxon scholarship.37 Nevertheless, concluding he: "None of this means, however, that the fall of the Roman Empire was anything other than a revolution. ... As we have seen, the process of fragmentation was protracted and violent." Heather is open to ideas of acculturation and, moreover, to acculturation as a two-way process: many Romans worked for the barbarians and for a variety of motives, just as the Goths had been drawn into Roman service.38 But in his model, barbarians did not simply assimilate; they actually invaded and destruction occurred. Military history is also enjoying a comeback, and as a consequence the late Roman army has been resurrected as a massive fighting machine, for example in two recent works by Michael Whitby.39

All this seems to me quite natural, even inevitable. After all, the story continues by reaction and reaction. However, I would also like to suggest, especially when examining Wolf Liebeschuetz's vigorous defense of the usefulness of the concept of decline,40 that here too we see in part a manifestation of the type of conservative populism that is currently evident in historiography and literary criticism in more modern periods. The historiography of Late Antiquity came late to debates long known to literary critics and reviewers, and even to debates about methodology among historians. Again, while Peter Brown may have brought about a sea change in the way people have written about Late Antiquity in our generation, it has been largely without overt discussion of historical method. Rhetorical criticism and the deconstruction of texts are methods that have entered scholarship in the field, especially in the work of younger American writers on Late Antiquity

36 Heather (2000) 31; see also on continuity versus rupture Wood (1998) 231.37 Also see for example Bowersock (1996).38 Heather (2000) 29.39 Whitby (1995); similarly, despite a "western collapse" section, Whitby

(2000) 40 Liebeschuetz (2001a).

ideologies and agendas in late antiquity studies 15

who are very aware of the power of language and the importance of representation.41 But serious debate about the nature of history and how to write it is noticeably lacking in their work.

Nevertheless, we are all affected by these debates, even when we may claim otherwise. The intensity of the general debate about history can be clearly seen in the reaction to a book published in 1997 by Richard Evans, 20th century historian. Germany and RegiusProfessor at Cambridge, and called In Defense of History.42 Evans's book is a relatively conservative defense of the claim (against the attacks of postmodernists) that history has to do with truth. Nevertheless, it offers a very well-balanced discussion of the issues, and if Evans is reluctant to go all the way down to relativity, he is not an unreconstructed conservative either: indeed, he ends with a chapter entitled "Objectivity and its limits". In that case, he has been bitterly attacked not only by postmodernists, so-called, but also by their opponents. In a 60-page afterword to the revised edition, he defends himself against critics from both the left and the right. Today, the former are not so much Marxists as postmodernists. The 'Right' is represented by neoconservative historians and biographers of the modern period such as Niall Ferguson, Andrew Roberts, Michael Burleigh and others, historians whose general approach – no-nonsense, massive documentation, political history, the return of the narrative and so on, has often been observed. Their counterparts are equally to be found among contemporary British reviewers and critics, for example Paul Johnson (also writing history), AuberonWaugh, Simon Jenkins and others, a group recently identified by John Carey, the Merton Professor of English in Oxford and a pro-lific reviewer and critic, as "conservative populists and anti-mod-ernists". Evans was clearly astonished, and no doubt flattered, to be attacked as forcefully by this group as by the more extreme postmodernists whom he gently criticized. This afterword is worth reading for those historians of late antiquity who believe they do or can stand aside from the vigorous and sometimes violent debates into which Evans enters.

Since I myself have recently been described as a cultural kin-

41 Particularly evident in the body of literature on asceticism and hagiography: see Cameron (2002) 180-83. Pilgrimages and places of pilgrimage, especially the sanctuaries of saints, constitute a field where archaeologists, historians and critics of texts can meet: see Cameron (2002) 176.

42 Evans (1997), (2000).

averil cameron16

ist, and since the conclusion to CAH XIV would certainly seem to reinforce this view,43 perhaps more is needed. All historians, myself included, must believe that writing history is a matter of integrity and faithfulness to the truth of the evidence; this is also Evans' position. But the naive realist's supposed anti-ideology will not do. It is not relativism to recognize that the questions we ask as historians are – at least in part – a question of what is going on in our own experience. Why else was the fall of the Soviet Union hailed as the "end of history"?44 Why did the European Science Foundation choose to sponsor a project on The Transformation of the RomanWorld, if not to explain the formation of Europe in the light of contemporary agendas, and why support the currently one about the formation of a European identity in the early modern period? Why has Marxist historiography been muted, to say the least, whereas not long ago Marxist agendas and Marxist ideology were so alive in the historiography of the later Roman Empire?

Why, indeed, is it that many late antique historians are so much more aware of the pitfalls of representation, if not that they have been influenced by the philosophical and literary movements of the last generation? The historian does not sit in his study immune to other influences. The prevailing models of ancient urbanism in the last half century were heavily influenced by theoretical writing, whether Weber, Polanyi or Marxist historiography. There may have been a move away from such influences, at least in Anglo-Saxon writing. But other influences have taken their place: late antique studies are not immune to the culture wars. We cannot bury our heads in the sand, appeal to 'health' and pretend that what happens in other disciplines, or in the world around us, is unimportant, and indeed, even if we claim to do so, we will not succeed.

Some agendas, such as pluralism, regionalism, fragmentation, the

43 CAH XIV, 972-81.44 Famous by Fukuyama (1992).45 See also Giardina (1999) 172-73; for discussion of Marxist interpretations in

late antique archeology see the memorable review article by C. Wickham, Wickham (1988). Ian Morris, in a preface to Finley (1999) xxv, traces the change from social and economic, though not Marxist, history to cultural history, noting that this is partly a generational problem and that many departments nevertheless still harbor aging (sic) " radical economic and social historians defending their turf against the cultural historians”.

ideologies and agendas in Late Antiquity Studies 17

questioning or redefining knowledge and creating new identities is at one and the same time political and historical.46

In advocating some of these I have been accused of following the Blairite Third Way. A deeper and fairer reading would not say that these are New Labor themes, but rather that they are the agendas of the day. In this sense, history cannot be anything other than political, in the sense that we ask the questions of the past that our own experience leads us to. to ask. If this is true of history, how much more is it true of archaeology? Even without considering the intellectual trends that I have described, all archaeologists are aware of the extent to which national, economic and cultural agendas inform the direction and choice of excavation, not to mention the interpretation of the results. Indeed, the weight of such influences is felt all the more strongly because money, local and national politics, and practicalities are all heavily involved. For historians, the pressure may be less direct, but they may be more insidious about it.

The title, "Ideologies and Agendas in Late Antique Studies" is timely. The events of 2001 dealt a severe blow to the Blair Project and the multicultural turn, along with the benign and optimistic view of cultural development rightly identified by Andrea Giardina as a major theme of Anglo-Saxon scholarship in the last generation. The end of communism is now seen to have brought with it a number of other problems. The histories of both Islam and Judaism, each an important component of the Harvard Guide's view of Late Antiquity, invite complex reassessment. It may be that we can expect to see the optimistic view of the 'long late antiquity' that has held the field in recent years replaced by a less benign model that is as attentive to conflict and division as to assimilation and cultural interaction. The nature of imperialism, not a prominent topic in recent Late Antiquity scholarship, will certainly require renewed attention, particularly in the case of the Eastern Empire. Not only does this point to a return to the older topics of late Roman government and economy, but this attention should also include consideration of the state structures of Byzantium. Indeed, there is a striking difference between those who approach late antiquity from the perspective of the early medieval West and those whose perspectives derive from the ancient world or from the East. This is

46 Hodges (1998) 4-5, also draws attention to the politics of the theme of 'transformation' of the Roman world.

averil cameron18

especially evident in archaeological writing, where researchers working mainly in the West are used to looking further chronologically. The questions they ask and the agendas they bring are often very different from those of historians whose work is more focused on the late Roman state itself. This was very clear during the meetings of the Transformation of the Roman World project, and it is equally clear in the publications resulting from it.47 All in all, the enormous amount of attention devoted to 'transformation' in late antiquity period, in the East and As in the West, the remarkable persistence of the Byzantine imperial endeavor usually goes unnoticed, as if the history of the Byzantine state is something entirely separate from late antiquity or the later Roman Empire.

A final note therefore. Much of current Late Antique archeology is concerned with cataloging the signs of the break or decay that marks the 'end of antiquity', which can be observed in many places, though not in all. But that very enterprise rests on pre-existing notions of periodization, and these notions of periodization are inexorably built into the very scope and coverage of a large number of archaeological publications. It seems to me that the questions that now arise for historians of late antiquity and archaeologists alike, and that I have just outlined, require a different kind of chronological and thematic division, which can trace the transition over a longer period of time and thus allow for a deeper understanding of its complexity. Many archaeologists are in the position, or imagine themselves to be in the position, that after the end of Late Antiquity they have little evidence from which to attempt some answers. But the archaeologist and the historian are actually complementary. In this period, perhaps more than any other, the historian needs the archaeologist, and the archaeologist needs the historian.

47 The debate for the West also remains particularly focused on cities; Ward-Perkins (2000) 315-19 points out that while the study of rural life, villages and the rural economy have for their own sake become a major theme on Eastern agendas, this has not yet happened for the West.

ideologies and agendas in late antique studies 19

Bibliography

Bowersock G. (1996) "The vanishing paradigm of the fall of Rome", Bulletin of the American Academy of Arts and Sciences 49 (1996) 29-43.

——— (1990) Hellenism in Late Antiquity (Ann Arbor, Michigan 1990). Bowersock G., Brown P. and Grabar O. eds. (1999) Late Antiquity. A guide to

Post-Classical World (Cambridge, Mass. 1999). Brandes W. (1999) “Byzantine cities in the seventh and eighth centuries – differ-

sources, different stories?”, in The idea and ideal of the city between Late Antiquity and the Early Middle Ages, ed. G. Brogiolo and B. Ward-Perkins (The Transformation of the Roman World 4) (Leiden 1999) 25-57.

Brogiolo G. and Ward-Perkins B. eds. (1999) The idea and ideal of the city between Late Antiquity and the Early Middle Ages (The Transformation of the Roman World4) (Leiden 1999).

Brogiolo G., Gauthier N. and Christie N. eds. (2000) Cities and their Territories between Late Antiquity and the Early Middle Ages (The Transformation of the Roman World9) (Leiden 2000).

Brown P. (1971) The World of Late Antiquity (London 1971).——— (1996) The Rise of Western Christendom: Triumph and Diversity (Oxford 1996).——— (2002) Poverty and Leadership in the Later Roman Empire (The Menahem Stern

Jerusalem Lectures) (Hanover 2002). Cameron Av. ed. (1995) The Byzantine and Early Islamic Near East: States, Resources

and Armies (Studier in Late Antiquity and Early Islam 1.3) (Princeton 1995).——— (1996) “Byzantines and Jews: some recent work on early Byzantium”, BMGS

20 (1996) 249-74. Cameron Av. and Garnsey P. eds. (1998) Cambridge Ancient History XIII. The late one

Empire, AD 337-435 (Cambridge 1998). Cameron Av., Ward-Perkins B. og Whitby M. edd. (2000a) Cambridge Ancient History

XIV. Late Antiquity: Empire and Successors, AD 425-600 (Cambridge 2000). Cameron Av. (2000b) "Conclusion, de aedificiis: the text and the realities", Ant Late 8

(2000) 177-80.——— (2002) "The 'Long' Late Antiquity: A Late Twentieth Century Model", in Klasse-

sics in Progress, red. T. P. Wiseman (Oxford 2002) 165-91.Christie N. og Loseby S. edd. (1996) Byer i overgang. Urban Evolution i Sen

Antiquity and the Early Middle Ages (Aldershot 1966) Durliat J. (1990) From the Ancient City to the Byzantine City. The Subsistence Problem (Rom

1990). Elsner J. (2000) “The Itinerarium Burdigalense: politics and salvation in geogra-

phy of Constantine's empire”, JRS 90 (2000) 181-95.Evans R. (1997) In Defense of History (London 1997, rev. ed. 2000).Finley M. (1973) The Ancient Economy (Berkeley og Los Angeles) 1973, rev. 1985,

1999).Fowden G. (1993) From Empire to Commonwealth. Monotheism's consequences in

Late Antiquity (Princeton 1993). Fowden E. Key (2000) The Barbarian Plain. Saint Sergius between Rome and Iran (Ber-

keley and Los Angeles 1999). Flusin B. (1993) Saint Anastase le Perse, 2 vols. (Paris 1993).Fukuyama F. (1992) The End of History and the Last Man (London 1992).Garnsey P., Hopkins K. and Whittaker, C. eds. Trade in the ancient economy (Cam-

bridge 1983).Geary P. (1999) "Barbarians and ethnicity", in Antiquity. A Guide to the Postclassic

averil cameron20

Verden, ed. G.W. Bowersock, Peter Brown og Oleg Grabar (Cambridge, Mass. 1999) 107-29.

Giardina A. (1999) "Esplosione di tardoantico", Studi Storici 40 (1999) 157-80.Harries J. (1999) Law and Empire in Late Antiquity (Cambridge 1999).Harries J. og Wood I., red. (1993), The Theodosian Code (London 1993).Heather P. (2000) "The western empire", CAH XIV (Cambridge 2000) 1-32.Hendy M. (1985) The Byzantine Monetary Economy, ca. 300-1450 (Cambridge 1985). Hodges R. (1998) "Henri Pirenne and the question of demand", i The Sixth Cen-

tury. Production, distribution and demand, ed. R. Hodges and W. Bowden (The Transformation of the Roman World 3) (Leiden 1998) 2-14.

Hodges R. and Bowden W. eds. (1998) The Sixth Century. Production, Distribution and Demand, (The Transformation of the Roman World 3) (Leiden 1998).

Hordern P. and Purcell N. (2000) The Corrupting Sea. A Survey of Mediterranean History (Oxford 2000).

Howard-Johnston J. og Heywood P. edd. (1999) The Cult of the Saints in Late Antiquity and the Early Middelalder (Oxford 1999).

Kazhdan A. P. udg. (1991) Oxford Dictionary of Byzantium, 3 bind. (New York 1991). Kelly C. (1998) "Emperors, Government and bureaucracy", CAH XIII (Cambridge)

1998) 138-83.——— (1999) “Empire building”, in Late Antiquity. A Guide to the Post-Classical World,

edd. G. Bowersock, P. Brown og O. Grabar (Cambridge, Mass 1999) 170-95.

Krueger D. (1996) Symeon the Holy Fool: Leontius's Life and the Late Antique City (Trans-formation of the Classical Heritage 25) (Berkeley 1996).

Lavan L. udg. (2001) Recent Research in Late Antique Urbanism, (JRA SupplementarySeries 42) (Rhode Island 2001).

Lee AD ed. (2000) Pagans and Christians in Late Antiquity. A Sourcebook (London 2000).

Lepelley C. (1979, 1981) The Cities of Roman Africa in the Lower Empire 2 vols (Paris 1979,1981).

Liebeschuetz J. H. W. G. (2001a) “Was Gibbon Politically Incorrect? The Use and Abuse of Concepts of 'Decline' in Later Roman History", in Recent Research in Late Antique Urbanism, ed. L. Lavan (JRA Supplementary Series 42) (RhodeIsland 2001) 233-38, reply to 238-45

——— (2001b) The Decline and Fall of the Roman City (Oxford 2001). Lim R. (1999) "Christian triumph and controversy", i Senantik. En guide til

Postklassisk verden, red. G. W. Bowersock, P. Brown og O. Grabar (Cam-bridge, Mass. 1999) 196-218.

Maas M. ed. (2000) Readings in Late Antiquity. A Sourcebook (London 2000). MacMullen R. (1964) "Social mobility and the Theodosian Code", JRS 54 (1964)

49-53.——— (1988) Corruption and the Decline of Rome (New Haven 1988).Mango M. (2000) “Building and Architecture”, CAH XIV (2000) 918-71.Marcone A. (2000) “ Late Antiquity and its periods”, Riv StorIt 112

(2000) 318-34.Mathisen R. og Sivan H. (red.) (1996) Shifting Frontiers in Late Antiquity (Aldershot)

1996). Miles R. udg. (1999) Constructing Identities in Late Antiquity (London 1999). Mitchell S. og Greatrex G. (red.) (2000) Ethnicity and Culture in Late Antiquity (Lon-

don 2000). Pohl W. and Reinitz H. (eds) (1998) Strategies of Distinction. The construction of Eth-

ideologies and agendas in late antique studies 21

nic Communities, 300-600 (The Transformation of the Roman World 2) (Leiden1998).

Rutgers L. (1995) The Jews in Late Antique Rome: Evidence for Cultural Interaction in the Roman Diaspora (Leiden 1995).

Shaw B. (1999) "War and violence", in Antiquity. A Guide to the Postclassical World, ed. G. W. Bowersock, P. Brown, and O. Grabar (Cambridge, Mass. 1999) 130–69.

Stroumsa, G. G. and Limor O. eds. (1996) Contra Iudaeos: ancient and medieval polemics between Christians and Jews (Tübingen 1996).

Valantasis R. udg. (2000) Religions of Late Antiquity in Practice (Princeton 2000). Ward-Perkins B. (2000) "Land, labor and settlement", CAH XIV (Cambridge 2000)

315-36.Whitby M. (1995) "Recruitment in Roman Armies from Justinian to Heraclius",

i The Byzantine and Early Islamic Near East: States, Resources and Armies, red. AverilCameron (Studies in Late Antiquity and Early Islam 1.3) (Princeton 1995)61-124.

——— (2000a) “The Army, ca. 420-602”, CAH XIV (Cambridge 2000) 288-314.——— (2000b) “Armies and Society in the Later Roman World”, CAH XIV (Cam-

bridge 2000) 469-85.Wickham C. (1988) "Marx, Sherlock Holmes and Late Roman commerce", JRS

78 (1988) 183-93. Wiseman T. P. ed. (2002), Classics in Progress (Oxford 2002). Wood I. (1998) “The Frontiers of Western Europe: Development East of the Rhine

in the sixth century”, in The Sixth Century. Production, distribution and demand, ed. R. Hodges and W. Bowden (The Transformation of the Roman World3) (Leiden 1998) 231-53.

this page is intentionally blank

archeology and late antique social structures 23

PART TWO SOCIAL LIFE

this page is intentionally blank

archeology and late antique social structures 25

ARCHEOLOGY AND LATE ANTI-SOCIAL STRUCTURES

JEAN-PIERRE SODINI

Abstract

The archaeological remains of late antique sites can be interpreted based on what they can tell us about ancient social structures. This is more straightforward when examining the social structures of the upper classes, which possessed the characteristics that allow them to be recognized as such. These characteristics occur on a Mediterranean basis and include lavishly decorated dwellings (in both urban and rural settings), monumental funerary structures in churches, splendid clothing, costly tableware and utensils, and rank in the form of jewelery such as gold brooches, fibulae or belt buckles. The middle class can also be traced in the cities (mostly in the form of artisans) and in the countryside, where small landowners and farmers could share similar lifestyles, characterized in some regions (such as the Near East and Asia Minor) by conspicuous levels of prosperity. However, the lives of these middle classes could change abruptly, plunging them into poverty and, as a result, making them difficult to trace archaeologically. Nevertheless, a judicious interpretation of the material remains, together with the evidence from documentary and epigraphic sources, to allow us to make some suggestions about the social structures of Late Antiquity.

Introduction

"Every city encloses at least two cities hostile to each other: one of the poor, one of the rich" Plato, Rep. 422

The goal of any young field archaeologist is not only to correctly identify and isolate excavated archaeological contexts, but also to understand past societies through the material remains contained within them. In this he hopes he will be more successful than the historian, who has fewer surviving sources at his disposal. These sources may include laws held throughout the ages and their application

L. Lavan og W. Bowden (red.) Theory and Practice in Late Antique Archaeology (Late Antique Archaeology 1 2003) (Leiden), s. 25-56

jean-pierre sodini26

tion is uncertain, lives of saints, epic poetry, or court chronicles focused on the deeds of an emperor living in a capital in the midst of a closed company of courtiers. The archaeologist, on the other hand, views his discipline as concerned with reality, as opposed to the rarefied life of the court or the lifestyle represented in elite writings. He or she is therefore proud to be accused of digging into the dustbins of the past.

As our hypothetical young archaeologist grows up, however, the story becomes more complicated. Once he has established his facts, she is forced to interpret them. The focus of his studies will shift from buildings to their occupants, raising questions regarding their status in the society they lived in. The answers from his findings also raise a host of unanswered questions. For example, if he has excavated a magnificent house or a magnificent mausoleum, or found jewelry or silver vessels in an undisturbed grave, it can be said that their owner was rich, perhaps a curialis. Alternatively, if the same house has been found in a poor condition, with worn pavements, narrowed or blocked doors and with pits and post holes cut into the floors of the rooms, the story of the occupants will be different and our archaeologist will have to resolve this. He will have to examine the various statuses of the residents and try to determine who was poor and who was rich, and perhaps whether the later residents were newcomers. These different extremes of wealth and poverty are, at least superficially, easy to determine. But even these seemingly simple definitions are not static or invariable, but instead can fluctuate between the 5th and 8th centuries, even within the same city. It is increasingly difficult to introduce degrees of differentiation between the wealthy, the middle class and the poor. What are the criteria for establishing these categories and distinguishing between them? Do we use diet and skeletal composition, the presence of utilitarian or luxury tools, or the size of the house to determine which category the occupants should be assigned to? In this simple case, what do we classify as a luxury home, yes-to-do apartment, or a modest home? Similarly, material culture is not necessarily a reflection of status or position within a social hierarchy. Shall we rank an ascetic monk, learned and of a rich background, among the poor? We are therefore constantly confronted with the possibility, indeed the probability, of significant errors of interpretation. In many cases we cannot determine the social status of a farmer (whether he is a freeholder, emphytheot, farmer or colonus),

archeology and late antique social structures 27

or acquired as a citizen. We cannot hope to retrieve individual stories; how, for example, could archeology trace the social ascension of the mother of Theodore Sykeotis, who rose from being a prostitute and innkeeper to being the wife of 'a leading citizen of Ankyra, the protector David'.1

Nevertheless, despite these limitations, it is possible in some cases to gain a greater understanding of the different social status of those we can trace in the archaeological record. The easiest of these two races is the aristocracy.

The elites

As Ross Balzaretti has written: "The archaeological study of power is difficult".2 However, I would add that it is easier to study power than to study its absence. It is certainly easier to discover wealth through archeology than it is to discover poverty; inherently, a mosaic is more likely to survive the archaeological process than a floor made of rammed earth. The rich are visible; they possess lands, houses, goods of every kind, even their dress shows their social importance. Their graves contain clothing tools that provide indisputable evidence of their status. We also have many historical signs of the presence of elites, some of them contained in legal texts, such as the Notitiae Dignitatum. These elites may include the head of the empire (emperors and their entourage and honorati, civil and military), the high clergy (patriarchs, bishops, abbots of great monasteries), the curiales, whose power was gradually concentrated in the hands of the wealthiest inhabitants (often called proteuons ), regardless of whether they were of Bouleutian origin or not. These last were also called possessores, a term whose exact definition is much disputed.3

Some of these honorati (for example Ekdikos or Defensor Civitatis, Pater Poleos), the bishop and possessores (or ktétores), appear frequently in the 6th century. imperial edicts and inscriptions.4

1 A good analysis of the Life of Théodore of Sykeon is provided by Mitchell(1993) II 122-50.

2 Balzaretti (1994) 104.3 The best explanation of this term can be found (pace Durliat (1996) and

Delmaire (1996)) in Laniado (1997) and (2002). See also Liebeschuetz (1996) and Holum (1996).

4 Keil and Wilhelm (1931) n. 197A; Feissel (1999).

jean-pierre sodini28

The characteristics of the elites

In all parts of the late antique empire there were huge imperial states. During the 5th century the church acquired extensive properties and there was a class of wealthy landowners in almost every town. But, as Paul van Ossel has noted, "property is a notion that archaeologists cannot study".5

In fact, archeology can rarely define the extent of estates on earth. A possible exception, rightly emphasized by van Ossel, is the so-called Langmauerbezirk, a wall surrounding a large number of rural settlements in the area between Trier and Bitburg.6 This large walled estate (220 km2, fig. 1), built under Valentin reign, clearly belonged to a single owner, who may well have been the emperor himself. This is indicated by the fact that the Primani, (troops belonging to the Imperial Guard) (comitatenses or pseudo-comitatenses), took part in the wall construction. However, it is an unusual solution to enclose such a stretch with a wall. Generally, the boundaries of an estate were preserved in notarial deeds or in records of tax payments kept by the fisherman. Many historians have tried to estimate the extent of these estates, which were not always in a single location, but were sometimes spread through different regions. The most convenient analysis for our purposes is that of J.-M. Carrié, who defines this type of land use as "citadine" land use (where land was held by people living in cities) or "curial" land use.7 These were important properties of varying size that were generally granted to farmers who paid rent. The increase in land taxes in the 4th century and the increase in land prices in the 6th century threatened the prosperity of the curiales, who did not have access to extra income (i.e. through the patronage of high-ranking officials) or could not farm directly with agriculture -natural workers. Those who succeeded in increasing their income from land became the proteuonts or possessors already mentioned. The emergence of a dynamic peasant middle class was also part of the weakening of the traditional curiality.

5 Van Ossel and Ouzoulias (2000); Balmelle and Van Ossel (2001).6 Van Ossel and Ouzoulias (2000).7 Carrie (1997).

archeology and late antique social structures 29

Fig. 1. Langmauerbezirk (Van Ossel P. (1992) 95 fig. 3). For fig. 2 see plates.

jean-pierre sodini30

Houses and Residences

We are better able to get an idea of ​​the wealth of the upper classes when we examine their houses. In recent decades, archaeologists have added some splendid discoveries to the known range of late Roman imperial residences that set the tone for this 'architettura di potenza'. These include the Palatine in Rome,8 the Maxence 'villa' on the Via Appia,9 just outside Rome, a possible example in Milan and further residences in Trêves, Sirmium, Thessalonica and Constantinople, as well as retirement homes (of which Split can be seen as a prototype ). The most magnificent example is Romuliana, the palace built by Galerius at his birthplace. The site, known for years as Gamzigrad, has been identified with certainty through the discovery of an archival vault inscribed with the words Felix Romuliana. Prior to the discovery of this inscription, the attribution of the site to Galerius and its identification as an imperial residence had been hotly contested. Among the spectacular finds from Gamzigrad relevant to this discussion are the pilasters (Fig. 2) with representations of emperors (two retired Augusti (seniores): Diocletian and Maximian, two Augusti: probably Galerius and Licinius, and two Caesars: Maximinus and Constantine). These pilasters may refer to the imperial college reconstituted at the conference of Carnuntum in November 308. On top of the Magura hill outside the walls were two mausoleums, accessed via a tetrapylon, in which Romula and her son Galerius was buried and deified. As Timothy Barnes has noted, the excavations provide more than just the location.10 They confirm the claim of Lactantius (Mort. Pers.) that as early as 305 Galerius had planned to retire after celebrating his vice-centenary on 1 March 312 .Galerius' son, Maximinus Daïa, born at Sarkamen, near Romuliana, also began building a palace at his birthplace, as well as two mausoleums for his mother and himself. When he was assassinated in 313, construction ceased, but he had managed to have his mother cremated and her ashes placed on top of a tomb found in the mausoleum. A jewelry magazine of 29 pieces of jewelry, discovered in a niche in the crypt, constitutes an additional class symbol that can

8 Augenti (1996) 9 Cullhed (1994).

10 Barnes (1996) 552.

archeology and late antique social structures 31

be associated with a particular person, a point to which we shall return.11

To the examples mentioned above can now be added the 'Palaisde la Trouille' in Arles, which may well have been Constantine's palace.12 The study of many of these palaces, including the Imperial Palace in Constantinople and the Palace of Galerius in Thessalonica, continues today. However, the absence of more comprehensive archaeological data means that there is still doubt about Bardill's recent interpretation of the palace in Constantinople.13

one. Governors Palaces and Praetoria Luke Lavan has recently compiled a catalog of buildings similar to the grandiose residences described above that have been identified as the residences of late antique governors.14

The identification of some of these buildings as governor's palaces is still questionable. The Praetorium at Gortyn, for example, seems in its earliest phase to have been a Hellenistic gymnasium connected to a stadium.15 In the Roman period this was replaced by the termae, which also functioned in connection with the stadium. In 382-383, after the earthquake of 365, the praes Ecumenius Dositheos Asklepiodotus erected a 'kainon praitorion' in honor of the emperors Gratian, Valentinian and Theodosius. This evidently relates to the court built on the NW corner of the termae already mentioned, and may even have been restored by Heraclius after the earthquake of 618. The existence of living quarters that could have housed the governor has not yet been firmly established.16

Other praetoria are even more controversial, as in the case of the huge 4th century. villa discovered at Cercadilla, 600 m. from the walls of Cordoba (fig. 3). Hidalgo originally identified it as a palatium of Maximianus Herculius, an attribution opposed by Arce, who suggested that it might be a governor's palace, a hypothesis that is also

11 Popovic and Tomovic (1998).12 Heijmans (1998).13 Bardill (1999). Another attempt to interpret the areas discovered so far, that

Mrs. Bolognesi and the French publishers of the second part of the book of ceremonies are in the process of a larger account of the terraces. The Turkish Archaeological Service has now carried out new excavations in the palace, and we may be able to get more convincing attributions to the various rooms.

14 Lavan (1999) and (2001).15 Di Vita (2000).16 Ibid. LVII-LXIX and 363-364: 'palazzetto' reuse of triconich could be

appropriate.

jean-pierre sodini32

Fig. 3. Cercadilla (Arce J. (1997) 295 fig. 1).

archeology and late antique social structures 33

open to question.17 In Caesarea two palaces have meanwhile been excavated. One of these, built on a promontory by Herod, was the governor's palace in the Roman period18, while the other was built while Caeserea was a Roman colony and housed the procuratorprovinciae in charge of the province's economic affairs. After the 4th century, the lower terraces of the Herodian palace were partially abandoned, although the remainder may have continued to serve as a castron for the dux in charge of military affairs. At the same time, the former praetorium of the procurator became the seat of the Byzantine governor. A recently discovered 5th century decree specifying "fees to be paid for specific judicial services and proceedings at the governor's civil court," confirms the palace's identification as a praetorium in Late Antiquity.19

High-ranking officers in the army may have used similar luxurious buildings (as already seen in Caesarea). At Mediana, near Naissus, a splendid villa may well have been the residence of a duxin-Constantine, although the complex has not been excavated in its entirety. The base of two porphyry statues occupies thediasemotatos (vir perfectissimus) Roimetalkés. According to AmmianusMarcellinus (XXVI 5, 1), under Valens' and Valentinian's sojurnin Naissus (364), both emperors sent their retinues to Mediana, while they themselves were apparently quartered intra-muros in Naissus.20 In the North Syrian basaltic zone, meanwhile, still prefer to consider that Qasr ibn Wardan (dated to 561–572) was almost certainly a military residence associated with nearby barracks, despite its recent interpretation as a proastion for a wealthy landlord.21

The wall construction and vaulting techniques are reminiscent of many places fortified by the Byzantine army along the banks of the Euphrates and East of Apamea.22 A low podium placed a little away from the main door may well have been the place where the officer stood when he reviewed the troops.23 Convincing arguments have been made for other praetoria in Palmyra and Caricin Grad (the so-called 'domusurbana').24

17 Hidalgo (1996); Arce J. (1997); Lavan (1999) 139-40.18 Netzer (1996); Burrell (1996).19 Patrich (1999) and (2001).20 Sodini (1997) 448; Srejovic (1993) 69-75.21 Gatier (2001) 105.22 Deichmann (1982) 727-35.23 Sodini (1997) 490.24 Sodini (1997) 485-87 and 445-46.

jean-pierre sodini34

Fig. 4. Akkale, plan of residential construction (Eyice (1986) 67 fig. 2).

Fig. 5. Akkale, partial view of housing construction and large vaulted cisterns (Hildand Hellenkemper (1990) 165-67 fig. 2).

archeology and late antique social structures 35

Turning our attention to Cilicia, we can notice in Akkale a small coastal settlement, 5 km from Elaiousa-Sebasté. This includes an Alavi-built complex of buildings (Figs. 4 and 5), the main building of which has a facade of 65 m with two wings projecting towards the sea at each end. A small cruciform building topped by a dome could be a pavilion or reception room. The remains of baths, substructures, a cistern and a wine press complete what appears to be an aristocratic residence connected to a small anchorage whose shape indicates a proasteion.25 At least in this case, the accuracy of this typological interpretation has been demonstrated. Two inscriptions, one discovered in the baths, and the other on a capital from the dwelling itself, give as owner gloriosissimus (endoxotatos) Illous. This is a perfectly appropriate term for the magister militum who was chief minister (patricius and magister officiorum) under Zeno. Therefore, we have both the identity of the owner of this proasteion and the date of its construction, the death of Illous in 488, providing a terminus.26

Private residences All aristocratic families had a city residence and/or a villa in the country, which bore some resemblance to the palaces already described. In Rome, the excavations of the Ospedale Militare, on the Celian Hill, have discovered some dwellings of particular interest.27

These include the house of Gaudentius (a neighbor and acquaintance of the Symmachii), whose name appears in an inscription in the triclinium of the complex. The statue of Antinous Casali, founded at the end of the 18th century. in a filling above the triclinium is associated with the Symmachii domus, and was probably placed in a niche in the triclinium. The domus in the center of the excavated area could be that of the Symmachii themselves. This magnificent house has a magnificent reception, to which an apse was added in the 4th century, and which is paved with opus sectile (Fig. 6). Between these houses and the church of S. Stefano Rotondo, under the Ospedaledell'Addolorata, excavations in the 17th and 18th centuries revealed the domus of Valerii. Although these excavations were largely unrecorded, ten statues or groups of statues were found, as well

25 Sodini (1997) 483.26 Feissel (1999).27 Pavolini et al. (1993); on the late residence in Rome cf. Guidobaldi (1986)

and (1999).

jean-pierre sodini36

such as a silver bottle with busts of Peter and Paul and a silver cup with an inscription called Valerii. These items along with a 4thc. bronze lamp in the form of a ship with Peter praying and Paul at the helm, provides the first evidence of the conversion of this old pagan family to Christianity.28

The house was burned during the Gothic Slaughter of 410, and no one could afford to buy it from Valerius Pinianus (Mella-nia's husband), due to the enormous size of the estate. In this case, archeology has allowed us to form an image of a Roman aristocratic house at the decisive moment of its owner's conversion to Christianity.29 As in previous cases, the presence of inscriptions on a pavement or on objects discovered nearby (as well as the documentary evidence) has been the decisive factor in identifying the owners. The scale and form of these luxurious residences indicate an architectural design for this type of domus, reminiscent of the description of Roman residences given by Olympiodorus of Thebes in his treatment of aristocratic houses in Rome: "Each of the great residences of Rome contains everything that is supposed to be found in a medium-sized city: a hippodrome, piazzas, temples, fountains and various types of bathing buildings ... A house is a city".30 Also outside the cities, aristocratic residences or proasteia were, to quote Ausonius ( III, 1,29), 'urbes in rure'. A study of the intra muros houses of Stobi, Histria, Aphrodisias, Apamea, Djemila, Thuburbo Majus or Ptolemais suggests that they all belonged to the wealthy. Andrew Poulter has painted a similar picture for 6th century cities, such as Nicopolisad Istrum, where the Byzantine city appears to have been occupied "by a few wealthy families and their dependents".31 Also in Egypt, aristocratic residences with a peristyle plan have been discovered in AbouMena, Marea, Kellis while Apion enjoyed it. a "palatial suburban villa".32

28 Brenk (1999).29 Other recently discovered dwellings, not mentioned by Guidobaldi, are i.a.

an example near the Arch of Constantine, on the Palatine Hill (for location see Claridge (1998) 120, fig. 50, n. 20; on the excavations see Hostetter (1994)), where many of the S. rooms fell into disuse in the first half of the 4th century, while the apsidal hall was converted into a funerary or religious building..

30 Blockley (1983) 152,31 Poulter (2000) 350-51,32 Alston (2002) 104-27.

archeology and late antique social structures 37

b. Episcopal Palaces In both East and West, the bishop emerged as a prominent political figure among the ruling classes.33 This is also indicated archaeologically by the occurrence of episcopal palaces as parts of extensive building complexes connected to cathedrals, which have been the subject of considerable interest after the 11th century. Christian Archaeological Congress in Lyon.34 At Byllis in Albania, current excavations reveal an enormous complex on the southern side of the cathedral with a series of annexes that, like in many other episcopal complexes (e.g. at Salona and Philippi), seem to have served an economic function. In other respects, the episcopal residence incorporates the same repertoire of architectural features found in other elite residences, such as triclinias, courtyards, gardens and fountains.

c. The houses of the pagan elites? Some years ago I refused to accept the identification of a series of magnificent houses in the Agora of Athens as the residences of the philosophers, with the possible exception of the house of Proclus.35

The subsequent discovery of the luxurious residence at Aphrodisias, which produced tondos of the ancient sages of Greece (Pindar, Pythagoras, Socrates, Aristotle and Apollonius) and also of their famous disciples, Alcibiades and Alexander, may well revive these attributions. There is, however, nothing to distinguish these houses from others of the generic aristocratic type that I have discussed above.36 Inscriptions on the pavements of the triclinia often allude to philosophy and ancient wisdom through the use of rather well-worn maxims of sophists and sophists. poets. Hanoune has also shown that some phrases may well reflect more personal choices, as for example in the houses of Cirta ('The just is his own law unto himself': Justus sibi lex est) and Bulla Regia ('Put your hope in yourself ').37

These phrases indicate that aristocratic pagans took common Christian maxims (God is your law, put your hope in God) and by changing a word, turned them into strong affirmations

33 There is a huge bibliography on this topic: see most recently Durliat (1996); Liebeschuetz (2001) and Brown (2002).

34 XIe CIAC (1989) where there are many articles on this subject in vol. I and II.

35 Sodini (1997) 463-65 (with bibliography on n. 120 and 121). 36 Sodini (1997) 474-77. 37 Sodini (1995) 181, n. 148.

jean-pierre sodini38

of the propagation of mankind against the Christian God. Balty has argued that a similar transformation is evident in the philosophical cycle found under the Cathedral of Apamea,38 where Christian imagery (even sometimes borrowed from pagan iconography, as in the example of the Apostolic College) was inverted to reflect pagan ideology. In some houses, such as in Carthage, the tone could be Christian or pagan (Omnia Dei sunt; agimur, non agimus), though using a vocabulary that reflected the same philosophical debates.

Regardless of the owner's identity or religious beliefs, all these dwellings are of a type present throughout the Mediterranean, in both urban and rural settings. They have the same repertoire of architectural devices, reflecting the same ideology, which mixed leisure and politics in an environment of halls and triclinias, fountains, nymphs and baths. They are decorated with mosaics, paintings and statues depicting mythological and philosophical themes (including portraits and the sentiments of philosophers and poets) or celebrating Hesiodic works and days with scenes of rural life and hunting. The statues, in particular, are indicative of the aristocrats' antiquarian taste, showing statues collected from the Hellenistic or earlier Roman periods, or sophisticated copies from the workshops of the Aphrodisiac sculptors and their followers. These were displayed in residences at various sites in Antioch, on the Esquiline in Rome, in Chiragan in Aquitania, and in El Ruedo in Spain.

Other status symbols

The elite, whether urban based, as was prevalent in the East and North Africa, or rural as in Britain, Gaul, Spain and Sicily, shared the same lifestyle and derived the majority of their income. from land. This lifestyle was also adopted by the ruling classes of the Vandals, Visigoths and Franks. At the end of the 5th century, a certain Steleco, whose name suggests German origin, was the owner of a villa that has been discovered near Mienne-Marboué (fig. 7).39 In the east, Ariobindus, a Goth who held high military office, acquired large estates and

38 Balty and Balty (1984); Balty (1989).39 Blanchard-Lemée (1983).

archeology and late antique social structures 39

possessed villages in the Cyrrhus area.40 Landowners, even in the Pars Orientis, may have lived on their estates and may have included not only the poorest of the curials but also the richer peasants. When one considers, for example, some of the huge houses in Djebel Zawiye (the Syrian limestone massif), at El Bara, and the magnificent mausolea with pyramidal roofs at Hass and El Bara (fig.8), dating to the 6th century to the conclusion that they belonged to rich landowners.41 Some of the gifts given to local churches also indicate the successful careers and lavish lifestyles enjoyed by some Syrians. About 578 and after 583 a certain Megas offered two silver jugs now in the Abegg Foundation-

Fig. 6. The triclinium is believed to be part of the residence of the Symmacchii on the Caelian Hill in Rome (Pavolini C. et al. (1993) 491 fig. 19).

40 Tompkins (1995).41 Sodini and Tate (1984).

jean-pierre sodini40

tion (E 37 and 38) (Fig. 9) and the Riha patent (E 35). These were found in the famous Kaper Koraon treasury, reconstructed some years ago by Marlia Mundell Mango.42 It was probably the same Megas that was recorded with stamps on three other pieces of the same treasure (E 33, 34, 36) , as being responsible for the quality of silver. Under Maurice he was elevated to the position of curator and was the recipient of a letter from Childebert (587–588). Why did he give all these pieces to the Kaper Koraon church? He may well have been a native of the village itself, or perhaps owned some land nearby. The same may be the case with other donors to Kaper Koraon, including Sergios, tribune and argyroprates (E 33, 34, 36) and themagistrianos Symeon (E1). John, the bishop of Kyrenia, another Syrian donor who contributed to the Phela treasure (E66), may have been born in Phela or had previously been a priest in the village.

Even in death, the elite maintained their superiority, which is evidenced by magnificent mausolea or by a tomb with a privileged position inside the church. Naturally, bishops, who were generally members of aristocratic families, would have the best seats inside the church or in the first rows of tombs near and around the apse, along with

42 Mundell Mango (1986).

Fig.7. Triclinium of Steleco's residence (Blanchard-Lemée M. (1983) pl. 47.2).

archeology and late antique social structures 41

priests and monks. However, senior members of the laity recommended by leading churchmen were also accepted, although saints such as St. of the tomb itself was also a sign of their social status. Likewise, although the late Roman dead were not buried in their full splendor or with their weapons as were the barbarians, they could be represented in this way on their sarcophagi, as with a recently published example found in the Theodosian walls (Fig. 10 ).44 The dead could also be represented in this way through paintings on the walls of tombs (such as in Thessalonica and Silistra).45 These representations are reminiscent of those on metal boxes, diptychs, on wall and floor mosaics, on manuscripts. , on triumphal monuments and on sculptures. Some metal dress ornaments are also symbolic of rank, such as the fibulae of

Fig. 8. El Bara, mausoleum (1974). For fig. 9-11 see plates.

43 Sodini (1986) 233.44 Deckers (1993); Deckers and Serdaroglu (1995).45 Kourkoutidou-Nicolaïdou (1997) pl. 37? Dimitrov (1962).

jean-pierre sodini42

the crossbow type, inscribed rings and the girdle or cingulum, many of which have been discovered in tombs. These items could sometimes be imperial gifts.46 The splendid jewelry and weapons in the tombs of barbarian princes, such as Childeric's at Tournai, or Gepiden's, Omharus, at Apahida I (fig. 11), probably fall into this category.47 These princes received titles of honor from the Roman emperors, giving an indication of the symbolic status of these implements.

Elite wealth is also well attested by silver vessels which were acquired for household use as well as being donated to churches, as we have already seen, and by the widespread appearance of gold jewelry, attributed to the Near East and dating from the 5th-7th centuries inclusive.48

In summary, therefore, it is clear that the elite elements of society maintain a dominant position in terms of their visibility in the archaeological record, as indeed they do in documentary sources.

The middle class

The urban middle class

Between the ever-present rich and the invisible poor, it is quite difficult to define a middle class. A family belonging to this group was certainly always vulnerable to the dangers of impoverishment. This was especially true of widows and orphans, helped by the church and ranked among the poor and needy, but who actually belonged to the fragile margin of the middle class, as is brilliantly shown by PeterBrown. The church recruited many of its priests and clerics, even bishops, from the middle classes, which included shopkeepers, artisans, low-ranking civil servants, soldiers and non-commissioned officers, as well as free peasants or tenants outside the church. cities. Many of these people could read and write, and well over half might have owned their houses or shops. They belonged to the populus orplebs, and some of them were members of the curia.49 They were among the 80,000 people in Constantinople who benefited from

46 Malte Johansen (1994).47 Périn (2000).48 Brun (1984); Baldini Lippolis (1999); Yeroulanou (1999).49 Sodini (1979) 118.

archeology and late antique social structures 43

annonae at the end of the 4th century. This number fell to 10,000 at the end of the 6th century. They were represented in the cities through various organisations. Most of them were members of guilds and participated in the activities of the circus factions.51 They contributed to the various munera, paid for the lighting of the streets or contributed to the building of the walls as in Constantinople or Tomis (where butchers participated in this activity).52 In economic terms, they were producers of goods that were exchanged in shops and markets and exported. For this reason, their skills were highly valued. Chosroes moved them from Antioch after the conquest of the city in 540 to stimulate Persian craftsmanship. He established them in a new city equipped with a hippodrome, giving a further indication of their involvement in factions. They may have represented a desirable source of wealth production in the eyes of a foreign sovereign. Their pride in their skills is evident from the mention of their jobs in their epitaphs, which are very specific about these activities, using expressions that indicate a high degree of craft specialization. These epitaphs were displayed in normal graves and are certainly not signs of poverty. There are examples of even richer graves which indicate a social level closer to the curiales. At Sardis the burial site of two high officers (ducenarii) of an imperial factory of shields and weapons established in the city in the early 4th century. reveals the aristocracy of artisans, organized along the same military hierarchy as those who worked in the imperial mints. The vaulted tomb of the second officer, Chrysanthios, who boasts of being a 'zographos' was painted by his own hand. Jewelers (argyropratai) were also very rich and were considered similar to bankers. Naucleroi (ship owners) had different social status depending on their level of success, and their popularity and wealth are often indicated by texts. They were associated with merchants (emporoi), whose activity was also lucrative. Artisans working with textiles were also among the wealthier citizens: the Tyre Necropolis has shown that eighty textile specialists constructed one or more monuments for themselves. They weren't the richest of the bulls, but they obviously belonged

50 Zuckermann (2000). For the plebs in Rome, see Purcell (1999).51 Zuckermann (2000) identifies the demoi of the factions with the demoi of

city, a view that has generally fallen out of favor following the work of Alan Cameron (1976). Liebeschuetz (2001) 203-20 does not connect guilds and circus factions (however see 277-78).

52 Sodini (1979) 107.

jean-pierre sodini44

to the 'moyenne bourgeoisie', whose highest rank was occupied by purple dyers.53

Substantial material traces survive regarding the professional activities of these middle classes. We have their stores in the big emboloi of the cities. Examples have been found in Scythopolis (BethShean) and in Beirut, where the numbers of the shops were marked on the pavement. The excavations at Sardis, which was burned by the Sassanids in 616, allow us (through Crawford's fine publication) to glimpse the interiors of these shops, which included restaurants, hardware stores, glass sellers and dyers (fig. 12). Shops engaged in the same activities tend to be clustered together, which reminds us of the presence of the flourishing artisan guilds, which seem to have had a particular influence at Sardis, where a violent strike in 459 ended with a sworn statement from the construction company. The finds in the shops of objects decorated with crosses or with menorahs probably indicate that the shopkeepers belonged to two communities: the Jews around the synagogue and the Christians further west. However, the economic situation of artisans and shopkeepers varied greatly in relation to the size and importance of the shop, the nature of the craft practiced (for example, politismes were inferior to argyropratai andchrysochooi54) and the location of the shop, whether in town and country. Artisans were also active in Egyptian cities, at least that is what the texts inform us, although archeology has not yet produced much evidence.55 There is doubt as to where the artisans and shopkeepers lived, although the texts mention that some did. in the room above the shop. The limited nature of this space suggests that if they lived in this location we can assume they were low income. In general, however, archaeological evidence for free craftsmen and traders is rather sparse and does not allow us to determine the position of the various crafts within a social hierarchy, for which we are obliged to use the evidence from texts, inscriptions and papyri. A large number of craftsmen are mentioned in funerary inscriptions, of which Korykos is of particular importance in this respect.56

53 Avramea (1991).54 Roueché (1995) 40-4155 Alston (2002) 128-84.56 Trombley (1987).

archeology and late antique social structures 45

Unfortunately, we do not have examples of the houses of the middle class and especially of the shopkeepers (probably due to a lack of suitable published excavations). As mentioned above, we are told by the sources that some lived above their shops, but this may not have been the general rule. Some public spaces were also gradually handed over to private residences as at Luni, and in a more coherent way at Cyrene, where the 3rd century agora was occupied by shops and a series of regularly built houses that may be connected to the shops.57 assume also reasonably that the members of the middle class may have lived outside the walls on the outskirts of the cities, as Poulter has noted in Nicopolis ad Istrum, where "the extramural area ... was densely inhabited, but by people who did not possessed the material wealth of intramural residents”.58

The rural middle class

Recent research in southern Greece has brought to light a series of 'bourgads' or villages built over the ruins of cities, such as Olympia59

or Gortyn, the former provincial capital of Crete. In these two cases the urban attributes of the site were rejected by the population, although Heraclius rebuilt an aqueduct at Gortyn. Small houses were clustered together and were inhabited by craftsmen (shown by glass furnaces, pottery workshops, metal workshops) and by farmers (shown by agricultural implements and equipment such as rotary kernels, olives and wine presses). At Gortyn there appears to be little difference between the mid-5th century houses. and those of this second half of the 7th century. At Olympia, however, the two phases give a different impression. The houses built before the earthquake of 551 (which probably destroyed the first late antique settlement) are composed of large rooms, although these are no longer arranged around a peristyle. After the earthquake, the houses were built between the second half of the 6th century. and the beginning of the 7th century (which saw Slavic occupation during the reign of Phocas) is smaller and is built of stone and clay (Fig. 13). They give the impression of impoverishment, but the inhabitants still belong to one

57 For Luni, se Ward Perkins (1981). For Cyrene se Sodini (1995) 189.58 Poulter (2000).59 Völling (2001).

jean-pierre sodini46

'middle class. Some social differentiation is shown in the tombs, but wealthy people are absent.

Archeology in Turkey has also revealed settlements that may be representative of middle-class housing. These are sometimes referred to as towns and include, for example, Mokissos and also Levisi. Part of this last place is gathered on the island of AghiosNikolaos and is known to have been a bishop's seat during the reign of Heraclius. These settlements are more commonly known as 'bourgades', of which good examples had been found at Alakisla, at Arif in Lycia and near Osmaniye in Caria).60 These latter settlements are reminiscent of the huge villages of Brad and el Bara in the limestone massif of northern Syria and those in the Negev. At least some may be the archaeological equivalent of the rich metrocomia near Antioch described by Libanius, such as Imma, or to the komai of the provinces of Palestine, Arabia and in Eastern Europe Dacia and Thracia.61

In the villages of the Syrian limestone massif and the basalt region we find a middle class of farmers whose lifestyle survives in two inscriptions from Apamene, recently republished by D. Feissel. These inscriptions record three cousins, two of whom were married to each other, a union not sanctioned by the church, who lived in a fortified house in a village. They learned enough to write two inscriptions in verse boasting of their family and their wealth, thereby providing a glimpse of the social and economic milieu of an Apamene peasantry proud of their origins and culture.62

The growing wealth of the Syrian villagers shows itself, as e.g. in Déhès, of changes in house construction. In the late 5th century and in the first half of the 6th century, carefully laid out square blocks were used, which involved the use of teams of specialist stoneworkers who used cranes and displayed considerable experience in stone cutting and sculpture. The peasants were no longer the builders of their houses, but instead paid specialized workers, either in money or in kind. We can also see differences in the size of the houses, which vary between the larger houses with six to ten rooms and smaller houses with five rooms or less. Can we go a step further and identify

60 Sodini (1997) 479; Ruggieri and Giordano (1996); Dagron (1979); Gatier (1994).

61 Dagron (1979).62 Feissel (1998).

archeology and late antique social structures 47

tify, as Tchalenko did for Behyo, 'villas' of the free peasants, farms of the tenants and the hoofs of the workers?63 The answer must be negative. When we try to determine whether the village was occupied by free peasants or by tenants, the same difficulties arise. Although we can assume that the largest houses may not belong to Totenes, for the rest it is difficult to draw a line between the houses of free peasants and tenants. Tchalenko tried to do this by using the village names - those beginning with beit/ba/ (greekepoikion) and those of Kafr (Greek komé) - to distinguish between the villages belonging to a single landowner and those inhabited by free people. However, these differences are blurred by the subsequent development of the villages. Some can e.g. having refused to pay taxes to their landlord and acquired freedom or became clients of a high-ranking army officer, while others, at the opposite extreme, may have given up their freedom to become the possession of a single landlord without adapting their names to new situation. In terms of their layout, there is not much difference between Baziher and a village like Dehes, although the former is surrounded by a wall and does not develop between the 4th and 6th c.64 The study of the development of the early Christian village , whether from a single property (the 'Bamuqqa type') or from an earlier village, depends primarily on archaeological evidence, although this is unlikely to resolve the issue of ownership of the latter.

It remains clear that there was a significant degree of social mobility in the 6th century. Syria and other regions. The wealthy, the bishops, the monks and the peasants all depended on each other for their welfare and that of their communities. Christianity in particular was a unifying force in villages, and not just those in Syria.

In the Negev, the development of the hinterland was due to considerable demographic pressure, and 'bourgads' and even cities flourished in the early Christian period. Their resources may not derive mainly from pilgrimage or from caravan traffic, but rather from cultivation. Vineyards were important and are well attested archaeologically by the holes after each vine and by the wine presses, as well as by texts such as the Life of Hilarion. This wine

63 Tchalenko (1953-58).64 Tate (1992) 213, 265-66, 283-84.

jean-pierre sodini48

was transported to Gaza and exported in the famous Gaza amphora to all parts of the Mediterranean, and was enjoyed at the court of the Merovingian kings as well as by the Hemmaberg clergy.65

But the archives found in Nessana and more recently in Petra allow us to gain a more refined view of life in these agrarian cities. In terms of land ownership, the land in Nessana was owned by freeholders (with "no hint of colonus or emphyteusis, except perhaps in three documents").66 At Petra, three brothers who shared inherited properties seem to lead a comfortable life, for example. The family had more than 5427 acres or 54.27 ha (134 acresx 40.5), which is about five times the land needed to feed a poor family. The fields were either 'leased' or 'farmed' by farm workers, the other being more dependent on the landlord.67

In Egypt and in Proconsularis we have evidence of potters working for landlords of large estates. Michael Mackensen tried to explain the presence of these terra sigillata pottery workshops in a rural environment from the beginning of the 4th century. to the end of the 7th He suggested that they belonged to landowners who also had at their command the means for the pottery's regional and trans-Mediterranean distribution. The potters would therefore be managers of the landlord's workshops. There is certainly evidence for such an arrangement in Egypt in the 3rd century, where contracts have been found between large landowners and amphora manufacturers.68

However, does this allow us to propose a similar situation for the sigillata potters of rural North Africa? Likewise, what does this tell us about the social status of potters working on the fringes of cities, in Africa and elsewhere, especially in the East? It remains difficult to define the social context of these crafts using archaeological evidence in isolation.

The society that emerged from the documents found in the earlier caves of Palestine III or from Egyptian papyri was therefore apparently a peaceful society where civil law was respected, with soldiers and civilians, landowners and peasants coexisting with each other.

65 Ladstätter (2002).66 Kraemer (1958).67 Koenen (1996) 183-84; Frösén (2002).68 Cockle (1981).

archeology and late antique social structures 49

The poor

As already mentioned, the middle classes were particularly vulnerable to economic hardship and could quickly become poor. For example, the weight of taxation can threaten the income of craftsmen. The Syrian Life of Symeon Stylites tells of many interventions by the saint in cases where the rich oppressed widows, orphans and the poor. One is of particular interest because it gives us the precise circumstances surrounding Symeon's intervention. In this case, a curialis from Antioch, who for one year (perhaps as pater civitatis) had received the city's revenue, decided to triple the annual tax the city received from the 'poor who dyed their skin red'.69 Among them came three hundred . to the saint who invain asked the curialis to return to the customary tax. The counselor refused and fell ill. He asked "all his villages and ... the priests in them" to persuade the saint to pray for him. In this case, however, God was inflexible and he died. We can also recall the joy described in the life of Joshua the Stylist when Anastasius suppressed the chrysargyron that had condemned so many artisans and shopkeepers to poverty. In the countryside too, taxes and rents were a permanent threat. Theodoret of Cyrus, as we have already mentioned, wrote to Ariobindus, the owner of the village of Sergitheum, to request that the farmers of this village, who could not provide the amount of olive oil requested as rent, be released from their obligations for one year. An inscription found on a mosaic at KefrZelih in Apamene has revealed part of a local chronicle which refers to a frost that killed the olive trees on 27 January 499, which may also have reduced the peasants to poverty. The most famous account of disasters combining plague, locusts and famine took place in Edessa in 499-502 and provides clear evidence of how easily a famine can break out that shakes the entire society to its foundations as e.g. 504-506.70

If we turn from the 'paupérisables' to the 'real' poor, we have little archaeological evidence, although they are ever present in the writings of the great leaders of the Church. Basil of Caesarea, Severus

69 Doran (1992) 135-137 §56. Their workshops already existed in the 1st AD. on the right bank of the Euphrates, in the direction of Ammanus, attested by the construction of the canal of the Fulani; see Feissel (1985). The generosity of a wealthy citizen paying wages to unemployed artisans also shows their vulnerability: IGLS V, No. 1999 (Mango (1986) 28-29).

70 Trombley and Watt (2000) para. 38-44; Garnsey (1996) 29-33; Tate (1989) 112-14.

jean-pierre sodini50

of Antioch and John the Almsgiver all insist unceasingly on the love and care of the poor. A pauper is even represented on the frame of the Megalopsychia mosaic found at Yakto near Antioch (fig. 14), which depicts familiar scenes and characters from the city. Texts and inscriptions also often mention hospices, hospitals and hostels for the poor, called xenodochia and ptochotropheia. An example, which celebrates the construction of a hospital in 511, was found in Firki and is now housed in the Museum of Ma'aret en Noman.71

Conclusion

Where there is abundant archaeological material related to the 7th and 8th centuries, the impression is (for example in Apamea, Gerasa and Scythopolis) that the class of possessors disappeared. Their houses were divided into smaller apartments or farms with agricultural equipment (such as the olive press in the 'Maison de l'Huilerie' in Salamisin Cyprus). However, there is an increasing number of shops and workshops (such as in Apamea's episkopeion), which indicates a certain 'triumph' of the middle class. Written sources may suggest a different picture, with ancient Christian families in Damascus and

Fig. 12. Shops at Sardis, recreation of Restaurant E1 (Stephens Crawford J. (1990) fig. 35).

71 Brown (2000) 35.

archeology and late antique social structures 51

Fig. 13. Plan and possible reconstruction of a house dated at the end of the 6th c; – Beginning of 7. (Völling (2001) 318 fig. 2). For fig. 14 see plates.

jean-pierre sodini52

in Edessa apparently remained influential, although this later period is beyond the scope of this article.

In conclusion, social structures are difficult to trace only through excavation. In some respects, the recognition of social structures is a by-product of archaeology, in that the real task is to interpret the development of the archaeological deposits and structures within their own particular context. We may notice different standards of living, some higher than others, in contemporary and successive contexts, although their significance is often difficult to interpret. In a normal excavation, social contrasts are not emphasized, objects do not in themselves symbolize wealth or poverty, and inscriptions are rare.

For the purposes of this paper I have had to draw on the entire corpus of Late Roman archeology and emphasize particular cases from which conclusions can be drawn, almost always by means of epigraphy. We discover palaces of famous emperors, governors or ministers, and on this basis we feel justified in attributing residences that have more or less the same features to this same aristocratic milieu. Similarly, jewelry and dress fittings can also indicate membership of the upper class. From this lifestyle we can suggest the origin of this wealth, which may derive from properties given to tenants or farmers or other activities whose nature can only be guessed. In relation to the merchants and artisans, the few examples we have show little evidence of luxury, and I have tried to interpret their social structure in accordance with what we know of the social status assigned to the various crafts and the new definitions given by Zuckermann. For the farmers, the situation is more complex because we do not have access to their relationship with the land. We may suspect that the difference between a small freeholder and a farmer may not have been marked, while the size and layout of the houses may give indications of social status and allow us to perceive a hierarchy among the peasantry. But we need criteria based on known landowners, as is the case at Nessana (because of the papyri evidence), which immediately gives us a much deeper understanding of these issues, or so we think.

There is still a gap between the material remains of anonymous individuals and their interpretation in relation to social structures. We need to address this problem using various hypotheses, which are increasingly based on statistics and models. However, our primary concern must always be to construct our theories of social structure using well-established facts.

archeology and late antique social structures 53

Bibliography

Alston R. (2002) The city in Roman and Byzantine Egypt (London and New York 2002) Arce J. (1997) "Emperadores, palacios y villae", Ant Tard 5 (1997) 293-302 Avramea A. (1991) "Crafts and textile trade in the Lower Eastern Empire-

numbers" in cultural and commercial exchanges between the Orient and the Greek world (Athens 1991) 23-29

Augenti A. (1996) The Palatine in the Middle Ages. Archeology and Topography, VI-XIII Centuries (Rome 1996)

Baldini Lippolis I. (1999) Goldsmithing in the Empire of Constantinople between the 4th and 7th Centuries (Bari1999)

Balty J. (1989) "Iconography and pagan reaction", in Blandinger P. Lévêque I (Besançon-Paris 1988) 17-32 (reprinted in Balty (1995) 275-89)

——— (1995) Ancient Mosaics from the Near East (Besançon-Paris 1995)Balty J. and J.-Ch. Balty (1984) "A philosophical program under the cathedral

of Apamea: The Neo-Platonic Ensemble of the Emperor Julian”, in Text and Image, Proceedings of the International Conference of Chantilly (13-15 Oct. 1982) (Paris 1984) 167-76 (reprinted in Balty (1995) 265-73)

Balzaretti R. (1994) "The Curtis, the archeology of sites of power", i La storiadell'alto medioevo italiano (VIe—Xe) alla luce dell'archeologia, red. R. Francovich og G. Noyé (Florence 1994) 99-108

Balmelle C. and Van Ossel P. (2001) “From Trier to Bordeaux. The Mark of the Elites in the Campaigns of Roman Gaul in the 4th and 5th Centuries”, in The Campaigns of Gaul in Late Antiquity, 4th Colloquium of the AGER Association, Montpellier 1998, ed. P. Ouzoulias, C. Pellecuer, C. Raynaud, P. van Ossel and P. Garmy (APDCA Antibes 2001) 533-52

Bardill J. (1999) "The Great Palace of the Byzantine Emperors and the Walkertrust Excavations", JRA 12 (1999) 216-30

Barnes T. (1996) "Emperors, panegyrics, prefects, provinces and palaces (284-317)", JRA 9 (1996) 552

Blanchard-Lemée M. (1983) "Late mosaics and the survival of villas in Gaulemmoyenne during the Merovingian period", in Mosaics, collection in tribute to H. Stern (Paris 1983) 75-80

Blockley, R. (1983) The Fragmentary Classicising Historians of the Later Roman Empire.Eunapius, Olympiodorus, Priscus og Malchus (ARCA Classical and Medieval Texts,Papers and Monographs 10) (Liverpool 1983)

Brenk B. (1999) "The Christianization of the Domus dei Valerii on the Celian Hill", i The Transformations of Urbs Roma in Late Antiquity, red. W. V. Harris (Journal of RomanArchaeology Supplementary Series 33) (Portsmouth 1999) 69-84

Brown K. (1984) Guldbrystkæden from the tidlige period in the Roman-Germanic Central Museum (Mainz 1984)

Brown P. (2002) Poverty and Leadership in the Later Roman Empire (The Menahem SternJerusalem Lectures) (Hanover og London 2002)

Burrell B. (1996) "Palace to Praetorium: The Romanization of Caesarea", in CaesareaMaritima. A Retrospective After Two Years, ed. A. Raban and K. G. Holum (Leiden-New York-Cologne 1996) 228-47

Cameron, Alan (1976) Circus Factions (Oxford 1976) Carrié J.-M. (1997) "The Social Arithmetic of the Agrarian Economy. Land Price,

Land Rent and Grain Prices in Roman-Byzantine Egypt” in Economiantique. Prices and price formation in ancient economies edd J. Andreau, P.Briant and R. Descat (Archaeology and History Interviews, EAHSBC 3) (Saint-Bertrand-de-Comminges 1997) 121-46

Claridge, A. (1998) Rom (Oxford Archaeological Guides) (Oxford 1998)

jean-pierre sodini54

Cockle H. (1981) "Pottery production in Roman Egypt: a new papyrus" JRS 71(1981) 87-97

Cullhed M. (1994) conservator Urbis suae, Studies in the Politics and Propaganda of the Emperor Maxentius (Stockholm 1994);

XIth CIAC (1989): Proceedings of the XIth int. for Christian Archeology (Lyon, Vienna, Geneva and Aosta, 1986) (Vatican State-Rome 1989)

Dagron G. (1979) "Between village and city: the 4th-7th century countryside in the East" Koinônia 3 (1979) 29-52

Deckers J.G. (1993) "The Hypogeum at Silivri Kapi in Istanbul", JbAC 36 (1993)140-63

Deckers J.G. og Serdaroglu U. (1995) "The Hypogeum at Silivri Kapi in Istanbul", Dokumenter fra XII. int. cong. kristen Archeologie Bonn 1991 (Vaticanstaden-Münster1995) 674-81

Delmaire R. (1996) “Cities and taxation in the Lower Empire. On the role of the curiale in the collection of taxes”, in The End of the Ancient City and the Beginning of the Medieval City, ed. C. Lepelley (Bari 1996) 59-70

Deichmann F.W. (1982) "Western building technology in Roman and Rhomian East", RömMitt 86 (1979) 473-527 genoptrykt i Rom, Ravenna, Konstantinopel, Nahe Osten (Wiesbaden 1982) 712-66

Dimitrov D. P. (1962) "The Decorative System and the Date of the Wall Paintings of the Ancient Tomb of Silistra", CahArch 12 (1962) 35-52

Di Vita A. (2000) Gortina 5.1 (Padua 2000) Doran R. (1992) The Lives of Simeon Stylites (Kalamazoo 1992) Durliat J. (1996) "Bishop and Municipal Administration in the Seventh Century", i La fin

of the ancient city and the beginning of the medieval city, ed. C. Lepelley (Bari 1996) 273-86

Feissel D. (1985) "Two Lists of Districts in Antioch Forced to Dig a Canal (73-74 A.D.)", Syria 62 (1985) 77-103

——— (1998) “Two epigrams from Apamene and the praise of endogamy in a sixth-century Syrian family”, in AETOS, Studies in honor of Cyril Mango, edd.I. Sevcenko and I. Hutter (Stuttgart and Leipzig 1998) 116-36

——— (1999) "Two Great Isaurian Families of the 5th Century According to Inscriptions from Cilicia Trachea", MiChA 5 (1999) 9-17

Festugière A. J. (ed. and translation) (1970) La vie de Théodore de Sykéon I-II (Subs. Hag.48) (Brussels 1970)

Frösén J. (2002) "Petra papyrus: information and significance", in Petra. A Forgotten and Rediscovered city. J. Frösén and Z. T. Fiema (Helsinki 2002) 18-24.

Garnsey P. (1996) Famine and Subsistence in the Graeco-Roman World (Paris 1996) Gatier P. L. (1994) "Villages of the proto-Byzantine Near East (4th-7th c.). Survey

régionale", in The Byzantine and Early Islamic Near East. II Agricultural and Settlement Patterns, eds G.R. King and A. Cameron (Princeton 1994) 17-48

——— (2001) “Big or Little Other Syria? For a Historical Geography of Proto-Byzantine Inner Syria”, in Conquest of the Steppe and Land Acquisition on the Dry Margins of the Fertile Crescent, ed. B. Geyer (TMO 36) (Lyon2001) 91-109

Guidobaldi F. (1986) "Single-Family Dwellings in Late Antique Rome", in Roman Society and the Late Antique Empire; Rome: Economic Policy, Urban Landscape II, ed. A. Giardina (Rome 1986) 165-237

——— (1999) “The late antique domus in Rome as 'sensors' of cultural and social transformations”, in The Transformations of Urbs Roma in Late Antiquity, ed.W. V. Harris (Journal of Roman Archeology Supplementary Series 33) (Portsmouth 1999) 53-68

Heijmans M. (1998) "The 'Palais de la Trouille' in Arles: Imperial Palace or Prefect's Palace?", Ant Tard 6 (1998) 209-23

archeology and late antique social structures 55

Hidalgo R. (1996) Public space and private space in the Palatine complex of Cercadilla (Cordoba): the central classroom and the baths (Sevilla 1996)

Holum K.G. (1996) "The Survival of the Bouleutics Class at Caesarea in Late Antiquity", in Caesarea Maritima. A look back after two millennia, ed. A. Rabanand K.G. Holum (Leiden-New York-Cologne 1996) 615-27

Hostetter E. (1994) "A late Roman Domus with apsidal hall on the Palatine 1989-1991 seasons", i Rome Papers (Journal of Roman Archaeology Supplementary Series 11) (Ann Arbor 1994) 131-81

Keil J. and Wilhelm A. (1931) Monumenta Asiae Minoris Antiqua, III: Denkmäler ausdem Rauhen Kilikien (Manchester 1931)

Koenen L. (1996) "The carbonized archive from Petra" JRA 9 (1996) 177-88Kourkoutidou-Nicolaïdou E. (1997) "From the Elysian Fields to the Christian

Paradise”, i The Transformation of the Roman World AD 400-900, red. L. Websterand M. Brown (London 1997) 128-42

Kraemer C. J. (1958) Ikke-litterære papyrus (Excavations at Nessana, volume III) (Princeton1958)

Ladstätter S. (2002) "The pottery of Hemmaberg", in Actes du VIIIe Congrès de l'AICM2Thessalonique 1999, ed. Ch. Bakirtzis (Thessalonica 2002) (in press).

