Settlement and Social Organization: The Merovingian Region of Metz

This book examines one region of north-eastern Gaul (that around Metz, in northeastern France and southern Germany) in the period between the end of the Roman Empire and the accession of Charlemagne. It is the first detailed regional study of a part of the Merovingian kingdoms to appear in English for almost twenty years. Using all available evidence, both documentary and archaeological, a new and multidisciplinary approach is adopted. The study deals with a broad range of historical themes. In particular, by looking at the reasons behind the creation of different forms of evidence, it examines how the different facets of social organization (ethnicity, gender, age and social hierarchy) were related intimately to each other and to contemporary settlement patterns of the region. As a result, it is argued that the Merovingian period was not one of slow 'transformation' from cRoman' to 'medieval' but was one of constant, dynamic social change and diversity, even between the recognized periods of dramatic upheaval.

Settlement and social organization

Settlement and social organization The Merovingian region of Metz

GUY HALSALL Birkbeck College, University of London

CAMBRIDGE UNIVERSITY PRESS

PUBLISHED BY THE PRESS SYNDICATE OF THE UNIVERSITY OF CAMBRIDGE

The Pitt Building, Trumpington Street, Cambridge, United Kingdom CAMBRIDGE UNIVERSITY PRESS

The Edinburgh Building, Cambridge CB2 2RU, UK 40 West 20th Street, New York NY 10011-4211, USA 477 Williamstown Road, Port Melbourne, VIC 3207, Australia Ruiz de Alarcon 13, 28014 Madrid, Spain Dock House, The Waterfront, Cape Town 8001, South Africa http://www.cambridge.org © Cambridge University Press 1995 This book is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 1995 First paperback edition 2002 A catalogue recordfor this book is available from the British Library Library of Congress Cataloguing in Publication data Halsall, Guy, Settlement and social organization: the Merovingian region of Metz/GuyHalsall. p. cm. ISBN 0 521 44256 7 1. Metz Region (France) - History. 2. Merovingians - France - Metz Region. 3. Metz Region (France) - Antiquities. 4. Excavations (Archaeology) - France - Metz Region. 5. Metz Region (France) - Social conditions. I. Title. DC801.M65H36 1995 936.4-dc20 94-31106 CIP ISBN 0 52144256 7 hardback ISBN 0 521521890 paperback

To Mum, Dad, Jon and Liz

Contents

List of illustrations Acknowledgements List of abbreviations Notes on the citation of sources 1

Introduction

Part I

Social organization

page xi xv xvii xx 1

Introduction 2 3 4

21

Social organization: descriptive analysis of the documentary evidence Creating a model: cemeteries of the Merovingian civitas of Metz

26 75

Testing the model: cemeteries outside the civitas of Metz

110

Part II Settlement Introduction

167

5

Rural settlement

175

6

Intermediate settlement: castra, vici, palaces and monasteries

7

Urbanism in Metz

199 214

Part III Conclusions 8

Town and country, c. 350 - c. 600

245

9

The later Merovingian period

262

Bibliography Index

283 301

IX

Illustrations

1.1 1.2 1.3 1.4 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 3.10 3.11 3.12 3.13 3.14 3.15 3.16 3.17 3.18 3.19 3.20 3.21 3.22 3.23

The civitas of Metz page 4 The 'region' of Metz 5 Physical geography of the region of Metz 6 Distribution of place-name suffixes, by commune, in the part of the civitas of Metz now in France 11 Places mentioned in chapter 3 75 Merovingian cemeteries in the civitas of Metz (including all possible sites) 76 78 Ennery cLes Trois Arbres': site-plan Numerical coding of artefact-types used in analyses described in chapters 3-4 80 Ennery 'Les Trois Arbres': artefact associations 81 Ennery cLes Trois Arbres': grave seriation, and sex predictions 82 Ennery cLes Trois Arbres': association between artefact-type and age group 84 Ennery 'Les Trois Arbres': predicting age and sex of disturbed graves 86 Ennery 'Les Trois Arbres': arrangement of the grave 87 89 Ennery cLes Trois Arbres': distribution of grave-types Ennery 'Les Trois Arbres': distribution of sexes 90 Ennery 'Les Trois Arbres': average number of grave-goods in male graves, by age group 92 Ennery 'Les Trois Arbres': numbers of artefact-types and grave-goods in adult male graves, by age group 93 95 Altheim cam Knopp': site-plan Berthelming 'Alt Schloss': site-plan - state of excavations in 1948 96 97 Bettborn cBienenzaun': site-plan Hayange Toret de Hamevillers': site-plan - state of excavations in 1910 101 Hayange Toret de Hamevillers': artefact associations 101 Hayange Toret de Hamevillers': grave seriation 102 Moyeuvre-Grande 'Kleiner Vogesenberg': site-plan 103 Walsheim 'am DorP: artefact associations 104 Wittersheim 'vor dem langen Zaun': site-plan 105 Lavish burials (1): graves with shield-boss, spur or item of horse-harness 107 XI

ILLUSTRATIONS

3.24 4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8 4.9 4.10 4.11 4.12 4.13 4.14 4.15 4.16 4.17 4.18 4.19 4.20 4.21 4.22 4.23 4.24 4.25 4.26 4.27

Lavish burials (2): graves with shield-boss, spur, item of horse-harness, bronze bowl, bucket, ango or balances and horse-burials Places mentioned in chapter 4 Classification of grave-types and number code for the ageing of skeletons Chaouilley caux Ecailles': site-plan Chaouilley 'aux Ecailles': artefact associations Chaouilley 'aux Ecailles': grave seriation Chaouilley 'aux Ecailles': average numbers of artefact-types, by age group Chaouilley 'aux Ecailles': cemetery organization - schematic plans Lavoye cLe Haie des Vaches': site-plan Lavoye 'Le Haie des Vaches', phase 1: late fifth and early sixth centuries Lavoye 'Le Haie des Vaches', phase 1: artefact associations Lavoye 'Le Haie des Vaches', phase 1: grave seriation Lavoye 'Le Haie des Vaches', phase 2: mid- to late sixth century Lavoye 'Le Haie des Vaches', phase 2: artefact associations Lavoye cLe Haie des Vaches', phase 2: grave types, and age and sex predictions Lavoye cLe Haie des Vaches', phase 2: 'points systems' for age predictions Lavoye cLe Haie des Vaches', phase 2: young adult and adolescent female burials Lavoye cLe Haie des Vaches', phase 2: cemetery organization, and distribution of adult males and females Lavoye 'Le Haie des Vaches', phase 2: distribution of grave-types Lavoye £Le Haie des Vaches', phase 2: numbers of artefacttypes and grave-goods in masculine graves Lavoye 'Le Haie des Vaches', phase 2: distribution of male 'prestige graves' Lavoye 'Le Haie des Vaches', phase 2: numbers of artefacttypes and grave-goods in feminine graves Lavoye 'Le Haie des Vaches', phase 2: distribution of female 'prestige graves' Lavoye 'Le Haie des Vaches', phase 3: seventh century Lavoye 'Le Haie des Vaches', phase 3: artefact associations Lavoye 'Le Haie des Vaches', phase 3: grave-types, and age and sex predictions Lavoye 'Le Haie des Vaches', phase 3: cemetery organization, and distribution of sexes Lavoye 'Le Haie des Vaches', phase 3: distribution of grave-types Xll

108 110 111 112 113 114 116 117 118 120 121 122 123 124 125 126 127 128 129 130 131 133 134 135 136 137 138 139

ILLUSTRATIONS

4.28 4.29 4.30 4.31 4.32 4.33 4.34 4.35 4.36 4.37 4.38 4.39 4.40 4.41 4.42 4.43 4.44 4.45 4.46 5.1 5.2 5.3 5.4 5.5 6.1 6.2 6.3 6.4 6.5 6.6 7.1 7.2 7.3 7.4 7.5

Lavoye cLe Haie des Vaches', phase 3: numbers of artefacttypes and grave-goods in masculine graves Lavoye cLe Haie des Vaches', phase 3: numbers of artefacttypes and grave-goods in feminine graves Lavoye cLe Haie des Vaches', phase 3: distribution of 'prestige graves' Dieue-sur-Meuse 'La Potence': site-plan Dieue-sur-Meuse £La Potence', phase 1: sixth century Dieue-sur-Meuse cLa Potence', phase 1: artefact associations Dieue-sur-Meuse 'La Potence', phase 1: grave seriation, and age and sex predictions Dieue-sur-Meuse 'La Potence', phase 1: distribution of grave-types Dieue-sur-Meuse 'La Potence', phase 1: comparison of gravesize and gender-specific grave-goods Dieue-sur-Meuse 'La Potence', phase 1: comparison of gravesize and number of artefact-types Dieue-sur-Meuse 'La Potence', phase 2: seventh century Audun-le-Tiche 'Bois de Butte': site-plan Audun-le-Tiche 'Bois de Butte': artefact associations Audun-le-Tiche 'Bois de Butte': grave seriation Audun-le-Tiche 'Bois de Butte': distribution of grave-types Audun-le-Tiche 'Bois de Butte': comparison of grave-size and gender-specific grave-goods Audun-le-Tiche 'Bois de Butte': comparison of grave-size, grave-goods and artefact-types Audun-le-Tiche 'Bois de Butte': cemetery organization, and prestige graves Audun-le-Tiche 'Bois de Butte': distribution of nails, potsherds, animal bones, shells and ashes Places mentioned in chapter 5 Roman settlement in the civitas of Metz The Seille valley survey Sixth-century and seventh-century cemeteries Eply-Raucourt, 'Haut de Villers' Places mentioned in chapter 6 Castra: Le Heraple Castra: Saarbriicken Castra: Sarrebourg Castra: Scarponne Castra: Tarquimpol Metz: sites mentioned in chapter 7 Metz: early Roman The Roman walls of Metz A seventeenth-century engraving of the small amphitheatre Archaeomagnetic intensity dating of tiles from St-Pierre-auxNonnains: results Xlll

140 141 142 143 144 145 146 147 148 149 150 151 153 154 155 157 158 159 161 176 178 179 185 187 199 203 204 204 206 207 215 216 220 221 223

ILLUSTRATIONS

7.6 7.7 7.8 7.9 7.10 7.11 7.12 8.1 9.1

Metz: c. 270 - c. 350 Metz: c. 350 - c. 400 Metz: c. 400 - c. 500 Metz: c. 500 - c. 550 Metz: c. 550 - c. 600 Metz: c. 600 - c. 700 Metz: c. 700 - c 800, and later Places mentioned in chapter 8 Places mentioned in chapter 9

xiv

225 226 229 232 234 235 237 245 262

Acknowledgements

This book is the barely recognizable descendant of an exceptionally erratic doctoral thesis (Halsall 1990). During the seven years, on and off, spent working on that thesis and this book, I have accumulated a number of intellectual and other debts which it is my pleasure to record here. Starting at home, I must express my thanks to my family and friends, who have always been ready with love and support when I needed them. Intellectually, my greatest debt is to Dr Edward James. I hope he will find in the ideas expressed here some reward for the personal and academic support he has given me, and for all his time, effort and, above all, patience during seven years as my undergraduate and post-graduate supervisor. I must also thank the examiners of my thesis, Bryan Ward-Perkins (Trinity College, Oxford) and Steve Roskams, for showing faith in a very crude preliminary sketch. I should, in particular, like to thank Steve, who has been a source of inspiration and encouragement since my first undergraduate days. I am pleased to record my thanks to the other medievalists of the University of York who have taught, helped and encouraged me over the years, in particular, Emeritus Professor Phillip Rahtz, Professor Barrie Dobson (now of Christ's College, Cambridge), Peter Biller, Tania Dickinson, Julian Richards and Peter Rycraft. In France my research benefited from the assistance of Mme M. ClermontJoly (Musee de l'Histoire du Fer, Jarville); M. J.-L. Coudrot and Mme M. Sary (Musees de Metz); Mme C. Aptel (Musee Historique Lorrain, Nancy); Mme D. Heckenbenner (Musee du Pays de Sarrebourg); Mme N. Dautremont, M. P. Thion, M. C. Lefebvre and M. P. Brunella (DAHL); Mile L. Roux (Archives departementales de la Moselle); and the staff of the Bibliotheque, Universite de Nancy II, and of the Mediatheque du Pontiffroy, Metz. In Germany I owe a huge debt of thanks to Professor Dr Frauke Stein (Universitat des Saarlandes), who gave me hospitality in Saarbriicken, as well as invaluable help with the area's cemetery archaeology, allowing me access to her excellent archive of notes on the subject. Back in the United Kingdom, Patrick Wormald (Christ Church, Oxford) and Paul Fouracre (Goldsmiths' College, London) have saved me from a number of errors. From the many, of omission and commission, which doubtless remain, it is more than usually necessary to dissociate all of the above.

XV

ACKNOWLEDGEMENTS

Finally, I should like to thank Dr Barry Coward and Dr Emma Dench, my colleagues in the Department of History at Birkbeck, for helping to keep me sane in the increasingly mad world of British higher education. Guy Halsall

XVI

Abbreviations

Primary sources Dipl Karol. + No. Ep. Aust. GC GDRF GM Gorze + No. Lex Ribv. LH LHF Lib.Ep.Mett. Pardessus + No. Pertz, Mayoral + No. Pertz, Royal + No. Pertz, Spuria + No. PLS Reg. Can. St-M. + No.

Charter in Miihlbacher (ed.) 1906, with its number in that collection. Epistulae Austrasiacae: Gundlach (ed.) 1957:403-70. Gregory of Tours, Liber in Gloria Confessorum: Krusch & Levison (eds.) 1969:294-370; Van Dam (trans.) 1988b. Gesta Dagoberti Regis Francorum: Krusch (ed.) 1888:396-425. Gregory of Tours, Liber in Gloria Martyrum: Krusch & Levison (eds.) 1969:34-111; Van Dam (trans.) 1988a. Charter of the Gorze cartulary with its number in d'Herbomez (ed.) 1898. Lex Ribvaria: Beyerle &c Buchner (eds.) 1951; Rivers (trans.) 1986:167-214. Gregory of Tours, Libri Historiarum: Krusch & Levison (eds.) 1951; Thorpe (trans.) 1974. Liber Historiae Francorum: Krusch (ed.) 1888:215328; Bachrach (trans.) 1973; Gerberding 1987:173ff (less complete but more accurate). Paul the Deacon, Liber de Episcopis Mettensibus: Waitz (ed.) 1878:260-70. Charter in Pardessus (ed.) 1843-9, with its number in that collection. Mayoral charter in Pertz (ed.) 1872. Royal charter in Pertz (ed.) 1872. Charter included in Pertz (ed.) 1872, Mayoral spuria section (pp.209-15). Pactus Legis Salicae: Eckhardt (ed.) 1962:1-267; Rivers (trans.) 1986:39-144; Drew (trans.) 1991:59156. Chrodegang, Regula Canonicorum: Pelt (ed.) 1937. Charter in the cartulary of St-Mihiel, with its number in Lesort (ed.) 1909-12. xvii

ABBREVIATIONS

TBLM VA VG VP VRAH VSM Wiss. + No.

Testament of Berthechramn of Le Mans: Busson & Ledru (eds.) 1901:101-41. Vita Arnulfi: Krusch (ed.) 1888:426-46. Vita Gaugerici: Krusch (ed.) 1896:852-8; Rouche (trans.) 1986. The page numbers used are those in Rouche's translation. Gregory, Vita Patrum: Krusch & Levison (eds.) 1969:211-94; James (trans.) 1991. Vita Romarici Abbatis Habandensis: Krusch (ed.) 1902:221-5. Gregory of Tours, Liber de Virtutibus Sancti Martini: Krusch & Levison (eds.) 1969:134-211; Van Dam 1993:199-303. Charter in the Wissembourg cartulary with its number in Glockner & Doll (ed.) 1979.

Archaeological Association Franfaise d'Archeologie Merovingienne. Direction des Antiquites Historiques de la Lorraine. Staatliches Konservatoramt Saarbriicken.

AFAM DAHL SKS

Periodicals, etc. Actes . . . Arch. Med. ASHAL VAustrasie BAR(B) BAR(S) Ber.RGZM Ber.SDS

XVlll

Actes des Xe Journees Internationales d'Archeologie Merovingienne. Metz 20-23 Octobre 1988 (Sarreguemines 1989). Archeologie Medievale. Annuaire de la Societe d'Histoire et d'Archeologie Lorraine. VAustrasie. Revue de Metz et de Lorraine. British Archaeological Reports (British Series). British Archaeological Reports (Supplementary Series). Bericht der romisch-germanischen Zentralmuseums Mainz (a) Bericht des Konservators der geschichtliche Denkmdler im Saargebiet (vols.1-5, 1923-34). (b) Bericht der staatlichen Denkmalpflege im Saarland. Beitrdge zur Archdologie und Kunstgeschichte (vols.6- , 1953- ).

ABBREVIATIONS

BLAFAM BSAHM BSAL BSCMHA BSL CAHA JGLGA JSAL JVEM KWZGK LB LFS LS LT MAM MGH MHVP MSAHM MSAL RAE REL TRHS WG

Bulletin de Liaison de I'Association Francaise d'Archeologie Merovingienne. Bulletin de la Societe d'Archeologie et d'Histoire de la Moselle. Bulletin de la Societe d'Archeologie Lorraine. Bulletin de la Societe pour la Conservation des Monuments Historiques d'Alsace. Bilan Scientifique Lorraine. Direction Regionale des Affaires Culturelles. Service Regional d'Archeologie. Cahiers d'Archeologie et d'Histoire d'Alsace. Jahrbuch der Gesellschaft fur Lothringische Geschichte und Altertumskunde (became ASHAL after 1918) Journal de la Societe d'Archeologie Lorraine (BSAL after 1900). Jahrbuch des Vereins fur Erdkunde zu Metz. Korrespondenzblatt der Westdeutschen Zeitschrift fur Geschichte und Kunst. Linckenheld 1933. Linckenheld 1932a. Linckenheld 1931. Linckenheld 1934. Memoires de I'Acadamie Nationale de Metz. Monumenta Germaniae Historica. Mitteilung des historische Vereins der Pfalz. Memoires de la Societe d'Archeologie et d'Histoire de la Moselle. Memoires de la Societe d'Archeologie Lorraine Revue Archeologique de I'Est et de Centre Est. REL 1901-3, vol. Ill (Ortsbeschreibung). Transactions of the Royal Historical Society. Wolfram & Gley 1931.

xix

Notes on the citation of sources

Primary sources Where the abbreviation of a primary source's title is separated from the reference by a colon, the reference is to page numbers in the edition used (in other words, the source is not divided into books, chapters or verses). Elsewhere, references to book (using upper case Roman numerals), chapter (using arabic numerals) and verse (using lower case Roman numerals) are used where appropriate.

Archaeological discoveries References to yearly reports of acquisitions of local museums (e.g. 'Dons faits au musee' in JSAL), annual summaries of discoveries in the region, to proces verbaux of society meetings, the annual 'chroniques' in Archeologie Medievale or 'informations archeologiques' in Gallia (and to reports in the separate journal Gallia Informations) are given simply as the year of the appropriate volume, and the page number in question. References to the BSCMHA in which the page number is preceded by (SC) refer to the proces verbaux of the seances du comite, where these have a separate pagination within the volume. Only separate articles on individual sites, or groups of sites, are given as a separate Harvard-system reference.

xx

1 Introduction

A farewell to the 'longue dureey Three hundred years is a long time; in important respects it was even longer in the middle ages. With average adult life expectancy in the forties a man might expect to die before he had seen his grandchildren. From a twentiethcentury standpoint, however, we are accustomed to see the three hundred years of the Merovingian period as a single entity. Writers jump back and forth between the sixth and eighth centuries, and from the seventh to the fifth, and in doing so often make similarly huge spatial as well as chronological leaps from one anecdote to the next, crossing hundreds of miles in a single bound.1 The flaws of such methodology can be highlighted by the simple expedient of adding 1,000 years to the dates in question. An early modern historian would be ridiculed who claimed that the evidence from fifteenth-century Provence, eighteenth-century Prussia and sixteenthcentury Spain could be thrown together to make a single point about the nature of lordship, the status of women or patterns of family life. Yet the early medieval historian has come to accept such aggregation of evidence as valid, and the particular testimony of individual documents more generally applicable, in this period. This 'melting pot' approach denies significant regional diversity or dynamic social change. As a result, we have grown used to thinking of the period c. 450 - c. 750 as one of 'transition', from the cRoman' world to the 'medieval'; of the 'passage' from antiquity to feudalism. The image is of slow, barely perceptible change down a single path. Certain debates have focused upon the point on this unilinear 'path' at which slavery 'gave way' to feudalism, or where the ancient world 'yielded place' to the medieval (Anderson 1974; Wickham 1984; Bois 1992; Murray 1983:221 for explicit 'pathway' imagery). From this perspective, our participants in this transition will not have seen much change in their own lifetimes, even though their experiences were very different from those of their remote ancestors or descendants. The image of 'passage' or 'transition' also assumes that the situation on either 1

This approach is so common as to defy any fair or meaningful selection. It is, for example, rife in Doehaerdt 1978 (see, for example, pp.85-127). Even as prominent an early medievalist as Janet Nelson has not always escaped its temptations (cp. Nelson 1990:72-4). Similar, or worse, cases of diachronicity could be drawn from many other pieces, such as Halsall 1989. Morris (1992:10-11) makes a similar criticism of classical social history, and Van Dam (1993:13) of the study of saints' cults.

