The Sixth Century: Production, Distribution and Demand 9004109803, 9789004109803

Table of contents :
Henri Pirenne and the question of demand in the sixth century / Richard Hodges 3
Reading Pirenne again / Paolo Delogu 15
The production and distribution of books in late Antiquity / Carlo Bertelli 41
The Migration Period: model history and treasure / Klavs Randsborg 61
Les conditions du commerce au VIe siècle / Jean Durliat 89
The destinies of the late Antique Italies: politico-economic developments of the sixth century / Federico Marazzi 119
Eastern Spain in the sixth century in the light of archaeology / Sonia Gutiérrez Lloret 161
Les echanges dans la Gaule du Nord au VIe siècle: une histoire en miettes / Stéphane Lebecq 185
Marseille and the Pirenne Thesis, I: Gregory of Tours, the Merovingian kings and 'un grand port' / S. T. Loseby 203
The frontiers of western Europe: developments east of the Rhine in the sixth century / Ian Wood 231
The Justinianic era of south Scandinavia: an archaeological view / Ulf Näsman 255
Overview: production, distribution and demand / Chris Wickham 279
Index 293

Citation preview

This collection of essays examines the sixth century A.D. from a new perspec­ tive. Being a result of the European Science Foundation’s programme devoted to the transformation of the Roman World, the authors examine the economic and social conditions of a century which has often been overlooked. The book takes a European overview, and includes studies by archaeologists and historians whom, in the course of the ESF project, have developed a lively dialogue focussing upon the issue of demand in the sixth century. An archaeologist poses many of the leading arguments in the first chapter, and an historian draws these themes together in the final one. The book includes a major review of the historio­ graphy of Henri Pirenne’s celebrated thesis devoted to the decline of the Roman empire and the beginnings of the Middle Ages. The majority of the essays, however, are regional studies approaching the subject with a new wide-angled, European vision. Richard Hodges is Professor in the School of World Art Studies and Museology, University of East Anglia and Director of the Prince of Wales’s Institute of Architec­ ture. His main archaeological excavations are at Butrint, Albania and San Vincenzo al Volturno, Italy. His books include Dark AgeEconomics ( 1982), The Anglo-Saxon Achievement (1989) and Light in the Dark (1997). William Bowden trained at the Institute of Archaeology, University of London before working as an archaeologist at the British School at Rome. He is currently a re­ search scholar at the British School at Athens studying Epirus in late antiquity.

TH E SIXTH CENTURY

THE TRANSFORMATION OF THE ROMAN WORLD A SCIENTIFIC PROGRAMME OF THE EUROPEAN SCIENCE FOUNDATION

Coordinators JA V IE R A R C E • EV A N G ELO S C H R Y S O S • IA N W O O D Team Leaders

Steering Committee

Gunilla Äkerström-Hougen Volker Bierbrauer Niels H annestad Przemystaw Urbañczyk Mario Mazza H .H . van Regieren Altena Heid Gjöstein Resi L. Cracco Ruggini

Miquel Barceló M ark Blackburn Gianpietro Brogiolo Alain Dierkens Richard Hodges M arco Mostert Patrick Périn W alter Pohl Frans Theuws Leslie Webster

Series Editor IA N W O O D VOLUME 3

T H E SIX T H CEN TU RY

H enri PIREtN N E Caricature of Pirenne from Pourquoi Pas?

THE SIXTH CENTURY ,

Production Distribution and Demand

EDITED BY

RICHARD HODGES AND

WILLIAM BOWDEN

BRILL LEIDEN • BOSTON • KÖLN 1998

This book is printed on acid-free paper.

Library o f C ongress C ataloging-in-Publication D ata The sixth century : production, distribution and demand / edited by Richard Hodges and William Bowden. p. cm. — (The transformation of the Roman world, ISSN 1386-4165 ; v. 3) Includes bibliographical references and index. ISBN 9004109803 (cloth : alk. paper) 1. Rome—Civilization—Foreign influences. 2. Rome—History-Germanic invasions, 3rd to 6th centuries. 3. Rome—Economic conditions. I. Hodges, Richard. II. Bowden, William, m . Series. DG504.S56 1998 937—dc21 97-40120 CIP

D ie D eutsche B ibliothek - C IP-Einheitsaufaahm e The sixth century : production, distribution and demand / ed. by Richard Hodges and William Hodges. - Leiden ; New York ; Köln : Brill, 1998 (The transformation of the Roman world ; Voi. 3) ISBN 90-04-10980-3

ISSN 1386-4165 ISBN 90 04 10980 3 © Copyright 1998 by Koninklijke Brill N . V., Leiden, The Netherlands AU rights reserved. No part o f this publication may be reproduced, translated, stored in a retrieved system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission from the publisher. Authorization to photocopy items for internal or personal use is granted by Brill provided that the appropriatefees are paid directly to The Copyright Clearance Center, 222 Rosewood Drive, Suite 910 Danvers M A 01923, USA.

Fees are subject to change. PRINTED IN THE NETHERLANDS

CONTENTS Henri Pirenne and the question of demand in the sixth cen tu ry ..................................................................... Richard Hodges

3

Reading Pirenne again................................................................. Paolo Delogu

15

The production and distribution of books in late Antiquity. . . Carlo Bertelli

41

The Migration Period: model history and trea su re................. Klavs Randsborg

61

Les conditions du commerce au Vie siècle............................... Jean Durliat

89

The destinies of the late Antique Italies: politico-economic developments of the sixth ce n tu ry ........ Federico Marazzi Eastern Spain in the sixth century in the light of archaeology. Sonia Gutiérrez Lloret

119 161

Les échangés dans la Gaule du Nord au VIe siècle: une histoire en m iettes............................................................ Stéphane Lebecq

185

Marseille and the Pirenne Thesis, I: Gregory of Tours, the Merovingian kings and “un grand port” ........................ S.T. Loseby

203

The frontiers of western Europe: developments east of the Rhine in the sixth century........................................... Ian Wood

231

The Justinianic era of south Scandinavia: an archaeological view ............................................................ Ulf Näsman

255

Overview: production, distributionand dem and....................... Chris Wickham

279

I n d e x ...........................................................................................

293

Fig. 1. Commemorative medallion of Henri Pirenne. (Courtesy the Director, Academia Belgica, Rome).

HENRI PIRENNE AND THE QUESTION OF DEMAND IN THE SIXTH CENTURY Richard Hodges

‘As I turn, however, to a new endeavour which is fraught with difficulty and is in fact extraordinarily hard to cope w ith ,.......I find myself stammering and shrinking as far from it as possible, as I weigh the chances that such things are now to be written by me as will seem neither credible nor possible to men of a later generation; and especially when the mighty stream of time renders the story somewhat ancient.’ Procopius, The Anecdota}

Introduction One theme which a project devoted to the ‘transformation of the Ro­ man world’ could not overlook is the question of production and dis­ tribution. The importance of economic history in the twentieth-century historiography of this period requires no introduction. The shadow of the Pirenne thesis is, if anything, as far-reaching today as it has ever been. A glance at recent studies in this field confirms this. Apart from historians of the transformation responding to this theme,12 numisma­ tists3 and archaeologists4 still find Pirenne’s overarching model seduc­ tive in parts. In some respects, Pirenne’s theme, rather like Edward Gibbon’s Decline and Fall, illthough its argument is conspicuously flawed, has stood the test of time because it binds together antiquity and the Middle Ages, southern and northern Europe, history and ma-

1 Procopius, The Anecdota, ed. and trans. H.B. Dewing (London, 1935), i. 4. 2 R. Van Dam, “The Pirenne thesis and fifth-century Gaul”, Fijih-Century Gauh a crisis of identify, eds. J. Drinkwater and Hugh Elton (Cambridge, 1992), pp. 321-33; A. Verhulst, “The origins of towns in the Low Countries and the Pirenne thesis”, Past and Present 122 (1989), pp. 3-35; G. Petralia, “A proposito dell’immortalita di ‘Maometto e Carlomagno’ (o di Costantino)”, Storica 1 (1995), pp. 37-87. 3 D.M. Metcalf, “The beginnings of coinage in the North Sea coasdands: a Pirennelike hypothesis”, The Twelfth Viking Congress. Birka Studies Voi 3, eds. B. Ambrosiani and H. Clarke (Stockholm, 1994), pp. 196-214; Thomas S. Noonan, “The Vikings in the East: Coins and Commerce”, The Twelfth Viking Congress. Birka Studies Voi. 3, ed. B. Am­ brosiani and H. Clarke (Stockholm, 1994), pp. 215-36. 4 R. Hodges and D. Whitehouse, Mahomet, Charlemagne et les origines de l’Europe (Paris, 1996).

4

RICHARD HODGES

terial culture. It bridges historical and geographical divisions. Needless to say, the thesis naturally serves the purposes of the European Science Foundation in this project. It is inappropriate to develop a critique of the durability of Pirenne’s thesis here. It is evident that many other historians were engaged in equally ground-breaking research as Henri Pirenne fashioned his rath­ er simplistic argument.5 This territory had already fascinated Werner Sombart and Max Weber, Alfons Dopsch and Christopher Dawson. At the same time, Pirenne made ineffective use of the available mate­ rial evidence, disregarding, for example, the work of Rostovzeff in the Mediterranean and Holwerda in the Netherlands. But, paradoxically, these flaws have worked to Pirenne’s advantage. Mohammed and Charle­ magne is a sketch wherein each of us can find some fulfillment in its holistic scope. But why does Pirenne’s unfinished sketch continue to hold such fas­ cination for us when its central tenets are so palpably flawed? The an­ swer has much to do with the present state of the historical sciences. In comparison with the ‘twenties and ‘thirties, there is a hugely in­ creased academic community concerned with this period of European history. The social dynamics of studying the period have changed. Ac­ ademic demands have altered as a result. We are all familiar with the individualized thrust of our disciplines, and the response to increased student numbers as Europe shifts from its emphasis upon manufactur­ ing to service industries. This is an age of deconstruction and inter­ disciplinary fissioning. The romance of Europe - as a historical theme and, simultaneously, an economic project - has been overshadowed in the post-war years by the preoccupation with dominance in some form or other. The individual’s part in this romance is in jeopardy, or so the thesis goes. In academic terms, the argument goes: no individual can master the source material; look at the flaws in Pirenne’s thesis. Not surprisingly, the politiceli trend of late twentieth-century Europe is nei­ ther global nor nationalistic in outlook, but regional. In every respect, our society has shunned the empires of our immediate past, and the unedifying taint of colonialism. Equally unsurprisingly, the trend in the historical sciences mirrors this, albeit, paradoxically, while greater use is made of the Internet’s World Wide Web! In the strict sense, then, the ESF is at odds with our times: it aims to sustain the romance of

5 Sec Petralia, “A proposito dell’immortalità di ‘Maometto e Carlomagno’ (o di Cos­ tantino)”.

HENRI PIRENNE AND THE QUESTION OF DEMAND

5

European collaboration. It brings together scholars from different na­ tionalities, who have a multiplicity of specialisms. It aims to discover some common ground beyond the parameters of the individual’s na­ tionalistic education and scholarly development. The barriers between disciplines as between nations are not easily broken down. In the hu­ manities, in particular, such common ground is not easily identified, let alone acceptable to all. This is why Pirenne’s thesis has such great status. It is a polyglot answer to the conception of our Europe. Pirenne’s thesis is a signpost for this group working within the frame­ work of the ESF project on ‘the transformation of the Roman world’. We defined ourselves as seeking to devise a post-Pirenne paradigm. It is an impossible objective, which most participants considered inappro­ priate. Let us be quite clear: a post-Pirenne paradigm would need to define an entirely new sequence of historical relations to explain the transformation of Roman Europe. Re-interpretation on this scale is be­ yond us. Taking account of the present state of scientific knowledge in this field, we can realistically expect only to analyze new data and new interpretations of those data which cumulatively, in a generation or two, may make it possible to produce a new paradigm. In any case, by then ‘the polyglot answer’ will be anachronistic as European union propels greater trans-European collaboration in our research. Having set our objective, we sought to define it more precisely by selecting time-slices where we might collaborate satisfactorily in inter­ disciplinary research. Three areas were identified where recent research in archaeology and history might be usefully developed: (i) the sixth century, (ii) the decades around AD 700, and (iii) the early ninth century. Recent ar­ chaeological excavations have transformed our understanding of these three time-slices, thereby providing historians concerned with-written and visual evidence to re-examine their sources. The emphasis upon time-slices might seem anachronistic in an age concerned with processes in the past. O f course, there is a danger of overlooking la longue durée, as we examine transformation in such com­ paratively brief periods. Yet each of these time-slices serves as a frame­ work for examining European change; for examining circumstances in northern as well as southern Europe. Examined on this geographical scale, extra-regional processes become the inevitable focus of our nor­ mal regional interests because these, as Pirenne comprehended, offer the means of explaining not only patterns, but also political action and, of course, transformations.

6

RICHARD HODGES

The Sixth Century almost by definition, colonial entanglement and struggle turn upon the dif­ ference between indigenous peoples and foreigners, natives and intruders, but recognition that this axis is fundamental should not obscure or marginalize the crucial fragmentation of knowledge and interests on both sides, the strug­ gles which always take place within both the metropolitan project of colo­ nialism and the indigenous project of appropriating or reacting to colonial intrusion.6

Pirenne forgot the sixth century. Late Roman patterns of production and distribution, in his view, had been set in the fourth or fifth centu­ ries.7 The seventh century, by contrast, witnessed the effective end of the imperial economy, and the effective reduction of imperial society to the courts of Constantinople, Ravenna and Rome.8 Pirenne’s thesis concerned a super-power in transformation: an empire with colonies, unaffected by indigenous struggles. The reality, as we now know, was very different. The transformation of the economy of the Roman em­ pire, in particular, is a complex history. To use the metaphor of Nicho­ las Thomas’ provocative study of the anthropology of exchange in the Pacific between Western colonial societies and indigenous peoples, this was a world of entangled, exchanges (my italics) not one predicated upon the simple “us/them” dichotomy between Westerners and Pacific Is­ landers. Being a historian of the age of Empire (Belgian, British, French, German and Russian), Pirenne, like the written sources of the later empire, favoured the “us/them” transformation of the Roman world system. But then, in the absence of multidisciplinary sources, Pirenne possessed no grasp of the pattern of demand on the eve of this transformation. Objects of entangled exchange, understandably, were beyond the scope of his research. The studies in this book seek to re­ dress this oversight. In this short introduction, using Thomas’ theme as a point of reference, we wish to illustrate the complex parameters of demand across the breadth of Europe in the course of this momen­ tous century. A glance at recent research on late Antiquity confirms Bowerstock’s contention that ‘the international revival of interest in Roman and

6 N. Thomas, Entangled Objects. Exchange, Material Culture, and Colonialism in the Pacific (Harvard, 1991), p. 205. 7 Van Dam, “The Pirenne thesis and fifth-century Gaul”. 8 M. Angold, “The shaping of the medieval Byzantine city”, Byzantinische Forschungen 10 (1985), pp. 3-4.

HENRI PIRENNE AND THE QUESTION OF DEMAND

7

Byzantine history between Constantine and the Arab conquest has transformed late Antiquity from an exotic and neglected field into one of the most vigorous and exciting areas of current research’.9 This re­ vival of interest bridges Europe. It concerns not only those, like Holy Men, who made history, but those who were, so-to-speak, denied it. The pattern of demand is no longer abstract, thanks to a wide range of historical and archaeological studies. The centrifugal demands of the Byzantine court are now well-chart­ ed.10 likewise, the changing character of the later Roman aristocracy has attracted a good deal of attention in recent years.11 Above all, there have been many studies dedicated to the power of the Church in late Antiquity. Peter Brown, in particular, has championed the rise of Christianity and Holy Men in this age,12 showing it (to cite one of his admirers) to be ‘a period of extraordinary interest, which witnessed the extinction of traditional paganism, the perversion of Christianity, the introduction of ideology as a test of loyal citizenship, the spread of in­ tolerance, institutionalized superstition and competitive asceticism’.13 By the later fifth century the Church was embedded not only in met­ ropolitan society but also in the countryside.14 Its powerful role in eco­ nomic strategies, especially in the field of reproduction strategies and inheritance, can no longer be doubted.15 Yet, undeniably, by AD 500 the transformation of European society exceeded any project devised by the Church. The sixth century is of particular interest because the imperial econ­ omy was still functioning despite the palpable collapse of imperial so­ ciety. It was an age when to quote Walter Pohl ‘the gentes could only define themselves versus the overwhelming reality of a polyethnic late 9 G. Bowerstock, “Review o fj. Matthews The Roman Empire of Ammianus," Journal of Roman Studies 80 (1990), p. 244. 10 M. McCormack, Eternal Victory (Cambridge, 1986); see also Angold, “The shaping of the medieval Byzantine city”. 11 F. Marazzi, “Il conflitto fra Leone III Isaurico e il papato fra il 725 e il 733, e il ‘definitivo’ inizio del medioevo a Roma: un ipotesi in discussione”, Papers of the British School at Rome LIX (1991), pp. 231-57. 12 P. Brown, The Cult of Saints (Chicago, 1981); Society and the Holy in Late Antiquity (Ber­ keley, 1982); Authority and the sacred: aspects of the Christianisation of the Roman World (Cam­ bridge, 1996). 13 C. Mango, “Review of G. Bradshaw, The Colour of Power*', Times Literary Supplement December (1989), pp.22-28. 14 C. Pietri, “Chiesa e communità locali nell’ occidente cristiano (TV-VI D.C.): L’es­ empio della Gallia”, Società romana e impero tardoantica. Voi 3. Le merci, gli insediamenti, ed. A Giardina (Rome, 1986), pp. 761-95. 15 J. Goody, The development of thefamily and marriage in Europe (Cambridge, 1983).

8

RICHARD HODGES

Roman state’.16 Inevitably, in terms of demand as well as production and distribution, the reality was ‘entangled’,17 more complicated than the unitary models proposed hitherto. This is most conspicuous in later Roman towns - centres of demand as well as production-distribution which were essentially alien to Germanic society. By AD 500, late-Roman towns in the West had witnessed extraordinary changes, most no­ tably the ruination of their great public monuments while they inher­ ited the roles of industrial vici as centres of production and distribu­ tion.18 As a result, in administrative terms, as Andrew Poulter put it ‘the poids of the sixth century bear little or no resemblance to the cities of the fourth, let alone the second century. They were essentially cen­ tres of imperial and ecclesiastical administration, and the civilian pop­ ulation appears to have been dependent upon, but largely excluded from, these Byzantine citadels. By the sixth century, it is doubtful if the urban population remembered or even understood the concept of ur­ ban self-government which had proved so attractive to the cities of the early Empire’.19 The emphasis was on fortification, great ecclesiastical monuments and, through investment in artisan activities, sustaining the economic umbilical cord with Constantinople. More precisely, the late Roman town-house with its apogee in the sixth century, embodied a hierarchy of access.20 Unlike the houses of the early Roman period, the late Roman aristocrat carefully defined the architectural context in which his public encounters occured. Great effort was invested in displaying a concentration of wealth in order to underscore the aristocratic nature of patronage as well as, in Simon Ellis’ view, to demonstrate overtly the ideology of the heroic host.21 The new taste and demands of late Roman society perplexed the ex-

16 W. Pohl, “Conceptions of ethnicity in Early Medieval studies”, Archaeologia Poiana 29 (1991), p. 40. 17 Thomas, Entangled Objects. 18 L. Saguì, “Produzione vetrarie a Roma tra tardoandco e alto medioevo”, La Storia economica di Roma nell’ alto Medioevo alta luce dei recenti scavi archeologici, eds. L. Paroli and P. Delogu (Rome, 1993), pp. 113-36. 19 A. Poulter, “The use and abuse of urbanism in the Danubian provinces during the Late Roman Empire”, The City in Late Antiquity, ed. J. Rich (London, 1992), p. 132; see also B. Ward-Perkins, “Urban continuity?”, Towns in Transition, ed. N. Christie and S.T. Loseby (Aldershot, 1996), pp 4-17. 20 S. EUis, “ The End of the Roman house”, American Journal ofArchaeology 92 (1988), 565-76; S. P. Ellis, “Power, architecture and decor: how the Late Roman aristocrat ap­ peared to his guests”, Roman Art in the Private Sphere, ed. E. K. Gazda (Ann Arbor, 1991), pp. 117-56. 21 Ellis, “ Power, architecture and decor”, pp. 123-30.

HENRI PIRENNE AND THE QUESTION OF DEMAND

9

cavators of San Giovanni di Ruoti in Basilicata, southern Italy.22 The plan of the Late Roman villa was altogether different in its morphology to its earlier imperial forebear. This quotation from the excavators’ account explains these differences without any reference to the ‘entan­ gled’ socio-economic circumstances of later antiquity: It is also possible that the plan of the villa points to new occupants with some un-Rom an social customs. The clearest sign is the way they disposed of their garbage. T he inhabitants of the previous villas followed the normal Roman practice of removing refuse from their buildings; but the occupants of the late villa at San Giovanni dumped their kitchen waste in the corridors and empty rooms, and immediately outside the entrances to the site. M ost of the midden piles date to the last part of the fifth century and to the beginning of the sixth, and belong to the second phase of this villa, but some go back to the begin­ ning of its first phase, ca. AD 400. Even the dining customs may have changed, for if the long narrow room with a mosaic floor at the northeast end of the site is a dining-room, as seems likely, then the inhabitants must have eaten there seated beside a long table, as was the practice of Germanic barbarians, rather than reclining around a low table in traditional Roman fashion, for there is no room there for a stibadium.23

Changing attitudes to diet as well as to refuse dispositi in late Antiquity, tempting though it is to attribute these to ethnicity, are more reason­ ably the result of the changing nature of society itself.24 Näsman, in his contribution to this book, cogently illustrates the close relations be­ tween Scandinavia and central Europe during this century. Concurrently, in southern Scandinavia, the relationship between the élite and the collective, as manifested in demand and taste, was also changing. Reviewing numerous excavations of timber dwellings from the later Iron Age, Frands Herschend showed that the long hall is a creation of the fourth to fifth centuries, and the norm by the sixth cen­ tury as far north as central Sweden.25 Herschend contends that ‘the hall was already the room of leadership, in an economic as well as a military sense, in the fifth century....The interesting thing is that, cen­ turies before we can talk of feudalism, the hall constitutes the room as a social space for the individual who in that room is the head of a nu­ clear family —a positive notion common to rich and poor - and not 22 A.M. Small and Robert J. Buck, The Excavations of San Giovanni di Ruoti (Toronto, 1994). 23 Small and Buck, The Excavations of San Giovanni di Ruoti, pp. 4-5. 24 R. Hodges, “The Late Roman setdement at San Vincenzo al Volturno”, San Vin­ cenzo al Volturno 2, ed. R. Hodges (London, 1995), pp. 127-28. 25 F. Herschend, “ The Origin of the Hall in Southern Scandinavia”, Tor 25 (1994), pp. 175-99.

10

RICHARD HODGES

just one of a team that runs an estate.... Some time during the late Ro­ man Iron Age, it became possible within the nuclear family to breed an individuality which acted for the collective and to make it a publicly accepted, social norm’.26 This individuality took other forms. The powerful, focused role of ideology in late Antiquity, evident in the social ascendency of the Holy Man, reached its apogee in the politically-inspired monuments of San­ ta Sophia in Constantinople and San Vitale in Ravenna. It is no less evident at Gamia Uppsala in central Sweden where a putative timber temple has been found adjacent to the Vendei burial mounds. What­ ever purpose this structure served, there is no doubting that it is a man­ ifestation of its age. Equally, with time, a similar structure will be found at Gudme on Funen - where a staggering quantity of goldwork asso­ ciated with ritual practices has been found in the environs of a major hall.27 In her analyses of the patterns of gold hoards from Gudme, Charlotte Fabech, concludes that the era of collective sacrifices of boo­ ty in bogs and wetlands ended in the fifth century. ‘From about 400 religious manifestations both on dry land and wedand (gold bracteates) are related to the settlement with manorial dwelling (jic)’.28 The tension between the entangled ethnic configurations of Europe, and the dialectic of the individual over the collective, fuelled demand. Craftsmen’s quarters are as much the stuff of the archaeology of later Roman ports and towns as town houses and churches. A glassmaker occupied the grandiose latrines of the theatre of the Cripta Balbi.29 Af­ ter identifying the wooden market stalls erected within the forum at Cherchel, Potter has tracked down the evidence for similar installations in many other North African cities of the period.30 Streets of shops serving the same purposes were erected in Sardis, Jerash and Palmyra (to name only a few examples).31 Perhaps the most evocative discovery,

26 Harschend, “The Origin of the Hall in Southern Scandinavia”, p. 195. 27 On temples see Ian Wood, “Pagan Religions and Superstitions east of the Rhine from the fifth to the ninth century”, After Empire, ed. G. Ausenda (Woodbridge, 1995), pp. 253-68; the archaeology of Gudme is cogendy described in The Archaeology of Gudme and Lundeborg, eds. P. O. Nielsen, K. Randsborg and H. Thrane (Copenhagen, 1994). 28 C. Fabech, “Reading society from the Cultural Landscape. South Scandinavia be­ tween Sacral and Political Power”, The Archaeology of Gudme ami Lundeborg, p. 175. 29 Saguì, “Produzione vetrarie a Roma tra tardoantico e alto medioevo”. 30 T.W. Potter, Towns in Late Antiquity, loi Caesarea and its context (Oxford, 1995); WardPerkins, “Urban continuity?”, pp. 11-15; A. Walmsley, “Byzantine Palestine and Arabia: Urban prosperity in Late Antiquity”, Towns in Transition, eds. N. Christie and S.T. Loseby (Aldershot, 1996), pp. 126-58. 31 J.S. Crawford, The Byzantine shops at Sardis (Harvard, 1990), pp. 107ff.

HENRI PIRENNE AND THE QUESTION OF DEMAND

11

illustrative of the complex economic transformation, occured at Sardis where shop El 3 was selling glass window panes on the eve of a devas­ tating fire in c. 616.32 Over 350 window panes and 350 goblets were found in the ruins, indicating the capacity to produce ample capital in nearby workshops as well as the shop-owners’ cash-flow capacity to purchase such a volume of merchandise. Finally, the illustration, one of many from the destruction of this extraordinary vignette of shop life in early seventh-century Sardis, reveals the nature of demand within the community, and the shopkeepers’ ability to supply this demand. Citizens of Sardis had the means to purchase house-fittings and table­ ware. Elsewhere in the Mediterranean the supply of commodities was by no means as buoyant, as Loseby’s chapter about Marseilles in this vol­ ume vividly illustrates. Production and distribution were inevitably af­ fected by demand. Demand fluctuated as the aspirations of the farflung Byzantine élite were snuffed out by increasing state expenditure on the military, and the impossibly high taxation caused by this. By AD 600 the geo-political implications for much of the old Roman Empire were verging on the catastrophic. Traditional control was ceded in the mountains of Italy, most of the Balkans and the inland areas of North Africa. Ancient towns in these areas were transformed in the course of a generation. Social complexity was reduced to a relic élite —local rep­ resentatives of state power33 —and, to judge from the archaeology, ur­ ban populations, albeit dwindling in numbers, sustaining in vain the shadow of Byzantine demand. The relic élite are scarsely a mystery to us, unlike the workforce occupying the shanty towns made within the ruined townscapes of the Mediterranean world.34 Invariably archaeol­ ogists have disregarded the latter as ‘squatters’, Slavs, Arabs and every other ethnic barbarian. But, given that in almost every case these poor quality structures - harbingers of Dark Age construction —are dated by Byzantine coins and Byzantine amphorae and tableware, it is not

32 Crawford, The Byzantine shops at Sardis, pp. 78-81; J. Henderson and M. Mundell Mango, “Glass at Medieval Constantinople. Preliminary scientific evidence” Constantino­ ple and its Hinterland, ed. C. Mango and G. Dagon (Aldershot, 1995), p. 338. 33 J. Haldon, “Military administration and bureaucracy: state demands and private interests,” Byzantinische Forschungen 19 (1993), p. 45. 34 See J. Haldon, Byzantium in the Seventh Century (Cambridge, 1990); Ellis, “The End of the Roman House”; Ward-Perkins, “Urban continuity?”, pp. 11-15; M.O.H. Carver, “Transitions to Islam: Urban roles in the East and South Mediterranean, fifth to tenth centuries AD”, Towns in Transition, eds. N. Christie and S.T. Loseby (Aldershot, 1996), pp. 184-212, esp. 194-95.

12

RICHARD HODGES

farfetched to propose that these were the homes of the artisanal class who hung on in hope of a revival of Byzantine fortunes. The migrants, it appears, occupied new niches within the relic classical landscape, and had recourse to minimal commodities. Such was the nature of in­ digenous, post-colonial demand in the entangled circumstances at the end of the sixth century in the Mediterranean basin. A vignette of comparable, misplaced aspirations has been highlight­ ed by recent archaeological investigations in Dark Age Britain. At the beginning of the sixth century, the tribes of post-Roman western Brit­ ain entertained a short-lived contact with the Mediterranean. The ob­ jects in demand were North African and East Mediterranean (B-ware) amphorae, North African tablewares and some (?) East Mediterranean glassware. The small assemblages of amphorae and tablewares have re­ ceived extraordinary detailed attention because, being found at Tintagel and Cadbury-Camelot, they are associated with the legend of King Arthur. First, the majority of the imported wares derive from the East Mediterranean; only a small fraction emanated from North Africa.35 This is consistent with the relative importance of East Mediterranean shipping vis-a-vis North African-based maritime activity by AD 500.36 Second, as all archaeologists have acknowledged, the quantities of sherds involved are tiny (e.g. 1495 B-ware amphorae sherds from Tintagel; 131 sherds from Cadbury Castle). In total the finds brought to light so far amount to no more than a few cargoes.37 Third, just as these sherds belong to a western British culture zone, distinct in its ma­ terial culture from the incipient ‘Anglo-Saxon culture’ of eastern En­ gland, so it is interesting to note the concurrent emergence of the hall house38 at a time when Anglo-Saxon settlements are noted for their absence of structural diversity manifested in dwellings for nuclear fam­ ily.39 In short, does the sixth-century world of King Arthur owe its last­ ing status to an incipient tension between the individual and the col35 L. Alcock, Cadbury Castle, Somerset. The early medieval archaeology (Cardiff, 1995), pp. 141-3; M. Fulford, “Byzantium and Britain: a Mediterranean perspective on post-Roman imports in western Britain and Ireland”, Medieval Archaeology 33 (1989), pp. 1-6; D. Williams and César Carreras, “North African Amphorae in Roman Britain: a re-apprais­ al,” Britannia XXVI (1995), pp. 240-1. 36 C. Panella, “Merci e scambi nel Mediterraneo tardoantico”, La Storia di Roma III/ ii, eds. A Carandini, L. C. Ruggini and A. Giardina (Turin, 1993), pp. 613-97. 37 Fulford, “Byzantium and Britain”. 38 Alcock, Cadbury Castle, Somerset, pp. 132-9. 39 R. Hodges, The Anglo-Saxon Achievement, (London, 1989), pp. 65ff; H. Hamerow, “Settelement mobility and the ‘Middle Saxon shift’: rural settlements and setdement pat­ terns in Anglo Saxon England”, Anglo-Saxon England 20 (1991), pp. 1-17.

HENRI PIRENNE AND THE QUESTION OF DEMAND

13

lective? Could it be a further illustration of the entangled circumstances in which demand for commodities played a part in the social transfor­ mation of the Roman world? These circumstances were short-lived. By the end of the century many of the ‘Anglo-Saxon’ tribes of southern and eastern England en­ tered into exchange relations with the Franks and other traders plying the Merovingian North Sea (see Näsman in this volume). The ‘AngloSaxon’ cemeteries after c.550 were affluent with imported Byzantine, Frankish and Scandinavian gifts to the gods. Such is their number that Vera Evison argued for a Frankish invasion of England at this time.40 This invasion hypothesis has long since been rejected, but the presence of these objects begs an explanation. Simply put, Anglo-Saxon society was almost certainly responding to the strains of a demand for primi­ tive valuables, as individuals sought to define themselves as late Roman aristocrats had, and as South Scandinavians were doing. From our standpoint the evidence remains largely speculative until the end of the sixth century when St. Augustine arrived in Kent, introducing new, Mediterranean, temples which were placed at the disposal of the par­ venu Anglo-Saxon élite for purposes in this life and for their privileged wellbeing, in common with Romano-Byzantine burial traditions, in death.41 At the same time the élite responded by constructing hall houses, such as had existed for more than a century in South Scandi­ navia, and for some time in western Britain. Concurrently, the élite also designated type A emporia, such as Ipswich on the river Gipping, where they could administer periodic markets.42 The sixth century marks the moment of utmost confusion amongst the polyethnic cultures of Europe before the emergence of a new geo­ political order in which the Franks, Lombards and Visigoths play a pre-eminent part. Patterns of demand bear witness to the confusion. Mediterranean society embarked upon a transformation while tradi­ tional consumption and demand patterns outlived the first, dramatic episodes of transformation. These entangled circumstances touched northern and western Europe in improbable ways. South Scandinavia, as the investigations at Gudme have shown, was drawn into the com-

40 V.I. Evison, The Fifth-Century Invasions South of the Thames (London, 1965). 41 R. Morris, Churches in the landscape, (London, 1989); D. Deliyannis, “Church burial in Anglo-Saxon England: the prerogative of kings”, Frühmittelalterliche Studien 29 (1995), pp. 96-119. 42 K. Wade, “Ipswich”, The rebirth of towns in the West, AD 700-1050, eds. R. Hodges and B. Hobley (London, 1988), pp. 93-100; Hodges, The Anglo-Saxon Achievement.

14

RICHARD HODGES

plex orbit of Byzantium just as King Arthur’s kingdom was. But it would be simplistic to attribute the ascendency of the individual over the collective to such direct interactions. Rather, these touched the complex peer-polity interactions between the myriad tribes of the re­ gion.43 Indeed, if there is any one conclusion to be drawn from these odd pieces taken from the geo-political jigsaw of Europe AD 500-600, it is that the transformation of the Roman world triggered an aston­ ishing diversity of responses. No doubt, instinctively aware of this as a result of his internment in a German prisoner-of-war camp in the tur­ bulent multinational circumstances of 1916-18, Pirenne chose to over­ look the sixth century in formulating his famous thesis. **** This volume attempts to examine these issues from a wide variety of angles: from Scandinavian and Mediterranean perspectives; from his­ torical and archaeological perspectives. Some of the essays deliberately have a wide scope; some are specific. The shadow of Pirenne affects some papers, such as Loseby’s, and others not at all. The strength of the ESF project is its pluralism; despite the difficulties of European in­ teraction, it provokes debate, even confrontation and communication of experiences. In preparing this volume I should like to warmly thank my research assistant, Will Bowden. My thanks also to Christian Mulberg, Max Spareboom and Vuokko Lepistö-Kirsila, who have been active secre­ taries and advocates of this programme at the ESF. My thanks, too, to all my colleagues, especially Javier Arce, Klavs Randsborg, Chris Wickham, and Ian Wood who have proved stimulating discussants, stern critics and firm friends in the course of the seminars at which we examined sixth-century Europe.

43 C. Renfrew, “Introduction, peer-polity interaction and socio-political change”, Peer-Polity Interaction and Socio-Political Change, eds. Colin Renfrew and John F. Cherry (Cambridge, 1986), pp. 1-18.

READING PIRENNE AGAIN

Paolo Delogu

In the thesis for which he is famous, Henri Pirenne interpreted the de­ cline of the ancient world and the beginning of the Middle Ages as the direct consequence of the Islamic expansion into the Mediterranean basin. The thesis was the result of two distinct historical considerations. Firstly, Pirenne maintained that the cultural and economic system of late Antiquity, based on the free movement of people, goods and ideas between the western and eastern provinces of the Empire, survived in outline until the seventh and eighth centuries. Secondly, he attributed the eventual crisis to an occurrence, such as the Muslim conquest, whose nature was ideological and military, rather than economic. The thesis had its most complete expression in Pirenne’s Mahomet et Charle­ magne, which appeared posthumously in 1937. Immediately after its publication, the book met with criticism of both its principle tenets. Serious questions were asked about the valid­ ity of the evidence used by Pirenne to demonstrate the cessation of traf­ fic across the Mediterranean following the Islamic conquests. As a con­ sequence, the crisis in the circulation of goods, men and ideas was de­ nied or otherwise contested, and the timing of the cessation of Medi­ terranean trade variously pushed back or brought forward. Above all, the very attribution of the disappearance of Christian shipping from the Mediterranean to a deliberate Islamic blockade was strongly con­ tested. In 1947, Maurice Lombard challenged Pirenne’s thesis, suggest­ ing rather that the arrival of Islam in the Mediterranean created con­ ditions favourable to a revival of commercial activity in the West, after a long period in which it had been stagnating.1 The various critiques to which Pirenne’s thesis has been subjected have revealed its many and intrinsic weaknesses. Yet despite questions concerning the sources upon which it depends and the treatment of 1 On the Pirenne thesis, see: A.F. Havinghurst, The Pirenne Thesis. Analysis, Criticism and Revision (Boston, 1958), which reviews most of the important earlier critical literature; also: B. Lyon, The Origins of the Middle Ages: Pirenne’s challenge to Gibbon (New York, 1972); P. Brown, “Mohammed and Charlemagne”, Daedalus (1974), pp. 25-33; R.S. Lopez,

16

PAOLO DELOGU

the evidence, the fact remains that Mahomet et Charlemagne represents a truly formidable attempt to formulate a unified reconstruction, in structural terms, of the transition from antiquity to the Middle Ages in the West. In its effort at unifying the diverse aspects of historical change, it has probably not yet been superseded. The alternative thesis proposed by Lombard was soon shown to be insufficient as an expla­ nation both of the crisis of the ancient economy and the emergence of the medieval one. At present, with a few exceptions, the research seems to argue against a unified explanation of the multiple aspects of the transition. Meanwhile, some of Pirenne’s ideas have found support in recently discovered evidence. For example, recent work has confirmed the cru­ cial position of the seventh century in the transformation of late An­ tiquity, in the West as in the East. Furthermore, the ever-expanding corpus of archaeological evidence available for the period reveals the Islamic conquest of North Africa to have impeded movement and trade in the Mediterranean, though admittedly not to the extent imag­ ined by Pirenne and probably only within the western Mediterranean.2 “Quarantanni dopo Pirenne”, La navigazione mediterranea nellyalto Medioevo (Settimane del C.l.SAM ., 25) (Spoleto, 1978), pp. 15-31; R. Hodges, D.B. Whitehouse, Mohammed, Char­ lemagne and the Origins of Europe. Archaeology and the Pirenne Thesis (London, 1983) (see also the revised French edition, Mahomet, Charlemagne et les origines de l’Europe (Paris, 1996)); D. Claude, “Der Handel in wesdichen Mittelmeer waehrend des Frühmittelalters”, Untersu­ chungen zu Handel und Verkehr der Vor-und Frühgeschichtlichen ^eit im Mittel-und NordEuropa 2 (Goettingen, 1985), p.9 ff.; G. Despy, A. Verhulst (eds.), Lafortune historiographique des thèses d’Henri Pirenne (Brussels, 1986); A. Verhulst, “Marchés, marchands et commerce au haut Moyen Age dans l’historiographie récente”, Mercati e mercanti nell’alto medioevo: i’ea euroasi­ atica e l’area mediterranea (Settimane del C.l.SAM ., 40) (Spoleto, 1993), pp. 23-43; G. Petralia, “A proposito dell’immortalità di ‘Maometto e Carlomagno’ (o di Costantino)”, Storica 1 (1995), pp. 37-87. Lombard’s thesis is developed in a number of works, above all: “Les bases monétaires d’une suprématie économique. L’or musulman di VIIe au XIe siècle”, Annales. Economies - Sociétés - Civilisations 2 (1947), pp. 143-160; and, “Mahomet et Char­ lemagne. Le problème économique”, Annales. Economies - Sociétés - Civilisations 3 (1948), 2 For example: R. Hodges, Dark Age Economies. The Origins of Towns and Trade. ÁD 6001000 (London, 1982); R. Hodges, “Dream Cities: Emporia and the End of the Dark Ag­ es”, Towns in Transition. Urban Evolution in Late Antiquity and the Early Middle Ages, eds. N. Christie and S. Loseby (Aldershot, 1996), pp. 289-305; A. Verhulst, “The Origins of Towns in the Low Countries and the Pirenne Thesis”, Past and Present 122 (1989), pp. 335; C. Panella, “Gli scambi nel Mediterraneo occidentale dal IV al VII secolo”, Hommes et richesses dans l’Empire byzantin. I: IV-VII siècle (Paris, 1989); C. Panella, “Merci e scambi nel Mediterraneo tardoantico”, Storia di Roma III. L’età tardoantica. IL I luoghi e le culture (Turin, 1993), pp. 613-697; P. Delogu, “The Rebirth of Rome in the 8th and 9th Cen­ tury”, The Rebirth of Towns in the West. AD 700-1050, eds. B. Hobley and R. Hodges (CBA Research Report 68) (London, 1988), pp. 32-42; P. Delogu, “La fine del mondo antico e l’inizio del medioevo: nuovi dati per un vecchio problema”, La storia dell’alto medioevo italiano (VI-X secolo) alla luce dell’archeologia (Convegno Intemazionale, Siena 2-6 dicembre 1992) eds. R. Francovich and G. Noyé (Florence, 1994), pp. 7-29 (with Chris Wickham’s “Con­ siderazioni conclusive”, pp. 741-759).

READING PIRENNE AGAIN

17

This is by no means sufficient to reinstate the Pirenne thesis as an ten­ able explanation of the process of transition. Yet these considerations do reveal the actuality, at least in part, of some of Pirenne’s hypotheses, while prompting us to reconsider the terms in which he elaborated his ideas. Possibly, the way of thinking followed by Pirenne to shape his thesis can be recognised as responsible for its fundamental weakness as a whole, perhaps more so than the nature and quality of the sources upon which he drew, or the evaluation of specific points. Investigation of the thought of past historians may aim to reconstruct something of the cultural climate in which they worked; to locate the historian in history, as it were. Alternatively, it may expose the factors which informed their historical views and reconstructions. Historio­ graphical analysis of this kind reveals what links us inextricably to the fathers of our discipline, while emphasising how different our present strategies of historical investigation are. In this way, what follows is an attempt to elucidate the key stages of the formulation and elaboration of the Pirenne thesis and to identify the essential theoretical concepts underlying it.

I In a well-known essay published in 1895 under the tide L’origine des con­ stitutions urbaines aux Moyen Age, Pirenne wrote: On sait que les villes romaines ont survécu à l’empire romain en Occident. Si dans l’extrême nord, sur les frontières germaniques, quelques-unes d’entre elles ont été détruites defond en comble, on s’aperçoit tout de suite, cependant, qu’après les in­ vasions, la plupart des cités restent debout. Il suffit de lire les textes du VF siècle pour voir que, dans ce temps là, la Gaule est encore un pays de villes. En dépit du desordre grandissant et de l’anarchie ménaçante, toute vie municipal n’est pas éteinte. On continue à insinuer les actes aux ‘gesta municipalia’. Ca et là, il est encorefait mention du ‘defensor civitatis’ ou des ‘curiales’. D’ailleurs, il subsiste quelque ac­ tivité commerciale et industrielle. Les droits de douane n’ont pas cessé defournir à l’état des revenus assez abondants. Grégoire de Tours vante la richesse des Verdunois; il áte la ‘negutiantium domus’ de Paris et parlefréquemment de marchands juifs et syriens. Il est manifeste que la Gaule se trouve encore, quand il écrit, en rélations suivies avec l’orient et que les ports de la Mediterranée n’ont pas encore perdu toute importance.

18

PAOLO DELOGU

Toutefois, cet état de chose ne pouvait durer. La vie économique s’éteint, en Gaule, comme s’éteint la vie littéraire, faute d’aliments. On voit l’or se raréfier peu à peu, puis disparaître complètement. Le système des échanges en nature tend à se substituer de plus en plus à celui de la circulation monétaire. Quand la Mediterranée est de­ venue un lac musulman, c’en estfait, et l’on entre alors décidément dans l’age ag­ ricole du Moyen Age. A l’époque carolingienne, l’argent atteint à lafois le maximum de sa valeur et le minimum de son emploi. La terre est maintenant la seule richesse connue, et dès lors se propagent victorieusement le système seigneurial et laféodalitéA In this passage, all the basic elements of the later thesis are already present: the survival of Roman towns and civic life after the Germanic invasions; and the continuation of strong ties (and especially trade links) between East and West, broken only by the Islamic expansion which precipitated the collapse of the ancient economy and the onset of the ‘agricultural epoch’ of the Middle Ages. When he wrote this text, Pirenne was thirty-five, and until then had published nothing but his Histoire de la constitution de la ville de Dînant (1889). Nonetheless, it seemed that he had already conceived the thesis that was to become the theme of his ultimate and most widely-known book, Mahomet et Charlemagne. Yet while the key elements of the thesis are present, Pirenne was still to investigate their inter-relation or to reconstruct the chain of agents and effects of historical change. At this stage, Pirenne’s interest was not directed at the transition from late Antiquity to the Middle Ages. Instead, he was concerned with the study of medieval cities from the eleventh century; and in par­ ticular in defining their social, economic and institutional character, with an understanding of the processes by which they emerged from the preceding “agricultural epoch”, during which it appeared that no comparable setdements existed. In his account of 1895, Pirenne took up a position against the theses advanced by various German scholars who identified the genetic factor of the medieval city with the juridical privilege of a given site (for example, a trading place or a fair), or a community (frequendy described as Markgemeinde, or village communi­ ty). Such privileges were credited with having caused migration to such a centre, thus encouraging the formation of a nucleated setdement, which was legally distinguished from rural setdements. In his opposi-3

3 H. Pirenne, “L’origine des constitutions urbaines au Moyen Age”, Revue Historique 57 (1895), pp. 57-98; 293-327, at p. 57.

READING PIRENNE AGAIN

19

tion, Pirenne maintained that the statutory definition of such settle­ ments came later, and that it was the consequence, rather than the cause, of a particular social change; one which was provoked by a new set of economic conditions. As Pirenne saw it, this transformation was marked by the appearance of a. class of merchants whose settlement patterns were determined by the movement of commercial traffic. Thus, permanent trading places were established, for example, near bridges, estuaries and cross-roads. This was determined not so much by local demand or supply of agricultural produce, but by the move­ ment of long-distance trade and by the structural necessities of its traf­ ficking. This kind of trade was an autonomous activity conducted by professional merchants; it therefore presented the requisites of an ur­ ban economy. According to Pirenne, medieval cities could not be de­ scribed simply as markets, as they were by other scholars, but as trad­ ing places of a particular kind, supported by long-distance trade and managed by a specialised class of professionals. Local trade could be conducted in such centres, of course, but it was by no means the rootcause of their formation or development. In 1895 Pirenne considered that professional merchants existed in the West as early as the ninth century; in the eleventh century a gen­ eral improvement in conditions led to an increase in the numbers of traders and the creation of an increasingly extensive network of settle­ ments, which became permanent centres of exchange and distribution, although the merchants continued to travel long distances in order to get the wares and supply the markets. Traders quickly formed a new social class, one which claimed a new legal status, eliciting new laws and rights. This class soon established itself as the dominant group within the settlements in which it conduct­ ed business and reflected its special legal condition on the settlement itself. Thus the mercantile class gave birth to the medieval city. In 1895 Pirenne did not attempt to explain the reasons why trade and the merchant class expanded during the eleventh century. He seems content to offer general references to an increased level of secu­ rity and stability under the Ottonian emperors and other sovereigns. He seems still less concerned to identify the laws of historical change, although on this matter at least he formulates an interesting concept which he describes as follows: Il est dans la vie des peuples des époques ou la marche de la civilisation semble se hater sous l’action d’idées et deformes nouvelles et où... incapables d’attentre que se

20

PAOLO DELO GU

soient transformées les institutions du passé, les hommes en créent d’autres qui les remplacent.4 Already in this passage we find an expression of the concept of discon­ tinuity in history that was to characterise Pirenne’s historical thought.

II Twenty-two years on, Pirenne was to return to the same themes in one of his most significant (though comparatively neglected) works, the His­ toire de l’Europe, which he began in February 1917, while interned in Germany.5 In this book, Pirenne gave substance to his ideas on the Middle Ages and the formation of Europe in that period. Political events and the institutional apparatus of the states were described in strict relation to the economic foundations and the development of European society from late Antiquity to the Renaissance. Pirenne also dealt with the mechanisms of transformation in history. The various phases of the development of society and economics were strictly connected in a chain of cause and effect. The transition from the ancient world to medieval society was described as follows: the Germanic migrations did not result in the break-up of Roman society - rather, the barbaric invaders preserved Roman institutions in order to take advantage of them; the institutional apparatus of the late-Roman Empire (and above all the organisation of the economy) was adopted and exploited by the Germanic kings for their own interests; and the organisation of landholding continued to follow established patterns, Germanic settlers simply installing themselves alongside Ro­ man proprietors. Towns survived as the centres of cultural, ecclesiasti­ cal and administrative life, as well as the terminals of long-distance ex­ change systems which ranged across the Mediterranean. Despite the Germanic invasions, the West continued to participate in the common

4 H. Pirenne, “L’origine des constitutions urbaines aux Moyen Age”, Revue Historique 53 (1893), pp. 53-83, at p. 62. 5 On Pirenne’s life and work, see: G. Gerardy, Henri Pirenne, sa vie et son oeuvre (Brus­ sels, 1962); J. Dhondt, “Henri Pirenne, historien des insdtutions urbaines”, Annali delta Fondazione Italiana per la Storia Amministrativa 3 (1996), pp. 81-129; B. Lyon, Henri Pirenne: a Biographical and Intellectual Study (Ghent, 1974); C. Violante, “Henri Pirenne e la grande guerra”, La Cultura 25 (1987), pp. 308-342.

READING PIRENNE AGAIN

21

civilisation which the Roman Empire had diffused throughout its prov­ inces. Only with the Islamic invasions did this unity begin to dissolve, with the occupation of the opposite coasts of the Mediterranean by two very different and hostile cultures. The West was separated from the East, from which until that time it had derived not only commodities, but more general cultural impulses. Trade became impossible, and this determines the extinction of markets based in towns and, consequently, the dissolution of the economic basis of urban life. Towns did not phys­ ically disappear, but changed, becoming settlements of a different kind which, with an economy founded on agriculture, were almost indistin­ guishable from rural centres. Deprived of one its of essential compo­ nents (long-distance trade and the social group which once served it), the West was forced to subsist, both economically and culturally, from its own resources. The fundamental premise of this reconstruction is the concept of a strict relationship between town, market and long distance trade, such as Pirenne had formulated in his studies of the Flemish medieval towns. In his opinion, the three elements were strictly interdependent, and consequently none could exist without the others, at least in evolution­ ary form. Applying this model to the crisis of the late-Roman world, Pirenne implicitly adopted his position in the debate about the nature of the ancient economy which was conducted between the end of the nineteenth century and the beginning of the twentieth century. His ideas opposed those of Karl Buecher, the brilliant and acute exponent of the German school of economic history, who interpreted the basis of the ancient economy in terms of Hausmrtschaft, a functional model according to which production and consumption occur within domes­ tic structures. In such an interpretation exchange has no essential or autonomous role. Adopting a contrary position, critics such as Eduard Meyer argued for the fully commercialised nature of the ancient econ­ omy.6 In fact, in the History of Europe, Pirenne made no reference to the debate. He took it for granted that the ancient economy consisted of merchants and markets, and that there was an interchange of commer­ cial traffic between the eastern and western provinces of the Roman world. Probably his position relied more on the numerous references

6 On the debate between Meyer and Buecher, see: M. Mazza, “Meyer vs. Buecher: il dibattito sull’economia antica nella storiografia tedesca tra Otto e Novecento”, Società e Storia 29 (1985), pp. 507-546.

22

PAOLO DELOGU

to the activity of merchants in western sources for the period from late Antiquity to the Merovingians - sources which had previously been used to make a similar case - than on any original analysis of the ev­ idence.7 On the other hand, Pirenne described the socio-economic structure of the Roman towns of the West with great discretion. He played down their potentialities as centres for an autonomous process of production and distribution, emphasising instead the role played by the state in supporting the conditions of urban society, even with regard to markets and trade. This is an important innovation, one which introduced a structural diversity between the towns of the later Roman Empire and the commercial towns of the Middle Ages. While Pirenne’s ideas on the nature of the ancient economy itself are yet to be fully developed at this stage, the criteria which he used in the History of Europe to assess the transition from ancient to medieval economics prefigure those in Mahomet et Charlemagne, though as yet they do not precisely coincide. The transition is described as the consequence of two concomitant processes, one of evolution, the other of disruption. For Pirenne, the Roman civilisation, and with it the organisation of the economy, soci­ ety and administration of late Antiquity, was not interrupted by the Germanic invasions. Rather, they survived them for the simple reason that the invaders did not possess a different, equally effective, system of social relations and institutions, with which to replace them. How­ ever, the structures of the late Roman Empire were subject to a pro­ gressive degeneration, or ‘barbarization’, resulting from the invaders’ inability to effectively maintain Roman systems of state administration and public authority. Thus the power of the Merovingian kings was undermined by mismanagement of state government and resources. At the same time, the powerful landed aristocracy, went on imposing control over rural society and encroaching on the functions of the state. With time it became the only force capable of providing for essential social needs like security and stability. Consequently, Merovingian so­ ciety was progressively ranged under the private power of the local ar­ istocracy in the characteristic form of the landed seigneurie. 7 Before Dopsch, evidence and theories employed by Pirenne in “Origine des con­ stituciones urbaines” were cited by: P. Scheffer-Boichorst, “Zur Geschichte der Syrer im Abendland”, Mitteilungen des Institutsfiir österreichischen Geschichtsforschung 6 (1885), pp. 521550; later, the theme was taken up by: L. Brehiér, “Les colonies d’orientaux en Occident au commencement du Moyen Age”, Byzantinische Zeitschrift 12 (1903), pp. 1-39.

READING PIRENNE AGAIN

23

Such a representation, reconciling the perceived continuity of condi­ tions and institutions with their internal and protracted deterioration, was not a new one. In large part, Pirenne must have derived it from the work of Fustel de Coulanges, although eventually he reached very different conclusions. What is original here is the denial of any tangible contribution on the part of Germanic culture to the transformation of the late Antique world, apart from its barbarous incapability to sustain complex institutional systems. Throughout the History of Europe. Pirenne’s treatment of Germanic culture is marked by a polemical at­ titude which in part resulted from the situation in which he was writ­ ing, confined in Germany, and from his personal aversion to a people and a culture which he perceived as responsible for the war. Later he was to temper his disdain; in Mahomet et Charlemagne he recognised an original and innovative Germanic civilisation, one capable of imposing order on society, at least within northern Europe.8 The degeneration which he attributed to the barbarity of the invad­ ers also involved the towns of the late-Roman West. Under the Roman Empire, the towns had been the residence of landed proprietors, state officials and merchants: the life of the towns and their economies were sustained by the machinery of the state. The economy of the late-Ro­ man town would have been limited to the production and consump­ tion of local resources in a restricted area, (the sort of economy that theorists like Buecher termed Stadtwirtschaft), had not the Roman Em­ pire brought about the extensive movement of people, commodities and ideas, incorporating the life and economy of each town within a vast network. When the Merovingians revealed their inability to administer these support systems, the towns of Roman Gaul began to decline; the network which sustained city-life and urban culture progressively re­ duced itself to trade links with the distant Mediterranean countries, a trade largely dependent on the enterprises of foreign merchants - the Syrians - who were still free to sail across the Mediterranean, whilst Frankish society shifted to an agricultural economy. As Pirenne wrote: Au milieu d’une société qui glissait vers le régime de la propriété seigneuriale, les villes s’étaient maintenues vivantes par le commerce et avec elles une bourgeoisie li­ bre.9 8 For example, in: H. Pirenne, Mahomet et Charlemagne (Paris-Brussels, 1937), p. 123, and in further sources cited below. 9 H. Pirenne, Histoire de l’Europe (Brussels, 1936), p. 39.

24

PAOLO DELO GU

When the Islamic invasions impeded the movement of eastern traders, the urban middle classes disappeared in the West, the towns lost their character of economic and cultural centres, and the landed aristocracy emerged as the only remaining force able to organise Merovingian so­ ciety. Thus, la conquête de la Mediterranée par les Musulmans devait précipiter l’évolution poli­ tique et sodale qui s’amorçait.101 So Pirenne explained the structural change through two distinct, though inter-connected, circumstances: gradual transformation gener­ ated from within; and eventual collapse dealt by external forces. The role of the latter was decisive, though its great consequence derives from its fusion with a pre-existing and independently activated process of decline. Further to this, Pirenne identified a progressive economic decline prior to the Islamic invasions, given that he stated that “La trasformation a du commencer dès le Ve siècle”} 1 The end of the Roman world and the origins of the social system that dominated Europe during the Dark Ages are explained therefore by a combined process of continu­ ous evolution (or rather involution), and sudden disruption.

Ill In the History of Europe, Pirenne approached the explanation of histor­ ical change also with rather different conceptual models. This occurs when he discusses the problem of the eleventh-century rebirth of towns and the revival of trade and commerce, a theme to which he returned repeatedly following his first essay of 1895 on the subject. Indeed, prior to the First World War, Pirenne’s reputation was attached above all to the history of medieval towns and of Belgium, his native country. Only later, was he to direct his energies to specific analyses of the problem of the late-Roman world. In the History of Europe Pirenne took a very different view of the re­ surgence of trade and the origins of the mercantile class to the one ex­ pressed in 1895. He modified his earlier conviction of the existence of

10 Pirenne, Histoire de l’Europe, p. 39. 11 Pirenne, Histoire de l’Europe, p. 62.

READING PIRENNE AGAIN

25

professional merchants during the “agricultural epoch” of European history before the eleventh century. Now he attributed the revival of trade to impulses coming from the peripheral fringes of the West: Ven­ ice and parts of southern Italy (where political and economic relations with Byzantium had been maintained despite Islamic supremacy with­ in the Mediterranean), as well as the Vikings raiding and trading be­ tween Byzantium and the shores of the North Sea. The combined ef­ fect of these two peripheral currents of traffic penetrated into continen­ tal western Europe reaching northern France and Flanders. This gave new subjects the opportunity to join the trading enterprises. They soon became numerous and active, and created a new social class of profes­ sionell merchants. This group had no immediate predecessors in the so­ ciety of the day, which until then had been marked by the conspicuous absence of any form of permanent trading activity managed by speci­ alised operators. The ancestors of the medieval merchants were the déracinés, the dropouts of the agrarian society. Here Pirenne reversed his opinions regarding this fundamental aspect of European socio-economic history: the appearance of the mercantile class which he saw' as the origin of the bourgeoisie. Previously he had interpreted these events in terms of an increase and acceleration of factors indigenous to westem-European society; now he was describing the sudden appearance of a totally new social group engaged in unprecedented forms of economic activity. Thus, in the same book, Pirenne explained historical change through contrasting lines of interpretation. At the end of antiquity ex­ ternal forces accelerated and exacerbated a pre-existing process of de­ cline, bringing about its eventual (though perhaps not inevitable) con­ clusion; by contrast, in the case of the eleventh-century transition from agricultural to commercial society, external forces are seen to intro­ duce de facto a set of conditions which determine the formation of a thoroughly new socio-economic equation. In this second case, change is more abrupt and discontinuity more evident. When and why Pirenne renounced his earlier ideas on commerce in the Carolingian and post-Carolingian period is unclear. The vitality of commerce in that epoch had been underlined by Alfons Dopsch in 1912-13, in his Wirtschaftsentwicklung der Karolingerzeit. Dopsch’s erudite account was admired by Pirenne, though he was not convinced by the treatment of the evidence upon which it was based.12 Perhaps it was 12 As, for example, in H. Pirenne, “Stages in the Social History of Capitalism”, Amer­ ican Historical Review 19 (1914), pp. 494-515.

PAOLO DELOGU

26

this which induced him to revoke his earlier evaluations of trade in the Carolingian period. In the Histoiy of Europe early medieval trade is de­ scribed as highly-localised activity, restricted to the Low Countries and parts of Italy, with neglegible importance in the economic system of the age. Pirenne appears to follow T. Inama-Sternegg in emphasising the overall prevalence of an economy driven by agriculture and organised within a framework set by seigneurial landlordship. However, aim of ac­ counting for the evidence for trade in the period, and in the interests of maintaining coherency with his earlier ideas on the crisis and col­ lapse of trade after the Islamic invasions, Pirenne endeavoured to ex­ plain the nature and role of exchange within the “agrarian economy”. Pirenne first considered the question in a paper presented to the In­ ternational Congress of Historical Science in London in 1913.13 There he maintained a case for the occasional, sporadic nature of early me­ dieval trade and its non-professional organisation. The volume of com­ modities was small and credit did not exist. Sites specifically dedicated to commercial activities, such as the portus of northern France and the Low Countries did not invalidate the scheme, because they were com­ paratively few; moreover they were located only in particular geo­ graphical sites and did not possess the nature of towns. In this way, Pirenne took issue even with Inama, who had argued for a commercial revival under Charlemagne. The portrait of the Carolingian economy drawn by Pirenne in the Histoiy of Europe was less optimistic; he delin­ eated a chain of economic consequence which deprived trade of real relevance. The disappearance of markets in the West following the closing down of the Mediterranean, extinguished in fact all forms of economic activity which involved the generation of profit through a process of exchange. Once profit ceased to be the prime mover of eco­ nomic activity, agriculture and other forms of production were geared to direct local consumption. In this situation, trade takes on an ‘abso­ lutely secondary’ importance. This resembles a version of the system described by German eco­ nomic theorists and in particular by Buecher, as a “domestic” or “closed economy”. Pirenne’s account, however, differs in attributing its inception to external forces rather than to a spontaneous re-shaping of medieval society. Moreover, he assumed that limited exchanges never ceased because no society can totally do without them. However, he 13 Supra n. 12.

READING PIRENNE AGAIN

27

could no longer explain the eleventh-century revival in terms of the ac­ celeration of a pre-existing activity, as he had in 1895. The impetus for the rebirth of the commercial economy must have been external to the agricultural society of the early Middle Ages; as he put it: Bref, l’histoire du commerce européen ne nous presente pas de tout, comme on aim­ erait le croire, le spectacle d’une belle croissance organique, faite à plaisir pour les amateurs d’évolutions,14 The contrast with his account of the late Antique crisis is clear. In the eleventh century it is discontinuity that prevails; the agents of change had to be external, as the “agrarian epoch” was devoid of indigenous potentiell.

IV This history of European development rests on theoretical concepts re­ ferring to economic systems, their organisation and historical develop­ ment. Pirenne derived them from a number of works which exerted wide-ranging influence on the historical and sociological thought of his time. According to these theories, the economic life of societies passes through a series of stages, each characterised by typical relations be­ tween production, distribution and consumption. Among the most widely-known and arguably most influential works of this time was Karl Buecher’s %ur Entstehung der Volkswirtschaft, published in 1893. Buecher maintained that in historical and ethnological experience, there are three principle stages of economic organisation, which he de­ fines respectively as ‘Hauswirtschaft’, 1Stadtwirtschaft'’ and ‘Volkswirtschaft?. Each stage is characterised by more and more complex and indirect relations between the production and consumption of commodities. Another important theory of economic stages was put forward by Werner Sombart in 1900. Sombart linked the stages of economic or­ ganisation to patterns of social interaction through which the satisfac­ tion of needs is realised; to forms of the institutional organisation of social relations, and to the social psychology of economics.15 Unlike 14 Pirenne, Histoire de l'Europe, p. 156. 15 For a discussion of the theories of Buecher and Sombart, see: B.F. Hoselitz, “The­ ories of Stages of Economic Growth”, B.F. Hoselitz et al., Theories of Economic Growth, (Glencoe Dlin., 1969), pp. 193-238.

28

PAOLO DELOGU

Buecher, Sombart did not argue that the theoretical sequence of the stages necessarily corresponds with the chronological order in which they appear in history. Pirenne did not adhere to any of these theoretical schemes, objecting above all to the manner in which the complex of historical reality is confounded by predetermined models and formulae. He expressed his disagreement with Beucher and Sombart on a number of occasions. He questioned the concepts of “domestic economy” and its relevance and usefulness when applied to late Antique and early medieval eco­ nomics; above all, he questioned the notion of the “town economy”, which Sombart and Beucher used to describe the foundation of the medieval town as an economic system based on direct exchange be­ tween producer and consumer and limited to the confines of the town and its immediate environs. While rejecting many of the specific claims made by the various ex­ ponents of economic theory in this period, Pirenne adopted much of their way of thinking and methodology. One can identify at least three points revealing their influence: the interest in the internal structures of economic systems as they appear in the course of history; the idea that systems can be interpreted as distinct phases of economic organi­ sation; and the attention reserved for the psychological motivations of economic activity. Pirenne’s interest in social psychology did not depend only on the influence of Sombart. In 1901, he expressed his admiration for the work of Karl Lamprecht, because of the attention Lamprecht paid to the “ferments psychiques de la vie économique” For a period the two schol­ ars were close friends, yet when later Lamprecht began to formulate his model of a succession of psychological phases in the historical de­ velopment of the German people, Pirenne communicated his serious doubts concerning Lamprecht’s work and ended in refuting his friend’s conclusions outright.1617 Pirenne’s mature attitude towards development and stages as funda­ mental problems of historical representation, was fully expressed in his address to the International Congress of Historical Science in London, 16 Pirenne’s letter to Lamprecht is cited by: B. Lyon, “The Letters of Henri Pirenne to Karl Lamprecht (1894-1915)”, Bulletin de la Commission Royale d’Histoire 132 (1966), pp. 161-231, esp. p. 212. 17 As Pirenne revealed in a letter to H. Sproemberg, cited by B. Lyon, Henri Pirenne: a Biographical and Intellectual Study, p. 379; see also: H. van Werveke, “Karl Lamprecht et Henri Pirenne”, Bulletin de la Commission Royale d’Histoire 138 (1972), pp. 39-60.

READING PIRENNE AGAIN

29

entitled Stages in the Social History of Capitalism. The intervention mainly aimed to demonstrate the capitalist nature of medieval commerce since its first appearance in the eleventh century; namely that its raison d’etre was the generation of surplus, independent of any necessity to provide for the essential needs of society. Buecher, Sombart and Max Weber had maintained that it was impossible to speak of anything resembling capitalism in the modern sense before the sixteenth century. Taking a different position, Pirenne set out to trace the evolution of European capitalism, its workings and attitudes, from the economic revival of the eleventh century to the industrial revolution of the eighteenth century. As he put it: I believe that, for each period into which our economic history may be divided, there is a distinct and separate class of capitalists. In other words, a group of capitalists of a given epoch does not springfrom the capitalist group of the preceding epoch. At every change of economic organisation, we find a breach of continuity. It is as if the capitalists who have up to that time been active, recognise that they are incapable of adapting themselves to con­ ditions which are evoked by needs hitherto unknown and which callfor methods hitherto unemployed... In their place arise new men... The permanence through the centuries of a capitalist class, the result of a continuous development and changing itself to suit changing circumstances, is not to be affirmed. On the contrary, there are as many classes of capitalists as there are epochs in economic history. That history does not present itself to the eye of the observer under the guise of an inclined plane; it resembles rather a staircase, every step of which rises abruptly above that which precedes it.18

Thus the history of capitalism (and with it the course of history in gen­ eral) seemed to Pirenne to develop and unfold through sequences of steps, each of which represents a distinct phase of economic and social life. Each successive phase was characterised by a new psychological disposition which directed and sustained economic activity, and which was the prerogative of an emergent group of individuals. The theories of the stages of economic activity and its psychological conditioning were thus brought together. Such a reading of the economic history of medieval Europe allowed Pirenne to reconcile his argument for the capitalist organisation of economics from the eleventh century with those of his opponents who refused to admit the existence of a true capitalist economy prior to the sixteenth century. He achieved this by describing the two periods of economic history as two separate distinct phases of a discontinuous evolution.

18 Pirenne, “Stages in the Social History of Capitalism”, p. 494.

30

PAOLO DELOGU

It is worth emphasising how Pirenne conceived the transition from one stage to the next. In his opinion, transition was effected by two sets of agents: the external conditions (or ‘circumstances’) under which eco­ nomic activity was carried out; and the psychology of the individuals and groups engaged in it. Psychology is not responsible for changing conditions; rather, it is interpreted as a pattern of ‘responses’ to altered circumstances. The type of response characterises a phase of historical development. The outstanding problem concerning the causes and rhythms of changes in “circumstances” was not addressed here. In the History of Europe Pirenne made use of the concepts described above as instruments of historical interpretation. This is evident both in his description of the agrarian system of the early Middle Ages and in the transition from this system to the commercial economy which followed. Dark Age society and economics were characterised by the wide­ spread renunciation of the generation of profit. This psychological at­ titude was determined by external forces and hardened into culture. On this basis, Pirenne reconstructed the social order of production as one in which subjection did not correspond to exploitation of the sub­ jected people; a social order capable of producing surplus but not in­ vestment. The psychological set is used to explain the lack of internal movement in a society which had no impulses capable of overcoming the prevailing socio-economic configuration. Pirenne examined the mental attitudes of the three orders of “agrarian” society - landed lords, the clergy and peasantry - concluding that it was not in the in­ terests of any party to disturb a socio-economic equilibrium by which the primary needs of subsistence, security and ‘status’ were satisfied. Given the lack of dynamism of the early medieval economy, the thrust of its had to come from must be sought outside its social structure; fur­ thermore, the very protagonists of new forms of economic activity, were necessarily individuals or groups with no place in the established social order and no interest in its preservation. The psychology of these agents was the convenient response to the changing circumstances, which were imported and imposed upon the medieval West. They en­ gaged in a new economic activity, which transformed the order of so­ ciety; an activity based on the quest for profit. The socio-psychological determination of new social and economic forms was not used by Pirenne to explain the transformation of the late-Roman economy. The new “circumstances” brought about by the Islamic conquests did not give rise to a new entrepreneurial class or set

READING PI RENNE AGAIN

31

of economic innovations; rather, society in the West was forced to adapt to a situation which had made such innovation impossible and useless. The external factors, rather than stimulating creative transfor­ mation, precipitated an economic recession; thus, bereft of trade, the society had to settle on simpler socio-economic forms that had been evolving within it for a considerable time. The transition from Antiq­ uity to the Middle Ages is consequently less innovative than the subse­ quent transition from the agrarian age to the mercantile one. Nonetheless, psychology and mental attitudes played a role also in Pirenne’s description of the demise of the ancient world. He insisted that the success of the Islamic invasions was due to the religious zeal of the Arab conquerors, a psychological and moral attitude which de­ termined their intentional subversion of the pre-existing order. By con­ trast, the Germanic invaders were prepared to acknowledge the inferi­ ority of their culture vis-à-vis the superiority of Roman culture; accord­ ing to Pirenne, it is precisely on account of this cultural mutability that the Germanic peoples missed the opportunity to transform the history of the world. In due course Pirenne was also to consider the economy of the lateRoman West in terms of a static structure. However, it is worth noting that not even in the History of Europe, did he care to explain the reasons for the change of “circumstances” whose “responses” he expounded at such length. Change appeared to be produced by chance; throughout his life, Pirenne frequently meditated on the role of chance in history.19

V Pirenne did not publish his History of Europe. Probably he was aware of the polemical bias that limited its reliability. Instead, he dedicated a number of separate studies to its central themes; in particular he re­ turned to the problem of the transition from antiquity to the Middle Ages in two brief studies, Mahomet et Charlemagne in 1922, and Un con­ traste économique: mérovingiens et carolingiens in 192 3.20 His intention in these essays was not to consider the crisis of the late-Roman economy 19 Lyon, Henri Pirenne: a Biographical and Intellectual Study, p. 267. On the role of chance in history, see: H. Pirenne, “La tache de l’historien”, Le Flambeau 14 (1931), pp. 5-22 esp. 6, 18. 20 Published in Revue Belge de Philologe et d’Histoire 1 (1922), pp. 77-86; and Ibid. 2 (1923), pp. 223-235, respectively.

32

PAOLO DELOGU

as such. Pirenne’s principle objective was rather to claim a crucial place for the Carolingian period in the history of Europe. He believed that during this period, the creative force of civilisation was relocated in the regions between the Seine and the Rhine whilst the Mediterra­ nean countries lost the creative and determinative role which had been theirs since antiquity, becoming a periphery and the frontier of Eu­ rope. It is in this dramatic disruption of an age-old settlement, conceived in terms of Universalgeschichte, that Pirenne located the beginnings of the history of modern Europe. In his view the geo-political settlement in­ augurated by the Carolingians had lasted until his own period and still characterised European civilisation. In the above-mentioned essay he only hinted at the two complemen­ tary tenets of his thesis on the transformation of the Roman world: the irrelevance of the barbarian invasions as the cause of the transforma­ tion and the decisive impact of the Islamic conquests in the crisis of the late-Antique Mediterranean world. A year later Pirenne returned to the subject. This could have been prompted by the new essay of Alfons Dopsch, Wirtschaftliche und soziale Grundlagen der europäischen Kulturentwicklung published between 1918 and 1919. Dopsch proposed a pattern of the continuous evolution of socio­ economic structures in the period between the fall of the Roman em­ pire and the emergence of the Carolingian. He maintained that no caesura or dramatic change could be perceived through the whole pe­ riod and consequently the Carolingian epoch did not constitute such a new set of historical circumstances as Pirenne believed. Pirenne held firm to his thesis, again addressing his arguments to the new position of the Carolingian period in every aspect of early medi­ eval civilisation. He set out to substantiate his argument in the field with which he felt most comfortable, which was also that worked by Dopsch: socio-economics. He framed his response to Dopsch as a com­ parison of Merovingian and Carolingian economics, in such a way as to reveal the diversity and originality of the latter. He described the economic systems of the two periods as two distinct models or types,utilising concepts with which he had experimented in the History of Europe, although now he made them more explicit.21

21 ‘A comparer les temps carolingiens aux temps mérovingiens, ce que l’on constante c’est un changement de type’: Pirenne, “Un contraste économique: mérovingiens et car­ olingiens”, p. 230.

READING PIRENNE AGAIN

33

These types were characterised by the same structural factors identified in 1895: towns and trade. Moreover, while Pirenne accepted that, as during the Roman Empire, agriculture represented the prevalent fac­ tor of the Merovingian economy, and land-ownership its staple, he maintained that the structure or, as he preferred it, the type, of econo­ my was characterised by the level and organisation of exchange mech­ anisms, as trade was not conducted independently of the agricultural activity but was thoroughly integrated with it. Its prevalent character­ istics were: the presence of a professional mercantile class, made up not only of eastern traders but also by indigenous merchants; the circula­ tion of more commonplace goods alongside high-status commodities; the movement of commercial traffic within inland Merovingian Gaul as well as along Mediterranean coastal routes; and the widespread availability of a strong currency. These factors reveal within the eco­ nomics of the Merovingian period a broad and permanent network of traffic; the pre-requisite for considering trade and exchange as fully in­ tegrated aspects of the economic structure. The situation under the Carolingians was very different. Exchange and the movement of goods (the existence of which is not denied by Pirenne) lost their former characteristics, taking on a very different ap­ pearance and function: Carolingian economics consisted of an mosaic of restricted and self-sufficient economies and of a largely inconsequen­ tial traffic of precious non-essential commodities; this latter was tran­ sient in its organisation, and was for the most part serviced by groups of individuals drawn from outside the agrarian society: adventurers and foreigners. Pirenne was ready to explain the appearance under these circumstances of markets and coinage: trading places were temporary, make-shift centres; coinage was rare and its circulation limited. All in­ dicators for the Carolingian period point to exchanges that were occa­ sional and sporadic or highly-localised in character. The exercise of historical dissection of the workings of economies, as they emerge from documentary sources, is prominent in this essay, which for this reason, is a significant product of Pirenne’s own histori­ ography. The psychological determinants of economic activity were now disregarded in favour of a functionalist analysis of systems. Yet the typological approach used by Pirenne rendered the descrip­ tion of structures even more rigid than it had been in the History of Eu­ rope', eventually this rigidity affected the very presentation of the late Antique and Merovingian economies. Internal movement and change became irrelevant, given that the type of economy was described and

34

PAOLO DELOGU

characterised by those aspects which appear to be constant and which remain so as long as the type itself survives. In this way, Pirenne gave up the progressive movement of Merovingian society and its economy on to the land and the consequent disintegration of the towns. He stat­ ed that the transition from Merovingian to Carolingian economics oc­ curred gradually, not as the result of changes in internal structures, but of the successive recessions caused by the progress of the Islamic con­ quest, which by its every turn provoked a significant decline in trade until its eventual extinction. Under the Carolingians, economics were redefined, though in rather a different sense than that intended by Inama-Sternegg and Dopsch: trade having become impossible, the econ­ omy was adapted to functioning without it, taking the static form that Pirenne had already described in the History of Europe. There are odd­ ities in this historical interpretation. Pirenne was interested in move­ ment. In a letter to Heinrich Sproemberg of 1931 he writes: Ce qui mefrappe surtout dans l’histoire ce sont les mouvements de masse que l’observation empirique révèle et qui m’apparaissent comme les réalités les plus scientifiquement observ­ ables de l’évolution historique,22

In his essay Mahomet et Charlemagne of 1922, Pirenne made a case for the continuity of the ancient socio-economic system through the Merovingian age and took issue against the customary division of his­ torians into opposing camps of ancient and medieval history. He re­ garded this practice as a misleading and intolerable barrier to a more sophisticated comprehension of the transition from Antiquity to the Middle Ages: De point de vue proprement historique, ce sontjustement les périodes intérmediaires, les péri­ odes de transition qui s’imposent surtout à l’attention, parce que c’est en elles que peuvent le mieux s’observer les changements sociaux quiforment l’objet même de l’histoire.23

Despite taking this line, in the following year he went on to describe the relationship between the Merovingian and Carolingian periods in terms which suggest precisely the same kind of historical rupture that he was refused to acknowledge for the fifth century. In constructing his ‘typology’ of socio-economic systems, Pirenne did not pay attention to the variations in the relative frequency and inten­

22 Lyon, Henri Pirenne: a Biographical and Intellectual Study, p. 379. 23 Pirenne, ‘Mahomet et Charlemagne’ , p. 210.

READING PIRENNE AGAIN

35

sity of historical events that may allow a description of social change in terms of continuity. Thus, all evidence for the ‘period of transition’ was lost in the historical representation; the “période intermediaire” is therefore presented as an obscure time lapse suspended between two historically-distinct systems, between which no links of continuity can be detected. Hence, the change in ‘circumstances’ must be ascribed to external and independent causes. On this basis the source of change was best considered the result of an event, rather than of a broader historical process. The Islamic invasion of the Mediterranean was well suited to the role of determinant factor in the decline of antiquity to the Dark Ages.

VI Pirenne’s meditation on the transformation of the Roman world found mature form in Mahomet et Charlemagne, the book which repeated the title of the former article. In this the hypotheses oudined by the essays of 1922 and 1923 were developed and supplemented with an abun­ dance of source material. Perhaps it was precisely the detailed treat­ ment of the evidence that led the book to encounter the harsh criticism of other scholars, who were quick to point out the weakness of its ten­ ets, above all those concerning the evolution of the economy and soci­ ety of late Antiquity and the supposed consequences of the Islamic con­ quests. In fact, Pirenne gathered a huge amount of evidence to give the ideas that he had already outlined a more solid foundation. However, he regularly neglected to pay attention to the geographical and chro­ nological context in which every piece of evidence occurred. He re­ marked on the irregular occurrence of structural indicators but did not care to delve deeply into the circumstances. He frequently chose one of several possible explanations of a given fact basing his conclusions not on rational evaluations but on pretentions or false statements. There is only one theoretical justification, albeit a contentious one, to this method; the silent assumption that the survival of individual as­ pects of late-Roman economics provides testimony of the continuity of the late-Roman economy in its entirety. Such presumption led Pirenne to neglect the symptoms of the progressive disintegration of late An­ tique systems which can be shown to predate the Arab conquests and even the Germanic invasions. By placing emphasis on this evidence,

36

PAOLO DELOGU

Pirenne’s critics presented a very different version of the crisis of late Antiquity, casting serious doubts on the historical legitimacy of the model he proposed. Similar reservations inflicted more limited damage to his account of the Carolingian period, in which Pirenne himself ac­ knowledged irregularities, though he endeavoured to integrate these into his thesis. Nevertheless, Mahomet et Charlemagne must not be cast as the simple restatement of themes previously dealt with in Un contraste économique', a restatement probably worsened by the fragile treatment of the evi­ dence. Despite the method and substance contained in this last work, one must acknowledge that the narrative is richer and the problems faced are in part new. The principle objective of Mahomet et Charlemagne was the same as in the essay of 1922; that is to establish a claim for the significant novelty of the Carolingian period. This Pirenne did by describing how in that period a new Europe was forged, not merely in economic terms, but also in political, social and cultural terms. The crisis of the ancient or­ der caused new political forces to emerge to supplant moribund pre­ decessors; this they were able to do by basing their authority on novel economic and cultural forms, more adapted to the altered circumstanc­ es of the day. Such forces were embodied by the Carolingians, the rep­ resentatives of Austrasian society, which Pirenne regarded as the leastRomanised society of Merovingian Gaul, and thus the least affected by the crises afflicting Roman institutions and economy. However, while the Carolingians progressively assumed exclusive authority throughout the Frankish kingdoms and the political hegemony over the Christian West, a parallel process was occurring with the Papacy, the only “Med­ iterranean” institution remaining in the West following the Arab con­ quest, as its protagonist. Abandoning their former ties with the Byzan­ tine East, the popes redirected their interests towards the north, finally throwing their political lot with the Carolingian kings in a formal alli­ ance. The importance of these parallel processes is derived from Pirenne’s realisation that the rising Europe consisted not only of crude economic or political quantities, but presupposed a unitary cultural ethos drawn from the traditions of the Roman church. In this version of the transition, the Islamic conquests did not simply result in the interruption of Mediterranean trade. In the causal chain the direct effect of the conquests was only that of robbing the Merov­ ingian kings of the financial foundation of their power. On a larger scale Islam contributed to the birth of Carolingian Europe by subtract-

READING PIRENNE AGAIN

37

ing Spain and North Africa from western Christendom. The political transformation that took place in Gaul derives its significance from these facts. At the same time Islam confined the activities of the Byz­ antine Empire to the Eastern Mediterranean and thus contributed to the isolation of the papacy from the East, and therefore to draw it pro­ gressively towards the north-western connection and the kingdom of Charlemagne. Pirenne stressed that: L ’empire de Charlemagne est le point d’aboutissement de la rupture, par l’Islam, de l’équilibre européen. S’il a pu se réaliser, c’est que, d’une part, la séparation de l’Orient d’avec l’Occident a limitée l’autorité du pape à l’Europe occidentale; et que, d’autre part, la conquête de l’Espagne ed de l’Afrique par l’Islam avaitfa it du roi des Francs le maître de l’Occident chrétien. Il est donc rigoureusement vrai de dire que, sans Mahomet, Char­ lemagne est inconcevable.24

Once again, one may recognise the application of a principle formed by Pirenne since 1913. New circumstances determined by external in­ fluences, result in the emergence of new historical subjects. The level on which this theory was applied and verified was now considerably more extensive and complex, since Pirenne was seeking here to de­ scribe not the emergence of a single social group, but of a new civilisa­ tion: an organic complex of political institutions, intellectual culture, lit­ eracy, learning, art and of connections between church and state. In Pirenne’s view, political initiative played a crucial role in the creation of the Carolingian civilisation, in as far as it promoted the circulation of individuals and ideas, the formulation and realisation of projects, the fusion of traditions and the creation of cultural institutions. In fact, in Mahomet et Charlemagne, Pirenne described the events which gave shape to the Carolingian world in terms of a coherent political evolution starting from late Merovingian times and aiming to reconstruct the whole Christian West into new forms. This was exactly the sort of passage which Pirenne declined to go into when he discussed the relation between the Merovingian and the Carolingian economies. This may have resulted from the fact that he did not consider Dark Age economy as an effect of innovation, but rather as an adaptation to worsened conditions; for him it was not worth examining it in terms of evolving mechanisms, but only in terms

24 Pirenne, M ahom et et Charlemagne, p. 210.

38

PAOLO DELOGU

of economic constraints in which the creative forces of the age were bound to operate. On the other hand, it will be observed that Pirenne’s interest in eco­ nomic and social organisation constitutes but a single aspect of the his­ torical panorama distilled in Mahomet et Charlemagne; this may be appar­ ent from, among other things, the position of chapters dealing with so­ cial and economic history after those which set forth the order of polit­ ical events and the creation of the state institutions. The formal structure of the thesis may indicate that, in the construction of the framework of society, Pirenne attributed pre-eminence to political will rather than economic determinants; at the same time, he regarded po­ litical and juridical institutions both as products of the culture in which they appear, and as instruments for the propagation of social ethos. Indeed Pirenne explicitly asserted his conviction that movement in history is caused by the human will, being the product of the conscious intention and action of individuals or social groups. He also held that success in history depends on the moral foundation of the action and on the personal skill of the individuals. This recalls the explanation he proposed for the success of the Islamic conquest, or the emphasis with which he underlined the degree of enterprise and ‘intelligence’ of the new capitalists, revealed in his eyes even by their assumed knowledge of foreign languages. All this does not conflict with the basic assumption that action is al­ ways firmly contained within the parameters set by economic condi­ tions. In Mahomet et Charlemagne, political and economic histories are tighdy interwoven. However, the former achieves pre-eminence in moulding the new European civilisation. Mahomet et Charlemagne is a his­ tory of the birth of Europe as a political and ideological entity, al­ though in the field of economics Dark Age Europe remained primor­ dial. It is worth remarking that Pirenne modified a number of controver­ sial statements he made in the History of Europe, in that he acknowl­ edged the Germanic elements present in Carolingian culture. In the rise of Europe He germanisme commence son role. Jusqu’ici la tradition romaine s’était continuée. Une civilisation romano-germanique va maintenant se dévelop­ per’.25 The focus of cultural innovation moved northwards exactly be­ cause in the continental regions of Europe, Romano-Germanic cross­ fertilisation was most diffuse and intense. 25 Pirenne, M ahomet et Charlemagne, p.211.