Laniado A. (1997) “Bouleutai et politeuomenoi”, ChrÉg 72 (1997) 130-44——— (2002) Research on municipal notables in the proto-Byzantine East (Paris 2002) Lavan L. (1999) “The residences of late antique governors: a gazetteer”, Ant Tard

7 (1999) 135-64——— (2001) "The praetoria of civil guvernors in Late Antiquity", i Recent Re-

søg i Late-Antique Urbanism, red. L. Lavan (Journal of Roman Archaeology Supplementary Series 42) (Portsmouth 2001) 39-56

Liebeschuetz W. (1996) "Civic finance in the Byzantine Period", ByzZeit 89 (1996)389-408

——— (2001) The Decline and Fall of the Roman City (Oxford 2001) Mango C. (1986) "Honorary Epigrams, Statues and Portraits in Byzantium", in

Tribute to N. Svoronos (Rethymno 1986) 23-25 ​​(genoptrykt i C. Mango Studieson Constantinople (Aldershot 1993) IX).

Malte Johansen I. (1994) "Rings, fibulae and buckles with imperial portraits and inscriptions", JRA 7 (1994) 223-42

Mitchell S. (1993) Anatolia, Land, Men, and Gods in Asia Minor, vol. II, The Rise of the Church (Oxford 1993)

Mundell Mango M. (1986) Silver from Early Byzantium. The Kaper Koraon and Related Treasures (Baltimore 1986)

Netzer E. (1996) "Fortified Palace", in Caesarea Maritima. A Retrospective AfterTwo Millennia, red. A. Raban and K. G. Holum (Leiden-New York-Cologne1996) 193-207

Neri V. (1998) I marginali nell’Occidente tardoantico (Bari 1998) Patlagean E. (1977) Economic Poverty and Social Poverty (Paris 1977) Patrich J. (1999) “The Warehouse Complex and the Palace of the Governor”, ​​in Caesarea Papers

2, udg. K. G. Holum, A. Raban og J. Patrich (Journal of Roman Archae-ology Supplementary Series 35) (Portsmouth 1999) 70-107

——— (2001) "Urban space in Caesarea Maritima, Israel" in Urban Centres and Rural Contexts in Late Antiquity, red. Th. S. Burns og J. W. Eadie (Michigan 2001) 77-110

Pavolini C. et al. (1993) "The Ancient Topography of the Celian Summit. The Excavations at the Military Hospital (1987-1992)", RömMitt 100 (1993) 443-505

Périn P. (2000) The Gold of the Barbary Princes. From the Caucasus to Gaul, 5th century AD. J.-C., Museum of National Antiquities, château Saint-Germain-en-Laye (Sept. 2000—Jan. 2001), Reiss-Museum Mannheim (Feb.-June 2001) Exhibition catalog (Paris 2000).

jean-pierre sodini56

Popovic I. and M. Tomovic (1998) "Gold Jewelery from the Imperial Mauso-leum at Sarkamen (Eastern Serbia)", Ant Tard 6 (1998) 287-31

Poulter, A. (2000) "The Roman to Byzantine transition in the Balkans: Preliminary results on Nicopolis and its hinterland", JRA 13 (2000) 346-58

Purcell N. (1999) "The populace of Rome in late antiquity", i The Transformationsof Urbs Roma in Late Antiquity, red. W. V. Harris (Journal of Roman Archae-ology Supplementary Series 35) (Portsmouth 1999) 135-61

Roueché C. (1995) "Aurarii in the Auditoria", ZPE 105 (1995) 37-70. Rougé J. (1990) "Aspects of poverty and its remedies in the 4th-5th centuries", i

Proceedings of the Constantinian Roman Academy, VII International Conference (Perugia 1990) 227-48

Ruggieri V. and Giordano F. (1996) "A Byzantine City at the Carian Site of Alakisla. Preliminary Report" OCP 62 (1996) 53-88

Sodini J.-P. (1979) "Urban Crafts in the Early Christian Era", Ktema 4 (1979)71-119

Sodini J.-P., Tate G., Bavant B., J.-L. Biscop and Orssaud D. (1980) "Dehes, Campaigns I-III", Syria 57 (1980) 1-308

Sodini J.-P. and G. Tate (1984) "Houses of the Roman and Byzantine Periods (2nd–6th Centuries) of the Limestone Massif of Northern Syria: Typological Study", in Actes du Colloquium Apamea of ​​Syria, Review of Archaeological Research 1973- 1979, Brussels 1980, ed. J. Balty (Excavations of Apamea in Syria, Miscellanea, fasc. 13) (Brussels 1984) 377-430

Sodini J.-P. (1986) "The 'Privileged Tombs' of the Christian East (with the exception of Diocesan Egypt)", in Privileged Burial from the 4th to 7th Centuries in the West. Proceedings of the Créteil Conference 1984, ed. Y. Duval and J.-Ch. Picard (Paris 1986) 233-43

——— (1995) “Habitat de l'antiquité tardive (1)”, Topoi 5 (1995) 151-218——— (1997) “Habitat de l'antiquité tardive (2)”, Topoi 7 (1997) 435 -577Srejovic D. (1993) Roman Imperial Cities and Palaces in Serbia (Belgade 1993)Stephens Crawford J. (1990) The Byzantine Shops at Sardis (Archaeological Exploration

af Sardes. Monografi 9) (Cambridge, Mass. og London 1990)Tate G. (1989) "Syria in the Byzantine Period: A Synthesis Essay" in Archaeology and

Histories of Syria II, red. J.-M. Denzer and W. Orthmann (Saarbrücken 1989) 97-116

——— (1992) The Campaigns in Northern Syria, from the 2nd to the 7th Centuries, I (Paris 1992) Tchalenko G. (1953-8) Ancient Villages in Northern Syria: The Bélus Massif at the Time

romaine, 3 vols. (Paris 1953-8) Tompkins I. G. (1995) 'Problems of dating and relevance in some letters from

Theodoret of Cyrrhus", Byzantion 65 (1995) 176-95Trombley F. R. (1987) "Korykos in Cilicia Trachis: the economy of a small littoral

by i senantikken (5th-6th century)—a précis", The Ancient History Bulletin 1.1(1987) 16-23

Trombley F. R. and J. W. Watt (2000) trans. The Chronicle of Pseudo-Joshua the Stylite(Liverpool Translation Texts for Historians 32) (Liverpool 2000)

Van Ossel P. og P. Ouzoulias (2000) "Rural settlement economy in Northern Gaulin the Late Empire: an overview and assessment", JRA 13 (2000) 133-60

Völling T. M. (2001) "The Last Christian Greeks and the First Pagan Slavs", in The Dark Centuries of Byzantium (Athens 2001) 303-323

Ward-Perkins B. (1981) "Two Byzantine Houses at Luni", PBSR 49 (1981) 91-98 Yeroulanou A. (1999) Diatrita (Athens 1999) Zuckerman C. (2000) "Circus, money and people. On registration .

of the Lower Empire”, REByz 58 (2000) 69-96

the construction of identities in post-Roman Albania 57

THE CONSTRUCTION OF IDENTITIES IN POST-ROMAN ALBANIA

WILLIAM BOWDEN

Abstract

Late Antique social structures in the provinces of Epirus Vetus and EpirusNova, as represented by urban centers, ecclesiastical authority, and Roman material culture, largely disappeared during the first half of the 7th century. Around this time, the practice of furnished burial was adopted by parts of the population. This transition has traditionally been interpreted as an expression of the ethnic identity of an indigenous population. However, this paper argues that these post-Roman cemeteries are a reflection of the more localized and fluid social structures that emerged in post-Roman times. Epirus rather than an attempt to construct or maintain ethnicities.

Late Antiquity Ethnicity and National Archaeologies

in the southern Balkans

In the late 6th and early 7th centuries, tribes of 'Slavonians' reportedly overran, among other areas, the provinces of Epirus Nova and Epirus Vetus, forcing a large part of the existing population to migrate elsewhere. The exact nature of this incursion and the extent of the subsequent 'Slavic' occupation have been matters of contention since Greece and Albania were established as nation-states in the 19th century. The demands of Albanian and Greek nationalism required the populations of both countries to be ethnically separate from those that surrounded them. them. In both countries, a process was envisaged in which successive incoming groups were assimilated into a dominant autochthonous culture, which thereby preserved its ethnic integrity. This gave rise to a very particular brand of cultural-historical archaeology, which played a key role in the establishment of ethnogenesis myths in both countries.1

1 In writing this paper I have greatly benefited from discussion with my colleague

L. Lavan og W. Bowden (red.) Theory and Practice in Late Antique Archaeology (Late Antique Archaeology 1 2003) (Leiden), s. 57-78

william bowden58

In both Greece and Albania this was of particular importance in relation to late antiquity and the post-Roman period. In Greece, archeology has so far played a largely passive role, providing illustrative material for a narrative history whose basic parameters are widely accepted by Greek citizens. Late Antiquity has not figured in this narrative to any great extent, and therefore, with the exception of early Christian and Byzantine churches, the archeology of the period has until recently been generally ignored in Greece.2 In Albania, the reverse is the case. During the communist rule of Enver Hoxha, who ruled Albania from 1945 until his death in 1985, Late Antiquity archeology was used to support a very specific narrative history of the Albanian people. Archaeological research also tended to reflect modern national borders. As a consequence, 'post-Roman Albania', although it may appear to be a very artificial construction (in the same way as 'Roman Norfolk'), is therefore a very real concept in Albanian archaeology.

The basic thesis of Albanian archeology has been relatively simple, namely one which asserts that the Albanians were the direct descendants of the Illyrians, a loose confederation of tribes which constituted the pre-Roman population of the western Balkans. According to the Albanian model, they had been subjugated by the slave-owning Romans but had never been fully assimilated into the Roman world and had consequently remained ethnically independent while under the Roman yoke. By the time the Roman 'occupation' ended, the Illyrians had reasserted their independence, reclaimed the sites of their former settlements, and revived pre-Roman forms of material culture and burial. Despite the historically attested presence of Slavic settlers in the region, members of this proto-Albanian

leagues from the Butrint project, and in particular the work of our Albanian colleague Etleva Nallbani, whose forthcoming doctoral dissertation promises to be a definitive statement on the Komani cemeteries. I am very grateful to Etleva for allowing me to quote her research on two important issues and for pointing out some of my mistakes. However, she is exempt from responsibility for those who remain for the opinions expressed. I am also very grateful to Neil Christie for his persuasive critique of an earlier draft of the paper and to James Crow for suggesting additional references. I am also grateful to the German Archaeological Institute in Athens for permission to use photographs from Heinrich Bulle's excavations in Corfu.

2 This situation is now changing and the late antique phases of a number of major sites in Greece are being reassessed. A new generation of Albanian scholars has likewise rejected the nationalist stance of their predecessors. For a more detailed discussion of the effects of Albanian and Greek nationalism on Late Antique archaeology, see Bowden (forthcoming), especially Chapter 2.

the construction of identities in post-Roman Albania 59

or the Arbër culture was considered to maintain an ethnic independence, occupying highland areas while the Slavs remained in the valleys.3

Arbër / Komani-Kruja-culture

The Albanian case for the emergence of a post-Roman independent Arbër population rested on a number of sites in central and northern Albania, including an extraordinary group of cemeteries, the largest of which were at Komani and Kruja (Fig. 1, see plates ). They are often referred to by Yugoslav archaeologists as representing the Komani-Kruja culture, although Albanian archaeologists generally refer to them in terms of the Arbër or proto-Albanian culture.4

These cemeteries were located in remote mountain areas, sometimes close to the fortified hilltop sites occupied in both the prehistoric and post-Roman periods. The burials were usually placed in tombs built from slabs of limestone, although occasionally a tent-like roof was taken, reminiscent of Roman 'cappucino' burials. Excavation reports tend to suggest that the tombs stood proud of the ground, although this seems to reflect the excavation technology. niques adopted rather than the way the tombs were constructed. This style of tomb construction, including the use of limestone slabs and pitched covers, is very similar to late Roman tombs in Albania.5

The occupants of the burials were buried with an extraordinary collection of objects. These could include gilded disk brooches and crescent-shaped gold earrings, which were probably imported from Byzantine Sicily, as well as local variants of the same items.

3 Bowden (2002) See also Wilkes (1992). Explicit statements about the Albanian model can be found in Anamali (1980); Anamali (1982). On the slaves of the Balkans Avramea (1997) and generally Curta (2001).

4 I will use the term Komani-Kruja throughout this article, although this does not imply adherence to the arguments of one group or the other. There is a considerable amount of literature on the cemeteries and associated buildings. For Komani(Kalasë see Dalmaces) see Spahiu (1971); (1980a). For Kruja see Anamali and Spahiu (1980). Other important cemeteries include Bukël (Anamali (1971)), Shurdah (Komata (1980)) and Lissus or Lezhë (Prendi (1980)). Popovi´ (1984) provides a useful summary of the evidence for the Komani-Kruja culture, and a further overview is provided by Anamali (1993). On the Aphiona cemetery in Corfu, see Bulle (1934) 213-40 and Bowden (forthcoming).

5 E.g. hos Butrint (Hodges, Saraçi et al. (1997) 228-29) 6 Examples of Anamali (1993).

william bowden60

away in Hungary, southern Italy, Constantinople, and the Crimea.7 Large, rectangular bronze fibulae are also found of a type found in the Danube provinces, Dalmatia, Thessalonica, and northern Italy.8 Bronze and iron pendants are also common, similar to examples that known from the Avar cemeteries. These appear to have functioned as a kind of chatelaine from which other objects were suspended.9 Other objects are paralleled in Avar and Slavic contexts as well as in Frankish and Lombard contexts, including the characteristic ray brooches of the period.10 Glass necklaces (sometimes including millefiori- pearls) is common, again of a type known from numerous contexts, for example in Northern Italy and Germany.11 Bronze rings, many of which were decorated with Christian cryptograms, were also often present.12

The graves also contain pottery, usually two-handled bottles and trilobal jugs made of a hard burnt brown fabric and painted with broad red stripes, similar to examples found in southern Italian cemeteries.13 With the exception of examples from the Aphiona cemetery in Corfu, finds of glass vessels are very rare.14 The excavators associate clothing and jewelry with female graves, while weapons are associated with male graves. Fibulae are common to both sexes; it is unclear whether this observation is based on real osteological differences between the sexes, or whether it has become a self-perpetuating axiomatic truth. The weapons are commonly axes, accompanied by arrowheads and short daggers. Spearheads and swords are unique, although they appear at Komani in a form that Popovi suggests finds parallels in Merovingian and Lombard contexts.15

7 Anamali (1972) 137; Popovi´ (1984) 219. For southern Italian examples see D'Angela (1982).

8 Se Popovi' (1984) 217-18.9 E.g. Fader (1980) Tab. XXII, 1-2 for Lezha. 10 E.g. Anamali og Spahiu (1980) Tab. VII, 11-12 til strålebrocher fra

Mug. See also Popovi' (1984) 221-22. There is some evidence that material from the cemeteries perceived as "Slavic" was deliberately suppressed during the communist period, although it is impossible to quantify to what extent this happened (E. Nallbani pers. comm.).

11 E.g. Anamali og Spahiu (1980) 75 for Kruja. For paralleller i Lombard-kontekster i Italien, that e.g. Tagliaferri (1990) 402-403.

12 Also Spahiu (1971) Tab. V, 12,13 F.eg. Anamali og Spahiu (1980) Tab. V, VI. For Italian examples of Arthur

(1998) with references.14 See Bulle (1934) 213-40.15 Popovi´ (1984) 221-22. For the division of the grave goods along gender lines,

that Spahiu (1980a) 40.

the construction of identities in post-Roman Albania 61

Despite the presence of so many grave goods, it is usually claimed that the tomb occupants are Christian, as evidenced by the Christian cryptograms on the rings and by the fact that an E-W orientation was sometimes, but not always, observed. This last point is ambiguous. Although in some cases (for example at Aphiona) considerable effort appears to have been made to maintain an E orientation, at Kruja 14 out of 28 graves had a N-S orientation, while 8 additional graves were oriented NW-SE . At Lezhë, 10 were oriented E-W, while another 24 were oriented NW-SE. At Dalmaces (Komani) the results are even less conclusive, as the alignment of the graves shows considerable variation with 8 graves running N-S, 11 running E-W, 11 running NE-SW and 7 running NW-SE.16

The cemeteries also have attached churches in some cases, or at least buildings interpreted as such, such as at Shurdahand Aphiona, although the relationship between these buildings and the tombs is inconclusive.17 While the presence of churches at the site is unlikely to be accidental. , they may reflect a later recognition of the significance of the location rather than a direct association. This is particularly likely in the case of Shurdah, where the remains of the settlement may well be dated to the late medieval period. The use of a furnished burial ritual may also question the Christianity of the inhabitants. This issue will be discussed further below.

Although the case for Christianity remains unproven, it is likely that the population represented in the cemeteries was Latin speaking or at least Latin literate, as indicated by the concentration of surviving Latin toponyms in the vicinity. This concentration of Latin place names is accompanied by a marked absence of the Slavic names that often occur elsewhere in Albania.18

As mentioned, the Albanian case argued that these cemeteries represented a surviving Illyrian culture which remained in contact with the Byzantine world and formed the root of today's Albanian population. However, the first documentary reference to the Albanians or Arbers appears in the 11th century. in the Alexiad of AnnaComnena.19 Based on parallels with the finds elsewhere, the comania

16 Anamali og Spahiu (1980) 79 (Kruja); Prendi (1980) 126-29 (Lezhë); Spahiu (1980a) 29-31.

17 Komata (1980) 113 (Shurdah); Bulle (1934) 216 (Aphiona).18 Popovi´ (1984) 211-13.19 Alexiad 4 (cit. Wilkes (1992) 279).

william bowden62

burials are usually dated to the late 6th, 7th and 8th centuries, leaving a large chronological gap in the Albanian model.19a

Another significant problem for the Albanian ethnogenesis argument was the relatively limited distribution of these cemeteries. This was resolved by Namik Bodinaku's identification of the so-called southernArbër culture, where burials containing similar objects to the Komani burials were found inserted in Bronze Age tumuli in southern Albania. A series of tumuli excavated at Piskova in the upper reaches of the Aoos river valley were of particular importance in this sense.20

Similar late antique or post-Roman burials have been found within prehistoric tumuli on the Greek side of the border, for example at Merope and Kato Pedina.21 This adoption of tumulus burial in the post-Roman period was also used to strengthen the case for Illyrian- Albanian continuity. However, the presence of the Komani-type cemetery at Aphiona in Corfu, beyond the current Albanian border, and more importantly, outside the area considered to have been 'Illyrian' in antiquity, was rarely discussed.

The reaction of non-Albanian archaeologists to this overtly nationalist interpretation of the Komani-Kruja cemeteries has been incomprehensibly negative. John Wilkes has written of the 'highly improbable reconstruction of Illyrian history of this period', although Nicholas Hammond was more sympathetic to the idea that the Albanians were the direct descendants of the Illyrians.22 Recently, some Albanian archaeologists were keen to distance yourself from the policies. of the communist regime, has also moved away from the former extremist position.23 However, the aim of the present paper is not to provide a detailed re-evaluation of the Komani-Kruja material or to comment on the relationship between the inhabitants of the cemeteries and the present-day Albanian population, but rather to suggest some alternative approaches to the problem of the emergence of furnished burial in the post-Roman period.

19a The dating of the objects in the Komani tombs is a complex subject which is somewhat beyond the scope of this paper, although it is clear that any detailed interpretation will depend on a more precise chronology of the finds. The best overview of the current chronology is provided by Popoviè (1984).

20 E. Nallabani pers. comm.; Bodinaku (1983).21 Andreou (1980); (1983); (1986); (1987).22 Wilkes (1992) 278; Hammond (2000). The Albanian's Illyrian ancestry is also

confirmed in The Times Guide to the Peoples of Europe (Fernández-Arnesto (1994) 217).23 e.g. Miraj and Zeqo (1993).

the construction of identities in post-Roman Albania 63

Archeology and Ethnicity

Discussion of ethnic groups in archeology has until recently been seen as a survival of outdated 'cultural-historical' archaeology. This term describes the perception of the human past in terms of the history of groups whose behavior is explained by reference to shared cultural norms rather than wider social, economic or environmental processes. Particular patterns of archaeological evidence were associated with particular 'peoples' who were often equated with past (and present) ethnic groups. Western archaeologists now view the history of their discipline in mainly evolutionary terms, with antiquarianism giving rise to cultural-historical archaeology, which in turn was supplanted by the 'new' or processual archaeology, itself subsequently rejected and revised in post-processual or interpretive theories. The idea of ​​progression through a series of stages of increasing sophistication is implicit and in many ways valid, but has imbued the discipline with a fear of quiescence or regression.

The rejection of cultural history in favor of alternative forms of social explanation has led to an ambivalence about the role of ethnicity in archaeology; the identification of ethnic groups through their material culture is irrevocably linked to the early 'cultural history' stage of archaeology's empirical development. Ethnographic work has also demonstrated the problems of equating cultural equality with ethnic groups.24 However, recent work has viewed ethnicity more positively, as an important aspect of social systems and as a dynamic element in economic and political structures, rather. than merely a passive reflection of a series of cultural norms. Ethnicity and ethnic groups have been removed from the area of ​​description and classification that characterized cultural-historical archaeology, and are now perceived as a framework for explanation and interpretation.25

In recent decades, the debate on ethnicity in anthropology has been dominated by two theoretical approaches, one that sees ethnicity in "primary terms" and the other that adopts an instrumentalist perspective. The primordial approach (which underlies traditional cultural-historical archaeology) sees tribal affiliation as an innate part of the human condition, driven by "givens of birth - 'blood',

24 Hodder (1982).25 Jones (1997) 11.

william bowden64

language, religion, territory and culture”.26 These form the primary identity of an individual, which transcends the other secondary identities that are subsequently acquired. Primordial attachments are involuntary and take precedence over the needs of particular social circumstances and situations. Strong criticism has been leveled at the original view. The main objections concern its inadequacy as an explanatory framework. The effects of so many of its defining elements (such as blood and language) are described as "intangible or ineffable", and ethnic identity is seen as unreasonably inherent in human nature. In recent years, therefore, the original perspective has faced challenges from a group of theoretical approaches that have been broadly defined as instrumentalist.

Instrumentalist views suggest that ethnic identity is not "given at birth" but is a dynamic and fluid response to changing situations (although under this umbrella there is considerable disagreement about the relative importance of individual action versus the constraints of social structures and cultural norms) . This has been particularly influenced by the work of Frederick Barth, who saw ethnicity as "the social organization of cultural differences". Barth focused on differentiation – how one group expresses its separation from another – rather than on the particular cultural characteristics of an ethnic group. Shared culture in Barth's model is a tool for the construction and maintenance of ethnic boundaries and a means by which the Other is defined.27 Barth further argued that the maintenance of ethnic boundaries can be seen as a way in which a group adapts to a particular social or ecological niche. Other authors such as Abner Cohenhave also emphasized that ethnic identity is not constant or given, but can be changed by individuals according to the needs of a particular situation. Likewise, the meaning of ethnic identity itself could be highly variable, with aspects of an individual's identity suppressed or emphasized according to circumstances.28 Geary applied this approach to early medieval Europe and concluded that "Early Medieval ethnicity should be seen as a subjective process by which individuals and groups identified themselves or others within specific situations and for specific purposes”.29

26 Jones (1997) 65.27 Barth (1969) in particular 15-16. See also Jenkins (1997) 12.28 Cohen (1974).29 Geary (1983) 16, cited by Shennan (1989)

the construction of identities in post-Roman Albania 65

The instrumentalist view has been criticized as reductionist in that it can make ethnicity no more than a strategy for responding to a particular situation, with culture as a tool that is expressed in economic and political relations. This is contradicted by the persistent failure of attempts at cultural integration, whereby "many individuals in fact stubbornly continue to ally with those with whom they have ethnically uniform ties, even though such alliances may run counter to patterns of group formation determined by common interests ".30 Instrumentalist some approaches also sometimes fail to recognize the complexity of social relations by, for example, assuming that common interests are perceived equally by members of a group (which in any case may be internally divided on gender or class levels) or that recognition of common interests will trigger joint action.

Do the Komani cemeteries constitute real evidence for the existence of a separate ethnic group or groups in post-Roman Albania? As with any isolated collection of archaeological material, the answer must be no, although this does not in itself undermine the validity of a question of this nature. Anthropological and ethnographic work has emphasized that material culture has a symbolic function at several different levels, of which group identity is only one aspect. It should therefore be possible to examine this material in terms of its dynamic role within a social strategy rather than as a passive representation of a set of ill-defined cultural norms. Ethnicity is not an explanation in itself, but can nevertheless form part of an explanatory framework. We must change the questions being asked of the Komani-Kruja material and instead ask precisely what purpose may have been served by the adoption of a ritual of furnished burial where none had previously existed.

Epirus in the 7th century

Any effort to review the Komani-Kruja material requires an understanding of the situation that prevailed in the area in the 7th century and what I mean by 'post-Roman'. The evidence, apart from what the cemeteries offer, is fragmentary and too wide open

30 Rex (1991) 11, cited by Jenkins (1997) 47. See also Jones (1997) 76-79.

william bowden66

interpretation. However, a number of themes can be broadly identified.31

1: City life was extinguished in AD 650. (and probably considerably earlier). While the late antique city was fundamentally different from the cities of the early empire, the former was recognizably derivative of the latter. Until about 550, the concept of classical urbanism was probably known to most of the population, although the reality was significantly removed from the ideal. By c. 650, although population centers remained within the precincts of the former Romantowns, their way of life in terms of material culture was indistinguishable from other centres, such as the hilltop sites briefly discussed above. These sites, which often reused the defenses of classical and Hellenistic predecessors, were certainly occupied from the 6th century. and possibly earlier. Excavations have produced quantities of coarse pottery, but relatively little in the way of recognizable imports. indeed, unequivocal 7th century occupation is known from only a handful of sites.31a Data are extremely limited for most of the Roman cities in central Albania, even in the case of Dyrrhachium (modern Durrës), which is believed to have remained in Byzantine hands in the 7th-9th century. Dyrrhachium is assumed by Albanian archaeologists to have been a rallying point for the Komani-Kruja population, although the nature of its occupation has not been established archaeologically.32 Although we should not imagine that the urban centers were completely torn away by barbarian attacks that a certain amount of evidence can be adduced to suggest a violent end to life in some settlements. In particular, the excavated areas of Onchesmos (modern Saranda) were found covered with a layer of burnt material dated by coins of Justin II and Sophia, which the excavator did not unreasonably associate with the documented Slavic attacks of 586/87.33

2: Contact with the remaining areas of the Eastern Roman world appears to have been limited. Although imported materials are rep-

31 These themes are expanded in Bowden (forthcoming).31a These sites will be discussed in detail in Bowden (forthcoming) and Bowden

and Hodges (forthcoming). 7th century levels have recently been identified at Butrintand at a site in its hinterland (see Bowden, Hodges et al. (forthcoming).

32 Wilkes (1992) 273-74. Evidence for occupation of the city during this period is almost non-existent, although see Tartari (1982) for the suggestion of a continuous am-fora sequence at Dyrrhachium.

33 Lako (1991) 157.

the construction of identities in post-Roman Albania 67

resented the Komani-Kruja cemeteries, their type and distribution suggest that trade and exchange lines had changed considerably. The large quantities of glass unique to the Aphiona cemetery and the unusual number of weapons found in the Komanigraves suggest that exchange mechanisms were subject to much more local variations. The manner in which the non-local material present in the burial assemblages was obtained is completely unknown. The supply of imported fine goods and other products declined in the early 7th century. like using coins. Although coins of Constantius II (641-68) are known, the latest coins to appear in any quantity date to the first two decades of Heraclius' reign.34

3: The Church's administrative structure appears to be fragmented. Although the documentary sources are problematic, they indicate that most of the bishops of Epirus Nova and Epirus Vetushad were forced to seek refuge on islands or in southern Italy, along with part of their congregations. In 604, 4 of the 5 remaining bishops of Epirus Vetus resided in Corfu, while Bishop John of Lissus was in Squillace in southern Italy when his church was in the hands of the barbarians.35 Although bishops from the Epirote provinces do. appears in episcopal lists from the 8th century, the references are probably anachronistic.

In summary, although the details and extent of these processes are debatable, there can be no doubt about the extent of the changes that took place in the later 6th and 7th centuries. These changes included the significant shift in settlement patterns indicated by the decline in urban settlement and the largely simultaneous increase in hilltop occupation, together with the undoubted lack of stability caused by the Slavic attacks (regardless of their scale and nature). It was in this context of a large-scale fragmentation of social structures that the Komani-Kruja cemeteries appear.

Reinterpretation of the Komani-Kruja Burials

The Komani-Kruja tombs represent the reintroduction of a furnished burial ritual in an area where furnished burial had not been practiced

34 coins of the late 7th and 8th centuries are represented by a single unprovenanced solidus of Constantine V (751-75) from the coin collection of the Archaeological Museum of Tirana (Spahiu (1980b) 385).

35 Chrysos (1981) 74-76.

william bowden68

for about 200 years. They were vehicles of conspicuous display and consumption, where there also seems to have been a marked difference between the sexes. Despite the presence of grave goods, for the reasons previously outlined, the burials are usually assumed to be Christian, although distinctly Christian restrictions against grave goods were considered to be of secondary importance to the need for display. It is this marked difference between the Komani-Kruja tombs and the simple unfurnished burials characteristic of the area in the 5th and 6th centuries that makes them so intriguing.

Within a theoretical framework that remains essentially that of cultural-historical archaeology, two basic explanations have been advanced for the Komani-Kruja tombs.

1. The cemeteries represent an existing population that survived the Slavic attacks and maintained an identity separate from the newcomers.

2. The cemeteries represent an incoming group, perhaps a Romanized Illyrian population from further north, who had been displaced by Slavic settlement.36

These arguments rest on the assumption inherent in cultural historical archeology that material culture is a manifestation of normative behavior within a particular group. In other words, the undertaker and the objects in the graves represent the customary practices of one group of people and not another. Within a classical cultural history framework, however, the rapid emergence of these objects and burial practices would suggest an invading or migrating group, thereby privileging the second explanation. This was unsatisfactory to archaeologists of the Communist period, requiring them to construct a model in which the objects in the tombs reflected established norms of Illyrian/Albanian material culture that were maintained in the Roman period and beyond. This tension between the demands of a nationalist ideology and the accepted parameters of cultural-historical archeology is one of the defining characteristics of Albanian archeology in this period.

The idea of ​​habitual practice, inherent in the primordialist definition of ethnicity and largely absent from the instrumentalist model, has been retained by recent more carefully nuanced definitions of ethnicity as represented by archaeological evidence. SianJones, in an attempt to reconcile primordialist and instrumentalist

36 As suggested by Wilkes (1992) 278.

the construction of identities in post-Roman Albania 69

explanations of ethnicity, suggests that "(those) cultural practices and representations that are objectified as symbols of ethnicity are derived from, and resonate with, the habitual practices and experiences of the people concerned, as well as reflecting the instrumental events of a particular situation .'37 Jones' model includes the concept of habitus - the unconscious acquisition of common or basic actions and attitudes as part of the socialization process within a particular cultural environment.38 This definition of ethnic symbols cannot include the Komani tombs as an expression of ethnic identity. -tity. If we adhere to the accepted view that the Komani tombs represent a Latin-literate Christian population, then a ritual of furnished burial seems unlikely to have reflected or emphasized a customary practice that would resonate with others members of society Similarly, in terms of material culture, the grave assemblages contain a wide variety of objects that formed part of a relatively homogeneous group of objects characteristic of non-Roman contexts in Europe in the 6th, 7th and 8th centuries. century. It is difficult to imagine these objects playing an active role in the construction of a non-barbaric Christian identity.

Indeed, in areas of Europe where a pseudo-Roman Christian identity was created and maintained, the means by which this was achieved varied widely. The survival of an epigraphic habit, for example, seems to have fulfilled a special role in several places. This is particularly notable in south-west Britain, where a large number of early Christian memorials are known, dating to the late 5th and 6th centuries, when the population faced considerable pressure from Germanic invaders to the east. The inscriptions appear in the context of places where quantities of imported Mediterranean pottery from the same period have been found.39

The example of Cornwall is particularly remarkable in that, unless we consider that the inscriptions and pottery reflect the tastes of a Romano-British refugee population, it seems that these phenomena do not represent the survival of a Romanized society, but the active construction of one where none had previously existed. Although this is not the occasion to approach this complex question, it seems possible that the people of south-west Britain

37 Jones (1997) 128.38 On the use of habitus in relation to ethnicity and archaeology, see also Shennan

(1989) 14–17.39 Epigraphic evidence is summarized in Todd (1987) 249–51. On imported

ceramics see Thomas (1981) and Fulford (1989).

william bowden70

attempted to construct boundaries between themselves and the Germanic 'Other' through the creation of a largely fictitious Roman identity. Similarly, an epigraphic habit was maintained in the city of Augusta Treverorum (modern Trier) on the Moselle, where an enormous number of inscriptions dating from the 5th and 8th centuries are known. In Trier, the association with the Roman past was also expressed through the apparent curation of terra sigillata vessels, which were probably removed from Roman graves.40

These pseudo-Roman features are conspicuously absent from the Komani-Kruja cemeteries. The Albanian argument, however, is that the tombs represent the revival of an Illyrian identity rather than a Roman one. The purported links between the material culture of the Illyrians and that represented in the tombs are dubious at best, and certain cases, such as the revival of ceramic styles, find little or no supporting evidence.41 The possibility that the Komani-Kruja people however, was deliberate to evoke the idea of ​​remote ancestry cannot be ruled out. The choice of places for settlement during this period is of interest in this respect. The reoccupation of prehistoric hill forts is a characteristic feature of Late Antiquity Europe, and although in many cases the choice of site was probably based on pragmatic rather than symbolic reasons, the significance attached to "ancient" sites should. not be overlooked. In times of social stress, the creation of belonging to the past, thereby establishing the primacy of occupation, has enormous resonance (a fact understood by the communist regime in Albania).

Reuse of Bronze Age tumuli may have some significance in this regard. Building associations with former burial grounds is not unusual. Reuse of older monuments such as burial mounds is a widely recognized facet of burial in post-Roman Europe and also occurs in prehistoric contexts. In Great Britain, William examined e.g. a sample of 71 Anglo-Saxon cemeteries, of which 38 (54%) showed evidence of monument reuse. In the Upper Thamesvalley the figure is even higher with around 60% of the 7th century. Anglo-Saxon cemeteries located next to barrows or other monuments. The "referential qualities" of earlier sites and monuments were also recognized in contemporary Christian contexts; for example, in Merovingian Gaul were both prehistoric and Roman monuments

40 Clemens (1998) has recovered terra sigillata with medieval graffiti from 8th century contexts. The inscriptions are cataloged by Merton (1990).

41 Nallbani and Gilkes (2000) 12.

the construction of identities in post-Roman Albania 71

suitable for new purposes.42 The expression of affinity with the past is a powerful tool in the construction of identities, especially in periods of instability. As Frank Furëdi puts it, "loss of control over the future stimulates a tenacious affirmation of history".43

Can the adoption of a richly furnished burial ritual be interpreted as an active strategy in the construction of a new group identity by a population that inhabited an uncertain and rapidly changing world? In the case of the Komani-Kruja cemeteries, this seems unlikely, and instead the reappearance of furnished burials may be more indicative of changes in internal social hierarchies and power structures, rather than the construction or maintenance of ethnic boundaries. Indeed, it may be more profitable to examine these tombs in terms of what they can tell us about the internal dynamics of the groups that created them. Although we might note Hodder's warning against a "simple association between social organization and burial", in the case of the Komani-Kruja cemeteries there is actually sufficient background evidence to suggest that the sudden appearance of furnished burial does indeed relate to profound social change.44

Although the cemeteries give the impression of homogeneity, there is actually considerable variation between the different assemblages, suggesting that the grave groups are responses to very local conditions. This is evident in the relationship between grave goods and gender, especially with regard to the gender division proposed by Albanian archaeologists. At Aphiona on Corfu, only nine of the nineteen burials excavated contained grave goods. All the objects corresponded to types associated with female burials in the Komani-Kruja cemeteries. There were no weapons present.45 If we accept the gender division proposed by the Albanian archaeologists (and the 10-9 division of the random sample excavated at Aphiona could suggest that they are plausible), it seems that in Aphiona is a need for consumption. and the display was articulated exclusively through female burials. But elsewhere these gender divisions seem extremely blurred in the published collections, where several different combinations of objects can be observed. At Dalmaces, axes and arrow/spearheads appear in combination with glass necklaces, knives, fibulae, earrings, rings and hanging ornaments, while at Lezhë they are to the same extent.

42 Williams (1997); Effros (2001). See also Harke (2001).43 Furëdi (1992) 3.44 Hodder (1982) 201.45 Bulle (1934).

william bowden72

variation can be seen. A similar degree of variation can be seen in the number of unfurnished burials in the cemeteries. That Lezhë out of 37 burials were only 4 unfurnished, while in Dalmaces there were 18 graves unfurnished, from a total of 40.46 Dalmaces therefore contained a similar proportion of unfurnished graves as the Aphiona cemetery, with the decisive difference that the Dalmaces sample contained 'male ' objects such as weapons, in contrast to Aphiona, where only 'female' dress items were found. A similar level of variation can be observed in the different orientations present in the burials (see above).

It is clear that even a cursory examination of the excavated samples of cemeteries reveals significant differences between the sites. To the variables mentioned above can be added factors such as the type of tomb construction, the occurrence of multiple burials, and the implication that some of the tombs were used as ossuaries. In Lezhë, for example, contained graves 19 and 35 4 skulls, while grave 31 contained 6 skulls. The presence of multiple skeletons in other graves suggests that the graves may have been opened on more than one occasion, further confusing the grave assemblages. Other variables are present, including the uncertain chronology of the tombs. At Shurdhah, it has been suggested that the unfurnished graves are later than those with grave goods, reflecting the church's increasing narrowness in relation to furnished burial.47 However, these unfurnished graves have a north-south orientation and were scattered among those with grave goods. In contrast, the tombs closest to the church of Shurdhah, which have an east-west orientation, are richly furnished.

A degree of variation between cemeteries is not in itself surprising given the relatively rapid development of a furnished burial rite in the 7th century. We can see these differences in the context of communities working within what was essentially an unknown medium, with a corresponding level of uncertainty and debate about what constituted "good" practice. This may have been further complicated by the Christianity of at least some of the participants, although as described below this Christianity is ambiguous at best. Likewise, we should not assume that all contemporary attitudes to death and burial were the same. The unfurnished burials could reflect conflicts regarding the adoption of an essentially pagan burial practice within

46 Spahiu (1980a); Prendi (1980).47 Komata (1980) 113.

the construction of identities in post-Roman Albania 73

a nominally Christian community, rather than reflecting the age, gender, poverty or ethnic identity of the residents.48 While, as Young has demonstrated, the presence of grave goods does not necessarily equal paganism, it is unlikely that they constituted an exaggeration of Christianity. 49

Conclusions: Ethnicity, status and participation

A person can manipulate customs if he becomes part of such agroup by adopting its current main symbols. He cannot manipulate others without being ready to be manipulated by them. He must pay the price of membership by participating in the group's symbolic activities and by some adherence to the group's goals.

Abner Cohen50

Did ethnicity really matter in the post-Roman period? It has been argued that its role was rarely important and that other aspects of the 'social code' were dominant.51 The demands of nationalist cultural-historical archeologies may have meant that it has been overemphasized, although this does not mean that it should be excluded. For example, Heinrich Härke's analysis of weapon burials in Anglo-Saxon cemeteries, which took into account epigenetic data tracing lines of descent within cemetery populations, suggests that weapons were included to demonstrate that the occupant of the grave was of Germanic descent, who apparently carried a certain status level.52

Genealogy was certainly important in the Germanic tribes, which e.g. appears in the royal pedigrees known from Anglo-Saxon and Gothic contexts.53 Nevertheless, the creation of lineages can be part of the construction of ethnic identity (such as the United States today), at least in an Anglo-Saxon context , lin-eage seems more related to the individual's status rather than their membership of a particular group.

This last point is of fundamental importance for the interpretation of the Albanian cemeteries. If the grave collections were expressed

48 Se Hodder (1982) 195-201.49 Young (1977).50 Cohen (1974) xiii.51 Daim (1998) 72.52 Härke (1992)53 Dumville (1976) for angelsaksiske genealogier; Heather (1991) 19-33 for Gothic

genealogist.

william bowden74

the tions of group identity, then the identity expressed is not that of an autochthonous Christian population trying to define itself in relation to an arrived pagan 'Other'. Regardless of whether the tomb dwellers professed Christianity or not, they operated in an essentially pagan environment and one that was prevalent throughout central and northern Europe in the 7th century. Migrant groups, including the Avars, Slavs, and Lombards, who were probably geographically closest to Epirus Nova, buried their dead with jewelry, weapons, and other goods. The Komani-Kruja graves were therefore part of a broad sphere of funerary practices concerned with individual rather than group identity, where hierarchies were expressed through the burial of high-status objects. The Komani-Kruja material is in fact consistent with patterns of status-related burial elsewhere, for example in Avar contexts, where great emphasis was placed on the presence of imported objects.54 The Albanian graves contain a wide variety of imported materials, especially gold earrings, which probably originated in Byzantine Sicily. Also in this context are the glass vessels which are unique to Aphionacemetery on Corfu.

Material culture was therefore not used as a medium through which difference was communicated, but rather as a means through which the Komani-Kruja people participated in a range of social relations. A primary function of these 'means of representation' was to create stability in a society. The post-Roman world was characterized by uncertainty and high mobility in terms of settlement or residence and in terms of identity and status.55 As we have seen, the pace changed within the 7th century. Epirus was fast. In the space of perhaps two generations, an urbanized or semi-urbanized society with recognizable links to the Roman past was transformed into one that was essentially rural, whose leadership structures had also changed dramatically. As I have argued elsewhere, power was individualized in late antique Epirus, a situation that may have originated in the patronage system of the late empire and which continued with the rise of powerful churchmen in the late 5th and 6th centuries.56 Epirus also became more and more peripheral to the empire's administrative structures. This perhaps created a situation where power coalesced around individuals

54 Daim (1998) 86.55 Daim (1998) 84.56 See Bowden (forthcoming).

the construction of identities in post-Roman Albania 75

rather than being focused on city or state. The result was a more fluid and localized network of relationships based on fluctuating loyalties and competition, which in terms of social and political cohesion was perhaps not unlike that of incoming migrant groups whose arrival undoubtedly exacerbated the process. times of political and demographic instability, almost unlimited opportunities may be open to anyone with a claim to authority and the means to enforce it".57

It is this context that may be key to understanding the Komani-Kruja cemeteries, framing the rapid emergence of a furnished burial ritual as a response to an entirely new and ever-changing set of social relations. The representation of individual identity and status in relation to peer groups proved to be more compelling than earlier restrictions in the Christian Church regarding furnished burial. If the occupants of the Komani-Kruja tombs were Christians, their religion was not an aspect of their identity that was apparently expressed within the funerary ritual (and perhaps not within other areas of social relations). Although Christian objects were placed in the tombs, as with the presence of Christian objects within the context of pagan tombs elsewhere in Europe, their symbolic value lay in their status as objects rather than as a means of religious representation.

In conclusion, it seems unlikely that expression of group or ethnic identity was a primary motivation behind the revival of the furnished burial in post-Roman Epirus Vetus and Epirus Nova. The Komani-Kruja people participated in a European medium of burial practices rather than constructing an identity that consciously expressed their difference from their neighbors. Attempts on the part of 'Roman' groups to emphasize their own diversity in relation to incoming populations emphasize their affinity with the classical past, its religion, epigraphic practices and material culture. However, the Komani-Kruja people did not try to maintain or construct boundaries to preserve their ethnic integrity. Instead, it was to participate in the localized and fluctuating social structures that emerged from the remains of Roman Epirus.

57 Scull (1992) 9 (referring to East Anglia in the 5th century).

william bowden76

Bibliography

Anamali S. (1971) "An early medieval cemetery of Bukël af Mirdita", Iliria1 (1971) 209-26.

——— (1972) "New Archaeological Discoveries on the Relations of the Albanians with Byzantium in the Early Middle Ages", Studia Albanica 9.1 (1972)135-41.

——— (1980) "Late Antiquity and Early Middle Ages in Albanian Research", Iliria 9-10 (1979-80) 5-21.

——— (1982) "The problem of the formation of the Albanian people in the light of archaeological data", Studia Albanica 19.2 (1982) 53-73.

——— (1993) “Goldsmiths, Byzantine Jewels in Albania: Komani”, XL Culture Course on Ravenna and Byzantine Art. International Seminar on: 'Albania from Late Antiquity to the Middle Ages, Aspects and Problems of Archeology and Art History' Ravenna, 29 April - 5 May 1993 (1993) 435-46.

Anamali S. and Spahiu H. (1980) "The Wooden Cemetery in Kruja", Iliria 9-10 (1979-80) 47-104.

Andreou E. (1980) "Meripi og Palipyrgis Pogoni", ArchDelt 35 (1980)303-307.

——— (1983) “Μερ�πη”, ArchDelt 38 (1983) 229-30.——— (1986) “Κ�τω Πεδιν� �αγ�ρ���υ 1) 1. —— (1987) “Παλη�πυργ�ς Πωγων��υ”, ArchDelt 42 (1987) 307-308.Arthur P. (1998) “Local Pottery in Naples and the Northern

and seventh centuries”, in Ceramica i Italia VI-VII Secolo, ed. L. Sagui (Florence 1998) 511-30.

Avramea A. (1997) The Peloponnese from the 4th to the 8th Centuries. Changes and Persistences (Paris1997).

Barth F. (1969) "Introduction" in Ethnic Groups and Boundaries: The Social Organization of Cultural Difference, ed. F. Barth (Oslo 1969) 9-38.

Bodinaku N. (1983) "Culture in the Early Medieval Albanian Cemetery of the Upper Lungina of Vjosa in the District of Përmet", Iliria 13.1 (1983) 241-50.

Bowden W. (forestående) Epirus Vetus: The Archaeology of a Late Antique Province (London2002)

Bowden W. and Hodges R. (forthcoming) “Balkan Ghosts? Nationalism and the question of rural continuity in Albania' in Landscapes of Change: The Evolution of the Countryside from Late Antiquity to the Early Middle Ages, eds N. Christie and S. Scott, (Aldershot forthcoming).

Bowden W., Hodges R., Lako K. et al. (2002) "Roman and Late Antique Butrint: Excavations and Survey 2000-2001", JRA 16 (2002) 151-80.

Bulle H. (1934) "Excavations near Aphiona on Corfu", AM 59 (1934) 147-240 υ κατ� τ�ν πρωτ��η�α-

din� ep��� (D-F ai)”, ’Hipir�tik r�nik 23 (1981) 6-111. Clemens

Shard from Trier", TrZ 61 (1998) 241-47. Cohen A. (1974) "Introduction", i Urban Ethnicity, red. A. Cohen (London 1974)

ix-xxiv.Curta F. (2001) The Making of the Slaves. History and archeology of the Lower Danube approx.

500-700 (Cambridge 2001). Daim F. (1998) “Archaeology, ethnicity and structures of identification: the

example of the Avars, Carantans and Moravians in the eighth century", in Strategies of Distinction. The Construction of Ethnic Communities, 300-800, ed.

the construction of identities in post-Roman Albania 77

W. Pohl og H. Reimitz (The Transformation of the Roman World 2) (Leiden, Boston og Köln 1998) 71-93.

D'Angela C. (1982) "An Early Medieval Cemetery at Merrine (Le)", Studi di Antichita3 (1982) 175-82.

Dumville D. (1976) "The Anglian collection of royal genealogies and regal lists",Anglo-Saxon England 5 (1976) 23-50.

Effros B. (2001) "Monuments and memory: repossessing ancient rests in EarlyMedieval Gaul", i Topographies of Power in the Early Middle Ages, red. M. DeJong, F. Theuws og C. Van Rhijn (Leiden, Boston og Köln 2001) (TheTransformation of the Roman World 6) 93-118.

Fernandez-Armesto P. (1994) ed. The Times Guide to the Peoples of Europe (London1994).

Fulford M. (1989) "Byzantium and Britain: a Mediterranean perspective on Post-Roman Mediterranean import in Western Britain and Ireland", Medieval Ar-chaeology 33 (1989) 1-6.

Furëdi F. (1992) Mythical Past, Elusive Future. History and Society in an Anxious Age, (London 1992).

Geary P. (1983) "Ethnic Identity as a Situational Construction in the Early Middle Ages", Notices of the Anthropological Society of Vienna 13 (1983) 15-26.

Hammond N. G. L. (2000) "Epirus through the Ages", The Anglo-Hellenic Review 22(2000) 9-10.

Härke H. (1992) "Changing symbols in a changing society: the Anglo-Saxon weapon burial rite in the seventh century", in The Age of Sutton Hoo, ed. M. O. H. Carver (Woodbridge 1992) 149-65.

——— (2001) “Cemeteries as places of power”, i Topographies of Power in the EarlyMiddel Ages, red. M. De Jong, F. Theuws og C. Van Rhijn (The Transformation of the Roman World 6) (Leiden, Boston og Köln 2001) 9-30.

Heather P. J. (1991) Goths and Romans 332-489 (Oxford 1991). Hodder I. (1982) Symbols in Action (Cambridge 1982). Hodges R., Saraçi G., Bowden W., Chiles P., Gilkes O., Lako K., Lane A., Martin

S., Mitchell J., Moreland J., O'Hara S., Pluciennik M. og Watson L. (1997) "Late-antique and Byzantine Butrint: Interim report on the port and its hinterland", JRA 10 (1997) 207-34.

Jenkins R (1997) Rethinking Ethnicity. Arguments and Explorations (London 1997). Jones S (1997) The Archeology of Ethnicity (London 1997). Komata D. (1980) "Arbërore Cemetery of Shurdah", Iliria 9-10 (1979-80) 105-22. Lako K. (1991) "The Paleochristian Basilica of Onhezm", Illyria 21.1-2 (1991) 123-86. Merten H. (1990) Catalog of the early Christian inscriptions of the episcopal cathedral and

Diözesanmuseum Trier (Catalogs and writings of the Episcopal Cathedral and Diözesanmuseum Trier; 1) (Trier 1990).

Miraj L. og Zeqo M. (1993) "Conceptual changes in Albanian archaeology", Antiquity 67 (254) (1993) 123-25.

Nallbani E. and Gilkes O. (2000) A Report on the Archive Materials Relating to Butrint for the Socialist Period (1945-1990), Report prepared for the Butrint Foundation.

Popovi´ V. (1984) "Byzantines, Slaves et autochthones dans les provinces de Prévalitaine et Nouvelle Épire", in Cities and Settlement in Proto-Byzantine Illyricum. Proceedings of colloquium organized by the French School in Rome. Rome, 12.-14. May, 1982, ed.V. Popovi' (Rome 1984) 181-243.

Prendi F. (1980) "A cemetery of arboreal culture in Lezhë", Iliria 9-10 (1979-80) 123-70.

william bowden78

Rex J. (1991) Ethnic Identity and Ethnic Mobilization in Britain (Warwick 1991) Scull C. J. (1992) "Before Sutton Hoo: Power and Community Structures in the Beginning

East Anglia" in The Age of Sutton Hoo, ed. M. O. H. Carver (Woodbridge 1992). Shennan S. (1989) "Introduction: archaeological approaches to cultural identity",

i Archaeological Approaches to Cultural Identity, red. S. Shennan (London 1989) 1-32.

Spahiu H. (1971) "Ancient finds from the medieval cemetery of the Dalmatian fortress", Iliria 1 (1971) 227-62.

——— (1980a) "The Arboretum Cemetery at the Dalmatian Castle", Iliria 9-10 (1979-80)23-46.

——— (1980b) “Byzantine coins of ca. V-XIII discovered in the territory of Albania", Iliria 9-10 (1979-80), 353-422.

Tagliaferri A. (1990) "Il ducato di Forum Iulii", i I Longobardi, ed. G. C. Menis (Milan 1990) 358-475.

Tartari F. (1982) "Amphorae and Archaeological Museums in Durrës", Iliria 12.2 (1982) 240-79.

Thomas C. (1981) A Preliminary List of Imported Pottery in Post-Roman Western Britain and Ireland (Redruth 1981).

Todd M. (1987) The South-West til AD 1000 (London-New York 1987).Wilkes J. (1992) The Illyrians (London 1992).Williams H. (1997) "Ancient landscapes and the dead: the genbrug af forhistorisk og

Roman Monuments as Early Anglo-Saxon Burial Sites”, Medieval Archaeology41 (1997) 1-32.

Young B. K. (1977) "Paganism, Christianization and Merovingian Funeral Rites", Medieval Archaeology 7 (1977) 5-81.

aspects of the transformation of the Roman domus 79

SOME ASPECTS OF THE TRANSFORMATION OF THE THEROMAN DOMUS BETWEEN THE LATE ANTIQUE AND

EARLY MIDDLE AGES

BARBARA POLCI

Abstract

This essay deals with some aspects of the transformation of the late Roman domus into the early medieval house and focuses on the spaces designed for reception and entertainment. First, I will consider the use and development of the reception areas of wealthy houses and their relationship to the growth of private patronage in Late Antiquity. Second, I will examine the transformation of this late antique model of elite housing into the new type of upper-class housing that emerged in early medieval Italy. In particular, I will focus on the transfer of the reception halls and banqueting chambers to the upper floor and on the social and architectural implications of this function.

Introduction

"Although churches remain for us from the Middle Ages, we do not have one private house: yet nothing is more meaningful than an old house."1 Thus lamented Andrea Carandini in his 1989 article on the subject of Europe's policy towards its cultural values. In recent years, however, new light has been shed on late antiquity and early medieval housing through a combination of important new archaeological investigations, new investigations into the social implications of house architecture and a re-examination of earlier excavations and literary sources.

In this essay I will examine some aspects of the transformation of the late Roman domus between the dissolution of Roman authority and the rise of the Carolingian Empire through a study of

1 Carandini (1989) 10.

L. Lavan og W. Bowden (red.) Theory and Practice in Late Antique Archaeology(Late Antique Archaeology 1 2003) (Leiden), s. 79-109

barbara polci80

surviving material and documentary evidence of elite lifestyles. In particular, I will focus on the spaces designed for reception and entertainment—the areas where meetings between elites and non-elites took place—to better understand the ways in which participants expressed and defined their power and status through their residences during this fundamental social process. change. The study is divided into two main parts. First, I will consider the use and development of the reception areas of wealthy houses and the relationship to the growth of private patronage in Late Antiquity. Second, I will examine the transition between this Late Antique model of elite housing and the new type of upper-class housing that emerged in early medieval Italy. In particular, I will focus on the transfer of reception halls and function rooms to the upper floor and on the social and architectural implications of this function. The aim is not only to clarify some crucial aspects of the architectural development and transformation of the late Romandomus, but also to arrive at a better understanding of the social, economic and cultural structures of the period in question.

Reception rooms in Late Antiquity: Use and development

The large number of accounts found in written sources testify to the great importance of dining and entertainment in the social and political life of the Roman aristocracy, which is also confirmed by the special importance given to reception areas in the Roman residences of the upper classes. Dinner would start mid-afternoon, but could well last into the night, accompanied by a wide variety of entertainment. The success of a banquet will depend not only on the quality, quantity and variety of food, but also on the elegance of its surroundings and the type of entertainment offered. Performances at dinner parties would often include music, song, dance and poetry performances and occasionally also acrobats, dramatic performances, mimes and pantomimes.2

In wealthy houses of the early and middle empires, this type of reception usually took place in the triclinium, which was generally the largest and most richly decorated room in the house. The triclinium was named after the three rectangular couches used to recline while eating. These sofas were placed around three sides of the room and were closely fitted together around a single central table,

2 Dunbabin (1996) 66-67. Cf. also Jones (1991).

aspects of the transformation of the Roman domus 81

one side of which was left for service. Strict social rules governed the arrangement of seating at dinner parties, which followed a hierarchical order dictated by differences in rank and status. The couch of honor was the lectus medius, opposite the empty side of the table with the right hand position assigned the highest 'consular' status (locus consularis). Next in importance came lectus summus, and lastly lectus imus. A nomenclator would announce the guests and show them to their couch and a place where they would recline crosswise, with their left elbow resting on a pillow and their feet at the foot of the couch.3

By the 3rd century A.D., literary sources, figurative representations and archaeological remains all point to an important change in fashion in the dining room, with the semicircular sofa known as the stibadium or sigma taking the place of the traditional three rectangular sofas. This dining arrangement is known from as early as the end of 1 century AD, when it was used exclusively for outdoor banquets, where the diners rested either directly on the ground or, more often, on a semicircular cushion. By the 4th century, however, dining on the stibadia was practiced for both open-air and regular indoor banquets and had supplanted the rectangular arrangement of triclinia, despite the occasional persistence of this traditional layout.4 One of the most striking archaeological signs of this change in the dining fashion in late antique prestigious residences is the large number of reception halls that are provided with an apse, where the path bath would be placed. The shape of the stibadi fell naturally to the shape of the apse, and the formal dining room thus developed as a combination of apse and rectangular or square hall.5

However, the semi-circular sofa could accommodate only a small number of diners - between 5 and 8 - while the traditional combination of three rectangular sofas could accommodate perhaps as many as 18 or 20 guests. It has been suggested that once the idea of ​​the diningon stibadia in an apsidal dining hall had become popular, an aristocrat wishing to host a larger number of guests would build the atriconch, a room with three apses, each providing space for a stibadium . this hypothesis, therefore the triconch design would have evolved naturally from the traditional use of three sofas

3 See Dunbabin (1991) 121-28 and Ellis (1991) 118-19.4 Dunbabin (1991) 128-36 and Rossiter (1991) 202-6. For outdoor banquets,

see also Salsa Prini Ricotti (1987) 135-84.5 For this view, see Ellis (1991) 119-20.

barbara polci82

triclinium, making larger gatherings possible.6 Rooms of this design are found in many prestigious residences of the late empire, among them the well-known examples of the villa at Desenzano on the shores of Lake Garda,7 the villa at Piazza Armerina8 (fig. 1) ) and the so-called Theoderic Palace at Ravenna (fig. 2.2).9

The ultimate luxury in this sense must have been the development of the aula trichora through the multiplication of niches, which enabled even more impressive gatherings of people. This was e.g. the case of the domus Cilonis, where the multi-apsidal refectory was later transformed into the church of Santa Balbinain Rom.10 Domus quae vocatur quinque accubita, a five-aps triclinium built in the episcopal complex of Ravenna by Bishop Neon in the mid-5th c. 11 may provide a further example, which was perhaps deliberately repeated on a smaller scale in the episcopal complex at Grado.12 The most impressive example of this type of multi-apsidal hall is the seven-apested chamber. at the House of Bacchus at Djemila in Numidia (fig. 2.1).13 A room of this size—almost 27 ml long, with apses over 4 m in diameter—provided sufficient space for about 50 guests reclining on stibadia. A very similar hall was founded in Constantinople on the northwest side of the Hippodrome. This has generally been recognized as the palace built for Lausus, a praepositussacri palatii of the first half of the 5th century, while the complex on its northern side has traditionally been identified with the palace of another praepositus, Antiochus.14 However, it has recently been argued , that both structures were different parts of the same complex – possibly the palace built for Antiochus, since his name was found on the bricks – and that the two chambers would have been used for the reception and entertainment of different ranks of guests.15

6 Dunbabin (1991) 129–30 and Ellis (1991) 119. For the origin of the triconch, cf. also Lavin (1962) 1–15.

7 Scagliarini Corlaita (1994) 43-58.8 Carandini et al. (1982); Wilson (1983).9 Johnson (1988) 79-85 og Baldini Lippolis (1996).

10 Guidobaldi (1986) 181-82.11 See De Angelis D'Ossat (1973) and Deichmann (1972) 94-112.12 Cf. De Angelis D'Ossat (1973) 269-73 and Lopreato (1988) 325-33.13 Lassus (1971).14 C.f. Muller Wiener (1977) 238-39 and Scagliarini Corlaita (1995) 865-66.15 Baldini Lippolis (1994) 298-2 The seven-ape hall has also been identi-

fied (erroneously in my opinion) as the Hall of the Nineteen Couches in the Great Palace of the Byzantine Emperors. See among others Verzone (1976) 41-47 and Bek (1983) 96.

aspects of the transformation of the Roman domus 83

Indeed, late Roman aristocrats increasingly needed to distinguish between the different classes of guests, which is clearly reflected by the presence of more specialized reception rooms in their residences. Whereas a triclinium in earlier houses would have been used both as a dining room and as an audience chamber where the lord of the house could conduct his business, most large houses built in the 4th century. and was later provided with at least two specialized reception rooms that served different functions.

Simon Ellis has interpreted these halls respectively as a private audience chamber where the dominus could meet with his clients, perhaps for greeting or 'good morning', and as a triclinium for

Fig. 1. Piazza Armerina. Plan of the villa. [C] = Corridor [G] = Grand Dining Hall [T] = Triclinium. After Ellis (1991).

barbara polci84

banquets.16 In some cases, such as in the 'Dux's palace' at Apollonia (fig. 3.3), the 'bishop's palace' at Aphrodisias (fig. 3.2) and the 'palace' above the theater at Ephesus (fig. 3.1). ,17 the audience hall would be built close to the street so that the less important customers could be excluded from the main part of the house, while more important customers and friends would be invited for dinner and entertainment in the main reception room, which was generally located outside the peristyle. In other cases, however, a magnificent dining room i

Fig. 2. Plan of three houses with magnificent dining rooms. [G] = large dining hall [T] = triclinium: [1] Djemila, House of Bacchus [2] Ravenna, Theodoric's so-called Palace [3] Mediana (Nià), villa. After Ellis (1991).

16 This concept was first expressed in Ellis (1985) and then further developed in Ellis (1988) and (1991).

17 For Apollonia, see Ellis (1985); for Aphrodisias, Erim (1969); for Ephesus, Keil (1932).

aspects of the transformation of the Roman domus 85

which, in order to receive the most important and influential friends, was placed on the main axis and entered through a separate entrance, thus securing privacy both for the rest of the house and for the meeting in question. Ellis has identified this second type of combination of two reception rooms at Theoderic's 'palace' in Ravenna, the house of the Bacchusat Djemila, the villa at Mediana near Nià and also in the villa at Piazza Armerina, where the 'great dining hall' in the form of an atriconch would have had no only a separate entrance, but also second, private peristyle.18

18 Ellis (1991) 120-22.

Fig. 3. Plan of three houses with audience chambers: [1] Ephesus, villa above the theater [2] Aphrodisias, 'Governor's Palace' [3] Apollonia, 'Dux's Palace'. After Ellis (1991).

barbara polci86

This trend appears to have reached its peak in the mid to late 4th century. century. phase of the villa at Littlecote in Wiltshire, where interestingly a large triconch hall was located away from the main villa building, almost isolated from the rest of the complex.19 This hall has been interpreted as part of an Orphic temenos, along with the adjacent structures , solely on the basis of the presence of a mosaic depicting Orpheus, which bears a vague resemblance to several other Orpheusmosaics known in Britain. However, the presence of a religious decoration related to Orpheus does not necessarily imply a place of worship: the use of images of Orpheus had actually become a common symbol of classical culture and a fashionable theme in the 3rd c.A.D., when Orpheus became associated with intellectual Christians with Christ.20 It is therefore more likely that the hall was simply a reception room located in a more secluded part of the complex, which enabled the owner of the house to conduct his business and social gatherings away from the domestic area .21

Nevertheless, literary sources clearly indicate that on some social occasions the two large public spaces of the house were also used together for two distinct stages of the same convivium: one for the reception and entertainment of guests before the meal, and the other where the guests would gather for the dinner itself.22 AroundA.D. 460, Sidonius Apollinaris records a visit to the villa of his friend Tonantius Ferreolus near Nimes, in southern Gaul, and describes some of the social activities that took place there:

I had hardly entered one or the other hall, then look! I found on the one hand opposing ballplayers stooping low amid the swirling developments of the disaster; in another quarter I heard the rattle of rattling dice-boxes, and of dice mingled with the rival shouts of the players; in another part, books in any number were ready to hand; you might have imagined yourself looking on the shelves of a professional scholar, or on the tiers of the Athenaeum, or on the towering presses of the booksellers. The arrangement was such that the manuscripts near the ladies' seats were of a devotional type, while those among the gentlemen's benches were works distinguished by Latin eloquence.23

19 Frere (1989).20 On this point see Murray (1981) 37-62.21 Toynbee (1981) and Ellis (1995) 173-75.22 Rossiter (1991) 200-202.23 Sidon. Ep. 2.9.4-5 (translation Anderson (1936) 453).

aspects of the transformation of the Roman domus 87

Sidonius also tells how some of the guests of the house sat in one corner of the room and engaged in a debate on the subject of Origen's writings until

a messenger approached from the head chef to tell us that the time for refreshment was near. He kept an eye on the progress of the hours as marked by the water clock, and when the fifth hour was just about to pass, it turned out that he had come just at the right time. The luncheon was at once short and sumptuous in the style of senators, who have an inherited and established practice of having plenty served in a few courses, though the meal is varied by having some of the meat fried and some stewed. While we sat over our wine, there were short stories, for amusement or instruction; they were started in two sets, bringing joy and edification respectively. To sum up, our entertainment was moral, elegant and sumptuous.24

A very similar account of aristocratic hospitality, but in an urban context, is found in Macrobius' account of the feast of Saturnalia.25

Macrobius describes a gathering of twelve guests, this time all men, who met before dinner in the bibliotheca of the house of the aristocrat Vettius Praetextatus.26 The gathering before dinner continued there for several hours, with the guests discussing the activities that were also permitted on feast days such as the proper treatment of slaves.27 But as Praetextatus himself notes, pre-dinner gatherings in the library were more usually an occasion for backgammon and drafts.28 Asin Sidonius' account, the debate was interrupted by a slave who announced that dinner was ready, after which the guests moved from the library to the triclinium to take their seats at the banquet.29

In both of these cases, therefore, the main reception rooms were not used in isolation, but were in fact closely connected. This could also have happened in most of the examples mentioned so far, as well as in other late Roman prestige residences that had more than one public room, such as in the villa of Patti Marina in Sicily30 or in the villas of Ecija and Rioseco de Soria in Spain.31 The rich domus

24 Sidon. Ep. 2.9.6 (translation Anderson (1936) 455-56).25 C.f. Rossiter (1991) 201.26 Macrob. Sat. 1.6.1 (translation Davies (1969) 49).27 Macrob. Sat. 1.7-24 (translation Davies (1969) 55-58).28 Makrob. Sat. 1.5.11 (translation Davies (1969) 47).29 Macrob. Sat. 1.24.22-25 (trans. Davies (1969) 158).30 Wilson (1990) 204-206.31 Gorges (1979) 374-75 and 403.

barbara polci88

recently excavated in Butrint, Albania, indicates the presence of a similar combination of two reception rooms in an urban context. The large triconch hall was possibly intended as a grand dining room, while the single-apex hall could alternatively have been used as an audience chamber or as a room for the entertainment of guests before dinner.32

Entertainment seems to have assumed greater importance on these occasions than previously. It has been suggested that one of the reasons for the adoption of the stibadium for indoor banquets could have been that this arrangement would leave more space available for entertainment, as with outdoor banquets where space restrictions did not apply.33 In the traditional rectangular triclinium, in fact, the space for service and entertainment must always have been limited and somewhat crowded. Music and a bit of dancing generally accompanied the dinner, but conversation within the group took precedence, and the idea of ​​the convivium was that the participants were their own entertainment. In contrast, the use of the semi-circular sofa kept the main area of ​​the room clear, leaving a practical large space in front of the diners. This triconch, where the entire central square could be used for service and entertainment, was even better suited for this purpose. Servants would use this area to bring the food from the entrance to a table which stood in the sofa's basket and was accessible to everyone without taking up the rest of the room. This arrangement would provide more space for entertainers, making it possible to host shows that were more extensive than before.34 Dining on the stibadia would also ensure a much better view of the performance taking place in the hall. At the same time, however, this arrangement would make the conversation in the group much less convenient. The entertainment took over and became one of the essential components of the dinner.

This phenomenon should not simply be considered a change in eating fashion, but actually has much deeper implications. Feeding

32 For the triconch house at Butrint, see Hodges et al. (2000) 248–50. For more examples of wealthy houses provided with grand reception rooms in both urban and rural contexts, see the impressive collection of examples presented in Sodini (1995) and (1997).

33 Dunbabin (1996) 77-78.34 Macrobius speaks of dancers performing ante triclinium, which may mean "in

in front of the sofas' but also 'outside the dining room': Makrob. Sat. 3.14.4. Also a passage by Olympiodorus mentions mimes performing 'before' the table: Olymp.FGH fr. 23, IV, 62, cited in Dunbabin (1996) 78, n. 49.

aspects of the transformation of the Roman domus 89

and entertaining friends and clients had always been one of the great social obligations of the Roman aristocrat, but became even more important in Late Antiquity. The decline of central government put an end to grand public spectacles and, at the same time, a growing trend towards the privatization of entertainment, which became a private concern of the upper classes. Members of the tea aristocracy filled part of this power vacuum, while their houses absorbed a number of public functions. This would accord with our current view of late Roman patronage, which was more autocratic and pervasive in character than before. Late antique aristocrats had greater political power than their earlier (and more) counterparts, and in the growing decline of central governments, they became the local power, taking almost complete control of the areas they dominated. As a result, many people, overwhelmed by the ever-increasing weight of taxes, were forced to resort to these local patrons and then became totally dependent on them.35 In the words of Peter Brown, "in the late empire all attempts to secure protection and redress. ofgrievances had to pass through a great man—a patronus".36 Social relations between patron and client became more formalized and ceremonial, and consequently also their architectural setting. The splendor of the entertainment that the host could offer would be one of the most important opportunities he had to impress his guests and customers, in the same way as the architecture of his house.