INTRODUCTION

side of the period of transition was, in some way at least, uniform or constant. This is, for example, a major flaw in Bois' (1992) analysis. This book attempts to confront such imagery. It is true that mid-eighthcentury society was very different from that of the mid-fifth. The 'passage' from one social organization to the other was not, however, pre-ordained. It is no mere truism to say that the 'passengers' themselves, the people of the fifth, sixth, seventh and eighth centuries, did not know where they were going. They experienced many changes in their lifetimes. Those who lived through the stress of the early eighth century must have been aware of changes. But so too were those who lived in the turbulent days of the end of the Roman Empire and the break-up of its social order, and those alive in the decades around 600 when the aristocracy was consolidating its power. There were, in short, many changes in social organization between 450 and 750.2 Each change was brought about by social struggle, but each did not in itself mark a 'step' on the 'road' to feudalism. There was no reason why at each juncture a different outcome could not have been produced, why our passengers could not have carried on in a completely different direction, or even set off back to a social structure similar to that of the late Roman period (as far as kings were concerned, this was usually what they thought they were doing). The path from antiquity to feudalism in north-eastern Gaul was a difficult, winding and tortuous one, cut by the people of the time with few or no sign-posts to help them. Society changed dramatically between 450 and 750, but no more so than it had done between 150 and 450 or than it was to do between 750 and 1050. Single processes of transformation are only discernible with considerable hindsight. When we define them, from our modern viewpoint, we mask many histories, especially when we study predetermined transformations on a broad geographical as well as chronological scale. Studies which concentrate upon describing specific changes in the patterns of landholding suffer when they attempt to explain them in terms purely of lordship and dependence; should we not also examine the effects of such processes on family structures, gender relations, the social roles of age, fictive kinship and so on, and vice versa? The compartmentalized areas of social history, such as women's history or the history of the family, suffer similarly in their analyses by exclusion of the wider patterns of socio-economic interaction. The brilliance of Bloch's (1962) study of feudal society derives in part from its breadth of scope. The approach adopted here has been at once to narrow and to broaden: on the one hand to narrow the geographical basis of the study, but, on the other, to broaden its thematic coverage, though it is still by no means comprehensive. In these general terms, the agenda is not dissimilar to Guy Bois' in his examination of Lournand (1992) or that of Wendy Davies' (1988) excellent study of Carolingian Brittany. Though focusing on important social changes around 600 AD, I have retained the long, 300-year time-span, in order both to put these changes in perspective and to show how society did not remain static before or after these interesting decades. Although they 2

Goetz 1993:50 makes a similar point; he too, nevertheless (ibid.:34-5), treats all 'Germanic' law between c. 500 and c. 750 together to make a single point about post-Roman slavery.

INTRODUCTION

exist3, regional studies of the immediately post-Roman centuries have not yet been used to stress regional variation in the chronology of social developments or local diversity of experience. The reasons for this are, I suggest, twofold. The first is a belief in a kind of cultural uniformity of the postRoman West, be it founded in a homogenous Germanic culture,4 the common instincts of 'heroic' society, or both (Hedeager 1992), or in a common Roman heritage (Fouracre 1992). The second is a belief in the primacy of written evidence, which is sparse by later standards. Just as, as mentioned, it has frequently been drawn from very different times in order to make up a 'complete5 picture, so it has also been trawled from widely differing places. This aggregation of evidence, of course, produces a picture of monolithic early medieval society, and so we return to where we started: the view of the Merovingian period as one when it took 300 years to move from one social organization to another. Unidisciplinary analyses have exaggerated the problems just outlined, but so have studies of particular forms of evidence (settlement archaeology, burial archaeology, saints' lives, legal material, etc). The restriction to a single evidential form has produced mutually exclusive explanations of social change. For example, some settlement studies (Percival 1976:ch.8) hypothesize about the nucleation of society around powerful people, whilst studies of burial evidence (e.g. Samson 1987a) stress social instability, and competition. It is easy enough to criticize previous attempts at analysis; what are we to put in their place? To begin with, a 'multi-disciplinary' approach has been adopted. Rather than skipping back and forth from documentary data to archaeological, each body of evidence has been analysed on its own merits. Where possible, even different sub-types of data have been examined separately (this is particularly true of cemetery, rural settlement and urban archaeological evidence). Data have been contextualized as rigidly as possible in time and space, and conclusions reached from each category of information are merged at a higher, more critical level. The premise behind this approach is that cultural practices are deliberate, meaningfully constituted and historically contingent. This is true whether we are talking of the naming of children, the design of a belt-set, the lay-out and furnishing of a grave, the plan of a house or settlement, the contents and structure of a law-code, charter or saint's vita, or the specific words chosen to express a writer's thoughts. The use of particular grave-goods in individual burials, or of different shapes, sizes and colours of ceramics in varying contexts is thus deemed to be significant and studied as such. Internal semantic analysis of texts is also held to be valid; as has long been argued (cp. Campbell 1979), writers did not use words at random. 3

4

For example James 1977; Kaiser 1973; Rouche 1979. There are very many more artefactually based archaeological regional surveys. For the region in question, see the invaluable Clermont-Joly 1978; Simmer 1987, which contains a bizarre attempt at synthesis; and Stein 1989b. This is an incredibly pervasive idea, although founded ultimately upon Germanist philosophies of the last century. As another unfair example pulled from a veritable sea of possible instances, see the 'unity of culture and ethos in Germanic society' (Burnell 1988:402). Archaeologically it is manifested in the idea of the Reihengrdberzivilisation, the 'row-grave-civilization', spreading from northern Italy to northern England. This term reduces widespread regions to (blatantly false) cultural uniformity simply on the grounds that their inhabitants buried their dead with grave-goods, hardly a historically or geographically restricted phenomenon.

INTRODUCTION

Transformations in the nature of society will thus be represented in a broad spectrum of cultural traces. This study attempts to show how different areas of human activity reflected the same processes of social change. Its theoretical starting point has been influenced by the post-processual archaeology (e.g. Hodder 1986; Miller 1985; Shanks & Tilley 1987; Tilley (ed.) 1991), but some early medieval historians have recently been advocating a more rigidly contextualized approach to documentary evidence, studying individual texts separately and seeing them similarly as actively, meaningfully constructed and historically contingent (Fouracre 1990; Wood 1992a). Ultimately, we should see all our evidence as 'textual', whether the text be formed by the marking of recognized signs on parchment, or the arrangement and display of understood material symbols in burials.

The study area and its geography The study focuses upon the Roman civitas of Metz (fig. 1.1), whose bounds, it seems, later crystallized as the diocese of Metz (see below). The evidence from the civitas was often, however, found to be insufficient to test ideas

civitas of Toul

Fig. 1.1

The civitas of Metz

INTRODUCTION

about social organization, so evidence was selected from a slightly wider area, the region of Metz of the title (fig. 1.2), extending south and west of the civitas. Evidence from north and east of the diocese was not discussed, partly because the Triererland has been thoroughly, and splendidly, studied (Bohner 1958; Ewig 1954), and partly because of physical geography. The region of Metz forms the north-eastern corner of the Paris basin, bordered to the north and east by the Hunsriick, Pfalzerwald and the Vosges (fig. 1.3). The Vosges, sandstone mountain ranges reaching up to 1,420m above sea level at their highest, drop gradually to the Plateau Lorrain, the foothills reaching down to the Sarre valley. In this region the highest peak is Le Donon (1,009m), traditionally seen as the boundary between the civitates of the Leuci, Tribocci and Mediomatrici. The Plateau Lorrain, gently rolling country, generally between 200 and 300m above sea level, is formed of Jurassic limestone. It descends gradually to the Moselle valley to the west and the Seille valley to the south. Around the Seille valley is the Saulnois, roughly corresponding to the Merovingian Pagus Salinensis (see below), an area involved in salt production since prehistory. The border between the civitates of Metz and Toul appears to have run along a series of low hills, but this division is not strongly marked geographically. The Seille valley is broad and shallow and, to the north of Metz, the Moselle valley appears to be a continuation of this feature. South of Metz, the Moselle valley is narrow and, particularly to the west, quite steep sided, the slopes being heavily wooded. Whereas the ascent from Metz eastwards to the Plateau Lorrain is relatively gradual, moving west from the city one

(PfalzerwalcP) Saarbriicken rreguemines

'= Limit of region from which evidence was taken.

Fig. 1.2

The 'region3 of Metz

• N 0

10 20 Km

INTRODUCTION

] Land 300-600m Q L a n d 600-900m J above sea l _ J above sea level. leveL

I

| Land 9 0 0 - 1 , 2 0 0 m • Land 1,200m • 3 above sea H above sea

lleveL

•

Fig. 1.3 Physical geography of the region of Metz

Km

INTRODUCTION

is immediately confronted with a steep climb up the Cote de Moselle, wooded and difficult to cross. To the west of the Cote de Moselle is the plateau of the Woevre, the Merovingian Pagus Wabrensis, broken up by meandering water-courses and etangs of varying sizes. Further west, the boundary between the civitas of Metz and those of Toul and Verdun is not particularly well marked geographically. South of the Woevre the border between the civitates of Metz and Toul is clearly marked by the valley of the Rupt de Mad. The boundaries of the civitas of Metz are thus geographically much less pronounced to the south and west than to the north and east. The Merovingian diocese forms more of a unity with the remainder of modern Lorraine than with the regions of Trier, Worms, Speyer or Strasbourg, and this is visible in the material culture of the Merovingian period. The cemeteries of Alsace and especially those of the hilly Triererland are rather different from those of the much lower-lying civitates of Metz, Toul and Verdun.

Historical background5 The end of empire Exactly when the effective power of the Western Empire ceased to be felt in Gallia Belgica is a matter of debate. There may be a case for seeing the beginning of the end as early as 388, with the suppression of the usurper Magnus Maximus. It is difficult to see much imperial activity in the region after that date, though that may result from a gap in our sources, which might have been filled by the lost Histories of Renatus Profuturus Frigeridus and Sulpicius Alexander (LH II.9). The extracts from Sulpicius quoted by Gregory of Tours hint that much of the government of the region had passed to the control of 'over-mighty' barbarian mercenaries, such as Arbogast, and that campaigns on the frontier could take on the form of licensed intertribal feuding. After the invasion of 406-7, during which, Fredegar (Chron. 11.60) claims, Metz was sacked, these suggested trends were exaggerated; barbarians and local 'tyrants' took the centre of the political stage. The extent to which Honorius' generals had restored stability by 420 should not be overestimated. Campaigns against the Franks were still necessary (as in 428) and the Burgundian kingdom, established by treaty in the north-east in 413, had to be destroyed in 436. By the 440s our fragmentary sources paint a picture of the dissolution of northern Gaul into a myriad of small local units, be they led by local Gallo-Romans (either the problematic Bacaudae or apparently legitimate local counts), officially sanctioned barbarian settlers (such as the 5

This account of the Merovingian history of the region need only be brief. For general histories see James 1982 and 1988a; Geary 1988; and, most recently and thoroughly, Wood 1994, which arrived just as the final touches were being put to this book. The essays in Wallace-Hadrill 1962 and 1975 retain their great value. For Austrasia itself, see Ewig 1952 and 1953; Cardot 1987; Picard 1988. Hermann (1963) summarized the state of research in many aspects of the history of the diocese of Metz in the early 1960s.

INTRODUCTION

Alans), or 'invading' barbarians such as Franks, Alamans and Saxons, to whom, though not technically barbarians, we might add Britons. By about 418 the administrative centre of Gaul had moved to Aries. The Trier mint ceased to strike for the emperors before 420. If any coins were minted there between 420 and 450 they were impoverished bronze cAe4' coins (Brenot 1991:172), and possibly a small series of silver imitation solidi which have sometimes been associated with Aetius (Lafaurie 1964), though King (1992) rejects this idea. In Gaul north of the Loire, the period between 420 and 451 can be characterized as political decentralization moderated by armed force. The military and diplomatic skill of the patrician Aetius gave a semblance of unity to the disintegrating edifice. Apart from Salvian's gloomy descriptions of the devastation of Trier and other cities, and of the hardships endured by the rural population, contemporary sources are almost silent about Belgica Prima. This phase was ended by Attila's invasion of Gaul in 451, which led to the famous sack of Metz on Easter Eve that year. The political events leading to the Hunnish defeat at the Catalaunian Fields, and those which followed the battle, culminating in the murders of Aetius and Valentinian III (4545), were decisive for Roman Gaul. Even more than those of Aetius, Aegidius' valiant attempts to maintain an effective imperial presence in Gaul between 454 and 464 can be described, borrowing Kapelle's (1979) excellent phrase, as 'government by punitive expedition'. It is, however, worth remembering that Aegidius was himself an 'illegitimate' leader for the last three years of his life. He was murdered in 464 and the Roman Empire north of the Loire came to an end.

The Frankish settlement In 443 the Franks sacked Trier. Perhaps in the late 450s they were campaigning in the Rhineland, driving Aegidius' troops, or more probably allies, from Cologne and Trier (LHF 8); in 463 they fought alongside Aegidius on the Loire, against the Visigoths (LH 11.18). James (1988a:64-77; 1988b) has argued that the extension of Frankish power into the Paris basin took place as early as the period between c. 460 and c. 481, under Childeric I. James' scenario, making the alleged Rex Romanorum Syagrius into a count of Soissons (1988a:71), accords well with the outline given above. Certainly the Moselle valley had come under Frankish influence by the later fifth century, whether through the medium of Frankish counts such as Arbogast of Trier (Sidonius Apollinaris Ep. IV. 17; Ep. Aust. 23), or of the petty kings of the Rhineland Franks (perhaps Arbogast himself was such a regalis). By 511 the Merovingian kings of the Salian Franks had acquired control of the region, which became part of the Teilreich of Clovis' eldest son Theuderic I. The idea that the Teilreiche resulted from the Merovingians treating their kingdom as partible family inheritance was challenged by Ian Wood (1977), who argued persuasively that the division was brought about by the political circumstances pertaining at Clovis' death. Theuderic had possibly been given a kingdom to rule during his father's lifetime. He was certainly old enough,

INTRODUCTION

as Gregory {LH III.l) makes clear. If the original promulgation of Salic Law is associated with Clovis, and there is no decisive reason to reject this traditional view, it may be that it was first issued in association with the Council of Orleans (511), just as the promulgation of Ripuarian Law is possibly to be associated with the Council of Clichy (626).6 The specification {PLS 47) that Salic Law applied between the Loire and the cCarbonarian Forest' has often (most recently Wood 1994:112) been used to date the code before 507, the battle of Vouille and Clovis' subsequent conquest of the rival Frankish kingdoms in the north-east. However, if we associate the earliest version of Salic Law with the Council of Orleans, the geographical provisions of clause 47 are less surprising. The bishops of what was to be Austrasia are not to be found among its signatories (Pontal 1989:47-58 and map 1); on the other hand, twenty out of thirty-two signatories are bishops of sees between the Loire and the Ardennes. With the former Gothic kingdom south of the Loire covered by the Breviary of Alaric, Clovis was perhaps promulgating a code for the remainder of his kingdom (southern Belgica II, Lugdunensis II and III, and Lugdunensis Senonia: the later Neustria). The lack of religious provisions in the Pactus would be explained by the promulgation, on the same occasion, of a series of canons from the council. If correct, this reading might suggest that the north-east (including the former rival Frankish kingdoms) was already ruled by Theuderic (this would be suggested by the signatories of the Council of Orleans in any case). If Theuderic did become sub-king in Austrasia by 511, Chlotild's political achievements on behalf of her sons, suggested by Wood (1977), were all the more remarkable. Whatever the case, by 511 the settlement of north-eastern Gaul by the Franks was long under way. Tracking the progress of Frankish colonization is extremely difficult (James 1979 remains essential). In the absence of written sources, archaeological cemetery evidence and toponymic data have been used. The archaeology of burials has given rise to two means of distinguishing Germanic settlers from indigenous Gallo-Romans. The first, physical anthropology, need not detain us long. Although, alarmingly, still recommended by the Director of the DAHL in 1988 (Burnouf 1988:114), and still popular in Merovingian archaeology, the scientific foundations of the approach are utterly invalid. The measurement of skulls, with longskulled, dolichocephalic individuals being Germans and round-skulled, brachycephalic persons being Gallo-Romans, still features in Merovingian cemetery reports (Simmer 1988) but has long been shown to be unjustifiable. The use of palaeopathological traits within cemetery-using populations to identify migrants and natives is similarly flawed, even in the rare instances where the results are statistically significant. Like many other aspects of Merovingian archaeology, it suffers from an inability to consider its premises. It is unlikely that there were two separate and physically distinct ethnic (let alone racial) groups within Merovingian north-eastern Gaul. Clearly 6

The most recent opinions of the date of Lex Ribvaria have rejected the early belief in a Carolingian, ninth-century date in favour of one in the reign of Dagobert (Rivers (trans.) 1986:7-11 and refs.). This could be placed in 622, when Dagobert was established as king in Austrasia, or in 631, when he placed his son, Sigibert III, on the throne at Metz. However, the events at Clichy, which established formal frontiers between Austrasia and Neustria and averted the threat of war, would seem to be a more plausible context.

INTRODUCTION

there were 'Roman' and 'German' elements but the two must have been inextricably mixed by the sixth century. In Lorraine, native Gallo-Romans were already adopting Germanic names and language (below, pp. 28-31). Similarly flawed, though even more widespread, is the use of burial customs to sketch out areas of Frankish settlement. Burial with lavish gravegoods has been assumed to be a 'Germanic' custom, whilst burial in a sarcophagus or with few or no grave-goods has been interpreted as Roman (Stein 1974). James (1979) rejected this interpretation, arguing instead that the spread of the custom represented the growth of Frankish lordship rather than the movement of Frankish people. Eight years later this article finally made an impression on archaeology in Lorraine, but was used to argue that there had been no Frankish settlement around Metz (Simmer 1987:392ff; 1988:155). Simmer agreed that Trankish' customs cannot be assumed to indicate Trankish' people but assumed that 'Roman' practices did reveal 'Roman' inhabitants; therefore there was no evidence of Franks in northern Lorraine! Even leaving aside the point that distribution maps of 'Frankish' and 'Roman' cemeteries, as has long been pointed out, reveal sixth- and seventh-century cemeteries respectively rather than different ethnic groups, the whole premise of the argument can be challenged. It can be shown that the introduction, around 400, of lavishly furnished burial into northern Gaul was not the result of the introduction of an incoming people's native funeral rites, but rather the product of social and political instability (Halsall 1992a; forthcoming:chs.5-6). The second general approach to the study of Frankish settlement has been through linguistic evidence, above all toponymic but also via the use of German words in documents. Place-names have been used in Lorraine, as elsewhere, to indicate the ethnicity of the local inhabitants. The suffixes -y, -ey, -ic(h) or -ac(h) are used to map the areas where Romans remained the dominant population element; -ing (or its derivative -ange), -ham and its derivatives, and other Germanic geographical names (-bach or -berg for example) have been used to plot areas of Frankish settlement. Places with the suffixes -ville or -court have usually been assumed to be hybrids or transitional names, resulting from the fusion of the two ethnic groups. James (1979:55-6) sounded warning notes about this approach, and the results of the study of early medieval place-names in England, where linguistic change was more significant, would rarely suggest that so straightforward an interpretation was justified. In Lorraine the conclusions drawn from toponymic evidence have often been simplistic. Records of place-names are uncommon before the eighth century. Even then, it is difficult to draw any direct conclusions. We do not know how long it took to change the dominant language of a region or to alter the usually accepted place-name. As examples we need look no further than Koln and Trier, cities which have lain in the heart of Germanic-speaking regions since the Merovingian period but which, twelve to fifteen centuries on, retain names clearly deriving from their Roman names, Colonia Agrippina and Augusta Treverorum. Even in the early modern period, place-names were still changing back and forth between German and Romance forms. Whether or not a settlement was given an -y or an -ing suffix may well have depended more upon the language 10

INTRODUCTION

spoken by the scribe than upon that of its inhabitants. Though the precise meanings of these suffixes were different, the extent to which they were interchangeable suggests that in practice they meant the same thing. The two languages coexisted happily, so that in 713 Weroald, an influential landowner in the middle Sarre valley (see ch.2) could grant to Wissembourg Abbey 'our villa at Chagambah [modern Waldhambach, Bas Rhin - a German place-name] which is called Distiagus [a Roman place-name, also recorded as Disciacus; Wiss. 192 and 256]'. Place-name construction may be more complex than is often supposed. Roman forms could be 'Germanized5 and vice versa. If the name BELLACUM recorded on some seventh-century coins is to be identified with Bellange (Heidrich 1974:84), we can see an example of this. The process of toponymic change was a long one, as has been noted. By the eighth century, in the western foothills of the Vosges, Germanic words for topographical features, such as bach or berg, and for settlements (e.g. hoba (farm)) or small areas (e.g. marca) were in use, but settlements retained Latin -acum and -villa suffixes (cp. Cilbociaga marca in 714 and Rimenuillare in 798; Wiss. 244 and 211 respectively). Place-names may, as James suggested for burial rites, result more from the ethnicity of the landowner than from that of local inhabitants. The location of place-names ending in -ing, -ange, -y, -ville or -court begs questions about the distinct groupings of these names (fig. 1.4). The

S>

a: -y and derivatives. b: -court. c: -ville and -viller &c. d: -ing. e: -ange.