READING PIRENNE AGAIN

39

Nothing could therefore be more inappropriate than to reduce the Pirenne Thesis to an essay on the transformation of the economic structures of the ancient world. But since this aspect is at least one con­ spicuous focus of Pirenne’s treatment of the theme, the conceptual lim­ itations which determined the thesis’ intrinsic weaknesses must be iden­ tified and explained. In doing this, it should be remembered that Pirenne drew upon a vast corpus of sources for social and economic history, and that he was ready to appraise the variety and irregularity with which such data appears; neither did he allow himself the conve­ nience of traditional historical périodisation or theoretical formalism.26 By so doing, he was able, for instance, to illustrate the continued ac­ tivity of trade routes long after the fall of the Roman empire. Similarly, he identified a number of indices of significant change between the sev­ enth and eighth centuries, the importance of which has been demon­ strated by recent research; he also illustrated the different roles that can be assumed by one form of economic activity across different periods or in different circumstances. Many of these perceptions are still useful, and certain of the prob­ lems addressed by Pirenne continue to occupy historical debate. De­ spite this, his contribution is coloured by the general presuppositions on the basis of which he marshalled his data and formulated his hy­ potheses. Pirenne was no student of economic or social theory. From scholars active in these disciplines he borrowed more in the way of su­ perficial influence and modes of reasoning than he did with regard to actual doctrines or canons for historical interpretation. It has been seen how Pirenne’s mode of representing the structure as a type, although more in a historical than a theoretical perspective, prevented him from adequately quantifying change occurring within the structure. Furthermore, in giving order and significance to his ma­ terial, Pirenne held two basic criteria: firstly, he assumed the perceived social and economic homogeneity of the whole of the West throughout the transition from late Antiquity to the early Middle Ages; and sec­ ondly he distinguished the successive economic systems in the history of Early Medieval Europe with reference to their level of commercial­ isation. Pirenne’s inter-connected analytical hypotheses kept him from a thoroueh awareness of the place of regional economies in the crisis of 26 This was noted by F.L. Ganshof, “Les grandes théories historiques de Henri Pirenne”, Henri Pirenne. Hommages et souvenirs 1 (Brussels, 1938), pp. 208-212.

40

PAOLO DELOGU

the antique world, though he recognised their existence in the Low Countries and in Italy. They also prevented him from realising that systems of exchange were not to be reduced to the opposition between long-distance and local trade; a third and original sort of trade might have existed founded on local and inter-regional exchange. Besides, Pirenne does not offer a single coherent theory of the rela­ tion between economics and social change. While for him economy provided the essential background within which societies .operate, he did not investigate the ways in which social operation modifies eco­ nomic structure. Consideration of these broader limitations may go further than the critique of individual statements and hypotheses in ex­ plaining why Pirenne is consistently called into question. Nevertheless, their very nature still prompts reflection and explains the continued standing of Pirenne’s historical research. [Translated by Matthew Moran]

THE PRODUCTION AND DISTRIBUTION OF BOOKS IN LATE ANTIQUITY Carlo Bertelli

Books are indicators of wealth in any time, but in late Antiquity they have a particular significance in this sense. They denote a level of lit­ eracy which is well beyond the merely functional and suggest a lifestyle consistent with that of the upper classes. Furthermore, in the sixth cen­ tury books were much more expensive than in the age of the printed page. Palaeography, codicology and the history of literature and cul­ ture have examined many important aspects of the late Antique book. In the context of the present volume, I will limit this report to sixthcentury book ownership and production as a material evidence, leaving aside the significance of the book in terms of spiritual or technical and artistic contents. It has been amply demonstrated in recent studies, that in the Byz­ antine world, and obviously even more in the West in the period under consideration, book-ownership had a quite élitist character.1 This so­ cial restriction was not a novelty. Although in the second century, Gellius had been offered some Greek books for sale at the harbour of Brundisium, which he found “full of wonders and tales”, a public popularity for books never existed in Antiquity. What we would term “popular literature” was in fact intended for the entertainment of the well-to-do. Furthermore, in the sixth century, if the circulation of books was lim­ ited by social factors in the East, in the West, an additional barrier was formed by the presence of large groups, in some cases militarily and politically dominant, for whom Latin was a foreign language, while the number of people who were able to read Greek was rapidly decreasing.

1 The traditional assumption that literacy was widespread in the Byzantine world (R. Browning, “Literacy in the Byzantine world”, Byzantine and Modem Greek Studies 4 (1978), pp. 39-54) was widely challenged at the Dumbarton Oaks Symposium of 1971: Byzantine Books and Bookmen: A Dumbarton Oaks Colloquium (Washington DC, 1975). See also the ex­ cellent survey by M. Mullett, “Writing in early medieval Byzantium”, The Uses of Literacy in Early Medieval Europe, ed. R. McKitterick (Cambridge, 1990), pp. 156-185.

42

CARLO BERTELO

As Cassiodorus put it, “Other people have arms, the Romans only el­ oquence”.2 Equally, we should not be misled by the sketchy and somehow “popu­ lar” features of some illustrated books of the fifth to sixth century. For example, many of the illustrations in the Ilias Ambrosiana, a codex pro­ duced in Alexandria at the turn of the fifth century, may appear to be a naive reduction to pure story telling of some of the most dramatic parts of the poem. However, the fact that illustrations of the same style as those of the Iliad, appear on some contemporary painted glass goblets, also from Alexandria but exported to Bactriana, points unequivocally to the taste of a rich international clientele.3 The illustrations of the Virgilius Romanus, a codex from Ravenna to which we will come later, were also considered “childish” by the learned Nicolai Kondakov, though the book was certainly not intended for children.4 The same applies to the illustrated codices of the Fables of Avianus5 or to the original of the Apocalypse from Trier,6 however crude the work of their copyists in the ninth and eleventh centuries may have been. Corresponding with the elementary character of the illustrations, the Latin reworking of Aesop’s fables by Avianus denotes that there was a public, who, although able to purchase a book rendered more expensive by its illustrations, were un­ willing or unable to read the fables in their original Greek. The material for book making had always been extremely expensive. W. H. Harris remarks that the cost of two obols for a single sheet of papyrus, documented in Egypt in the second century A.D., if com­ pared to the contemporary daily wage of six obols for a skilled labour­ er, would equal the price of thirty to thirty five dollars for a sheet of 2 Cassiodorus, Variae, ed. Th. Mommsen, Monumenta Germanica Histórica, Auctores An­ tiquissimi 12 (Berlin, 1882), IX, 21, pp. 286-87. 3 R. Bianchi Bandineüi, Hellenistic-Byzantine Miniatures of the Iliad (Ilias Ambrosiana) (Ol­ ten, 1955); F. Coarelli, “Studi miscellanei”, Seminario di Archeologia e Storia deW Arte Greca e Romana dell’ Università di Roma / (Rome, 1961), pp. 29-41. 4 N. Kondakof, Histoire de V art byzantin (Paris, 1886) (reprinted New York, 1970 ), pp. 74-75.

5 Most likely to be a sixth-century production, the manuscript is known from a copy made in Limoges c. 1030 by the monk Ademaras; see: G. Thiele, Die illuminierte latinische Aesop in der Handschrift des Ademar Codex Vossianus lat. oct. 15 (de Vries, Suppl. I ll ) (Leyden, 1905); A. Cameron, Macrobius, Avienus and Avianus, Classical Quarterly n. s., XVII (1967), pp. 385-96, dates the work of Avianus around 430. According to Thiele the text may always have been illustrated; A. Goldschmidt, An early manuscript of the Aesop Fables of Avianus (Princeton, 1947), ascribes the model of Ademaras to the sixth century and lo­ cates its production to the monastery of Saint-Pierre in Vienne. 6 R. Laufner and P. Klein, Trier Apokalypse, Vollständige Faksimile Ausgabe im Originalfor­ mat des codex 31 der Stadtbibliothek Trier, I, Faksimile, II (Graz, 1972).

THE PRODUCTION AND DISTRIBUTION OF BOOKS IN ANTIQUITY

43

paper today.7 The cheapest and the most common writing materials were wood tablets. In the Roman fortress of Vindolanda (c. A. D. 100) the remains of discarded correspondence have been found in the de­ bris. They were very thin pieces of new wood which had been folded over and tied round the middle, before being sent to the local officers.8 These leaf-tablets, as their finder and illustrator has called them, were apparently the means of transmitting orders and information which was then in use in the army. However, no similar material can be sup­ posed in more refined correspondence. Saint Augustine, writing from Hippo, asks his correspondent to excuse him for using parchment in­ stead of papyrus and declares that he would have giadly written the letter on his ivory tablets, if he had not sent them already to somebody else.9 Papyrus was utilised in the Byzantine chancery for centuries. After the Byzantine reconquest, Ravenna remained the sole administrative center in the West which regularly used Egyptian papyrus for its records.10 The Visigothic kings made also use of papyrus whenever it was available. Their last deed on papyrus is dated to the year 692, at a time when Calif Abd al’ Malik had banned the export of this material in defiance of the aggressive politics of Justinian II.11 However, in spite of this blockade, a few years later in 716, the abbey of Corbie was granted a tax exemption for buying just fifty sheets of Egyptian papyrus which were included among a number of other luxury items from the Orient.12 It is interesting to note the small amount of this precious writ­ 7 W. H. Harris, Literacy in Antiquity (Cambridge, Mass, and London, 1989). On the price of parchment in Byzantium, see: N. G. Wilson, “Books and readers in Byzantium”, Byiantine Books and Bookmen:A Dumbarton Oaks Colloquium(Washington D.C., 1975), pp. 1-16. 8 A. K. Bowman and J. D. Thomas, Vindolanda: The Latin Writing-Tablets (London, 1983), pp. 29-31. 9 Augustine, epistula ad Romanos, 15. 1 = J.P. Migne, Patrologie Latmae Cumiso Completas, xxxiii (Paris, 1865), pp. 80 and ff: non haec epistula sic inopiae chartae indicai ut membrana saltern abundare testetur. Tabellas ebúrneas, quas habeo avúnculo tuo cum Metis misi. Tu enim buie pelliculae facilius ignoscas, quia deferti non potuit quod ei scripsi The present tense ‘habeo’ should indicate that the ivory tablets were to be returned to their owner. Cf. K. K. Hulley, “Light cast by saint Jerome on certain palaeographical points”, Harvard Studies in Classic Philology 54 (1943), p. 84; E. Ams, La technique du livre d’après saintJérôme (Paris, 1953), p. 27, gives other examples. 10 L. Santifaller, Beiträge zur Geschichte der Beschreibstoffe im Mittelalter I (Graz, 1953). 11 J. Irigoin, “Les débuts de 1’ emploi du papier à Byzance”, Byzantinische Zeitschrift 46 (1953), pp. 314-319; J. O. Tjaeder, “Die nichtliterarischen latinischen Papyri Italiens aus der Zeit 445-700”, Acta Instituti Romani Sueciae series in 4°, 19, pp. 1-3 (Lund, 1954 and 1955; Stockholm, 1982); G. Geraci, “La papirologia e i papiri ravennati”, Tesori nascosti (Exhibition catalogue - Ravenna 1991)(Milan, 1991), pp. 39-44, and files by U. Zaccarini, pp. 46-52. 12 A. R. Lewis, Naval Power and Trade in the Mediterranean, 500-1000 (Princeton, 1951), p. 92, passim.

44

CARLO BERTELLI

ing material which the abbey was prepared to buy. While papyrus be­ came rare in the West, in the Orient it remained a less expensive writ­ ing material than parchment, and as such it was still in use in Anatolia in the tenth century.13 Symptoms of the progressive neglect in the upper classes of the clas­ sic paideia, with the consequently diminished interest in books of literary contents, are apparent from the contents of late papyrus codices writ­ ten in Latin and found in Egypt. A significant early presence is that of Virgil, often in bilingual texts, which is an obvious indication of the desire to maintain ties with Rome. Bilingual editions of Virgil are found in limited quantity from the second century.14 They continue to be present among the findings of papyrus codices ascribed to the sec­ ond half of the fourth century15 and to the fifth.16 Later on, juridical texts prevail, possibly demonstrating a change of interests, passing from a humanistic choice to a more pragmatic one. These Latin books do not suggest the leisurely land owner in his villa but rather the notary, the lawyer or the municipal officer who needs Latin to read the law or for dealing with people from the Latin speaking provinces. Cicero’s Orationes appear on papyri from Egypt in the fifth century17 while Latin juridical texts reach also in Egypt the sixth century.18 Ac­ cording to Lowe, they were all “written in an important centre of legal studies in the Eastern part of the Roman Empire, probably at Byzan­ tium”. These fragments, to which many others can be added, prove an in­ terest in Latin texts which precedes the first campaign of Justinian in Italy (540) and predates some celebrated events about the official res­ toration of Latin at the court, such as, for instance, the speech deliv­ ered by the poet Corippus at the accession of Justin II.19 There is no doubt that the use of Latin in the Egyptian examples had no ideolog­ ical motivations. It cannot be taken as sign of revival, as it is due only

13 For Arethas, ep. 40, sec: Arethae scripta minora I, cd. L. G. Westerink (Leipzig, 1968), p. 297; also Mullet, “Writing in early medieval Byzantium”, p. 158. I+ E.A. Lowe, Codices latini antiquiores: A Palaeographical Guide to Latin Manuscripts Prior to the Mnth Century (I-XI), plus Supplement (Oxford, 1935-71), 1816, 1832, 1833. 15 Codices latini antiquiores, Supplement, 1813. 16 eg. Codices lahm antiquiores XI, 1651, 1652, 1653; Supplement, 1832 17 - Codices latini antiquiores, Supplement, 1839. 18 See Codices latini antiquiores, XI, 1645, 1711 (first half of the sixth century), 1713 (A.D. 529-535), 1722, 1723 (post A.D. 533), ** 1037. 19 Corippus. In laudem Iustini Augusti minoris, ed. A. Cameron (London, 1976). Cfr. J.J. O ’Donell, “Liberius the Patrician”, Traditio 37 (1981), pp. 31-72.

THE PRODUCTION AND DISTRIBUTION OF BOOKS IN ANTIQUITY

45

to the persistent ties with the Byzantine state and with its Latin corpus of laws. B. Bischoff and V. Brown were able to reconstruct a now dispersed papyrus codex, dated by Lowe to the fourth-fifth century, and written in Italy, which gathered together Cicero, the Psalms and the Carmen de Alceste.20 We have unfortunately no hint about when this miscellany, which was obviously made for personal use, came to Africa and with whom. Apparently it belonged to somebody who did not suffer the problems of a Saint Jerome, who was split between the seduction of Cicero and his love for the Bible. After a hiatus during the first years of the Vandal occupation, the school in Carthage reopened, under the auspices of king Gunthamuns, who died in 496. On that occasion the master of grammar Felicianus was called to Carthage to “bring back to Africa the letters which had emigrated before”,21 while the Anthobgia Latina preserves a number of names of the last Latin poets before the Arab conquest. O f course “school” in this period did not mean a compulsory scho­ lastic system for all the children of the country. The school continued to attract the children of the middle and upper classes, especially those who were destined for a career as notaries, masters and even priests. It is hard to estimate how many books they required, although it is un­ likely to have been very many. The royal decree concerning Felicianus mentions the institution of his salary, but is silent about any school equipment. The monastic rules provide some insight about the provision of books in the case of a very specific religious school. A detailed model stressing the role of the book is given by the Regub Magistri. In winter­ time, it prescribes that monks and pupils should read for three hours a day. During these three hours, monks were divided into groups of ten, of which nine had to listen to the remaining one who read aloud. At the same time children were also grouped in tens and are thought to have written on their tablets under the direction of a master. Illiter­ ate adults were required to gather in a group of fifty to be instructed

20 Codices latini antiquiores, Supplement, at XI, 165). 21 P. Riché, Education et culture dans 1’ Occident barbare, Vie- Ville siècles (Paris, 1962) (Sec­ ond edition, 1972 ), pp. 76-77; P. Langlois, “Dracontius”, Reallexikonfiir Antike und Chris­ tentum, Sachwörterbuch zur Auseinandersetzung des Christentums mit der antiken Welt, ed. T. Klauser (Stuttgart, 1950-), 6 (1959), pp. 197-210; J. M. Diaz de Bustamante, Draconcio y sus connina profana (Santiago de Compostela, 1978).

46

CARLO BERTELLI

in reading and writing, while those who did not know the Psalms by heart had to exercise their memory. Following this three-hour period, all went to the religious service. In summertime, between meal and sunset, they were required to read, listen to readers, teach or repeat the Psalms. At the end of the “course” they had to demonstrate to the abbot that they knew the Psalms by heart. The same applied to the canticles and other divine texts.22 It does not seem that a high degree of literacy was required or envisaged and as for books, only the Psalms are discussed, with no mention of the Bible or the Gospels. The Rule of Saint Benedict suggests a greater quantity of books. “If someone wishes to read after meal when all the others rest, he is al­ lowed to do so, on condition that he will not disturb the others... In the period of Lent every monk shall receive a book that he shall read entirely and orderly. But those who are unwilling or unable to read, shall do some other work”.23 There is no hint as to whether these books were home produced, purchased, inherited or otherwise. Nonetheless, some interesting information can be gleaned from the Benedictine text. The reading was individual and monks were allowed to read silently. The books were kept in a special place under control with a system of distribution to the individual readers. There were also illiterate members of the community. In fact, as we know from the Rule, the application to enter the monastery could be signed by a third party if the applicant was unable to put his own signature on the form. The rule of the Master however, suggests fewer books, while there is a certain insistence on the tablets. The latter were the general me­ dium for the transmission of the written word. They were not made for permanent record. The script could be erased and the tablet re­ used. This was particularly the case with waxed tablets, which were the most common, on which the scribe could quickly incise his text with a stylus and erase it when it was obsolete. The norm, in the time of Saint Jerome, was to write a text first on a tablet and then to give it to a scribe for transcription on papyrus or parchment.24 The words left on the tablets were destined to vanish after that they had completed their temporary function, the tablet effectively being a form of dictation.

22 Regula magislri, c.50 (8-17) and (62-69): A. de Vogué, La Règle du Maître (Paris, 1964), pp. 225 and 235. 23 Regula Benedicti, cxlviii, ed. A. de Vogué (Paris, 1972), pp. 181-82. 24 Ams, La technique du livre d’après saint Jérôme, p. 31.

THE PRODUCTION AND DISTRIBUTION OF BOOKS IN ANTIQUITY

47

Tablets remained the cheapest instrument for writing. At the same time they were a very personal implement and many carried them at­ tached to their belt. This personal touch gave them a certain distinc­ tion. To receive the gift of some tablets was appreciated in the same way as a present of personal writing paper today: “I like the affection­ ate thought more than the use to which I will put this present”, wrote a student of Hildesheim to his friend who had sent him some tablets.25 Tablets could also become dangerous weapons, as was discovered by a young pupil in Corvey, who was killed with some tablets wielded by one of his comrades.26 When Saint Patrick appeared in Ireland ctim tabulis in manibus scriptis, he risked being lynched with his companions, because the people took the wooden tablets to be some kind of men­ acing weapon.27 A set of tablets found at Springmount bog, now in the National Mu­ seum in Dublin, are about 75 x 210 mm., and around 7 mm. thick. They date to c. 600 and are inscribed with a text of Psalms 30-32. They seem to have been lashed together as a group of six, with their waxed sides together.28 Given the practical use of tablets from Antiquity to the end of the early Middle Ages, the appearance of ivory tablets suggests an addi­ tional distinction. It is clear that such luxurious items can have be­ longed only to the very rich members of the bookish, or at least literate, élite. Their value is indicated by the fact that they embellish a quite primary instrument, and moreover, a very humble and common one. The use of ivory diptychs is well documented in the fourth century, but it is only in the sixth that we have consular diptychs from the East, the last western specimen being dated 476. Consular ivory diptychs

25 H. Südendorf, Registrimi oder Merkwürdigen Urkunden III, 2 (Leipzig, 1949-54). See: P. Riche, Les écoles et l’ enseignement dans l’ Occident chrétien de la fin du Ve siècle au milieu du Xle siècle (Paris, 1979), p. 222. 26 Vita Anscharii 4, Monumenta Germaniae Histórica Scriptores II, 692: Riché, Les écoles et l’enseignement dans l’Occident chrétien de lafin du Ve siècle au milieu du Xle siècle, p. 222. 27 J. Stevenson, “Literacy in Ireland”, The Uses of Literacy in Early Medieval Europe, ed. McKitterick, pp. 11-35, at p. 20. 28 Stevenson, “Literacy in Ireland”. Tablets were used for painting ‘itineraria’ and geographic surveys, a practice frequently referred to by St Jerome. Abbot Adomnàn cop­ ied the maps of the Holy Places from the tablets brought by Arculf: "primo in tabulis describenti fideli et indubitabili rumatime dictavit; quae nunc in membranis brevi textu scribuntuiDe loris sanctis, ed. and transi. D. Meehan and L. Biler (Dublin, 1958), p. 36; ‘has ...figuras... justa exemplar quod mihi ... sanctus Areufus in paginólafigurami cerata depinximuL, pp. 46-47.

48

CARLO BERTELLI

from Constantinople appear from 506 to 540, prior to the effective ab­ olition of the consulship in 541.29 We may consider these precious objects as an expression both of the crystallisation of intellectual activity, and of the high regard in which it was held. The two cónsules ordinarli were chosen directly by the em­ peror and their magistrature was valid for the whole empire. Their first task, after their nomination, was to read, in front of the emperor, on the first day of the year, a public speech, the gratiarum actio, which they were supposed to have written in advance. After that they proceeded to the hippodrome, and by waving the mappa, a handkerchief with a solar representation, they opened the races. This gesture was the most frequent in the representations carved on the external faces of the diptychs. The inner sides, however, were waxed and ready to be used for writing. The recipient of the diptychs could utilise this precious diary throughout the year, while the inscriptions on the exterior sides give the name of the consul and his titles, although Justinian had his dip­ tychs inscribed with the dedication to the members of the senate, as did one of the last consuls, Philoxenos, the first to translate the Latin senatus in the Greek gerusia. These dedications point to a large production of ivory diptychs, though we do not know how many of the senators were the actual re­ cipients of such sumptuous gifts. Mamertinus, quoted by Chastagnol, remarks “in consulatu honos sine labore susdpitur”, but that honour could be very expensive.30 The identity of the donors and the destinataries of the church dip­ tychs is not touched upon. Religious diptychs normally bore the names of the local saints and bishops to be commemorated during the mass.31 They could be made of two plaques or of five, on the model of some imperial diptychs. A splendid ivory diptych of which only the two cen­ tral plaques remain, both in Berlin, shows the remains of a monogram rightly interpreted as that of Archbishop Maximian of Ravenna. The inner faces of the same tablets bear traces of a sixth-century inscription in which the names of the saints of Ravenna have been recognized.32 29 For a general survey, see: F. V. Volbach, Elfenbeinarbeiten der Spätantike und desfrühen Mittelalters (Mainz, 1952). 30 A. Chastagnol, L’ évolution politique, sociale et économique du monde romain, 284-363 (third edition, Paris, 1993), p. 219. 31 See F.W. Volbach, Reallexikon zur Deutschen Kunstgeschichte IV (Stuttgart, 1958), cols. 50-60; O. Stegmüller, ReallexikonJur Antike und Christentum, Sachwörterbuch zw Auseinanderset­ zung des Christentums mit der antiken Welt III (1957) cols. 1138-49. 32 A. Arnulf, “Die Namenliste des Berliner Christus-Maria Diptychon”, JahrbuchJtir Antike und Christentum XXXVI (1993), pp. 134-39. The Ravenna diptych provided the

THE PRODUCTION AND DISTRIBUTION OF BOOKS IN ANTIQUITY

49

In their turn the images carved on the external faces of the same dip­ tych are stylistically very close to those carved on the ivory throne ded­ icated by the same archbishop. Possibly the throne was not intended as a seat, but for the exhibition of the gospels during a synod, or pos­ sibly even a council, according to the ambitious plan of both Justinian and Maximian.33 The transition from roll to codex has been widely explored. Follow­ ing the pioneer work of C. H. Roberts and the assessment of E. G. Turner, who rightly compared the invention of the codex to the most creative efforts of mankind,34 we are well aware of the role of the Christian religion in the diffusion of this new form of script. However, it would be a mistake to consider this evolution only in the terms of book making and not in terms of the general practice of scripture. Christians were keen for the transmission of the word and memorizing and knowing by heart was considered more important than the ability to read the written word. Two frescoes in Pompeii introduce the onlooker to the variety of writing instruments in two wealthy Roman houses. The first fresco rep­ resents a still life.35 However, contrary to European paintings of the

model for the distinctive five-part diptych of the late sixth-century used in the ninth-cen­ tury binding of the Gospels of Saint-Lupicin, Jura (Paris, Bibliothèque Nationale, ms. lat. 9384; D. Gaborit Chopin, Byzance, V art byzantin dans les collections publiquesfrançaises. Lou­ vre: exhibition catalogue (Paris, 1992-93), n. 27, pp. 74-77). The influence of the reliefs of Maximian’s throne is evident from a fragmentary diptych from Marseille (Byzance, l'art byzantin dans les collections publiques françaises, n. 28, pp. 77-78); and in the scenes on the sides of the main panel in the gospels from Saint-Lupicin. An ivory plaque of unknown provenance now in Lyon is very close to the throne both in style and iconography {Byz­ ance, l'art byzantin dans les collections publiques françaises, n. 25, p. 72). Precisely how these models came to exert such a wide-ranging influence was unclear until a fragmentary ivo­ ry pyxis was discovered at Medinah El Fayoum (now Ann Arbor, Kesley Museum), close to the centre of the ivory craft and market, namely Alexandria. See A. Cutler, The Craft of Ivory. Source, Techniques and Uses in the Mediterranean World A.D. 200-1400 (Washington, 1985). 33 R. Farioli Campati, Splendori di Bisanzio (Exhibition catalogue, Ravenna 1990) (Mi­ lan, 1990), p. 253, and colour plates. Litde attention has been paid to the similarity of style between the ornamentili carving of the throne with the Alexandrian tradition and the carving of the roof trusses in Justinian’s basilica on Mount Sinai. 34 C. H. Roberts, “The Codex”, Proceedings of the British Academy 40 (1955), pp. 169204; and more recently: G. E. Turner, The Typology of the Early Codex (Philadelphia, 1977). 35 W. Helbig, Wandgemälde der vom Vesuv verschütteten Städte Cämpaniens (Leipzig, 1868); Corpus Inscriptionum Latinarum, ed. T. Mommsen, (Berlin, 1883), 4, 1174 (p. 7*); W. K. Princhett, “The Attic Stelai”, Hesperia 25 (1956), pp. 178-327, at pp. 220 and fT.; E. G. Turner, Greek Manuscripts of the Ancient World (Oxford, 1971), pi. 9.

50

CARLO BERTELLI

same subject from the seventeenth century, conceived to celebrate the vanity of ephemeral things, this painting was designed to comment on the high standing of the proprietors of the house, as only the owners of huge estates would have required the complex system of administra­ tion indicated by the items scattered on the desk. Together with a purse of money, there is an open roll, laying on the table upside down, a wooden poliptych composed of six writing tables, and a plaque (tabula ansata) on which some notes have been traced. A label is attached to the roll, hinting that it comes from an archive in which it shall be cor­ rectly replaced after use. The other fresco introduces a different symbolism of writing.36 Here a girl and a young man are represented. He is holding a roll which, as its label proves, has been taken from the shelves of his library. The girl touches her lower lip with the stylus, with an absent look as if she were searching for inspiration. In her left hand she holds a poliptych of writ­ ing tablets composed of four elements; eight sheets in all. She is at­ tempting something creative, and indicating to the viewer that what she intends to write will be very different from the routine work exhib­ ited in the previous fresco. The couple are not seeking to demonstrate their wealth, but affirm their membership in the class of the best, and cultivated, people. As we can see, a poliptych of wooden tablets is very similar to the quire of a codex, the only difference being that of material. The pas­ sage from poliptych to codex must have been logical to people who had the habit of writing on tablets. This transition corresponded to the new fashion, in the upper classes, of the practice of amending and ed­ iting a book directly, abandoning the old practice of entrusting this type of activity to a slave or to a paid professional. Subscriptions to literary codices of the sixth century prove the spreading of this bookish interest.37 Besides the fact that wooden poliptychs could never be as volumi­ nous as papyrus codices, there were other economic considerations. For instance, as T. C. Skeats has demonstrated, the codex offered a 36 House of Julia Felix: W. Helbig, 1726 = Corpus Inscriptionum Latinarum, iv 1775b and Suppl. 37 Some signatures are those of members of the senatorial class. Nordenfalk at­ tributes the proliferation of decoration in initials and other parts of the manuscript to this trend for ‘do-it-yourself books. However, it is also clear that such elaboration was not always intended for the private gratification of the scribe. See: C. H. Roberts, “The Codex”, Proceedings of the British Academy 40 (1955); and more recently: G. E. Turner, The Typology of the Early Codex (Philadelphia, 1977).

THE PRODUCTION AND DISTRIBUTION OF BOOKS IN ANTIQUITY

51

saving of 44% of the paper. The actual writing could be done quire by quire, and consequendy more than one scribe could be involved at the same time. On this basis Skeats calculated a saving of up to 26% in the cost of writing.38 Other factors which were in favour of the codex were not envisaged when this form of book was adopted but became apparent over time. In contrast to the column of script of the roll, the page of the codex offered all the opportunities of a window in a modern computer. As on a screen, it was possible to gather, in the same place, the text, the footnotes, the commentaries, the corrections, even the images. Quires could be numbered, signs could be inserted between two pages, the be­ ginning and the end of the text were assured by the firm hold of the binding. By the sixth century, the codex had existed for almost five hundred years and it had proved to be the ideal instrument for the col­ lection of a corpus of laws and a necessity for the orthodox and unor­ thodox churches. O f course the codex “codified” the script; nevertheless it did not make it as blocked as it was in the columns of the roll. Just as the dis­ position in quires of this new form of book may echo of the writer’s thought, or of the dictating voice on the waxed table, so the wide mar­ gins of the codex allowed corrections and notes, while initials caught the attention, numbered quires made quotations easier, and a variety in size and distribution of the illustrations made the book more attrac­ tive. In Umberto Eco’s words, we could refer to the codex as an opera aperta, open to endless revitalisation. The spread of the codex allowed book production to be envisaged on a scale never previously attempted. Since Constantine had ordered the only Christian library of his day, that in Caesaraea, to simulta­ neously produce fifty codices for the churches of Constantinople, it was clear that the new type of book allowed large-scale production with a resulting high degree of uniformity. The latter quality was of course important for the maintenance of orthodoxy in liturgical and doctrinal texts, but it was also a necessity imposed by the market. 38 T.C. Skeats, “The length of the papyrus roll and the cost advantage of the Co­ dex”, ^eitschrißßtr Papyrologie und Epigraphik, XLV. 1 (1982), pp. 169-75. Scribes’ rates of pay were fixed by Diocletian {‘de pretiis rerum venatium, col. VII, 39-41): thus, for 100 lines of the very best writing, a scribe could earn twenty-five denarii; for 100 lines of writing of the next best quality, 20 denarii; for the writing of a petition or legal document, a notary was paid ten denarii per 100 lines. Egyptian monasteries generated extra revenue by copying manuscripts. See: P. Ladeuze, Etude sur le cénobitisme pakhomien pendant le IVe siècle et la première moitié du Ve (Louvain-Paris, 1898), pp. 322-23.

52

CARLO BERTELLI

An edition of fifty items was exceptional, but it was also possible in a more limited number to achieve a standard production and also cut the costs by simultaneously utilising more than one scribe, as well as binders and possibly illuminators. A door had been opened to the book market. In classical times the publication of a book had followed a rather exclusive pattern. A new work could be offered by the author to some friend or sponsor, or it could be publicly read, or in some cases depos­ ited in a public library. Thereafter, whoever wished to become the owner of a copy had to hire a scribe, give orders to a literate slave, or copy himself from a borrowed example. This system led to each sam­ ple having a highly individual nature, to the point than when Pliny la­ mented the negligence of copyists who had departed from the exacti­ tude of the prototype of the illustrated treatise by Crateuas he suggest­ ed that the new copies should be amended by going back to a specia­ lised medical garden for a new start. The prescription was against the formation of a transmission from model to copies; in fact each copy would have been an independent original work.39 Prior to the existence of efficient copyists, public libraries had ful­ filled the need to go back to the prototype, providing editors with a number of examples by which to control their new edition. In 534, Securus Melior Felix, a school master who practiced his trade on the Forum Traiani, amended Martianus Capella’s Nuptiae Mercurii et Philologiae with the help of his disciple Deuterius but unfor­ tunately checking his work, as he states, against very defective exam­ ples.40 The proximity of his working place to the famous Bibliotheca Ulpia had been of little help in this case. Already at the time when the Historia Augusta was composed, many books from that library had been transferred to the baths of Diocletian.41 The contents of the Ulpian li­ brary had become legendary. One of the authors of the Historia, writing under the pseudonym of Flavius Volpiscus Syracusanus, mentions an ivory book, of which he even gives the class-number in the library’s shelves.

39 Pliny, Naturali* Historia, XXV. 4 40 J. Préaux, “Securus Melior Felix, 1’ ultime ‘orator urbis Romae’ ”, Corona gratiarum, Mélanges offerts à E. Dekkers (volume 2) (Bruges, 1975), pp. 101-121. +1 Auctores Historiae Augustas, ed. E. Hohl, II (Leipzig, 1927), Vita Probi, pp. 202-21. Cassiodorus, Institutiones, ed. Mynors (Oxford, 1937), p. 3.

THE PRODUCTION AND DISTRIBUTION OF BOOKS IN ANTIQUITY

53

In the fourth century, the catalogue of the regions gives the number of twenty-eight libraries in Rome.42 Many were still extant in the sixth century before the Gothic wars, although in what condition we do not know. Almost all the public libraries of Rome were attached to a tem­ ple and they must have suffered from the measures taken by Theodo­ sius against pagan worship after the defeat of Eugenius. A library is a collection of fragile material, particularly if the majority of its books are made of papyrus and therefore require constant care. We may pre­ sume that the closing of the temples was one of the causes of the loss of many texts by classic authors. Attacks against the properties of temples quickly followed, as dem­ onstrated by the marvellous collections of old columns and capitals in the fifth-century basilicas of Rome and in front of San Lorenzo in Mi­ lan. Deliberate smashing of pagan monuments also affected the impe­ rial fisc, as proved by the destruction of the statues in the cave of Sperlonga brought forth by the monks who in 511 had installed themselves in the former villa of the Emperor Tiberius. It has been suggested that the collections of books in the villas of Campania were transferred to the newly founded monasteries.43 There is no evidence for this supposition, nor it seems was there any intention to keep an eye on things reputedly pagan and at least distracting. Maintaining a library is expensive and there were not the conditions for imposing an extra expenditure on a monastery. It is certain that the long campaign of Justinian against the Goths inflicted a terrible blow to all the cultural structures of the ancient capital and of many municipal towns. We have a very authoritative witness for this disaster in Cassiodorus. In his introduction to the Institutiones, Cassiodorus tells how in the same year 534 when Securus Melior Felix was editing Martianus Capella, he had tried with the help of Pope Agapitus to build a library and a school for Christian studies in Rome, on the model of that of Nisibis. This would be a new institution which would rival the public teaching of secular literature. However, due to the wars and

42 A. Nordh, ed., Libellus de regionibus urbis Romae (Skriftcr Svenska Institutet i Rom, III, 80) (Rome, 1949). Also, C. Wendel, “Bibliothek”, Reallexikonfid Antike und Christentum, Sachwörterbuch zur Auseinandersetzung des Christentums mit der antiken Welt 1, (1954), pp. 24446; K. Wessel, “Bibliothek”, ReaUexikon zur byzntinischen Kunst, eds. K. Wessel and M. Resdé (Stuttgart, 1966), 1, pp. 612-15. 43 M. Cagiano de Azevedo, IX Settimana di Studi sull’alto medioevo, Spoleto 1951 (Spoleto, 1952), pp. 449ÍT.

54

CARLO BERTELO

troubles which “had devastated the kingdom of Italy”, the project fell through.44 Pope Gregory the Great organised a library attached to the Lateran palace. Its setting was splendid, and its walls were decorated with fres­ coes among which a superb portrait of Saint Augustine, bowed on his writing desk as an ancient author, survives.45 But Rome had become more a dispensary of its stock of books than a producer. The suggestion that the celebrated fragment of the Gospels of Saint Augustine of Can­ terbury were produced in Rome, and not in South Italy, seems to be still an open question.46 Parchment remained expensive. Gregory of Tours recounts that Chilperic, king of Neustria, had sent instructions to all the cities of his kingdom, saying that the letters which he had added to the Latin al­ phabet should be taught to boys in school, and that within older books, the old characters should be erased with pumice-stone and the new ones inserted. Gregory, who compared the king to Nero and Herod may not be a very objective reporter, but in any case it is interesting to notice that he did not accuse the king of getting rid of the old books and.substituting them with new ones, but attributed to him a more eco­ nomic approach, one which would seem justified to his readers, while the making of entirely new books made of virgin material would have looked very extravagant.47 The Ostrogothic king in the Italian peninsula, Theoderic, did not write theological treatises in the same way as Chilperic, nor attempted to reform the alphabet. However, he engaged in a strong cultural pro­ gram, in which books played a major role. We may presume that the change of the capital from Milan to Ravenna had provoked the transfer of books from one town to the oth­ er. However, in Milan the senators were more commuters than resi­ dents and if there was a cultivated layer of citizens, and therefore of books, it must have been rather thin. Milan had a school of rhetoric and obviously at least one public library, but nonetheless, Augustine

44 Cassiodorus, Institutiones, ed. Mynors (Oxford, 1937), p. 3. 45 Ph. Lauer, “Les fouilles du Sancta Sanctorum au Latran”, Mélanges d’ archéologie et d’ histoire 20 (1900), pp. 251-87. 46 F. Wormald, The Miniatures in the Gospels of St Augustine, Corpus Christi College MS 286 (Cambridge, 1954); Wormald proposes a south-Italian provenance. 47 Gregory of Tours, Historia Francorum V. 44, 46, ed. B. Krusch (Hannover, 1885), Monumenta Germaniae Histórica, Scriptores Rerum Merovingicarum, I, pp. 253-54.

THE PRODUCTION AND DISTRIBUTION OF BOOKS IN ANTIQUITY

55

found retirement in a villa of Casciacum more convenient than the us­ age of the private or public libraries of Milan.48 The reduction in the number of public libraries, which were trequendy susceptible to fires and other calamities causing damage which would have been too expensive to repair, coincided with the opportu­ nities offered by an editorial system which made possible a standard level of production and an increase in the quantity of volumes issued at a time. At the same time that there was a new demand for books, there was a capital without public libraries, and a new ruling class of German extraction without experience in the bookish world. The Gothic court provided employment for intellectuals, who were engaged in an enormous work of translating from Greek to Latin and of in­ structing the Goths in the traditional disciplines of Roman govern­ ment. As well as summoning Boethius and Cassiodorus to his court, Theoderic supported the national church of the Goths, not only promoting prestigious buildings, but also endowing the cathedral with splendid purple codices of the Gospels, in which the old Latin translation which preceded the Vulgate was adopted, perhaps as a means of distinguish­ ing the Arian Gothic priests from the Roman Catholics.49 The extant codices prove Theoderic’s intention of equaling Constantine’s enter­ prises in the promotion of sacred purple books. No other western centre witnessed such an intense activity in book production and in the book market as Ravenna at the time of the Ostrogothic kings. It is under Theoderic that we see the first signature of a book-merchant and publisher. His name was Viliaric, a man known in contemporary records as bokareis, the Gothic for librarius. In his sta­ tion was confectus the codex of Paulus Orosius which is now one of the jewels of the Laurenziana library in Florence.50 48 See: H. Chadwick, Augustine (London, 1986), p. 34. 49 O. Tjaeder, “Der Codex Argenteus in Uppsala und der Buchmeister Viliaric in Ravenna”, Studia gotica. Die eisenzeitlichm Verbindungen zwischen Schweden und Südosteuropa. Vor­ träge beim Gotensymposion in Statens Historika Museum Stockholm 1970, ed. U. E. Hagberg (Stockholm, 1972), pp. 144-64. 50 According to Tjaeder, the subscriptions with Gothic names in the pap. Tjaeder 34vd, may represent other members of the same stationary office of Viliaric. See also C. Nordenfalk, Die spätantiken ^ierbuchstaben (Stockholm, 1970), pp. 100-101; Nordenfalk re­ fers to Roman or Romanized stationarii: Gaudiosus, who had a shop by San Pietro in Vincoli in Rome (see also: D. de Bruyne, “Gaudiosus, un vieux libraire romain”, Revue Bénédictine 30 (1913), pp. 343-45); Ursicinüs, a layman who in 517 described himself‘¿erf. ecclesiae veronensis’ of the Biblioteca Capitolare, Verona (cod. xxxviii 836) (Codices latini antiquores IV. 494).

56

CARLO BERTELLI

The choice of author reflects a wise publishing policy. Paulus Orosius was the historian who had attributed the barbaric invasions to the de­ sign of Divine Providence and no other text would have been more appealing to a Gothic public able to read Latin. Other books published in Ravenna in the same period, mirror the anxiety of the king in pre­ serving the Roman knowledge in the application of boundary-laws. A letter by Cassiodorus, written in the name of king Theoderic, praises the profession of the surveyor to whom a collection of texts, the Corpus Agrimensorum, was dedicated.51 At the same time, while lawyers were engaged in the keeping of Ro­ man laws, other members of the best society could appease their curi­ osity in medical matter with the illustrated edition of Sextus Placitus, a collection of medical oddities with a mixture of magic formulae, leg­ ends and anecdotes. It must have figured as a “popular” substitute to the learned treatise of Dioscurides, of which a sumptuous edition had been offered in Constantinople in 512 to the princess Anicia Juliana.52 The books mentioned above are not manuals for ordinary use. They are all splendidly decorated. The Paulus Orosius is known as one of the first cases of ornamented initials in a western manuscript,53 while the first of the two texts of the Agrimensores preserved in Wolfenbuettel, both bound together in a single volume, is a masterpiece of monumental script from Ravenna and is introduced by a solemn frontispiece with the outstanding image of a naked philosopher. The medical treatises collected under the name of Sextus Placitus are known from two cod­ ices made in Sicily in the late twelfth century, both after the same mod­ el, and they betray in their illustrations the attempt at interpreting the

51 H. Butzmann, Corpus Agnmenosrum Romanorum. Codex Arcerianus A der Herzog-AugustBibliothek zu Wolfenbuettel (cod. Guelf. 36.23 A) (Lugduni Batavorum, 1970); The sixth-cen­ tury Ravennate model of a Carolingian manuscript in the Vatican library has been iden­ tified by F. Muetherich, “Der Karolingische Agrimensoren Codex in Rom”, Aachener Kunstblätter 45 (1974), pp. 59-74. On the royal books produced in Ravenna: H. Butz­ mann, Corpus, pp. 21-22; O. von Friesen and A. Grape, Codex Argenteus Uppsalensis phototypice editus (Uppsala, 1928); G. Cavallo, “La cultura a Ravenna tra Corte e Chiesa”, Le sedi della cultura nell’ Emilia Romagna. L’alto Medioevo, ed. V. Fumagalli (Milan, 1983), pp. 29-51. On the royal books produced in Ravenna: H. Butzmann, Corpus, pp. 21-22; O. von Friesen and A. Grape, Codex Argenteus Uppsalensis phototypice editus (Uppsala, 1928); G. Cavallo, “La cultura a Ravenna tra Corte e Chiesa”, Le sedi della cultura nell’ Emilia Ro­ magna. L'alto Medioevo, ed. V. Fumagalli (Milan, 1983), pp. 29-51. 52 Florence, Biblioteca Laurenziana, cod.65.1: Nordenfalk, fwrbuchstaben, pp. 99-100. 53 See for instance the misrepresented Mercurius, who, in the image of the “herba ‘immolwn’ "should appear as the discoverer of this herb, following ‘Odyssaea’ 10, pp. 302306: H. Biedermann, Medicina magica. Metaphysische Heilmethoden in spätantiken und mittelalter­ lichen Handschriften (Graz, 1972), pl. 20 (from Vienna, NB, cod. med. 93, fol. 61v).

THE PRODUCTION AND DISTRIBUTION OF BOOKS IN ANTIQUITY

57

iconography of ancient divinities of whom the illuminator had lost any recognition.54 If in the case of the Agrimensores a royal commission is apparent, in the other examples a private initiative can be safely envisaged. The handsome Paulus Orosius was produced to be sold and the rich medical manuscripts are of too little practical use to attribute them to a state program of encouragement of the sciences. All these codices show a strong interest in luxurious books. The same inclination towards prestigious editions is apparent in the Holy Lands. We notice a remarkable uniformity in such stately purple codices as those of Genesis or the Gospels, which were produced in Palestine at the end of the fifth century and in the first decades of the sixth.55 Written in Greek in golden letters, with a marked pre-emi­ nence of the illustration over the script, they were not made for the local readers, but rather destined for the rich customers who visited a region which had never before had a reputation for being a centre of book production, nor of being particularly rich, prior to the invasion of pilgrims in late Antiquity. As a late copy of the legends ofjoseph has revealed under the expert scrutiny of Otto Pächt, the models and the painters came from two nearby regions, namely from Alexandria and from Syria.56 Alexandria, the home of a huge Jewish colony, the town where the Bible had been put into Greek, the great center of papyrus and book trade, must have developed a stock of illustrations of Jewish legends at a very early stage. It was in Alexandria that between the last years of the fifth century and the beginning of the sixth, an illustrated book of Genesis was produced which offered the most influential iconographie cycle of the entire west­ ern middle ages.57 We can now be certain, after the inquiries conduct­ ed by Kurt Weitzmann, that in the thirteenth century the volume, now known as the Cotton Genesis, was in Venice,58 but other copies must have travelled extensively and long before.

54 H. Gerstinger,

D io sk u rid ie s. C odex V in dobon en sis m ed. g r. 1 der O esterreich isch en N a tio n a l­

b ib lio th ek . K o m m en ta rb a n d z u d er F a k sim iliea u g a b e

(Codices phototypice impressi, 12) (Graz,

i " 2). 55 G. Cavallo

e t a i , C odex p u rp u reu s R o ssa n en sis , V ollstän dige F a k sim ile-A u sg a b e im O rig in a l­

(Graz-Rome, 1985). 56 J. and O. Pächt, “An unknown Cycle of Illustrations of the Life ofjoseph”, C a h iers arch éologiqu es 7 (1954), pp. 39-4i). 57 K. Weitzmann and L. Kessler, T h e C otton G en esis (Princeton, li)86). 58 Weitzmann and Kessler, T h e C o tto n G en esis.

fo r m a t d er H a n d sc h rift

58

CARLO BERTELLI

The era of the secular book was over. While precious book covers, like that made of gold which Pope Gregory the Great sent to Queen Theodelind, or those in silver, found in Anatolia and now in the col­ lection of Dumbarton Oaks, demonstrate highly significant investments in liturgical books, literary books for individual reading became in­ creasingly rare. The passage from roll to codex would have required new classements in the old libraries. Few librarians had been willing or able to change their shelves and most had remained faithful to the roll until it was too late. Consequendy, many texts which were never transferred from roll to codex were lost. If papyrus was costly, parchment had always been expensive. We do not need to go to Chilperic’s Gaul to find pages washed and erased to make way for a new text. A case is provided by the codex Palat. lat. 24, in the Vatican Library, which had been writ­ ten in the third century and, after washing and erasing, was written again in the sixth to seventh century.59 Christians, who were the most enthusiastic users of codices, for a long time did not have their own libraries. The only exceptions were for a long time the library founded in Caesaraea by Orígenes, who was educated in Alexandria, and later the one in Nisibis which was taken as a model by Cassiodorus. It was at Caesaraea that Constantine had commissioned his fifty volumes mentioned above. In 529, the council of Vaison, presided by Caesarius from Arles, es­ tablished that, on the model already practised in Italy, every rural par­ ish had to teach the Psalms and the Holy Scripture to the faithful, thereby preparing the unmarried literate youth for recruitment to the priesthood.60 It is unlikely, however, that those courses were supported by libraries of any size. The Psalms were learnt by heart and we do not know whether individual parishes owned a complete edition of all the four gospels, or, if so, in what translation. Caesarius informs us in the same passage that the situation in Italy was better as far as religious education was concerned, though a parish school could be equipped with just a Psalter. It is also interesting that the church took the role of a grammar school, although at a low level of literacy.

59 J. Irigoin, “Les textes grecs circulant dans le Nord d’ Italie au Ve et Vie siècles. Attestations littéraires et témoignages paléographiques”, Teoderico e i Goti tra Oriente e Oc­ cidente, ed. A. Carile (Ravenna, 1995), pp. 390-400. 60 Council of Vaison, c. 1: ed. C. De Clercq, Les Canons des Corniles Mérovingienes, 6e7e siècles (Paris, 1989) 148 A, p. 78.

THE PRODUCTION AND DISTRIBUTION OF BOOKS IN ANTIQUITY

59

Before the end of the sixth century, the Gothic wars, and subsequently the Lombard invasion, were responsible for the fate of the whole Ital­ ian book heritage, including public and private libraries, chapter hous­ es and monasteries. Even Monte Cassino fell a victim of the last bar­ baric invasion. However, prior to this, even in better times, both the church and the Goths had been highly selective in their choice of texts worty of being transmitted. In terms of the intellectual climate in the Gothic kingdom, Ennodius gives us a number of names of the rhetoricians and grammarians active in Theoderic’s reign.61 They had little to do with the old paideia, for grammar and rhetoric were disciplines considered necessary to the art of government. The king loved to think of himself “as a philosopher under a purple mande”,62 but the choice of Greek texts transcribed or translated in Ravenna has its emphasis on the practical.63 The general attitude of the Theoderic’s cultural policy suggests that it is unlikely that the codex known as the Vtrgilitis Romanus (Vadcanus Latinus 3867) was made during his reign, though its bald letters point to Ravenna and to the sixth century.64 An illustration such as that of the banquet given by Queen Dido to Aeneas and his companions, in which the Trojans appear as distincdy Mediterranean, both in their at­ tire and look, while the waiters are blond and long-haired, makes it rather hard to suppose that such a monumental codex would have pleased a Gothic customer. The text was never corrected and probably not even read, until the codex reached the abbey of Saint Denis in the ninth century. It was possibly a show piece, a monumental tribute to the Latín national poet, and it has been suggested that its aim was to celebrate, in the palace of Ravenna, the restitution of the capital to the Empire and the reunification of the Trojans in Constantinople with the Trojans in Rome.65 It is interesting to note how this volume was produced. While in the case of the Bucolics the ruling traced with a point on the parchment 61 In Paraenensis Didascalia, Monumenta Germamae Histórica, Auctores Antiquissimi, VII, p. 310. 62 Cassiodorus, Variae, LX, 24. 63 Irigoin, “Les textes grecs circulant dans le Nord d’ Italie au Ve et Vie siècles. At­ testations littéraires et témoignages paléographiques”, p. 396. 64 G. Cavallo,, “Libri e continuità della cultura in età barbarica”, Magistra Borboritas, eds. G. Pugliese Carratelli et al (Milan, 1984), pp. 603-62. 65 C. Bertelli, “Le illustrazioni del Virgilio Romano nel contesto storico e artistico”, Vergilius Romanas, volume a commento dell’ edizione facsimile del cod. Vat. lat. 3867 (Codices e Vaticanis selecti, LXVI), ed. I. Lana (Milan, 1986), pp. 171-215, at p. 189.

60

CARLO BERTELLI

helped the painter in his elementary compositions, for the other more ambitions illustrations, the solution was to provide the painters with separated bifolios, so that they could work at the same time as the scribes66. To facilitate his task, the painter, who used a repertory of stock fig­ ures, employed a grid based on the golden section, as had been utilised for the Theoderician mosaics in Saint Apollinare Nuovo.67 If the codex belonged to the palace, it would be a rare case of a secular book kept in the treasury just as the most precious evangeliaries were in the churches. It is a tribute to the authority of poetry, but also a testimony, in its ingenious and simple construction and in its un­ checked mistakes, of an awe which is far removed from the critical at­ titude of the classic age. The bridge between the ancient learning and the new basic demand of culture was erected successfully by Cassiodorus in the monastery which he founded at Squillace, one of his properties in Calabria. The dispersal of the monastery’s library after his death contributed for cen­ turies to the expansion of knowledge and to the production of new books in many western centres, from Iona to Bamberg, Nonantola and Monte Cassino.

66 F.W. Deichmann, Ravenna, Hauptstadt des Abendlandes, Wiesbaden 1968, Kom­ mentare, 1, pp. 182 and ff., fig. 121 and ff. 67 A. Momigliano, Cassiodorus and Italian Culture of his Time, Proceedings of the Brit­ ish Academy 41 (1955), pp. 207-245; T.D. Kendrick et al, Evangeliorum qrnttor Codec Lindisfamensis, (Oltun and Lausanne, 1960), pp. 47-60; R.L.S. Bruce-Mitford, ibidem, pp. 142157; G. Cavallo, Ricerche sulk, maiscola bilica (Florence, 1966), pp. 110-112.

THE MIGRATION PERIOD: MODEL HISTORY AND TREASURE Klavs Randsborg

urging them to make sudden raids into the Em peror’s territory, so that they too m ight be in a position to exact a price for p eace...’ Procopius, The Anecdota}

1. Basics on treasure Late-Antique/Migration-period Europe (c. AD 400-600/(350-650)), when studied through the written sources of the age, centres around a triad of politics, alliances, and warfare, which tend to be concerned with rights and wealth. This includes treasure or portable wealth. Treasure is manifested in the form of hoarded silver and gold, some­ times as coin, usually Roman siliquae and solidi. Indeed, the hoards, which, for example, reflect past political alliances, and the material ev­ idence of warfare, such as the weaponry from sacrifices and burials, supply the archaeological parallels for those aspects of study which have been established on the basis of the written sources. The afore­ mentioned parameters of behaviour were highly important to ancient society, which was dominated by concerns over politics, alliances, war­ fare, wealth, and religion. Ancient society was also concerned to invest in these matters materially (as well as emphasize them in many other ways). However, while it might be said that the written sources of the age tend to focus on issues such as honour and beliefs, they are virtu­ ally mute about values and, in particular, the needs of daily life. Settlement-archaeology and material culture relating to the charac­ ter and intensity of agriculture12 as well as the social and economic di1 Procopius, T h e A n ecd o ta , ed. H.B. Dewing (London, 1935), xi. 6. 2 See, for example, K. Randsborg, T h e F irs t M ille n n iu m A .D . in E u ro p e a n d the M e d ite r­ ran ean . A n A rch a eo lo g ica l E ssa y (Cambridge, 1991), p. 29ff; also for much of the following, E. Bourgeois, “Evolution du peuplement humain dans la haute valéc du Doubs à partir des donées polliniques”, L 'A g e du F er d a n s le J u r a . A c te s du 1 5 e colloqu e de lA ss o c ia tio n fr a n ç a ise p o r l'étu d e d e l'Â g e du F er, eds. G. Kaenel et P. Curdy (Pontalier and Yverdon-lesbains,1992), pp. 15-20.

62

KLAVS RANDSBORG

mensions are seldom described in the written sources. This is especially the case in the Germanic world and other non-Roman realms of the Migration period. Only the Late Roman sources voice the concerns of the state in terms of taxation, revenue, and landed interests, throwing light on the economics, if only in an indirect way. The level of production in Late Roman society was no doubt high, but, on the whole, regionally orientated. Archaeology nevertheless in­ dicates some international movement of goods, of both a utilitarian and prestige nature, a process which had occurred since the Bronze Age and before, involving the transportation of luxuries across enor­ mous distances in Europe. It is important to note, however, that the flow of objects and ideas was not uniform, nor only conditioned by the interests of the culturally stronger, usually southern, part of Europe. The centuries of Classical Antiquity display many examples of rejections of Mediterranean norms on the part of, for example, the Celtic and, in particular, the Germanic regions of the Continent. The rise of particular barbarian art-styles is a strong testimony to this phenomenon, which is also indicated by the nature of Roman imports, which in the Germanic parts of Europe were almost exclusively artefact types fitting readily into the local cul­ ture. Furthermore, in contrast to earlier, prehistoric, periods, such lux­ uries were only rarely imitated. In the main, the substantial variation in patterns of emission of var­ ious types of coins in the (later) Roman Empire, evident from the finds, probably reflects different ways of paying administrators, soldiers, pub­ lic debt, etc. (rather than direct economic concerns per se).3 In the “ex­ portation” of coins, other factors were also relevant. Smaller denomi­ nations of gold coins were common among the Franks, but practically unknown, for example, in Scandinavia. Roman silver coins, by con­ trast, also of lower value, were quite widespread. Such observations, in particular if based on smaller samples of coins, make the archaeological picture highly complicated and difficult to disentangle. In addition, large hoards (and treasuries) often contain long and highly varying “tails” of earlier coin emissions, from which a small sample (for a mod­ est find) might equally be drawn, thereby producing a different pattern of stock and/or distribution.

3 Cf. Randsborg, The First Millennium AT), in Europe and the Mediterranean. An Archaeological Essay, p. 132.

THE MIGRATION PERIOD MODEL HISTORY AND TREASURE

63

Detailed, but often somewhat confusing and “over-statistical” studies, especially of the Scandinavian finds, utilise data relating to mint, socalled die-links, weight, wear, hoard context, etc. of the Roman solidi as well as related aspects of the wide range of other gold items within archaeological hoards.4 The surprisingly common die-links of the solidi found in Scandinavia (of various areas, and periods) correlate in num­ ber with the amount of coin in the various areas/periods. Thus, there is no particular region of import with a pronounced representation of die-links.5 The die-links would appear to indicate some direct contact between Roman leadership and high-ranking Germans. The Germans primarily valued Roman coins for their metal con­ tent; as a result these coins had a long life. It is a mistake, therefore, to use the Roman coins as economic indicators of long-distance trade, as one might with the thousands of Arabic dirhems of the Viking Age. High quality silver denarii, mainly of the second century AD, were still in circulation in large numbers in the late fifth century, to judge from their occurence in hoards and graves (such as that of the Frankish King Childeric, who died in AD 482), together with contemporary solidi.6 The changing distributions of solidi in regions beyond the Empire, therefore most probably reflect the changing politics and military rela­ tions between the Empire and the barbarian kingdoms of Europe, as well as relations within these kingdoms.7 Warfare was doubtless the prime generator of royal/tribal and derivate treasure in the Migration period, perhaps even for the Romans. Treasure seemingly flowed between centres, such as the Byzantine court at Ravenna on the one hand or, on the other, Gudme in Den­ mark, the most powerful political centre of fourth- and fifth-century Scandinavia. To some extent this wealth was retained at the centres, although eventually often used for various investments (such as lavish churches, military expeditions, long-distance exchanges of luxuries, princely gifts and feasting, and various other enterprises). Thus, trea­ sure was redistributed. Archaeologically, it has been recorded even from the “parish”-level of society, notably in the form of grave-goods. This is no great surprise, since the prime source of wealth (as well as 4 c.g., O. Kyhlbcrg, “Late Roman and Byzantine solidi: an archaeological analysis of coins and hoards”, E x ca va tio n s a t H e lg ö 10 (Stockholm, 1986), pp. 13-26. 5 Data from J.M. Fagerlie, L a te R o m a n a n d B y za n tin e S o lid i F in d s in S w eden a n d D en m a rk (New York, 1967). 6 Cf. H.J. Eggers, E in fü h ru n g in d ie V orgeschicte (Munich, 1959), p. 173. 7 Data from S. Mosser, A B ib lio g ra p h y o f B y za n tin e C oin H o a rd s (New York, 1935).

64

KLAVS RANDSBORG

of subsistence) remained the land, cultivated or otherwise, with its do­ mesticates and other biological as well as mineral resources. In short, treasure was a means both of creating energy (in terms of resources and items), and of converting these from one form to another. In this short essay, I wish to examine a number of themes relating to hoards and treasure, showing how these have some bearing upon contemporary interpretations of the celebrated Pirenne thesis.

2. Modelsfor the history of hoarding The Roman policy towards the Germanic and other advanced barbar­ ian interfaces - the Europe of Kingdoms - was neither uniform nor stable. In the third/fourth centuries AD, centre-periphery relations, re­ inforced by social and economic links, were negotiated between the Empire, which was then in the stronger position, and favoured Ger­ man partners. Inter-Germanic exchange, which was equally important, was widespread, organized and intensive, although structured into lev­ els like luxuries, materials, and foodstuffs. In the late fourth century, the relations on the northern frontier shifted towards a “War (or mere threat of war)/Treasure (or wealth) syndrome”, unrelated to aquisition by commerce. The clearest exam­ ples of this are found in the important Middle Danubian basin, where the occurrence of local gold treasures exactly mirrors the periods of po­ litical unrest with the Empire.8 In the crucial sixth century, however, (after the crises of the fifth century and the transformation of the West­ ern Roman Empire), the relationships between the Eastern Empire and traditionally friendly groups - such as the Franks - may have been governed by treaties, providing regular payments, or “gifts” in gold to the Franks, albeit in relatively modest amounts. Similarly, both the Early and the Late Imperial periods (0-400 AD) display a correlation between the appearance, and possibly even the quantity, of Roman “exports”/Germanic “imports” (mobile high status treasure and wealth in the form of precious coins, drinking glasses, bronze vessels, textiles, weapons, etc.) and peaceful political relations across the frontiers (Limes). The first wave of Roman solidi arrived in Scandinavia, for ex­

8 A. Kiss, “Die Goldfunde des Karpatenbeckens vom 5. - 10. Jahrhundert: Angaben zu den Vergleichsmöglichkeiten der schrifdichen und archäologischen Quellen”, Acta Archaeologica Academiae Scientiamm Hungancae 38 (1986), pp. 105-145.

THE MIGRATION PERIOD MODEL HISTORY AND TREASURE

65

ample at Gudme on Funen, and in Saxony. There was also much Scandinavian contact with south-eastern Europe (and the Visigoths) outside the Empire, traceable, for instance, in the significant interna­ tional distribution of drinking glasses. The mid-fifth century saw the second and major wave of solidi in the North, often die-linked, and which, archaeologically, in particular, are known from the eastern margins of Scandinavia (where these coins, in addition to Nordic jew­ ellery, were worn as trinkets). The end of the century marked the rise of the Franks (who conquered Alamannia after AD 496/506 and Visigothic Gaul in AD 507), while numbers of Norseman and Saxons ar­ rived and settled in England. The imperial solidi disappeared from north-eastern central Europe during the period between AD 350 and AD 450/500, while very few were still entering Scandinavia in the early sixth century (see below, section 3, Fig. 6). In western Europe, solidi were not uncommon, even after the reign ofJustinian. A similar process of gradual isolation (from both Imperial control and political and military communication) of the landlocked regions of trans-Alpine, non-Mediterranean Europe is not­ ed in the case of the Balkans, although the Avars and the Romanians, for example, as well as people in the Ukraine, did possess a few sev­ enth-century Byzantine solidi. During the seventh century, in the easternmost Mediterranean, and in the Near East, the Persian wars with the Roman Empire of late An­ tiquity were followed by the Islamic assault, representing a dramatic change of supreme political, military and ideological power in a social and economic region which was, in many ways, self contained and, in some respects, already separated from the west.9 Concomitantly, there were Slav (and Avar) pressures on Constantinople. The links between the Germanic kingdoms (including Gaul), and Byzantium were sev­ ered, with the consequence that gold, for example, disappeared from the West. In Byzantium, in this period, the monetary economy de­ clined (for example, the army and bureaucrats were paid in grants of land, rather than coinage), although new standards, both gold and cop­ per, were still issued. One of the consequences of this process was the eventual rise of the new silver coinages of the West (as well as in the Islamic East).

9 Cf. P. Pcntz, The Invisible Conquest. The Ontogenesis of Sixth- and Seventh- Century Syria (Copenhagen, 1992).

66

KLAVS RANDSBORG.

3. Treasure in action In this section, developing the theme described above, I wish to make a number of observations on the archaeological material of the Migra­ tion period, in particular from the Germanic realms (including Scan­ dinavia), which may serve to illuminate some of the main points: Fig. 1. (a) The dates o f Middle Danubian gold finds (the latest ones being clearly post-imperial in period), in comparison with (b) the Roman and Byz­ antine Imperial payments to barbarian princes known from historical sourc­ es. A strong correlation can be noted between the two.10

10 After Kiss, “Die Goldfunde des Karpatenbeckens vom 5.-10. Jahrhundert: An­ gaben zu den Vergleichsmöglichkeiten der schriftlichen und archäologischen Quellen”.

THE MIGRATION PERIOD MODEL HISTORY AND TREASURE

67

Fig. 1. (b)

Fig. 2. (a) The distribution of selected Late Roman (around AD 400) types o f eastern drinking glasses (mainly from barbarian, usually Germanic graves); showing the notable spread between the Black Sea and Scandinavia. This is compared with (b) the contemporaneous westward spread (on the Continent) o f Nordic and (derived) German relief brooches o f the (mid) later Migration period (around AD 500).11 The two maps illustrated in this figure, reflect the change from a general (both eastern and western) flow o f artefacts to, and from, Scandinavia before AD 450 or 500, and the exclusively western flow from the late sixth century onwards, indeed, the “westernization” o f the re­ gion concomitent to the expansion o f the Slavs in eastern Europe.

11

After U. Näsman,

G la s och h a n d el I sen rom ersk tid och fo lk u a n d rin g stid . E n

n E k e to rp -II, Ö la n d , S verige

(Uppsala, 1984).

Studie h in g g la s

68

KLAVS RANDSBORG

Fig. 2. (a)

THE MIGRATION PERIOD MODEL HISTORY AND TREASURE

Fig. 2. (b)

69

70

KLAVS RANDSBORG

Fig. 3.

Fig. 3. Hoards with Byzantine gold coins from the period between the late fifth and the mid-fourteenth century: (a) chronological distribution; (b) distri­ bution according to size, as indicated by number o f finds with, respectively, less than fifty and fifty or more coins, in (Emp) the Eastern Roman or Byz­ antine Empire (to the east of the Adriatic Sea), (N) regions to the north of the empire, (W) Mediterranean regions to the west of the heartlands o f the empire, (NW) north-western Europe, and (Scand.) Scandinavia; (c) and (d) changing frequencies (in percentages) of hoards from the above-mentioned regions (here Scandinavia is included among the finds from, respectively, north-western Europe (AD 450-550) and the regions to the north of the Byz­ antine Empire.12 The changing fortunes (and rates of emissions), as well as the progressive shrinkage o f the sphere of influence of the Eastern Empire, should be noted.

12 After Randsborg, The First Millennium A.D. in Europe and the Mediterranean. An Archaeologicai Essay; source (data) Mosser, A Bibliography of Byzantine Coin Hoards.

THE MIGRATION PERIOD MODEL HISTORY AND TREASURE

71

Fig. 4.

Fig. 4. The correlation between the numbers o f die-links and the numbers o f solidi from various parts of Scandinavia in the fifth century A D .13

13 Source data taken from Fagerlie, Late Roman and Byzantine Solidi Finds in Sweden and Denmark.

72

KIAVS RANDSBORG

Fig. 5. Fig. 5. A sample of the contents o f the grave o f the Frankish King Childeric, who died in AD 482, including the chronological distribution o f the secondcentury silver denarii and the fifth-century (mainly third quarter) solidi.14

14 After Eggers, Einfiihrung in die Vorgeschichte.

the

Mi g r a t i o n

p e r i o d m o d e l h is t o r y a n d t r e a s u r e

73

Fig. 6. Examples o f regional diagrams, showing the chronological distribu­ tion o f solidi and other coins (gold, if not otherwise indicated)15

Fig. 6. (a) All o f the “Free Germany” o f the Imperial period, showing a notably sharp decline in the import o f solidi from the late fifth century AD.

15 Sources of data for Fig. 6 are as follows: (a) S. Bolin, Fynden av romerska mynt I detjria Germanien. Studier i romersk och äldre germansk historia (Lund, 1926); (b) various data-sets, in­ cluding Fagerlie, Late Roman and Byzantine Solidi Finds in Sweden and Denmark; Kyhlberg “Late Roman and Byzantine solidi: an archaeological analysis of coins and hoards”; (c) P.C.J.A Boeles, Friesland tot de elfde eeuw. %ijn voor- en vroege geschiedenis (s’-Gravenhage, 1951); (d) W. Knapke, “Aurei und Solidivorkommen an der Südküste der Ostee”, Acta Archaeologica XII (1941), pp.79-118; idem, “Aurei und Solidivorkommen am Mare Baltikum”, Acta Archaeo­ logica XIV (1943), pp.55-65; (e) J. Werner, Münzdatierte austrasische Grabfunde, Germanische Denkmäler der Völkerwanderungzeit. Serie A, 3 (1935); (f) L Huszár, “Das Münzmaterial der Völkerwanderungzeit im Mitteieren Donaubecken”, Acta Archaeologica Academias Scientiarum Hungaricae 5 (1955), pp.61-109; (g) Knapke, “Aureiund Solidivorkommen an der Südküste der Ostee”; idem, “Aurei und Solidivorkommen am Mare Baltikum”.

74

KLAVS RANDSBORG

Fig. 6. (b) Scandinavia (Sweden & Denmark; very few coins have been found in Norway): there is a notable similarity with the general pattern shown in (a).

Fig. 6. (c) Frisia: there is a strong similarity with the pattern in the Rhineland shown below (e), including many late (sixth-century) solidi.

THE MIGRATION PERIOD MODEL HISTORY AND TREASURE

75

Fig. 6. (d) Thuringia: the relatively few very late (sixth-century AD) solidi is rem­ iniscent of the above general “Free Germany”-Scandinavia pattern, but is not at variance with the supposed, although tenuous, links with the Ostrogoths and Italy.

Fig. 6. (e) The Rhinelands: only graves are included (Ostrogothic silver coins are also shown). T he Reihengräber phenomenon (including rich personal gravegoods) is restricted to the period AD 450/500-650/700. In particular the high pro­ portion of late (sixth- to seventh-century) gold coins should be noted.

76

KLAVS RANDSBORG

Fig. 6. (f) Eastern Hungary: silver and base-metal coins are also included. The relatively common, late (sixth-century) gold coins are particularly noteworthy.

Fig. 6. (g) Ukraine (etc.): The many late (sixth-century) gold coins are particularly noteworthy. The statistics displayed in this figure are produced on the basis of the contents of hoards, both large and small, and of single coins, occasionally from graves. The diagrams should not be read as direct representations of fluctuations in exchange —therefore communication —between the Empire and its barbarian interface, ex­ cept, perhaps, for the point of the cesession of the export/im port of solidi. Inter­ tribal affairs are also relevant, as well as a num ber of archaeological factors, main­ ly involving the changing visibility of the coins in the antiquarian records.

THE MIGRATION PERIOD MODEL HISTORY AND TREASURE

77

Fig. 7. Exam ples o f the chronological distribution o f base-metal coins are given (cf. Fig. 6 (f)), for: (a) Constantinople (the church o f St Polyeuktos, Saraçhane). (b) Corinth.16

Fig. 7. (a)

16 (a) after Randsborg, The First Millennium AD. in Europe and the Mediterranean. An Ar­ chaeological Essay; data source, M. Harrison, Excavations at Saraçhane in Istanbul, 1 (Princeton, 1986); (b) after Randsborg, The First Millennium A.D. in Europe and the Mediterranean. An Ar­ chaeological Essay; data source, K.M. Edwards, Coins 1 8 9 6 - 1 9 2 9 . Corinth, 6 (Cambridge, Mas, 1933).

78

KLAVS RANDSBORG

Fig. 7. (b) Fig. 8. Frequencies of public (and private) investments in church building and major-repair are shown,17 illustrating patterns of investment both at the core of the (Eastern) Empire and in its semi-core regions (although not always within the proper boundaries of the Empire), for: (a) Constantinople. (b) Syria, with particularly numerous churches, etc. (“Monuments”) from the sixth century AD, when the region developed its strength, and economic and cultural independence from the west. (c) Various cities in Italy.

17

After Randsborg, T h e F irs t M ille n n iu m A .D . in E u ro p e a n d th e M ed iterra n ea n . A n A rch a e­ (a) data source; W. Müller-Wiener, B ild le x ik o n z u r T op o g ra p h ie Is ta n b u ls. B y z a n tio n -K o n sta n tin o p o lis b is zu m B egin n des 1 7 . J a h rh u n d erts (Tübingen, 1977); (b) data source; G. Tchalenko, V illages a n tiq u es de la S yrie du N o rd . L e m a s s if du B è lu s à la époqu e rom ain e , 1-3 (Paris, 1953-58); (c) data source; B. Ward-Perldns, F rom C la ss ic a l A n tiq u ity to th e M id d le A g es. U rb a n P u b lic B u ild in g in N o rth ern a n d C en tra l I ta ly A D 3 0 0 - 8 5 0 (Oxford, 1984). o lo g ica l E ssa y ;

THE MIGRATION PERIOD MODEL HISTORY AND TREASURE

Tig. 8. (a)

Fig. 8. (b)

79

80

KLAVS RANDSBORG

Fig. 8. (c)

Fig. 9. Examples are shown of the investment in grave-goods, indicated by Lombard graves in Italy.18 The case demonstrates wealth in action at the “parish” level and the unity of Continental Germanic burial customs, as well as a desire to signal the particular patchworks of family and wealth structures in the newly-won regions.

18 After L. Jorgensen, “Castel Trosino and Nocera Umbra. A Chronological and So­ cial Analysis of Family Burial Practices in Lombard Italy (6th-8th Cent. A.D.)”, Acta Archaeologica 62 (1991), pp.1-58; cf. H. Steuer, “Proposals on the Social Structure of the Merovingian Empire”, The Birth of Europe. Archaeology and Social Development in the First Mil­ lennium A.D., ed. K.Randsborg (Rome, 1989), pp.100-122.

THE MIGRATION PERIOD MODEL HISTORY AND TREASURE

Fig. 9.

81

82

KLAVS RANDSBORG

Therefore, in the fifth century, to judge from the archaeological records, close links were nursed between Germanic Scandinavia and both the western and the eastern parts of the Continent. A particularly strong flow of solidi (and probably other treasure as well) arrived in the second and third quarters of the fifth century, especially in the latter. This wealth no doubt included parts of cash payments (cf. the many die-linked coins) to Germanic kings and their allies and followers (even in Scandinavia), who were participating in either the assaults on the Empire (including those of the Huns), or in its defence. In the early sixth century it is possible to note: (1) A link between the Justinianic reconquest of the central and, partly, the western Mediterranean, and renewed alliances between the Eastern Empire and, e.g., the Franks, the latter being old federates of the Western Empire, and, together with the Alamanni, important sup­ pliers of officers and soldiers to the Late Roman army.19 (2) A rather fragile Ostrogothic-Thuringian-Scandinavian connec­ tion (with further additions and ramifications). Scandinavia also had close relationships with Anglo-Saxon England, Frisia, and also, no doubt, neighbouring Saxony (see Fig. 2, (b)). In addition, the Scandi­ navians dominated the Baltic. The outcome of this unstable political situation left the Franks in the dominant position in western and cen­ tral Europe (particularly after the arrival of the Lombards in Italy), and eventually served both to isolate the North from the Byzantine Empire (indicated by the nearly total absence of late solidi in Scandinavia) and to integrate it into the West (cf. Section 3, Fig. 2, above). Furthermore, the non-German Slavs established a linguistic social buffer in Eastern Europe after c. 500 AD. A final blow to the long-standing treaty relationships between the eastern Empire and various barbarian kingdoms - mainly but not ex­ clusively of Germanic extraction - was dealt by the Persians and, in particular, by Islam in the seventh century. Western (unlike Eastern) Europe, according to the solidi (and other) archaeological evidence, only remained in contact with the Eastern Empire till the early seventh century, the period of the fined wars with Persia over the Middle East, shordy before the rise of Islam and the latter’s definitive conquest of the same valuable regions. This explains the progressive decline of the 19 Cf. Ian Wood’s contribution to the present volume.

THE MIGRATION PERIOD MODEL HISTORY AND TREASURE

83

gold content in the Frankish so-called moneyer coins of the mid- to late seventh century, as well as the eventual rise of the north-west Eu­ ropean silver coinage around AD 700.

4. Interlude: Gudme-lundeborg To further illuminate the above, it is useful to examine the royal centre of Gudme in Denmark. Gudme is a massive site of perhaps a hundred farmsteads, on south-eastern Funen. Together with the nearby tempo­ rary coastal emporium at Lundeborg, it constitutes a major barbarian Late Roman and Migration period setdement complex.20 It raises the question whether the emporium, a highly important feature of the Carolingian period, originated in the barbarian North, on the periphery of the Roman world, in a period when Roman villas, via and major towns were still functioning in the northwestern parts of the Empire. In any case there is, little doubt about the remarkable, and contem­ poraneous, increase in coastal production sites, both large and small, around the year AD 700 in all areas of north-western Europe.21 These were perhaps harbingers of the larger Carolingian emporia which pos­ sess evidence of having been planned enterprises with, for example, gridded street, and plot systems. In Denmark, the production and trad­ ing sites of this period (the earliest Viking Age) were linked to the cre­ ation of a new type of large farmstead with an augmented level of pro­ duction. Furthermore, estates comprising several units of production (farmsteads, etc.), were probably widely established during the same general period.22 Pollen-analyses demonstrate that the area which was taken up by fields and pastures expanded significantly from around AD 700 onwards, while the forests shrank concomitantly. In addition, ce­ real pollen appears far more commonly than in earlier phases, when there is a dominance of pollen from plants growing in pastures. This

20 P.O.Nielsen, K. Randsborg & H. Thrane, eds.

T h e A rch aeology o f G u dm e a n d L u n d e­

(Copenhagen, 1994). 21 Cf. J. Ulriksen, “Danish sites and settlements with a maritime context, AD 2001200”, A n tiq u ity 68 (1994), no.261, pp.797-811, Fig. 3. 22 Cf., for example, J. Callmer, “A contribution to the prehistory and early history of the South Scandinavian manor. The Bjäresjö investigations”, T h e A rch aeology o f th e C u l­ tu r a l L a n d sca p e. F ie ld W o rk a n d R esearch in a S ou th S w e d ish R u r a l R egion , eds. L. Larsson, J. Callmer & B. Stjernquist, Acta Archaeologica Lundensia. Series in 4’. No. 19. (1992), pp.411-457. borg. P a p ers P resen ted a t a C onference a t S ven d b o rg O cto b er 1 9 9 1

84

KLAVS RANDSBORG

is a general pattern throughout the north-western parts of the Conti­ nent. The Gudme complex of the Migration period might well be viewed as a lone demesne or proto-demesne centre, foreshadowing later de­ velopments, and perhaps inspired by major Late Roman rural settle­ ments or villas. Gudme, however, is not the only such centre in the country or general region. Furthermore, the double-complex of Gudme-Lundeborg might suggest that the production and trading sites of the Carolingian period acted as outlets and conversion points for the rural and other products of the demesnes. (More contentiously, Gudme-Lundeborg might even be termed a “pirates’ nest”, to put it into a military perspective. Minor expeditionary fleets are known to have roamed the waters of the Baltic and the North Sea). In form, Gudme (etymologically, the “Home of the Gods”) was an oversized village, consisting of a large number of farmsteads scattered (in groups) over more than one sq.km of prime agricultural land. A large, and unusually strongly-built, longhouse (47 by 10 m) may con­ stitute the hall of the royal homestead. From a military point of view, Gudme alone was able to muster at least two hundred warriors - the equavalent of six ships - at very short notice. In an age when expedi­ tionary armies were counted in hundreds rather than thousands, it would have been very difficult for enemy forces coming from the out­ side, to overrun the settlement. In addition, nearby villages would pro­ vide help, also at short notice. Hoards and a high number of single gold- and silver-finds, including many Roman coins, as well as a huge amount of other metal artefacts, fibulae among others, are scattered over the whole area of the Gudme settlement. Several of the hoards, and many of the other fine artefacts, were found within the confines of house structures (for example, in post holes), while the golden items were sometimes in rather small buildings, workshops. It is the rich finds of gold and silver, in particular, which sets Gudme apart from other villages or demesne centres. The fibulae suggest that the heyday of Gudme was in the fourth and, in particular, the fifth century. By the Carolingian period, Gudme was a rather or­ dinary village. It is also noteworthy that the Gudme setdement (note its heathen sacral name), is surrounded by other localities with heathen sacred names, which lie within a couple of kilometres of it. Although these names are imprecisely dated, they indicate the apparent religious significance of Gudme. Incidentally, a couple of kilometres to the east

THE MIGRATION PERIOD MODEL HISTORY AND TREASURE

85

of the settlement, at Mollegârdsmarken, half-way to the coast, is a very large cemetery. Some four kilometres to the east of Gudme is Lundeborg, a settle­ ment site around 900 m in length, which lies along the coast of the Greater Belt. Lundeborg has extensive traces of production, construc­ tion and repair including small workshops. In addition, weights have been found, as well as one hundred small pictoral gold-foils, of a prob­ ably ritual nature. The occupation of Lundeborg begins in the third century but the site is particularly prominent in the fourth century, and survives into the fifth, while a number of activities are even later. In most archaeological respects, Lundeborg is difficult to distinguish from the smaller emporia of the Carolingian period and, as indicated above, probably served the same functions, although the ceramic material is predominantly local and regional. The range of foreign items (mostly Roman silver coins, bronzes and some glasses) corresponds to that of the gifts found in Late Roman graves, for instance at Mollegârds­ marken. The rich Roman imports in Denmark are highly restricted in type and were probably political gifts rather than items which were traded commercially. Thus, the Lundeborg emporium and landing-site reveals both cultural aspects of its own particular period and the nature of exchange in the northern Germanic world during the Late Roman and the Late Antique periods. The Gudme-Lundeborg complex and a range of similar but smaller, and often less wealthy, sites provide a strong testimony to the impor­ tance of the Southern Scandinavian region, which was an island of sta­ bility and influence during the period of desolation of the western Em­ pire, in the fourth and fifth centuries AD. Indeed, the north is a highly important archaeological stepping stone in the making of the Late and post-Roman history of Europe and the Mediterranean.,

5. Pirenne and beyond Judging by the development in Gaul, the Roman setdement of the Late Imperial period, both rural and urban, was in decline.23 However, al­ ready by the sixth century, the number of new rural foundations in the Rhineland, and northern Gaul, regions crucial for the Merovingian 23 For this, and that which follows, see Randsborg, The First Millennium A.D. in Europe and the Mediterranean. An Archaeological Essay.