Moving up: The transfer of reception rooms to

Upper floor

Despite these important developments, the late Roman domus was still closely related to and derived from the traditional Romandomus. Mainly organized around a central courtyard, it was generally single-storey, and even where an upper storey existed, reception areas were usually found at ground level. In Italy, however, from the 5th century. and thereafter some fundamental changes occurred in the layout of the traditional Roman house. Although the opulent houses built at an earlier time continued to be occupied,

35 Ellis (1988).36 Brown (1971) 37.

barbara polci90

the 'classical' peristyle house gradually disappeared, and a new model for elite housing slowly began to emerge.37

Our knowledge of elite residences during this period of transformation is still quite limited and based on scant archaeological evidence and a few mentions in literary sources. Therefore, it is particularly difficult, if not impossible, to outline its development in the way that is possible for its earlier counterpart. However, an examination of the available data allows us to identify some interesting features and make some preliminary remarks about them. One of the most notable changes was the relocation of the main living areas from the ground to the upper floor.

Episcopal complex

Already in the 6th century reception rooms are found on the upper level in a number of episcopal complexes. In Ravenna, for example, a dining room was built that was used as a mensa canonica above the vivarium (a series of arches, niches and openings used to house various species of birds and fish for the bishop's table), probably under Bishop Maximian (546 -565) ).38 At Aquileia, the 4th century episcopal complex was rebuilt after a great fire, perhaps related to Attila's onslaught, and it seems likely that one of the halls used in the earlier phase of the complex , was converted into a warehouse, with columns possibly supporting a large reception hall on the upper floor.39 Similarly, in Geneva, a second floor hall was built between the end of the 5th century. and the middle of the 6th century after a fire.40 At Poreµ, a large chamber on the ground floor was converted into a warehouse, while columns were again used to support the large, impressive hall located on the upper floor.41 Also episcopium 'par excellence', the Lateran, was a two-story building from at least the 7th century. When the Ravenna army went there to rescue Sergius I, they found that both the upper and lower doors (tam inferiores quamque superiors) to the patriarchate had been closed.42 The evidence regarding the presence

37 For the end of the peristyle house, see Ellis (1988).38 Deichmann (1972) 109-10.39 Bertacchi (1985) 375-83.40 Bonnet (1986) 72-73.41 Matej´iµ (1995) 84-89. The presence of a hall on the upper floor is also postu-

late to the bishopric of Salona, ​​although the evidence in this case is much less clear: cf. Bertacchi (1985) 410-12.

42 LP I 373 .

aspects of the transformation of the Roman domus 91

of halls on the upper floors is not very well defined in this early phase, but by the middle of the 8th century, when Zacharias built a triclinium on the upper floor of the tower building, it is clear that they were a feature of the Lateran Palace. 43

However, the organization of the space into two or more floors and the relocation of living activities to the upper floors should not be considered to be exclusive to episcopal residences. It is more likely that the common presence of these features in the episcopal context can be explained as a result of the higher survival rate of these complexes due to the continuity of their use. This does not mean that the same phenomena did not also occur in secular house architecture, but only that the available evidence for them between the 5th and 9th centuries. is extremely spotty. In this sense, the residences of bishops can therefore be considered one of the most valuable indicators of how late Roman elite housing was gradually transformed into its early medieval form.

Second floor reception rooms in secular housing

In relation to a lay context, it is particularly interesting to take a closer look at the archaeological sequence at the villa complex excavated at San Giovanni di Ruoti, near Potenza, in Italy's southern region of Basilicata (fig. 4).44 Excavations have shown that around the year 400 e. NOK , a new villa was built over the remains of an earlier one that had been abandoned in the early 3rd century. Its main features were a bath building and a large apsidal hall on the south-east side, which would have been used by the lord of the house for his social gatherings. Around 460 AD the apsidal hall collapsed and some of the other buildings were damaged, probably after an earthquake. The villa was reconstructed and practically doubled in size. However, the new complex was not arranged around a peristyle, and its main units were separated only by narrow passages. The apsidal hall was demolished and replaced with a new one located on an upper floor at the northern end of the site. In addition, several new buildings were erected, such as a tower and antechambers adjacent to the apsidal hall. Some parts of the complex were then covered with mosaics. Most of the new buildings had two floors, with the most important rooms located on the upper floor, as was the case with the new great apsidal hall.

43 LP I 432.44 Small and Buck (1994).

barbara pollci92

Fig. 4. John of Ruoti. Reconstruction of the villa, approx. 460-535. After Small and Buck (1994).

aspects of the transformation of the Roman domus 93

Visitors wishing to enter the hall would have entered the complex from a porticoed entrance on the north side, close to a small room identified as a porter's lodge. They would then have reached the upper floor by means of the stairs, finally arriving at the antechambers of the hall without approaching the central part of the complex.

'Palazzetto' excavated at Monte Barro, near Lecco, in northern Italy, and dated to the first half of the 6th century. is also interesting in this context.45 In this case, the building was arranged symmetrically around three sides of a large square courtyard. However, the main room of the complex appears to have been located on the top floor of the building, above a large room probably used for food storage. Remarkably, fragments of painted plaster and important metal finds, such as a bronze hanging crown, have been found during the excavations, suggesting that the chamber on the upper floor was used as a kind of aula regia by the owner of the house.

In urban contexts, the archaeological evidence for the early phase of this transformation is even rarer. This may be because the stock of dwellings that survived from antiquity was more than sufficient for the needs of a much reduced population. In the case of Rome, both literary and archaeological evidence shows that the surplus of Roman houses continued to be used throughout our period, and that housing was also installed in other types of pre-existing structures, such as porticoes and other public buildings. fact that we have no evidence of newly built houses until the 9th century. should also be considered in this context. The city's population may not have needed to build new houses, simply because they could still use the large number of existing ones.46

Early medieval residence in the See of Ravenna

We have more information about urban dwellings from the early Middle Ages from written sources from the 7th-8th century. century. forward. Of particular importance is the evidence from the Codex traditionum ecclesiae ravennatis, also known as the Codice Bavaro. It consists of a collection of early medieval leases from the 7th to the 10th century, which contain very detailed descriptions of residences belonging to the bishopric of Ravenna.47 We shall

45 Brogiolo and Castelletti (1991).46 Coates-Stephens (1996).47 Cf. Cagiano (1972a) and (1972b).

barbara polci94

take a closer look at some of these documents, especially those relating to Rimini, where the oldest residence in Codexis recorded a house rented between 688 and 705 by a certain Johannes vicariusnumeri ariminensium.48 It was a two-story building with a canapha anda. calidarium on the ground floor and a triclinium and cubiculi duos on the upper floor. Next to the house there was a courtyard with an anor yard and a well. The house was covered with brick and the ground floor was built of stone. The material used for the construction of the upper floor is not specified, although it is likely that wood was used. The house rented by the two sisters Borsa and Susanna between 748 and 769 was very similar. This was also on two floors with acanapha on the ground floor, three bedrooms and a triclinium on the upper floor and a curte et orto.49 A further house rented almost a century later (between 834 and 847) in the same city also had three bedrooms upstairs and a canapha et coquinola et caldariolo on the ground floor.50

A similar house, although one with greater pretensions, was rented to the magister militum Mauricius and his wife Petronia in the early 10th c. 51. The building overlooked the forum and was located on the platea publica que pergit a pusterula sancti Thomae apostolic a quadrubioubi est petra ociosa. It included a courtyard, an orchard with a well and a tower. On the ground floor was a canapha and a station (perhaps a shop), while the kitchen was in a separate building. On the upper floor were a triclinium and five bedrooms. The house rented to Johannesand Anna in the 19th century. was also of interest as it included a balneocum vasos posturios duos cisternas duas and a porch.52

Other documents included in the Codex as well as in other collections describe very similar houses in a number of other places, for example in Lucca, where a number of documents date between 757 and the 11th century. refers to two-story houses of the same type as the Riminesi building.53 Many later documents of a similar nature could also refer to houses built at an earlier time. This was probably the case with the Ravennate document from 975, which described two stone houses, both two-storey, which also had a bath,54 and possibly

48 Cagiano (1972a) 135-36 og Cagiano (1972b) 178, n. 54.49 Cagiano (1972a) 136 og Cagiano (1972b) 178, n. 55.50 Cagiano (1972a) 136 og Cagiano (1972b) 178, n. 59.51 Cagiano (1972a) 136 og Cagiano (1972b) 178, n. 61.52 Cagiano (1972a) 137 og Cagiano (1972b) 178, n. 63.53 Belli Barsali (1973) 488-99 og 542-52.54 Cagiano (1972a) 138.

aspects of the transformation of the Roman domus 95

be true of the 10th century two-story dwellings known in Bari, 55 Rome, 56

Naples57 and elsewhere. The picture that emerges from these descriptions is quite

consistent. The type of house that replaced the ancient Roman domus generally consisted of a main building, often of two stories, described using terms such as casa solarata and casa terrinata orpedeplana to refer respectively to a two-story building and a one-story house. The one-story house was generally simpler and had little functional division of space, while the two-story solariata was slightly more complex. Houses were usually rectangular in shape and were built in recycled stone and brick. In some cases, only the ground floor was built of stone, with the upper floor built of wood or clay. Roofing consisted of clay tiles or shingles, as evidenced by the frequent use of such expressions as cum tegulis et imbricibus and cum scindolis.58 The distribution of the different rooms was similar in most of the examples described: the ground floor was generally devoted to agricultural activities, shops and stables and for domestic activity, as indicated by the quite common presence of a canapha, a kind of warehouse, and of a coquina, the kitchen, often consisting of a basic open hearth. Sometimes the kitchen was separate from the real house, perhaps for security reasons, and was shared between different households. Ground floors often contained wells, cisterns and, in some rare cases, a balneum or calidarium, possibly a single room provided with a basic drainage system. Living activities were limited to the upper floor, with a triclinium or a sala and cubiculaor camere that opened the main room. Since sala and triclinium or cubicula and camere are never mentioned together in the same document, they may refer to the same room. There are no references to stairs (internal or external) leading to the upper floor. Backyards or orchards for growing vegetables and fruits are also often mentioned.

Early medieval residence in Rome

The image from the textual evidence is confirmed by the impressive remains of the two houses excavated at the Forum of Nerbaby Riccardo Santangeli Valenzani, the best preserved residential area

55 Guillou (1976) 145 .56 Hubert (1990) 176-84.57 Arthur (1991) 768-99.58 Cagiano (1972a) 365 and Cagiano (1972b) 178, n. 54.

barbarapolici96

structures so far revealed from early medieval Italy (fig. 5).59 Dated by the excavator to the 9th century, both houses were irregular rectangular structures, built in stone, with walls made of carved blocks of peperino used from the Forum. The first house, built on the street front, has an exceptionally well-preserved vestibule, built in blocks of peperino with high-quality brick and mortar inserts that form the facade. The excavation also revealed traces of an external wooden staircase leading to an upper floor. A well and a cesspool were included in this structure as well as a latrine located on the first floor. Nothing remains of the building's interior partitions, except for a large post

Fig. 5. Rome, Forum of Nerva. Reconstruction of the excavated houses. After Santangeli (2000).

59 Santangeli Valenzani (1997) 64-70 and Santangeli Valenzani (2000).

aspects of the transformation of the Roman domus 97

hole in the floor, interpreted as the location of a supporting pier. Regarding the first floor, there is nothing to indicate whether the construction was made of wood or brick. But this was the place where the cubicles, the bedrooms of the household and possibly a room or some other kind of hall must have been located. On both sides of the building, the excavations revealed two open areas within the enclosure wall. Apart from the cesspool, this area was probably used for livestock, as evidenced by the remains of a tension point in the wall and part of a drinking trough. This second house is not as well preserved, but still contains some interesting features. The brickwork of this building was in peperino with patches of recycled bricks. The ground floor room was vaulted, a feature probably contemporary with the original construction. This second house also has traces of an external staircase on the west wall leading to a first floor.

Santangeli Valenzani has convincingly argued that the two structures are representative of a relatively well-defined type of residential building characteristic of the houses of the wealthier classes of the time. This picture is also confirmed by the domus recently excavated by Stefano Coccia's team on Via Iugaria, at the western end of the Basilica Julia, and by Etienne Hubert's detailed analysis of documentary sources related to dwellings of a slightly later phase, which seem to confirm that the upper floors were generally associated with aristocratic residences.60

It is important to note that by the beginning of the 9th century, the relocation of the ofriclinia to the upper floor had become quite popular, not only in the residences of people of some means, but also, on a more impressive scale, in proper palace structures. The Lateran Palace, as we have mentioned above, was equipped with a triclinium on the upper floor of Zacharias as early as the middle of the 8th century. Two more magnificent triclinia were added to it by Leo III (795-816). The earlier of the two, used by 799, was the so-called Aula Leonina, a triconch hall located on the upper floor of the Patriarchium, used as a banqueting hall by the Pope and Cardinal on certain feasts. The Liber Pontificalis claims that this triclinium was larger than similar earlier halls in the Lateran and was magnificently decorated with porphyry and white columns and paved with precious marble.61 Shortly after 800, Leo III built another

60 Hubert (1990) 176-79.61 LP II 3-4.

barbarapolici98

triclinium at the Lateran, which was also located on the top floor of the palace. It consisted of a hall, fully 68 m long, projecting from the northern flank of the basilica, and with a primary central apse and five conchs on each side.62 The same pope also decorated the papal complex in the Vatican with a large triconch triclinium, located , again, on the top floor of the complex.63

"A Lombard architecture"?

The palace built in Salerno by the Lombard duke Arechis II (758-787) also consisted of a two-story building.64 The Aula regia, located on the piano nobile, was accessed via a monumental external staircase and was directly connected to the still-surviving palace . chapel. Archaeological investigations have also shown that the ground floor probably has the form of a loggia. A close parallel to the Ducal Palace of Salerno is found in the palatial structure found in the monastery of San Vincenzo al Volturno, interpreted by the excavators as an area for receiving and housing the monastery's distinguished guests (fig. 6).65 This was also a two-story structure, the ground floor consisting of three simple basements, probably used for stables and other residential purposes, such as accommodation for storage rooms or warehouses. An external staircase gave access to the upper level, where richly decorated apartments were located.

One must obviously be careful in generalizing on the basis of so few examples. It seems clear, however, that the 8.-9. The palatial structures described here, albeit on a more impressive scale, were very similar to their contemporary, more modest counterparts. While elite residences in the ancient world were characterized by a one-story construction arranged around an open peristyle, early medieval elite residences were very compact architectural units, divided into two stories. The ground floor was generally reserved for domestic activities, stables and warehouses and had a much smaller number of rooms than the late antique elite domus. Residential activities, meanwhile, were concentrated on the top floor, which was clearly separated from the service area.

62 LP II 11.63 LP II 8.64 Cagiano (1971) 14-15; Delogu (1977) 42-50 and Peduto et al. (1988).65 Hodges and Mithen (1993) and Hodges, Gibson and Mitchell (1997) 251-

61.

aspects of the transformation of the Roman domus 99

Fig. 6. St. Vincent to Volturno. Reconstruction of the palace, approx. 800-2 After Hodges, Gibson and Mitchell (1997).

Cagiano De Azevedo interpreted this new tendency to live mainly on the first floor as something that was typically Lombard. According to this hypothesis, the Lombards wanted to live in open spaces and simple buildings and preferred to live on the first floor rather than

barbara polci100

the ground floor because the upper level was more airy and cool.66

However, as we have seen in some of the examples presented above, this type of arrangement is not only attested in Lombard contexts, but actually seems to have been quite widespread throughout the Italian territory. It is also attested before the Lombard conquest of Italy, as in the cases of San Giovanni di Ruoti and the palazzetto at Monte Barro or also at the episcopal complex in Ravenna.

Cagiano de Azevedo's theory in favor of a Lombard culture, characterized by a vertical orientation of dwellings, in contrast to the Roman or Byzantine tradition, where dwellings were usually one-story, is made even more unlikely by the existence of some interesting examples of double -storey houses in territories that cannot be safely defined as belonging to a 'Germanic' cultural tradition. On the basis of this evidence, other scholars have instead characterized the origin of this new style as a product of Byzantine influence.67

A similar kind of arrangement is actually found in the East in earlier and more common domestic structures. For example, documents in the form of papyri from Egypt provide interesting information about the organization of space across different floors of buildings in residences from the 1st to the 3rd centuries. AD At the bottom of the house there was usually a cella (kellios), while dining rooms (triclinoi or symposioi) and bedrooms (akkubitois) were generally on the upper floor.68 In the account in St. Mark's Gospel (Mark 14:12-16) ), followed closely after St. Luke, the Last Supper was celebrated in a large upper chamber, then known as the cenaculum. The room in which the fire of Pentecost appeared to the apostles was also located on the upper floor, while the building in which the apostles and Mary lived after the crucifixion also had an upper room, anageion or hyperoon, as mentioned in Acts 1:13-14. Excavations in Palestine have brought to light a number of dwellings dating to the 1st and 2nd centuries AD, reminiscent of those described in the Gospel. In Jerusalem, the remains of the so-called 'Burnt House', for example, indicate that

66 As Cagiano De Azevedo ((1974) 34) put it: "they wanted open spaces and simple buildings to live in, but they did not try to live on the ground floor anymore ... life used to take place in the rooms on the first floor and no longer on the ground floor, probably because more airy and fresh. In this sense it is possible to speak of a Lombard architecture.".

67 Guillou (1976) 151-54.68 Ellis (2000) 96.

aspects of the transformation of the Roman domus 101

domestic activities took place on the ground floor, while living rooms were located on the upper floor.69 Also in Jerusalem, the house known as 'The Great Palace' had storage rooms in the lower part of the house, while living rooms and a large reception were built in the upper part.70 At the city of Pella, meanwhile, on the eastern side of the Jordan River, the excavators found a house with storerooms and stables on the lower level and traces of living quarters on the upper floor.71 Other interesting examples of a later date come from southern Syria: the 5th c. at Il Medjdel, for example, had a single large square room on the ground floor, probably used as a stable, while the upper floor was divided into four rooms and was accessed via an external staircase (fig. 7).72 The so-called palace in Bostra also has a triconical reception room on the upper floor (fig. 8).73 Many other examples could be added to this list. However, my main concern here is not to present the housing pattern in the central and eastern Mediterranean. I will limit myself to the observation that the widespread adoption of second-story housing in the central and eastern Mediterranean was probably part of a process in which urban space became more ruralized, in which building types already known in a rural context gradually was introduced in the city.

But while I would reject Cagiano de Azevedo's claim in favor of the Germanic origin of this architecture in Italy, neither would I claim that it was the product of Eastern or North African influence. I believe that the origin of this phenomenon is most likely to be found in other practical reasons, rather than within ideas about the impact of foreign influences and ethnic groups, as has been the case in the past. For early medieval Naples, Paul Arthur argued for two-story houses due to the need to house a growing population in a limited urban space. He suggests that after a decline in population size in the 4th-6th of the inhabitants increased again, which explains the common use of two-storey buildings.74 However, this explanation is less applicable in many other contexts. Safely in

69 Hirschfield (1995) 58,70 Hirschfield (1995) 59-60,71 Hirschfield (1995) 69-70,72 Ellis (2000) 91-93,73 Ellis (2000) 93,74 Arthur (1991) 768-69.

barbara polci102

Fig. 7. Il Medjdel. Plan and erection of the house. After Ellis (2000).

aspects of the transformation of the Roman domus 103

Fig. 8. Bostra. Plan and elevation of the 'Palace'. After Ellis (2000).

barbara polci104

in Rome's case it is hardly credible. Despite still being the largest city in the western Mediterranean, Rome's population has declined quite dramatically. While Procopius's tale that there were only 500 men left in Rome at the end of the Gothic Wars sounds impossibly pessimistic, the estimated number is still quite small and will not explain the development of this new housing model.75

It could be suggested that, at least in some cases, the transfer of reception halls to the upper floor was due to defensive considerations: in Aquileia and Geneva, for example, the hall on the upper floor was built after a fire. Gregory of Tours mentions the case of Angers, where Bishop Andoveus built a dining room, a solarium, on top of the city walls, perhaps on a tower.76 It is also worth noting that at the end of the 4th century there were country villas such as e.g. represented in the famous mosaic of Dominus Iulius at Carthage, increasingly assumed a fortified aspect, and walled enclosures became a more common feature.

Nevertheless, a number of other important aspects should be taken into account, especially regarding the urban context. First, one should consider the process of ruralization of the city. Animals, as mentioned above, were kept on site. The relatively common presence of backyards and orchards in the sources, attested archaeologically by 'dark soil' deposits, indicates the presence of cultivated areas in a domestic context, providing vegetables and other foodstuffs for the family. It is also possible that in some cases the vertical growth of the city could also have been due to the problems of sewerage and drainage systems and water supply. Flooding and its disastrous consequences were well known to the ancient Romans, but seem to have become more acute in the 6th and 7th centuries. than they had been previously.77 This was possibly due in part to political fragmentation: difficulties with water and drainage now faced weaker governments, often without the necessary financial and technical means to maintain the old Roman system. It is likely that problems associated with drainage and control of sewage and waste disposal are increasingly leading to the vertical growth of the city. The transfer of residential areas to the upper floors of housing could perhaps be considered one of the consequences of this rural formation and general impoverishment of the city. In this new settlement pattern, the land

75 Hodges og Whitehouse (1983) 48-52,76 Cagiano (1974) 649,77 Squatriti (1998) 66-96.

aspects of the transformation of the Roman domus 105

floor was used for agricultural and domestic activities and was no longer considered the best location for the living areas of the house.

Finally, one should consider the simplicity of these homes. To judge from their austerity, the desire to emulate an elite modus vive appears to have diminished considerably. This trend could be seen as having both an economic and an ideological dimension. On the one hand, these changes can be considered a consequence of the lack of resources that followed the collapse of the ancient Roman tax system. This situation may have led to the emergence of a new social system, characterized by the decline or disappearance of intermediate social classes, and therefore polarized between a rich and powerful minority and the rest of the population. On the other hand, it could be suggested that another consequence of the transformation of the hierarchical structure may have been the emergence of new ideological factors, characteristic of early medieval societies, which sanctioned the display of wealth mostly in religious contexts.78

It can therefore be argued that aristocrats from the early Middle Ages may have preferred to express their status and prestige more through the construction and foundation of monasteries and churches than through their own residences. In the words of Chris Wickham, "what really changed was also the rhetoric of construction, which, in the absence of the leading role of a truly rich public power, could easily fall sharply in scale."79

Conclusion

The purpose of this article was to analyze some aspects of the transformation of elite residences in Italy in the period between the dissolution of the Roman imperial system and the rise of the Carolingian Empire. One of my primary concerns was to look at the ways in which their catchment areas of affluent residences were defined and articulated to express the power, wealth and status of their owners.

In the first part of the article I outlined the evidence for a relatively formal and ceremonial domestic architecture in the later Roman Empire, characterized by the presence of very elaborate reception facilities, used as spectacular settings for reception and entertaining

78 Jf. Wickham (1989).79 Wickham (1989) 142.

barbara polci106

customers and guests. In the second part, I examined the new model of aristocratic housing that gradually began to appear in Italy after the dissolution of the Roman Empire. The analysis of the limited archaeological and documentary evidence has shown a rather different picture than that identified for Late Antiquity. Early medieval elite residences were compact architectural units divided into two floors. Domestic activities were concentrated on the ground floor, while the upper floor was clearly separated and reserved for the owners' living space. The transfer of living activities to the upper floor has been identified as one of the most characteristic features of the process of transformation of the Roman domus into the early medieval house. It is proposed here that the origin of this feature can be linked to defensive strategies and, at least for urban contexts, to the increasingly ruralization of cities, characterized by the presence of agricultural activities and also by the lack of water mains maintenance and drainage systems and decline in efficient waste disposal, which gradually led to the vertical growth of the city. In this sense, the relocation of residential areas to the upper floors of dwellings could therefore be considered another consequence of the disintegration of urban administration in the late Roman and post-classical world.

aspects of the transformation of the Roman domus 107

Bibliography

Arthur P. (1991) "Naples: a case of urban survival in the early Middle Ages?", MEFRM 103 (1991) 759-84

Baldini Lippolis I. (1994) "Houses and Palaces in Constantinople Between the 4th and 6th Centuries", Cultural Courses on Ravenna and Byzantine Art 41 (1994) 279-311

——— (1996) "Arrangement and Decoration of the Imperial Palace at Ravenna", Cultural Courses on Ravenna and Byzantine Art 43 (1996) 1-31

Bavant B. (1989) "Lifestyle and Urban Habitat in Central Byzantine Italy (6th-7th Centuries)", MEFRM 101 (1989) 465-532

Bek L. (1983) "Questiones Conviviales: the idea of ​​the triclinium and the staging of convivial ceremony from Rome to Byzantium", AnalRom 13 (1983) 81-107

Belli Barsali I. (1973) "The topography of Lucca in the VIII-X centuries", i Proceedings of the V International Congress of Studies on the Early Middle Ages (Lucca, 1971) (Spoleto 1973) 461-554

Bertacchi L. (1985) "Contributions to the study of early Christian episcopal palaces: the cases of Aquileia, Parenzo and Salona", AquilNost 56 (1985) 361-412

Bonario M. (1983) "Society Music from the Ancient Latin World to the Middle Ages", in Social Performances from Classical Antiquity to the Italian Courts of the 15th Century (Viterbo 1983)119-47

Bonnet C. (1986) "The Reception Rooms of the Episcopal Group in Geneva", RACrist65 (1986) 71-86

Brogiolo G.P and L. Castelletti (1991) eds. Archeology at Monte Barro I. The Great Building and the Towers (Lecco 1991)

Cagiano de Azevedo M. (1971) "Does a Longobard Architecture and Town Planning Exist?", in Proceedings of the International Conference on the Theme: The Civilization of the Longobards in Europe (Rome and Cividale 1971) (Rome 1971) 1-41

——— (1972a) "Early Medieval Buildings on the Northern Adriatic Coast", AquilNost43 (1972) 133-46

——— (1972b) "Le case descritte dal Codex traditionsum Ecclesiae Ravennatis", Rendiconti dell'Accademia Nazionale dei Lincei 27 (1972) 159-81

——— (1974) "Urban Aspects of Early Medieval Cities", in Weeks of the Italian Center of Studies for the Early Middle Ages XXI (Spoleto 1974) 641-77

Carandini A. (1989) "A Politics of Archaeology", Times Literary Supplement 4524 (1989)10

Carandini A., Ricci A. and de Vos M. (1982) Philosophical. Villaen in Piazza Armerina (Palermo 1982)

Coates-Stephens R. (1996) "Dwelling in Early Medieval Rome, AD 500-1000", PBSR65 (1996) 177-232

De Angelis D'Ossat G. (1973) "On the destroyed hall of five accubita in Ravenna", Cultural Courses on Ravenna and Byzantine Art 20 (1973) 263-73

Deichmann F. W. (1972) "Studies on the Lost Ravenna": Felix Ravenna 103-4(1972) 61-112

Delogu P. (1977) Myth of a Southern City (Salerno, VIII-XI Centuries) (New Middle Ages2) (Naples 1977)

Dunbabin K. M. D. (1991) "Triclinium and stibadium", i Dining in a ClassicalContext, red. W.J Slater (Ann Arbor 1991) 121-48

——— (1996) “Communal Spaces: Dining and Entertaining in the Roman Villa”, JRA 9 (1996) 66-80

Ellis S. (1985) "The "Palace of the Dux" at Apollonia and Related Houses", in Cyrenaicain Antiquity, ed. G. Barker, J. Lloyd and J. Reynolds (BAR International Series236) (Oxford 1985) 15-26

barbara polci108

——— (1988) “The end of the Roman house”, AJA 92 (1988) 565-76——— (1991) “Power, architecture and interior design: how the late Roman aristocrat

appeared to his guests”, in Roman Art in the Private Sphere, ed. E. Gazda (AnnArbor 1991) 117-34

——— (1997) “Late Antiquity dining: architecture, interior design and behaviour”, inDomestic Space in the Roman World: Pompeii and Beyond, ed. R. Lawrence and A. Wallace-Hadrill (Journal of Roman Archeology Supplementary Series 22)(Portsmouth 1997) 41-51

——— (2000) Roman Housing (London 2000) Erim K. T. (1969) “The Ninth Excavation Campaign at Aphrodisias in Caria

1969”, Turk Arkeoloji Dergisi 18 (1969) 87-110Gorges J. P. (1979) Hispano-Roman villas. Inventory and archaeological issues.

(Paris 1979) Guidobaldi F. (1986) "Single-Family Dwellings in Late Antique Rome",

in Roman Society and the Late Antique Empire II, Rome: Politics, Economy, Urban Landscape, ed. A. Giardina (Bari 1986) 165-240

Guillou A. (1976) "The resting place in Byzantine Italy: Exarchate, Sicily, Catepanato (VI-XI centuries), in Proceedings of the international colloquium of medieval archeology (Palermo-Erice, 20-22 September 1974) (Palermo 1976) 140 -54

Hirschfeld Y. (1995) The Palestinian Settlement in the Roman-Byzantine Period (Jerusalem1995)

Hodges R., Bowden W., Gilkes O. and Lako K. (2000) "Late Roman Butrint, Albania:survey and excavations, 1994-98", Archeologia Medievale 27 (2000) 241-57

Hodges R., S. Gibson and J. Mitchell (1997) "The making of a monastic city. The architecture of San Vincenzo al Volturno in the ninth century", PBSR 65(1997) 233-86

Hodges R. and Mithen S. (1993) "The 'Southern Church': a late Roman funerary church (San Vincenzo Minore) and the Hall of Distinguished Guests", in San Vincenzoal Volturno 1, ed. R. Hodges (Archaeological Monograph of the British Schoolat Rome 7) (London 1993) 123-90

Hodges R. og Whitehouse D. (1983) Mohammed, Charlemagne and the Origins of Europe:Archaeology and the Pirenne Thesis (London 1983)

Hubert E. (1990) Urban Space and Habitat in Rome from the Tenth Century to the End of the Thirteenth Century (Rome 1990)

Johnson M. J. (1988) "Toward a history of Theoderic's building program", DOP42 (1988) 73-96

Jones C. P (1991) "Dinner theatre", i Dining in a Classical Context, red. W. J Slater (Ann Arbor 1991) 185-98

Lassus J. (1971) "The room with seven apses of Djemila-Cuicul", AntAfr 5 (1971) 193-207

Lavin I. (1962) "The Lord's House: Aspects of the Role of the Palace Triclinia in the Architecture of Late Antiquity and the Early Middle Ages", ArtB 44 (1962) 1-27

Lopreato P. (1988) "The Excavation of the Bishopric of Grado", Antichità AltoAdriatiche 32(1988) 325-33

Matej'iµ I. (1995) "The bishop's palace in Pore". Results of Recent Exploration and Restoration", Hortus Artium Medievalium 1 (1995) 84-89

Miller, M. C. (1991-92) "The Development of the Archbishop's Residence at Ravenna", Felix Ravenna 141 (1991-92) 145-73

Müller-Wiener W. (1977) Pictorial Dictionary of the Topography of Istanbul (Tübingen 1977) Murray S. C. (1981) Rebirth and Afterlife. A Study of the Transfiguration of Some Pagans

Pictorial Material in Early Christian Funerary Art (BAR International Series 100) (Oxford1981)

aspects of the transformation of the Roman domus 109

Peduto P. et al. (1988) "An Access to the History of Salerno: Stratigraphies and Materials in the Longobard Palace Area", Rassegna Storica Salernitana 10 (1988) 9-63

Rossiter J. J. (1991) "Convivium and villa in Late Antiquity", i Dining in a ClassicalContext, red. W. J. Slater (Ann Arbor 1991) 199-214

Salsa Prini Ricotti E. (1987) "The Significance of Water in the Roman Garden Triclinia", in Ancient Roman Villa Gardens, ed. E. MacDougall (Washington 1987) 135-84

Santangeli Valenzani R. (1997) "House-building and urban aristocracy in Roman in the early Middle Ages", in 1st National Congress of Medieval Archaeology, ed. S. Gelichi (Pisa 1997) 64-70

——— (2000) “Residential construction in early medieval Rome”, in Early MedievalRome and the Christian West, ed. J. Smith (Leiden 2000) 101-12

Scagliarini Corlaita D. (1994) “The Villa in Desenzano. Architectural and decorative events", in Studies on the Roman villa of Desenzano (Milan 1994) 43-58

——— (1995) "Polygonal Environments in Late Antique Residential Architecture", Cultural Courses on Ravenna and Byzantine Art 42 (1995) 837-73

Small A. M. og Buck R. J (1994) The Excavations of San Giovanni di Ruoti (Toronto1994)

Sodini J. P (1995) “Habitat de l'antiquité tardive (1)”, Topoi 5/1 (1995) 151-218——— (1997) “Habitat de l'antiquité tardive (2)”, Topoi 7/ 2 (1997) 435-577Squatriti P. Water and Society in Early Medieval Italy, AD 400-1000 (Cambridge 1998)Toynbee J. M. C. (1981) “Apollo, Beasts and Seasons: some thoughts on the Littlecote

mosaic", Britannia 12 (1981) 1-5Verzone P. (1976) "The Destruction of the Imperial Palaces of Rome and Ravenna and

restructuring of the Lateran Palace in the 9th century in relation to that of Constantinople", in Rome and the Carolingian Age. Proceedings of the Study Days, 3-8 May 1976 (Rome 1976) 39-54

Wickham C. (1989) "Italy and the Early Middle Ages", in The Birth of Europe. Archeology and Social Development in the First Millennium AD, ed. K. Randsborg (AnalectaRomana Instituti Danici Suppl. XVI) (Rome 1989) 140-51

Wilson R. J. A. (1983) Piazza Armerina (London 1983)——— (1990) Sicily under the Roman Empire: the architecture of a Roman province, 36 BC.

AD 535 (Warminster 1990)

this page is intentionally blank

aspects of the transformation of the Roman domus 111

PART 3 ECONOMICS

this page is intentionally blank

late antiques trade 113

LATE ANCIENT TRADE: RESEARCH METHODOLOGY

& MARKET PRACTICES

SEAN A. KINGSLEY

Abstract

Trade patterns offer diverse insights into ideology, politics and social orders between the 4th and 7th centuries AD. While different trade models have been formulated for the Western Mediterranean, the East is incomparably very poorly understood. This paper examines trade in the East, with particular reference to Byzantine Palestine, by considering three key themes: rural production, urban economies, and overseas trade as represented by shipwrecks and pottery. It is argued that more focused research strategies are needed to maximize available sources.

Introduction

The study of late antique exchange requires the negotiation of a bewildering and diverse array of data, from obscure historical texts to ship cargoes deep beneath the Mediterranean. That trade played an important role in society, such as generating revenue to pay government taxes, is uncontroversial. However, when seeking more precision on important topics, there are still many open questions. What was the content of the economic portfolios of the eastern Mediterranean provinces in Late Antiquity? What products were the basis of regional economic stability and prosperity? How far did specific agricultural and industrial products travel and in what quantities? What strata of society benefited from exchange: was it only the landed elite or village communities as well? Were the merchants of Late Antiquity state agents or private entrepreneurs?1 Surprisingly

1 This article incorporates research carried out for my D.Phil thesis at the University of Oxford (Kingsley (1999a)). I am eternally grateful to my inspirational supervisors, the late John Lloyd and Bryan Ward-Perkins, and equally to ClaudineDauphin, for ongoing support and advice. Thanks to Luke Lavan for inviting me to present my thoughts on this subject at the Oxford meeting of 'LateAntique Archaeology' in March 2001.

L. Lavan og W. Bowden (red.) Theory and Practice in Late Antique Archaeology(Late Antique Archaeology 1 2003) (Leiden), s. 113-138

sean a. kingsley114

topics of this nature remain under investigation. This is despite the potential of trade as a window into the social structures and ideologies of Late Antiquity, from the role of the humble farmer to the economic policy of the state.2

Part of the reason these questions have been neglected is that the methodology for studying ancient Mediterranean trade is still underdeveloped. In this article, however, I will not attempt a total critique of current approaches, but focus on three key topics for trade studies: rural production, urban economies, and the study of overseas exchange through Late Antique shipwrecks and pottery studies. For each topic, I will examine how appropriate methods or field practices can significantly advance our knowledge. For the sake of clarity, this discussion is limited to the Eastern Mediterranean, with particular reference to Byzantine Palestine, where investigations and excavations over the last 20 years have yielded a wealth of data, the extent of which is so far unmatched elsewhere in the East. 3

The East, stretching from modern Turkey to Egypt, has seen very little research on late antique macroeconomics. In the western Mediterranean, on the other hand, trade models have been discussed for decades. 4

It is widely recognized that Rome was heavily dependent on a continuous flow of North African grain and oil, and that the state's development of sea routes facilitated open trade. This is reflected in the archaeological record in the form of African red slipware produced in the area of ​​modern Tunisia.5 Simon Keay's research has suggested that much surplus state oil purchased as tax from Tunisia was sold commercially from state-owned olive oil depots in Rome. The penetration of oil via Rome was thus commercial, but deliberately facilitated by state initiatives.6 Elsewhere in the West, David Mattingly has examined African oil production and export in great detail, Michael Fulford has discussed economic interdependence in Carthage, Simon Loseby has assessed trade flows for late antique Marseilles and Clementine

2 For the importance of trade in understanding Roman society, see Storey (1999). Wickham (1988 and forthcoming) continues to argue forcefully that exchange is at the heart of understanding late antique society.

3 For a discussion of the sources, including 2930 sites of Byzantine date in Israel, see Dauphin (1998); also Kingsley and Decker (2001).

4 e.g. Jones (1974); Hopkins (1983b).5 For current information on forms of ARS production, cf. Mackensen (1998a),

(1998b).6 Keay (1984) 414, 424.

late antique shop 115

Among other things, Panella has investigated import patterns for Italy. Most of this work has been summarized by Paul Reynolds.7

If the Roman economy is an academic battleground in the West, as Keith Hopkins noted in the introduction to Trade in the Ancient Economy,8 in the Eastern Mediterranean it is an academic wasteland. In general terms, research strategies used for urban excavations and landscape surveys tend to be far less sophisticated than in the West, generally not being formulated in ways that are systematic or sensitive enough to reveal economic patterns. Where useful information has been forthcoming, it is more by chance discovery than by design.9 This situation is frustrating because it was the eastern provinces that were the beating heart of the later Roman Empire between the early 4th and mid-7th .century.10 It is the purpose of this paper to examine some of the ways in which appropriate methodologies can unlock the potential of this area and form the basis of a new understanding of Eastern trade.

Rural districts

'Primitivist' ideology has long envisioned late antique rural life as depressed and overtaxed, with peasants trapped in a social caste system that prevented social fluidity.11 Much of this gloomy background to late antiquity derives from A. H. M. Jones's interpretation of historical texts, written before archaeological evidence became widely available. For example, using papyri from Antaeopolis and Ravenna, both of which date from the reign of Justinian, Jones calculated that the land tax absorbed over a quarter of gross land produce, and that ordinary landowners paid three times more tax in the 6th century. than theirs

7 Mattingly (1998); Fulford (1980); (1983); Loseby (1998); Panella (1986); Reynolds (1995).

8 Hopkins (1983a) ix.9 The few exceptions include amphora and fine-ware quantification at Pella i

Jordan: Watson (1992); and Alexandria in Egypt: Majcherek (1992). For amphibious excavation and survey work at Caesarea, Israel: Adan-Bayewitz (1986), Oleson et al. (1994), Blakely (1996), Tomber (1999); and in Dor, Israel: Kingsley and Raveh (1996), Gibson et al. (1999). The work of Abadie-Reynal and Sodini (1992) at Argosin in the Peloponnese is also relevant in this context.

10 Ward-Perkins (2001) 167.11 For the roots of this theory, Jones (1964) 817, 822-23, (1974) 87. For an example

of agreement, see Hendy (1989) 4, 6.

sean a. kingsley116

predecessors in the 1st century B.C.12 The primitivist model saw agriculture not only as depressed and overtaxed but also unable to sell a surplus even if it produced one. Trade goods circulated only within a regional context, hindered by the difficulties and costs of overland transport.13 Proponents of this argument often cite Gregory of Nazianzus's 4th century. account of how agricultural surpluses produced in Caesarea in Cappadocia (located 250 km inland) were unprofitable because the infrastructure for long-distance exchange did not exist that far from the sea.14 The impression that few imports penetrated the countryside is also suggested by a less well-known text.According to Synesius of Cyrene, writing in the second half of the 4th or early 5th century, the peasants living inland in southern Cyreneica had never encountered marine life. After observing salted fish spilled from an Egyptian amphora, Synesius says that some peasants, saying they were the corpses of evil snakes, jumped up and fled, suspecting that the spikes were as dangerous as the venom of a snake's fang .'15 Although. these texts are few and far between, they give the impression that rural areas were not engaged in meaningful levels of trade. However, it is increasingly possible to study rural settlements archaeologically and find good structural and ceramic evidence that provides an alternative source for assessing these questions.

Sumaqa i Carmel, Israel - and case study

Recently published results of major excavations and surveys at Sumaqa in Israel provide a rare, detailed insight into the kind of rural villages where the majority of people would have lived and worked in Late Antiquity.16 It provides a useful example of how current field priorities and site methodology can used to expand and actually

12 Jones (1974) 83. This comment in no way diminishes the recognition that Jones remains one of the finest minds of our time in understanding the holistic complexity of Late Antiquity. Would he have adopted archaeological sources if they had been widely available?

13 Inflexible arguments regarding the prohibitive costs of land transport still limit the acceptance of theories of large-scale intraregional trade. For a comprehensive demolition of the primitivist view, see Laurence (1998).

14 As discussed in Hendy (1989) 7.15 Synesius Ep. 148 (trans. Fitzgerald (1926) 243).16 Excavated from 1983-95 by Professor Shimon Dar of Bar Ilan University:

Men (1999).

late antique shop 117

change our view of economic life in the countryside. Although it is only 11 km inland from the Mediterranean coast, the village is located at 340 m above sea level and is hidden among the dense canyons and steep slopes of Carmel. Natural communication routes into and out of the village are particularly poor (Fig. 1, see plates). Sumaqa's inhabitants (estimated at a maximum of 937 people) are believed to have been Jewish, based on the discovery of a synagogue, a burial cave whose facade incorporates an incised menorah on each side, a mikveh (ritual purification bath), and Talmudic references to various permanent residents rabbis.17 In terms of architecture and decoration, the village is unspectacular. Structures are built of local limestone, marble is very rare, colored mosaics are lacking and no inscriptions have been found. A traditional interpretation of such a settlement would lead to the conclusion that Sumaqa must have been a poor village whose economy revolved around subsistence agriculture.

Excavations give a very different picture. Within the village's surrounding catchment (60% of which is terraced), 5 wine presses, used from the 3rd century, are recorded. until at least the 6th AD Their collection vessels have a total liquid capacity of 46,500 liters (equivalent to about 2300 local amphorae), which it is estimated would have yielded an annual surplus of 64% after meeting local demands. -trial at Sumaqa centered on at least 12 workshops, which resemble wine presses in the plan of their plastered upper floors (7.5 x 13.4 m max.) and their lower collection vats (4.0 x 4.6 m max. ) (Fig. 2, see plates). But these installations are characterized by the presence of large cylindrical stones up to 2.64 m long with an average weight of 750 kg. Moving these stones would have required the effort of 5 people. There are different opinions about the product that is processed under these pebbles. But their impressive scale, together with the fact that their associated vats could hold 7,000–10,000 litres, certainly shows that these workshops were designed to crush a tough agricultural product that yielded abundant liquid.19

Regardless of the identity of the mystery product associated with these workshops, there is no doubt that a profit was made

17 Dar (1999) 17-34, 38-40, 139,18 Dar (1999) 100-107; Kingsley (1999a) 74-75.19 Dar (1999) 77-94, including the unlikely argument that local fauna and

flora was processed into dyes. Such work would not have required large vessels or heavy stone rollers.

sean a. kingsley118

for trade. This is evident in the ceramic record which, although dominated by local bag-shaped amphorae (LR5 makes up 92% of the total), includes a remarkably wide range of imports from North Africa (Keay XXVG), Egypt (painted Coptic ware) and Eastern Mediterranean (LR1, LR2, LR3 and Beirut amphorae). Of these imports, LR3 is numerically dominant (6% of all amphorae). The 755 fragments of imported fine dining bowl fragments found at Sumaqa (Cypriot red slip: 59%; Phocaean Red slip: 26%; African red slip: 6%) are equally revealing.20

This evidence shows that Sumaqa's villagers lived well above the subsistence level and used at least some of the profits from the sale of their agricultural produce to purchase semi-luxury foods and 'exotic' tableware. Such trade was almost certainly carried out at least in part through money exchange: 439 coins from 4-7 The A.D. date has been found from the site, struck in Antioch, Alexandria, Thessalonica, Constantinople, Nicomedia, and Rome.21 That villagers chose to identify with a pan-Eastern Mediterranean material culture is particularly interesting; it contradicts the conservative belief in Israeli research that Roman and Byzantine Jewish communities followed rigid rules of ritual purity (halakhah), which are particularly strict with regard to imports.22 The excavations and interpretation of the material culture at Sumaqa suggest an economically comfortable entity inhabited by a fairly 'liberal' section of Jews who, through trade, integrated themselves into a very extensive cultural area. This evidence, combined with the impressive manufacturing facilities, does not suggest a society suffocated by heavy taxation. On the contrary, at Sumaqa we can imagine a situation where tax demands were perhaps a positive force.23 Villagers apparently responded to tax pressures by intensifying production, creating larger crop surpluses that could be sold commercially in coastal markets; some of the income from this was reinvested in the infrastructure of agricultural production. The distribution of both imported amphorae and imported fine wares (Fig. 3) throughout Israel indicates that Sumaqa is far from an isolated case.

Archaeological research has necessitated a radical rethinking of the traditional view of natural communities in rural areas.

20 Kingsley (1999b) 264-67, tables 11-17.21 Kindler (1999) 347.22 Adan-Bayewitz (1993) 229-31; Zevulun and Olenik (1979).23 As generally suggested by Whittaker and Garnsey (1998) 277.

late antique trade 119

Fig. 3. Distribution map of Phocaean Red Slip imports (5th to mid-7th century AD) in Israel (excluding the Negev). These occur in 135 sites between Shelomi in the north (site 6), Hammat Gader in the east (site 59) and Ma'on to the south (site 223). PRS is represented in 14 small towns and at least 38 rural districts (villages, farms, estates, field guard cabins, cave dwellings). (S. Kingsley 1998). For fig. 4-6 see plates.

sean a. kingsley120

Sumaqa emphasizes that field recording of rural settlements should not be limited to architecture, but must include surveys of productive structures and excavations with detailed studies of ceramic assemblages if we are to develop an accurate idea of ​​the economic character of a given region.

It is interesting that new historical research is now moving in the same direction as archaeological work. Recent textual studies of the role of periodical markets and research in rural areas loaded into ancient Egypt, using papyrus and numismatics, both concluded that most Roman and Byzantine peasant households were less self-sufficient and considerably more involved in trade than is traditionally claimed. This view may be supported by Libanius' reference to coinage as being a standard medium of exchange among peasants in the 4th century. Antioch.24 Although coinage entered circulation via public wages or state payments to contractors, it appears to have become the preferred private and public store of value in the Late Antique East. In Egypt, for example, farmers, stonemasons, unskilled workers and grape pickers all paid with gold coins.25 The argument that Late Antiquity production was inward-looking and that farmers were an unhappy lot is no longer tenable; a new approach that promotes greater social and economic fluidity is gaining momentum within current scholarship. Although it would be naïve and irresponsible to claim that levels of peasant 'income', indebtedness, disorder or satisfaction were stable throughout the Eastern Mediterranean, research has certainly turned away from the primitivist position outlined above.

Byøkonomi

Turning from the rural periphery to the heart of Roman civilization, the nature and structure of urban commerce is as yet barely explored in the eastern Mediterranean. The 'consumer city' debate has long raged in the West26, but has hardly figured in the East. In the primitivist view of the Roman economy, 'consumer cities' lived parasitically on the surrounding countryside, acting as centers where they landed

24 Liebeschuetz (1972) 83-84.25 Banaji (1992) 73-74, 97.26 Whittaker (1990), (1993). See especially. Parkins (1997), (1998).

late antique trade 121

elite used their wealth. Cities certainly served as residential centers for an elite who derived their wealth from rural satellite communities: Libanius informs us that some estates in the Antioch region "are farmed for the city and the income from them belongs to the city that owns them" 27 private individuals as well , such as Evangelius of Caesarea from the 6th century Palaestina Prima may own entire villages, such as Porphyreon in modern Haifa, which he bought for 300 pounds of gold.28 However, it also appears from both literary and archaeological sources that eastern cities were centers of trade and industry.

City trade

Historical texts confirm that seasonal markets and fairs were held on Christian festival days in Mambre, Ptolemais, Botna and Jerusalem in Palestine.29 Choricius, for example, refers to skenai (temporary stalls) set up by traders during St. the Sergius Festival in Gaza, which was 'really laden with goods both for the rich and those of moderate means'.30 Such festivals testify to a symbiotic relationship fostered between town and country, where both town and village products were sought after and bought by all walks of life. Nevertheless, references to markets of this kind, or to urban retail trade, are few and far between; archaeology, one might hope, could give a better idea of ​​the extent and nature of these activities. Unfortunately, excavation strategies in the Eastern Mediterranean have neglected to study commercial structures that could provide an idea of ​​the nature and extent of urban retail trade. Instead, they have and still concentrate on recording monumental public buildings that shy away from unglamorous utilitarian structures. Excavations of colonnaded streets complete with their shops are rare; only at Scythopolis has there been a real attempt to include an assessment of their development as part of a wider urban research strategy. The excavation in the early 7th century. However, shops at Sardis in Turkey, complete with their contents, demonstrate the importance of understanding the scale and evolution of these structures. Annex stores have also been neglected; those unveiled at Caesarea Maritima i

27 Libanius, Or. 50.473 (translation Norman (1977) 63).28 Procopius, Anecdota 30 (translation Dewing (1960) 17).29 de Ligt (1993) 73.30 Choricius, Laud. Brand 1.83 (transl. de Ligt (1993) 73).

sean a. kingsley122

Israel is a testament to the benefits of open field excavation that aims to understand urban neighborhoods holistically rather than seeking the easy targets of churches and bathhouses.31

Byindustri

Industry was a conspicuous element in urban life throughout the East, and is mentioned widely in textual sources. Papyrological evidence suggests that in late antique Egypt some craft specialization was limited to cities (bleachers, dyers, linen weavers, bakers, silversmiths and glassmakers) and that at least 15–20% of household heads were engaged in urban industrial production. 32 At coastal Korykos in Cilicia, 40% of those registered in a sample of 456 funerary inscriptions worked in clothing production and small handicrafts, in shipping, wine importation, pottery production and retail trade.33 Here it is possible that the percentage is exaggerated, as part of the population may never have received inscribed tombstones. However, it testifies to the importance of this part of the population.

Again, it is difficult to ascertain the extent and nature of these activities through archaeology. First, few urban archaeologists are actually interested in these topics; serious discussions, such as those about the bone and ivory workshops of Alexandria, are rare.34 Second, there are difficulties with the nature and survival of the evidence, since not all productive activities left lasting physical remains. Outstanding preservation from the beginning of the 7th century. the shopping street at Sardis emphasizes this: color processing in this area was indicated by crucibles found next to accumulations of color pigment over 1 m high.35 When it survives, archaeological evidence for production is more often structural, such as furnaces and vessels from a 6th c. .tinctoria (a cloth dyeing installation) uncovered at Jerash.36 These structures have been particularly badly neglected by monument-hunting excavation strategies; they are not normally located unless there is large-scale open area excavation

31 Sardis: Stephens Crawford (1990); Caesarea Maritima: Patrich (1996).32 Bagnall (1993) 86. For higher percentages at Oxyrhynchus and Panopolis,

se Alston (1998) 83-84.33 Trombley (1987) 18-21.34 Alexandria: Rodziewickz (1998).35 Stephens Crawford (1990).36 Uscatescu og Martin-Bueno (1997) 77.

late antique shop 123

are carried out and are in any case more likely to be located on the outskirts of cities, away from the core areas of public construction, and so be missed.

In fact, the study of urban industry really requires a total rethinking of field practice. The most appropriate way to study it is through-the-city field survey. This technique, almost unknown in the eastern Mediterranean, involves systematic field walking of urban areas, as is done in the countryside, but in a more intensive way. It is particularly effective in recovering diagnostic industrial waste, such as spills and slags, which can be mapped and considered alongside recovered, datable ceramics. Geophysical work can also be used to pick up kiln sites and other features. The urban survey thus provides a new framework in which to place industrial remains and decide on suitable excavation strategies. Until it is widely used in the East, urban excavation will not generate much serious economic data; it will continue to provide us with only political or religious information, the result of research strategies oriented towards restoring monumental buildings of this type.

Urban Cloth Production and Dyeing in Palæstina - Et casestudie

So far I have pointed to tensions between the interpretation of textual and archaeological sources for the late antique economy. But skepticism about the value of textual information is most appropriate for the study of rural areas, where textual references are few, where papyrological documentation is exceptional, and where archaeological data often seem much richer. In suburban settings, this is not always the case. Some of the shortcomings of existing archaeological data have been pointed out above; In contrast, textual sources sometimes exist in useful detail due to the urban focus of classical literary culture. Much can be gained through careful combinations of archaeological and textual evidence, although one must allow each to tell us about different subjects in its own way, rather than trying to force the evidence to agree. The evidence of a specialist practice from Palestine, clothing production and dyeing, can be used as a case study.

First, textual sources give us a useful idea of ​​the minimum scale of clothing production and dyeing. In Palestine, Talmudic literature refers to purple dye production in Diospolis;37 Caesarea

37 Schwartz (1991) 173.

sean a. kingsley124

and Neapolis are also listed as exporters of purple cloth in the Expositio Totius Mundi et Gentium, believed to date to AD 359. The same source suggests that cloth-making in Scythopolis was best known in Palestine; it refers to abundant clothing production in this city and its export worldwide. To this list of places we can add Ascalon: an inscription found on a tabula ansata from KhirbetDah-Dah in North Hebron, dated to AD 527, refers to the 'shed of Zenonos the tanner of Ascalon', who presumably learned his trade in the city his. birth.39 This compensates for the fragmentary nature of our archaeological evidence, which is effectively limited to two sites. At Dor, a complex of rock-cut basins, water channels and courtyards from the 4th and 6th centuries, sheltered along the north harbor, is plausibly associated with the city's long tradition of purple-dye production40, which dates back to the 5th century. B.C. However, in the absence of any texts, it is doubtful that this page would have been interpreted in this way. More substantial remains of the mid-5th century. A.D., covering 250-300 m2, has been excavated next to the synagogue of Constantia, the maritime city of Gaza (Fig. 4).41

The size of the industrial area here suggests that it saw high capital investment, either by the state or private investors, with the fabrication processing pigment sourced from the Negev, Sinai, Italy and Greece.

Historical texts also provide us with important information about the organization of production. They suggest that at least some city operations were initially state-controlled (though challenged by private enterprise). Thus an edict of AD 374 in the Theodosian Code warned that anyone illegally harboring linen workers at Scythopolis would be fined 5 lbs of gold.42 The underlying assumption is that the dyeworks were staffed by state employees whose employment rights were tightly regulated. Another law from the same source, issued in 385, decreed that "if anyone should dare to usurp the use of a boat assigned to the compulsory public service of purple dye collection and for the collection of shellfish, he shall be held responsible for the payment of two pounds of gold".43 In the first half of the 5th century, the state's attempt to monopolize

38 Expositio Totius Mundi et Gentium 31 (oversat. Rougé (1966) 165).39 Amit (1991) 162.40 Raban (1995) 301.41 Ovadiah (1969) 196-97.42 Cod. Theod. 10.20.8 (translation Pharr (1952) 286).43 Cod. Theod. 20.10.12 (Pharr's translation (1952) 287).

late antique trade 125

the purple dye industry may well have become unenforceable. Another edict in the Theodosian Code, dated 436, refers to 300 pounds of purple silk which had been 'dyed in secret dyeing operations'. The state was so disturbed by this illegal activity that it took the extraordinary measure of sending one in seven men from the secretary's office, one in six men from the general tax office, and one in five men from the registrar's office to dye-work in Phenicia to investigate this corruption. 44 It is uncertain whether centralized control over the production of purple dyes was regained, but this study reeks of desperation in a world where even the most specialized industry was increasingly commercialized.

In this combination of archaeological, documentary and epigraphic evidence, the sources are not in competition, but rather provide complementary information that can give an indication of the complexity of the structure of urban trade and industry between the 4th and 7th centuries. This testifies to the importance of considering all types of source material in an urban context and not succumbing to the temptation to privilege archeology as was once done for textual sources.

Overseas exchange

Shipwreck archaeology

If rural and urban agricultural and industrial production were locked into Mediterranean macroeconomic development, how can we assess the nature and extent of the latter? The study and excavation of shipwrecks throughout the Mediterranean provide exciting opportunities to investigate short-term trade history. Theoretically, the study of a solid sample of cargoes should allow a reconstruction of which regional products spread among the Mediterranean markets, of how they were distributed, and in what quantities. The optimal preservation of organic materials on wrecks promises to end the ongoing debate about the content of specific amphora types. It also enables analysis of households, primarily the personal containers used by ship's seafarers and stored in the galley area.

44 Cod. Theodore. 10.20.18 (translation Pharr (1952) 288).

sean a. kingsley126

ship's provenance must be determined; this is based on the assumption that the crews used goods manufactured in their home countries.

The potential of shipwreck archeology is currently very marginalised. This is because marine archaeologists, unlike everyone else in the wider discipline of archaeology, are still defined solely by the medium they work in. Specialists who study towns, villages, landscape development, water systems, churches or villas usually concentrate on a clearly defined subject or historical period; marine archaeologists tend to concentrate on excavating wrecks rather than applying the results of their investigations to wider historical debates. Part of the explanation behind this strange reality is the cumbersome artificial breathing equipment required to operate underwater, which tends to exclude and limit personnel in this field of study. Similarly, the discoveries of underwater archeology are more dependent on the vagaries of their environment for their survival and discovery. As a result, underwater archaeologists tend to be more evidence-focused and less period-specific than their counterparts on land.

The rationale behind wreck analysis may seem obvious, but alongside the basic assumption that ships are tools for studying trade patterns lies a very poorly developed set of research goals in the Mediterranean.45 Wreck research has proven chronologically broad in the sites selected for analysis, and limited in strategy formulation. .Currently, we have no archaeological evidence for shipping Egyptian wheat to Constantinople (estimated at 31,200 tons annually, or 624 shipments),46 or for shipmasters flouting state laws by transporting commercial cargo alongside state shipments, as suggested in historical texts. Where are the great annona ships of Late Antiquity that staved off famine in Constantinople? Similarly, we have no idea how Syrian, Cypriot and Greek oils and wines were distributed by ship, apart from the evidence of the Yassi Ada wreck of 626 AD. it is clearly foolhardy to pin down all our hypotheses from the early 7th century. shipping on a single wreck.47 Other products incl

45 For the potential of shipwreck studies, see Gibbins (1990), Parker (1990), Kingsley (2000).

46 Calculated in Kingsley and Decker (2001) 2.47 Demonstrated by van Alfen's reassessment of LR1 amphorae from the Yassi

The Ada Wreck (1996), which suggested that, as opposed to being a commercial cargo, the transport amphorae may have been collected from various ecclesiastical properties as annona militaris.

late antiques trade 127

Egyptian papyrus and most metal raw materials are completely invisible within the available maritime archaeological record.48

Three factors must be addressed before the potential of shipwreck archeology can be fully realized. First, archaeologists need to become less intuitive and more analytical in survey and excavation design. By asking specific questions (such as how did the Vandal occupation of North Africa affect trade flows, or how did maritime trade change in the wake of the Justinian Plague? ), a proactive approach can replace what is effectively a process of pot-luck in the procurement of data. Second, a wider range of sites with varied preservation needs to be examined to properly consider shipping of different dates and grades. It has long been proven that it is not only contiguous hulls with intact cargoes that provide useful trade-related data, but even widely dispersed locations where cargoes are intermingled in ship graveyards.49 If we wish to interpret late antique patterns based solely on aesthetic appeal accumulations of intact amphorae and well-preserved underlying hulls, we will have a long wait. Third, publication rates must improve unless shipwreck archeology is to remain tangential to studies of late antiquities. Wrecks in Greek waters have been unofficially estimated at around 2000, which undoubtedly includes dozens of ships from the 4th to 7th years. The Israel Antiquities Authority has investigated about 200 wreck and cargo concentrations. Although the databases from both of these countries offer massive potential for radically rewriting the history of long-distance trade in the East, no late antique site has yet been published by these government agencies.50

Dor, Israel (Southern Carmel Coast) - A case study

The excavation of a small merchant ship at Dor, Israel, in 1999 exemplifies how important new information can be extracted from even relatively poorly preserved Late Antiquity wrecks. After a preliminary investigation in 1991, I deliberately chose this site (Dor D) for excavation to test the interpretive potential of a partially coherent

48 The Marzamemi wreck off the coast of Sicily, which contained a shipment of Thessalian marble, provides an example of another type of cargo that was carried around the Mediterranean (Kapitän (1980)).

49 Parker (1979); Kingsley (2000) 67.50 Regarding the retention of very valuable primary data, I do not consider-

publication of wreck sites off Israel as more problematic than the delayed release of the Dead Sea Scrolls. Cf. Kingsley (2000).

sean a. kingsley128

ent Byzantine wreck (with limited wood protection and no intact pottery vessels) lying in shallow water off Israel. Despite the ship's state of preservation (Fig. 5), we expected to locate some parts of the hull clamped under the 346 stone ballast blocks. We also expected to recover a significant sample of sherds, consisting of the original transport amphorae from between the wreck and the shore (where they would have been washed by wave action), despite the surface density of amphorae being as low as 8 sherds per m2 .51

Both hypotheses turned out to be correct. A strict policy of total pottery extraction, from three 1 meter wide trenches cut along the flanks of the site, recovered 89 kg of amphorae and household items. Quantification showed that 90% of the load probably consisted of bag-shaped amphorae made in southern Palestine (LR5; Fig. 6) and that 7% was LR4 from the Gaza/Ascalon regions. The variety of amphora subtypes and clay materials shows that at least some of the cargo amphorae had been reused. Over 500 LR4 and LR5 sherds were found that had a coating, indicating a wine content.52 Grape seeds were also found adhering to pitch on one sherd. This single sherd, discovered only through a policy of total sherd recovery, is the first unequivocal evidence that the LR5 amphora carried some form of the wines for which the Holy Land was known in Late Antiquity. Another reused jar shard was characterized by a hole filled with a lead plug, probably originally used to vent gas from fermentation.53

Analysis of the ballast carried out by Dr. Eleni Morrisseau of the GeologicalSurvey Department in Nicosia has confirmed a provenance in northwestern Cyprus. The few rims of LR1 amphorae from the wreck, interpreted as coming from the galley stores, have been independently tested by Dr. David Williams from Southampton University as he is also from Cyprus. The pottery evidence, the radio-carbon readings and the tap-and-tap technology used to build the wooden vessel conjure up a picture of a 15-20m trading vessel of Cypriot origin carrying empty Palestinian amphorae

51 For the Dor D study, see Kingsley and Raveh (1994); Kingsley and Raveh (1996) 64-66. The 1999 excavation is described in Kingsley (2002).

52 Kingsley (2001) 50-51.53 Vent holes present on the shoulders of Palestinian LR4 and LR5 amphorae

believed to have been used to release excess carbon dioxide into the glasses as the wine contents continued to ferment.

late antique trade 129

back to the Holy Land for replenishment in the second half of the 6th century. The implications of these findings for the economic history of the later 6th century. is currently being assessed. Although poorly preserved, the Dor D shipwreck was quickly excavated in just two weeks and exemplifies the great potential of a type of site traditionally ignored by 'marine archaeologists' in favor of coherent wrecks.

Quantification of ceramics

Despite widespread enthusiasm for marine archaeology, it is unlikely that sufficient sites will be recorded or published within the next 50 years to clarify the extent of trade in specific products and trade routes unless a visionary institute devotes its energies to produce a master data base of late antique Mediterranean shipwrecks.54

This is frustrating in the contemporary academic climate, which regards quantitative analysis as one of the best ways to escape hollow generalizations about the relative magnitudes of interregional trade.

Conventional pottery studies remain our best source for determining which strata of society consumed imported products and the amount of material involved in the trade. For example, except for a dense array of Palestinian wine amphorae scattered in wreckage along Israel's coast, our export path runs cold along more distant sea routes. A few loads are known from Turkey, a 5th century bag-shaped amphora has been fished up off Corfu, while the 6th century wreck of La Palud off southern France contained a smaller consignment of LR4 and LR5 from Gaza/Askalon- regions and northern Palestine.55 In contrast, a detailed examination of survey and excavation reports provides a relatively comprehensive picture of the geographic extent and changing distribution over time. shipped goods. Although we could examine this problem using any major type of eastern Mediterranean amphora, the point can be well illustrated using the Palestinian evidence. Instead of representing examples of small-

54 I am puzzled why the Institute of Nautical Archeology at Texas A& M University, which has been studying the waters off Turkey for decades, has not compiled a database of the hundreds of sites identified; cf. for example, Rosloff (1981). The threat of inadvertently informing 'treasure hunters' of locations is a weak excuse.

55 Discussed in Kingsley (2001) 51-55. For the important 6th century wreck of LaPalud, see Long and Volpe (1998).

sean a. kingsley130

scale trade in the odd amphora exchanged as a political gift or carried west as a pilgrimage souvenir, the evidence from 29 quantified pottery deposits from sites located between Sinai and Spain is indicative of large-scale export of Palestinian wines.56

Pottery quantification is an extremely time-consuming method that measures the relative frequency of different types of wares in assemblages and is implemented by counting or weighing sherds or by measuring the total rim or base circumference of individual vessel types.57 Quantified deposits available. suggests that the amphorae from the Gaza and Ashkelon regions already reached Beirut in significant quantities by the second quarter of the 4th c.58. It is tempting to interpret the development of this trade as driven by newly regulated state demands for taxes, which were probably transported in kind at this date.

The major trade in Palestinian wine only took off in the early 5th century and lasted for another 250 years (Fig. 7). In many cases, occurrences of Palestinian amphorae are higher than currently recognized. They accounted for as much as 45% of all imported amphorae in the mid-5th century. century. Carthage; 20% at 6th century Argos; 12% in Arles in southern France; up to 20% at the end of 6. or the beginning of the 7th. century. Marseille; 16% in Naples; and 11% in Rome around 600-50.59 Palestinian wines that reached Spain, Gaul and Italy after the fall of the Roman West can only be explained as evidence of significant trade and not state-regulated exchange. These Palestinian wines comprised just one element in the flow of products distributed from the East in Late Antiquity. Similarly, the wines and oils of Syria, Cyprus, Greece, and Turkey occur far off the main lines of annona civica's supply routes, and so continuously in the West after the fall of Rome that their circulation can only have been driven by market forces. How to over-

56 Kingsley (1999a) tables 12–13.57 Despite general consensus among scholars that pottery quantification is

crucial to both assessing trade patterns and assessing with greater precision when specific sub-forms of specific pottery types proliferated and declined, it is instructive that in the recently published papers from a conference on late antique ceramics between Syria and Jordan (Villeneuve and Jordan) Watson (2001)), only three of the 27 papers include any discussion of quantification: Abadie-Reynal (2001) fig.1; Hayes (2001) 281-82; Uscatescu (2001) fig. 1.

58 Reynolds (1997-98) 54.59 Carthage: Riley (1981) 90; Argos: Abadie (1989) 54; Arles: Congress and

Leguilloux (1991); Marseille; Bonifay (1986) 303-304; Napoli: Arthur (1985) 255;Rom: Whitehouse et al. (1985) 186.

late antiques trade 131

the maritime trade cannot be explained away as a reflection of the food resources of famine-stricken cities whose populations have outgrown the capacity of their landscape to support them; large shipments for these purposes would indeed have been incredibly inefficient. Instead, I believe that the spread of Eastern wine and oil throughout the West and elsewhere reflects a broad consumer taste that has so far been little studied.

Conclusion

In its entirety, the term "trade" covers a wide range of activities; in this paper I have deliberately emphasized the importance of inter-regional exchange at the intermediate level. Without a doubt the annona

Fig. 7. Distribution map of Palestinian amphorae from Gaza/Ashkelon (CLR4) and central and northern Israel (CLR5) found outside the production area between the 5th and mid-7th century. (S. Kingsley 1999)

sean a. kingsley132

militaris, 60 annona civica and even ecclesiastical gifts all played a very influential role in Late Antiquity. However, there is good evidence that trade was far more extensive than this and touched and integrated far more sectors of society than our history books indicate. The ever-increasing stream of archaeological data suggests that we are doing antiquity an injustice by assuming that ancient civilizations were innocent of the broad cycles that drove rural production, as Sir MosesFinley believed. During Late Antiquity, many farmers keeping vines in the fields of Palestine and Syria appear to have acted in response to commercial forces as well as fiscal demands.

Unfortunately, these points are still not fully appreciated. In MichaelMaas's recently published Readings in Late Antiquity,61 which is no doubt a standard textbook for many university students, we hear nothing of the complexities of rural production, but only descriptions of how natural disasters—locusts, famine, and plague—exacerbated the unhappy state. of rural areas. The reasons for this lie partly in the failure to synthesize existing information in a meaningful way, but also in the fact that archaeologists have not adopted appropriate research models and field methodology. Landscape archaeologists working in the Eastern Mediterranean need to pay more attention to regional economic structures; this means studying agricultural and industrial sites and the relative amounts of local and imported pottery found in sherds, rather than focusing on geographically dominant 'sites'. Far more studies of pottery assemblages from rural areas are needed to reveal the extent to which imports penetrated the country. Since at least 60% of society, and almost certainly much more in most eastern Mediterranean provinces, was based on the land,62 major studies of village life, especially in Egypt, Syria and Turkey, could radically change our current rather one-dimensional view. .Within cities, project managers should seek to excavate specific neighborhoods rather than uncover individual public buildings and should consider intensive field survey work. The historical texts of the period are still greatly neglected; I'm sure there are very informative narratives that contain lively trading anecdotes waiting to be discovered. Similarly, Egyptian papyri also promise to unlock the silent world of farmers, trade and urban commerce. The current plethora of reports on

60 For an example of LR2 associated with military trade, see Karagiorgou (2001). 61 Maas (2000) 58-68. 62 Bagnall and Frier (1994) 55 estimate that around 3 million of Roman Egypt's

4.75 million inhabitants (63%) were based in the countryside.

late antiques trade 133

excavated sites from the 4th to the 7th century. Date coming from Egypt will also give a significant boost to the subject. Finally, traditional distribution studies, complimented by amphora quantification, still provide the best insight into long-distance exchange structures, especially when we accept the incomplete nature of the archaeological and historical records. Having recognized that many of the economic products of Late Antiquity cannot be discovered, we must maximize what is available: in short, infinite numbers of broken pottery sherds.