Fig. 1.4

Distribution of place-name suffixes, by commune, in the part of the civitas of Metz now in France

11

INTRODUCTION

distribution of 'Germanic5 place-names may give an idea of the general area within which Germanic languages came (by whatever date) to predominate, but little more. The idea that -ville and -court names are hybrid is questioned by the fact that these do not lie between zones of 'Roman' and 'German' names but are rather separated from the latter by the former. Figure 1.4 reveals micro-regional place-name 'dialects', local preferences for one of this array of suffixes meaning 'settlement'. The region's place-names reveal only the long-term end results of Frankish settlement. Documentary studies do, however, show that Simmer's idea (1987:395) that 'Mosellan' dialect is pre-Roman and unrelated to Frankish settlement is simply untenable. The transition from Latin to German forms can, as mentioned, be revealed as still under way in the eighth century and later. The exact routes and chronology of Frankish settlement cannot then be examined, but the premise upon which such an image of migration is based may be mistaken. As will be argued (ch.2), early medieval ethnicity was more than a simple matter of the geographical origins of one's ancestors. Ethnicities could be adopted and abandoned in the social confusion surrounding the western Empire's collapse, and important light can be shed upon the means by which the inhabitants of the region of Metz 'became' Frankish. What is certain is that the Frankish and Alamannic migrations contributed significantly to this social confusion.

Metz becomes a 'capital' In the sixth century the Merovingian kings of Austrasia transferred their principal seat from Reims to Metz. The date of this needs fuller consideration. Fredegar (Chron. 111.29) tells us that Metz was a 'capital' from the death of Clovis I (511), but could well have based his statement upon the fact that Metz was the chief Austrasian royal residence in his own day (the seventh century). Gregory states that in the 561 division of the kingdom, Sigibert of Austrasia had his seat at Reims (LH IV.22). By 566, however, he had moved to Metz, where he married Brunechildis (Venantius, Carm. VI. 1). The Vita Goaris Confessoris Rhenani (c. 9) refers to Sigibert's court at Metz. Theudebald I had had some connection with Metz, having called a synod there for the election of the bishop of Clermont in 551 (L//IV.7). Sigibert may have wanted to identify himself with the first Austrasian royal dynasty by choosing as his principal urban residence a royal seat of the last representative of that line. When Bishop Cautinus of Clermont (the successful claimant at the synod of 551) died in 571, Sigibert had his successor, Avitus, brought to Metz and consecrated in the royal presence (LH IV.35), by then well established in the city. In 566 Venantius (Carm. III. 13) referred to Metz as patriae ducis, indicating its supremacy within the realm, and the establishment of the royal seat at Metz must be placed in that decade. From then on, the civitas remained in the Teilreich of Austrasia and the town's importance grew rapidly. A church council was held there in 590 (L//X.19), to try Bishop Egidius of Reims, and we can snatch glimpses of life in the area from Gregory of Tours. An outbreak of dysentery swept the 12

INTRODUCTION

area in 587 (LH IX.13) and Frankish troops from Champagne, on their way to join an invasion of Italy, sacked the city in 590. McNamara, Halborg and Whatley (McNamara et al. (trans.) 1992:138-9) propose an ingenious explanation for this attack, but their introduction to the Vita Glodessindis is factually unreliable. The survival of three comparatively unimportant letters to bishops of Metz in the Epistulae Austrasiacae, otherwise generally a diplomatic collection of letters between the Austrasian and east Roman courts, suggests that the collection originated in royal archives at Metz. In 596, when Childebert II died, his elder son Theudebert II received Austrasia c with his seat at Metz' (Fredegar, Chron. IV. 16). On Theuderic IFs death at Metz (Vita Columbani 1.29; Fredegar, Chron. IV.39), of dysentery according to Fredegar, Brunechildis was also at the 'capital' with his sons (ibid. IV.40). In 631 Dagobert collected his army for a Thuringian campaign at Metz (ibid. IV.74) and in the following year the importance of the city was underlined when Dagobert placed his young son Sigibert on the throne of Austrasia, with 'Metz as his headquarters' (ibid. IV.75). On Dagobert's death, Sigibert's share of the royal treasure was presented to him and inventoried at Metz (ibid. IV.85). Wettinus' Vita Galli (c. 22) states that Sigibert was married there, and built a nunnery in Metz. Later tradition (Sigibert of Gembloux's mid-eleventh-century Vita S. Sigiberti Regis) claims that he was buried near Metz at the church of St Martin, the modern St-Martin-les-Metz. Sigibert of Gembloux would have us believe that the king built this church but, as Gauthier (1986:52) points out, the Vita Romarici also refers to a church of St Martin beyond the walls in 612. The church must have been situated outside the town since, after praying there, Romaric returned to the city (regressus ad civitatem; VRAH 3).

The later Merovingian period The period between c. 566 and c. 651 seems to have been the cgolden age' of Metz's political importance. This impression may, however, result from the unevenness of our evidence. We have hardly any evidence for the narrative history of the town between the death of Sigibert III and the end of the Merovingian dynasty. In 651, after Sigibert Ill's death (Gerberding 1987: 50ff for the date), Grimoald, the Austrasian mayor of the palace, put his son, Childebert 'the Adopted' on the throne. As Gerberding (1987:ch.4 and refs.) shows, Grimoald held his position for perhaps as long as six years before being ousted by his opponents and the Neustrian Clovis II (Gauthier 1980:265 reached a similar conclusion). In 675, after a disastrous attempt to rule with 'absolute powers' in the style of his sixth-century predecessors, the Austrasian king, Childeric II, was assassinated. Whether or not messengers were dispatched to Dagobert II in Ireland immediately after the murder, the Austrasians managed without a king for about six months between autumn 675 and April 676, 'government' passing (probably briefly) to the mayor, Wulfoald, and then to the Arnulfings Pippin 'IF and Martin. The region and its magnate families certainly played an important part in seventh-century history. The Arnulfings held estates in the region 13

INTRODUCTION

(Hlawitschka 1962). One of Grimoald's supporters, the deacon AdalgiselGrimo, held estates which clustered between Metz and Verdun (Levison 1948b). Possible opponents of Grimoald, and his affinal relatives the Arnulfings, may also have held estates in the civitas of Metz. Fredegar (Chron. IV.86; IV.88) tells how one of Grimoald's chief rivals for mayoral power between 638 and 641 was 'Otto the son of Uro the domesticus'; Grimoald had him killed by Leuthar the duke of the Alamans (who in turn may have been related to the family which founded and endowed St-Pierre-aux-Nonnains in Metz; Gauthier 1980:295-7; below ch.7). A charter of 669-70 (Pertz, Royal 29) shows Childeric II reorganizing the forests of the Pippinid monastery of StavelotMalmedy. The men detailed to mark out new boundaries were the domesticus Hodo and Duke Gundoin (Gerberding 1987:123-4). Hodo may well have been related to the Otto killed on Grimoald's orders; the Annales Mettenses Priores (s.a.688) allege that a Gundoin killed the Arnulfing Ansegisel. The families of a Gundoin and of a Count Audoin (in some charters called Otto) are shown by the Wissembourg cartulary to have held extensive lands in the Seille and Sarre valleys (below ch.2; ch.5). It does not seem at all unreasonable to identify the two royal officials of 669-70 with the fathers (both deceased by 699-700) of the families who gave land to Wissembourg (Ewig 1953:137 n. 152; Ebling 1974:64-6, 167-8). As we shall see (ch.2), these two families seem to have been, if not related, then closely associated. Probably allied, and related by marriage, to these families was that of Count Wulfoald, a probable descendant of the Mayor Wulfoald who died shortly after Childeric II. This family possessed a significant base in the south-west and southern edge of the region of Metz, and in 707/8 Wulfoald constructed a castrum at St-Mihiel against the Arnulfings. This was later confiscated and made into a monastery (Gauthier 1980:397). That these families still controlled an important territorial base in the early eighth century is, if it were needed, further evidence that the rise of the Arnulfings was by no means assured before Charles MartePs supremacy. This idea is further supported in a local context by Gerberding's (1987:139 n. 39) argument that the region of Metz was one of the last areas of Austrasia to fall into line behind Charles Martel, not doing so until after the battle of Vinchy. The stress of the second decade of the eighth century and the Arnulfings5 eventual victory, are perhaps attested by the rash of charters of Charles MartePs opponents to Wissembourg Abbey. The sons of Gundoin and Weroald son of Audoin donated almost all of their estates to the abbey, whilst their associate, Chrodoin, entered into a possibly protective relation of dependence upon it (cp. Wiss. 195). The connection with Metz, so important to Carolingian propaganda (Oexle 1967), probably stems from the town's continuing political importance, rather than any genuine claim to be the Arnulfings' ancestral and political heartland. The Arnulfings' questionable standing in the region can be further examined through a study of Messin episcopal politics.

Ecclesiastical history The region's ecclesiastical history is thoroughly covered by the work of Nancy Gauthier (1980). Werner (1979:155 n. 22) draws attention to one 14

INTRODUCTION

bishop apparently omitted from the episcopal lists, an Abbot Gundulf, whom Venantius (Vita Radegundis 13) says became bishop of Metz. Using the problematic methodology of basing hypothetical relationships upon the supposed existence of 'leading names' within families, some tenuous possibilities emerge. Bishop Gundulf may have been connected to Gregory of Tours' maternal great-uncle, Duke Gundulf (almost certainly Gundulf the subregulus, to whom St Arnulf was entrusted for his education; VA 3). He would then be associated with a family which prospered in the service of the Austrasian royal house. The next two prelates were possibly members of the royal family. Paul the Deacon (Lib.Ep.Mett.:264) alleges that Bishop Agiulf was descended from one of Clovis I's daughters. Elziere (1993: n. 45), drawing on dubious Carolingian sources, creates a link with Theudebert I's wife, Deuteria. Agiulf's successor was his nephew (or grandson), Arnoald. After him, Papolus could (though the name is common) have been the Count Papolus whom Venantius (Carm. VI.8) mentions in the royal entourage on the Moselle, or at least one of his relatives. It is possible that the see based upon the royal capital was reserved for faithful servants of the king or members of their families. In 616, Berthechramn of Le Mans bequeathed certain unidentifiable estates to Arnulf of Metz. These are said to have been held and ravaged 'with iniquity' by cEgulf and Arnoald' and given to Berthechramn by Chlothar II (TBLM:129; Gauthier 1980:376-8). These two are without doubt Agiulf and Arnoald, the bishops of Metz. The bishop of Le Mans was being ordered by Chlothar II to return confiscated lands to the see of Metz. The reigns of Bishops Arnoald and Papolus have to be compressed into a short period between 601, when Agiulf is attested in a letter of Gregory the Great to Brunechildis, and Arnulf's appointment in 614 (Gauthier 1980:377 for the date). But this was a time of great political turmoil. Gauthier (ibid.) suggests that the see of Metz was vacant from the time of Chlothar II's invasion of Austrasia (613) until after the Council of Paris (614). Agiulf and Arnoald's 'iniquitous' seizure of land demonstrates their active partisanship of Brunichildis during the civil wars, and, as outlined, Papolus may also have supported Brunechildis. Perhaps he was driven out, and possibly killed, when Chlothar II invaded Austrasia and wiped out Theuderic II's family in 613. We can then see Arnulf, one of the ringleaders of the faction which invited Chlothar into Austrasia, eventually being rewarded with the diocese of Metz. In expelling Brunechildis' party, Chlothar returned certain 'ravaged' estates to Berthechramn of Le Mans, another distant relative. Arnulf's appointment continued the tradition of having a trusted supporter as bishop of Metz, and he was soon appointed to be one of the young Dagobert I's nutritores. Arnulf may not have been popular in Metz. His departure from the city for a hermitage in the Vosges may not quite be straightforward evidence of his sanctity (Prinz 1965:141, 144 suggests political disgrace; Cracco Ruggini 1992:131 disagrees). His vita stresses repeated strife with local secular magnates and indeed with the kings, and even records that one local magnate actively opposed the translation of Arnulf's remains back to Metz.7 Parisse 7

It has recently been suggested that the Vita Arnulfi, rather than being written by a contemporary of the saint, as the text claims, was instead written in the early eighth century and associated with the Carolingians (Wood 1988:371). Given that Pippin 'IF and his family began their association with the

15

INTRODUCTION

(1987:41) suggests that Arnulf's burial at the church of the Holy Apostles, away from previous, and subsequent, bishops (at St-Simphorian) could be taken as evidence of some friction. We might interpret the Vita Arnulfi in line with Fouracre's (1990) model, as a Life composed, from a safe chronological distance, to smooth over local difficulties by rewriting local history and an unpopular bishop's place therein. Hence the stress laid by the vita's author on the king's and the citizens' wailings and remonstrations on Arnulf's departure from his see. The cult of Arnulf developed only slowly (Parisse 1987:41). The church where he was buried was not usually known as St Arnulf's until the Carolingian period. Arnulf's son Chlodulf was buried there in the late seventh century, probably at about the same time as the earliest Arnulfing privilege for the church (Pertz, Mayoral 2). The family's first major use of the shrine came in the crisis period of the first decades of the eighth century when Drogo son of Pippin CIF was buried there. The tradition of political appointments to the see continued. Arnulf's successor, Goeric-Abbo, was one of the mid-seventh-century churchmen who served their apprenticeship at the court of Dagobert I, of whom Eligius of Noyon and Desiderius of Cahors are perhaps the most famous. Abbo was one of the three recipients of the testament of Dagobert I (GDRF 39). The next bishop, Godo, may have been an advisor of Sigibert III (cp. Gauthier 1980:390). Gauthier doubts the identification, seeing the Godo in question rather as the contemporary bishop of Verdun. However, it is plausible that a bishop of Metz advised a king with such strong connections with the city as Sigibert III. Chlodulf, Arnulf's son, later became bishop of Metz, despite a reputation for impiety in his youth (Lib.Ep.Mett.:264-5), something else suggesting that the stories circulating in Metz about the Arnulfings were not uniformly favourable. He too was of the circle of Desiderius of Cahors and Goeric-Abbo (Gauthier 1980:391) and was appointed some time during the 650s,8 thus, again following Gerberding (1987:ch.4), probably during the period when his cousin, Grimoald, had taken power in Austrasia. If his appointment was associated with Grimoald, he nevertheless seems to have survived his cousin's downfall. Chlodulf's successor Abbo II may have been in attendance on Clovis III, witnessing three of his charters (Pardessus 355, 431, 451), but the Bishop Abbo in question is more likely to have been the contemporary bishop of Troyes (Gauthier 1980:394). Nevertheless, the next bishop, Aptatus, may well have been one of Clovis Ill's referendaries ( b i d )

At the very end of the period, Bishop Chrodegang, later to play a prominent part in the usurpation of Pippin the Short, had earlier been referendary to Pippin's father, Charles Martel (Lib.Ep.Mett.:267). Chrodegang's predecessor Sigibald had been involved in the land transaction in 707/8 (Gauthier 1980:397) which allowed the anti-Arnulfing Count Wulfoald to create a fortified outpost at St-Mihiel. Metz's position in the heartland of Austrasian

8

cult of Arnulf in the late seventh century (see below), it is odd that the vita makes no mention of the names or descendants of Arnulf's two sons. This lacuna makes more sense if the text was indeed written in the 660s to 680s5 when Arnulf's family was in the political wilderness, and its association with the mayor Grimoald not something to be stressed. See also ch.9 n. 1, below. According to the Liber de virtutibus sanctae Geretrudis (c. 2) he was bishop by 17 March 659.

16

INTRODUCTION

opposition to the Arnulfings, and its tardiness in joining Charles Martel, have already been noted. Sigibald lived long, and Paul the Deacon (Lib.Ep.Mett.:267) cannot conceal that he was much venerated in the area. All this must have been somewhat inconvenient for Charles Martel, who narrowly predeceased Sigibald. It may have been with some relief that Carloman placed his father's trusted follower, Chrodegang, on the episcopal throne. Chrodegang's subsequent refounding of Sigibald's cell of St-Paul, as St-Nabor (now St-Avoid; ibid.:268), takes on added meaning.

Administration It has been suggested that the civitas framework did not survive in northeastern Gaul (James (trans.) 1991:106 n. 4). Though cited as such, Gregory of Tours' description of Nicetius as bishop of the Treveri rather than as bishop of Trier does not seem to be particularly conclusive evidence; the use of the tribal name rather than that of its city-capital might indeed prove that, on the contrary, the old administrative system founded on pre-Roman tribal units was still very much in force. The problem is perhaps best examined chronologically. In the late sixth and early seventh centuries, both Metz (Fredegar, Chron. IV.38) and Trier (L//VIII.15) are recorded as having dependent territorial which surely equate at least vaguely to the old civitates. More problematic is the question of whether or not a single count of the civitas existed in Metz. It has been assumed that one did, but it must be conceded that the assumption is founded upon no explicit testimony. Documentary data from Metz, most notably and surprisingly the Vita Arnulfi, make no mention of a count of Metz. It might be supposed that the institution did not survive in the north-east. Instead, Gregory of Tours' reference to a duke at Metz (LH VIII.21) could lead us to suppose that the old civitas was incorporated in a wider area, possibly equating to modern Lorraine, governed by a duke. However, Gogo's letter to Petrus (Ep.Aust. 22) refers to a 'leader of the citizens'. This must be some kind of secular official, if not the count then perhaps a defensor civitatis. If the latter office had survived at Metz, it seems odd that that of comes had not. A count existed at Trier in the 470s (Sidonius, Carm. IV. 17; Ep.Aust. 23) but his precise role is obscure (Barnwell 1992:35-6). There were, nevertheless, counts north of the Loire at Rouen, Meaux and Cambrai in the 580s (L//VI.31, VIII.18; VG:283). Venantius Fortunatus mentions two counts in Austrasia (Carm. VI.8, VII. 16). Frankish law refers on several occasions to counts or grafiones. There were certainly counts, evidently broadly equivalent to southern comites civitatum in the heart of the Austrasian kingdom (Murray 1986). Nevertheless, counts of civitates in the north-east are rarely mentioned as such in our sources, dukes certainly being more prominent. If there was a weakening of the civitas unit, this would support the hypothesis of social and administrative confusion and discontinuity between the late Roman and Merovingian periods in north-eastern Gaul. In the seventh century, we find that the civitas of Metz had divided into pagi and that, by this period, these were more important than the old civitas 17

INTRODUCTION

unit. From written sources, we know of the Pagus Salinensis (the Seille valley), the Pagus Sarroensis (the Sarre valley), the Pagus Mosellensis (the Moselle valley, north of Metz) and the Pagus Scarponensis (the area north and west of Scarponne). In later sources, a Pagus Mettensis around Metz itself is mentioned, but in at least one case this seems to be an error for Pagus Mosellensis. The Pagus Aquilensis around the river Eichel in the far east of the region does not seem to have become established as a formal administrative unit during the Merovingian period. Similarly the Bliesgau (the Blies valley in the modern Saarland) is not attested in Merovingian documents. A Count Willibert is mentioned in the Pagus Sarroensis in the early eighth century (Wiss. 192). The pagi were further divided into hundreds, administered by centenarii; Chardoin the centenarius, apparently one of Willibert's officials, is a frequent witness to early eighth-century Wissembourg charters. Some of these pagi can be shown to spread across later diocesan boundaries, perhaps supporting the notion that the civitas structure had broken down in the north, at least by the seventh century. It seems that the old civitas, by this date, survived only as an ecclesiastical unit.

18

Parti Social organization

Introduction

Part I concerns the 'ordering5 of communities, its basis and the mutual human relationships which go to make it up. This introduction aims to present a way of seeing early medieval society, and a provisional vocabulary for use in this study. This does not claim to be a general, fully formulated theory in the manner of Runciman's (1983; 1989) of social evolution, or Mann's (1986) of power. It simply represents the stage which my heuristic system, my 'mental map3 of the workings of society (which, like anyone else's, is constantly developing), had reached when this study was conducted; it is a statement of the framework within which the latter was carried out. Making this framework explicit should nevertheless make it easier to engage critically with the ideas expressed here. Social relations are governed by the interplay of identities. An individual may adopt one or more of a number of different identities in relationships with any other individual, choosing, where appropriate, to accentuate similar or different identities depending upon the situation, selectively emphasizing or playing down that which either unites or divides (cp. Bourdieu 1977:41). Identities may relate to one's position in a social hierarchy of ranks or classes (see below), one's family, one's age, or one's gender, religion or ethnicity. The nature of any of these identities may be ascribed by other people, or created or chosen by those sharing that identity. When differentiation is stressed, identities raise and play upon barriers between people. For ease of description, I have divided such barriers into vertical, horizontal and diagonal. Horizontal barriers are those between people of different ranks or classes (see below); vertical barriers are those based upon gender differences cutting down through the social hierarchy; age difference can be perceived as a diagonal barrier; kinship identities form another kind of diagonal barrier, between different families. Ethnicity and religion might be seen as similarly 'diagonal' but I have chosen to treat ethnic identities as a separate category. The barriers created by the adoption of an identity should not be viewed as constant. At any one place or time, the nature of certain barriers may be affected by the others. Gender or kin-group identity might differ greatly on one side or other of specified horizontal barriers. The barriers between particular identities may be negated by stressing another common identity. Thus, for example, gender difference may be played down in favour of common rank or class (see below), age or kin-group; difference in rank or class may be modified by stressing common age, gender, kin-group or ethnic identity. Some social groupings are created to modify the effect of certain 21

SOCIAL ORGANIZATION

barriers, usually by raising others, thus creating unity amongst one group by excluding others. The creation of lordship, patronage, dependence or protection networks is one such example. Fictive kinship (see below, ch.2) helps to create other groupings. The nature and strength of the barriers define the space within which an individual can act to further his or her aims, and indeed to alter the barriers' strength and nature. This relationship between the possibilities of action and the nature and knowledge of the barriers limiting it equates to some extent to Bourdieu's (1977) notion of the habitus. Social action is then a question of playing within (but also with) the rules which define these barriers. A person's identity is thus formed by his or her position on the 'map' of society, formed by the network of barriers. This is, however, in a state of constant renegotiation. Especially when the nature of a barrier is not institutionalized, it is formed by the accumulation of all known previous interactions between particular different identities; in other words, where there is no institutionalization it is governed purely by the habitus. It can never be exactly reproduced and all social interaction affects the nature of the barriers between individuals; births, marriages, deaths cause major redefinitions of the 'social map'. Hence we cannot accept the imagery discussed in chapter 1, wherein a single change takes effect over three centuries. As Shanks and Tilley say (1987:212), 'it is stability rather than social change that needs explaining'.