86

KLAVS RANDSBORG

and later developments, exceeded that of the abandoned settlements. A substantial number of the settlement sites of the High Middle Ages in this region were actually founded during the seventh to eighth cen­ turies. Truly urban life, however, did not return to the former Roman cities and towns until the tenth century. Proto-urban setdements, socalled emporia (or ports-of-trade), based on production and exchange, were established already in the late seventh century, or by the Carolingian period, mainly on coasts and river banks. In Free Germany, such sites are known already from the third/fourth century onwards, exemplified by the Gudme-Lundeborg complex described in the pre­ vious section. It is evident from Scandinavia’s shifting relations with the different parts of Europe - west, south and east, that the major sixth-century changes in communication, as well as the rise of the, initially relatively isolated, Slavonic agrarian societies, in eastern Europe, and various de­ velopments on the Steppes, also played a role - albeit indirecdy - in the formation of western Europe. Communication in both the North and the West was now clearly centred on Gaul and its tenuous links with the Mediterranean regions, such as Lombard and Byzantine Italy. From all this, it would seem that the development of the Early Me­ dieval, or Carolingian western Europe was contingent on the erosion and political and military collapse of the Western Roman Empire, rather than on the rise of Islam. Thus, archaeology maintains a tradi­ tional, pre-Pirenne, stand on this fundamental issue. *** The development in Late Roman and Late-Antique Syria (c. AD 200600), at the other end of the “Charlemagne-Mohammed” axis, was one of substantial rural and urban growth which remained unchecked until the period of the Persian Wars of the early seventh century. Islam seems, therefore, to have conquered some of the most profitable re­ gions of the eastern Empire. Early Islam, under the Ummayad dynas­ ty, made sub-Roman Damascus the capital of the Caliphate, but it failed to maintain the momentum of this development. In the mid­ eighth century, attention moved towards Mesopotamia, where the co­ lossal Abbasid capital of Baghdad was founded. Links between Islam and Byzantium were never completely severed, and were also maintained with Christian western Europe, although archaeologically this remains largely uninvestigated. Charlemagne’s Ele-

THE MIGRATION PERIOD MODEL HISTORY AND TREASURE

87

phant, for example, the famous gift from Harun-al-Rashid, arrived via Italy. Other corridors, although marginal, between Western Europe and the Near East, and, in particular, Byzantium, were re-opened in the late eighth century, by way of the North and Russia. Finally, the east-west route, linking the Frankish/German and western Slavonic lands with both south-eastern Europe and the landmass of later Russia, should also be borne in mind. From these observations, it emerges that Islam, although politically and militarily highly successful, could not reverse the slow decline of the newly conquered former Near Eastern possessions of the Eastern Byzantine Empire. Where Islam entered the lands, societies and com­ munities of the former Eastern Roman Empire - economically and po­ litically now isolated in Byzantine Aegean and Anatolia - it managed at most to halt the dedine in the regional settlement and economy. However, the communications and international market of the com­ monwealth of Islam were important in the long run, even for Christian Europe. By around AD 800, the mutual benefits of such world-systems of luxury and other trade, incorporating even Eastern Africa, India and China, must have been noted everywhere. It is therefore this period, the late eighth to early ninth centuries, rather than the seventh century, with the fall of the Roman Near East, which is the one in which the impact of Islam was truly felt in the West, as a distant driving force and global middle-man, stimulating Carolingian politics and luxury-generated economies, and therefore the rise of Western Europe. In summary, a “Pirenne Thesis” or “Pirenne Model” of today, drawing upon the archaeological evidence unavailable sixty years ago would stress: (1) the importance of the rise and development of the Roman Empire. (2) the eastward drift of the main centres of gravity (in terms of popu­ lation, production, political interest, etc.), of the Empires of the first millennium AD, as exemplified by the capitals of Rome, Constantino­ ple, Antioch and Damascus, Baghdad, and Samarra. (3) the emergence of regional polities in the slipstream of these devel­ opments. (4) the perseverance, and indeed the growth, of local populations and economies, freed from imperial domination and destruction. Alliances and warfare determined a series of historical events and re­ lationships, but were seemingly rarely decisive in themselves. Rather,

88

KLAVS RANDSBORG

ideological factors, in particular, religious beliefs, conceptions of ho­ nour, the power of luxury goods, and the rights to land and other means of production (as clearly emphasized by the material cultural re­ alities) shaped the patchwork of early west European societies. Through the need for alliances (including those of the church), and the quest for luxuries, combined with a rise around AD 700 in local de­ mand for craft products, links were forged with non-regional, even “world-systems”, of exchange. Eastern Europe, however, largely re­ mained in the slipstream until the tenth century. Pirenne seems to have regarded the integrated nature of western Eu­ rope from the sixth century onwards as of marginal importance. He based his thesis on the earlier western Europe of Classical Antiquity, a period in which the continent was composed of a number of highly disparate elements. In much the same way he seems to have ignored the particular development of eastern Europe, in which the southern regions remained a world apart, linked to Byzantium.

LES CONDITIONS DU COMMERCE AU Vie SIÈCLE Jean Durliat

D ’excellentes études ont rassemblé et discuté toutes les sources écrites actuellement disponibles pour une histoire du commerce et les archéo­ logues apportent chaque jour des informations nouvelles qui com­ plètent le tableau.1Nous savons que les marchands circulaient à travers l’ensemble du monde connu de l’époque, que les produits les plus divers étaient à la disposition de ceux qui pouvaient les payer, que l’ac­ tivité était importante dans les ports. Bélisaire rencontra une foule de marchands à son arrivée dans Carthage2 et le pape Grégoire le Grand n’a aucune difficulté pour adresser, à tout moment, une lettre dans n’importe quelle région du monde méditerranéen. Mais une question reste posée. Quel était le volume des biens trans­ portés par rapport à la production totale? Dépassait-il ce que l’on con­ naît, avec une précision relative, pour le XVIe siècle, ou les échanges étaient incontestablement développés et où, pourtant, le grand com­ merce du blé ne concernait guère plus de 1% de la production des pays

1 Parmi les nombreuses recherches consacrées au commerce pendant le Vie siècle, voir, en particulier, L. Ruggini, E co n o m ia e so cietà n e ll’ 'I ta lia a n n o n a ria ’. R a p p o r tifr a a g rico ltu ra e com m ercio d a l I V a l V I secolo d . C r (Milan, 1961); D. Claude, D ie b y za n tin isc h e S ta d t im 6 . J a h rh u n d e rt (Munich, 1969); D. Claude, D e r H a n d e l im w estlick en M itte lm e e r w ä h ren d des F rü h ­ m itte la lte rs (Göttdngen, 1985); D. Claude, “Aspekte des Binnenhandels im Merowingerre­ ich auf Grund der Schriftquellen”, U ntersuchungen z u H a n d e l u n d V erkehr d er v o r- u n d frü h g e ­ sch ich tlich en £ e it in M itte l- u n d N o rd eu ro p a , éd. K. Düwel, H. Jahnkuhn, H. Siems, D. Timpe, (Göttingen, 1985), pp. 9-99; R. Hodges et D. Whitehouse, M o h a m m ed , C h arlem agn e a n d th e O rig in s o f E u ro p e, (Londres, 1983); E x ca va tio n s a t C arth age. T h e B ritis h M issio n . 1 , 2 : T h e A ven u e d u P ré sid e n t B o u rg u ib a S a ia m b o ; T h e P o ttery a n d oth er ceram ic O bjects, éd. M. G. Fulford et D. P. S. Peacock, (Sheffield, 1984); M. Hendy, S tu d ies in th e B y za n tin e M o n e ta ry E con om y C . 3 0 0 - 1 4 5 0 , (Cambridge, 1985); S o cietà rom an a e Im pero ta rd o a n tico . 1 , Is titu tu z io m , ceti, econ­ o m ie 3 . L e m erci, g li in sed ia m en ti, éd. A. Giardina, (Roma-Bari, 1986); H o m m es e t rich esses d a n s l ’ E m p ire b y za n tin , 1 , I V - V I I siècle, (Paris, 1989); K. Randsborg, T h e F ir s t M ille n n iu m A D in E u ro p e a n d th e M ed iterra n ea n : a n a rch a eo lo g ica l E ssa y , (Cambridge, 1991); H. Siems, H a n d e l u n d W u ch er im S p ie g e lfrü lh m itte la lte rlic h e r R ech tsqu ellen , (Hanovre, 1992); L a sto ria econ om ica d i R o m a n e ll’a lto M ed io e v o a lla lu ce d e i recen ti s c a v i a rch eo lo g ici éd. L. Parali et P. Delogu, (Flo­ rence, 1993). 2 Procope, D e b ello va n d a lico , éd. G. Wirth, (Leipzig, 1962), 3, 20, 22.

90

JEAN DURLIAT

méditerranéens?3 Ne faut-il pas distinguer entre le nombre des march­ ands et les quantités transportées, car les embarcations étaient de petite taille, les routes souvent en mauvais état et la majorité de la population vivait à la limite de l’indigence, constamment menacée par la famine, sans compter les conséquences des épidémies et des guerres? Or le récit le plus circonstancié, le plus beau trésor monétaire, la plus belle fouille d’un port ou d’un dépotoir souffrent, quand on tente une histoire quantitative, d’une même faiblesse: sont-ils représentatifs de la situation dans l’ensemble du monde “romain”, conçu comme l’ensemble des pays - et de leurs marges - qui avaient en commun un passé et l’in­ fluence plus ou moins forte de la mer Méditerranée? Combien de blé, de tissus, de céramiques ou d’épices circulaient réellemment, et —ce qui est peut-être le plus important -, jusqu’à quelle distance? Pour se faire une idée plus précise de la place réelle occupée par le commerce dans la vie économique et sociale du Vie siècle, il n’est peut-être pas inutile de rappeler les conditions dans lesquelles il était pratiqué, d’abord les conditions concrètes de l’échange, puis les effets de la ponction publique, enfin les rapports entre l’offre et la demande. Elles sont suffisamment originales pour retenir l’attention et fournir un cadre dans lequel chaque document prend sa place et devient signifi­ catif de la situation générale, dans un domaine particulier. * * *

Le commerce est évidemment tributaire des conditions de la circula­ tion. Notre connaissance des moyens de transport est des plus réduite, ce qui conduit à élargir le champ chronologique de l’enquête. D’après l’édit de Dioclétien, les charrettes portaient habituellement une charge de 1.200 livres, soit quelque 400 kg,4 ce qui est peu, et rien ne laisse supposer que les teclntiques de construction aient évolué depuis le début du IVe siècle.5 Elles circulaient parfois à grande distance, ce

3

F. Braudel, C iv ilisa tio n m a térielle, E con om ie e t C a p ita lism e . X V e - X V I I I siècle. I. L es stru ctu res (Paris, 1979), p. 102. + M. Giacchero, E d ic tu m D io c le tia n i et collegaru m de p r e tiis rerum ven aliu m , 1 , E d ictu m , (Gênes, 1974), 17, 3. 5 Une enquête sur les outils dans les Balkans montre qu’il est impossible de dater un objet quand il n’est pas trouvé dans un contexte archéologique lui-même daté: G. Ostuni, L es o u tils d a n s les B a lk a n s du m oyen âge à n os jo u r s , (Paris, 1986). Ce qui est vrai de tous les outils l’était des charrettes avec lesquels on les fabriquait. d u q u o tid ien ,

LES CONDITIONS DU COMMERCE AU VIe SIÈCLE

91

qu’atteste, au Vile siècle, les transports pour l’abbaye de Corbie.6 Même si les faveurs royales pour cette dernière n’existaient pas au Vie siècle, d’autres transporteurs devaient utiliser les mêmes méthodes, puisque le port de Marseille recevait beaucoup d’huile.7 Mais les routes n’étaient toujours de bonne qualité et l’on avait souvent recours aux ânes,8aux mulets ou aux chameaux.9 Les bateaux annonaires, qui ne devaient pas se distinguer beaucoup des bateaux de commerce, avaient une capacité de 2.000 à 20.000 muids (15 à 150 tonnes),10 mais il n’est pas sûr que la moyenne d’entre eux jaugeait environ 10.000 muids (75 tonnes). La loi invoquée à ce sujet fixe seulement un barème: tout nou­ veau naviculaire consacrera le revenu fiscal de 50 juga à la construc­ tion de bateaux pour un tonnage de 10.000 muids; il est clair qu’une fraction de 50 juga sera affectée à un bateau de moins de 10.000 muids et un multiple à un plus gros.11Justinien réquisitionna des bateaux de 3.000 à 50.000 médimnes (22 à 330 tonnes) pour transporter les troupes en Afrique.12 Comme l’empereur avait sans doute retenu les plus gros porteurs disponibles, ceux-ci ne devaient donc pas dépasser 300 à 400 tonnes. Quand Jean Moschos présente comme exceptionnel un bateau de 50.000 muids, il confirme cette impression.13 Mais les ca­

6 L. Levillain, E x a m en c ritiq u e des ch a rtes m érovin gien n es e t carolin gien n es de V abbaye d e C orbie, (Paris, 1902) , P. J. no 15. 7 G regorii ep isco p i T u ro n en sis lib r i h isto ria ru m decent, éd. B. Krusch et W. Levison, {M o n u ­ m en ta G erm an iae H istó ric a , S crip to res R eru m M ero vin g ica ru m 1, 1) (Hanovre, 1937-1951), 5, 5. 8 M. Giacchero, E d ic tu m D io c le tia n i, 17, 5. 9 M. Giacchero, E d ic tu m D io c le tia n i, 17, 4. 10 Toutes les sources sur ce thème sont citées dans Claude, D e r H a n d e l im w estlich en M itte lm e e r w ä h ren d d es F rü h m ittela lters l, pp. 52-58. C odex T h eodosian u s, éd. Th. Mommsen et P. Ewald, (Berlin, 1905), 2, L eges n ovellae a d T h eodosian u m p ertin en tes, 8: FEtat peut réquisi­ tionner des bateaux de 2.000 muids et plus pour l’annone. V ie d e S ym éon le fo u e t vie de J e a n d e C h ypre , éd., trad, commentaire A.-J. Festugière en collaboration avec L. Ryden, (Paris, 1974), p. 8. Pour l’interprétation de ce passage, J. Durliat, D e la v ille a n tiq u e à la v ille b y za n tin e . L e p ro b lèm e d es su b sista n ces , (Rome, 1990), p. 240149. 11 Codex Theodosianus, 13, 5, 14: Excusandis videlicet pro denum milium modiorum luitione quinquagenis numero jugis in annonaria praestatione. J. Rougé, “Quelques aspects de la navigation en Méditerranée au Ve siècle et dans la première moitié du Vie siede”, C a h iers d ’h isto ire 6, (1962 p. 133 parle de 50.000 muids, pour l’époque romaine. 12 Procope, D e b e lb v a n d a lico , 1, 11, 13 13 Jean Moschus, P ra tu m sp iritu a le , dans P a tro lo g ia graeca 87, 83, 190. On retrouve les indications données les gros bateaux annonaires du Haut-Empire. Que la taille moyenne de ces derniers ait diminué à notre époque ne prouve pas un recul dans l’art de la con­ struction navale. 11 faut en effet se souvenir que 1 ’annone fut détournée de Rome vers Constantinople. Or, jusqu’à la construction, par Justimen, des immenses greniers de Tenédos, il fallait remonter le détroit des Dardanelles, dont les courants étaient forts, ce qui imposait l’usage de bateaux plus petits.

92

JEAN DURLIAT

pacités indiquées par diverses sources sont très variables, de même que la taille des épaves retrouvées - de 40 à 200 ou 300 tonnes —, ce qui interdit de conclure à une taille moyenne et, encore moins, de proposer la moindre estimation du rythme des rotations, donc des quantités transportées. En outre, il convient de distinguer entre les gros porteurs et la foule des petits bateaux ou même des barques qui effectuaient les transports à courte distance, souvent du simple cabotage.14 Enfin, il faudrait connaître les bateaux qui circulaient sur les fleuves navigables, le Nil et le Tibre,15 évidemment, mais aussi le Pô, la Seine16 et autres. Si les capacites des moyens de transport n’apportent aucune infor­ mation générale, l’étude leur coût marque les limites strictes imposées aux diverses formes de commerce. D’après l’édit de Dioclétien, alors que le blé vaut 100 deniers pour un muid militaire,17 soit 50 deniers p8our un muid ordinaire de près de 9 litres, ou quelque 6,5 kg, compte tenu de la densité du blé,18 le transport par chariot de 1.200 livres (400 kg) revient à 20 deniers pour une lieue, soit environ 1, 5 km.19 400 kg équivalant à 400: 6,5 = 60 muids, leur valeur est de 60 x 30 = 1.800 deniers. Le prix augmente donc de 1,1 % par lieue (0,75 % par km) et double au bout de 90 lieues (150 km). On pourrait calculer un même tarif pour le vin commun qui vaut approximativement le double du blé, car le poids du contenant compense sa valeur supérieure.20 Le coût du transport terrestre était donc prohibitif pour les produits de base et de faible valeur. Par contre, le transport maritime était beaucoup plus avantageux. La traversée de la Méditerranée, d’Orient en Lusitanie, 14 Exemple de barques de cabotage: J. Durliat, “Taxes sur l’entrée des marchandises dans la cité de Gzrafef-Cagliari à l’époque byzantine (582-602)”, D u m b a rto n O a k s P a p e rs, 36 (1982), pp. 2-3. 15 Pour les bateaux annonaires qui remontaient le Tibre d’Ostie ou de Porto à Rome (n aves a m n ic i) , tirés par des bœufs, nous disposons d’une indication: leur taille était de 40 cu pae (1.040 muids, environ .7,5 tonnes): N o v e lla i V a len tin ia n a e (dans C odex T h eo d o sia n u s, T h e o d o sia n i L ib r i X V I cum C on sH tu tion ibu s S irm o n d ia m s, eds T. Mommsen, P. Meyer, e t a l, 2) 29, 3. 16 V ita G enovefae v irg in is p a risie n s is, éd. B. Krusch, (M on u m en ta G erm an iae H istó ric a , S c rip tores R eru m M ero vin g ica ru m 3)(Hanovre, 1896), 35, avec le commentaire de M. Heinzelmann, dans M. Heinzelmann et J.-C. Poulin, L e s v ie s an cien n es d e sa in te G en eviève d e P a ris . É tu d e s critiq u es, (Paris, 1986), pp. 98-99. Des bateaux qui pouvaient être bloqués par des arbres ne devaient pas être de grande taille. 17 M. Giacchero, E d ictu m D io c le tia n i, 1,1. 18 Sur le rapport entre le muid militaire et le muid ordinaire, Durliat, D e la v ille a n ­ tiq u e à la v ille b y za n tin e , p. 494. 19 M. Giacchero, E d ic tu m D io c le tia n i, 17 3. Information confirmée par C odex T h eodosia n u s 8, 5, 28. Comme il est seulement question de donner des ordres de grandeur, tous les nombres sont arrondis. 20 Durliat, D e la v ille a n tiq u e à la v ille b y za n tin e , pp. 505-9.

LES CONDITIONS DU COMMERCE AU VIe SIÈCLE

93

ne coûtait que 24 deniers pour 1 muid militaire, soit 12 deniers pour un muid courant.21 Le prix du blé ou du vin étaient majorés de 24 %. D ’Alexandrie à Byzance, les frais n’étaient que de 6 deniers, soit un accroissement du prix de 12 %.22 Cette indication prouve que les nom­ bres donnés par l’édit correspondent à la réalité du Vie siècle. En effet les naviculaires qui transportent le blé annonaire sur ce trajet rece­ vaient, sous le règne de Justinien, une commission de 10 %, puisqu’ils transportaient 800.0000 d’artabes (24.000.000 de muids) valant 800.000 sous,23 moyennant une somme forfaitaire de 80.000 sous,24 soit 10 %. La faible différence s’explique aisément par le fait que leurs bateaux étaient constamment pleins, qu’ils n’avaient pas à chercher ou attendre les clients, et qu’ils faisaient toujours le même trajet. Ces in­ formations sont sans doute transposables dans les régions nordiques. En effet, pour atteindre la Lusitanie, les marins devaient affronter l’océan sans que les prix soient majorés. Même si les bateaux étaient plus petits et les frais légèrement supérieurs, l’ordre de grandeur ne devait guère varier, par exemple dans la Mer du Nord. Une différence très nette s’impose donc entre le commerce terrestre et le commerce maritime. Le premier est nécessairement réduit et ne peut transporter des produits de faible valeur à de grandes distances, alors que le second est capable de tout conduire d’un port à un autre, quelle que soit la distance et quelle que soit la marchandise. Mais la structure des prix introduit une contrainte supplémentaire. Toutes les études portant sur les prix romano-byzantins montrent qu’ils sont identiques, dans les mêmes conditions, d’un bout à l’autre du monde connu de l’époque, non seulement à un moment donné, mais à long terme.25 Comme elles ne distinguent pas toujours entre prix 21 M. Giacchero, E d ic tu m D io c le tia n i, 35, 17. 22 M. Giacchero, E d ic tu m D io c le tia n i, 35, 3. 23 Sur le prix du blé en sous, voir ci-dessous. 24 Édit 13, 7, dans C orpu s ju r is c iv ilis , 3, N o v e lla e , éd. R Schoell et G. Kroll, (Berlin, 1895). 25 Pour l’époque romaine, M. Corbier, “Dévalution et évolution des prix (Ier-IIIe siè­ cle)”, R evu e n u m ism a tiq u e, 27 (1985), pp. 69-106. Pour notre époque, E. Patlagean, P a u vreté écon om iqu e e t p a u v re té so c ia le à B y za n c e ( P - 7 e siècle), (Paris, 1977), pp. 377-422; C. Morrisson, “Monnaie et prix à Byzance du Ve au VIIe siècle”, H o m m es et rich esses d a n s l ’E m p ire b y za n tin , 1 , I V 1- V IIe sü c le , pp. 240-60; Durliat, D e la v ille a n tiq u e à la v ille b y za n tin , pp. 491-512. Pour la période byzantine, J.-C. Cheynet, É. Malamut, C. Morrisson, “Prix et salaires à Byzance (Xe-XIe siècle), H o m m es e t rich esses 2 , pp. 339-74. Les courbes réalisées par Ruggini, E co n o m ia e t so cietà n e ll’ “'Ita lia a n n o n a ria ”, pp. 364-68, sont intéressantes par la richesse de la docu­ mentation, mais ne sont pas significatives, car les nombres ne sont pas critiqués. Nous ignorons tout des prix dans les royaumes germaniques, sauf dans l’Italie ostrogothique, où ils n’ont pas changé. Les prix des oliviers, donnés par les T a b le tte s A lb e rtin i (éd. C. Cour­ tois, L. Leschi, C. Perrat, C. Saumagne, (Paris, 1932), doc. 3, 7, 9, 14, 18), posent plus de questions qu’ils n’aident à en résoudre. Mais l’Afrique reconquise n’a apparemment connu aucune évolution importante, dans ce domaine, preuve que les prix n’avaient guère changé. On voit mal comment les souverains germaniques auraient provoqué de grands change­ ments avant l’instauration de monnayages indépendants dans les années 580.

94

JEAN DURLIAT

publics et prix du marché, ou entre prix donnés par les lois et prix réellement pratiqués, il reste à prouver que toutes ces indications sont réellement convergentes. Une étude exhaustive des prix des denrées annonaires, en particulier du blé - pour lequel nous disposons d’une série suffisamment longue pour être significative -confirme cette im­ pression, à condition de prendre quelques précautions. Par exemple, le prix du blé est de 30 muids (2 qx) pour 1 sou, tant dans les tarifs offi­ ciels et les valeurs sur le marché, en année moyenne.26 Les papyrus d’Égypte donnent à eux seuls une liste significative, correspondant strictement aux informations glanées dans le reste du monde connu.27 Mais il est très sensible à la conjoncture. Si la récolte est excellente, il peut diminuer de moitié.28 Quand l’année est mauvaise, il monte très vite et triple, voire quadruple, quand la famine s’installe, et, lorsqu’une source dévoile les variations annuelles dans un lieu donné, comme à Édesse au tournant du Ve et du Vie siècle, on constate l’extrême sen­ sibilité des prix au volume de la production annuelles.29 En outre, les prix publics de coemptio,30 doivent être analysés avec soin, en tenant compte des méthodes particulières de la comptabilité publique. Ainsi un prix de 40 muids pour 1 sou inclut les frais de transport pour les bénéficiaires qui le touchent: l’État leur doit 30 muids pour 1 sou et il ne leur restera que 30 muids quand ils seront arrivés à l’endroit où ils doivent le consommer, mais l’administration leur accorde 33 % au titre du transport, y compris les éventuelles pertes en cours de route.31 Enfin le commerce s’effectuait presque sans usage de la monnaie car l’État en contrôlait strictement la circulation. L’exportation des métaux précieux, monnayés ou non, est interdite pour ne pas diminuer le stock

26 Durliat, D e la v ille a n tiq u e à la v ille b y za n tin e , pp. 497-502. 27 A. C. Johnson et L. C. West, B y za n tin e E g y p t E con om ie S tu d ies, (Princeton, 1948), pp. 175-98. Il faut se défaire du préjugé selon lequel l’Égypte serait particulière. Seule sa documentation diffère de ce qui est conservé ailleurs, par la richesse de ses archives. Mais partout où l’on a conservé des papyrus, ils sont de même nature que ceux d’Égypte. D’autre part, ils sont rédigés conformément aux lois générales de l’Empire. Il est parti­ culièrement remarquable que, dans le domaine où cette région se distingue le plus des autres, la produedon du blé, les prix soient identiques à ceux qui sont cités ailleurs. 28 A n o n ym i V a lesia n i p a r s p o sterio r, éd. Th. Mommsen, (Berlin, 1891) {M on u m en ta G erm a n ia e H istó ric a , A u cto res a n tiq u is sim i 9) p. 18. 29 Le fait est particulièrement visible, car il est bien décrit, à Edesse, entre 499 et 505 (.In certi a u cto ris p seu d o -d io n ysia n u m vu lgo d ictu m , éd. J.-B. Chabot, (Louvain, 1927), trad. J.-B. Chabot, (Louvain, 1949) pp. 177-217). Cf. Durliat, D e la v ille a n tiq u e à la v ille b y za n tin e , pp. 406-20. 30 Sur la coem ptio, voir ci-dessous. 31 N o v e lla e V alen tin ian ae 13. Cf. Durliat, D e la v ille a n tiq u e à la v ille b y za n tin e , p. 499.

LES CONDITIONS DU COMMERCE AU VIe SIÈCLE

95

disponible.32 11 en découle que les trouvailles monétaires constituent un mauvais guide pour une histoire du commerce international. A l’in­ térieur de l’Empire, la circulation des monnaies est limitée aux besoins personnels des marchands circulant sur les routes, durant leurs dé­ placements. La suite du texte, relative aux marins, n’est pas très claire, mais la loi limite leur droit à transporter de la monnaie, sans tenir compte de celle qui servirait à leur commerce: comme leurs collègues qui empruntent les routes, ils ne transportent pas de monnaie pour leurs échanges.33 D ’ailleurs la Loi des Rhodiens prévoit le cas où les marchandises sont perdues ainsi que celui où un passager voyage avec de l’argent,34 mais elle ne dit rien de celui que le capitaine emmènerait pour ses affaires. C’est donc qu’elles s’effectuaient normalement sans transport de fonds. Dans ces conditions, l’absence de monnaies d’or frappées hors de la capitale ou de la région dans laquelle elles ont été trouvées devient significative: ainsi on ne trouve en Afrique byzantine que des monnaies de Carthage ou de Constantinople,35 dans l’IUyricum que des monnaies de la capitale ou de la préfecture du prétoire,36 et, à Ostie et Porto, presque exclusivement des pièces frappées à Rome.37 Le commerce s’effectue sans transfert de fonds. Le même phénomène se retrouve dans les royaumes. Par exemple, le petit nom­ bre de pièces byzantines découvertes en Gaule au Vie siècle, en dehors des côtes méditerranéennes, s’explique ainsi. Quant à celles qui ont été découvertes dans le nord-est, elles proviennent plutôt des soldes touchées par des mercenaires ou de tributs que d’échanges commer­ ciaux.38 Mais le phénomène se complique du fait que, dans les années 32 C odex Iu stin ia n u s, éd. P. Krüger, C orpu s Iu r is C iv ilis II, (Berlin, 1892-5), 4, 63, 2. Non seulement il est interdit de payer en or les achats effectués auprès des barbares, mais les marchands doivent essayer de leur en tirer par la ruse. Les peines sont très lourdes contre les fonctionnaires qui ne sanctionneraient pas les contrevenants. l’Empire était mercan­ tiliste. 33 C odex T h eo d o sia n u s 9,23, 1. 34 N o m o s R h ô d iô n n a u tik o s. T h e R h o d ia n S e a -L a w , éd. W. Ashbumer, (Oxford, 1909), 3, pp. 13-14. 35 R. Guéry, C. Morrisson, H. Slim, R ech erch es arch éologiqu es fra n c o -tu n isie n n e s à R ou gga , 3. L e tréso r d e m on n aies b y za n tin e s , (Rome, 1982), pp. 62, 63, 71. De même on n’a trouvé que 3 monnaies d’or africaines hors d’Afrique 36 V. Popovic, “Les témoins archéologiques des invasions avaro-slaves dans l’Dlyricum byzantin”, M éla n g es de l ’É co le fr a n ç a is e de R o m e. A n tiq u ité , 87 (1975), pp. 445-504. 37 E. Spagnoli, “alcune riflessioni sulla circolazione monetaria in epoca tardoantica a Ostia (Pianabella) e a Porto: i rinvenimenti dagli scavi 1988-1991”, L a sto ria econ om ica d i R o m a , pp. 249-50. 38 J. Lafaurie et C. Morrisson, “La pénétration des monnaies byzantines en Gaule mérovingienne et visgothique du Vie au VIIIe siècle”, R evu e N u m ism a tiq u e , 29 (1987), carte 2.

96

JEAN DURLIAT

580, les royaumes d’Espagne, de Gaule et d’Angleterre se sont dotés d’une monnaie nationale et ont repris à leur compte la politique im­ périale. Les monnaies étrangères disparaissent presque totalement dans les royames et l’Empire n’admet pas la circulation de pièces barbares sur son sol.39 Ces conditions donnent la clé pour interpréter aussi bien les infor­ mations tirées des textes que celles de l’archéologie, à condition de dis­ tinguer soigneusement entre les types de commerce et les types de pro­ duits et de tenir compte autant du silence des sources que de leurs in­ dications positives. Le commerce local ou régional était vivant. Chaque bourg avait son marché où les paysans et les citadins échangeaient leurs produits. Bélisaire ordonnait à ses soldats d’acheter ce dont ils avaient besoin, pendant la campagne d’Afrique, ce qui suppose l’exist­ ence de personnes prêtes à leur céder ce qu’elles auraient apporté au marché.40 En outre la ville de Sullectum, en Byzacène fut prise par des soldats cachés sous les chariots des paysans qui se rendaient au marché.41 à Edesse, l’auteur anonyme de la chronique locale donne une véritable mercuriale des denrées.42 Les Perses, qui assiègent la ville, autorisent le commerce sous les murs d’Amid.43 Le témoignage le plus précis sur le commerce local est fourni par le tarif de Cagliari qui révèle la convergence vers la ville des mulets, des chariots et des barques chargées de subsistances,44 tandis que les papyrus donnent de longues listes de produits artisanaux.45 De même, des échanges variés et constants unissaient les deux rives d’une même mer, comme l’Ar39 Les monnaies byzantines en Gaule deviennent très rares (Lafaurie-Morrisson, “La pénétration de monnaies byzantines en Gaule mérovingienne et visigothique du Vie au VIIIe siècle”, carte 3). De même, on ne trouve guère de monnaies visigothiques hors du royaume et de monnaies étrangères dans le royaume (X. Barrai I Altet, L a circu la tio n des m on n aies s u b e s et visig o tiq u es. C o n trib u tio n à l ’h isto ire économ ique du royau m e visigota (Munich, 1976), p. 153). La lettre de Grégoire le Grand, dans laquelle il demande des produits plutôt que de la monnaie aux domaines du patrimoine de Gaule signifie que la monnaie étrangère n’a pas cours dans l’Empire. G regorii I P a p a e registru m ep istu la ru m , éd. E. Ewald et L. M. Hartmann, 2 t., (Berlin, 1887-1889) {M on u m en ta G erm an iae H istó ric a Ep. 1-2) 6, 10. Cf. J. Durliat, “Moneta e Stato”, L a cu ltu ra b iz a n tin a . O ggetti e m essagio , 4 , M o n e ta ed econ om ia , ed. A. Guillou (Roma, 1986), pp. 168-78. Contra P. Grierson, “The P a trim o n iu m P e tri in illis p a rtib u s and the pseudo-imperial coinage in Frankish Gaul”, R em ie belge de n u ­ m ism a tiq u e , 105 (1959), pp. 95-111. 40 Procope, D e B ello V an dalico 3, 16, 14. 41 Procope, D e B ello V an dalico 3, 16, 11. 42 In certi au ctoris p seu d o -d io n ysia n u m vulgo d ictu m , éd. Chabot, pp. 177-217. 43 H isto ria ecclesia stica ja c h a r m e rh etori vu lgo a d s c rip ta , trad. E. W. Brooks, 2. (Louvain, 1953), 7, 5. 44 Durliat, “Taxes sur l’entrée des marchandises dans la cité de G a rn ies” . 45 Johnson-West, B y za n tin e E g y p t: E con om ie S tu d ies , pp. 107-51

LES CONDITIONS DU COMMERCE AU VIe SIÈCLE

97

golide et l’Asie Mineure,46 ou les côtes italiennes et istriennes.47 Cepen­ dant le commerce entre cités de l’intérieur était presque inexistant. La chronique d’Edesse ne parle d’aucun marchand capable de soulager les misères de la population. Les seules mesures portent sur la distribu­ tion des réserves locales, en particulier par réquisition chez les posses­ seurs de blé.48 Dans ce cas, l’argument e silentio est pleinement valable, car, lorsque l’armée arrive, l’approvisionnement suit et notre source le mentionne avec précision; mais il s’agit de transports publics, non de commerce.49 La progression de la peste, en 543, livre un indice sup­ plémentaire. D’après plusieurs auteurs, le mal avança le long de la côte, par sauts de puce, sans pénétrer profondément dans les terres.50 C ’est la preuve que les échanges étaient réduits, dès qu’on quittait les côtes. De même, en Italie, l’abondance des sources permet de dresser une longue liste de famines qui ont frappé la péninsule.51 Elles sont ponctuelles, touchant une ville ou une petite région, et jamais les sub­ sistances ne circulent. Ce sont les hommes qui fuient vers les régions plus favorisées.52 Le coût des transports et surtout le prix moyen iden­ tique du blé sur toute l’étendue de l’Empire interdisaient de créer des circuits de distribution qui auraient dû être réorganisés chaque année, en fonction de la conjoncture dans les diverses régions.

46 C. Abadie-Raynal, “Céramique et commerce dans le bassin égéen du IVe au VIIe siècle”, Hommes et richesses dans l’Empire byzantin, pp. 143-60. 47 Magni Aurelxi Cassiodori Variarum libri XII, éd. A. J. Fridh, (Turnhout, 1973), 1, 34. En Italie, Théodoric interdit le commerce pour conserver sur place les ressources dis­ ponibles. Mais il fait seulement surveiller les côtes, preuve que le commerce était limité à ces régions. Magni Aurelii Cassiodori Variarum libri XII, 12, 22: Théodoric rappelle qu'il est plus sûr de “vendre” au fisc royal (c’est une coemptio) qu’aux marchands étrangers dont la présence n’est pas garantie. La plupart des autres mentions rapportées par Claude, Der Handel im westlichen Mittelmeer während des FrümittelaÜers, pp. 71-5 se rapportent à des transports publics. Le commerce existait entre les côtes italiennes et les régions voisines, mais il était limité. Quand, par exemple, Théodoric limite les exportations de blé {Magni Aurelii Cassiodori Variarum libri XII, 1, 34), c’est uniquement pour que les besoins publics {expensae publicad) soient couverts avant que les bateaux étrangers {peregrinae noues) ne puis­ sent exporter. Mais peregrinus signifie “étranger à la cité”. Ces bateaux peuvent venir d’une ville italienne voisine. Les crises frumentaires en Gaule ostrogothique sont com­ pensées autant par l’annone que par les marchands privés (Ruggini, Economia e società nell’”Italia annonaria”, p. 267). 48 Incerti auctoris pseudo-dionysianum vulgo dictum, éd. Chabot, pp. 196-7. Cf. Durliat, De la ville antique à la ville byzantine, pp. 409-16 49 Incerti auctoris pseudo-dionysianum vulgo dictum, éd. Chabot, pp. 210, 224. 50 Procope, De bello persico, éd. G. Wirth, (Leipzig, 1961), II, 22, 8-9. Corripus, Johannide, éd. J. Diggle et F.R.D. Goodyear, (Cambridge, 1970), 3, 344. Jean d’Ephèse. Historiae ecclesiasticae pars tertia, trad. E. W. Brooks, (Louvain, 1952), 238. 51 Ruggini, Economia e società nell’”Italia annonaria”, pp. 466-89. 52 Procope, De bello gothico VI, 20, 15-21.

98

JEAN DURLIAT

Le grand commerce était soumis à des contraintes aussi rigides. Les produits de consommation courante, ainsi que tous les produits dis­ ponibles localement, en étaient exclus.53 Nous verrons que la totalité du blé fourni aux villes venait des environs ou était donné par l’annone publique. Le vin et l’huile posent une question intéressante, car il faut distinguer plusieurs cas. Dans la région méditerranéenne, où leur pro­ duction est partout facile, les produits de consommation courante étaient consommés sur place: puisque leur prix était partout le même, il était inutile et non rentable de les transporter à plus de quelques ki­ lomètres, par voie de terre ou à quelques dizaines de kilomètres, par voie d’eau. Jean l’Aumônier refusa de boire le vin de Palestine qu’on lui servait parce qu’il voulait, comme les pauvres54 Alexandrins con­ somme le vin du lac Maréotide.55 L’abondance des amphores étrangères et leur fort pourcentage parmi la totalité des tessons dans certains dépotoirs, semble contredire cette affirmation. Mais il faut rap­ peler d’abord qu’on les trouve surtout dans les ports et rarement à l’in­ térieur des terres, signe que ce commerce s’arrêtait souvent au bord de l’eau. En outre, une amphore n’est pas chère56 et ne servait qu’une fois pour le transport à grande distance. Par contre, une fois arrivée à des­ tination, on l’utilisait un nombre indéterminé de fois pour du vin ou de l’huile locaux, ce qui fait baisser d’autant le rapport entre les im­ portations et la consommation locale. Même avec la moitié d’am­ phores importées, la part du vin étranger ne devait guère dépasser 10% de ce qui était bu, presque exclusivement dans les ports. Cette quantité représentait sans doute les crus que les riches achetaient pour leur qualité.57 D’autre part, le vin devient une denrée de luxe dans les ré­ gions nordiques où il n’est produit que dans certaines régions particu­ lières. Ainsi le vin de la région parisienne avait-il un statut différent de celui du lac Maréotide, proche des crus syriens et palestiniens qui ar­

53 Cf. Claude, D er H a n d e l in w estlich en M itte lm e e r w ä h ren d d es F rü m ittela lters, p. 73 qui cite et commente toutes les mentions de commerce du blé. 54 D faut entendre par pauvre, l’immense majorité de la population qui disposait uniquement de quoi subvenir a ses besoins élémentaires. 55 H. Delehaye, “Une vie inédite de saint Jean l’Aumônier”, A n a le c ta b o lla n ix a n a , 45 (1927), 10; trad, dans V. Jean l’aumônier, 327. D dit: ‘verse-moi du vin du lac Maréotide dont le goût n’ait rien de rare et dont l’achat soit bon marche’. 56 Prix des amphores dans Johnson-West, B y za n tin e E g y p t E con om ie S tu d ies , pp. 188189. Cf. Durliat, D e la v ille a n tiq u e à la v ille b y za n tin e , p. 530122. 57 II faut aussi faire intervenir la notion de « fret d’aller» sur laquelle nous revien­ drons.

LES CONDITIONS DU COMMERCE AU VIe SIÈCLE

99

rivaient dans les riches maisons de Carthage ou de Rome.58 fl suffit, pour se convaincre de ce que les produits de luxe étaient les seuls à circuler, de relire l’Expositio totius mundi, qui décrit la situation générale de l’Empire au IVe siècle, et correspondait sans doute à la situation au Vie siècle. La liste des produits importés en Égypte con­ firme la présence exclusive de produits de luxe, ayant une grande valeur et qui pouvaient donc venir de très loin.59 A l’autre bout du monde méditerranéen, le diplôme pour Corbie révèle une situation identique, avec cette particularité, normale pour un monastère du nord de la Gaule, que l’huile tient une grande place.60 Ce diplôme est im­ portant car il reflète la situation des greniers du fisc, donc la nature des échanges dans le port. Au Vie siècle, les seuls produits dont la valeur n’était pas nécessairement élevée, et qui faisaient l’objet d’un grand commerce, étaient ceux qui ne se trouvaient qu’en des lieux particu­ liers, comme le sel, dans certaines regions, ou les produits miniers, comme l’atteste un passage de la vie de Jean l’Aumônier sur lequel nous reviendrons. * * *

Si le commerce était limité aux secteurs que nous venons de décrire, comment vivaient l’armée et les grandes villes, grosses consommatrices de subsistances introuvables, pour l’essentiel, dans les environs? En ef­ fet, il ne faut pas oublier une donnée constante de l’économie avant la révolution agricole qui commença au XVIIIe siècle, à savoir que la nourriture d’un citadin absorbe ce que 8 ou 9 paysans ne consomment pas, ne livrent pas pour l’impôt ou ne vendent pas aux voisins, comme

58 Pour les Parisiens, le vin devait être un produit de consommation courante, mais, pour les Anglais, Saxons ou autres, c’était un produit de luxe, justifiant un transport à grande distance, puisque la vigne ne poussait pas chez eux. Mais la culture spéculative de la vigne ne se faisait qu’au bord de l’eau, de la Seine et de ses affluents (voir, pour une époque postérieure, J. Durliat, “La vigne et le vin dans la région parisienne au début du IXe siècle d’après le polyptyque d’Irminon”, L e M o y e n A g e, 1968, pp. 387-419). 59 E x p o sitio to tiu s m u n d i e t g en tiu m , éd. trad. J. Rougé, (Paris, 1966). U est remarquable que le blé figure parmi les richesses des provinces (index, s. v. F ru m en tu m ), non parmi ce qu’elles exportent, sauf en Sicile(c. 65). Mais n’est-ce pas un produit annonaire? L’auteur dit expréssement que l’Égypte fournit la capitale et l’armée en blé, donc en blé public (c. 36). Pour l’Égypte, Johnson-West, B y za n tin e E g y p t: E con om ie S tu d ies, pp. 140-151: U n’est fait mention que de produits de luxe, perles, tissus fins... ou huiles et vins, sans aucun doute de grande qualité. 60 Cf. n. 6.

100

JEAN DURLIAT

les prêtres, les artisans et les commerçants locaux.61 Même si les esti­ mations ne sont pas très sûres, Constantinople comptait au moins 400.000 habitants,62 Alexandrie et Antioche en avaient, chacune, plus de 100.000 et Rome, bien que déchue, sans doute plus de 50.000.63 Leur alimentation requérait respectivement - et au minimum 3.000.000, 800.000 et 400.000 paysans qui leur auraient livré tout ce qu’ils ne consommaient pas. Aucune société ni aucun ensemble de so­ ciétés n’était capable dé satisfaire de tels besoins de manière régulière, à une époque où - nous les verrons - les grandes sociétés commerciales n’existaient pas. L’Etat se substituait donc aux particuliers. La ponction publique était considérable. Divers calculs permettent d’établir qu’elle absorbait environ 20 % de la production de l’Empires.64 Il suffira de rappeler que l’aroure égyptien (1/4 ha) de terre arable versait 1,25 artabe (3,75 muids ou quelque 25 kg) de blé au titre de l’annone romaine puis constantinopolitaine,65 et que l’annone représentait la moitié de l’impôt total payé par l’Égypte.66 La charge totale était donc d’environ 25 x 2 x 4 = 200 kg pour un hectare. Com­ me toutes les estimations donnent un rendement approximatif de 10 qx à l’hectare, la proportion est bien de 20 %, dont la moitié en blé. Cela explique qu’il restait peu de choses pour le commerce, surtout pour le commerce international. Tout ce que nous savons des royaumes germaniques montre qu’ils n’ont rien changé à l’organisation

61 Et cela, bien que presque tout le monde, à la campagne, ait possédé au moins un jardin pour assurer une partie de sa nourriture. 62 L’annone d’Égypte livrait de quoi nourrir plus de 600.000 personnes (Durliat, D e la v ille a n tiq u e à la v ille b y za n tin e , pp. 259-260) mais les archéologues hésitent à admettre un nombre si élevé. Cependant, même si on suppose une affectation importante à des villes comme Antioche ou Thessalonique, la population de la capitale in tra m u ros ne pou­ vait être inférieure à 400.000 habitants, auxquels s’ajoutaient ceux des nombreuses villes de la banlieue. 63 Ce sont là des minima absolus, inférieurs à la réalité (Durliat, D e la v ille a n tiq u e à la v ille b y za n tin e , pp. 380-381; 338-339). Pour Rome, R. Krautheimer, R o m e. P ro file q f a C ity , 3 1 2 - 1 3 0 8 , (Princeton, 1980), propose, un peu au hasard, le nombre de 90.000 ha­ bitants à l’époque de Grégoire le Grand. Mais toutes les études rassemblées dans R o m a insistent sur le maintien des activités à un assez haut niveau. Il ne faut pas oublier que les constructions d’églises, monastères et hospices ont remplacé celles des temples, ther­ mes ou autres. 64 J. Durliat, L es ren tiers de l ’im p ô t. R echerches su r le sfin a n c e s m u n icip a les d a n s la P a ys O rie n tis a u I V e (Vienne, 1993), pp. 24-7. 65 C a ta lo g u e g én éra l d es a n tiq u ité s égyptien n es du m u sée d u C a ire. P a p y ru s g recs d ’époqu e b y z a n ­ tin e, éd.J. Maspéro, (Paris, 1911), 67057, addenda, p. 204. 66 Durliat, “Moneta e Stato”, pp. 193-96.

LES CONDITIONS DU COMMERCE AU VIe SIÈCLE

101

fiscale romaine et que l’ordre de grandeur de la ponction publique était identique à celui de l’Empire.67 Pour éviter toute confusion entre les échanges commerciaux et les prélèvement fiscaux, il convient de rappeler rapidement les postes con­ cernés par ceux-ci. Le premier bénéficiaire de l’impôt est l’administra­ tion centrale, qui doit assurer le salaire des fonctionnaires, l’entretien de la cour mais aussi le ravitaillement de la capitale et une partie du ravitaillement des grandes villes. Le second est l’armée dont il faut pay­ er les soldes, l’équipement ainsi que les besoins pendant les campagnes militaires. Puis viennent les cités qui utilisent le reste des sommes levées pour leurs dépenses propres, comme l’entretien des bâtiments et une partie de l’approvisionnement des marchés. Enfin, il ne faut pas oublier les Eglises dont les ressources proviennent, pour l’essentiel d’affectation de revenus publics, avec un budget spécifique, ce que rappelle Justini­ en, pour l’Empire, et que prouve l’analyse de la documentation occi­ dentale.68 Depuis que les évêques ont été placés à la tête de l’adminis­ tration municipale, les sources -surtout ecclésiastiques- ne distinguent pas toujours entre les deux budgets qu’ils administrent -celui de la cité et celui de son Eglise-, pour donner l’impression que toute la cité est régie par son prélat, habilement présenté comme un simple pasteur d’âmes, alors qu’il gérait aussi les affaires civiles.69 La multitude des be­ soins à couvrir explique la grande diversité des produits concernés. L’armée avait besoin d’armes, fabriquées dans les arsenaux, et de vête­ ments,70 sans compter les denrées et leur transport,71 l’entretien des bâ­ timents publics et religieux réclamait des pierres et de la chaux,72 et la 67 J. Durliat, L e s fin a n c e s p u b liq u e s d e D io clétien a u x C arolin gien s (2 8 4 - 8 8 8 % (Sigmaringen, 1990), pp. 203-5. 68 N o v e lla e J u s tin ia n i 7, 2: ‘La différence est faible entre entre le Sacerdoce et l’Empire, de même qu’entre les biens sacrés et ceux qui appartiennent à la collectivité ou à l’Etat, puisque les libéralités du pouvoir impérial fournissent constamment aux très saintes Ég­ lises la totalité de leurs ressources et de leur prospérité’. Cf. C odex Iu stin ia n u s 1, 2, 12. 69 Quand, par exemple, l’évêque d’une cité italienne indéterminée fournit 1.500 sous à l’armée de passage, sur ordre de la cour ostrogothique, il ne puise pas dans les ressou­ rces de son Église, mais dans celles de la cité (Fridh, M a g n i A u re lii C a ssio d o ri V ariaru m lib r i X I I , 2, 8.). D faut constamment se méfier de la redoutable habileté rhétorique des auteurs du Vie siècle. 70 Sur la v e stís m ilita r is et son importance, car elle concernait tout l’équipement des soldats, J. Delmaire. Largesses sacrées et res privata. L \aerariu m im p é ria l e t son a d m in istra tio n du I V e a u VIe siècle , (Rome, 1989), pp. 332-46. 71 In c e rti a u c to r isp seu d o -d io n ysia n u m vu lgo d ictu m , p. 207 et 211: la cité d’Édesse doit cuire 630.000 muids de blé, soit 42.000 qx. pour l’armée, et le blé vient parfois de loin, même d’Égypte (p. 217). 72 Sur la chaux, CTh 14, 6, 1 et 3. Le transport de la chaux se retrouve au Ville siècle, en Bavière: L e x B a ia w a rio ru m , éd. E. von Schwind, (Hanovre, 1892) {M on u m en ta G erm a n ia e H istó ric a Leges, 5, 2), 1,13. Noter que les transports de chaux sont effectués en partie jusqu’à la ville, donc pour satisfaire un besoin public.

102

JEAN DURLIAT

population attendait du blé ou ses substituts, comme les légumes secs,73 ainsi que des vêtements pour les pauvres, pour le clergé et autres béné­ ficiaires. Le principal piège, pour nous, provient de ce que l’administration utilise des méthodes originales. Afin de tenir des comptes homogènes, tout est exprimé en monnaie par les services de la comptabilité centrale et transmis, sous cette forme, à l’administration locale des gouverneurs ou des comtes. Ainsi les naviculaires d’Alexandrie reçoivent 80.000 sous de commission, pour le transport du blé annonaîre.74 Mais la dépense est, le plus souvent, effectuée en nature. Quand une cité doit une certaine somme au titre de l’annone, elle verse l’équivalent en blé, au tarif public de 30 muids pour 1 sou.75 Quand l’armée reçoit du blé, la somme correspondante est imputée à son compte, mais elle va cher­ cher les céréales dans un grenier public, au même tarif, majoré d’une commission pour couvrir les frais de transport.76 Ainsi l’État manipule de la monnaie de compte, mais la dépense est effectuée par affectation directe d’une certaine quantité de produits. Les royaumes germaniques connaissaient les mêmes méthodes de gestion.77 Par ce moyen, l’essen­ tiel des paiements publics échappait au commerce. Le fait, bien connu pour l’annone, vaut pour l’ensemble des budgets. Les impôts, exprimés en monnaie de compte,78 étaient versés pour partie en monnaie, pour partie en produits divers et pour partie sous forme de services.79 Les bénéficiaires touchaient leur dû de la même manière.80 Ainsi la dépense publique évitait le commerce et, comme elle représentait 20% de la production, une grande quantité - sinon la plus grande quantité-

73 Les autorités de Thessalonique envoient des bateaux “acheter”, c’est-à-dire obte­ nir par cœ m p tio , des légumes secs le long des côtes -toujours à proximité de la mer- de Grèce (L es p lu s an cien s recu eils d e m ira cles d e s a in t D ém étriu s, éd. P. Lemerle, 1, L e tex te, (Paris, 1979),§ 254 et 258). 74 Édit 13,7,dans N o v e lla e J u s tin ia n i, appendices. 75 Durliat, D e la v ille a n tiq u e à la v ille b y za n tin e , pp. 497-502. 76 Novellai V alen tin iae 13. Maspero, C atologu e g én éra l d es a n tiq u ité s égyptien n es du m u sée du C a ire. P a p yru s grecs d ’époqu e b y za n tin e , 67320. 77 E p isto la d e fis c o B arcin on en si, dans C o n cilio s visig o d o s, éd. J. Vivès, (Barcelone-Madrid, 1963), p. 54. 78 Par exemple, N o v e lla e V alen tin iae 36 donne à la fois la somme inscrite au budget de l’État et la quantité de viande que les contribuables livreront. 79 C odex T h eodosian u s 14, 6, 1 et 3 montre, par exemple, que le transport de la chaux constituait une part d’impôt payée sous forme d’un service. Les “services” effectués à l’époque carolingienne donnaient lieu à la même opération: tant de jours remplaçaient le versement d’un certain nombre de deniers. 80 Nous avons vu que les militaires reçoivent des vêtements et de la nourriture, au moins antant que de l’argent.

LES CONDITIONS DU COMMERCE AU VIe SIÈCLE

103

des produits circulant dans l’Empire et les royaumes échappait aux marchands, d’autant plus que l’État disposait d’une méthode simple pour couvrir ses dépenses extraordinaires: la coemptio (synônè en grec). Le terme est connu au Vie, et même au Vile siècle.81 Son applica­ tion ne correspond pas à un appel au marché par l’imposition de ventes forcées.82 fl s’agit d’une procédure purement fiscale qui permet­ tait d’exiger un produit particulier quand les circonstances l’exigeaient, par exemple lors du passage d’une armée ou pendant une famine. Procope en donne une excellente illustration, à travers un cas limite, peutêtre exagéré, puisqu’il présente les décisions de Justinien comme autant d’exactions: [L’administration] oblige les propriétaires fonciers à entretenir l’armée en proportion de l’im pôt dû par chacun. Les versements ne sont pas effectués en fonction des possibilités offertes par la saison où on les exiger mais en fonction de ce que l’administration croit possible et décide, sans chercher à savoir si les propriétaires produisent les biens exigés. D en résulte, pour ces m alheureux, l’obligation d ’importer les biens nécessaires aux soldats et aux chevaux, en achetant tout à des prix souvent très supérieurs, et cela dans des régions qui peuvent être éloignées, pour les transporter là où l’armée se trouve; ils doivent les distribuer aux responsables de l’arm ée non pas selon l’usage com m un à tous les hom m es, mais selon les norm es de ces derniers. C ’est ce qu’on appelle synônè.83

L’impôt est bien payé en nature, en fonction des besoins particuliers d’une armée en marche. L’obligation de recourir au marché parce que les exigences de l’État sont disproportionnées par rapport aux possibilités locales est présentée comme un abus inadmissible. Effectivement, les sources n’en donnent aucun autre exemple.84 En général, la coemptio

13 et Maspero, C atologu e g én éra l des a n tiq u ité s égyptien n es d u m usée du 67320 donnent des tarifs de coem ptio. Le terme apparut encore dans L e lib e r p o n tx fica lis, éd L. Duschesne, (Paris, 1886), 1, 366, sous le pontificat de Jean V (685-686). 82 Durliat, L es ren tiers d e l ’im p ô t. R ech erch es su r les fin a n c e s m u n icip a les d a n s la P a rs O rien tis a u IV e , pp. 131-4. L’édit de Dioclétien est en fait un tarif complet de coem ptio, indiquant à quel prix on doit lever tous les produits et de combien la charge fiscale sera réduite par ces versements. 83 Procope, A n ecd o ta , éd. G. Wirth, (Leipzig, 1963), 23, 11-13. 84 C ’est d’après cette description concrète qu’il faut interpréter l’ordre donné à l’évêque italien par Théoderic (cf. n. 69): il devra non pas fournir des sous, mais les pro­ duits dont l’armée a besoin, à concurrence de 1.500 sous. 81

N o v e lla i V a len tin ia e

C a ire. P a y ru sg re c s d ’époqu e b y za n tin e ,

104

JEAN DURLIAT

fonctionnait sans recours aux achats auprès des paysans ou des mar­ chands. Grégoire le Grand en donne un exemple précis.85 Elle était largement utilisée et des opérations qui apparaissent comme des achats sont en fait des coemptiones. Quand les autorités de Thessalonique en­ voient des bateaux “acheter” des légumes secs sur les côtes de la mer Égée, il est peu vraisemblable qu’ils aient eu recours au marché.86 Des magistrats ne se livrent pas au commerce; ils réquisitionnent dans les greniers publics. Le rôle de la monnaie dans les échanges était donc beaucoup plus faible qu’il n’y paraît au premier abord, ce qui rend anachronique la tentative d’appliquer à notre époque les formules qui établissent un lien entre le niveau des prix et la quantité de monnaie disponible. D’ailleurs aucun État n’avait de politique monétaire, au sens économique du terme. Les États disposaient de plusieurs moyens quand ils voulaient trans­ porter les produits. Les animaux étaient conduits à pied par des mem­ bres de la corporation chargés de les distribuer, comme les porcs qui étaient donnés à la population de Rome.87 Faute de sources, nous de­ vons supposer qu’il en allait de même, au Vie siècle, pour les distribu­ tions de Constantinople et pour tous les autres transports du même type. Le blé public était transporté par des flottes spécialisées, constru­ ites sur ordre du pouvoir par des naviculaires-armateurs, qui consac­ raient une partie de l’impôt perçu à la fabrication et à l’entretien de la flotte annonaire. Une loi nous apprend que l’impôt de 50 unités fiscales (juga) était affecté à la construction de bateaux d’une capacité de 10.000 muids.88 Nous savons aussi que les naviculaires-transporteurs touchaient une commission qui représentait 10 % du blé transporté. Les bateaux étaient alors exclusivement destinés au transport public, avec des contraintes qui assuraient leur rentabilité maximale.89 11 ex­

85 Ewald - Hartmann, Gregorii I Papae registrimi epistularum, 1, 42: le blé sera levé selon les pretia publica, avec des mesures particulières et ils doit être reçu (fiumenta debeant accipij, non acheté. 86 Cf. n. 72 87 Durliat, De la ville antique à la ville byzantine, pp. 74-80. 88 Codex Theodosianus 13, 5, 14. Le jugum rapporte environ 2,5 sous d’impôt, dont 1,25 pour l’annone et 50 juga rapportent donc 62,5 sous (Durliat, Les rentiers de l’impôt. Recher­ ches sur lesfinances municipales dans la Pars Orientis au IVe, pp. 24-26). Nous savons, par la loi des Rhodiens (Ashbumer, Nomos Rhôdiôn nautikos. The Rhodian Sea-Law, 2, 16), qu’une capacité de 1.000 muids valait 50 sous, ce qui représente 500 sous pour 10.000 muids. Un bateau annonaire était donc amorti en 8 ans. Mais il faut tenir compte des frais d’en­ tretien. Une fois encore, les tarifs donnent des ordres de grandeur comparables aux prix du marché. 89 Édit 13, 7, dans Novellae Justimani, appendices.

LES CONDITIONS DU COMMERCE AU VI e SIÈCLE

105

istait des flottes de guerre, dont témoigne la construction, par Théodoric, de 1.000 dromones90 mais, pour les transports exceptionnels, les ba­ teaux marchands étaient réquisitionnés, ce que fit Justinien pour at­ taquer l’Afrique vandale,91 comme Théodoric pour assurer l’approvi­ sionnement de son armée en Provence.92 Les réquisitions ne concernaient pas uniquement l’armée puisque sainte Geneviève en prit au moins onze pour aller chercher de quoi nourrir les Parisiens as­ siégés.93 Mais on utilisait le plus souvent les services d’embarcations privées qui s’acquittaient de leur impôt sous forme de transports pour l’administration, comme les bateliers du Nil.94 Sur terre, les produits étaient acheminés par des paysans qui préféraient effectuer des trajets avec leurs chariots plutôt que de donner une part de leurs maigres ré­ coltes. Le terme technique de la législation romaine pour désigner le paiement de l’impôt sous forme de transports est angaria. Ces services étaient réglementés. L’administration n’avait pas le droit de demander plus que le paysan ne devait au titre de l’impôt, ni de surcharger son chariot, ni de trop l’éloigner de chez lui. Les produits étaient unique­ ment ceux qui servaient aux besoins de l’État; parmi eux, les subsis­ tances pour l’armée et les matériaux pour les bâtiments publics ten­ aient une place prépondérante.95 Le terme et la pratique furent con­ tinués par les royaumes.96 Pour stocker les produits, avant de les redistribuer, des greniers étaient construits dans les villages et les villes.97 D’immenses greniers, à Ténédos, recevaient le blé d’Égypte en attendant que des bateaux plus petits le transportent jusqu’à Constantinople, quand les vents fa-

90 Fridh, M a g n i A u re lii C a ssio d o ri V ariaru m lib r i X I I , 5, 16. Ce sont des bateaux publics servant au transport de Pannone et à la défense des côtes (Claude, D e r H a n d e l im w estlich en M itte lm e e r w ä h ren d d es F rü h m ittela lters , p. 49). 91 Procope, De Bello Vandalico, X I,13-14. 92 Par exemple, Fridh, M a g n i A u re lii C a ssio d o ri V ariaru m lib r i X I I , 3, 41 93 V ita G en ovefae u irgin is p a risie n s is, éd. B. Krusch, (Hanovre, 1896) (M on u m en ta G erm a m a e H istó ric a , S crip to res R eru m M ero vin g ica ru m , 3), 35. 94 Contrat pour le transport du blé annonaire sur le Nil par un batelier: T h e O x yrh yn ch u s P a p y ri, 1, éd. B. P. Grenfell et A. S. Hunt, (Londres, 1898), no. 87. 95 C odex T h eo d o sia n u s 8, 5, p a s s im . 96 L eges V isigoth oru m , éd. K. Zeumer, (Hanovre-Leipzig, 1902) (M on u m en ta G erm an iae H istó ric a Leges 1) 5, 5, 2 et 12, 1, 2. L e x B a ia w a rio ru m , éd. E. von Schwind {M on u m en ta G erm a n ia e H istó ric a Leges 5, 2) 1, 13: comme dans l’Antiquité, les transports servent, en particulier, aux bâtiments urbains (cf n. 72). Les autres sont utilisés pour l’Église, qui est une institution publique. 97 Sur les greniers, C. Alzon, P ro b lèm es re la tifs à la lo ca tio n d es en trepôts en d ro it rom ain , (Paris, 1963).

106

JEAN DURLIAT

vorables permettaient de remonter le détroit des Dardanelles.98 La cor­ respondance de Grégoire le Grand montre leur organisation pour l’annone romaine à la fin du Vie siècle. Le blé est concentré sur le lieu de production, en Sicile; bien qu’ils soient appelés horrea ecclesiastica, ils servent à des opérations publiques.99 Une partie est sans doute affectée sur place, aux besoins locaux. Une autre est livrée aux fonctionnaires, pour les besoins généraux de l’Empire.100 Le reste part pour Rome sur des bateaux annonaires où il est conservé avant d’être distribué. Ainsi apparaît la grande souplesse du système. L’administration sait de com­ bien elle dispose, fait verser tout ce qui est dû par tous les contribuables et affecte diverses quantités à des besoins permanents ou exceptionnels, en fonction des circonstances. Le responsable, l’horrearius, doit tenir les comptes, vérifier que les rentrées sont correctement effectuées et ex­ iger des justificatifs avant de livrer ce qu’on lui demande.101 Dans l’in­ térieur des terres, les transports étaient si difficiles que le blé, le vin, l’huile et les autres produits servaient uniquement aux besoins locaux et à ceux de l’armée. Mais le grenier constituait des réserves pour at­ ténuer les effets des variations climatiques, comme à Edesse, où des dis­ tributions furent organisées pour les citadins, mais aussi pour les ruraux venus, de manière paradoxale, chercher en ville la nourriture que leurs champs ne produisaient pas.102 Le long des cours d’eau, il était possible d’aller chercher de la nourriture assez loin.103 Dans les ports maritimes, les bateaux amenaient de plus grosses quantités, puisées dans les grands greniers des régions qui produisaient tel ou tel produit plus facilement. C’est pourquoi toutes les grandes villes sont des ports. Rome faisait ap­ pel à la Sicile, mais aussi à l’Égypte.104 Alexandrie s’approvisionnait en Sicile quand la récolte locale était mauvaise.105 Toutes les villes situées sur la route du blé destiné à Constantinople profitaient de détourne­ ment en cas de pénurie, même si, pour notre époque, les sources l’at­ testent uniquement pour Thessalonique,106 et la capitale puisait dans 98 Procópe, Anecdota, 23, 11, 14 99 Ewald - Hartmann, G regorii I P a p a e registrim i ep istu la ru m , index, s.v. 100 Grenfell - Hunt, T h e O xyrh yn ch u s P a p y ri, 1, 16, no. 1920, pour un exemple de versement à des soldats. 101 Durliat, D e la v ille a n tiq u e à la v ille b y za n tin e , pp. 464-71. 102 In c e rti a u cto ris p seu d o -d io n ysia n u m vu lgo d ictu m , pp. 195-6. 103 Geneviève se rend à Arcis-sur-Aube: cf n. 93. 104 L’envoi de blé égyptien à Rome apparaît pour la dernière fois dans les sources sous le pontificat de Benoît 1er (575-579) {Le liber pontificala, p. 308). 105 Festugière-Ryden, V ie d e S ym éon le fo u e t v ie d e J e a n d e C h ypre, c. 11. 106 Lemerle, L e s p lu s a n cien s recu eils d e m ira cles d e s a in t D ém étriu s, 1, 8 et 13. Commen­ taire dans Durliat, D e la v ille a n tiq u e à la v ille b y za n tin e , pp. 390-9.

LES CONDITIONS DU COMMERCE AU VIe SIÈCLE

107

tous les greniers de l’Empire. L’État byzantin assurait ainsi la continu­ ité de l’alimentation dans les très grandes villes, incapables de subvenir à leurs besoins par le seul recours au marché. L’État ostrogothique fa­ isait de même en interconnectant les greniers publics.107 Malgré l’ab­ sence de sources, il est vraisemblable que les autres royaumes avaient continué la tradition romaine. Mais, comme leurs villes étaient plus pe­ tites, les besoins étaient moins importants. Malgré son rôle prépondérant, l’État ne satisfaisait pas toute la de­ mande. La majorité des villes n’avait qu’un grenier destiné aux besoins publics et à l’alimentation pendant une soudure difficile. Le reste ve­ nait du marché local ou régional. Seules les grandes et très grandes villes étaient incapables d’obtenir ainsi la subsistance de base, le blé, et une part variable des autres, comme le vin, l’huile et la viande de porc. Mais la viande de bœuf,108 et tous les produits frais étaient fournis par le commerce. D suffit de songer au nombre de jardins mentionnés dans les villes ou à leur périphérie pour en être persuadé, 109ou de songer aux marchés urbains qui apparaissent dans les sources.110 En outre, les tissus, sauf ceux qui étaient destinés à l’armée, les matériaux de con­ struction, la vaisselle et autres produits étaient achetés par les particu­ liers. Les marchands étaient seulement incapables de créer les grandes compagnies sans lesquelles la continuité de l’approvisionnement aurait été impossible. D ’où vient cette incapacité? * * *

L’histoire du commerce à l’époque pré-capitaliste se situe dans un cad­ re de pénurie et le Vie siècle ne fait pas exception. Le faste des cours, la taille et le luxe des plus belles églises ou la richesse de quelques per­ sonnes ne doit pas dissimuler la misère de la grande majorité de la pop­ ulation. Or l’histoire économique traite des phénomènes de masse, non des exceptions qui alimentent une mince frange de l’activité en général, et du commerce en particulier. Imaginer un niveau de développement

107 Fridh, M a g n i A u r e lii C a ssio d o ri V arioru m lib r i X I I , 27-28. 108 Mention des b o a rii de Rome dans N o v e lla e V alen tin iae, 36. 109 Jardins dans la ville de Rome, par exemple dans la charte du pape Serge pour Sainte-Suzanne (texte dans L e ü b er p o n tific a te , pp. 379-80). 110 Les marchés n’ont pas été étudiés mais les sources en mentionnent un grand nombre dans les villes. Les prix du marché à Édesse supposent qu’il en existait au moins un.

108

JEAN DURLIAT

semblable à celui que connut l’Europe occidentale à partir du XHIe siècle serait un anachronisme grossier. Cette pénurie apparaît d’abord à travers les rares indications rela­ tives aux salaires. Rares sont ceux qui atteignent ou dépassent 1 sou par mois, soit 12 sous par an,111 à une époque où le blé coûtait 1 sou les 2 qx (30 muids).112 Chez les paysans, une estimation aboutit ap­ proximativement au même résultat. Le jugum correspond à l’exploita­ tion paysanne moyenne; son imposition, qui est de 2,5 sous par an, représente environ 20% de la production, ce qui donne une production de 12,5 sous et un revenu, après impôt, de 10 sous, qui ne doit pas être loin de la réalité, si on enlève la semence et si on ajoute le petit élevage, la pêche ou la cueillette. Comme la consommation moyenne est, pour toutes les époques où on peut la calculer, de l’ordre de 2 qx par personne et par an, le pain absorbait au moins la moitié du salaire, pour une famille de 5 à 6 per­ sonnes. Mais il ne s’agit là que d’une moyenne car, au moment de la soudure ou pendant les mauvaises années, le prix du blé doublait rap­ idement, voire triplait, ce qui provoquait une disette, ou augmentait encore plus, et c’e3tait la famine. Or, dans les économies tradition­ nelles, les mauvaises années revenaient souvent.113 Les bonnes années ne compensaient pas les périodes de crise, au moins pour les paysans qui formaient plus de 80% de la population, car les prix s’effondraient, faute de clients et de débouchés ex­ térieurs.114 Une anecdote, rapportée par Procope, illustre ce fait: [Les empereurs] avaient établi des relais de poste à raison tantôt de huit, tantôt de m oins, mais pas m oins de cinq en général, sur la distance d ’un jour de marche pour un hom m e sans bagages. [...] T ous les propriétaires fonciers, et en particulier ceux dont les dom aines se trouvaient à l’intérieur des terres,

111 Johnson-West, Byzantine Egypt: Economie Studies, pp. 194-7. C. Morrisson, “Mon­ naie et prix à Byzance du Ve au VIIe siècle”, pp. 255-6. Ces indications sont difficiles à interpréter car les notaires touchent des commissions, et la conversion des prix exprimés en monnaie de cuivre en monnaie d’or est souvent approximative. Il apparaît cependant que seuls les ouvriers spécialisés dépassent 1 sou par mois, comme le tailleur de pierre qui gagne 1 carat par jour (environ 2 sous par mois), le charpentier qui en gagne l ' / 3 ou le calfat qui en touche 1 V2. Le porteur d’eau qui gagne 3 sous par an est-il nourrir 112 Durliat, De la ville antique à la ville byzantine, pp. 497-502. 113 Ruggini, Economia e società neU’”Italia annonaria”, pp. 466-89, pour l’Italie. 114 Les prix des périodes d’abondance sont faibles. En Italie, ils chutent de 50 % (60 muids pour 1 sou): Anonymi Valesiani pars posterior, éd. Th. Mommsen, (Berlin, 1891) {Mon­ umenta Germaniae Histórica Auctores Antiquissimi 9), 73.

LES CONDITIONS DU COMMERCE AU VIe SIÈCLE

109

en tiraient une prospérité tout à fait considérable. En effet ils gagnaient beau­ coup d ’argent en vendant chaque année au fisc les produits de leurs cham ps pour la nourriture des chevaux et des palefreniers. H en résultait que le fisc encaissait chaque année l’argent dû par chacun115 tandis que les fournisseurs de l’État se voyaient rendre l’argent sur-le-cham p.116

Quand la poste fut supprimée, les propriétaires furent ruinés car ils devaient toujours verser de l’argent, mais n’avaient plus de clients. De même, quand la récolte était excédentaire, les acheteurs faisaient dé­ faut. Ce texte fait à nouveau apparaître la différence entre l’intérieur et les régions côtières. Dans les premières, le seul client était l’État et l’arrêt de ses achats, ou la perception de l’impôt en monnaie et non en nature, provoquait une crise de surproduction. Or les régions de l’in­ térieur sont plus significatives de la situation économique générale que les minces bandes côtières. Quant aux nobles,117 ils ne participent en rien à la production de biens destinés au commerce. Ils vivent des revenus de leurs terres, sans qu’on constate, chez eux, la moindre préoccupation de l’investissement productif,118 et des salaires qu’ils touchent en tant que fonctionnaires. La liaison entre possession du sol et service de l’État apparaît nette­ ment dans toutes les sources, mais plus particulièrement à Ravenne, qù les personnes qui possèdent des terres, les reçoivent du souverain ou prennent celles de l’Église de Ravenne en bails emphytéotiques, sont le plus souvent des militaires ou des fonctionnaires.119 Or l’exercice du 115 Pour le paiement de l’impôt. 116 Procope, A n ecd o ta , 30, 3, 7. Procope décrit un cas de q u a si-c o e m p tio puisque l’ar­ gent gagné en vendant à l’État rapporte de quoi lui payer l’impôt. Les marchands sont absents. Ce système par lequel les pièces ne font que passer dans les caisses des con­ tribuables constitue un “circuit court” de la monnaie qui va du Trésor au Trésor, moy­ ennant un petit détour. 117 Sur la notion de noblesse, J. Matthews, W estern A risto cra cies a n d Im p e ria l C ou rt. A . D . 3 6 4 - 4 2 5 , (Oxford, 1975); Actes du colloque, N o b ilita s . F u n ktion u n d R ep rä sen ta tio n des A d e ls in A lteu ro p a (sous presse). Il manque une étude de la noblesse dans l’empire proto­ byzantin. 1,8 Les papyrus et les autres documents, tels la correspondance de Grégoire le Grand, montrent un souci de maintenir les terres en bon état, mais aucun souci de l’in­ vestissement productif. Il faudrait discuter de près les exceptions, réelles ou apparentes, par exemple, Ewald - Hartmann, G regorii I P a p a e reg istrim i ep istu la ru m , 2, 38, dans laquelle le pape demande de reconvertir les bouviers en agriculteurs. Il faudrait surtout distinguer entre ses activités comme chef d’exploitation et comme administrateur. 119 D ie n ich tlitera n isch en latein isch en P a p y r ii Ita lie n s a u s d er £ e it 4 4 5 - 7 0 0 , éd. J.-O. Tjader, 1, (Lund, 1955); 3, (Lund, 1982), p a s s im . T. S. Brown, G en tlem en a n d O fficers. Im p e ria l A d ­ m in istra tio n a n d A risto c ra tie P o w e r in B y z a n tin e I ta ly A .D . 5 5 4 - 8 0 0 , (Rome, 1984), en tire des conclusions excessives sur la place de l’armée dans la vie sociale, qui ne nous intéressent pas ici.

110

JEAN DURLIAT

commerce leur est interdit par la loi puisqu’un marchand qui entre dans la curie doit vendre son entreprise et acheter des terres avec cet argent.120 Ceux qui sont nobles par la naissance, par les honneurs et par leurs patrimoines, ne doivent pas exercer un commerce dangereux pour les villes, mais laisser les plébéiens et les commerçants vendre et acheter entre eux.121 Ces dispositions rappellent celles du Digeste, qui, sous le Haut-Empire, interdisaient aux décurions de gêner le marché par la vente intempestive de blé du grenier municipal.122 D leur était certainement interdit d’utiliser les stocks publics déposés dans leurs gre­ niers pour des ventes privées. Les clercs ne sont autlorisés à échanger que pour le bien de leur Eglise ou pour des affaires de faible impor­ tance.123 D’ailleurs le commerce est mal vu des élites, auxquelles ap­ partiennent les clercs: les marchands sont considérés comme tous les humbles qui travaillent de leurs mains et ne se consacrent pas à l’otium. Quand ils justifient leur action par un argument qui nous paraît parfaitement recevable, ils ne convainquent pas les nobles de l’époque.124 Les nobles apparaissent donc comme des consommateurs -surtout de produits de luxe125 - et non comme des agents du com­ merce. Leur seule intervention peut se situer dans la fourniture de capitaux par le biais de prêts maritimes entre un bailleur de fonds et un capi­ taine. La Loi des Rhodiens ne précise pas la condition du premier,126 12, 1, 72. 4, 63, 3. Noter le lien entre noblesse, naissance, fonctions publiques et fortune foncière. 122 C orpu s J u r is C iv ilis, 1, D ig e sta , éd. Th. Mommsen (Berlin 1827), 48, 12, 3; 50, 1, 8; 50, 8, 7. 123 Claude, D e r H a n d e l im w estlich en M itte lm e e r w ä h ren d d es F rü h m ittelaU ers , pp. 236-9; Siems, H a n d e l u n d W u ch er im S p ieg el frü h m itte la lte rlic h e r R ech tsq u ellen , pp. 666-78. 124 Saint Augustin. E n a rra tio n es in p sa lm o s L X V , éd. E. Dekkers et J. Fraipont, (Tumhout, 1956). 1,17 reprend l’argument généralement défendu par les marchands de son temps quand il fait dire à l’un d’eux qu’il exerce une activité utile puisqu’il apporte des marchandises de loin dans des lieux où elles sont absentes, et que le péché se trouve dans les abus de certains marchands, non dans le commerce (noter que le commerce a pour but de procurer ce qui manque sur place, non de fournir des débouchés à la production locale pour la valoriser). Mais l’ensemble de la littérature, laïque ou ecclésiastique, qui reprend les conceptions nobiliaires, méprise les commerçants. Cf. Claude, D e r H a n d e l im w estlich en M itte lm e e r w ä h ren d des F rü h m ittela lters, pp. 200-4; Siems, H a n d e l u n d W u ch er im S p ie g e l frü h m itte la lte rlic h e r R ech tsqu ellen , pp. 679-95. 125 Le commerce des marbres, par exemple, ne concerne que les nobles et les clercs. Sur ce commerce, J.P. Sodini, “Le commerce des marbres à l’époque protobyzantine”, H o m m es e t rich esses d a n s l ’E m p ire b y za n tin , I, I V e - V i t siècle , pp. 186-7. Les nombreuses cartes de cet article illustrent bien le fait que les marbres s’éloignent rarement de la mer ou des fleuves navigables. 126 Ashburner, N o m o s R h o d io n n a u tik o s. T h e R h o d ia n S e a -L a w , 3, 17-18. 120 121

C odex T h eo d o sia m is

C odex J u stin ia n u s

LES CONDITIONS DU COMMERCE AU VIe SIÈCLE

111

mais, d’après une lettre de Grégoire le Grand, dans laquelle le prêteur est un vir magnifiais,127 nous apprenons que certains nobles s’en­ gageaient dans ces opérations. Cependant les autres exemples connus mettent en scène des individus qui ne sont pas des nobles.128 L’inter­ vention des Églises pose une question intéressante. Grégoire le Grand n’intervient pas directement dans les affaires mais se contente de venir en aide à des marchands en difficulté, comme il le ferait pour toute autre personne.129 L’Église d’Alexandrie aurait eu une flotte et, en par­ ticulier, 13 bateaux qui firent naufrage dans la mer Adriatique avec une cargaison d’une très grande valeur: 34 kentenaria, soit 3.400 livres d’or ou près de 250.000 sous.130 Mais l’auteur ne mentionne explicite­ ment que des légumes secs, des vêtements et de la monnaie. Ce sont les produits que l’Église de Ravenne tire de ses patrimoines.131 En out­ re, les capitaines sont des pronaukléroi, terme grec qui signifie, dans la langue administrative, chef des naviculaires, c’est-à-dire des marins qui assurent les transports publics. Surtout, Jean est réconforté, après cette catastrophe, par les curiales (oi apo)132 d’Alexandrie dans une pièce de l’évêché (sèkréton) qui fait manifestement office de curie.133 La flotte de l’Église d’Alexandrie est donc la flotte qui assure le transport des produits fiscaux auxquels la cité a droit, comme la flotte de Ravenne transporte les produits fiscaux destinés à l’Église de la cité. L’idée que les évêques participeraient à de grandes actions commerciales est d’ailleurs en contradiction avec la législation civile et religieuse qui leur interdit ce type d’activité. Pas plus que les nobles, les Églises ne parti127 Ewald - Hartmann, Gregorii I Papae registrimi epistulanan 9, 10. 128 Ewald - Hartmann, G regorii I P a p a e reg istrim i e p istu la n a n 3, 55; 4, 43; 9, 40. Cf. Claude, D e r H a n d e l im w estlich en M itte b n e e r w ä h ren d d es F rü h m ittela lters, pp. 220-5. 129 D faudrait commenter attentivement les lettres citées à la n. précédente, ainsi que les textes analysés par Claude, D e r H a n d e l im w estlich en M itte lm e e r w ä h ren d d es F rü h m ittela lters, pp. 231-6. 130 Festugière, V ie de S ym éon le f o u e t v ie d e J e a n d e C h ypre, c. 28. Ce nombre est in­ vraisemblable, comme souvent, dans les sources hagiographiques. 131 Agnellus, L ib e r p o n tific a lis ecclesiae R a ven n a tis, éd. O. Holder-Egger, S crip to res rerum la n g o b a rd ica ru m e t ita lic a ru m sa ec. VI-IX„ (Hanovre, 1878), p.350: les bateaux de l’Église sont chargés de blé, de légumes, de peaux, de vêtements, de vases en argent et en or et de 31.000 sous. Ces produits, tirés de ses domaines, sont des revenus publics affectés par l’État aux besoins des Églises. 132 Oi apo est une expression technique pour désigner les curiales. Quand l’auteur prétend que la moitié de la ville vint dans le sekreton, il prend manifestement la partie, les curiales, pour le tout, la population, car il est impossible de faire rentrer une telle foule dans une salle, si grande soit elle. Ce procédé est très fréquent dans les sources. 133 Une étude exhaustive des sources montrerait que non seulement l’évêque est de­ venu le chef de l’administration municipale mais que la curie se réunit dans des locaux épiscopaux.