Within such a broad approach to late antiquities, the researcher must be prepared to design innovative research methodologies, synthesize archeology with history, and embrace comparative sources from the past 200 years that enable us to avoid the pitfalls of invalid moderns. analogies. Such an interdisciplinary approach constitutes, in my opinion, the most effective agenda for the early 21st century.

sean a. kingsley134

Bibliography

Abadie C. (1989) "The ProtoByzantine Amphoras of Argos (4th-6th Centuries)", i Research on Byzantine Ceramics, red. V. Déroche og J.-M. Spieser (BCHSuppl. 18) (Paris 1989) 47-56.

Abadie-Reynal C. (2001) “Middle Eastern Imports to Argos (4.-7.

centuries)”, in Byzantine and Proto-Islamic Pottery of Syria-Jordan (4th-8th centuries AD). Proceedings of the symposium held in Amman on 3, 4 and 5 December 1994, ed. E. Villeneuve and P.M. Watson (Beirut 2001) 283-87.

Abadie-Reynal C. and Sodini J.-P. (1992) Early Christian Pottery from Thasos (Aliki, Delkos, Ancient Excavations) (Athens 1992).

Adan-Bayewitz D. (1986) "The ceramics of the late Byzantine building and its implications (Stratum 4)", in Excavations at Caesarea Maritima 1975, 1976, 1979. Final Report, ed. L. I. Levine and E. Netzer (Qedem 21) (Jerusalem 1986) 90-129.

——— (1993) Common Pottery in Roman Galilee. A study of local trade (Ramat-Gan1993).

Alston, R. (1998) "Trade and the city in Roman Egypt", i Trade, Traders and the Ancient City, red. H. Parkins og C. Smith (London 1998) 168-202.

Amit D. (1991) "Khirbet Dah-Dah", ESI 9 (1991) 162-63.Arthur P. (1985) "Naples: notes on the economy of a Dark Age city", in Papers in

Italian archeology 4. Part 4. Classical and medieval archaeology, ed. C. Malone and S. Stoddart (BAR International Series 246) (Oxford 1985) 247-60.

Bagnall R. S. (1993) Egypt in Late Antiquity (Princeton 1993).Bagnall R. S. og Frier B. W. (1994) The Demography of Roman Egypt (Cambridge)

1994). Banaji J. (1992) Rural Communities in the Late Empire AD 300-700: Monetary and Eco-

nomic Aspects (D.Phil Thesis, University of Oxford 1992). Blakely J. A. (1996) "Toward the study of economics at Caesarea Maritima", i

Maritime caesarean section. A look back after two millennia, ed. A. Raban and K. Holum (Leiden 1996) 327-4

Bonifay M. (1986) "Observations of late amphorae in Marseilles from the excavations of the Bourse (1980-1984)", RANarb 19 (1986) 269-305.

Congès G. and Leguilloux M. (1991) "A Late Antiquity Dump in the Esplanade District of Arles", RANarb 24 (1991) 201-34.

Dar, S. (1999) Sumaqa. A Roman and Byzantine Jewish Village on Mount Carmel, Israel (BAR International Series 815) (Oxford 1999).

Dauphin C. (1998) Byzantine Palestine. Stands and Populations, Volumes 1–3 (BARInternational Series 726) (Oxford 1998).

Dewing H. B. overs. Procopius VI. The Anecdota or Secret History (3. udg.) (HarvardUniversity Press 1960).

Fitzgerald A. transl. The Letters of Synesius of Cyrene (Oxford 1926). Fulford M. (1980) “Carthage: overseas trade and the political economy, c. AD 400-

700", Reading Medieval Studies 6 (1980) 68-80. Gibbins D. (1990) "Analytical approaches in maritime archaeology: a Mediterra-

nean perspektiv", Antiquity 64 (1990) 376-89. Gibson S., Kingsley S. og Clarke J. (1999) "Town and country in the Southern

Carmel: report on the landscape archeology project at Dor (LAPD)”, Levant31 (1999) 71-121.

Hayes J. W. (2001) "Late Roman fine wares and their successors: a Mediterranean-nean Byzantine perspective (with reference to the Syro-Jordanian situation)", in Byzantine and Proto-Islamic Pottery of Syria-Jordan (4th-8th centuries CE .Kr.) .

late antiques trade 135

AD). Proceedings of the symposium held in Amman on 3, 4 and 5 December 1994, ed. E. Villeneuve and P.M. Watson (Beirut 2001) 275-82.

Hendy M. F. (1989) The Economy, Fiscal Administration and Coinage of Byzantium (Northampton 1989).

Hopkins K. (1983a) "Introduktion", i Trade in the Ancient Economy, red. P. Garnsey, K. Hopkins og C. R. Whittaker (London 1983a) ix-xxv.

——— (1983b) “Models, Ships and Staples”, in Trade and Famine in Classical Antiquity, ed. P. Garnsey, K. Hopkins and C.R. Whittaker (Cambridge 1983b) 84-109.

Jones A. H. M. (1964) The Later Roman Empire 284-602, vol. 2 (Oxford 1964).——— (1974) The Roman Economy (Oxford 1974). Kapitän G. (1980) “Architectural elements for a basilica from the naval wreck of

VI Century of Marzamemi (Syracuse)", XXVII Ravenna and Byzantine Culture Course (1980) 71-136.

Keay S. J. (1984) Late Roman Amphorae in the Western Mediterranean. A Typology and Economic Study: The Catalan Evidence (BAR International Series 196) (Oxford 1984).

Karagiorgou O. (2001) "LR2: a container for the military annona on the Danubian Border?", i Economy and Exchange in the East Mediterranean under Late Antiquity, red. S. Kingsley og M. Decker (Oxford 2001) 129-66.

Kindler A. (1999) "Summary of twelve years of numismatic finds in the excavations of Sumaqa (1983-1995)", in Sumaqa. A Roman and Byzantine Jewish Village on Mount Carmel, Israel, ed. S. Dar (BAR International Series 815) (Oxford 1999) 347-61.

Kingsley S. (1999a). Specialized production and long-distance trade in Byzantine Paletine (D.Phil thesis, Univ. of Oxford 1999).

——— (1999b) “The Sumaqa pottery assemblage: classification and quantification”, in Sumaqa. A Roman and Byzantine Jewish Village at Carmel (BAR International Series 815) (Oxford 1999) 263-330.

——— (2000) “Marine archeology in Israel: an undeveloped historical resource”, Bulletin of the Anglo-Israel Archaeological Society 18 (2000) 57–72.

——— (2001) “The Economic Impact of the Palestinian Wine Trade in Late Antiq-uity”, in Economy and Exchange in the East Mediterranean during Late Antiquity, edd.S. Kingsley and M. Decker (Oxford 2001) 44-68.

——— (2002) A Sixth Century A.D. Merchant Vessel in Dor Harbour, Israel (Dor D)—Final Excavation Report (BAR International Series forthcoming) (Oxford 2002).

Kingsley S. og Decker M. (2001) "Nyt Rom, nye teorier om interregional udveksling. An introduction to the East Mediterranean economy in LateAntiquity”, i Economy and Exchange in the East Mediterranean during Late Antiq-uity, red. S. Kingsley og M. Decker (Oxford 2001) 1-27.

Kingsley S. A. and Raveh, K. (1994) "Stone anchors from Byzantine contexts in the Port of Dor, Israel", IJNA 23 (1994) 1-12.

——— (1996) The Ancient Port and Anchorage of Dor, Israel. Results of the Underwater Surveys, 1976-1991 (BAR International Series 626) (Oxford 1996).

Laurence R. (1998) "Land transport in Roman Italy: costs, practices and economics", in Trade, Traders and the Ancient City, ed. H. Parkins and C. Smith (London 1998) 129-48.

Liebeschuetz J.H.W.G. (1972) Antioch. City and Imperial Administration in the Later Roman Empire (Oxford 1972).

Ligt L. de (1993) Fairs and Markets in the Roman Empire (Amsterdam 1993).

Port Cros (Var). Preliminary note”, in Excavations in Marseille. Furniture (I'm-

sean a. kingsley136

7th century AD AD), ed. M. Bonifay, M.-B. Carre and Y. Rigoir (Etudes Massaliètes) (Paris 1998) 317-42.

Loseby S. T. (1998) "Marseille and the Pirenne Thesis, I: Gregory of Tours, the Merovingian kings, and 'un grand port", i The Sixth Century. Produktion, distribution og efterspørgsel, red. R. Hodges og W. Bowden (Transformation of the Roman World 3) (Leiden 1998) 203-29.

Maas M. (2000) Readings in Late Antiquity (London 2000). Mackensen M. (1998a) "New Evidence for Central Tunisian Red Slipware with

stamped decoration”, JRA 11 (1998) 355-70.——— (1998b) “Centres of African red slipware production in Tunisia from

late 5th to 7th century”, in Ceramics in Italy: VI-VII Centuries. Proceedings of the Conference in Honor of John W. Hayes, Rome, 11.-13. May 1995, ed. L. Saguì (Florence 1998) 23-39.

Majcherek G. (1992) "Den senromerske keramik fra sector 'G' (Alexandria 1986-1987)", EtTrav 16 (1992) 81-117.

Mattingly D. J. (1988) "The olive boom. Oil surplus, wealth and power in RomanTripolitania", LibSt 19 (1988) 21-41.

Norman A. F. oversættelse. Libanius, Selected Works (Cambridge, Mass. og London 1977). Oleson J. P., Fitzgerald M. A., Sherwood A. N. og Sidebotham S. E. (1994) The

Ports of Caesarea Maritima. Results of the Caesarea Ancient Harbor Excavation Project 1980-85. Volume II: the find and the ship (BAR International Series 594) (Oxford 1994).

Ovadiah A. (1969) "Excavations in the Area of ​​the Old Synagogue at Gaza (Preliminary Report)", IEJ 19 (1969) 193-98.

Panella C. (1986) "Amphorae, Production Centers and Preferential Markets", in Societaromana and the Late Antique Empire III Merci, the Settlements, ed. A. Giardina (Rome and Bari 1986) 251-72.

Parker A. J. (1979) "Method and Madness: Wreck hunting in lavt vand", inProgress in Underwater Science, vol. 4, udg. J. C. Gamble og J. D. George (London 1979) 7-27.

——— (1990) "Classical antiquity: the maritime dimension", Antiquity 64 (1990)335-46.

Parkins H. (1997) "The 'consumer city' domesticated? Den romerske by i eliteøkonomiske strategier”, i romersk urbanisme. Beyond the Consumer City, red. H. M. Parkins (London 1997) 83-111.

——— (1998) “Time for Change? Shaping the Future of the Ancient Economy”, inTrade, Traders and the Ancient City, ed. H. Parkins and C. Smith (London 1998) 1-15.

Patrich J. (1996) "Stores and Granaries in Caesarea Maritima", in CaesareaMaritima. A look back after two millennia, ed. A. Raban and K. Holum (Leiden 1996) 146-76.

Pharr C. (1952) oversættelse. Theodosian Code and Romans and the Simondian Constitutions (Princeton 1952).

Raban A. (1995) "Dor-Yam: maritime and coastal installations at Dor in their geomorphological and stratigraphic context", in Excavations at Dor, Final Report, Volume IA. Areas A and C: Introduction and Stratigraphy, ed. E. Stern (QEDEM Reports 1) (Jerusalem 1995) 285-354.

Reynolds P. (1995) Trade in the Western Mediterranean, AD 400-700: the Ceramic Evi-dence (BAR International Series 604) (Oxford 1995).

——— (1997-98) "Pottery production and economic exchange in second century Berytus: some preliminary observations on ceramic trends from quantified ceramic deposits from the Aub-Leverhulme excavations at Beirut", Berytus43 (1997-98) 35-110.

late antique trade 137

Riley J. A. (1981) "The pottery from the cisterns 1977.1, 1977.2, 1977.3", i Excavations at Carthage 1977, Udført af University of Michigan. vol. 6, udg. J.H. Humphrey (Ann Arbor 1981) 85-124.

Rodziewicz E. (1998) "Archaeological evidence for bone and ivory carvings in Alexandria", in Trade and crafts in Hellenistic and Roman Alexandria, ed.J.-Y. Keser (BCH Suppl. 33) (Paris 1998) 136-58.

Rosloff J. (1981) "INA's 1990 Turkish underwater survey", IJNA 10 (1981) 277-86.

Rougé J. overs. Exposition of the whole world and nations (Paris 1966). Schwartz J. J. (1991) Lod (Lydda), Israel (BAR International Series 571) (Oxford)

1991).Stephens Crawford J. (1990) The Byzantine Shops at Sardis (Massachusetts 1990).Storey G. R. (1999) "Archaeology and Roman society: integrating textual and

archaeological data", Journal of Archaeological Research 7 (1999) 203-48.Tomber R. (1999) "Pottery from the Sediments of the Inner Harbor (Area I14)",

i Caesarea Papers 2, red. K. Holum, A. Raban og J. Patrich (Portsmouth, Rhode Island) 295-322.

Trombley F. R. (1987) "Korykos in Cilicia Trachis: the economy of a small coastal town in Late Antiquity (saec. V-VI) - a précis", The Ancient History Bulletin 1 (1987)16-23.

Uscatescu A. (2001) "The contribution of the excavations of macellum (Jerash, Jordan) to the knowledge of Late Byzantine ceramics from Gérasa", in Byzantine and Proto-Islamic Ceramics of Syria-Jordan (IV.-VIII. .). Proceedings of the conference held in Amman on 3, 4 and 5 December 1994, ed. E. Villeneuve and P. M. Watson (Beirut 2001) 59-76.

Uscatescu A. and Martin-Bueno M. (1997) "The Macellum of Gerasa (Jerash, Jordan): from a marketplace to an industrial area", BASOR 307 (1997) 67-88.

Van Alfen P. G. (1996) "New light on the 7th century. Yassi Ada shipwreck. Capacities and standard sizes of LRA1 amphorae", JRA 9 (1996) 189-213.

Villeneuve E. and Watson P. M. (2001) Byzantine and Proto-Islamic Pottery in Syria-Jordan (4th-8th Centuries AD). Proceedings of the symposium held in Amman on 3, 4 and 5 December 1994 (Beirut 2001).

Ward-Perkins B. (2001) "Specialisation, trade and prosperity: an overview of the economy of the late antique eastern Mediterranean", in Economy and Exchange in the East Mediterranean during Late Antiquity, ed. S. Kingsley and M. Decker (Oxford 2001) 167-78.

Watson P. (1992) "Change in foreign and regional economic relations with Pella in the seventh century AD: the ceramic evidence", in La Syrie de Byzance à l'IslamVIIe-VIIIe siecles, ed. P. Canivet and J.-P. Rey-Coquais (Damascus 1992) 233-48.

Whitehouse, D., Constantini, L., Guidobaldi, F., Passi, S., Pensabene, P., Pratt, S., Reece, R., and Reese, D. (1985).

Whittaker C. R. (1990) "The consumer city revisited: the vicus and the city", JRA3 (1990) 110-18.

Whittaker C. R. og Garnsey P. (1998) "Landsliv i det senere romerske imperium", i The Cambridge Ancient History, bind XIII. The Late Empire, A.D. 337-425, red. A. Cameron og P. Garnsey (Cambridge 1998) 277-311.

Wickham C. (1988) "Marx, Sherlock Holmes and Late Roman commerce", JRS78 (1988) 183-93.

——— (forthcoming) “Trade and Exchange, 550-750. The view from the west",

sean a. kingsley138

in Late Antiquity and Early Islam. Fifth workshop. Trade and Exchange in the Late Antique and Early Islamic Near East (New Jersey, forthcoming).

Zevulun U. and Olenik Y. (1979) Function and Design in the Talmudic Period (2nd ed.) (Tel Aviv 1979).

coins and the late Roman economy 139

COINS AND THE LATE ROMAN ECONOMY

RICHARD REECE

Abstract

Coins provide a source of information about the Roman economy that is not available in the written sources. They can be studied either as products of the imperial administration - coins as struck, or as a reflection of coinage - coins as found. The coins struck are well described in standard reference works, and only a few caveats need to be added. The found coins constitute an area of ​​study that is in its early stages, and a fairly basic overview of the available information and methodology is needed. Based on these topics, a number of questions can be asked for further consideration.

Introduction

In many early societies, taxes, goods, services and profits took place along theoretical accounting lines or at the very practical level of exchange of goods. In the Roman Empire, however, the emperor produced coins, which made it possible to realize all notions and goods in the form of small lumps of metal. The extent to which this medium of exchange was used or enforced can only be known with certainty for a very small number of times and places when the written documents survive. Fully monetized exchange should never be assumed. This means that Roman coins cannot be used directly as evidence of absolute levels of economic activity. Nevertheless, it is possible to consider the patterns of coin production, supply and loss through time and space, in the hope that this will provide important clues to at least part of this story.

The use of coins to study the economy has two main elements. First, the nature and production of the coins themselves allows us to examine what the state intended to do, or thought it was doing, as coin production volumes varied through time and place. times that reflect payments or intended payments for government actions

L. Lavan og W. Bowden (red.) Theory and Practice in Late Antique Archaeology(Late Antique Archaeology 1 2003) (Leiden), s. 139-168

richard reece140

meant politics. Second, the coins found can provide information about how public spending worked in practice and how the wider public used coin. Coinage as produced in the later empire is therefore a direct testimony to the overall functioning of the imperial 'state' economy. Coins spent, lost and found are evidence of the activities of parts of the wider economy, from provinces, to cities, to districts, to households and even to individuals. All we have to do is gather the evidence and then interpret it.

Møntproduktion: Coinage as Struck

Coinage as produced is reasonably well recorded, studied and published in the volumes Roman Imperial Coinage1 and Moneta Im-perii Byzantini.2 Unfortunately, coinage is thought to be a very abstruse and complex field of study, and few historians have mined the corpus for their own purposes. This is partly due to the fact that, although the corpus is a near-perfect guide to coins produced inqualitative terms, there is little guidance in terms of minted quantities, which would be much more interesting for studying state intentions. In practical terms, this means that it is difficult for a reader to determine which numbers were produced in bulk and circulated to all coin-using persons in the empire, and which numbers may never have been seen by more than a few privileged people in the imperial capital New Year's Day. This is extremely frustrating for historians interested in the relative production of coins in different periods of the empire.

Even if we knew the total number of coins that survived for each issue, it is not possible to extrapolate relative production levels from this. Unfortunately, it seems likely that the bulk of some gold issues from the mid-6th century. from the imperial coinage was transferred to barbarian tribes to the north to ensure peace. These groups may have used the coins to purchase goods in frontier markets, in which case the coins would have circulated back into the empire and been lost and hoarded there, thus making them available to us for study. On the other hand, they may have melted down the coins, with the result that these coins have been erased from the archaeological record.

1 Mattingly, Sydenham et al. (1923 ff.).2 Hahn (1973).

coins and late Roman economy 141

No doubt a few of the same issues circulated in Constantinople, but their small surviving numbers would give a completely false picture of the minted quantity.

Alternatively, it may be possible to estimate production volumes by comparing all coins in a given issue and thus counting the number of different dies involved in production, as on average dies are likely to wear out at a similar rate. Another problem could be opted out and these dies count as well. The relative size of the two questions, e.g. the first regnal year and the last regnal year, could then be specified. But historians usually want to convert relative numbers into absolute numbers—how many tons of gold? - and it is currently a very dangerous subject. Moreover, two points combine to exclude the historian from the work of counting and discourage the numismatist. Compiling a photograph of all known examples of a coin issue is a huge amount of work that can take years. Comparing each photograph to all others to determine how many dies were used increases both the effort and the time required. The violent views that have been expressed regarding assessments of the magnitude of the issues, a perfectly reasonable and correct thing to try to do, have unfortunately closed the topic for now. Anyone wishing to pursue the subject should consult the highly reasonable study by de Callatay.3 The net result of all this is that the study of coins as struck is a very sadly underutilized resource. Perhaps we should turn instead to the study of coins as found.

Coin Supply: Interpretation of coin distributions from coins like

Found

The study of coins as found should be divided into the study of coin loss and the study of coin supply. Coin supply seeks to reconstruct the overall chronological pattern of coins brought into a place or region; coin loss seeks to investigate the date on which coins delivered to a site fell out of use. Coinage, on an empire-wide and regional level, allows us to tentatively approximate the workings of the state; it gives us a perspective that coins, when struck, cannot give: how state wages were used in practice, and it allows us to try to discover other forces important in shaping the distribution of change.

3 of Callatay (1995).

richard reece142

Understand bullion and small change

To understand the coinage of the Roman Empire, one must be aware of the difference between gold and copper coins. By the end of the 4th century, coins were either worth gold or silver, or they were made of impure copper and had virtually no intrinsic value. The relationship between them is not known in quantitative terms. If copper coinage was such an insignificant part of the amount of value available, why was so much of it produced in the later empire? My understanding of it is that while everything that mattered was done in gold or silver coins - payment of soldiers and civil servants - this valuable gold baron had to be lured back to the treasury with copper bait. The soldier was paid and this was a constant concern of the state. The soldier was eager to use his coin, but was not allowed to receive change if he took gold or silver to a store as payment for a lower value. It therefore had to take the gold or silver to a money changer and receive usable copper in return, whereby both were happy. The state had earned its gold bullion back, the soldier could use his coin.

For this to work, gold and perhaps silver were produced at comitatensian mints, which moved with the emperor or struck coins when he was present.4 Such coins were sent by the state to places where it was needed. At its simplest, this was where the loyalty of the army and civil servants had to be maintained. It makes sense to assume that where the largest amount of gold was sent, it would be necessary to provide the largest amount of copper per bait to bring it back to the treasury. Gold and silver would have performed a complete cycle – coinage – treasury – paymasters – civil servants – public – tax collectors – treasury – coinage. The only gold found will be that which escaped the cycle by being mounted as jewelry, through burial as treasure that was not recovered, or simply by being lost. Copper will not perform a cycle. It will be minted and sent out and will stay where it arrives, oiling the wheels of the bullion cycle because it has no value outside of a bullion-backed economy. If proof of this were needed, the distribution of gold and copper within and without the empire is clear. Gold travels far, while late Roman copper practically stops at the border. Gold has its own intrinsic value, but for copper to have any value, it must be validated with gold or silver.

4 Mattingly, Sydenham et al. (1923 ff.) volumes VIII and X.

coins and late Roman economy 143

We thus have a theory of a connection between the location of the paid civil servants and the distribution of copper coins. On a broad scale, this theory seems to be supported by the evidence. Late Roman copper is most common in frontier provinces such as Pannonia, the Rhineland and Britain, where there are large armies. The idea that bags were sent out from the mint to non-military locations founders because of the variety of what is being sent and because of the fact that it is unlikely that provincial mints sent coins to Rome. There was no point in sending out or even releasing copper from the mint if there was no gold available to back it. While studies of (gold and silver) coins as struck can potentially tell us about how much the government intended to spend at a given time, studies of coin supply (from copper coins found in situ) can tell us where the government's wages actually were were used. Again, it must be emphasized that a large number of copper coins during a period in a given region represents income generation, not wealth. It was and is entirely possible to handle complex transactions and debts without large amounts of change. Copper coins therefore tell nothing about the state of the economy, rather only about the workings of the state, which created the coins of the time. Needless to say, it is not reasonable a priori to assume that the state constituted the largest part of the economy, and to equate state activity seen in coins with structurally determining economic activity.

Room and Supply

On closer inspection, the connection between copper coin supply and government employees seems somewhat more complex. This can be seen in Rome through the large collection of coins that come from the Forum and Palatine. Rome itself was a prolific mint throughout the 4th century. 53 of the coins from this group were struck between 305 and 317 AD. On 21 of these coins the coin is illegible. This leaves 32, of which only 9 were struck outside Rome (three from Ostia, one each from Lyon, Arles, Carthage, Aquileia, Antioch and Alexandria). On the other hand, 1051 coins from the Forum and Palatine areas were struck between 364 AD 378. Of these, 870 had illegible mintmarks. Of the 181 from known mints, there were 55 from Rome and 126 from other locations (Arles 17, Aquileia 28, Siscia 43, Thessa-lonica 20, Constantinople 6, Antioch 5, and one each from Trier, Lyon, Serdica, Heraclea, Nicomedia , Cyzicus and Alexandria). Therefore struck off the coins found in a number of places in Rome.

richard reece144

Between 305 and 317 AD the majority were beaten in the city. However, of the coins lost in Rome struck between AD 364 and 378, more than three times as many came from mints other than originating in the mint of Rome. What are we to make of this?

The key to this problem is an understanding of the balance between supply and arrival. Supply represents directed policy, while arrival represents undirected operation. In the early part of Constantinethere's reign, there is clearly good political reason for coins to stick to their issuing emperor's sphere of influence, as armies rarely moved between these spheres. Of the coins struck from 305 to 317, Rome has thus primarily minted coins in Rome. After 317, when politics changed dramatically, a large number of Western coins are lost in the East. For example, in Jerusalem, out of 14 coins struck between 310 and 330, there are 8 legible mintmarks from the West (Arles 1, Ticinum 2, Rome5) and 4 from the East (Heraclea 1, Cyzicus 1, Antioch 2). The pattern of this small number is repeated in several other places. If we are reluctant to assume the positive supply of coins struck in Rome from 310 to 317 to the Jerusalem area, then the most likely background for this pattern is the movement of Constantine, his armies and their purses to the east after 323. alone the coinage evidence, it is impossible to determine whether the bulk of Rome's coins of 305 to 317 were purposefully released in the city at the time of their striking, or whether Rome provided a coin pool of relatively limited circulation in central Italy, some of which returned to Rome of the areflux movement.

The issue of coins of 364 to 378 is much more difficult to resolve neatly. If the Constantinian model is used, there was a significant troop movement from the Balkans to Rome in these years or soon after (Rom 55, Siscia 43). The problem here is that coins from Siscia, while saturating the Balkans, also moved strongly elsewhere. At Carnuntum on the Danube, and so accessible by water both east and west, a sample of 713 coins of 364 to 378 shows 443 from Siscia.5 The movements of the coins from Siscia can be seen in part in diagram form for Malta, Rome, Conimbriga and Carnuntum6 and in detailed tables for Great Britain.7

5 Hahn (1976).6 Reece (1982b).7 Reece (1978).

coins and late Roman economy 145

UK and supply

Another interesting region to consider is the UK. There seems at present no point in citing an individual site list in Britain, because the 4th c. coin that appears to be homogeneous. This is based on work done 25 years ago and since then, although the number of coin finds has increased dramatically, no further detailed studies have been undertaken. The conclusion reached was that the coinage of the islands came from a single continental source after being thoroughly mixed. This is nonsense, of course, but it is what the evidence suggests. The second point is the common appearance in Britain of only two of the 48 different coin marks that Siscia used as batch control marks between 364 and 378.8 These marks do not appear in the center of Rome, for example, and are not common in Carnuntum.

It is currently unknown what troop movements, if any, or other imperial actions could explain the coins from Siscia in Rome or the mix of coins found in Britain. An alternative, probably better, explanation is that in the relative peace of the 4th century, a lot of small changes were moving around the empire without direction. The Siscia coins in Rome can only be called supply if a provincial mint sacked a sample coin and delivered it to a capital city. Rome itself produced plenty of coins during that period, and again such a supply seems highly unlikely. It also seems a good explanation for Britain, rather than imagining that a shifting group of mints from Trierand Lyon to Aquileia and Siscia faithfully collected a portion of their production and sent it, regardless of distance and price, to some distant province, who were unable to request this service.

Coin supply and site histories

Recent excavations in the Mediterranean, which have been published at a reasonable rate, tend to have lists of coins found and sometimes have comments on the significance of those coins. This is a fairly recent departure and the evidence is very thin and very patchy. The appearance of coins in excavation reports can vary from the mention of a few examples that helped date certain features,

8 Carson, Hill and Kent (1960) Part 2, Nos. 1408-29.

richard reece146

for a complete list. Comments can range from two sentences to two pages. Commentary on coins from excavations is reasonably well developed in Britain, but few other places have adopted the practice.

From the optimistic tone of the introduction to this article, one might think that coins could comment directly on the economic success or failure of the place where they were found. Variations in the coin supply (that is, simple lists of coins from excavations without context) are often taken to equate with the economic health of a place or region. Unfortunately, that cannot be the case. That might be the case if the coins found were all gold or silver, but they are not - they are almost all copper. As mentioned above, these coins were issued by the state in order to recover imperial wages of gold and silver; their level of supply to a region therefore varied widely. As such, their presence in a region is not directly related to levels of prosperity, but only to levels of income generation. Despite this, it is possible to use coin supply as a guide to levels of economic activity in individual locations, if not regions. However, a number of methodological stages must be taken into account.

What does the total number of coins on a site mean?: The value of coins used and coins found: gold, silver and bronze

First, it seems very likely that all the coins found in any one place represent only a small fraction of the monetary wealth present at any one time in antiquity. This is because when an economy uses several different metals for coinage and produces several denominations, most of the value resides in the largest denominations, which are usually minted in precious metals. This is exemplified by Depeyrot's work in France, where the number of copper coins from 1650 to 1910 was often over half of the total number of coins struck annually.9 There were as many or more copper coins than there were gold and silver coins together. However, the normal value of copper coins was between one and three percent of the total money struck, although very low gold and silver in the 1880s brought the relative value of copper to its highest level - about 15% of the produced lots of money in the decade. The normal total value of copper coins was only about 1 to 3% or all of the money produced.

9 Depeyrot (1979).

coins and late Roman economy 147

In the Roman Empire, there is good evidence that not only was bullion widely used to settle major government debts, such as payments to the military, but also in the later empire that the daily use of gold was increasing.10 The problem is, that coins that were lost and have now been found in excavations are mainly of bronze and copper and biased towards the lowest values. Thus even 7,000 copper finds, more than are published for most sites, cannot give any indication of the annual income of a landowner or administrator because they represent the value of only one or two gold coins in total.

This gives rise to two points: 1. Coins from excavations cannot be interpreted in a way that pro-

shows a simple financial history for that location.2. When interpreted, such coins can only provide information

on their own terms. These will almost certainly differ from the expressions expected by historians accustomed to texts written by the middle and upper strata of society.

The importance of regional context

Despite what has been said here, it is possible to achieve a significant, but relative, measure of economic activity in a place. This does not come from measuring the total number of coins supplied to it, but from estimating how relatively common/uncommon this is compared to the regional background. The presence in one place of a large number of coins of Constantine the Great as sole ruler could be taken to show that there was a flourishing market economy in the years 324 to 337, while a smaller number of coins of Valentinian and his family could indicate a decline in the economy from 364 to 378. The number of years available for issuing coins is almost equal so it cannot explain the difference. Perhaps the number of coins produced in the two periods was therefore markedly different.

Fortunately, this does not matter. Even if mint production could be accurately measured, there is no reason, and many reasons against, to assume that each issue of coin was supplied equally to all parts of the empire. What we need to know is not the level of production, but the level of local supply. We need to know this because it could be that every site for 100 miles around has more coins of Constantine than Valentinian. If this is the case, it is a case

10 Banaji (1996).

richard reece148

provincial economies, or provincial supply, rather than a local peculiarity. Before we can interpret the coins from our excavation, we need to know the pattern formed by nearby sites, perhaps the provincial pattern, so that we can isolate peculiarities from a regional background. In other words, if all nearby sites show a similar number of coins found for Constantine and Valentinian, and our excavation shows fewer coins of Valentinian, then there is something peculiar to our site, and so something that needs to be explained, perhaps in non-economic terms.

The moral here is that each site must be interpreted against its local or provincial background. If this is not done, there is a danger of seeing the coins as indicative of events in the site's history, when in fact they reflect the vagaries of the coin supply for a large area of ​​the empire. However, due to the very thin coverage of good coin lists of the empire, we still have little idea of ​​which coins supplied which areas, and also where the borders between different areas were. A critic could indignantly point to the coin lists from Sardis,11 and Histria,12 and say that it was perfectly possible to tell which mints supplied these places and how the supply changed over time.

However, these are areas rather than different individual locations. Are Sardis and Histria, judging from the coins they have been supplied, in the same supply area? Although the material in these two cases is published, the question cannot be answered without hard work, and that work has not been done. Even when the work is done, we will only be able to discuss the differences and similarities between two points on the map, rather than two areas. There is no particular reason why the two cities should define the coin supply to the areas around them, unless that is to say that the cities were the means by which the coin entered the hinterland. This in itself would be interesting to determine. If the opposite were true, in that the surrounding area differed qualitatively in coin loss from the city, then it provides a different set of interests and opportunities.

A hard line would be that it will be a few more years before enough Mediterranean coin lists are published to give us a single Mediterranean background, let alone separate backgrounds for Italy, Asia Minor, North Africa, Greece or the Pales-

11 Buttrey et al. (1981).12 Preda and Nubar (1973).

coins and the late Roman economy 149

thaw Harsh attitudes and exclamation points can make for lively reading, but they don't help the subject move forward. If we heed the warning that there are dangers of misinterpretation ahead, we could proceed cautiously and find much that can be safely done.

An example resulting from recent work is SaxonShore Fort at Reculver in Kent. The fort was substantially built or rebuilt in the late 2nd century. to the beginning of 3. century. AD The standard pattern of coin losses across Britain is very low in the first half of the 3rd century. These coins appear in surprising numbers in Reculver. The presence of these unusual coins marks the site as unusually important for that period.13

Coin loss: Methodological prerequisites

The most important historical application of the study of coin losses is the study of revenue generation, where large changes can sometimes correlate with large economic changes in different regions. However, it is a more complex area than coin supply and requires a strong methodological framework.

The importance of archaeological context

The establishment of real coin losses during different periods of occupation depends on an accurate assessment of the context in which the coins were found. Without dated contexts for coins, one cannot study the frequency of coins lost over time, only broad patterns of supply to the site. Thus, although a site may have many 4th c coins, it is quite possible that these arrived and were lost much later when they were still in circulation, giving a false impression of coin loss. Unfortunately, the number of places where it is possible to study coins in context is currently extremely limited; most places come from Britain, in the Mediterranean there are only a handful, and of large places we only have Rome, Carthage and Jerusalem.

A further interpretation problem is posed by the physical processes in the formation of archaeological sites. Even if you have coins in dated contexts, you still have the problem of residuality - that coins in different layers may have been badly mixed, by digging tendon holes into earlier levels or through the actions of animals. This is not

13 Reece (forthcoming).

richard reece150

a serious problem in each location; however, it does occur, and is a phenomenon that you should be aware of. The excavations of the Crypta Balbi site in Rome provide us with an example of such a residency and a plausible method of discovering it.14

The Roman coins from this site occur mainly in contexts dated later than the 4th century. Many people take this to show that 4th century coins continued to circulate for up to 500 years after they were minted. This suggests that they have not read the reports. The authors make this suggestion one of their first inquiries, and they definitely demonstrate that the coins are found in deposits long after they were struck. They show a gradual decline in the number of coins deposited by deposit and question whether this actually shows continued use. The same work was done on the Roman pottery from the 4th it happened. From these studies, diagrams were drawn for both pottery and coins, showing that the rate of disappearance is the same for the coins and the broken pottery. Quite reasonably, the authors insist that if anyone wants to claim the continued use of the coins, they must also accept the continued use of worn and broken pottery. The case for continued use is practically rejected. At Crypta Balbi, the Roman coins and Roman pottery occur less and less frequently as the strata accumulate; the most obvious conclusion is that both are part of recycled junk and that the coins did not continue to circulate long after they had ceased to be supplied.

Interpretation of coin loss

Coins on Websites - Exchanges or Piles of Garbage?

Although most interest in coin losses is focused on overall site-wide patterns, the distribution of coin deposits across sites also raises some interesting questions. At this point it is best to point out that it is extremely difficult to interpret patterns of coins deposited at the site as related to exchange activities such as market stalls. To be available for discovery, coins must be deposited. This could either mean that they were accidentally lost and not recovered,

14 Deaf (1989).

coins and the late Roman economy 151

purposefully deposited for a reason, as at a shrine, or simply discarded. They will be at their find because whatever adventures they may have had after leaving a wallet or money box, their journey ended at the find and they were not moved afterwards. Some people are happy with the idea of ​​coins being spent. deposited for a reason. Coins can be placed in a shrine on a street corner or dropped into the cabbage leaves of a busy market and swept up at the end of the day. However, they get excited when it's suggested than anyone ever throws money away or, less emotionally, discards it.

Almost all excavated and published coins originate from the uncovering of the monuments beloved of Mediterranean excavators. In the cities they are market places, great houses, basilicas, gates and so on, while in the countryside they are forts with military inscriptions or villas with luxurious mosaic pavements. By studying archaeological contexts, one could reasonably assess whether this was related to the use of money at these sites, and then hope to identify places of exchange, such as shops and stalls. However, without knowing the precise context of finds, which is the usual situation, it is also reasonable to interpret the occurrence of coins as being a form of waste disposal in urban centers. 19th century industrialists in northern England summed up the feeling that you could only make money by getting your hands dirty in the saying "Wherethere's muck there's brass". Archaeologists may need to reverse the saying, as it seems likely that lost or discarded coins are markers of debris. If this is the case, coins will usually be concentrated in largely unexcavated waste sites on the fringes of sites. For example, if there are five coins in every truckload of household waste, then ten coins found on the road surface of a gateway suggest. very messy entrance to a city. On this basis, the ups and downs of coins at any site, whether viewed against a provincial background or not, trace clutter, waste and disorganization, or their absence. A number of coins found in the same area of ​​an excavation may therefore indicate a pile of rubbish rather than the site of a thriving market stall.

Site without coins

What about sites that don't show much, if any, coin loss? What does the absence of coins mean and what does the few coins found on such sites mean? The most recently published page that approaches this

richard reece152

subject of great clarity and excellent technique, is the late villa at San Giovanni di Ruoti in southern Italy.15 This large and thoroughly excavated villa was inhabited for much of a 500-year period and produced 41 coins ranging from the Roman Republic to around 400 AD The greatest period of the villa's life, from 400 to 520, is characterized by substantial, even palatial, buildings and a large center with large quantities of ceramics. No coins minted in this period can be identified, and only 15 out of 41 coins cannot be firmly placed in the years 1 to 400 AD. appearance. This leaves only 6 coins that could even belong to the 5th century. This detail is given to show that there are places in the middle of the Mediterranean with high status and flourishing economies that hardly participate in the use and loss of small change. At SanGiovanni this pattern, followed by the site throughout its history, continues into the 5th century. even when strong activity is quite evident from the excavated record. Why this should be the case is uncertain; only when additional sites are excavated will it be clear whether this is an accident or perhaps related to special circumstances of the use of coin in a site of this type.

Mønttab and City History - Carthage

Hitherto it has been argued that the only way to approach the economic history of places is by considering their coin supply in relation to the surrounding region. Coin loss generally measures relative revenue generation. But when coin loss is real, derived from coins found in dated deposits, it can occasionally be used as a guide to the comparative history of different periods in the same place. This can be done when the life of the coin is very long; this makes it possible to compare the relative use of coins between different periods in the same place, as old coins that were lost very late must have been in circulation for a long time. This is useful for studying places that we have other reasons to believe underwent periods of great economic change, as has been suspected for Carthage after the Vandal conquest.

Carthage provides important evidence for the interpretation of

15 Small and Buck (1994), Simpson (1997), Reece (1997).

coins and the late Roman economy 153

coin finds highly relevant to any discussion of sites between A.D. 300 and 600. Here we not only have an idea of ​​the overall pattern of coins delivered to the site, but also the placement of coins within dated deposits and then their loss pattern.

The key issue in Carthage can be summed up in two sentences:1. A good number of coins from the 4th century were lost at Carthage2. Very few coins were lost in the 4th century. at Carthage.

There is no contradiction between these two sentences because they constitute two very different statements. The difference between them must be understood to get the full value out of coin finds. Item 1 comments on the number of coins found in the British excavations at Carthage which were minted in the 4th century. (and were probably mostly delivered to Carthage within a few decades). Item 2 comments on the number of coins found (and therefore lost) in 4th century deposits. These are two completely different statements with very different implications.

Point 1 suggests that there was a fair coin supply to Carthage in the 4th century but little more. If this had not been the case, would get 4thc. coins could ever have been lost there unless they were imported in bulk long after they were struck. There is no reason to assume this, and what evidence there is from elsewhere speaks against it. Point 2 deals with coin losses and allows for many suggestions about coinage in 4th century Carthage about the running of the city and by extension, the lifespan of coins in use. The coin lists from the various excavations show that many coins from the 4th century reached the city and probably circulated there. We are unable to compare Carthage's offerings with a regional context and therefore do not even have a locally relative measure of the city's economic history. However, most of what we know from archeology and from written sources suggests that Carthage was a thriving city in the 4th century. It's only a minor stretch to propose a coin-powered exchange of goods that runs at a comfortably high level for those with money. Of course, this is a hypothesis.

If we were to take the evidence of low coin loss at face value, as an index of economic activity, we would have to declare 4th century Carthage to be a near-deserted disaster area, because there are actually very few 4th century coins . levels. Few would believe this second interpretation - it ignores the significant buildings

richard reece154

4th century and all written texts; most would accept the image of a thriving city and ask for an explanation for the scarcity of coins. But if it were suggested that coins were valued, that they were in such dynamic use that they rarely hit the ground, and that when they did the ground was clean enough to find them and pick them up again, few would argue. This seems like a perfectly reasonable explanation. Lowcoin losses when you know coins are being spent can actually indicate success rather than failure.

The pattern of coin supply in Carthage changes after 400. From about 400 to 480 the average date of coins in subsequent strata does not become later. The pattern known before 400 then begins with the beginning of the coinage of the Vandals around 480, followed by the Byzantine reconquest. Again, in the absence of a wider regional supply background, it is not possible to estimate the city's relative prosperity compared to the rest of Africa. However, deposits from the 5th century up to about 470 continue to show low coin loss, although these coins are mainly up to 100 years old and more. Again, a traditional explanation of low coin losses could be attempted, whereby the problems in the years around 400 unsettled trade in Carthage and a death blow was inflicted by the Vandal conquest. This is what caused the lowcoin loss in the 5th century; there were few coins to drop and no change to lose them. This argument obviously requires an acceptance of the 4th century. poverty in the city.

Another obstacle to this argument is the context in which the coins of the 4th century. went missing. Only half of the coins minted in 330 to 345 were sealed in repositories, lost without recovery, around 510. This means that the other half of the coins, together with surrounding issues from the 4th century, were still available to be lost in the 6th century .. when they are lost in significant numbers. Thus they survived the 5th century somewhere, allowing them to continue in use afterwards and to form a significant part of coin losses up to about 600. Catastrophe in the 430s would have led to the abandonment of coins and coinage - and I think I can point to this very phenomenon elsewhere. This shows that there was indeed continued coinage, with low coin loss, into the 5th century. I therefore think it must be allowed the same meaning in the 5th century. The coins lost in the 5th century are basically the same as those lost in the 4th century. because the 5th century question from the eastern Mediterranean remains firmly in the eastern Mediterranean, even when areas like Sicily and Malta are still a full part of the empire. Thus at Carthage, a combination of a good knowledge of the city from the

coins and the late Roman economy 155

chaeology and texts and an unusual pattern of coin loss allow us to establish a strong continuity of coinage over some two hundred years of the city's history as it experienced major changes in political regime.

Admittedly, this story has been constructed very selectively from the coins found in the British excavations.16 Clearly, other sites in Carthage should be brought into play, for comparison or contrast, and ideally others should North African cities are considered as well. Unfortunately, many of the coins from other excavations have not yet been published, and those that have cannot be put back into the deposits where they were found based on the information reported. The thought that other North African cities could be considered also raises a sad smile, because with one or two exceptions, no other good coin lists have yet been published. The main exception is the Moroccan site of Zilil, studied by my long-time numismatist Georges Depeyrot.17 It is very comforting to be able to say that he has also independently worked on the idea of ​​coin life by examining the coins in each stratigraphic unit, and our ideas are very similar . Unfortunately, the latest dated levels at Zilil are shortly after 378, so the site cannot confirm the ideas given above after that date.

Politically motivated coin loss

Not all patterns of coin loss relate to economic pressure. The coins from Kenyon's excavations in Jerusalem yield about 2340 identifiable coins, ranging from about 300 B.C. to A.D. 1960/18

The general loss rate was therefore about one coin per year for over 2000 years. There were times of greater coinage loss, such as the last century BC. or the 18th century A.D. and there were almost gaps in the series. But three years stand out because they account for 209 coins - about 10% of the total. The years are 67, 68 and 68 AD. - once these dates are given, everything becomes clear; these are of course the coins of the Jewish revolt and are unusual because they were produced locally in quantities because their message was strongly against the occupying power and the people who issued

16 Reece (1984), (1995).17 Depeyrot (1999).18 Reece et al. (coming).

richard reece156

they lost their fight for independence. This meant that a large number of coins with little intrinsic value had no future and were dangerous possessions; the simplest explanation for their occurrence at archaeological levels is that they were quickly discarded. Here we can see a clear example of an increase in the discarding of coins for political and not economic reasons. Although no examples of this type of discard have been identified so far for Late Antiquity, it is possible that they will be in the future.

Major changes in monetization

When coinage systems change, denominations fall into disuse, and lower value base metal coins may lose their symbolic value. They have only their intrinsic value, which is somewhat less than a metal disk, which at least has a use to which it can be put. Demonetized coins can be given to children to help them play stores, they can be used as counters for games, or they can be left somewhere and forgotten. Some of the most commonly found coins in northwestern Europe, whether as single finds or as magazines, are poor quality ray coins from the period AD 260 to 274. The simple explanation for their number in the archaeological record is that they lost their value and were set aside or discarded. The reasons why such changes occurred can vary - in many cases it seems to be related to periods of economic stress - sometimes local, sometimes felt in many regions. Late Antiquity presents several examples of large-scale changes in revenue generation that will be considered here.

Britain in the late 3rd century.

In later Roman Britain, Verulamium is close to the average type of town in its coin finds. At Verulamium comes a large coin loss at the end of the 3rd century, the period with very low value radiate coins. It could be that the late 3rd century was a testing time for many of the cities of Britain, especially those to the east, and there was doubt as to whether the coin economy of the cities would survive. Indeed it did, although in most cases it never returned to full strength. If only we had some good city coin lists from France, we might see similar events at work there, allowing us to connect the essential information in Britain with the poorly recorded Mediterranean.

coins and the late Roman economy 157

The end of coinage in the 5th century west?

For Late Antiquity, one of the most interesting questions is the general cessation of low value coin losses in the western Mediterranean area in the 5th century. Such a cessation cannot be proven by a simple coin list that gives all coins found at the site in the order in which they were struck. The absence of coins struck after 410 does not necessarily show the end of coinage in 411. It merely shows the end of ​​the supply that is common to the area. Cessation of coinage can only be suggested by listing coins from layers above the first coin in 410 in the sequence and thus comparing them with coins from earlier periods, while looking at pottery in the same layers for evidence of remains. This kind of information exists for very few places - so again we are limited to Rome, Carthage, Jerusalem and places in Britain.

Rome: Slow down or stop?

I think that in Rome in the early 5th century. coinage slowed or stopped, leading to a major episode of discards. We have few coin finds published from Rome, but what there is points in the same direction. Of particular importance in this case is the CryptaBalbi, where the excavators have produced a complete coin list and a discussion of the 212 coins in their contexts, which is unique in Rome.19 There are also 77 coins from several seasons of excavation at the Schola Praeconum, 20 and 5231 miscellaneous finds from the clearing of the Forum and Palatine areas, although in most cases these lack further context.21 Frankfurt numismatists are currently working on two other groups of coins from Rome, again without firm context, but their results are not yet complete.

In the coins of the Forum and Palatine clearing, several deposits were kept separate, and one from the Basilica Aemilia contained a complete hoard, which is able to show us what coins were in circulation around the year 410. A provenance in the basilica seems acceptable, because the deposit included fragments of marble floors, molten bronze and lumps of charred wood. This is consistent with the evidence from the surviving floor

19 Sorda (1989).20 Whitehouse et al. (1982); (1985).21 Reece (1982a).

richard reece158

basilica where there are irregular spots from pools of bronze and curved edges where marble has been chipped away. This destruction is attributed to Alaric's sack of Rome in 410. The treasure is described as complete because it was still covered with threads of coarse linen that had almost certainly been burned with it; corrosion products from the copper coins had been absorbed into the ash from the wires.

When the hoard was cleaned and the coins identified, there were 100 coins or fragments of coins, the latest of which was one of PriscusAttalus from the year 409. The rest of the hoard dated back to c. 356 and then ended up with only one or two earlier copies. The importance of this hoard obviously lies in its demonstration of the coin pool in use around the year 410. However, the coins marked from the basilica showed broadly the same chronological distribution as those in the hoard, with numbers extending back. to about 350, with very few coins before that. It is very likely that they also represent coins in circulation in 410, when the basilica is believed to have been destroyed. They can even represent the coin used by the money changers in this place, in the center of the forum.

Collections of coins found elsewhere in the Forum and Palatineareas could be divided into two main groups. A category concentrated on coins from the Republic up to the 4th century, consisting of fairly uniform losses over about 600 years. The second category was very similar to both the Basilica Aemilia hoard and the general finds from this area. No group had a number of coins struck after about 420, although single examples, and even medieval and later coins, had been incorporated into several repositories over time. Looking at the finds in Basilica Aemilia, the composite picture throughout the Forum and Palatine area suggests a gentle and uniform loss of coinage throughout the Roman period, ending with significant loss hoards in the first twenty years of the 5th century. After that coin loss when it happened was very low.

A further group of coins comes from the Schola Praeconum. This site, regardless of the name and purpose of the structures, consisted mainly of a rubbish dump of animal bones, pottery and a few coins. These deposits are in close proximity to the deposits from the clearing of the Palatine and Basilica Aemilia, which included a large number from the later 4th century. coins. But the coins here belong to the 5th century, with one or two earlier issues from the later 4th century, and stop at about 450. The pottery has been variously dated to about the end of the

coins and the late Roman economy 159

5th c. or shortly after 450.22 The key point is that 5th c. issues are present in small numbers, but very few 4th c. issues have survived and those that have are insecure and worn. This suggests that the large loss or discard in Rome shortly after 400 marks a break in general coinage in the city, which was reestablished rather weakly later in the century. The remaining coins at Crypta Balbiseem confirm the suggestion that widespread coinage was not re-established after the early 5th century.

UK

If the scarce numbers of 408 to 410 are removed, then the composition of the magazine from Basilica Aemilia is very similar to the coins found in recent deposits at some sites in Britain. In Britain, few would now doubt the equation of the number of coins from the later 4th century, in terminal waste deposits, with an end to coinage and the accumulation of waste on sites that were being abandoned. In Cirencester, coin losses after these deposits are unknown, and it is therefore reasonable to assume that coinage ceased after c. 415. The clearest evidence so far (at present unpublished) comes from Ware in Hertfordshire. The average date of coins in subsequent strata becomes progressively later, as might be expected. But after the first layers are dated with coins to about 400, the coins that do appear seem to be almost a random selection, and the average date of each deposit seems completely erratic. I assume that after the end of coin supply, coins that occur in later layers are mainly recycled waste.

Africa

At Carthage there are few coins in 5th century deposits, and none of these correspond to the wholesale deposit of 4th century coins as in Rome. The discarding of 4. century coinage in Carthage occurs in the middle of the 6th century. after the Byzantine reconquest, as will be discussed below. Zilil, in Morocco, follows the British rather than the Roman pattern.

Summary and interpretation of Western websites

So in summary: coinage in Cirencester and Zilil ends shortly after 400. I think there may be a serious blow to coinage after 410 in Rome,

22 Reynolds (1995).

richard reece160

with several almost coinless decades. In Rome there are later deposits of coins, notably the Crypta Balbi and the ScholaPraeconum, but in these deposits the coinage is from the 4th century. is much rarer; they are basically 5th or 6th c groups with a few earlier coins. At Carthage the coin loss continues at its low level without interruption. It seems clear that Britain did not have a monetary economy after 400, but Carthage did, as did Rome, albeit in a weaker way. Nevertheless, there is still an overall pattern of a collapsing coin supply that needs to be explained.

Coin in a system

Perhaps the important point to draw from all this is that the significance lies not in the presence or absence of individual coins, but in the existence of a system in which such coins had an agreed value. This nuances our interpretation: I think most people assume the continued use of gold, certainly in Italy in the 5th century, and probably also in Spain and North Africa. If gold was in use, then the backing of atoken currency such as copper was in place; however, the evidence we have suggests that it was apparently not functional. Does this mean that gold had become a commodity with coins as units of goods, so that the greater - gold - was seen as having no relation to the smaller - copper - other than through their intrinsic values? Even then it would have been possible to produce copper nuggets, which in a purely internal sense related to the gold piece, but this was not done.

I don't understand this. I accept that the state is necessary to validate the issue, the size, weight and fineness of gold pieces. The many outbreaks of copying copper coins throughout the empire, but especially towards the borders, show that the local population was quite capable of taking the production of change into their own hands when the official supply was cut off. If acceptable and accepted gold was available in the western Mediterranean area in the 5th century, why were local copper coins not produced to facilitate buying and selling (emendi et vendendi utilitas), as Anonymus DeRebus Bellicis (I.6) states as a main purpose of money?

Monetization in the East: Lack of copper coins and copying

Roman coinage in the 5th century East shows no signs of demonetization. In Jerusalem, the use of small copper coins runs strongly throughout

coins and the late Roman economy 161

in the 1st millennium AD Indeed, in Egypt there appears to have been a period of scarcity of copper coins, when the demand exceeded the supply, judging from the copies of coins and the molds for copies. Dr. H. C. Noeske is currently studying such forms from Alexandria, and has published a brief survey of copies from site finds.23 There are unstudied examples in the Flinders Petrie Museum at University College London, and examples now in the National Museum in Warsaw from Edfou (Apollonopolis Magna ) has been published.24

The latest coin used to make impressions of the Warsaw form belonged to about the year 400, but Noeske's copies continue with late Roman and Aksumite coins until the reform of Anastasius.

The coins from such molds are very thin pieces of cast copper. They resemble the normal struck small coins of Vandal Carthagein shape and size, but are often totally illegible. The overprinted forms from 4th and 5th century coins strongly support the evidence from 5th and 6th century strata at Carthage that the coinage, when used in the 5th century, included a high proportion of issues over 100 years old . I have the impression, and I cannot say it more definitely, that where new coins appear in the 5th century, there is a low point in the third quarter of the century. This may be the point where copying takes place, rendered unnecessary by the larger coins struck after the reform of Anastasius.

I suppose it is still possible at this point in our ignorance to propose a completely different alternative where the copying takes place after the reform where change ceased to be provided by the state. This interpretation would be consistent with the work of Noeske, who has found copies of coins up to the pre-reform issues of Anastasius. We need a thorough re-examination of a large group of the 5th and early 6th centuries. finds to clarify some of the many uncertainties. It may be that the place to look will be Corinth, where all the small late coins were collected in the original publication as they needed further study. Although there is no hope of collecting the deposits they came from, the time may be ripe for a full reissue. Current excavations, which are carefully published, could be used to suggest contexts for the various coin types.

23 Noeske (1998).24 Krzyzanowska (1988).

richard reece162

East and West

This issue of late copies once again highlights a difference between East and West. In the eastern Mediterranean, lack of regular supplies of small money may well have been remedied by local production. However, the absence of supplies of copper coins to the western area did not provoke any significant reaction. There is therefore a core area where coins are needed, wanted and domestically produced if necessary and another area towards the West where ceramics and other goods are traded but which do not seem to need change and certainly do nothing to compensate for it , when it is not delivered. The kingdom of the Vandals lay at the crossroads between the two areas. Unlike Malta to the immediate north, it produced its own copper coinage from about 480 onwards; there therefore seems to be a strange way in which the rebel kingdom kept modern Tunisia attached to the Greek East, when it has always been thought of as part of the Latin West.

The Visigothic Kingdom of Spain and Provence had no such effect. The Visigoths seem to have been responsible for a good portion of the gold coinage produced in their area in the 5th century, but they never produced any extensive copper coinage that has been identical until now. tified. In Italy, however, the Ostrogoths produced a good gold coin age in the imperial name and tradition, along with some silver and some very good, but rare, copper coins. Editions such as those of Athalaric are of good size and fairly distinct in design, so they have never been relegated to the 'uncertain 5th to 6th century' category. It seems likely that it was the Ostrogoths who produced marks 42 and 83 (possibly denarii or nummi) of quite fresh issues of Vespasian, Titus and Domitian (AD 69–81); the only convincing explanation for these is that a large hoard of Early Imperial copper was discovered and judiciously reused. But the coins were struck in mints in Rome and Ravenna from the 550s to the beginning of the 7th century. are rare assit finds and can hardly be used to suggest that a functioning coin economy still existed. Heraclius' copper issues in Rome and especially in Sicily suggest a rather ham-fisted last gasp than a market-driven flourishing coinage.

coins and late Roman economy 163

Value collisions

Site finds can sometimes give us information about the effects of the collision between two economic systems that have different price ranges and coin values.

Jerusalem

Post-antique coins from Jerusalem deserve future study and especially comparison with other major groups in the area when these become available. The history of the area can be written to some extent from the coins found, but there are anomalies that make it clear that the coins say something different than the texts. Thus the Byzantine coins stop in the 7th century. while they continue in Greece and large parts of Asia Minor - that's how it is. But the coinage of the 12th century consists of a number of Islamic issues, very few issues from the Frankish west, and only one or two issues of the Crusader states. It may be significant that the Islamic issues are entirely copper, while the Franks are silver, as copper coins were known in the West at the time. Perhaps the economy of Jerusa-lem, which had a continuous tradition of the use of small money from at least the 2nd century B.C., and the economic ideas of the newcomers formed a mismatch that could only be resolved by the use of local small money. Future studies of the context of the Jerusalem coins may show continued use of Hasmonean and Herodian small coins well beyond the strike period. This is because the Roman system, which went down to only a quarter As (the quadrants), lacked the smallest coin used in daily trade, the minutum. This comes to us directly from the Gospels (Mark 12.42), where the widow gives two minutes (Greek lepta) 'which form a quadrant'.

The Byzantine Conquest and Western Coinage

A very similar case is demonstrated in Carthage in the 6th century. A.D. Carthage drifted from the central Roman coinage system, and much else, with the establishment of the Vandal Empire around 430, but was brought into the Eastern Mediterranean coinage system with the Byzantine reconquest around 540. From the 470s to the 530s, the Vandal state produced a large amount of copper coins. The overwhelming value lost was the single nummus, a very small, thin and crudely made coin. There were some larger designations

richard reece164

nations, but they are rarely encountered. As they do not occur in the 'expected' ratio that their value might suggest, it seems likely that they did not circulate as freely as the small coins. If all other factors were equal (number in circulation, usage times, and soon), then I would expect a coin with a value of 1 to be lost about five times more frequently than a coin with a value of 5. When testable, this idea looks appear to hold true. In Carthage, however, the nummus seems to be the standard coin for market purposes.

During this period, the central authority in Constantinople issued copper coins of up to 40 nummi in value. Ones and twos were known at some mints, but the standard coins found usually ranged from fives to forties. There is no doubt about that, because the value is the most obvious feature on the reverse side of the coin. A is one, I is 10, K is 20 and M is the big 40 nummi piece. The Byzantine occupation of Carthage brought the Vandal system, based on nummus, into collision with the Byzantine system, practically making noise. Again, as in Jerusalem, a metropolitan system had a higher value system than the provinces. This separation continued well into the 19th century, when Britain had to strike half and third specifically for new Mediterranean colonies such as Malta, where a lower price system than that of Britain survived.

In 6th century Carthage, a Byzantine mint was quickly opened, striking a number of Justinian single nummi, but the North African system quickly succumbed to the inflated values ​​of the Byzantine state. Vandal issues fall out of use and are kept in magazines containing only a few current issues, giving them their undoubted date. The killing of 4. BC coins also seem to occur mainly around this time (520 to 570), and may therefore be related to these changes. After Justinian's death, when the effects of reconquest seem to be strongly felt, exchange seems to depend on a much smaller number of higher value coins. By 580 individual nummi were a thing of the past, and by 620 the economy had come into line with central authority.

In these two examples, the results of the collision were different. The price and coinage of the Vandal kingdom was practically destroyed by the Byzantine reconquest. This appears clear from the nature of the finds at the site because coin values ​​certainly changed and I believe this is a reflection of price changes. In the case of the Crusades, the incoming economy seems to have

coins and the late Roman economy 165

been cast by the home (Arab) economy because the coins in use appear to have continuity. Since coinage in Carthage appears to have declined sharply after the Byzantine arrival and before the Arab conquest, it may be that the effect of the Byzantines was to destroy what had been a thriving market economy. The smaller effect of the Crusades suggests that the violent military disruption had a less violent effect on the markets. As always, this must be judged on the basis of coin finds, and these, as always, reflect daily market transactions. No logical extrapolation can be made to the great state economies of the modern period because so many models can be built that all agree with this small piece of material evidence.

Conclusion

The interpretation of coin finds requires considerable methodological caution. However, there are many significant questions for the study of the late antique economy that can only be resolved through numismatics. To summarize: coin supply from coins as found can give us essential information about how state payments worked in practice and provide a measure of a site's relative economic history in relation to its region. Coin losses can tell us about the relative economic history of two or more periods at the same site, as well as about changes in regional monetization and the interaction of different monetary systems. Unfortunately, there are probably less than 20 coin reports from the entire Mediterranean area that list enough coins in sufficient detail to provide a basis for expanding this type of work. Despite this, a start can be made and more reports should be published soon. I hope I have shown that the form of issue is crucial, because details about the context in which the coins were discovered add dramatically to the information that can be extracted from crude lists. Nevertheless, where information on stratigraphic relationships is not available, this must not deter workers from publication because there is still much work to be done on simple lists.

richard reece166

Recognitions

I am grateful to the editors who restructured the paper to make better sense, and to Paul Arthur, Kevin Butcher, Georges Depeyrot, Pe-ter Guest, Paul Reynolds and Jonathan Williams for reading through the manuscript and making many helpful suggestions.

coins and the late Roman economy 167

Bibliography

Banaji J. (1996) "The Circulation of Gold as an Index of Prosperity in the Central and Eastern Mediterranean in Late Antiquity" in Coin Finds and Coin Use in theRoman World, ed. C. E. King and D. G. Wigg (Studien zu Fundmünzen derAntike 10) (Berlin 1996) 41-54

Buttrey T.V. et al. (1981) Greek, Roman and Islamic Coinage from Sardis (Cambridge(Messe) and London 1981)

Carson R. A. G., Hill P. V. og Kent J. P. C. (1960) Late Roman Bronze Coinage (London 1960)

de Callatay F. (1995) "Calculating old coin production: seeking a balance", Numismatic Chronicle 155, (1995) 289-312

Depeyrot G. (1979) “On the old money supply: the example from the XVII.-XX. centuries” in Symposium numismatico de Barcelona I, ed. J. P. Gurt and A.M. Balaguer (Barcelona 1979) 29-38

——— (1999) Colonia Iulia Constantia Zilil; Study of Cash (Collection of the French School of Rome 250) (Paris 1999)

Hahn W. (1973) The Coinage of the Byzantine Empire (Vienna 1973), revised edition in English, forthcoming

——— (1976) Coin finds of the Roman period in Austria, III/1, Carnuntum, (Wien1976) 156-62

Krzyzanowska A. (1988) "A mold for casting coins found in Egypt", RocznikMuzej Narodne XXXII (1988) 81-91

Mattingly H., Sydenham E.A., Carson R.A.G., Sutherland C.H.V. og BurnettA. M. (1923-94) Roman Imperial Coinage bind I til X (31 f.Kr. til 491 e.Kr.) (Lon-don 1923-1994)

Noeske H. C. (1998) "On the Cast Imitations of Axumite Bronze Coins in Egypt and Palæstina", ΘΕΜΕΛΙΑ (Late Antiquity and Coptic Studies 3) edd.M Krause og S Schaten, (Wiesbaden 1998) 249-65

Preda C. and Nubar H. (1973) History III: Monetary Discoveries 1914-1970 (Bucharest1973)

Reece R. (1978) "Bronze Coinage in Roman Britain and the Western Provinces" in Scripta Nummaria Romana, Essays Presented to Humphrey Sutherland, ed. R.A.G.Carson and C.M. Kraay (London 1978) 124-42

——— (1982a) "A Collection of Coins from the Center of Rome", PBSR 50 (1982)116-45

——— (1982b) "Roman Coinage in the Western Mediterranean: A Quantitative Approach" Opus I/2 (1982) 341-50

——— (1984) "Coins" in Excavations at Carthage: The British Mission, bind I, 1, edd.H. Hurst og S. Roskams (Sheffield 1984) 171-81

——— (1994) "Coins" in Excavations at Carthage: The British Mission bind II, 1, udg. H. Hurst (Oxford 1994) 249-60

——— (1997) “Coinage and the Economy in the Later Phases of San Giovanni” in Simpson (1997) 83-87

Reece R. et al. (forthcoming), "The Coins" in Excavations in Jerusalem 1961-67, ed.K. Prague, (British Academy, London, forthcoming)

Reece R. (forthcoming), Coin Report in Forthcoming Reculver Excavation Report by B. Philp

Reynolds P. (1995). Trade in the Western Mediterranean, AD 400-700: The CeramicEvidence (BAR International Series 604) (Oxford 1995), Appendices D19, D20,327-32.

richard reece168

Simpson C. J. (1997) The Excavations of San Giovanni di Ruoti II (Toronto 1997) Small A. M. and R. J. Buck (1994) The Excavations of San Giovanni di Ruoti I, (Toronto)

1994)Sorda S. (1989) ed. Coins in archaeological contexts. Examples from the excavations in Rome (Instituto

di numismatica, studie e materiale 2) (Rom 1989) Whitehouse D. et al. (1982) "The School of Heralds I", PBSR 50 (1982) 53-101——— (1985) "The School of Heralds II", PBSR 53 (1985) 163-210

coins and the late Roman economy 169

DEL FOURTOPOGRAPHI

this page is intentionally blank

you ancient bytopography 171

LATE ANCIENT URBAN TOPOGRAPHY: FROM ARCHITECTURE TO HUMAN SPACE

LUKE LAVAN

Abstract

Writing about the late antique city is dominated by topographical mapping, architectural studies and site syntheses. These approaches operate within a conception of space as the location of points in an imaginary grid. However, there are other notions of space. This article proposes a move from the location and description of physical remains to an investigation of human spatiality. It also seeks to re-establish the study of topography at a general, rather than site-focused, level. The limitations of existing approaches and the needs for late antique evidence are examined. An alternative topographical approach is proposed, based on studying 'activity spaces' (human activities in their total material surroundings) rather than simple buildings. This gives particular prominence to texts, but seeks to combine all forms of evidence. The methodological issues involved in doing this are considered. Possible implications for archaeological field practice are also explored.

Where did we lose the late antique city?*

One of the most evocative 'late antique' experiences one can have is walking around the ruins of an extensively excavated Mediterranean city, where much of the site has been preserved largely as it was in the final years of antiquity. Amidst the remains of churches, colonnades and bathhouses, one feels very close to the human events remembered by late antique writers, which so often took place in urban settings. This immediacy is unfortunately lost back in the library when reading academic studies of urban monuments. The liberation this brings about is partly a result of the complexity of urban data. However, it is also due to a scientific desire to master this

* This paper was written before the advent of Allison (2001), which covers some of the same topics in an early Roman context and contains very interesting bibliography.

L. Lavan og W. Bowden (red.) Theory and Practice in Late Antique Archaeology (Late Antique Archaeology 1 2003) (Leiden), s. 171-195

luke lavan172

information rather than synthesizing it, and especially from the use of spatial concepts and methods that impoverish writing about Late Antiquity cities. This leaves them empty and cold, like physical descriptions and line drawings, without life. Reconstruction drawings of these buildings are common, but few are drawn in use; therefore, not only scholars but also the general public have little or no idea what the ancient ancient cities looked or felt like. This is not a brilliant situation to be in after decades of excavation and library research. In this paper, I argue that it is possible to expand our understanding of late antique urban topography by rethinking the spatial concepts and methodology that we use, and especially by implanting a human perspective at the center of our writing.

'Space' outside late antique studies

Within the humanities and social sciences, there is a wide range of approaches to the study of space. This work has evolved from multiple disciplines, creating a wide variety of concepts; these have either been well defined or simply assumed according to the priorities of different researchers. The best established tradition conceives of space as the arrangement of points in a physical grid. Although present in the ancient world and other pre-industrial civilizations, it received its classical exposition in the works of Descartes and Newton; this concept of space has often been associated with traditional regional geography, in terms of, for example, the study of settlement patterns and the location of industry. Another tradition emphasizes the importance of using spatial concepts based on human action; this has been explored by geographers interested in developing a sense of place or in describing the perceived geographies and spatial vocabularies carried around in the minds of different individuals. A third line of thought sees space not so much as a subject in itself, but as an inherently separable part of human action through which all aspects of our lives are constructed and negotiated; this is particularly associated with concepts of 'social space' developed by the philosopher Lefebvre. Many concepts from the second and third traditions outlined here have been tested in archaeology, including in Roman cities.1

1 My description of space concepts is a gross simplification, but useful for the purpose of this article: for current definitions of space used in hu-

senantik bytopography 173

'The room' and the late antique city

The most systematic spatial study of the late antique city is that of Dietrich Claude, who in his Die byzantinische Stadt (1969) devoted almost 100 pages to Topographie. This work sought to list monument types built or maintained in Mediterranean cities in the 6th century. A.D., using archaeological and literary sources. In its time, it gave researchers a first glimpse of the special nature of cities in this period. Since this book was published, there has been an enormous amount of new archaeological work which has greatly changed our understanding of urban history. In early medieval archaeology, in Western Europe, this has included a great interest in the broad topographical development of cities.2 Within the more classical-looking cities of the Late Antique East or West up to the mid-5th century, few researchers have followed in Claude's footsteps and written about topography in a systematic, holistic way.3 Broadly speaking, three types of spatial writing have dominated research: that done by self-conscious topographers, that done in recording architecture, and that which is carried out as part of site syntheses and guidebooks. All of these exist within the first concept of space outlined above: that of mapping physical phenomena against an absolute grid.