Vocabulary The barriers between individuals, raised by stressing a particular identity, and the extent to which they can be overcome in social strategies, are based upon many factors. Social actors can play with any number of these. A vocabulary must now be proposed, defining certain of these factors, and other concepts of use in discussing early medieval social history. Status: the standing, based upon mutually-accepted norms, of any individual in any sphere, with relationship to any other individual. Status can only correctly be used in this broad sense; the frequent use, by some archaeologists and historians (cp. Filmer-Sankey 1992:49; Amory 1993:14), of status to mean rank is erroneous and misleading. Value and worth: certain types of people were buried with lavish gravegoods, or were accorded high legal standing, but the reasons for this made it misleading simply to talk of them as having high status or prestige (q.v.) within the community (below, chs. 2-4 and 8-9). It was important to distinguish between 'passive' and 'active' forms of high status, for which, respectively, the terms value and worth were used. By the former is meant status ascribed by others according to external, 'institutionalized' definitions of an individual's social role, and its value to other community members. Value is often ascribed to those whose actions can be profitably controlled

22

INTRODUCTION

by others. 'Active' status, worth, means status acquired by an individual from his or her actions or acquisition of skills. Those with worth control their own actions and are free to make decisions. Both forms of status may bring prestige (q.v.) but it is important to note the differing foundations of this high status. In particular, once someone ceases to occupy a role defined as of value, they lose the high status it brought, unless they used it actively to create worth for themselves. Power: the ability to achieve aims, vis-a-vis other individuals. Prestige: positive status, giving possible power to the individual. Privilege: the right to power over certain, often defined, individuals. Prestige may lead people to expect others to conform to their wishes; those with privilege have an accepted right to expect this. Mention of people with defined power over other people leads to discussion of forms of status based upon membership of certain kinds of social grouping. Rank: status based upon a position in a legally-recognized hierarchy. Slave, semi-free and free are examples of ranks. Class: class will be used here to describe groups of people who occupy a recognized position in society but not one which is necessarily legally recognized. This position may be one within a loose hierarchy; the de facto powerful, the rich and influential, may constitute a class. On the other hand a class may relate to other dimensions of social organization; various kinds of craftsmen, or religious groups, may form classes. It is important here to distinguish 'hierarchical class', that is informal ranking, from other forms of class. Weber's (1982:69) definition of class is useful (the italics are my own): 'Class situation' means the typical probability of 1. procuring goods 2. gaining a position in life and 3. finding inner satisfaction, a probability which derives from the relative control of goods and skills and from their income-producing uses within a given economic order. 'Class' means all persons in the same class situation.

The Weberian concept of 'status group3 has been ignored. Such groups are discussed as particular forms of class. It is also important to distinguish between cclass in itself and 'class for itself, that is to say between cases where groups of people simply find themselves, de facto, in a similar position on the 'social map', within a similar 'class situation', and those wherein people have a clear consciousness of a class identity and of a common strategy in maintaining or improving their class situation. Here, when considering social hierarchies, there will frequently be several 'classes' within each 'rank'.

23

SOCIAL ORGANIZATION

A working distinction in the terminology for ruling classes must also be proposed: Aristocracy: 'rule by the best', that is to say the most powerful, people. Such power has no necessary legal or constitutional base and may be acquired and lost from one generation to another, or even within one lifetime. Nobility: rule by, in our terms, a rank, acquired by birth, and thus transmitted from one generation to another. The privilege of being a noble can only be lost in exceptional circumstances. The aim of most aristocracies is to turn themselves into a nobility. Though the two are not, it must be stressed, mutually exclusive, these types of ruling group can be distinguished by a number of oppositions: Aristocracy : Nobility Class : Rank Prestige : Privilege Status may also be defined by one's position, gender or age. Such status may be formally institutionalized or left informal. At the other end of the social hierarchy, the following definition is adopted: Peasant: my definition of peasant is based upon that of Shanin (1987:25): peasants are 'small agricultural producers, who, with the help of simple equipment and the labour of their families produce mostly for their own consumption, direct or indirect, and for the fulfilment of obligations to the holders of political and economic power [my italics]' (Shanin 1987:3). Whether or not the above terms and their usage find general acceptance, it is to be hoped that the use of more explicitly defined terms will result in greater clarity and precision. They are, in Hodder's (1986:18) phrase, 'provisional but precise'.

Methodology The foregoing discussion of the barriers within society and provisional vocabulary represents, it must be stressed again, an analytical tool, a 'way of seeing', and not an end in itself. Our concerns must be: 1. To examine the relative strengths of the barriers raised by the adoption of identities, and the weight placed upon them by contemporaries. 2. To consider if, how and why these relative strengths and weights changed through time. 3. To consider the ways in which they changed in different sections of society. Were, for example, gender identities stronger or weaker, or built upon different bases, among the aristocracy? If so, why? 4. To look at the ways in which barriers were crossed. 5. The implication of (4) is that if we can identify different groups 24

INTRODUCTION

within society and place them within the schema outlined above, we must examine the relationships between them and both describe and explain any temporal changes in these. 6. In particular we must look at how powerful groups established and maintained their position. This, as outlined above, means examining their relationships with those whom they dominated, as well as with those who, theoretically, dominated them: the state, the king. This agenda incorporates an essentially two-stage analysis. In the first (parts 1 and 2), the results of the study of different, separate forms of data are described (this stage, it must be made clear, is analytical in itself); in the second (part 3), a composite description of Merovingian society and the ways in which it worked and changed (in so far as changes are perceived by us) is given, and this is then explained. The problems of early medieval data rarely if ever allow straightforward value-neutral 'reportage5 (cp. Runciman 1983:ch.2) at any but the most banal level. What I have used as 'reports' to build a descriptive picture (whence, in turn, explanation proceeds) will not be accepted as such, in every respect, by every other observer of Merovingian society. I have often had to allow the existence of equally plausible alternative descriptive conclusions, and have tried to keep these alternatives open for as long as possible before deciding on one over another, usually as a result of the higher-level comparison of data from different disciplines mentioned in chapter 1. In the second phase, explanation has sometimes preceded description; for example, it has been necessary to explain why certain burial customs were adopted before descriptive conclusions about the nature of society could be proposed (ch.8). In chapter 2 evidence from written data is analysed to make up a provisional descriptive model of society at various stages during the Merovingian period. The data have, wherever possible, been restricted to those from the region of Metz. In the earlier part of the period, however, this has not always been possible. Laws, chronicle references to the north of Gaul and so on have thus been employed to create a general 'northern Gaulish' back-drop against which the patchier specifically local data can be set. Evidence particular to other regions of Gaul, and from periods other than those under consideration, have not been incorporated into the descriptive picture. Where such evidence has been used I hope it is made clear that it is used purely as comparanda. It is nowhere implied that one can usefully 'fill in the gaps' in the description of fifth-, sixth-, seventh- and eighth-century north-eastern Gallic society in the region of Metz by straightforward incorporation of information from other times and places. Analogies with apparently similar experiences in other regions during different eras have, where I felt they were interesting or instructive, been traced; no one-to-one correlation is implicit and no general, consistent attempt is made to compare and contrast society and social change in Merovingian northern Lorraine with that elsewhere in the early medieval world. I leave such comparison to other researchers.

25

2 Social organization: descriptive analysis of the documentary evidence

The ethnic division1 The kingdoms of the post-Roman West derive much of their individuality from their division into two societies, 'Roman' and 'barbarian'. It is, however, a mistake to assume that ethnic distinction between 'Frank' and 'Roman' represented an immutable binary polarity. Ethnicity is, above all, an identity. This section will examine what purposes the adoption of ethnic identity served in Merovingian north-east Gaul. Sixth-century evidence for ethnic distinction comes from only a handful of sources: law, name-giving and anecdotal references in narrative and hagiographic works. The legal material is the most useful.2 As a statement of social theory, Salic Law may always have been, for want of a better word, a 'Neustrian' code, but even if this was the case, Austrasian Law probably derived either from Salic Law itself or from fundamentally analogous sources. Examination of the legal depiction of the ethnic division must begin with a consideration of the nature of the law itself. Did it set out norms only for a single people, as an ethnic or personal law-code, or was it territorial? If, as has usually been assumed, it was the former, specifically dealing with Salian Franks, our enquiry runs into serious difficulty. We must conclude that its compilers were not concerned with the norms of non-Frankish north Gallic society. The case for the 'personal' nature of Salic Law has, however, been weakened over recent decades by the realization that it may have drawn greatly upon Roman vulgar law, rather than ancient Frankish tradition imported intact from Germany (Wallace-Hadrill 1957; Collins 1983:24-31). Salic Law's promulgation thus has to be more than a simple record of the customs of a particular 'tribe' now settled on new lands (cp. Wormald 1977). Whilst acknowledging the many and dangerous pitfalls of using the Pactus Legis Salicae for any detailed semantic analysis (Murray 1983:ch.8), this problem may be elucidated by examining the use of ethnic labels within the legal sources. 1 2

Amory 1993 is now an excellent introduction to this topic, with a very useful bibliography. Throughout this work the laws have been treated as social theoretical 'treatises', rather than as necessarily effective legislation. Wickham 1992 (esp. p.233) makes excellent points about this and the positive ways in which Frankish law-codes may nevertheless be used. The analysis of cemeteries below will, furthermore, show that the ways in which the law-makers theorized about social order were interestingly similar to those in which local communities did so.

26

SOCIAL ORGANIZATION: DOCUMENTARY EVIDENCE

Pactus Legis Salicae and the edicts appended to it legislate for Romans as well as Franks. That the clauses mentioning Romans relate (with one important exception) to crimes against the person led to the assumption that Romans only impinged upon 'Frankish' law where they interacted violently with Franks. Otherwise they lived by Roman law. This superficially reasonable deduction is questioned by two things. First, the exception mentioned (PLS 16.v) is a law about Romans burning the property of other Romans. Here, at least, Frankish law-makers felt competent to set out legal procedures for crimes not involving Franks at all. The second objection concerns the fact that, before Childebert II's edicts, the term francus (with two exceptions) only occurs in the same limited contexts as romanus. One might as easily deduce that Franks only appeared in Salic Law where they interacted violently with Romans. Ethnic personality was important in crimes against the person and personal freedom (murder, abduction, confinement, enslavement). There are no laws about Romans stealing from Franks or vice versa, about mixed marriages between Franks and Romans, and so on. The implication seems to be that in most cases everyone was subject to the same law. Most titles, including almost all those on theft, marriage, injury and inheritance, are entirely impersonal, beginning cif anyone' (si quis). Exceptions to the general impersonality of the laws concern those specifying crimes committed by or against ingenui ('freemen'). Interestingly, the bulk of laws mentioning freemen or freewomen relates to the same range of crimes as those mentioning 'Franks' or 'Romans'. Before the 590s the term occurs primarily in relation to violent crime (murder, robbery, castration) or crimes against personal freedom (abduction, confinement, enslavement, obstruction). Elsewhere, not surprisingly, the word usually occurs in laws mentioning the half-free and unfree (setting down different penalties for free and unfree, penalizing crimes by the free on the unfree and vice versa, forbidding marital or criminal unions between free and unfree, etc.). Again this legal personality is mainly important in crimes against the person or in the definition of degrees of freedom. On the basis of this unsurprising conclusion one might deduce that the word meant any free person, Frank or Roman. This would definitely suggest that Salic Law was territorial. However, in other cases (above all PLS 39.v and 42.iv), it is made clear that the term 'freeman' does not include Romans. An equation of 'freeman' with 'Frank' is suggested by PLS 4O.v, which refers to 'the sort of crime for which a freeman or Frank is held liable for 8,000 denarii', and indirectly by the laws on castration, the only laws on injury to specify a legal personality for the victim. In PLS 29.xvii the victim is a 'freeman' and in PLS 71 it is either a Frank or an antrustio (royal bodyguard, by which, as we shall see, is usually meant a Frank). Other indications perhaps point to the identification of 'freemen' and 'freewomen' with Franks, but for now the important conclusion is that the restricted and specific use of ingenuus implies that the bulk of the law applied equally to all other non-slave members of the population, surely including the half-free and Romans (who are legally equated in PLS 42.iv). Again we reach the conclusion that the theory of Salic Law applied territorially, to all the inhabitants of the Merovingian realms north of the Loire. This is supported by the specification of geographical limits to 27

SOCIAL ORGANIZATION

the area within which the law was applicable (PLS 47), and by references to the sale of slaves abroad. Similar arguments have been put forward for the territorial nature of 'barbarian' law (Collins 1983:24-31, contra King 1980, for the Visigoths; Amory 1993 for the Burgundians). This conclusion is important, since it removes the necessity to assume a lost corpus of Roman law for the non-Frankish population of north Gaul. Salic Law represents sixth-century legal theory for all the region's inhabitants, from which we can study legal concepts of ethnicity and the social roles of 'Romans' and 'Franks'. The 'sixty-five-title text' of the Pactus is fairly consistent in its treatment of ethnicity. This may be because it indeed originated as a single treatise, rather than, like Burgundian Law (Amory 1993), being built up from a number of edicts of different date. The impression is given that the Franks already formed about half of the population, but we must consider what Frankish identity implied. Did it mean that a person was a settler from former Free Germany, or a descendant of such settlers, or did it mean something looser? The law clearly gave Franks legal privilege. Franks were valued at twice the wergild of equivalent members of the Roman population. The adoption of Frankish ethnic identity thus raised a kind of horizontal barrier, and could be used to obtain differential, and preferential, legal treatment. If the equation of ingenuus and francus is accepted, some indications point to a military role for the Franks. Two laws (PLS 26.i and 63) mention 'freemen' in the army. To be 'Frankish' may then have meant liability for military service. This would not be surprising. The Frankish kings who subjugated northern Gaul in the fifth century drew their military strength partly from an old-style tribal levy of freemen, and partly from the bands of warriors whom they attracted, and who subsequently adopted Frankish ethnicity. Those Franks who migrated across the Rhine and settled in northern Gaul had always been subject to military service; those who adopted Frankish ethnicity as a result of fighting on the winning Frankish side doubtless regarded themselves as subject to the same dues. What of the Romans? PLS 41 identifies two main kinds of free Roman: the landholder (in turn subdivided into members of the royal court and other landholders), and the tributarius, the rent- (or tax-) payer. This suggests that 'Roman' lands were held, as before, in return for rent, the landlord keeping a portion and passing on the remainder. Perhaps the distinction between Roman and Frank in northern Gaul was that between those who paid tax and those who did military service.3 This identification, which, it must be admitted, rests upon the equation of ingenui and Franks, may be supported by the fact that the highest class of Roman, the landholder at the royal court, is specified to be a member of the convivia regis (the 'king's dining friends') rather than the trustis regis (the royal retinue) like the highest grade of 'freeman'. The distinction was probably always a little blurred. An edict appended to the Pactus adds a further class, the Roman soldier, but this itself implies that Romans did not usually fight. 3

In Lombard Italy this kind of effective distinction between 'Roman' tax-payer and 'barbarian* soldier is more clearly suggested.

28

SOCIAL ORGANIZATION: DOCUMENTARY EVIDENCE

Ethnic identity may thus have highlighted a functional difference between Romans and Franks. The ethnic barrier between them does not seem to have been particularly strong. Salic Law, as mentioned, does not regulate mixed marriages (unlike seventh-century Frankish law). In spite of their higher wergilds there is no indication that the Franks were lords over the Romans. The laws give an impression of parallel societies rather than of an immigrant ethnic group dominating a subject indigenous population. Nor do they give any impression that sixth-century Merovingian kings had much interest in preserving the rigidity of the ethnic barrier. One might have expected that the kings would have tried to maintain the amount of taxable land by restricting the Romans' ability to avoid taxation by 'becoming' Frankish. That they did not implies that Frankish tax exemption was only partial. The personal names which occur in sixth-century sources, to whom we might add the bishops recorded in later Messin episcopal lists, shed important light upon ethnicity. The bishops have Graeco-Roman names almost uniformly until the accession of Agiulf (some time between 580 and 601; Parisse 1990:11). The two possible exceptions are Gunsolinus, who must have governed the see in the late fifth or very early sixth century, and the mysterious mid-late sixth-century bishop Gundulf (see ch.l). However, the predominance of Roman names in the church should not surprise us. Many of these had Christian associations with the apostles, evangelists or early martyrs. The church may have recruited its clergy and especially its bishops from 'Romans', or people destined for, or entering, the church may have been given, or adopted, Roman names. Another source of names is the handful of letters from the region, particularly that written by Gogo to Bishop Petrus (c. 580; Ep.Aust. 22). This is an interesting snap-shot of Messin notables. Of the seven named ecclesiastics, four have Roman names while three have German. The one possible secular official, Theodemund ('leader of the citizens'), has a German name, which may be significant. The people from the region named by Gregory are mainly secular officials and have German names. One, however, Gundulf,4 whilst having a German name, is revealed to be none other than Gregory of Tours' maternal great-uncle. Our sixth-century sources reveal more or less equal numbers of Roman and Frankish names, but there seems to be a functional difference: churchmen with Roman names, secular aristocrats with Frankish. However, the trend appears to be towards Frankish names. Gundulf's case makes this abundantly clear. Even to a member of what was, as his nephew Gregory was so fond of pointing out, one of Gaul's leading senatorial families, it was politically advantageous to have a Frankish name. The impression given by the main body of sixth-century legislation and by name-giving customs can be compared with the narrative and hagiographic sources' anecdotal detail. Almost all such sources come from further south which is unfortunate, but in some respects these writings acquire the characteristics of third-party observation and derive some interest therefrom. The use of the word Frank in Gregory's works is instructive; it is a word which appears surprisingly rarely. When used it is commonly as half of the 4

His association with the region comes mainly from his identification with the Gundulf who brought up the young St Arnulf, but also from Gregory's statement that he came from Childebert II's court.