112

JEAN DURLIAT

cipaient au commerce, mais faisaient acheminer - par des bateaux tra­ vaillant pour elles dans les mêmes conditions que pour l’État — ce qu’elles percevaient sur leurs patrimoines ou recevaient des greniers publics, sans avoir recours au marché. De même que les greniers ec­ clésiastiques sont soit ceux des Églises soit ceux des cités, dont le prélat a la responsabilité, les bateaux des Eglises servent à la circulation de tout ce que l’évêque reçoit, aussi bien comme chef religieux que com­ me chef de l’administration municipale. L’importance de la circulation des produits destinés à l’Etat ou aux cités donne l’illusion d’un grand commerce, alors qu’il s’agit de transports publics. L’ensemble des ba­ teaux qui avaient des liens avec les Eglises - et auxquels elles faisaient le plus souvent appel pour leurs transports - étaient dits recommandés, sans que, pour autant, elles aient possédé des flottes ou participé aux affaires.134 Le commerce ne concernait donc aucune catégorie de puis­ sants, laïques ou ecclésiastiques. L’exament des contrats, qui constituent une forme archaïque de commenda, fait apparaître qu’ils étaient signés pour un seul trajet, avec partage des pertes et des profits au retour.135 Un seul article de ce code maritime envisage le cas où ‘quelque chose arrive à l’un des bateaux’, et indique que la collaboration portait parfois sur plusieurs navires.136 Le plus souvent, les marchands louaient une partie d’un bateau, pour une petite cargaison.137 Les marchands ne disposaient pas de gros cap­ itaux et, si certains d’entre eux devenaient fort riches, comme les deux frères Elijah et Théodore qui possédaient une luxueuse demeure à Mélitène,138 ou atteignaient des postes aussi importants que l’évêché de Paris,139 la majorité d’entre eux ne possédait qu’un bateau et ne dis­ posait que de fonds propres limités. Les miracles de saint Artémios, bien qu’ils soient du milieu du Vile siècle, décrivent sans aucun doute une situation stable depuis longtemps. Ils nous présentent uniquement des individus ou des familles trafiquant de manière strictement in­

134 Claude, Der Handel im westlichen Mittelmeer während des Frühmittelalters, pp. 234-235. 135 Ashburner, Nomos RhSdiSn nautikos. The Rhodian Sea-Law, 3, 17-18. Cf. Claude, Der Handel im westlichen Mittelmeer während des Frühmittelalters, pp. 224-226. 136 Ashburner, Nomos RhSdion nautikos. The Rhodian Sea-Law, 3, 21. 137 Ashburner, Nomos RhSdion nautikos. The Rhodian Sea-Law, 3, 23. 138 Jean d’Éphèse, L iv e s o f th e eastern S a in ts, éd. E. W. Brooks, dans P a tro lo g ia o r ie n ta ls 18. (Turnhout, 1974), p. 576. Cet exemple révèle le souci d’ennoblissement plus que d’in­ vestissement chez les marchands enrichis. 139 G regorii e p isco p i T u ro n en sis lib r i h isto ria n u m decerti, éd. B. Krusch et W. Levison (Ha­ novre, 1951) (M on u m en ta G erm an ica H istó ric a S crip to res R e n a n M ero vin g ica ru m 1) 10, 26.

LES CONDITIONS DU COMMERCE AU VIe SIÈCLE

113

dépendante.140 Les seuls cas connus d’action collective concernent les regroupements de marchands étrangers dans une ville, uniquement pour s’adresser à l’évêque du lieu, comme à Mérida,141 ou les commu­ nautés juives.142 Les marchands de Verdun, qui se portent garants d’un emprunt et touchent les intérêts de la somme prêtée par le roi, n’ont en commun que leur participation à une opération financière.143 Rien ne dit qu’ils soient associés dans leur travail. En outre, l’absence de système bancaire leur interdisait d’emprunter.144 Quand ils connais­ saient de graves difficultés, ils s’en remettaient à des particuliers ou à l’Église.145 Leur position sociale était en général particulièrement frag­ ile. Ds dépendaient des nobles et des évêques pour les transports publics et la vente des produits de luxe qui constituaient une grande partie de leurs activités, pour la défense de leurs intérêts, puisqu’ils étaient sou­ mis aux règlements municipaux146 et qu’ils devaient solliciter l’inter­ vention de l’évêque, chef de la cité, dans les moments difficiles.147 En­ fin, ils étaient soumis à de nombreuses contraintes sans disposer d’une législation qui défende leurs intérêts.148 Dans ces conditions, ils ne disposaient ni de correspondants capables de suivre leurs affaires loin de chez eux ou de prospecter les marchés,

éd. A. Papadopoulo-Kerameus, S b o m ik grech eskikh n à zd a n n y c h (Saint-Pétersbourg, 1909), c. 5: un négociant de Chios vi­ ent souvent à Constantinople; c. 27: un marchand vit dans le sanctuaire du saint avant de partir pour la Gaule; c. 32: un marchand de vin blessé pendant une tempête est soigné par un médecin; c. 35: un Rhodien, propriétaire ‘de son propre bateau’ reste dans le sanctuaire pendant que ses enfants continuent à commercer. 141 El lib ro d e la v id a s de lo s sa n to sp a d res de M é r id a , (Mérida, 1988), 4, 3, 1-3. 142 Claude, D e r H a n d e l im w estlich en M itte lm e e r w ä h ren d des F rü h m ittela lters, pp. 187-190. 143 G regorii ep isco p i T u ro n en sis lib r i h isto ria n u m decem 3, 34. Nous avons ici un exemple de marchands disposant de sommes importantes en monnaie. 144 Pour l’époque romaine, J. Andreau, L a vie fin a n c iè re d a n s U m on de ro m a in . L es m étiers d e m a n ieu rs d 9œ rgent (IV e siècle au. J - C A l t siècle a p . J - C f (Rome, 1987). Pour le Bas-Em­ pire, J. Andreau, “Declino e morte dei mestieri bancari nel Mediterraneo occidentale”, S o cietà rom an a e Im p ero ta rd o a n tico , 1, éd. A. Giardina, pp. 601-16. Pour notre époque, ex­ emples de prêts entre particuliers, sans recours à des banquiers, Claude, “Aspekte des Binnenhandels im Merovingerreich auf Grund der Schriftquellen”, pp. 81-2. La n. précédente montre que les marchands non seulement finançaient eux-mêmes leurs opérations, mais disposaient parfois de liquidités. 145 Cf. n. 128. 146 En particulier, l’évêque contrôlait les poids et mesures, Festugière-Ryden, V ie de S ym éon le f o u e t v ie de J e a n d e C h ypre , c. 11. 147 Cf. n. 30. 148 Siems, H a n d e l u n d W u ch er im S p ieg elfrü h m itte la lte rlic h e r R ech tsq u ellen , cherche en vain une législation sur le commerce dans les lois germaniques. Elle était tout aussi absente du droit romain. Les échanges donnaient lieu à des règlements dont la L o i d es R h o d ien s nous a conservé un exemple, non à des lois. 140

M ir a c u li sa n c ii A rte m ii,

bogoslavU dkh tek sto v I V - X V bekov ,

114

JEAN DURLIAT

ni de marges pour acheter sans vendre, ni d’informations sur les ré­ gions qu’ils ne connaissaient pas directement.149 Tout cela explique la primauté de la demande sur l’offre. Les marchands ne partent pas ex­ porter une production locale qui cherche des clients, mais vont dans des lieux où ils sont sûrs de vendre. C’est pou5rquoi ils viennent dans les provinces avec des marchandises, et circulent sans monnaie.150 Dans les sources, ils partent toujours dans la même direction, comme le gallodromos, le capitaine égyptien qui se rendait constamment en Gaule151 ou comme les nombreux marchands qui sont bien connus dans le port où ils arrivent parce qu’ils font constamment le même tra­ jet, tels le juif qui vient chez des amis à Carthage,152 l’oriental Antiochus qui réside à Naples depuis longtemps153 ou les marchands qui sont connus à Mérida.154 Et, pour être sûrs de vendre leur cargaison au retour, ils vont chercher ce qui est demandé sur place. Le moteur du commerce n’est pas l’exportation des produits locaux, mais l’impor­ tation de ceux qui manquent sur place. D’ailleurs, ce que nous avons dit des faibles surplus, exclut le besoin d’exporter. La population se privait d’une partie de ce qu’elle aurait consommé pour obtenir des produits qui n’existaient pas sur place et qui étaient indispensables. Cela explique le faible usage de la monnaie. Les marins ont quelques pièces de cuivre dans leur poche pour leurs frais personnels et on les retrouve dans les fouilles,155 mais le capitaine n’a pas d’or car il n’en a pas besoin. D vendra tout pour acheter le maximum de ce qu’il écoul­ era sur le marché de sa région. Les Vénitiens échangeaient directement leur sel contre la nourriture dont la population avait besoin et désig­ 149 Claude, D e r H a n d e l im w estlich en M itte lm e e r w ä h ren d d es F rü h m ittela lters, pp. 216-17. 150 C odex T h eo d o sia n u s 9, 23, 1. 151 Festugière-Ryden, Vie de S ym éon le f o u e t vie d e J e a n d e C h ypre, c. 36. De même le constantìnopolìtain qui se rend en Gaule semble bien connaître la route (M ira c u li sa n c ii A rte m ii, éd. A. Papadopoulo-Kerameus, c. 27) et un marchand rhodien reste à Constan­ tinople dans le sanctuaire du saint pendant que ses fils conduisent son bateau et le visitent quand ils reviennent dans la capitale (M ira c u li sa n c ii A rte m ii, éd. A. Papadopoulo-Ker­ ameus, c. 35). Le marchand qui se rend miraculeusement en Bretagne s’arrête au retour chez un ami, dans la Pentapole (cf. n. 158). 152 D o c trin a la c o b i n u per b a p tiz a ti, éd. N. Bonwetsch, Abhandlungen der kgl. Gesell­ schaft d. Wissenschaften. Zu Gottingen, phil.-hist. K l. N. F. XII, (Berlin, 1910), 5, 20. Le juif ne sait s’il s’arrêtera en Afrique ou s’il ira jusqu’en Gaule, mais il est suffisamment connu à Carthage pour qu’il y soit venu souvent. Peut-être allait-il en Gaule uniquement quand il n’avait pas tout vendu en Afrique. 153 Procope, D e bello g oth ico, 5 (1), 8, 21. 154 E l lib ro de la s v id a s de los sa n to s p a d re s de M é r id a , 4, 3. 133 Nous avons vu que la monnaie d’or ne quitte guère la région où elle est frappée. Par contre, la monnaie de cuivre venue de tout l’Empire est présente, au moins dans les ports.

LES CONDITIONS DU COMMERCE AU VIe SIÈCLE

115

naient les blocs de sel comme de la moneta victuatis.156 Théodore de Constantinople échangeait toute sa cargaison contre de l’étain et du plomb en Bretagne.157 Le marchand à qui Jean l’Aumônier donna un bateau chargé de 20.000 muids de blé fit un voyage miraculeux et des affaires si extraordinaires, que les Bretons lui auraient proposé de le payer pour moitié en étain et pour moitié en monnaie car la famine était telle que le blé aurait valu 1 sou le muid.158 L’auteur introduit ce paiement en or uniquement pour insister sur la grandeur du miracle. Ensuite, l’étain se serait miraculeusement transformé en argent, ce qui revient à dire que le marchand a échangé une partie de son étain con­ tre de l’argent espagnol. La personne qui découvrit ce nouveau miracle était ‘un associé en affaires’ et ‘un ami’ qui lui demandait’ de l’étain. C ’est donc une personne chez qui il s’arrête régulièrement et qui at­ tend ce métal. Comme souvent, dans les récits hagiographiques, tout correspond strictement à la réalité, sauf les traits merveilleux, précisé­ ment pour les rendre acceptables par le lecteur. Le marchand n’est pas ailé en Bretagne pour vendre du blé, mais pour acheter de l’étain. Le but premier du commerce est donc de procurer les biens indis­ pensables qui manquent sur place, comme le sel, les produits miniers ou - pour les riches - les produits de luxe, non de vendre d’hypothé­ tiques “surplus”. Ce que le marchand emporte constitue un “fret d’al­ ler” destiné à payer ce qu’il veut ramener, pour satisfaire la demande qui est première par rapport à l’offre. Comme aujourd’hui, les achats et les ventes s’équilibraient à long terme, mais le statut du marçhand était radicalement différent. Au lieu d’être un agent actif qui oriente la production, en suscitant la demande et en suscitant l’offre, il était pure­ ment passif.159 Cette constatation conduit à s’interroger sur la signifi­ cation des trouvailles archéologiques. Les milliers d’amphores vinaires découvertes à Carthage ne révèlent sans doute pas un dynamisme par­ ticulier de la viticulture orientale, mais, à la fois, le goût des Carthag­ inois pour le vin oriental et le besoin, en Orient, de produits africains dont nous ignorons la nature. En outre, les marchands devaient ajouter quelques amphores, pour la clientèle portuaire, aux soieries, aux épices et autres biens inconnus, attendus dans toute l’Afrique. De même, la 156 Fridh, M a g ra A u r e k i C a ssio d o ri V arioru m lib r i X I I , 12, 24, 6. 157 Io a n n is m on ach i L ib e r d e m ira cu lis, cd. M. HofFerer, (Wurzbourg, 1884), c. 26. 158 Festugière-Ryden, V ie d e S ym éon le fo u e t vie d e J e a n d e C h ypre, c. 8. Commentaire de ce passage dans Durliat, D e la v ille a n tiq u e à la v ille b y za n tin e , p. 240149. Encore une fois le prix est excessif. 159 Cf. n. 124

116

JEAN DURLIAT

céramique africaine fut développée pour satisfaire des besoins locaux indéterminés, en la vendant à des fournisseurs, dispersés le long des côtes et des fleuves dans tout le monde méditerranéen, et qui concen­ traient les productions de l’intérieur. En effet, comme les amphores et la céramique ne s’éloignaient guère de l’eau, elles constituaient vraisemblablement un simple complément pour des cargaisons qui n’ont pas laissé de traces, comme le papyrus, les dattes, et même l’huile, sans doute transvasée dans des tonneaux puisqu’elle circulait sans que nous ayons trouvé les amphores, dans tout l’ouest et le nord de la Gaule.160 La persistance d’un commerce en Gaule, sans tessons à Marseille, conduit même à se demander si le tonneau n’est pas dev­ enu le contenant prépondérant, y compris sur la mer, dans le courant du Vile siècle. Mais nous entrons là dans des questions si difficiles que les réponses sont très hypothétiques.161 * * *

La réflexion sur les conditions du commerce rend compte des par­ ticularités aussi diverses que le faible usage de la monnaie dans les échanges, l’existence de prix uniformes sur toute l’étendue des Etats ou

160 Pour la circulation des ces produits en Gaule, voir le diplôme pour Corbie (n. 6). Exemples de cartes donnant la répartition des amphores et céramiques africaines dans S. Tortorella, “La ceramica da mensa africana dal IV al VII secolo D.C.”, dans S ocietà rom an a e Im pero ta rd o a n tico , 3, éd. A. Giardina, pp. 211-25; C. Pavolini, “La circolazione delle lucerne in terra sigillata africana”, dans S o cietà rom an a e Im pero ta rd o a n tico , 3, éd. A. Giardina, pp. 241-249; A. Carignani, “La distribuzione delle anfore africane tra III e VII secolo”, dans S ocietà rom an a e Im pero ta rd o a n tico , 3, éd. A. Giardina, pp. 273-7. Ainsi s’explique la contradiction majeure entre les données des textes et de l’archéologie. Les premiers attestent que les bateaux -sauf ceux de l’annone- transportaient plusieurs pro­ duits (Claude, D e r H a n d e l im w estlich en M itte lm e e r w ä h ren d des F rü h m ittela lters , pp. 227-9; cf. Ashburner, N o m o s R h ò d iò n n a u tik o s. T h e R h o d ia n S ea -L a w , 3, 22-23: le capitaine a le droit de compléter la cargaison si le premier marchand n’a pas de quoi remplir le bateau avec ses produits). La seconde, qui se fonde uniquement, par nécessité, sur les restes non périssables, donne l’impression de grands courants d’échanges pour un petit nombre de produits transportés en masse, bien que -pour donner un exemple imaginaire- 100.000 amphores pour un siècle ne représentent que 1.000 amphores par an, et 2 amphores par bateau, si on suppose qùe 250 bateaux seulement ne font que 2 trajets par an. O r c’est de loin inférieur au minimum absolu pour un grand port comme Carthage, ouvert quelque 250 jours par an. Cependant, les calculs sont très aléatoires car, d’une part, 100 tessons peuvent appartenir à 10 amphores, d’autre part, nous n’avons qu’une part infime des tessons enfouis dans le sol et, enfin, nous ignorons la proportion des bateaux chargés d’amphores orientales parmi tous ceux qui visitaient Carthage. 161 Seule, la connaissance intime de toutes les sources écrites et de tous les champs de fouille permettrait, si une telle entreprise est un jour réalisable, d’harmoniser les don­ nées de ces deux grandes catégories de documents.

LES CONDITIONS DU COMMERCE AU VIe SIÈCLE

117

la présence de grandes villes sans qu’on trouve la trace de grandes so­ ciétés commerciales. Elle fait apparaître une situation relativement uni­ forme dans un monde encore fortement marqué par la survivance du passé romain et aide à mettre de l’ordre dans les indications, trop sou­ vent anecdotiques, et rarement quantifiées, des sources sur le com­ merce au Vie siècle. Elle justifie, sans les modifier beaucoup, les con­ clusions auxquelles sont parvenus les historiens, après un examen ri­ goureux et exhaustif des sources écrites.162 Partout les relations entre les petites villes et les campagnes donnaient lieu à des échanges qui as­ suraient l’essentiel de l’approvisionnement pour les premières, et de la fourniture en produits artisanaux pour les secondes. Mais, dès que la taille des agglomérations dépassait un certain seuil, difficile à préciser, la part des livraisons publiques augmentait rapidement, jusqu’à représenter la grande majorité des denrées non périssables, dans les très grandes villes d’Orient et, au moins, à Rome. Partout, les greniers municipaux assuraient les soudures difficiles et apportaient une aide pendant les années de disette ou de famine, au point que les paysans d’alentour venaient chercher du pain en ville. Le commerce autre que local était limité aux régions maritimes, pour des produits de valeur marchande faible ou moyenne. Grégoire de Naziance dit formellement et de manière générale, au IVe siècle, ce que nous constatons à Édesse au Vie siècle: la ville de Césarée de Cappadoce ne peut rien importer des régions éloignées car elle se trouve à l’intérieur des terres; le com­ merce terrestre est impossible.163 Seuls les produits de luxe et ceux qui étaient indispensables, bien qu’on les trouve uniquement dans des lieux précis, comme le sel ou les métaux, circulaient à grande distance, sur terre ou sur mer. Les marchands ne jouaient qu’un rôle d’intermédi­ aires, sans susciter la création de centres de grande production artisa­ nale ou de vastes zones d’agriculture spéculative. C’est pourquoi ils n’étaient guère considérés et ont suivi, plus qu’ils ne les ont provoqués, les changements dans les goûts, et donc la nature de la demande, ap­ parus au Vile siècle, avec le déclin de l’idéal urbain, du genre de vie et des échanges qui lui étaient associés.

162 On a noté la convergence presque totale entre les conclusions de Claude, Der Handel im westlichen Mittelmeer während des Friihmittelalters, et celles qui sont proposées ici. 163 Grégoire de Na2Ìance, In laudem Basilii magni. Patrologia graeca 36. 541.

THE DESTINIES OF THE LATE ANTIQUE IT AT.1RSPOLITICO-ECONOMIC DEVELOPMENTS OF THE SIXTH CENTURY. Federico Marazzi

1. Sudden trauma and the “Langue-durée” “It is by now widely accepted that the Lombard invasion of 569 rep­ resents a hiatus. This is acknowledged in every general profile of the history of Italy. The importance of the Lombard stimulus to the his­ torical evolution of Italy is also equally taken for granted....therefore the erudite modem researcher....who is more indifferent to the nation­ alistic judgements which dominated during the age of the Risorgimen­ to, is always occupied more with the factual reconstruction of the po­ litical structure of the Lombard kingdom and with a minute and pre­ cise reconstruction of historical events (...) As well as the innovations of knowledge, the conjunction of archaeology and history also imposes the need to consider facts which might otherwise be taken for granted, which a long tradition has left in the background of dominant histo­ riographic views and scholastic traditions. This applies in particular to the autonomous dimension and to the “longue durée” of the history of the Germanic peoples, which have been overlooked in favour of a his­ torical perspective in which they appear only at the moment of the great invasions when they challenged the Imperial Roman structure. We consider today with greater clarity than previously, the fact that, while the conquest of Alboin represents certainly a hiatus in the history of Italy - in terms of the splitting of an unified territory, the settling of a dominant political power hostile to the imperial authority of Byzan­ tium, and the revolution of the judicial, administrative and fiscal sys­ tems - , it also represents, above all, a hiatus in the history of the Lom­ bards”.1

1 P. Cammarosano, “Tradizione, storiografia e storia dei Longobardi: un cenno in­ troduttivo”, Langobardm, eds. S. Gasparri and P. Cammarosano (Udine, 1990), pp.VIX X I, X XI.

120

FEDERICO MARAZZI

This was how Paolo Cammarosano synthesised the arrival and settle­ ment of the Lombards in Italy. The Lombards, he argued, represent a decisive moment of redefinition of the geo-political identity and cul­ ture of Italy. Recent research helps us to understand those aspects in which the dynamics of integration directly prevail upon those of coun­ teraction, in both the relationship between Roman and non-Roman el­ ements, and in the chronological articulation of the penetration of the Italian territory and of the settlement within its interior. Traditionally, however, research in these areas is largely directed towards the com­ prehension of the consequences for Italy, starting from the moment of that fateful crossing of the Julian Alps made by Alboin and his people in 568. The question of the evolution of the Lombard military con­ quest, for example, is largely considered as a means by which to un­ derstand the diffusion within the territory of new forms of settlement and government, and the definition of juridical characteristics of rela­ tionships between people, as well as eventual changes in the structure of production and of the exchange of goods. Few, however, have considered the extent to which the appearance of the Lombards has exaggerated characteristics which already be­ longed to the physionomy of late Roman and Gothic Italy. In essence, the Lombards were inserted dramatically into processes that had al­ ready started, although obviously the importance of the Lombard in­ fluence on the radical changes which were occurring must not be un­ derstated. The radical nature of the Lombards’ arrival in some regions of the peninsula should also cause us to reflect on the susceptibility of these regions to this infiltration, and to the establishment of population nuclei and politico-territorial rule which differed from previous expe­ rience of Romano-German cohabitation in Italy during the fifth and sixth centuries (during the dominion of Odoacer and the Ostrogothic rule). In summary, either, the “Scandinavian teeth” of the Lombards tore up the meat of an Italy that was offered as a homogeneously struc­ tured body, or else they managed to sink them in places where, for structural reasons, the resistance was minimal. This question has re­ cently been clearly posed for the first time by Paolo Delogu,2 who con­ siders the critical evaluation of the “Lombard phenomenon” to be nec­ essarily linked to the critical evaluation of the changes known in Italian society in the course of late Antiquity. These changes appear most 2 P. Delogu, “Longobardi e romani: altre congetture”, Langobàrdia, eds. Gasparri and Cammarosano, pp. 111-67, in particular p. 145

THE DESTINIES OF THE LATE ANTIQUE ITALÏES

121

markedly in the evolution of urban and rural settlement patterns (iato sensu, in the appearance and distribution of wealth). In this case it is worth reporting in full the principal assumption of the interpretation given by Delogu: Significant changes can be discerned in the material organisation of the towns and the houses. Until the end of the fourth century, the data is multi­ faceted and sometimes contradictory; the evidence of the deterioration of public buildings is countered by traces of new construction and sometimes sumptuous private building. However, at the end of the fourth century, the lament of Saint Ambrose on the cadaverous, semi-destroyed towns of the Po region, in spite of the intent with which it was written, marks, if not the crisis of antique urbanism, the beginning of an era in which the physionomy of the city changes markedly.3

The collected data - which has seen important additions in the years following Delogu’s publication in 1990 - demonstrates that the fifth century saw the clear assertion of two concomitant phenomena, which should not be either combined or confused with one another. On the one hand, there was a growth of the importance of some elements of the urban fabric and a contemporary overshadowing of the others, and on the other, a general impoverishment and a widespread simplifica­ tion of residential structures (in terms of architecture, planning, and re­ lation to the service infrastructure). Delogu continued: These transformations precede the Lombard invasion and singularly denote a deterioration of urban organisation already occurring in the late-Roman city.... Together they comprise a block of evidence, however heterogeneous and discontinuous, which suggests that prior to, and independently of, the Lombard invasions, the Italian towns underwent, particularly during the fifth and sixth centuries, a process of regression, the gravity of which varied from place to place. In the more serious cases this culminated in the cessation of habitation in various ways and periods, which often resulted in the reduction of the urban area, with abandonm ent and decline of previously populated quarters.

In general, one finds a degeneration of the gridded plan, the cessation of maintenance work, and more or less radical transformations of the residential buildings.4

3 Delogu, “Longobardi e romani: altre congetture”, p. 146. 4 Delogu, “Longobardi e romani: altre congetture”, pp. 147-8.

122

FEDERICO MARAZZI

As stated previously, overall deterioration, and structural and function­ al changes of setdements and\or their various parts, should be treated always as distinct categories, even if the first of the two appears to be the prevailing characteristic. The changes in Italian urban settlements between the fifth century and the first half of the sixth century were accompanied by an evolu­ tion of rural setdement, which showed a clear drop in the quality of architecture and construction and the same tendency to simplification in terms of function and plan which one finds in the urban environ­ ment. However, in the fifth century at least, there was also the more frequent occurrence of sites which show signs of a sometimes remark­ able opulence. Rural contexts demonstrate more clearly that the de­ cline of the material quality of setdements does not correspond to their abandonment, but instead is an indication of the transformation of ag­ riculture in the period of late Antiquity.5 In the main, independendy organised rural labour (even if often juridically dependent), was based on the farm unit, and linked, according to geographical context, to scattered settlements. The result was that the village formations re­ duced the preponderance of the villa element, underwent intensive growth and introduced a subsequent element of “rustication” of the countryside. A better understanding of late Antique material, and a reduction of the propensity to consider the villa and the slave planta­ tion (the success of which was limited in space as well as in time) as “the” model of success of antique agriculture, has brought a more bal­ anced valuation of the characteristics of late Antique Italian agricul­ ture. Nevertheless, during the first half of the sixth century, the rural settlement also shows signs of a decline of the material conditions of life. The matters examined above can therefore be summarised in the following points:

5 For a summary of the problem of the numbers of rural sites surviving in the period of late Antiquity and the development of their material culture, see T. Lewit, Agricultural production in the Roman Economy, AD 200-400 (Oxford, 1991); for a comprehensive analysis of the development of the structure of agricultural production in late Antique Italy, with particular emphasis on the central southern areas, see D. Vera, “Dalla ‘villa perfeta’ alia villa di Palladio”, Part 1 , Athenaeum 83/1 (1995), pp. 189-211; Part 2, Athenaeum 83/2 (1995), pp. 331-56.

THE DESTINIES OF THE LATE ANTIQUE IT ALIES

123

1) The transformation (although we could also say the disintegration), of the fabric of classical urban settlement in late Antique Italy, oc­ curred as of the fifth century. 2) This transformation is expressed both in changes in the hierarchy of space within the urban environment, and in a general decline both of the quality of private residential accommodation and of the general quality of the infrastructure. Although difficult to quantify, it is also a sign of a widespread urban demographic decline. 3) In rural areas, the same phenomenon of the material impoverish­ ment of settlement, appears to be more gradual and imperceptible. However, this does not appear to be connected to a macroscopic phe­ nomenon of depopulation. If the evidence is viewed in this way, it allows us to understand in prin­ ciple, that the arrival of the Lombards does not constitute an element which activated, but one that only deepened, these modifications to Italian settlement structure and the weakening of the economic re­ sources available for its maintenance. This underlines the need for a different historicisation of these modifications. In addition, they are misleading unless viewed through a magnifying glass which allows us a greater geographical precision.

2. Italian politico-economic anatomy of thefifth century: winners and losers We must turn back around a hundred and fifty years before 568 to briefly recap on the evolution of the political situation at the start of the fifth century, which I believe had serious repercussions for econom­ ic conditions in Italy, and therefore on the availability of resources for investment. The crisis of the winter of 406-407 was going to change profoundly the face of the western Roman Empire. The crossing of the Rhine and the invasion of Gaul by the Vandals, Alans, Burgundians and Sueves, determined the final loss of territorial integrity within the western prov­ inces. In the case of Britain the loss was total, while in Gaul and the Spanish peninsula it was more partial, and varied over a period of

124

FEDERICO MARAZZI

time.6 In the same years there was also a serious crisis in Italy, more precisely in the years 405-411, due the presence in the peninsula, firstly of the Ostrogoths under Radagaisus, defeated by Stilicho in 406, and then the Visigoths. The Visigothic assault on Rome was only one ep­ isode, albeit of enormous relevance from a political and psychological point of view, at the start of a long series of disturbances caused by their presence. The successive migration of the Visigoths into firsdy Spain (414), and then Aquitaine (418), established an enclave which could not be eliminated from Imperial territory, and one which the government was forced to come to terms with. In 429, the Vandals moved from Spain and transferred to Africa. They were granted offi­ cial recognition by the Imperial government in 435 by means of a trea­ ty which established King Genserie as an Imperial federate first in Nu­ midia and Mauretania, and then in 442 in Africa Proconsularis. From 440 the Vandals continuously raided Sicily until 468, after which for eight years the island was effectively under Vandal control.7 In addi­ tion, from 433, the internal regions of Illyricum, were occupied almost without interruption, firstly by the Huns and then by the Ostrogoths. To these facts must be added the reality of the, by now complete, division of the Empire, operating from 395 and widened after the elim­ ination of Stilicho in 408, appointed by Theodsius to the control of the two principes pueri, Honorius and Arcadius. This determined not only a separation of administration but also a recurring conflict, particularly in the zones of direct contact between the two partes imperii, such as Il­ lyricum.8 The Empire, however, did not submit passively to the great crises of the start of the fifth century, and those which followed. A political

6 On the situation in Gaul, see the work of J. Harries, Sidonius Apollinaris and thefall of Rome, AD 407-485 (Oxford, 1994), particularly the considerations in the summaries, pp. 243-52; I. Wood, The Merovingian Kingdoms 450-751 (London-New York, 1994), chap. 1; for the situation in the Iberian peninsula see, J. Arce, El ultimo siglo de la Hispania romana (Madrid, 1986). For a recent re-examinadon of the concurrence of polidcal and eco­ nomic-administrative causes of the ceding of the paw Occidentis see H. Montgomery, “The parting of the ways. Byzantium and Italy in the Fifth Century”, Aspects of late Antiquity, eds. L. Rydén and J.O . Rosenqvist, Swedish Research Institute in Istanbul, Transactions 4, (1993), pp. 11-20. 7 For the chronology, see S. Mazzarino, L’Impero Romano, II, pp. 794-807 and G. Zec­ chini, Aezio: l’ultima difesa dell’Occidente romano (Rome, 1983); for the events in Sicily, see R.J.A. Wilson, Sicily under the Roman Empire. The Archaeology of a Roman Province, 36 BC - AD 535 (Warminster, 1990), pp. 330-1. 8 S. Mazzarino, Stilicone. La crisi imperiale dopo Teodosio (Milan, 1990), 2, chap.II.

THE DESTINIES OF THE LATE ANTIQUE ITALIES

125

stance existed, which did not consider the entrance of the gentes to Im­ perial territory as something necessarily harmful, to be eradicated, but rather as something which, if appropriately managed, would bring both military and, indirectly, economic, advantages to the Empire, guaranteeing the presence of a new force of soldiers and labour.9 This was the political line of the great minds, who, following the disaster of Adrianople (378) and the start of an established presence of gentes inside imperial territory, had begun to conceive a firm but flexible strategy of containment of the movements and the placing of the barbarians with­ in the frontier. This policy was followed by Theodosius (died 395), Stilicho (died 408), Constantius III (died 421) and Aetius (died 454). In summary, Gaul and the Iberian peninsula were not lost overnight, and even Britain was the objective of Roman military expeditions up till the 440s. This was not however the only political line or always that which prevailed. From the moment that the Rhine limes (and also those of the Danube) were no longer secure, the Italian province found itself in the front line. This provoked reactions different from those which might be termed Stilicho-Aetian. There existed that which could be defined as a contrary faction, represented by parts of the Imperial fam­ ily, and certainly rooted close to the court bureaucracy and a good pàrt of the Italian senatorial class. Their attitudes oscillated between ex­ treme anti-barbarianism (which brought attempts at resolute action against the barbarians, almost always with unhappy results), and a siege mentality which produced the sensational result of the transferrai

9 The main axis of the Aetian policy had always been the control of the Gallic sector, with the objective of re-establishing the stability of the limes. In the internal areas of the Empire the policy of Aetius had remained that of circumscribing the pockets of barbarian setdement, through the technique of the foedus, which had also been die policy of Stilicho and of Constantius III. The Gallic policy of Aetius aimed to retain the Roman nature of Gaul, through the defusing of the conflict caused by the presence of the Germans and other external populations. The policy of integrating the catholic bishopric with the power of the state was utilised in this tense situation with the intention of stabilising the Roman nature of Gaul and allowing a diffusion of power, as opposed to attempting to break the external forces and adopting measures of extreme severity by which to confront the rural masses in revolt. It is interesting to consider in this sense (Zecchini, Aezio: l’ul­ tima difesa dell’Occidente romano, chap.DÍ), the different strategy followed by Aetius in the confrontations with the Burgundians settled around the Rhine frontier, who after their military defeat were massacred between 436 and 437 and later deported to Savoy, and the tactic of containment employed against the Visigoths at Tolosa between 436 and 439.

126

FEDERICO MARAZZI

of the imperial residence from Milan to Ravenna.10 At the same time, there was a tendency to look to the pars (Mentis, where the anti-barbar­ ian policy had produced considerable successes, as the central source of the defence of the romanitas. All these events had the effect of producing two results, which were closely interconnected, namely an expensive state of permanent war and a progressive erosion of fiscal income, which was caused by the loss of territory (temporary and or permanent) and by the impossibility for some provinces of finding sufficient resources to make payments to the central government, due to the immediate effects of war. To this situation, should also be added the fact that the evolution of the polit­ ical situation in the first twenty-five years of the fifth century had caused other vices already endemic in the fourth century to become chronic. These vices were inherent in the functioning of the collection and redistribution of resources. The late Roman fiscal system had in­ troduced the revolutionary new idea of transforming the tributum paid by the subiecü to the conquerors into a tax paid by all the citizens of the Empire in exchange for the services and the protection assured by the central government. This sort of “social pact” nevertheless pro­ voked some serious imbalances. To achieve the objectives mentioned, the late-Roman fiscal system developed the tendency (even if not al­ ways intentionally) of decreasing the autonomous control of the town over the fiscal yield from its dependent territory. Essentially, the State sought to direct the fruits of tax collection as much as possible towards the Pretorian Praefectura, (in money or in kind), to the end of being able to supply more rapidly the army, civil service and court. The burden of collection was eventually delegated to the curiales who also gained the responsibility of ensuring that the sum corresponded with the esti­ mate fixed by central government. However, the great landowners (frequently high civil servants, important military personnel or sena­ tors), exploited their position to evade the tax which was requested from every citizen, by seeking to avoid the checking of their taxable income in local offices, in order to obtain favourable conditions or reach agreements with central government. The tablet of Trinitapoli, in northern Puglia, edited by Andrea Giardina and Francesco Grelle. 10 L. Cracco Ruggini, “Milano da ‘metropoli’ degli Insubri a capitale deU’Impero: una vicende di mille anni”, Milano Capitale dell’Impero romano, 286-402 d C. (Milan, 1990), pp. 17-23, in particular pp. 21-2; V. Neri, “Verso Ravenna capitale: Roma, Ravenna e la residenze imperiali tardoantiche”, Storia di Ravenna, I, L’evo antico, ed. G. Susini (Venice, 1990), pp. 535-84, in particular pp. 535-40.

THE DESTINIES OF THE LATE ANTIQUE IT ALIES

127

demonstrates how, already in the last quarter of the fourth century, among the actores of the great landowners, there was widespread boy­ cotting of the functionaries responsible for the collection of the trib­ ute.11 The difficult political conditions in Italy during the first decades of the fifth century12 choked the regular flow of taxes towards the centre, and the imperial government found itself in a position of weakness, from which it was unable to impose the collection of taxes. The state, therefore, to guarantee an acceptable income in these emergency situ­ ations, accepted compromises that in some cases smacked of a surren­ der to private interests and to the privileged conditions of great landowners. In Italy, then, there were repeated measures for the remission of tax­ es in the areas hit by the events of war13 and there is the sensation that these progressively deepening habits, already in place in the preceding decades, tended to aid above all the great possessores, who were offered more favourable conditions for the payment of tax.14 These took the form of the possibility of paying the taxes in one instalment, as opposed to three payments in a year and, above all, the authorisation —granted ad personam - to deal with the problem of payment of taxes directly with the central offices of the Pretorian Prefect or the comes sacrarum largitionum, or via the provincial governors. This assumed the character of a privilege when granted ad personam, and generally applied to persons of a certain social standing. In fact, it was consent to organise the sum of payment to the fisc without the need to pass before the inspectio to calculate the effective validity and income of the single properties. This type of solution generally involved payments in cash rather than in kind. The system, which took the name of autopractorium or autopragia was condemned in a disposition sent to the west in 399 and to the east in

11 A. Giardina and F. Grelle, “La tavola di Trinitapoli: una nuova costituzione di Valendniano I”, MEFRA 95, (1983), pp. 249-303. 12 Synthesis in G.A. Cecconi, Governo imperiale ed élites dirigenti nell’Italia tardoantica. Problemi di storia politico-amministrativa (Como, 1994), in particular chap. VI. 13 Codex Theodosiams Theodosiani Libri XVI cum Constitutionibus Sirmondianis, eds. T. Mommsen, P. Meyer et al (Berlin, 1905) XI, 28, 4; 7; 12 (for Italy), and see also those cited by Wilson for Sicily. 14 Qn these problems see, S. Giglio, Il tardo impero d’Occidente e il suo Senato (Naples, 1990), in particular p. 119 and pp. 161-5.

128

FEDERICO MARAZZI

40915 and, in an inextricable link of cause and effect, fed the phenom­ enon of the patrocinium, as the commendatio to persons who negotiated more favourable conditions than normal, created a relationship with the tax authorities which could present more of an advantage for indi­ viduals of low status used to facing the impact of the fiscal burden only through their own efforts. In conclusion, leaders such as Aetius tried to maintain, where pos­ sible, the presence of non-Italians in the higher echelons of civil ad­

15 Respectively, Codex Theodosianus XI, 7, 15 and XI, 22, 4. The extent to which autopragia was established between the late fourth and the start of the fifth century has been variously interpreted. The “classical” point of view of Remondon, based essentially on the examination of the Egyptian documentation, which associated it with the defini­ tive surrender of the imperial government to the pressure of the senatorial aristocracy and the connected dissolution of civic control of the territory (“L’Egypte au V siècle de nôtre ère: les sources papyrologiques et leurs problèmes”, Atti dell’X l Congresso Intemazio­ nale di Papirologia (Milan, 1966), pp. 135-48) has been recendy reappraised by Giglio, Il tardo Impero d}Occidente e il suo senato, pp. 151-69, without however reference to the previous scholar. The more cautious position articulated by Gascou (“Les grands domains, la cité, l’état en Egypte byzantine”, Travaux et Mémoires du Centre de Recherche dyHistoire et Civilisation de Byzance, 9 (1985), pp. 1-89), - always based on the Egyptian data - seems rather to support a process of hegemonisation, on the part of the possessores, within the composition of local administrative structures, rather than an eradication of these structures. The result was therefore that the possessions of the great land owners became worlds closed to direct governmental control, and many setdements tended to transform from ordinary territorial centres into subordinate satellites of dominant complexes. The attention of Cerati, Caractère annonaire et assiette de Vimpòtfoncier au Bas-Empire, Paris 1975, is centred on the increasing habit of paying in cash the annonae between the late fourth century and the start of the fifth. This process shows two aspects, on the one hand demonstrating the restored credibility of the currency, but on the other, the fact that this was the chan­ nel by which it was easier for the great landowners to obtain favourable conditions in the calculation of the compulsory taxes and, contemporaneously, a lessening of the ob­ ligation to provide recruits for the army. However, the preference for gold coinage, on the part of the imperial administration, for the payment of taxes, would have continually increased the margins of advantage for those of means, who were more able to utilise this type of currency. Nevertheless, it should not be forgotten that we are using extreme­ ly fragmentary data to follow the development of this process during the course of the fifth century. For an overview of the favourable conditions that the potentes often man­ aged to obtain as regards payment of taxes in late Antiquity, see MacMullen, “Tax pres­ sure in the Roman Empire”, Latomus 46/4 (1987), pp. 737-54. The analysis of Lewitt, Agricultural Production in àie Roman Economy, pp. 71-83, is also interesting in this regard, re­ counting how the change in the fiscal strategy on the part of the imperial government at the start of the fifth century, which produced also the effect of the granting of “auto­ pragia”, was part of a more wide-ranging system, aimed at rendering more flexible the estimate of tax to be paid by all the contributors. This happened after a long period during the fourth century, during the course of which it was attempted to encompass the basis of taxable income as much as possible and maintain the stability of the parameters of the basic valuations, to the end of guaranteeing more secure sources of income. The price of this, however, was evidently extensive evasion, and hints of rebellion, if not open revolt, of the population against their fiscal obligations.

THE DESTINIES OF THE LATE ANTIQUE ITALIES

129

ministration, thus maintaining the bond of common interest between the province and the centre. This must have been accompanied by an effort to facilitate the transmission of financial resources towards the Pretorian Prefect, from which then depended the supply to the armies and the imperial administration in its various forms. However, as we have just seen, there also existed legislation of a con­ trary type, which tended instead to safeguard the great estates (includ­ ing the imperial estates) from the systemátic payment of fiscal contri­ butions. This contributed to the paralysis of the autonomous political initiative of the pars Occidentis and obstructed the redistribution, in the form of services and military defence, of the fiscal exaction. This leg­ islation was, moreover, expression of a tendency that, on a political lev­ el, was expressed in a more rigid defence of Jiomanitas. This determined the unmanageable nature, in the West, of a coherent relationship with the territories outside Italy, and pushed Italy into a policy of withdraw­ ing from them. This also depended on the fact that these views were shared by the greater part of the italian senatorial aristocracy, whose economic interests and political affiliations were mainly directed to­ wards the Mediterranean, gravitating around Sicily, Africa and the East.16 We are aware therefore, that the political orientation which envis­ aged a situation of active equilibrium regarding western affairs, which involved concessions to the gentes but retained the prestige and strength of the Empire in managing the contradictions of the Western provinc­ es, fundamentally failed in favour of a prevailing integrated barbarophobia, the impracticality of which resulted in a passive rather than an active policy.

16 The presence has been noted of something profoundly incongruous and almost unconsciously self-destructive in the line of conduct followed by the aristocracy of the late Antique Italian possessors, since the withdrawal of resources to the state, if it gave the undoubtful advantage of maintaining the private wealth of these persons, weakened the political framework of which they formed a conspicuous element. In substance, the diverging effects produced by the intertwining of conflicting political and economic in­ terests in the dominant social categories, can be seen as one of the determining compo­ nents of the structural changes of late-imperial Italian society (cf. P. Brown,' The World of late Antiquity (London, 1971), in particular pp. 34-40; P. Anderson, Dall’antichità alfeudal­ ismo (Milan, 1978) pp.87-90, and C.J. Wickham, “The other transition: from the ancient world to feudalism”, Past and Present 103 (1984), pp. 3-36, in particular pp. 15-16).

130

FEDERICO MARAZZI

3. Socio-economic effects on Italy of the political disarticulation of the West. In essence, the fiscal system introduced by Diocletian and perfected during the course of the fourth century had allowed the draining of immense resources towards the centre. Above all, it had introduced the fundamental institutional novelty of reducing the territory of Italy - with the exception of Rome, which enjoyed a special status —to the level of every other province of the Empire. This has been defined as the “provincialisation” of Italy, in which the peninsula lost its privi­ leged position in terms of reinvestment of resources.17 The stability of the frontier had maintained the stability of the resources which most importantly allowed the payment of the armies, which became pro­ gressively filled with elements supplied from non-Roman gentes. There was also maintained to some extent a promotional policy that, al­ though more imperial than civic, expended part of the resources on the territories that had produced them. In fifth-century Italy, the situation had changed. It is impossible to quantify the extent by which revenues were diminished, but certainly the damage must have been significant. In this situation, important and unresolved contradictions, inherent within in the system (fiscal eva­ sion; draft evasion; the weight of fiscal demands upon the contributors), caused a financial crisis which external factors brought to a critical threshold. In the fifth century, local reinvestment of resources entered into an irreversible crisis.18 As Peter Brown pointed out in a recent study,19 this was the result of an administrative and cultural turnaround. The cities progressively lost their power of negotiation with the centre, to the ad­ vantage of the peripheral branches of the central administration. They also lost their power of dialogue with the centre, w,hich became mo­ nopolised by the Episcopal authority, when present. With rare excep­

17 A. Giardina, “Le due Italie nella forma tarda dell’impero”, Sociatà Romana e Impero Tardoantico, ed. A. Giardina, I (Rome-Bari, 1986), pp. 1-36; Idem. “La formazione dell’Italia provinciale”, Storia di Roma, III/I, eds. A. Carandini, L. Cracco Ruggini and A. Giardina (Turin, 1993), pp. 50-68. 18 On the importance in the late Roman system, of the regular function of the fiscal levy in the sphere of the processes of redistribution of wealth, cf. the reflections of Wick­ ham, “The other transition: From the ancient world to feudalism’*pp. 9-19; idem, “L’Ita­ lia e l’alto medioevo”, Archeologia Medievale 15 (1988), pp. 105-124. 19 P. Brown, Potere e Cristianesimo nella tarda antichità (Rome-Bari, 1995), in particular pp. 113-69; the problem has also been lucidly and memorably examined by Anderson, Dall’antichità alfeudalismo, pp. 79-80.

THE DESTINIES OF THE LATE ANTIQUE ITALI ES

131

tions, there was no longer any real investment in the infrastructure which, already overstretched in the fourth century, became completely plethoric in the fifth, due to there being less money and no longer the intention to spend it in the same way.20 The structural change produced in Italy during the course of the dis­ turbances of the fifth century was that of the major increase in accu­ mulated and hoarded wealth (by the lay potentes, by the church, and by the central government), in proportion to the wealth in circulation. More precisely, for reasons which I have outlined, the quantity of wealth reinvested in the territory in the fashion of the Imperial Roman state, in various ways under the principate and dominate (e.g. the great programmes of public works and the maintenance of a bureaucracy large enough to redistribute locally a significant part of the costs of their upkeep),21 diminished to a point at which it was no longer rele­ vant. These “positive” elements remained functional only near the centres of power, whereas the only element which remained operative in local seats was that of tax collection which the State could not re­ nounce. Only the passing of the armies, in certain circumstances, could be, paradoxically, a means of redistributing the wealth that the cost of maintaining these same armies required. In this case, however, it was a “selective” redistribution, which benefited only those who were in­ volved in the production and related commerce of supplying the troops. Furthermore, this was a redistribution whose materialisation at a local level was governed by circumstance (the carrying out of a campaign of war), and was by no means a form of investment in the territory. 20 If this was the undoubted tendency, the consequences were not uniform, since the reactive capacity of single local communities must also be considered. This was also de­ termined, as I will allude to later, by wealth produced in other ways, not only by state reinvestment in loco, but by the contacts periodically established by community members with the central authority, and by the geographic location of the towns in respect of the development of the political situation. AU these different factors also caused in individual cases, variations in the level of conservation of the vitality of the ideology of civic pride. Cf. also C. Lepelley, “Un éloge nostalgique de la cité classique dans les “Variae” de Cassiodore”, Haut Moyen-Age, culture, éducation et société. Études offerts à Pierre Riché (La GarenneColombe, 1990), pp. 33-48, on the consequences and the development of the de-urban­ isation of Bruttimi at the start of the sixth century, and on the limits of the reactive ca­ pacity of the civic ideology to the crisis of the urban institutions. 21 In this sense, the growth of the importance of the bishops in the “representation” of local communities - linked with the carrying out of a series of civic functions within the towns, constituted an undoubted advantage for the imperial administration, in that it cut the costs of peripheral bureaucracy, which also caused a reduction of the job op­ portunities offered.

132

FEDERICO MARAZZI

Therefore, in summary, there was less wealth, but that which existed had few reasons or possibilities by which to circulate in Italy outside the circuit controlled by the potentes, who were at this time able to ac­ cumulate a higher percentage of the overall available wealth. The al­ most permanent state of instability that Italy knew from the death of Majorian (461) until the deposition of Romulus Augustulus (476) could have determined the shifting fortunes between persons and groups, but certainly did not allow any process of change to begin. This may have augmented the headlessness of Italian politics and the polarisation of its dynamics around persons and groups, rather than around coherent initiatives of the central government. Contemporaneously, the political interests of the more significant (and therefore economically stronger) Italian social categories, including the great churches of Rome and Ravenna and the Western Imperial court, became progressively at­ tracted towards the eastern half of the Mediterranean where romanitas was immune from the menaces occupying the West. In the fifth cen­ tury, in political-administrative terms, Roman Italy suffered from the draining of its resources and of withdrawal to the Mediterranean, hav­ ing lost its centuries-old function as the centre of aggregation of the West. 4. The Gothic period: counterbalances and accelerations of the politico-economic dynamics of thefifth century. The Gothic period attenuated the radical nature of these tendencies in three essential ways. Firstly, the largely peaceful phase in Italy that extended for around forty years from the seizure of power by Theoderic (493) until the start of the war with the Romans (535) and the contemporary political expansion towards Dalmatia, Noricum and Provence, constituted a restoration of the equilibrium in comparison with the recent past. Secondly, the actions of Theoderic were charac­ terised by a coherence and continuity that returned the rhythm of the fiscal yield to normal and allowed a policy of investments and inter­ ventions by the central power in the Italian territory, or at least the expression of a desire in this direction.22 Thirdly, and by no means last, the different structure of the army in the Gothic period must be recognised, in respect of that of the Roman army in the preceding hun­ 22 C. La Rocca, “Una prudente maschera ‘antiqua’. La politica edilizia di Teoderico”, Teoderico il Grande e i Goti d’Italia, (Spoleto, 1993), pp. 451-515.

THE DESTINIES OF THE LATE ANTIQUE ITALIES

133

dred years. Whether one opines to the traditional theories that suggest the Goths were effectively given the land according to the rules of the hospitalitas, or those proposed within recent years by Goffart,23 who sug­ gests rather that the Goths received a third of the sum collected in tax­ es, the Gothic army constituted a stable presence in the Italian territo­ ry, and one which was certainly more homogeneously established in relation to the finances of the kingdom. Nevertheless, the four decades of Gothic rule did not, and could not, modify the characteristics of the economic trend established during the first half of the fifth century. On the 30th October 552, Teias, the last king of the Goths, was killed in battle at the foot of Vesuvius, and it could be said that Italy was once again in the hands of the Roman Empire. But over which Italy, or rather which of the Italies, was Imperial control re-estab­ lished? One of the more interesting documents of the history of late Antique Italy relates exactly to the moment of pacification of Italy at the end of the Romano-Gothic war. This is the so-called Pragmatic Sanction written by Justinian at the request of Pope Vigilius,24 written not so much to reorganise the administration of the Italian territories (at least the regiones suburbicariae), as much as to reassert the basic rules without which any re-establishment of imperial authority would remain purely theoretical. O f the twenty six sections of this document (the first is of a general introductory character) at least fifteen are occupied in various ways with protection of the rights of property and a further six with the restoration of normality in the various aspects of the exactions of taxes. The two decades of war determined two factors:—a prolongedfragmentation of the geographical unity of Italy (as well as a break­ down of the territorial coherence of the provinces within it), concen­ trated in the northern and western Po area (in addition to north-west­ ern Tuscany), the areas where the Goths twice managed (in 541 and 552),25 to regroup their forces before renewing conflict with the Impe­ rial armies. The fragmentation according to these geographical lines was also the result of a reconquest of Italy by a romanitas brought from 23 W. Goffart, Barbarians and Romans, AD 418-584. The techniques of accommodation (Princeton, 1980), in particular chapters II and III. 24 Constitutio Pragmatica, Corpus Iuris Civilis, III, Novellen, eds. R. Schoell and G. Kroll (Hildesheim, 1993), App. VII, pp. 799-802. 25 Procopius. De Bello Gothico II, 30 and IV, 35.

134

FEDERICO MARAZZI

the east. This was therefore an impression of Italian romanitas, the cen­ tre of gravity of which, had by now entirely moved to the East, via the channels of Mediterranean connections. - a potential negation of the “romanitas” of Italy, implicit in the nature of the war. It was clearly impossible to conclude the clash with a renewed form of cohabitation between Goths and Romans. This factor was ex­ acerbated in the second period of the war (542-552), and was contem­ poraneously cause and effect of the “subversive” measures carried out by Totila towards the great properties (senatorial, ecclesiastical etc) and the colonii and slave workforce existing within them. The territorial fragmentation and the destruction of the social balance of the peninsula eradicated the two cardinal factors on which the fun­ damental equilibrium of Roman Italy was based: - the concentration of wealth in Roman, or largely Romanised hands; - the control of the territory and its resources through the fiscal levy. In reality, Totila’s aims had been pardy achieved in spite of and not­ withstanding his defeat. The weakening of the northern and western frontiers of Italy caused by the Frankish incursions, frequent from 537 onwards,26 rendered the Roman reconquest practically impossible in this sector. Moreover, the exacerbation of the Romano-Gothic conflict had caused the elimination or emigration of the senatorial aristocracy from Italy who, notwithstanding their conflicting relations with the central power, nevertheless possessed a deeply rooted culture and ex­ perience of government and controlled (and created) significant accu­ mulations of wealth, in the form of land, fixed possessions, and liquid capital. These two factors were moreover so combined as to be indis­ tinguishable. Furthermore, the last clause of the Sanctio Pragmatica (the twenty-seventh, Ut qui voluerint ad praesentiam imperatori navigare, non im26 The incursions of the Franks became an endemic occurence in the political reality of north-western Italy. These incursions, in some moments of the Gothic war, were se­ rious expeditions resulting in more or less prolonged phases of occupation of the north­ western part of the peninsula as far as Milan and Pavia (and sometimes reaching even the eastern regions of the Po valley), and contemporary with the moments of the highest military efforts of the Goths against the Romans. Following the invasions of the Lom­ bards, the north-western areas were repeatedly visited by the Frankish armies until around 590, in a political sphere in which expansionist Frankish aims against the Lom­ bards (or, at least, the wish for an hegemony upon them), existed contemporaneously with elements who shared interests with the Romans, who paid the Franks to keep the Lombards occupied on two fronts (cf.Wood, The Merovingian Kingdoms, pp. 164-9). The fact was, in any case, that the Romans were not in a position to establish stable control over the western Po region.

THE DESTINIES OF THE LATE ANTIQUE IT ALIES

135

pedianiui) does in fact presuppose that the Italian viri gloriossissimi ac mag­ nifici senatores were by this time coopted into the senate of Constantino­ ple and their incidental presence iri Italy constituted a temporary commoratio, pro reparandis possessionibus....cum dominis absentibus recreari possessiones aut competentem mereri culturam difficile sit.21 It should also be added that the Italian fiscal yield - notwithstanding that a praefectura praetorio per Italiam had been re-established - in the pre­ ceding twenty years when it had been possible for the Romans to or­ ganise and collect it, had been spent principally for reasons of war. In summary, in Italy, by 555, there was no longer expendable cap­ ital, by which I mean not just that expendable by the State, but also that which had been accumulated in a parasitic fashion by the potentes, which would have been then some help for Italy, despite how greatly it had hindered the formulation of a self-determined policy for the West in the crucial decades between the end of the fourth and the first half of the fifth century. Exceptions to this may be seen in the cases of the two Episcopal seats of Rome and Ravenna, which, having never ceased to function, are often found playing the role of political inter­ mediary in structuring the re-establishment of Imperial rule in Italy. These appear as autocthonous centres of administration, primarily as points of reception for the flow of people, by land and sea, activated by political- religious and economic interests stemming from their high institutional position within the Empire and the widespread range of their patrimonial interests. This was an interesting “anachronism” that in some way determined the character of the intellectual and material culture of Rome, Ravenna and their connected regions during the course of the seventh century.2728 The re-establishing of a widespread “autochthonous accumulation of capitals” would have necessitated recreating an aristocracy which

27 On this argument, the most exhaustive and up-to-date treatment is that of T.S.Brown, Gentlemen and Officers. Imperial Administration and Aristocratic Power in Byzantine It­ aly, AD. 554-800 (London, 1984), in particular chapter II. However, see also S.J.B. Barnish, “Transformation and survival in the western senatorial aristocracy, c. AD. 400700”, Papers of the British School at Rome 56 (1988), pp. 120-55, in particular pp. 149-55. 28 F.Marazzi, “Il conflitto fra Leone III Isaurico e il papato e il “definitivo” inizio del medioevo a Roma: un’ipotesi in discussione”, Papers of the British School at Rome 59 (1991), pp.231-57; idem, “Roma, il Lazio, il Mediterraneo: relazioni fra economia e po­ litica fra VII e IX secolo”, La Storia economica di Roma nell’Alto Medioeoo alla luci dei dati archeofogici, eds.' P. Delogu e L. Paroli (Florence, 1993), pp. 267-86.

136

FEDERICO MARAZZI

would be deeply involved in the execution of military functions and the “provincial” administration of an Italy reconnected to the Empire. The small, but significant, difference, in respect of the past, was that Italy would no longer be treated as an independent organism, but sim­ ply as one tier of an external system of government.

5. What to buy and how much to produce? Economic inferencesfrom some aspects of Italian material culture during thefifth and sixth centuries. Besides its catastrophist implications in terms of population and settlement, and of its extreme standpoint on the ground of overall es­ timates of productive processes in late Antiquity,29 few now question the substantial validity of Andrea Carandini’s suggestion that, starting from the third century, Italy was more importer than exporter of main­ stream goods (wine, oil) and goods of specialised production and serial manufacture (certain productions of ceramics for table and transporta­ tion), or at least no longer dominated the production of these types of goods.30 The presence or absence of imported goods implies two apparendy tautological considerations, which in reality differ slightly from one an­ other. Firstly, the purchase of these products presupposes the existence of economic means, for reasons which are peculiar for individual areas and periods. Secondly, however, the lack of purchases of imported goods, while it can certainly be the sign of limited economic means, can be considered also as the result of a different use of resources. I will therefore attempt to show the interrelation between these two fac­ tors, within the limits of this paper. It should be remembered that the Vandal invasion of Africa certain­ ly had an effect on the production or distribution of African red slip

29 See the recent evaluation of the discussion of Carandini’s synthesis and the em­ phasis on the peculiarity of the productive structure of late Antique Italy in Vera, “Dalla villa ‘perfecta’ alla villa di Palladio”, I, in particular pp.201-10. On the repercussions of the changes of the economic-productive structure on the relationship between man and environment in late Antiquity, see G. Traina. Ambiente and paesaggi di Roma antica, (Rome, 1990), in particular pp. 100-*l 09. 30 A. Carandini, L’ultima civiltà sepolta o del massimo oggetto desueto, secondo un archeologo, in Storia di Roma, a c. di A. Schiavone, voi. III/z, L’età tardoantica. I luoghi e le culture, eds. A. Carandini, L. Cracco Ruggini e A. Giardina, (Torino 1993), pp. 11-40.

THE DESTINIES OF THE LATE ANTIQUE ITALIES

137

ware, but not a radically disruptive one, rather affecting the state-di­ rected shipping of food to Rome.31 A range of forms existed which were subjected to modifications and restructuring, which appear to re­ flect “internal” changes in the evolution of taste. This can be con­ strued as a sign of an enduring vitality of production and appears to have been repeated in a significant way during the heart of the period of Vandal domination, between the middle and the end of the fifth century. The range of forms appears to demonstrate signs of recession only in the first quarter of the sixth century.32 However, what clearly changed, starting from the first half of the fifth century, was the way in which the African products were received in Italy. In this period one starts to define a process of differentiation of the distribution of African red slip ware which becomes very clear in the second half of the same century, during which it progressively ceased to penetrate the interior of much of the Italian peninsula. Regions affected included the Po region and those areas situated on lines of communication connect­ ed with the basin of the Po, particularly those on its northern side, in relation to areas that centred on Milan.33 It is interesting to note the 31 This indicates an initial prevalence of factors of Mediterranean economic integra­ tion over those of fragmenting politics, during the establishment of Vandal control on Africa. S. Tortorella, “African red slip ware: cronologia e distribuzione”, Colloquio in onore di John Hayes. Ceramica in Italia: VI-VII secolo, eds. E.Fentress and L.Sagui, Roma 11-13 maggio 1995, (in press). As will be seen in the following notes, many of the statistics discussed here, come from the lectures given at the conference held in Rome in honour of John Hayes in May 1995, the Acts of which are still in press. Again I am deeply grateful to Elizabeth Fentress and Lucia Sagui for allowing me to utilise this information prior to its publication. 32 Regarding this argument, much still needs to be done for an integrated reading of the data of the written and archaeological sources, for example that which concerns the change of the political position of the Vandal kingdom in the Mediterranean after the death of Genserie (477). For the details see, C. Panella, “Gli scambi nel Mediterra­ neo occidentale dal IV al VII secolo, dal punto di vista di alcune ‘merci’”, Hommes et richesses dans VEcnpxre Byzantin, I, IV-VII siècle (Paris, 1989), pp. 129-41; Idem, “Merci e scambi nel Mediterraneo tardoantico”, Storia di Roma, III/2, eds. A. Carandini, L. Cracco Ruggini, and A. Giardina, (Turin, 1993), pp. 613-97, in particular pp. 641-54; Tortarella, “African red slip ware: cronologia e distribuzione”, and M.Mackensen, “Centres of African Red Slip ware production in Tunisia from the late fifth to the seventh centu­ ry”, Colloquio in onore di John Hayes. Ceramica in Italia: VI - VII secolo. 33 In general, one can say that the Italian areas of major diffusion of ceramics imported from Africa and the East were Sardinia, Corsica, Sicily (above all the southern coast, but the picture is distorted by the small quantities of data available for other areas), both coasts of Calabria, Basilicata (with a significant presence of imported ma­ terial around the large villas of the inland area, cf. H. Di Giuseppe, “Insediamenti rurali della Basilicata interna tra la romanizzazione e l’età tardoantica: materiali per una tip­ ologia”, Epigrafia e territorio, 1995, in press), the coastal areas of Puglia and, in the interior, the territories of Canosa and Venosa, Naples, Rome, the coastal areas of Campania and

138

FEDERICO MARAZZI

almost total disappearance, during the fifth century, of African mate­ rial from the zone on the other side of Milan, which runs towards the Alpine passes and from there, into Gaul.34 It is also interesting to note that during the fifth century, in areas in which the diffusion of fine ta­ ble wares and amphora of African provenance is more conspicuous, they are associated with similar materiell of eastern provenance (from Asia Minor, Cyprus, Syria, Palestine and Egypt). Although “one finds in the West, a market already saturated with African products” this material sometimes appears in significant proportions.35 Similar con­ siderations are suggested by the patterns of distribution of lamps of Af­ rican and oriental production.36 Significandy, during the same period (second half of the fifth centu­ ry, first quarter of the sixth), in various areas of Italy, highly distinctive local productions appear, imitating those of Africa. They are distribut­ ed in well defined areas, and generally are associated with imports. However, in some cases (in the higher Po Lombard region - Milan and

Lazio, the coastal sites of Tuscia, Liguria, the coastal sites of Piceno, the area around Ravenna and the Po-Veneto plain along the coast and along the course of the Po (cf. P. Arthur, H.L. Patterson, “Ceramics and early Medieval central and southern Italy: ‘a pot­ ted history”’, La storia d’Italia nell’alto medioevo italiano (VI-X secolo) alla luce dell’archeologia, eds. R.Francovich and G.Noyé (Florence, 1994), pp. 409-41 passim; Tortorella, “African red slip ware: cronologia e distribuzione”, cit.; M.P. Lavizzari Pedrazzini, “Oggetti d’uso: produzione e commerci”, Milano Capitale dell’Impero Romano, 286-402 d.C. ed. G. Senachiesa (Milan, 1990) pp. 362-405. Particularly interesting is the data relevant to the Sabina, where, from the start of the fifth century, the ridge of mountains that separates the Tiber valley (in direct contact with Rome) from the Rieti valley (a typical Appenine internal basin) became the watershed that delimited the penetration of imported ceram­ ics, that continued to be present on the Rome side throughout the first half of the sixth century (P. Roberts & H. Patterson, “New light on dark age Sabina”, Colloquio in onore di John Hayes. Ceramica in Italia: VI - VII secolo; Sabina Tiberina remained an integrated part of the suburbanitas of Rome, linked to the consumption of the abnormal market of Rome, until at least the time of Theoderic (cf. F. Marazzi, “L’insediamento nel suburbio di Roma dal IV all’VIII secolo” ballettino dell’Istituto Storico Italiano per il Medio Eoo 94 (1988), pp. 256-313, and T. Leggio, “Forme di insediamento nella Sabina e nel Reatino” ibid.y 95 (1989), pp. 165-201). 34 See the case of Aosta: R. Mollo Mezzena, “Augusta Pretoria e il suo territorio”, Archeologo in valle d’Aosta dal Neolitico alla caduta dell’Impero Romano, 3500 a.C- V sec.d.C. (Quart-Aosta, 1988), pp. 63-138, in particular pp. 129-31; Mollo Mezzena, “Augusta Praetoria tardoantica. Viabilità e territorio”, Felic Temporis Reparatio, eds. G. Sena Chiesa and E. Arslan (Milan, 1992), pp. 273-320. 35 Panella, Gli scambi nel Mediterraneo Occidentale, pp. 135-6. 36 C. Pavolini, “Le Lucerne in Italia nel VI-VII secolo d.C.; alcuni contesti signifi­ cativi”, Colloquio in onore di John Hayes. Ceramica in Italia: VI - VII secolo.

THE DESTINIES OF THE LATE ANTIQUE ITALI ES

139

Brescia37 - in central west Emilia,38 in northern Campania and in southern Molise -ager Falernus and Venafro,39 and in the interior of Basilicata and in Cilento40), recent work has demonstrated that the pic­ ture is different, or at least that the locally manufactured material is preponderant in comparison with the imported material, particularly in the examples from rural areas. In some of the rural areas (Monte Massico - ager Falernus —and Calle di Tricarico - inner Basilicata) points of production have been identified. Interestingly, it has been shown (for example in Rome) that these ceramics which imitated terra sigillata, occupied a niche market for objects absent in the African pro­ duction, such as small, deep, table ware forms for soup, which were probably cheaper items, while for the more costly material such as the grand table wares, there is a monopoly of imports.41 The appearance in urban residential areas (such as Naples) of clibani, starting from the late fifth century, also seems worthy of note, for reasons which I shall return to. This sort of small portable oven used by housewives for the preparation of bread, was previously found only in rural environ­ ments.42 37 Fontana, “Le imitazioni delle sigillata africana e le ceramiche da mensa italiche tardo-antiche”, Colloquio in onore di John Hayes. Ceramica in Italia: VI - VII secolo. 38 S. Gelichi, N. Giordani, “Produzioni artigianali dai pozzi-deposito”, Il tesoro nel pozzo. Pozzi deposito e tesaurizzazioni nell’antica Emilia, ed. S. Gelichi (Modena, 1994), pp. 75-132. ^ ____ 39 Fontana,-tpPkiL and Arthur-Patterson, “Ceramics and early Medieval centrai and southern Italy: {a potted history’”. 40 Fontana, op. cit. and Di Giuseppe, “Insediamenti rurali della Basilicata interna tra la romanizzazione e l’età tardoantica: materiali per una tipologia”. 41 Interestingly the presence has been noticed at Carthage, starting from the third or fourth decade of the fifth century, of material imitating the African sigillata produced however, mainly in Sardinia and Sicily. Contemporaneously, in Italy, imports of socalled “cooking wares” from the same production areas as the African red slip ware, ceased. Therefore, the concurrence between the level of manufacture of the “cooking wares” in one of the areas of production, and the “drop” in imports of these wares in comparison with the continuing importation of African red slip, assumes transmarine di­ mensions, reinforcing the image of the permeability of coastal trade (Panella, “Gli scambi nel Mediterraneo occidentale dal IV al VII secolo, dal punto di visto di alcune ‘merci’”, pp. 134-5). Also significant in this sense is the identification of the commercial distribu­ tion networks of the so-called “late Gallic sigillata”, produced in Provence and in Langue­ doc, which enjoyed a widespread coastal distribution towards Catalonia, Lazio and Rome, crossing to Corsica and Elba, and Morocco, as well as diffusion towards Liguria via small-scale coastal trade (cf. Y. and J. Rigoir, “Les dérivées des sigillées palaeo-chrétiennes”, Colloquio in onore di John Hayes. Ceramica in Italia: VI - VII secolo). 42 P. Arthur, “Local pottery in Naplea and northern Campania in the sixth and sev­ enth century”, Colloquio in onore di John Hayes. Ceramica in Italia: VI - VII secolo; on the form in generati and its distribution, A.L. Cubberley, J.A. Lloyd, P.C. Roberts, “Testa and cli­ bani: the baking covers of classical Italy”, Papers of the British School at Rome 56, (1988), pp. 98-119.

140

FEDERICO MARAZZI

In actual fact, from a general point of view, African red slip wares (or other production line ceramics of Mediterranean origin), were proba­ bly not per se more expensive than other types of production. Indeed, the cost of their manufacture - if we exclude the work on particularly elaborate pieces - must have been competitive. This is confirmed by the fact that, notwithstanding the presence of local productions which were potentially competitive at the “lower” end of the market, the dis­ tribution of the African and Oriental sigillata, where it is present on a large scale, is not limited only to areas of luxury consumerism (al­ though a tendency towards this, should not be entirely ruled out). If, therefore, the circulation of African and Oriental terra sigillata (be­ sides that of amphorae) in the interior of peninsula and northern Italy, became sporadic in the fifth century, one could identify a structural change in the conditions of transport and receipt of the merchandise, which caused a radical differentiation of habits and/or the availability to the consumer, in comparison to the regions where these products are widely distributed. All this signifies is that there must have fre­ quently occurred situations in which the cost of transporting the ce­ ramics from the coast to the inland areas increased, as opposed to an increase in the cost of transporting them from Africa or the East to the coastal cities which received them. The problem is to understand to what extent this depended on either an objective impoverishment of the inland markets and on increasing difficulties at reaching them, or a fundamental change in the tastes and needs of those who supported these markets. I consider these two phenomena to be manifestations of a radical economic and social change in the inland regions, as I will clarify in section 7. In the sixth century the patterns of distribution already described were maintained. The nature of these patterns had been defined in the first half of the fifth century. The new factors, however, were a more marked quantitative reduction of imported goods, both from Af­ rica and the East, and a more evident geographic concentration of their distribution. In terms of quantity, the local productions (and therefore the local distribution patterns) now really appear to start to take precedence over the imports. This transformation starts to be­ come apparent in the Gothic period. At present it is still difficult to understand if this depends on a decrease in the productions at source or from a lesser degree of integration between main market areas,

THE DESTINIES OF THE LATE ANTIQUE IT ALIES

141

thereby giving a greater importance to short range distribution net­ works.41*43 An important point pertains to local products, which appear in the same affluent zones as the imported products. Up until now, these have been recognised and studied mainly in the central southern re­ gions of Italy. There appears to be a progressive departure, in terms of typology and production, from the forms of African red slip, and one sees the development of types of ceramics that, besides displaying familiar elements of the terra sigillata chiara (for example, the use of lids splattered with red-orange colour) demonstrates other profoundly dif­ ferent characteristics. These include a prevalence of closed forms, and the covering of the whole body of the pot (when present) in decorative motifs, frequendy grouped in bands variously placed on the vessel. The sixth-century material shows clear signs of an evolution in the use of ceramics and in aesthetic tastes. Within this evolution there are indications of simplification of the processes of production and decora­ tion. It is interesting that these products also belong to a Mediterra­ nean koinè, but of a more markedly eastern Mediterranean character, with recognisable regional variations.44 Therefore, although the present level of knowledge dictates caution, it appears that, while a basic picture persists of a number of areas of Italy remaining open to the dynamic of exchange of ordinary consum­ er products that we see operating during the first half of the fifth cen­ tury, there are interesting signs in the late sixth century of a further turn towards the Eastern Mediterranean as the main point of reference of this same dynamic. 41 Panella, Mera e scambi nel Mediterraneo tardoantico, pp. 648-54. With Wickham, L’Ita­ lia e l’alto medioevo, one can interpret this state of things as the fall-out, at this point, of the decline of income from provincial taxation. I believe, as I have explained, that the cause of this effect must be placed two or three generations before, and that it manifested itself between the end of the fifth and the start of the sixth century. It was the first “vis­ ible” consequence of the political fragmentation which became more marked with the stabilisation of the “Romano-barbaric” kingdoms. 44 The variations in production are reflected in the variety of products, which ap­ pears to be confirmed for example, by finds from the Abruzzo of “Orecchio type” pot­ tery, with its Egyptian parallels; those of Luni which can be compared with the insular or southern areas, and which are connected also to possible fluctuations in the importa­ tion of material from the East, and by comparisons of Calabrian and Sicilian material with contemporary finds from Greece and the Balkans (F.Cuteri, “Ceramica a bande rossa”, Colloquio in onore di J.Hayes. Ceramica in Italia: VI - VII secolo). For ceramics of the so-called “Crecchio type”, which at present is the best studied class of this type of pro­ duction, see Dall’Egitto, copto all’Abruzzo bizantino. I Bizantini in Abruzzo (sec. VI-VII), eds. A.R. Staffa and W. Pellegrini (Moschiano-Crecchio, 1993), pp. 45-48. The data from Crecchio is reinforced, in a certain sense, by the significant incidence of importations of terra sigillata from Asia Minor, until the disappearance of the settlement in the mid sev­ enth century (ibid. p. 31).

142

FEDERICO MARAZZI

6. Ghosts or people amongst the “semirutarum cadavera urbium”. It is difficult not to refer to the destiny of the urban centres of Italy between the fourth and sixth century in terms of “decline”. A work of around ten years ago, by Bryan Ward Perkins45 described and identi­ fied geographically within the Italian peninsula, the phases of the pro­ gressive drying up of the private and state EDepyeaia within the cities of Italy during late Antiquity. However, in recent years, the archaeo­ logical research and theorisation on the destiny of the late Antique city has started to propose models relative to the idea of a city that func­ tioned on different principles from those of the classical city. Once again, it is instructive to read Peter Brown, who succeeds in showing the profound structural changes of the balances of power in the cities of the Eastern Mediterranean between the fourth and fifth centuries, without making us lose sight for a moment of the teeming vitality of these urban centres.46 There is no doubt that the crisis of the munic­ ipal systems, the exacerbation of the bureaucratic centralism of the structure of the Empire, the rise in the rule of the bishops in terms of secular administration and the relationship between centre and periph­ ery, and the rise of the demands of military protection for the urban nuclei, all had profound effects on the redistribution of resources avail­ able for the life of the city. The central government, which was not in a position to restore the resources to the cities, restricted itself to in­ tervening (and in reality mainly in cases of emergency) to maintain de­ fences and other parts of the infrastructure essential for survival (such as the aqueducts). It was clearly able to invest also in the parts of the infrastructure essential for the functions of government, but it is evident that it was unable to do much more. Procopius of Caesarea’s De Aedificiis, which describes the point of view of Imperial intervention, is par­ ticularly instructive, conveying to us the image - vaguely disquieting at first sight - of cities as empty boxes of which the unique distinguishable element was really the container, that is to say the walls which defend­ ed it, the strength of which was maintained by the sovereign. The walls constitute —following the success or otherwise of the city within them - the sign of the existence of life of the city and therefore of the Imperial domination of the region held by the city. One must be wary

+5 B. Ward Perkins, From Classical Antiquity to the Middle Ages. Urban public building in Northern and Central Italy, AD 300-850 (Oxford, 1984). ■*® Brown, Potere e cristianesimo nella tarda antichità, (Roma-Bari 1995).

THE DESTINIES OF THE LATE ANTIQUE ITALIEN

143

of a pre-conceived notion of the stance of the Imperial government, as being ideologically against the autonomous vitality of the cities; instead, the imperial interventions were in principle directed to enable the re­ covery of this vitality. An example, again from Procopius,47 shows us how the completion of the wall and of the aqueduct of Caputvada in Byzacena had been the start of a new phase of the life of the local com­ munity: ‘And the rustics have thrown aside the plough and lead the existence of a community, no longer going the round of country tasks but living a city life’, the first distinguishing feature of which was that of oyopaÇetV. The citizens could spend time taking part in public af­ fairs in the forum, and dedicate themselves to the activities of com­ merce. In the new conditions of security and urban dignity, the citi­ zens of Caputvada could in this way finally ‘conduct all the other af­ fairs which pertain to the dignity of a city’.48 The interest of the Imperial government in aspects that were not solely,concerned with defence and the furnishing of water, indicates the exceptional nature of the city in question. This was the case at Justiniana Prima in Dardania,49 in the native land of Justinian, where the Em­ peror was involved not only in the construction of the walls and the aqueduct, but ¿liso in the construction of churches, buildings for local magistrates, porticoed streets, markets, fountains, streets, baths and shops. Similar comparable efforts can be seen in a few other cities, the prosperity and magnificence of which was directly connected to the demonstration of Imperial power. These include Constantinople —the manifestation of the present imperial power - and Rome —the mani­ festation of that which was past. For Rome, at least on paper, it was considered necessary, in 554, to instigate measures in defence of the monumental and intellectual patrimony (through the conferment of prebends to orators, grammarians, doctors and jurisprudents).50 In other circumstances, that is; in the situations which constitute the majority of cases, the attitude of the Imperial government must have been that of leaving the local community to a sort of “do it yourself’, with the difference, compared with the past, that the resources left in the local coffers by the prevailing legislative trends, were much smaller 47 Procopius, De Aedificiis VI. vi. 14-16. 48 See in this sense, the formularised analogy of the image of the ideal urban life represented in Cassiodorus, Vanae 31. (527) for the city of Bruttium, on which see Lepelley, “Un éloge nostalgique de la cité classique dans les Variae de Cassiodore”, cit. 49 Procopius, De Aedificiis IV. i. 20-25. 50 Constitutif) Pragmatica, cit., 22 and 25.

144

FEDERICO MARAZZI

than in the age of the Principate. In principle, this state of affairs did not necessarily damage the assets of the individual. Certainly it did not damage the assets of the local churches, which, from the fourth century, were instead progressively becoming beneficiaries of a fiscal regime which was generally more favourable towards them than it was towards the average imperial subject.51 In an uninterrupted rise be­ tween the fifth and sixth centuries, they were defined as privileged in­ struments of central power for the administration of popular consensus in the urban centres.52 The real economic dynamism of the local communities was the search for links directly to the central power on the part of the local leading figures and the polarised force of popular consensus under the control of the Episcopal seats. This was no longer counter-balanced in late Antiquity (for reasons at the same time political, institutional and cultural) by the possibility or the will to express a civic pride, the purpose of which was really the control of the urban institutions. The Empire had passed, with the political and legislative changes of the fourth century (which sealed long-standing transformations) from the Empire of the cities to the Empire of the provinces, of which the cities were no longer complete cells, but rather osmotic barriers between the pressure of the centre and that of the plebes. The establishment of ab­ solute Imperial power, the spread of Christianity and the placing of the figure of the Emperor on the summit of the ecclesia of the subjects, had put city and state into a relationship which was more confrontational than in the past. It was this, in a certain sense, which Santo Mazzarino provocatively termed the “démocratisation” of late Antique society.53 The theory behind this hypothesis is built on a concrete vision of the spatial distribution of the late Antique metropolitan centres - in the eastern Mediterranean in particular - which it has been possible to ex­ amine, bearing in mind the changing social and cultural conditions of the era. The “destructured” and teeming cities of Anatolia, Syria and North Africa of the fourth, fifth and sixth centuries, lived “metastasising” the Hippodamean urban fabric of Hellenism and the principate,

51 G. Barone-Adesi, “Il ruolo sociale dei patrimoni ecclesiastici nel Codice Teodosiano”, Bulletino dell’Istituto di Diritto Romano “Vittorio Scialoia”, 83 (1980), pp. 221-45; Idem. “D sistema giustìniano delle proprietà ecclesiastiche”, La proprietà e le proprietà, ed. E. Cor­ tese (Milan, 1988), pp. 75-120. 52 Brown, Potere e cristianesimo nella tarda antichità, at. 53 This concept was expressed, for the first time, in S. Mazzarino, Aspetti sociali del IV secolo, (Roma, 1951).

THE DESTINIES OF THE LATE ANTIQUE IT ALIES

145

polarising it in extreme ways, around the commercial quarters and the cult centres which were often placed in eccentric positions in respect of the traditional centre of the cities. This produced a change in the nature of urbanism, even when it did not involve the abandonment and ruin of the cardinal elements of the classical urban model: fora, monumental streets, theatres etc. This has been associated with the idea of a tout court decline which can no longer remain entirely unchal­ lenged.54 In this rich and fecund area of contradictions, the walls stand out clearly, because they “signify” the unique fact which was of true im­ portance for the geography of power: that of the defence of the exist­ ence of the city, which was in any case an essential nerve centre for the territorial co-ordination of the functioning of the Empire. The archaeologists who occupy themselves specifically with this as­ pect, have rightly emphasised the importance of the establishment of a Christianised city, or of Christian spaces inside the city,55 but it would be semantically incorrect to derive from this, the existence of an urbanism of the Christian city. This emerged, for example in the West, only when the ecclesiastical power did not merely act as a “fifth column” of political power but instead was obliged to autonomously establish its own techniques of territorial domination.56

54 C. Foss, Ephesus after antiquity: a late Antique, Byzantine and Turkish city (Cambridge, 1979); J.S. Crawford, The Byzantine shops at Sardis (Harvard, 1990); T.W. Potter, Towns in late Antiquity: Iol Caesarea and its context (Sheffield, 1995). General considerations in W.Liebeschuetz, “The end of the ancient city”, The City in Late Antiquity, ed. J. Rich (LondonNew York, 1992), pp. 1-48. More recently has appeared as a broad survey on the topic N. Christie - S.T. Loseby ed., Towns in Transition. Urban Evolution in Late Antiquity and the Early Middle Ages, (Aldershot, 1996). 55 P. Testini, G. Cantino Wataghin, L. Pani Ermini, “La cattedrale in Italia”, Actes du Xle Congrès International dArchéologie Chrétienne, I, (Rome-Vatican City, 1989), pp. 5-57.; L. Reekmans, “L’implantation monumentale chrétienne dans la paysage urbain de Rome de 300 à 850”, Actes du Xle Congrès International dArchéologie Chrétienne, II, (DATE), pp. 861-915; V. Saxer, “L’utilisation par la liturgie de l’espace urbain et suburbain: l’ex­ emple de Rome dans l’Antiquité et l’Haut Moyen-Age”, Actes du Xle Congrès International dArchéologie Chrétienne, II, pp.917-1033; U.M. Fasola, V. Fiocchi Nicolai, “Le necropoli durante la formazione della città cristiana ”, Actes du Xle Congrès International dArchéologie Chrétienne, II, pp. 1153-205; J. Guyon, “Roma, Emerge la città cristiana”, in Storia di Roma, III/2, (Turin, 1993), eds. A. Carandini, L. Cracco Ruggini, A. Giardina, pp. 53-68. 56 G. Tabacco, “La città vescovile nell’alto medioevo”, Modelli di città. Strutture efun­ zioni politiche, ed. P. Rossi (Turin, 1987), pp. 327-345; F. Marazzi, “Le‘città nuove’ pon­ tificie e l’insediamento laziale nel IX secolo”, La storia dell’alto medioevo italiano (V - Xsecolo) alla luce dell’archeologia, eds. R. Francovich and G. Noyé, (Florence, 1994), pp. 251-77.