Topographic mapping

Topographic mapping, more commonly known simply as 'topography', is an academic practice that has grown into classical archeology out of geography. It usually involves extensive recording of the broad outlines of places, from their natural surroundings to

manities and social sciences, with bibliography see Curry (1996) and Simonsen(1996). The discussion in Alston (2002) 12-43 is also useful. For a review of spatial approaches in Roman archeology see Laurence (1997).

2 See for example Brogiolo and Gelichi (1998).3 Claude's work significantly studied not only construction work but occasionally

are considered relevant literary sources for monument use. German scholarship has continued to study topography, though not in as systematic a manner as Claude:Bauer (1996); Bruhl (1979, 1990). W. Müller-Wiener's Bildlexikon zur TopographieIstanbuls (1977) compiles available evidence for Constantinople but does not seek to synthesize it. Mango (1990) does this, for monuments and large-scale urban changes, though not everyday life, as see Mango (1981-82). E. Zanini's chapter on "Le città dell'impero," in Introduzione all'Archeologia Bizantina (1994) 117-71 is the only work to preserve Claude's ambition for a general treatment of Late Antiquity, although it organizes material in terms of urban syntheses rather than as Claude did, by theme.

luke lavan174

the location of monuments in them. In this, it fits traditional dictionary definitions of topography as "the location of physical phenomena" or "where things are", though not newer concepts such as "sacred topography", which treat topography as equivalent to "spatial narrative".4 Topographic mapping is particularly cher. -have been into classical urban archeology because it seems to give greater order to the complexity of the ancient city, and because there is still so much to do from it. It is seen as a necessary prerequisite for all further fieldwork. Much of this work is carried out as an integral part of excavation campaigns and not as independent research. Other studies are carried out as library research: locating monuments from old excavation reports, textual studies or even scrutinizing place names. In both kinds of work, topographers have a secondary historical object. This usually involves three areas of interest: the first is to relate the implantation of cities and buildings to relief and other natural features; the second is an attempt to explain the motives behind the design of newly planned urban areas; the third seeks to account for the surroundings and design of individual buildings in a new or existing urban environment. The latter approach was that used by Krautheimer in the introduction to his Three Christian Capitals, and seems to be implicit in the work of the Topographie Chrétienne series and C.-R. Bruhl's Palatium und Civitas.5 For the Republican and early Imperial period, this work can be very successful, providing an opportunity to study concepts of urban design and ideal forms in a period of colonial foundations and urban growth; it also helps one to think through the intentions of the city architects from this period, where monumental ideological programs and symbolic intentions often seem to have played an important role in the design and relative position of many monuments.6

Unfortunately, this work is somewhat less useful for Late Antiquity, despite Krautheimer's early optimism.7 The 4th to 6th c.

4 Oxford English Dictionary "the practice of describing a particular place ... delineating the features of a locality.. the study of local distribution" - Simpson and Weiner (1989)257.

5 Krautheimer "Introduktion" i Three Christian Capitals: Topography and Politics(1983) 1-5. Gauthier og Duval et al. (Paris 1986 f.). Bruhl (1975, 1990).

6 Urban planning: see Ward-Perkins (1974) and Owens (1989) but note critical reflections in Laurence (1994) 12-19 and more recently articles in Fentress (2000).W. L. Macdonald himself sees the placement of early imperial street monuments as planned in relation to architectural criteria: (1996), e.g. 5-31

7 Krautheimer (1983) 5.

you ancient bytopography 175

so far fewer city foundations than the 2nd B.C. to 3 AD New towns were small and contained only a limited number of structures. Orthogonal planning of new urban neighborhoods in existing cities was rare. there is certainly some information to be gleaned from the choice of sites, but inertia and practicality intervene in situations of reuse to make this less useful as a source of intentionality.9 Furthermore, new buildings may show a lack of care in their placement compared to early imperial monuments .10

There was probably far less clear ideological planning in Late Antiquity. Krautheimer argued that some larger churches in imperial capitals were meaningfully located: he believed that Constantine's churches in Rome were built on the periphery because they sought to avoid conflict with the pagan senate in the center of Rome; San Lorenzoat Milan was built by Ambrose outside the city, as a new focus, in opposition to the Arian-dominated court. But since he wrote, his conclusions seem less certain. The location of Constantine's church on the edge of Rome can now fit into a pattern observed elsewhere, whereby churches were built away from urban centers, on available land; his conclusions about Milan have been undermined by changes in San Lorenzo's chronology. Furthermore, Topographie Chrétienne's considerable effort to determine the factors behind the placement of churches rather sadly concluded that pragmatism rather than ideology played a large, if not the largest, role in the placement of many churches, especially those in the cities.11 Gisella Cantino Wataghin has also pointed out that ideology-

8 For new towns see Zanini in this volume. The use of orthogonal grids for new urban areas was rare: 5th century urban expansion at Carthage continued the existing orthogonal grid: Canadian Carthage Team (1985) 21. New 5th century neighborhoods at Sitifis and Scythopolis were ordered but not orthogonal: February (1965); Foersterand Tsafrir (1987-88) 40-41. Tsafrir and Foerster (1997) 104-105.

9 This is not to say that topographical surveys that consider reuse cannot produce important historical information: see Leone in this volume. Urban architectural design was not entirely over: the continued placement of some street monuments in relation to their wider architectural context is argued by Bauer (1996) 363-94.

10 Mango (1976) 30-35 does this in relation to the cathedral of Gerasa. For less ideological planning in relation to the early empire: Gauthier (1999)195-98, 202. However, fortifications are more carefully placed in relation to natural relief than early imperial defenses had been.

11 Gauthier (1999) 198-202 and Cantino Wataghin note in this volume that churches often seem to be located due to the availability of building plots, lack of funds

luke lavan176

cal justifications for the placement of cemeteries on new sites seem to have been largely made post-eventum, as late as the 7th or 8th century. Similarly, at the level of individual monuments, aesthetics and intentionality in the use of spolia can often fall flat when subject to detailed examination.12

Architectural studies

Architectural studies of individual monuments tell us what large-scale topographical work does not: how cities were organized physically on a human scale. The study of period architecture, especially churches, is of course very popular; it forms the late antique branch of what was old-fashioned classical archaeology. Plans and layouts of individual buildings, often obtained through clearance excavation, form the basis for a technological and aesthetic study of architecture. The focus of most architectural studies is to enumerate the details of structural features rather than a functional catalogue. interpretation of these buildings. This work tends to favor studying 'architecture' as an end in itself, rather than studying 'buildings' as a source of economic, social or spatial history.13 Functional labels are given to parts of buildings and to buildings as a whole , for 'churches', 'baths', 'houses' and other structures. In the early imperial period, some architectural historians such as Pierre Gros have done this on a systematic basis and tried to get a clear idea of ​​what the words templum, basilica and curia meant in ancient textual sources.14 For the study of late antique churches, it has been more difficult to develop a detailed idea about the different functions of the structures. Some authors have tried seriously to relate the interior of the church to function. Unfortunately, the absence of early liturgies means that the interpretation of church plans in terms of function is largely guesswork; despite a century of research even some senior researchers, such as

or because of the gift of an aristocratic house that may have previously served as a house church.

12 Krautheimer (1983) 2-5. Gauthier (1999) 198-99. San Lorenzo, now 5. c:Rossignani (1990) 137-38, 138 and 138-39. On burials see Cantino Wataghin (1999)147-80. On spolia see: Ward-Perkins (1999) 225-44 and Leggio and Coates-Stephen in this volume.

13 For examples of this kind of work at its best, see Duval (1995-1998), Cheva-lier (1995) and Michel (2001), which is truly high art. When it is most boring: Lavan (1999)135-67.

14 See Gros (1996), at the beginning of each section.

senantik bytopography 177

Mango and Poulter, believe that this may be an exercise in futility.15 Nevertheless, architectural studies continue to provide us with most of our knowledge of late antique urban topography.

Site syntheses

Another part of research in Late Antiquity that could be described as topographical is the compilation of evidence syntheses for urban areas. Here, Clive Foss' work has dominated and defined the field. Typically, this work proceeds by first providing a textual and epigraphic discussion of a site's political and religious history, then detailing standing monuments by type, or as part of a tour. As such, it is related to the guidebook tradition; Foss's work on Ephesus and Sardis is indeed ideal for this purpose. Again, this work mainly has an architectural focus on buildings and structures, with individual functional labels usually given to structures due to their form. Unfortunately, there is no Antiquity equivalent to the systematic and detailed topographic writing that has been carried out from archaeological evidence in the early Roman period for Pompeii and Herculaneum, although late antique urban landscapes in many areas of the eastern and southern Mediterranean are sufficiently well preserved to make this possible; many have had no significant building since the 6th century and may have statue bases and graffiti still in situ.16

Evoking the Late Antique City: Towards a Human Spatiality?

The three approaches outlined above essentially exist within a conception of space as a grid. Each represents a valid, well-developed, and useful aspect of late antique studies. However, I think this work has also inadvertently encouraged a wider impression

15 Mango (1976) 9-11. Poulter (1994) 249. A. Michel's monograph contains a serious discussion relating archaeological evidence to liturgical use and the overall functions of the complexes; it is careful and good work. Bowden (2001) 56-68 instead wants to explore the social interpretation of churches as monumental architecture; this kind of cross-cultural approach is valid, but most strongly in a Mediterranean context when it is firmly connected to the information of texts: see Lim(1999) 266-81 for a similar study, using entirely written sources.

16 Examples of C. Foss' work: (1976), (1979), (1996), (1998). Pompeii and Herculaneum: Laurence (1994), Wallace-Hadrill (1994), Zanker (1998). For the level of preservation in late antique cities in Asia Minor, see Roueché (1999) or Smith (1999) 170-73 on the location of statue bases.

luke lavan178

of the late antique city as a human historical space, which strongly conditions our perception of urban life in this period, but exceeds the real limits of its methodology. The labels given to buildings, syntheses, guidebooks and architectural commentaries suggest an account of urban topography where structures have single functions, such as for a 'baths', 'church', 'civil basilica'; these labels are given implicitly and their basis is often not properly discussed.16a It is not surprising that from this information one might wish to derive a general or regional understanding of the late antique city as a lived historical space; However, I think this would be intellectually flawed. This article seeks to establish the limitations of existing work in attempting to develop a human understanding of the urban landscape. It also suggests that to move in this direction we need to adopt both a new methodology and a new evidence base.

Limitations of established approaches

It may seem obvious that architectural studies of cities and site syntheses give us a good idea of ​​urban topography as experienced by Late Antiquity. I would like to argue that this is not the case by presenting here a number of limitations in moving towards a human understanding of urban space from existing work. This involves writing about topography as a narrative of human space, as one might write about it in geography, but as an explicitly historical exercise that actively reconstructs the past. This means reconstructing a defuncthuman spatial environment that is dead and gone, rather than describing surviving buildings and structures in modern architectural terms. A number of obvious problems arise in doing this from existing approaches: many important buildings have been neglected or missed; the functions of other structures have been misread or oversimplified; architecturally invisible activities have often been completely excluded from discussion.

Buildings neglected / Buildings missing

Studies of architecture tend to concentrate on new buildings built in Late Antiquity, such as churches. In doing so, they neglect older structures where important traditional activities may still not be

16a For a broad discussion of this problem see Allison (2001) 185-92.

senantik bytopography 179

continue, such as civil basilicas. The study of 'period architecture', such as churches and baptisteries, tends to favor more distinctive buildings. In Late Antiquity, this creates a serious problem because the architecture of many urban elements, including secular public buildings, is less clearly defined than in the first three centuries AD; the difficulty of interpreting the 'principia' at JustinianaPrima or the structure known as the 'xenodocheion' at Teurnia inNoricum from their plan alone are just two examples of this.17 Second, Late Antiquity saw the extensive reuse of pre-existing urban areas— new buildings were often inserted in the elderly; in these circumstances, functional identifications may be difficult or impossible from surviving plans.

In fact, the identification of administrative, caratative and even some religious buildings often seems to depend only on the existence of insitu inscriptions - for example, the tax office in Caesarea Maritima or the synagogue in the house of Kyrios Leontis in Scythopolis, were both identified only by mosaic inscriptions.18 Alternatively identifications can only be made by recovering related artefact assemblages. Thus, Andrew Poulter has been able to interpret an unusual ancient building type found on the Danube as a horreum, through the preservation of the contents of one of these structures, destroyed by fire; prior to his discovery, these buildings had been misinterpreted as grain mills. Contemplation of the excavated architectural remains alone would not have provided a correct identification here. 19

It should also be borne in mind that the importance of colonnades as covered shopping centers has only really been appreciated since the excavation of a set of shops adjacent to one of these colonnades at Sardis, complete with their daily contents, preserved at the sudden collapse of an upper floor, which sealed goods, weighing apparatus, etc., in each of the rooms.20 Needless to say, the continued clearance excavation of Late Antique sites in the Levant, despite rich evidence of destruction containing objects in situ, is catastrophic. All

17 'Principia' in Justiniana Prima: see discussion with bibliography in Lavan(1999) 143-44. Xenodocheion in Teurnia: this identification is not based on inscriptions. It rests on the idea that separate entrances for each of the eight rooms in this structure, which adjoins a church, may indicate a hospice: Glaser (1990).

18 Skattekontoret i Caesarea: Holum (1995). Synagogue i Scythopolis: Zori (1963).

19 Horrea at Dichin, Bulgaria and nearby: Poulter (1999).20 Sardis shops: Stephens Crawford (1990).

luke lavan180

what this procedure does is to bring to light even more sterilized late antique architecture, of which we probably have too much.

Features: Multiple Not simple; Recycling for new purposes

Architectural studies also tend to concentrate on the primary functions of new buildings, disregarding the possibility that these spaces may have had more than one function. It is also true that traditional activities that once took place in classical structures can now take place in the alternative architectural settings provided by new building types, as secondary functions.21 In some cases, the architecture of architecture may even have nothing to do with the use of which it was used for most of its history. This is important in the Balkans, where it appears that there were sometimes inconsistencies between the aspirations of the late Roman state for public buildings and the needs and desires of the communities they sought to impose upon themselves.

St. George complex at Serdica, from the 4th century. or later provides a good example. This appears from the plan to be a bathing building; its main chamber is also heated by a generous hypocaust; However, because the hypocaust was never connected to any flues, it has been postulated that it was intended for drainage and that the structure was originally an imperial audience hall, a function for which the parallels are not strong but perhaps just possible. More likely, the structure was built as a bath but was not completed; instead, it could have served a number of functions, including storage, before eventually being converted into a church.22 This situation is actually not so remarkable for the Balkans; here an effort seems to have been made, through imperial charities and the activity of the army, to maintain urban centers, with public works such as fortifications and especially new foundations such as Diocletianopolis, Gorsium-Herculia or the short-lived Justiniana Prima. Broader economic indicators of the region, such as the nature of rural settlement or even the condition of urban housing, suggest that it did not have the prosperity necessary to sustain these projects without imperial funding.23

21 T. P. Wiseman makes a similar point related to public performances of early imperial Rome: they could occur in a variety of settings. Thus, even in principle, one should not try to limit human activity to the framework provided by well-defined public architecture: Wiseman (1983) 151-55.

22 Serdica St George complex: Lavan (1999) 144-45.23 The Kaiserthermen in Trier provides another example: Here an ambitious

first phase (late 3rd / early 4th century) was never completed. The terms were eventually

you ancient bytopography 181

Architecturally invisible activities

Activities that do not have a distinctive architectural framework are rendered invisible by architectural studies. This is a serious problem for the late antique city. To be sure, street processions of governors, aristocrats, or religious groups did more to articulate social structure than mosaics, colonnades, or elaborate baths in private houses; But only now are researchers beginning to look for the often small epigraphic evidence for processions or street rituals in the late antique urban environment.24

Thus, existing approaches can give us a very fragmentary view of ancient topography in human terms, where much is missing and where some things are misinterpreted. In fact, architectural descriptions with fixed functional labels are not even a complete description of the ancient material world, but produce an aspatial environment that is empty of objects as well as being empty of people.

Alternative spatial methods

Theoretical alternatives from archaeology

There are, of course, a variety of methods available in the archeology of other periods that have been used to reconstruct spatial environments in a more human way. Chief among these is the long tradition of site interpretation present in prehistoric archaeology; here, finds in and around buildings with ground surfaces are often related to use, rather than simply to their abandonment; Quite understandably, excavators seek to reconstruct the activities to which these artefacts and their distribution relate, thereby understanding the functions of buildings and spaces. Where artifact distributions appear very highly structured and relatively undisturbed, excavators have sought to define "activity areas" and even match these dispersions to human activities through careful ethnological observations of similar patterns of debris created in comparable situations today. Like

ally completed to a more modest design: Fontaine (2001) 122-34, with bibliography. Diocletianopolis: Madkarov, (1993). Gorsium: Fitz and Fedak (1993). On the artificial nature of urbanism on the Danube in the 6th century, possibly supported by the state: Poulter (1992).

24 Roueché (1999b).

luke lavan182

topographic mapping, this approach seeks to provide an account of the physical environment, but focuses attention on a human scale. It is explicitly concerned with the historical reconstruction of human action, the function of buildings and artefacts rather than the study of architecture or objects. for their own sake, which is still the dominant approach in late antique studies.25

More recently, archaeologists have sought to develop more complex assessments of ancient space by adapting methods current in geography and elsewhere. These scholars have focused on the past-spatial experience of historical human subjects rather than the material record alone; they aim to produce an 'insider' understanding of how people perceived and behaved in and around monumental structures, rather than simply attempting to describe them in terms of labels for contemporary 'objective' categories. Such work often uses anthropological analogy as an interpretive aid or uses a number of concepts popular in social theory such as phenomenology. Individual and conflicting perceptions and the construction of spatial alternatives are important areas of inquiry. Using ideas of 'social space' enables these different concepts to be related to other contested social structures.26

The theoretical needs of Late Antiquity urbanism

I believe that some, but not all, of this work is useful for late antique studies. Spatial studies are actually so underdeveloped in Late Antiquity that at present there seems no point in developing very complex notions of space. In fact, many approaches that are modern at the moment require a level of information about spatial practice that we do not have and may never have. They are drawn from geography or cultural studies, where direct observation and recording of the world according to the latest theoretical criteria is possible. In contrast, urban historians must spend 90% of their time reconstructing a fragmentary account of the basics of the material world. The reality is that late antique studies still need to develop

25 This is often known as 'spatial archaeology', for the history of which see Gowlett (1997). For state of the art see other papers in the same volume. The adaptation of artefact dispersions to the waste patterns observed during ethnographic observations of activity areas is part of 'intermediate area theory': Binford (1983) esp. 95-192.

26 Many of these works are discussed in Laurence (1997) 7 n.1, to which may be added Johnson (1996) and (1993). For an exploration of individual perceptions of Roman monumental architecture see Revell (2000).

you ancient bytopography 183

a basic knowledge of the everyday use of some of its most common spaces.

In this situation, I would suggest that it is currently best to adopt a modified version of an approach generated by archaeological fieldwork itself; competent excavators are well used to dealing with human use of space, although many show little interest in exploring these issues away from the confines of their own sites. I suggest adopting a modified version of the prehistoric interpretation of the site. But to fit the rich but complex urban data we have from Late Antiquity, this approach needs to change. First, it must be extended to incorporate texts of some kind that are recognized by all but the most repentant archaeologist to be indispensable to understanding life in Roman cities.27 Second, it must be broad enough to not only to take into account the functions of architectural structures, but to include the many important activities that people living in a Mediterranean environment quite naturally carry out away from buildings, in open spaces. Finally, I also believe that it should be used to inform a general topographical approach, rather than simply a basis for considering the features of individual sites. Rather than acting as a means of interpreting the archaeological record, it would become the main focus of interest. This amounts to a reversal of prehistoric priorities - reflecting the reality that it is disciplined interpretation and not the data that really matters in Mediterranean classical archaeology.28

A possible access

Here I want to propose a framework that addresses these concerns to some extent. It is not offered as an exclusionary approach

27 On the necessity of including textual sources in our accounts of the city as well as archeology and conflict in this debate, see the review of positions in Laurence(1997) 9-11 and articles in the same volume and comments in Laurence (1994) 8 .See also Allison (1997a) and (1997b).

28 It is typical of historical and especially Mediterranean Roman archeology that careful textually delimited interpretation is the main focus of interest notdata - who in Mediterranean science wants to hear about alternative cultural identities, possible religious interpretations, elites or top-ranked settlement centers -tres: we want to hear about being Roman/Germanic, Christianity, curiales and cities. The idea that archeology can find a real competitive interpretive alternative to this for the Mediterranean world is unfounded. This is not to say that data does not matter: rather, it is to emphasize that the quality of data in relation to the questions asked is crucial: the interest in data is not absolute when we have so much of it.

luke lavan184

others – just one approach that will have its limitations. I'll be happy if it just sparks discussion. But it is time that new methodologies of space were applied to the late antique city, which allows us to explore it better as a historical cityscape, full of people, rather than as a lifeless ruin.

My approach is based on traditional site interpretation, but differs from it in four ways:

First, I will group 'areas of activity' into larger units, 'spaces of activity', such as horrea or praetoria (buildings) or (non-architectural) law courtor adventus. They would consist of two elements: i) the spatial functions (the activities themselves/what people are doing) ii) their material surroundings (from the bodies of the people present and their clothing to any buildings involved)

Secondly, I will use the term 'activity space' much more generally than 'activity areas'. They would serve not only as conceptual tools for interpreting sites, but would be the basis for general descriptions of human space, components of a historical narrative of an ancient topography.28a

Third, I would change the source base for the reconstruction of human activities within these spaces: artefact collections would be fine, supplemented by in-situ inscriptions, but instead of ethnographic analogy I would use descriptions of activities retold by contemporary texts.

Fourth, as far as possible, I will base ideas about the organization of space on those described in contemporary textual sources.29

This methodology represents little more than an attempt to import into our discipline an explicitly historical study of human spatiality common elsewhere in archaeology. It establishes human action as its focus, but retains the traditional topographical goal of providing a coherent narrative description of the spatial world for its own sake. It is distinctive in its desire to establish a general framework for action.

28a Thus, activity spaces such as 'praetorium', 'law court', 'curia' and 'adventus' within an overall topographical narrative would occupy a similar position to the simple functional labels given to buildings in guidebook narratives; here, however, the topographical narrative I propose seeks to describe a synthetically reconstructed city, rather than simply being a commentary on one place.

29 It is easy to find terms in the texts for categories dealing with religious, political and some aspects of social life; Ancient writers defined much of their own conceptions and practices for these activities. For economic activities and broad spatial concepts used to study 'urbanism', which we are somewhat more interested in than the old ones, separate spatial categories can be used where necessary - e.g. 'garbage-bish tip', 'pottery workshop'.

you ancient bytopography 185

counts of late antique urban topography.29a There has been curiously little discussion of this, especially among archaeologists, despite the application of many theories of space to early imperial monument types and sites. On the other hand, historians, such as recently Richard Lim on Late Antiquity Rome or Christopher Haas on Alexandria, seem more interested in writing about topography in a general way.30 It is particularly important to work on a general level in Late Antiquity on due to the nature of the city certificate. Although the preservation of the site for the end of antiquity is good, we have no 4th or 5th century. Pompeii; archaeological information is often fragmentary, and literary sources, although arguably richer than for the early empire, are widely scattered. Our detailed knowledge of the city in this period must come from carefully integrating sources, from a wider geographical area, in a general but regionally sensitive narrative: there is no alternative.

By explicitly using texts, this approach goes some way to providing an 'insider' perspective on the ancient space; however, the use of texts as a basis for the monument function is nothing new. Good architectural studies have always appreciated the need to use texts.30a

Indeed, scholars already use texts as the source of most functional identifications of buildings—in labeling structures such as basilicae, curiae, or agorae. Nevertheless, it sometimes escapes us that archaeologists do not actually discover basilicae, curiae, or praetoria, but nameless, empty monumental structures.31 Indeed, we conceptually depend on texts, epigraphy, and analogies from them to provide functional labels for most political , social and religious buildings.32

29a This approach is close to, but differs from, the study of place, which is more concerned with individual localities, rather than general classes of human space.

30 Laurence's (1994) monograph on Pompeii reflects this. It is unusually rich in the many approaches it employs; however, it is not always clear how the frameworks of different chapters can be related to each other. Zanker e.g. (1998) write in a systematic way about 'urban landscapes', albeit on the remains of individual cities, rather than for general classifications of human space. Lim (1999); Haas (1997). Writing about sacred topography as a coherent historical narrative is now fashionable among historians - see for example the section on topoi in Markus (1990).

30a Simplistic use of architectural labels based on texts has led to errors, as see Allison (2001) 185–94. Allison rightly displays a strong degree of skepticism towards this practice in relation to domestic architecture and shop identification; however, it is important not to overdo this skepticism; Label identifications for public buildings such as basilicas (in the west), curiae, forums and thermae are generally much more secure, being based on strong textual references and inscriptions.

31 Usually the findings relate to abandonment; stone floors are swept in use.32 See Balty (1991), who shows that identification of city council chambers

luke lavan186

However, my approach tries to make the basis for identification through these sources rigorous and explicit. Where we can identify unusual building types through mosaic inscriptions or well-founded textual identifications, it is reasonable to accept them as a solid core of good material, as a basis for making secondary analogies elsewhere. When we do not have a firm basis for discussing building functions on the individual plots, we will do best not to give any interpretations of this kind at all and let silence prevail. Lack of rigor in identifying the functions of buildings has caused great confusion in the historical debate in the past, especially when identifications based only on loose analogy, without clear textual or epigraphic foundations, have entered the wider literature. In particular, the character of imperial palaces and the praetoria of civilian governors has been obscured by localized, sometimes sensational reasoning.33

Potential objections

Texts and Archaeology: Incompatible Sources?

There are currently many examples of scholars attempting to combine texts and archeology in studies of the ancient city. The work of Andrew Wallace-Hadrill, Ray Laurence and others on the urban environment of Pompeii and Herculaneum has sought to understand the use of a wide range of urban spaces through the spatial concepts and practices revealed in textual evidence. Simon Ellis has done much the same in his studies of late antique dwellings.34

Interestingly, Ray Laurence has expressed concern that he has produced "an academic heresy", at least in part due to his combination of textual and archaeological evidence.35 This was a valued methodological taboo for an earlier generation of archaeologists who regularly were presented with anachronistic interpretations of archaeological levels in terms of familiar historical events, rather than more mundane processes of site formation. However, I do not at all suggest that you try to create texts, epig-

in the West is largely made possible by a coin identification of the curiasenatus in Rome and inscriptions from the curia of Timgad: 12-15, 73.

33 On praetoria see Lavan (1999) and (2001a). On imperial palaces see Duval (1992). In the case of praetoria, some correlation of function and architecture occurs when identifications are limited to those based on inscriptions.

34 Wallace-Hadrill (1994); Laurence (1994); Laurence (1997); Ellis (1988).35 Laurence (1994) ix.

you ancient bytopography 187

raphy and archeology artificially agree, despite their separate methods. Rather, I want to argue that urban topography, especially the study of political, social and religious spaces, is a legitimate meeting place for all these sources.36

It must be emphasized, to the chagrin of some field archaeologists, that archeology is often not a superior source to textual evidence for topographical reconstruction - it is just one source among many that should not be unduly privileged. As Laurence has noted, both are equally a cultural product of the Roman world; the relationship between archaeological remains and past behavior is not straightforward, as with the texts.37 At an ancient site we find only ruins, not ancient topography, despite their seductive materiality. Much of the ancient city is missing in their testimony, especially in relation to the construction of political and social activities; these leave little unambiguous artifactual evidence that can help us interpret cellular buildings or general hall structures that housed them. Ruins at ancient sites represent partial and fragmentary data that will never produce a narrative to replace the spatial world described by Ammianus and Libanius.

In practice, as the works of Balty, Laurence, Ellis and others show, texts and archeology can often be related to precisely the same questions of function, location, etc., which neither source can fully answer on its own. It is important to point out that when I argue for the integration of textual sources, I am not talking about using rhetorical descriptions of cities such as can be found in Menander Rhetoror Procopius's de aedificiis, although these are important on a symbolic level; I am talking about using the countless anecdotes, legal references and papyri that recall everyday activities and their surroundings in Mediterranean cities.38 The integration of sources on a practical level

36 This was recognized by Krautheimer (1983) 1-5.37 Laurence (1994) 8. The functions of some urban elements, such as

tions and places of production are much more easily identified by analogy: this is not something I wish to dispute. Archeology is also close to the only source we have for economic topography.

38 I am talking here about the Mediterranean. The integration of archeology in Northern Europe with a small number of texts is another question. Nevertheless, it should not be assumed that the function of buildings and cities in the 4th century was less 'Roman' here than in the Mediterranean; this should be demonstrated or silence should prevail. For the different types of site descriptions contained in Procopius's de aedificiis and their responsible use by archaeologists, see Cameron; some of these are difficult, some more straightforward: Cameron (2000)179.

luke lavan188

should be very aware of their different value for different kinds of questions. For example, the location of the praetorium in the cities, the functions of their courtyards, and the architectural significance of the term praetorium are very different issues for which different types of evidence have different values. Some questions will always be better answered by archaeology, others will only ever be answered by texts.39 This integration of texts and archeology is best done at the level of general or regional synthetic discussions before being critically applied to individual sites. It is at the former level that a meaningful historical discussion lies, where there is literary evidence in sufficient quantities to enable the combination with archaeology.

Connected spaces

Another major objection to writing about "spatial functions plus material surroundings" is that such an approach disrupts a coherent appreciation of space, because human activities overlap spatially while architecture does not. However, this problem is largely avoided in practice - most activity spaces in the ancient city were spatially separated and occupied different building types. On the other hand, a study of activity spaces rather than buildings ensures that spaces that have no fixed architectural framework, such as those that take place in the street, will not be missed; nor will multiple functions in a single building be ignored.

Relativism

Perhaps the most serious are objections derived from broader theoretical considerations. Our postmodern experience certainly teaches us that a single account of human activities is impossible because of the conflicting perceptions of different people involved; an attempt to describe a single account of events may even add up to a positivist assumption of thought that attempts to suppress the thoughts of others. This chimes with a view popular among some archaeologists that texts 'merely' represent the distorted views of an elite, rather than reflecting a largely consensual wider consciousness. Such skepticism is valid for the early imperial period, in the newly conquered

39 See Alston (1997) 25-40 for houses based on Egyptian papyri. There are some aspects of his reconstruction that archeology could never reveal.

you ancient bytopography 189

provinces, but is not so rewarding in the Mediterranean, from the 3rd century. forth.40

In my own research, I have not found perceptual pluralism particularly useful in relation to late antique urban topography; I have tried hard to find perceptual conflicts about space in the late antique city. It certainly existed over religious affairs and over the relative importance of religious to secular topography, as Richard Lim, Chris-topher Haas and Annabel Wharton have demonstrated; multiple view-points, if not conflict, existed over some other conceptual issues, such as the meaning of spolia, as articles in this volume demonstrate. However, there seems generally to have been little consensus about the nature of the day-to-day functions of political and social spaces; alternative voices do not seem to be traceable.41 Janet Lindblom has reached a somewhat similar conclusion in her work on women and social space in early Byzantium; differences are recorded between some behavior as recorded in anecdotes and codes of behavior for women as recommended in texts; but male and female authors of these works show no differences between them, either in their conception of how women should speak in public or in their behavioral anecdotes.42

It is important to emphasize that Late Antiquity was not post-Reformation or post-modern Northern Europe, where deep perceptual conflicts have been based on increasingly exaggerated identities. In Late Antiquity, considerable consensus means that a single functional account of urban topography, based on ancient texts, does indeed seem possible—with the caveat that such an approach should be based firmly on our experience of the data rather than on any particular philosophical position.

40 For the Early Imperial period see Webster and Cooper (1996).41 Lim (1999); Haas (1997); Wharton (1995); see also Maier (1995). I have not

found it possible to generate alternative views from the texts about the nature of the adventus, forums, council chambers, praetoria, courts, prisons, etc. The hostile views of the clergy about entertainment buildings should be treated as quite unrepresentative, since they continued to attract crowds without problems: Segal (1985-89) 158-159; Ward-Perkins (1984) 92-118.

42 Lindblom (2001) 175, plus discussion on the paper day.

luke lavan190

Conclusions

The approach outlined in this article represents only one response to the difficulties of writing about human space in the late antique city. As suggested earlier, it is likely that political and social spaces have the most to gain from it; in my second paper in this volume, I consider the application of 'spaces of activity' to the political topography of the ancient ancient city, which I argue could in some ways change our appreciation of urban life in this period. The topography of religious conflict and dissent probably needs a different approach with a more flexible methodology. An 'economic topography' can use very little textual information. However, the construction of systematic topographical accounts of human use of space can be more than an end in itself. It could have wider benefits.

First, it could provide a good intellectual context for justifying modern archaeological records to the wider historical community. rather, it is intended to highlight an interpretive gap between urban excavation results and historically meaningful research. Many scavengers working on Roman cities in the eastern Mediterranean still simply do not see the point of using modern recording techniques to solve what they consider to be interesting historical questions. They have long consoled themselves that, while they were not archeological wizards, they were at least good historians of topography who knew the wood from the trees, especially with regard to public buildings. However, if high quality archaeological fieldwork could be undertaken to contribute more strongly to a general level of topographical writing there might be more pressure for change; it must be demonstrated, in a historically meaningful arena, that the recording of architecture as built upon which excavators build is less important to understanding the functions of public buildings than recording repairs, minor epigraphic evidence and artefact assemblages.43

Conversely, a general understanding of 'activity spaces' within the ancient city of Telaten could also provide a basis for improving the excavation

43 I do not wish to provide a bibliography of 'bad' excavations at this point. However, for a study of the impact of modern logging on a traditional clearing excavation see Ratté (2001).

senantik bytopography 191

strategies. A decade of targeted inquiry-based work could provide a better understanding of the late antique city than a century of uncovering prominent public monuments. This would involve reversing the current relationship between field archeology and synthetic history. Over the past thirty years, archeology has heroically rewritten the history of the Late Antique economy, land settlement, and the broad condition of cities, largely by producing new data that act as "wrecking bullets," as Mark Whittow puts it. Discoveries in the area have shown unexpected economic and demographic vitality in many urban centers during this period. For the public space of monumental Roman cities, archeology is now well past the stage of discovery. Most progress will now come from reconsidering the evidence we have, whether archaeological, textual or epigraphic, and asking better informed questions in the field.

Recognitions

I thank Penelope Allison, William Bowden, Simon Ellis, Helen Gittos, Carlos Machado, Gregoire Poccardi, Jean-Pierre Sodini and BryanWard-Perkins for reading this text. They are, of course, exempt from the opinions expressed in it, from which they differ; errors remain my own. I am also grateful to the Leverhulme Trust for funding my stay in Paris during the academic year 2001-2002 during which these two articles were written.

luke lavan192

Bibliography

Alston R. (1997) "Houses and households in Roman Egypt", i Domestic Space inthe Roman World: Pompeii and Beyond (JRA Supplementary Series 22) udg. R. Laurence og A. Wallace-Hadrill (Portsmouth, Rhode Island 1997) 25-40.

Balty J. Ch. (1991) Curia Ordinis. Research into Ancient Architecture and Town Planning in the Provincial Curia of the Roman World (Brussels 1991).

Bauer F. A. (1996) City, place and monument in Late Antiquity. Studies of the layout of public space in the late antique cities of Rome, Constantinople and Ephesus (Mainz 1996).

Binford L.R. (1983) In Pursuit of the Past. Decoding the Archaeological Record (London1983).

Bowden W. (2001) “A New Urban Elite? Church builders and church building inlate ancient Epirus”, in Recent Research in Late Antique Urbanism (JRA Supplementary Series 42) ed. L. Lavan (Portsmouth, Rhode Island 2001) 56-68

Brogiolo G. P. og Gelichi J. (1997) Byen i den tidlige italienske middelalder: arkæologi og historie (Rom 1998).

Bruhl C.-R. (1990) Palace and City, Studies in the Profantopography of Late Antique Cities I, Gallien (Köln og Wien 1975).

——— (1990) Palatium and Civitas, Studies on the Profane Topography of Late Antique Civitates IIBelgica I, both Germania and Raetia II (Cologne 1990).

Cameron A. (2000) "Conclusion", AntTard 8 (2000) de aedificiis: teksten til Procopius og dokumentariske realityter", 177-80.

Canadian Carthage Team (1985) "Excavations in the Canadian section," CEDACCarthage 6 (1985) 21.

Cantino Wataghin G. (1999) "The Ideology of urban burials" i The Idea and Idealof the Town between Late Antiquity and the Early Middelalder edd. G. P. Brogiolo og B. Ward-Perkins (The Transformation of the Roman World 4) (Leiden, Boston og Köln 1999) 147-80.

Chevalier P. (1995) Salona II : Ecclesiae Dalmatiae (Rom 1995). Curry M. (1996) "On space and spatial practice in contemporary geography" i

C. Earle, K. Mathews og M. S. Kenzer (red.), Concepts in Human Geography (Lanham, Maryland 1996) 3-32.

Duval N. (1992) "The Palace of Milan among the Imperial Residences of the Lower Empire", in Milan, Capital of the Roman Empire, (Milan 8-11 March 1990) edd. G. SenaChiesa and E. A. Arslan (Milan 1992) 138-46.

——— (1995-98) ed. The first Christian monuments in France. I-III (Paris 1995-98).

Ellis S. P. (1997) "Late Antiquity dining: architecture, furniture and behaviour", in Domestic Space in the Roman World: Pompeii and Beyond (JRA SupplementarySeries 22) edd. R. Laurence and A. Wallace-Hadrill (Portsmouth, Rhode Island 1997) 41-52.

——— (1988) “The end of the Roman house,” AJA 92 (1988) 565–76. Fentress E. (2000) ed. Romanization and the city: creation, transformations and failures

(JRA Supplementary Series 38) (Portsmouth, Rhode Island 2000) February P.-A. (1965) Excavations of Sétif. The Christian basilicas of the northwest quarter (Paris

1965). Fitz J. and Fedak J. (1993) "From Roman Gorsium to Late Antique Herculia: a

summary of recent work at Tác (NE Pannonia)", JRA 6 (1993) 271-73. Foerster G. and Tsafrir Y. (1987-88) "The Bet Shean project: center of ancient

Betshean-nord,” Excavations and Surveys in Israel 6 (1987-88) 24-43.Foss C. (1976) Byzantine and Turkish Sardis (Cambridge, Mass 1976).

senantik bytopography 193

——— (1979) Ephesus after Antiquity: A Late Roman, Byzantine and Turkish City (Cam-bridge 1979).

——— (1996) "The cities of Byzantine Pamphylia in the Byzantine age" in Cities, fortresses and villages in Byzantine Asia Minor, C. Foss (Aldershot, England 1996) article no.4.

——— (1998) “Syria in Transition A.D. 550–750: An Archaeological Approach,” DOP 52 (1998) 189–269.

Gauthier N. and Duval N. et al. ed. (1986 f.), Christian Topography of the Cities of Gaul from the Origin to the Mid-8th Century, several volumes (Paris 1986 f.).

Gauthier N. (1999) "La topographie chrétienne entre idéologie et pragmatisme", in The idea and ideal of the city between Late Antiquity and the Early Middle Ages. G. P. Brogiolo and B. Ward-Perkins (The Transformation of the Roman World 4) (Leiden, Boston and Cologne 1999) 195-210.

Glaser F. (1990) "The episcopal building complex in Teurnia: a Xenodocheion", Carinthia I 180 (1990) 89-94.

Gowlett J. A. J. (1997) "High definition archaeology: threads through the past", World Archaeology 29.2 (1997) High Definition Archaeology 151-71.

Gros P. (1996) Roman Architecture from the Early 3rd Century BC. J-C. at the end of the High Empire vol.1 Public monuments (Paris 1996).

Haas C. (1997) Alexandria in Late Antiquity. Topography and Social Conflict (Baltimore and London 1997).

Holum K. G. (1995) "Inscriptions from the Imperial Tax Office in Byzantine Caesarea. Palaestinae", in Roman and Byzantine Near East: Some Recent Archaeo-logical Research, ed. J.H. Humphrey (JRA Supplementary Series 14) (Ann Arbor, Michigan 1995) 333-46.

Johnson M. (1993) Housing Culture: Traditional Architecture in an English Countryside (Washington and London 1993).

——— (1996) An Archaeology of Capitalism (Oxford 1996).Krautheimer R. (1983) Three Christian Capitals: Topography and Politics (London 1983).Laurence R. (1994) Roman Pompeii: Space and Society (London 1994). ——— (1997) "Space and Text", i Domestic Space in the Roman World: Pompeji og

Beyond, red. R. Laurence og A. Wallace-Hadrill (JRA Supplementary Series22) (Portsmouth, Rhode Island 1997) 7-14.

Laurence R. og Wallace-Hadrill A. (1997) (red.) Domestic Space in the Roman World: Pompeii and Beyond (JRA Supplementary Series 22) (Portsmouth, Rhode Island1997) 7-14.

Lavan L. (1999) "The Residences of Late Antique Governors: A Gazetteer", AntTard7 (1999) 135-67.

——— (2001) "The praetoria of civil guvernors in late antiquity", i Recent Re-search in Late Antique Urbanism, red. L. Lavan (JRA Supplerende bind 42) (Portsmouth, Rhode Island 2001) 39-56.

Lim R. (1999) "People as power: games, munificence and contested topography", i The Transformations of Urbs Roma in Late Antiquity ed. W. V. Harris (JRA Supplementary Series 33) (Portsmouth, Rhode Island 1999) 266-81.

Lindblom J. (2001) "Women in social space in 6th century Byzantium", i XX. International Congress of Byzantine Studies pre-acts III. Gratis kommunikation, red. V. Déroche et al. (Paris 2001) 175.

Macdonald W. L. (1996) The Architecture of the Roman Empire II: An Urban Appraisal (New Haven 1996).

Madkarov K. (1993) Diocletianopol Tom. 1. (Sofia 1993). Maier H. O. (1995) "The topography of heresy and dissent in late-fourth century

Rom”, History 44.2 (1995) 231-49.

luke lavan194

Mango C. (1976) Byzantine Architecture (New York 1976).——— (1981-82) "Daily Life in Early Byzantium," in XVI Internationaler

Byzantinist Congress Vienna 4.-9. October 1981 filer (Yearbook of Austrian Byzantine Studies 31-32) (Vienna 1981-82) 337-53.

——— (1990) The Urban Development of Constantinople (IV-VII Centuries) (2nd ed.) (Paris1990).

Markus R. (1990) The End of Ancient Christianity (Cambridge 1990).Michel A. (2001) The Byzantine and Umayyad Churches in Jordan (the Provinces of Arabia)

and Palestine). 5.-8. century. Architectural typology and liturgical arrangements (Senantikbibliotek 2) (Turnhout 2001).

Müller-Wiener W. (1977) Bildlexikon zur Topographie Istanbuls (Tübingen 1977). Owens E. J. (1989) The city in the Greek and Roman world (London 1989) Petit P. (1955) Libanius et la vie Municipale à Antioche au IVe century AD J.-C. (Paris

1955). Poulter A. (1994) "Churches in space: the early Byzantine city of Nicopolis", i

Churches Built in Ancient Times, Painter K. (ed.) (London 1994) 249-68.——— (1992) “The Uses and Abuses of Urbanism in the Danube Provinces of

senere Romerriget”, i The City in Late Antiquity (Leicester—Nottingham Studiesin Ancient Society bind 3). udg. J. Rich (London 1992) 136-44.

——— (1999) “Gradishte near Dichin: a new late Roman fortress on the Lower Danube”, in The Limes on the Lower Danube from Diocletian to Heraclius, ed. G. von Bulow and A. Milcheva (Sofia 1999) 207-27.

Ratté C. (2001) "New Research on the Urban Development of Aphrodisias in Late Antiquity", in Urbanism in Western Asia Minor. New studies of Aphrodisias, Ephesos, Hierapolis, Pergamon, Perge and Xanthos (JRA Supplementary Volume 45) ed. D.Parrish (Portsmouth, Rhode Island 2001)

Revell L. (2000) "Skabelsen af ​​flere identiteter i Roman Italica", i TRAC99Proceedings of the Ninth Annual Theoretical Archaeology Conference, red. G. Fincham, G. Harrison, R. Rodgers Holland og L. Revell (Oxford 2000) 1-7.

Rossignani M. P. (1990) "The Basilica of S. Lorenzo", "The colonnade" and "The Foundations and the Roman amphitheatre", in Milan, capital of the Roman Empire, 286-402 AD, Catalog of the exhibition (Milan-Palazzo Reale , 24 January/22 April 1990), ed. G. Sena Chiesa and E. A. Arslan (Milan 1990) 137-38, 138 and 138-39.

Roueché C. (1999) "Looking for late antique ceremonial: Ephesos and Aphrodisias", in 100 years of Austrian research at Ephesus. Files from the symposium Vienna 1995, (archaeological research 1. DenkschrWien 260) ed. H. Friesinger and F. Krinzinger (Vienna 1999) 161-68.

Segal A. "Theatres in Ancient Palestine during the Roman-Byzantine Period (Anhistorical-Archaeological Survey)", Scripta Classica Israelica 8-9 (1985-89) 158-59.

Simonsen K. (1996) "What kind of space in what kind of social theory?", Progressin Human Geography, 20.4 (1996) 494-512.

Smith R. R. R. (1999) "Late Antique Portraiture in a Public Context: Honorary Statutory at Aphrodisias in Caria, A.D. 300-600," JRS 89 (1999) 155-89.

Stephens Crawford J. (1990) The Byzantine Shops at Sardis (Archaeological Exploration of Sardis. Monograph 9) (Cambridge, Mass. and London 1990).

Tsafrir Y. and Foerster, G. (1997) "Urbanism at Scythopolis-Bet Shean in the Fourth to Seventh Centuries", DOP 51 (1997) 85-146.

Wallace-Hadrill A. (1994) Houses and Society at Pompeii and Herculaneum (London 1994). Ward-Perkins B. (1984) From Classical Antiquity to the Early Middle Ages. City public

Building in Northern and Central Italy AD 300-850 (Oxford 1984).——— (1999) “Reusing the Architectural Heritage of the Past, entre idéologie et

senantik bytopography 195

pragmatism,” in The idea and ideal of the city between Late Antiquity and the Early Middle Ages edd. G. P. Brogiolo and B. Ward-Perkins (The Transformation of the Roman World 4) (Leiden, Boston and Cologne 1999) 225-44.

Ward-Perkins J. B. (1974) Cities in Ancient Greece and Italy: Planning in Classical Antiquity (New York 1974).

Webster J. and Cooper N. eds. Roman Imperialism: Postcolonial Perspectives (LeicesterArchaeology Monographs 3) (Leicester 1996).

Wharton A. J. (1995) Refiguring the Post-Classical City: Dura Europos, Jerash, Jerusalem og Ravenna (Cambridge 1995).

Wiseman T. P. (1983) "Efterord: borgerlivets teater", i I. M. Barton (red.)Roman Public Buildings (Exeter Studies in History 20) (Exeter 1983) 151-55.

Zanini E. (1994) Introduction to Byzantine Archeology (Rome 1994) Zanker P. (1998) Pompeii: Public and Private Life (Revealing Antiquity 11) (Cambridge,

dimensions and London 1998) trans. DL Schneider, originally published in German as Pompeii: Cityscape and Taste of Living (Mainz 1995).

Zori N. (1963) "The house of Kyrios Leontis at Beth Shean," IEJ 13 (1963) 123-34.

enrico zanini196

THE URBAN IDEAL AND URBAN PLANNING INBYZANTINE NEW CITIES IN THE 6TH CENTURY AD

ENRICO ZANINI

Abstract

Archaeological evidence and historical texts can be combined to discern some of the physical characteristics of Byzantine cities in the 6th century. A.D. They can also be used to define a common conceptual model, like Byzantine people in the 6th century. remembered when they uttered the word polis, either when defining an existing settlement or when planning a new city. This paper deals in particular with four cities that were rebuilt or completely rebuilt during this time. The survival of the idea of ​​the city in the 6th century. The East also seems to be important for understanding the preservation and renewal of this concept in the Western Mediterranean, after the dark centuries of the early Middle Ages.

Introduction

The development of the ideal of the city between the 5th and 7th centuries is a critical issue for the transformation of the city from Late Antiquity to the Early Middle Ages.1 Much of the debate about the fate of urban centers in late antiquity revolves around the question of what connoted a 'city' in Late Antiquity and in the first centuries of the Middle Ages. Another major interest is how the 'conceptual image' of the city during this period differed between the different regions of the pre-Roman world.

There is, I think, a consensus that the 6th century was decisive for the deepening of regional divisions. In the West and in Central Europe, this century saw probably the greatest political, economic and physical crises affecting the city. In the Mediterranean regions and in

1 See the proceedings of the international seminar The Idea and Ideal of Townbetween Late Antiquity and the Early Middle Ages, coordinated by G. P. Brogiolo and B. Ward-Perkins in the European Science Foundation project The Transformation of the Roman World.

L. Lavan og W. Bowden (red.) Theory and Practice in Late Antique Archaeology(Late Antique Archaeology 1 2003) (Leiden), s. 196-223

urban ideal and urban planning in Byzantine cities 197

Especially in those countries which had been longer under the Byzantines, the city generally held its position better in every respect. But within this picture, regional and sub-regional variations were of course present. For example, the crisis in Britain was more severe for some cities than for others;2 the continuity of cities in Byzantine Italy3 compares poorly with the flourishing of cities in ancient Syria. G. Dragon has described the Byzantine Empire in the 6th century. as still a "mosaic of cities"; in the cities of the eastern Mediterranean this observation seems to be confirmed and even reinforced by archaeological investigations. These reveal that the Byzantine city at the end of the 6th and the beginning of the 7th century was still in a very healthy state and was able to survive disasters of both natural and human origin.4 And this sees appear to be true for both central and peripheral areas.5

The Urban Ideal in the Sixth Century: Texts and

Archaeology

The continuity seen in many eastern cities raises a further question. Was life in these centers merely a survival of sites, buildings, infrastructure, and administrative and economic models inherited from antiquity? Did it merely represent an exotic survival of classical urbanism against a broader negative pattern of change occurring more rapidly and dramatically in the West? Or did the classical ideal of the city—with its bond of civitas and civilitas linking the city to urban civilization—continue to be a critical and integral part of the culture of people living in Byzantine territories?

It goes without saying that there is no single answer to such a complex question. Nevertheless, I believe that a significant contribution can be made by comparing written sources and archaeological data. This involves an experiment in 'cognitive archaeology' to try to understand the ideal of the city shared by Byzantines in the 6th century. In the terminology of linguistics, the question revolves around

2 An important example of continuity in urban life, albeit in profoundly altered forms and spaces, is found at the site of Wroxeter: finally Barker (1997); but see Ward-Perkins' Critical Reflections (1996) 9-13.

3 Zanini (1998) 105-208; Brogiolo (1999).4 Antioch and Apamea are instructive cases here, since an impressive array of

natural disasters and deadly events failed to put an end to their lives; see Balty (1986); Condolences (2000).

5 Whittow (1990) 13-20; Walmsley (1996); Tsafrir-Foerster (1997).

enrico zanini198

relationship between 'signifier' and 'signified'; in the linguistic code of the time - the language, as Ferdinand de Saussure would say - was the word 'city' associated with a specific, well-defined image, or was it rather just a remnant of past experiences with a definition of convention, lacking correspondence to the reality of the time? In other words, when a 6th century Byzantine uttered the word polis – or civitas, if he was educated enough to know Latin – a clear and commonly shared image formed in the minds of his listeners, or was it rather an indistinct one image that everyone saw differently?

It is relatively easy to consult the written sources, but there is a great risk of misunderstandings. If one collates references derived from the historical and geographical texts of the time, one finds that the term 'city' is used for over a thousand different places, often very different in size, structure, economic importance, administrative status, etc. In this respect, the aedificiis of Procopius of Caesarea is probably the most important source for the 6th century, despite the oft-discussed limitations of its use.6 The places where Procopius uses the word polis are, of course, countless; it is therefore unlikely that the word has a single unequivocal meaning throughout the text.7

Nevertheless, one can come fairly close to a model of Procopius' urban ideal by considering his description of the restoration of Helenopolis in Bithynia.8 In this passage, Justinian makes several reconstructions, which Procopius lists in a certain order, presumably according to their recognized priority. The first monument that was erected as a rebuild was an aqueduct that was supposed to supply houses, but also – and perhaps above all – public baths. In fact, the Constantinian baths were rebuilt and another complex was built anew. Then some new churches were built, of which we do not know the exact number; these were then followed by a palace, porticoes, and a residence of magistrates, until the place assumed, in the words of Proco-pius, 'the appearance of a prosperous city.'

It is evidently not easy to determine how much of Procopius's tale is reality and how much is exaggeration to celebrate the ruin of his sovereign. This element often appears in the aedificiis and

6 On the complex problems associated with the use of this text, in many ways unusual, see finally Roques (2000), who collects and synthesizes the previous bibliography, and other authors in the same volume.

7 It would be interesting to see the current interest in the text generate a study of this specific aspect.

8 Procop. Aed., 5.2.1-5.

urban ideal and urban planning in Byzantine cities 199

gives the work the character of a panegyric. It is equally difficult to say whether this report by Procopius actually describes a real building campaign carried out by Justinian, or by Anastasius or Justin I, or whether it should rather be considered one of the narrative topoi often found in this work . 9 Nevertheless, the frequency with which one finds elsewhere in the work references to aqueducts, cisterns, baths, public buildings, churches and residences of public authorities leads one to believe that these elements - along with fortification walls, of course - were , characterized a polis in the collective ideal or, better, in the conceptual framework, from the 6th century Byzantines.

This impression becomes even stronger if we look at two of Justinian's short stories. These texts are of an entirely different character, but are probably not entirely independent of the aedificiis. They contain descriptions of the important places of a city in terms almost identical to those used by Procopius. By defining tasks for the praetors in Lycaonia10 and Thrace11, the emperor obliged them to restore and preserve public monuments. First the names became aqueducts, and then roads, bridges, fortifications, harbors, etc. The similarity between these legal texts and Procopiusis is not surprising, since the historian seems to have drawn on material coming directly from the imperial chancellery. However, we can accept that the above-mentioned elements continued to characterize the urban concept expressed by the ruling class in the 6th c.12.

Confirmation of this view is available from the epigraphy. In a catalog of inscriptions referring to buildings built by Justinian, recently produced by D. Feissel,13 at least one inscription almost repeats the word, formula used in the Novellae and in Procopius' treatise.14

9 It is easy to note that the elements indicated by Procopius largely coincide, with the obvious exception of churches, with those considered by Pausanias to be the essential characteristics of a settlement that is fully entitled to be called a city.

10 Nov., 25.4.11. Nov., 26,412 The same picture can be drawn from the urban building regulations in

time; for example, mentions Nov. 25.4 the necessity to demolish the private interventions built in areas set aside for certain public functions. Theses on architecture and urbanism give the same impression; the so-called "Treatise on Urbanism" of Julian of Ascalon lays down the technical specifications of laws which we may suppose to have been much more numerous and much more detailed than those we know from written tradition; see Saliou (1994) and (1996).

13 Feissel (2000) 91, no. 15.14 Here it is worth mentioning a short inscription from Terracina, Italy, which,

in the second half or even at the end of the 7th century, expressly referring to maintenance

enrico zanini200

The analysis of these texts makes it possible to draw some conclusions about the ideal of the city and its physical reality at this time. We can see, as would be expected, a mixture of traditional and innovative European characteristics. Roads lined with porticoes, aqueducts, baths and administrative buildings were traditional elements of antiquity that continued to exist in cities of Justinian's empire. Other traditional elements, such as entertainment buildings, have disappeared. On the other hand, the urban landscape was enriched with two new elements - fortress walls and churches. These are not exactly new to the 6th century, but their presence in the city was definitely 'normalized' during this period. Perhaps the best illustration of the emergence of a new Byzantine urban identity comes again from Procopius's text; it is the well-known passage about the city of Melitene,15 in Armenia. This city developed in the age of Trajan on the site of a former legionary castrum. It was then gradually expanded with churches, markets, roads, porticoes, baths and 'everything that contributes to the adornment of a great city', finally being completed by the construction of powerful walls in two building campaigns, under Anastasius and Justinian.

This is, at least in essence, the ideal of the Byzantine city as suggested by the written sources. It can now be compared and checked against the available archaeological data for the Byzantine cities of the 6th century. It is generally accepted that in the 6th century the eastern cities of ancient tradition still had an 'urban' appearance in the classical sense of the term. The great capitals of the empire, Constantinople, Antioch, Jerusalem, the other major cities, from Apamea to Carthage, and smaller cities in the regions along the eastern coast of the Mediterranean, exhibit unbroken continuity in their urban standards, as is regularly confirmed by archaeological investigations. This is all too obvious for our experiment in 'cognitive archaeology'. We are entering the realm of tautology: if Procopius (and the Imperial Chancellery and the editors of the epigraphic formulas) described what they sated and called it a city, then there was no distinction between the traditional definition and reality. More complex and more useful for our experiment are the so-called Justinianian new cities, i.e. those

operations in the forum of the ancient city. This suggests that a perception of the physical continuity and symbolic value of traditional monumental and privileged areas of the city was still alive even in the periphery of the empire at a time of pressing difficulties; see Zanini (1998) 186-87.

15 Aed., 3.4.15-20.

urban ideal and urban planning in Byzantine cities 201

cities built ex novo or completely renovated under Justinian, or in the first half of the 6th century, if we are not to trust Procopius's panegyric. The very existence of a phenomenon of new urbanism calls for some reflection. By this time in the late antique West, and even in much of Italy, urban centers – and with them the ideal of the city itself – were in serious danger of extinction. In the eastern part of the Greco-Roman world, the birth of a new generation of cities testifies to the continuity of the urban tradition here and, at the same time, to the formulation of a new ideal of the city in the proto-Byzantine Empire. There are a large number of cities for which the literary sources suggest foundation or extensive renovation as a result of Justinian's direct intervention16; this correlates with a large number of dynastic names of cities that refer to the emperor himself or to Augusta Theodora.17

Recently, their number has been increased by several places recorded by epigraphy.18 The total number could now be close to thirty. Moreover, we can add to this corpus a large number of sites which, although they bear no trace of the Justinian intervention in their names, can be counted as new cities in the 6th century on the basis of archaeological field results. However, in this article I will focus on four new cities from which we can discover something useful for our investigation.

The new cities

The four cities in question are: Justiniana Prima (Cari´in Grad) in Serbia, Dara in Mesopotamia, Resafa-Sergiopolis and Zenobia-Halebiyye, both in Syria (Pl. 1). It is true that they are far from each other and are located in provinces with different histories, but they seem to have many elements in common. First, they were all built or greatly altered during the 6th century. in places that had never before been occupied by an urban settlement; secondly we

16 Jones (1964) 719; Haldon (1999) 9.17 Jones (1964) 719 points out how Justinian's reign marked a sharp rise

in the number of cities taking imperial or dynastic names (there are, by comparison, only nine known cities named Anastasiopolis and only four called Zenonopolis). This fits well in the context of a well-consolidated Hellenistic tradition in the eastern provinces of the empire, especially in Late Antiquity.

18 Feissel (2000) 83 and note 18-19.

enrico zanini202

know from textual sources the foundation date of each and often some details of the construction phases; finally, they all had very short urban lives, being founded and abandoned in less than a century. The individual histories of these towns are sufficiently well known to be briefly sketched here.

well

The oldest of the four cities is Dara, on the southeastern border of the empire, which was founded between 504 and 506 by order of Emperor Anastasius (Fig. 2). The sources19 contain detailed reports on the phases of its construction. These provide the best basis for reconstructing the complex institutional, economic and strategic mechanisms involved in the creation of a new city at this time. The texts tell us of the procedures followed: the selection of a suitable site, protected by the mountains behind and with abundant fresh water from the river Cordes; the acquisition of the land, originally owned by the Church of Amida; an architect's dispatch

19 Zach. Myt. HE 6.6 (red. Brooks (1953)); The Chronicle of Joshua the Stylite (red. Wright (1882)) 70.

Fig. 1. Map of the Mediterranean basin showing the location of the four cities mentioned in the text.

urban ideal and urban planning in Byzantine cities 203

from the capital with a preliminary draft of the town plan; and the tax-exempt status offered to the citizens involved in the construction work. Procopius20 writes in detail about the restoration of the walls and the infrastructure associated with them, as well as about the general renovation of the city plan that Justinian undertook a few decades later.21 The site, which has never been the subject of systematic excavations, can today only be partially investigated by archeology . Nevertheless, it is still fairly easy to distinguish

Fig. 2. Sketch plan of Dara.

20 Aed. 2.1.11-27; 2.2.1-15.21 The credibility of Procopius' description of Justinian's building at Dara

has been strongly questioned in an important article by B. Croke and J. Crow (1983), which marked the beginning of a phase of disagreement about the general reliability of the aedificiis. Their conclusions were later challenged on historical grounds by M. Whitby (1986), while a series of field investigations by Italian archaeologists (Furlan (1984), (1988) and (1994), Zanini (1990)) have provided new possible evidence for , that there was a Justinian construction phase at Dara.

enrico zanini204

triangular original plan with three corners, corresponding to as many hills, connected by the impressive remains of the walls. The internal organization of the city is characterized by two bridges, an ingenious hydraulic system that controls and distributes the water from the river crossing the city, an aqueduct, two cisterns, a probable public building and a possible religious structure.

Zenobia-Halebiyye

Halebiyye is similar in many ways to Dara and is also an important stronghold on the border with Mesopotamia (Fig. 3). The city was built in Justinian times, probably in a series of building operations between 530 and 550, although there may have been an earlier Anastasian phase. It was built on the site of ancient Zenobia, which had been founded three centuries earlier, by the queen of Palmyra of the same name; it had fallen after the Roman conquest of her kingdom.22 In this case, too, the physical topography of the place was decisive for the city's planning. The site is located on the right bank of the Euphrates, against a steep hill, and dominates what in antiquity was one of the most important fords across the river. Its walls are still largely preserved, marked by sharply projecting rectangular towers at regular intervals. A citadel rises on top of the hill, and just below are the remains of what appears to have been a large public building, perhaps a praetorium, the stronghold's military headquarters. The plan of the city is interesting: the center was marked by the intersection of the two main roads, one of which connected the two city gates, while the other connected the citadel to the river bank, where there may have been a small port. The settlement was crossed by a dense, orthogonal road network, still little studied, within which were located the significant remains of at least two Christian basilicas.23 The military role of the city was further enhanced by the presence of a fortified settlement on the hill on the opposite wide, known today as Zalebiyye and in all probability corresponding to the Byzantine fort of Annoukas, mentioned by Procopius.24 It is likely that this was part of a joint system

22 Sources and archaeological data are described in detail in Lauffray (1983) and de' Maffei (1990).

23 Lauffray (1991) 35-47.24 de' Maffei (1990). 172-77

ur

ba

n id

ea

l an

d u

rba

n fly

nn

ing

i b

back

ntin

e c

itie

s205

Fig. 3. Plan of Zenobia-Halebiyye, modified from Mango (1978).

enrico zanini206

of view and defense of the fords that crossed the Euphrates.25

Resafa-Sergiopolis

The third case is Resafa-Sergiopolis, which is geographically close to Halebiyye (Fig. 4). This was actually a pre-existing settlement, a modest castrum on the eastern limes of the Roman Empire, which became famous as the site of the martyrdom of Saint Sergius, one of the most revered saints of Eastern Christianity.26 It became a center of pilgrimage and which early in the first quarter of the 5th century was provided with a martyrium to host the remains of the saint. At the end of the same century, a first phase of the renovation began; according to a new interpretation by G. Brands, the first large martyr church (the so-called Basilica A) was built at this time.27

A few decades later, Justinian's strategists judged the site to be a key center for control of an important part of the eastern limes; this area was then fought over by the Byzantines and Persians, as well as being subject to expansionist moves by local nomadic populations. Therefore, within a few years, the pilgrimage center was transformed into a significant fortified city, where three urban functions coexisted: an administrative and economic center for an important area,28 a fortified city and a place of worship and pilgrimage. The works may have been planned by the same architects, and they are very likely carried out by the same workforce that a few years later planned and built the walls of Zenobia-Halebiyye, as these walls show similarities with Resafain's materials, construction techniques and structural solutions used.

25 On the routes followed by the Persian armies in their repeated attacks on the Byzantine frontier see de' Maffei (1987); Fowden (1999) 60–67.

26 The sources regarding the birth and development of Resafa have been collected and discussed in detail by Fowden (1999) passim.

27 Brands' conclusions, come from his thesis carried out at Freie Univ. ofBerlin (1994) and to be published (Resafa 6. Die Bauornamentik von Resafa-Sergiupolis); they are summarized and discussed by Fowden (1999) 82-83. Brands' interpretation of Basilica A is based on archaeological data previously discussed by Ulbert (1986).

28 It is worth remembering how the process of transformation of the city's political and administrative roles began late in the reign of Anastasius, when the relics of Saint Sergius were moved to Constantinople. At the same time, the bishop of Resafa was given the rank of metropolitan, indicating that perhaps Resafa was now regarded as an emblem of Byzantine urban civilization in relation to the Arab 'barbarians'; the relevant sources are collected and discussed in Fowden (1999)64-66, footnotes 24-30.

urban ideal and urban planning in Byzantine cities 207

Here, too, the ruins impress, from the circuit of fortifications, wonderfully preserved throughout its length,29 to the four churches discovered so far30 to the underground cisterns, fed by an ingenious system that channeled and filtered water from sediment.31

Justiniana first

The final case to consider (and from its name certainly the most resonant) is that of Justiniana Prima, convincingly linked to the site of Cari´in Grad in Serbia (fig. 5). The city, founded by Justin-ian on the site of the remote village of Tauresium in NorthernIllyricum, his birthplace, is actually the least mentioned of the four cities by ancient sources. Nevertheless, a number of legal texts, esp

Fig. 4. Plan of Resafa, modified after Karnapp (1976).

29 Karnapp (1976).30 The large bibliography on the archaeological investigations at Resafa is compiled

in Ulbert (1989) For the latest excavations see Konrad (1999).31 Brinker (1991).

enrico zanini208

Fig. 5. Plan of Justiniana Prima— Cari´in Grad, modified after Duval (1984). For fig. 6-11 see plates.

urban ideal and urban planning in Byzantine cities 209

Novellae 11 and 131 of A.D. 535 and A.D. c. The site is on a flat hilltop and benefits from a slightly higher area for the Acropolis; here was the bishop's church with a number of adjacent buildings and a possible palace complex, perhaps the seat of the ecclesiastical or civil administration. Below the acropolis was the upper city, organized around a cardo and decumanus. At their intersection there was a circular square, while houses and churches were built all around. South of the upper city, originally outside the walls but later included in the defensive perimeter, stood two churches and a bathing complex. More religious and civil buildings, a bathing complex and at least two churches were located beyond the outer walls.

Traditional urban characteristics

These cities represent four individual cases, but also belong to a single typology. Indeed, from the evidence we have so far, the urban institutions at each of the four archaeological sites seem to share several characteristics, making them an interesting case study of the fate of the city during this period. First of all, I would say that all these cities can be considered part of the classical, Greco-Hellenistic and Roman urban tradition: their general layout, infrastructure, monuments and public and private areas correspond broadly to the general physical features that usually associated with the typical functions of an old city. In the classical age and until late antiquity, a city was the place where public authority resided: a city council,34 a provincial governor belonging to the administrative hierarchy of the empire, a military commander a bishop. It was often the place where all these authorities functioned, as in Antioch, either separate from each other or grouped together

32 Maksimviµ (1980); Zanini (1988).33 For the history of the excavations at Cari´in Grad-Justiniana Prima, see

Kondiµ—Popoviµ (1977); for the latest studies from the joint Franco-Serbian missions, see Cari´in Grad I (1984) and Cari´in Grad II (1990), with notes by Duval (1996).

34 On the complex issue of the institutional and operational continuity of curiales in the public administration of cities in Late Antiquity: Whittow (1990) 4-12; Brandes (1999) 29-31; Liebeschuetz (2001) 104–36.

enrico zanini210

together in different ways.35 As a result, a city had to include a certain number of public buildings that were either intended for public use or to give the impression of a recognizable seat of authority.

As for the headquarters of the authorities, we are spoiled for choice, because the four new cities contain a kind of anthology of what is now traditionally called the 'archaeology of power'. At JustinianaPrima we have an acropolis designed to host an authority of the highest rank. It could have been the praetorian prefect per Illyricum (if we prefer to expand the role of our city to that of the new capital of northern Illyricum - which it actually became under the Justinian laws at the time36) or the bishop (if we have a more restrictive view on the city's privileges)37, which was addressed directly by Justinian's novels, and whose authority was recognized as being equivalent to that of the Metropolitan of Thessalonica.38 We must also mention the principle identified after the Franco-Serbian excavations in the southern corner of the upper city . It had been built in the foundation phase, and in the following decades it was the subject of a series of building campaigns to enlarge and transform it.39

In Dara, the seats of the civil and military authorities are less precisely defined, but a number of archaeological indications and a careful reading of the sources seem to point to a large building (Fig. 6) located in the center of the city, on a large artificial terrace, in whose depths could have been a terrible prison.40 In Zenobia there is no doubt about the seat of the military authority. This found its natural headquarters in the buildings of the fortified citadel and in a large monumental building (Fig. 7) located in the northwestern part of the city, which was structurally included in the circuit of fortifications.41 This is a complex building, both militarily functional and able to visually communicate the authority of the ruling power; the building reveals through the refinement of its architectural features that it was much more than a simple garrison headquarters. At the time of the study conducted by M. Dunand and J. Lauffray in 1945,

35 On continuity and changes in the city's legal status and political prominence in the 5th and 6th centuries. AD see Liebeschuez (2001).