29

SOCIAL ORGANIZATION

compound regnum francorum or rex francorum. There are notable changes in the word's usage in different parts of the Histories. In Book II, not surprisingly, the Franks are simply a barbarian people with their own leaders. When relating the history of the late fifth and sixth centuries Gregory used the word in military contexts (LH 11.18-19) and contexts of domination and lordship (LH 11.35; IV.4; X.31). In Books III-V, the main usage of francus (and of its pejorative equivalent barbarus) is military and political. It appears routinely in battle and campaign descriptions. Before the Thuringian campaign, Theuderic I of Austrasia summoned 'the Franks' (LH III.7); cthe Franks' threatened to desert Theuderic for his brothers if he did not lead them on campaign (LH III. 11); during Chlothar I's disastrous Saxon war it was 'the Franks' who refused to make peace and clamoured for battle (L//IV.14). When describing his own time, however, Gregory used francus in situations where it implied political power: 'the Franks' asked Sigibert I to take over Neustria in 575 (L//IV.51); they attended the Neustrian court and brought wedding presents for Rigunth (LH VI.45); the Treaty of Andelot was made with the consent of cthe Franks' (LH IX.20). The army, however, was still the exercitus francorum (LH IX.25; X.3), though that may have been because it was the instrument of the regnum francorum. As Irsigler (1969) argued, Gregory's overwhelmingly politico-military use of the ethnic label might support the idea that to be Frankish was to be a warrior aristocrat. However, Gregory twice refers to Austrasia as Francia (LH I V.I 4 and IV. 16), and he may have considered that the bulk of the northern population was Frankish. In Paris a mob of 'Franks' gathered outside the church where Praetextatus of Rouen was being tried (L//V.18). Fredegundis was confronted in Rouen by a deputation of 'the senior Franks of the area' (seniores illius loci franci; L//VIII.31). Tournai was inhabited by 'Franks' (Tornacenses franci; LHX.27). The Franks who are mentioned in political contexts are qualified as utiliores (LH IV.22) or meliores (LH VI.45). A boy miraculously cured near Carignan was 'the son of a certain Frank, most noble among his people' (LH VIII. 16; see below). Chilperic I talked of acting for the glory and renown of 'the Frankish people' (L//VI.2). There were also 'Frankish' customs and rituals (ZJ7VII.29; VII.32; X.27). Perhaps most interestingly, Gregory (LH 111.36) tells us that the Franks of Trier despised Parthenius the tax-collector because 'in the time of King Theudebert he had inflicted tribute upon them'. More specifically (LH VII. 15), in Paris, Audo's life was in danger because, 'along with Mummolus the prefect, he had subjected to the public tribute [i.e. tax] many Franks who in the time of King Childebert [I] had been senior free men'. These passages again suggest that Franks were not subject to at least some taxation. Their anger took the form of righteous indignation whereas southern, Roman tax riots were based upon injustice or breach of specific, previously granted immunity. The implicit equation of 'freedom' with partial tax exemption and Frankish identity (LH VII. 15) is interesting. Francus in Gregory's works seems, therefore, to be a term used on several levels. At one level it meant the 'Frankish people', the inhabitants of northern and north-eastern Gaul. These Franks seem to have been involved in political 30

SOCIAL ORGANIZATION: DOCUMENTARY EVIDENCE

and military affairs and to have been free from certain taxes, and this led to the use of the word in a more restricted sense to mean the northern 'Frankish' aristocracy. It is difficult to see the 'Romans' of the north in Gregory's works. He never specifically mentioned such non-Frankish people. This may be because to him all northerners were Franks, and thus Frankishness was a general regional identity, used when no more specific information was forthcoming. One embassy included 'Warinar [or Warmar] the Frank and Firminus the Arvernian' (L//IV.40), whilst another comprised 'Bodegisil the son of Mummolenus from Soissons, Evantius the son of Dynamius from Aries, and Grippo the Frank' (LH X.2). The Vita Gaugerici adds some information for the neighbouring region of Trier. Gaugeric, though having a German name, was a Roman by birth, his parents, Gaudentius and Austadiola, being neither rich nor poor (VG:281). However the vita's later seventh-century author appears to say that their home, the oppidum of Yvois (Carignan), belonged to 'Germans'. Here, the ethnic divide was clearly noteworthy, but it does not seem that Gaugeric's family were in any way subordinate because of their ethnicity. Nonetheless, they gave their son a German name. Gregory's information on the Roman population further south cannot be used to close the gaps in our knowledge about the north-east. The situation south of the Loire was very different. Salic Law did not apply there, and there was no extensive Frankish settlement. Here some Romans performed military service, as numerous references to the men of individual civitates under arms demonstrate. There are indications that not everyone was liable to military service in the south (LH VII.24; VII.42), but how such liability was assessed is mysterious. Tenants were apparently not subject to military service; those of St-Martin, Tours, were not accustomed to serve because they were 'the men of St Martin' (LH VII.42). This perhaps compares with Salic Law's distinction between landholders and tributarii but beyond that it is difficult to go. In the absence of a substantial 'Frankish' population the burdens placed upon Roman landholders may have been different in the south. Whether or not they were subject to military levies, they were certainly subject to tax (LH V.28). Changes in the nature of the ethnic division and of the social groupings thereby created begin to become apparent towards the end of the sixth century. The Decrees of Childebert II use ethnic labels in slightly different ways from earlier legislation. At clause 3.i, a law concerning latrones (probably hardened criminals), 'Franks' are differentiated from 'lesser people', which might support the interpretation of Franks as aristocrats, or might simply reflect the higher legal status of Franks. In clause 3.vii Salian Franks are differentiated from Romans in the matter of working on Sundays. This distinction, in a law not concerning crimes against the person, would have been unusual in earlier law. Lex Ribvaria, probably issued in 626 as an Austrasian code, is almost certainly territorial. In this code there are significant differences from earlier practice in the use of ethnic terms. Gone are references to Roman landholders or tributarii. The chapter about murder (Lex Ribv. 40) only assigns specific identities to foreigners; everyone else is a 'Ripuarian'. Importantly, clause 68.ii mentions Romans being liable to various 31

SOCIAL ORGANIZATION

forms of royal service, which might or might not include military duties. If so, any previous ethnic differentiation in this matter had broken down. Clauses 61.viii, 61.x-xi, 61.xix, 69.ii and 90 specifically equate Romans with king's freedmen or church freedmen. 'Roman' status has less of an 'ethnic' sense than before and implies rather a degree of personal freedom {Lex Ribv. 64). All Romans are dependent upon Ripuarian freemen, who speak for them at law and pay fines on their behalf, just as with the half-free. Perhaps not unexpectedly, clause 61.xi is the first legislation about mixed marriages. In cases of mixed marriage the offspring descend to the lower, Roman legal status. The equation of Roman ethnicity with 'freed' status is explicit. These are the only clauses to mention Romans. On the other hand the use of 'freeman' and 'Ripuarian' is widespread throughout the code and in relation to a far wider range of crimes than was the case with the use of 'freeman' and 'Frank' in Salic Law. A further difference from earlier law, however, is the specification that Roman law could be followed in cases involving people of 'Roman' status {Lex Ribv. 64.ii). In other cases, 'Roman law' is specifically associated with the church, and church freedmen {Lex Ribv. 61). The nature of ethnic identity has changed. It now refers more to legal status and degree of freedom, and is less the functional difference which it had been in the sixth century. In the seventh century far more personal names are known to us. The first point to note is that after Agiulf 's accession to the bishopric of Metz (between 580 and 601) only three bishops had Roman names during the remainder of the Merovingian period (Papolus, c. 612; Aptatus and Felix at the end of the seventh century). The Wissembourg cartulary furnishes a long list of landholders, their wives and children. With a handful of exceptions (e.g. Bonifatius, Benedict, Constantine, Justinus, Petrus) they have German names. By the end of the Merovingian period almost every north-easterner known to us has a Frankish name, even clerics. Indeed, by this time it is clear that all the free, lay inhabitants of the north regarded themselves as Franks; Romans were the people who lived south of the Loire (cp. Vita Eligii 11.20). The term francus still retained its association with the militarily and politically powerful, however (Gerberding 1987:167-8 for the Liber Historiae Francorurns use of the word), though, as we shall see, with slightly different emphasis. By the early eighth century, there is evidence that much of the region of Metz was now Frankish-speaking. Germanic place-names and names of topographical features and settlement types are increasingly recorded in the charter evidence from the east and north of the diocese (above, ch.l, and below Part II, for further discussion). This surely implies that during the fifth to seventh centuries these areas had been settled sufficiently densely by people from beyond the Rhine for major cultural changes to be effected. It is worth remembering this when rightly considering the often fictive nature of early medieval ethnic identities, and the socio-political purposes which their adoption, or abandonment, could serve. There were, in the north-east, many Franks who, simply enough, actually were descendants of people who migrated into Gaul from beyond the Rhine!

32

SOCIAL ORGANIZATION: DOCUMENTARY EVIDENCE

Horizontal barriers (1). The late fifth and sixth centuries It must be conceded that a description of Merovingian society in northern Gaul builds on uncertain foundations. There is a huge gulf between, on the one hand, the many useful studies of late Roman governmental legislation and policy, and of literary sources, and, on the other, regional studies of the realities of late Roman western provincial society. In northern Gaul there are almost no late imperial written sources or epigraphic data, which prevents documentary historical studies of the type carried out for other late antique, particularly eastern, provinces. The vast corpus of data from the many excavated settlements and fewer well-recorded cemeteries has, unfortunately, not been tapped to drive social historical surveys, even for the early Roman period.5 Work has tended to be descriptive, concerned with villa typology, Romanization, religion, economics and manufacture (Baudoux et al 1990). There are, however, indications that some local archaeologists are beginning to move towards more sophisticated terms of analysis (Hoerner 1986; Burnand 1990). For the late Roman period there is a significant gap between what we know of the central imperial government's policies and statements and the literate elite's ideologies, and what this actually meant 'on the ground' in Belgica Prima. As stated, this makes a shaky foundation from which to study post-Roman social change on a local basis. Nonetheless, some provisional points can be made. The late Roman nobility Traditionally, the late Roman period saw the creation of a rigid, sharply defined social hierarchy with a small, immensely powerful nobility at its apex, standing in a clearly delineated relationship to the central government. In return for involvement in the extraction of the surplus necessary to maintain the imperial administrative superstructure, the nobility was rewarded with profuse titles and privileges and supported by legislation stratifying society yet more rigidly. In turn the nobility found the imperial administration to be one of its most important sources of wealth and authority. It was thus able to build up local power based upon large latifundia. In northern Gaul, this view can be disputed (Burnand 1990:211-32 is a thoughtprovoking antidote; see ch.8). Salvian's diatribe (esp. Governance V.5, V.8), so long one of the mainstays of the traditional view, has been reinterpreted more critically (cp. Drinkwater 1992), as showing more subtle struggles for local power than those simply based upon increasing opposition between oppressive aristocrats and oppressed coloni. The early Merovingian aristocracy The dearth of written evidence between the fourth and the sixth centuries makes it difficult to study the development of this social organization. When 5

Hingley 1989 on Roman Britain is an example of what might be done with the settlement data; Roman Britain is indeed better served all round than its neighbouring provinces.

33

SOCIAL ORGANIZATION

Salvian was writing in the 440s northern society was clearly in turmoil. Many Roman nobles, like Salvian himself, had fled to the south or even further (Mathisen 1992:228-30; Hunt 1992). Light can be shed on one aspect of this process. One feature of the late imperial government's symbiotic relationship with its regional aristocracy was patronage, the multiplicity of offices within the administration, which served to maintain the stability of the social order, as well as being, for some, a passport to high noble rank and consequent financial and fiscal privilege and immunity. Some curial offices, to be sure, were more expensive than profitable, becoming difficult to fill even in the fourth century. These doubtless vanished in the chaos of the fifth. They and the other, lucrative, posts depended, however, on their situation within a provincial and diocesan framework and thus on the maintenance of the wider imperial administration. As was seen in chapter 1, this kind of administrative structure ceased to be effective in northern Gaul by about 420. Even the efficient presence of the state, necessary for the maintenance of status based upon imperially bestowed rank and titles, and, indeed, of scattered extensive lands, increasingly ceased to be felt after 388. One response of the central government to this type of process may have been the institution of the comes civitatis, the count of the city, an attempt to invest local administrative and military power in the hands of a single magnate. Barnwell (1992:35-6) demonstrates the very weakness of the evidence for such a supposition, but the most plausible context for the creation of the comites civitatum remains the collapse of the Roman Empire. The counts, of unspecified type, glimpsed in accounts of the fifth century (Paul, Arbogast) look no different from other local warlords. They were, however, nominally at least, acting in the name of Rome and provided regional power bases which could support, and in turn be shored up by, military expeditions from the imperial bases in the south-east. Nevertheless, the withdrawal of effective government meant a significant reduction of imperial patronage.6 The local elite lost an important means of cementing its power; a vital link between it and the Empire was removed. This may have been a factor which led to the emigration of magnate families. The power of those remaining could no longer be drawn from old sources; new ones needed to be found. When documentary evidence again becomes available in the mid-sixth century the picture of society is quite different. Obviously, the settlement of the Salian and Ripuarian Franks, and possibly the Alamans too, had caused great disruption. The personnel of the upper social strata had been changed and at a lower level a new element had to be incorporated into the rural population. This cannot but have added to the upheaval caused by the disintegration of the Roman state. The social hierarchy which emerged from this turmoil has long been a matter for debate, largely concerned with the nature of the aristocracy. Was there an hereditary nobility in the sixth century, as, for example, Irsigler (1969) argued, or was there no formal aristocratic rank? This is an historical minefield which has justly been worked through by a number of theses (Bergengruen 1958; Irsigler 1969; Grahn-Hoek 1976). Any discussion of 6

I am grateful to Dr Peter Heather for pointing out to me the scale of this process. 34

SOCIAL ORGANIZATION: DOCUMENTARY EVIDENCE

social organization must deal with this problem but there is no space here to do more than sprint through the minefield. The Pactus Legis Salicae recognizes no hereditary noble rank among the Frankish population (Grahn-Hoek 1976:27-54). Instead all 'freemen' have a wergild of 200 solidi. Higher wergilds were assigned according to age or gender, as will be outlined, but there was only a single set of wergilds for the free population. The only other people with higher wergilds were those holding royal office (as grafio (count) or sacerbaro), the sworn members of the royal retinue (trustis), and men on active service. Clearly there must have been de facto differences in the wealth and prestige of 'freemen' even if, de jure, they were all created equal (Grahn-Hoek 1976:37). Nonetheless, this lack of legal distinction is extremely telling and arguments from it cannot be dismissed as based upon nineteenth-century beliefs in early Germanic democracy (cp. James 1982:130). Wickham (1992:233) makes an important contribution to this debate when he draws attention to the PLS' obsession with rural society (in which, it might be added, it differs significantly from seventh-century Lex Ribvaria). As he says, it is difficult to envisage such a treatise showing these concerns in a society where legal theorists needed frequently to consider 'the society of large landowners' (ibid.). However, we must not ignore (as Halsall 1990 and 1992b does) the existence of a Roman population. As has been mentioned, Salic Law set out a legal distinction between two types of free Roman: landholders and rentpayers. The 'landholders' seem to have been the equivalents of Frankish 'freemen'. The only group legally separated from this general body was that of landholders in the convivia regis; again social distinction was brought about by a direct link with the king. It might be possible to argue that the Franks, the 'freemen', were themselves aristocrats, as Irsigler (1969) argued was the case with the Franks who appear in Gregory's writings. If this were the case, the legal distinctions would give the group most of the characteristics of a nobility. However, this would imply a very different nature of Frankish settlement from that envisaged by Irsigler and, indeed, by most German scholarship. The settlement would become very similar to that of the Visigoths in southern Gaul and Spain: a warrior elite battening on to a native population. However, there are significant problems. There is no legal discussion of the relationship between Roman and Frank such as one might expect if most Romans were subject to Frankish lords (PLS 39.v is a possible though not probable exception). More importantly, if the legal use of francus did mean nobleman, where are the lower strata of Frankish society? Surely there must have been some non-noble Franks? The extent of linguistic change effected in the east and north of the region strongly suggests this. We are left with two alternatives: either there was no legally defined Frankish nobility, or all Franks were noble before the law. For now the former option appears preferable. It seems simplest to assume that the legally-defined body of free Franks included all classes of wealth and power from smallholders to prestigious families with extensive estates. The vague distinction between 'lower classes' and 'upper classes' in an undated capitulary appended to the Pactus Legis Salicae (PLS 102) bears this out. Even if one takes the the alternative 35

SOCIAL ORGANIZATION

interpretation, and, as should have become clear, there is evidence to support it, one must accept that the barrier between this Trankish' 'nobility' and the rest of the population was not particularly rigid. James (1982:130) argues that the narrative sources show a powerful sixthcentury aristocracy in existence, whose power 'was by no means dependent upon the king'. Gregory of Tours' writings do indeed depict men such as Guntramn Boso, Rauching, Berthefried and Ursio as extremely wealthy, with numerous followers and widespread landholdings. However, this picture does not contradict the impression drawn from Salic Law. The power of these magnates, or optimates as they are called in the Pactus pro tenore pads (524-58; Grahn-Hoek 1976:55 n. 248 for the dating) was extremely uncertain and based essentially upon the prestige of the individual. When Count Bodigisil died of old age (585) Gregory thought it noteworthy that he passed on his estate 'without having to forfeit any of it' (LH VIII.22). Unfortunately, no location is given for the story; Bodigisil may have died in the far south, though his estate could have lain elsewhere. When Waddo was killed near Poitiers (589) his son had to go to the king before he could enter into possession of his father's estate (LH 1X35). These stories suggest that the power of the Frankish elite, south of the Loire as in the north, was based upon gifts from the king, of fiscal resources. Anecdotes like these thus argue against aristocratic power being independent of the king, and underline just how uncertain the transmission of power was from one generation of 'Franks' to the next. The army, the political assembly of the Franks, could dictate a king's 'foreign policy', threaten to betray him for his brothers (LH 111.11) and even manhandle him if he refused to act as they wished (LH IV. 14). Nevertheless, even these stories underline the fact that, individually, the members of the army relied upon the king and his wars to bring them treasure and the chance to demonstrate their worth and to try to obtain offices through good service. The relationship between aristocracy and king can be illuminated if we consider the question of taxation. Sixth-century Merovingian rulers attempted to preserve late Roman taxation (Durliat 1990). As we have seen, Franks appear to have been at least partly exempt. This may have sprung from the Roman army's exemption from taxation; if the Franks were the army then they were not liable to the capitatio. The mechanics of sixthcentury Merovingian taxation in the north, and its efficiency, are largely concealed from us. Indications from the south are that taxation was levied on all landholders, royal officials (counts and their subordinates) and appointees being responsible for its levying. In the absence of small denomination coinage it was probably levied in kind (confirmed, in relation to the Limousin, by LH V.28), as had been much of late Roman taxation, or in bullion, perhaps using old coins still in circulation. Royal officers had the responsibility for turning these levies into bullion (Goffart 1982b). The efficiency of taxation may not have been high. The Franks' violent resistance to attempts to tax them, and anecdotal evidence from the south, where there was more administrative continuity, have been used to suggest general opposition to taxation and ever-increasing immunities (Goffart 1982a: 15; Wickham 1984:21). We should perhaps not be misled by this evidence. Parthenius was 36

SOCIAL ORGANIZATION: DOCUMENTARY EVIDENCE

stoned, and Audo driven into sanctuary, because they tried to tax a social group not deemed liable for taxation. The nature of Gregory's other tales implies that acquiescence in taxation was more the norm; the revolts were against unfair taxation rather than against taxation toute courte. Similar points can be made about military service. This is not surprising; the levying of taxes and the harnessing of manpower for military service were the two most important ways in which the royal government impinged upon the lives of its subjects. Here again we see that all free Franks were liable to such service. To adapt a phrase of Kiefer's (1972:55), to be Frankish was to be capable of fighting; a man must fight because he is a Frank. The association of Frankishness with military power has already been noted. That the army was a political as well as military assembly is shown not only by references in Gregory's works but also by the fact that the edicts of Childebert II are dated on 1 March, the date of the annual military muster, the Marchfield. Neither the narrative sources, nor the hagiography, nor the laws give any indication that military service was differentiated according to wealth or social status, as it would be later, let alone restricted to the aristocracy. Aristocrats who held the offices of count, which involved marshalling and disciplining (cp. LHV1.31) levies, or of duke, which involved leading armies, were the only ones with any institutionalized role in the raising of armed forces. The only other military aristocracy was composed of the antrustiones, the members of the royal trustis or retinue. In all cases royal appointment or a direct relationship to the king were stressed. Beyond that, the theory was that the army was the levy of all free Franks - and all Franks had the right to serve (something to which we shall return). Institutionally, as with taxation, the aristocracy was by-passed or harnessed to the king. In questions of military service, early medieval kings across Europe long strove to maintain such a system, or at least the fiction of one (cp., for Italy, Wickham 1981:136-40), even in the heyday of military feudalism (cp. Henry IPs 1181 Assize of Arms; Douglas & Greenaway (trans.) 1981:449-51). Another aspect of the relationship between king and aristocracy concerns royal marriage. As is well known, the Merovingians of the sixth century drew their wives from two sources: foreign princesses and slave or servant women. The two groups are well represented by the protagonist queens in the wars of the late sixth century, Brunechildis the Visigothic princess, and Fredegundis the ex-slave. Merovingian kings did not marry their aristocrats' daughters or sisters, though they did sometimes promote the relatives of their low-born wives (such as Berthechramn, a relative of Chlothar Ps wives Aregundis and Ingundis, who became bishop of Bordeaux). This refusal to unite in marriage with aristocratic families, and the actions of certain rulers and regents such as Brunechildis herself suggest it was 'policy', can be seen as a correlation of the kings' attempts to by-pass the de facto powerful in the extraction of surplus or manpower, and of their attempts to frustrate the creation of an institutionalized aristocratic rank, a nobility. It may have been behind the anger of the Franks at Theudebert Ps marriage to Deuteria, a Roman noblewoman (LH 111.27). As far as we can tell the inverse was also the case; Merovingian princesses did not marry into aristocratic families. One possible exception concerns the eighth-century tradition recorded by 37

SOCIAL ORGANIZATION

Paul the Deacon (Lib.Ep.Mett.:264) that Bishop Agiulf of Metz was descended from a daughter of Clovis I who had married a Roman senator. Though problematic, this tale may be based upon fact (though, as mentioned, Elziere 1993: n. 45, makes him a relative of Deuteria). In the early phases of his reign, Clovis may have felt the need to make alliances with the more powerful native aristocrats, as, in an obviously garbled story, Procopius records (Wars V.12.xii-xv). By the reigns of his sons, this need had passed. Patronage is a final aspect of the relationship between kings and their aristocrats. Mention has been made of the dramatic reduction in imperial patronage in the fifth century. This, as stated, weakened the link between the imperial government and the local aristocracy. But what had been a weakness and a failure for the western emperors could be made into a strength and a success by the Frankish kings. Above all, sixth-century kings, unlike fifth-century emperors, ruled Gaul from Gaul, and could make their presence felt, when necessary, in any part of their smaller realms, enabling skilful use of patronage. Titles such as duke or count were few and worth competing for, particularly as they brought substantial local administrative and legal power, and control over fiscal lands, which could be exploited. Such titles were not, however, permanent and did not bring hereditary rank, unlike Roman high offices. The bearers of these titles were tied directly to the king, who thus held all the aces. By clever regular redistribution of comital and ducal offices, the king could both ensure that his patronage was spread widely and evenly and prevent any one faction or group of families from acquiring too much power. Such distributions and redistributions also enabled kings to keep control over their more distant fiscal lands. When offices were hereditary, later experience would show that kings would have enormous difficulty in preventing their lands from falling under the control of the local count or duke (cp. Reuter 1991 :esp. 85-7, for the case of Carolingian Germany). Contemporary terminology is also illuminating. The appearance of the Frankish word leudes, and of the Latin terms optimates, potentes and even viri magnificentissimi obtimates in the sixth-century capitularies (PLS106, issued by Chilperic I), does not necessarily argue against these conclusions (cp. Grahn-Hoek 1976:55ff). As has been repeatedly pointed out (cp. Bergengruen 1958:ch.2), none of these titles says anything about security of tenure or whether or not such status was hereditary or granted. In the Histories Gregory uses nobilis thirteen times to describe people. In eleven cases it is used to describe Romans, people living south of the Loire (including one Italian Ostrogothic context; L//III.31) or the Merovingian royal house. In only one instance (LH VIII. 16) is it used to describe a Frank (Francus cuiusdam et nobilissimi in gente sua). Given Gregory's usual modus operandi in discussing politically sensitive issues,7 it may well be that he is talking of an unnamed Merovingian here. Genicot (1962:52) was quite wrong to say that the Histories are full of nobiles; they are not, and where nobles 7

These characteristically (see Wood 1993) include cryptic circumlocutions only understood by reference to the rest of the text, here the other uses of nobilis in a Frankish context; the situation of such discussions in reported speech; and the authentication of statements by their association with men or women of sanctity, in this case Vulfolaic the Lombard.