146

FEDERICO MARAZZI

This metamorphosis or, to use an efficacious term of Lellia Ruggini,57 ‘pseudomorphosis’ of the ancient city, occurred in Italy, between the fifth and sixth centuries, in the context of the dramatic changes expe­ rienced by the western Empire after the collapse of the limes on the 31st December, 406. The crisis of the Italian city was contemporane­ ously, a crisis of identity and a grave crisis of material resources. How­ ever, the two crises, which certainly precipitated profound turmoil in a centuries-old order with its own established rules, should not be con­ fused with each other. The various parts of Italy reacted differently to the changed political conditions. In those areas denoted by the extreme rarefaction or the disappearance of imported consumer goods, the decline of the Roman “city of stone” appears earlier and more extensive, to the point of it being possible to hypothesise widespread phenomena of abandonment or disintegration of urban sites.58 The crisis of the cities in these cases - for example those of the Appenine interior of Abruzzo, Molise, Apulia, Campania59 or western Emilia and the Tuscan-Emilian Appe-

57 L. Cracco Ruggini, “La città romana dell’età imperiale”, Modelli di città. Strutture efunzioni politiche, ed. Rossi, pp. 127-52, in particular, pp. 146-50. I personally believe that, perhaps swayed by the examination of the case of Sicily, Cracco Ruggini has too resolutely maintained her position, stressing the generalised weakness of civic life of the late Antique city. 58 The starting point for the analysis of the problem is always Delogu, “Romani e Longobardi. Altre congetture”, in particular pp. 145-67. 59 A.R. Staffa, “L’Abruzzo fra tarda antichità e alto medioevo: le fonti archeolog­ iche”, Archeologia Medievale 19 (1992), pp. 789-853; A.R. Staffa et al., “Progetto Valle del Pescara. Terzo rapporto preliminare di attività (1991-1995)”, Archeologia Medievale 22 (1995), pp. 291-342; F. Grelle, G. Volpe, “Geografia economica e amministrativa della Puglia nella tarda antichità”, Culto e insediamenti micaelici neWItaìia Meridionalef a tarda anti­ chità e medioevo (Bari, 1994), pp. 15-81; F. Cambi, “Paesaggi di Etruria e di Puglia”, Storia di Roma, III/2, pp. 229-54, in particular pp. 243-49; R. Hodges, “Il declino e la caduta: San Vincenzo al Volturno”, Storia di Roma, III/2, pp. 225-78; R. Hodges and C. Wick­ ham, “Incastellamento and after: the Biferno Valley AD 600-1400”, in G. Barker, A Mediterranean Valley (Leicester, 1995); E. Migliarlo, “A proposto di CTh IX, 30, 1-5: al­ cune reflessioni sul paesaggio italico tardoandeo”, Archeologia Medievale 22 (1995), pp. 47586, which is very useful and precise in defining the presence of the great imperial saltus, as not along the Adriatic coast between Abruzzo and Apulia. However, in my opinion, it is mistaken in its mechanical transferal of the late Roman saltus into the gualdi of the Lombard period, since although by chance they appear in the same areas, they are gen­ erated by historical reasons of their own, and are separated by the rise of the internal frontiers between the Lombards and the Romans and between various Lombard poten­ tates, which I will consider in the last paragraph of this present contribution.

THE DESTINIES OF THE LATE ANTIQUE ITALIES

147

nine,60 parts of Piemonte61 and the sub-alpine valleys of the high Atesine region62 — was profound enough to overwhelm or oppress strongly even the dynamic elements peculiar to late Antique urbanism which we have already examined. As stated previously, this cannot be hastily interpreted as an overall impoverishment of the areas in ques­ tion or as a sign of total depopulation. However, it certainly indicates a structural modification of the relationship between these areas and the remaining parts of Italy, which was characterised by a strong ten­ dency to marginalisation (or isolation). The true picture of these areas in late Antiquity still needs to be fully defined by archaeology, perhaps largely because these areas lack the elements by which urban phenom­ ena are usually defined. However, where long distance commercial connections are still evi­ dent in the fifth and at the start of the sixth century (such as the areas of Lombàrdia and of Romagna for which there is sufficient data),63 an urban facies is apparent which shows clearer and more constant mate­ rial traces. The prevailing impression is that of a strong polarisation of building typologies. There are a few known great residences and cult buildings, that in terms of design, material (building techniques and materials uti­ 60 M. Catarsi dall’Aglio, I Longobardi in Emilia Occidentale (Sala Baganza-Parma, 1993) and S. Gelichi, “Pozzi-deposito e tesaurizzioni nell’antica Regio V ili Aemilia”, in Il tesoro nel pozzo, pp. 15-48. 61 C. La Rocca, “Fuit ciuitas prisco in tempore, trasformazione dei municipia abbandonati dell’Italia Occidentale nel secolo XI”, in La contessa Adelaide e la società del secolo XI\ in Segusium 32 (1992), pp. 103-40, in particular pp. 108-20; G. Cantino Wataghin, “L’edilizia abitava tardoantica e altomedievale nell’Italia nord occidentale. “Status quaestionis\ Edil­ izia residenziale tra V e Vili secolo (Seminari sul tardoantico e ValJtomedioevo in Italia centrosettentri­ onale, IV), ed. G.P. Brogiolo (Mantua, 1994), pp. 89-102. 62 L. Dal Ri and G. Rizzi, “L’edilizia residenziale in Alto Adige tra V e VIII secolo”, in Ediäzia residenziale tra V e Vili secolo, pp. 135-48. 63 G.P. Brogiolo, “L’Edilizia residenziale tra V e V ili secolo: un’ introduzione”, Edil­ izia residenziale tra V e Vili secolo, pp. 7-13, G.P. Brogiolo, “Edilizia residenziale in Lom­ bardia (V-VIH secolo)”, Edilizia residenziale tra V e Vili secolo, pp. 103-13; S. Gelichi, “L’edilizia residenziale in Romagna tra V-VIII secolo”, Edilizia residenziale tra V e Vili secolo, pp. 157-67. Similar situations to those found in Emilia-Romagna and Lombardy contexts are emerging from the extensive work carried out in the urban sites of the Abruzzo region by Andrea Staffa, such as Truentum and Pescara (“Scavi nel centro storico di Pescara, 1: primi elementi per una ricostrutzione dell’assetto antico ed altomedievale dell’abitato di Ostia Atemi”, Archeologia Medievale 18 (1991), pp. 201-367; idem., “Scavi medievali in Abruzzo”, Conferenza italiana di archeologia medievale 1995 (forthcoming), as well as the situation which has been demonstrated in Rome in the area of the Crypta Balbi and its environs (D. Manacorda, “Roma. I monumenti cadono in rovina”, Storia di Roma, III/2, pp. 93-104; idem, “Trasformazioni dell’abitato nel Campo Marzio: l’area della ‘Pqrticus Minucia’”, La Storia economica di Roma neWalto medioevo, pp. 31-52.

148

FEDERICO MARAZZI

lised) and infrastructure (service annexes and finishing of buildings) demonstrate luxury characteristics, although they also indulged in the practice (motivated by both cultural circumstances and expediency)64 of the use of spolia as a construction material. However, the transformation of the general character of ordinary habitation is also relevant. Wood appears to have been prevalent in the area of the Po plain, but also appears to have been widely utilised elsewhere in comparison with previous phases. Buildings using wood and stone and/or bricks (both mainly reutilised) which show a marked simplification of construction techniques also appear widely. The spe­ cialised organisation of labour of the Roman period changed to an or­ ganisation that was articulated principally through simplified restora­ tion and new construction. This wooden construction appeared main­ ly in forms that called for limited specialised skills acquired in a family or community atmosphere (rather than within a specialised workshop). This characteristic of ordinary late-antique construction is a sign of various phenomena. It suggests a changing relationship towards the city inherited from the preceding centuries, that appears both as ag­ gression towards the structures of its buildings and as a selective move­ ment against the functions (habitation, cult, entertainment, public service), which the buildings put out of use could no longer fulfil.65 This therefore, is a factor of the vitality of the urban organism and of its population, since it is indubitable that what is subtracted from one part is replaced in another. The cessation of the functioning of the service buildings that serve social functions which are very characteristic of the life of an urban community (theatres, baths, schools, basilicas, markets, libraries, non-Christian cult buildings, etc.), is also often marked by their substitution with different buildings that have similar functions (for example Christian churches which replace pagan cult buildings). However, on other occasions determined functions are abandoned en­ tirely. We must ask, albeit somewhat contentiously, whether urban life remained the same without the baths or the theatre and whether this indicates either a material impoverishment or alternatively, profoundly changed socio-cultural demands that transfer resources to some ends 64 See, for example, S. De Lachenal, Spolia, (Milano 1995). 65 On this aspect, C. La Rocca, “Public buildings and urban change in northern Italy in the early medieval period”, The City in Late Antiquity, ed. Rich, pp. 161-80. There remains still a critical, and still largely unresolved, point; that of the relationship between the incidence of spoliation, an already widespread phenomenom in the fifth century, and the types of réutilisation (structural rebuilding, decorative embellishment, the use for lime and cement, etc) of the different sorts of material.

THE DESTINIES OF THE LATE ANTIQUE ITALIES

149

rather than to others. As I have said, in the case of Italy, it appears extremely difficult to separate the two aspects. Nevertheless, there is a strong indication of something which could define a widespread “progressive weakness” within the overall metamorphic changes in the urban fabric. The buildings which show signs of reuse, in the majority of cases are in fact stripped of material and distorted in their internal divisions, but in a way that is, however, largely “passive” and weak in respect of the urban structure of the inherited city. We are faced there­ fore with occupation of buildings of which the preceding material structure is no longer used or usable in its entirety (whatever the rea­ sons), but that remain, in a more or less advanced state of ruin, still strongly conditioning the order of the new phase. That this was determined by widespread economic weakness and not by free choice, is verified by the fact that, when investments of im­ portance were undertaken (such as the work of the dominant classes, of the church and of the sovereign powers that came to control Italy between the fourth century and the first half of the sixth century) they led to a real restructuring of the urban spaces. The “classicism” or “anti-classicism” of this is also open to question, but the intention be­ hind it cannot be doubted. Limiting myself to only civil building, I shall discuss the generic cases of the great urban residences of Rome,66 of Ostia,67 and of late-antique Cagliari,68 which in some cases saw a complete restructuring of the pre-existing fabric (even of public buildings in some cases in Rome) ac­ cording to a defined project, sometimes going as far as modifying the asset of a high-quality road network. Presumably, analogous situations can be seen at Ravenna (the so-called “Palace of Theoderic”)69 and at 66 F. Guidobaldi, “L’edilizia abitava unifamiliare nella Roman tardoantica”, Società Romana e Impero tardoantica, pp. 165-238; idem, “Roma. D tessuto abitavo, le ‘domus’ e I ‘tituli’”, Storia di Roma, III/2, pp. 69-83; C. Pavolini et ai, “Il Celio. Ascesa e caduta di un quartiere delFantica Roma”, Archeo 11/3 (1996), pp. 34-41. Regarding Guidobaldi’s idea of a defìntive crisis of intensively inhabited buildings in fourth-century Rome, in favour of a prevalence of single family unit buildings, I believe that the present state of knowledge does not yet permit generalisations of this kind. What is certain, however, is that the expansion, between the fourth and fifth centuries, of luxury single family houses (the great domus) is a sign that the privately wealthy, could impose their will on the urban fabric of the city with a freedom comparable to the late republican period. 67 C. Pavolini, “L’edilizia commerciale e l’edilizia abitava nel contesto di Ostia tar­ doantica”, Società Romana e Impero Tardoantica, II, pp. 239-97. 68 M.A. Mongiu, “Cagliari e la sua conurbazione tra tardoantico e alto medioevo”, Il surburbio delle città in Sardegna: persistenze e trasformazioni (Convegni di studio sull’archeologia tardoromana e altomedievale in Sardegna, III), (Taranto, 1989), pp. 89-124. 69 Gelichi L’edilizia residenziale in Romagna, cit

150

FEDERICO MARAZZI

Milan (the domus of piazza Duomo).70 If we then consider the work of urban remodelling determined by the foundation of important Chris­ tian cult buildings (bearing in mind the obvious example of Rome, but also Luni, Aquileia, Ravenna, Naples, Paestum, Porto Torres) we can see situations analogous to those of the great cities of the East or of North Africa, where the definitive triumph of Christianity determined a forma urbis imposed around a centre different from that of the classical period.71 Unfortunately, the Italian archaeological data does not yet al­ low hypothesis on the evolution of commercial zones and infrastruc­ ture, which could suggest that the late Antique cities involved in these changes took the semblance of “madinas” ante litteram, normally asso­ ciated with the Arab period.72 In summary, the urban landscape of the part of Italy in which links with the Mediterranean commercial networks were still present and thriving between the fifth and the first half of the sixth century, allows us to identify two principal critical points: (i) the circulation of imported goods excludes demographic collapses and presupposes local market dynamics sufficiently active, in terms of retail business, to justify long distance transport and its cost; (ii) the diffusion of phenomena of “squatting” over large urban areas, and the limitation of real interventions of urban restructuring to the work of restricted social groups, politically at the summit of local soci­ ety, are unquestionable signs of both a cultural change and a marked concentration of financial resources for any sizeable enterprise in few hands. These circumstances may also have caused, as has been shown, an extreme simplification in the activity of building and therefore pre­ suppose a perceptible reduction of occupation of this environment.73 70 Brogiolo and Gelichi, op. cit., note 63. 71 For a synthesis of the problem see Potter, Towns in Late Antiquity: lot Caesarea and its Context, and E. Zanini, Introduzione all’archeologia bizantina, (Roma, 1995), in particular pp. 136-144. 72 Still hesitantly advanced by H. Kennedy, “From polis to medina: urban change in late Antique and early islamic Syria”, Past and Present 106 (1985), pp. 3-27, the apparendy “heretical” hypothesis, was restated more decisively in P. Pentz, The invisible conquest: The ontogenesis of sixth- and seventh- century Syria (Copenhagen, 1992), and in the synthesis of Potter Towns in Late Antiquity: Iol Caesarea and its Context, on the basis of intensive discussion of the subject between the late 1980s and early 1990s. 73 The spread in urban contexts of the so-called clibani- household bread ovens, which I have highlighted above, can probably be considered as a further element in the extension, in an urban environment, of self-sufficient sections of domestic economy

THE DESTINIES OF THE LATE ANTIQUE ITALIES

151

In areas such as central and southern Tuscany,74Basilicata and Cilen­ to,75 and vast areas of Sicily,76 the apparent reduction of the role of many cities (although few real urban excavations have been carried out) was counterbalanced by the importance of the agrarian economy, which continued to thrive, reflecting the expanded suburbanites of Rome (until the end of the fifth/beginning of the sixth century), and the direct connection to Mediterranean-wide exchange networks. In areas such as those mentioned above, the survival of urban sites is of a selective nature, associated with the vitality of rural settlement, whose peculiar­ ities are becoming more and more comprehensible due to the progress of investigations. The phenomenon of the great villas of the fourth and fifth centuries becomes easier to understand, as the typologies of the distinctive archaeological materials of late Antiquity become more re­ fined. From this material it is possible to determine the effective form of these settlements, which certainly decreased in number and under­ went a process of material regression, although at a much slower pace than that envisaged ten years ago. The great villas therefore follow the progressive languishing of the politico-economic significance of Italy.77 In some areas of northern Italy, which can be considered as suburbia of Milan and Aquileia, the dynamics of settlement outside the dominant city suggest mechanisms similar to those observed in the towns, but on a wider scale, and therefore more easily comprehended In these cases, the undeniable, although declining vitality of the dominant city in the course of the fifth century, determined a pseudomorphosis not only of

74 Cambi, Paesaggi d‘Etruria e di Puglia. 75 S.J. Bamish, “Pigs, plebians and potentes: Rome’s economic hinterland, c.350-600 AD”, Papers of the British School at Rome 55 (1987), pp. 157-85; H. Di Giuseppe, “Insedia­ menti rurali della Basilicata interna tra la romanizzazione e l’età tardoantica: materiali per una tipologia”, Epigraphia e territorio, 1995 (forthcoming); Marazzi, “L’insediamento nel suburbio di Roma tra IV and V ili secolo”. 76 L. Cracco Ruggini, “Sicilia III/IV secolo: il volto della non-città”, Kokakalos, 2829 (1982-1983), pp. 477-515; Wilson, Sicily under the Roman Empire. One notes in Sicily, according to the conclusions of Wilson, a city-country and coast-interior fusion, which was one of the elements that appeared to diminish in other areas of mainland Italy. 77 Lewit, Agricultural production in the Roman economy, and Vera, “Dalla “villa perfecta’ alla villa di Palladio”. In his examination of the great vólias of Sicily, Wilson considers that, in a picture of constant vitality of the rural settlement between the fourth and fifth cen­ tury, there was a major difficulty, for these complexes, in the maintaining of the levels of prosperity and magnificence evident in the fourth century. This is evident in the cases of the villas of Patti, Favara, and, apparently, also Piazza Armerina (Sicily under, the Roman Empire, pp. 204-6 and 334-46).

152

FEDERICO MARAZZI

the urban centre,78 but also of the surrounding territory which, inside the complex - and still partly elusive - dynamic of settlement contrac­ tion demonstrates that there still existed the possibility of a long term investment in the form of suburban and rural cult centres.79 7. From the “two Italies” of Diocletian to the two Italies of Totila: explicit and implicit geopolitics. Let us briefly recap on the discussion up to now. Within the frame­ work of the “provincialisation” of the peninsula between the end of the third and the start of the fourth century, two Italies became defined within the diocese of Italy, both under the title vicarius. One of these regions comprised the Po region, until the northern Marche, while the other was ‘a type of vast extension of the Urbs’.80 If it is accepted and understood that the peninsula and island provinces were consolidated under the control of the vicarius Urbis in order to facilitate the transpor­ tation of the urban annona to Rome,8182it is possible that the administra­ tive area of the Po (which extended effectively to Raetia and Noricum) was defined in order to create a buffer zone for the system of admin­ istration which saw Milan function as the western imperial seat, and northern Italy as the “backstreet” of the Rhenish and Danubian limit­ es?2 In addition to this administrative system, the changes of the first half of the fifth century resulted in a clear division of the peninsula be­ tween areas which, in the economic and political collapse during these

78 I shall not dwell here on the well-documented phenomena of the christianisation —and on its monumental aspects —of Milan and Aquileia: see Milano Capitale dell’Impero Romano, pp. 91-150 and 209-26. Analogous siuations have been proposed in at least two territorial contexts of southern Italy, that, however, do not stricdy depend on a dominant urban centre, namely Abruzzo (A.R. Staffa, “L’Abruzzo fra tarda antichità e alto medi­ oevo: le fonti archeologiche”, Archeologia Medievale 19 (1992), pp. 789-853) and Sicily (Wil­ son, Sicily under the Roman Empire). 79 M. Buora, “Continuità e discontinuità degli insediamenti in Aquileia”, Il territorio tra tardoantico e alto medioeoo. Metodi di indagine e risultati (Seminari sul tardoantico e l’alto medioeoo nell’area alpina e padana, III), eds. G.P. Brogiolo and L. Castelletti, (Florence, 1992), pp. 73-84; M. Sannazzaro, “Costituzione e sviluppo di centri religiosi cristiani nelle cam­ pagne lombarde: problemi topografici e archeologici”, Il temitonio tra tardo antico e altomedieoo, ac. di G. P. Brogido and L. Castelletti, Firenze 1992, pp. 61-72. 80 Giardina, “Le due Italie nella forma tarda dell’impero”, pp. 9-11. 81 Barnish, “Pigs, plebians and potentes: Rome’s economie hinterland, c.350-600 AD” Papers of the British School at Rome, 55 (1987), pp. 157-185. J. Durliat, De la oille antique à la oille byzantine. Le problème des subsistances (Rome, 1990). 82 L. Ciacco Ruggini, “Milano da ‘metropoli’ degli Insubri a capitale dell’Impero: una vicenda di mille anni”, in Milano capitale dell’ Impero Romano, pp. 17-23.

THE DESTINIES OF THE LATE ANTIQUE ITALIES

153

decades, managed to maintain urban (and sub-urban) vitality in terms of economy, production and Mediterranean-wide trading links, and ar­ eas, which instead experienced a more drastic impoverishment, imply­ ing a more tangible exclusion from the outside world. A number of other factors may also have had significant impact on the directions according to which these “further” late Antique two Ital­ ics were delineated during the fifth century. These included the relo­ cation of the court to Ravenna, the connection (in a subordinate posi­ tion) of the western branch of the Theodosian dynasty to the eastern one, and the prevailing orientation of a Romano-southern-Sicilian axis of the politico-economic and patrimonial interests of the great senato­ rial and aristocratic families. Although it is difficult to quantify the proportions of cause and ef­ fect, the marginalisation of the internal areas and the general collapse of urban settlement that we have seen would have caused a generalised economic recession in these same areas, as well as upheavals in the so­ cial composition of the populations and therefore changes in their hab­ its and tastes. The disappearance of imported ceramics from these ter­ ritories during the course of the fifth century, could in actual fact be the more clearly evident sign of this type of transformation. While there are inevitable margins of imprecision in this scheme, it is possible to say that the first of the two areas comprised the greater part of the Italian coastal regions (with the possible exception of tracts of the coast of central Tuscany), the principal routes linking the two coasts and the areas affected by them, significant portions of the coastal plains of Sicily, Sardinia, Calabria, Lucania, and southern Apulia, as well as penetrating the Po plain (principally on its northern half) along the axis of the Po, as far as Milan. The second area comprised the interior areas of Appenine Italy, sections of central and northern Tus­ cany, the more westerly regions of the Po plain, and the pre-Alpine valleys of Trentino/Alto Adige. The dynamics of the war between the Goths and Romans, the re­ conquest of Italy from the East, and the definition of the areas of Ro­ man resistance in the phases of retreat during the conflict, together with the transformation of north-west Italy into a chequerboard of frontiers, firstly between Goths and Franks, and later Romans and Franks, and finally the political dependence of Italy on the East, radi­ calised this territorial polarisation of the peninsula, which was defined during the course of the fifth century.

154

FEDERICO MARAZZI

In summary, the Italies defined between the fifth and sixth centuries constituted a mosaic of spaces, bearing little relation to the administra­ tive system of Italy, and no longer officially sanctioned by territorial subdivisions of the peninsula now reattached to the Roman Empire. Prior to a detailed examination of the political geography of the var­ ious phases of the Romano-Gothic war, through which to aid compre­ hension of the actual means of control and management utilised by the two contenders in the various parts of the Italian territory, I would like to briefly consider some data furnished by Procopius, relating to the final phases of the war. Procopius tells that the Goths who survived the battle of Morts Lactarius, sought refuge at Pavia,83 and we know that a series of cities in Lombardy, such as Crema, Brescia, Pavia and Verona still remained in the hands of groups of Goths for some years after 553.84 Procopius also explicitly states, that in the final phases of the war ‘some towns of Liguria, the Cottian Alps and most of Venetia’ had been paying tribute to the Franks; adding also that ‘the Goths in­ deed had a few fortresses left in Venetia, while the Romans held the coast towns; but the Franks had brought all the others under their sway’.85 Therefore, northern Italy, particularly its northern and west­ ern regions, was, and remained, a sort of “grey area” of highly fluctu­ ating dimensions, politically disjointed and difficult for the Roman armies to penetrate. While the Roman army did not stint in its at­ tempts to return these areas to imperial control, Totila, despite his de­ feat, had achieved the result of making visible a process, the roots of which stretched back at least a century. This process was the “deMediterraneanisation” of a part of Italy and its division along a broken and imperceptible line. An examination of the politics of the Franks on either side of the Alps, indicates that the mountain chain was by comparison a lesser barrier.

83 Procopius, Bell. Goth. rV,35. 84 Agathias, Hist. 1.1; 6;15;20; Greg.Tour., Hist.Franc. III. 32; IV. 9; Fredegarii, Chron. 11,62; III,44;50. North-western Tuscany was kept under Imperial control with difficulty; following the overthrow at Vesuvius, groups of Goths held Lucca again (Agat.,//úí.,I,16). 85 Procopius, Bell.Goth.,IV,34. We still lack a complete examination-from the Italian point of view-of the policy of the Franks towards Italy, although for a beginning in this direction see, P. Lamma, “Richerche sulla storia e la cultura del VI secolo, II, Oriente e Occidente nell’opera storica di Agazia”, Oriente e Occidente nell’alto medioevo. Studi storici sulle due civiltà (Padua, 1968), pp. 90-131.

THE DESTINIES OF THE LATE ANTIQUE IT ALIES

155

8. “Blitzkrieg” and tactical warfare of the Lombards in Italy I will conclude my contribution with some observations on the articu­ lation of the first phase of the Lombard penetration into Italy. Some years ago, in a synthesis of the rule of the Lombards86 Paolo Delogu opined that the invasion of Alboin, in its initial phase proceeded “ac­ cording to a regular itinerary from Savogna on the Isonzo, to Forum Iulii (Cividale), taking the Via Postumia until Verona; from there along the Via Gallica to reach Milan, where (the king) before the 3rd Sep­ tember 569, followed westwards into northern Piedmont. The towns situated on these roads - Ceneda, Treviso, Vicenza, Verona, Brescia, Bergamo, Milan —were occupied without meeting resistance.(...) By the start of 570 all the region between the Alps and the Po was con­ quered.” The first resistance of the Romans is vouchsafed between 570 and 572 at Pavia, and it concentrates henceforth, until around 590, around the cities of the middle Po, Brescello and Mantua. These areas along the Po, above all those south of the same river and closer to the Adriatic, are also those which show the major penetration of im­ ported Mediterranean ceramics during the late fifth and early sixth centuries. If the recent hypotheses are correct, the formation of autonomous Lombard nuclei in central-southern Italy, would have been generated by the settlement of warriors in these areas by the Romans for military purposes. These warriors initially operated in concord with the Impe­ rial government until the elimination of King Cleph in 5 72.87 The loss of control over them by the Imperial command could correspond to the phase in which the Lombard royalty acquired the capacity to co­ ordinate the actions of the Lombard gens in Italy, during the final phase of the reign of Authari and still more during that of Agilulf. It is how­ ever important to underline that not all these autonomous Lombard entities could be successful. Perugia - on the axis linking Ravenna and Rome - and Classe, for example, never gained any autonomy of ac­ tion, and in fact soon rescinded to Roman control. However, the des­ tinies of Spoleto and Benevento and of the Lombard groups stationed in the interior of Tuscany, from Lucchesia, to Casentino and to Mugel-

86 P. Delogu, “D regno longobardo”, Storia d’Italia, I, Longobardi e Bizantini, ed. G. Galasso (Turin, 1980), pp. 3-216, in particular pp. 13-14. 87 Delogu, “D regno longobardo”; S. Gasparri, “Il ducato e pricipato di Benevento”, Storia del Mezzogiorno, I I / 1, ed. G. Galasso and R. Romeo (Naples, 1988), pp. 83-146.

156

FEDERICO MARAZZI

lo, to the Chiana valley, were different, and progressively acquired a real expansive dynamism. In substance, the successes of the Lombard territorial expansions in Italy, in the course of the first twenty years of permanent occupation of the peninsula, appear to have been encouraged, other than by the military activism of the Lombards, by the yielding of sectors of “Ro­ man” Italy, thereby indicating structural weaknesses in the control of the territory. This yielding manifested itself in different ways. In the northern parts of the Po plain, evidently the reorganisation of Roman control of the territory never happened after the war with the Goths, or at least did not allow the preparation of real fronts of resistance. The south-central Appenine areas were more remote in respect of the belt of coastal centres in which the Roman society of this Italian prov­ ince of the Empire was concentrated, looking to its real political sum­ mit in the east. The Lombard presence could be considered in some way as a particularly brilliant chemical reagent which demonstrated these structural characteristics of late Roman Italy, whose differences had been heightened by the reannexation “from the East” to the world of the Empire. In both the geographical contexts which I have indicated, the weak­ ness of Roman territorial control was the result of nearly two hundred years of diminishing investment and of progressive distancing from participation in the world of exchange of products and people who were left living in the coastal cities. In any case it would be unrealistic to say that the intrusion of the Lombard element into the Italian panorama did not introduce forma­ tive elements of its own, which affected all ways of life within the areas that entered into the orbit of the new invaders. Clearly, the first of these was the realisation of a division of the pen­ insula that had been until then only tangible in extreme moments of the conflict between Goths and Romans. The frontiers on the interior of the peninsula became a fact. The exceptional results of recent ar­ chaeological research now permit a clear vision of the structural mod­ ifications in settlement geography determined by this, demonstrated in the formation of belts of fortified sites on both sides of the frontier, as well as belts of territory that could almost be defined as “no-mans land” in which the population was extremely dispersed. As we learn from the later documentation of the eighth and ninth centuries, great agglomerations of fiscal land, both royal and ducal, were frequently formed in these, areas, together with frequent tracts of forest or uncul­

THE DESTINIES OF THE LATE ANTIQUE ITALIES

157

tivated land (the gualdi), which we later find combined in the patrimo­ nies conferred on the great Benedictine monasteries founded and pro­ tected by the same sovereign authority.88 At the same time, it is now understood that, notwithstanding the substantial similarities between rural settlement forms on the two sides of the frontier, and their continuation within the framework of sites de­ fined in the late Roman period, these same frontiers created a basis for the growth of organised labour systems and extraction of tax revenues within the countryside, that demonstrate differing juridical character­ istics on the two sides of the frontier.89 Furthermore, the late sixth-early seventh-century occupation phases of the urban centres of northern Italy which rescinded to Lombard control, show developments in the form and appearance of settlement, which are not just the radicalisation of tendencies seen over the pre­ ceding century and a half, but the emergence of something new. In the words of Gian Pietro Brogiolo, in Lombard cities such as Milan and Brescia, “more disruptive phenomena manifest themselves (in this phase): diffusion of very poor building types, accompanied by a partial ruralisation of the urban centres”. In urban environments, we see the appearance of building types which in the Roman period are encoun­ tered only in marginal areas, such as houses of earth and wood, with 88 For the Roman-Lombard frontier in Emilia, see S. Gelichi, “Pozzi-deposito e te­ saurizzazioni nell’antica Regio V ili Aemilia”, in II tesoro nel pozzo, pp. 15-48; idem., “Ter­ ritori di confine in età longobarda: l’ager mutinensis”, Città, castelli, campagne nei territori di frontiera (secoli VI-VII), pp. 135-48; for the Romano-Lombard “times” in Liguria, see N. Christie “The times bizantino reviewed: the defence of Liguria, A.D. 568-643”, Rivista di Studi Liguri, 55 (1989), pp. 5-38. for the Roman-Lombard frontier in Abruzzo, A.R. Staffa, “Una terra di frontiera: Abruzzo e Molise fra VI e VII secolo”, Città, castelli, cam­ pagne nei territori di frontiera (secoli VI-VII), pp. 187-238; idem, “Un quadro di referimento per Castel Trosino; presenze longobarde fra Marche ed Abruzzo”, La necropoli longobarda di Castel Trosino: Bizantini e Longobardi nelle Marche (Milan, 1995), pp. 93-124; on defining the frontier in Tuscany, W. Kurze, C. Citter, “La Toscana”, Città, castelli, campagne nei territori difrontiera (secoli VI-VII), pp. 159-86; for the Sabina, T. Leggio, “Forme di insedi­ amento nella Sabina a nel Reatino. Alcune considerazioni”, Bulletino dell Istituto Storico Italiano Medio Evo 95 (1989), pp. 165-201. On the siting of the gualdi in respect of the Adriatic frontier between Benevento and Spoleto, J.M. Martin, La Pouille du VI au XII siede {Rome, 1993), in particular pp. 177-9, 183-5 and 194-6; F. Marazzi, “San Vincenzo al Volturno tra VIII and IX secolo: il percorso della grande crescita (un’indagine com­ parativa con le altre grandi fondazioni benedettine italiane)”, San Vincenzo al Volturno. Cul­ tura istituzioni, economia (VIII-XII secolo), ed. F. Marazzi (Abbey of Montecassino, 1996), op. 38-90; on the gualdi in the territory of Spoleto, E. Migliano, “A proposito di CTh IX, 30, 1-5” e A. Baroni, “Problemi di topografia agraria fra tarda antichità e alto me­ dioevo “Gualdus” nella documentazione farfense”, Athenaeum 82/2, (1994), pp. 437-458. 89 The ‘classic’ reference for this material is V. Fumagalli, Coloni e signori nell'Italia settentrionale (Bologna, 1978).

158

FEDERICO MARAZZI

horizontal beams inserted into the ground and half-buried huts, which have close parallels with northern European types. There were also structures of stone/brick and wood, established inside spaces formed by the decline and abandonment of previous buildings. These differed from previous re-occupations, that they were orientated and developed in a manner independent of the ruins of the pre-existing buildings.90 In this sense the cities of the “Roman” side of Italy demonstrate a stronger continuity from the preceding period (although the archaeo­ logical evidence is much more fragmentary). This is reflected also by the continued flow of importation of manufactured goods, from the east and from North Africa, throughout the sixth century and well into the seventh. However, we also see the results of the social transforma­ tions started by the abandonment of the peninsula by part of the late Roman ruling classes (and therefore the removal of the great capital which they commanded) which I have discussed previously. The dis­ appearance (at least until proof to the contrary) of the grand luxury residential buildings and the concurrent rise of ecclesiastical settlements to the status of undisputed protagonists of the urban scene, represents two sides of the same coin. The widespread appearance of burial intra urbem, in the case of Rome, dating from between the second half of the sixth and the first half of the seventh century, has recently been intelligently located with­ in this general framework.91 As well as being itself subversive in terms of hierarchy and use of space, within the classical city (such as for ex­ ample the burials in the baths of Caracalla and next to the Colosseum), this phenomenon could be effectively the sign of the transfer of power and civic socio-cultural pre-eminence to new hands, which in turn shaped a new geography. Similarly, excavations made next to the exedra of the Crypt of Balbus, also in Rome, related to a precise strati- < graphic sequence dating from between the end of the sixth and the start of the eighth century, have revealed an incredibly rich horizon of material culture. This included numerous waste products of high qual­ ity craftsmanship, relating to the production of bone, ivory and metal objects, reflecting the equally high level consumerism, represented by 90 Brogiolo, “L’edilizia residenziale tra V e V ili secolo: un introduzione”, cit., pp. 7-13, esp. p. 10. 91 R. Meneghini, R. Santageli Valenzani, “Sepolture intramuranee e paesaggio ur­ bano a Roma tra V e VII secolo”, La storia economica di Roma, pp. 89-111; idem, “Sepolture intramurane a Roma fra V e VII see. d.C.: aggiornamenti e considerazioni”, Archeobgia Medievale 22 (1995), pp. 283-90.

THE DESTINIES OF THE LATE ANTIQUE IT ALIES

159

the very rich presence of imported ceramics and by the widespread presence of coinage, both bronze and gold. The deposit is almost cer­ tainly connected to an adjacent monastery (San Lorenzo in Pallacinis) and, although chronologically slightly later than the period which con­ cerns us here,92 could be evidence of the hegemony of the ecclesiastical institutions over the Roman-Byzantine towns of Italy, in terms of con­ trol of the greater concentrations of wealth (and the possibility of its mobilisation), in this phase of redefinition of the social equilibrium which followed the decapitation of the old ruling class. Igitur cum rex Alboin (...) from the Julian Alps, partem Italiae contemplatiis est, to the moment of undertaking the descent into the Fruili plain and the occupation of that which the Romans could not manage to defend, he was unaware that he was establishing a division of Italy that poten­ tially already existed. In a certain sense, the great transformation of sixth century Italy made real the dialectic between the two-fold voca­ tions of the Italian peninsula to be both continental and Mediterra­ nean, the ground for which was, in some ways, laid by the transforma­ tions of the late Empire. From this point onwards, this dialectic would not be expressed from the position of an imperial hegemony, which was that of the ancient world, but from a less comfortable and some­ times painful position, driven by radically different interests and prior­ ities which were directed from elsewhere. [Translated by William Bowden]

92 L. Saguì, “Distribuzione a Roma: la Crypta Balbi”, Colloquio in onore di John Hayes', idem, “Crypta Balbi (Roma): conclusione delle indagini archeologiche nell’esedra del monumento romano. Relazione preliminare”, Archeologia Medievale 20 (1993), pp. 409-18.

EASTERN SPAIN IN THE SIXTH CENTURY IN THE LIGHT OF ARCHAEOLOGY Sonia Gutiérrez Lloret

Throughout Spain and the rest of Europe there has been an increasing interest in the transition from late Antiquity to the early Middle Ages. A sign of this has been the development of many studies on the more diverse materials and varied literary evidence of the period between the fifth and the tenth centuries. This interest has been expressed through a crucial and long-standing historiographic debate about the spontaneity and the specificity of the early Middle Ages, or more pre­ cisely about the continuity or discontinuity of this transition, which is clearly an ideological debate.1 Even when interpretations differ widely, the catastrophic approach is softened in favour of a slow and discontinuous transformation of the old political, social and economic structures, which is apparent in dif­ ferent forms and velocities in the various Mediterranean regions.2 Within this “lunghissima tarda antichità”,3 the sixth century lies at the culminant point in the transformation of the economic model that had 1 Some recent reflections on this subject are by R. Hodges and D. Whitehouse, Mo­ hammed, Charlemagne and the origins of Europe (London, 1983); C. Wickham, “The Other Transition: from Ancient World to feudalism” Past and Present 103 (1984), pp. 3-36 (= Studia histórica, Historia Medieval. VII, 1989, pp. 7-35, Spán. trans.); K. Randisborg, The First Millennium A.D. in Europe and the Mediterranean. An Archaeological Essay (Cambridge, 1989); A. Carandini, “L’ultima civiltà sepolta o del massimo oggetto desueto, secondo un archeologo” Storia di Roma, III/2 .1 luoghi e le culture, eds. A. Carandini, L. Cracco Rug­ gini and A. Giardina (Turin, 1993), pp. 11-38; P. Delogu, “La fine del mondo antico e l’inizio del medioevo: nuovi dati per un vecchio problema”, La storia dell'Alto medioevo ital­ iano (VI-X secoli) alla luce dell'Archeologia, eds. R. Francovich and G. Noyé (Florence, 1994), pp. 7-32, as well as “Considerazioni conclusive” by C. Wickham in this same volume, pp. 741-59. 2 R. Hodges and D. Whitehouse, “Maometto, Carlo Magno e altri”, Opus I I / 1, (1983), p. 256; C. Panella, “Merci e scambi nel Mediterraneo tardoanrico”, Storia di Ro­ ma, III/2, pp. 613-97; Delogu, “La fine del mondo antico e l’inizio del medioevo: nuovi dad per un vecchio problema”, p. 8. Cf., an exact opposite view that stresses the catastrophism of the ancient world can be seen in Carandini, “L’ultima civiltà sepolta o del massimo oggetto desueto, secondo un archeologo”, p. 12. 3 F. Marazzi, “Roma, il Lazio, il Mediterraneo: relazioni tra economia e politica dal VII al IX secolo”, in La Storia economica di Roma nell'Alto Medioevo alla luce dei recenti scavi archeologici eds. L. Paroli and P. Delogu (Florence, 1993), pp. 267-87.

Fig. 1. Spain in the sixth century - Byzantine possessions and principle settlements.

162 SONIA GUTIERREZ LLORET

EASTERN SPAIN IN THE SIXTH CENTURY

163

characterized the Mediterranean during the early Empire and which by the second century A.D. can be considered at a close.4 This was a market economy based on standardized forms of production, set up to satisfy a massive demand, through a “worldwide” distribution.5 Clearly the above-mentioned system underwent a substantial and uneven pro­ cess of destructurization, which culminated in the seventh century - at least in the case of the western Mediterranean —together with a de­ crease in demand and a crisis of the state (which had hitherto kept most of the goods in circulation through a system of public transport).6 The aim of this paper is neither to analyse this crisis nor in any way to present an overall synthesis of Spanish archaeological investigations into the sixth century. Instead, my purpose is to discuss the archaeo­ logical implications of transformation, and debate some of the issues that are presently at stake. Undoubtedly, the historiographic debate about the end of Antiquity has recently been enriched and enlivened by the use of archaeology as an historic source, particularly in the wake of the publication of the first serious and rigorous results. These gave new elements of understanding to an era whose history cannot be com­ prehended without the assistance of archaeological data. Although the kind of information provided by archaeological surveys is limited when it comes to a global historical understanding of the early Middle Ages, an archaeological explanation is relevant for understanding the various changes in the production of goods and their circulation, as well as for studying market evolutions.7 From this point of view, archaeological data help to clarify those aspects especially related to the role of trade, the social organization of territory, the relationship of town to coun­ tryside, and population transformation. 4 Panella, “Merci e scambi nel Mediterraneo tardoantico”, p. 614. 5 A. Carandini, “Il mondo della tarda antichità visto attraverso le merci” Società ro­ mana e impèro tardoantico —Le merci. Gli insediamenti ed. A. Giardina, III. (Bari, 1986), pp. 320, and in general the whole book; C. Panella, “ Gli scambi nel Mediterraneo occiden­ tale dal IV al VII secolo” Hommes et richesses dans l’Empire byzantin, IVe - Vile siècle (Paris, 1989), pp. 129-141; C. Wickham “Marx, Sherlock Holmes and the Late Roman Com­ merce” Journal of Roman Studies, 78 (1988), pp. 183-193; Panella, “Merci e scambi nel Mediterraneo tardoantico”, p. 617 and ibi. 6 Carandini, “L’ultima civiltà sepolta o del massimo oggetto desueto, secondo un ar­ cheologo”, p. 25 and Ibi.; Wickham, “Marx, Sherlock Holmes and the Late Roman Commerce”, pass.; H. Patterson, “Un aspetto dell’economia di Roma e della campagna romana nell’altomedioevo: l’evidenza della ceramica” La Storia economica di Roma nell’Alto Medioevo alla luce dà recenti scavi archeologia, eds. L. Paroli and P. Delogu (Florence, 1993), pp. 105-24. 7 C. Wickham, “L’Italia e l’alto medioevo”, Archaeologia medievale XV (1988), pp. 10524.

164

SONIA GUTIÉRREZ LLORET

With this in mind, two questions can be formulated, both concerned with trading in the Iberian peninsula during the sixth century. Firstly, what goods were imported, and secondly, where were they consumed or distributed? Through archaeology it is possible to ascertain the cir­ culation of only certain types of goods; those which have been pre­ served in stratified deposits. Undoubtedly, the most useful of these is pottery; more precisely amphorae (containers exclusively used for food transportation), cooking and table wares, oil-lamps, as well as other goods, often luxury items, which came from various parts of the Med­ iterranean.8 One of the first elements to bear in mind when analysing the prov­ enance of the ceramics imported into Hispania is their own economic heterogeneity. This is evident in the different commercial activities of the interior territories of the Peninsula on the one hand, and of those along the Mediterranean coast on the other.9Whereas in the sixth cen­ tury, internal market demand was covered by local fine wares (late Ter­ ra Sigillata Hispánico), and by those imported from the south Gallic re­ gions (the Narbonnensis productions) or derivatives of them,10 the mar­ kets along the Mediterranean coasts were clearly dominated by the Af­ rican production, which had begun in the second century A.D. and lasted until the seventh century.11 This predominance of trade can be seen for the most part in harbour cities, notably Tarraco, and in coast8 Panella, “Merci e scambi nel Mediterraneo tardoantico”, fn. 2, p. 617; X. Aquilué, “Comentaris entom a la presencia de les ceràmiques de producció africana a Tarraco”, Miscellània Arqueológica a Josep M. Recasene (Tarragona, 1992), pp. 25-33. 9 For a debate concerning the commercial heterogeneity in Hispania during the sixth century; L. Olmo, “El Reino visigodo de Toledo y los territorios bizantinos. Datos sobre la heterogeneidad de la Península Ibérica”, Coloquio Hispano Italiano de Arqueología Medieval (Granada, 1992), pp. 185-98; cf. S. Gutiérrez Lloret, “Production and trade of local and regional pottery in early medieval Spain (7th-9th centuries: the experience of the south­ east of the Iberian Peninsula”, Medieval Europe 92. Pre-printed Papers, voi. 5 (York, 1992), pp. 77-91 {Boletín de Arqueología Medieval 6, 1992, pp. 9-22). 10 See, among others, L. Caballero Zoreda and L. C. Juan Tovar “Terra Sigillata Hispánica Brûlante”, Empúries, 45-46 (1983-84), pp. 154-193; L. Caballero Zoreda, “Cerámicas de época visigoda y postvisigoda de las provincias de Cáceres, Madrid y Se­ govia”, Boletín de Arqueología Medieval 3 (1989), pp. 75-108, particularly p. 86 and fol.; M. Orfila, “Terra Sigillata Hispánica Tardía Meridional”, Archivo Español de Arqueología 66, (1993), pp. 125-47. 11 O f the abundant bibliography on African imported pottery, I will mention only a few of the most recent syntheses: S. Keay, Late Roman Amphorae in the Western Mediterranean. A typology and Economic Study: the Catalan Evidence, 2 vols (Oxford, 1984); X. Aquilué, Las cerámicas africanas de la ciudad romana de Baetulo (Hispania Tarraconensis), BAR Inter. Ser., 337 (Oxford, 1987); R. Járrega, Cerámicasfinas tardorromanasy del Mediterráneo oriental en España. Estado de la cuestión (Madrid, 1991); P. Reynolds, Settlement and Pottery in the Vinalopó Valley (Alicante, Spain), A. D. 400-700, BAR Inter. Ser., 588 (Oxford, 1993).

EASTERN SPAIN IN THE SIXTH CENTURY

165

al commercial centres with rural traditions. In Tarraco, African pottery from various middens of the fifth century constitutes about 75% of the total amount. This percentage is comparable to that for amphorae be­ lieved to be African, from the coastal setdements of Punta del Arenal en Jávea (Alicante).12 In the sixth century, the eastern coastal regions of the Iberian Pen­ insula continued to have strong trading relations with those of the cen­ tral Mediterranean. Archaeologically speaking, these trade connections are characterized by an almost total predominance of table wares, lamps, and amphorae, used mosdy for oil or for preserving fish from Tunisia.13 As well as this African domination of the market, in the lev­ els of the first half of the sixth century, there is also evidence of a lim­ ited presence of table wares of eastern origin (Cyprian, Egyptian, or Phocis),14 although always in small amounts and in urban environ-

12 X. Aquilué, “Comentaris entom a la presencia de les ceràmiques de producció africana a Tarraco”, p. 27; J. Bolufer, “Les céramiques tardanes importades (segles IV - VII de la N. E.) del jaciment romà de la Punta de l’Arenal” (Xàbia, Marina Alta), III Reunió d\'Arqueología Cristiana Hispánica (Barcelona, 1994), p. 375-89, especially p. 375. The total amount of imported African goods is comparable in all respects to that found in Italy (Carandini, “L’ultima civiltà sepolta o del massimo oggetto desueto, secondo un ar­ cheologo”, p. 19; Panella, “Merci e scambi nel Mediterraneo tardoantico”, p. 650, n. 148). 13 The African supremacy in the oil market in the first half of the sixth century is a well known fact in the eastern Tarragonensis region, where African amphorae represent more than 79% of the whole (S. Keay, “La importación de vino y aceite en la Tarra­ conense oriental en la antigüedad”, El vi a l’antiquitat, Economia,.producció i comerç al Mediterroni occidental, (Badalona, 1987), pp. 383-95). It can also be found at the coastal site of Punta de l’ilia en Cullerà, and in the same city of Valencia, due to the abundance of amphorae types Keay XXXV, XXXVI, LV, LXI and LXII. The only oil coming from the East, which can be documented by the amphorae Keay LIII, comes from Syria. Cf. I. Garcia and M. Rosselló, “Las ánforas tardorromanas de Punta de L’ilia de Cullerà”, Trabajos Varios del STR, 89, (Valencia, 1992), pp. 639-61 and J. Blasco et al, “Assaig de sintesi del panorama cerámic de la ciutad de Valencia a l’Antiguitat tardana”, III Reunió dArqueologia Cristiana Hispánica (Barcelona, 1994), pp. 357-73, especially p. 363. 14 For Phocean pottery on the Mediterranean coast of Hispania see R. Mendez, “Cerámica tipo Late Roman C'en Cartagena”, fyrenae 19-20 (1983-84), pp. 147-56; J. Ni­ eto, “Algunos datos sobre las importaciones de cerámica Phocean Red Slip en la Península Iberica”, Papers in Iberian Archaeology, eds. T.F.C. Blagg, R.F.I. Jones and S.J. Keay (Ox­ ford, 1984), pp. 540-51 and J. A. de la Sierra, “Cerámicas foceas de Barniz rojo (Late Roman C) en el vedle del Guadalquivir y estrecho de Gilbratar” III Reunió dArqueologia Cristiana Hispánica (Barcelona, 1994), pp. 391-5, as well as the aforementioned R. Járrega (1991). For Egyptian and oriental imports see R. Jarrega an d j. F. Clariana, “Ceramica xipriota i egipcia-B tardörromana a la comarca del Maresmc”, III Reunió dArqueologia Cristiana Hispánica, pp. 333-7.

166

SONIA GUTIÉRREZ LLORET

ments.15 This evidence is surely an indication of the relatively abun­ dant importation of eastern wines, especially from Palestine and the Black Sea.16 It testifies to the appeal of these wines to certain urban sectors, especially the newly rich religious milieus, rather than to a par­ ticularly intense and generalized level of trade with the eastern Medi­ terranean.17 There are also other oriental products that are well represented in the urban harbours and coastal setdements, as for example late unguen­ taria. By the sixth or the beginning of the seventh century, these were distributed throughout the various regions of the Levantine coast, more precisely at Punta de Tilla (Cullerà), Alicante and Carthagena.18 How­ ever, one must not overestimate the importance of these pieces as in15 The distribution of the Late Roman C in Hispania spans the period between the second half of the fifth c. and the first of the sixth, and is not widespread (Blasco et al, “Assaig de sintesi del panorama ceràmic de la ciutad de Valencia a PAntiguitat tardana”, p. 359). Its disappearance by the second half of the sixth century from certain areas of the Iberian Peninsula, for example the coastal part of Catalonia, has been linked by some authors to a supposed commercial crisis coinciding with the Byzantine conquest of the remaining part of the Iberian Peninsula (Nieto, “Algunos datos sobre las importaciones de cerámica Phocean Red Slip en la Península Ibérica”, p. 547; J. M. Nolla, “Excavaciones recientes en la citidadela de Roses, El edificio bajo-imperial”, Papers in Iberian Archaeology, BAR Inter. Ser., 193 (ii), (Oxford, 1984), pp. 430-64, 447). However, most recent works have showed that this inflection is shared by the whole eastern coast of Hispania, and that its commer­ cialization is related to the Vandal dominion of Africa rather than the Byzantine one (Panella, “Merci e scambi nel Mediterraneo tardoantico”, p. 645 and fol.). 16 The importation of eastern wine well into the sixth century can be found in the coastal site of Punta de Pilla in Cullerà, Valencia, with amphorae type Keay LTV and LXVI from Palestine and Keay LXV from the Black Sea (I. Garcia and M. Rosselló, 1992, “Las ánforas tardorromanas de Punta de Lilla de Cullerà”, p. 655), and also in the city of Valencia (Blasco et al, “Assaig de sintesi del panorama ceràmic de la ciutat de Valencia al’Antiguitat tardana”, p. 363); cf. with the imports of oriental amphorae in the second quarter of the fifth c. in the dumping grounds of Vila-roma of Tarragona (Un abocadof del segle Vd.C, en elfirum provincial deTàrraco, Taller Escola d’Arqueología (Tarra­ gona, 1989), p. 276 and fol.). For the Tarragona region in general see S. Keay, “La im­ portación de vino y aceite en la Tarraconense orientili en la antigüedad”, pp. 386-7. 17 One should not lose sight of the possible role of Carthage as a redistributor of eastern products coming from a trans-mediterranean trade (Panella, “Merci e scambi nel Mediterraneo tardoantico”, p. 649 and fol.). For this particular issue and trade in gen­ eral, see the interesting work by Jean Durliat in this volume. 18 For the late unguentaria distribution along the Hispanic coasts see I. Garcia and M. Rosselló, “Late Roman Unguentarium: ungüéntanos cristianos de la Antigüedad tardía procedentes de Punta de Pilla de Cullerà, Valencia”, Archivo Español de Arqueología 66, (1993), pp. 295-300.; P. Reynolds, Settlement and Pottery in the Vmalopó Valley (Alicante, Spain), A. D. 400-700, p. 144; M. C. Berrocal, “Late Roman Unguentarium in Carthago Nova (Cartagena)”, XXIII Congreso Nacional de Arqueología (Elche, 1995, pp. 119-28). Its Palestinian origin proposed by J. W. Hayes has been recently revised bÿ the same author, who has suggested a possible provenance from South Turkey (Pamphylia) in Colloquio in onore di John Hayes. Ceramica in Italia: VP VII secolo (Roma, maggio 1995, in press).

EASTERN SPAIN IN THE SIXTH CENTURY

167

dicators of a regular trade with the eastern Mediterranean, given the fact that their use seems to have been mostly liturgical (as containers for consecrated water as well as vinegar). Their religious use means that the distribution of these unguentaria is part of a specific long-dis­ tance trade that has nothing to do with foodstuffs or of other related goods. In fact, the distribution of these wares may have been in answer to religious demands. This can be inferred by their notable presence in Punta de l’Eia de Cullerà, a coastal monastic settlement.19 Although at the end of the fifth century and during the first half of the sixth century the overall production of the coastal areas appears to have been dominated by African trade, already the first symptoms of an increasing economic self-sufficiency were beginning to appear. Panella has interpreted this phenomenon, independently from the vol­ ume of trade or the degree of dependency on the outside world, as the predominance of local products over those imported.20 This general move towards a self-contained economy is, according to Panella, what characterizes the second half of the sixth century and basically the whole of the following century. It is not that there is a noticeable re­ duction in importation of table wares and Tunisian amphorae, whose presence was always very high in the coastal setdements during the whole of the sixth century, but rather that common African pottery de­ creases while there is a cautious increase of local production. With ref­ erence to this, Aquilué points out that in the important harbour of Tarraco, the volume of common African pottery went from 16.6% at the end of the first half of the fifth century A.D. to hardly 3.5% of the total amount by the beginning of the sixth.21 According to Aquilué, as re­ gards Tarraco, the complete disappearance of common African pottery in the sixth century must be connected with the increase of the locad small pottery, production of which finally replaced the imported prod­ ucts.22 19 For the monastic characteristics of this establishment see E. IJobregat “Las cruces de Punta de l’ilia (Cullerà)”, Trabajos varios del S.I.P., 89, (1992), pp. 663-70, especially p. 667, and M. Rosselló, “Punta de l’ilia en Cullerà (Valencia): un posible establecimiento monástico del siglo VI d.C.”, IV Reunió d1'Arqueología Cristiana Hispánica (Lisboa, 1992, pp. 151-62) 20 Panella, “Merci e scambi nel Mediterraneo tardoantico”, p. 615. 21 The first figure comes from the dumping ground of Vila-roma (A.D.440-50) and the second from the one in Sant Llorcnç (A.D.475-525) according to Aquilué, “Comentaris entorn a la presencia de les ceràmiques de producció africana a Tarraco”, pp. 244-46. 22 Aquilué, “Comentaris entom a la presencia de les ceràmiques de producció afri­ cana a Tarraco”, p. 28, n. 34 and 35; Taller Escola d’Arqueología, Un abocador del segle V d. C., en elfinan provincial de Tárraco, pp. 244-46.

168

SONIA GUTIÉRREZ LLORET

Like Tarraco, in many other ports and well-connected inland cities, as well as in many coastal settlements with semi-urban features, there is evidence of a appreciable increase in the production of common local ceramics. Obv'ously, the repetoire of forms varies enormously from one producing centre to another, but, generally speaking, this pottery shows a high technical quality (refined clay, complex decorative design, good turning and high-temperature firing) and a certain standard of specialization, which suggests a diversified demand and a structured market. The prevailing wares are cooking vessels fired in reductive conditions (pots and saucepans) and a wide variety of table service pot­ tery and containers. The latter includes jugs and bottles, basins, vessels with pouring spouts and bowls of various shapes and sizes. In some cases, especially in the centres of production further inland, copies of African burnished table pottery were made. In the sixth century, one of the characteristics of local production was the proliferation of deco­ ration that was ingrained, grooved, combed, imprinted, and, to a lesser extent, painted in red and white. This, together with burnished and finely polished surfaces, gives the finished products a high quality. The most fully studied groups come from various rural settlements in Ibiza (the Balearic Islands),23 from the city of Valencia,24 the Vinalopò Val­ ley in Alicante,25 Carthagena and Tolmo de Minateda in Hellín (Al­ bacete).26 Evidence also exists of other types of production , much simpler and often made by hand or on a short wheel, which did not demand very sophisticated technology (maintaining limited formal variety, rough clay and low firing temperatures in reductive conditions). Cooking ves23 J. Ramon, El Baix Imped i l’època bizantina a les Ules pitiüses (Eivissa, 1986). 24 Blasco et al, “A ssaig de sintesi del panorama ceràmic de la ciutad de Valencia a l’Andguitat tardana”. 25 Reynolds, Settlement and Pottery in the Vinalopò Valley (Alicante, Spain), A. D. 400-700. 26 For Carthagena see M. Laiz and E. Ruiz, “Cerámicas de cocina de los siglos VVII en Cartagena (C/ Orcel -D. Gil)”, Antigüedady Cristianismo, V (1988), pp. 265-302. The excavations carried out in the Augustan theatre area have provided an important sequence of elements dated precisely and successively between the fifth and the sixth cen­ turies. Study is presently being carried out although it is possible to consult a preview of the final results in S. Ramallo et al, “Informe sobre las excavationes arqueológicas rea­ lizadas en el solar de la casa palacio de la Condesa Peralta (Cartagena)”, Memorias de Arqueología 4 (Murcia, 1993), pp. 130-7). Tolmo de Minateda is an urban setdement, which has been populated uninterruptedly from the sixth to the ninth centuries, and studies of it are presendy being published. For the setdement see L. Abad et al, “El Tolmo de Minateda (Hellín, Albacete). Nuevas perspectivas en el panorama arqueológico del sureste peninsular”, Arqueología en Albacete. Jornadas de Arqueología de la UAM . eds. J. Blánquez Pérez, R. Sanz Gamo and Ma. T. Musat Hévas (Madrid, 1993), pp. 145-76.

EASTERN SPAIN IN THE SIXTH CENTURY

169

sels are predominant. These include semi-spherical or frustum-conic shaped saucepans with solid half-moon handles and S-shaped pans, though flat lids and, to a lesser degree, large storage containers have also been documented. Hand-moulded pottery had coexisted with turned pottery since late Imperial times (second and third centuries), at least in rural areas, although the percentage is small in comparison with other common products, either local or imported, such as, for ex­ ample, African cooking pottery. Recently, greater archaeological atten­ tion given to late Roman contexts has shown the increasing impor­ tance of these wares from the fourth century A.D. onwards. In the same way as Carthage, in the excavated contexts on the Mediterra­ nean coast of the Iberian Peninsula hand-made goods are already sig­ nificant by the fifth century, and reach a considerable proportion in the sixth century A.D.27 The same phenomenon can be observed in various parts of the west­ ern Mediterranean and seems to be demonstrated in a generalized pro­ duction tendency which inclines towards self-sufficiency. Starting from the fourth century, in Porto Torres (Sardinia) there is a notable increase in the local production. This is characterized by cooking containers with typical half-moon handles, which reached a peak of production and de­ velopment by the fifth and sixth centuries.28 Similar developments occur in various parts of Campania, where in certain levels of the sixth and part of the seventh century, there is evidence of pottery turned on a slow wheel, with semi-lunar handles decorated with digital marks, reminiscent

27 By the end of the first half of the fifth century, in Vila-roma (Tarragona) “rough” pottery represents 51.5 % of common pottery (Taller Escola d’Arqueología, Un abocador del segle V d. C., en el firum provincial de Tàrraco, p. 427), and constitutes a characteristic element in the hilltown settlements of the fifth century in the Vinalopó Valley (Alicante), to be more precise pottery of Reynolds group 5 (P. Reynolds, “Ceramica tardorromana modelada a mano de carácter local, regional y de importación en la Provincia de Ali­ cante” Lucentwn IV (1984), pp. 245-67). Some typical forms from the sixth and seventh centuries in the south-eastern part of the Peninsula can be seen in S. Gutiérrez Lloret, “La ceràmica tosca a mano de los niveles tardíos de Begastn (siglos VI and VII): avance preliminar”, Antigüedady Cristianismo I (2 ed., 1994), p. 145-54. It is a general phenome­ non, noticeable in the town of Carthage, where hand-made pottery represents 20% of the total amount after the Byzantine conquest (M.G. Fulford and D. P. Peacock, Exca­ vations at Carthage: the British mission, Volume I, 2. - The Avenue du President Habib Bourguiba, Salammbô: The pottery and other ceramic objectsfrom the site, (Sheffield, 1984), p. 167). 28 F. Villedieu, Turns Libisonis, Fouilles d’un site romain tardif à Porto Torres, Sardaigne (Ox­ ford, 1984), p. 155 and fol.

170

SONIA GUTIÉRREZ LLORET

of Iron Age ceramics implying an important change in the predominant cookery traditions, which prevailed until the middle of the fifth century.29 The existence of these hand-made wares during the sixth and sev­ enth centuries on the African Mediterranean shores is poorly docu­ mented, not so much because they have not been unearthed in exca­ vations but rather through lack of identification, and therefore a diffi­ culty of chronological attribution.30 However, by careful reading of old archaeological documentation it is possible to identify similar frag­ ments of late Terra sigillata africana of type D in the Roman castellani of El Benian in Morocco,31or in the late Roman and Byzantine levels of Sétif,32 Hippone,33 or the thermae of Rusguniae.34 On the coast of the Iberian Peninsula, especially in the south-east, early discoveries of hand-made pottery have enabled scholars to fill a missing link in ceramic production from the seventh to the tenth cen­ turies. This gap was created by the supposed non-existence of ceramic material attributable to this period, a fact that was immediately inter­ preted, in historical terms, as irrefutable evidence as to the abandon­ ment of several late Roman settlements, both urban and rural, thereby indicating a deep demographic crisis.35 No doubt the reader is aware that the beginning of this crisis, was initially attributed to the third cen­ tury, but has now been established as dating from the middle of the seventh century. The re-dating of this break is not surprising given that 29 Examples from Capua have been published by P. Arthur “Scavo in proprietà Car­ rillo, S.M.C.V.: contributo per una conoscenza di Capua tardo-antica” Archeologia Medi­ evale, XTV (1987), pp. 517-34, illus. 6 and 7. There exists a debate on the common pot­ tery of Northern Campania of the sixth-seventh centuries, discussed in Colloquio in onore di John Hayes, op. cit., fn. 18, (in press). 30 In this regard R. Rebuffat’s (pp. 156-8) observations about the impossibility of da­ ting the moulded cooking pottery of Bu Njem, in Tripolitania, are relevant. He points out that these kinds of productions appear in the whole of the western Mediterranean and can be as much prehistoric as contemporary. Cf. Rebuffat, “Bu Njem, 1970”, Libya Antiqua, VI-VII (1969-70), pp. 107-69. 31 M. Tarradell, “El Benian, castellum romano entre Tánger y Tetuán” Tamuda, I, (1953), pp. 302-9. 32 P. A. Février, A. Gaspary and R. Guéri, Fouilles de Sétif (1959-65). Quartier nordouest, rempart et cirque (Alger, 1970); R. Guèry, La necropole orientale de Sitifis (Sétif, Algérie). Fouilles de 1966-67 (Paris, 1985); E. Fentress (ed.), Fouilles de Sétif, 1977-1984, V supple­ ment to Bulletin d’Archeologie Algérienne (Chèraga, 1991). 33 J. P. Morel, “Céramiques d’Hippone”, Bulletin d’Archeologie Algérienne I (1962-65), pp. 107-37. 34 R. Guèry, “Les thermes du sud-est de Rusguniae (Tarnet foust). Rapport provi­ soire: fouilles 1964”, Bulletin d’Archeologie Algérienne, I (1962-65), pp. 21-41. 35 A similar debate has been put forward for the Roman case by Patterson, “Un as­ petto dell’economia di Roma e della campagna romana nell’altomedioevo: l’evidenza della ceramica”, op. àt., fn. 6 especially pp. 309-10.

EASTERN SPAIN IN THE SIXTH CENTURY

171

it was based mainly on the disappearance of the commerce in fine pot­ tery.36 The redefinition of the chronology is due to the enormous im­ provements in our knowledge of these products, achieved mainly through the important work of John Hayes, transforming African cook­ ing pottery into a “type fossil” for late Antiquity.37 The development of contextual studies, a method of properly anal­ ysing the remains of imported pottery and establishing its real chrono­ logical value, has enabled scholars to demonstrate that this gap in ma­ terial and population is, in most cases, only apparent and is the result of a lack of archaeological identification, which therefore made it im­ possible to define any continuity in population. Thus, specific types of hand-made and turned pottery, as well as imported Tunisian ceramics and local high-quality products, already existing in the sixth century, reached an important quantitative development by the seventh and eighth centuries, when they overlapped with the first Islamic wares. Despite their great formal and functional conservatism, the possibility of access to long chronological sequences has allowed scholars to iden­ tify slow morphological transformations in the pottery, now that new elements of chronological discrimination are also available. In the case of the Iberian Peninsula and possibly the Maghreb coastline, the rela­ tion between hand-made pottery of late Roman tradition, made in the Byzantine or Visigothic eras, and early Islamic production seems indis­ putable and may reflect demographic continuity and a transmission of cultural traditions that have enriched and rendered more complex, the process of Islamization of these territories.38 If it is true that the increment in production of hand-made pottery had been a phenomenon on both sides of the western Mediterranean coast since at least the fifth century, then it is necessary to establish the significance of this increasing presence in terms of the market, and 36 As has been pointed out by Patterson, “Un aspetto dell’economia di Roma e della campagna romana nelTaltomedioevo: l’evidenza della ceramica”, p. 309. 37 J. Hayes, Late Roman Pottery, (London, 1972). 38 Reference to studies concerning hand-made pottery in early Islamic contexts is far too large for it to be entirely contained in this note; however, for recent studies concern­ ing the south-eastern region of the Iberian Peninsula, with reference to its close proximity to North Africa, see: cf. M. Acién and R. Martinez, “Cerámica islámica arcaica del sur­ este de al-Andalus”, Boletín de Arqueología Medieval Española III (1989), pp. 123-35; S. Gutiérrez Lloret, “La cerámica paleoandalusí del sureste peninsular (Tudmir): produc­ ción y distribución (siglos VII al X)”, I Encuentro de Arqueología y Patrimonio: la ceràmica altomedieval en el sur de al-Andalus, ed. A. Malpica Cuello (Granada, 1993). There are refer­ ences to the African hand-made pottery of similar chronology in N. Benco, The Early Medieval Pottery Industry at al-Basra, Moroco, BAR Inter. Ser., 341 (Oxford, 1978) and in Fentress, Fouilles de Sètif, 1977-1984, p. 194 and fol., 206 and fol.

172

SONIA GUTIÉRREZ LLORET

whether it carries with it any commercial implications. Generally, these ceramics tend to be considered complementary and secondary prod­ ucts. They were made in household environments and only distributed locally or, at the most, regionally. They represented only a specific and marginal segment of the demand.39 With this in mind, their coexist­ ence with other wares of an industrial nature -both imported and localwithin a complex market system, are interpreted, in economic terms, as archaeological evidence of a return to household and artisanal pro­ ductive formulas in rural areas. The return to this type of production appears, at least in the sixth century, and in the greater part of the seventh, to run parallel to, and be contemporary with, the survival of a rudimentary market economy in the harbour and urban districts, where the population density and the social fabric permitted the sur­ vival of a certain level of demand.40 However, this explanation could be subject to two different argu­ ments: first, recent archaeological studies have proved that these goods formed part of a far ranging shipping business, at least in several parts of the central and western Mediterranean;41 second, hand-made pot­ tery not only appears in rural contexts but also appears consistently in numerous inland cities - at least in the south-east of the Iberian Pen­ insula such as Begastri,42 Tolmo de Minateda43 or Lacipo44 - and in 39 This kind of approach is advocated by D. P. S. Peacock, Pottery in the Roman world. An ethnoarchaeological approach (London, 1982), in particular that which regards the “house­ hold industry” (defined as home production and mosdy carried out by women), of a com­ plementary and seasonal nature, which did not require any specific space or mechanical system. 40 It was interpreted in this way, at least in the case of southern Italy, by P. Arthur and D. Whitehouse, “La ceramica delTItalia meridionale: produzione e mercato tra V e X secolo”, Archeologia Medievale IX (1982), pp. 39-46. 41 According to Peacock, Pottery in the Roman world. An ethnoarchaeological approach, p. 17 and fol., production relating to the “household industry” model may have retained a cer­ tain degree of regional diffusion because of their cooking connotation, satisfying a specific part of the consumption demand in more complex markets, although usually such de­ mand is not too far from the centre of production. For the maritime trade of hand-made pottery, see some examples found in Carthage, notably the productions Hand-made, fab­ rics 1.1 and 1.2, coming respectively from the Islands of Pantelleria, Sardinia and the Eolian Islands, according to Fulford and Peacock, Excavations at Carthage: the British mission, Volume I, 2. pp. 157-61. 42 The “Cabezo de Rohenas” in Cehegin (Murcia); Gutiérrez Lloret, “La cerámica tosca a mano de los niveles tardíos de Begastri (siglos VI y VII): avance preliminar”, op. cit fn. 27. 43 Tolmo de Minatcda¡ in Hellin (Albacete); Abad et al, “El Tolmo de Minateda (Hellín, Albacete). Nuevas perspectivas en el panorama arqueológico del sureste penin­ sular”, op. cit, fn. 25. 44 Lacipo in Casares (Málaga); R. Puertas Tricas, “Excavaciones arqueológicas en Lacipo (Casares, Màlaga). Campañas de 1975 y 1976”, Excavaciones Arqueológicas en España 125, (1982), p. 267, flus. 186.

EASTERN SPAIN IN THE SIXTH CENTURY

173

many harbour cities or those strictly related to ports such as Barcino,45 Tarraco,46 Valentia47 and Illici,48 as well as in areas close to the coast such as Punta de TArenal in Javèa,49 Baños de la Reina in Calpe,50 Benalúa in Alicante,51 Sanitja in Menorca,52 or Can Sorá in Ibiza.53 Along the Mediterranean coastline of the Iberian Peninsula, many of these hand-made wares are believed to have been imported, as at Carthage, although this probable foreign provenance cannot always be proved.54 In various sites around Alicante, such as the coastal settlements of Benalúa, Baños de la Reina and Punta de l’Arnal, there is evidence of imported pottery of form 8 (hand-made, fabric 1.2) identi-

45 References to fragments without a stratigraphical context in R. Jarrega, “Notas sobre una forma cerámica: aportación al estudio de la transición del mundo romano al medieval en el este de Hispanid\ I Congreso de Arqueología Medieval Española, I (Saragossa, 1986), pp. 305-13. 46 TED’A, Un abocador del segle V d. C., en elfirum provincial de Tarraco, pp. 233-46. 47 Blasco et al, “Assaig de sintesi del panorama ceràmic de la ciutad de Valencia a PAntiguitat tardana”, p. 369 pi. 9. 48 Colonia Iulia Illici Augusta, Alcudia in Elche (Alicante). P. Reynolds, “Ceràmica tardorromana modelada a mano de carácter local, regional y de importación en la Provin­ cia de Alicante”; S. Gutiérrez Uoret, Cerámica común paleoandalusi del sur de Alicante (Ali­ cante, 1988), p. 50 and fol. 49 Bolufer, “Les céramiques tardanes importades (segles IV - VII de la N. E.) del jaciment romà de la Punta de l’Arenal”, p. 380. 50 Mentioned by Bolufer, “Les céramiques tardanes importades (segles IV - VII de la N. E.) del yaciment romà de la Punta de YArenai”, p. 380. The site has been excavated several times between 1986 and 1995, and is presently the object of a study directed by J. M. Abascal Palazòn. Cf. L. Abad et al, “Els Banys de la Reina, Calp, La Marina Alta”, Excavacions arqueólogiques de sabament a la Comunitat Valenciana 1984-88, II (Valéncia, 1990), pp. 34-6. 51 Reynolds, “Cerámica tardoromana modelada a mano de carácter local, regional y de importación en la Provincia de Alicante” and Settlement and Pottery in the Vinalopó Valley (Alicante, Spain), A. D. 400-700; F. Sala and A. Ronda, “Excavaciones arqueológicas en Benalúa” Historia de b Ciudad de Alicante, I, eds. R. Azuar Ruiz an d j. Hinojosa Montalvo (Alicante, 1990), pp. 287-312. 52 M. C. Rita, “Anforas africanas del Bajo Impero romano en el yacimento ar­ queológico de Sanitja (Menorca)” III Reunió d3Arqueobgia Cristiana Hispánica (Barcelona, 1994), pp. 321-32, especially p. 323 and 330, n. SC2/79/5. 53 Ramón, El Baix Imperi i l’època bizantina a les illes pitiüses, illus. 15. 54 In my opinion, the hand-made production of Ibiza is arguably an example of importation. These ceramics, which totally replaced African cooking pottery during the second half of the fourth century, are distinguishable by their inclusion of abundant gold­ en mica, and are considered to be of probable African origin (Ramón, El Baix Imperi i l’època bizantina a les illes pitiüses, p. 33). However, the quality of the clay seems to be more related to Reynolds group 5, a type which is very abundant in the south-eastern part of Hispania (Murcia and Alicante), and characterized by a coastal distribution which pene­ trated down the well-connected river valleys. Reynolds considers them of local origin (“Cerámica tardorromana modelada a mano de carácter local, regional y de importación en la Provincia de Alicante”, p. 253).

174

SONIA GUTIÉRREZ LLORET

fied at Carthage (P. Reynolds, group 2, form 2.1),55 together with sup­ posedly regional wares of Reynolds group 5, characterised by semispherical saucepans and basins.56 Other authors, however, are more cautious when interpreting morphological similarities as symptoms of trading activity, preferring to explain the resemblance as the exporta­ tion of a certain model rather than as actual pieces. This is the case in Tarraco, where a resemblance in form can be seen between the groups Vila-roma 7.8 and 7.13 and those wares discovered in Carthage. What is arguable is the identity of production due to the differences observed in the tempering agents.57 In this sense, hand-made or slow-turned moulded pottery is considered a typically regional or local product, that belongs to a craft tradition common to both Mediterranean coasts.58 Its development perfectly reflects this increasing impulse to­ wards domestic consumption that, according to Panella, characterizes the productive environment of the western Mediterranean beginning in the second half of the sixth century.59 In certain examples, like those of the coastal sites of Alicante, the resemblance is so clear, not only in terms of form but also in terms of workmanship and finishing (the polishing and the “spatulato” - the use of the spatula -o f the external surfaces), that it suggests close similari­ ties of production with those containers found at Carthage, (form 8, fabric 1.2). In many cases these coastal settlements date back to the early Empire and, although their original sizes are unknown, they are believed to have covered an area similar to that covered by the pre­ served remains.60 Judging by the volume of imported goods (pottery, building materials, luxury goods, etc.), they seem to have been pros­ perous economic centres that combined industriell activities with those 55 Reynolds, “Cerámica tardorromana modelada a mano de carácter local, regional y de importación en la Provincia de Alicante”, (group 2, form 2.1, p. 250); Bolufer, “Les céramiques tardanes importades (segles IV —VII de la N. E.) del jaciment romà de la Punta de l’Arenal”, p. 380 and illus. 3, 1-3. 56 For- Reynolds group 5, see above, fn. 53. 57 Taller Escole d’Arqueología, Un abocador del segle V d. C., en elforum provincial de Tar­ raco, pp. 145-6 and pp. 233-46, especially p. 236. 58 Taller Escole d’Arqueología, Un abocador del segle V d. C., en elforum provincial de Tar­ raco, p. 245; Acién and Martínez, “Cerámica islámica arcaica del sureste de al-Andalus”, p. 134. 59 See above, fn. 20. 60 With the exceptions of Baños del la Reina, where regular excavations have recent­ ly been started, and Benalúa, which between 1988 and 1989 was the object of a joint project between the Municipality and the University of Alicante to systematically sound the area, the sites lack any sort of unitary project of archaeological intervention, and thus the same material comes from private collections and old excavations, or from preserva­ tion works carried out in different periods under different directors.

EASTERN SPAIN IN THE SIXTH CENTURY

175

of trade, such as the processing of fishery products and the manufac­ ture of glass,6162which were still active at the beginning of the seventh century.52 However no structures are known to us within these late set­ tlements. All signs of permanent occupation which have come down to us are indirect and often derive from the ditches and enormous dump­ ing grounds that have obliterated the existing structures, as occurs at Punta de l’ilia.63 Whatever the case, the characteristics of these centres suggest a con­ siderable degree of economic self-sufficiency based on a significant au­ tonomy of trade, where the importation of hand-made pottery pro­ duced elsewhere in the central Mediterranean may be considered as part of a secondary trading system, within a network of alimentary products and consumer goods, through which the sea trade of the sixth century was organized. The fact that the importation of hand-made goods was by preference practiced on the coast, and to a lesser extent in those valleys better connected with the sea,64 suggests a maritime distribution unable to penetrate the hinterland, and closely tied to those trade circuits through which other imported goods, mosdy Afri­ can, were distributed.65 The low shipping cost of this cooking ware re61 Both Punta de l’lila and Baños de la Reina demonstrate peculiar deposits and compartments engraved into the cliff face on the seashore. These have been interpreted as fish farms and connected with food preservation activities (L. Abad, “Arqueología ro­ mana del Pais Valenciano: panorama y perspectivas” Arqueología del Pais Valenciano: pan­ oramay perspectivas (Elche, 1985), pp. 337-82, especially p. 370). Although usually this in­ dustrial exploitation is related to the period of the early Empire, archaeological evidence of salt-working sites in adjoining areas on the Murcia coast, suggest that this industry continued up until at least the sixth century. (S. Ramallo, “Envases desalazón en el Bajo Imperio (I)” VI Congreso Internacional de Arqueología Submarina (Madrid, 1982), p. 435-43. Glass manufacture in Benalúa is documented by the scoriae of the dumping grounds (L. Abad, “La romanización”, Historia de la dudad de Alicante, I (Alicante, 1990), pp. 119-148, especially pp. 144-6). 62 This can be inferred from the most recent African Sigillata and, in the case of Punta del Arenal, by the discovery of a lyre shaped, bronze belt plate of the seventh century, (during the Visigothic period) (Bolufer, “Les céramiques tardanes importades (segles IV - VII de la N. E.) del jaciment romà de la Punta de TArenal”, p. 375, n. 4). 63 Bolufer, “Les céramiques tardanes importades (segles IV - VII de la N. E.) del jaciment romà de la Punta de l’Arenai”, p. 375. 64 P. Reynolds mentions several sites along the Vinalopó Valley, the furthest of which is about 20 km from the coast (Reynolds, “Ceramica tardorromana modelada a mamo de carácter local, regional y de importación en la Provincia de Alicante”, p. 250). 65 The only known evidence of sea transportation of this hand-made pottery comes from a ship that sank in Cala Reona near Cabo de Palos. The boat, as yet only published in a preliminary form, transported amphorae and salting spatheia dated between the fifth and sixth centuries. Within the wreck, the remains of many hand-made vessels were found, interpreted as being pottery for on-board use. Cf. J. Pinedo and A. Pérez, “El yacimento subacuático tardorrpmano de Cala Reona. Estudio preliminar” , Antigüedady Cristianismo, V ili (1991), pp. 391-407.

17 6

SONIA GUTIÉRREZ LLORET

mains the only reason that might justify the circulation of a product that presumably had little mercantile value. This would also explain (if its foreign origin is accepted), why it did not penetrate further inland. To redistribute these goods overland would have increased the costs enormously, thus the demand for cooking pottery would tend to be supplied by local production. Consequently, in those inland settle­ ments, both rural and urban, the types of cooking wares were domi­ nated by hand-made products, except that, in contrast to what has so far been observed in the coastal settlements, here it is clear that they are locally produced or, at most, have only a regional distribution, con­ nected to the main arteries of land communication (the main roads, river valleys, etc.). These local products, which mostly correspond to Reynolds groups 5 and 7, span the chronological horizon from the fifth to the eighth centuries. They overlap with the manufactured goods of the Islamic period, in terms of the forms, the clay and the various treat­ ments used on the vessels. These local hand-made goods do not display any form of homoge­ neous distribution. Although they are typical of the rural and urban contexts of the Carthaginian coastal provinces in the sixth and seventh centuries,66 they are totally absent from contemporary levels at the im­ portant harbour city of Carthago Nova (Carthagena, in Murcia). This was the capital city of the Byzantine territories of Hispania from the middle of the sixth century until its destruction by the Visigoths, around 621-623.67 In the second half of the fourth century A.D., Carthagena had begun the process of economic recovery. This consist­ ed, as was the case in other hispanic cities, of a great public works pro­ gramme. A functional transformation of the city was instigated, con­

66 There is evidence of local hand-made pottery of the sixth and seventh centuries in several urban settlements in the southern part of the Carthaginian province (cf. Gutiérrez Lloret, “La cerámica paleoandalusi del sureste peninsular (Tudmir): produc­ ción y distribución (siglos VII al X)”). The difference between urban and rural settlements rests on the production and consumption of local cooking pottery turned on the wheel which is always much greater in urban centres. In rural settlements, local hand­ made cooking pottery continued to be used almost exclusively, especially in those places furthest from the cities. 67 Isidoras of Seville, Etjmologiarum, ed. and trans. J. Oroz Reta and M.A. Marcos Casquero (1983) XV, 1, 67.

EASTERN SPAIN IN THE SIXTH CENTURY

177

verting old monumental areas into the commercial heart of the har­ bour district.68 In the Byzantine period of Carthagena, the market for cooking wares was dominated by high-quality local products. These were rounded, wheel-turned pots directly influenced by oriental models, which dem­ onstrate morphological parallels with Byzantine Italy.69 There were large pans, both hand-moulded and turned on a slow wheel, and used certainly for the baking of bread,70 as well as African table pottery (Hayes 91 B, 99, 105-109) and oil amphorae of the same origin, espe­ cially of Keay’s type LXII.71 This scarcity of hand-made pottery, both imported and local, has important implications. Indeed, the predomi­ nance in Carthagena of turned cooking ceramics between the second half of the sixth century and the first half of the seventh, seems to in­ dicate that the city retained its traditional demand for production and consumption owing to its strong sea-borne trade which continued up until its destruction by the Visigoths. Already, by the end of the sixth century this market structure no longer existed either in the remaining inland settlements, or in the urban centres. 68 The sector including the Augustan theatre was re-urbanized between the middle and the end of the fifth century. Part of its structure and its architectural elements were reused, in order to construct a large commercial building with numerous tabemae, that was already functioning by die sixth century. This compound was demolished at an un­ specified moment in the second half of the same century to make way for new buildings which became the commercial centre under Byzantine rule. This in turn was destroyed following the capture of the city by the Visigoths. For Carthagena in general see S. Ramallo, La dudad romana de Carthago Nova: la documentación arqueológica (Murcia, 1989). An appraisal of Byzantine levels can be seen in D. Laiz et al, “Perspectivas arqueológicas sobre la presencia bizantina en Cartagena”, Actas de las VIIIJomadas sobre Bizancio, Veleia, ser. min. 2, eds. P. Bádenas andJ.M . Egea (Vitoria, 1993), pp. 119-33. However, since this study dates from 1988, the date of the conference, it is better viewed in the light of the work of Ramallo et al, “Informe sobre las excavaciones arqueológicas realizadas en el solar de la casa palacio de la Condesa Peralta (Cartagena)”. 69 Cf., among others, F. Fiumi and L. Prati, “Note sulla ceramica comune”, Ravenna e il Porto di Classe, ed. G. Bermond Montanari (Bologna, 1983), pp. 118-26; H. Patterson and P. Roberts, (La produzione di ceramica comune bizantina), “Fornaci altomedioevali ad Otranto. Note preliminari”, Archeologia Medievale XIX (1992), pp. 110-14; D. Petrone and E. Siena, (La produzione di ceramica comune), “Una fornace d’età bizantina a Cas­ tellana di Pianella (Pe)”, Archeologia Medievale XXI (1994), pp. 272-6). 70 For the pans and plates used for baking bread according to the method in texto coctus indicated by Isidorous of Seville (Etjmologiarum, XX, 2, 15), and its continuation into the Islamic period, see S. Gutiérrez Lloret, “Panes, hogazas y fogones portátiles. Dos formas cerámicas destinadas a la cocción del pan en al-Andalusi el hornilla (tannúr) y el plato (itábaq)” Lucentum IX-X (1990-91), pp. 161-75. 71 I j5\\7. and Ruiz, “Cerámicas de cocina de los siglos V-VII en Cartagena (C/ Orcel -D. Gil)”; Ramallo et al, “Informe sobre las excavaciones arqueológicas realizadas en el solar de la casa palacio de la Condesa Peralta (Cartagena)”.