36 Maksimoviµ (1980)37 Lemerle (1954).38 Nov. 11 and Nov. 131.39 Cari´in Grade II (1990) 123-60.40 Furlan (1988).41 Lauffray (1983) 121-23; de' Maffei (1990) 146-47.

urban ideal and urban planning in Byzantine cities 211

remains of the episcopal complex were still easily identifiable on the ground; this was located near or even physically connected to the so-called East Church and was immediately next to the intersection of the two main road axes in the city.42 In Resafa, the characteristics of the archeology of power, especially in terms of the civil and military seats are concerned , appears at first to be less neatly characterized; However, it is worth considering two complexes: the first is the so-called Basilica A, where the relics of Saint Sergius were kept and whose large adjacent buildings could have hosted the episcopal seat; the other is the complex of the tetraconch church, of a type found elsewhere in Syria and northern Mesopotamia, which has been associated with important seats of ecclesiastical power.43

A traditional city worthy of its name should also include a basic infrastructure linked to the quality of daily community life: roads, aqueducts, public baths, etc. These requirements are also largely met by our four cities. At Justiniana Prima, the only city in our group for which we know the plan well, the almost geometrical city center is marked by a circular square surrounded by porticoes: this square marks a fundamental stage in the ideal route between the main gate (south) and the acropolis. It follows a pattern typical of the late antique and proto-Byzantine world and reminds one – mutatis mutandis – of the row of forums along the axis of the via regiain Constantinople.44 All the roads of Justiniana Prima had porticoes (fig.8 ), which is one of the features that give its urban planning a monumental and somewhat artificial appearance, noted by many scholars.45 As we have already seen, its layout is essentially orthogonal, with only a few variations due to the necessity of ​to adapt to the country. characteristic of the site.46 The same planning rigor cannot be found in the road system in the lower part of the city, but this was annexed later, after the middle of the 6th century. The same regular layout is found at Halebiyye, where orthogonal axes were followed even more systematically, although this may reflect a reuse and adaptation of

42 Lauffray (1991) 49-80.43 Kleinbauer (1973).44 On the importance of this pivotal point in the planning of the layout of Justiniana

Prima and on the possible existence of a theoretical plan conceived in Constantinople and later adapted to the site, see Vasiµ (1990), reconsidered in Sodini (1993) 145; critical observations in Duval (1996) 326-27.

45 See finally Duval (1996), with synthesis and discussion of earlier bibliography. 46 See supra p. 207.

enrico zanini212

former Roman road system.47 At Dara and Resafa, the layout is unclear, partly because we actually know very little about them. At Dara, the presence of a well-organized road network is implied by the presence of three bridges (two inside and one just outside the walls) that made it possible to cross the Cordes River. In the case of Resafa, the very location of the gates - and the design of the road network - seems to be strongly conditioned by the Christian monuments that predate the urbanization of the site.

But the elements of infrastructure that the architects and engineers from the 6th century. more attention was paid to those associated with the provision and management of urban water supply, through the use of aqueducts, cisterns and distribution systems. Justiniana Prima was fed by a 30 kilometer long aqueduct which entered the city from the south with a long haul . 48 A castellumaquae was placed immediately within the walls, where the aqueduct entered, and from which pipelines supplied the various quarters of the site and the baths in the lower part of the city (fig. 9). We have no certain information about the existence of aqueducts and cisterns in Halebiyye, but the city is located on the banks of the river, and in those happy times, without problems of pollution, it was unnecessary to make any great efforts, since the great father The Euphrates provided everything the water needed. The water of the river was certainly used, through an ingenious raising system, to feed the great baths erected in the north-eastern part of the city, near the walls along the Euphrates.49 Procopius, in his precise account of the work done in this city, attributes the baths to two well-known architects from Constantinople: John of Byzantium and Isidore of Miletus the Younger.50

Matters were much more complex for Dara and Resafa. There was also a river at Dara, but it was much less manageable than the placid Euphrates, and above all, its flow was seasonal. The RiverCordes was completely dry during the long Syrian summer and wild and violent during the rainy season. In this case, the imperial architects did their best by devising a composite system that utilized the river waters to build up sufficient water reserves. Upstream from

47 Lauffray (1991) 35-47.48 Petroviµ (1970).49 Lauffray (1991) 113-29.50 Aed. 2.8.25.51 Furlan (1984).

urban ideal and urban planning in Byzantine cities 213

a large dam51 was built in the city, which made it possible to slow down the flow of water, thus filling a reservoir to be used during the siege. The water from the river was led into the city through special sluice gates (Fig. 10) protected by grates for obvious safety reasons. Within this structure was an ingenious system of movable stone barriers which regulated the flow and distributed it through a series of secondary conduits. Along the southern part of the walls, another sluice allowed the water to leave the city. Here, another adjustable dam next to the proteichisma allowed an additional reserve of water, used to water animals in the event of a siege, to be built up; this was stored between the main wall and the proteichisma.52 This complicated technique solved the problem of water supply during the rainy season, but could do nothing against the long summer drought, when the RiverCordes ran completely dry. To overcome this problem, Justinian's mechanikoi dug a long aqueduct into the rock, collecting the water from several sources and bringing it to a large covered cistern in the center of the city, close to the western walls. Traces of the aqueduct are still visible today and can be followed for several hundred meters along the Cordes valley.53

At Resafa, right in the middle of the Syrian desert, a similar solution was adopted: the water from what must have been a large catchment area was brought into the city through a series of very long canals. The water was cleaned of sand by a complicated system of filters and settling tanks, and was then stored in huge underground cisterns with parallel vessels (Fig. 11). These cisterns, still largely well preserved near the southwest corner of the walls, closely resemble the huge Justinian underground cisterns at Constantinople in some architectural details, such as their superbly woven brick vaults, in the materials used and in their construction techniques. This is another proof of the existence of connections between center and periphery, here within the specific area of ​​architecture.

One could go on to enumerate the traditional urban characteristics of these new cities in the 6th century; but here I just want to emphasize an additional aspect which I consider to be particularly significant. Certainly in the case of Justiniana Prima, but very likely

52 The function of the complex system of managing the water of the river Cordes is described by Procopius (Aed. 2.2.1-21), whose words find exact correspondence with the archaeological remains still visible on the ground; see Furlan (1984). .

53 Furlan (1995).

enrico zanini214

Also in Resafa's case, the city had a suburb, just like the cities in the classical tradition.54 I think that this can be seen as a strong sign of conceptual continuity in the classical relationship between acity and its territory, which in this respect can be argued against the urban model proposed for this area of ​​the Balkans at this time by Andrew Poulter, of an administrative center which largely excluded the city's inhabitants.55 Like the Hellenistic and Roman cities, Justiniana Prima had a suburb, that is, a transitional zone between the city ​​and what is not the city; critically, this was built at or around the same time as the rest of the city. This transition zone was arranged, typically like those in ancient cities, along one or more of the main road axes leaving the city. Here were located public buildings, sometimes of remarkable importance. These were, in our case, mainly religious, but there were also other public buildings, such as the so-called extra-urban baths, outside the eastern gate of the lower city. It goes without saying that the suburb of Justiniana Prima (and even more so that of Resafa), has little to do with the suburban areas of the large and medium cities of antiquity, at least in their dimensional scale, but the simple fact that it became considered to normally plan a structured suburban area from the beginning shows us, I think, the cultural importance that the founding of a new city in a region recently brought back under Roman control and civilitas, as assumed in the 6th century. Constantinople.

En ny type by

In addition to traditional urban characteristics, our four cities also show new features that somehow remove them from the classic city canon and allow us to see them as a new type of city. With some simplification, these characteristics can be summarized as follows: as cities, they are small, fortified, Christian and imperial.

In terms of dimensions, a simple comparison (Fig. 12) of the urban areas of our four centers with the urban areas of a few large and medium-sized cities in the ancient tradition (Apamea or Nicaea or Gerasa, to name just a few) shows that they the former is much smaller. This observation risks being both abstract and arbitrary, then

54 There was probably not a suburb at Dara and Zenobia where defensive considerations clearly prevailed.

55 Poulter (1992).

urban ideal and urban planning in Byzantine cities 215

Fig. 12. Sketch plans of some old and new Byzantine cities redrawn to the same scale.

it is clear that in ancient times there were also small towns. On the other hand, if a city like Justiniana Prima, artificial and festive but nevertheless designed to play a central administrative role in the region, was planned with the dimensions of a Roman legionary castrum, the choice must somehow be important . Therefore, if we measure the new cities in relation to the parameter of physical dimension, they are remarkable heirs to the classical tradition; if dimensional reduction implies decadence, then we are certainly facing decadence. But equating dimensional reduction with a simplification of the city's ideal is perhaps a little too banal. It is true that the new cities were not large, but it is also true that, like the late antique cities with an established urban tradition, they concentrated several functions within themselves which defined their relationship with the surrounding territory. You others

enrico zanini216

words, our cities appear to have been new types in the traditional tripartite classification of Roman and late Roman settlement into the city, the countryside, and the defensive center. They were at the same time administrative centers, nodes of a defense system and fixed points in the ecclesiastical geography of their respective regions.

In relation to this, the close relationship between each of our cities and the border is particularly important. It is enough to look at their geographical positions to understand this. Naturally, local conditions determined the individual aspects of each city, but for all their differences, Justiniana Prima, Dara, Resafa and Zenobia are all border towns in one way or another. In some cases the defensive character is clearly dominant, while in others strategic and administrative aspects predominate. However, all four cities represent the cultural vanguard of Byzantine urban civilization, facing the territories across the border and their different peoples.56 Here a window opens onto the cultural and religious perception of the border in the Byzantine world. Although this question is beyond the scope of my article, it is worth recalling a short and significant passage in Procopius57, where the historian wrote about the renovation of the fortresses of the Danube limes by Justinian and that the emperor, through his interventions, not only gave them back their defensive role, but also turned them into real small towns, densely populated. This statement fully conveys the richness of the meaning associated with the ideal of the city at this time.

Like late antique cities, but even more so because of their location along the frontier, our new cities were naturally defended by walls. In some cases the walls had the appearance of monumental, symbolic enclosures, as in the acropolis and in some respects also in the upper part of Justiniana Prima; in others (at Dara and Halebiyye obviously, but also at Resafa) the defensive character clearly prevailed. In all cases, however, the construction of a circuit of fortifications seems to be a fundamental element of the new city. The walls not only defended urban settlement, but at the same time closed the city jurisdiction into a conceptual and physical boundary. They also related functionally to the surrounding areas that the city ruled and defended,

56 We have already pointed out how Resafa is often described in the sources as located in the 'barbarian area'; see supra n. 26.

57 Aed. 4.1.2, 24-30.

urban ideal and urban planning in Byzantine cities 217

in return receives their financial support. On the whole, I would say that here we see the beginning of a conceptual shift from city, to defensive city, to city that defends a territory. This finds exact correspondence in the terminological shift from polis-civitas to castrum; an extraordinarily important shift for the understanding of the development of Byzantine cities, in the centuries from the early to the central Middle Ages.

Attention to the walls of the new cities went far beyond what was necessary on a purely functional level;58 they seem to have been the subject of specific and precise planning. For Darawe has precise written accounts59 which reveal in detail the procedure followed in the design and construction of the original Justinianicwall circuit. For Zenobia, the accounts we have are perhaps less detailed, but we are certain of the involvement of two important imperial architects. For Resafa we have no conclusive evidence, but the presence here of the same workforce and probably of the same pool of designers who worked at Zenobia is at least probable given the absolutely identical construction techniques used and very strong similarities in the overall design of the walls. We have no textual references to the building of the walls of Justiniana Prima, but also in this case the hypothesis of a direct intervention by architects from the capital is supported by their similarity to the walls of the other settlements and by the use of the same construction technique ( rows). of stone alternating with layers of brick), which was common in Constantinople from the age of Theodosius II and was still widespread in the 6th century.

But in addition to being small and walled, the Byzantine new city was also - and perhaps above all - a Christian city, and it was during the 6th century, and especially in the decades of Justinian's rule, that the Christianization of ​​the urban areas that had begun in the time of Constantine, two centuries earlier, came to an end. This phenomenon is particularly evident in Constantinople, where a simple count of the surviving monuments lends credence to Procopius' account that in the age of Justinian (or if we prefer the first half of

58 It should be noted that this idea of ​​the meaning of the urban perimeter is markedly different from the simple functional idea that characterized Italy, where the care and renovation of city walls in Late Antiquity is rarely mentioned in written or epigraphic sources; see Christie (2001).

59 Zach. Myt. HE 6.6 (red. Brooks (1953)); The Chronicle of Joshua the Stylite (red. Wright (1882)) 70; Procop. Aed 2.1.11-27, 2.2.1-15.

enrico zanini218

in the 6th century) many more churches were built than had been built since the founding of the city. A similar phenomenon is visible at JustinianaPrima, which is possibly even more significant given its size. Here, archaeological investigations have so far revealed the existence of eight churches in an urban area of ​​less than five hectares, giving the city a density of churches that can only be compared to Rome in the Middle Ages. What is also striking about Justiniana Prima's churches, apart from their number, is the extraordinary variety of types.60 This suggests complex cultural influences, almost an architectural cosmopolitanism, and is, I think, a strong indication of the rich, complex meaning . , as the design and building of this city had for contemporaries. A similar picture emerges for Resafa, where Christianity was a decisive element in the birth of the city in the 5th century. and became - in the age of Anastasius and even more so in that of Justinian - a key element in the city's urban structure and fundamental character. For Zenobia we have already pointed out the central position of the episcopal complex. Close by is at least one other large church.61 The location of buildings for the Christian cult in Dara is less clear, but also in this case we can easily surmise the presence of several important religious buildings. There are scattered remains of a number of churches. , one of which can hypothetically be placed in the southwest quadrant of the city, close to the river and the alleged praetorium.

The last, but certainly not least, innovative character common to the Byzantine new cities of the 6th century is that they were all founded and developed under direct imperial orders. The direct involvement of the emperor and his closest officials in the restoration and construction of public buildings is a common phenomenon throughout Late Antiquity, especially in the East. This phenomenon became widespread during the 6th century. and was not limited to new cities. But in the work of these cities we may see a coherent, almost centralized standard unit of notions that govern the foundations of new cities and their development. This is possibly the most important point.

Literary and epigraphic sources suggest that not only was there an explicit imperial decision behind the foundation of new cities and all the operations to restore the old cities or their walls in the 6th century, but there was also what we today will call an 'arrangement team'.

60 Duval (1984).61 Se supra, side 211.

urban ideal and urban planning in Byzantine cities 219

chitects', tasked with planning, designing and realizing each project.62 The way in which interventions were carried out seems to have been repeated almost identically in each case and to have followed a sequence: the central decision; the general plan prepared by a pool of architects in Constantinople, based on very accurate information about the nature of the site; the actual realization on site of the works to be carried out, carried out either by the same architects or, more usually, by other, lower-ranking colleagues. In this regard, the case of Dara is again particularly illuminating. The extraordinary works of water engineering which characterize Justinian's intervention at Dara were attributed by Procopius to Anthemius of Tralles and Isidore of Miletus; these famous architects were responsible for the construction of the Hagia Sophia in Constantinople. On the basis of information brought to the capital, the two masters devised a solution to the recurring flooding of the river and probably gave general guidelines for the execution of the project. But these great characters did not waste their time on a difficult and perilous journey into the distance. the province of Mesopotamia. It was therefore another mechanikos, Crises of Alexandria, otherwise unknown but obviously well acquainted with the problems to be faced, who was responsible for the realization of the project.

Something similar happened in Zenobia. Here, according to Procopius, John of Byzantium and Isidore of Miletus the Younger were responsible for the second phase of the construction work. The latter owes his place in architectural history to designing the new dome of Saint Sophia, after the original collapsed in 558. Procopius expressly states that both were very young at the time of the works at Zenobia, so it is not unreasonable to believe that they could at this time has taken responsibility for the execution of the works on the ground, possibly a fundamental step in the course honorum for a 6th century Byzantine architect. The same arrangement could possibly be indicated by epigraphy, which reveals that Isidore was the builder of the fortifications at Chalcis in Syria at about the same time as he was busy in Zenobia.63 Mekanikos Victorinus,64 is in many respects a similar figure. He is otherwise unknown, but two

62 Regarding architects in the Byzantine world, the observations of Downey (1946-48) are still generally valid, but see also the more recent adjustments of Saliou(1996) 84-91.

63 Fourdrin, Feissel (1994).64 Feissel (1988).

enrico zanini220

inscriptions found in Albania indicate that he was responsible for the fortifications of Byllis and for the majority of the defense systems in Thrace, Moesia, Epirus and Illyricum. All we have to do now is to use our imagination a little, and since the site of Cari´in Grad is right in the middle of the area where this architect was supposedly active, to imagine Victorinus designing or rather executing the plan for Justiniana Fine.

It is, of course, not about attributing this or that work to this or that architect. What should be emphasized is that the involvement of imperial architects is general; this seems to be tantamount to a standardization of government interventions in these cities, both in the case of total renovation and especially in the case of new foundations. When the decision was made in Constantinople to build a new city in one of the provinces of the empire, reference was made to a perfectly defined conceptual model; this could be used as a basis for a well-articulated overall plan that must be adapted to the specific needs of each settlement. It was probably this kind of conceptual model that was a 6th century Byzantine. remembered when he uttered the word polis whether he was drawing up the plan of a new settlement, or more simply defining the place where he lived with his community. If such a conceptual model was so clearly defined in the 6th century. East, the same was certainly not true at the same time in the West, where old architectural and urban conceptions had long been in decay. Nevertheless, the survival of the city in the Byzantine world, and to an even greater extent the survival of the ideal of the city in Byzantine material culture, may well have been a key factor in the preservation of this ideal also in Western culture.65 I believe, that it represents one of the most important factors to explain the rebirth of cities in the West after the truly dark centuries of the early Middle Ages.

65 For the role of Constantinople and its monuments in this process, see Ward-Perkins (2000). It would be interesting in this regard to study the role that the preservation of the classical idea of ​​the city in the Byzantine world (or even the dissemination of this ideal and concrete projects) could have had in the rapid development of urban planning in the early Islamic world, after the conquests. See on this the preliminary study by Hillenbrand (1999).

urban ideal and urban planning in Byzantine cities 221

Bibliography

Balty J. C. (1986) "Apameia in the sixth century. Archaeological evidence for the wealth of a city", in Men and Wealth in the Byzantine Empire, I, IV.-VII. century (Paris1989) 79-96

Barker P. (1997) The Baths Basilica Wroxeter: Excavations 1966-90 (London 1997) Brandes W. (1999) “Byzantine Cities in the Seventh and Eighth Centuries – Various-

sources, different stories?”, in The idea and ideal of the city between Late Antiquity and the Early Middle Ages, ed. G. P. Brogiolo, B. Ward-Perkins (Leiden, Boston and Cologne 1999) 25-57

Brandes W. (2000) Haldon J., "Cities, tax and transformation: state, cities and their hinterlands in the Eastern Roman world, c. 500-800", in Cities and their territories between Late Antiquity and the Early Middle Ages, ed. GP Brogiolo, N. Gauthier, N. Christie (Leiden, Boston and Cologne 2000) 141-72

Brinker W. (1991) "On the water supply to Resafa-Sergiupolis", DM 5 (1991)119-46

Brogiolo G. P. (1999) "Ideas of the City in Italy during the Transition from Antiquity to the Middle Ages", in The Idea and Ideal of the City Between Late Antiquity and the Early Middle Ages, ed. G. P. Brogiolo, B. Ward-Perkins (Leiden, Boston and Cologne 1999) 99-126

Brooks E. W. (1953) Historia ecclesiastica attributed to Zechariah the rhetor in general (Louvain 1953) Cari´in Grad I (1984) “Cari´in Grad I: Les basilicas B et J de Cari´in Grad; four

notable objects from Cari´in Grad; the treasure of Hajdu´ka Vodenica”, edd.N. Duval, V. Popoviµ (Belgrade-Rome 1984)

Cari'in Grade II (1990) "Cari'in Grade II: the southwestern quarter of the upper town", ed. B. Bavant, V. Kondiµ, J.-M. Spieser (Belgrade-Rome 1990)

Christie N. J. (2001) "War and order: urban remodeling and defensive strategy in Late Roman Italy", i Recent Research in Late Antique Urbanism, red. L. Lavan (Portsmouth, Rhode Island 2001) 107-22

Croke B., Crow J. (1983) "Procopius and Dara", JRS 73 (1983) 143-59de' Maffei F. (1987) "Justinian's Works on the Eastern Limes: Monuments and Sources",

i Major Papers of the 17th International Byzantine Congress (New Rochelle-NewYork 1987) 237-98

——— (1990) "Zenobia and Annoukas: Justinian's Fortifications on the Middle Euphrates. Phase of the Encroachments and Date", in Constantinople and the Art of the Eastern Provinces (Rome 1990) 135-228

Downey G. (1946-48) "Byzantine Architects: Their Education and Methods", Byzantion18 (1946-48) 99-118

Duval N. (1984) "The Religious Architecture of Tsaritchin Grad in the Context of Eastern Illyricum in the 6th Century", in Cities and Settlement in Proto-Byzantine Illyricum (Rome 1984) 399-481

——— (1996) “The Urban Planning of Cari´in Grad. An Artificial City and Its Ceremonial Buildings: A Local Specificity or a Decisive Step in the Typology of Military Principia?”, AntTard 4 (1996) 325-39

Feissel D. (1988) "The Architect Victorinos and Justinian's Fortification in the Balkan Provinces", Bulletin de la Société Nationale des Antiquaires de France (1988) 136-46

——— (2000) "Justinian's Buildings in the Testimony of Procope and Epigraphy", AntTard 8 (2000) 81-104

Fourdrin J.-P., Feissel D. (1994) "A city gate built at Chalcis in Syria by Isidore of Miletus the Younger", TravMém 12 (1994) 299-307

enrico zanini222

Fowden E. K. (1999) The barbarian plain. Saint Sergius between Rome and Iran (Berkelei-Los Angeles-London 1999)

Furlan I. (1984) Checks in Dara (Padova 1984)——— (1988) “Oikema katagheion. En problematisk struktur i Dara”, i Milion.

Studies and Research in Byzantine Art, I (Rome 1988) 105-27——— (1995) "Cisterne a Dara", in Profane Art and Sacred Art in Byzantium, ed. A. Iacobini,

E. Zanini (Roma 1995) 51-64Haldon J. (1999) "The idea of ​​the town in the Byzantine Empire", i The Idea and

the ideal of the city between Late Antiquity and the Early Middle Ages, ed. G. P. Brogiolo, B. Ward-Perkins (Leiden, Boston and Cologne 1999) 1-24

Hillenbrand R. (1999) "'Anjar and Early Islamic Urbanism", i The Idea and the Ideal of the Town between Late Antiquity and the Early Middle Age, red. G.P. Brogiolo,B. Ward-Perkins (Leiden, Boston og Köln 1999) 59-98

Jones A. H. M. (1937) The Cities of the Eastern Roman Empire (Oxford 1937)——— (1964) The Later Roman Empire. 284-602 (Oxford 1964) Karnapp W. (1976) Die Stadtmauer von Resafa in Syria (Berlin 1976) Kleinbauer W. E. (1973) “The origin and function of the nave tetraconch churches

in Syria and Northern Mesopotamia", DOP 27 (1973) 91-114. Kondiµ V., Popoviµ V. (1977) Cari'in Grad, a Fortified Settlement in Byzantine Illyricum (Belgrade)

1977) Kondoleon C. (2000) Antioch: the Lost Ancient City (Princeton 2000) Konrad M. (1999) “Research on the Roman and Early Byzantine Frontier in the North

Syria”, JRA 12 (1999) 392-410Lauffray J. (1983) Halabiyya-Zenobia, stronghold of the Eastern Limes and Upper Mesopotamia

in the 6th century, I. The frontier duchies of Mesopotamia and the fortifications of Zenobia (Paris1983)

——— (1991) Halabiyya-Zenobia, Stronghold of the Eastern Limes and Upper Mesopotamia in the 6th Century, II. Public, Religious, Private and Funerary Architecture (Paris 1991)

Lavan L. (2001) "The late antique city: a bibliographic essay", i Recent Researchin Late Antique Urbanism, red. L. Lavan (Portsmouth, Rhode Island 2001) 9-26

Lemerle P. (1954) "Invasions and Migrations in the Balkans from the Late Roman Period to the 8th Century", RevHist 221 (1954) 265-308

Liebeschuetz J. H. W. G. The Decline and Fall of the Roman City (Oxford 2001) Maksimoviµ L. (1980) "Northern Illyricum in the 6th Century", Proceedings of the Byzantine Institute

19 (1980) 17-58Mango C. (1978) Byzantine Architecture (Milan 1978) Petroviμ N. (1970) "The Vodovodo of Love", Starinar n.s. 20 (1970) 289-98Pollard N. (2000) Soldiers, Cities and Civilians in Roman Syria (Ann Arbor, Michigan).

2000). Poulter A. (1992) “The Uses and Abuses of Urbanism in the Danube Provinces of

the later Roman Empire” in The City in Late Antiquity, ed. J. Rich (Leicester-Nottingham Studies in Ancient Society 3) (London 1992) 99-135.

Roques D. (2000) "Les constructions de Justinien de Procope de Césarée", AntTard8 (2000) The 'de aedificiis' of Procope: the text and documentary realities 31-43

Saliou C. (1994) Building Laws (Beiruit 1994)——— (1996) Julien d’Ascalon’s Urban Planning Treatise. Law and Architecture in Palestine

in the 6th century, (Paris 1996) Sodini J.-P. (1993) “Archaeology's contribution to knowledge of the world

Byzantine (4th-7th century), DOP 47 (1993) 139-84 Spieser J.-M. (1989) “The Development of the Byzantine City in the Early Christian Period

to iconoclasm", i Men and wealth in the Byzantine Empire, I, IV.-VII. århundrede (Paris 1989) 97-106

urban ideal and urban planning in Byzantine cities 223

Tsafrir Y, Foerster G. (1997) "Urbanism at Scythopolis—Betshean in the Fourth to Seventh Centuries", DOP 51 (1997) 85-146.

Ulbert T. (1996) Die Basilika des Heiligen Kreuzes in Resafa-Sergiupolis (Mainz 1986).——— (1989) "Cities and Fortification on the Euphrates in the Paleo-Christian Period",

in Syria's Archeology and History, 2, ed. J.-M. Dentzer, W. Orthmann (Saarbrücken 1989) 283-96

Vasiµ È. (1990) "The Town Plan of the Upper Town: an attempt at reconstruction", in Cari´in Grade II: the south-western district of the Upper Town, ed. B. Bavant, V. Kondiµ, J.-M. Spieser (Belgrade and Rome 1990) 307-15

Walmsley A. (1996) "Byzantine Palestine and Arabia: urban prosperity in Late Antiquity", in Cities in Transition. Urban Development in Late Antiquity and the Early Middle Ages, ed. N. Christie, S. T. Loseby (Aldershot 1996) 126-58

Ward-Perkins B. (1996) "Urban Continuity?", in Towns in Transition. Urban Development in Late Antiquity and the Early Middle Ages, ed. N. Christie, S.T. Loseby (Aldershot 1996) 4-17

——— (2000) “Constantinople: a city and its ideological territory”, in Cities and their Territories between Late Antiquity and the Early Middle Ages, ed. GP Brogiolo, N. Gauthier, N. Christie (Leiden, Boston and Cologne 2000) 325-46

Whitby M. (1986) "Procopius and the development of Roman defenses in UpperMesopotamia", in The Defense of the Roman and Byzantine East edd. P. Freeman, D. Kennedy (Oxford 1986) 727-35.

Whittow M. (1990) "Government in the late Roman and early Byzantine city: a continuous history", PastPres 129 (1990) 3-29

Wright W. (1882) ed. The Chronicle of Joshua the Stylite, composed in Syriac A.D. 507 (Cambridge 1882)

Zanini E. (1988) "Boundary and Border: The Danube Limes in the Sixth Century", in Million. Studies and Research in Byzantine Art, I (Rome 1988) 257-73

Zanini E. (1990) “The Walls of Dara. Materials for a Stratigraphic Analysis', in Constantinople and the Art of the Eastern Provinces (Rome 1990) 229-64

Zanini E. (1994) Introduction to Byzantine Archeology (Rome 1994) Zanini E. (1998) Byzantine Italy. Territory, settlements and economy of the province

Byzantine of Italy (VI-VIII centuries) (Bari 1998)

gisella cantino wataghin224

CHRISTIAN TOPOGRAPHY IN THE LATE ANTIQUE CITY: RECENT RESULTS AND OPEN ISSUES

GISELLA CANTINO WATAGHIN

Abstract

The study of Christian topography has become increasingly subtle in its assessment of both textual and archaeological evidence. Despite this, the model established in the 1980s, with an intramural episcopal church and extramural cemetery churches, still seems to be valid. However, this image remains provisional; it must not be assumed where the evidence is poor. There are still many questions that can be investigated. Excavations carried out to a high standard may be one way of addressing these. Nevertheless, urban data remain fragmentary, so historical discussions must remain at the model level.

Christian Topography: Birth and Development

Christian topography was established as a research area in late antiquity and early medieval archeology in the early 1970s. It developed against a background of growing interest in urban history, especially in the transition from the Roman to the medieval city. At this time, researchers began to systematically investigate the distribution of early Christian churches in late antique cities, their relationship to the urban context and their influence on its early medieval development.1 Following Bognetti's pioneering remarks, a number of researchers, including Février and Cagiano, established the basic outlines of the subject; the work of the French research group Topographie Chrétienne, which has been operating since 1973, provided an example of method.2 At the same time, the practice of regular excavations in urban contexts, first introduced by British archaeology, gained wide, if not general, acceptance.3

1 Wataghin Canteen (1992); Brogiolo and Gelichi (1998).2 Bognetti (1959); February (1974); Cagiano de Azevedo (1974).3 Galinie and Randoin (1979); Hudson (1981); Biddle and Hudson (1983); Carver

L. Lavan og W. Bowden (red.) Theory and Practice in Late Antique Archaeology(Late Antique Archaeology 1 2003) (Leiden), s. 224-256

Christian Topography in the Late Antique City 225

Since the 1970s, remarkable progress has been possible thanks to the integrated use of archaeological and written sources. Texts, whether documentary, narrative or hagiographic, have been - and still are - reinterpreted in the light of more conscious and subtle readings, as can be seen in the published volumes of TopographieChrétienne.4 Archaeological sources have also been significantly improved, not only in quantity, but more importantly, in quality. Perhaps the best example of an improvement in quality comes from the investigations carried out from the early 1970s in the cathedral of Geneva (fig. 1, see the plates).5 From the beginning, this project used advanced methods of stratigraphic excavation and masonry analysis; from the work carried out it was possible to reconstruct the development of the episcopal group through many phases, from the foundation of the first church at the end of the 4th century. (fig. 2), through numerous enlargements, transformations and reconstructions from the 5th and 6th centuries until the foundation of the present cathedral around 1160.6

The same method has been used elsewhere to re-evaluate early excavations; this has been done at the church of La Madeleine7, whose complexity was not discovered during archaeological work carried out at the beginning of the 20th century; a comparison between the plan from 1977 (fig. 3) and the plan from 1914/18 (fig. 4) makes clear enough the differences in the method involved (fig. 5).

Geneva provides a model for recent work, but there are many other equally significant examples of contemporary practice in Christian archaeology: recent discoveries in cities of various sizes and importance – such as Lyon, Aosta, Grenoble,8 to name but a few – have allowed a fairly detailed idea of ​​the distribution of early churches to be established.9 Nevertheless, the archaeological evidence remains fragmentary; the total amount of data is small compared to the number of cities affected, as is evident from a quantitative study carried out a few years ago.10

(1983); Carver (1987); Carver (1993); Fehring (1996); Brogiolo and Gelichi (1986).4 Gauthier, Picard, Beaujard and Prévot ed. (1986-2000).5 Bonnet (1979).6 Bonnet (1993); Bonnet (1997).7 Bonnet (1977).8 Lyon: Reynaud (1998); Aosta: Bonnet and Perinetti (1986); Grenoble:

Collardelle (1986); Collardelle (1995); Baucheron, Gabayet and Montjoye (1998).9 The first Christian monuments (1995-98). Gauthier, Picard, Beaujard and Prevot

edd. (1986-2000).10 Cantino Wataghin, Gurt Esparraguera and Guyon (1996).

gisella cantino wataghin226

Establishing a model

A landmark in the study of Christian topography was the 11th International Congress of Christian Archaeology; this was held in Lyon in 1986 and its main theme was "The Bishop and his cathedral".11 It was the occasion for a comprehensive study of the latest knowledge in this field, which led to the general acceptance of a model for the development. of Christian topography which had first been proposed by the scholars of Topographie Chrétienne.12 According

Fig. 2. Geneve, Episcopal group, North Cathedral, fase 1 (Bonnet 1993).

11 Proceedings of the XI Congress (1989).12 Pietri (1976).

Christian Topography in the Late Antique City 227

Fig. 3. Geneva, La Madeleine: map (Bonnet 1977).

that's it

c

in

tin

o w

at a

gh

i 228

Fig. 4. Geneva, La Madeleine, map 1914-18 (Bonnet 1977).

ch

rist

ian

to

pog

ra

phy

i t

one day

A'et

ntiq

ue t

Oh

n229 Fig. 5. Geneva, La Madeleine, masonry stratigraphy (Bonnet 1977).

gisella cantino wataghin230

to this model, the common pattern of early Christian topography was bipolar, consisting on the one hand of an urban church, conventionally known as the 'Episcopal' church, and on the other of a variable number of extramural churches. These churches had different, complementary functions: the episcopal or parish church was the place for formal worship, liturgical use, general education and baptismal education, with a baptistery and associated episcopal or clerical quarters; extramural churches were intended for burial practices and the cult of martyrs, which could also have involved the presence of a baptistery.13 The model seems to be valid for cities over a large area, from Rome (fig. 6) to Milan (fig. 7) and Aquileia, from Geneva (Fig. 8) to Lyon (Fig. 9) and at Aosta (Fig. 10), Tarragona, Grenoble (Fig. 11) and Paris (Fig. 12), to name but a few of a large number of examples.14

As regards chronology, there is no evidence of the foundation of churches before the age of Constantine, not even where a congregation, though as large as that of Rome, existed in earlier times. The presence of a bishop seems to be an essential condition for the building of churches, either in urban areas or outside the walls. It can be inferred that Christian topography was the result of a variety of factors, apart from the spread of Christianity and the consequent growth of society; these included the church's position as an institution, the church's financial resources, and last but not least, the quest for visibility, which was crucial to the emergence of a monumental Christian architecture.

There is not enough evidence to say that the location of the episcopal church in the city followed any specific rule. It could be placed in a central position or on the outskirts of the city, close to the fortress wall (a position once thought to be canonical, but represented by less than 40% of cases), or it could be located in a number of intermediate positions (fig. 13-14).15 In principle, neither the chronology of the church's foundation nor the possibility of

13 Cantino Wataghin, Gurt Esparraguera and Guyon (1996). 14 Aquileia: Testini, Cantino Wataghin and Pani Ermini (1989); Geneva: Bon-

net (1986); Lyon: Reynaud (1998); and Aosta: Bonnet and Perinetti (1986), Tarragona: Godoy Fernández (1995); Grenoble: Collardelle (1986); Paris: Duval, Perin and Picard (1992).

15 Testini, Cantino Wataghin and Pani Ermini (1989); Duval (1991); Guyon (1995); Guyon, Boissavit Camus and Souilhac (1995); Cantino Wataghin, GurtEsparraguera and Guyon (1996).

ch

rist

ian

to

pog

ra

phy

i t

one day

A'et

ntiq

ue t

Oh

n231

Fig. 6. Rom, Constantinian Christian Foundations (Krautheimer 1980).

that's it

c

in

tin

o w

at a

gh

i 232

Fig. 7. Milan, kristen topographers (Testini, Cantino Wataghin og Pani Ermini 1989).

Christian Topography in the Late Antique City 233

fig. 8. Geneva, Christian topography (Bonnet 1986).

gisella cantino wataghin234

Fig. 9. Lyon, Kristen topography (Reynaud 1986).

ch

rist

ian

to

pog

ra

phy

i t

one day

A'et

ntiq

ue t

Oh

n235

Fig. 10. Aosta, Christian topography (Bonnet and Perinetti 1986).

that's it

c

in

tin

o w

at a

gh

i 236

Fig. 11. Grenoble, Christian Topography (The First Christian Monuments in France 1995).

Christian Topography in the Late Antique City 237

Fig. 12. Paris, Christian topography (Duval, Perin and Picard 1992).

gisella cantino wataghin238

Fig. 13a. Location of the Cathedral of Metz and Fréjus (Duval 1991).

Christian Topography in the Late Antique City 239

Fig. 13b. Location of the cathedral of Chalon-sur-Saone and Angers (Duval 1991).

gisella cantino wataghin240

Fig. 13c. Location of the cathedral in Grenoble and Tours (Duval 1991).

Christian Topography in the Late Antique City 241

Fig. 14a. Location of the cathedral in Reims and Aix-en-Provence (Duval 1991).

gisella cantino wataghin242

Fig. 14b. the location of the cathedral of Aosta and Arles (Duval 1991).

Christian Topography in the Late Antique City 243

a pre-existing domus ecclesiae affects its location; its position each time seems to have been due to fortuitous factors, such as the needs of the community and the availability of a suitable building site to build a church. However, written and epigraphic sources are explicit about the role that rich donors played in the founding of churches; their contribution generally consisted in the relinquishment of a property to be converted into or replaced by a church: there are many examples either of the reuse of a domus for the construction of a church or of the presence of a domus beneath it.16

Due to their function, cemetery churches and martyria were in principle built within Roman necropoleis, which were located outside the walls in accordance with Roman law; it was here that the tombs of the martyrs were located and also where Christian cemeteries developed; exceptionally, different locations resulted from the desire to honor the site of a martyrdom, as is the case of the church built in the amphitheater of Tarragona (fig. 15). 17 In Constantinian Rome, there seems to have been a desire to mark the main roads out of the city with Christian churches. 18 Also in Milan, the location of the cemetery churches and the martyrdom make it founded between the 4th and 5th c. by Ambrose or his successor Simplicianus does not. depend on the location of honored graves; on the contrary, Ambrose moved objects from the martyrs "invented" in different places into the churches already built in the suburbs (St. Gervasius and Protasius in the Basilica Ambrosiana, St. Nazarius in the Basilica Apostolorum). This choice may imply a desire to create a pattern that evoked the cross or to build a wall of sanctuaries protecting the city, intentions expressed by later writers.19

It is nothing new to say that the Christian topography was a strong factor in the city's transformation. Its influence was felt in two main respects: firstly, churches became the new focus of the urban network, as meeting places for civil society; secondly, they were a unique architectural component of the urban landscape, all the more so as this landscape increasingly came to consist of low wooden buildings, which became a common feature of

16 Cantino Wataghin (1999b).17 Godoy Fernandez (1995).18 Krautheimer (1983).19 Orselli (1996);Gauthier (1999).

that's it

c

in

tin

o w

at a

gh

i 244

Fig. 15. Tarragona, kirke i amfiteatret, plan (Godoy Fernández 1995).

Christian Topography in the Late Antique City 245

European cities in the 6th-7th century c.20 The transformation of the city also affected the suburbs, where cemetery churches grew in number with the spread of the cult of saints and with the consequent increase in the number of pious foundations. Monastic communities were often attached to churches for pious reasons, and the tomb or the relics worshiped in the church gave rise to pilgrimages; these had a range more or less commensurate with the fame of the saint, whose feast was the occasion of a market.

Open questions

Although the broad overview of the nature of Christian topography seems quite clear, the number of open questions is very high. For a large number of towns, no archaeological evidence exists to allow traditions regarding the material origins of the local church to be verified. In other cases, the chronology of known archaeological remains and/or their interpretation is still under discussion. To give just a few examples, we can remind ourselves of the ongoing debate on the assumption that the Baptistery of Florence (Fig. 16) is a late antique foundation: a possibility that, unfortunately, cannot be proven by stratigraphic arguments .21 In Milan, a new hypothesis has recently been put forward about the organization and relative chronology of the episcopal group (figs. 17-18);22 this is at once attractive but questionable: it suggests a complete overview of the earliest , little-known buildings; however, it is based only on observations of the baptism of Saint-Stephen, taking as a model the buildings of Bishop Theodorus of Aquileia; but at present there is no evidence that these buildings actually played a comparable role. For Luni, a recent article on the cathedral (which dates from the early 5th century) (fig. 19) has suggested that the domus below in its final phase served as a domus ecclesiae; this reopens the debate about the churches' dependence on older, informal church sites. 23

The discovery of a privileged burial from the 5th century in Geneva's northern cathedral (Fig. 20) raises the question of the possible influence of

20 Ward-Perkins (1988); Brogiolo and Gelichi (1998).21 Cantino Wataghin, Cecchelli and Pani Ermini (2001).22 Lusuardi Siena (1997), Cantino Wataghin, Cecchelli and Pani Ermini (2001).23 Lusuardi Siena and Sannazaro (1995).

gisella cantino wataghin246

cathedral on the entry of funerals into city centres.24

Clearly, there is still much work to be done; but for the results to be reliable, some methodological conditions must be met. When archaeological investigations are carried out, they should be on a large scale following the 'extensive excavation' method; experience shows that soundings and trenches can raise more questions than they answer, if their results are not actually misleading. A full exploitation of all the possibilities of current archaeological procedures requires the integration of stratigraphic excavations and masonry studies with scientific methods of analysis; these must be carefully managed within a strong interdisciplinary framework. Evidence must be collected and recorded to the highest possible standards to allow interpretations to be checked and revised long after the formal end of field work. That is also

24 Bonnet (1993); Cantino Wataghin (1999a).

Fig. 16a. Florence, kristen topographers (Testini, Cantino Wataghin og Pani Ermini (1989).

Christian Topography in the Late Antique City 247

Fig. 16b. Florence, Cathedral and Baptistery in Late Antiquity, according to Cardini (Cardini 1996).

that's it

c

in

tin

o w

at a

gh

i 248

Fig. 17. Milan, episcopal group, according to De Capitani.

ch

rist

ian

to

pog

ra

phy

i t

one day

A'et

ntiq

ue t

Oh

n249 Fig. 18. Milan, episcopal group, according to Lusuardi (Lusuardi Siena 1997).

gisella cantino wataghin250

Fig. 19. Luni, plan of the cathedral and the dome (Lusuardi Siena and Sannazaro1995).

Christian Topography in the Late Antique City 251

important that archaeological and written sources are given equal importance, although each must be considered on their own terms; discrepancies between them may mean that one or perhaps both have not been correctly understood.

Even if all the practical and cultural problems that hinder large-scale excavations in urban contexts and buildings in use could be overcome, the evidence would still remain fragmentary. Interpretation therefore still needs to concentrate on discussing models. Without questioning the general validity of the current model, from this point of view one cannot hide the risk of it becoming a kind of comfort that can be applied arbitrarily in any situation. This prevents peculiarities from appearing and hinders the construction

Fig. 20. Geneva, north cathedral in the 5th century (Bonnet 1993).

that's it

c

in

tin

o w

at a

gh

i 252

Fig. 21. London, plan of the early medieval city (Thomas 1981).

Christian Topography in the Late Antique City 253

complementary or even alternative models. Some 'exceptions' to the model should be examined more carefully, such as the displacement of the episcopal church from its original location, or the apparently extramural location of some cathedrals, as in Florence; ​​or the union in the same church of parish and cemetery functions, as in Concordia or in many diocesan centers in Sardinia.25

Moreover, the current model of Christian topography was built on the evidence of Roman cities, which were Christianized in Late Antiquity, when Roman urban structures were still recognizable. It does not take into account, except in cases where the destructuring of urban areas started earlier, or where particular local religious and civil conditions were important. On the other hand, one must consider that the Paleo-Christian pattern outlined by the extant model still seems to be effective in British cities26 where Christian topographies developed after the collapse of the Roman urban structure. We may see in this fact a sign of the strength of tradition and therefore the validity of the pattern itself, as well as the effectiveness of the Roman mission to Britain launched by Pope Gregory the Great. Here we should conclude that the Paleo-Christian pattern had become a model that could be imitated from the premises that had created it: but by this time Late Antiquity was certainly over.

Recognitions

I would like to thank Luke Lavan and Bryan Ward-Perkins for improving the English text of this text.

25 Testini, Cantino Wataghin og Pani Ermini (1989).26 Thomas (1981); Maler (1989).

gisella cantino wataghin254

Bibliography

Proceedings of the XI International Congress of Christian Archaeology (Lyon, Wien, Grenoble, Genève og Aosta 21.-28. september 1986) (Vatican City 1989)

The first Christian monuments in France 1-3 (Paris 1995-98) Baucheron Fr., Gabayet F., de Montjoye A. (1998) About the episcopal group of Grenoble.

Deux millénaire d'histoire (Lyon 1998)Biddle M., Hudson D. (1973) The Future of London's Past: a Survey of the Archaeological

Implications of Planning and Development in the Nation's Capital (Worcester 1973) Bognetti G. P. (1959) “Problems of Method and Objects of Inquiry in the History of

Italian Cities in the Early Middle Ages”, in The City in the Early Middle Ages (VI Week CISAM) (Spoleto 1959) 59-87

Bonnet Ch. (1977) Madeleine's First Christian Buildings in Geneva (Geneva 1977)——— (1979) “New Archaeological Investigation of the Chapel's Basement”, in Cathedral

Saint-Pierre of Geneva. The Chapel of the Maccabees (Geneva 1979) 77-96——— (1986) Geneva in the Early Christian Era (Geneva 1986)——— (1993) The Excavations of the Former Episcopal Group in Geneva (1976-1993) (Geneva)

1993)——— (1997) Around the Church. Archaeological excavations in Geneva 1967-1997 (Geneva

1997) Bonnet Ch., Perinetti R. (1986) Aosta. The first Christian monuments (Aosta 1986) Bonnet Ch., Privati ​​B. (1991) Le temple de Saint-Gervais (Geneva 1991) Brogiolo G. P., Gauthier N., Christie N. eds. (2000) Cities and their territories are

between Late Antiquity and the Early Middle Ages (Leiden, Boston and Cologne 2000).Brogiolo G. P., Gelichi S. (1998) La città nell'alto medioevo italiano (Bari 1998)Brogiolo G. P., Ward-Perkins B. edd. (1999) The idea and ideal of the city between

Late Antiquity and the Early Middle Ages (The Transformation of the Roman World4) (Leiden, Boston and Cologne 1999)

Cagiano de Azevedo M. (1974) "Urban Aspects of Early Medieval Cities", in Urban Topography and Urban Life in the Early Middle Ages in the West, (XXI Uge CISAM) (Spoleto 1974) 641-77

Cantino Wataghin G. (1992) "Late antique town planning and Christian topography: terms of a problem", in Felix temporis reparatio. Proceedings of the International Archaeological Conference "Milan Capital of the Roman Empire" (Milan 8-11 March 1990), edd.G. Sena Chiesa and E. A. Arslan (Milan 1992) 171-92

——— (1999a) "The Ideology of Urban Burials", i Brogiolo G.P., Ward-PerkinsB. edd. (1999) 147-80

——— (1999b) “...For at dette hus må blive hemmeligholdt i kirken for Herren Kristus. The Christian reuse of ancient buildings between late antiquity and the early Middle Ages", i Ideologies and practices of reuse in the early Middle Ages, (46 CISAM Week) (Spoleto 1999) 673-749

Cantino Wataghin G., Cecchelli M., Pani Ermini L. (2001) "The baptistery in the fabric of the late antique and early medieval city in Italy", in The baptistery in Italy. Aspects and problems, Proceedings of the VIII National Congress of Christian Archeology (Genova, Sarzana, Albenga, Finale Ligure, Ventimiglia 21-26 September 1998) (Florence 2001) 231-65

Cantino Wataghin G., Gurt Esparraguera J. M., Guyon J. (1996) "The topography of the Christian civitas between the 4th and 6th centuries", in Early Medieval Towns in the Western Mediterranean, Ravello, 22-24 September 1994, ed. G. P. Brogiolo (Mantova 1996) 17-41

Carver M. O. H. (1983) "Forty French cities: an essay on archaeological site evaluation and historical objectives", OJA 2 (1983) 339-78

Christian Topography in the Late Antique City 255

——— (1987) Underneath English Towns (London 1987)——— (1993) Arguments in Stone. Archaeological research and the European city in

First Millennium (Oxford 1993) Colardelle R. (1986) Grenoble in the Early Christian Era (Paris 1986)——— (1995) “Eglise Saint-Laurent”, in The first Christian monuments in France.

In, Southeast and Corsica, ed. N. Duval (Paris 1995) 239-44 Duval N. (1991) "The ecclesia space of the Christian community", in Naissance des

Christian arts (Paris 1991) 50-69 Duval N., Perin P., Picard J.-Ch. (1992) "Paris", in Christian Topography of the Cities

Gallien, VIII, Ecclesiastical Province of Sens, red. N. Gauthier og J.-Ch Picard (Paris 1992) 97-129

Fehring G. (1996) Urban archeology in Germany (Darmstadt 1996) february P.-A. (1974) "Permanence and legacy of antique in topography

de l'Occident durant le Moyen Age", in Urban Topography and Town Life in the Early Middle Ages in the West (XXI CISAM Week) (Spoleto 1974) 41-138

Galinié H., Randoin B. (1979) Land Archives of Tours. The Survival and Future of Urban Archeology (Tours 1979)

Gauthier N. (1999) “Christian topography between ideology and pragmatism”, in Brogiolo G. P., Ward-Perkins B. eds. (1999) 195-210

Gauthier N., Picard J.-Ch., Beaujard B., Prévot Fr. ed. (1986-2000) Christian topography of the cities of Gaul from its origin to the middle of the 8th century (Paris 1986-2000)

Godoy Fernández C. (1995) Archeology and Liturgy. Latin American Churches (4th to 8th Centuries) (Barcelona 1995)

Guyon J. (1992) "Establishment of episcopal councils and early Christian ensembles in Roman cities. The cases of Gaul and Italy", in Simposi Internacional sobre lesEglésies de Sant Pere de Terrassa (1991), Acts (Terrassa 1992) 17-37

Guyon J., Boissavit Camus B., Souilhac V. (1992) "Christian topography of agglomerations", in Cities and ancient urban areas of southwestern Gaul, history and archaeology. Second Aquitania colloquium (Bordeaux 1990), (VI supplement to Aquitania) 391-430

Hudson P. (1981) Urban archeology and research planning: the example of Pavia (Florence 1981)

Lusuardi Siena S. (1997) "The Cathedral Group", in The City and its Memory. The tradition from St. Ambrose (Milan 1997) 36-37

Krautheimer R. (1983) Three Christian capitals. Topography and Politics (Berkeley, LosAngeles and London 1983)

Lusuardi Siena S., Sannazaro M. (1995) "The excavations in the area of ​​the Luni Cathedral: from the private use of the space to public religious buildings", in Splendida civitasnostra. Archaeological Investigations in Honor of Antonio Frova, ed. G. Cavalieri Manasse and E. Roffia (Rome 1995) 191-216

Orselli A. M. (1996) "Consciousness and images of the town in the sources between the 5th and 9th centuries", in Early Medieval Towns in the Western Mediterranean, Ravello, 22-24. September 1994, ed. G. P. Brogiolo (Mantua 1996) 9-16

Maler K. S. (1989) "Recent discoveries in Britain", in Acts 2031-71Pietri Ch. (1976) “Notes on the Christian Topography of the Cities of Gaul

between the Loire and the Rhine (from the origin to the 7th century)", Revue d'Histoire de l'Eglise de France 62 (1976) 189-204

Reynaud J.-Fr. (1998) Lugdunum christianum. Lyon from the 4th to the 8th Century: Topography, Necropolis and Religious Buildings (Paris 1998)

Testini P., Cantino Wataghin G., Pani Ermini L. (1989), "The Cathedral in Italy", in Proceedings of the XI International Congress of Christian Archeology (Lyon, Vienna, Grenoble, Geneva and Aosta 21-28 September 1986). ) ) (Vatican State 1989) 5-229

gisella cantino wataghin256

Thomas A. C. (1981) Christianity in Roman Britain til AD 500 (London 1981) Ward-Perkins B. (1988) "The towns of Northern Italy: Genfødsel eller fornyelse?", i

Rebirth of cities in the west. AD 700-1050, ed. R. Hodges and B. Hobley (London 1988) 16-27

topographies of production in North African cities 257

TOPOGRAPHIES OF PRODUCTION IN NORTH AFRICAN TOWNS DURING THE VANDAL AND

BYZANTINE PERIODS

ANNA LEONE

Abstract

An important characteristic of North African cities in Late Antiquity is the emergence of structures related to craft production in unusual settings, often in former public buildings. In this article, I argue for developing a study of this sector that looks not only at products, such as ceramics, but also at productive structures and their wider urban location. Archaeological evidence from Tunisia and Tripolitania is analysed, dating from the Van Valley, Byzantine and also occasionally early Islamic times, and relating primarily to murex dyeing, fish salting, olive oil production and pottery making. Lime kilns are also considered.

Introduction*

The concept of 'production topographies' refers to the spatial organization of production sites and their location in urban areas.1 It is a topic that has received almost no attention in Late Antiquity urbanism, despite the importance of production activities in the transformation of classical public buildings.2 Production topographies have attracted few studies even for the early Roman period; the most important has been Jean-Paul Morel,

* While working on this paper I learned that Yvon Thébert died. I am eternally grateful to him for his help in orienting me at the start of my studies in North Africa and for many useful suggestions. I dedicate this paper to his memory.

1 All locations mentioned in this article are located with reference to the latest BarringtonAtlas (hereafter BA). Fig. 8 on the plates also gives information on all the Tunisian places mentioned in the text. The term "late antiquity" will here include both the 'Vandal' and 'Byzantine' periods. However, the definition of this term is still a matter of discussion: see for example Giardina (1999).

2 For general accounts of urban production sites in North Africa see Potter(1995) 64-102 and Wilson (2001), although topography was not the main focus of these studies. An attempt to reconstruct "production topographies" for the Van valley and later periods has recently begun at the site of Uchi Maius, where several olive presses (including one established in the forum) have been recorded, from the 5th century. AD and onwards - see Vismara (1999) 73-74.

L. Lavan og W. Bowden (red.) Theory and Practice in Late Antique Archaeology (Late Antique Archaeology 1 2003) (Leiden), s. 257-287

anna leone258

in Rome. In this work, Morel attempted to identify how production spaces were organized and established a topography of production for later Republican and Classical Rome.3 The main elements to be considered seem to be: the overall location of craft/manufacturing activities, such as e.g. they were in the center or on the periphery of the city; if they were kept separate from residential areas; and if they showed any tendency to localize in certain places in a city, or if they had a completely random distribution.

In North Africa, widespread structural evidence of craft production in urban centers is a noted feature of Vandaland Byzantine periods. The purpose of this paper is twofold. First, I will examine how the location of places given over to production changed in urban areas of Tunisia and Tripolitania, from the 5th to the 7th century. A.D., and occasionally beyond; I will consider not only the broad location of production sites in urban areas, but also the kinds of structures in which such activities took place, often former public buildings. Second, I will consider what factors can explain the changes in this period, from the reuse of particular building types to the general location of new types of production activities in urban centers. I will not consider the spatial organization of production sites during the Roman period, but will focus on the transformations that occurred during the Vandal and Byzantine periods. I will also concentrate on change or continuity in the location of late antique production sites, those that arise in new urban or architectural settings; therefore, I will not analyze in detail existing production sites established in earlier times, such as the northern part of the circular harbor or the Magon quarter of Carthage.4 Finally, because this article is primarily concerned with topography, I will not consider here the economic aspect of urban production, which would be too broad for this paper and its focus.

3 Morel (1987) 127-28. Morel also analyzed the tabernae, the markets, the horrea, etc. These elements are often not easily recognizable in late antique cities. For Pompeith there have been a number of studies, such as Majeske (1972) analyzing bakeries and Curtis (1984), on the saltfish industry.

4 The northern part of the circular harbor was used as a manufacturing site, from classical times to the 7th century, with several major phases of change; for detailed analysis: Hurst (1994). Magon neighborhood: mainly Rakob (1991).

topographies of production in North African cities 259

Limitations of the evidence

Before looking more closely at the archaeological evidence, it is necessary to define large areas of doubt and uncertainty. Available information is in many cases fragmentary or of poor quality, which greatly affects the use to which it can be put. In general, recent excavations are reliable and their chronologies are well defined; difficulties, especially with regard to dating, arise from excavations carried out in the 19th and early 20th centuries. These excavations were mainly interested in recording the Roman and Punic phases and removed the evidence of later occupation from many sites. For this reason, the topographical information we can obtain is sometimes fragmentary; it will be difficult for future work to fill the gap because much relevant archaeological evidence has been irretrievably lost. However, new excavations specifically aimed at knowledge of the Vandal, Byzantine and Islamic phases will certainly make it possible to establish a better understanding of urban change. Complex site histories also make it difficult to distinguish production structures of different dates. For example, many sites were continuously occupied or abandoned for a short period after the Arab conquest and then reoccupied, in both cases often inhabited until the 9th/10th century AD.

To make matters worse, there has been no research into the typological development of production structures. For example, the differences between Byzantine and Islamic period olive presses have not been seriously studied. At present, there appears to be similarity between Byzantine and Islamic archaeological evidence. This suggests continuity in production technology from at least the 6th to the 8th/10th. century, when many former urban areas were definitively abandoned. However, it is also possible that Byzantine presses remained in use many centuries after they were built; these structures

5 Serious studies of the post-classical history of African sites have only recently begun. For a first approach see Gelichi and Milanese (1998) 477-80, who provide a synthesis of the main places where continuity or abandonment has been recorded. In Rougga (Tunisia) and Sétif (Algeria) continuity can be seen; the latter did not have a complete re-urbanization until the 11th century. In contrast, at Cherchel (Algeria) evidence of a hiatus has been recorded—Gelichi and Milanese (1998) 478. Another interesting site is Sidi Ghrib, abandoned in Byzantine times and then reoccupied in the Islamic period, perhaps as a fondouk ( rest-house) found mainly along the caravan routes), with the insertion of a forge: Ennabli and Neuru (1994a), (1994b). Significant for the amount of information is also Belalis Maior—Mahjoubi (1978).

anna leone260

was impressive and could survive for a long time.6 Only through further excavations will this subject move forward.

Production in the cities of late antique Africa

North Africa was one of the richest provinces of the Roman Empire: a great variety of goods were produced here. Of the many production activities that characterized the North African economy, the most important that archeology could identify were murex dyeing, fish salting, olive oil and pottery making.7 Olive presses and pottery kilns are relatively well recorded. In contrast, it is not always easy to identify murex staining or fish salting. Nevertheless, evidence for all these production activities will be examined here to provide a rounded overview of the available information.

Murex dyeing and fish salting

Garum production and murex dyeing were both very important to the economy of African coastal towns. They are mainly attested by historical texts.8 For garum, distribution patterns of used ceramic vessels can also be considered an important source that sheds light on the extent of trade.9

Despite a significant number of texts relating to murex dyeing activity in North Africa, few structures directly associated with it have ever been excavated. Most evidence for murex extraction comes from

6 General survey of press technology and consideration of continuity: see Frankel(1999) 176. General information on historical sources on technology used for olive oil production see Humphrey et al. (1998) 159–61.

7 Textile production was probably an important element of the North African economy, although it was difficult to identify archaeologically; however, it is attested in Carthage, in the northern part of the circular harbor: Hurst (1994).

8 Historical sources recording murex dye production exist for many periods: from Pliny (HN 9.126) to Notitia Dignitatum occ. 11.69 (dated to late 4th/early 5th century AD). For a general collection of sources see Drine (2000) 93-94; see also Humphey et al. (1998) 360. For a collection of sources related to garum production see Curtis (1991) 6-15.

9 It has been argued that "African Grande" amphorae were probably used to transport garum. For later forms, the content is not determined. For an experiment see Ben Lazreg et al. (1995) 128.

topographies of production in North African cities 261

the presence of murex shells, such as those reused in mortar in the Byzantine wall of Lepcis Magna10. At Meninx (fig. 8 no. 13), a number of vessels have been recorded in the Vandal period, possibly related to murex dye extraction, although certain examples of this type of production are not known. The area is located on the coast, about 250 m south of the city forum. This activity stopped here immediately after the Byzantine conquest.11 It should also be pointed out that to date we have no reliable documentation of murex dyeing activity in the Byzantine period for Tunisia and Tripolitania. Although locational analysis is not possible here, it should be noted that structures for murex dyeing in cities in all periods were generally located along the coast, as attested for example at Constantia (mid-5th AD) and probably Dor (AD 4-6) in Palestine .12

The major period of garum production in North Africa does not begin until the end of the 1st century. AD It falls off somewhat in the 3rd century; according to Curtis, some of the salteries went completely out of use or reduced their production "drastically" at this time.13 Archaeological evidence of garum production is known from the Imperial period, for example in Neapolis (fig. 8 no. 5); here, in the Nymfarum Domus, a complex was in use from the second half of the 1st century A.D., but was gradually abandoned from the 3rd to the mid-4th century. 14 AD Archaeologists have found no African garum production center after the 5th century, the latest in use being the salteries at Lixus and possibly Kouass, Tahadart and Tipasa.15 Only in Carthage have production sites been suggested: it has been assumed that basins have been found in room 16 in the northern sector, a production site in the circular port, was used for garum16 and in rue 2 mars 1934, in

10 Romanelli (1970) 14. See also Blanc (1958).11 Isle of Djerba—BA map 35 Tripolitana C1. General survey of the territory:

Fentress (2000). The production site was identified from aerial photographs, which revealed a theatre, circus, forum, amphitheater and a series of basins stretched along the coast: Drine (2000) 87. Two areas have been excavated. In general about the excavation, see: Drine (2000) 89-93. The proposed chronology is based on ceramic evidence - see Fontana (2000).

12 See Kingsley, in this volume: pp. 124.13 Curtis (1991) 69.14 Sternberg (2000) 137-38. Today Nabeul BA map 32 Carthago G4.15 Curtis (1991) 69.16 Hurst (1994) 94. Certain elements related to garum production were not found

here; the main manufacturing activity recorded was clothing processing. Hurst suggested the site as an imperial gynaecium (an official clothing factory).

anna leone262

in the same area, another 6 basins were found, for which the same use is proposed.17 In both cases, the constructions were located close to the coast and near the port. For the Byzantine period, certain archaeological evidence regarding garum production and salting of fish in Tripolitania or Tunisia has not been clearly recorded until now. Curtis has suggested that during this period North African garum production experienced a serious crisis.18 It may be that the production of garum in North Africa declined during the Byzantine period, although it is possible that the lack of evidence could be due to the confused and partial archaeological data available.

Olive oil production

Two main elements of the North African economy in Late Antiquity were the production of olive oil and the manufacture of ceramics, which were highly interrelated activities. The structures associated with them are massive and easily identifiable and should therefore be considered in detail. Unfortunately, evidence regarding olive oil production in urban areas is particularly poorly dated, and evidence can be attributed to either the Byzantine or the early Islamic periods. This uncertain situation makes interpretation difficult. However, the current theme is the reuse of classic public buildings and public spaces.

The vandal proof

The only press dated to the Vandal period is the one recently recorded by Uchi Maius (fig. 1 and fig. 8 no. 3). In the forum here was a production center for petroleum oil, in use from the second half of the 5th century until the end of the Vandal period.19 This chronological range corresponds to an increase in North African commercial activity

17 Hurst (1994) 97. This identification is uncertain; it was first suggested that the basins should be used for cloth processing; it was later argued that their use of garum cannot be ruled out. For difficulties in distinguishing between structures used for garum production and those used for murex dyeing, see Curtis (1991) 65-66.

18 Curtis (1991) 69: "The decline in salting...in Zeugitania may have reflected the crisis...at the same time...in Spain".

19 Located in northern Tunisia, today Henchir ed Douimas—BA map 32Carthago D4 Gelichi and Milanese (1998) and (1995); Vismara (1999) 73-74: In the area of ​​the forum, a residential quarter was also developed.

topographies of production in North African cities 263

Fig. 1. Plan of Uchi Maius—Location of the principal city monuments: 1. Citadel;2. Forum; 3. Marabout; 4. The cisterns; 5. City walls; 6. Aqueduct - from Gelichi and Milanese (1995)

anna leone264

confirmed by the ceramics. For example, African pottery from the same period shows the creation of new forms and a wide distribution throughout the Mediterranean;20 a similar trend can be seen in the amphorae, often containers for the olive oil itself. In the Vandal period, especially at the end of the 5th century, new forms were created, primarily characterized by both wide-bodied amphorae and very thin spatheia.21

Byzantine and Islamic Press

Evidence for the Byzantine and early Islamic periods is more problematic, as most of the archaeological remains cannot be closely dated. I will first consider all presses that can reasonably be dated to Byzantine times, followed by the uncertain evidence. It should be pointed out that most of the evidence for which it is possible to identify a chronological range concerns the second half of the 6th century. AD and forward.

Dated presses Starting from the north, in Zeugitania, at Asadi22 the Christian basilica was expanded in the second quarter of the 6th century and an olive press was placed in an adjacent room, which was probably abandoned after the Islamic conquest.23 Unfortunately here limited information on the place. precludes further comment on the precise topographical situation. Belalis Maior (fig. 8 no7) is similar; here an olive press was built in a room attached to a small building. This has been identified as a Christian cult structure, built in the 6th century. in the western part of the city.24 Around this building, a larger production and residential quarter was built, which was also in use during the Islamic period.25

20 Tortorella (1998) and Mackensen (1998).21 Spatheia are long and thin, which favors the storage and transport of more

estate. General synthesis on late amphora production: Keay (1998) and Panellaand Saguì (2001).

22 Modern Jedidiah, located about 8 km northeast of Segermes. See BA map32 Carthage F4

23 Bejaoui (1986) and Ben Abed et al. (1993) 491: the olive press adjoined the vestibule and the sacristy of the church. The olive press was certainly in use in the second phase of the basilica.

24 Nu Henchir el Fauar,: BA, cort 32 Carthago D3. Mahjoubi (1978) 246-54 og Mahjoubi (1983).

25 In this phase, another press was probably inserted in the sapse of the religious building: Mahjoubi (1978) 254 n. 674.

topographies of production in North African cities 265

Regarding the topography of the early Roman city, it can be suggested that the structure was located in a peripheral sector. The church building was built over an earlier structure, identified by Mahjoubi as a possible sanctuary, perhaps dedicated to Ba'al-Hammon-Saturn;26

The presence of an olive press, related to a church and located on the periphery of the city, has been recorded at Sufetula (fig.8 no10) in Byzacena. Here, adjacent to the Basilica of Saint Gervasius, Protasius and Tryphon (Basilica V), an olive press was built across the main street that crosses the town (Fig. 3). In this sector, during the Byzantine period, two fortresses and a late bath27 were also built (fig. 2). Duval has suggested that the installation of the press took place in a late phase, probably in the 7th century, due to the higher level of the floor.28 Here, too, the production activities were located next to the church inside a residential area and was protected by the presence of fortified structures.29

A connection between a Christian religious complex and an olive oil production site has also been recorded at Mactaris (fig.8 no9), in Byzacena. Here, a structure called the Basilica Juvenes, located near the ancient forum of the city, was converted into a church in the 4th century. In the Byzantine period, the church was partially restored, and at the same time an olive press was established in the old peristyle.30 It has been suggested that the complex was in use until the 11th century, when it was used for housing. The area is located in the center of the city, close to a Christian cemetery.31 Similar evidence has also been recorded at Lepcis Magna, where in the 6th-7th an olive

26 For a detailed study of the underlying complex and its identification, see Mahjoubi (1978) 246-51.

27 Today Sbeitla. BA map 33 Theveste-Hadrumetum D2. Duval (1964) 103 and (1999).

28 Duval (1999) 931. He suggests that this part of the city survived for a long time, probably until the 10th century.

29 For a general survey of Sufetula in the late phase see Duval (1964) and more recently Duval (1990) 512-14.

30 BA map 33 Theveste-Hadrumetum D1. Picard (1957) 130 and Bourgeois (1984-85) 188.

31 In the same area was a building marked "édifice Chatelain", probably a late edifice à auges from the 4th century. or later. Today it has been completely destroyed. See Picard (1955-56) 174 and Picard (1957) 137. For questions regarding edifices see Duval N. and Duval Y. (1972), Duval (1976), Duval (1979), and Nestori (1980-82).

anna leone266

Fig. 2. Plan of late Sufetula - from Duval (1990)

topographies of production in North African cities 267

Fig. 3. Plan of the church and the olive press – from Duval (1999)

anna leone268

press was annexed to the church in the old forum, within the Byzantine city wall.32

Badly dated presses Badly dated presses are mainly located in reused public buildings. At Lepcis Magna, a late olive press was placed in the gymnasium of the Hadrianic baths, near the Severan colonnade.33 At Sabratha, in the same province, an olive mill is recorded immediately within the Byzantine defensive wall.34 At Thurburbo Maius (fig. 8no 6) was The Capitol and the Temple of Mercury reused, and an entirely new manufacturing district has been recorded (Fig. 4). Evidence of habitation/manufacturing has been found along the eastern side of the Capitolium street, where rooms with large jars for storage have been found in the floor.35 In the basement of the Capitol and nearby, olive presses were found. The area appears to have been in use in the 7th century. and in the Islamic period; a chronological reference point is provided by a 7th century. coin hoard, containing early 7th century coins of Phocas and Heraclius. A late chronology is also confirmed by the presence of ARS pottery, decorated with Christian symbols.36 Islamic occupation of the site is indicated by some pottery fragments that can probably be dated to the 11th c.37. The description of the excavation is not very much

32 De Miro (1996) and (1997). The olive press was built against the eastern wall of the basilica. It was originally dated to the 5th century. A.D.—De Miro (1996)—but was then dated to the 6th century; it was possibly also in use in the 7th century - DeMiro (1997). For the pottery found and the dating evidence, see De Miro (1998)171.