38

SOCIAL ORGANIZATION: DOCUMENTARY EVIDENCE

do appear they are either Merovingian royalty or southerners. If Gregory shared his younger contemporary Isidore of Seville's definition of nobilis, 'one whose name and lineage are known' {Etymologies X.I84), it is surely significant that the only term which could imply an hereditary nobility is largely absent in Frankish contexts. This is underlined by the fact that hardly any of the Frankish counts mentioned by Gregory can be shown to have had a father of similar status. The Soissonais Bodegisil (not to be confused with the Bodigisil who died of old age) and Bobo his brother are exceptional among northern Frankish aristocrats in that Gregory knew their civitas of origin and the name of their father. How did all this theory work in practice in the north-east? Ultimate dependence upon the king is shown by the careers of the principal Austrasian aristocratic faction mentioned in Gregory's Histories: Guntramn Boso, Rauching, Ursio and Berthefried. The 587 'Revolt of the Dukes' was led by these men, of whom at least two were associated with the region of Metz. The family of Guntramn Boso's wife was connected with Metz (L//VIII.21); Ursio held estates in the Woevre (L//IX.9). Although such men could acquire vast lands and treasures, their continued well-being relied entirely upon the maintenance of good relations with their master. When they fell foul of the king they had no option but to take to a sanctuary (LH IX.8) or a stronghold (LH IX.9) and plead for forgiveness. Either way the end was near. The careers of men like these show that whilst it was possible to hold out for some time (cp. LH VI. 1 and ff (strongholds); LH V.4 and ff, and VII.21 and ff (sanctuary)) and sometimes to slip into and out of royal favour with comparative ease, the withdrawal of royal support was usually fatal. Boso, Rauching, Ursio and Berthefried all met similarly bloody ends (LH IX.9-10; IX.12). Their property was confiscated by the fisc, and their families exiled. The removal of royal favour was frequently arbitrary. After the revolt of these magnates, Gregory tells us, 'certain dukes were removed from their dukedoms, and others were appointed to replace them' (LH IX.12). In Metz itself, a certain Magnovald was summarily executed on Childebert IPs orders and no one knew the reason why (LH VIII.36). His property too was appropriated by the treasury. In these cases we see in action the redistribution of royal patronage, and the subsequent reestablishment of control over fiscal lands. Though there were clearly powerful men in sixth-century Austrasia their power was by no means guaranteed. Samson (1987b:288) has concluded from an analysis of the Libri Historiarum that perhaps as many as 40% of the highest levels of office-holder (dukes, counts, patricians, etc.) could expect to die violently. The non-aristocratic free in the sixth century Sixth-century people had as much difficulty as us in drawing a firm line between aristocrats and the rest of the free population; in our terms a rigid horizontal barrier did not exist. The nature of the elite obviously has important repercussions for the discussion of the less powerful classes. The study of the non-aristocratic Merovingian free has suffered from the tendency towards aggregation of evidence (castigated in chapter 1). This study aims 39

SOCIAL ORGANIZATION

to describe a society in constant change, and one very different from that of central and southern Gaul, the areas best served by written data. Southern Gaul is exceptional in post-Roman Europe. The chaos of the transition from Roman province to barbarian kingdom there was less, and the continuity of social organization greater, than in almost any other western region. It is unwise to generalize from this area's evidence (the survival of which is itself unusual). Where emphasis has been laid upon change we run into the tendency towards teleology, equally common in Marxist and non-Marxist histories. For the Roman period we have plentiful statements, by the government and its nobility, about the nature of the rural population, coloni and slave tenants, and their relations with their landlords. For the Carolingian period we have voluminous legislation and statements of ideology, with the additional benefit of plentiful data about specific regions and estates, and so can say much about the late eighth-, ninth- and tenth-century peasantry. It has been possible to detect similarities between the late Roman and Carolingian situations (cp. Percival 1969; Goffart 1972 is a sophisticated critique). Researchers have assumed that the state of affairs during the less well-documented Merovingian period was somewhere mid-way between. The Roman tenurial system was evolving into medieval serfdom. A social formation based upon slavery and/or taxation was at some point (exactly when being the subject of debate) giving way to a feudal social formation. cIt is very easy to be misled into seeing all these processes as teleological, viewing 'fully formed' military feudalism and the seigneuries as the logical and inevitable result of the economic relations of private property' (Wickham 1984:29). Like all analyses which take predetermined categories of social formation and the progression from one to another as a starting point rather than a descriptive result of the study of social change, even Wickham does not entirely escape his own strictures. The necessary corollary of envisaging society in a perpetual state of redefinition and renegotiation of the barriers on the 'social map' is that there can be no inevitability about social change. That analogous situations existed in c. 300 and c. 800 which could plausibly be linked by a single and direct process of 'evolution' is no reason to suppose that there should not have existed in c. 550 a social formation radically different from both. It should come as no surprise, given the foregoing discussion of the aristocracy, that those of the free who could not be included in that category are found generally to have lacked the classic relations of dependence on their more powerful fellows. The Roman tributarii of Salic Law are the single exception. In the context of the law (PL5 41), tributarius clearly retains its old Roman sense: a payer of tax and rent to a landholder. Among 'freemen' there is no mention of rent-paying or of being in the service of others. The impression given of local politics is of decisions made by the local community. Struggles for local power within this arena are made from a, de jure at least, more or less equal starting point. Legislation about the formation of illegal armed bands talks of contubernia (PLS 42, 43), which other use of the word would argue were gangs of equals (pace Irsigler 1969). The accept-

40

SOCIAL ORGANIZATION: DOCUMENTARY EVIDENCE

ance or otherwise of a newcomer into a community depended upon the goodwill of all community members equally (PLS 45). As has been argued, the theories of taxation aimed at raising taxes from all landholders of whatever size of property. Military service was, as suggested, levied 'horizontally' on all 'freemen' rather than Vertically' down a chain of lords, dependants and retainers. The Pactus Legis Salicae's concern with protecting freemen in the army (PLS 26.i, 63, 69) manifests a desire by the royal legislators to encourage people to perform their duties, but, at the same time, to exercise their rights. Armed assemblies were, obviously, occasions where personal rivalries could burst into violence and old scores be settled, as a law of Chlothar I (PLS 91) acknowledges (and as, later, Fredegar, Chron. IV.55, shows), and weaker freemen might thus need protection and encouragement to attend. Sixth-century legislation maintains, similarly, equality before the law. It does this not simply by assigning a single set of wergilds, but by implying that royal law is applicable everywhere within the realm. Every freeman is answerable to the king's law and has access to his justice, which is everywhere administered by his officials, albeit assisted by shadowy figures called rachimbergs. The nature of these 'speakers of the law' is unclear but an equation with modern jurymen is plausible (Rivers (trans.) 1986:12). Whatever the case, in theory, the law, like taxation and military service (as far as can be judged), applied equally across the free population; the only mediation between freeman and king was through royal officers. The free population, in our terms, constituted at most three ranks, Roman rent-payers, Roman landholders and Franks. Within this amorphous mass there were many classes of more or less powerful. At the more powerful end of the spectrum were those whose relationship with the king accorded them temporary privilege. At the other were those, like the tributarily who were dependent upon more powerful men, but among the Frankish population these were not yet recognized as a separate category. Beyond this, the only institutionalized attempts to group the free population were those using vertical and diagonal barriers: gender, age and family. The half-free in early Merovingian society There were ranks below those of the free: the half-free (laeti) and the unfree, slaves. Evidence for these groups is even scarcer than for the non-aristocratic free, who made up the bulk of the population. The laeti are particularly obscure and can only be approached via references in the laws. Freeing a slave usually made him or her a 'freedman' or 'freedwoman', technically a 'half-free' person, linked in a direct relation of dependence to the manumitter. If we are searching for a dependent population or peasantry then it may be to the laeti that we must look. The scarcity of overt references to them hardly implies that they were numerous, but, as suggested earlier, the bulk of impersonal (si quis) legislation may include them. Their social role is obscure. In the Roman period laeti had been freed prisoners settled upon the land in return for agricultural and military services. It is not difficult to

41

SOCIAL ORGANIZATION

see how the name came to be applied to freed slaves, particularly if the latter had been captured during raids. If laeti were settled on, and worked, portions of estates, the name is yet more suitable. As for military service, the law (PLS 26.i) makes it clear that half-free men could be found in the army with their masters, though it is not clear whether they came to fight or, as with certain married ceorls in the seventh-century Northumbrian army (Bede, Ecclesiastical History IV.22), just to fetch and carry for their masters. The latter is perhaps more likely. Like the Roman laeti, the sixth-century halffree were probably required to perform agricultural and military duties. Related to the laeti who performed military duties may have been the pueri encountered both in Salic Law and in the narratives. It is occasionally assumed that these are slaves, the application of the word 'boy' to male slaves being analogous to the situation in the antebellum United States. In the legal sources the word puer means, literally, a male sub-adult except, it seems, when talking of pueri regis, usually translated as 'king's servants' (PLS 13.vii, 54, 117). These pueri regis are equated with freedmen or laeti and so are the pueri mentioned in PLS 42.iv (where the word regis may have fallen from the text), where they are also equivalent to 'Romans'. The puer is then distinguished from a 'freeman' but that term, as we have seen, seems to have been rather more specific than is often presumed. The law-makers also saw some reason to separate pueri from slaves and freedmen. Where we meet pueri in Gregory's Histories (cp. VII.42), they are retainers; their actions rarely suggest that they were in any way unfree. An equivalence with the ministeriales ('unfree knights') of central medieval Germany might be suggested, but it is not the only interpretation. If laeti lived apart from their masters, farming separate lands, this might explain why kings and their legislators felt less need to regulate their interactions with their masters and other members of the free population than those of slaves. Perhaps, however, it is safer to assume that there were simply not very many of them. Early Merovingian slavery* The question of the numbers of the unfree is vexed. Bloch (1947:30-1) surmised that the wars attending the end of the Empire in the West greatly increased the number of slaves, and this premise is still widely accepted (cp. Bonnassie 1991b:2; Wickham 1984:4-5). Unfortunately, it is not based upon any evidence. Fifth- and sixth-century warfare doubtless produced its unfair share of human booty (cp. LHIX. 18). Literature is full of references to captives and their ransom (cp. in a local context, Lib.Ep.Mett.:263). But the scale of warfare was smaller than it had been and, within Gaul after 536 at least, it was waged within a functioning political system. There were usually means of tracking down and ransoming enslaved relatives (LH III. 15). The law (PLS 39.ii-iv) envisaged mechanisms whereby freemen and kidnapped slaves sold abroad might be redeemed. Plundering raids may well have served more to redistribute an existing unfree population than to create substantial num8

For slavery in the late antique and medieval world see Bonnassie 1991b; Dockes 1982; Finley 1980; Verlinden 1955 (Book II, ch.l for Merovingian Gaul), and bibliographies.

42

SOCIAL ORGANIZATION: DOCUMENTARY EVIDENCE

bers of new slaves. Enslaved captives brought a short distance over the frontier can have hoped for escape back home or for release during the next counter-raid launched by their own side. A small raiding army could do more damage to its enemies by freeing the latter's slaves, and make a more profitable return by 'enslaving' free people simply to hold them to ransom. It would split its own strength by the need to guard captive slaves day and night, and both slow its speed and increase its vulnerability to attack by trying to shepherd large numbers of human beings across the countryside. Sixth-century material suggests that the prime sources of slaves were largescale operations outside Gaul (e.g. the invasion of Thuringia) and foreign slavers, both mechanisms which could operate in reverse (as in Breton, Basque and Lombard attacks, or as in the PLS reference to selling slaves abroad), as well as penal slavery and hereditary slavery. It is doubtful whether any of these means did more than replace slaves lost to manumission (as reward for good service, as death-bed piety, or, probably rarely, through ecclesiastical intervention) or mortality (which must have been higher, and fertility less, among the unfree than the free). Sixth-century and later legislation suggests that the manumission of a rival's slaves was a tactic in local power struggles, particularly as a freed slave became a dependant of his or her manumitter. The lack of references to the half-free in sixth-century law suggests that most slaves made the transition directly to free status. Halffreedom was an unwanted complication in sixth-century northern Gaulish society, with its paucity of rigid horizontal barriers. Many slaves will have acquired their freedom in fifth-century northern Gaul, with its political chaos and uncertainty, the local changes in patterns of authority and the flight of many landlords, replacement of others and attendant upheavals. This could be done either by running away, by obtaining freedom from their owner's rivals, which in this era must have been even more common than in the sixth century, by joining barbarian warbands or the bands of rival local masters (sufficient slaves did this for it to become a cliche, and even for illegitimate authority to be dubbed slave rebellion), or, for agricultural slaves, simply by continuing to farm the same land and claiming free status. No source of this period gives a clear indication, let alone states outright, that the numbers of the unfree were increasing. It seems far more likely that slavery in northern Gaul was on the decrease. The social and economic role of slavery is, of course, directly linked to the number of slaves. There are few enough anecdotal references to slaves in northern Gaul, and the issue is blurred by the mistranslation of puer as 'slave'. Those which survive, and those from further south in the Loire valley (where slavery may have been more common), tend to situate slaves in the households of their owners or otherwise in direct attendance upon them, rather than living separately, farming distinct parts of the estate, or in slavebarracks. Our primary evidence from the north, the legal material, gives a similar impression. The royal legislators' concern with 'mixed marriages' between slaves and free is perhaps more likely to have arisen in a society where small numbers of slaves lived in their owners' households than in one where slaves led separate existences. Such slaves probably worked (or more 43

SOCIAL ORGANIZATION

probably helped to work) their owners' lands and did household chores. The laws also enumerate slaves involved in more specific skilled work: blacksmiths, swineherds, vinedressers and grooms (PLS 35.ix). Nothing contradicts the impression of numerically few slaves living in close proximity to their owners. In this kind of situation slave-free relations were frequent, everyday and intimate, carried on within a very restricted arena. Legally, a slave nonetheless remained a 'non-person', under the absolute control of the head of the household and subject to whatever discipline he wished to impose. Slaves were simply chattels: cif anyone steals another's slave or maidservant, horse or ox . . . let him be liable for thirty-five solidi (PLS lO.i). They could be sold to another owner with or without their spouses and children. There was no legal recognition of slave marriage. The slave was, as Patterson (1982) terms it, a 'genealogical isolate'. Children born to slaves were also slaves and any free or half-free person who married a slave joined his or her partner in unfree condition. Being completely at the mercy of another person in, to use Patterson's (1982) phrase, ca relation of domination' could and probably did result in horrific mental and physical abuses, especially where the owner was a powerful aristocrat with many slaves. In Francia, unlike Visigothic Spain, there were no laws limiting slave-owners' power to inflict physical punishments on their slaves, but it is likely that with most slave-owners living under the same roof as perhaps one or two slaves, free-unfree relationships were, if not harmonious, at least rather more peaceful. The intimacy of these relationships is further supported by tales, albeit from other regions, where arbitrary violence against a slave resulted in retaliation by the slave and the owner's injury and even death (LHVIIA7). This could suggest, though it is admittedly far from being the only interpretation, that routine violence against slaves was not the norm. One would not expect it to be in the kind of small-scale, domestic slavery suggested here. The Pactus Legis Salicae is unclear on the question of slaves having possessions. In places (e.g. PLS 12) it implies that a slave could pay fines. However, PLS 40 makes it explicit that this can only be done with the master's consent, presumably because the slave ultimately has no possessions. PLS 35.viii specifies that the slave's master was responsible for heavy fines. More important, however, is the PLS9 repeated stress upon physical punishment for slaves, and torture as a means of establishing guilt. Early Merovingian slaves doubtless saw more of their masters and mistresses than had many Roman slaves, but, as their numbers fell, so their conditions seem to have improved. The realities of social organization produced a discrepancy between the institutional theory of the laws and the practical exigencies of the habitus. Thus, according to this interpretation, the best analogy is with an extreme, brutal version of 'Upstairs Downstairs'. This did not, of course, mean that slaves were content with their lot. Slave resistance did not take the form of armed revolt. Dispersed in ones and twos in their owners' houses, organization would be difficult; even if it were not, whatever the short-lived satisfaction it produced, violence brought little chance of lasting success and the near-certainty of harsh retribution. Resist-

44

SOCIAL ORGANIZATION: DOCUMENTARY EVIDENCE

ance thus followed lines well attested in other slave-owning societies. Laws bear witness to slaves selling their owners' produce or goods (PLS lO.v, where it is equated with theft), undoubtedly pretending to be ignorant of the legal niceties which dictated that they could own no property. Early medieval literature contains the same caricature of the slave as idle, drunken and stupid, so well known to students of black slavery in America or of nineteenth-century Russian serfdom. The reasons behind this caricature doubtless centre on the slave's best means of resisting enforced labour: getting drunk, feigning ignorance and working as slowly as possible. Less subtle resistance included flight (if the slave could run far enough, as Attalus did (LH III. 15), or escape detection for long enough, this could be successful), or, as mentioned, attempting to gain manumission from the owner's rivals. Flight into sanctuary could bring about enforced manumission, even if it meant becoming a dependant of the church. Sixth-century society was structured according to very few definite horizontal barriers. Between the two poles of slavery and royal service, the social hierarchy seems to have been fluid.

Horizontal barriers (2). The seventh and early eighth centuries The later Merovingian aristocracy In the seventh century our evidence becomes more plentiful, and by 700 we begin to have local 'documents of practice', charters. The picture of earlier seventh-century society is different from that of sixth-century. Seventhcentury law continued to deny legal recognition to any noble rank, but reveals vitally important transformations in the nature of the aristocracy and its relationships with the king and the other hierarchical ranks and classes. Lex Ribvaria, whilst often modelled on the Pactus Legis Salicae, differs notably in several respects. Certain lords have immunity (Lex Ribv. 68), some freemen have their own bands of dependants (homo cum satellitibus suis; ibid. 45.iii), and mention is made of freemen in the service of others (ingenuus in obsequio; ibid. 35). Some of these developments can be paralleled in other seventh-century legislation, such as Chlothar II's 614 Edict of Paris, which similarly mentions immunities for aristocrats (c. 14). Furthermore, Lex Ribvaria (68.ii-iii) makes it clear that not only freedmen but 'Romans' too all now have a lord responsible for them. Examination of contemporary, early seventh-century terminology reveals similarly slow but important changes in the same direction. In our major source, Fredegar's Chronicle, the word nobilis remains rare and notable by its absence in a Frankish context until the end of Dagobert I's reign. Fredegar uses the term to describe the magnates of other groups such as the Burgundians (Chron. IV.27), Lombards (IV.34) and Bavarians (IV.52). Within Gaul the existence of the senatorial aristocracy in the south and of the

45

SOCIAL ORGANIZATION

burgundaefarones in Burgundy9 suggests that the Franks were alone in not formally recognizing a noble rank. Instead, their magnates are described as proceres, primati, optimates or leudes just as before. In 636, however, Fredegar describes Aega as cof noble birth' {eratque genere nobele; Chron. IV.80), and the losses during Sigibert IIPs 639 Thuringian campaign included many of cthe bravest of his noble warriors' {nouilium fortissemi pugnatoris seo; Chron. IV.87). As becomes clear from Gerberding's (1987:167-8) analysis, the author of the early eighth-century Liber Historiae Francorum uses the word franci to mean the aristocracy, in, superficially, a similar fashion to Gregory of Tours. However, it is clear that his (or her) franci are rather different from Gregory's. They are representatives of powerful and wellknown families, controlling vast estates, numerous dependants and private monasteries, more or less independently of royal control. Later in the eighth century this is underlined by Fredegar's continuator. Glossing the earlier Liber Historiae Francorum^ he describes as noble Bodilo {uno franco nobile; Continuation 2), Leudesius (Leudesio filio Erchinoaldi nobile; ibid.) and Ansegisel {quondam franco nobile; ibid. 3), all franci in the LHF. Whereas the carnage of the civil wars between Childebert II's sons was described in terms of cmany strong warriors' being killed {Chron. IV.31), that in Charles Martel's wars for supremacy is described as the loss of noblemen {non modicum ibi perpessus est damnum de viris strenuis atque nobilibus; Continuation 9). It has often been remarked (Sprandel 1961:57-8; Graus 1965:ch.3; Prinz 1967:532) that seventh-century hagiography differs from sixth-century in that the saints are repeatedly described as of high and even noble birth (cp., in the present context, VA 1, where Arnulf is described as prosapie genitus francorum altus satis et nobilis parentibus; Cracco Ruggini 1992). Whatever the precise interpretation, it appears to support the conclusions drawn from Fredegar and suggests, as Sprandel (1961:57-8) said, a clear class identity in the later period. That Leudesius was recorded as 'son of Erchinoald' and Pippin as 'son of Ansegisel' is also significant. Frankish magnates' lineage, and thus their 'nobility', was becoming better known by the later years of the century. It is no coincidence that historians have been able to produce prosopographies of northern, Frankish aristocrats in this period, showing the tenure of offices within particular families (Ebling 1974). As before, the security of land tenure is a good index of aristocratic power. Here we must address the significance of the fact that from the early seventh century, growing numbers of documents survive regulating the ownership and transmission of land. The earliest version of the Vita Waldradae seems to be based upon the text of a charter of 596-612 (and is thus the oldest surviving fragment from the region of Metz),10 there are at least two surviving charters in favour of Metz cathedral (Gauthier 1980:387 n. 85) and Holy 9 10

Burgundian law, unlike Frankish, distinguished nobles - nobiles - from other freemen even in the early sixth century. See Liber Constitutionum 2.ii. The standard edition of the Vita Waldradae is Ada Sanctorum, Mai II, pp.51-2. There are problems with this text, however, which looks like a later (Carolingian?) expansion of a very bald earlier life (existing in a Utrecht MS), edited in Mabillon's Acta Sanctorum Ordinis S. Benedicti. From its style and structure, the latter seems clearly to be based upon an early seventh-century charter.