178

SONIA GUTIÉRREZ LLORET

A brief analysis of trade during the sixth and seventh centuries shows that the disappearance of imported pottery and the proliferation of lo­ cally produced wares geared towards self-consumption, represent an effective archaeological indicator of the destruction of the economic system which had hitherto characterized the Mediterranean world of the Imperial period and the political system which had sustained it.72 The breakdown of the commercial markets, which had been based on an intense and far-reaching sea trade, was slow and heterogeneous. In­ deed, the scale and the speed at which it occurred differed, depending on the locations (whether coastal or inland), on the population (wheth­ er urban or rural), and on the differing political climate. This process was part of a global trend of economic recession, a centuries-old ten­ dency towards impoverishment whose point of inflection falls by the second half of the sixth century and the first of the seventh. Prior to this date —during the fifth century and the first half of the sixth century - the south-eastern coastal areas of Hispania participated in a flourishing maritime trade in much the same way as the rest of the Mediterranean world. This resulted in an uniformity of ceramic wares and presumably the circulation of goods as well, which can be observed within the coastal regions at both ends of the Mediterra­ nean.73 In south-eastern Hispania in particular, this situation is dem­ onstrated by the presence of imported goods, from both Africa, (oil am­ phorae, cooking wares and oil-lamps) and the Orient, (wine and, to a lesser extent, cooking wares), coexisting with the common local pottery. This latter production reflected a certain technical level of specializa­ tion, adapted to a diversified demand, and to a market organized, to a major degree, by urban élites. At the same time some wares, which were somewhat poorer and rougher, and characterized by an unsophisticated technology, (hand­ made of coarse clay containing substantial amounts of degreasing agents) start to appear in a systematic way, at least as far as archaeo­ logical surveys on the African and Hispanic coasts are concerned. Their technical features, which clearly denote a “primitivism” similar to proto-historic wares, together with their increasingly frequent ap­ pearance in late Roman archaeological surveys, were initially interpret­ 72 Concerning the historic interpretation of the decline in the importation of goods in the Mediterranean see besides the bibliography mentioned in the fh. 1, 5 and 7, the debate started by P. Delogu “Longobardi e Romani: altre congetture” in Langobardia, eds. S. Gasparri and P. Cammarosano (Udine, 1990), pp. 111-68. 73 Panella, “Merci e scambi nel Mediterraneo tardoantico”, especially p. 616 and fol.

EASTERN SPAIN IN THE SIXTH CENTURY

179

ed as being “technologically backward”, and representing a “cultural decline”, or certainly a productive crisis. Preliminary studies at Carthage, which were soon supported by other such studies in other regions of the western Mediterranean, indicated their coexistence with “industrial” products, such as the burnished table pottery. The reason for the success of these wares must have been that they suited a new demand for cooking wares;74 Soon they completely replaced African kitchen pottery on the local market. For this reason and independently from the interregional sea-borne distribution of some of these wares, the increasing frequency of hand-made pottery in the archaeological surveys into the first half of the sixth century must be interpreted as indicating an ever increasing tendency towards a domestic mode of production, although, in general, a market system was still functioning. It appears that, unlike the situation in territories under Roman in­ fluence,75 between the end of the fifth and the first half of the sixth century, there is no clear evidence of any other type of distribution of imported goods (amphorae, table pottery and oil-lamps) either in the cities and the industrial and commercial coastal centres,767or in the ru­ ral areas. This includes both the flourishing villae11and the small hilltop settlements which started to proliferate from the fifth century onwards. During the first half of the sixth century, these received African table 74 In the absence of sophisticated cooking systems, these manufactured articles an­ swered the need to obtain fire'-resitant pottery even from inadequate clay such as calcar­ eous clay. If volcanic clay (which is rich in kaolin and so enables production of thinly moulded vessels fired at high temperature) was unavailable, then the most elementary solution was to use thick and abundant inclusions that made it possible to fire the clay at relatively low temperatures (700-800) and ensure that the results were not too rigid and were able to withstand thermal shocks. Cf. El-Hraiki, Recherche ethno-archeologique sur la céramique du Maroc (Lyon, 1989), p. 365 and fol.; G. Aliprandi and M. Milanese, La ceramica europea. Introduzione alla tecnologìa, alla storia e all’arte (Genoa, 1986), p. 80. 75 H. Patterson, “Un aspetto dell’economia di Roma e della campagna romana nell’altomedioevo: l’evidenza della ceramica”, p. 309. 76 Within this category are the settlements of Punta de l’Arenal, Baños de la Reina, Benalúa in the Alicante region and those of Azohia, Castellar de Mazarrón in the Murcia region, the Island of Fraile de Aguilas, whose proximity to the ports and natural harbours suggest the possibility of direct trade, as asserts R. Mendez and S. Ramallo, “Cerámicas tardías (ss. IV-VII) de Carthago Nova y su entorno” Antigüedady Cristianismo, II (1985), pp. 231-80. 77 Some examples of late Imperial rural setdements are now known, which have large and even sumptuous residential buildings (rooms with mosaics, thermae, etc.) and extensive industrial systems, populated until very late times. These sites, which sometimes date back to the Early Empire, were transformed in the fifth century, often related to the building of turcularia for making vinegar. There is evidence of this in the village of Vilauba (Camós, Gerona) and in the settlements of Villaricos (Mula, Murcia) and Caña­ da Joana (Crevillente, Alicante). Keay links this to the increasing isolation of the urban

180

SONIA GUTIÉRREZ LLORET

wares on a regular basis,78 although in smaller quantities and showing less variety than in the coastal centres, where the imported hand-made pottery arrived and coexisted with the local production. Archaeological surveys suggest that in the hilltop settlements, the local hand-made pot­ tery predominated. This uneven distribution shows that the sea-borne trade of hand-made goods was not a particularly significant phenom­ enon in the markets, and only affected, to a greater or lesser degree, the coastal areas which retained a fluid network of alimentary trans­ portation. In other words, hand-made imported wares arrived as cargo and satisfied a demand, which in the internal markets was met by local production which was much cheaper, and therefore not redistributed. The second half of the sixth century in Hispania was marked polit­ ically by the Byzantine occupation of a wide tract of territory in the south-eastern part of the peninsula, as well as the Balearic Islands (in the possession of the Empire since 534). This political presence has conditioned archaeological studies enormously. It has often been used to explain the difference in the pace of trade, observed in the territories of the Tarragona province, (which remained a late Visigothic domin­ ion), and in the Carthaginian and Andalusian coastal regions, which were under Byzantine rule. In the former, there may have been a break in the importation of African wares, both of amphorae and table pottery,79 conforming with a decrease in African exports to the western Mediterranean which took occurred following the Byzantine conquest

markets, in comparison with the rural world which was much more self-sufficient (Keay, “La importación de vino y aceite en la Tarraconense oriental en la antigüedad”, p. 387). For these rural settlements see A. Roura et al, La villa romana de Vilauba (Camôs). Estudi d’un assentament rural (Campanyes de 1979-85), ser. mon. 8, Centre d’Investigacions Arqueologiques de Girona (Girona, 1988); M. Lechuga and M. Amante, “El yacimento romano de los Villaricos (Mula, Murcia). Aproximación al estudio de un establecimento rural de epoca romana en la Region de Murcia”, Antigüedady Cristianismo, Vili, (1991), pp. 363-389; J. Trelis, “Aproximación a la transición del mundo tardoantiguo al islàmico en las comarcas meridionales del Pais Valenciano: el ejemplo de Crevillente (Alicante)”, IV Congreso de Arqueología Medieval Española, II (Alicante, 1993), pp. 309-15. 78 For the Late Roman hilltop setdements, see S. Gutiérrez Uoret, “El poblamiento tardorromano en Alicante a través de los testimonios materiales: estado de la cuestión y perspectivas “, Antigüedady Cristianismo, V (1988), pp. 323-37, especially 329 and fol. 79 Cf. S. Keay, Late Roman Amphorae in the Western Mediterranean. A typology and Economic Study: the Catalan Evidence, p. 417 and fol; idem., “Decline or continuity?, The coastal econ­ omy of the Conventus Tarraconensis from the Fourth Century until the Late Sixth Cen­ tury”, Papers in Iberian Archaeology, BAR Inter. Ser., 193 (ii) (Oxford, 1984), pp. 552-77; Idem, “La importación de vino y aceite en la Tarraconense oriental en la antigüedad”, p. 388.

EASTERN SPAIN IN THE SIXTH CENTURY

181

of North Africa.80 Presumably, this fall in trade also affected the distri­ bution of oriental goods such as the Late Roman C tableware, which stopped arriving on the Catalan coast at about this time.8182 The situation in the coastal regions under Byzantine rule was differ­ ent. Here it is possible to find African sigillata of the second half of the sixth century and the first half of the seventh century (Hayes 91 C and D, 99, 101, 105, 106, 107 or 109) together with oil amphorae (espe­ cially Keay LXI and LXII) and spatheia. The distribution of these wares, which extended even into the interior of Murcia or Albac­ ete,182seems to testify to a steady commerce with North Africa until at least the second quarter of the sixth century. It has even been suggested that the settlements of the Byzantine province might even have encour­ aged this trade, centred on the important port of Carthagena and on the other coastal centres, redistributing the goods into the hinterland.83 At the same time problems appeared concerning the distribution of im­ ported products into the rural setdements away from the coast. Al­ though there is evidence of these African goods in the better connected inland cities, their proportion is relevant compared to the other types of locally produced fine table wares, that often replaced and certainly imitated the African sigillata.84 For this reason trade relations in the Byz­ antine regions must not be overestimated. Instead, there was a marked fall in trade, with few imported goods and an uneven redistribution. With these archaeological data in mind, it seems odd that the Byzan­ tine conquest of Africa should cause a decrease in the exportation of 80 S. Tortorella, “La ceramica fine de mensa africana dal IV al VII secolo d. C.”, Le merci, gU insediamenti. Società Romana e Impero Tardoantico (Roma, 1986), pp. 211-25, es­ pecially p. 220. At present, the research indicates that the lack of African table pottery in the last two thirds of the sixth century along the Hispania coast, shown in map 7, re­ flects a lack of archaeological knowledge for this area, rather than a genuine absence. 81 For the end of oriental table ware imports, see above, n. 15. 82 In Tolmo de Minateda (Hellín, Albacete) amphorae type Keay LXII and African sigillata of Hayes form 105, also documented in Begastri, (Cehegin, Murcia), has been found. For a comparison of later forms of African sigillata distributed in the south-east part of the Peninsula see the catalogue by R. Járrega, Cerámicasfinas tardorromanasy del Medi­ terráneo oriental en España. Estado de ¡a cuestión, especially p. 46 and fol. 83 R. Mendez and S. Ramallo, “Cerámicas tardías (ss. IV-VII) de Carthago Nova y su entorno”, pp. 266-7. E. Serrano, “Cerámicas con motivos estampados halladas en el te­ atro romano de Malaga”, Mainake VIII-IX (1986-87) (1988), pp. 201-13, esp. 210; R. Járrega, “Notas sobre la importación de cerámicas finas norteafricanas (sigillata clara D) en la costa oriental de Hispania durante el siglo VI a inicios del VII d.C.”, Congreso de Arqueología Medieval Española U (Madrid, 1987), pp. 337-44. 84 This is well documented in the late levels of Tolmo de Minateda, where types of Terra Sigillata Hispánica Meridional, as M. Orfila, “Terra Sigillata Hispánica Tardía Me­ ridional”, defines them, seem to be more abundant than the African forms, given the fact that both products are rare compared to common pottery at the end of the sixth and during the seventh centuries.

182

SONIA GUTIÉRREZ LLORET

the African sigillata to south-eastern Hispania. This trade was on a larger scale than, and differed from, that which has been observed in other areas of the western Mediterranean. In fact, this trade nuance of the second half of the sixth century was not particularly radical in the Tarraconensis coastal areas of the late Visigoth dominion. Recent surveys have pointed out that often in the Catalan coastal setdements, decora­ tive elements of Hayes style E (ii) are more frequendy to be found, whereas the forms of later African sigillata, such as Hayes 104 B, 105, 107 and 109, which are typical of the end of the sixth and the begin­ ning of the seventh centuries, are to be found at Roses, San Marti d’Empúries, Barcelona and Tarragona.85 These new data disprove the former assertion that maritime trade relations were broken off between Byzantine Africa and those Mediterranean areas then forming part of the Visigoth kingdom of Toledo, after the creation of the Byzantine province of Hispania. It suggests, as Aquilué and also Tortorella have recendy pointed out, that trade and political relations at this time pro­ ceeded along different paths.86 The same impression is formed from analysing the major African produced import of oil. This trade was still practised at the end of the sixth century in Carthagena, judging by the abundance of amphorae type Keay LXII.87 These containers, together with frequent spatheia, of­ ten better represented in contexts of the late sixth and the early part of the seventh centuries, as at Tolmo de Minateda for example, seem to have found their way in quite large quantities to the royal city of Recópolis (Guadalaja), founded by the Visigoth king Leovigildo in the year 578.88 Their presence, which was by then, totally independent 85 R. Járrega, “Notas sobre la importación de cerámicas finas norteafiieanas (sigillata clara D) en la costa oriental de Hispania durante el siglo VI a inicios del VII d.C.”, p. 341; Aquilué, “Comentaris entorn a la presencia de les cerámiques de prodúcelo africana a Tarraco”, p. 28 and fol; idem, “Cerámicas decoradas africanas procedentes de la Torre de la Audiencia (Tarragona)”, Empuñes, 48-50/1 (1986/89), pp. 26-35, especially p. 34. 86 Aquilué, “Comentaris entorn a la presencia de les cerámiques de producció afri­ cana a Tarraco”, p. 29; S. Tortorella “ARS: cronologia e distribuzione”, Colloquio in onore dì John Hayes, in press. 87 Its presence in the period of the destruction and flight from the city, in the first quarter of the sixth century, seems constant. See a brief reference by S. Ramallo et at, “Informe sobre las excavaciones arqueológicas realizadas en el solar de la casa palacio de la Condesa Peralta (Cartagena)”, p. 132, and confirmed by unpublished data kindly given by S. Ramallo, director of archaeological excavations in the Carthagena theatre, and by M. C. Berrocal who studies the late amphorae of Carthago Nova and its environs. 88 See, for example, some materials in C.E.V.P.P., “Cerámicas de época visigoda en la Península Ibérica. Precedentes y perduraciones”, A ceramica Medieval no Mediterráneo Ocidental (Lisbon, 1991), pp. 49-67, especially p. 50, ill. 8, 19-23. The site of Recopolis is at present being systematically excavated by L. Olmo Enciso, who I thank for his many suggestions.

EASTERN SPAIN IN THE SIXTH CENTURY

183

from the importation of fine pottery, again contradicts the argument that military hostility in Hispania between the Byzantines and the Visigoths caused the commercial isolation of the latter, reflected in the absence of wine and oil imports.89 Furthermore, this suggests the exist­ ence of an overland commercial redistribution network of those ali­ mentary products in daily use, such as oil, in spite of the high cost of transport.90 The apparent importation of foreign oil could indicate ei­ ther a special taste for this African product or a, perhaps temporary, local production shortage, although in this particular instance, one must wonder why it was brought such a long distance. Both supposi­ tions raise issues which require further thought. Also of note is the fact that during the seventh century these containers appear in several coastal and inland settlements, although the possibility of their réutili­ sation should not be discounted, as these amphorae were initially made for transportation but have also been found used as domestic contain­ ers91 and within warehouses.92 The chronological milestone of the conquest of Carthagena by the Visigoths (c. 621-623), marks the definitive end of the importation of fine goods in urban harbour districts. It was at this time, as in other parts of the western Mediterranean, that the classical “type fossils” cease, and an apparent gap opens up in the pottery continuity chain which has only recently started to be resolved thanks to contextual studies. In the second half of the seventh century, the inland market appears to be have been clearly dominated by local production. How­ ever, certain high class pottery traditions were retained, such as turned cooking pottery with fine surfaces which are occasionally burnished in­ ternally, together with common pottery of good workmanship, polished and treated with a broad knife. These wares which pertain to the Visigothic era are often predominant in the most run-down urban environ89 Kcay, “La importación de vino y aceite en la Tarraconense oriental en la anti­ güedad”, p. 388-94. 90 See J. Durliat in this volume. 91 This has been discovered at Tolmo de Minateda in domestic surroundings, dating from the second half of the seventh century. Here an example of amphora type Keay LXII has been found in association with a hearth and various cooking vessels and con­ tainers. Some of the most significant materials can be seen in “El yacimiento urbano tardio del Tolmo de Minateda (Hellin, Albacete, Spain)”, Bulletin n. 5 deh Association pour l'Antiquité tardwe. 92 See, for example, the warehouse situated at the basilica of Fomells on the north coast of the island of Menorca. Cf. P. de Palol, “Els diposits d’amfores de vi i d’oli en els ambits basilicals cristians”, El vi a I’antiquitat, Economia, producció i comerç al Mediterrani occidental, (Barcelona, 1987), pp. 419-36.

184

SONIA GUTIÉRREZ LLORET

ments. In the cities, their quality tends to diminish proportionally with the increase of rough, hand-made pottery. The increase of the latter, which predominanted in rural markets after the end of the sixth cen­ tury, can be linked with the breakdown of the centres of production, organized for, and dependent on the cities. The substitution of proper workshops with isolated domestic workspaces resulted in the reduction of the repertoire of forms and in the extension of those technological features typical of culinary pottery, such as coarse clay, manufacture by hand and low firing temperatures. Practically all the different types of pottery were subject to these changes, regardless of their use. The crisis in the economic model based on Mediterranean-wide commercial interchanges swept away with it many of the regional products which had a certain level of specialization of production, and which still responded to the demands of a complex structured urban market. Within this state of generalized de-commercialization, the only systems that were able to survive, and some certainly did, were pre­ cisely those that adapted to local and regional market demands. These economic systems were less complex and less exacting, but still ade­ quate and much steadier. This regionalization and progressive evolu­ tion towards more and more rudimentary forms of production is evi­ dence of a definitive breakdown of any market economy in the seventh century. In the Iberian Peninsula, it was not before the middle of the next century that incipient symptoms of recovery can be detected. The reconstruction of the market was a slow process and was associated with the development of the Islamic state. It was supported by a social organisation totally different from other regions in the western Medi­ terranean. This, however, is another story; that of the diversities of ear­ ly medieval Europe.

LES ECHANGES DANS LA GAULE DU NORD AU VIe SIECLE : UNE HISTOIRE EN MIETTES Stéphane Lebecq

L’historien qui se penche sur le problème de l’économie d’échanges dans le nord de la Gaule au cours du VIe siècle a au prime abord la sensation d’un certain vide: l’apogée romain, avec son économie in­ tégrée, est loin derrière lui; et ce n’est guère avant le début du VIIe siècle que les recherches récentes situent l’avènement d’un renouveau commercial dans le bassin des mers du Nord de l’Europe susceptible d’entraîner derrière lui l’ensemble des pays riverains.1 Les sources écrites sont rares,2 et elles sont de nature littéraire plus que diplomatique ou juridique. Leurs auteurs (principalement Gré­ goire de Tours, Venantius Fortunatus, Jonas de Bobbio et quelques au­ tres hagiographes) sont tous des hommes d’Eglise, le plus souvent issus de milieux aristocratiques romains ou fortement romanisés, dont les yeux se tournent plus volontiers vers l’horizon méditerranéen et les éventuelles survivances de la grandeur antique que vers les brumes d’un Septentrion abandonné à la domination barbare. De toutes façons, non seulement ils ne sont guère préoccupés par la question de la production et de la distribution des biens matériels, mais ils tendent de surcroît à mépriser le monde des marchands. Quant à l’archéologie, elle n’a pas encore révélé le moindre embry­ on d’emporium avant l’extrême fin du VIe siècle; et si la découverte de biens mobiliers dans les fouilles de nécropoles et d’habitats permet 1 Voir par exemple le livre fondateur de R. Hodges, Dark Age Economies. The origins of towns and trade (AD 600-1000) (Londres, 1982); ou encore, S. Lebecq, “Première esquisse d’une économie médiévale (VIIe-IXe siècle)”, L’économie médiévale, P. Contamine dir. (Par­ is, 1993), pp. 53-81. 2 On en trouvera l’inventaire quasi exhaustif dans A. Verhulst, “Der Handel im Merowingerreich : Gesamtdarstellung nach schriftlichen Quellen”, Early Medieval Studies, 2 (1970), pp. 2-54; D. Claude, “Aspekte des Binnenhandels im Merowingerreich auf Gr­ und der Schriftquellen”, Untersuchungen zu Handel und Verkehr der vor- undJmhgeschichtlichen Zeit in Mittel- und Nordeuropa, Teil III, Der Handel des frühen Mittelalters, éd. K. Düwel, H. Jankuhn, H. Siems, D.K. Dimpe (Göttingen 1985), pp. 9-99; et P. Johanek, “Der Aussenhandel des Frankenreiches der Merowingerzeit nach Norden und Osten im Spiegel der Schriftquellen”, ibid., pp. 214-254.

186

STÉPHANE LEBECQ,

d’envisager des déplacements significatifs d’objets depuis des lieux de production qui restent souvent hypothétiques, elle ne permet guère de savoir si le mode de distribution s’inscrivait dans le cadre d’une écon­ omie commerciale, dans le cadre du troc, ou dans celui de l’échange primitif fondé sur le don et le contre-don. Si l’on en croit Philip Gri­ erson, même la monnaie, très présente dans les horizons archéologi­ ques protomédiévaux, peut n’avoir pas eu dans les contrées septentri­ onales les plus germanisées d’usage à proprement parler commercial.3 Peut-on dans ces conditions tenter de brosser un tableau de l’écon­ omie d’échanges dans la Gaule septentrionale au cours du VIe siècle? Ce tableau, reflet de l’inégalité et de la précarité des sources, paraîtra nécessairement fragmentaire.

Le témoignage des sources littéraires et l’apparente survivance de l’échange gallo-romain Même si (un peu par facilité pour ne pas se priver d’un bon nombre d’informations offertes par les oeuvres historiographiques et ha­ giographiques de Grégoire de Tours et par la Vita Columbani de Jonas de Bobbio) on considère que le nord de la Gaule s’étend jusqu’à la moyenne et la basse vallée de la Loire, on ne trouve dans les sources littéraires que peu d’informations relatives à des trafics à longue dis­ tance. Cependant on peut considérer grâce à Grégoire de Tours qu’ex­ istaient encore à la fin du VIe siècle des liaisons entre certaines cités de la Gaule septentrionale et les plus lointains marchés de production méditerranéens ou orientaux, d’où venaient par exemple les vins de Gaza et de Laodicée présents sur le marché de Tours,4 ou l’huile et les épices présents sur le marché de Paris.5 3 Voir P. Grierson, “Commerce in the Dark Ages : a critique of the evidence”, Trans­ actions of the Royal Historical Society 5e série, 9 (1959), pp. 123-140; Idem, “La fonction so­ ciale de la monnaie en Angleterre aux VIIe-VIIIe siècles”, Moneta e Scambi nell’alto medioeoo (Settimane di studio del Centro italiano di studi sull’alto Medioevo, Vili) (Spoleto, 1961), pp. 341362. Ces deux articles pionniers ont été reproduits dans P. Grierson, Dark Age Numismatics. Selected Studies (Londres, 1979). Voir enfin P. Grierson et M. Blackburn, Medieval European Coinage, the Early Middle Ages (5th-10th centuries) (Cambridge, 1986). 4 Grégoire de Tours, Decern Libri Historiarum, VII 29, éd. R. Buchner, Gregor von Tours. Zehn Bücher Geschichte, vol. 2, Darmstadt, 1970: p. 126: '’ad requisendo potentiora vina, Laticina videlicet adque Gazitina’. 5 Grégoire de Tours, Decem libri Historiarum, VIII 33, vol.2, p.208 : ‘quidam (negotiator) im prumptuario est ingressus, adsumptoque oleo hoc ceteris quae necessaria erant, abscessit, lamine scus cupella olei derelicto’-, et Ibid. VI 5, vol.2, p.8 : “Iudaeus quidam Priscos nomes, qui ei (=le roi Chilperic) ad species quoemendasfamiliaris erat, advenif’.

LES ECHANGES DANS LA GAULE DU NORD AU VIe SIÈCLE

187

Et on peut considérer grâce à la très crédible Vita Columbaniy renforcée par quelques traditions hagiographiques irlandaises nimbées de merveilleux, dont la date de rédaction n’est de surcroît pas toujours facile à apprécier, qu’un trafic régulier - de vin en particulier - existait dès le VIe siècle entre la Gaule, spécialement les bouches de la Loire, et le Nord-Ouest britannique, depuis l’Irlande jusqu’aux Hébrides : la Vita de Ciaran de Clonmacnoise (mort vers 545) parle de marchands de vin de Gaule en Irlande; celle de Columba (mort en 597), dans la version d’Adomnan d’Iona (fin du VIIe siècle), évoque la présence dans le monastère des Hébrides de navigateurs venus de Gaule; et Jonas de Bobbio raconte comment Colomban, voulant rentrer dans sa patrie dans les toutes premières années du VIIe siècle, trouva sur la Loire en aval de Nantes un navire marchand en provenance d’Irlande qui était prêt à appareiller pour le voyage de retour.6 Ces divers auteurs nous montrent également l’animation des voies intérieures de la Gaule, spécialement des rivières qui facilitaient les li­ aisons régionales entre les différentes cités : ainsi naviguait-on sur l’Es­ caut, sur la Seine, sur la Marne, sur la Loire.7 Ainsi sait-on de manière plus explicite que la Meuse était riche de la circulation des navires et des biens, qu’il se pratiquait un trafic de vin sur la Loire entre Orléans et Tours, et un trafic de sel sur la Moselle entre Metz et Trêves.8 6 Vita Ciarani, éd. C. Plummer, Vitae Sanctorum Hibemiae, I, p. 14: Kmercatares cum vino Gallomm venerunt ad. S.Kïaranum; Adomnan of Iona. Vita Columbae Abbatis, I 28, éd. A. et M. Anderson, Adomnan3s Life of Columba, 2e éd. (Oxford, 1991), et trad. angl. R. Sharpe (Harmondsworth, 1995), p. 132 : ‘Gallici nautae de Galliarum provinciis\ Vita Columbani par Jonas de Bobbio, I 23, éd. B. Krusch, Monumenta Germanice Histórica, Scriptores Rerum Merovingicarum, IV, pp. 97-98 : ‘Reperta ergo navis quae Scottorum commercia vexeraL..'. Commentaires et critique des sources hagiographiques par E. James, “Ireland and western Gaul in the Merovingian period”, Ireland in Early Medieval Europe, Mélanges Kathleen Hughes, éd. D. Whitelock (Cambridge, 1982), pp. 362-386, en part. pp. 375-378. Voir également, P. Johanek, “Der Aussenhandel des Frankenreiches der Merowingerzeit nach Norden und Osten im Spiegel der Schrift quellen”, cité supra n. 2 pp. 227-228; et S. Lebecq, “La Neustrie et la mer”, La Neustrie. Les pays au nord de la Loire de 650 à 850, éd. H. Atsma, 2 voi. (Sigmaringen, 1989), vol.l, pp. 405-440, p. 431. 7 Grégoire de Tours, Decern Libri Historiarum, II 40, vol.l, p. 136 (dans la bouche de Clovis : ‘per Scaldemfluvium navigarerri)\ Ibid, VI 25, voi.2, p. 44 (‘nam tantum mandatone Sygona Matronaque àrea Paùsius intulerunt, ut inter àvitatem et basilicam sancii Laurenti nau/ragia saepe contingerenf); Jonas de Bobbio. Vita Columbani, I 22, pp. 96-97. 8 Venantius Fortunatus, Carminum libri, VII 4, éd. F.Leo, Monumenta Germaniae Histó­ rica, Auctores Antiquissimi, IV, p. 155 (‘Mosa dulce sonans,...triplice merceferax : alite, pisce, rat¿)\ Grégoire de Tours, Decem Libri Historiarum, VII 46, voi. 2, p. 152 (‘Cristoforus negotiator ad Aurelianensem urbem abiit...comparato vino et lintribus irwectó); Grégoire de Tours, Vvrtutes Mar­ tini, IV 29, éd. B. Krusch, Monumenta Germaniae Histórica, Scriptores Rerum Merovingicarum, I, p. 656 et trad, angl., R. Van Dam, Saints and their Miracles in Late Antique Gaul (Princeton, 1993), p. 296 (voici l’histoire qui est arrivée à un Treuericus negotiator : ‘dum Metis salmi negotiasset et ad pontem Mettis adplicuisset, dicit : (SDomne Martine, me et puncellos quos habeo et navicellam meam tibí commendo... Irwenimus nos ante portam Trevericarri’).

188

STÉPHANE LEBECQ

D’ailleurs plusieurs de ces cités avaient leur propre portus, souvent situé à proximité d’un pont ou d’une porte de la ville : ainsi Metz, Trêves, Amiens, Tours, Nantes.9 A lire de près telle ou telle anecdote rapportée par Grégoire de Tours, il nous est possible de reconstituer la petite entreprise de certains de ces marchands. Le Tourangeau Christophe, qui s’en est allé à Orléans acheter un vin destiné à être revendu à Tours, l’a chargé sur des barques avec l’aide de ses esclaves saxons;10 et c’est également avec sa barque et ses esclaves que le marchand de Trêves est allé prendre livraison de son sel sur le port de Metz.11 Ainsi est-on enclin à penser que dominait la petite entreprise individuelle, dans laquelle le capital du marchand consistait en une ou plusieurs embarcations, appelées sui­ vant les cas et les auteurs naves, navicellae, Untres, scafae ou encore rates,12 et en quelques esclaves, presque systématiquement appelés par Gré­ goire pueri ou puricelli. 13 Si l’on ne voit pas dans les sources du VIe siècle que ces marchands fussent au service exclusif de telle ou telle institution - palatiale ou ecclésiale, par exemple —, il est vraisemblable que, tel le Juif Priscus qui était le familiaris du roi Chilpéric dont il était le four­ nisseur attitré,14 les uns et les autres avaient quelques clients privilégiés. Il semble que dans bien des villes ces marchands constituaient des groupes homogènes et parfois solidaires: ainsi à Nantes, où, suivant Venantius Fortunatus, l’évêque Germain de Paris eut affaire aux ‘negotiatores civitatis Namneticae'; ainsi à Verdun, où, suivant Grégoire de Tours, l’évêque Desideratus fut aidé par ‘Ai negutium exercentes’ ou encore Hlli negutia exercentes'1 , ainsi sans doute à Orléans, où, si le roi Gontran fut, toujours suivant Grégoire, accueilli solennellement en 585 ''bine lingua Syrorum, bine Latinorum, bine etiam ipsorum Iudaeorum’, on a quelque raison

9 Pour Metz et Trêves (‘ad ponton Mettis' et ‘ante portant Trevericani), mêmes références que ci-dessus note 8. Pour le portus fluminis d’Amiens, voyez Grégoire de Tours, Virtutes Martini, I 17, p. 598, trad.Van Dam, Saints and their Miracles in Late Antique Gaul, pp. 215216. Pour Tours et Nantes, Jonas de Bobbio, Vita Columbani, I, 22 et 23, mêmes référenc­ es que ci-dessus note 7. 10 Grégoire de Tours, Decent Libri Historiamm, VII 46. Voir ci-dessus note 8. 11 Grégoire de Tours, Virtutes Martini, IV 29. Voir ci-dessus note 8. 12 Naves (Grégoire de Tours, Virtutes Martini, I 17 ou II 17; Jonas de Bobbio, Vita Columbani, I 23); navicellaé (Grégoire de Tours, Virtutes Martini, IV 29); Untres (Grégoire de Tours, Decem Libri Historiamm, VII 46); scafae Jonas de Bobbio, Vita Columbani, I 21-22); rates (Venantius Fortunatus, Carminum Libri, VII 4). 13 Pueri (Grégoire de Tours, Decem Libri Historiamm, VH 46); puricelli (Grégoire de Tours, Virtutes Martini, IV 29) : mêmes références que supra note 8. 14 Grégoire de Tours, Decem Libri Historiamm, VI 5, vol.2, p.8, passage cit supra note 5.

LES ECHANGES DANS LA GAULE DU NORD AU VIe SIÈCLE

189

de penser que s’y trouvaient des communautés de marchands syriens et de marchands juifs.15 Il est vraisemblable dans le dernier cas que les différents groupes ethniques constituaient des communautés juridique­ ment homogènes dans l’exercice de leur profession comme dans le pro­ tocole des cérémonies officielles. D’ailleurs les marchands verdunois évoqués par Grégoire paraissent constituer une véritable collégialité (pourquoi pas un collegium à l’ancienne?), puisqu’ils se sont portés col­ lectivement garants d’un prêt de sept mille aurei que le roi Théodebert (534-548) a consenti à l’évêque de la ville.16 Il n’y a pas qu’à Orléans qu’on entrevoit la présence de marchands orientaux. Il y en avait également à Tours, où le marchand de vin Christophe, qui faisait des affaires avec son beau-père installé à Or­ léans, était sans doute d’origine grecque ou syrienne; et à Paris, où d’une part le ‘negotiator genere Syrus’ Eusebius, ayant acheté le siège épis­ copal vers 591, s’empressa d’y placer de nombreux agents ‘degenere suo’, et où d’autre part le Juif Priscus, sollicité par Chilpéric de se convertir au christianisme, réclama un délai, le temps que son fils épousât une certaine iMassiliensim Hebraeam\ 17 Ainsi apparaît-il à travers ces trois ex­ emples que les marchands orientaux avaient entre eux des relations fa­ miliales, matrimoniales ou de clientélisme, dont on a de bonnes raisons de penser qu’elles venaient renforcer les relations d’affaires.18 On entrevoit donc par quel canal les vins de Palestine ou de l’Egée, les huiles et les épices de la Méditerranée gagnaient la Gaule du Nord: les marchands orientaux de Paris, de Tours et d’ailleurs avaient, tel le Juif Priscus, des correspondants dans les ports du Midi où étaient débarquées les précieuses cargaisons. Il est donc possible, au vu des quelques fragments d’oeuvres littéraires que nous avons passés en ré­ vue, que le grand commerce se soit perpétué en Gaule du Nord

15 Vcnantius Fortunadas, Vita Germani episcopi Parisiaci, XLVII, éd. B. Krusch, Monu­ menta Germaniae Histórica, Scriptores Rerum Merovingicarum, VII, p. 402; Grégoire de Tours, Decern Libri Historiamm, III 34, vol. 1, p. 186; et Grégoire de Tours, ibid., V ili 1, voi. 2, p. 160. 16 Voir S. Lebecq, “Grégoire de Tours et la vie d’échanges dans la Gaule du Vie siècle”, Grégoire de Tours et l’espace gaulois actes du congrès de Tours, éditées par N. Gauth­ ier et H. Galinié, Tours 1997, pp. 169-176. 17 Grégoire de Tours, Decem Libri Historiamm, VII 46 (voir ci-dessus note 8); Grégoire de Tours, Decem Libri Historiamm, X 26, vol. 2, p. 388); Grégoire de Tours, Decem Libri Historiamm, VI 5 et 17 (vol. 2, pp. 8 et 34). 18 Voir sur cette question S. Lebecq, “Le devenir économique de la cité dans la Gaule des Ve-IXe siècles”, La fin de la dté antique et le début de la cité médiévale, de la fin du Hle siècle à l’avènement de Charlemagne, actes du colloque de Nanterre, édités par C. Lepelley (Bari, 1996), pp. 287-310, en part. p. 292.

190

STÉPHANE LEBECÇ)

jusqu’en plein VIe siècle - peut-être même au-delà - dans les structures qu’il avait héritées de l’Antiquité tardive. Mais il est tout aussi vraisem­ blable que ce commerce-là ne concernait qu’une clientèle rare et privi­ légiée, où se côtoyaient aristocrates gallo-romains perpétuant des hab­ itudes de luxe héritées de l’Antiquité, parvenus barbares enclins à la parade sociale, et églises soucieuses de se ravitailler en ces produits précieux qu’exigeait la liturgie. Des échanges à caractère régional, presque rien n’est dit dans les sources écrites, si ce n’est quelques informations sur l’échange des pro­ duits de première nécessité (comme le sel) ou de consommation cou­ rante (comme le vin), et sur les difficultés du ravitaillement dans les périodes de disette, où les marchands (revendeurs de campagnes ou fournisseurs des marchés citadins) se livraient volontiers, comme s’en indigne Grégoire de Tours, à la spéculation.19 D’éventuels trafics le long des littoraux des mers du Nord ou à des­ tination d’un outre-mer septentrional, presque rien n’est dit, en dehors de rares allusions - glanées uniquement dans les Vitae des saints irlan­ dais - au trafic de vin entre la basse vallée de la Loire et la lointaine Hibemia. Peut-être est-ce parce que, tout entiers entre les mains de Bar­ bares celtes ou germaniques, ces trafics échappent totalement aux yeux d’observateurs dont le regard reste rivé sur l’horizon méditerranéen; ou alors, parce que, se développant d’une manière empirique, en de­ hors de toute structure clairement identifiable, ils demeurent invisibles. L’archéologie en dira-t-elle un peu plus? Viendra-t-elle confirmer les suggestions des textes sur la relative permanence de l’héritage antique en matière de grand commerce? Permettra-t-elle de préciser nos con­ naissances sur les échanges à caractère régional? Lèvera-t-elle un peu le voile sur les communications maritimes entre les côtes de la Gaule et les outre-mers septentrionaux ?

Les réponses de l’archêobgie En matière de grand commerce hérité de l’Antiquité, les textes sug­ géraient la présence dans les marchés urbains du Nord de certaines productions telles que les épices d’Orient, les vins de la Méditerranée

19 Pour cette question, qui ne concerne que marginalement mon propos, je renvoie à ce que j ’ai écrit dans “Grégoire de Tours et la vie d’échanges dans la Gaule du VIe siècle”, art. cité, ci-dessus note 16.

LES ECHANGES DANS LA GAULE DU NORD AU VI e SIÈCLE

191

et l’huile d’Afrique du Nord. Produits périssables, ils ne peuvent appa­ raître dans les fouilles archéologiques que par l’intermédiaire de leurs éventuels containers. Or les fouilles réalisées dans les sites de la Gaule septentrionale n’ont guère permis de retrouver ces céramiques médi­ terranéennes du VIe siècle —amphores le plus souvent -, porteuses de vin, d’huile, voire parfois de garum, qui ont été retrouvées avec une rel­ ative profusion en Italie, en Espagne, à Marseille, ou encore en Irlande et dans le Sud-Ouest britannique:20 les rares fragments découverts à Bordeaux ou à l’île Lavret dans les Côtes d’Armor permettent tout sim­ plement d’imaginer une route littorale conduisant des Colonnes d’Hercule aux mers de l’Europe du Nord-Ouest, sans escales marchandes significatives sur les côtes de Gaule.21 Aura-t-on plus de chance avec un autre type de mobilier, comme les “bronzes coptes” originaires de Méditerranée orientale sinon d’Egypte même, qu’on a souvent retrouvés dans des tombes datées de la période 500-560? En réalité, si on les trouve assez nombreux (une centaine d’exemplaires au total) en Italie, en Catalogne, et surtout dans l’Alle­ magne du Sud-Ouest et dans l’Angleterre du Sud-Est, on n’en a dé­ couvert que de très rares specimens dans le bassin de la Seine (d’une part près de Sens, d’autre part près de Rouen).22 Il semble donc 20 Sur cette question, voir, entre autres, P. Arthur, “Amphorae and the Byzantine World”, Recherches sur les amphores grecques, éd. J.Y. Empereur et Y. Garlan (Athènes, 1986), pp. 655-660; J.W. Hayes, Late Roman Pottery (Londres, 1972); et, plus précisément, C. Th­ omas, “Imported Pottery in Dark-Age Western Britain”, Medieval Archaeology 3 (1959), pp. 89-111 ; et Idem, A Provisional List of Imported Pottery in Post-Roman Western Britain and Ireland (Institute of Cornish Studies, Special Report VII) (Redruth, 1981). 21 Voir B. Wattier, “Les amphores funéraires de la nécropole de Saint-Seurin de Bor­ deaux”, L’Irformation de VHistoire de lArt, 18 (1973), pp. 113-118 (cité et mis en contexte par M. Bonifay dans “Observations sur les amphores tardives à Marseille d’après les fouilles de la Bourse (1980-1984)”, Revue Archéologique de Narbonnaise 19 (1986), pp. 269-305, en part. p. 281). Et P.R. Giot et G. Querre, “Le tesson d’amphore B2 de Pile Lavret (Bréhat, Côtes-du-Nord) et le problème des importations”, Revue Archéologique de l’Ouest, 2 (1985), pp. 95-100. M. Fulford souligne le caractère “aberrant” de cette découverte, dans “Byz­ antium and Britain : a Mediterranean perspective on Post-Roman Mediterranean Imports in Western Britain and Ireland”, Medieval Archaeology 3 (1989), pp. 1-6, en part. p. 3. 22 Voir J. Werner, “Zur Ausfuhr koptischen Bronzegeschirrs ins Abendland während des 6. und 7. Jahrhunderts”, Vierteljahrschrift Jur Sozial- und Wirtschaftsgeschichte 42 (1955), pp.353-356; et Idem, “Femhandel und Naturalwirtschaft im östlichen Merowingerreich nach archäologischen und numismatischen Zeugnissen”, Moneta e Scambi nell’alto Medioevo (Settimane di studio del Centro Italiano di studi sull’alto Medioevo, Vili) (Spoleto, 1961), pp. 557618. A compléter par R. Hodges et D. Whitehouse, Mohammed, Charlemagne and the origins of Europe. Archaeology and the Pirenne Thesis (Londres, 1983), en part. pp. 88-91; et par P. Perin, “A propos des vases de bronze “coptes” du VIIe siècle en Europe de l’Ouest: le pichet de Bardouville (Seine-Maritime)”, Cahiers Archéologiques 40 (1992), pp. 35-50.

192

STÉPHANE LEBECQ,

d’après l’archéologie que les grands trafics en provenance de la Médi­ terranée effleuraient la Gaule du Nord (en particulier le long des côtes occidentales et le long de l’axe rhénan) plus qu’ils ne s’y arrêtaient: voilà qui confirmerait l’idée que les quelques liens préservés par les marchands orientaux de nos sources écrites n’étaient que de fragiles survivances. * * *

Le sel et le vin des trafics régionaux ou interrégionaux entrevus grâce à Grégoire de Tours n’ont pas plus laissé de traces archéologiques di­ rectes. Force est donc, une fois de plus, de nous tourner vers ces quelques artefacts que nous donne à connaître la fouille, comme les pro­ duits de la céramique, de la verroterie ou de la métallurgie.23 En mat­ ière de production de poterie, on sait que le nord de la Gaule avait vu s’installer à la fin du IVe et dans la première moitié du Ve siècle la prépondérance de la céramique sigillée d’Argonne, dont les ateliers, sis à la lisière de la Champagne et de la Lorraine, étaient capables de rav­ itailler, jusqu’à 300 km à la ronde, aussi bien les camps du limes que les cités les plus importantes. Passé le milieu du Ve siècle, à une époque où le limes, peu à peu vidé de ses garnisons, n’en recevait plus, certaines cités continuèrent d’être approvisionnées, de Maastricht à Auxerre, et de Cologne à Rouen. Par contre, il apparaît que dans la deuxième moitié du VIe siècle c’est dans un rayon maximal de 100 à 200 km autour du foyer de production qu’était diffusée la céramique d’Ar­ gonne: privée des commandes institutionnelles de l’Empire et des cités, elle ne faisait donc plus l’objet que d’une distribution d’ampleur régio­ nale.24 Dans cette mesure, la production et la distribution de la céramique argonnaise devenait comparable à celles que réalisaient dé­ sormais la plupart des ateliers de céramique tournée de qualité du nord de la Gaule, comme ceux de Picardie, de Flandre, de l’Eifel, de la

23 Je renvoie d’emblée à l’étude récapitulative de H. Roth, “Zur Handel der Merow­ ingerzeit auf Grund ausgewählter archäologischer Quellen”, Untersuchungen zu Handel und Verkehr der vor- undfrühgeschichtlichen Zfit in Mittel- und Nordeuropa, Teil III, Der Handel des frühen Mittelalters, éd. K. Düwel et ai, pp. 161-191. 24 Voir les Actes du Colloque de Boulogne-Outreau sur La céramique du P au Xe siècle dans l’Europe du Nord-Ouest, édité dans les Travaux du groupe de recherches et d’études sur la céramique dans le Nord-Pas-de-Calais (Arras-Saint-Josse, 1993), en particulier les con­ tributions de Marie LrufFeau-Libre, “La céramique de l’Antiquité tardive dans le nord de la France”, pp. 91-105; et surtout de Didier Bayard, “La céramique dans le nord de la Gaule à la fin de l’Antiquité (de la fin du IVe au VIe siècle). Présentation générale”, pp. 107-128.

LES ECHANGES DANS LA GAULE DU NORD AU VIe SIÈCLE

193

Campine, ou du Limbourg.25 L’heure était à une diffusion à l’horizon microrégional, souvent limité à une cinquantaine de km autour du foy­ er de production: il faudrait attendre les VIIe-VIIIe siècles pour que certains fours de l’Eifel (Mayen), du bassin de Cologne (Badorf), ou de la basse vallée de la Seine (La Londe) retrouvent un rayonnement véri­ tablement macrorégional, voire international, spécialement en direc­ tion des pays du Nord.26 Cette distribution à caractère régional s’intégrait-elle dans des cir­ cuits véritablement commerciaux, où l’on faisait usage de la monnaie comme instrument régulier des paiements? Ce n’est pas sûr, car on a retrouvé dans le nord de la Gaule, dans un contexte généralement funéraire, d’assez nombreuses petites balances à deux plateaux - ainsi en Picardie à Arcy-Sainte-Restitue, en Normandie à Envermeu, en Champagne à Vouciennes, en Hainaut à Harmignies27 - qui suggèrent

25 Voir différentes contributions au colloque sur La céramique du V au Xe siècle cité cidessus, en part. Bayard, “La céramique dans le nord de la Gaule à la fin de l’Antiquité (de la fin du IVe au Vie siècle). Présentation générale”, ibid., pp. 124-126; D. Bayard et S. Thouvenot, “Etude de la céramique du haut Moyen Age (Vc-Xc siècles) dans le dé­ partement de l’Aisne (France): premier bilan”, ibid., pp. 291-340; Y. Hollevoet, “Céramiques d’habitats mérovingiens et carolingiens dans la région d’Oudenburg (Flan­ dre occidentale, Belgique)”, ibid., pp. 195-207; A. Verhoeven, “L’évolution de la céramique aux Pays-Bas méridionaux avant l’an mil”, ibid., pp. 209-215 (où l’on ap­ préciera la très remarquable carte de la page 213 sur les ateliers de potiers mérovingiens et carolingiens à diffusion régionale); et P. Demolon et F. Verhaeghe, “La céramique du Ve au Xe siècle dans le nord de la France et la Flandre belge: état de la question”, ibid., pp. 385-407; S. Lebecq, “Synthèse du colloque”, ibid., pp. 408-413, en part. pp. 410-411. 26 Voir W. Van Es et W.J.H. Verwers éd., Excavations at Dorestad I. The Harbour : Hoogstraat I (Amersfoort, 1980), en particulier “Early-Medieval Pottery”, pp. 56-111); R. Hodges, avec la collaboration de J.F. Cherry, The Hamwih Pottery : the local and imported wares from 3 0 years 'excavations at Middle Saxon Southampton and their European context, Council for British Archaeology, Research Report 37 (Londres, 1981); D.R.M. Gainster, M. Redknap et H.-H. Wegner éd., Zur Keramik des Mittelalters und der beginnenden Neuzeit im Rhein­ land, British Archaeological Reports, International Series 440 (Oxford, 1988), (en parti­ culier M. Redknap, “Medieval Pottery Production at Mayen : recent advances, current problems”, pp. 3-37). 27 J. Werner, “Waage und Geld in der Merowingerzeit”, Sitzungsberichte Bayern Aka­ demie Wissenschaft, Phil.-Hist.Klasse, I, (1954), pp. 3-40; idem, “Femhandel und Natural­ wirtschaft im östlichen Merowingerreich nach archäologischen und numismatischen Zeugnissen”; et surtout H. Steuer, “Gewichtsgeldwirtschaften im frühgeschichtlichen Eu­ ropa - Feinwaagen und Gewichte als Quellen zur Währungsgeschichte”, Untersuchungen zu Handel und Verkehr der vor- undfrühgeschichtlichen Z^t in Mittel- und Nordeuropa, Teil IV, Der Handel der Karolinger- und Wikingerzeit, éd. K. Düwel, et al., Göttingen 1987, pp. 405-527 (cet article considérable dépasse largement le cadre chronologique suggéré par le titre du volume: y est faite en particulier la récapitulation des découvertes de balances d’époque mérovingienne, spécialement du VIe siècle, p. 519). Sur les découvertes picardes, voir F. Vallet, “Economie, techniques et commerce”, La Picardie, berceau de la France, catalogue d’ex­ position (Amiens-Soissons, 1986), pp. 231-233.

194

STÉPHANE LEBECÇ)

l’existence, dans une vaste zone qui longe le front de forte colonisation germanique, d’une économie intermédiaire entre l’économie mon­ nayée (au Sud) et la Naturalwirtschaft (au Nord-Est), zone intermédiaire dans laquelle, si l’on utilisait éventuellement la monnaie, c’était plus pour son poids de métal précieux que pour le crédit que prétendait lui garantir la puissance émettrice. Il est une découverte archéologique qui suggère très bien ce que pouvaient être le métier, la fortune et le rayonnement d’un de ces ar­ tisans de la Gaule septentrionale, c’est la tombe n°10, dite “tombe du forgeron”, de la nécropole d’Hérouvillette, dans la plaine de Caen.28 L’individu avait été enterré avec plusieurs pièces de monnaie —la plu­ part contenues dans une bourse accrochée à sa ceinture, mais une placée dans la bouche - datées de la période 525-550, ce qui donne une idée non seulement de sa fortune, mais aussi, bien sûr, de la péri­ ode (milieu du VIe siècle) où il a été inhumé; il avait des armes, sin­ gulièrement une lance, une hache, une épée et un scramasax, ce qui prouve qu’il était libre; enfin il avait à ses pieds un abondant matériel significatif de son activité professionnelle: des outils (marteaux, ten­ ailles, ciseaux, limes, balance), mais aussi quelques morceaux de métal brut, d’alliages ou de scories (peut-être la matière même sur laquelle il travaillait au moment de sa mort). Cette découverte exceptionnelle fait écho à celle des trop rares - et souvent tardifs —foyers de production (forges aussi bien qu’ateliers de potiers ou de verriers) qui ont été iden­ tifiés dans la Gaule septentrionale. Elle prouve qu’il existait une catégo­ rie d’artisans comparable à ce que les sources écrites nous avaient fait entrevoir des marchands: des individus libres, travaillant sans doute “à la commande”, riches de leur activité, et dont le rayonnement dépas­ sait sûrement l’horizon purement local. Ce rayonnement atteignait-il l’horizon international? C’est une au­ tre histoire. Qu’on ait retrouvé dans l’Angleterre méridionale nombre de joyaux d’orfèvrerie cloisonnée de grenats originaires de la région de Trêves, ou des coupes de verre moulé originaires de Picardie orien­ tale,29 signifie qu’il y eut des contacts pour le moins , épisodiques, mais pas nécessairement qu’il y eut un commerce organisé. 28 Voir J. Decacns, “Un nouveau cimetière du haut Moyen Age en Normandie: Hèrouvillette”, Archéologie Médiévale 1 (1971), pp. 1-125; ou le bon résumé d’ E. James, The Franks (Oxford, 1988), pp. 207-208. 29 Sur le travail du grenat (lui-même originaire du Sud-Ouest de la Bohême) à Trêves, voir B. Arrhenius, Merovingian Garnet Jewellery : Emergence and Social Implications (Stockholm, 1985). Sur la diffusion du verre picard: voir Vallet, “Economie, techniques et commerce” cité supra n. 27.

LES ECHANGES DANS LA GAULE DU NORD AU VIe SIÈCLE

195

Des échanges maritimes sur les rivages de la Gaule septentrionale au cours du VIe siècle? Mises à part les allusions (rares et tardives) des textes hagiographiques sur les liaisons entre la Gaule atlantique et les lointaines contrées celt­ iques de l’Europe septentrionale, il n’y avait rien dans les sources écrites du VIe siècle qui révélât une quelconque activité maritime au nord de la Gaule, rien en particulier sur d’éventuelles relations économiques entre les littoraux gaulois et l’Angleterre. D’ailleurs les noms de Boulogne (Gesoriacum, Bonomia) ou de Lillebonne (Iuliobona), ports florissants sous le haut Empire et déclinants depuis deux siècles, n’apparaissent nulle part. Et si l’on trouve plusieurs occurrences du nom de Rouen (Rotomagus), il n’y est jamais question d’activités march­ andes. Est-ce à dire que la Gaule du Nord n’avait plus d’activité mar­ itime au VIe siècle? Ou est-ce à dire que nos auteurs, qui, surtout préoccupés par les souvenirs de la grandeur romaine, avaient les yeux rivés sur l’horizon méditerranéen, ne voyaient rien de ce qui se passait au Nord? Je crois qu’il faut avant tout retenir cette seconde explication: le monde maritime septentrional était presque totalement étranger à la cosmographie d’un Grégoire de Tours ou d’un Venantius Fortunatus. N’empêche qu’en évoquant le raid du Danois ChbchHaickus sur les côtes du nord de la Gaule vers 520-525,30 Grégoire suggère une expli­ cation à cette étrangeté des milieux littoraux du Nord, en même temps qu’à un possible ralentissement des échanges maritimes: les mouve­ ments des peuples marins et la piraterie, endémiques depuis deux siè­ cles, ont assurément rendu plus difficiles les communications pacifiques entre les deux rives du Channel. Mais ont-ils pour autant aboli toute vie de relations? Sûrement pas. La piraterie en effet a été longtemps associée à la migration des peuples marins qui, partis des côtes de la Germanie (depuis le Jutland jusqu’au delta du Rhin), a abouti à leur installation, spécialement dans le courant d’un long Ve siècle, sur les côtes de l’Est et du Sud-britanniques et, d’une façon plus discontinue, sur celles du continent. Que ces mouvements aient été contrôlés par Rome, qui procédait à la distribution des populations engagées dans la

30 Grégoire de Tours, Decerti Libri Historiamm, III 3, vol. 1, p. 146. Sur la vraisembla­ ble équivalence entre ce Chlochilaichus et le Hygelac du poème Beowulf, voir G.N. Garmonsway, J. Simpson et H. E. Davidson, Beowulf and its Analogues (Londres, 1980), pp. 112-113.

196

STÉPHANE LEBECÇ)

défense du litus saxonicum, ou au contraire lui échapper totalement,31 ce qui est sûr c’est que l’archéologie témoigne de l’intensité de la vie de relations, (1) dans un premier temps, entre les régions d’origine des mi­ grants et les régions de leur installation, (2) dans un deuxième temps, entre les diverses régions d’installation, spécialement entre les deux rives du Channel. Ainsi dans les nécropoles littorales du Ponthieu, entre la basse vallée de la Canche et la basse vallée de la Somme (Waben, Vron, Nouvion),32 et dans les nécropoles de la Basse-Normandie, en particulier celles de la plaine de Caen (Frénouville, Giberville, Hérouvillette, Saint-Martin-de-Fontenay, Sannerville),33 qui, les unes comme les au31 Dans une bibliographie immense qui reste dominée à mes yeux par L. Musset, Les Invasions. Les vagues germaniques, 3e éd. complétée par S. Lebecq, (Paris, 1994), et par J.N.L. Myres, The English Settlements, 2e éd., (Oxford, 1989), on doit mentionner les ap­ proches récentes de I. Wood, “The Channel from the 4th to the 7th centuries”, Maritime Celts, Frisians and Saxons, actes du colloque d’Oxford, Council for Brithish Archaeology, Research Report 71, éd. S. McGrail (Londres, 1990), pp. 93-97; M. Welch, “Contacts across the Channel between the 5th and the 7th centuries : a review of the archaeological evidence”, Studien zur Sachsenforschung, 1 (1991), pp. 261-270; et S. Lebecq, “Entre Manche et Mer du Nord, entre Grande-Bretagne et continent, les relations à travers le détroit dans les premiers siècles médiévaux”, Les champs relationnels en Europe du Nord et du NordOuest des origjines à lafin du premier Empire, actes du colloque de Calais (Calais, 1994), pp. 29-43. 32 Sur les nécropoles du Ponthieu, voir C. Seillier et P. Demolon, éd. Le Nord de la France de Thiodose à Charles Martel, catalogue d’exposition, (Lille, 1983), et, plus spéciale­ ment, D. Piton, La nécropole de Nouvion-en-Ponthieu (Berck-sur-Mer, 1985); C. Seillier, “Développement topographique et caractères généraux de la nécropole de Vron (Som­ me)”, Archéologie Médiévale 16 (1986), pp. 7-32; idem, “Les tombes de transition du ci­ metière germanique de Vron (Somme)”, Jahrbuch des Römisch-Germanischen ^entralmuseums Mainz 36 (1989), pp. 599-634; et idem, “Céramique de type anglo-saxon du cimetière de Waben (Pas-de-Calais)”, Antiquités Nationales 21 (1989), pp. 83-89. 33 Pour les nécropoles de Basse-Normandie, voir les études générales de C. Pilet, “Quelques témoignages de la présence anglo-saxonne dans le Calvados (Basse-Nor­ mandie, France)”, Frühmittelalterliche Studien, 13 (1979), pp. 357-381; et de C. Lorren, “Des Saxons en Basse-Normandie au VIe siècle? A propos de quelques découvertes archéologiques funéraires faites récemment dans la basse vallée de l’Orne”, Studien zur Sachsenforschung 2 (1980), pp. 231-259. Et voir les monographies de Decaens, “Un nou­ veau cimetière du haut Moyen Age en Normandie: Hérouvillette,,; C. Pilet, “Un nou­ veau témoin de la présence anglo-saxonne en Basse-Normandie à l’époque mérovingi­ enne: la fibule cupelliforme de Vierville (Manche)”, Archéologie Médiévale 7 (1977), pp. 8393; idem, La nécropole de Frénouville, British Archaeological Reports, International Series 83, 3 vol. (Oxford, 1980); idem, “Le mobilier anglo-saxon de la nécropole de Frénouville”, Centenaire de Vabbè Cochet, actes du colloque de Rouen (Rouen, 1978), pp. 441-463; idem, A ciel ouvert, treize siècles de vie (VIe siècle av. J.C.-VIt siècle ap. J.C.). La nécropole de Saint-Martin-deFontenay (Calvados) (Paris, 1987); idem éd., La nécropole de Saint-Martin-de-Fontenay (Cahados). Recherches sur le peuplement de la plaine de Caen du V s. av. J.C. au Vile s. ap. J.C., supplément à Gallia no 54 (Paris, 1994); idem éd., “Les nécropoles de Giberville (Calvados), fin Vefin VIIe siècles ap. J.C.”, Archéologie médiévale 20 (1990), pp. 3-140; idem éd., “Le village de Sanerville. Fin de la période gauloise-fin du Vile siècle ap. J.C .”, Archéologie médiévale 22 (1992), pp. 1-189.

LES ECHANGES DANS LA GAULE DU NORD AU VIe SIÈCLE

197

tres, ont été occupées de la fin du IVe jusqu’au VIIe siècle, les premières générations (IVe siècle et première moitié du Ve) ont été in­ humées avec un matériel dont une partie venait de la Basse-Saxe et du Schleswig-Holstein (ainsi certains types de fibules ansées ou de fibules en arbalètes), les générations suivantes (deuxième moitié du Ve et VIe siècle) ont été inhumées avec un matériel dont les plus sûrs correspon­ dants se trouvent dans le Sud britannique, par exemple dans le Kent ou Fîle de Wight (ainsi les disc brooches ou fibules rondes, les button brooch­ es ou fibules cupelliformes - spécialement celles à décor de masque hu­ main - , ou encore les urnes de céramique non tournée à décor linéaire). Que de tels échanges aient pu avoir lieu suppose l’existence d’un système de communications reposant sur des hommes, des bateaux, des fieux de débarquement, peut-être certains instruments de l’échange. Or, pour le coeur du VIe siècle, les textes sont muets sur les intermédiaires humains; ils sont presque muets sur les bateaux (il faut lire Gildas pour voir suggérée dans les mers du Nord la distinction en­ tre les curucae de peau des Scots, les cyules de bois des Saxons et les longae naves des Bretons)34; et comme fieux de débarquement ils ne mention­ nent que les ports fluviaux évoqués tout à l’heure. Sans doute les navires les plus utilisés par les Germains des mers du Nord restaient-ils au VIe siècle à peu près conformes aux types de ba­ teaux longs de l’époque migratoire - mus à la seule force des bras com­ me le dit encore clairement Procope35 -, dont l’épave de Nydam (fin du IVe siècle) reste la plus connue: les pierres gravées de Stenkyrka et de H(ggeby (Gotland, Ve-VIe siècles) suggèrent qu’ils étaient encore utilisés dans la Baltique, et les figures de proue en forme de dragon trouvées dans le fit de l’Escaut sont des témoignages significatifs de leur utilisation dans les eaux plus occidentales, encore au VIe siècle.36 34 Gildas, De Excidio Britonum, c.19 et c.23, éd. M. Winterbottom (Londres/Chichester, 1978), pp.- 94 et 97. Si Ton reconnaît aisément derrière les currucae les ancêtres des actuels curraghs, Gildas précise (c. 23, p. 97) que les cyules (cf. le vieil-anglais ceolé, quille navire) des Saxons correspondent à “nostra longae naves”. Voir S. Lebecq, “The Economy of the Northern Seas (Vth-VIIth centuries),,J à paraître dans le vol.l de la New Cambridge Medieval History of Europe. 35 Procope, Histoire des Guerres, VIII 20, éd. H.B. Dewing, (Londres/Cambridge (Ns) 1954), vol. 5, p. 260. 36 Voir D. Ellmers, “Die Schiffe der Angelsachsen”, Sachsen und Angelsachsen, cata­ logue d’exposition, éd. C. Ahrens (Hambourg, 1978), pp. 495-509; O. Crumlin-Pedersen, “The boats and ships of the Angles and Jutes”, Maritime Celts, Frisians and Saxons, actes du colloque d’Oxford, éd. McGrail, pp. 98-116. Sur les figures de proue en forme de dragon de la vallée de l’Escaut (et conservées au British Museum ainsi qu’au Musée mar­ itime d’Anvers), voir H. Vierck, “The origin and date of the ship’s figure head from Moerzeke-Mariekerke, Antwerp”, Helinium 10 (1970), pp. 139-149.

198

STÉPHANE LEBECQ

D’ailleurs le grand navire de Sutton Hoo, qui avait déjà une longue vie derrière lui lorsqu’il a été enterré vers 620-635 dans la tombe n°l de la nécropole royale, appartient à ce type, même s’il avait déjà subi quelques modifications de façon à recevoir un gréement.37 Ces navires, qui avaient un profil longitudinal strictement symétrique, et qui avaient un fond relativement plat, pouvaient être échoués partout, ou presque: cela signifie que les sites de débarquement étaient nombreux, en particulier sur les rivages sableux des côtes méridionales des mers du Nord. Il devient du même coup hasardeux de vouloir identifier, sinon de véritables ports d’envergure, du moins des lieux de débarquement et de traite internationale entre le IIIe siècle, qui vit l’éti­ olement des ports romains du Nord, et le tournant du VIe et du VIIe siècle, qui vit un réel renouveau de l’activité portuaire. De ce point de vue, il n’y a rien de comparable en Occident à ce que les archéologues danois ont trouvé dans le complexe (peut-être royal) de Gudme-Lundeborg, sur la côte orientale de la Fionie, qui fut occupé et abrita un trafic intense - entre autres des métaux les plus précieux - des environs de 200 aux environs de 600 après J.-C.38 Il n’est pas impossible que certains postes de traite soient identifiés sur les côtes de la Gaule sep­ tentrionale, mais une forte concentration d’objets importés dans un site littoral, comme celui de Bénouville près de Caen, en bordure occiden­ tale de l’embouchure de l’Orne, suffit-elle à y reconnaître un “comp­ toir anglo-saxon”?39 En tout cas, aucun des peuples nouvellement installés dans les con­ trées maritimes ne frappait encore monnaie : tous continuaient à vivre dans la dépendance du standard antique imité par les monétaires de l’intérieur de la Gaule - si du moins le besoin de numéraire se faisait

37 Voir A. Care Evans et R. Bruce-Mitford, “The Ship”, The Sutton Hoo Ship Burial, vol.l, Excavations, background, the Ship, dating and inventory, éd. R. Bruce-Mitford (Londres, 1975), pp. 345-435; ou encore A. Care Evans, The Sutton Hoo Ship Burial (Londres, 1986), en part. pp. 27-28. 38 H. Thrane, “Das Gudme-Problem und die Gudme-Untersuchung”, Frühmittelalter­ liche Studien, 21 (1987), pp. 1-48; S. Jensen et M. Watt, “Trading sites and central places”, Digging into the Past. 25 Years ofArchaeology in Denmark, éd. S. Hvass et B. Storgaard (Copcnhague/Aarhus, 1993), pp. 195-201; H. Clarke et B. Ambrosiani, Towns in the Viking Age (Leicester, 1995), en part, pp.50-51 et 202. 39 Hypothèse formulée par C. Riet, dans “Les nécropoles de Giberville (Calvados), fin Ve-fm VIIe siècles ap. J.C.”, p. 33 et dans “Le village de SanerviUe. Fin de la période gauloisefin du VIIe siècle ap. J.C .”, p. 40.

LES ECHANGES DANS LA GAULE DU NORD AU VIe SIÈCLE

199

sentir chez eux, ce qui n’est pas évident. Car les échanges alors pra­ tiqués entre le Nord de la Gaule et le Sud de la Grande-Bretagne paraissent avoir procédé de contacts diffusé, de troc, et peut-être de la pratique du don et du contre-don, qui faisait partie des bons usages diplomatiques et matrimoniaux, singulièrement dans le milieu des élites, le seul que la documentation écrite du temps éclaire quelque peu:40 car celle-ci s’intéressait plus que tout à la première tentative de christianisation de la cour royale du Kent sous influence franque, après le mariage vers 580 du roi Aethelbert avec Bertha, fille du roi franc Caribert.41 Assurément, la tombe à navire n°l de Sutton Hoo, avec son trésor de monnaies gauloises assemblé vers 625 (qui a tout d’une collection artificiellement constituée), et avec son mobilier d’origine aussi bien franque qu’orientale et scandinave, constitue le témoin privilégié de cette première phase d’échanges, de caractère pacifique mais non encore commercialé, réalisés de part et d’autre du détroit.42 Et son point d’aboutissement.

Vers 600, l’avènement d’une nouvelle économie commerciale sur les rivages de la Gaule du Nord Car vers 600 tout change. On entre assurément dans une phase d’échanges commerciaux structurés, avec des sites à vocation authen­ tiquement portuaire (les premiers wiks ou emporia), avec de la monnaie vouée aux nécessités de l’échange, et avec des intermédiaires spécial­ isés. C’est vers 600 en effet que les peuples marins du Nord, voisins et partenaires septentrionaux des Francs, se mirent à frapper monnaie. Si le trésor d’Escharen, enfoui vers 600 dans la basse vallée de la Meuse, montre le poids persistant du standard antique jusque dans les confins de la Gaule et de la Frise, puisqu’on y trouve, parmi ses 66 pièces d’or, d’authentiques sous d’or de Byzance et des imitations de sous ou de trientes frappés dans les ateliers de la Gaule méridionale et moyenne, il montre aussi quelques specimens d’imitations frisonnes des trientes gau­ lois. C ’est à ce moment en effet que les Frisons des bouches du Rhin 40 Q u’on se reporte aux études de P. Grierson citées supra note 3. 41 Je me contente de renvoyer à l’étude pénétrante de I. Wood, The Merovingian North Sea (Alingsâs, 1983), en part. pp. 15-16. 42 Voir les études de R. Bruce Mitford et d’A. Care Evans évoqués supra note 37.

200

STÉPHANE LEBECÇ)

commencèrent à imiter les monnaies de Maastricht et inaugurèrent la série connue sous le nom de trientes du type de Dronrijp (du nom du site de Frise où un important trésor a été découvert en 1876). C’est à ce moment ainsi que les Anglo-Saxons (dans le Kent, à Londres) se mirent à frapper leurs premiers thrymsas, imitations, eux aussi, des tri­ entes mérovingiens.43 C’est vers 600 également qu’apparurent les premiers véritables emporia, à commencer par Walcheren-Domburg, dans l’archipel zélandais, au débouché conjoint de l’Escaut, de la Meuse et du Rhin dans la mer du Nord, où ont été retrouvées (malheureusement en dehors de tout con­ texte scientifique) des structures de bois, un abondant mobilier céramique et métallurgique, et surtout de nombreuses monnaies ven­ ues, dès la fin du VIe siècle, de toute la Gaule du Nord, et bientôt d’Angleterre.44 Peut-être Walcheren-Domburg fut-il dès l’origine en contact avec le site east-anglien, contemporain d’après les fouilles ré­ centes, d’Ipswich, où ont été retrouvées de nombreuses céramiques d’origine rhénane, flamande et mosane.45 Dans tous les cas WalcherenDomburg, qui avait déjà joué ce rôle sous le haut Empire romain, mais qui avait disparu sous les coups de la deuxième transgression dunkerquienne, apparaît comme l’ultime marche-pied, pour tous les marins venus du grand delta rhéno-mosan, en direction du Sud-Est britan­ nique. Non seulement le renouveau commercial promut-il une nouvelle génération de ports de commerce, mais il contribua à la réanimation des anciennes cités les mieux placées pour en profiter, spécialement celles qui étaient installées au fond des estuaires: ainsi Londres qui, sui­ vant Bède le Vénérable, était déjà au moment où y débarqua en 604 l’évêque Mellitus, un emporium fréquenté, par “multi populi terra manque

43 Voir J. Lafaurie, “Le trésor d’Escharen aux Pays-Bas”, Revue Numismatique 2 (1960), pp. 153-210. 44 Voir S. Lebecq, Marchands et Navigateurs Frisons du haut Moyen Age, 2 vol. (T-ill«*, 1983), en part, vol.l, pp. 50-54; et P. Grierson et M. Blackburn, Medieval European Coinage, vol. 1, pp. 135-138 et 159-164. 45 Voir S. Lebecq, “L’emporium de Walcheren-Domburg: une mise en perspective”, Peasants and Townsmen in Medieval Europe. Studia in honorem Adriaan Verludst, éd. J-M. Duvosquel et E. Thoen (Gand ,1995), pp. 73-89. Voir K. Wade, “Ipswich”, The Rebirth of Towns in the West (AJ).700-1050), éd. R. Hodges et B. Hobley, Council for British Ar­ chaeology, Report 68 (Londres, 1988), pp. 93-100, en part. p. 96; ou encore R. Hodges, The Anglo-Saxon Achievement. Archaeology and the Beginnings of English Society (Londres, 1989), pp. 97-101.

LES ECHANGES DANS LA GAULE DU NORD AU VIe SIÈCLE

201

venientes”46 (témoignage que les fouilles du Strand peut-être le Ludenwich des textes, paraissent devoir confirmer)47; ainsi Rouen, où monnaie fut frappée dès la fin du VIe siècle;48 ainsi Nantes, d’où partait vers 600 une ligne de navigation en direction de l’Irlande et de l’Ouest britan­ nique - si du moins nous faisons confiance à Jonas de Bobbio et aux hagiographes irlandais évoqués plus haut.49 C ’est enfin vers 600 que fut frappé sur les bords de la Canche (QVANT1Ä) l’un des trientes retrouvés dans la grande tombe à navire de Sutton Hoo. Il n’y était pas encore question de wik, mais, dès 630-640, des monnaies furent frappées à WIC IN PONTIO, c’est-à-dire à 11Wik en Ponthieu”, avant qu’apparussent définitivement, sur les monnaies comme dans les sources écrites, le nom de Quentovic, “le wik de la Can­ che”, qui deviendrait aux VIIe-IXe siècles le principal point de com­ munication entre la Gaule du Nord et l’Angleterre du Sud.5051Preuve, s’il en est, du tournant décisif des années 600, qui virent le début de la structuration des grands échanges entre la Gaule du Nord et ses out­ re-mers septentrionaux. Ce qui est frappant, c’est que tous les monétaires du VIIe siècle qui laissèrent leur nom sur les monnaies de Quentovic avaient un nom plus anglo-saxon que franc: DAGULFUS, DUTTA, ANGLUS, DONNA, ELA...N Comme si l’apparition de Quen­ tovic, et peut-être même au-delà la réanimation marchande des lit­ toraux de la Gaule du Nord, était venue moins des Francs que de leurs partenaires maritimes, Anglo-Saxons ici, Celtes là-bas, et ailleurs Frisons. * * *

46 Bèdc, Historia Ecclesiastica Gentis Anglorum, II 3, éd. B. Colgrave et R.A.B. Mynors, (Oxford, 1969), p. 142. 47 Voir en particluier A. Vince, Saxon London. An Archaeological Investigation, (Londres, 1990), en part. pp. 13-17. 48 Voir J. Lafaurie, “Trouvailles de monnaies franques et mérovingiennes en SeineMaritime (Ve-VIIIe siècles)”, Histoire et numismatique en Haute-Normandie, cahier des Annales de Normandie 12 A, (Caen, 1980), pp. 93-107, en part. p. 95; S. Lebecq, “La Neustrie et la mer”, La Neustrie. Les pays au nord de la Loire de 650 à 850, actes du colloque de Rouen, éd. A. Atsma, vol.l (Sigmaringen, 1989), pp. 405-440, en part. p. 412. On attend avec impatience la publication du travail de Jacques Le Maho, Rouen et tes Vikings. De la cité carolingienne à la ville normande, où sera discutée l’hypothèse du développement d’un quar­ tier marchand dans le suburbium de Rouen. 49 Voir ci-dessus note 6. 50 Dans une littérature abondante, je me contente de renvoyer à S. Lebecq, “Quen­ tovic : un état de la question”, Studien zar Sachsenforschung 8 (1993), pp. 73-82. 51 Voyez V. Zedelius, “Zur Münzprägung von Quentowic”, Studien zar Sachsenforschung 7 (1991), pp. 367-377.

202

STÉPHANE LEBECÇ)

Au final, que sait-on de l’économie d’échanges dans la Gaule du Nord au long du VIe siècle? D’une part, de grands échanges orientés vers le Sud, dans le cadre d’une économie commerciale et monétaire héritée du temps de la grandeur romaine, qui, pour quelque temps encore, re­ lient à la Méditerranée de rares cités (Paris, Orléans, Tours...) habitées par une clientèle privilégiée, et dont seules, ou presque seules, nous parlent les sources écrites. D’autre part des échanges significatifs en di­ rection du Nord, en particulier de la Grande-Bretagne, dont on a quelque raison de penser qu’avant 600 l’essentiel s’inscrivait plus dans l’ordre de contacts multiformes liés aux usages sociaux (don et contredon, échanges diplomatiques et matrimoniaux, peut-être troc) que dans l’ordre d’une économie à proprement parler commerciale, et dont seule, ou presque seule, nous entretient l’archéologie. Entre les deux, des échanges intérieurs animés par des marchands et des artisans au rayonnement régional, travaillant pour la plupart en marge de l’écon­ omie monétaire, et dont les rares indices nous sont fournis moins par les textes que par la fouille. En attendant l’éveil, au tournant du VIe et du VIIe siècle, d’une véri­ table activité commerciale au nord de la Gaule, dont les initiateurs furent sans doute moins les Francs que les peuples marins du Nord, mais qui entraînerait à terme dans un nouveau développement les campagnes, les périphéries urbaines et les vieilles cités d’un immense hinterland, amenant le basculement du sud au nord de ses horizons économiques, l’histoire des échanges dans la Gaule du Nord au VIe siècle est assurément, à l’image de la précarité des sources qui la doc­ umente, une histoire en fragments, une histoire en miettes, où se su­ perposent et éventuellement s’enchevêtrent les différents niveaux de la Geldwirtschaft et de la Naturalwirtschaft: on comprend dans ces conditions que le VIe siècle ait été “le siècle oublié par Pirenne”.

MARSEILLE AND THE PIRENNE THESIS, I: GREGORY OF TOURS, THE MEROVINGIAN KINGS, AND “UN GRAND PORT”1 S.T. Loseby

Introduction Gregory of Tours, the late sixth-century Frankish historian and hagiographer, never went to Marseille, as far as we know. Indeed, if, as seems likely, the manuscript reading which places the house of his mother Ar­ mentaria at Chalon-sur-Saône in Burgundy, and not at Cavaillon in Provence, is to be preferred, then there is no reason to imagine that the bishop of Tours even ventured into the Midi, let alone as far as the shores of the Mediterranean.2 The Merovingian kings likewise seem to have eschewed first-hand experience of the delights of Provence, that milieu of sophisticated senators and philosophising judges, whose polite society is shunned by Abbot Domnolus of Paris in one of Gregory’s more obvious jokes.3 But if one could ask Gregory, or one of his Merovingian kings, to think of a port, I suspect that, however unfamil­ iar it was to them personally, Marseille would still have sprung instinc­ tively to their minds. Carolingian rulers and their chroniclers show scarcely any more interest in Provence than their Merovingian prede­ cessors, but their unfamiliarity with the south, I would suggest, is con­ ceptual as well as literal. While it is not entirely clear which port they

1 Earlier drafts of this paper, or of elements within it, were given in 1994 at confer­ ences and seminars in Edinburgh, Oxford, Leeds, London and Birmingham during my tenure of a British Academy Postdoctoral Research Fellowship. I am grateful to those present on these occasions and to Ruth Featherstone for their many helpful comments, and to the Academy for its support. 2 Gregory of Tours, De oirtutibus sancti Martini episcopi, III.60 ed. B. Krusch, Monumenta Germaniae Histórica Scriptores Rerum Merovingicantm 1.2, (Hannover, 1885), pp. 584-661. Cf. Idem., Gloria confessorum, 84 ed. B. Krusch, Monumenta Germaniae Histórica Scriptores Rerum Merovingicanm 1.2, (Hanover, 1885), pp. 744-820). For discussion see R. Van Dam, Saints and their miracles in late Antique Gaul (Princeton, 1993), p. 283, n. 93. 3 Gregory of Tours, Historiae, VI. 9 ed. B. Krusch and W. Levison, Monumenta Germa­ niae Histórica Scriptores Rerum Merovingicarum 1.1, (Hannover, 1937-51). As is often the case in Gregory’s writings, the precise butt of this joke is nevertheless far from clear.

204

S.T. LOSEBY

would have named in response to our hypothetical question, it seems unlikely that their answer would have been Marseille. This, of course, is the contrast between the Merovingian and Carolingian periods which made early medieval Marseille an important el­ ement in the thinking of Henri Pirenne. In Mahomet et Charlemagne, the most evolved (if still incomplete) formulation of his celebrated thesis, Marseille recurs as a choice illustration for Pirenne’s version of the transformation of the Roman world. In the first place, in support of his contention of the persistence of the exchange-networks of Antiquity well beyond the fall of the western Roman Empire, he assiduously as­ sembles documentary (and, in passing, numismatic) evidence, indica­ tive of trade via Marseille. Indeed, he employs a little licence to claim for Marseille one or two additional anecdotes of an explicitly or im­ plicitly commercial flavour which are not specifically located there in the primary sources.4 Even into the seventh century, he claims, ‘Marseille nous donne tout à fait l’impression d’un grand port’.5 Sec­ ondly, and in contrast, he asserts that such documentary evidence for trade via Marseille comes abruptly to a halt after the first decades of the eighth century, a silence which for him echoes the arrival of the Arabs off the coast of Provence. By 720 Marseille was doomed, be­ cause the Mediterranean had become a barrier: ‘la mer nourricière s’est fermée devant lui’.6 The economic geography of Frankia was henceforth reorientated towards the north, and towards the heartlands of Carolingian power. Whilst remaining respectful of the breadth and ambition of Pirenne’s vision, which has dominated subsequent discussions of early medieval exchange-networks to a considerable, perhaps even an un­ healthy extent, historians have since been concerned to criticise and to modify his thesis in various ways. The ammunition at their disposal has in recent years been greatly reinforced by the availability of ever-increasing quantities of archaeological evidence, a resource scarcely available to Pirenne himself. One influential work of synthesis in par­ ticular - subtided Archaeology and the Pirenne thesis - provides a succinct and cogent statement of what I think still passes for current orthodoxy.7 * H. Pirenne, Mahomet et Charlemagne, 3rd. ed. (Brussels, 1937). Examples can be found on p. 80 (Syrians), and p. 98 (slaves). It is a measure of Pirenne’s influence upon subsequent scholarship that some of his assumptions have since tended to acquire the status of fact. 5 Pirenne, Mahomet d Charlemagne, p.77. 6 Pirenne, Mahomet et Charlemagne, p.168.

MARSEILLE AND THE PIRENNE THESIS I

205

Long-distance Mediterranean exchange-networks did indeed survive the fall of the Roman Empire in the West, but were nevertheless in progressive decline, a decline which had reached terminal proportions by the end of the sixth century. The effect of the rapid progress of the Arabs around the southern shores of the Mediterranean was at most to deliver the coup de grâce to an economic network which was already as good as dead. However, both this and other recent works of synthe­ sis have paid no regard to the evidence from Marseille.78 This omission is odd, firsdy because of the abundance and interest of the archaeolog­ ical material recovered from the port in the last quarter of a century, and secondly because his perception of the fluctuating fortunes of Marseille is such a significant component of Pirenne’s analysis. For whatever one chooses to make of Pirenne’s thesis, in one sense he was right: Marseille should be a touchstone in any evaluation of post-Roman Mediterranean trade. This paper seeks to reconsider Pirenne’s first, positive contention about Marseille, that it continued to flourish through the sixth century and into the seventh.9 His account and in­ terpretation of its subsequent demise will be left for reconsideration in a second paper. Before turning to the new archaeological evidence from Marseille, and then trying to tie it back into the other categories of data, some general geographical background will help to explain its peculiar significance.

The geographical context Marseille is located some forty kilometres from the present-day mouth of the Grand Rhone. The historic heart of the city is its sheltered, deep-water harbour, known today as the Vieux-Port, a superb natural anchorage which sufficed to meet the needs of commercial shipping until as recently as the early nineteenth century. However, this mari­ time advantage is offset to landward by the city’s tiny and infertile hin­ 7 R. Hodges and D. Whitehouse, Mohammed, Charlemagne and the origins of Europe. Ar­ chaeology and the Pirenne thesis (London, 1983). 8 This is a failing common both to critiques of the Pirenne thesis and general discus­ sions of the ancient and medieval Mediterranean economy. A notable exception is K. Randsborg, “Barbarians, classical antiquity, and the rise of western Europe”, Past and Present 137 (1992), pp. 8-24. 9 For a more comprehensive presentation of the evidence discussed in this paper, with extended and additional argumentation, see S.T. Loseby, Marseille in Late Antiquity and the early Middle Ages (unpublished Oxford D.Phil thesis, 1993), especially chs. 4-5.