33 Wilson (2001) n. 41.34 Wilson (1999).35 BA map 32 Carthago E4. Archaeological data are not clearly reported. Arabic

pottery from the 11th century A.D. seems to have been recorded in the filling of 12 silos in the ground to the southwest between the Forum and the Capitol, Poinssot and Lantier (1925) LXXVII.

36 The description is unclear, Poinssot and Lantier (1925) LXXIX: “...In the houses of the late period (located to the east on the other side of the rue du Capitole) the remains of the large Christian dishes with glazed red glaze have been found.and decorated with stamped motifs, ox, rooster, dove, cross sometimes enclosed in a heart, and a certain number of sherds, some decorated with cuts, others with reliefs of paintings...”. The coins were found in one house: Poinssot and Lantier(1925) LXXIX. Poinssot and Lantier (1925) LXXXIV suggests that these houses (de basse epoque) were occupied in the Byzantine period; on the other hand, near the Capitol, they seem to have found pottery from the Islamic period, suggesting occupation at this time. From the report it is difficult to define which structures were Byzantine and Islamic.

37 Poinssot and Lantier (1925), Maurin (1967) and Frend (1983). The dating of

topographies of production in North African cities 269

Fig. 4. Proverb Maius - Olive press in the Capitol

precisely, and it is not possible, for example, to understand when tea olive presses were put into the basement of the Capitol. Theoretically, they could belong to any period after the Capitol ceased to be used as a pagan temple.

Ceramic production

Important evidence also comes from pottery production. The two main dated production sites reuse large bath complexes for pottery kilns. They are located in two different provinces: the first producing fingernails at Uthina (Fig. 8 no. 4) in Zeugitania, not far from Carthage, and the second, at Leptiminus (Fig. 8 no. 12), producing amphorae, on coast of the province of Byzacena, close to the supposed capital

some of these presses are complicated because sites were often occupied or reoccupied after the Islamic conquest, as at the Bulla Regia. Here, in the Baths of Julia Memmia, are parts of a press dated to the 9th century. was recorded recently: Broise and Thébert (1993) 391.

anna leone270

Hadrumetum (modern Sousse). In Uthina, two different areas of pottery production have been identified: one in the recycled baths of Laberii and another on the periphery of the city; both kilns produced African red pottery and were in use from the second half of the 5th century. until the end of the 7th century A.D.38 (Fig. 5-6).

Leptiminus (fig. 7) was already an important production center in Roman times. Kilns producing "African Grande" amphorae have been recorded throughout the city, and the production of African red pottery has been suggested.39 Recent excavations have recorded evidence of industrial activities in the reused East Baths in the 6th and 7th centuries. century. At this time, the structure housed a large production site with activities including butchery, metalworking and glass production;40 the most significant evidence is for amphora production, with forms Keay LXI, VIIIA, Keay LXII A and E produced here, the latter probably being one of the latest produced in North Africa following the chronology proposed by Keay.41 Re-use of baths for pottery production has also been recorded in Carthage in the Antonine Baths.42 It is clear that buildings of this type were particularly adapted to such a function as they were strong structures with numerous tanks, and the insertion of furnaces was facilitated by the presence of fireproof building materials in the heating system.

It is also significant that the two examples of new production sites mentioned here are both located along or close to the coast. The Antonine Baths at Carthage (Fig. 8 No. 1) are close to the harbor front

38 BA 32 Carthage F2. According to Barraud et al. (1998) 146. Main forms made in LaberiiBaths: Hayes 96,97,99B,98,99A: Mackensen (1998).

39 BA 33 Theveste-Hadrumetum G1; Mattingly et al. (2000).40 This evidence and the material found in the excavation suggest a very

organized place of production, perhaps state-controlled. The complex on the north side of the circular harbor of Carthage has been similarly interpreted; this was in use throughout the Byzantine period; imperial control has been suggested here - Hurst(1994) 68 and 93.

41 For the pottery from the excavation see Dore (2001). On these amphora forms see generally Keay (1998). Their presence may indicate olive oil production around the city. Keay has suggested that they were used to transport oil; but there is no direct evidence. They occur most frequently on the Spanish coast, but are common in Italy and Gaul.

42 Lézine (1968) 67-73, see especially 69. Unfortunately, we do not have detailed information on the pottery types produced in this part of the city. A pottery kiln was also put into a bath at Tiddis at some point in the Middle Ages—Lassus(1959) 191.

topographies of production in North African cities 271

Fig.

5.

Pla

n off

Uth

if-

from

Bar

iron

et

al.

(199

8)

anna leone272

Fig. 6. Uthina—Baths of Laberii with location of the pottery kilns—from Barraudet al. (1998)

topographies of production in North African cities 273

and the East Baths in Leptiminus are close to the ports.43 Additional information from both cities also seems to support a pattern of association. In Leptiminus, Stirling,44, anticipating data from the field survey, they noted that Roman ovens ringed the outskirts of Leptiminus; but from Vandal and especially Byzantine times they were mainly located along the coast.45 A similar trend has also been recorded in Carthage, where an increase in the number of pottery kilns found along the coast is recorded at the end of the 6th/7th century. century. A.D.: pottery kilns were found in the vaulted warehouses and on the quay of the rectangular harbor46 and also on the island in the middle of the circular harbor.47 The presence of kilns here was closely associated with residential areas. The general situation recorded in the port area of ​​Carthage continued in the Islamic period. Lepcis Magna provides similar evidence for the early Islamic period. At this point, a pottery kiln is connected to houses from the 9th-10th century. century have been recorded in the Flavian temple, inside the Byzantine wall, near the coast48, and additional ovens associated with houses from the same period have been found in the city's port. 49

Calcareous

One more class of structures must be considered: lime kilns. These represent production as they were used to make lime for mortar and plaster used in construction work. Lime kilns are taken up in both public and private complexes. They have mainly been assigned to the Byzantine and Islamic periods, although they are

43 At Uthina, the only site in the country with useful information, pottery kilns were mainly located along a main road.

44 Stirling (2001) 68-69.45 This evidence is only from coastal towns; that inside the country looks different: Pea-

cock et al. (1989), Peacock et al. (1990).46 Humphrey (1980) 99 notes the similarity of changes in the two ports;

both were invaded by graves.47 Hurst (1992) 88-89 and Humphrey (1980) 98. It is not determined whether

the ovens date to the 6th or 7th century A.D.48 Fiandra (1974-75).49 Information provided by M. Munzi and E. Cirelli at the seminar "Transizione

of the Cities of North Africa between Late Antiquity and the Early Middle Ages” held at the University of Siena, 17 March 2001. See also General Communication in Fiandra (1997) 251 and in Laronde(1994) 997.

anna leone274

not well dated. A good example comes from Sabratha in Tripolitania, where a domestic quarter was built around the Temple of the Unknown Divinity, probably in Byzantine times. Lime kilns have been found in the northwestern corner of the temple area.50 The temple was probably abandoned after an earthquake in 365 AD, as evidenced by archaeological evidence. and later reoccupied. At Leptiminus (fig. 8 no. 12) in Byzacena, a large number of lime kilns were built over part of the cemetery at site 10, probably in Late Antiquity or Islamic times.51 Another kiln was built in Bulla Regia (fig. 8 no. 2) in the 'Maisonde la Chasse', although the chronology is uncertain.52 Similar evidence has also been recorded in the forum of Uchi Maius (fig. 8 no. 3),53 where a lime kiln was established sometime in the 6th century. The forum here had already been transformed into an olive oil production area in the Vandal period (see above), and had thus probably lost its function as a public space (fig.1). Finally, lime kilns were recorded at Bararus (Fig. 8 no. 11), again in the city's forum, probably dating to the Islamic period; as previously mentioned, this sector was inhabited until the 8th-9th century. century. A.D.54 Overall, the only important spatial factors affecting the location of lime kilns seem to be the presence of inhabited areas, although they often favor the city center, close to or in imposing buildings; both public and private structures were invaded by them.

Patterns of placement and possible explanations

This article aimed to explore changes in the location of manufacturing activities and suggest possible reasons for them. In the introductory section, significant problems were identified as the intra-urban/peripheral location of craft activities, their proximity to residential areas and the nature of their distribution in cities: either structured

50 Joly and Tommasello (1984) 8. Late graves were also found in the same area.51 Ben Lazreg and Mattingly (1992) 177, fig. 1 179, 190, 195, 201.52 This information was provided by Y. Thébert, whom I thank. He suggested

that the oven could be Islamic. The site is located in the Medjerda valley, near modern Jendouba: BA map 32 Carthago C3.

53 Gelichi og Milanese (1998), (1995); Vismara (1999) 74. Uchi Maius er nu Henchir ed Douimas: BA map 32 Carthago D4.

54 Guéry (1981) 99. Insertion of lime kilns is not definitely dated. Now HenchirRougga, in Byzacena: BA map 33 Theveste-Hadrumetum G2.

topographies of production in North African cities 275

or randomly. It has not been possible to examine all production activities in these terms. Garum and murex staining activities did not provide much useful information; however, olive presses, pottery kilns and lime kilns provided enough data to allow some patterns of placement/association to be suggested.

First, there was a tendency for olive oil production, especially in the later Byzantine period, to occur interspersed with residences within the city walls, often close to a fortified complex and close to a church, to which they were often annexed. Second, pottery kilns were found in coastal cities, where they were often located along the coast, although only clearly recorded at Leptiminus and possibly at Carthage. Pottery kilns also sometimes reused bathhouses. Finally, lime kilns were associated with late inhabited areas and appear to be distributed in both public and private buildings in city centers or imposing structures elsewhere.

In proposing possible explanations for the distribution of production activities in non-traditional locations, there seem to be three main alternatives. The first is that they represent an orderly development of the late Roman city, with the pragmatic adaptation of redundant public buildings to house productive activities, positioned in relation to traditional commercial pressures; the other alternative is that new sites for craft production relate to new priorities and planned changes in urban centres. The third is that they represent a phase of dissolution in the structure and function of the urban settlement. In this process, they may have taken on some of the roles and characteristics of rural sites, such as processing agricultural products.

Ordered Evolution

The first alternative seems to relate best to the Vandal and Byzantine periods. During one or both of these periods, the traditional manufacturing activities of pottery, and possibly fabric dyeing and garum production, continued to be present in the cities as they had been for several hundred years. It is very likely that these activities were still governed by the same commercial local pressures that shaped them during the Roman period. In this regard, the possible connection between pottery kilns and the coast in the Byzantine period should be mentioned; this could relate to the necessity of maintaining

anna leone276

easy communication and possibly also the tendency to bottle goods just before export.55 The use of bathhouses for furnaces from the Byzantine period onwards is also interesting. These buildings offered ready-made fireproof structures and could be adapted pragmatically. Artisans may have chosen these sites because privatization of a large number of public buildings and public spaces provided many advantageous sites that could be exploited.

The hypothesis of a privatization of public space is based on the connection between the beginning of the transformation of public areas in the classical cities into private spaces and the occupation of Tunisia by the vandals.56 The occupation of the forum of Uchi Maius with housing at this time is not an isolated case.57 Unfortunately, dated archaeological evidence is rare, and it can only be said that the phenomenon of private occupation of public lands has been recorded at various sites in Tunisia from the 5th century. forward. For example, the forum of Belalis Maior58 was transformed into a residential area; similar evidence is recorded along the cardo maximus at Carthage59 or in the forum at Bararus.60 Many traditional public buildings were abandoned and destroyed, then converted to serve three different main functions: as areas of private residence, as funerary complexes, such as the theater at Carthage,61 or as production sites,

55 It is unlikely that the location close to the sea coast is related to the need to obtain salt water for pottery production, as this has been equally important in the early Roman period. The importance of salt water in pottery making in North Africa has been suggested by Peacock (1984b) 263–64.

56 The privatization of public space provides a legal context for the establishment of production sites within Late Antiquity. Several studies related to the 'privatization' of public buildings have been produced recently in an attempt to reconstruct the stages of this phenomenon. Book 15 of the Codex Theodosianus provides many legal references to the control of public structures: Janvier (1969). For a general synthesis see Saradi Mendelovici (1990). For a compilation of the sources: Cantino Wataghin (1999) 735. Unfortunately we have no evidence for the time of the Vandals.

57 A residential quarter also developed around the area of ​​the forum: Vismara(1999) 73-74.

58 Mahjoubi (1978) 252. See Thébert (1983) 111.59 Deneauve and Villedieu (1977) 98; Deneauve and Villedieu (1979) 151: re-

is dated to the 5th, 6th or 7th century AD. Not all public buildings in Carthage lost their original function. We know, for example, that while the theater and the bath were abandoned, the circus and amphitheater continued to be in use.

60 Guéry (1981) 97: evidence of a floor in a possible house dated between the 4th or 5th century. has been admitted to the forum. So at the end of the 6th century. a house was built against the front hall and sometime in the 7th century a fortress was built.

61 Destruction after the capture of the Vandals was followed by poor housing, then graves: Picard and Baillon (1992) 13.

topographies of production in North African cities 277

as in the case of Uchi Maius. It is difficult to reconstruct the process of transformation of these areas/structures during the vandal period. It is possible that there was an ordered transfer of public lands to private ownership as part of the Vandal state's reorganization of land tenure, characterized by extensive confiscations and redistributions,62 which took place in the 5th century. Alternatively (and more likely), buildings may simply have been abandoned and their ownership usurped. Unfortunately, textual sources do not allow us to clarify this. In any case, a large number of potential buildings became available in an already existing and still densely inhabited urban structure. It is therefore likely that at least some of the unusual sites of craft production relate to the pragmatic adaptation of old urban environments to new priorities; however, this does not necessarily signal any kind of economic change in the roles of cities.

En the byorganization?

Other aspects of the productive evidence cannot be explained in terms of a pragmatic development of the classical city, but suggest that elements of a new urban organization were developing. The connection between religious complexes and olive oil production structures, often adjacent, falls into this category, especially from the end of the 6th century. AD This could indicate that there was a conscious connection between the press and the church. Recent studies have shown a connection between productive centers (for ceramics, glass, bricks, etc.) and churches in different parts of the Roman Empire, such as Torcello, Thebes (Teassaglia), Florence, Cornus (Sardinia), Cari´in Grad (Serbia). 63

62 Courtois (1955) 278 suggests that these redistributions were widespread. In the Vita Fulgentii (1) it is specifically stated that major confiscations and reorganizations took place in Zeugitania. However, Procopius (Vand. 2.6.9) makes no geographical distinction, but (Vand. 1.5.12) clearly attests that all Vandal properties are exempt from taxes and reports that Geiseric destroyed all documents relating to tax collection that had been in use in late Roman times (Vand. 2.8.25). It is not clear from this whether tax payments continued under the Vandals or not; Courtois(1955) 258 suggests that they probably continued. We do not know whether production activities, located in former public buildings/areas, were state-controlled or not, or whether a state-controlled production system survived into this century. Archaeological evidence in one case seems to indicate the absence of a centralized organization, such as the circular port of Carthage, a focal point for the state's oil collection and redistribution from at least the 4th century. A.D.-Peña (1998) - was abandoned.

63 For a detailed collection of all these data see Martorelli (1999).

anna leone278

This evidence cannot rule out the idea that the church played a significant role in the organization of both production activities and public space, at least in the later Byzantine period, when this association seems to have been more common. This connection between the church and the olive oil presses could be interpreted as an indication that olive oil production took place in a well-regulated manner during the Byzantine period. The association of Byzantine presses with not only churches but fortresses also suggests that, at least in this period, production may have been deliberately regulated in urban settings by political or religious authorities. Governmental regulation of artisanal activity is also suggested by the well-organized production centers recorded at Carthage and Leptiminus. It is possible that a significant amount of olive oil was still destined for export. Furthermore, it has also been suggested that the movement of olive presses in urban areas was probably linked to an increased necessity to control oil production in rural areas.64 Only further work, regarding the extent and organization of production in rural areas and on oil amphorae exported from Africa to other regions during this period will help advance this question further.

Urban decay

So far it has been suggested that changes in production location can be related to fairly structured processes. However, at least some of the evidence could also be seen as reflecting disintegration in the structure and function of urban settlements. The placement of lime kilns in public and private structures in city centers, their lack of respect for the traditional urban framework of antiquity and their association with residential areas suggest that they were associated with the spoliation (burning marble for lime) and indeed the destruction of the monumental fabric of African cities for primarily residential construction, not significant public or private projects. Unfortunately, imprecise dating does not allow us to assign a larger number of lime kilns to either the Byzantine or Islamic periods, which prevents further analysis; Still, lime kilns appear to represent positive evidence of decay in the urban fabric.

The location of plants for the processing of agricultural products in urban areas, mainly from the end of the 6th century, was

64 Roskams (1994) 4.

topographies of production in North African cities 279

not very common before Late Antiquity and seems to be at least partly related to changes in the function of urban settlements. Nevertheless, it seems that some presses, especially those well dated to the Byzantine period, can be associated with churches. Later, in the 8th/9th century it seems that cities were still occupied by olive oil presses, spread over urban areas, as for example at Thurburbo Maius or Belalis Maior. Recently, there have been e.g. found a screw press in use during the Islamic period in the so-called 'Carthagena Basilica' in Carthage.65 Two other elements belonging to the same type of press were found on Mount Byrsa in the Antonine Baths, suggesting that olive oil was perhaps also produced in these parts of Carthage during the Islamic period. This evidence supports a possible less orderly view of olive oil production, with household units in the Arab period. As such, the location of these presses could be interpreted as indicating a progressive convergence of the roles of villages and towns. It is also possible that cultivation zones developed in urban areas. In Carthage, archaeological evidence of cultivated sectors in the Byzantine period on the outskirts of the city66 seems to suggest the existence of small sectors, rather than extensive cultivation, probably for the self-sufficiency of the inhabitants.67

Major changes in the urban structure can also be detected from the later Byzantine period. Cities now seem to have been reduced to small inhabited nuclei characterized by the presence of a church and a place of production, often associated with a fortified complex.68

65 Ennabli L. (2000) 129-30. The screw press is unknown until this date in Tunisia and Tripolitania, although it is possible that screw presses may have been used before the Islamic period; in Morocco they are recorded from the 3rd century A.D. It is very difficult to reconstruct the organization of Islamic Carthage, mainly due to the loss of evidence caused by clearance excavators in the earlier 20th century. Islamic evidence for the city has been compiled by Vitelli (1981).

66 e.g.: see Av. Bourguiba Salamboo neighborhood, on the southern periphery of the city, close to the city wall, where a production complex, probably related to a house, was built in the early 7th century: Hurst and Roskams (1984).

67 The presence of cultivated plots in Rome has recently been recorded in excavations of the imperial forums. See generally and for further bibliography: Meneghini (2000a), (2000b), Santangeli Valenzani (2000).

68 This evidence has been clearly described for Sufetula—Duval (1990) 512–13. The cores may include horrea as in the case, for example, of the Basilica Juvenes inMactar. We have no evidence to determine the official status of such sites, thought they appear to be structured; it is possible that they were meeting places and occasional places of exchange.

in

already

leo

it is

280

Fig. 7. Plan of Leptiminus - from Ben Lazreg and Mattingly (1992). For fig. 8 see plates.

topographies of production in North African cities 281

phenomenon is recorded especially from the later Byzantine periods, attested by the case of Sbeitla (fig.2) and probably continued in the early Islamic period; it can perhaps be explained as a consequence of uncertainty and the desire to protect both machines and goods produced. It is also possible that the location of pottery kilns along the coast at Leptiminus and Carthage may also be due to the manufacturers' desire to be close to the fortified enclosure, as the city is contracted to a narrow zone along the coast (fig.7).69 Kl. Lepcis Magnaearly Islamic kilns have been recorded in the Flavian temple, located within the Byzantine wall close to the coast.70 At both Carthage, Lepcis and possibly in Leptiminus, the location of the pottery kilns along the coast also corresponds to the presence of inhabited sectors in the same zones, which suggests that production and residential areas were gradually mixed together, both located along the coasts and the landing sites.

Conclusions

The evidence considered here has shown that there are patterns of shifting production activities in Late Antique urban landscapes. Nevertheless, two points must be taken into account when assessing their importance for late antique urbanism. First, although similar urban phenomena were experienced in many parts of the Mediterranean and Late Antiquity, the chronologies of these changes vary greatly and are best considered primarily in a regional or subregional context.71 Second, it must be emphasized that individual cities often experienced quite different processes of destructuring and restructuring; this limits the extent to which generalizations can be made. Thus, the comments here relate exclusively to Tunisia and

69 It is not possible to exclude this aspect (as the evidence of the relics probably being transferred from the Basilica Maiorum - located in the northern suburb of the city - in the 7th century to the southern part of the city to the church of Bir El Knissia seems to suggest—Stevens (1995)—was associated with the presence of an important inhabited area exhibited around the ports in this century Destruction in the 19th/20th century, especially in the southern part of Carthage, does not allow further data acquisition to assess this evidence.

70 Fiandra (1974-75). See above on 'Pottery Kilns'. Here, too, the Byzantine city was located along the coast, by the ports.

71 Gelichi og Milanese (1998) 478.

give leone282

Tripolitania, and is provisionally advanced, and only as far as the fragmentary and poorly dated archaeological evidence allows. Here, further detailed investigation and excavation work is needed to advance the subject. Studies in other regions, particularly Asia Minor and the Levant, where late antique and early medieval urban landscapes are relatively well preserved, would also help to contextualize the African evidence.

Because of these limitations, it is difficult to draw firm conclusions. However, the North African evidence suggests that the establishment of production sites in unusual settings should not be universally interpreted as part of urban decay. The location of production centers may still relate to the practical commercial choices made by artisans over several hundred years. Others may reflect changes in the political organization of production, as may be seen in the connection between olive presses and churches, or desire for security, as seen in their proximity to fortifications.

Recognitions

I would like to thank L. Lavan, L. Spera, C. Panella, D. Palombi, M. Brizzi for our discussions. Thanks are also due to D. J. Mattingly and E. Fentress for their suggestions and to N. Duval, J. Dore, S. Fontana, and A. Wilson for their comments on draft copies. A. Wilson and L. Lavan also kindly helped me with the editing of the English text.

topographies of production in North African cities 283

Bibliography

Barraud D., Bonifay M., Dridi F., Pichonneau J.F. (1998) "The Ceramic Industry of Late Antiquity", in Oudhna (Uthina). The rediscovery of an ancient city in Tunisia, ed. H. Ben Hassen, L. Maurin (Bordeaux, Paris and Tunis 1998)139-67

The Barrington Atlas of the Greek and Roman World, red. R. J. A. Talbert (Princeton 2001)

Bejaoui F. (1986) "Discoveries of Christian Archeology in Tunisia", in Proceedings of the XI International Congress of Christian Archeology (September 1986) vol 2 (Lyon, Vienne, Grenoble, Geneva, Aosta and Rome 1989) 1927-60

Ben Abed A., Bonifay M., Fixot M. (1993) "The (Tunisian) Jedidiah", MEFRA 105.1 (1993)486-92

Ben Lazreg N. and Mattingly D. (1992) Leptiminus (Lamta): a Roman port city in Tunisia. Report No. 1, (JRA Supplementary Volume 4) (Ann Arbor, Michigan 1992)

Ben Lazreg N., Bonifay M., Drine A., Trousset P. (1995) "Production and Marketing of Salsa Cements from Ancient Africa", in Ancient and Medieval North Africa. VI Coll. int. African Production and Exports (Paris 1995) 103-132

Blanc A.C. (1958) "Remains of purple wares at Leptis Magna and MonteCirceo", in the Roman port of Leptis Magna. (Bulletin for the Study Center for Architectural History 13), ed. R. Bartoccini (Rome 1958) 187-210

Broise H. and Thébert Y. (1993). Franco-Tunisian Archaeological Research at BullaRegia. II. The architectures. 1. The Memmian thermal baths, (Collection of the French School in Rome 28.2) (Rome 1993)

Bourgeois C. (1984-85) "The wells of the House of Venus and another well at Mactar", AntAfr 21 (1984-85) 179-81

Cantino Wataghin G. (1999) "...Ut haec Christo Domino in ecclesiam consecratur. The Christian reuse of ancient buildings between Late Antiquity and the Middle Ages", Study Weeks of the Italian Center for the Study of the High Middle Ages XLVI. Ideologies and Practices of Reemployment in the Early Middle Ages, vol 2 (Spoleto 1999) 673-751.

Courtois Ch.(1955) Les Vandales et l'Afrique (Paris 1955)Curtis R. I. (1984) "The salted fish industry of Pompeii", Archaeology 37.6 (1984)

58-59, 74-75——— (1991) Garum and Salsamenta. Production and trade in Materia Medica (studies

on Ancient Medicine 3) (Leiden 1991) De Miro E. (1996) “Preliminary report on the results of the archaeological mission

fra University of Messina i Leptis Magna", LibAnt n.s. 2 (1996) 199——— (1997) "Missione archeologica dell'Università di Messina a Leptis Magna

1996", LibAnt n.s.3 (1997) 246-47——— (1998) "Missione archeologica dell'Università di Messina a Leptis Magna",

LibAnt n.s. 4 (1998) 170-71 Deneauve J. og Villedieu F. (1977) “The Cardo Maximus and the buildings located

east of the road", AntAfr 11: 95-130——— (1979) "Cardo Maximus and the buildings east of the road", in BYRSA I.

Reports preliminary des fouilles, ed. S. Lancel (Rome 1979) 143-76Dore J. (2001) "The large ceramic deposits after the abandonment of the East Baths",

in Stirling et al. (2001a) 75-98Drine A.2000 “The excavations of Meninx. Results of the campaigns in 1997 and 1998",

Romanian Africa 13 (2000) 87-94

anna leone284

Duval N. (1964) "Observations on the Late Town Planning of Sufetula", CahTun 12 (1964)87-103

——— (1976) “Again the trug monuments of Africa: Tébessa, Khalia, Hr.Faraoun”, MEFRA 88 (1976) 929-59

——— (1979) “Form and Identification. Questions about method. On Trough Monuments and Triconches in North Africa”, MEFRA 91 (1979) 1015-19

——— (1990) “Sufetula: the history of a Roman city in the upper steppe in the light of recent research”, Africa in the Roman West (1st century BC-4th century AD) (Roma 1990 ) 495-526

——— (1999) "Church V (of Saints Gervais-Protais and Tryphon) in Sbeitla (Sufetula) Tunisia", MEFRA 111.2 (1999) 927-89

Duval N. and Duval Y., "False basilicas (and false martyrs): some trough buildings in Africa", MEFRA 84 (1972) 675-719

Duval N. and Lézine A. (1959) "Christian necropolis and underground baptism at Carthage", CahArch 10 (1959) 71-147

Ennabli A. and Neuru L. (1994a) "Excavation of the Roman Villa at Sidi Ghrib, Tunisia 1985-1992", Echos du Monde Classique/Classical Views 38, n.s. 13 (1994) 207-20

——— (1994b) "Recent Excavations at Sidi Ghrib: 1994", Archaeological News 19.1-4 (1994) 28-30

Ennabli L. (2000) The Basilica of Carthagena and the Site of the Seven Monks of Gafsa. New Christian buildings in Carthage (Paris 2000)

Fentress E. (2000) "The Djerba Survey: Settlement in the Punic and the Romanperiod", Africa Romana 13 (2000) 73-85

Fiandra E. (1974-75) “The Ruins of the Flavian Temple of Leptis Magna. 1) Vicende from the 4th to the 9th century AD, LibAnt 11-12 (1974-75) 147-50

——— (1997) "Italo-Libyan Archaeological Mission "Tempio Flavio" in Leptis Magna. Report on the activity carried out in 1996", LibAnt n.s.3 (1997) 249-51

Fontana S. (2000) "A 6th-century refuse site from Meninx: the end of purple production and material culture in Djerba in the early Byzantine period", Afrika Romana 13 (2000) 95-114

Frankel R. (1999) Wine and Oil Production in Antiquity in Israel and Other Mediterranean Countries (Sheffield 1999)

Frend W. H. (1983) "The end of Byzantine North Africa: some evidence of transition", in Proceedings of the II International Conference on the History and Archeology of North Africa (BCTH 19) (Grenoble 1983) 387-97

Gelichi S. and Milanese M. (1995) “Uchi Maius: the citadel and the forum. Preliminary Report on the 1995 Excavation Campaign', in Uchi Maius I. Excavations and Epigraphic Research in Tunisia, eds M. Khanoussi, A. Mastino (Sassari 1995) 49-82

——— (1998) "Problems in the Transition to the Middle Ages in Ifriqya: First Results from the Archaeological Excavations at Uchi Maius (Teboursouk, Beja)", Africa Romana 12 (1998) 457-86.

Giardina A. (1999) Esplosione di Tardoantico (Como 1999). Guéry R. (1981) "The occupation of Rougga (Bararus) according to the stratigraphy of

forum", BCTH 17B (1981) 91-99Humphrey J. H. (1980) "Vandal and Byzantine Carthage: some new archaeologi-

cal evidence”, i. (red.), New Light on Ancient Carthage, red. Griffith Pedley J. (AnnArbor, Michigan 1980) 85-120

Humphrey J.H., Oleson J.P., Sherwood A.N. (1998) Greek and Roman Technology:a Sourcebook (London-New York 1998)

topographies of production in North African cities 285

Hurst H. (1992) "Admiralty Island, the Circular Harbor and Avenue Bourguiba", in To save Carthage, (UNESCO/INAA) (Tunis 1992) 79-96

——— (1994) Excavations at Carthage. The British Mission. II.1. The Circular Harbour, North Site. The site and finds other than pottery (Oxford 1994)

Janvier Y. (1969), Legislation of the Lower Roman Empire on Public Buildings (Aix enProvence 1969)

Joly E. and Tommasello F. (1984) The Temple of an Unknown Divinity of Sabratha (Monographies of Libyan Archeology 18) (Rome 1984)

Keay S. (1984) Late Roman Amphorae in the Western Mediterranean. A typology and economic study. The Catalan Evidence, 2 vols (BAR Int. Ser. 196) (Oxford 1984)

——— (1998) “African Amphorae”, in Saguì (1998) 141-55Laronde A. (1994) “New Archaeological Research in the Port of Lepcis

Magna", CRAI (1994) 991-1006Lassus J. (1959) "Chronicle: Punic, Roman and Christian Archeology in 1957",

RAfr 103 (1959) 169-92 Lézine A. (1968) Carthage—Utique. Studies in Architecture and Urban Planning (Paris 1968) Mackensen M. (1998) “Centres of African Red Slip Ware in Tunisia from the late 5th

to the 7th century AD”, in Saguì (1998) 23-39Majeske B. J. B. (1972) Bakeries, bakers and bread in Pompeii. A study in social and

Economic History (Ann Arbor, Michigan 1972) Mahjoubi A. (1978) Research in History and Archeology at Henchir el Faouar: the City

Belalitani Maiores (Tunis 1978)——— (1983) “The City of Belalitan Maiores. Example of duration and trans-

formation d'urbanisme antique", Afrika Romana 1 (1983) 63-71 Martorelli A. (1999) "Reflections on productive activities in late antiquity and

early middle ages: is there an ecclesiastical craft? ”, RACrist 75.1-2 (1999)571-96

Mattingly D. J., Stone D., Stirling L., Ben Lazreg N. (2000) “Leptiminus (Tuni-sia). En producentby?”, i Economies beyond Agriculture in the Ancient World, red. Mattingly D. J., Salmon J. (London 2000) 66-89

Maurin L. (1967) "Thuburbo Majus and the Peace Vandal", CahTun 15 (1967) 225-54

Meneghini R. (2000a) “Rome. Early Medieval Structures and Urban Layout between the Seventh and Eighth Regions", Medieval Archeology 27 (2000) 303-10

——— (2000b) “The Imperial Forums of the Post-Classical Age: The Forums of Nerva, Cesare and Pace”, in Crypta Balbi. Imperial forums. Urban Archeology in Rome and the Restoration of the Great Jubilee, ed. S. Baiani, M. Ghilardi (Rome 2000) 79-82

Morel J. P. (1987) "The topography of crafts and commerce in Ancient Rome", in L'URBS. Urban space and history (1st century BC-3rd century AD). (Rome 1987) 127-55

Nestori A. (1980-82), "A hypothesis about African buildings "à auges", RendPontAc 53-54 (1980/82) 313-23.

Notitia dignitatum, ed. O. Seeck (Berlin 1876) Panella C. and Saguì L. (2001) “Consumption and Production in Late Antiquity Rome

and Early Middle Ages: goods and contexts.

Peacock D. P. S. (1984a) "The amphorae: typology and chronology", i The Av.du President Habib Bourguiba, Salambo: the Pottery and Other Ceramic Objects from theSite, red. M.G. Fulford, D.P.S. Peacock (Sheffield 1984) 116-40.

Peacock D. P. S. (1984b) "Seawater, salt and pottery", in The Av. you President Habib

anna leone286

Bourguiba, Salambo: Pottery and Other Ceramic Objects from the Site, ed. M.G.Fulford, D.P.S. Peacock (Sheffield 1984) 263-64

Peacock D. P. S., Bejaoui F., Ben Lazreg N. (1989) "Roman amphora production in the Sahel region of Tunisia", in Roman amphorae and economic history: ten years of research (Rome 1989) 179-222

Peacock D. P. S., Bejaoui F., Ben Lazreg N. (1990), "Roman Pottery Production in Central Tunisia", JRA 3 (1990) 59-84

Peña J. T. (1998) "The mobilization of state olive oil in Roman North Africa: the evidence from the late 4th century. ostraca from Carthage", in Carthage Papers. The economy of the early colony, water supply, a public bath and the mobilization of state olive oil. edd. J. T. Peña, J. J. Rossiter, A. I. Wilson, C. Wells, M. Carroll, J. Freed & D. Godden (JRA Supplementary Volume 28) (Portsmouth, RhodeIsland 1998) 117-238

Picard G. Ch. (1955-56) "Session of the North Africa Commission 13. februar 1956", i BCTH (1955-56) 172-85.

——— (1957) "Mactaritan Civilization", i Carthago 8 (1957) 1-156. Picard G.Ch., Baillon M. (1992) "Le théâtre romaine de Carthage", i Actes du Ve

international conference on the history and archeology of North Africa. Shows, harbor life, religions (Paris 1990) 11-27

Pliny H. N. (1955) Pliny the Elder, Natural History IX, E. De Saint-Denis trans. (Paris 1955)

Poinssot L. og Lantier R. (1925) "Session of the Commission for North Africa 10. February 1925", i BCTH (1925) LXXI-LXXXVIII.

Potter T. W. (1995) Towns in Late Antiquity: Iol Caesarea and its Context (Sheffield 1995). Procopius, Vand. Procope, La guerre contre les vandales, D. Roques transl. (Paris 1990) Rakob F. (1991) "Die römischen Bauperioden", in J. Holst, Th. Krauss, M.

Mackensen, F. Rakob, K. Rheidt, O. Teschauer, M. Vegas, I. Wiblé, A. Wolff, Carthage I. The German Excavations at Carthage (Mainz 1991) 242-51.

Romanelli P. (1970) Topography of Roman Africa (Encyclopedia of Ancient Art) (Rom1970).

Roskams S. (1994) "The urban transition in the Maghreb", in Early Medieval Cities in the Western Mediterranean. Documenti di archeologia 10, ed. GP Brogiolo (Ravello1994) 43-54

Saguì L. (1998) Pottery in Italy: VI-VII Centuries. Proceedings of the Conference in Honor of John Hayes (Rome 1995, Florence 1998).

Santangeli Valenzani R. (2000) "The Imperial Forums of the Post-Classical Age: The Forums of NervaCesares and Pace", in Crypta Balbi. Imperial forums. Urban archeology and restoration interventions in the great jubilee, ed. S. Baiani, M. Ghilardi (Rome 2000) 82-89

Saradi Mendelovici H. (1990) "Christian Attitudes to Pagan Monuments in Late Antiquity and Their Legacy in Later Byzantine Centuries", DOP 44 (1990)47-61.

Sternmberg M. (2000) "Data on products manufactured in a pharmacy in Neapolis (Nabeul, Tunisia)", MEFRA 112.1 (2000) 135-53.

Stirling L., med assistance fra Mattingly D. J., Ben Lazreg N. (2001) "The EastBaths and their industrial re-use in Late Antiquity: 1992 excavations", in Leptiminus (Lamta): Rapport nr. 2, The East Baths, Venus Mosaic, Cemeteries and otherstudies, (JRA Supplementary Volume 40), red. Stirling L., Mattingly D.J. og Ben Lazreg N. (Portsmouth, Rhode Island 2001) 29-74

Thébert Y. (1983) "Urban Development in the Eastern Provinces of Late Roman Africa", OPUS 2 (1983) 99-130

topographies of production in North African cities 287

Tortorella S. (1998) "The African sigillata in Italy in the sixth and seventh centuries AD: problems of chronology and distribution", in Saguì (1998) 43-69

Vismara C. (1999) “Aetas succedit Aetati. Reuse of Classical Elements in African Cities up to the Islamic Age', in Acts X Day for Archaeology. The Past Reimagined Continuity and Recovery from Antiquity to the Present (Genova 1999) 69-85.

Life of Fulgent (1929) Life of Saint Fulgence of Ruspe, oversat. Lapeyre G.S. (Paris 1929)

Vitelli C. (1981) Islamic Carthage. The archaeological, historical and ceramic. Evidence (CEDAC. Dossier 2) (Tunis 1981).

Wilson A. I. (1999) "Commerce and industry in Roman Sabratha", LibStud 30 (1999) 29-52.

——— (2001) "Urban economys of late antique Cyrenaica", i Economy and Exchangein the East Mediterranean during Late Antiquity, red. S. Kingsley, M. Decker (Oxford2001) 28-43.

richard bayliss288

ARCHAEOLOGICAL SURVEY AND VISUALIZATION: THE VIEW FROM BYZANTIUM

RICHARD BAYLISS

Abstract

This paper will explore the relationship between the methods used in archeology to record buildings and landscapes and the development of computer visualization technology, with specific reference to Late Antiquity and Byzantine archaeology. The main objective is to highlight the usefulness of visualization technologies as the critical solutions for the presentation and examination of survey data and to mutually demonstrate that meaningful computer visualizations benefit from a secure underpinning of archaeological survey data. My intention is to show that computer visualizations should not be seen exclusively as a tool for archaeological reconstruction, but rather can be seen as an integral part of the interpretation process. The first of two case studies is the Anastasian Wall Project, a study of massive linear fortification in Turkish Thrace, which presented a number of practical problems both in terms of data capture and representation. The applicability of computer-based techniques for the integration and visualization of various types of survey data will be further explored through the study of Alacami, a multi-period Byzantine church in Cilicia, before a final commentary on the current and potential roles of computer visualization in Byzantine archaeology.

Introduction: The ethics of visualization

For most of us, our knowledge of the role of computer visualization in archeology has been formulated by exposure to the high-budget animated sequences that are now an essential part of any respectable TV documentary. The sterile, polygonal and uninhabited computer-generated scenes that audiences experienced in the early 1990s have been replaced by deeply textured ancient environments occupied by realistic (or at least mobile) characters. This

L. Lavan og W. Bowden (red.) Theory and Practice in Late Antique Archaeology (Late Antique Archaeology 1 2003) (Leiden), s. 288-313

archaeological investigation and visualization: Byzantium 289

has had a huge impact on public perception and expectations of archeology in general.

The vital link between archaeological interpretation and the representation of an archaeological site is therefore increasingly under threat as graphic ideals take precedence over historical reality. The pursuit of a perceived reality in computer visualization has also inevitably led to further compromises in the historical and archaeological integrity of computer visualization methods. Likewise, computer visualization in archeology has developed at a pace that has effectively outpaced considerable academic debate about the ethical and theoretical basis for its utility.1 The amazing capabilities of computer reconstruction and visualization techniques and the speed of their development have therefore initiated something of an abandonment of conscience over questions of authenticity and ethics about the final representation. In Alan Sorrell's famously sinister illustrations of reconstructed landscapes, areas of uncertainty were obscured with either thick smog or a deeper gloom. Attempts to mediate such uncertainties within virtual worlds or computer visualization have so far been limited.2 Although some now actively encourage the abandonment of realism, recent publications on virtual archeology clearly show that many others continue to seek this justification.3

To further compound the problem, visualization techniques have now attained a level of complexity beyond the ability of many archaeologists to either understand or apply, and responsibility for their implementation and development has moved more firmly into the realm of computer science and graphics. industry.4 Therefore, we tend to find that computer visualization technologies are used as a supplement or side project of archaeological

1 For the main issues of archaeological visualization see Miller and Richards (1995), Ryan (1996) and Eiteljorg (2000). The most active proponent of theory in archaeological visualization has been Mark Gillings (see various references in bibliography, including work with G. Goodrick). My thanks to James Crow and GlynGoodrick for their comments on final drafts of this paper.

2 Sorrell (1981) and his reconstruction painting of the 9th-11th century Constantinople in Talbot Rice (1969) 120. The exception is the work with hyperreality in VR, see Goodrick and Gillings (2000).

3 Forte and Siliotti (1997), Barceló, Forte and Sanders (1998) and various articles in Fisher and Unwin (2002).

4 This is demonstrated by the complexity and diversity of data collection and visualization techniques outlined in Barceló (1998).

richard bayliss290

logical research rather than as an integral part. This marginalization of archaeological research agendas has done much to exacerbate academic skepticism about the potential role of these techniques in 'serious' archaeology. Interestingly, this problem was recognized as far back as 1993 by the then outgoing president of the Computer Applications in Archeology (CAA) association.5 These were the halcyon days of computer archeology when geographic information systems (GIS) were the answer to every archaeologist's prayers. and future archaeologists should greatly benefit from full immersion in virtual worlds.6 Such utopian ideals have recently diminished, and while GIS has undoubtedly found its niche in archaeology, the status of computer visualization and virtual reality (VR) is as yet unproven.7

This article aims to introduce the concepts, techniques and ethics of computer visualization to Late Antiquity and Byzantine archaeologists through the critical discussion of two Byzantine period fieldwork projects in which computer-based methods have formed an integral part: the Anastasian Wall Project and the Alacami Project. The context and surviving composition of these structures could not be more different: one, a landscape study of monumental linear engineering surviving in deep forests, the other a highly detailed study of a multi-period basilica on the edge of a Turkish city. The survey methods for both of these projects will be outlined and I will show how computer visualizations can and should be derived from this primary archaeological data. Through the use of these case studies, I will also emphasize the importance of ethical manufacturing in computer visualization and at the same time show how such techniques can be used as basic interpretive elements in a research program. Finally, I hope that this paper will demonstrate how effective results can be achieved using relatively simple techniques and software, enabling archaeologists themselves to take a more significant role in their implementation.

5 Reilly (1995).6 Miller (1996).7 For a critical look at utopianism in computer archeology see Huggett (2000).

archaeological investigation and visualization: Byzantium 291

Anastasian Wall Project

In 400 AD the wealth and population of Emperor Constantine's new city on the Bosphorus rapidly approached that of Rome. As the city's population grew, so did the residents' demands and expectations. This led to a proliferation of construction activities, which in just 200 years led to the construction of a palimpsest of wealthy public and private structures in the city. Early in the 5th century, the city's Theodosian walls separated populous Constantinople from its immediate hinterland and provided a lasting defense against would-be assailants. The worsening situation in the Balkans led Emperor Anastasius (491-518) to have an additional outer fortification built about 65 km west of the city wall. The Anastasian Wall is over 56 km long and crossed the hills of central Thrace from the Black Sea to the Sea of ​​Marmara. It was undoubtedly a massive feat of engineering, requiring considerable human and financial resources, calculated to protect the city and its hinterland from anything but a naval attack (Fig. 1).

The Anastasian Wall Project began in 1994 with the aim of producing a record of this great early 6th century. outer fortification. However, this detailed study was part of a wider strategy in which the study of the city's hinterland would contribute to a better understanding of the city itself. In addition to the study of the wall, project researchers therefore began an investigation of the water supply system of the Byzantine city, extensive evidence of which survives near the wall. The pre-Constantinian water supply system was clearly inadequate to sustain the rapidly expanding city, and thus a new system was begun under Constantius II, which within a hundred years probably rivaled Rome as the most extensive aqueduct system in the ancient world.9 The most important branch of the system - whose completion was celebrated during the reign of Emperor Valens - was a staggering 250 km long, almost three times the length of Rome's longest aqueduct, the Aqua Marcia. The remnants of this endeavor survive in the dense forests of Thrace

8 The main publication on the Anastasian Wall is by Crow and Ricci (1997). Fieldwork reports have appeared from 1995-2001 in the Bulletin of British ByzantineStudies and Anatolian Archaeology. See also the website http://museums.ncl.ac.uk/long_walls/index.html.

9 Bayliss and Crow (2000) and Bono, Crow and Bayliss (2001). This research was carried out in collaboration with Professor Paolo Bono (Hydrogeologist) from "LaSapienza" University, Rome.

richard bayliss292

in the form of impressive bridges, underground canals and several long and deep tunnels. Within the city, the canal crossed the water bridge known as the Aqueduct of Valens, which, at nearly a kilometer long, is still a prominent feature of modern Istanbul. Within the city walls, nearly a hundred Byzantine cisterns have been identified, including three huge open-air reservoirs and several substantial covered cisterns near the city core. This research into the city's water supply was a natural continuation of the Anastasian Wall Project, especially in terms of survey methodology, which was necessary to provide precision over the vast geographical area of ​​Turkish Thrace.

Survey method

In 1994, while standing atop a high visible hill known as Kuâkaya in central Turkish Thrace, the directors of the AnastasianWall Project surveyed the scene below. To the north and south, the wall could be seen winding off towards the Black Sea and the Sea of ​​Marmara respectively. 'How do we begin to investigate such a thing?', they asked. But what they were looking at was not Wallitself, but its familiar signature in the dense forests: an elevated band of trees growing along the top of the wall mound, visible as a dark green linear shadow in the otherwise monolithic panorama (Fig. 2 ) ).

The difficulties presented by this rugged forested terrain had contributed significantly to the limited nature of previous research and publications on the wall and water supply system. Therefore, our understanding of their composition and context was piecemeal and superficial. Despite the limited budget, small fieldwork and the often underestimated problems of working abroad, the team was nevertheless able to implement an integrated method to produce an archaeological record of the remains of these significant works of engineering. These included terrestrial (Total Station) and GPS (Global Positioning System) survey techniques, hydrogeological survey and hydrological analysis of water channels and headwaters, database and GIS (Geographical Information System) resource management, geophysical survey and more traditional techniques where appropriate. In addition, by applying computer visualization techniques to our survey data, we have for the first time been able to see the physical impact of the Wall on its surrounding topographical context (Fig. 3).

archaeological investigation and visualization: Byzantium 293

The field methodology of the project has evolved significantly since the project started in 1994. The survey started with modest ambitions with detailed total station surveys in two relatively accessible areas, at Derviâ KapÌ near the middle of the wall and towards the northern end in the Evcik sector (Fig. 1). In both areas, the survey team systematically recorded surviving facing blocks and the core of the wall. Where neither was visible, the top and bottom of the wall's earthwork were recorded, unless accessibility was significantly confounded by vegetation, in which case a single line of points simply marking the route of the wall was recorded. At the same time, a condition survey was carried out and referenced to the Total Station survey to produce a complete assessment document of the Wall within the survey zone.

In addition to the wall itself, a number of outcrops were examined around Derviâ KapÌ, where a significant outer ditch and countershear were recorded in detail (Fig. 4). No further defenses were found on the northern part of the Wall, where the ground in front of the Wall sloped away considerably, which probably made extensions redundant. Both surveys were continued in subsequent seasons as forest clearance by the Ministry of Forestry (Orman BakanlÌÆÌ) opened up new areas for surveys. The same method of detailed position and condition survey after deforestation was applied to several other zones along the wall. On only two occasions did the research objectives require independent forest clearing, notably for the investigation of a surviving wall fort, Büyük Bedesten, carried out in collaboration with a local workforce in 1998.

The increased availability and improved efficiency of GPS equipment throughout the 1990s has allowed us to integrate the disparate and disjointed terrestrial survey zones on a single coordinate system to centimeter accuracy.10 In addition, it has proven essential for the illumination of the water supply system. , which required reliable measurement of the relative height of water channels over very large distances. Thus, we have been able to transform the entire position dataset of both the Anastasian Wall and the water supply system into a WGS-84 coordinate system. These geographic coordinates can therefore be projected onto either the UniversalTransverse Mercator (UTM) Zone 35 North grid used in the Turkish

10 Perhaps the most significant development was the removal of SelectiveAvailability (the deliberate degradation of GPS signals) in May 2000, meaning that civilian users of GPS can now achieve much higher accuracy in positioning than was previously possible (Satirapod, Rizos and Wang (2001)).

richard bayliss294

ish National Mapping Programme, or on the Mediterranean Network, used on the 1943 British topographic map series of Thrace (1:25000).11

Visualization on the Anastasian Wall

About 75% of the Anastasian Wall is buried deep within the forests of central Turkish Thrace, making it extremely difficult to conceptualize the physical relationship between the monument and its landscape context. The data collection strategy outlined above was clearly only one step towards the production of a legible record of the wall and water supply system, but the presentation of the data in a meaningful way remained problematic. A simple line on a small-scale map would clearly not allow adequate representation of the wall's structure and detail. On the other hand, it would require a series of fragmented and potentially confusing plans to show the details of the monument on a larger scale. Although both techniques were used with relative success, it was believed that neither gave a satisfactory overall impression of the Wall.

In addition to the above, the opportunities to obtain general perspective photographs of the Wall were extremely limited. Our photographic archive mainly comprises close-ups of masonry fragments protruding from deep undergrowth, with occasional wide panoramas of the Wall's most visible signature: the aforementioned elevated row of trees (Fig. 2). However, the digital data collection strategies have allowed us to use computer visualization techniques to provide solutions. An oblique aerial view of a rendered digital terrain model (DTM) clearly allows a wider view of the Wall in relation to its surrounding landscape (Fig. 3).

Map, survey and model The main challenge in producing a landscape visualization of the wall was to achieve appropriate levels of detail for both the immediate wall corridor and the wider topographic context. The detailed data was collected over three seasons (1994-6) using TotalStation surveying instruments. This included a complete record of the surviving remains of the wall, both face and core, along a 3 km stretch extending either side of the major road junction at DerviâKapÌ, together with a detailed survey of the immediate topography-

11 British Library reference: MAPS 43986.1 (M.D.R. 629).

archaeological investigation and visualization: Byzantium 295

the Ish context of the wall, including its outbuildings (Fig. 4). This was our most intensive study zone on the wall and thus provided a viable data set for visualization. A wider context of about 16 km was subsequently digitized using intelligent point selection from a single 1:25000 map of the region.

LSS survey software was used to generate digital terrain models (such as triangulated irregular networks or TINs) from both the map and ground-based data.12 This data was transferred directly to AutoCAD®

and geo-referencing using common control points within the two datasets. The potential variation between the quality of the data from the 1:25000 maps and the quality of the microtopographic point data closes

Fig. 1. Map of Thrace showing the Anastasian wall and the long-distance water supply system. For fig. 2-3 see plates.

12 LSS survey software is produced by McCarthy Taylor Partnership (www.mccarthytaylor.com).

richard bayliss296

to the Wall was of course a serious consideration, but the boundaries between the two were unexpectedly contiguous. It was then possible to use the study of the surviving facing blocks and core as footprints to model the Wall (fig. 3).13

The model of the wall remains relatively crude, as the priority was not to reconstruct, but to provide a context for data that had been collected directly from the surviving monument. The height and appearance of the wall remains a major question and presented the only element of 'rebuilding' in the model. There are very few surviving linear fortifications from which comparisons could be drawn, but many 6th century city circuits are known throughout

Fig. 4. Topographic survey of the Anastasian wall at Derviâ KapÌ (100 m grid). For fig. 5-10 see plates.

13 These post-survey techniques were devised and implemented by the author with assistance from Glyn Goodrick and Mark Jackson of the Department of Archaeology, Newcastle University. For a detailed explanation of the creation and visualization of 3d models, see Barceló (1998).

archaeological investigation and visualization: Byzantium 297

eastern Mediterranean. After considerable consideration, a height of approx. 12 m therefore adopted based on near-modern parallels Resafa and Dara on the eastern border.14 The location of windows/loopholes on the towers is also an unknown factor, and although several significant gateways and placards have been found at various locations in the central and northern sector, they are too few to propose a consistent system. It was only recently that the remains of merlons were found in the rubble, allowing us to suggest that the wall was crenellated: a feature that might have previously been assumed but not substantiated. These elements were therefore largely omitted from the model.

To obtain approximate photorealistic textures, a standard "ashlarwall" texture map was composed from photographs of the wall and applied to all visible surfaces of the wall model using Accurender®

raytrace and radiosity software, which work directly in AutoCAD®.15 Environmental components of the model, such as landscape texture, lighting and background, were defined and renderings were produced from a variety of viewpoints.16 The processed images provide an impression of the wall's impact on its landscape context that is clearly not bound by the aforementioned limitations of scale. In addition, the viewer gets an oblique perspective on the construction of the forts and an understanding of the different tower forms found on this stretch of wall. The viewer is thereby presented with a perspective of the Wall, which today is impossible to achieve through other airways due to the dense forest cover.

Limitations of the wall model There are several representational limitations of this image that need to be considered. Firstly, it must be remembered that the topographic survey data represent a modern land surface, so the banks will have been eroded and the ditches gradually filled in since their original construction. The juxtaposition of these data with the Wall - reconstructed to a hypothetical 6th century. phase – therefore creates something of a visual illusion, an uncomfortable combination of modern reality with speculation about the past. Another area

14 Karnapp (1976), Croke and Crow (1983).15 Accurender is produced by Robert McNeel & Associates (www.

accurender.com).16 For the principles of raytrace and radiosity rendering see Barceló (1998) 21.

The meaning and implementation of photorealism is discussed in Chalmersand Stoddart (1996).

richard bayliss298

of difficulties relate to the wider composition of this modern landscape and the integration of high-resolution survey data with low-resolution map data. The survey data represent the surface morphology at ground level. However, the map data is derived from aerial photogrammetry and is therefore more representative of the forest canopy than the ground surface. This factor also made it difficult to decide on a suitable surface texture for the landscape, and in the end a generic grass-like texture was adopted, giving the landscape the unfortunate appearance of a freshly cut lawn, but avoiding over-complicating the visualization. The compositional elements in the visualization are therefore completely abstract, somewhat contradictory and perhaps a little misleading. However, given the intention of the visualization project (to provide a broad perspective of the Wall in a landscape context stripped of vegetation), the effect of these discrepancies is relatively marginal.

The fort in the forest Similar techniques were used to produce a visualization from the results of a microtopographic survey of the best-preserved forton the Wall, known as Büyük Bedesten (Fig. 5). However, in this case the visualization of the fort was designed solely as a means of providing a representation of the survey data and no reconstruction was attempted. Therefore, this visualization functions essentially as a primary archaeological record and not as an attempt to provide interpretation (Fig. 6).

The main elements of the visualization are the digital terrain model generated from the survey data, the surviving walls of the fort, the fallen blocks from the collapsed walls and the wider topographical context. Masonry was only modeled to its actual surviving height and was provided with photo-realistic textures from photo-graphic documentation on site. The positions of the fallen blocks were recorded with a single point during the study. The random block effect seen on the visualization was achieved by creating multiple versions of a solid block with different rotations and randomly applying them to the detected positions. The forested background was created using Vistapro 3.10 and represents the area for which survey data were not collected.

The surrounding landscape (with the significant scale and north arrow) is the only element of the visualization that can be described as fictional. The image therefore does not demonstrate what the premight looked like, but instead clearly shows the surviving elements

archaeological investigation and visualization: Byzantium 299

from an informative oblique and aerial perspective. In this sense, the results can be compared to an aerial reconnaissance simulation, except that the operator effectively gains control over the weather, lighting, viewpoints, and other atmospheric variables to create the appropriate conditions for optimal visibility.

The survey therefore acts as a filter for the visualization, allowing the removal of visually distracting elements, such as the variable texture of the vegetation and recent human impact on the site. The result is that relevant and relevant elements such as the surviving masonry and fallen blocks of the fort are made clear, while the texture of the landscape is homogenized to give a clear picture of the archaeological remains.17

The form of landscape visualization outlined here constitutes a very different visual medium from the traditional contour or hachure plot, the latter being popular in archeology for its arguable effectiveness in representing artificial earthworks in a complex natural landscape.18 The visualization of ​​the digital terrain model directly from the survey data is arguably a less subjective method than the hachure survey, although this depends to some extent on the data collection strategy. However, it is the immediate readability of such computer visualizations that underlines their intrinsic value. They generally cannot be used as scalable metric illustrations from which the viewer can obtain dimensions, so their role remains primarily as informative complementary media to the hachure or contour plan. Likewise, the effectiveness of the visualization as a didactic tool depends on the medium in which it is displayed. and a computer-based presentation of visualizations is considerably more effective than the printed images in this publication.

17 Clarification at the expense of realism in the representation of an archaeological object or site is of course a well-established technique of archaeological illustration. In lithic or ceramic drawing, for example, attention is primarily directed to the features that illuminate the object's manufacturing process or those that can be described as decoration, with less attention to the overall variations in texture or tone (see Pringle and Molding (1997) 22 ).

18 The argument for hachures is advanced by Bowden (1999) 65-7, while the benefits of computer visualization in landscape representation are argued by Fletcher and Spicer (1992).

richard bayliss 300

Primary archaeological data and the introduction of end-

user interaction: the Alacami 3d model

The critical emphasis on primary archaeological data in computer visualization is fully illustrated in this second study as we move beyond Constantinople to the eastern province of Cilicia near the modern city of Adana. On the outskirts of the nearby market town of Kadirliis an elegant abandoned mosque, known as Alacami (Fig. 7). The Temoskeen was created by adapting a medieval chapel built within the empty shell of an early Byzantine basilica, which itself was erected on the site of a former tomb. This is the last surviving monument of the ancient Roman and Byzantine city of Flavia, and its remains present a microcosm of the city's history. A digital survey of the structure was carried out in 1997, and the data was used not only to produce a series of plans and elevations, but also to create a detailed three-dimensional reconstruction of the church in its primary phase (Figs. 8-10) ).19

Alacami is an ideal subject for the implementation of computer visualization techniques, since the walls of the basilica still stand at the height of the cornice, and there is reliable evidence of the superstructure and the decoration of the interior. Therefore, the speculative element in the reconstruction is relatively minimal. The motivations for the production of a computer model were then twofold. First, in a way similar to the visualization of the Wall Fort, the modeling project allowed us to effectively filter out the later modifications of the original building, which confused the conceptualization of the original form. Likewise, we were also able to reinstate original elements for which there was conclusive evidence, as for example in the case of the rectory.20 Secondly, an extensive series of mosaics and opus sectile pavements were excavated in the 1960s both internally and externally by the building. Mary Gough recorded these pavements in a series of remarkable watercolors at the time of excavation, but the mosaics themselves were subsequently reburied. The addition of these images to the visualization scheme

19 The main publication is Bayliss (1997). The results from the 1997 season can be found in Bayliss (1999a,b) or on the website http://museums.ncl.ac.uk/alacami/index.htm. For earlier discussion of the Alacami 3d model see Gillings (1999) and Goodrick and Gillings (2000).

20 A chancel is a vertical extension of the nave above the height of the aisles to provide elevated lighting.

archaeological investigation and visualization: Byzantium 301

has enabled us to begin to appreciate the overall composition of the decorative program. The importance of a mosaic is often judged on the basis of its art-historical or aesthetic qualities, and the importance of context can therefore be marginalized. Displaying mosaic in a modeled architectural space emphasizes the importance of location, spatial significance and visibility in the decoration of the interior space. This is especially relevant in a church where different architectural zones are charged with meaning by the liturgy and processional movement that takes place within them. Such visualization strategies potentially provide an engaging and stimulating medium for addressing these issues.

The composition of the Alacami model The model is undoubtedly a highly effective visual method of presenting the primary data from the surveys and the excavation. In the same way, the reconstruction process forced us to decide on questions regarding the composition of the individual elements of the building. This was especially the case when considering the form of the superstructure, an issue often overlooked by archaeological studies where the primary medium of interpretation is the two-dimensional ground plan.21

The primary archaeological data used in the model came from:1. ground plan and topographic survey of the site.2. photogrammetric recording of surviving walls, used (as surface de-

cals) to the modeled surfaces.3. excavations to clarify some details of the building - the western steps,

underground chamber and the choir area.4. digitized maps (1:1000 maps) to provide topographical data for the lo-

cal environment.5. on-site photo documentation of photo-realistic textures.6. watercolor illustrations of the excavated mosaics.

This example clearly shows how the individual elements of the 3d model can be explicitly derived from the results of on-site

21 The need to move beyond the 2d plane is emphasized by a number of researchers (eg Gillings (2000)) as one of the main advantages that computer visualization technology has to offer to archaeological interpretation. Others (eg Eiteljorg (2000)) have also commented on the use of modeling projects as a means of structuring and interpreting the mechanics of a building. Some VR systems are under development that allow the structural elements of a VR model to be modified and scaled by an end user, see Ryan and Roberts (1997).

richard bayliss302

examination. The photogrammetric elevations in particular improve the visualizations in a number of significant ways. On an immediate visual level, they allow the actual surface composition of the surviving wall to be incorporated into the model. More importantly, they also provide a visual mechanism to demonstrate the surviving extent of the walls, thereby showing the scale of the reconstructed contents. This is also an effective method of presenting the details of elevation drawings and of demonstrating their physical location in relation to the structure, which is often difficult to convey in two-dimensional paper drawing.22

Virtual Reality and "Bubbleworld" So far we have mainly looked at static visualizations produced from complex 3d models. These are single image renderings processed from a predetermined point of view. If a sequence of such views is produced by moving either the camera or the target viewpoint, a series of single images can be combined to produce an animation. Rendering a single image is usually a lengthy process, although this varies greatly depending on the power of the computer used. The production of a 20 second animation will nevertheless require a significant amount of processing time. This explains the problem of facilitating end-user interaction, in which each frame must be processed continuously as the viewer makes choices about which parts of the model to explore. Relatively smooth simulated motion requires processing at least 30 individual frames for each second of the animation. Therefore, virtual reality (VR) models that allow some degree of interactivity are generally much more primitive and crude than preprogrammed high-quality animations, due to the limitations imposed by the time it takes to process the images.23

In the case of Alacami, when attempts were made to convert a non-textured version of the Alacami 3d model into VRML (standard VR format), the resulting VR model could be explored on an extremely powerful computer at a dizzying speed of one hit every 40 minutes. To provide a degree of interactivity, without

22 This can be a much more significant problem when dealing with an extensive decorative program covering all interior surfaces. A very successful example of this technique applied in such circumstances is the visualization work done on the Egyptian tomb of Sen-nedjem by Terras (1999).

23 See, for example, the cumbersome VR models in the Multimedia Zone on the BBC's History website, http://www.bbc.co.uk/history/multimedia_zone/3ds/index.shtml (last accessed 2 January).

archaeological survey and visualization: Byzantium 303

having to reduce the level of detail in the model considerably, an apanoramic viewing system, known as Bubbleworld, was used.24

A number of fixed positions are defined in the model, and from each location a series of single-frame renderings are produced to achieve full 360 degree coverage from each position. The images from each location are seamlessly stitched together and warped to produce a single image of the entire 360-degree field of view. When the viewer is in a certain location, they only see a small part of the image. Rotation of the viewpoint therefore gives the impression that a 3d space is being occupied. Further movement in the model is facilitated by connections or 'hot-spots' between different Bubbleworld locations. Although the exploration of the model is effectively limited to a series of fixed points, the viewer is still presented with choices about where to look and which path to follow. The Bubbleworld technique therefore allows the details incorporated in the model to be maintained within a semi-dynamic visualization system. At the same time, Bubbleworld files are pre-rendered single images and can therefore be relatively small and thus served over the Internet.

Realism and authenticity in the Alacami model Despite the level of detail provided by the quality of the surviving material at Alacami, the visualizations retain the sterility typical of this type of technique. However, it is worth bearing in mind that this 3d model was constructed between 1994 and 1998, mainly using the DOS-based AutoCAD® Release 10 software. More recent models of similar subjects demonstrate the level of detail and texture that can now be simulated, as for example in the reconstruction of the church of St. Symeon the Younger near Antakya (anc. Antioch).25 Although advances in software enable us to produce more visually convincing results, we should be aware that the more realistic a visualization appears, the more it can seem like a authentic portrayal, although many elements remain a matter of speculation. This problem is readily apparent in the contrast between the Alacami model, which is almost intact, and that of St. The Church of Symeon, for which the evidence is extremely fragmentary. The bigger

24 Goodrick (1999), Goodrick and Gillings (2000), Dykes (2002). These techniques are more commonly used with real-world panoramic images to provide "virtual tours" of homes on the real estate market or luxury vacation rentals.

25 Condolences (2001) 217-221.

richard bayliss304

attention to texture and lighting undoubtedly makes the latter seem more realistic, despite the more limited nature of the primary evidence.

Despite the obvious reluctance of many model designers to indicate the extent of speculative content in a visualization, there are many techniques that can be successfully employed, for example brightening or fading the less certain elements.26 In the case of Alacami 3d. model, the Internet has been the primary dissemination medium, and therefore more could have been done to utilize hypermedia technology to provide additional information about the composition of the 3d model. In particular, the visualizations could have been coded as image maps and embedded with spatially located data that includes photo documentation or explanatory material appropriate to the site. In this way, each zone of the image could have an easily accessible explanation of the archaeological interpretation, should the viewer wish to see it. The ability to provide this kind of intra-site GIS documentation as a fundamental component of a visualization is one of the great strengths of hypermedia and its integration into dynamic virtual reality systems is likely to be one of the biggest advances over the next few years in this field.27

Software advances may tempt us to add additional elements of realism to a 3d visualization. Yet there are aspects of Alacami that we will simply never know, and therefore we should not attempt to reconstruct, for example how either the interior or exterior of the walls were rendered, how the choir area was composed, and how the interior was fitted and decorated.28 Currently, technology would also allow us - if we wanted - to populate the church with a clergy and congregation and to provide a more sensual experience by filling the church with Byzantine chant.29 At present, all these elements must

26 Eiteljorg (2000) fig. 8.27 Brown, Kidner and Ware (2002); Hackley (2002). See e.g. el-

tronic indexing system based on 3d photogrammetry used for documentation of the monastery of St. Dionysios on Mt Athos, http://www.gisdevelopment.net/events/isprs/2001/ts6/isprs6004pf.htm (last accessed January 2).

28 Here, the archaeologist responsible for the integrity of the model should define the thresholds for authentic representation. For example, it is probably reasonable to reconstruct an entire tiled roof from a few tile finds, as long as the shape of the superstructure is known, as was the case with Alacami. Careful and minimal use of comparative evidence also plays an obvious role in this process.

29 The 3d reconstruction of the Romanesque cathedral in Santiago in the north

archaeological investigation and visualization: Byzantium 305

still be envisioned, and we must question how far we want to pursue such goals. As it stands, the Alacami 3d model serves primarily as an integrated display medium for the primary archaeological data. The reconstructed elements are structurally significant, but based on solid in situ evidence, and there are few attempts to improve the model further in the quest for realism.

Thoughts on the future: Computer visualization and

Byzantine Archaeology

The sites in this study are not the only examples of computer visualization in Late Antiquity and Byzantine archeology that have been attempted to date. Probably the best known applications of visualization techniques to Byzantine monuments are those produced by Tayfun Öner for his Byzantium1200 website.30

The site has a collection of stunning computer generated images of the Byzantine monuments of Constantinople as they might have appeared in 1200 AD. Most are based on 3d models produced by Öner, an engineer who has worked on several other archaeological reconstruction projects . His work is given academic credibility through the involvement of the leading Byzantine historian and topographer Albrecht Berger. For several monuments, Öner has bravely attempted complete reconstructions, although the supporting evidence varies considerably in both quantity and quality. The model for the Church of the Holy Apostles, for example, is based both on the historical sources and the surviving churches that it inspired. In contrast, the Church of SS Karpos and Papylos has been reconstructed in its entirety based solely on the surviving substructure. The considerable effort that has undoubtedly gone into this work is beyond criticism, especially as the site clearly shows the work in progress. However, it is worth pointing out that the academic potential of this work could have been further realized if the images were accompanied by subjective assessment and contemplation of the nature and authenticity of their content. Since pre-

western spain presents a great example of detail and realism that can be stranded with current technology, fittingly in this case with a church interior. An interactive multimedia CD-ROM is available via the publisher's website www.fbarrie.org, but sample images and an important review are presented by Gerrard (2001).

30 http://www.byzantium1200.org.

richard bayliss306

sends a significant opportunity to promote discussion and debate about the urban structure of Constantinople, especially considering Öner's ultimate intention, which is to map his structures on a digital terrain model derived from Muller-Wiener's topographic map of the city.31

The creation of a 'virtual' Constantinople as a medium for debate would undoubtedly both benefit and facilitate the study of the city's Byzantine topography, especially if accompanying qualitative and bibliographical information were incorporated in a form in which it could be both consulted and interrogated as a dynamic spatial information system. Istanbul, which has a population of about 11 million people, was the capital of a Mediterranean empire for about 17 centuries. Its continued occupation over this long period means that our understanding of the historical topography of the early Byzantine city is extremely fragmentary, composed primarily of historical sources and the archaeological evidence of the later Byzantine, Ottoman and modern Turkish city. It is therefore extremely difficult to visualize the early Byzantine city, not only in terms of the appearance and function of individual buildings, but also to perceive the spatial dynamics of the living city and to appreciate the multi-vocality and interoperability of its places and spaces .