46

SOCIAL ORGANIZATION: DOCUMENTARY EVIDENCE

Apostles (Pertz, Mayoral 2), and the Wissembourg cartulary includes a number of seventh-century documents. Several objections may be raised against the significance of the appearance of such documents only in the seventh century. Literacy may well have been higher in the sixth century, and the use of written documents in law and the administration was widespread (Wood 1986; 1990). The date of the oldest surviving written records may then depend only upon the date of the foundation of monasteries; older charters, written on papyrus, have simply been lost. Neither explanation is entirely satisfactory. Chance loss can never be ignored in the early middle ages, but the extent and completeness of the documentary blank across sixth-century northern Gaul cannot be explained solely as an 'accident of survival5, by appeal to accidental fires and marauding warbands. Moreover, when we have Merovingian papyri which have survived 1,350 years since c. 625, a further fifty years is neither here nor there. We are not in any case talking simply of the survival of originals; even charters copied by later compilers (like the Waldrada charter mentioned earlier) go back no further than c. 600. The fact that our documents do largely come from seventh-century foundations is more interesting, but again not entirely satisfactory. Metz cathedral and Holy Apostles existed in the fifth and sixth centuries too. The difference may be explained by the fact that the seventh-century foundations were probably different in nature from earlier ones, being aristocratic rather than episcopal. This change itself requires some explanation, though, and is probably significant. A final, more satisfactory explanation would attribute the appearance of charters and other documents to the break-down of municipal archives (gesta (or acta) municipalia) in the seventh century. Documents were now kept by individuals or institutions, rather than being kept in central municipal archives. This and the collapse of the gesta led to the creation of the later seventh-century formularies, the Roman-style documents preserved therein providing the link between Merovingian deeds and late Roman prototypes (Classen 1955; 1956). But there are problems here too. This change from communal to private in the traditions of keeping documents would itself be significant and demand explanation. Though the Epistulae Austrasiacae probably originated in a royal archive in Metz and the trial of Egidius of Reims shows royal and episcopal archives in action, there is little or no other reference to municipal acta in the north, and Wood (1986:13-14) shows that the acta further south in Le Mans had broken down by 616. It is difficult to attest the continuation of any other late Roman urban institutions in northern Gaul between c. 400 and c. 600, making the continuation of gesta municipalia questionable. Furthermore, why should the archives have begun to die out only in the seventh century? Consideration of a broader range of evidence questions their survival still further (ch.9, below). Early seventh-century Lex Ribvaria manifests a particular concern with written documents. Whereas transfers of land in the Pactus Legis Salicae are verified by witnesses, and governed by precise rituals (cp. PLS 46, 50), Ripuarian Law talks of producing authentic charters or deeds (Lex Ribv. 59, 62, 63, 70). Dowers were confirmed by written document (ibid. 41) and 47

SOCIAL ORGANIZATION

slaves manumitted by charter {ibid. 61). Fouracre (1986), though showing that written evidence was far from being the only admissible form of proof, demonstrates this use of, and demand for, written documents in practice. The formularies, which appear in northern Gaul in the late seventh century, include precisely such things as charters (Marculf II.2-6), and deeds of immunity {ibid. 1.3-4), manumission {ibid. 11.32—3) and dower {ibid. 11.15). The chronological correlation between the first detailed legislation about charters, the earliest surviving Frankish examples of such documents, and the appearance in the north of formularies argues that the increase in the survival of documents is significant, representing real social change. Fairly widespread literacy and use of written administrative documents in the sixth century did not lead to a perceived need to preserve documents for their own sake. There does not seem to have been an idea that written deeds were of long-term use in matters of property and landholding (which may be similar to the Roman situation; Classen 1955:34-5). This interpretation can be challenged but it will be retained as a working hypothesis, and receive further attention in chapter 9. The seventh century thus saw an increasing concern to keep written records of property ownership and transfer. This, I suggest, represents a fundamental change in society. As was mentioned above, the bestowal of fiscal lands in the sixth century seems to have been a temporary and personal contract, revocable at will by the king, and subject to royal confirmation when the original grantee died. Even at other levels of society, the use of public rituals and living witnesses manifests a particularly temporary view of property ownership. Anyone who called a royal charter a forgery and could not back up the accusation was subject to either capital punishment or penal slavery {Lex Ribv. 59.iii). Land held by charter was now distinguished from other forms of landed property. If a deed confirmed a landowner and his or her descendants in perpetual possession of a specified estate, then, as long as any conditions set down in the original deed pertained and were adhered to, perpetual ownership was exactly what was meant. As Wormald (1984:22-3) has argued for early Anglo-Saxon England, the appearance of the charter in the seventh century represents more permanent ownership of, and the right to alienate, at least certain types of land. Security of land tenure was increased and, with it, the independence of the landholding elite. The attendant changes in views of heredity, property, time and one's place in the landscape should by no means be ignored. Increased aristocratic independence is also suggested by the fact that (even though the law discouraged the practice) lords evidently answered for their retainers in certain legal circumstances {Lex Ribv. 35). Chlothar II's 614 Edict of Paris, which shows several similarities to Lex Ribvaria in this, reveals that some lay and ecclesiastical lords had acquired rights to administer justice on their estates (c. 14) and were involved in the appointment of judges (c. 19). What was earlier termed the 'horizontal' application of royal law was breaking down. Consideration of the mechanics of taxation and military service reveals similar processes. Seventh-century taxation is difficult to study but by the early decades of the century it seems that local landlords had appropriated many forms of Roman taxation, passing on much more 48

SOCIAL ORGANIZATION: DOCUMENTARY EVIDENCE

limited dues to the king (Goffart 1982a). Some lords acquired immunity from certain dues and public duties such as the maintenance of royal agents or the hospitality due to people on their way to the royal court {Lex Ribv. 68). The decline of taxation is perhaps not surprising given the spread of 'Frankish' ethnicity and the association of Frankishness with tax exemption (Marculf 1.19; Goffart 1982a:21; James 1988a:193; Durliat 1990:109 challenges this). The documents collected in Marculf's formulary show the realities of immunity (e.g. Marculf 1.3-4). The army, too, is less often spoken of as the army of the Franks, until Charles Mattel's and Pippin the Short's campaigns against Gascons, Saracens, Lombards, Alamans and Bavarians in the early eighth century. References to military service in the narratives are few and oblique. As before, though, there is a sense in which taxation and military service were alternative obligations.11 In Dagobert's 631 campaign against the Wends, launched from Metz, the Austrasian army was joined by Burgundian and Neustrian scarae (bands) of selected men, under their dukes and counts (Fredegar, Cbron. IV.74). Whether or not this is indicative of a more socially selective form of military service is unclear. By the later seventh century, individual aristocrats, such as Pippin CIP, raised exerciti (Fredegar, Continuation 3 and 5) from their supporters (socii) in Austrasia. Ebroin, the villain of the early chapters of the Continuation of Fredegar's Chronicle, raised 'a mighty force of followers' (multo comitatu excercituum; ibid. 2) for his invasion of Francia (by which, as before, is meant the north). Immunities which forbade the king's agents from entering estates left the lord to raise any necessary military contingent; such at least is the impression given by the fact that the immunist could collect the fine for non-fulfilment of military service. This led to the ability to raise substantial forces from friends and followers and to a 'privatization' of the army, spelling the end of 'horizontally' levied military service. Armies were raised Vertically' down a chain of lords and their dependants. Thus, as stated, the losses suffered in Charles Martel's wars against the Frisians and Neustrians (715), were discussed in terms of noblemen rather than 'strong men', such as had been those lost, a hundred years before, in the battle of Toul (612). When, in 741, Pippin the Short campaigned in Burgundy, his army was composed of ca multitude of magnates and a great band of their followers' {multitudine primatum et agminum satellitum plurimorum; ibid. 24). During the seventh century, as is well known, royal authority broke down and, with it, kingly control of patronage, though warning notes have been sounded against taking the weakness of the later Merovingians too far.12 By the second half of the century patronage was controlled as much by the mayors of the palace as by the kings. The situation reached its apogee in 708-15, when Pippin 'IF installed his grandson Theudoald as puppet mayor of the palace to the puppet King Dagobert III (LHF 50; Fredegar, Continu11

12

Cp. Fredegar's Chronicle IV.74, where the Saxons offer, in effect, to perform military service in defending the Franks' Wendish frontier instead of paying the tribute of 500 cows hitherto due annually to the fisc. Inter alia, Fouracre 1990; Fouracre 1986 makes valuable points about the importance of the royal court; Gerberding 1987:63 and 109-14 has significant implications for the power of Clovis II and Childebert III respectively; Wood 1992b:21-2.

49

SOCIAL ORGANIZATION

ation 7; James 1982:151). The extent to which aristocratic families now controlled ecclesiastical patronage in the north, hitherto seemingly shared between the king, the local community and the church itself, is well known, through many studies of Merovingian Kirchen- and Klosterpolitik (cp. Prinz 1965). Seventh-century politics after Dagobert Fs death (638) were increasingly dominated, as sixth-century politics had never been, by struggles for power and control over the royal courts between rival aristocratic factions. In this context it is not surprising that the relationships between king and aristocracy, or nobility as at least a part of it was by now, were the inverse of those of the sixth century. We saw earlier how kings such as Childebert II could arbitrarily strip of their power and even execute aristocrats like Magnovald. When Childeric II tried to revive such powers in 675 and ordered Bodilo, significantly termed a noble by Fredegar's continuator, to be beaten, he was swiftly assassinated along with his pregnant wife {LHF 45; Fredegar, Continuation 2). In chapter 1 we noted that this assassination resulted in an effective interregnum of six months. Earlier, the Austrasian mayor, Grimoald, had placed his own son, Childebert, on the throne for six years or so. Although, as Gerberding (1987:83) rightly stresses, the nobility still needed a king to provide a figurehead and focus, the nature of their relationship with, and need for, the king differed profoundly from that of their sixth-century predecessors. This is shown by even the most careful studies of the seventhcentury nobility, concerned to rectify the traditional picture of late Merovingian impotence. The LHF's author glossed Gregory's earlier account of Merovingian royal politics to make the kings consult with their franci before major undertakings (Gerberding 1987:33-5). This is clearly an early eighthcentury perspective on the nature of politics. In the region of Metz, it is only from the late seventh century that we can move from the theory of the laws, the circumstantial evidence of the narratives and saints' lives, and from the discussion of general developments on a wide geographical basis, to look at the realities of landholding and aristocratic power. Before the end of the Merovingian period, our evidence comes almost exclusively from the Wissembourg cartulary. Sprandel (1961:54-5), working from the witness-lists of these documents, concluded that they testified to the existence of a class of free commoners under changing leadership, but his conclusion needs some modification. The geographical spread of the estates suggests that some families at least could be described as potentes qui per divisa possident, to borrow Chlothar II's phrase (Edict of Paris, c. 19). Gundoin's family gave extensive lands in five villae13 around Einville in the Sanon valley, as well as in the upper Sarre valley at Biberkirch (Wiss. 223, 226, 228-9, 239-40, dated between 699 and 715). The contemporary family of Weroald, son of Count Audoin, held estates in several villae between the middle Sarre and the Vosges (Wiss. 231, 233, 241-3, 256, dated between 700 and 721). Similarly, the family of Chrodoin (d. before 21 March 726-7), whose dealings with Wissembourg can be traced well into the Carolingian era, gave estates in fairly widespread villae in the Sarre valley, the 13

See below, ch.5, for discussion of the meaning of this word.

50

SOCIAL ORGANIZATION: DOCUMENTARY EVIDENCE

Vosges foothills and possibly further afield (Wiss. 194-5, 224, 227, 232, 247, 257, dated between 712 and 736). The witnesses of these three families' charters show no very clear patterns of association. Constantine, Dructimund, Frumoald and Charoin witnessed charters of both Chrodoin and Weroald in 712. Dructimund and Charoin witnessed two of the Gundoinings' charters three years later. Chardoin and Theudo also witnessed these charters (and those of the Gundoinings for 699) but are attested in the witness-list of a 713 charter issued by Weroald. In 718 they witnessed a charter of Chrodoin and were joined by Baldoin, who had witnessed a charter of Weroald in 713 and one of the Gundoinings of 706-7. These witnesses were not involved in any exclusive ties with any of the major families. This, presumably, led to Sprandel's conclusions. However, there are problems. The witness-lists seem to be the most problematic parts of the Wissembourg documents. That of Wiss. 246 (Wissembourg, 29 June 775) includes the somewhat suspicious pair Romanus and Francus. More suspiciously, the witness-list to Wiss. 235 (Wissembourg, 1 Dec. 741) contains, after the first three names, those of the witnesses to the previous charter in the collection, issued twenty-nine years earlier. None of these men is otherwise known after 715. More importantly, there are reasons to believe that Ermbert and Otto (the sons of Gundoin), Weroald (son of Audoin) and Chrodoin (son of Petrus) were members of a particular noble faction, if not more closely related. In addition to political historical indications (see ch.l), Audoin is in one charter referred to as Otdo, like the son of Gundoin. The latter is occasionally referred to as Ottun or Odon, suggesting that the two shared the same name. The three families also held lands in many of the same villae. Apart from Audoin, two other counts are associated with them. Counts Willibert and Adalchard signed Weroald's charters in 713 and 721 respectively. In 720, Adalchard also held as benefice the lands which Gundoin's sons gave to Wissembourg. It can be proposed that we are dealing here with closely allied families, or even with different lines of the same high-prestige family. As discussed in chapter 1, this group of families may have provided the principal local opposition to the Pippinids and Arnulfings in the later seventh century. The charters reveal the power of people like the Gundoinings. In, for example, Wiss. 240, Ermbert son of Gundoin gave ten named properties to Wissembourg; in Wiss. 194 Chrodoin son of Petrus gave nine named properties; the forged foundation charter of St-Mihiel (St-M. 1) records that the Gundoinings' postulated relative, Count Wulfoald, held twenty or so estates. Bequests included slaves (see below) and churches: the Gundoinings controlled the church of SS Martin and Hilary at Einville-au-Jard; Chrodoin owned St-Martin, Kirch-Berg and Waldhambach church; Samuel and Benedict, sons of Chroccus, held St-George, villa Teurino and St-Remigius, Bourscheid. The splendidly detailed charter Gorze 19 (below, ch.5), illustrates the income which the control of local churches produced for aristocrats by the later eighth century. Aristocrats like Duke Theotchar (Wiss. 213) and Count Wulfoald (St-M. 1) also controlled forms of manufacture, such as salt-pans on the middle Seille. One feature which is worth stressing, however, is the absence of aristocratic rural monastic foundations in the region of Metz, which thus 51

SOCIAL ORGANIZATION

contrasts significantly with neighbouring areas. St-Mihiel, twelve kilometres from the south-western extremities of the diocese of Metz, is the exception, founded on lands confiscated from Count Wulfoald (Gauthier 1980:397; Dipl. KaroL 8). All other known rural houses (St-Avoid and Gorze), were founded on episcopal initiative. Only urban houses, such as St-Pierre-aux-Nonnains and Ste-Glossinde (but not, pace Gerberding 1987:100, St-Arnoul), can be argued to have been aristocratic foundations. Benefices appear in the region of Metz in 718-20 (Wiss. 195, dated Rimsdorf, 18 May 718; Wiss. 267, Biberkirch, 12 July 720). The term beneficium appears as a sudden cluster of references (in 718-21 and 736-9 west and east of the Vosges respectively), followed by a sparse but steady trickle thereafter. The use of benefices was, it seems, well established by the time that Gorze Abbey was founded in the 740s. The significance of this requires attention. The relatively sudden appearance of the term14 might quite simply stem from the introduction of a new model charter into the scriptorium. Beneficium is a Roman term (cp. Classen 1955:20) and may be found in royal charters of the early seventh century, but in a very different sense, as 'benefit' or as the 'benevolence' of a gift (cp. Pertz, Royal 85). This is its usual sense in Marculf. However, in practice, the term begins to be used in a new way, to describe a form of landholding, at about this time across northern Gaul. The earliest such usage in Pertz's collection is Mayoral 10 (dated 2 December 719). By the middle of the eighth century, references begin to the lex beneficium (Gorze 7, dated 1 Jan. 761). Legal references to benefices also begin at this time (cp. Pippin I 'Decree concerning Marriage Issued at Compiegne', 757, c. 9; Boretius (ed.) 1883:38). Major secular landholders, dukes and counts gave land to lesser freemen as benefices (Wiss. 9, 10, 58, 60, 162, dated 735-84); this was not a peculiarly ecclesiastical landholding arrangement. I would argue that the appearance of beneficium at this time reflects a need to describe adequately a particular relation of landownership, and probably a new one. This is, again, open to debate, but it may be retained as a hypothesis and tested against other evidence in chapter 9. The vocabulary of the benefice reflected the personal rather than economic nature of the transaction, the intimate relationship between grantee and grantor and, nevertheless, the dominant position of the latter. This created a landholding context into which the terminology of commendation could be fitted. The often associated term precaria similarly has Roman origins. It appears in Wiss. 229 (Wissembourg, January 706/7), and is already found in Marculf's collection and elsewhere, albeit not always in quite the sense it has in the eighth century. It is usually a simple arrangement whereby a donor receives back the usufruct of the land given. It has been suggested that this, especially where the gift is to a 'family monastery', was simply a means of protecting a family patrimony against its disintegration through divided inheritance. This interpretation perhaps gives added significance to the spate of precarial grants by the Gundoinings and their allies in the critical period 14

We have twenty-seven or so Merovingian charters from the region which predate its appearance, and Marculf's formulary, though usually using beneficium in its earlier sense (see below), uses it twice to mean this kind of land-tenure; Marculf 11.39-40.

52

SOCIAL ORGANIZATION: DOCUMENTARY EVIDENCE

700-20. This relationship appears much more rarely outside ecclesiastical contexts. Also visible on the local level is the break-down of royal patronage in the later Merovingian period. The Wissembourg cartulary reveals two or three counts of (or in) the Pagus Sarroensis, a duke in the Pagus Salinensis and another count (Audoin, the father of Weroald) whose territorial jurisdiction is unclear. In the early Merovingian period, there was one count per civitas (and, as we noted, even this institution may not have existed at Metz), and dukes ruled even broader areas. It was noted that, as a result of the paucity of offices and of frequent redistributions of titles, the kings held significant power over their aristocracies. By the last century of the Merovingian period such offices had proliferated, and some families (such as the Etichonids in Alsace) had managed to acquire control over certain of them. In the region of Metz, Adalchard, count in the Pagus Sarroensis in the second and third decades of the eighth century, may have been related to Count Adalramn, an official on the other side of the Vosges in the same decade.15 The Wissembourg documents also show that deceased counts were referred to by their old title, perhaps suggesting that such titles had become more permanent than in the sixth century. Offices and titles had multiplied, and control of their distribution had slipped from the king to his senior palatine officials and even into the hands of individual families. Throughout the later Merovingian period the nature of the elite was constantly developing. By 750 the local aristocracy had acquired an impressive degree of authority over land and men. Their power reached deep into the daily lives and even patterns of worship of lesser folk, and their control of estates and tenants was assured and independent. As a result, the personnel of the aristocracy remained more constant and, from the early seventh century, a group identity was fostered; legal recognition and privilege was gradually accorded to the elite. During the last century of the Merovingian period in the region of Metz, it seems reasonable to talk of at least the more powerful aristocrats as a nobility. The free in the seventh and early eighth centuries To recapitulate, seventh-century legal material gives much more recognition to the existence of freemen in relations of dependence upon other freemen. We have already encountered the satellites and ingenui in obsequio of Lex Rihvaria. These seem to be followers in immediate attendance upon their lord. These clauses of Lex Rihvaria give legal recognition to aristocrats' creation of dependent followings outside the usual structures of political power. This might also be suggested by the general absence of contubernia in Lex 15

Wiss. 41. Ebling (1974:32), despite suggesting no relationship, makes Adalramn a count of the Pagus Sarroensis, which would make kinship even more plausible. However, this relies upon a location of the land-grant in question {villa Audoino) west of the Vosges, perhaps at Einville-au-Jard (Meurthe-et-Moselle), Audviller (Moselle) or Ottwiller (Bas Rhin). GlSckner & Doll choose Einvilleau-Jard out of these but argue persuasively that, although Einville was a villa Audoino, the subject of Wiss. 41 is to be located at Edesheim near Landau. Einville-au-Jard was in any case in the Pagus Salinensis. Audviller and Ottwiller are not ruled out.