206

S.T. LOSEBY

terland. The exiguous basin in which Marseille lies is barricaded in be­ hind a continuous ring of hills, rising up barely ten kilometres or so from the rear of the Vieux-Port. For this reason, noted the geographer Strabo, writing in the early first century AD, the inhabitants of Marseille have ‘been compelled to look to the sea rather than the land’ for their livelihoods.10 Marseille may be in Provence, but it is not really of Provence: the inadequate integration of the city and the region re­ mains a problem today. Instead, Marseille looks away from Provence to the far greater commercial hinterland opened up by its proximity to the Rhone corridor, one of the very few, and perhaps the easiest, of the natural north-south routes through the Pyrenees-Alps-Carpathians mountain system of the European continent, and a vital artery of ex­ change and communications between the Mediterranean and northern Europe. Marseille may be some distance from the nearest of the Rhone mouths, but the hazards of the delta area, (low-lying, with poor visibil­ ity, sundry sandbanks and constantly shifting channels), have been such as to prevent the establishment of any permanendy viable port within or closer to the corridor. Rivals to Marseille’s pre-eminence have pe­ riodically emerged, such as Arles in tandem with an outport at Fos in the Roman period, or, in the eleventh century, St-Gilles.11 But over the longue durée Marseille, which requires a minimum of infrastructural support, and is not direcdy susceptible to the caprices of the river, has seen them all off to remain (he principal seaport for exchange via the Rhone corridor. The mediation of interregional traffic was the original idea behind the foundation of Marseille by colonists from the Greek city-state of Phocaea in Asia Minor in around 600 BC. The Phocaeans came late to the vogue for colonisation and they had to venture further; Marseille was the first Greek colony in southern Gaul, and would remain the greatest.12 The site of Marseille is also characteristic, in its landward

10 Strabo, Geographica, IV. 1.5, ed. and tr. H.L. Jones, 8 vols. (Cambridge, Ma., 196982). 11 For some brief comments on the viability of Arles as a seaport, cf. S.T. Loseby, “Marseille: a late Antique success story Journal of Roman Studies 82 (1992), pp. 165-85, at 181-2, and, more generally Loseby, Marseille in Late Antiquity and the early Middle Ages; S.T. Loseby, “Arles in Late Antiquity: Gallula Roma Arelas and urbs Genesii”, Towns in tran­ sition: urban evolution in late Antiquity and the early middle ages, eds. N. Christie and S.T. Loseby (Aldershot, 1996), pp. 45-70. 12 For Greek Marseille see most recendy the many papers in Marseille grecque et la Gaule (.Études massaliètes, 3), eds. M. Bats, G. Bertucchi, G. Congés and H. Tréziny (Lattes/ Aix-en-Provence, 1992), with anterior bibliography.

MARSEILLE AND THE PIRENNE THESIS I

207

isolation and maritime focus, of Phocaean colonisation, for the Phocaeans were, in Aristotle’s words, ‘practisers of emporia’.13 Marseille duly satisfies the criteria used to define that category of settlement known variously as an emporium, gateway community, or port of trade.14 It will, for example, have generally been a provider of guaran­ teed landing, offloading and storage facilities; it has a large population but a small political area; it is primarily concerned with wholesale rath­ er than retail trade. But most important of all, it is situated at a passage point between two natural and cultural regions - in this case the Med­ iterranean and northern Europe - ‘separated not only by geography, but also by their civilisations, such that products commonplace in the one can assume considerable value in the other’.15 The latter quota­ tion, which is of course the guiding principle of long-distance trade, is taken from a recent discussion of the emporial function of our Phocae­ an colony in the years after its foundation. I would argue that it applies equally well to Marseille in the sixth and seventh centuries AD. There is a tendency in early medieval economic history to see emporia as phenomena of the northern seas16 : Hamwic and Dublin, Hedeby and Birka, Quentovic and Dorestad are the places which spring most readi­ ly to mind in this context. But down in the Mediterranean Marseille was also a functioning emporium, at least in the Merovingian period. Long-distance trade had been the raison d’etre of Marseille from its foun­ dation and, devoid as it is of any significant hinterland, the port has always depended, for its periods of greatest prosperity and significance, on through traffic rather than the production or distribution of any lo­ cal agricultural surplus or on manufacturing industry. In the nine­ teenth century, for example, Marseille waxed fat on the French colo­ nisation of Africa, only to suffer subsequently during the long process of disengagement. And in the age of Gregory of Tours, the ability of Marseille to function as an emporium is an equally valuable pointer to the vitality of contemporary interregional exchange. For the fortunes

13 Aristotle, quoted in Athenaeus of Naucratis, Deipnosophistae, XIII, 576a, ed. and tr. C.B. Gulick, (London/New York, 1927-41). ,+ K. Polanyi, “Ports of trade in early societies”, Journal of Economic History 23 (1963), pp. 30-45. 15 J.-P. Morel, “Marseille, ville phocéenne”, Marseille 160 (1991), pp. 10-13, at p. 12 (my translation). 16 e.g. R. Hodges, Dark AgeEconomics: the origins of towns and trade AD 600-1000 (Lon­ don, 1982), pp. 47-86.

208

S.T. LOSEBY

of Marseille can indeed be used in precisely the way Pirenne used them, as will become apparent in moving from context and theory to practice. The archaeological evidence It is neither feasible nor appropriate for me to attempt here a compre­ hensive survey of the late antique and early medieval archaeological data recovered from Marseille in recent years. Instead, I will merely pick out one or two features of particular interest for summary consid­ eration. The first is a general point about late antique Marseille. The vast majority of the many archaeological operations carried out across the city over the last twenty years or so, latterly by the urban archae­ ological unit constituted in 1985, have, as so often in urban environ­ ments, been rescue excavations, limited in time, scope and resources, and responding to an agenda dictated by developers rather than ar­ chaeologists. 17 But the basic evidence of occupation recovered in these excavations nevertheless reveals one significant feature of fifth- and sixth-century Marseille. While many cities in Gaul were contracting sharply in size in Late Antiquity, Marseille was not. There are various indications that its Hellenistic wall-circuit, enclosing an area of around 50 ha (the precise course of its northern side remains uncertain), re­ mained the basis of the city’s late antique defences. The most obvious is the fore-wall recovered in excavations at the Bourse site, clearly de­ vised to conform to and protect the Hellenistic defences at the north­ eastern corner of the port, and dated to the fifth century.18 The recent discoveries of late antique occupation levels on sites right across the area enclosed by these walls meanwhile suggests no significant reduc­ tion in the scale of intra-mural setdement before the end of the sixth century.19 Indeed, the Bourse excavations have indicated some subur­

17 The results of these excavations are excellently presented in summary form in Marseille, itinéraire d’une mémoire: ànq années d’archéologie municipale, eds. L.-F. Gantés and M. Moliner (Marseille, 1990), Musées de Marseille, Le temps des découvertes: Marseille, de Prods à la reine Jeanne, (Marseille, 1993). 18 M. Bonifay, “Éléments d’évolution des céramiques de l’antiquité tardive à Marseille d’après les fouilles de la Bourse (1980-1981)”, Revue archéologique de Narbonnaise 16 (1983), pp. 285-346, at pp. 287-9; P.-A. Février, “Aux origines de quelques villes médiévales du Midi de la Gaule”, Rivista di studi liguri 49 (1983), pp. 316-35, at pp. 327-8. 19 Gantés and Moliner, Marseille, itinéraire d’une mémoire: cinq armées d’archéologie munici­ pale, especially pp. 93-6; A. Hesnard, “Une nouvelle fouille du port de Marseille, place Jules-Verne”, Comptes-rendues de l’academie des inscriptions et belles-lettres (1994), pp. 195-217.

MARSEILLE AND THE PIRENNE THESIS I

209

ban expansion of fifth-seventh-century date around and ultimately onto the shallow and silting watercourse outside the east gate of the city, a suburb which the Ste-Barbe/Puget III excavations, 200 metres north of the Corne, suggest may have extended outside the walls for some distance away from the port.20 It would of course be premature to push this evidence too far. The material recovered in the various intra-mural excavations in particular has not yet been fully published, and is likely in many cases to remain difficult to interpret in default of opportunities for further or more extensive excavation. Nevertheless, the apparent continuities of occupation and walled area, and the de­ velopment of a new extra-mural suburb of artisanal character contrast sharply with the evidence from most cities in late antique Gaul of im­ ploding landscapes characterised by reduced enceintes enclosing frac­ tions of rapidly contracting settlements.21 The details may remain open to debate. But in purely spatial and comparative terms post-Roman Marseille was indeed the ‘grand port’ which Pirenne envisaged. Alongside this general, if impressionistic, picture of the relative vital­ ity of Marseille in the fifth and sixth centuries, can be set the specific evidence afforded by the one extraordinary site where the detailed and extensive excavation of archaeological deposits of this period has been possible: the Bourse. A full report of these excavations has yet to ap­ pear, but the sequence and the material found in one area in particu­ lar, the Come, has been published in a series of excellent and expedi­ tious articles.22 The Come had been the site of an inner harbour, cre­ ated out of a marshy creek at the north-east comer of the Vieux-Port, and defined by a stone quay erected in the course of the first century AD. However, this harbour had always been susceptible to silting be­ cause of the outfall of a nearby spring. By the third century elaborate

20 Bourse suburb: R. Guéry, “Le port antique de Marseille”, Marseille grecque et la Gaule [Études massaliètes, 3) eds. M. Bats, G. Bertucchi, G. Congés and H. Tréziny (Lattes/ Aix-en-Provence, 1992, pp. 109-21. Puget III: Gantés and Moliner, Marseille, itinéraire d’une mémoire: cinq années d’archéologie municipale, pp. 53-8. 21 The best overview of late Antique towns in Gaul is P-A. Février, “Vetera et nova: le poids du passé, les germes de l’avenir”, Histoire de la France urbaine, I, ed. G. Duby (Paris, 1980), pp. 393-493. 22 M. Bonifay, “Éléments d’évolution des céramiques de l’antiquité tardive à Marseille d’après les fouilles de la Bourse (1980-1981)”; idem, “Observations sur les am­ phores tardives à Marseille d’après les fouilles de la Bourse (1980-1984)”, Revue archéologique de Narbonnaise 19 (1986), pp. 269-305; D. Foy and M. Bonifay, “Eléments d’évolution des verreries de l’antiquité tardive à Marseille d’après les fouilles de la Bourse (1980)”, Revue archéologique de Narbonnaise 17 (1984), pp. 289-308.

210

S.T. LOSEBY

Fig. 1. Come, main excavation: proportions of pottery by form and origin. After Bonifay, “Céramiques”, p. 304, fig. 13.

Fig. 2. Come, main excavation: proportions of fine wares only. Proportions derived from sherd counts. Only significant categories are shown. All périodisation is necessarily approximate. After Bonifay, “Céramiques”, p. 304, fig. 14.

MARSEILLE AND THE PIRENNE THESIS I

211

Fîg. 3. Come, sondages: proportions of pottery by form and origin. After Cavaillès-Llopis, uCeramiques”, p. 171, fig. 8.

Fig. 4. Come, sondages: proportions of fine wares only. Proportions derived from sherd counts. Only significant categories are shown. All périodisation is necessarily appro­ ximate. After Cavaillès-Liopis, “Céramiques”, p. 171, fig. 8.

212

S.T. LOSEBY

efforts to contain this problem had broken down.23 The continuing ac­ cumulation of alluvial deposits and the tipping of rubbish thereafter combined to create what the analysis of fauna and flora finds show to have been a fetid dump, circled by swarms of flies, and picked over by at least three species of vulture.24 But in the fifth and sixth centuries this singularly noxious site went through at least two phases of redevel­ opment. The mud was dredged and new quays erected, all with the clear intention of restoring it as a functioning watercourse. It was only in the seventh century that the unequal battle against the silt was once again abandoned, and the adjacent suburban occupation encroached onto the alluvium.25 Again there are difficulties in putting this data in context: the reasons for the revival of this inner harbour cannot be clear without further knowledge of the state of similar harbour facilities elsewhere around the port. Even so, it looks like a primafeme indication of commercial vitality. This impression is borne out by the ceramic assemblages recovered from the well-stratified fifth-, sixth- and seventh-century deposits in sampled areas of the Come (Figs. 1-4).26 The fine ware pottery falls into two main categories. One is the ware which goes by a variety of names including dérivées des sigillées paléochrétiennes (DSP), grey/orange stamped ware, and even, in older reports, the misleading Visigothic ware. Variants of this ware were produced between the end of the fourth and the seventh centuries at various centres in southern Gaul from Aquitaine to Provence, very probably including Marseille.27 But 23 Guéry, “Le port antique de Marseille”; M. Euzennat, “Fouilles de la Bourse à Marseille”, Comptes-rendus de l’academie des inscriptions et belles-lettres (1976), pp. 529-52, esp. pp. 545-50.

24 L. Jourdan, Lafaune du site gallo-romain et paléochrétien de la Bourse, (Paris, 1976), esp. summary at pp. 302-4. 25 Description, phasing, and dating in Bonifay, “Éléments d’évolution des céramiques de l’antiquité tardive à Marseille d’après les fouilles de la Bourse (19801981)”, esp. pp. 290-7, 303-26, with modifications in Bonifay, “Observations sur les am­ phores tardives à Marseille d’après les fouilles de la Bourse (1980-1984)”, pp. 270-1. 26 See previous note and for the smaller pottery sample obtained in a keyhole exca­ vation on the other side of the port, M.-T. Cavaillès-Llopis, “Céramiques de l’antiquité tardive à Marseille”, Documents d’archéologie méridionale 9 (1986), pp. 167-95.1 will make no serious attempt here to compare these assemblages with other Mediterranean contexts; for discussion along these lines see Loseby, Marseille in Late Antiquity and the early Middle Ages. 27 J. Rigoir, “Les sigillées paléochrétiennes grises et orangées”, Gallia 26 (1967), pp. 177-244, is the basic discussion. Among the many subsequent updates see e.g. J. and Y. Rigoir, “Les dérivées des sigillées dans la moitié sud de la France”, SFECAG, Actes du Congrès de Reims (1985), pp. 49-56; “La céramique du haut moyen-âge en France mérid­ ional: éléments comparatifs et essai d’interprétation”, La ceramica medievale nel Mediterraneo occidentale, Siena/Faenza 1984 (Florence, 1986), pp. 27-54, at pp. 40-2.

MARSEILLE AND THE PIRENNE THESIS I

213

while it is becoming increasingly apparent that some DSP production reached export markets in Spain and Italy, it seems primarily to have enjoyed an intra-regional distribution, unlike the main imported fine ware found at the Bourse, African Red Slip (ARS), the dominant in­ ternational pottery of late Antiquity. In contrast, the main eastern fine ware of this period, Late Roman C, is represented in the Corne de­ posits until the early seventh century, but only by a few statistically in­ significant sherds.28 This dominance by ARS of the imported fine ware assemblage at Marseille is typical of western Mediterranean sites.29 What is more unusual is its persistence. There is no sign of any decline in the supply of ARS to Marseille before the period when Gregory of Tours was writing, in the late sixth century. Indeed, the percentages of ARS in the sampled levels from the Corne are generally rising, wheth­ er as a proportion of the whole pottery assemblage or of the fine wares alone. Finally, the range of imported common wares represented in the fifth-seventh century levels at the Come is also worthy of notice, not because they arrive in any significant quantity, but rather because they come from regions of production scattered all over the Mediterra­ nean.30 These, like the few sherds of Late Roman C, hint at the en­ during economic integration of the Mediterranean and at the potential diversity of Marseille’s commercial relationships in a way which the other ceramic evidence does not. The amphora data recovered from the Come gives a similarly pos­ itive and internally consistent impression of the vitality of exchange via Marseille in the post-Roman period (Fig. 5).31 While fine-ware pottery is a traded commodity in its own right, amphorae were, of course, con­ tainers, generally for foodstuffs, and in particular for oil, wine, and var­ ious nutritious varieties of that favourite ancient condiment, fish-sauce.

28 Bonifay, “Éléments d’évolution des céramiques de l’antiquité tardive à Marseille d’après les fouilles de la Bourse (1980-1981)”, p. 322, and more generally F. Mayet and M. Picon, “Une sigillée phocéenne tardive (“Late Roman C ware”) et sa diffusion en occident”, Figlino 7 (1986), pp. 129-42. 29 Concise overview in S. Tortorella, “La ceramica fine da mensa africana del IV id VII secolo d.C.”, Società romana e impero tardoantico, III, le merci, gli insediamenti, ed. A.Giardina, (Rome/Bari, 1986), pp. 211-26. 30 “Importations de céramiques communes méditerranéennes dans le midi de la Gaule”, A ceramica medieval no Mediterráneo occidental, Lisbon 1987 (Lisbon, 1991), pp. 27-47. Eighteen different types of imported common wares are attested at Marseille. 31 Bonifay, “Observations sur les amphores tardives à Marseille d’après les fouilles de la Bourse (1980-1984)”. For an overview of late Antique amphora production and distribution see C. Panella, “Le anfore tardoantiche: centri di produzione e mercati preferentiali”, Società romana e impero tardoantico, III, pp. 251-84.

S.T. LOSEBY

214

African

C.425-C.450

C.575-C.625

(Period 1) all frags, r/h/b

(Period 2B) all frags, r/h/b

seventh century (Period 3) all frags, r/h/b

25% 43%

22% 37% 1% 15% 21% 4%

61%

97%

90%

23%

47% 24%

2%

10%

15% 1%

12% 17%

1%

424

2333

138

2085

Eastern Spanish Keay LU (Italian?) 0% 25% Indeterminate Residual 2% Sample Size

4749

42

Fig. 5. Come, main excavation: percentages of amphorae by region of production. Derived from Bonifay, “Amphores”, pp. 297-302-4, with minor modifications (for more detailed observations on the amphora assemblage cf. Loseby, “Marseille”, p. 185).

Although amphora imports from Spanish and Italian sources are pe­ tering out in the fifth century, these containers were continuing to ar­ rive at Marseille in the sixth century from Africa and from various ar­ eas of the eastern Mediterranean from the Aegean to the Nile, for con­ venience lumped together here under a broad ‘eastern’ classification.32 Nevertheless, African amphorae were increasingly dominant, and by the time of Gregory of Tours they enjoyed the lion’s share of the mar­ ket. This trend culminated in the virtual disappearance of eastern am­ phora imports from the latest, seventh-century levels at the Bourse, im­ plying some contraction, or at least change, in the pattern of exchange via Marseille. Now there are all sorts of difficulties with archaeological evidence of this type, even with well-stratified, carefully - excavated and meticu­ lously - analysed assemblages such as these. It is not my intention here to worry about questions of phasing or counting methods, but rather to highlight the potential pitfalls of ostensibly coherent statistical evi­ dence. The first is obvious. The above percentages are indicative only of relative and not absolute volumes of trade. African amphorae may, for example, represent around one-fifth of the mid-fifth-century sample 32 For more detailed discussions of amphora types see M. Bonifay and F. Villedieu, “Importations d’amphores orientales en Gaule (Ve-VIIe siècle)”, Recherches sur ta céramique byzantine, eds. V. Déroche andJ.-M. Speiser (Paris, 1989), pp. 17-46; M. Bonifay and D. Pieri, “Amphores du Ve au VIIe siècle à Marseille: nouvelles données sur la typologie et le contenu”, Journal of Roman Archaeology 8 (1995), pp. 94-120.

MARSEILLE AND THE PIRENNE THESIS I

215

compared with nine-tenths of its seventh-century equivalent, but this tells us nothing reliable about the real quantities of African imports ar­ riving at Marseille in either period. A little less obvious perhaps, but slightly more controllable, is the point that these figures measure con­ tainers and not commodities. Since the capacities of African amphorae were far greater than those of their eastern equivalents, the relative vol­ ume of the imported goods therein will tend to be underestimated. One approximate calculation has for example suggested that although eastern containers predominate in the mid-fifth-century sample, the quantities of African and eastern goods represented thereby, would be roughly equivalent.33 A further complication is introduced by the like­ lihood that some indefinable proportion of these eastern amphorae are reaching Marseille not via direct contact with their regions of produc­ tion but via intermediary ports, indeed via African ones. This unpleas­ ant possibility is implicit (in reverse) in Gregory of Tours’ account of a Frankish embassy to Constantinople venturing via Carthage.34 It has been given new force in recent years by the discovery of wrecks such as those at Port-Cros in the Des d’Hyères or in the Anse St-Gervais off Fos, dating from the sixth and early seventh centuries respectively.35 The cargoes of these unfortunate vessels included both eastern and Af­ rican amphorae. Knowing roughly where an amphora is manufactured is one thing. Postulating direct exchange between production centre and findspot is another. These problems of interpretation could be expanded upon or mul­ tiplied. But they pale into insignificance beside the more serious limi­ tations of the archaeological evidence. For the fundamental problem is that most of the exchange which the documentary sources show to have been carried on via Merovingian Marseille is what one might call ‘archaeologically challenged’, because either the commodities them­ selves, for example papyrus or slaves, or the containers in which they were transported, such as boxes, skins or barrels, are generally irre­ trievable by excavation. In short, archaeology privileges exchange ei­ ther in, or of, ceramics, a problem which haunts any attempt to recon­ struct late antique exchange-networks on the basis of excavated evi33 Bonifay and Villedieu, “Importations d’amphores orientales en Gaule (Ve-VIIe siè­ cle)”, pp. 37-9. 34 Gregory of Tours, Historiae X.2. 35 L. Long and G. Volpe, “Lo scavo del relitto tardoantico della Palud (Isola di PortCros, Francia)”, Vetera Christianorum 31 (1994), pp. 211-33; Gallia Informations, 1987-88, i, 12.

216

S.T. LOSEBY

dence alone. In the case of Marseille this has two important conse­ quences. First, the archaeology only shows trade in one direction, from the Mediterranean to Frankia. Gaul had stopped exporting goods in amphorae as early as the fourth century; as previously stated, the var­ ious southern Gallic varieties of DSP fine ware, although produced throughout late Antiquity, enjoyed a regional rather than an interna­ tional distribution. For knowledge of any reciprocal traffic via Marseille we therefore have to rely on documentary sources which, here at least, afford only limited help. Secondly, as the ARS workshops go out of production in the course of the seventh century, Marseille (and indeed Provence as a whole) is plunged into an archaeological void punctuated only by the odd Carolingian chancel fragment or sherd of imported pottery such as Forum Ware, and by local common ware manufactures which cannot at present be dated with any precision. The disappear­ ance of good diagnostic pottery evidence in the course of the seventh century reduces in particular the contribution which the archaeology of Marseille can make to the testing of the second part of Pirenne’s thesis. But it also hampers consideration of the latter part of its sup­ posed prosperity under the Merovingians. Our ceramic-led perception of post-Roman patterns of exchange makes it tempting, because only one side of its activity is archaeologically visible, to regard the ability of Marseille to function as an emporium in this period as limited. More specifically, it makes the disappearance of African ceramic imports seem emblematic of the failing fortunes of Marseille. It must be admit­ ted that the latter is scarcely a positive sign. But it must also be remem­ bered that the archaeology can tell only part of the story. Leaving this pessimistic litany of caveats and concerns to one side, the archaeological evidence does provide some additional support for Pirenne’s interpretation of Marseille in the Merovingian period as a ‘grand port’. Firstly and specifically, the statistics derived from the ar­ chaeological data can, for all their flaws, be used to illustrate something more positive about exchange via Marseille in the time of Gregory of Tours than the mere fact of its continuing existence. The best evidence here lies not in the amphorae, analysis of which is beset by sundry vari­ ables, but in the pottery (Figs. 1-4). The steady rise in the proportions of ARS in sampled levels at Marseille is significant because pottery is a traded commodity in its own right, and in statistical terms this is therefore a discrete set of data. Although we cannot determine the size of the Marseille pottery market in any given period, the increasing share of it captured by ARS imports shows not just the enduring via-

MARSEILLE AND THE PIRENNE THESIS I

217

bility of this traffic, but also its growing popularity in relation to locallyproduced fine and common wares.36 This would seem to be an index of the ready availability and affordability of ARS, and of its continuing importation to Marseille in some quantity down to the period around 600. Secondly, and more generally, the archaeology of Marseille in this period suggests that the port was insulated from the general trend which saw most towns shrink considerably in extent, and provides one potential explanation for this in the endurance of Marseille’s Mediter­ ranean trade-links, and especially those with Africa. For while the his­ tory of eastern amphora imports to Marseille is broadly consistent with that suggested by the contemporary assemblages excavated elsewhere around the western Mediterranean, the persistence into the seventh century of African imports to Marseille seems less typical. The appar­ ent decline in African imports on Spanish and Italian sites after the Byzantine reconquest of Africa in 533 has encouraged the interpreta­ tion of the latter episode as the catalyst for a reduction in African trade with other regions of the western Mediterranean, in part because the reimposition of the annona would have led to the requisitioning by the state of much of the available agricultural surplus, bringing ‘Africa into the eastern circuit as a peripheral rather than a central region’.37 But against a general background of decline, the exceptional persistence of imports to Marseille suggests that there was still some surplus African agricultural production available for distribution to western consumers and indeed that Frankia may have become a preferred market for such African exports as were continuing in the late sixth and seventh cen­ turies.38 In short, the archaeology of Marseille suggests that Pirenne 36 Indeed, the steady deterioration in the quality of DSP makes it increasingly diffi­ cult in later levels to distinguish it from local common wares: Bonifay, “Éléments d’évo­ lution des céramiques de l’antiquité tardive à Marseille d’après les fouilles de la Bourse (1980-1981)”, pp. 332-4; Cavaillès-Uopis, “Céramiques de l’antiquité tardive à Marseille”, pp. 177-80, esp. figs. 15-17. For a general summary, “Importations de céramiques communes méditerranéennes dans le midi de la Gaule”, p. 42, suggesting evidence for the continuation of DSP manufacture into the seventh century is confined to Marseille. 37 C. Wickham, “Marx, Sherlock Holmes, and late Roman commerce”, Journal of Roman Studies 78 (1988), pp. 183-93, at p. 193. This is an especially sophisticated version of a widely-held opinion, e.g. Hodges and Whitehouse, Mohammed, Charlemagne and the Or­ igins of Europe, pp.28-30; Tortorella, “La ceramica fine da mensa africana del IV al VII secolo d.C.”, pp.220-2; Panella, “Le anfore tardoantiche: centri di produzione e mercati preferentiali”, pp. 263-5, and, more generally, C. Panella, “Le merci: produzioni, itiner­ ari, e destini”, in Società romana e impero tardoantico, III, pp. 431-59, especially pp. 456-7. 38 For some brief speculation as to the reasons, Loseby, “Marseille: a late Antique success story?”, pp. 182-3.

218

S.T. LOSEBY

was right in the importance he attached to the port. But it also suggests that Marseille may have been rather exceptional. If we now return to the categories of evidence which Pirenne did have at his disposal, these impressions can be further strengthened.

The written record The crucial source for the commercial vitality of Merovingian Marseille is Gregory of Tours. It must nevertheless be admitted that in terms of direct evidence for the mediation of exchange via Marseille, Gregory is not that much use. Indeed, on the only occasion when he mentions the origins of a commercial vessel putting into the port - a ship from Spain - he maddeningly neglects to identify what it was car­ rying, except to observe that it was ‘the usual cargo’, preferring to focus his attention on the plague which the inhabitants of Marseille unwit­ tingly introduced into their homes with their purchases.39 And the spe­ cific (but unprovenanced) commodities which he does name in connec­ tion with Marseille number but three. Two of these, oil and an enig­ matic liquamen, perhaps some sort of fish-sauce, arrive at the port in a single shipment, seventy jars of which are promptly stolen.40 The im­ port of the third emerges only in a jibe. Stung by an accusatory letter from his eminent suffragan, Bishop Felix of Nantes, Gregory observes that if only Felix were bishop of Marseille instead, then oil and other (sadly unspecified) goods would no longer be shipped there, only pa­ pyrus, so as to give the libellous Felix more scope for his defamations of the good, that is to say the likes of Gregory and his family.41 And that is the sum total of Gregory’s specific references. To build up a list of commodities exchanged via Merovingian Marseille we must have resource to other documents. Like Gregory, Marculf s Formulary and the grants to the great northern Frankish monastic houses of St-Denis and Corbie of privileges at the Provençal ports of Marseille and Fos lay particular emphasis on the traffic in oil.42 However, the famous 39 Gregory of Tours, Historiae IX.22. 40 Gregory of Tours, Historiae IV.43. 41 Gregory of Tours, Historiae V.5. 42 Marculf, Formulae, Supplementum, 1, ed. K.Zeumer, Monumenta Germaniae Histórica Formulae I (Hannover, 1886), pp. 107-12. St-Denis: Gesta Dagoberti 18, ed. B. Krusch, Monumenta Germaniae Histórica Scriptores Rerum Merovingicarum II (Hannover, 1888), pp. 40125; Diplomata regumfiancorum e stirpe merowingica, nos. 61, 67, 82, ed. G.H. Pertz, Monumenta Germaniae Histórica Diplomata I (Hannover, 1872); Corbie: Diplomata regumfiancorum e stirpe merowingica, no. 86.

MARSEILLE AND THE PIRENNE THESIS I

219

Corbie diploma of 716 also highlights a cornucopia of other merchan­ dise available at Fos43 , a veritable delicatessen of herbs and spices, fruits and nuts, in quantities large and small, as well as Spanish skins, papyrus and, of course, fish-sauce. Altogether this remarkable docu­ ment is indicative of the importation into Frankia of an astonishing range of exotica from around the Mediterranean.44 The written sourc­ es, like the archaeology, are regrettably less informative about ex­ change via Marseille in the opposite direction, from the northern world to the Mediterranean, although there are recurring indications of a lively traffic in slaves.45 There exists, in short, a reasonable range of documentary evidence for the long-distance movement of goods via Provence which, when it is specific, tends to be centred upon Marseille.46 In matters of detail the relationship between the written record and the archaeology re­ mains uneasy, although there is one noteworthy case where they dove­ tail with remarkable precision; the Gaza wine mentioned by Gregory finds its archaeological counterpart in the increasing proportions of the chestnut-coloured Carthage Late Roman Amphora type 4 present in the late sixth-century assemblage at the Bourse, a type of amphora

43 For simplicity’s sake I am assuming here that the information pertaining to Fos can be extrapolated to Marseille without serious risk of error. 44 Pirenne made much of the Corbie diploma in particular: Mahomet et Charlemagne, pp. 89-91. My use of these privileges within the context of this paper is admittedly anach­ ronistic, since they fall in the period between the mid-seventh and early eighth centuries, running from the reign of Dagobert I down to that of Chilperic II. But I think it is rea­ sonable to use the information they contain as the best illustration of the range of com­ modities likely to have been imported through the Provençal ports in the preceding cen­ tury. I will be returning in more detail to these privileges and the problems with their interpretation in a later paper. 45 For Pirenne, Marseille was the great slave-market of the period: Mahomet et Char­ lemagne, p.99. See Vita Boniti, 3, ed. B. Krusch, Monumenta Germaniae Histórica Scriptores Re­ rum Merovingicarum VI (Hannover, 1913), pp. 110-39; Vita Eligii, 1.10 ed. B. Krusch, Mon­ umenta Germaniae Histórica Scriptores Rerum Merovingicarum IV (Hannover, 1902), 634-741; Gregory I, Registrimi epistulanm, VI. 10, ed. D. Norberg, 2 vols., Corpus christianorum series latina, 140 and 140A (Tumhout, 1982). Stricdy speaking, the last two sources do not spec­ ify precisely where the events they describe took place, although Marseille is the most likely location in both cases. I make no attempt here to piece together all the documen­ tary data for exchange via Marseille. For collections of the textual evidence for trade in this period, see D. Schwärzei, Handel und Verkehr des Merowingerreiches nach den schriftlichen Queden (Marburg, 1983) and more generally D. Claude, Der Handel im westlichen Mittelmeer während des Frühmittelalters (Untersuchungen zu Handel und Verkehr der vor- undfrühgeschichtlichen Zeit in Mittel- und Nordeuropa, 2) (Göttingen, 1991). 46 In this sense Pirenne’s occasional assumption that any commercial episode docu­ mented in Provence must have taken place at Marseille, though misleading, is under­ standable.

220

S.T. LOSEBY

thought to have carried wine from the Gaza region. Indeed, its ship­ ment inland is confirmed by the presence of fragments of such ampho­ rae in sixth-century levels on the avenue A. Max site in Lyon, which, very tidily, is precisely where Gregory records its regular donation to the church of St.Mary for use in the sacrament.47 But this exceptional coincidence between archaeological and historical data is offset by much larger discrepancies. Spanish trade, for example, is mentioned in the written sources, but leaves no trace in the archaeology; African imports, on the other hand, dominate the archaeological record, but are never explicitly documented. All this highlights once again the lim­ itations of both categories of evidence, and the necessity of assessing the exchange networks of this period from an interdisciplinary stand­ point. This is not the place to discuss such dislocations in detail. But it should be observed that although the two types of data do not neces­ sarily match, then the cumulative effect of the discrepancy as far as the commercial vitality of Marseille is concerned is, even so, a positive one. This effect is heightened if we return to the written sources, and in particular to Gregory of Tours, in whose works it is not so much the direct references to the trade of Marseille (the extent of which, as we have seen, is somewhat disappointing) but the indirect indications of the peculiar significance of Marseille within late sixth-century Frankia which are the most telling. According to Raymond Van Dam ‘any dis­ cussion of Merovingian Gaul based primarily [as most are] on the writ­ ings of Gregory of Tours will resolutely reflect his own parochial world’.48 If so, then Marseille was part of that world. Gregory may nev­ er have gone anywhere near Marseille, but he knew plenty of people who had; his deacon Agiulf, his priest John, and his maternal uncle Gundulf to name but three. Agiulf had been relic-collecting at Rome; John went to Marseille to trade; Gundulf was the agent of a Frankish king.49 Between them they exemplify the interconnected ways - travel, 47 For Gaza wine at the Bourse and elsewhere in Gaul, Bonifay and Villedieu, “Im­ portations d’amphores orientales en Gaule (Ve-VIIe siècle)”, pp. 20-21, 27-30, and for the important Adolphe Max site in general, F. Villedieu, Lyon St-Jean: lesfouilles de l’avenue Adolphe Max {Documents d’arckéologie en Rhône-Alpes, 3) (Lyon, 1990). Documentary referenc­ es: Gregory of Tours, Historiae VII, 29; Gloria Confessorum, 64. It needed a miracle to stop a member of the Lyonnais clergy from replacing the offering of Gaza wine with teethjarring vinegar, and reserving the former for his own enjoyment. 48 R. Van Dam, Leadership and community in late Antique Gaul (Berkeley, 1985), p. 201. 49 For Agiulf, Gregory of Tours, Historiae X.1; Gloria Martyrum, 82, ed. B. Krusch, Monumenta Germaniae Histórica Scriptores Rerum Merouingicarum 1.2 (Hannover, 1885), 484561; John: Gregory of Tours, Vita Patrum, VIII.6, ed. B. Krusch, Monumenta Germaniae Histórica Scriptores Rerum Merovingicarum 1.2 (Hannover, 1885), 661-744; Gundulf: Gregory of Tours, Historiae VI. 11.

MARSEILLE AND THE PIRENNE THESIS I

221

trade, and high politics —in which Marseille was tied into the Merov­ ingian polity, and through which the port imposed itself upon the con­ sciousness of Gregory up in Tours. In the writings of the bishop of Tours and in other documentary sources of the late sixth and early seventh centuries, such as the letters of his namesake Gregory the Great, the normal route for travellers be­ tween northern Europe and the Mediterranean is via Provence and the Rhone corridor.50 In almost all specified cases their journey involved taking ship to or from Marseille. It was the habitual port of passage for a steady flow of ambassadors, churchmen, missionaries, pilgrims, even the occasional pretender, whose movements reflect the dominant long­ distance trade axis of the day. And although he had never been on such a journey himself, the dominance of that route is firmly fixed in Gregory’s mind, for twice in the Histories, his mental map of the Merovingian world finds instinctive expression at a level beyond the mere description of actual events. The first example is in the afore­ mentioned jibe at Felix, when Marseille occurs to Gregory as the nat­ ural commercial context in which to set the bishop of Nantes’ imagined need for endless reams of papyrus.51 The second occurs when, on its return from Constantinople, a Frankish embassy achieves a spectacu­ larly unhappy landfall at Agde, and Gregory feels the need to explain why it had not put in at Marseille.52 For that, it would appear, was the normal thing to do. Seen from Gregory’s perspective at the heart of Frankia, Marseille was the obvious Mediterranean port-of-call, wheth­ er for political travellers or commercial traffic. By comparison, the na­ scent trade-networks of the northern seas have not seriously impinged themselves upon his consciousness.53 For Gregory, Marseille was the Merovingian port. For him, as for Pirenne, the reorientation of the commercial geography of Frankia from south to north has not yet tak­ en place. Gregory’s world view is reflected in the attitudes of the Merovingian kings. Like him, they hardly ever venture south in person. But they are 50 e.g. Gregory of Tours, Historiae VI.2, VI, 6; X .l; X.2; Gloria martyrum, 82; Gregory I, Registrimi epistularum, VI.52; IX.209; XI.41, etc. 51 Gregory of Tours, Historiae V.5. 52 Gregory of Tours, Historiae VI.2. 53 Fleeting references to “the son of a king of Kent”, Danish sea-raiders, and interestfree loans to the merchants of Verdun appear to be as far as Gregory”s imagination or interest stretches in this direction: Historiae IV.26; IX.26; III.3; III.34. But for the very real activities of the Merovingians in the North Sea area see I.N. Wood, The Merovingian North Sea (Alingsás, 1983).

222

S.T. LOSEBY

similarly aware of the enduring significance of Marseille. For the par­ ticular nature of their economic interests in the port we again need to look beyond Gregory’s narrative, where the motivations behind their actions are largely implicit. It is possible to identify two specific mech­ anisms established by the Merovingian kings in an effort to profit from their main Mediterranean outport. The first emerges in the relatively explicit information afforded by seventh-century sources. These show Marseille and the nearby port of Fos to be at the head of a chain of royal toll-stations extending up the Rhone corridor.54 Although it is impossible to determine empirically the value of customs-dues to the royal fisc, the rarity with which either the right to collect the tolls or immunity from them was granted away, not to mention the apparent enthusiasm of the Merovingians for increasing the number of locations at which they might be levied, suggests that the operation of this system was potentially lucrative.55 Those in charge of the toll-station at Marseille, the telonarii, are variously described in grants to St-Denis as adores regii, indices publia, or viri inlustri.56 They appear to be men of some status, directly answerable to the king, and also responsible for the administration of a cellariumfisci attached to the toll-station. A sim­ ilar institution is attested at Fos in the Corbie diploma of 716.57 The charters show the cellaria at Marseille and Fos to be royal ware­ houses of some description. Their exact workings, nevertheless, remain rather mysterious. They could be for the holding of goods in bond, for the storage pending collection of goods on which due toll has already been paid or remitted, or simply for goods purchased by royal agents on behalf of the crown, or indeed, and most likely, a combination of all three.58 But three things about the cellaria seem clear enough. Firstly, they are attested only at the Provençal ports. Secondly, the quantities

54 Gesta Dagoberti, 18; cf. Marculf, Formulae Supplementum, I. 55 R. Kaiser, “Steuer und Zoll in der Merovvingerzeit”, Francia 7 (1979), pp. 1-18. 56 Diplomata regíanfrancorum e stirpe merowingica, nos. 61, 82. 57 Diplomata regimi francorum e stirpe merowingica, no. 86. If the cataplus mentioned by Gregory at Marseille in Historiae. IV.43 is identical with a cellarium, as seems probable, then this institution can safely be pushed back into the sixth century. 58 H. Pirenne, “Le cellarium fisci: une institution économique des temps mérov­ ingiens”, reprinted in his collected essays, Histoire économique de l’occident médiéval (Paris, 1951), pp. 104-12. More convincing is F.-L. Ganshof, “Les bureaux de tonlieu de Marseille et de Fos: contribution à l’histoire des institutions de la monarchie franque”, Études historiques à la mémoire de JV. Didier (Paris, 1960), pp. 125-33. The potential affinities between these cellaria and the Byzantine apotiieke are intriguing, though the precise oper­ ations of the latter are equally debatable: see M.F. Hendy, Studies in the Byzantine monetary economy c.300-1450 (Cambridge, 1985), pp. 626-34, 654-62.

MARSEILLE AND THE PIRENNE THESIS I

223

of merchandise specified in the charters —or at least the quantities of oil - are such that even if they are taken with a substantial pinch of salt, we still need to be thinking of a cellarium more in terms of a ware­ house than a lockable cupboard in the office of the actores regii from which the occasional precious amphora of imported extra virgin olive oil might be produced for the very fortunate. Thirdly, whatever their precise workings, the cellaria seem indicative of a system of some so­ phistication, presided over by officials of rank, and intended to maxi­ mise both the exaction of tolls and the efficient provision of imported goods to the crown and its chosen beneficiaries. All this implies that either the tolls, or the commodities, or indeed both, were of sufficient value to justify royal maintenance of such a mechanism of control. The numismatic evidence from Marseille, referred to by Pirenne only in passing, via the works of his friend Maurice Prou, tells a similar story.59 Here again can be perceived the development by the Merov­ ingians of a particular mechanism of control and profit, centred upon Marseille. In late sixth-century Frankia, the minting of an increasingly debased gold coinage was generally left up to the local initiative of in­ dividual moneyers, and struck neither in the name of the Frankish kings, nor of the eastern emperor.60 This is not the case at Marseille, where the 570s saw the emergence of the distinctive gold coinage known as ‘quasi-imperial’. This was struck in the name of the Byzan­ tine emperor, but conformed to the lighter Germanic gold standard.61 The simultaneous and die-linked appearance of this coinage in c.575 at Marseille and other mints on the lower and middle Rhone (princi­ pally Arles, Uzès and Viviers) argues strongly for its direction by a cen­ tral authority, an impression reinforced by its regularity of weight, the absence of moneyers’ names, and the restriction of its minting to cities, all exceptional for Frankish coins of this period. The subsequent history of this coinage is dominated by and in all probability directed from Marseille. From the reign of Chlothar II (613-29) onwards, this coinage 59 Pirenne, Mahomet et Charlemagne, pp. 108-9. 60 P. Grierson and M. Blackburn, Medieval European Coinage, I: the Early Middle Ages (5th-10,h centuries) (Cambridge, 1986), pp. 111-28. 61 S.E. Rigold, “An imperial coinage in southern Gaul in the sixth and seventh cen­ turies?”, Numismatic Chronicle, 6th ser. 14 (1954), pp. 93-133; J. Lafaurie, “Les monnaies de Marseille du VIe au VIIIe siècle”, Bulletin de ta Société Française de Numismatique 36 (1981), pp. 68-73; Grierson and Blackburn, Medieval European Coinage, I: the early Middle Ages (5**110th centuries) pp. 128-31. But see especially M. Hendy, “From public to private: the western barbarian coinages as a mirror of the disintegration of late Roman state struc­ tures”, Viator 19 (1988), pp. 29-78, at pp. 68-70.

224

S.T. LOSEBY

is minted in the name of the Merovingian kings and not the Byzantine emperor, but it is in all other respects identical, and, in the numismatic context of seventh-century Frankia, equally anomalous. Again, some fundamental things about the quasi-imperial and royal coinages of late sixth- and seventh-century Marseille seem clear. In the first place, they are of high value, and can only have been of practical monetary use in major transactions.62 Secondly, they were intended for use within northern Europe, as their clearly-marked Germanic weight standard demonstrates. Confirmation that these solidi Gallicani were not acceptable around the Mediterranean appears in a letter of Gregory the Great, who in 595 demanded that rents from the papal estates in Provence be paid in kind not coin, because ‘in our land Gallic solidi cannot be spent’.63 The distribution pattern of these coinages corrob­ orates this further. It corresponds closely to the dendritic river-system of France, and extends up into Frisia and across the Channel into south-east England.64 Finally, the appearance and persistence, through much of the seventh century, of the quasi-imperial and royal coinages suggests that control of the lower Rhone valley money-supply was sufficiendy lucrative to make the issuing of a distinct and carefully-man­ aged coinage worthwhile here, when elsewhere in Frankia it was not. But this last point begs the crucial question of by whom precisely it was managed. The quasi-imperials have, with good reason, been described as ‘the most puzzling group of coins struck in Merovingian Gaul’.65 The key to this puzzle lies in the identity of the central minting author­ ity, for on that must hinge the problem of why these coins should be struck in the name of the emperor. It cannot be a civic authority, since the coinage was introduced simultaneously at several centres. But nor

62 It is noteworthy in this respect that while the rest of Frankia minted only tremisses (one-third of a solidus) in this period, the quasi-imperial and royal coinages retained the solidus as well as the smaller denomination. 63 Gregory I, Registrimi epistularum VI. 10. Gregory may have found this out the hard way: cf. Registrimi epistularum III.33, of two years earlier, where he fulsomely accepts pay­ ments from Provence. 64 Lafaurie, “Les monnaies de Marseille du VIe au VIIIe siècle”, p. 72, for a distri­ bution map of the Marseille gold coinages. 65 P. Grierson, “The patrimonium Petri in illis partibus and the pseudo-imperial coinage in Frankish Gaul”, Revue belge de numismatique 105 (1959), pp. 95-111, at. p. 95. This is by far the most serious attempt to put the coins in context. Its conclusions are nevertheless retracted in Grierson and Blackburn, Medieval European Coinage, I: the Early Middle Ages (5,h10th centuries), p.130; but cf. Hendy, “From public to private: the western barbarian coinages as a mirror of the disintegration of late Roman state structures”, p. 70.

MARSEILLE AND THE PIRENNE THESIS I

225

can it be a single royal authority since, in the divided Merovingian pol­ ity of the 570s, the main issuing centres pertain to different kingdoms. The political division of the main mint sites has usually been seen as the problem. Instead, I think it may provide a clue to the solution. Happily, the first appearance of the quasi-imperials can be put into a political context, since it occurs in precisely the period to which Gre­ gory of Tours devotes the bulk of his Histories. And here we return to those indirect indications of Marseille’s significance which underpin the mechanisms of control outlined above. In the Histories the recurrent power-struggles in Marseille operate on two levels. The first is that of local animosities and conflicts among the learned Provençal élite. The intensity of their rivalries is evident, even if the reasons which underlie them frequently are not. This is pardy because in the Histories these conflicts are articulated only within the wider context of the panFrankish second level, that of the close and complex political and social relations between the magnates of Marseille and their lords and mas­ ters at the royal courts.66 In Gregory’s narrative and to a lesser extent in the poems of his friend and contemporary Venantius Fortunatus, the patricius Dynamius, Bishop Theodore and the lesser actors on the com­ paratively well-lit stage of late sixth-century Marseille stand out not as passive bystanders on the periphery of the Merovingian world, bending this way and that with the winds of political change, but as men inex­ tricably embroiled (on different sides) in the high-court power-struggles at the centre. Their peripheral location may enable them to scheme with some success, but their freedom of action is limited, and its exer­ cise is risky: a duke is always likely to be sent in by the vigilant kings, and neither patrician nor bishop can resist the central power for long. The reach of the Merovingian kings comfortably extends as far as Marseille. But Gregory’s narrative reveals more than that, as can be seen if the high politics is divorced from the local dimension in which it is embedded in the Histories. In his peculiarly episodic way Gregory of Tours shows how, in the internecine and often frankly bewildering game of contemporary Merovingian power-politics, Marseille was a bargaining counter of sin­ gular importance. Splicing the episodes together, and excluding the lo­ cal dimension, the first element of note is the peculiar treatment of 66 The main episodes are Historiae IV.43; VI.7; VI. 11, though minor but significant allusions to some of the leading characters involved in them and to Marseille lurk throughout the Histories. For detailed discussion see Loseby, Marseille in Late Antiquity and the early Middle Ages, ch. 5.3.

226

S.T. LOSEBY

Marseille in the divisio regni which followed the death of Chlothar I in 561. Although the bulk of Provence fell to Guntram, Marseille went to Sigibert.67 Now, if the mechanics of the various partitions of the Merovingian kingdom are difficult enough, the imperatives behind them frequently defy explanation. But in the case of Marseille, a cityterritory of negligible size and less agricultural potential, the raison d’etre of this peculiar arrangement must be sought in the profits to be made from its trade. This is further brought out in the concession to Guntram of half of Marseille which was made by the magnates of Sigibert’s heir, the minor Childebert II, ‘on his father’s death’.68 This was clearly the price of Guntram’s goodwill, sealed by a treaty at Pompierre in 577.69 It is equally clear, I think, that this refers not to any territorial division of Marseille or of the lands dependent upon it, but to the fi­ nancial division of what makes the port of such capital importance to the Frankish kings, namely the receipts from tolls on its trade. But in 581 a coup in Childebert’s kingdom sees a new faction of dominant magnates reverse existing policy by abandoning the arrangement with Guntram in favour of an alliance with Childebert’s other surviving roy­ al uncle, Chilperic.70 Childebert duly demands the conceded half of Marseille back, highlighting how central it was to the original deal.71 When Guntram refuses, he strives to assert his control by force, but with only temporary success. However, in 584, Gregory baldly reports that Guntram restored Marseille wholly to Childebert, of his own ac­ cord; the explanation is to be found two chapters further back in Gre­ gory’s narrative, where Childebert’s ambassadors cite Guntram’s con­ trol over Marseille and his harbouring of fugitives from Childebert’s kingdom as the reasons for their enmity.72 Not surprisingly, therefore, the return of Marseille is soon followed by a renewal of the old alliance between Childebert and Guntram.73 If Marseille is not among the cit­ 67 The best (though I think slightly erroneous) discussion of the impact of this division on Provence is E. Ewig, “Die fränkischen Teilungen und Teilreiche (511-613)”, reprinted in E. Ewig, Spätantikes und Fränkisches Gallien (Zurich/Munich, 1976), pp. 114-71, at pp. 135-8. Sigibert’s dominions form the so-called “Austrasian corridor” discussed for exam­ ple in R. Buchner, Die Provence in merowingischer £eit: Verfassung - Wirtschaft - Kultur (Stut­ tgart, 1933), pp. 10-11. But it is doubtful whether this division is best envisaged in terri­ torial terms. 68 Gregory of Tours, Historiae VI. 11. 69 Gregory of Tours, Historiae V. 17. 70 Gregory of Tours, Historiae VI. 1; VI.3. 71 Gregory of Tours, Historiae VI. 11. 72 Gregory of Tours, Historiae VI.33; VI.31. 73 Gregory of Tours, Historiae VI.41.

MARSEILLE AND THE PIRENNE THESIS I

227

ies mentioned in the text of the treaty which the two kings finally con­ cluded at Andelot in 58774, then this is because its return was of pri­ mary importance, a necessary precondition for any such agreement. The close watch which the Merovingian kings exercised over Marseille, and the central significance which the port held in their con­ flicts in the 580s, is explicable only in terms of the enduring commer­ cial vitality noted by Pirenne from the written sources, and now re­ emphasised by the archaeology. The institution of the cellariumfisci rep­ resents one attempt by the Merovingian kings to maximise their ex­ ploitation of this resource. To my mind, the emergence of the quasi­ imperial coinage in precisely the period described by Gregory of Tours is another. The problem of responsibility for this coinage would be sim­ plified is we could admit the possibility of economic co-operation be­ tween kingdoms in tapping the traffic plying the lower Rhone valley trade-route. Gregory’s narrative shows that the necessary political con­ ditions for such an arrangement - in the form of an alliance between Guntram and Childebert - had indeed existed in the latter part of the 570s, the very period in which the quasi-imperials were introduced.75 The existence of such a compromise would then provide one possible explanation for the decision to mint the coins in the name of the east­ ern emperor.76 And royal control from the outset would help to ac­ count for the seamless transition under the sole rule of Chlothar II from the quasi-imperial to the royal coinage. No-one would doubt that the royal coinage of the seventh century was controlled by the Merov­ ingian kings. The same surely goes for its quasi-imperial antecedent, otherwise identical in every respect. In any event, the name on the coins matters less than the destination of the profits, which, as the re­ peated impositions of Merovingian authority indicate, is always likely to have been the crown. And if such a deal had been struck between kingdoms, then no wonder Marseille was right at the heart of Merov­ ingian politics in the time of Gregory of Tours.

74 Gregory of Tours, Historiae IX.20. 75 These two kingdoms between them include all the mint-sites at which the quasiimperial coinages were struck. 76 This is a rather more specific version of the “compromise” solution suggested by Lafaurie, “Les monnaies de Marseille du VIe au Vili'' siècle”, p. 71, and followed in Grierson and Blackburn, Medieval European Coinage, I: the early Middle Ages (5th-10,h centu­ ries), p. 130.

228

S.T. LOSEBY

Conclusion: ‘Un grand port’ The new archaeological evidence combines with the reconsideration of the old documentary and numismatic data to suggest that Pirenne’s in­ stincts about the significance of Marseille in the post-Roman period, and about the commercial basis of its vitality, were entirely justified. Marseille flourished in the age of Gregory of Tours and on into the seventh century because it dominated a still active trade-route between Frankia and the Mediterranean, and because it could function effec­ tively as an emporium, the role for which it was founded, and which over the longue durée has marked its periods of greatest importance and prosperity. But Marseille cannot have been prospering in a commercial vacuum. Its dynamism therefore suggests that, to misquote Mark Twain, rumours of the death of Mediterranean trade before 600 have been gready exaggerated. And if we prefer to emphasise the complex­ ity of post-Roman Mediterranean exchange-networks, comprehension of which ‘would require the mind of one of those chess-players who can play fifty games at once while blindfolded’,77 this need not pre­ clude our recognition of the Merovingian kings as grandmasters, and of Marseille as their preferred opening. The scale of Merovingian pow­ er and the sophistication of their administration has until recendy been underestimated by historians, their attention distracted by the lethal combination of Gregory of Tours’ very individual historiographical agenda and the dexterity of Carolingian propaganda.78 The Merovin­ gians, like all early medieval rulers, needed to maximise their exploita­ tion of available resources, and this dictated the close integration of Marseille into the Merovingian polity and the development of partic­ ular mechanisms of control and profit centred on the port.79 None of this makes sense unless there was something worth exploiting, and in the case of Marseille that something can only have been trade: the ar­

77 C. Wickham, “L’Italia e l’alto medioevo”, Archeologia Medievale XV (1988), pp. 10524, at p. 111. 78 A much more positive view of the Merovingians is for example offered in I.N. Wood, The Merovingian Kingdoms, 450-751 (London, 1994). 79 For a stimulating discussion of early medieval kingship and ‘economic policy’, see J.R. Maddicott, “Trade, industry and the wealth of King Alfred”, Past and Presad 123 (1989), pp. 3-51, with the subsequent critiques by R. Balzaretti andJ.L. Nelson and re­ sponse by Maddicott in “Debate: trade, industry and the wealth of King Alfred”, Past and Present, 135 (1992), pp. 142-188. Like the latter, I think that early medieval kings were keen ‘to don the purple of commerce’ and would take vigorous steps to do so if they considered the rewards sufficiendy worth their while.

MARSEILLE AND THE PIRENNE THESIS I

229

chaeological evidence now affirms that this was indeed the case. Gre­ gory of Tours may never have gone to Marseille, but he knew a great deal, indeed perhaps all too much, about the port, its magnates, and its importance in Merovingian politics. In his day Marseille was indeed a ‘grand port’. Whether it would remain so for as long as Pirenne imagined will be the subject of a second paper.

THE FRONTIERS OF WESTERN EUROPE: DEVELOPMENTS EAST OF THE RHINE IN THE SIXTH CENTURY Ian Wood

The search for continuity after the fall of thé West Roman Empire has been pursued at many levels: over a wide geographical scale on the one hand,1 and through the minutiae of government on the other.2 Nevertheless the question of continuity or change affects so large a can­ vas, or perhaps so many canvases, that complete landscapes can be for­ gotten. However many instances of continuity the historian may find, the Empire in the West vanished in the fifth century. In the Balkans imperial control was at best patchy, and after the mid sixth century rarely as extensive as that.3 In the Middle East, south of the Taurus, it survived until the second quarter of the seventh century, while south of the Mediterranean, Roman Africa lasted in some measure, with a not-so-brief interlude under the Vandals, until the 690s. If one looks beyond what had been the frontiers of the Roman Empire, the changes are equally significant, though here one is looking not at a world sud­ denly made subject to barbarian rule, but at one which became per­ meable to traditions of the South and West in new ways.4 Although the Franks made a distinction between the eastern kingdom, Austrasia, and the lands east of the Rhine,5 their rule extended far to the east, and, while it is impossible to be specific about where any border lay, Frankish hegemony in the sixth century included Thuringia in the 1 H. Pireime, Mohammed and Charlemagne (London, 1939); R. Hodges and D. Whitehouse,^Mohammed, Charlemagne and the Origins ofEurope. Archaeology and the Priemte Thesis (Lon­ don, 1983). 2 P.S. Barnwell, Emperors, Prefects and Kings: the Raman West 395-565 (London, 1992). 3 H. Wolfram, Geschichte der Goten, 2nd ed. (Munich, 1980); H. Wolfram, Die Geburt Mitteleuropas (Vienna, 1987); P. Heather, Goths and Romans 332-489 (Oxford, 1991); W. Liebeschuetz, Barbarians and Bishops: Army, Church and State rii the Age of Arcadius and Chry­ sostom (Oxford, 1990); W. Pohl, Die Awaren (Munich, 1988); M. Whitby, The Emperor Mau­ rice and his Historian: Theophylact Sxmocatta on Persian and Balkan Wars (Oxford, 1988), pp. 55191. 4 On old permeability, L. Hedeager, “Empire, frontier and the barbarian hinterland. Rome and Northern Europe from AD 1-400”, in M. Rowlands, M.T. Larsen and K. Kristiansen, eds., Centre and Periphery in the Ancient World (Cambridge, 1987), pp. 125-40; P. Brown, The Rise of Western Christendom (Oxford, 1996), pp. 15-16. 5 W. Fritze, Untersuchungen zurjriihslawischen undfiihfränkischen Geschichte bis ins 7. Jahr­ hundert (Frankfurt, 1994), pp. 242-4, 265-7.

232

IAN WOOD

North and Bavaria in the South. Yet this eastern hegemony of the Merovingians had reached its limits by the second quarter of the sev­ enth century, when the edge of the Frankish world came to be marked not by a geographical barrier, but by the culture and language of the Slavs and the Avars. By then the Danube was no longer a frontier, but a waterway running through the heart of the Avar world. Because the history of sixth-century Frankish expansion in the East, and its seventhcentury check at the hands of the Avars and Slavs, is a fragmentary one, it is useful, yet again, to put the fragments together,6 to provide a context for some of the patterns of exchange in the sixth century. The history of Childeric I (d. c.481) provides a reasonable starting point. According to Gregory of Tours, Childeric was rejected as ruler by the Franks, because of his sexual behaviour, and he fled to the king­ dom of the Thuringians, where he was received by king Bisinus. There he remained until advised that it was safe to return. On his return he was followed by Basina, Bisinus’s wife.7 That Childeric went into exile in Thuringia, and that he married a Thuringian wife, is perfectly pos­ sible. That the lady was already married to his host is perhaps less than likely. The story may be the first of a series of untruths about Thurin­ gia peddled by the bishop of Tours. The name ‘Basena’ is engraved on a silver spoon from Weimar, dated to the early sixth century,8 which suggests that it did not go out of fashion in Thuringia - as one might have expected had Bisinus’ queen eloped with Childeric.What­ ever the truth, Bisinus, Pisen, or even Fisud, is known from other sources.9 In none of them is he portrayed as a cuckolded husband. Per­ 6 This exercise has been made immeasurably easier with the publication of Fritze’s 1952 dissertation, Untersuchungen zurfrühslawischen undfrühfränkischen Geschichte bis ins 7. Jahr­ hundert. 7 Gregory of Tours, Decent Libri Historiarum II. 12, ed. B. Krusch and W. Levison, Monumenta Germaniae Histórica, Scriptores Rerum Merovingicanan 1. 1 (1951); see also the elab­ oration in Fredegar, III 11-12, ed. B. Krusch, Monumenta Germanica Histórica, Scriptores Re­ rum Meromngicarum 2 (Hannover, 1888). 8 Paulys Real Encyclopädie der classichen Altertumswissenschqfl, part 2, vol 11, 2nd ed. (Stut­ tgart, 1936), p. 642; B. Schmidt, Die späte Völkerwandersungzeit in Mitteldeutschland, Veröffent­ lichung des LandesmuseumsJur Vorgeschichte Halle, voi. 25 (Berlin, 1970), p. 84, plate 91, 1; S. Hauser, Spätantike undfrühbyzantinische Siberlöffel. Bemerkungen zur Produktion von Luxusgütem im 5. Und 7.Jahrhundert, JahrbuchJur Antike und Christentum, Erganzungsband 19 (Münster, 1992), p. 26; on the function of such spoons (for food rather than liturgical use) see M. Schmauder, “LöfTel”, Spätantike undfrühes Mittelalter: Ausgewählte Denkmäler im Rheinischen Landesmuseum Bonn, ed. J. Engermann and C.B. Rüger (Cologne, 1991), pp. 290-294.1 am indebted to Michael Schmauder for advice on the interpretation of the ‘Basena’ spoon. 9 Venandus Fortunatus, Vita Radegundis 2, ed. B. Krusch, Monumenta Germaniae Histó­ rica, Auctores Antiquissimi 4. 2 (Berlin, 1885); Origo Gentis Langobardorum 4, ed. G. Waitz, Monumenta Germaniae Historia, Scriptores Renan Langobardicarum et Italicanan (Hannover, 1878); Historia Langobardorum Codicis Gotham 4-5, ed. Waitz, Monumenta Germaniae Historia, Scriptores Rerum Langobardicarum et Italicanan.

THE FRONTIERS OF WESTERN EUROPE

233

haps more reliable and more significant for the future is Gregory’s comment that Childeric’s son, Clovis (c.481-511) defeated the Thuringians, in the tenth year of his reign, that is c.491.10 This victory might be linked to a passage in Procopius, commenting on Thuringian fear of the Franks after the death of the first barbarian ruler of Italy, Odoacer.11 In this context it may be significant that, according to one source, Odoacer’s father was a Thuringian.12 That Odoacer’s death might have had repercussions north of the Danube can be inferred from other evidence. His interest in the area can be seen most clearly in his dealings with, or rather conquest of, the Rugian kingdom of Feletheus, based on the north bank of the river, opposite the failing Roman province of Noricum.13 It was an interest which continued, and was probably enhanced by his Italian successor, the Ostrogoth Theodoric. The latter drew several of the peoples north of the Danube into a web of alliances,14 and he attempted to use his contacts north of the Danube, among them the kings of the Heniles, the Warni and the Thuringians, to put pressure on the Frankish king Clovis, whose expansionist ambitions were becoming ever clearer, and to avert a war with the Visigoths in 507.15 In the event, the alliances were of little use, and Clovis defeated Alaric at the battle of Voulon.16

10 Gregory of Tours, Decern Libri Historiarum II. 27 11 Procopius, Wars V. 12. 21, ed. H.B. Dewing (London 1914-38) 12 H. Castri dus, “Zur Geschichte der Heermeister des Westreichs”, Mitteilungen des Instituts ßir österreichische Geschichtsforschung 92 (1984), p. 25; A. Demandi, “The Osmosis of Late Roman and Germanic Aristocracies”, in E. Chrysos and A. Schwarcz, eds., Das Reich und die Barbaren (Vienna, 1989), pp. 81-2: more generally on the problems of de­ scribing Odoacer, I.N. Wood, “The Ethnogenesis of the Burgundians”, p. 63, in H. Wol­ fram and W. Pohl, ed. Typen der Ethnogenese und besonderer Berücksichtigung der Bayern, voi. 1 (Vienna, 1990). 13 Eugippius, Vita Seuerini 44.2, ed. P. Régerat, Sources Chrétiennes 374 (Paris, 1991); Paul the Deacon I. 19, ed. Waitz, Monumenta Germaniae Historia, Scriptores Rerum Longobardicarum et Italicarum. On Feletheus and his kingdom see also Eugippius, Vita Seoerini 8. 1; 22. 2; 31. 1; 40. 1; 42. 1; 44. 1; Wolfram, Die Geburt Mitteleuropas, pp. 63-8. 14 For Theodoric’s marriage alliances, Procopius, Wars V. 12. 21-2; Anonymus Valesianus II. 63, 68, 70, ed. J. Moreau, Excerpta Valesiana (Leipzig, 1961); Jordanes, Getica 58 (297-9), ed. T. Mommsen, Monumenta Germaniae Histórica, Auctores Antiquissimi 5 (Berlin, 1882); Cassiodorus, Variae IV. 1, ed. T. Mommsen, Monumenta Germaniae Histórica, Auctores Antiquissimi 12 (Berlin, 1894). The chronology of these marriages is, however, important, with some taking place in the decade or more prior to 507 and others taking place there­ after: see, for example, J. Moorhead, Theodoric in Italy (Oxford, 1992), pp. 51-4, 192-3, with n. 87. 15 Cassiodorus, Variae III. 3; see also Variae III. 1, 2 and 4, addressed respectively to Alaric II, Gundobad and Clovis; see also Variae IV. 2 to the king of the Heniles. 16 Following the identification of R.A. Gerberding, The Rise of the Carolingians and the Liber Historiae Francorum (Oxford, 1987), p. 41.

234

IAN WOOD

This, however, led not to an abandonment of Ostrogothic policy north of the Danube, but rather to a strengthening of it. In particular Theodoric married his niece, Amalberga, to the Thuringian king Hermenefred.17 The marriage of Amalberga to Hermenefred is one indication that Thuringia in the second decade of the sixth century was no backwater. Amalberga herself seems to have been well educated.18 This, of course, need not imply that she married in to a family with any cultural pre­ tensions: the Merovingian princess, Chlodoswintha, correspondent of Nicetius of Trier, was sent ofT to marry to Lombard Alboin,19 a king seemingly in the heroic rather than civilised mold.20 Nevertheless, in describing the childhood of Hermenefred’s niece, the later Merovin­ gian queen and saint, Radegund, Venantius Fortunatus implies the ex­ istence of a Thuringian royal court.21 The cultural significance of that court may be indicated by the high-quality metal work which has been associated with the Königshoj,? 2 One might even guess that Christianity had made considerable headway at the court, before Thuringia was in­ corporated into the Frankish state. At the very latest, it must have reached Thuringia with the arrival of Amalberga. It is notable that no other source suggests that Radegund was brought up as anything other than a Christian. Perhaps more significant is the sixth-century ‘litur­ gical’ spoon, apparently of Byzantine origin, inscribed with the name ‘Basena’.23 This may even be an indication that the Thuringians were christianised before they were incorporated into the Frankish state. It is notable that no source suggests that Radegund was brought up as anything other than a Christian.

17 Cassiodorus, Variae IV. 1; Procopius, Wars V. 12. 21-2; Anonymous Valesianus II. 70; Moorhead, Theodoric in Italy, pp. 192-3, with n. 87. 18 Cassiodorus, Variae IV. 1; Moorhead, Theodoric in Italy, p. 88 19 Epistolae Austrasiacae 8, ed. W. Gundlach, Monumenta Germamae Histórica, Epistolae Merowingici et Karolini Aevi 1 (Berlin, 1892); Gregory, Decern Libri Historiamm IV. 3, 41. 20 Paul the Deacon, Historia Langobardorum I. 27. 21 Venantius Fortunatus, De Excidio Thoringiae, ed. F. Leo, Monumenta Germamae His­ tórica, Auctores Antiquissimi 4. 1 (Berlin, 1881), Appendix 1. 22 B. Schmidt, “Die Thüringer”, p. 544, in B. Krüger, ed., Die Germanen (Berlin, 1986). Ulf Näsman tells me that the evidence for glass production points in the same direction: see, G. Behm-Blancke, Gesellschaft und Kunst der Germanen. Die Thüringer und ihre Welt (Dresden, 1982), and B. Schmidt, “Hermunduren-Wamen-Thüringen. Zur Ge­ schichte des 3. bis 6. Jahrhunderts in Saalegebiet”,Jahresschriftfiir Mitteldeutsche Vorgeschichte 65 (1982), pp. 173-215. 23 Paulys Real Encyclopädie der classischen Altertumswissenschaft, part 2, vol. 11, 2nd ed. (Stuttgart, 1936), p. 642. But see above n.8.

THE FRONTIERS OF WESTERN EUROPE

235

The history of the Frankish take-over of Thuringia begins after Clovis’ death in c.511. Following Gregory of Tours Hermenefred murdered his brother Batterie, father of Radegund.24 Driven on by his wife, Theodoric’s niece Amalberga, he turned against another brother, Berthecar, offering Clovis’s son, the east Frankish king Theuderic I (511-533), half his kingdom if he would help remove him. Theuderic consequent­ ly killed Berthecar, but received nothing for his pains.25 Some years later, in c.531, Theuderic turned to his half-brother, Chlothar I (511561), and together they destroyed the Thuringian kingdom.26 Hermenefred’s son, Amalfred, escaped to Italy and thence to Constantinople, where he became a general of some note.27 His niece Radegund was taken captive, and married to Chlothar.28 Gregory’s account of Hermenefred’s dealing with his brothers is al­ most certainly fallacious.29 In particular, sources relating to Radegund consistently imply that she was on good terms with her uncle Herme­ nefred, her father’s supposed murderer,30 and depict her as extremely close to Hermenefred’s son, Amalfred.31 Gregory’s depiction of the bloodthirsty Hermenefred and his Ostrogothic wife seems to have much in common with his depiction of the Burgundian king Gundobad, whom he also sees as a fratricide,32 although a better-informed source rather undermines this picture.33 Probably Gregory’s accounts are a reflection of Frankish propaganda,34 but it is difficult to know

24 Gregory of Tours, Decern Libri Historiamm III. 4. 25 Gregory of Tours, Decern Libri Historiarum III. 4. 26 Gregory of Tours, Decern Libri Historiamm, III 7; Procopius supports a date subse­ quent to the death of Theodoric in 526, Wars V. 13. 1. 27 Procopius, Wars VIII. 25. 11-13; see also Venantius Fortunatus, De Excidio Thuringiae (= Appendix 1) on Radegund’s attempts to make contact with Amalfred. 28 Gregory, Decern Libri Historiamm III. 7. 29 B. Schmidt, “Die Thüringer” p. 504, n. 155, plausibly dismisses the story as an attempt to legitimise the Frankish conquest of Thuringia. 30 Venantius Fortunatus, Vita Radegundis 2 (3); id., carni. Vili. 1. 23-24; id., De Excidio Thoringiae 1. 50; id., Ad Artachin (Appendix III), 1. 7. 31 Venantius Fortunatus, carni. V ili 1, 24; id., De Excidio Thoringiae presents the fic­ tion of being written by Radegund to Amalfred; id., Ad Artachin is a lament, also purport­ edly written by Radegund, on Amalfred’s death. 32 Gregory, Decern Libri Historiamm II. 28, 32-3. Amalberga can be compared to a number of bloodthirsty women active in the pages of Gregory: see Clovis’ own sister, Audofleda, Decern Libri Historiamm III. 31, and the second wife of Sigismund, Decern Libri Historiarum III. 5. 33 Avitus of Vienne, ep. 5, ed. R. Peiper, Monumenta Germaniae Histórica, Auctores An­ tiquissimi VI (2) (Berlin, 1883). 34 Schmidt, “Die Thüringer”, p. 504, n. 155.

236

IAN WOOD

what his friend Venantius Fortunatus, who was himself a confidant of Radegund, would have made of them. The Frankish move against the Thuringians in c.531, like their de­ struction of the Burgundian kingdom in 534, can plausibly be set in the context of declining Ostrogothic power, following the death of Theodoric in 526.35 The treatment of both the Thuringian and Bur­ gundian kingdoms by the Merovingians was held against the Franks in diplomatic circles.36 Nevertheless, although Radegund’s brother was executed following a Thuringian revolt,37 and although the Thuring­ ians consorted with the Saxons in a revolt against Chlothar in 556,38 they remained an integral part of the east Frankish powerblock, prob­ ably being called upon to provide troops for the king of Austrasia, as did other peoples east of the Rhine,39 until the revolt of Radulf, dux of Thuringia, against the court of Sigibert III in 639.40 Related to the Thuringians were the Warni, who seem to have been based in the Mecklenburg region:41 indeed the two peoples were later subject to a single law-code, the Lex Anglmum, Werinorum hoc est Thuringorum.42 In the early sixth century their king Hermegisl was important enough to have as his queen Theudechild, sister of Theudebert I (533548). Hermegisl seems to have regarded the alliance as being of con­ siderable importance: according to Procopius, before his death, he told his son, Radigis, to marry his step-mother, in order to preserve the Frankish alliance.434 Although Radigis was forced to abandon his Frankish bride, a kingdom of the Wami is said to have lasted until 595.“ Remaining somewhat more independent of the Frankish empire were the Saxons.45 Yet they were made subject by Theuderic, at some point 35 Schmidt, “Die Thüringer”, p. 504, 545. 36 Procopius, Wars VI. 28. 17. 37 Venantius Fortunatus, Vita Radegundis 12. 38 Gregory, Decern libri Historiarum IV. 10; Marius of Avenches, s.a. 556, ed., J. Favrod, La Chronique de Marius d’Avenches (455-581) (Lausanne, 1991). 39 Gregory, Decern Libri Historiarum IV. 49, 50; Fredegar IV. 38, 40; I.N. Wood, The Merovingian Kingdoms, 450-751 (London, 1994), p. 164 40 Fredegar IV 87; Fritze, Untersuchungen zurßühslawischen undJrühßänkischen Geschichte bis ins 7. Jahrhundert, p. 248. 41 It may be that Wami were to be found in more than one region. Unfortunately the written sources are less than helpful on this issue. 42 ed. K. von Richthofen, Monumenta Germaniae Histórica, Leges 5 (Hannover, 1875-89). 43 Procopius, Wars VIII. 20. 11-21. 44 Fritze, Untersuchungen zurßühslawischen undfiühfiänkischen Geschichte, p. 15, citing Fre­ degar IV. 15. 45 I.N. Wood, The Merovingian North Sea (Alingsás, 1983), pp. 9-10.

THE FRONTIERS OF WESTERN EUROPE

237

before 53S.46 They revolted in 555, probably taking advantage of the death of Theuderic’s grandson, Theudebald (548-55), and the conse­ quent transfer of power.47 They revolted again in 556, when they were joined by the Thuringians.48 Following the Frankish response to the revolt of 556 they were forced to pay an annual tribute of 500 cows.49 Clearly, they tried to throw off this burden at various moments, in the period between 561 and 575,50 in c.605,51 and again between 622 and 629.52 It was, however, only in the 630s that they exchanged their trib­ utary status for an obligation to protect the eastern frontier of the Franks against the Wends.53 In this tributary period they also lost some land to the Sueves. When a substantial body of Saxons accompanied Alboin in his trek into Italy, Sigibert I (561-75) reallocated their lands to a group of Sueves.54 When these same Saxons became disenchanted with the Lombard kingdom they tried to return to their homelands, receiving permission from Sigibert, but unexpectedly meeting consid­ erable opposition from the Suevic incomers.55 The land which lay between the Franks and the Saxons scarcely fea­ tures in the written record before the late seventh century. Even those parts of the lower Rhineland which were under Frankish control are rarely mentioned. Theudebert I had to deal with a Danish raid, during the reign of his father, Theuderic I. The defeat and death of the Dan­ ish king Chlochilaich, or Hygelac, may have had significant repercus­ sions in Denmark itself: no other named Danish king appears in the historical record for another century and a half. It would, however, be unwise to conclude much from this silence. Chlochilaich’s raid is one of the rare moments that the lower Rhine, a region much closer to 46 Implied by Gregory, Decern Libri Historiarían IV. 10. 47 Marius, s.a. 555; Gregory, Decern Libri Historiarían IV. 10, 14, 16. 48 Marius, s.a. 556; Gregory, Decern Libri Historiarían IV. 10, 14, 16; Fredegar, IV 74. 49 Gregory, Decern Libri Historiarían, IV 14; Fredegar IV. 74. 50 Liber Historiae Francorum 31, ed. B. Krusch, Monumenta Germaniae Histórica, Scriptores Remm Merovingicarum 2 (Hannover, 1888). 51 An additional uprising shortly after 600 is mentioned by Paul the Deacon, Historia Langobardorum IV. 31. 52 Liber Historiae Francorum 41. 53 Fredegar IV 74; Fritze, Untersuchungen zurfrühslawischen undßühfränJdschen Geschichte, p. 105 sees this as a mark of considerable Frankish weakness. 54 Gregory of Tours, Decern Libri Historiarum IV. 42; V. 15. See also Paul the Deacon, Historia Langobardorum II. 6. If Gregory and Paul are not mistaken in naming Chlothar and Sigibert as the kings who organised the Suevic settlement, then Sigibert must have been active in Austrasia before his father’s death. 55 Paul the Deacon, Historia Langobardorum III. 5-7; see also Gregory, Decern Libri His­ toriarum V. 15.

238

IAN WOOD

Francia and its writers than Denmark, appears in our written sources before the late seventh century. Even the origins of the Frisian state remain obscure until then.56 While the Franks themselves already controlled Cologne and Zülpich by the end of the fifth century, further south the Rhineland was in the hands of the Alamans. This originally powerful Germanic confederation was defeated by Clovis, probably in a series of cam­ paigns, culminating in the acquisition of their land, the centre of which appears to have lain in modern Schwaben, probably towards the year 506.57 Some Alaman survivors crossed the Alps to join Theodoric’s Goths in Italy.58 Thereafter Alamannia may have been integrated into the powerblock east of the Rhine controlled by the Austrasians,59 while remaining, like Thuringia, under the immediate control of its own duc­ es. Frankish involvement to the area east of the Alamans, in the region of the upper Danube, is attested, if only sketchily. At the very least, it is possible to assume Merovingian involvement in the development of the duchy of Bavaria.6061This is most clearly stated in the preface to the Leges Baiwariorum.^ The Agilulfing family, which provided the duces of Bavaria, seems to have been Frankish and to have been established in power by the Merovingians.62 In the narrative sources we find Theudebald disposing of an unwanted wife, Walderada, daughter of the Lom­ bard king Waccho, on dux Garibald.63 One of Garibaldi daughters would later marry the Lombard dux, Evin of Trent.64 Meanwhile the 56 I.N. Wood, The Merovingian Kingdoms, 450-751, pp. 160-1. 57 On the problems of the chronology of Clovis’ Alaman wars see especially A. Van de Vyver, “L'unique victoire contre les Alamans et la conversion de Clovis en 506”, Revue Belge de Philologie et d’Histoire 17 (1938), pp.793-813. 58 Ennodius, Panegyricus dicto Theodorico, XV (72-3), ed. F. Vogel, Monumenta Germaniae Histórica, Auctores Antiquissimi 7 (Berlin, 1885); Cassiodorus, Variae II. 41; XII. 28. 59 Fritze, Untersuchungen zur fiiihslawischen undJrühfränkischen Geschichte, p. 243, righdy draws attention to the uncertainty of the position of Alamannia: On the evidence of Agathias I, 4, see A. Cameron, “Agathias on the early Merovingians”, Annali della Scuola Normale Superiore di Pisa, Lettere, Storia e Filosofia, 2nd series, 37 (1968), pp. 100-1, 107, 121. 60 On the history of Bavaria in this period, M. Spindler, ed., Handbuch der bayerischen Geschichte, 1 Das Alte Bayern: Das Stammesherzogtum, 3rd ed. (Munich, 1975), pp. 101-17. See also Wolfram, Die Geburt Mitteleuropas, pp. 91-5. 61 Lex Baiwariorum, prol, ed. E. de Schwind, Monumenta Germania Historia, Leges, 5 (Hannover, 1926); Wood, The Merovingian Kingdoms, 450-751, pp. 116-7. 62 E. Zöllner, “Die Herkunft der Agilulfinger”, Mitteilungen des InstitutsJur österreichische Geschichtsforschung 69 (1951), 245-64; Fritze, Untersuchungen zw fiiihslawischen und fiühfiränkischen Geschichte, p. 78; Spindler, ed., Handbuch der bayerischen Geschichte, 1, pp. 102-3. 68 Paul, Historia Langobardorum I. 21; Gregory of Tours, Decem Libri Historiarum IV. 9. M Paul, Historia Langobardorum III. 30.

THE FRONTIERS OF WESTERN EUROPE

239

Lombard king Authari asked to marry Garibald’s daughter, Theudelinda.65 This potential marriage connection may have suggest­ ed a degree of independence from the Franks which the Merovingians had no wish to accept: following Paul the Deacon the Franks soon made difficulties for Garibald, and Theudelinda and her brother, Gundoald, fled to Italy, where she married Authari.66 Subsequently the Austrasian king Childebert II (575-596) elevated another member of the Agilulfing family in Bavaria: according to Paul he appointed Tassi­ lo as king of the Bavarians:67 it is by no means clear, however, that a Frankish monarch would have conceded such a tide. Moreover, in about 630 the Merovingian king Dagobert I (623-639) could rely on the Bavarians to massacre a group of Bulgars, who had been setded among them, following defeat at the hands of the Avars.68 Meanwhile, connections between the Lombards and the Bavarian Agilulfings con­ tinued. One Bavarian ruler married the daughter of Gisulf II of Friuli, sometime after 610.69 As for the area of Rhaetia which lay between Bavaria and Alamannia, there is a litde to be learnt about its relations with the Frankish kingdom from Jonas of Bobbio’s account of Columbanus’ period at Bregenz, which it depicts as being under the authority of Theudebert II (596-612),70 and from the fragmentary Vita Galli ve­ tustissima, which describes the region of the Bodensee in the early sev­ enth century.71 In order to fill out the picture of Merovingian interests in the Alpine region, it is necessary to look at Frankish interventions in north-eastern Italy, since the Austrasians controlled Venetia for substantial parts of the mid to late sixth century, and Frankish dominance of Venetia nat­ urally implies control of certain Alpine passes. This Merovingian activ­ ity south of the Alps, which is recorded almost entirely in non-Frankish sources, needs to be understood in terms of the relations between the Franks, the Byzantines and the barbarian invaders of Italy, first the Os­ trogoths and then the Lombards. Already in 535, the Emperor Justin­

65 Paul, Historia Langobardorum III. 30. 66 Paul, Historia Langobardorum III. 30; Fritze, Untersuchungen zurfiühslawischen undfrühfränkischen Geschichte, pp. 75-6. 67 Paul, Historia Langobardorum IV. 7. 68 Fredegar IV. 72. 69 Paul, Historia Langobardorum IV. 37. 70 Jonas, Vita Columbani I. 27, ed. B. Krusch, Monumenta Germaniae Histórica, Scriptores Rerum Merovingicarum 4 (Hannover, 1902). 71 Vita Galli vetustissima, new ed. I. Müller, “Die älteste Gallusvita”, Zeitschrift Jiir Schweizerische Kirchengeschichte 66 (1972), pp. 209-49.