Therefore, it is not only for the purposes of data presentation and interpretation that computer visualization can influence Byzantine archaeology. Current trends in the archaeological theory of urbanism pave the way for the use of VR to address questions regarding the spatial and experiential composition of such complex sites. In her book, The Urban Image of Augustan Rome, Diane Favro introduced us to a visually rich, descriptive 'walkthrough' of a virtual Rome, both before and after the Augustan transformation of the city. This approach drew on the pioneering work of urban geography theorists such as Kevin Lynch, which has also been applied to the study of late antique transformations in the built environment.32 In his attempts to visualize buildings in wider contexts and imagine urban vistas and vantage points, Favro emphasized the future of VR's role as a medium to facilitate such

31 Müller-Wiener (1977). It is unfortunate that the momentum behind the site has dropped significantly, with no new input for almost two years.

32 See Favro (1996), Lynch (1960, 1984), Wharton (1995), Bayliss (1999b), in addition to other work evaluating or inspired by Lynch's approaches to the study of ancient urbanism, such as Yegül (1994), Ellis (1995) ), Carl, Kemp, et al. (2000).

archaeological survey and visualization: Byzantium 307

33 Her vision is now being implemented as part of the Rome Reborn project, within which she plays a key role. This began in 1995 with the aim of producing contextualized computer reconstructions of ancient and medieval Rome and is now producing some significant results.34

Increasingly, computer software will allow us to provide complete reconstructions of structures and places and to introduce ever more compelling elements of realism. Hardware development and improved access to specialized VR centers will also encourage the development of more immersive VR experiences.35 However, if we as archaeologists maintain research-led goals for visualization projects, our virtual worlds may not be very realistic, but they will be usable . Therefore, instead of following the technological trend towards full immersion, we might spend more time thinking more carefully about how computer visualization can enhance our research.

Considerable debate has surrounded the terminology of VR, but not enough thought has yet been given to its potential impact on archaeological research.36 With any human-computer interface that claims to be virtual reality, the question remains as to whether reality is actually represented. The vigorous nature of academic debate on any particular question of historical authenticity should constantly remind us that there is no single known past that everyone would either accept or acknowledge. In this context, the considerable body of work produced by those concerned with the ethics of experimental reconstruction should not be (but often is) overlooked.37 However many may strive for complete immersion in a virtual constructor-synthetic/hyper-realistic environment, we are ultimately only immersing our 21st century selves along with our own cultural baggage and preconceptions. Work must therefore continue with the theoretical justification for VR and computer visualization, but perhaps more

33 Although generally seen as an analytical application of GIS, viewhed analysis has recently been approached more reflexively through the use of three-dimensional visualization systems (Gillings and Goodrick 1997).

34 Frischer, Favro et al. (1998). See also http://www.aud.ucla.edu/~favro/rome-reborn/ (last accessed January 2) and http://www.cvrlab.org/ (last accessed January 2). Ongoing work at Sagalassos in Pisidia is progressing in the same direction with some impressive results, see Matens, Legrand et al. (1998).

35 Forte (1998).36 Gillings (2000, 2002).37 See e.g. various articles in Stone and Plane (1999), especially the

introduction.

richard bayliss308

Specifically, visualization must be integrated as a medium for interpretation and structured thinking in our research programs.38

Some attempts have been made to solve this problem by exploring the possibilities of creating adaptable virtual worlds, not only VR constructions that can be modified by the user, but models that can be transformed to investigate how different social groups may have perceived sites and spaces .39 Approaching questions of accessibility, visibility and perception using such techniques has obvious value for Byzantine archaeology, especially, for example, in relation to church architecture and the perception of interior design and decoration.

The difficult union between archaeological theory and VR technology is a significant response to doubts about the role of VR in academic research, especially regarding the problematic areas of reconstruction, realism and embodiment. Quantifiable success in these areas has been extremely limited to date, with more attention paid to the rhetorical value of VR as an integral part of the survey process rather than its actual benefits to ongoing archaeological research.

The case studies presented here have focused on the role of computer visualization, not as a compelling medium for producing virtual worlds or reconstruction illustrations, but as a means of providing engaging solutions to traditional problems of visualizing the primary archaeological record. The potential for data integration and filtering as outlined in both studies has clear implications for Late Antiquity and Byzantine archaeology, which is characterized by complex multi-period standing structures that often require a variety of data acquisition techniques for the acquisition process. The role of survey-based computer visualization as a complementary medium for the interpretation and realization of archaeological sites is therefore well established and has been demonstrated here as an integral component of research programs involving both broad landscapes and complex multi-period structures. However, computer visualizations of this nature should not be seen as either end products or inferior

38 Examples of VR applications used to directly address research questions can be found in the recent research at the Neolithic monuments at Avebury by Goodrickand Gillings (2000) and at Thornborough by Goodrick and Harding (2000).

39 See, for example, the alternative versions of the Santa Maria Maggiore model in Frischer, Favro et al. (1998) and the Gummiding technique of Goodrick and Gillings (2000).

archaeological investigation and visualization: Byzantium 309

doubles of the original constructions.40 They provide a means of engaging with the archaeological record in an otherwise unattainable way and as such generate their own dialogues and biographies. All the visualizations presented here are themselves original constructions, based more on the remains as they survive today than on any consideration of how they might have looked or been perceived. If some would call this virtual reality, then it must be accepted that the reality of these works, both in terms of their source data and the viewer's perceptions, is completely contemporary.

Glossary of conditions

Digital terrain model (DTM / DEM / TIN ) – a method of storing elevation data digitally, primarily used for contour interpolation or three-dimensional display. Data is most often stored either in a uniform grid (Digital Elevation Model ), or as irregular point elevations at critical points connected by lines to construct a triangulated surface (TriangulatedIrregular Network).

Geographic Information System (GIS) - a suite of computer-based tools used to collect, manage, analyze, and display spatial data.

Global Positioning System (GPS) - a system for providing precise location based on data transmitted from satellites. Developed by the US military, GPS is now widely used in surveying and navigation situations.

Intelligent point selection – a non-random selection of spatial points used to describe a topographic surface based on a defined and subjective set of criteria, for example the tops and bottoms of slopes. Used for creating triangular irregular networks.

Micro-topographic survey - topographic survey where surface morphology is recorded in high detail.

Photogrammetry – uses scaled overlapping stereo images to produce planimetric and topographic maps primarily of the Earth's surface, but also used for features in the built environment, particularly building heights. Photogrammetry makes use of control points by which photographs are recorded for the location of features identified in terrestrial or building surveys.

Photorealism – used in the context of computer visualization to define rendered images of a quality that can approximate a photographic image.

Rendering – generating a two-dimensional static image or animation frame from a three-dimensional computer-generated model.

40 Gillings (1999).

richard bayliss310

Total Station—(electronic tachometer) electronic digital theodolite capable of measuring distances as well as horizontal and vertical angles.

Universal Transverse Mercator (UTM) – a series of 120 coordinate systems based on the Transverse Mercator projection and designed to provide worldwide coverage.

VRML - an open standard for 3D multimedia and shared virtual worlds on the Internet.

WGS-84—World Geodetic Reference System of 1984 is a global datum widely used as a base datum for processing and converting data from one datum to any other datum. The Global Positioning System (GPS) is based on this datum.

archaeological investigation and visualization: Byzantium 311

Bibliography

Barceló J. (1998) “Visualizing what could be: an introduction to virtual reality techniques in archaeology”, in Virtual reality in archaeology, ed. J. A. Barceló et al. (BAR international series 843) (Oxford 1998) 9-35.

Barceló J., Forte M. og Sanders D. (1998) Virtual reality in archeology (Oxford 1998).Bayliss R. (1997) "The Alacami in Kadirli: transformations of a sacred monument",

AnatSt 47 (1997) 57-87.——— (1999a) “A place with duration in the cemetery at Flavias”, in Olba II =

Proceedings of the I. International Symposium on Cilician Archaeology, red. S. Durugonul og M. Durukan (Mersin University 1999) 483-98.

——— (1999b) “Usurping the cityscape: the experience of ritual topography in late antique Near Eastern cities”, in TRAC 98, ed. P. Baker et al. (Oxford 1999) 59-71.

Bayliss R. and Crow J. (2000) "The Fortification and Water Supply System of Constantinople", Antiquity 74 (2000) 25-26.

Bono P., Crow J. and Bayliss R. (2001) "Constantinople's water supply: archeology and hydrogeology of an early medieval city", Environmental Geol-ogy 40 11/12 (2001) 1325-33.

Bowden M. (1999) Unraveling the Landscape: an Inquisitive Approach to Archaeology (Stroud, Glouc. 1999).

Brown I., Kidner D. and Ware, J. (2002) “Multi-resolution virtual environments as a visualization front end for GIS”, in Virtual Reality in Geography, ed. P. Fisher and D. Unwin (New York 2002) 144-62.

Carl P., Kemp B., et al. (2000) "Were cities built as images?", Cambridge Archaeologi-cal Journal 10/2 (2000), 327-65.

Chalmers A. and Stoddart S. (1996) "Photorealistic graphics for the visualization of archaeological site reconstructions", in Imaging the past, ed. T. Higgins et al. (British Museum Occasional Papers no. 114) (London 1996) 85-94.

Croke B. og Crow J. (1983) "Procopius and Dara", JRS 73 (1983) 143-59. Crow J. og Ricci A. (1997) "Investigating the hinterland of Constantinople: in-

interim report on the Anastasian Wall Project", JRA 10 (1997) 235-62. Dykes J. (2002) "Creating information-rich virtual environments with geo-refer-

enced digital panoramic imagery", i Virtual Reality in Geography, red. P. Fisherand D. Unwin (New York 2002) 68-92.

Eiteljorg H. (2000) "The compelling computer image - a double-edged sword", Internet Archeology 8 (2000) http://intarch.ac.uk/journal/issue8/eiteljorg_toc.html (last accessed 1 December).

Ellis S. (1995) "Prologue to a Study of Roman Urban Form", in Theoretical RomanArchaeology: Second Conference Proceedings, ed. P. Rush (Aldershot 1995) 59-71.Favro D. (1996) The Urban Image of Augustan Rome (Cambridge 1996).

Fisher P. and Unwin D. (2002) Virtual Reality in Geography (New York 2002). Fletcher M. and Spicer D. (1992) "The display and analysis of ridge-and-furrow

from topographically searched data”, in Archeology and the Information Age, ed.P. Reilly and S. Rahtz (London & New York 1992) 97-122.

Forte M. (1998) "About virtual archaeology: disorders, cognitive interactions andvirtuality", i Virtual reality in archaeology, red. J. Barceló et al. (BAR international serie 843) (Oxford 1998) 247-59.

Forte M. and Siliotti A. (1997) Virtual archaeology: great discoveries brought to life through virtual reality (London 1997).

richard bayliss312

Frischer B., Favro D., et al. (1998) "Virtual reality and ancient Rome: UCLACultural VR Lab's Santa Maria Maggiore Project", in Virtual reality in archae-ology, ed. J. Barceló et al. (BAR international series 843) (Oxford 1998) 155-62.

Gerrard C. (2001) “Reconstruction of Master Mateo's Stone Choir”, InternetArchaeology 11 (2001) http://intarch.ac.uk/journal/issue11/reviews/gerrard.htm (last accessed 2 January).

Gillings M. and Goodrick G. (1997) "Sensuous and reflexive GIS: exploringvisualisation and VRML", Internet Archeology 1 (1997) http://intarch.york.ac.uk/journal/issue1/gillings_index.html (last accessed Jan .02).

Gillings M. (1999) "Engaging place: exploring the potential of VR in experientiallandscape studies", in Archeology in the Age of the Internet: CAA 97, ed. L. Dingwall et al. (BAR international series 750) (Oxford, Eng. 1999) 247-54.

——— (2000) “Planes, elevations and virtual worlds: developing techniques for the routine construction of hyperreal simulations”, in Virtual reality in ar-chaeology: CAA 1998, ed. J. Barceló et al. (BAR international series 843) (Oxford 2000) 59-69.

——— (2002) “Virtual Archaeologies and the Hyper-Real”, in Virtual Reality inGeography, ed. P. Fisher and D. Unwin (New York 2002) 17-34.

Goodrick G. (1999) "VRML, virtual reality and visualization: the best tool for thejob?", i Archaeology in the age of the internet: CAA 97, red. L. Dingwall et al. (BAR international serie 750) (Oxford 1999) 266-68.

Goodrick G. and Gillings M. (2000) “Constructs, simulations and hyperreal worlds: the role of Virtual Reality (VR) in archaeological research”, in On the Theory and Practice of Archaeological Computing, ed. G. Lock and K. Brown (Oxford2000) 41-58.

Goodrick G. og Harding J. (2000) "Virtual Reality at the Neolithic monumentcomplex of Thornborough, North Yorkshire", i Virtual reality in archaeology:CAA 1998, red. J. Barceló et al. (BAR international serie 843) (Oxford 2000)115-19.

Haklay M. (2002) "Virtual reality and GIS", i Virtual Reality in Geography, red. P. Fisher og D. Unwin (New York 2002) 47-57.

Huggett J (2000) "Computers and archaeological cultural change", in On the theory and practice of archaeological computing, ed. G. Lock and K. Brown (Oxford2000) 5-22.

Karnapp W. (1976) Die Stadtmauer von Resafa in Syrien (Berlin 1976).Kondoleon C. (2001) Antioch: The Lost Ancient City (Princeton 2001).Lynch K. (1960) The Image of the City (Cambridge, Mass. 1960).——— (1984) "Reconsidering The Image of the City", i Cities of the Mind, red. L.

Rodwin and R. Hollister (New York 1984). Matens F., Legrand J., et al. (1998) “Computer Aided Design and Archeology at

Sagalassos: methodology and possibilities for CAD reconstructions of archaeological sites", in Virtual reality in archaeology, ed. J. Barceló et al. (BAR international series 843) (Oxford 1998) 205-12.

Miller P. and Richards J. (1995) "The good, the bad and the downright misleading: archaeological adoption of computer visualization", in Computer Applications and Quantitative Methods in Archeology 1994, ed. J. Huggett and N. Ryan (BAR International Series 600) (Glasgow 1995) 19-26.

Miller P. (1996) "Digging deep: GIS in the city", i Interfacing the past: CAA95, edd.H. Kamermans og K. Fennema (Analecta praehistorica Leidensia, 28 bind II) (Leiden, Holland 1996) 369-78.

Müller-Wiener W. (1977) Billedordbog om Istanbuls topografi (Tübingen 1977).

archaeological survey and visualization: Byzantium 313

Pringle M. and Molding M. (1997) "Applications of Virtual Reality, and associated information technology, for the illustration of archaeological material", Graphic Archeology (1997), 22-33.

Reilly P. (1995) "A management consultant's view of the current state of CAA and some thoughts on its possible future", in Computer Applications and Quantitative Methods in Archeology 1993, ed. J. Wilcock et al. (BAR international series, vol. 598) (Oxford 1995) 1-6.

Ryan N. (1996) "Computer-based visualizations of the past: technical 'realism' and historical credibility", in Imaging the past, ed. T. Higgins et al. (British Museum Occasional Papers no. 114) (London 1996) 95-108.

Ryan N. and Roberts J. (1997) "Alternative archaeological representations within virtual worlds", http://www.cs.ukc.ac.uk/people/staff/nsr/arch/vrsig97/ (last accessed Jan 02).

Satirapod C., Rizos C. and Wang. J. (2001) "GPS single point positioning with SA turned off: how accurate can we get?", Survey Review 36/282 (October 2001).

Sorrell A. (1981) Reconstructing the Past (London 1981). Stone P. og Plane P. (1999) The Constructed Past: Experimental Archaeology, Education

and the Public (One World Archaeology 36) (London 1999). Terras M. (1999) “A Virtual Tomb for Kelvingrove: Virtual Reality, Archaeology

and Education”, Internet Archeology 7 (1999) http://intarch.ac.uk/journal/issue7/terras_index.html (last accessed 2 January).

Talbot Rice, D. (1969) The Dark Ages (London 1969).Wharton A. J. (1995) Refiguring the Post-Classical City (Cambridge 1995).Yegül, F. K. (1994) "The street experience of ancient Ephesus", i Streets: kritisk

perspectives on public space, ed. Z. Çelik et al. (London 1994) 95-110.

luke lavan314

THE POLITICAL TOPOGRAPHY OF THE LATEANT CITY: SPACE OF ACTIVITY IN PRACTICE

LUKE LAVAN

Abstract

The political topography of the late antique city is a topic that has been largely neglected. This is largely because research has concentrated on new buildings, mainly churches, rather than the reuse of older structures, and because textual evidence has been neglected. This article examines aspects of political topography in terms of 'spaces of activity'. This involves studying discrete units of human activity, plus their material surroundings, whether or not this involves a specific building type. All source types are used to create a general narrative, which here mainly concerns cities in the eastern and central Mediterranean. Emphasis is placed on changes that this approach can bring to our understanding of urban life.

Introduction

In my first paper I suggested that the study of 'spaces of activity' might be particularly suited to political and social aspects of the late antique city. In this second article I consider the application of this approach to political topography, i.e. the spaces where institutionalized governmental activities took place. performed. Some of these, such as the aspraetoria or city council chambers, had distinct architectural settings, and others, such as provincial courts, did not. In fact, I do not want here to describe our general knowledge of these activity spaces fully in accordance with the method described in my first article.1 Rather, I want to use a series of case studies to examine how this approach, in particular the establishment of a clear idea on spatial function through all available source types, can change our historical understanding of the late antique city. To do this I will examine: spaces that are

1 I am currently trying this elsewhere for praetoria, town council chambers, agorai and other political spaces.

L. Lavan og W. Bowden (red.) Theory and Practice in Late Antique Archaeology (Late Antique Archaeology 1 2003) (Leiden), s. 314-337

the political topography of the late antique city 315

particularly difficult to identify architecturally; spaces that have been unduly neglected because they occupy old rather than new buildings; spaces that can be easily misinterpreted by an architectural approach; spaces where important secondary functions have been missed; non-architectural spaces that deserve to be considered more in their urban context. All these subjects will be considered on a synthetic level and relate to the monumental cities of Late Antiquity, from the middle of the 3rd to the 6th century. A.D., mainly in the central and eastern Mediterranean.

Space missing

Pretoria

Probably the most neglected political space in the late antique city is the civil praetorium.2 This was the official residence and administrative headquarters of the civil governor of a province and was typically located in his metropolis. It has received far less attention than it deserves, as its architectural remains have proven difficult to identify archaeologically. This is largely because its shape derives much of its character from high-status residential architecture, but also largely because praetoria often reused earlier monumental structures in built-up urban areas.3 Epigraphic finds seem to be the best way to make identifications , and from examples selected on this basis a certain level of architectural coherence seems to emerge: they appear to be distinguished from ordinary houses by the normal presence of large courtyards, offices, sanctuaries and sometimes large audience halls opening onto to the street, features which interestingly reflect the composition of the Principia building in a military fortress, where the pretoria is divided into this administrative building and a completely residential structure.

It is ancient texts that give the clearest idea of ​​the spatial functions of these complexes. These make it clear that until the early 7th century praetoria were both the official residences of governors and their administrative headquarters. They could include prisons, tax offices, archives and horrea, as well as being social centers in honor of

2 For all this see Lavan (1999) and (2001a).3 Re-use of earlier structures: known through archeology at Gortyn, Ptolemais

and Caesarea Maritima (twice) and through texts in Constantinople and Antioch (twice): Lavan (1999) 145-47, 153-58 and id. (2001a) 43-45.

luke lavan316

a province. It was the center of much civic as well as provincial administration, as the governor now oversaw urban politics to an unprecedented degree, through the authorization of most building works, the appointment of magistrates and sometimes direct intervention in the provision of civic services.4

Significantly, it appears from the combination of information from both literary sources and archeology that the preferred location for the praetoria in Late Antiquity was in the forum/agora rather than on the outskirts of the city, as had been more common in the early imperial period. ; examples of this location occur in Constantinople, twice in Antioch, in Gortyn, and perhaps in Athens, which were chosen as sites for new praetoriums in this period, as well as in Carthage. the governors themselves appeared at the heart of bourgeois political life. The praetorium, with its prisons, ceremonial courts, offices, archives and audience halls, represents the most developed public building ever located on the ancient agora. Set against the development of the Roman provincial administration, which was now increasingly drawn from the local population, this makes some sense. Provincial and civic government were merging, marking the most mature phase of political life seen in the Roman Empire. In this light, the location of the praetoria on the agorai can be seen as a reflection of its monumental history. The Palace of the Giants on the Agora in Athens therefore no longer needs to be seen as a breach of the monumental history of this square; rather, it can be considered to mark the latest phase of its development as a civic meeting place, where a comprehensive center of city government and direct channel to the imperial court took its place at the heart of political life in the city.6

4 Involvement of governors in community life: Lewin (2001); Liebeschuetz (1972) 131-33, 165-67; Bowman (1971) 90, 97 and elsewhere.

5 locations in Pretoria: Gortyn at Agora; Constantinople at Leo Forum: Lydus Mag. 2:20-21; Antioch about the Hellenistic agora and Valens Forum: Malalas 10.10, 13.4, 13.30. See discussion in Downey (1961) 621-30. Athens, Palace of Giants: Frantz (1988): Although not identified from its inscription, it is a possible candidate as a civilian praetorium due to the similarity of its plan to the Palace of Dux Ripae at Dura Europos.

6 This was true both for civil servants and ultimately provincial governors: see Lavan(2001b) 45-117. For this integration of officials within provincial social structures Banaji (2002) 115–33.

the political topography of the late antique city 317

Traditional spaces neglected

out now

Some activity spaces have been neglected due to their continued use of traditional civic architecture dating to earlier periods: most prominent among these are the town council meetings (curiae/bouleuteria) and the fora/agorai. The fora/agorai were the primary social and political centers of Roman cities, but are widely perceived to be in decline in Late Antiquity. I would suggest that this is an exaggeration. It is the result of too much concentration on the most obvious archaeological evidence, that to abandonment; this has overshadowed other classes of evidence, including the evidence of archeology for continuity of use and for traditional functions. Although evidence of abandonment is dramatic, it is often poorly dated; generalized datings to Late Antiquity simplify the chronology of change.7 Indeed, there is much literary evidence for the survival of fora/agorai as social and commercial centers in the 4th and early 5th centuries. 8 At this time, court hearings continued to be held here;9

imperial letters could be read to the public in the forum.10 Until the second quarter of the 5th century. honorary statues were still consecrated in fora/agorai in considerable numbers in many parts of the Mediterranean.11 There is also much textual, epigraphic and archaeological evidence not only of repair in this period but of the building of new fora. Great public works were carried out in Rome under Diocletian and various city prefects; in Constantinople newfora were built under Constantine, Theodosius I, Arcadius and Leo; in Antioch a new forum was built under Valens. In provincial capitals such as Corinth, Ephesus, Side, Aphrodisias and Corinth,

7 Poor dating of abandonment: eg from Africa in Potter (1995) 66-71.8 Fora/agorai as social centers 4th/early 5th c—a selection: Nicomedia Lib.

Or. 1,65-74, Soz. 4.16.12; Ephesus ACO 1.1.5.14; Ankara Lib. Ep. 298; AntiochLib. Or. 29.2-4; Alexandria Rufinus HE 2.22-30, Athanasius Hist. Year. 55-56; General John Chrys. in Matt. 1.17, Sozom. HE 7,6, 8,18). Arles cited at the end of this section. Agorai as shopping centers at the time of Julian: Sozom. HE 5.9, Theoderet HE 3.9.

9 trials: Trier Dionisotti (1982). 104-105; Carthage Augustine Conf. 6.9; Ptolemais Synesius Ep. 41.

10 imperial letters read out in forum/agora: Trier (4th century) Dionisotti (1982) 104-105; Ephesus (A.D. 431): ACO 1.1.5.14; Alexandria (A.D. 482): Zach Mytilene HE5.7, id. vita Severi 41-42.

11 4th c./early 5th c. statues in forum/agora at Tarraco, Cordoba, Corinth, Aphrodisias, Ephesus, Antioch and Lepcis Magna: Lavan (2001b) 300-301.

luke lavan318

fora and agorai saw major repairs.12 In contrast, there are very few fora/agorai in the eastern and central Mediterranean where abandonment can be definitively dated to the 4th or early 5th century. This leads to the conclusion that up to the beginning or middle of the 5th century. one cannot speak of a relative decline in the forum/agora, especially not a loss of its function. Architectural decay occurred before this time in some regions, such as Gaul and Spain, but even here it may be related to wider urban changes, not a decline in the functions of these spaces. This is demonstrated by a later 5th century. honorable dedication of the lapsed provincial forum of Tarraco; this at least makes credible literary evidence from the same years attesting to very traditional activities carried out in the Arles forum, which archeology has shown to have been in a similar state of decay.13

Banner/Bouleuteria

Town council meetings were a critical occasion in the political life of the Roman city. There has been a major monographic study of curiae/bouleuteria architecture: Jean Balty's Curia Ordinis. The focus of this work, however, is architecture as built; therefore, it contains little information about the later Roman period. Nevertheless, texts and epigraphs show that meetings of city councils remained critical events in late antique cities until at least the mid-5th century. It was here that a city's curiales discussed the financing and organization of public services and construction works, the collection of taxes, embassies and honors voted before members and outsiders. Although in some regions this came into being in the 4th century. rather a matter of reactions to the directives of curators and governors

12 Restoration and new construction at the fora/agorai in the 4th century: Rome under various city prefects: Ward-Perkins (1984) 42; Constantinople for new fora of Constantine, Theodosius I and Arcadius see Müller-Wiener (1977). In the first half of the 5th new forums continue to be built in Constantinople until Leo: Mango (1993). Antiochunder Valens: Malalas 1.30 p.m. Restoration in provincial capitals: Corinth Kent (1966)165-67, nos. 504-505; Ephesus Foss (1979) 82 n.70; Page Robert (1951) no. 219a; Aphrodisias Roueché (1989) no. 22. Restoration in common towns: Italy occupied by literature and epigraphy: Lib. Bridge. 1.186 (Naples restoration by Constantine), CIL 5.7781 (Albegna); Africa CIL 8.24095 (Capasa), ILAlg 1.2107 (Madauros), IRT246 and IRT 543 (Lepcis Magna).

13 Evidence of function can be separated from evidence of appearance if the monumental decay is total, not particular. Tarraco dedications continue until A.D. 468/472 (RIT 94-97 and 100), despite general decay from about the middle of the 5th century. century: Taller Escola d’Arqueologia (1989). Arles: Sit. Apol. Ep. 1.11.7; V. Caesarii 1.31, 2.30, 2.39. See Gauthier (1999) 204 and Loseby (1996) 55 for this issue.

the political topography of the late antique city 319

than sovereign decision-making, the curia still met, in fully half of Late Antiquity.14

The non-architectural aspect of this space was arguably as important to its physical appearance as the building in which the meetings were held. A decline in numbers is well documented, but was relative. Timgad had about 200 councilors in the 360s; Oxyrhynchusse seems to have had at least 50+ in 370. Libanius speaks of a city where only one councilor remains, but the disappearance of small municipalities was not new. More significant was the reduction of Antioch's 600 by several hundreds in the mid-4th century. Nevertheless, legislation, inscriptions and correspondence suggest that numbers remained at or above a working minimum until the 430s. In Africa and Asia Minor, until the middle of the 5th century, traditional magistrates were still filled completely in at least some cities.15 In the West, councilors often sat around three sides of a rectangle on wooden chairs - in the East they sat in an arch on the bouleuterion's stone seats. Many still wore formal togae or himatia, although some of those of imperial rank wore the chlamys.16 Business was orderly, and motions, speeches, and acclamations were recorded in minutes, of which we have examples from two cities in Egypt. Ten senior decurions, which had a semi-constitutional basis at the time, decided much; they may have sat around the president and the notaries in the center of the arch. In the East, a wider public audience could also participate.17 Honorary statues and portraits, significant inscriptions and, in the beginning, altars might decorate the space where they met.18 City councils as an institution lost executive control in the East by AD 500. latest, being

14 On city council proceedings, see Jones (1964) vol. 1 724-63, vol. 2 1298-1313; Lepelley (Paris 1979) vol 1; Petit (1955) and Bowman (1971).

15 Magistrates in the 4th-beginning of the 5th Africa: Lepelley (1979) vol 1 149-69. Magistrates in the East, 4th and about 5th c.: Lewin (c. 1995) 92-95, with someone performing all offices twice in Stratoniceia, prob. 5th century: Varinlioglu (1988).

16 statues of civic notables in Late Antiquity reflect both types of dress. Lydus Mag. 3.49, speaks of curials from the later 5th century still wearing the toga. Chlamys: Lib.22.6.

17 Theoretical numbers: Jones (1964) 724, 1298; Balty (1991) 7. Timgad: Chastagnol (1978). Oxyrhynchus: P. Oxy. 17.2110. A councillor: Lib. Ep. 696. Reforms of Julian add 200 to Antioch: Julian Misop. 367d. 600 down to rhetorical 60 at Antioch: Lib. Or. 2.33, 48.3. Minutes from Egypt: Bowman (1979)32-34; van Minnen and Worp (1989). Official records are known in early 5th century Africa and throughout the West into the 6th century. and further: Lepelley (1979) vol 1163, Liebeschuetz (2001) 131, 136. Public at Antioch: Lib. Or. 2.36, 31.36-40; Alexandria A.D. 416: Cod. Theodore. 16.2.42.

18 statues: eg Roueché (1989) no. 56; Bartocini (1950) 36-37; Malalas 14.8.Portraits: Lib. Or. 2.10, 42.43.

luke lavan320

replaced by selections of high-ranking notables. In the West after the middle of the 6th century, the acts of town councils are attested by nothing more than the formulas of land transfers; here too transferred power to leading men and the bishop.19

The architectural setting for city council meetings, in traditional curiae and bouleuteria buildings, was equally institutionalized. Surprisingly, evidence for the continued use or abandonment of curia/bouleuterion buildings and related structures during Late Antiquity has not been examined. The "notables" headed by the bishop are only attested to meet in episcopal palaces, perhaps the natural home of this body, but not in the traditional curia/bouleuterion; this makes a study of the use of town council chambers politically important at the end of town council meetings.20 Epigraphic evidence from Africa reveals that restorations of curiae did indeed take place in many common towns during the 4th century. A re-examination of the buildings themselves shows a similar pattern of repair and remodeling. For example, at Lepcis Magna the curia was moved into a new structure, and that at Sabratha was rebuilt.21 Repairs in spolia can be traced to a number of bouleuteria in Asia Minor, in provincial capitals and ordinary cities.22 The small number of dated repairs is 4th c. .; they preserve the original layout of these well-defined buildings in their entirety. This should be taken primarily as evidence for the continued meeting of the council, especially if there is epigraphic evidence elsewhere in the city for the boule's existence. It is also clear that many bouleuteri in this region were spared spoliation in the later 4th and 5th centuries, while neighboring temples were razed to their foundations. Demolitions and rebuilding of bouleuteriato other features in Asia Minor and the East, when dated, seem to relate mainly to the 6th century; an absence of reuse for other purposes in the 5th century. in this region may also indicate continued use at this time. However, there is so far no positive evidence for the maintenance of

19 Transition to notables: Laniado (2002) and Liebeschuetz (2001) 104-36 .20 Notables meet in the bishop's secretory at Mopsuestia, (A.D. 550) Mansi 9.275-

90; possibly secretly in Anastasiapolis, late 6th century: Life of Theodore of Sykeon 76, 78. Possibly in Hadrianopolis, Paphlagonia, where an imperial letter to the city's notables was published in the bishop's secreton under Justin I or Justinian: Feissel and Kaygusuz (1985).

21 Epigraphic evidence for repairs in 7 cities in Africa: Lepelley (1979) vol. 2.295. Lepcis: Balty (1991) 39-42; Sabratha and Rome Reconstructed: Bartoccini (1950) 29-35; Bartoli (1963).

22 Repair: Nysa, Ephesus, Heracleia am Latmos, Sagalassos; differential spoliation at Ephesus, Termessos and probably at Priene. I will publish this research soon.

the political topography of the late antique city 321

nance for the 6th century and there is no example at all of Justinianic interest in council chambers in Procopius's de aedificiis showing the Senate House in Constantinople, suggesting that traditional council meetings in the East ceased to take place at this time. 23

Space misinterpreted

decide

Surely nothing could be more material, more architectural a monument than a statue. Here it seems that we can engage directly in the interpretation of its physical features on the basis of typological comparisons or by considering texts found on its basis.24

However, such an approach risks serious misinterpretation. I would suggest that to properly understand Late Antique statues of governors or emperors, one must first understand their ancient function, from a wider body of source material, including texts. From literary sources it is clear that imperial statues and other images were both quasi-religious objects and symbols of the emperor's power. As such, they were the focus of religious rituals that expressed loyalty to the emperor. They were received and copied in the provinces as part of the recognition of a new ruler and were used by governors to directly support their own actions. The traditional honorific function of imperial statues was still sometimes present, but was now greatly overshadowed by sacred and symbolic use.25

Statues of late antique governors resemble, at first glance, imperial statues. They represent the largest group of statue dedications after emperors, often the majority of the rest. Examples of actual torsos known from the Aegean usually seem to reflect the Constantinopolitan dress with military cloaks. Surely these statues were also 'symbols of government power'? In a study of the only three complete examples known, from Aphrodisias and Ephesus,

23 city council chambers converted to new functions: Aphrodisias Roueché (1989) no. 43; Ptolemais Goodchild and Kraeling (1962) 92; Scythopolis (demolished) Foerster and Tsafrir (1988-89) 18-19. Salona Clairmont (1975) 54-56, a possible example.

24 On late antique statues see Smith (1999) and (1997). For epigrams from the governor's statues see Robert (1948).

25 Use of imperial images of governors in tribunals and all places where authority is required: Severian de Mundi creatione Or. 6.5 and the depictions of the Rossanogospel, mentioned above. Religious rituals surrounding imperial statues, up to at least A.D. 425: Socrates HE 6.18; Cod. Theodore. 15.4.1 (A.D. 429); Cod. Just. 1.24.3 (AD 439). See Delmaire (1996) 46-47 for quasi-religious status.

luke lavan322

R. R. R. Smith has suggested that the grave facial features of these statues can be seen as communicating the virtues of office alluded to by the inscriptions on their bases. Although Smith does not go this far, it could be easy to proceed from these observations to describe these statues as symbols of governmental power, similar to images of the emperor.26

The study of all the sources in a wider context reveals that they are indeed very different from imperial statues. First, no religious attributes are ever attributed to images of governors; nor was it treason to attack them. Furthermore, there are no known examples of the use of such an image to strengthen a governor's power/authority. The distribution of where the statue bases were erected and an examination of who erected them is also most revealing. In Africa Proconsularis and Italy, where many senior senators served, statues, especially those erected by small towns, appoint governors as patrons. Identical monuments were also erected by client cities in the senatorial houses of Rome. This suggests that although the influence of the governor was acknowledged in these monuments, this was in response to what individual governors might do next for their clients, rather than an attempt to appease them in office.27 The one body that one could expect the pressure, involuntary dedication is the provincial assembly. However, examples of such dedications are actually very rare, in all parts of the empire, indicating that they were not automatic at all; Examples dedicated by members of a governor's staff are even rarer.28 This suggests that statues of governors are not symbols of governmental power, but rather traditional monuments of honor—rewards for good behavior in office, which is in fact what the texts on their foundations say. Ammianus makes it clear that at the end of the 4th century. the senatorial class had by no means lost or changed its traditional taste for seeking the honor of a gilded statue.29

This fits well with our broader picture of civilian governors who ruled largely by consent and usually do not appear to have had decisive in-

26 Interpretation of statues of governors: Smith (1999) 183-87.27 Governors named as patrons outside provincial capitals in pre-Vandal

Africa, at Madauros, Bulla Regia, Gigithis, Sabratha and Cuicul: see entry in Lepelley vol 2. For Italy see the catalog in Ward-Perkins (1984) 231-35 - which also shows statues erected by client cities in Rome.

28 Horster (1998) 49-51 The coercion of a provincial delegation is recorded in Ammianus 30.5.8-10 as having been attempted by Petronius Probus, under Valentinian I. He failed, although he was praetorian prefect, not allowed consularis or praesesularis. .

29 Ammianus 14.6.8, 21.10.6.

the political topography of the late antique city 323

erciv power.30 An alternative explanation for the predominance of statues erected to governors is that they represent a continuation of the civic honor tradition of statue dedication; they seem to reflect the governors' assimilation of great responsibility for civic life from the early 4th century, in a process by which they had come to look more like a mayor than Julius Caesar in Gaul or Vespasian in Judaea, the all-powerful representative for the Roman government in the provinces. Indeed, the physical appearance of the statue and the rhetorical aspect of the texts on its base seem to have little to do with its function. As Smith broadly acknowledges, most aspects of the appearance of the statues that have been identified from Aphrodisias and Ephesus seem to be shared with representations of notables who were not governors, and to be a reflection of wider court style; the rhetoric of the dedicatory inscriptions is likewise unspecified, reflecting the absolutist moral rhetoric seen in government laws. Thus, the appearance of the sculptures does not reflect specific characteristics of the governorship, but a more general image of power derived from the imperial court.

Further interpretive difficulties may arise for other monuments believed by modern scholars to relate to attempts to communicate 'power'. Noel Duval has suggested that the interest in an 'architecture of power' in imperial residential architecture may have its origins in contemporary experience with the architectural experiments of the 20th century. dictators.31 It may be anachronistic when applied to Late Antiquity. From the texts, it appears that imperial rank, formal symbols of office, visible clients, wealth, and distinction in classical learning were the keys to most political and social meetings, rather than the shape of one's audience hall. Reception rooms could certainly have helped to signify social power, especially in relation to clients; However, we should consider whether this architecture was primarily in communication power, or was partly more of a fashionable backdrop.

30 Governors appear to have ruled essentially by consent, with usually no day-to-day access to overwhelming power: Lendon (1997) 231–36 and Brown (1992) 22–25. Many governors were assassinated in the exercise of their functions: Amm. Marc. 14.7.6; Lib. Or. 20.3; Theophanes Chronology year 5891; Socrates, HE 7.14; John of NikiuChronicle 89.35. Praetoria attacked or burned: Caesarea: Malalas 18,119. Thessalonica (attempted): Malchus Frag 20:11-12. Antioch: Malalas 16.6; Constantinople A.D.532: Malalas 18.71. Other governors took over their metropolitan bishops or local aristocracy and lost badly: Lavan (2001b) 55.

31 Duval (1986) 465. See for example Brown (1992).

luke lavan324

Activity room with neglected secondary functions

The political and social functions of the churches

The political and social functions of late antique churches have been overlooked in favor of their religious role, the dominant factor behind their architectural design. This despite the importance of this subject for the understanding of the rise of the Church in the period. Theforum/agora certainly loses its prominent political and social position in the mid to later 5th century. By the middle of the 6th century, new forums were rare and were only a fraction of the size of their predecessors; some had fallen into disrepair, in otherwise monumental cities; many were built over.32 Statue dedications were now rare or non-existent. What has been less studied is what exactly happened to the spatial functions that once took place here - did they disappear entirely or were they moved elsewhere?

The answer probably lies partly in the growth of political and social activities around cathedrals and other churches, from the early years of the 5th century. From this time, imperial letters were read aloud in churches. From the second half of the 5th century, inscribed copies of imperial decrees, including some on secular affairs, are found at St Mary's and St John's cathedrals in Ephesus.33 Imperial statues, often displayed in the forum, traditionally provided an official focus for ceremonies proclaiming professed loyalty to the emperor and provided sanctuary for those who seized them as supplicants. The emperor was now honored by the prayers of the faithful within the church walls, where a shrine had also become available. The governor Andronicos was keenly aware of the implications of this practice of pre-episcopal power in his struggles with Synesius, metropolitan bishop of Ptolemais; he hammered his ordinances of denying asylum in churches to the very doors of the city's cathedral.34

32 France/gorai of the 6th century small: Justinian Prima (ca. 28m diameter); Arif, Lycia (approx. 15mx20m). Forum/agorai built over by churches: Aix: Guild et al. (1988) 17-64 Diana Veteranorum Duval (1977) 847-73; Pergamon Dörpfeld (1902) 16f and (1904) 114f; Knidos: Özgan (1998) 208 fig. 1. Decay and fora rebuilt in Africa, Italy and elsewhere: Potter (1995) 63-9 Decay at Apamaea, Gerasa, Petra: Liebeschuetz (2000) 43, 48,

33 imperial/royal letters read in the Church: Gen: John Chry. in ep. SEC. at Thess. 3.4; Edessa (possible) Chronicle of Joshua Stylites 34 (A.D. 498/99); Constantinople: Chron. pas. Olympiad 352 (A.D. 628); Vandal Africa: Victor of Vita 2.38. Their exhibition: Feissel (1999) 121-32.

34 Asylum by statues: Cod. Theodore. 9.44 (386 AD) = Cod. Just. 1.25. Shrine

the political topography of the late antique city 325

Court audiences before bishops had also developed in increasing importance, at least from the later 4th century, and took place either in episcopal residences or in or before churches. As noted above, meetings in the 6th century, which had replaced the boule, could now take place in the episcopal palace.35 Procopius reveals that, at least in major cities in the mid-6th century. the agora remained an important centre, perhaps still the main focus of social life.36 But the antechambers of the churches had also become popular meeting places; beggars could now be found here as well as in the agora.37 The existence of atria at several churches within a single city may therefore well have been a significant factor in the demise of many fora/agorai. However, it is only through an understanding of the functions of both these spaces that their development can be linked.

Non-architectural activity spaces

The law courts

By moving away from architecture to human action, one is able to consider non-architectural spaces of activity, such as those that took place on the street rather than in buildings, or in a variety of architectural settings. One of these non-architectural spaces was the provincial governor's court. In the life of a provincial capital it was of equal importance to the praetorium. The function of the court was not only to deliver the highest legal judgment in a given province, but it was also a place where social structures were published.

in churches: Cod Theod. 9.45 (392-432 AD). Prayers for the emperor in the church: continued from Tert. Apol. 30; well established in the 4th century: Ewig (1976) 3-71. Andronikos: Synesius Ep. 42.

35 Legal Audiences of Bishops: Jones (1964) vol. 1480-81. Before a church in Alexandria: Leontius N. v. Jo. Eleem 5. Governor perhaps emulating this practice in Edessa: Chronicle of Joshua the Stylite 29. Notables meet in secret of the bishop: see note 20 above.

36 Agora as an important element of a prosperous city, mid-6th century year: Procop. Aed.2.10.19-22, 3.4.18, 6.5.10-11, 6.6.13-16. Church and agora both as places to socialize: Procop. Anecdote 26.11. Merchants in the Atrium of the Cathedral of Edessa: Chronicle of Joshua the Stylite 32.

37 Beggars in the agora late 5th century to early 7th century: Constantinople Procop. Anecdota24.7; Alexandria Leontius N. v. Jo. Eleem 27. Also about churches: Rome Procop.Anecdota 26.28-30; Constantinople Procop. Anecdota 26.28-30; Edessa Chronicle of the Stylite Joshua 43; Alexandria Haas (1997) 253. This of course relates largely to the charitable activities of the Church, but is still an important change in social patterns of life.

luke lavan326

licly defined, in the battles of the well-to-do and through the selective use of torture, according to a person's social status as one of the honest or humiliores. Legal glasses would attract a lot of attention for this and other reasons; in the writings of Synesius and Libanius they represent a key event in the social and political life of a provincial capital; the court seems to have been a characteristic feature of the metropoleis, since in many and perhaps most provinces the old judicial journey seems to have ended.38 Libanius lists the courts, among porticoes, libraries, and temples, as being the most important among the attributes of Nicomedia, the metropolis of Bithynia, after the earthquake of 358; it was also one of the three places in Alexandria, along with the theater and the bouleuterion, from which parabalani was banned in AD 416.39.

The substantive prominence of the court has also been unfairly overlooked. The architectural settings of late antique law courts varied: they were held in civil basilicas, in bathhouses, and there is one example of a governor's court held in a church, raising the possibility that very "Roman" activities could have been carried out out of not -traditional spaces in Late Antiquity.40 However, the law courts were most strongly visually defined by their non-architectural aspects. It is possible to reconstruct these in great detail from texts, mainly of 4th century date and depictions of 5th and 6th century date. The texts give a lot of detail about the people usually present at public hearings: the governor, who sat alone or with honoraria; thecornicularius, who presided over the court; the assessors, who gave the governor legal advice; the exceptions, take notes; the lawyers, the witnesses and the accused. We hear at least the basics of the procedures followed: speeches by lawyers, inquisitorial questions by the governor, and the use of torture, which is recorded vividly.41 Image-

38 See Lavan(2001b) 51-53 for the likely end of the governor's assize tour in most regions.

39 Se jordskælvsbeskrivelsen Libanius Or. 61,7-10. Alexandria A.D. 416: Cod. Theod. 16.2.42.

40 Courts in bath buildings in Byzantium and Antioch in the 3rd century: Feisseland Gascou (1995) 67-72, 78-89. In Edessa in winter baths in 497/98: Chronicle ofJoshua Stylites 43. A court built into the 'Byzantine City Bath' at Ptolemais may have served this purpose: Kraeling (1962) 166; it may also have the hall of the western bath building at Scythopolis, which is covered with smaller mosaic dedications of governors: Mazor (1987-88) 14-17. Law court held in/before the church in Edessa 497/98: Chronicle of Joshua Stylites 43.

the political topography of the late antique city 327

tions from the 5th and 6th centuries show the governor on a throne or bench, with a court or a footstool. He is consistently flanked by guards. In the 'Christ and Barabbas' court scene from the Rossano Gospels, these soldiers have imperial pretensions. In three depictions, a table with a tablecloth (twice white) is placed before the governor, probably to receive the imperial portraits that gave the court authority; twice the cloth itself is decorated with them.42 The governor, in almost all depictions, grasps a scroll, probably his codicil, his letter of appointment from the emperor. The wearing military robes, with diamonds of different colors, also give him and his officials imperial rank. Other aspects of the court, such as an exceptor taking notes on writing tablets (twice) and a group of lawyers in togas (once) are also depicted.43

Although we have no secure depictions of honorati sitting on the governor's bench in the 6th century, there is much in common between the literary and surprisingly consistent pictorial evidence. This suggests that the visual image of the court was sharply defined in contemporary consciousness. Architecture receives almost no attention in these depictions. It thus seems clear that the non-architectural aspects of the courts were most significant in its operation - the building used could vary; it conferred none of the authority that official imperial portraits brought, and inspired little of the horror provided by the crests, racks, and canes of the governor's torturers.44

Adventen

Another non-architectural activity space that has actually received a lot of attention for both its spatial function and appearance is theadventus. This was the welcoming ritual by which cities greeted in-

41 Textual sources on 4th century courts: Dionisotti (1982) 104-105, Lanata (1973), Jones (1964) vol.1 517-522 Lavan (2001b) Libanius Or. 33 and 56. Honorati sit with the Governor: Lib. Or. 56.4 see Liebeschuetz (1972) 190. Torture: eg Dionisotti (1982) 105; Jerome Ep. 1; Synesius Ep. 41.

42 Such tables, with codicils bearing imperial portraits, appear in the Notitiadignitatum as part of the symbols of office of several governors: Berger (1981) fig.1, 23, 24, 46, 48.

43 For visual details of the governor's court see not only the Rossano Gospels: Loerke(1961) 171-95, but other depictions such as the situla in the British Museum, London:no. 267 or the depiction of Joseph as governor of Egypt on Archbishop Maximian's cathedra in Ravenna. Torture apparatus, 4th century: Dionisotti, op. cit. 105, lines 75 and 119. Jerome Ep. 1. Synesius Ep. 41.

44 Like the priest, the bread and a communion cup could be the most important spatial elements in defining a church in the consciousness of Late Antiquity, rather than columns, domes or mosaics: e.g. Athanasius Apol.c.Ar. 11; Socrates HE 1.27.

luke lavan328

would-be dignitaries galore, whether emperor, governor or bishop. It is attested throughout Late Antiquity, and an example of an adventus to receive imperial images is even recorded in AD 787. In this ritual greeting, communities could demonstrate their appreciation or voice grievances through coordinated song. Through careful choreography and precedence, they could also illustrate and confirm the relative social characters of their city. As such, it was one of the most important political activities that took place in a city, which could define a city's relationship with a new ruler. Unpopular governors or bishops sought to avoid the adventus by entering secretly, even though such a gesture could lead to violent public protest.

Texts give us lots of spatial information about material objects used in adventus, such as the dress of the reception party, the burning of incense, and held candles, and sometimes even tell us about hymn singing. The order of reception for different classes of people is also known and seems to have been standardized to some extent. Here it cannot be claimed that a preoccupation with buildings has caused the neglect of either the function of this ritual or its non-architectural material aspect. The adventus of both governors and emperors has been well researched, and it is not necessary to repeat this work here.46 Rather, I wish to suggest that the separation of textual and archaeological studies has led to a neglect of the urban architectural framework for this and other public ceremonies. Despite not being surrounded by a discrete building, the adventus seems to have had a strong relationship with the gates, colonnades and other monuments of a city. For the imperial adventus, and probably that of governors, the route was defined in advance and specially decorated.47 Events unfolded at different stages depending on where the procession was. Usually, it seems that at least some of the councillors, clergy and some crowds could greet dignitaries from outside the city, perhaps from a fixed point such as the boundary of the urban area. His passage at the gate would be very congested, as it was a crossing of the spomoerium of the city. People either came out to meet the governor or greeted him right inside the city, from roadsides and rooftops.

45 AD 787: Mansi 12.1013.46 General introduction: S. MacCormack (1972); Delmaire (1996) 42-45, som

lists other relevant bibliography.47 Imperial adventus route planned and decorated: Proclus Or. 9.148 Acclamations to a governor just after passing the gates: Lib. Or. 56.15, 20.17,

the political topography of the late antique city 329

tional status and could be recorded.49 In fact, late antique inscribed acclamations have been recorded on a gate at Hierapolis, in Phrygia, and on the Golden Gate in Constantinople.50 At some point in the ritual, governors were expected to descend from their ceremonial chariots to greet members of the city council; in Antioch this occurred before the bouleuterion, where members of the city council were appointed to meet him.51 In cities where members of the imperial family were present, officials were also expected to visit the palaces as part of their adventus, and if their cortège passed without doing so, this was interpreted by the public as a great miss.52

It is clear that the adventus had a non-arbitrary relationship to a number of key points outside and inside a city; this opens up the possibility that aspects of the monumentality of late antique cities may indeed be related to these important processions and other aspects of public ceremonial. Charlotte Roueché has recently pointed to a range of epigraphic evidence for late antique public ceremonies at Ephesus and Aphrodisias; the acclamatory inscriptions she has studied seem to occur in very definite places, in contiguous groups; they may well represent the positions where various forms of acclamation actually took place.53

I would argue that the monumentality of the Embolos of Ephesus, with its exceptional concentration of governor statues and late antique decrees, could also be interpreted ceremonially in relation to the adventus of the proconsul, the most important secular ceremonial event in the life of this city. A 5th century proconsul on his first visit would probably arrive via the harbor (fig.1 of Ida Leggio's paper, no. 9) and from there towards the bouleuterion (no. 26), the most critical stop before ending up at his praetorium. 54 He wanted to process down first

25.19. Folk at Rooves: Wallis (1915) 586; 143, 95–96. Board

49 Standardized acclamations: Seeck (1906) 84-101. Acclamations recorded: Cod. Theod. 1.16.6 (331 AD) = Cod. Just. 1,140.3; Collection Avellana 14 (A.D. 418); ACO 1.4201 n.274 (c. A.D. 433).

50 Inscribed on gates: Roueché (1999a) 131-36.51 Governors dismount to greet the council of Antioch in the council chamber:

Libanius Or. 46.40.52 Ammianus 14.7.10; 14.11.14.53 Roueché (1999b)

had the right from Caracalla to receive the proconsul before all other cities onward

luke lavan330

the broad Arcadiane (No. 10), in his official carriage. This would have been an ideal processional space where welcoming crowds could gather to catch a glimpse of the proconsul and greet his arrival. I do not think it is a coincidence that this great monumental street, at least partly late antique, connects the city center to the harbor rather than to one of the gates; it seems to have been built to allow a processional entry from this direction. Acclamations may have been given and inscribed on this great avenue or in the theater (no. 11) to which it led.55 From the theater square the governor would then pass down the street on the east side of the Lower Agora. Here Roueché has identified a series of spatially structured imperial acclamation inscriptions from the early 7th century. date; she suggests a possible explanation is that they may have been associated with the reception of imperial portraits. It may well be that even as late as the 5th century the notions given by the emperor to the governor and carried by him into the province were formally received here.56 Near this point a 5th century governor would have had to to abandon his coach and go up the pedestrian Embolos (No. 21) on foot to reach the bouleuterion. Climbing Embolos he would have come face to face with a series of monumentalized imperial decrees granting favors to the city (at no. 16). After this he would have been forced to confront a remarkable display of statue after statue of late antique governors (at No. 21), hitherto unparalleled in the Mediterranean, reminders of his predecessors who had behaved well enough to be honored; one can imagine his companions since the docks, the leading honors of the city, providing a running commentary. These men could make or break a governor's tenure; they would have been glad to remind him of their power as he approached his meeting with the city fathers at the bouleuterion itself (no. 26), on the Upper Agora.57

The reconstruction of Adventus offered here is beyond doubt

arrival; this privilege may have been jealously guarded at the time the Digest was compiled. Ulpian de officio proconsulis book 1 in Digest 1.16.7. The port would probably have been the natural port of entry if it came from Constantinople or from Farfield.

55 Roueché (1999a); for a recent study of the Arcadiane see Schneider (1999).56 Roueché: (1999b) 163. Performances given by the emperor: Severian de Mundi creatione

Or. 6.5 (PG 56 489)57 Embolos: for the archeology of this area see: Bauer (1996) 284-89. Arch

blocking Embolos, undated: IvE 2.587: epigraphic and stylistic arguments have been advanced for a date in AD 459, but it could be earlier; Bammer (1976-77) statues are certainly more frequent towards the arch, suggesting that they were placed here for pedestrian appreciation.

the political topography of the late antique city 331

tentative, even suggestive. However, other secular street processions took place in late antique cities, such as those of aristocrats in coaches followed by clients, or of governors with their officials walking from their praetorium to the courts.58 The location of such rituals in an urban context may be to explain much of how the late antique streets work in the future; they can provide an understanding of the context of street monuments and graffiti, which are still poorly understood but, as Charlotte Roueché has demonstrated, far from arbitrary in their placement. Likewise, it might one day be possible to understand the eventual encroachment of monumental streets in terms of changing or ending old procession patterns.

Conclusion

The political "spaces of activity" presented in this article have of course not been exhaustively explored. Many other everyday activities and rituals could be mentioned, such as those relating to public entertainments. The sketches provided here have been provided to illustrate the potential that concepts of human space have to change the way we think about the late antique city. My conclusions are both methodological and historical.

In terms of methodology, it is clear that much of our knowledge of late antique urban topography has simply relied on crude patterns in the survival of archaeological evidence; To improve this, we must concentrate on actively writing about topography, not only through the collection of evidence, but through argumentation, critically evaluating different kinds of sources to create a human spatial narrative for Late Antiquity. The use of texts as well as archeology to pursue this can enable important aspects of urban life to be recovered from neglect or obscurity: such as the courts or the social functions of the churches; without these we cannot have a balanced appreciation of the late antique city. By consciously weighing different types of evidence, we can further change our understanding of seemingly familiar issues such as the decline of the forum/agora. The combination of texts and archeology also allows a living

58 Aristocratic Chariot Processions with Clients: Alexandria A.D. 415: Damascius fr. 104; Rome, 4th century: Ammianus 14.6.16-17; Antioch, Libanius Or.3.13 – implying that it is shameful to have only 20 people following.

luke lavan332

evocation of human experience in the 4th to 6th centuries. Potentially, the disciplined organization of anecdotal textual references may one day allow both high places and sheds in late antique cities to be filled not only with activities, people and objects, but also with sounds and smells.59

On a general historical level, an examination of spatial function, through-texts and evidence of the use of buildings reveals more continuity with the Roman past; this is an important counterweight to studies of Christianity; traditional political activities, carried out in ancient architectural settings, must not be neglected; nor those transferred to new architectural frameworks as secondary functions must be ignored. An explicit historical method also helps to establish a clear periodization in the urban organization of the late antique city. Although Mediterranean cities emerge from the 3rd century. at a new more modest level they preserve a very stable Roman civic monumentality from the time of Diocletian up to the early 5th century; this time sees the increasing Christianization of the city. Then, from the middle of the 5th century, the traditional civil political buildings are first neglected and then rebuilt; however, secular public buildings of a non-political nature, such as baths and street monuments, remained important in the late antique city up to the 7th century. The end of the bouleuteria andagorai can be seen as equivalent to the earlier decline of the stadiums and gymnasium. Both mark very specific changes in the priorities of people who lived in already old urban environments but wanted to reshape them for their own use.60

Recognitions

I thank Penelope Allison, William Bowden, Simon Ellis, Helen Gittos, Carlos Machado, Gregoire Poccardi, Jean-Pierre Sodini and BryanWard-Perkins for reading this text. They are, of course, exempt from the opinions expressed in it, from which they differ; errors remain my own. I am also grateful to the Leverhulme Trust for funding my stay in Paris during the academic year 2001-2002 during which these two articles were written.

59 Locating animal carcass dumps is a start: for extra mural dumps in Gaza: Sozom. HE 5.9, for dumps inside the city see Mango (1976) 30-35.

60 Here I must emphasize again that I am talking about the Mediterranean, especially the central and eastern regions.

the political topography of the late antique city 333

Bibliography

Abbreviations

ACO = E. Schwartz (1922f.) ed. Proceedings of the Ecumenical Council (Berlin and Leipzig 1922f.).

CSEL = Corpus Scriptorum Ecclesiasticorum LatinorumCIL = Corpus Inscriptionum LatinarumIRT = J. M. Reynolds and J. B. Ward-Perkins (1952), Inscriptions of Roman

Tripolitania (Rome 1952) IvE = Die Inschriften von Ephesos (Bonn 1979-1981) Mansi = J. D. Mansi (1797) ed. The New and Comprehensive Collection of the Holy Council Volume 12

(Paris 1901; Graz 1960) (2nd edition; original edition 1797).PG = Patrialogia GraecaPL = Patrialogia LatinaPO = Patrialogia Orientalis

Editions

Ammianus = J. C. Rolfe (1935-39) overs. Ammianus Marcellinus i tre bind (London and Cambridge, Mass. 1935-39).

Athanasius Apol. c. he = H. G. Opitz (1935-38) udg. Athanasius works (Berlin and Leipzig 1935-1938) vol. 2:1 87-168

Athanasius Hist. he = H. G. Opitz (1935-38) udg. Athanasius works (Berlin and Leipzig 1935-1938) vol. 2.1 183-230

Augustine Conf. = W. Yates (1946) ed. and translation. St Augustine's Confessions in TwoVolumes (London and Cambridge, Mass. 1946).

Chronicle of Joshua Stylites = A. Luther (1997) overs. The Syriac Chronicle of Joshua Stylites (Berlin 1997).

Chronicon Pascale = Mi. Whitby and Ma. Whitby (1989) translation. Chronicon Paschale284-628 AD (Translated Texts for Historians Volume 5) (Liverpool 1989).

Cod. Theodore. = T. Mommsen and P. Meyer et al. (1905), Theodosian Libri XVI with the Sirmondian Constitutions (Berlin 1905)

Torsk Read = P. Krüger (1887) edg. Corpus Juris Civilis volume 2: Codex Justinianus (Berlin 1887).

Collection Avellana = Avellana Collection - Epistles from other emperors and popes... Avellanaquae is called the Collection (CSEL 35)

Damascius = C. Zintzen (1967) ed., Damascii, Vitae Isidori reliquiae (Hildesheim 1967).Digest = Th. Mommsen (1870) ed., Justinian, Digesta seu Pandecetae (Berlin 1870).Jerome Ep.= PL 22John Chry. and Matt. = PG 56-57John Chry. i ep. sec. ad Thess. = PG 62Lydus Mag. = T. F. Carney (1971) ed. and trans. John the Lydian on the Magistrates

of the Roman Constitution (from magistratibus) (Coronado Press, no place 1971). Chronicle of John of Nikiu

don 1916).Julian Misop. = W. C. Wright (1913) ed. and translation. Julian Works Volume 1 (London

and Cambridge, Mass. 1913) Leontius N. v. Jo. Eleem = E. Dawes and N. H. Baynes (1948) ed. and translation. From

luke lavan334

Three Byzantine Saints: Contemporary Biography of St. Daniel the Stylite, St. Theodoreof Sykeon and St. John the Almsgiver (London 1948)

Lib. Eller = R. Foerster (1903-22) ed. Libanius Opera bind to cry. (Leipzig 1903-22). Lib. Ep. = R. Foerster (1903-22) ed. Libanius Opera bind to cry. (Leipzig 1903-22). Lib. Bro = (ed.) L. Duchesne (1886-1957) Le Liber Pontificalis 3 volume (Bibliothéque

French schools in Athens and Rome) (Paris 1886-1957). Life of Theodore of Sykeon = A. Festugière (1970) ed. Life of Théodore de Sykéon (2 vols., SubsidiaHagiographica 3) (Brussels 1970).

Malalas = E. Jeffreys, M. Jeffreys og R. Scott (1986) red. The Chronicle of JohnMalalas (Melbourne 1986).

Malchus, Testimonia = R. C. Blockley (1983) transl. og udg. The Fragmentary ClassicisingHistorians of the Later Roman Empire vol. 2 (ARCA 10) (Liverpool 1983).

Job announcement = O. Seeck (1962) ed. The information about the dignitaries is added to the information about the city of Constantinople and the tiles of the provinces (Berlin 1876) (reprinted Frankfurt 1962).

Proclus Or. = PG 65Procop. Aed. = H. Dewing (1954) ed. Procopius. Buildings (London and Cambridge,

Lot. 1954).Procop. Anecdote = (ed.) H. Dewing, Procopius. The Anecdote or the Secret History (London

and Cambridge, Mass. 1960). HE = PG21 Severian on the creation of the world Or. 6 = PG 56 Pg. Apollo Ep. = W. B. Anderson (1936-65) ed. and translation. Sidonius, Poems and

letters (London and Cambridge, Mass. 1936-65) Socrates HE = PG 67 Sozom. HE = J. Bidez and G. H. Hansen (1960) ed. Sozomenus, Church History

(Greek Christian Writers Volume 50) (Berlin, 1960) Synesius Ep. = A. Garzya (1989) ed. Operate the Synesios of Cyrene. E-pistol operates he

(Torino 1989) Tert. Apol. = G. H. Rendall (1931) trans. Tertullian: Apology and Spectacles (Loeb

250) (London and Cambridge, Mass. 1931) Theoderet HE = PG 83 Theophanes Chronographia = C. Mango and R. Scott (1997) ed. Chronicle of

Theophanes the Confessor (Oxford, 1997) Tzetes Scholia i Lycophron Alexander = E. Scheer (1881) ed., Lycophron Alexandra

Vol. 2 (Berlin 1881)) V. Caesarii = W. E. Klingshirn (1984) transl. Caesarius af Arles: Liv, Testamente, Breve

(Liverpool 1984) Victor of Vita = J. Moorhead (1992) ed. and translation. Victor of Vita: History of the Vandals

Persecution (Translated Texts for Historians vol.10) (Liverpool 1992). Zach Mytilene HE = F. Hamilton and E. Brooks (1899) transl. The Syrian Chronicle

known as Zachariah of Mitylene (London 1899). Zach Mytilene v. Sev. = PO 2

Secondary sources

Balty J. Ch. (1991) Curia Ordinis. Research into Ancient Architecture and Town Planning in the Provincial Curia of the Roman World (Brussels 1991).

Bammer A. (1976-77) "A late antique gate building from Ephesus", ÖJh 51 (1976-77) supplement 119-22.

the political topography of the late antique city 335

Banaji J. (2001) Agrarian Change in Late Antiquity. Goldwork and Aristocratic Dominance (Oxford 2001).

Bartoccini R. (1950) "The curia of Sabratha," Quaderni di Archeologia della Libia 1 (1950)29-35

Bartoli A (1963) Curia senatus. Lo scavo et il restauro (Rome 1963) Bauer F. A. (1996) City, Square and Monument in Late Antiquity. Equipment studies

of public space in the late antique cities of Rome, Constantinople and Ephesus (Mainz 1996).

Berger PC (1981) The Insignia of the Notitia Dignitatum (New York and London1981).

Bowman A. K. (1971) The Town Councils of Roman Egypt (Toronto 1971).Brown P. R. L. (1992) Power and Persuasion in Late Antiquity: into a Christian Empire

(London 1992) Chastagnol A. (1978) L'Album Municipal de Timgad (Antiquity III, 22) (Bonn 1978) Clairmont C.W. et al. (1975) Excavations in Salona, ​​Yugoslavia (1969-1972) (Park

Ridge, New Jersey 1975). Delmaire R. (1996) "Some Aspects of Communal Life in the Lower Empire", i

Splendidissima civitas: studies of Roman history in tribute to Francois Jacques, edd.A. Chastagnol, S. Demougin and C. Lepelley (Paris 1996) 42-45.

Dionisotti A. C. (1982) "From the school days of Ausonius? A school book and its relations," JRS 72 (1982) 83-125.

Dörpfeld W. (1902) "The Work at Pergamon 1900-1901: The Buildings", AM27(1902) 16-43.

——— (1904) "The Work at Pergamon 1902-1903: The Buildings", AM 29 (1904)114-51.

Downey G. (1961) A History of Antioch in Syria from Seleucus to the Arab Conquest (Princeton1961).

——— (1977) "La 'basilique' de Zana (Diana Veteranorum)," MEFRA 89 (1977) 847-73.

Duval N. (1986) "Is there a 'palatial structure' specific to Late Antiquity", in The palatial system in the East, in Greece and in Rome, Proceedings of the Colloquium of Strasbourg (TCRPOGA 9) Strasbourg, 1985 ed. E. Levy (Strasbourg 1986) 463-90.

Ewig E. (1976) "On the Christian Idea of ​​Kings in the Early Middle Ages", in Late Antiquity and Frankish Gaul vol 1 (Munich 1976) 3-71.

Feissel D. (1999) "Administrative Epigraphy and Urban Topography: The Location of Inscribed Deeds in Proto-Byzantine Ephesus (4th-6th Centuries)", in Efeso paleocristianae byzantina, ed. R. Pillinger et al. (Vienna 1999) 121-32.

Feissel D. and Gascou J. (1995) "Unpublished Roman Archival Documents from the Middle Euphrates," JSav (1995) 67-78.

Feissel D. og Kaygusuz I. (1985) "A Sixth-Century Imperial Warrant", TravMém9 (1985) 397-419.

Foerster G. and Tsafrir Y. (1988/89) "Hebrew University expedition" in "Bet Sheanproject - 1988," Excavations and Surveys in Israel 7-8 (1988-89) 15-32.

Foss C. (1979) Ephesus after Antiquity (Cambridge 1979) Frantz A. (1988) The Athenian Agora, vol. 24: Senantik (Princeton 1988) Gauthier N. (1999) "Christian Topography between Ideology and Pragmatism",

in The Idea and Ideal of the City between Late Antiquity and the Early Middle Ages. G. P. Brogiolo and B. Ward-Perkins (The Transformation of the Roman World 4) (Leiden, Boston and Cologne 1999) 195-210.

Goodchild R. G. and Kraeling C. H. (1962) "The 'Odeon' and Related Structures", in Ptolemais: City of the Libyan Pentapolis, C. H. Kraeling (Chicago 1962) 91-95.

Guild R et al. (1988) "Saint-Sauveur d'Aix-en-Provence", CAF 143 (1988) 17-64.

luke lavan336

Haas C. (1997) Alexandria in Late Antiquity (Baltimore 1997) Horster M. (1998) "Ehrungen's Splendid Statthalter", AntTard 6 (1998) 37-59. flights.Kent J. H. (1966) Corinth vol. 8 part 3, The Inscriptions 1925-1950 (Princeton 1966) Kraeling C. H. (1962) Ptolemais: City of the Libyan Pentapolis (Chicago 1962) Laniado A. (2002) Research in Municipal Notables in the Proto-Byzantine Paris Empire

2002). Lavan L. (1999) "The residences of late antique guvernors: a gazetteer", AntTard

7 (1999) 135-67.——— (2001a) "The civil governors' praetoria in Late Antiquity", in Nylig-

search in Late Antique Urbanism, ed. L. Lavan (JRA Supplementary Volume 42) (Portsmouth, Rhode Island 2001) 39-56.

——— (2001b) Provincial Capitals of Late Antiquity (PhD thesis Nottingham 2001)Lendon J. E. (1997) Empire of Honour. The Art of Government in the Roman World (Ox-

ford 1997) Lepelley C. (1979) Les cités de l'Afrique romaine au Bas-Empire (Paris 1979) 2 vols Lewin A. (c. 1995) Popular Assemblies and Political Struggle in the Cities of the Roman Empire (Florence)

c. 1995).——— (2001) “Public Urban Building from Constantine to Julian: the Epigraphic

bevis”, i Recent Research in Late Antique Urbanism, red. L. Lavan (JRA Supplementary Volume 42) (Portsmouth, Rhode Island 2001) 27-37.

Liebeschuetz J.H.W.G. (1972) Antioch. City and Imperial Administration in the Later Roman Empire (Oxford 1972).

——— (2000) "Late Late Antiquity in the Cities of the Roman Near East", MediterraneoAntico 3.1 (2000) 43-75.

——— (2001) The Decline and Fall of the Roman City (Oxford 2001). Loerke W. C. (1961) "The miniatures of the trial in the Rossano Gospels," ArtB

43 (1961) 171-95. Loseby S. T. (1996) "Arles in Late Antiquity: Gallula Roma Arelas and the City of Genesis".

in Towns in Transition: Urban Evolution in Late Antiquity and the Early Middelalder, red. N. Christie og S. T. Loseby (Aldershot 1996) 45-70.

MacCormack S. (1972