240

IAN WOOD

ian had invited the Franks to join the war against the Goths, to which they agreed.72 This intervention led the Gothic king Vitigis to cede ter­ ritory in southern Gaul to the Franks, in return for which they secredy changed sides and surreptitiously supplied the Goths with troops.73 Among the kings supplying troops was the east Frankish ruler Theude­ bert I, who is said to have sent Burgundians, who were instructed to present themselves as if acting of their own accord.74 According to Pro­ copius, Theudebert was concerned to conquer Italy for himself.75 This intervention of Theudebert came to nothing because his troops were struck with dysentery and had to be withdrawn.76 Nevertheless, again following Procopius, he showed increasing ambition, racing horses (in imperial fashion) in Arles, and putting his image on his gold coins, something which even the Persians did not do.77 More important, while Totila was retaking land from the Byzantines in southern Italy, Theudebert was able to take control of Venetia.78 By the time that Theudebert died in 548, he had subjected to Frankish domination Lig­ uria (essentially modern Lombardy), the Cottian Alps and most of Ve-

72 Procopius, Wars V. 5. 8. 73 Procopius, WarsV. 11. 17-18, 28; V. 13. 16; Marius, s.a. 538. 74 Procopius, Wars VI. 12. 38-9. There are some problems with this account. Ac­ cording to Procopius, Wars VI. 25. 9 the Burgundian troops in Italy in 539 behaved in a pagan fashion, sacrificing women and children: according to Marius of Avenches, s.a.. 538 Milanese priests and senators were killed in the churches of Milan. Averil Cameron is inclined to accept the Byzantine sources at face value, see “Agathias on the early Merovingians”, pp. 139-40. There is, however, a problem in that such “pagan” behav­ iour does not fit with anything else known about the Burgundians, and does fit rather better with what is known about the Alamans: Agathias I. 7: see Cameron, “Agathias on the early Merovingians”, pp. 103, 109, 129. It may be that the chief difficulty concerns the definition of Burgundian. Although later Austrasian kings did not hold land in Bur­ gundy, it is fairly clear that Theudebert and Theudebald did: Marius of Avenches, s.a. 534. What is unclear is which section of the Burgundian kingdom Theudebert took over. Geographically the most logical area would have been the northern part of the kingdom, where there had certainly been Alamanic settlement. It is possible that the “Burgundi­ ans” of this region (of Alaman stock?) had not been as deeply christianised as those of the Rhone and Saône-basin cides, and that it was from these that Theudebert recruited his troops. 75 Procopius, Wars VI 25. 1-4; Cameron, “Agathias on the early Merovingians”, p. 124, points out that Agathias does not share this view of Merovingian policy. Also on Theudeberts policies see Fritze, Untersuchungen zurfrühslawischen undfiühfränkischen Geschiehte, p. 73, 194-203; also R. Collins, “Theodebert I, ‘Rex Magnus Francorum*”, Ideal and Reality in Frankish and Anglo-Saxon Society, ed. P. Wormald (Oxford, 1983), pp. 7-33. 76 Procopius, Wars VI. 25. 17-18; V. 26. 1; Marius of Avenches, s.a. 539. 77 Procopius, Wars VII. 33. 5-6. 78 Procopius, Wars VII. 33. 7; these gains are mentioned retrospectively by Marius of Avenches, s.a. 556.

THE FRONTIERS OF WESTERN EUROPE

241

netia, except for a few forts which were held by the Goths and the coast which was in Byzantine hands.79 The implicadons of Austrasian activity in northern Italy for our un­ derstanding of Bavaria and of the eastern Alps in general are consid­ erable. The control of Venetia in particular is likely to have been ef­ fected through the eastern Alpine passes, even though our evidence for the interest of Theudebert and his son Theudebald in the Danube ba­ sin is fragmentary. Certainly the Franks held Lienz in the Drautal.80 More grandiose, but plausible, are the claims made by Theudebert, in writing to Justinian, that he had jurisdiction from the Danube and the limes of Pannonia to the Ocean,81 by which he presumably means the North Sea. Indeed Theudebert’s claims, although diplomatic propa­ ganda, provide us with our best evidence for Frankish hegemony east of the Rhine in the mid sixth century. Less plausible are the rumours circulating in Byzantium and recorded by Agathias, that Theudebert planned to march along the Danube, and attack Constantinople it­ self.82 Any ambitions the Franks held in that direction will not have been encouraged by the Lombard move into Pannonia.83 Despite diplomatic protests by Justinian,84 Theudebald held on to Venetia after his father’s death, and his troops refused to let Narses pass through the region.85 Even after the defeat of Totila at Busta Gal­ lorum in 552, the Austrasians intervened to prevent the Byzantines tak­ ing Verona from the Goths.86 The Byzantines, nevertheless, continued to think that the Franks intended to conquer Italy for themselves,87 and this assumption was in some degree confirmed by the Italian cam­ paigns fought by Theudebald’s generals, the Alaman brothers Buccelin 79 Procopius, Wars VIII. 24. 6-10. 80 Paul, Historia Langobardorum II. 4. On later Frankish interest in the east Alpine val­ leys see Fritze, Untersuchungen zurfrühslawischen undfrühfränkischen Geschichte, p. 76; Pohl, Die Awaren, pp. 149-50. 81 Epistolar Austrasiacae 20; Fritze, Untersuchungen zurfrühslawischen undfrühfränkischen Ge­ schichte, pp. 213-5, makes unnecesarily heavy weather of this passage: there is absolutely no need to interpret Oceanus as meaning Adriatic. 82 Agathias I. 4; see Cameron, “Agathias on the early Merovingians”, pp. 100-1, 107, 122-3. 83 Paul the Deacon, Historia Langobardorum I. 22; see also Cameron, “Agathias on the early Merovingians”, p. 122, following, E. Stein, Histoire du Bas Empire, vol. 2, trans. J.R. Palanque (Amsterdam 1949), p. 528. On the entry of the Lombards into Pannonia see also N. Christie, The Lombards (Oxford 1995), pp. 31-68. 84 Procopius, Wars VIII. 24. 11 85 Procopius, Wars VIII. 26. 18-19. 86 Procopius, Wars VIII. 33. 5. 87 Procopius, Wars VIII. 34. 18.

242

IAN WOOD

and Leutharius, and Aming.88 For the present argument, none of these campaigns would have been feasible without very strong lines of com­ munication between northern Italy and Austrasia. The campaigns of Buccelin mark a legendary highpoint of Merov­ ingian intervention in Italy, but they were no means the last of the ma­ jor invasions. They were, however, followed by the loss of the lands conquered by Theudebert.89 Nevertheless, at some point between the Lombard invasion of Italy in 568 and the death of Sigibert I in 575, the Franks invaded the Val di Non, taking Nano, and attacking Trent, where the Frankish general Chramnichis was defeated and killed by the Lombard dux Evin.90 The possibility of a renewal of Theudebert’s policy was, however, temporarily halted with Sigibert’s murder. There­ after, although there were substantial Frankish attacks on Italy, largely in the north-west of the peninsula,91 they tended to be associated with imperial diplomacy intended to dislodge the Lombards.92 In time the Franks themselves opted increasingly for diplomatic relations with the Lombards, which for a while seem to have allowed them to exert some sort of superior status over them.93 In the long run the history of Merovingian expansion to the south east was not checked simply by the growing confidence of the Lom­ bards. Arguably more important was the appearance of the Slavs, which was followed by the emergence of an Avar kingdom, based in

88 The problems of establishing Buccelin’s career are substantial: he was clearly in­ volved in campaigns in 553/4: Agathias I. 7; see Cameron, ‘Agathias on the early Merovingians’, pp. 102-3, 108-9, 125-6; additional information on these campaigns may also be found in Marius of Avenches, s.a. 555; Gregory of Tours, Decern Libri Historiarum IV. 9; Menander, fragment 3, ed., R.C. Blockley, The history of Menander the Guardsman (Liverpool, 1985). Gregory provides an apparent duplicate of his description of this cam­ paign in Decern Libri Historiarum III. 32, and in ascribing to Buccelin victories (in reality won in 553/4) he is followed by Paul the Deacon, Historia Langobardorum II. 2. It is not, however, impossible that Buccelin was one of the leaders of Theudebert’s army in 538 or 539, since he is recorded as campaigning for Theudebert in Italy by Jonas, Vita Jo­ hannis Abbatis Reomaensis 15, ed. B. Krusch, Monumenta Germaniae Histórica, Scriptores Rerum Meroivingicarum 3 (Hannover 1896). It should be noted that Marius’ death-date for Buc­ celin, s.a. 555, falls in the Austrasian reign of Chlothar I. 89 Marius, s.a. 556 records a successful Frankish invasion, followed by a defeat and the loss of the Italian lands. Despite the problems in the chronology of Marius’ chronicle, it is likely that this loss followed, and was related to, the death of Theudebald. 90 Paul the Deacon, Historia Langobardorum III. 9. 91 Wood, The Merovingian Kingdoms, 450-751, pp. 167-8. 92 P. Goubert, Byzance avant ITslam, vol. 2, Byzance et l’Occident (Paris 1955), pp. 1213, 19-20; see also B. Bachrach, The Anatomy of a Little War (Boulder 1994), pp. 27-9; Whitby, The Emperor Maurice and his Historian, p. 12. 93 e.g. Paul the Deacon, Historia Langobardorum IV. 40.

THE FRONTIERS OF WESTERN EUROPE

243

what had been Roman Pannonia,94 and which in turn provided an umbrella for further Slavic activity.95 The earliest Slavic settlements in the lower Danube region appear to date from the late fifth to early sixth century.96 Already in the 530s the Slavs were presenting problems for Byzantium along the Danube frontier.97 The Avar threat emerged some twenty years later.98 Even before the Lombard migration to Italy from what had been Pannonia Alboin had allied himself with the Av­ ars.99 When the migration took place Alboin left Pannonia to the Av­ ars, on the understanding that if the Lombards were forced out of the peninsula they could return to their Pannonian territories.100 The Av­ ars, however, had created trouble further to the north-west before they settled in Pannonia. Apparently, after disagreements with Justinian in 562, they launched an unsuccessful attack on the Franks, possibly on Thuringia.101 Subsequently, when the Emperor Justin II, on his acces­ sion, refused to make gifts to the Avars, they went off and ravaged the land of the Franks, apparently defeating and capturing the Austrasian king Sigibert, who had to be ransomed.102 Perhaps as a result of this a treaty was made between the Avars and the Franks, who had to sup­ ply their attackers with food.103 Thereafter there seems to have been a lull in Avar-Frankish gela­ tions. By the end of the century, however, the Lombards under Agilulf renewed their peace with the Avars,104 and shortly after, perhaps in

94 Pohl, Die Awaren, pp. 70-6 95 Pohl, Die Awaren, pp. 147-52; Whitby, The Emperor Maurice and his Historian, pp. 55-6. 96 M Comsa, “Die Slawen im karpatisch-donaulandischen Raum im 6.-7. Jahrhun­ dert”, Zeitschriftf ir Archäologie 7 (1973), pp. 197-228; id., “Einige Betrachtungen über die Ereignisse im 6.-7. Jahrhundert an der unteren Donau”, Slavia Antiqua 21 (1974), pp. 6585. 97 Procopius, Wars VII. 14. 2; on attacks as early as 512, Fritze, Untersuchungen zur frühslawischen undfrühfränkischen Geschichte, pp. 9, 11. 98 Whitby, The Emperor Maurice and his Historian, p. 85. 99 Menander, fragment 12, 1-2; Paul the Deacon, Historia Langobardorum I. 27. 100 Paul the Deacon, Historia Langobardorum II. 7. 101 Blockley, The History ofMenander the Guardsman, p. 267, n. 151, combines, Gregory of Tours, Decern Libri Historiarum IV. 23, with Paul the Deacon, Historia Langobardorum II. 10. It is not clear whether Blockley’s reconstruction is correct, not least because Paul mentions two Frankish victories; Gregory two Frankish defeats, or possibly one victory and one defeat, {Decern Libri Historiarum IV. 23; IV 29). Fritze, Untersuchungen zur frühslawischen undfrühfränkischen Geschickte, pp. 16, 20, dates these wars with the Franks to 561 and 567; Pohl, Die Awaren, p. 45. 102 Menander, fragment 8. Blockley associates this with Gregory, Decern Libri Histori­ arum IV. 29: Pohl, Die Awaren, pp. 45-6. 103 Menander, fragment 11. 104 Paul the Deacon, Historia Langobardorum IV. 4; Pohl, Die Awaren, p. 149.

244

IAN WOOD

c.596, Brunhild, acting as regent for Theudebert II and Theuderic II, bought off the Avars, following an attack launched from Pannonia on Thuringia.105 Within a few years, in 601, the Franks were asking the emperor Maurice to pay them to attack the Avars, but he was unwill­ ing to provide any financial incentive.106 In time, however, the Franks were asked instead to join an alliance which included the Avars and the Lombards. This alliance began when the Avar khagan approached the Lombard Agilulf.107 Thereafter the Avars suggested extending the alliance to include the Franks.108 One result was a combined Avar, Slav and Lombard attack on Istria,109 and there was a subsequent joint Lombard-Slav attack on the imperial-held cities of Cremona, Mantua, Valdoria and Brescello. Avar alliances, however, did not last long, and C.610 the khagan launched his own attack on Lombard Friuli.110 Parallel to this history of the Avars was that of the Slavs. The chro­ nology of their expansion into Carinthia, Bohemia and eastern Thur­ ingia is unclear: the evidence is chiefly archaeological, and Slav archae­ ology is notoriously problematic. The land between the Saale and the Elbe still seems to have belonged to Thuringia in 561.111 One event which may have given the Slavs their opportunity was the destruction of a kingdom of the Wami by the Franks in 595.112 There is some ar­ chaeological material to suggest that Slavs reached Bohemia before the mid sixth century,113 but Germanic finds continue through that peri­ od.114 The first literary evidence for Slavs settled in Bohemia only comes in Fredegar’s account of the career of Samo, in other words to the 620s and 630s.115 Samo was a Frankish merchant, who led a resis105 Paul the Deacon, Historia Langobardorum IV. 11; Pohl, Die Awaren, p. 150. 106 Theophylact Simocatta VI. 3, 6, ed. and trans. M. and M. Whitby, The History of Theophylact Simocatta (Oxford, 1986). 107 Paul the Deacon, Historia Langobardorum TV. 12; Pohl, Die Awaren, p. 159. 108 Paul the Deacon, Historia Langobardorum IV. 24. 109 Paul the Deacon, Historia Langobardorum TV. 24. 110 Paul the Deacon, Historia Langobardorum IV. 37; Fritze, Untersuchungen zur frühslawischen undfrühfränkischen Geschichte, p. 82 suggests that the Avar attack on the Lombards was the outcome of an alliance with Theuderic II, who was currently at war with his brother Theudebert II, an ally of the Lombards; Pohl, Die Awaren, 238-40. 111 Fritze, Untersuchungen zur frühslawischen und frühfränkischen Geschichte, p. 15, citing Paul the Deacon, Historia Langobardorum IL 10. 112 Fritze, Untersuchungen zurfrühslawischen undfrühfränkischen Geschichte, p. 15, cidng Fredegar IV. 15. The geographical situation of this kingdom of the Wami is, however, a matter for surmise. 113 Christie, The Lombards, pp. 19-20. 114 J. Tejral, “Abriss der Entwicklung in Mähren während der Völkerwanderungs­ zeit”, AU-Thüringen 14 (1977), pp. 244-57. 115 Fredegar IV. 48, 68; Pohl, Die Awaren, pp. 256-61. Again, however, Fredegar’s geography is imprecise.

THE FRONTIERS OF WESTERN EUROPE

245

tance movement of Slavs against Avar oppression. Having established Slav independence he was then insulted by the envoys of the Frankish king Dagobert, and consequendy fought against an alliance of Franks, Alamans and Lombards, defeating the Frankish contingent, but being defeated by the Lombards. The uprising against the Avars may be dat­ ed to 623 and the war against Dagobert to 631.116 Less certain is the extent of Samo’s kingdom. It can reasonably be seen as being centred on Bohemia, and may have stretched as far as Carinthia in the south and the middle Elbe in the north.117 The evidence relating to Carin­ thia, however, is by no means clear cut. That a one-time flourishing Roman and then Romano-Gothic society failed in the course of the sixth century is suggested by the abandonment of urban sites.118 Slav setdement is attested archaeologically119 and thereafter, somewhat sketchily, in the written sources.120 Nevertheless the region continued to have links with the Lombard kingdom: in the course of the early seventh century, the Lombard duces Taso and Cracco were able to take tribute from the Slavs of ^ellia, apparendy on the border of Carinthia and the region continued to be tributary to the Lombards until C.740.121 Even so, the emergence under Samo in the 620s, of a sizeable Slav kingdom may be taken as marking an end to one, Frankish, phase of development to the east of the Rhine.122 Frankish expansion east of the Rhine created a new political geog­ raphy. It is necessary, however, to ask whether the developments which took place in the region during the course of this eastwards expansion were anything other than cosmetic. Although the region had been out­ side the Roman Empire, the frontier had always been permeable.

116 Fritze, Untersuchungen zurfiühslawischen undfiühfränkischen Geschichte, pp. 86, 88, gives dates of 623 for the uprising and 631/2 for the war: see also Pohl, Die Awaren, pp. 256-8. 117 Fritze, Untersuchungen zurfiühslawischen undfivhfiiùdàschen Geschichte, pp. 89-99 argues for a kingdom stretching from Carinthia to the Elbe. See the modifications of Pohl, Die Awaren, pp. 260-1. 1,8 Pohl, Die Awaren, pp. 119-20. 119 Pohl, Die Awaren, pp. 119-20: Christie, The Lombards, pp. 67-8. 120 Pohl, Die Awarm, p. 149. 121 Paul the Deacon, Historia Langobardontm, IV 38; Pohl, Die Awaren, p. 259: Christie, The Lombards, p. 94. 122 Fritze, Untersuchungen zur fiühslawischen undfiühfiänkischen Geschichte, pp. 103-7 and 283-5 sees Dagobert’s defeat at the hand of Samo as spelling disaster for the Merovin­ gians, much as subsequent historians have linked the failure of the Merovingians to the defeat of Sigibert III by Radulf in 639; J.M . Wallace-Hadrill, The Long-Haired Kings (Lon­ don, 1962), p. 234: although both versions oversimplify the history of Merovingian de­ cline, they do indicate significant moments of Merovingian failure.

246

IAN WOOD

Kings and magnates had been the recipients of high status gifts sent from the Empire.123 Tradesmen, Roman and Germanic, had crossed the frontiers, while Germanic soldiers of fortune had entered the Em­ pire, to make a career in the army, before returning home.124 So much at least is suggested by the archaeology, and by occasional references in our written sources. Nor was diplomatic activity in the lands to the north of the Danube anything new, even though the circumstances of the fifth and sixth centuries made it more than usually important.125 There is in fact a danger that the wealth of sources relating to this re­ gion in the fifth and sixth centuries may delude us into imagining that we are observing new developments, whereas in the Germanic world, although not that of the Slavs, we may only be seeing culminations of processes already well under way. It is not even clear whether the Ger­ manic world had become more open during the migration period than it had been before. That it was remarkably open in the fifth and sixth centuries is implied by numerous distribution maps of various styles of jewellery and types of precious stone. Such precious stones even crop up in the written record: thus Theodoric wrote to the Æstii (or Estones) about amber from the Baltic.126 Yet more tantalising evidence relating to connections between Constantinople, the Danube region and the Baltic comes in the strange story of the appointment of a new king of the Herules, involving a legation from the Danube region to Denmark, whither one branch of the tribe had recently migrated.127

123 Hedeager, “Empire, frontier and the barbarian hinterland. Rome and Northern Europe from AD 1-400”, pp. 125-40; L. Hedeager, Iron-Age Societies (Oxford, 1992), pp. 155-61; J.M. Fagerlie, Late Roman and Byzantine Solidi found in Sweden and Denmark (New York, 1967); O. Kyhlberg, “Late Roman and Byzantine Solidi: an archaeological anal­ ysis of coins and hoards”, Excavations at Helgö X, ed. B.E. Hovén, et a i (Stockholm, 1986), pp. 19-20; B. Arrhenius, “Connections between Scandinavia and the East Roman Em­ pire in the Migration Period”, From the Baltic to the Black Sea: Studies in medieval archaeology, eds. D. Austin and L. Alcock (London, 1990), pp. 126-9; U. Näsman, “Vendei Period Glass from Eketorp-II”, Acta Archaeologica 55 (1986), pp. 55-116, esp. at pp. 97-8. 124 Arrhenius, “Connections between Scandinavia and the East Roman Empire in the Migration Period”, pp. 118-9, 132, 134-5. 125 Whitby, The Emperor Maurice and his Historian, p. 243 states that the late sixth cen­ tury was ‘a period when the Romans showed an uncharacteristic interest in the peoples beyond the frontiers’. One should not, however, forget the importance of the embassies to Attila, as described by Priscus. 126 Cassiodorus, Variae V. 2. 127 Procopius, Wars VI. 15. On this episode see A. Ellegârd, “Who were the Eruli?”, Scandia 53 (1987), pp. 5-34, who points out that there is no evidence that the Herules originated in Scandinavia.

THE FRONTIERS OF WESTERN EUROPE

247

In certain respects there were new developments in the fifth and sixth centuries. Even if the novelty and the extent of the barbarian migra­ tions have been exaggerated,128 they led to a radically altered percep­ tion of the state of Europe, and to a new political geography. Already in the fifth century, Attila’s empire must have forged new relations be­ tween a whole set of peoples who were subject to him. Meanwhile re­ ligious changes were also being effected. The Christian Goths seem to have been responsible, directly and indirectly, for the evangelisation of a number of peoples north of the Danube, including the Rugi.129 Such vast changes are not always well documented. Following the establish­ ment of the Ostrogothic state in Italy, however, the Germanic world, as it existed east of the Rhine and north of the Danube after the death of Attila, comes more clearly into view. Famously, Theodoric the Great had a policy of creating marriage alliances.130 Nor was it only marriages which created bonds of kinship: Theodoric adopted a Herule king as a son in arms.131 Theodoric’s use of family bonds to create a policy is well-known, specifically because it was commented on by contemporaries. He was, however, by no means the only Germanic king of the period to pursue a policy of marriage alliances.132 Even the Merovingians, who were for much of the time unconcerned about the status of their own wives, exploited marriage on occasion.133 Clovis’ sister married Theodoric,134 and Clovis himself married a Burgundian princess.135 His eldest son Theuderic I married the Burgundian Suavegotha.136 Theuderic’s sister married the Visig128 See the revisionist argument of W. Goffart, Barbarians and R o m a n s: Techniques o f Accommodation AD 418-584 (Princeton, 1980), chapter 1. 129 On Rugían arianism, Eugippius, Vita Seoerini 4. 12; 5. 2; 8. 1; on the likelihood that they were converted by the Goths, L. Schmidt, Die Ostgermanen, 2nd ed. (Munich, 1941), p. 121. 130 For Theodoric’s marriage alliances, Procopius, Wars V. 12. 21-2; Anonymus Valesianus II. 63, 68, 70; Jordanes, Getica 58 (297-9), ed. T. Mommsen, Monumenta Germaniae Histórica, Auctores Antiquissimi 5 (Berlin, 1882): Cassiodorus, Variae IV. 1. Among recent comment, S. Krautschick, “Die Familie der Könige in Spätantike und Frühmittelalter”, Das Reich und die Barbaren, ed. Chrysos and Schwarcz., p. 111. 131 Cassiodorus, Variae IV. 2; Moorhead, Theodoric in Italy, p. 193. 132 For full coverage of the marriage alliances of the period, including a genealogy, Krautschick, “Die Familie der Könige in Spätandke und Frühmittelalter”, pp. 110-20, 126-36. 133 e.g. Gregory of Tours, Decem Libri Historiarum IV. 27; Krautschick, “Die Familie der Könige in Spätandke und Frühmittelalter”, pp. 115-6. 134 Anonymus Valesianus II. 63; Jordanes, Getica 57 (295); Gregory of Tours, Decem Libri Historiarum III. 31. 135 Gregory of Tours, Decem Libri Historiarum II. 28. 136 Gregory of Tours, Decem Libri Historiarum III. 5.

248

IAN WOOD

othic king Amalaric.137 Theuderic’s son Theudebert I married the Lombard princess Wisigard,138 and his daughter, Theudechild, mar­ ried first Hermegisl, king of the Warni, and then Hermegisl’s son, Radigis.139 Theudebert’s son, Theudebald, married Walderada, Wisigard’s sister.140 Merovingian marriage into other royal families is more pronounced among the kings of Austrasia than among the Frankish kings of Neustria or Burgundy. Thus Sigibert married the Visigothic princess, Brunhild,141 his daughter Ingund the Visigothic prince Hermenegild,142 his grandson Theuderic II (596-613) was betrothed to a Visigothic princess, Ermenberga,143 and a great-granddaughter was betrothed to the Lombard king Adaloald.144 Almost as important were the marriages of Lombard kings. Waccho had three wives: the first, Ranicunda, was the daughter of Bisinus, the second, Austregusa, was the daughter of a king of the Gepids and the third, Silinga, was a daughter of a king of the Heniles.145 By Austre­ gusa, Waccho had two daughters, Wisegard, who married Theudebert I, and Wuldetrada, who married Theudebald, who passed her on to the Bavarian dux Garibald.146 Among later Lombard rulers Alboin married Chlodoswintha, daughter of the Frankish king Chlothar I,147 and when she died, he married Rosamund, daughter of Cunimund, king of the Gepids.148 Unfortunately Alboin had been responsible for the death of Cunimund, and (supposedly) for turning his skull into a goblet, from which he forced his wife to drink. Not surprisingly she ar­ ranged his assassination.149 Subsequently Authari asked to marry a sis­ ter of the Frankish king Childebert II: he was offered the hand of an-

137 Gregory of Tours, Decent Libri Historiarum III. 1, 10; Procopius, Wars V. 13. 4. 138 Gregory of Tours, Decern libri Historiarum III. 20, 27, 33; Paul, Historia Langobardontm I. 21. 139 Procopius, Wars VIII. 20. 11-41. 140 Gregory, Decern Libri Historiarum IV. 9; Paul, Historia Langobardorum I. 21. 141 Gregory, Decern Libri Historiarum IV. 27. 142 Gregory, Decern Libri Historiarum IV. 38; V. 38. 143 Fredegar IV. 30. 144 Paul, Historia Langobardorum IV. 30. 145 Paul, Historia Langobardorum I. 21; Origo Gentis Langobardorum 4; Historia Langobar­ dorum Codicis Gotham 4; Krautschick, “Die Familie der Könige in Spätantike und Frühmittelalter”, pp. 116-8. 146 Gregory of Tours, Decem Libri Historiarum IV. 9. 147 Gregory of Tours, Decem Libri Historiarum IV. 3, 41; Paul, Historia Langobardorum I. 27. 148 Paul, Historia Langobardorum I. 27. 149 Paul, Historia Langobardorum II. 28; compare Menander Protector, fr 12, 1

THE FRONTIERS OF WESTERN EUROPE

249

other Chlodoswintha,150 but ultimately he married the sister of Garibald of Bavaria,151 who had himself married Walderada, Waccho’s daughter and Theudebald’s cast off,152 and whose daughters married Evin of Trent,153 and the Lombard kings Authari and Agilulf.154 In some respects marital diplomacy was as important after the death of Theodoric the Great as it had been before. It simply lacked a single guiding hand. Between the late fifth and early seventh centuries a number of major diplomatic developments had, therefore, taken place in the lands to the east of the Rhine. Initially there was an attempt by Theodoric to inte­ grate the peoples east of the Rhine within a world of Ostrogothic di­ plomacy. This policy was a counterbalance to or, perhaps, was coun­ terbalanced by Frankish ambitions. In the end it failed to check those ambitions, and the Franks took over Alamannia and Thuringia, and cast their eyes even further to the south east. Theudebert could claim, with some justification that his authority extended from the borders of Pannonia to the North Sea. If anything this was an understatement, in that, by the end of his reign he also controlled large areas of territory south of the Alps, including Venetia. Theudebert’s ambitions came to nothing, partly because his succes­ sors were less successful than he was, and partly because of the chang­ ing political situation in the north-west Balkans. In particular the movement of the Lombards, first into Pannonia and then into Italy radically altered the political scene. This was a movement to which the Franks were remarkably sensitive from very early on: Merovingian contacts with the Lombards come, not in Italy, but before the migra­ tion of the latter from Pannonia, in the marriage of Chlothar Fs daughter Chlodoswintha to Alboin. Yet, while the movements of the Lombards provided a significant check to Merovingian ambitions, it was ultimately the Avars in Pannonia, and perhaps more important the Slavs, on the borders of Thuringia, in Bohemia and in Carinthia, who drew a political line, north-south, across central Europe.

150 Gregory of Tours, Decern Libri Historiarum IX. 16, 20, 25; Paul the Deacon, Historia Langobardorum III. 28. 151 Paul, Historia Langobardorum III. 30. 152 Paul, Historia Langobardorum I. 21. 153 Paul, Historia Langobardorum III. 10. 154 Paul, Historia Langobardorum III. 30, 35.

250

IAN WOOD

It is worth asking whether this political history was in any way reflected in patterns of distribution and exchange.155 International diplomacy automatically carried with it implications for the distribution of goods. The successor states of the West Roman Empire could not compete with the vast diplomatic expenditure of the Roman Empire, even in its Byzantine guise. Nevertheless Theodoric did his best to emulate the imperial model.156 Merely to adopt a prince as a son in arms involved the significant transfer of weapons.157 Mar­ riages will have occasioned the presentation of even more considerable gifts. There is a strong likelihood that the development of Thuringian metalwork took place in the context of Amalaberga’s marriage to Hermenefred.158 Crucial developments in Thuringian metalwork seem to have de­ pended on external influences.159 The qualitative changes in the jew­ ellery produced in the late fifth and sixth centuries seem to have been directly related to contacts with the Ostrogoths. Nor were these con­ tacts between the Ostrogoths and the peoples of the basin of the Saale one-sided. Theodoric was the recipient of furs, blond-haired boys and damascined swords from the Warni, close neighbours of the Thuringians.160 Developments in metalwork can, therefore, be seen as a reflection of the diplomatic world of Theodoric, and some of the products belong specifically to the culture of the Thuringian court. By contrast the Frankish take-over is marked, at a lower social level, by the introduc­ tion of Frankish metalwork into Thuringia.161 Further, the settlement of Frankish warriors in the region may be indicated by grave finds in 155 For a general archaeological survey of the cultural impact of imperial diplomacy see H. Vierck, “Imitato imperii und interpretatio Germanica vor der Wikingerzeit”, Les Pays du Nord et Byzance, ed. R. Zeitler, (Uppsala 1981), pp. 64-113. 156 c.f. Cassiodorus, Variae I. 46: II 40 157 Cassiodorus, Variae IV. 2; Moorhead, Theodoric in Italy, p. 193. On the question of such adoptions, Fritze, Untersuchungen zurfrühslawischen undfrühfränkischen Geschichte bis ins 7. Jahrhundert, pp. 204-7: D. Claude, “Zur Begründung familiärer Beziehungen zwischen dem Kaiser und barbarischen Herrschen”, Das Reich und die Barbaren, eds. Chrysos and Schwarcz, pp. 31-9. 158 D. Claude, “Universale und partikulare Züge in der Politik Theoderichs”, Francia 6 (1978), p. 34, n. 108: B. Schmidt, “Theoderich der Grosse und die damaszierten Schw­ erter der Thüringer”, Ausgrabungen und Funde 14 (1969), p. 38. 159 Schmidt, “Die Thüringer”, pp. 535-6, 544. 160 Cassiodorus, Variae V. 1; Schmidt, “Theoderich der Grosse und die damaszierten Schwerter der Thüringer”, pp. 38-40. Schmidt is, however, a little too keen to amalgam­ ate the Thuringians and the Wami, who were a separate, if related, people. 161 Schmidt, “Die Thüringer”, p. 547.

THE FRONTIERS OF WESTERN EUROPE

251

places of particular strategic importance.162 Thuringian archaeology, thus, not only reflects the integration of the region within the world of Ostrogothic diplomacy: it also reflects the quite different integration within the area of Merovingian hegemony. Nor is it only archaeology that suggests changing distribution pat­ terns. The annual payment of 500 cows as tribute from the Saxons to the Franks, is in itself evidence of a very significant exercise in distri­ bution.163 Almost inevitably it is not something that is clearly reflected in the archaeological evidence. Although the region east of the Rhine was no stranger to external influence, it is possible to see in the written and archaeological sources for the late fifth and sixth centuries some­ thing more than a simple redrawing of political boundaries following the fall of the Western Empire. Patterns of distribution and exchange were affected by the eastwards expansion of the Franks, and subse­ quently by the weakening of Frankish power and the arrival of Slavs. The border of Frankish Europe came to rest in Thuringia, an area which had previously been a part of Germania Libera. It was not merely Ostrogothic diplomacy and Frankish aggression which determined the internal geography of Europe. The Slavs were another, equally important, factor. Not that one should make too much of the cultural and linguistic factors which distinguished the Slavs from their neighbours in noting their impact. The Frankish his­ torian Fredegar, certainly, was unmoved by such ‘ethnic’ distinc­ tions.164 In terms of the history of the distribution of goods, the evi­ dence here points in slighdy contradictory ways. Samo, the architect of Slav resistence to the Franks, was a Frank himself, from Sens, in­ volved in trade with the Slavs.165 In addition, the fact that the outbreak of hostilities between Samo’s kingdom and the Franks was caused by Slav mistreatment of Frankish merchants,166 suggests that up into the 630s, at least, trade between Franks and Slavs was not uncommon. Further, the trade route from Thuringia to Mainz was certainly known

162 Schmidt, “Die Thüringer”, p. 547. 163 Gregory, Decern libri Historiancm IV. 10, 14; Fredegar IV. 74. 164 Fritze, Untersuchungen zurJrühslawischen undfiühfränkischen Geschichte bis ins 7. Jahrhun­ dert, pp. 278-85. Fritze’s comments must, of course, be understood in the context of a reaction to Nazi ethnic ideology: they nevertheless represent a very sensitive reading of Fredegar’s text. 165 Fredegar IV. 48. 166 Fredegar IV 68: Fritze, Untersuchungen zurfrühslawischen undJrühfränkischen Geschichte bis ins 7. Jahrhundert, pp. 87-9; Pohl, Die Awaren, pp. 256-9.

252

IAN WOOD

by Slavs in the mid eighth century.167 It may be that Frankish weapons were traded.168 Equally, it is likely that there was a significant slave trade, but slaves can rarely be identified as such in the archaeological record. It is enough, however, to remember that the word ‘slave’ is de­ rived from the name Slav.169 It has even been assumed, very plausibly, that Samo himself was a slave-trader.170 It has, nevertheless, been argued that there are indications of a breakdown in the distribution of certain goods, caused by Slav expan­ sion. The argument that Merovingian trade with the Ostpreußen may have been broken by the Slavs171 is, however, disproved by recent finds at Hacki in Poland,172 and there is no reason to lay the disruption of the circulation of Byzantine gold coins into east, central and northern Europe, at their door.173 Yet, while Slav aggression may not have been responsible for any rupture of links between western and central Eu­ rope, the comparative material poverty of early Slav sites does suggest that contacts did decline - and it may even allow one to speculate that the decline in contact between Francia and Italy on the one hand and north-central Europe on the other was linked to a relatively egalitarian

167 Eigil, Vita Sturmi 7, cd. P. Engelbert, Die Vita Sturmi des Eigil von Fulda (Marburg, 1968). 168 Fritze, Untersuchungen zurfrühslawischen undfrühfränkischen Geschichte bis ins 7. Jahrhun-* dert, p. 87. 169 Pohl, Die Awaren, p. 196. 170 Fredegar, IV 48; Brown, The Rise of Western Christendom, pp. 259-60. 171 Fritze, Untersuchungen zurfrühslawischen undfiühfränkischen Geschichte bis ins 7. Jahrhun­ dert, p. 16, with p. 305, n. 121. 172 Z. Kobylinski and Z. Hensel, “Imports or local products? Trace element analysis of copper-alloy artefacts from Hacki, Bialystock Province, Poland”, Archaeologia Poiana 31 (1993), pp. 129-40. I am grateful to Professor Kobylinski for drawing my attention to this material. 173 Fritze, Untersuchungen zurfrühslawischen undfrühfränkischen Geschichte bis ins 7. Jahrhun­ dert, p. 306, n. 124; O. Kyhlberg, “Late Roman and Byzatine solidi”, Excavations at Helgö X, p. 64 raises the possibility that the expansion of the’Slavs cut oiF the supply of Byz­ antine gold to the north: p. 65 argues that the very last Nordic hoards (which continue to c. 568-88, p. 57) originate from a western source. The comments on a Burgundian Empire (p. 64) and a Lombard invasion of northern Italy in 534 (p. 65) unfortunately have no evidential basis. See, however, Arrhenius, “Connections between Scandinavia and the East Roman Empire in the Migration Period”, p. 135, which downplays the possibility of a disruption of connections between Byzantium and Scandinavia before the Viking Age, and Fagerlie, “Late Roman and Byzantine Solidi found in Sweden and Den­ mark”, p. 174: ‘There is no apparent justification for the recurrent belief that invasions of central Europe by Slavs and Avars in the mid-sixth century severed the contact be­ tween the Baltic and the South and brought about the end of the solidi import. The solidi import ended because its source had run dry and internal disturbances in Scandinavia made it impossible to find and exploit alternative sources.’ See also Klavs Randsborg’s chapter in this volume.

THE FRONTIERS OF WESTERN EUROPE

253

social structure among the Slavs.174 To the south, the evacuation of Pannonia and Carinthia and the settlement of Slavs in the eastern Alps marks the beginnings of a truly Dark Age in that region, despite Lom­ bard ability to take tribute from the Slavs of gettici.175 Yet further south, Istria was devastated by the Avars and Slavs in the opening years of the seventh century,176 and it was still thought of as an area of devas­ tation in the eighth.177 By contrast with many of the Slav regions, the centre of the Avar kingdom could boast considerable wealth, but, de­ spite incursions into the West and the occasional taking of tribute, the economic links of the Avars seem to have been chiefly to the East.178 At the very least the development of patterns of distribution between the Germanic successor states and central and eastern Europe is likely to have been altered by the westward expansion of the Slavonic and Avar peoples. The political and economic developments which took place east of the Rhine cannot compete in importance with the vast shifts between the Mediterranean and the North Sea which took place in the early Middle Ages. They do, however, have their importance in the history of Europe and the formation of its internili boundaries. They also have an importance because they show how complex those shifts in politics and in economic distribution could be. Ostrogothic diplomacy affected the region in ways different from Frankish conquest, politically as well as economically. Merovingian aggression destroyed Theodoric the Great’s policies: Slav expansion put a check to Merovingian aggres­ sion, all in little over a century, in a region most of which had never been part of the Roman Empire.179

174 For this see the forthcoming contribution of P. Urbanczyk to the work of the ESF team headed by Franz Theuws. 175 Paul the Deacon, Historia Langobardorum IV. 38; Pohl, Die Awaren, p. 259. 176 Paul the Deacon, Historia Langobardorum IV. 24; Pohl, Die Awaren, pp. 159-60. 177 Die Kosmographie des Aethicus, ed. O. Prinz, Monumenta Germaniae Histórica, Quellen zur Geistesgeschichte des Mittelalters 14 (Munich, 1993), p. 233. 178 This, at least, is the picture presented by Pohl, Die Awaren, pp. 195-9 179 This paper was drafted while I was fortunate enough to be a Fellow-in-Residence at the Netherlands Institute for Advanced Study. I am particularly grateful to Professor Zbigniew Kobylinski and to Dr. Ulf Näsman for supplying me with references to recent literature relating to the archaeology of north Central Europe.

THE JUSTINIANIC ERA OF SOUTH SCANDINAVIA: AN ARCHAEOLOGICAL VIEW Ulf Näsman

Introduction The Scandinavian peoples lived, during the Roman period, at the pe­ riphery of the world which was elucidated by the geographical and po­ litical knowledge of Mediterranean civilisations. Therefore, they often appear to many historians as very distant actors in the historical play. However, the self-esteem of the inhabitants of Scandinavia caused them to place themselves, with pride and self-confidence, at the centre of the world. This is revealed by both their material culture and their mythology. Only a sparse trickle of information about the peoples in the North reached the Roman world. Few of these glimpses of knowledge were actually recorded, and unfortunately little has survived to the present day. Therefore the possibilities for the historian to properly understand in what way, if at all, these regions were part of the Roman world, are extremely limited. It is similarly difficult for historians to study the transformations of barbarian Scandinavia that led to the eventual ap­ pearance of three main polities, the Danes, the Norwegians, and the Swedes. According to the written sources this did not happen until the Viking Age. Subsequently they became three Christian medieval king­ doms: Denmark, Norway and Sweden. The written sources about early Scandinavia are few, and without underestimating the great value of the information found in, for exam­ ple, Jordanes, Procopius, and Gregory of Tours, I must emphasise that the reconstruction of Scandinavian societies and their transformations during the first millennium A.D. has to be based on a foundation of material culture, and consequently, it is only archaeology that can give Scandinavian cultures their proper place on the scene of the trans­ formed Roman Europe.

256

ULF NÄSMAN

Scandinavia - a Constituent of Europe One may ask, whether Migration period Scandinavia is of any in­ terest to the European scholarly audience of today. Is it really relevant to study Scandinavian developments within the framework of the pro­ gramme ‘The Transformation of the Roman World’? Clearly, it is! Someone once said that the development in peripheral Scandinavia is a sensitive thermometer revealing what is going on in the European body. This, however, is not a good metaphor, since Scandinavia, un­ like a thermometer, is a constituent part of the same body. Therefore, I favour a comparison with the weather forecasts on television: to un­ derstand the changes in the climate in southern and western Europe you must know also the routes that cyclones take over the Norwegian Sea, as well as the location of depressions in Russia. Weather charts have to cover the whole of Europe and North Africa, as must archae­ ological and historical maps when studying long-term trajectories and wide themes such as ‘The Transformation of the Roman World’. There are obvious reasons to study Scandinavia in a European set­ ting. Firstly, the texts of classical authors demonstrate that in some pe­ riods an interest in Scandinavia existed, in the Roman empire, and in the Byzantine empire, as well as in some of the Germanic successor kingdoms. Without an archaeological study of the material culture of Scandinavia we shall never fully comprehend the causes of this fluctu­ ating interest. Secondly, the geographical as well as the economic position of the various Scandinavian regions renders them, I believe, very sensitive to changes in the European economic exchange systems. Studies of Scan­ dinavian external contacts will thus be important contributions to the understanding of the exchange system between the barbarians and the Roman Empire and its provinces, as well as throwing light upon the network that, in the early Middle Ages, connected the Germanic Suc­ cessor Kingdoms all over Europe with one another, as well as with Byzantium. One important aspect of the problems of production, dis­ tribution, and demand, is the prelude of urbanism outside the former Roman Empire, and here Scandinavian archaeology has much to offer those looking for a post-Pirenne paradigm. Thirdly, the incorporation of Scandinavia into the Catholic Chris­ tian area at the end of the first millennium A.D. was an important con­ stituent of the transformation of the Roman world into medieval West­ ern Europe. Furthermore, in South Scandinavia at any rate, the pro­

THE JUSTINIANIC ERA OF SOUTH SCANDINAVIA

257

cess of inclusion within the Mediterranean sphere was already well ad­ vanced in the era of Justinian. Fourthly, Scandinavian archaeology has a very rich and varied se­ lection of Migration period material to study, and research in this pe­ riod is currently progressing rapidly. Scholars studying other Germanic polities will find excellent material for comparative studies in the Scan­ dinavian material culture. In summary, stated modestly, the late “pre­ history” of South Scandinavia is not by any means the least interesting part of European archaeology.

Three Theses In order to substantiate the three theses I will make in my paper, I shall look at the development in South Scandinavia from the third to the seventh century. Archaeology cannot work within a time span as limited as the forty years of the Emperor Justinian’s rule, A.D. 527565. To be able to examine the second third of the sixth century at all, archaeologists have to start not later than the late Roman Iron Age (c. AD. 150/160-375/400), cover the entire Migration period (c. 375/ 400-530/560, known in Denmark as the early Germanic Iron Age) and cannot end before the seventh century, well into the Merovingian pe­ riod (from c. 530/560 to the late eighth century, known as the late Germanic Iron Age in Denmark). This is obviously far removed from the Justinianic era, but in archaeology a long-term perspective is nec­ essary. The three theses are: i) The Mediterranean interest in Scandinavian affairs during the Mi­ gration period reflects Scandinavian participation in continental events. ii) A fundamental break of long-distance exchange routes in the sixth century directed South Scandinavia towards the western parts of Eu­ rope. iii) Scandinavia was an integrated part of a pan-Germanic network of social and economic communication, which in the sixth century con­ tributed to the formation of a Danish kingdom, not very different from other Germanic kingdoms.

258

ULF NÄSMAN

The Mediterranean interest in Scandinavia Traditional explanations of the Mediterranean interest in Scandina­ vian afTairs all have trade as the main factor, and Jordanes’ statement about the fur trade of the Svear is often quoted. O f course the supply of luxuries such as furs and amber was an obvious reason to record geographical knowledge in itineraries which described some routes to the North Sea and the Baltic as well as the routes within Scandinavia. For example, the so-called peoples’ catalogue of Jordanes may be the preserved remains of a couple of lost itineraries.1 But trade in luxuries is not sufficient to explain the periodic occur­ rence of written sources about Scandinavia. Parallel with military re­ connaissance, trade may have contributed to the first interest of the ex­ panding Roman Empire in Scandinavian matters (such as that of Pliny, Tacitus, and Ptolemy), but a lack of exchange cannot explain the si­ lence during the third and fourth centuries. Archaeology can, on the contrary, demonstrate frequent contacts with the Roman provinces in the west, as well as the east, during theses historically dark centuries.2 It is not until the period of the Great Migrations that light is again shed on Scandinavian matters. In the Migration and early Merovingian pe­ riods, merchants, soldiers, and immigrants are the plausible sources of accounts by Byzantine, Gothic, Frankish, and Anglo-Saxon authors. The establishment of a written record of myths of Scandinavian origin, in some Germanic successor states, explains some of the sources, but the participation of Scandinavian troops in contemporary wars, and raids and migrations by Scandinavians also contributed to the renewed interest. Gregory of Tours gives us the most explicit example of a Scan­ dinavian raid into former Roman territory, the raid of the Danish king Hygelac at the lower Rhine in the early sixth century, and we must not forget the raiding Jutes nor the migrations of Heruli.3 Procopius’ 1 Jordanes, Gttica, trans. C. C. Mierow The Gothic history ofJordanes (Princeton, 1915). —Cf. R. Hackmann, Die Goten und Skandinavien (Berlin, 1970); J. Svennung, ‘Jordanes und die gotische Stammsage”, Studia Gotica, ed. U.E. Hagberg (Stockholm, 1972), pp. 20-56. 2 U.L. Hansen, Römischer Import im Norden (Copenhagen, 1987); U.L. Hansen, Himlingoje - Seeland - Europa (Copenhagen, 1995). 3 On Danish raids on the Merovingian realm, see I. Wood, The Merovingian North Sea (Alingsás, 1983) 6fF. On the Jutes, see for example, S. Chadwick Hawkes, “Anglo-Saxon Kent c. 427-725”, Archaeology in Kent to AD 1500, ed. P.E. Leach (London, 1982), pp. 6478; B. Yorke, Kings and kingdoms ofearly Anglo-Saxon England (London, 1990). On the Heruli, see L. Schmidt, Geschickte der deutschen Stämme bis zum Ausgang der Völkerwanderungszeit. Die Ostgermanen (München, 1934) 548ÍF; Die Germanen II cd. B. Kruger (Berlin, 1983); H. Wol­ fram, History of the Goths (Los Angeles, 1988). Hypercritical is A. Ellegârd, “Who where the Ertili?” Scandia 53/1 (1987) pp. 5-34.

THE JUSTINIANIC ERA OF SOUTH SCANDINAVIA

259

story about the Heruli, who travelled to and from Thule, demonstrates Scandinavian participation in a major Migration period event in the reversed perspective as receivers of immigrants.4 Furthermore, it is in this period that, for the first time, we receive some information about the Danes in the works of Jordanes, Procopius, and Gregory of Tours. The imported prestige goods, which still characterised the display of wealth in Scandinavian societies, provide archaeological evidence of continued continental contacts.5 However, in the sixth century, Scandinavia was still a remote pe­ riphery of the Merovingian sphere of interest, and by the seventh cen­ tury the written sources are silent again. Archaeology shows, however, that the traditional links with central and western Europe were unbro­ ken.6 Significantly, the period is labelled ‘the Merovingian Iron Age’ by Finnish and Norwegian archaeologists. The source situation changed again in the eighth century, as a result of the Carolingian ex­ pansion into northern Europe. Ongendus is the second Danish rex to be mentioned in a written source.7 A rather detailed understanding of Scandinavian geography is revealed, following both new and close trade relations between Scandinavia and western Europe, but prima­ rily as an accompaniment to the growing political and military impor­ tance of the Danish kingdom.8 From the late ninth century the first certain Scandinavian itinerary is preserved, the account of Otter’s jour­ ney from north Norway to Hedeby in South Jylland.9 However, the final literary elucidation of the Viking North came with Adam of Bre­ men in the eleventh century,10 symbolising as it were the final integra­ tion of Denmark into Catholic western Europe.

4 Procopius, De bello Gothico II. 15, cd. H.B. Dewing, (Cambridge, Mass, 1914-28). 5 U. Näsman, Glas och Handel i senromersk tid ochfolkvandringstid (Uppsala, 1984); L. Hedeager, Iron-Age Societies (Oxford, 1992). 6 H. Vierck, Hayo, “Imitado imperii und Interpretado Germanica vor der Wikingerzeit”, Les Pays du Nord et Byzance, ed. R. Zeitler (Uppsala, 1981), pp. 64-113; B. Arrhenius, Merovingian garnetjewellery (Stockholm, 1985); U. Näsman, “Vendei period glass from Eketorp-II, Öland, Sweden”, Acta Archaeologica 55 (1984 [1986]) pp. 55-116; H. Steuer, “Helm und Ringschwert”, Studien zur Sachsenforschung 6 (1987), pp. 189-236. 7 Alcuin, Vita Willibrordi, trans. C.H. Talbot, The Anglo-Saxon missionaries in Germany (London, 1954). 8 P. Sawyer, Kings and Vikings (London, 1982); K. Randsborg, The Viking Age in Den­ mark. Theformation of a state (London, 1980); E. Roesdahl, The Vikings (London, 1991). 9 N. Lund, Two voyagers at the Court of King AlfredThe ventures of Ohthere and Wulfstan (York, 1984). 10 Adam of Bremen, History of the Archbishops of Hamburg-Bremen, trans. FJ. Tschan (New York, 1975).

260

ULF NÄSMAN

In the Migration period, Germanic peoples pretending to originate from the North were important players on the Mediterranean stage. Therefore, for this reason, it seems natural that we find an interest in the peoples of Scandinavia. However, it was not only in this indirect fashion, as putative ancestors of important gentes, that Scandinavians were interesting. Archaeology indicates that Scandinavia received its share of the Mi­ gration period turmoil. The circumstantial archaeological evidence is rich and certainly of historical significance, despite problems of inter­ pretation. Troops of Scandinavian origin presumably took part in the warfare between barbarians and Romans. The evidence consists of Scandinavian solidus hoards of the fifth and sixth centuries which are often interpreted as the remains of soldiers’ pay, earned primarily dur­ ing the wars in the Danubian basin and brought home by Scandina­ vian warrior bands.1112In an important paper on the solidus treasures, the late Joachim Werner related their existence to the building of the ring-forts of Öland and their similarity to some late Roman fortifica­ tions. Today this may be doubted, but the excavations at the ring-fort at Eketorp revealed the remains of a gate which probably had been equipped with a port-cullis, clearly based on a late Roman prototype.10 The archaeological evidence of widespread warfare in the late Romanearly Migration period consists of hundreds of hill-forts in mountainous areas and ring-forts and dykes in the plains.13 In south Scandinavian waters, the invasion defence also included the protection of Qords and bays by ship-blockages.14 The well-known finds of sacrificed spoils-ofwar, provide ample evidence of the many battles fought in south Scan­ 11 J.M. Fagerlie, Late Roman and Byzantine solidifound in Sweden and Denmark (New York, 1957); K. Godlowski, “Zur Frage der völkerwanderungszeitlichen Besiedlung in Pom­ mern”, Studien zur Sachserforschung 2 (1980), pp. 63-106; F. Herschend, “Tvä studier i öländska guldfynd”, Tor 18 (1978-1979) pp. 33-294; O. Kyhlberg, “Late Roman and Byzantine solidi”, Excavations at Helgö X. Coins, iron and gold, ed. A. Lundström and H. Clarke (Stockholm, 1986) pp. 13-126 (unfortunately, his historical analysis in section 4.5 is full of mistakes); cf. Klavs Randsborg in this volume. 12 J- Werner, “Zu den auf Oland und Gotland gefundenen byzantinschen Goldmün­ zen”, Fomvärmen 44 (1949) pp. 257-86; and see critical discussion in U. Näsman, “The gates of Eketorp-II. To the question of Roman prototypes of the Oland ringforts”, The Birth of Europe: Archaeology and social development in thefirst millennium AD. ed. K. Randsborg (Rome, 1989), pp. 129-39. 13 See §§28-31 in the entry Burg in Reallexikon der Germanischen Altertumskunde 4/1-2 (Berlin/New York, 1978); J. Engström, Torsburgen, (Uppsala, 1984); summary by M. Olausson, Det inneslutna rummet (Stockholm, 1995). 14 F. Rieck, “Aspects of coastal defence in Denmark”, Aspects of maritime Scandinavia AD 200-1200, ed. O. Crumlin-Pedersen (Roskilde, 1991), pp. 83-96.

THE JUSTINIANIC ERA OF SOUTH SCANDINAVIA

261

dinavia between the late second and the end of the fifth century A.D.15 The dates of the different source material demonstrates an increasing frequency of warfare from the second century A.D. The evidence of warfare culminates in the fifth century and declines rapidly in the sixth century (Fig. 1).

Fig. 1. A rough sketch of the frequency o f ‘war indicators’ in South Scandinavia during the first millenium A.D: fortifications, ship blockages, booty sacrifices (shaded area), and written sources (weapon graves are excluded because they reflect social rather than military circumstances). After U. Näsman, “The Iron Age Graves of Öland - Representative of What?”, Prehistoric Graves as a Source ofInformation, ed. B. Sqemquist (Stockholm, 1994), pp. 15-30.

The letters preserved in Cassiodorus’ Variae show that Theoderic the Great maintained at least sporadic diplomatic contact with powers far into the north,16 and a man called Roduulf who spent his last years at the Gothic court in Italy was once king of the Ranii of West Scandi­ navia.17 Archaeology also, can show that Scandinavia was an integrat­ ed part of a social network, linking Germanic kingdoms to one another

15 J. Dkjær, IUerup Âdal 1-2. Die Lanzen und Sperre (Ârhus/Hojbjerg, 1991); J. Ilkjær, IlierupÂdal 3-4. Die Gürtel (Ârhus/Hojbjerg, 1993); C. Fabech, “Booty sacrifices in South­ ern Scandinavia: a reassessment”, Sacred and profaned (Oxford, 1991), pp. 88-94. 16 Cassiodorus, Variae, trans. SJ.B. Bamish (Liverpool, 1992); cf. Ian Wood’s contri­ bution in this volume. 17 Jordanes, Getica, ch. 24. Cf. Wolfram, History of the Goths, p. 326.

262

ULF NÄSMAN

and having connections to the Roman and Byzantine world as well, and in fact the distribution maps of significant types of objects can give a much more detailed picture of the relations between Scandinavian and continental powers. In the Empire, gold medallions were used as diplomatic gifts by the emperors. Some were even given to Germanic allies, and here they were soon imitated.18 Both Roman and barbaric medallions came as far as Scandinavia. The Scandinavian examples probably came indirecdy as gifts from other Germanic rulers, who had rapidly adopted many elements of Roman imperial ceremonial. Obviously, the pictorial content of the Roman medallions expressed imperial propaganda, so the content of the barbaric imitations rapidly changed to give an inter­ prétatif) germanica of the symbolism. In this transformation, certainly more than the pictorial expression of imperial ideology spread to the barbaric leaders. In south Scandinavia, domestic goldsmiths soon de­ veloped the characteristic Nordic gold bracteates, the pictorial content of which quickly detached itself from the Roman prototypes and de­ veloped into a unique conveyor of Scandinavian religious and ideolog­ ical propaganda, clearly an interpretatio nordica.l9 The wide distribution of Nordic bracteates in Europe reveals certainly that the Scandinavian culture in the Migration period influenced large parts of Germanic Eu­ rope. The finds demonstrate in which parts of the Germanic social and political network that Scandinavians were most active; the European distribution of Nordic relief-brooches, women’s high-status jewellery, shows a similar pattem.20 In a similar way, high-status imports from the continent reveal that the life-style found in other Germanic societ­ ies made a strong impact on the leading echelons of Scandinavian pol­ ities. In particular, helmets and sword fittings which would hardly have

18 The largest find of medallions, some of which are barbaric copies, was found at Simleul-Silvaniei/Szilagy-Somlyo: R. Noll, Vom Altertum zum Mittelalter (Vienna, 1974) 73fT.; cf. W. Pohl, “Die Gepiden und die Gentes an der mitderen Donau nach dem Zer­ fall des Attilareiches”, Die Völker an der mittleren und unteren Donau imfünften und sechstenJahr­ hundert, ed. H. Wolfram & F. Daim (Vienna, 1980), pp. 239-305. 19 A. Andren, “Guld och makt - en tolkning av de skandinaviska guldbrakteatemas funktion”, Samfundsorganisation og regional variation. Norden i romerskjemalder ogfolkevandringstid, ed. Ch. Fabech &J. Ringtved (Hojbjerg/Arhus, 1991) pp. 245-56; M. Axboe, “Guld og guder i folkevandringsdden”, Samfundsorganisation og regional variation. Norden i romersk jernalder ogfolkevandringstid, ed. Ch. Fabech &J. Ringtved, pp. 187-202; M. Axboe, “Gudmc and the gold bracteates”, The archaeology of Gudme and Lundeborg, ed. P.O. Nicken et al. (Copenhagen , 1994) pp. 68-77. 20 For distribution maps of bracteates and relief-brooches, see maps 10-11 with ref­ erences in Näsman, Glas och handel i senromersk tid ochfolkvandringstid.

THE JUSTINIANIC ERA OF SOUTH SCANDINAVIA

263

spread as normal merchandise can be used to demonstrate the direct links between some prominent individuals and princely seats in conti­ nental Europe.21 Indeed, one of the important changes which followed in the wake of the fall of the Western Roman Empire was the new cultural and political network that developed between regions on both sides of the fallen limes. This network included Scandinavia, (or to be more precise, those regions in which farming societies dominated, i.e. Germanic Scandinavia and the south-western parts of Finno-Ugrian Finland). The importance of the contacts to other Germanic courts have, in my opinion, not been sufficiendy considered in the discussion of internal social development in Scandinavia. In south Scandinavia in particular, where archaeology can demonstrate the presence of creative and active societies, the effect of close contacts to Frisians, Saxons, Vami, AngloSaxons, Franks, Alamanni, Thuringians, Heruli, Lombards, Gepids, and Goths must have been considerable. The foundations on which Viking Age Denmark rested were laid out in the fifth to seventh cen­ turies, and many of the building bricks were certainly of continental Germanic or late-Roman origin.

The Long-Distance Exchange Routes to Scandinavia In the Migration period, the import of prestige goods to Scandinavia continued the pattern of the late-Roman Iron Age. Glass vessels are typical luxuries and as such they represent ancient distribution patterns in Europe. They are probably the most representative commodity, and the provenance of glass vessels can often be established quite accurate­ ly. Found as vessels in graves or as shards in settlements, both qualita­ tive and quantitative information can be deduced on the basis of glass finds. Therefore, a study of changes in the frequency and distribution of glass vessels in a long-term perspective will elucidate important as­ pects of the European development of production, distribution and de­ mand. For example, the numerous glass vessels decorated with cut ovals or facets that date to the fourth-fifth centuries represent an im­ port via northern Poland from the Black Sea region as demonstrated 21 C. Fabech, “Reading society from the cultural landscape. South Scandinavia be­ tween sacral and political power”, The archaeology of Gudme and Lundeborg, ed. P.O. Nielsen et aL, pp. 169-83; with references to Arrhenius, Merovingian garnet jewellery, and Steuer, “Helm und Ringschwert”.

264

ULF NÄSMAN

by the distribution map (Fig. 2).22 Most of the glass vessels, which are found in Scandinavian, come from grave finds, but recendy the num­ ber of new vessels found as shards on setdement sites has increased considerably. Glass vessels were also continuously imported from the western provinces.23 During the late-Roman period, various types of vessels were imported from the West, but during the Scandinavian Mi­ gration period, the assortment is reduced and the material is dominat­ ed by tall beakers with a small foot and looped trailing on the body.24 The workshops of the production of these so-called Snartemo beakers are unfortunately unidentified, but scholars agree that they should be sought west of the Elbe and north of the Alps. It should also be noted that a vessel type which is very widespread in the northern Merovin­ gian regions and in Anglo-Saxon England, the so-called Kempston cone beaker, is very rare in Scandinavia.25 The distribution map clear­ ly shows that the two glass types probably have different origins (Fig. 3). A change in funerary ritual explains the relative lack of finds in south Scandinavia; but since all new finds come from settlement exca­ vations, there is no reason to believe that the wealthy in Denmark had fewer glass vessels in the Migration period than during the late-Roman period. This opinion is further compounded by the many hoards con­ taining imported gold and silver that have been found in South Scan­ dinavia.26 It can be concluded that in the fourth and fifth century, a considerable import of prestige goods came to Scandinavia from both the west and the east, i.e. from provinces both at the Rhine and at the Black Sea littoral. However, the connections to Gaul seem to be only 22 Näsman, Glas och handel i senromersk tid ochfolkvandñngstid, E. Straume, Gläser mit Facettenschliffaus skandiruwishen Gräbern des 4. und 5. Jahrhunderts n.Chr (Oslo, 1987). 23 Hansen, Römischer Import im Norden. 24 Näsman, Glas och handel i senromersk tid ochfolkvandñngstid, 66ff; B> Stjernquist, “Glass from the settlement of Gärdlösa, southern Sweden”, Meddelandenfron Lunds universitets Historiska museum. NS 6 (1985-1986), pp. 139-66; E. Straume, “Die Gläser mit Fadenauflage aus der Völkerwanderungszeit in Norwegen”, Kölner JahrbuchJur Vor- und Frühgeschichte 22 (1989), pp. 187-92. 25 U. Koch, Der Runde Berg bei Urach 6. Die Glas- und Edelsteirifunde aus den Plangrabungen 1967-1983. 1-2 (Heidelberg/Sigmaringen, 1987); V.I. Evison, “Some Vendei, Viking and Saxon glass”, Trade and exchange in prehistory. Studies in honour of Berta Stjernquist, ed. B. Hàrdh et al. (Lund, 1988), pp. 237-45. 26 Silver: E. Munksgaard, “Spätantikes Silber”, Frühmittelalterliche Studien 21 (1987), pp. 82-4 & pl. 16; C. Fabech, “Sjörup - an old problem in a new light”, Meddelandenfrän Lunds Universitets Historiska Museum. New series 8 (1989-1990), pp. 101-19. Gold: J. Hines, “Ritual hoarding in Migration-period Scandinavia”, Proceedings of the Prehistoric Society 55 (1989), pp. 193-205; L. Hedeager, “Gulddepoteme fra ældre germanertid”, Samfiindsorganisation og regional variation. Norden i romerskjemalder ogfolkevandñngstid, ed. C. Fabech and J. Ringtved (Hojbjerg/Ârhus, 1991), pp. 203-12 (summary).

THE JUSTINIANIC ERA OF SOUTH SCANDINAVIA

265

Fig. 2. The distribution of cut glass vessels of probable eastern origin during the Late Roman and Migration periods (mainly from the fourth-fifth centuries A.D.): cylindrical beakers with cut ovals, conical beakers with cut ovals or facets, and cut beakers with a foot. For references, see H. Günter Rau, “Körpergräber mit Glasbeigaben des 4. nachchristlichen Jahrhunderts im Oder-Weichsel-Raum”, Acta Praekistorica et Archaeologica 3 (1972), pp. 109-214; E. Sträume, Gläser mit Facettenschliffaus skandinavishen+Gräbem des 4. und 5. Jahrhunderts n. Chr., (Oslo, 1987); Nasman, Glas och handel I senromersk tid ochfolkvandring Idem, “Om fjärrhandel i Sydskandinaviens yngre jämälder”, Hikuin 16 (1990), pp. 89-118, fig. 3, summary pp.l55f.

266

ULF NÄSMAN

Fig. 3. The distribution of beakers with looped trailing on the body. Dot: thick beakers with a foot, mainly of the so-called Snartemo type, and a few of the so-called Kapusany type. Triangle: conical beakers without a foot, mainly of the so-called Kempston type. For references, see fig. 4 in Näsman, “Om fjärrhandel i Sydskandinaviens yngre jämälder”; B. Stemquist, “Glass from the settlement of Gärdlösa, southern Sweden”, M e d d e la n d e n f i a n L u n d s u n iv e rsite ts H is to r ik a m u sseu m , N S 6, (1985-1986), pp. 139-166; U. Koch, D e r R u n d e B e rg b e i U rach 6 . D ie G la s - u n d E d e lste in fu n d e a u s den P la n g ra b u n g en 1 9 6 7 1 9 8 3 . 1 - 2 , (Heidelburg/Sigmaringen, 1987); E. Straume, “ Die Gläser mit Fadenauflage aus der Völkerwanderungszeit in Norwegen”, K ö ln er J a h rb u c h J u r V or - u n d F rü h gesch ich te 22 (1989), pp. 187-192.

THE JUSTINIANIC ERA OF SOUTH SCANDINAVIA

267

Fig. 4. The distribution of claw beakers. Solid dots are certain claw beakers, open dots are uncertain finds. For references, see V. Evison, “Anglo-Saxon claw beakers”, Archaeologia 17 (1982), pp. 43-76; Koch, Der Runde Berg bei Urach 6. Die Glas - und Edelsteinjunde aus den Plangrabungen 1967-1983. 1-2; Näsman, “Om fjärrhandel i Sydskandinaviens yngre jäm älder”.

268

ULF NÄSMAN

indirect, and there are still very few indications of contacts with Great Britain. The import of cut glass from the south-east ebbed away at the end of the fifth century. A discontinuation of the traditional links between the Baltic and Black Sea littorals as well as a break of the routes to the Danubian basin is also reflected in other material sources, and can probably be explained by changed political and ethnic circumstances in the Ukrainian steppe, in the forest zone north of the Carpathians, and in the Danubian Basin. The rise and fall of the Hunnic realm, the subsequent Avar invasions, the concomitant disintegration of German­ ic polities, and the accompanying formation of Slav territories in cen­ tral and eastern Europe are also contributing circumstances.27 The fact that vessels of the Snartemo type have never been found in the context of the Merovingian (Late Germanic) period in Scandi­ navia indicates that the imports ceased in the early sixth century. The production of Frankish glass vessels in the mid-sixth century, e.g. ves­ sels of Kempston type, hardly reached Scandinavia in large numbers. Whether the surmised temporary decrease in prestige goods trade is, in reality, an erroneous conclusion based on an unrepresentative sam­ ple, or whether the decrease reflects historic events, such as for exam­ ple, the Merovingian expansion east of the Rhine, must await further studies of both the typochronology of the period and the relations be­ tween South Scandinavia on the one side and Alamannia and Thur­ ingia on the other, before and after the Merovingian conquest.28 From the beginning of the Merovingian period in Scandinavia (Le. from the mid-sixth century) the amount of imported prestige goods in­ creases rapidly again, including glass vessels produced in Frankish and now plausibly, also Anglo-Saxon workshops. Among the types are claw beakers, squat jars, palm cups, bell beakérs, and bag beakers.29 The distribution map of the wide-spread claw beakers demonstrates the vir­ tual total lack of finds in the Slav and Avar regions. Close connections between Anglo-Saxon England and the Rhineland are demonstrated, in particular between Kent and Scandinavia (Fig. 4). Changes in burial

27 28 29 man,

References in Näsman, Glas och handel i senromersk tid ochfolkvandrmgstid, 106ÍF. See the essay by Ian Wood in this volume. V.I. Evison, “Anglo-Saxon claw beakers”, Archaeobgia 17 (1982), pp. 43-76; Näs­ “Vendei period glass from Eketorp-II, Öland, Sweden”.

THE JUSTINIANIC ERA OF SOUTH SCANDINAVIA

269

customs leave only central Sweden and Godand with a rich collection of wealthy graves and consequently many finds of glass imports. Until recently Denmark lacked claw beakers, but now settlement excavations are producing an increasing number of shards. Scandinavia was already linked to a pan-European network of lux­ ury goods exchange in the Bronze Age and the Early Iron Age. Until as late as the third century A.D., this exchange seems, in Scandinavia, to have been based on an exchange system in which the residences of chieftains served as focal points of tribal distribution patterns. The ar­ chaeological identification of these places is usually based on rich grave finds with imported prestige goods, in Denmark, for example, Hoby on Lolland and Himlingoje on Zealand from the Early and Late Ro­ man Iron Age, respectively.30 However, in one case, the buildings of a magnate’s residence from the late Roman and Early Migration periods have been excavated, at Dankirke in Jutland.31 In the third century, the first central-place appears at Gudme on Funen with an adjacent landing place at Lundeborg. During the Migration period it grew into an important centre with an obviously significant role in regional pro­ duction and distribution as well as in long-distance exchange.32 From the fifth century onwards, the number of central places and landing places increased rapidly in South Scandinavia. These centres were not urban communities, but provide evidence of a new type of centralised economy. They were places where the agrarian surplus was mobilised for use in the construction of a more complex social system. Presum­ ably, the resources of the incipient Danish kingdom were created at these centres. They also fulfilled important religious and political func­ tions. They can be regarded as South Scandinavian pre-urban centres, and the small finds (jewellery, etc) give rich evidence of their links to

30 Hansen, Römischer Impart im Norden, 31 H.J. Hansen, “Dankirke: affluence in late Iron Age Denmark”, The Birth of Europe: Archaeology and social development in thefirst millennium A D ., ed. K. Randsborg (Rome, 1989), pp. 123-28; H.J. Hansen, “Dankirke”, Kami ( 1988-89), pp. 201-47; Zusammenfassung. 32 H. Thrane, “Das Gudme-Problem und die Gudme-Untersuchung”, Frühmittelalter­ liche Studien 21 (1987) pp. 1-48 & pi. 1-16; H. Thrane, “Das Reichtumszemtrum Gudme in der Völkerwanderungszeit Fünens”, Das historische Horizont der Götterbilder-Amulette aus der Übergangsepche von der Spätantike zum Frühmittelalter, ed. K. Hauck (Göttingen, 1992), pp. 299-380; L. Jorgensen, “The find material from the settlement of Gudme II - composi­ tion and interpretation”, The archaeology of Gudme and Lundeborg} ed. P.O. Nielsen et a i, pp. 53-63.

270

ULF NÄSMAN

continental Germanic centres.33 Later, in the early eighth century, the next phase appears to have occurred, with the establishment of the ear­ liest known Scandinavian proto-town, Ribe in Jylland which was founded as a permanently inhabited market and craft centre c. A.D.720.34 The breakdown of the north-south transit routes through central and eastern Europe in the post-Roman era would have had decisive consequences regarding the economic orientation of Scandinavian pol­ ities. Communication to the Mediterranean cultures was from this point in time until the Viking Age limited to the routes through west­ ern Europe. A significant effect was the increasingly important position of South Scandinavia both as a link between Scandinavia proper and the Continent, and as a key to the channels between the Baltic and the North Sea. An increasing pressure on the regions at the Belts and the Sound was certainly one factor which triggered the formation of the Danish kingdom and which was recognised by Jordanes, Procopius, and Gregory of Tours.35 The Scandinavian routes to central Europe and the Mediterranean crossed the territories of the Varni, Thuringians, Alamanni and others, in other words, east of the Franks. After the Merovingian conquests in the early sixth century, the Merovingian realm became a dominating power which could in reality control the Scandinavian supply of luxuries and other commodities from the south. The shift in Scandinavian orientation is important as the back­ ground of the ensuing development of urbanisation and rural econo­ my, a development that by and large followed that of north-west Eu­ rope, modelled primarily on the economy of the northern provinces of

33 U. Näsman, “Some comments on the symposium ‘Social organization and region­ al variation, Sandbjerg Manor, April 1989’”, Samfimdsfundsorganisation og regional variation, ed. C. Fabech &J. Ringtved (Hojbjerg, 1989) pp. 328-33; T. Nilsson, “Stentinget”, K um l (1990), pp. 119-32; S. Jensen and M. Watt, “Trading sites and central places”, Digging into the past, ed. S. Hvass & B. Storgaard (Kobenhavn/ Hojbjerg, 1993), pp. 195-201; J. Ulriksen, “Danish sites and setdements with a maritime context AD 200-1200”, A ntiquity 68 (1994). pp. 797-811; L. Jorgensen, “Stormandssæder og skattefund i 3. —12. ârhundrede”, Fortid og ñutid (1995), pp. 83-110. 34 S. Jensen, The Vikings ofRibe (Ribe, 1992); C. Feveile, “The latest news from Viking Age Ribe”, Developments around the Baltic and the North Sea in the Viking Age. The Twelfth Viking Congress, eds. B. Ambrosiani & H. Clarke (Stockholm, 1994), pp. 91-9. 33 I have arrived at a result similar to the analysis by Olaf Olsen in his paper “Royal power in Viking Age Denmark”, Les mondes normands, ed. H. Galinié (Caen, 1989), pp. 27-32, but I believe that the geopolitical situation was ripe already in the fourth to fifth centuries, two to three hundred years earlier than Oben suggests.

THE JUSTINIANIC ERA OF SOUTH SCANDINAVIA

271

the Frankish realm.36 It led eventually to the formation of Viking Age Scandinavia. In Sture Bolin’s famous revision of the Pirenne thesis, Scandinavian tradesmen contributed to the change of the political and economic geography of western Europe during the Carolingian peri­ od.37 However, this pattern of communication had in fact already de­ veloped prior to the Arab conquests in the western Mediterranean. The shifting to the west of the trade routes from Scandinavia has, as far as I can see, little to do directly with events in the Mediterranean or in the West. The shift was caused by the barrier across eastern Eu­ rope, raised by the establishment of Slav polities during the surge of the ‘Great Migrations’. The changes between the Baltic and the Black Sea are, from a Scandinavian point of view, as important as the chang­ es between the North Sea and the Mediterranean. Therefore, I would suggest, Scandinavian archaeology will be able to contribute to our construction of a post-Pirenne paradigm in a Pan-European perspec­ tive.

The Formation of a Danish Kingdom Any attempt to answer questions about the Danish ethnogenesis fades away in the Early Iron Age. What we know for sure is that a gens called the Danes was mentioned for the first time in the sixth century. Gre­ gory of Tours called their leader rex, so they were obviously led by a sovereign, what we would call a king. The authors of the Early Roman Iron Age mention several tribes in what seems to be South Scandina­ via. In the sources of the sixth century, the number of tribes seems to have become much smaller. Among the most prominent are the Danes. Since written information is not available to elucidate the de­ velopment during the intervening centuries, we do not know whether the same development of tribal confederation took place here, as among the west Germanic peoples, among whom important new pol­ ities appeared: the Franks as a confederacy of tribes belonging to the 36 U. Näsman, “Det syvende ärhundrede”, Fra stamme til stat i Danmark. 2 Hnudingesamjund og kongemagt, cd. P. Mortensen & B. Rasmussen (Hojbjerg, 1991), pp. 165-77; L. Jorgensen, “Stormandssaeder og skattefund i 3. - 12. ärhundrede”, who perhaps takes his analogy to the Frankish economic and political system a little too far. 37 S. Bohn, “Mohammed, Charlemagne and Ruric”, The Scandinavian economic history review 1 (1954), pp. 5-39. For further debate on the Pirenne-Bolin theses, see R. Hodges, Dark Age Economics (London, 1982); R. Hodges & D. Whitehouse, Mohammed, Charlemagne & the Origjms of Europe (London, 1983).

272

ULF NÄSMAN

tribal group of the Istvaeones, the Alamanni, who grew out of a Suebic group, and the enigmatic Saxons at the root of the Cimbrian peninsu­ la, the formation of whom seems to be based on a group of peoples sharing the Nerthus cult.38 However, in the same way as the appear­ ance of Franks, Alamanni and Saxons was the result of tribal confed­ eration, one may suggest that the Danes, in the course of the second to sixth centuries, gradually achieved a more or less total hegemony in South Scandinavia, in a confederacy that came to embrace large parts of what later became Denmark. This hypothesis can be transferred to the analyses of material culture published by Danish archaeologists.39 For the Early Roman Iron Age, a number of small archaeological regions can be distinguished, perhaps as many as fifteen in the present Schleswig, Denmark and Scania. For the Late Roman Iron Age, the number has decreased, and in the be­ ginning of the Migration period one can barely discern more than sev­ en regions. The regions of South Scandinavia in the Late Roman — Early Migration period were of varying size. If they represent indepen­ dent polities, in size they correspond reasonably well to what Tribal Hidage tells us about polities in Anglo-Saxon England.40 The two pol­ ities of Jylland (ca. 15,000 sq. km each) correspond roughly to Mercia and East Anglia. Skâne (11,239 sq. km) should probably be divided into a south-western and an eastern region, which would correspond roughly to Kent and the South Saxons, respectively. Sjælland (7,543 sq. km) is the same size as Kent. Fyn (3,482 sq. km) is comparable to Anglo-Saxon kingdoms like the Lindesfarona and the East Saxons. Lol­ land and Falster (1,797 sq. km) correspond to small Anglo-Saxon units like the Hendrica, and Bornholm (587 sq. km) is comparable in size to the Pecsætna (the Peak dwellers). The core of Thuringian culture in the

38 R. Wenskus, Stammesbildung und Verfassung (Cologne, 1977); E. Zöllner, Geschichte der Franken bis zur Mitte des 6. Jahrhundert (München, 1970) 2f; R. Christlein, Die Alamannen. Archäologie eines ¡ebendigenVoÚces (Stuttgart, 1979), p. 22; C. Ahrens, “Vorbemerkungen zur Archäologie und Geschichte der Sachsen und Angelsachsen”, and A. Genrich, “Ur­ sprung und Verbreitung der Altsachsen bis zum 5. Jahrhundert”, both in Sachsen und An­ gelsachsen, ed. C. Ahrens (Hamburg-Harburg: 1978), pp. 17-43, pp. 43-50; I. Wood, The Merovingian Kingdoms 450-571 (London, 1994), 35fF. 39 For example, J. Ringtved, “Jyske gravfund fra yngre romertid og aeldre germanertid”, Kami (1986), pp. 95-231; K.H. Nielsen, “Centrum og periferi i 6.-8. ärh”, Fra stamme til slot i Danmark. 2 Hmdingesamfiind og kungemagt, ed. P. Mortensen & B. Rasmussen (Ârhus/Hojbjerg, 1991), pp. 127-54. 40 D. Hill, An atlas ofAnglo-Saxon England (Oxford, 1981), fig. 136ff.; Yorke, Kings and kingdoms of early Anglo-Saxon England, pp.lOf.

THE JUSTINIAN!C ERA OF SOUTH SCANDINAVIA

273

fifth-sixth century covered an area of ca. 15,000 sq. km., and including more peripheral parts the extent of the kingdom was at least 22.000 sq. km.41 Therefore, the scale of the polities of South Scandinavia that I suggest interacted in the formation of the Danish hegemony seems right. In the sixth century and in the Merovingian period only two to three distinguishable regions appear (plus the island Bornholm). Present-day Denmark is only c. 42,900 sq. km large, so a Danish kingdom of this size is comparable to the assessment of Wessex in the Tribal Hidage: c. 48,480 sq. km. In the Viking Age and the High Middle Ages region­ al traits are still discernable indicating that in the long-lasting process of the Danish kingdom formation, differences between east and west Denmark endured.42 There are indications in both written sources and in material culture that the Danes in the sixth century had won hegemony in South Scan­ dinavia, and in the Merovingian period there is no reason to doubt that the Danes led the most powerful kingdom in South Scandinavia. Thus the sixth century seems to be one of the more marked dividing lines drawn in Danish history. In the traditional archaeological narra­ tive, analysis of the period was influenced by the general historical view of the Migration period as a period of collapse, and consequendy it was described as afflicted by a deep economic and political crisis as well as by devastating warfare,43 with setdement abandonment and population decrease believed to have been the result. However, ad­ vances in setdement archaeology and landscape studies have provided a new picture of the long-term evolution of the South Scandinavian cultural landscape.44 The rural setdement pattern changed in the pe­ riod around c. A.D.200 and many small sites combined to form larger nucleated setdements consisting of large well-built farmsteads. The lay­ out of farm buildings and the organisation of farmsteads into smaller and larger hamlets remained relatively stable for a long period, until

41 B. Schmidt, “Die Thüringer”, Die Germanen II, ed. B. Krüger (Berlin, 1983), pp. 502-48. 42 A. Andrén, “Städer och kungamakt”, Scandia 49/1 (1983), pp. 31-76; summary p. 159f. 43 Critical review of the arguments in U. Näsman, “Den folkvandringstida Pkrisen i Sydskandinavien”, Folkevandringstiden i Norden. En krisetid mellem aldre ogyngre jemalder?, ed. U. Näsman & J. Lund (Arhus, 1988) pp. 227-55. 44 B.E. Berglund, (cd.), The cultural landscape during 6000years in southern Sweden —the Tstad Project (Lund, 1991).

274

ULF NÄSMAN

the next change around c. A.D.700.45 Palaeoecological and archaeo­ logical studies of the land-use indicate relatively stable subsistence pro­ duction in all main rural districts.46 However, in contrast to the relative economic stability in the period before and after the sixth century, other sources reveal substantial so­ cial and political change. Funerary rituals changed and the display of wealth in burials ceased entirely in some regions, and in others became eventually a rare phenomenon. In South Scandinavia, rich burials are, with a few exceptions, absent in the sixth century.47 The thousand year-old practice of ritual hoarding changed from wet to dry contexts, from lakes and bogs to settlement sites (Fig. 5). The many offerings of spoils-of-war characteristic of the second to the fifth centuries reflect, it is surmised, the widespread warfare of the period. These offerings stop around AD. 500.48 The change of hoarding behaviour has been interpreted as a shift from a collective cult of fertility gods with cere­ monies at bogs and lakes, to religious ceremonials that took place at the residences of charismatic individuals.49 Most interestingly, contemporary pictorial art also reveals a change, a shift in the propaganda of the domestic elite. Nordic art in the Late Roman Iron Age and early Migration period presented numinous mes­ sages propagating the ideas of an ‘old society’ but from now on it il­ lustrated an ideology that is clearly related to a new political power. An early group of Gotlandic picture stones and the contemporary gold bracteates convey highly symbolic and conservative messages of elevat­ ed gods. The new evolution is illustrated by later groups of Gotlandic picture stones as well as by gold foil figures and stamped figures on helmets. They have an obvious narrative content, depicting gods, he­ roes, and the ceremonial behaviour of the war lords’ retainers. The early motifs strongly reflect the late-Roman influence on Germanic art 45 S. Hvass, “The status of the Iron Age settlement in Denmark”, Arheologie en landschap, ed. M. Bierma et al. (Groningen, 1988), pp. 97-132; S. Hvass, “The Iron Age and the Viking Period: Setdement”, Digging into the past, ed. S. Hvass & B. Storgaard (Copen­ hagen, 1993), pp. 187-94. 46 Considerable desertions of settlement are however found in a few regions, for in­ stance in Angeln —K-H. WiUroth, Untersuchungen zur Besiedlungsgeschichte der Landschaften An­ geln und Schwansen (Neumünster, 1992). 47 Ringtved, “Jyske gravfund fra yngre romertid og ældre germanertid”, Hedeager, Iron-Age Societies. 48 See note 15. 49 Fabech, “Booty sacrifices in Southern Scandinavia: a reassessment”; idem., “Read­ ing society from the cultural landscape. South Scandinavia between sacral and political power”.

THE JUSTINIAN!C ERA OF SOUTH SCANDINAVIA

275

rs « s 5 a e? w bC_O « C¡ « C & -S Ö-5

■*SJ T 3 . ~d Up § n iÜlus ti

S3-73 § T jl" 3 S’C « ■g frS-

iDy lwÏ13l— ? *i CS/3 S S 3 ^ TÎ -C .

£

4J

Fi ill

J ££ e0!s-8 *c *5 o *» g-e a5 c-S^ ,2

ití

■SsV ■g g I oa w*** ¿ I c o*3 «

ñQJ «