53

SOCIAL ORGANIZATION

Ribvaria; they are replaced by references to hariraida, which may imply more a leader and followers than a band of equals. If land was given to satellites or ingenui in obsequio as reward for service, the relationship between grantor and grantee seems to have remained informal. We can examine seventh- and eighth-century freemen through the, admittedly problematic, medium of the charters. The Wissembourg documents, as Sprandel noted, do not permit any conclusion that the witnesses of the Gundoinings', Weroald's or Chrodoin's charters stood in any formal or exclusive relationship of dependence upon them. What we may have is evidence of an aristocratic faction and its supporters, those over whom they had the power and influence to call to an assembly in their support; it does not look like a group of lords and their followers and retainers. Economic independence is shown by the fact that Charoin/Heriwin (who witnessed charters of, amongst others, Weroald and Ermbert son of Gundoin between 712 and 717) was himself able to dispose of his paternal alod and extensive Alsatian lands in favour of Wissembourg (Wiss. 18, 27 Feb. 725/6; Wiss. 1, 15 June 742.). Other references to land of paternal and maternal inheritance might suggest that even the less powerful donors to Wissembourg (and, later, to Gorze) were in possession of landed inheritances of which they had the power to dispose at will (but see Reynolds 1991 for important reservations about this kind of conclusion). The charters suggest that before the mid-eighth century, the usual means of land transfer, apart from dower and dowry, were sale and outright gift. Even these donors, however, were hardly paupers. Further down the social scale we find references to tenants (accolae) who may have been free. These drop out of the formulae of the Wissembourg charters for Lorraine by the 730s, though they persist into the 740s in the Wissembourg cartulary's 'Rhenish' documents16 and until the end of the century in the Gorze charters. In this cartulary, from about 800, they are replaced by references to lands c held' by X (again, the freedom of X and the nature of the tenure is usually left unclear), to X holding land as benefice, and to mansi ingenuiles (cp. Gorze 19). It is difficult to know on what terms lands were 'held' in this period. Rents expressed as money could mean all kinds of movable goods and produce (cp. Campbell 1989:27, and Gorze 19 as a practical example); servitium could simply mean rent. In the Merovingian documents of the region of Metz, the mention of angaria (messenger services) in Wiss. 267 is the first reference to rent in services, and is unique. Formal, specified services attached to land tenure appear gradually through the eighth century. The freemen of Gorze 19 (770) paid rent in kind and were responsible for the upkeep of ten perches of enclosure. The use of the terms jornales (the land which could be ploughed in a day's work) and carradas (possible cartloads of produce) to measure land might hint at labour services. In the Wissembourg charters, both terms only become common after 740, and especially after the 780s, during which decade they were used particularly frequently. The services listed in the mideighth-century Lex Baiouariorum (1.13), which include sowing and har16

They reappear there in a burst of references in 774-5.

54

SOCIAL ORGANIZATION: DOCUMENTARY EVIDENCE

vesting andecingae of arable and reaping one arpent of meadow, support this interpretation. In 795 Deodatus agreed that the rent due from his benefice in St-Julien-les-Gorze, Higmarega (St-Baussant) and Sontonega (PXonville) would be that he (or at least his men!) would plough, sow and reap one;ornalis of arable (Gorze 34, dated St-Julien-les-Gorze, 21 Dec. 795). One of the most important implications of the use of term beneficium was that the personal nature of the transaction gave it some degree of 'honour' and marked the holder of a benefice out from the occupier of a simple free tenement. The Romans constitute a further, albeit shadowy, element among the free population. However, as mentioned, in the seventh century, Lex Ribvaria made the lesser legal status of Romans much more rigid, and equated them with king's or church freedmen; mixed marriages were regulated, offspring descending to the Roman status. 'Roman' status is very frequently equated with manumission of slaves and half-free so that Roman law is largely associated with the half-free. A slave might be freed, but making him or her a Roman denied equality with his or her previous owners. The unfree classes in later Merovingian society The sources give the impression that the social groups not considered fully free multiplied in the later Merovingian period. We may begin with seventhcentury law. Rent-payers {tributarii; Lex Ribv. 65) were a kind of manumitted freedmen, with a low, thirty-six-so/zd^s wergild. This might support the idea, mooted earlier, that laeti were settled on tenements away from the estate centre. Nevertheless, this new rent-paying category is interesting. The half-free remain much as before, although two new classes, king's freedmen {homo regius) and church freedmen {ecclesiasticus', or homo ecclesiasticus) are added (cp. ibid. 61.xi); their wergild is 100 solidi {ibid. 9-10), the same as a Roman in sixth- and seventh-century law {ibid. 64.ii), or the earlier pueri regis. The extent to which the half-free are dependants of others is made even more explicit than in the PLS. By 797 a document of manumission (Wiss. 68) refers to freedom from servitium or libertinitatis obsequium (the service of a freedman; the formula is from Marculf 11.32). Much of the law concerning the half-free relates to manumission, proofs of half-free status and accusations of illegal manumission, or legal procedure pertaining to them {Lex Ribv. 60-l.i-viii, 64-5). The peculium, the belongings of a half-free man or woman, or the fruits of their labour, remained the property of those upon whom they depended {ibid. 61.i, for church freedmen). If king's or church freedmen died without heirs, their property passed to the fisc or the church {ibid. 6O.iv, 61.iv, 64.i). There appears to be a far wider range of law pertaining to these categories, although the nature of the barrier separating them from the free seems to have related purely to legal procedure. The half-free were a dependent, peasant population, but, to all intents and purposes, it seems that their lives might differ little from that of legally 'freer' inhabitants of rural north-east Gaul. The pueri of the PLS and sixth-century narrative sources are almost entirely absent from Lex Ribvaria. The one appearance of the pueri regis relates to the killing of a count of such status {Lex Ribv. 54.ii). Though 55

SOCIAL ORGANIZATION

modified, it is very similar to PLS 54.ii, and possibly something of a fossil. However, VA 12 refers to apuer in the service of a Frankish magnate called Noddilo. Here the puer is a household retainer and relative; there is no implication of servile status. On the other hand, the pueri encountered by Gaugeric near Famars (VG:286) are indubitably slaves. The 'boys' (pueri) of the sixth century have been replaced by the 'men' {homines) of the seventh, not simply king's and church freedmen, but also freemen in obsequio and satellites.

Lex Ribvaria s discussion of slaves is different from that of the Pactus Legis Salicae. Royal and ecclesiastical slaves are distinguished from others (Lex Ribv. 61.ix, 61.xx). As before, certain laws demonstrate that the slave was still viewed as a form of property (ibid. 75.iff, 78), but it may be that the institutionalized lot of the slave was improving. There is no mention of a slave suffering corporal punishment; it seems clearer in Lex Ribvaria that slaves will pay fines, whilst, if they are believed by their owners to be innocent, their master may swear an oath on their behalf. Only in cases of the murder of a slave by a slave, or of the injury of a freeman by a slave is it made explicit that the lord pays the fine (ibid. 23, 29). The fine in both cases is thirty-six solidi, which might be seen as more than a slave could typically pay; the other fines mentioned are well under this amount. This is very different to the procedures envisaged in the Pactus Legis Salicae (PLS 35). If the master cannot tell the slave's guilt or otherwise, he must deliver the slave up to ordeal by boiling water. Here is the major difference between royal or ecclesiastical slaves, and other slaves. The difference between king's slaves and other slaves in the PLS is simply that the king's property was more valuable, whereas Childebert IFs Decrees (3.vi) maintained no difference between church or fiscal slaves and any other; in Lex Ribvaria the former may swear oaths and answer in court themselves. The distinction concerns legal rights. Although slaves were now considered capable of paying small fines without their masters' consent, no one was permitted to buy anything from a slave (Lex Ribv. 77); their peculium was ultimately the property of their masters. There is detailed legislation about intermarriage between all these different unfree and half-free classes, and between them and the free (Lex Ribv. 61.ixff). As before, slave-free marriage results in both partners, and their descendants, becoming slaves. In all cases of intermarriage, it is now specified that any offspring will descend to the lower legal status, but, unlike earlier law, Lex Ribvaria (61.x) allows for instances where the higher status of one of the parties is retained for the duration of their lifetime. The family of a freewoman who wishes to marry a slave may offer her the choice of 'sword or spindle': choosing the sword forced the woman to kill the slave; choosing the spindle meant she remained with him in slavery (Lex Ribv. 61.xviii). The important difference between seventh- and sixth-century law in this respect is Lex Ribvaria's concern with the status of descendants. Lex Ribvaria is far more concerned with manumission than the PLS (Lex Ribvaria 60, 61.i-viii, 64), and, as mentioned, with subsequent proofs of free, 'Roman', 'rent-paying' or freed status. A slave may now be freed by charter, 56

SOCIAL ORGANIZATION: DOCUMENTARY EVIDENCE

as well as by 'penny-throw'. Seventh-century law lacks the extended enumeration of different types of slave, found in the PLS, which may be significant. Lex Ribvaria allows the suggestion that perhaps the conditions of slaves had slightly improved. There can be little doubt that the legal difference between slave and free remained the most important institutionalized barrier in society. With a more varied, and evidently more important, group of 'half-free' classes between them and the free, and a perhaps more blurred boundary between some at least of the half-free and the lower free classes, perhaps this barrier was becoming less important in daily life. This impression might be tested by examining Marculf 's formulary and the charters. Marculf's documents shed a very immediate light on the theory of the laws. The deeds confirm the absolute authority of the owner of the slave. this slave I now give over to you to possess and from this day you freely have the power to have, hold and do whatever you want to him/her (Marculf 11.22) if I do not do this, or if I wish to leave your service in any way, or to petition another lord, or to take on his business, you have licence to impose whatever kind of discipline you want on me, or to sell me or to do whatever you want with me. (ibid. 11.28).

Patterson's 'relation of domination' is still clearly attested. The documents, as shown above, show that sale of these individuals and complete rights over their persons was still practised, and presumably slaves could be and were sold away from their families. Marculf also includes deeds of manumission (e.g. Marculf 11.32—4). These usually imply that manumission made a slave a freeman. The phrase commonly used is that the slave will be cfree as though born of free parents'. No mention is made of half-freedom or freed status as such, though reference to dependence upon churches or 'anyone else you choose' (ibid. 11.32) shows that it was common for freed slaves to enter into some other protective relation of dependence. The charters attest a wide array of unfree categories: mancipia, servi, ancillae, accolae (who might be free; see above), vasalles, puellae, homines, liberti and liti are all mentioned as evidently dependent groups of people in seventhand eighth-century charters of the region. This might confirm the diversity suggested in Lex Ribvaria. The complete absence of references to coloni in the Merovingian region of Metz is worth noting, however. The liti encountered in Gorze charters of the later eighth century (they disappear by the ninth) are probably the half-free of the law-codes; these charters constitute their only appearance outside the legal material in the region of Metz. There is one reference to liberti (Wiss. 45, of 20 June 719); they are granted with their peculium, confirming the laws' statements about this. 'Thepuellae (Wiss. 17, of 19 June 739), homines (Wiss. 12, of 731-9, and 52, of 27 May 742) and vasalles (Wiss. 17), met with in earlier eighth-century 'Rhenish' Wissembourg charters, all seem to be household servants. Their degree of legal freedom is unspecified; they could be free or slave. That they are granted with the domus in which they work does not necessarily imply unfree status. It might simply be a right to their work that was being bestowed. 57

SOCIAL ORGANIZATION

More can be said about the mancipia of the charters. This category is that which appears most commonly in the documents; by the ninth century it has all but completely replaced the others. A large number of mancipia are named. That they and servi are frequently named whereas accolae never are might suggest that rights over the person and his or her descendants are being bestowed in these cases, while in the case of accolae it is rights to land and the rent or labour due from it which is being given. This might suggest that accolae are indeed in some sense 'freer' than mancipia or servi, whatever the de facto similarities of their condition. In the charters mancipia seems sometimes to be used as a noun incorporating male and female slaves {servi and ancillae respectively). Servus and ancilla seem usually to be used as singular nouns; neither is as common as mancipium. Mancipia could be given away either with the land which they farmed or without it. It is the latter (for example, Wiss. 241, of 18 June 737) which goes furthest to proving a real lack of freedom. Gift with the land might merely indicate a gift of labour and other dues from the land worked by the specified mancipia', gift of mancipia away from land indicates the gift of human beings and some fuller degree of control over their persons. Wiss. 183, of 769-84, records the sale of a slave called Amalbert for two pounds, showing such transactions in practice. That mancipia held no property outright continues to be attested by gifts of mancipia and servi with their peculia (cp. Wiss. 1, 13, 15, 16, 46, 54, 96, 100 and 125, ranging in date from 695 to 788). There are some indications of an amelioration of status. The mancipia given away in Wiss. 241 are listed by family: for example, Chroacar and Gaila his wife {cum uxore suae Gailane), and their children {cum infantis eorum). Grouping of male and female names within other lists of mancipia suggests that this may have been common practice. Children too are grouped with their parents. In 788 a 'Rhenish' Wissembourg charter (Wiss. 102) refers to a grant of mancipia cum uiris et mulieribus uel infantibus. Age and sex were now to be distinguished within the general category of mancipia. All this suggests that the attitude towards slave families in late Merovingian north-eastern Gaul was not, in practice, quite as harsh as Bonnassie argues (1991b:23-5). His argument, though, derives solely from the 'barbarian' lawcodes (to which he adopts a singularly diachronic, 'aggregating' approach). His argument is certainly plausible in the earlier period, but even the laws suggest (as argued above) some slight improvement in legal conceptual status between the sixth and seventh centuries. There are, nevertheless, early Carolingian instances of mancipia being granted away from their family (Wiss. 210, of 24 Nov. 783, a gift of four female mancipia), showing that unfree families could still be broken up by their owners. The listing of children may have less to do with any more 'humane' recognition of slave families than with a concern to make clear the unfree status of the descendants of mancipia, such as is attested in Lex Ribvaria and in Marculf's formulary. But even so, practical exigencies had forced a recognition of slave families; the slave is no longer quite Patterson's (1982) 'genealogical isolate'.

58

SOCIAL ORGANIZATION: DOCUMENTARY EVIDENCE

The Wissembourg cartulary includes references to manumission, closely modelled upon Marculf 's formulae. Wiss. 68 (?Ottwiller, nr Drulingen, Bas Rhin, 7 Jan. 797), from the region of Metz, bestows freedom upon eight servi vel ancillae. These are freed from servitium and libertinitatis obsequium (as noted above), and are to be subject to God alone. Nevertheless, Wissembourg Abbey is given mundium (legal guardianship) over them all and they are to pay the abbey four denarii a year, at Martinmas. Not unexpectedly, these slaves did not make a straightforward transition to fully free status. Freed slaves would have been extremely poor and would have found life difficult without a powerful protector. By the later eighth century, there are also references to specified services to be performed by mancipia. Wiss. 63 (24 June 774) specifies that mancipia perform three days' work per week, with a further forty days per year on major works if necessary {Lex Alamannorum 22.i specified similar service in the early eighth century). By this time, those mancipia who lived away from the estate centre on unfree tenements had their owner's demands on their time and labour restricted by custom. By 830, Thiethar and Wolfgart, two mancipia in Biberkirch (Bas Rhin) were granted freedom from all labour services in return for two denarii per annum, and Wissembourg Abbey's agent at Waldhambach was to go to their house to collect it (Wiss. 51, and cp. Wiss. 198). This is a significant improvement in the slave's lot, and shows how far the barrier between slave and free had been blurred by the late eighth century. By this time it seems to have largely concerned legal procedure. In practical economic terms, the poor free, the half-free and the unfree were merging into a single dependent peasant population. Hence it is no surprise that by the ninth century, in our charters, mancipia is almost the only term used to describe dependants. In the region of Metz, this change would would appear to date to the period c. 775800, which tallies with Kuchenbuch's (1978:355-82) conclusions on the neighbouring region around Prum. Goetz (1993) has usefully highlighted the opposition of legal, juridical, 'conceptual' treatments of slavery, and practical, everyday socio-economic exigencies and de facto status. Which of these determined the conditions of slaves? Without adopting any crosy' view of early medieval slavery, it must be said that much work on slaves has been excessively concerned with legal formulations and has perhaps painted too uniformly bleak a picture. As has been argued above, the habitus probably moderated the legal status of slaves considerably even in the sixth century; Lex Ribvaria hints at a weakening of these institutionalized barriers in the seventh century. Nevertheless, that the diversity of different unfree and semi-free classes attested in seventh-century law is also witnessed by documents of practice, the charters, argues strongly that the institutionalized, legal identities were of primary importance until the late eighth century. The terminological change noted above might suggest a final triumph of socio-economic similarity over juridical difference, of social assimilation over legal distinction, in the decades around 800. One cannot call it a triumph of serfdom over slavery; the legalities of unfree condition remained in force, and were exploited for a long time

59

SOCIAL ORGANIZATION

to come. Such a straightforward, binary opposition is, as Goetz argues, too simplistic.

'Vertical' and 'diagonal' barriers Kinship, marriage and the family Throughout Merovingian Austrasia the kin-group was of dominant importance in defining social organization. The subject has been thoroughly covered in the literature, albeit usually from a diachronic and general European point of view17 and so does not require detailed treatment here. In the earlier part of our period, the Pactus Legis Salicae defines the formal kingroup widely, including the father's and the mother's relatives. This kingroup was supposed to help in the payment of fines (PLS 58) and, conversely, the law delineated very carefully how, and in what order, these people would inherit family property, or receive shares of fines paid to the kin (PLS 44, 59, 68; Murray 1983:chs.9-15). But it is always important to remember Bourdieu's (1977:33ff) useful distinction between official and practical kin. The laws are concerned with official kin. The 'practical' kin may have been much smaller.18 Nevertheless, official kin were important. Gregory (LH X.27) shows how involvement in feud could ripple outwards through the kin-group from the immediate 'practical kin'. In the sixth century we should also note the power of the local community, as witnessed by PLS 45, concerning migrants. Marriage was clearly important as a social strategy in such a society, especially since maternal as well as paternal relatives became part of one's 'official kin'. Marriage would have been a key means of securing allies and protection. The Pactus says little about it, except that betrothal took place as public declaration, and penalized a man who broke off such an engagement (PLS 65a). The laws for betrothal to a widow (PLS 44, 100) underline the point that it is public betrothal which is important. That family approval is vital, as one would expect in these circumstances, is also suggested by these clauses and by those dealing with illegal procurement of children for marriage (PLS 99) and raptus (illegal abduction). Little mention is made of divorce. This will be discussed further below. Presumably, once a man and woman were betrothed, the important alliance was made, and was held to be in force whether or not the parties lived together, as long as neither took another partner. At the end of the century, Childebert IPs decrees (l.ii), deal with incest, which is punishable, in the case of a son marrying his father's wife, with death. The degrees within which incest is prohibited are quite narrow, however. This contrasts with other legal definitions of kin, and may 17 18

Herlihy 1985 contains a useful introductory bibliography; Theis 1976 is interesting; Murray 1983 is essential. This is the conclusion of Theis' (1976) interesting study. There are obvious problems, though. It is unsurprising that authors of saints' lives were unaware of their subject's more distant kin. They almost always knew their subjects (if at all) in their ecclesiastical or monastic setting, and late in life. Where a hagiographer came from the same family as his subject, as with Altfrid's Vita Liudgeri (Lebecq 1990), a deeper knowledge of the kin-group is, not unexpectedly, demonstrated.

60

SOCIAL ORGANIZATION: DOCUMENTARY EVIDENCE

restrict marriage only within 'practical kin'; restricting marriage within 'official kin' would have been too prohibitive in sixth-century rural communities. In seventh-century Lex Ribvaria^ there are important changes in the legal concept of the family. Whereas PLS 58, concerned with what happened when a man could no longer pay a fine, envisaged the formal, ritualized, widening of the circle of his living kin brought in to help pay the fine, Lex Ribvaria (c. 12) says 'let his descendants pay for three generations'. Though the notion of three degrees of kinship remains constant, there is an important change from the kin-group broadly defined in the present to that more narrowly defined but projected into the future. This parallels the law's concern with the legal status of the offspring of mixed marriages, and also the increased concern with projection into the future suggested above, with relationship to the new practice of keeping documents. Lex Ribvaria has few clauses about marriage and betrothal, though it does add an increased concern with written deeds of dower, as noted. At the very end of our period, Pippin I issued two capitularies about marriage (Decree at Compiegne, 757, and Decree at Verberie, 758-68; Boretius (ed.) 1883:3741) which were largely concerned with detailed prohibition of incestuous marriages, and, in contrast to Childebert II's edicts, these are quite broadly defined, being prohibited between partners related within three generations. The collective will of the later Merovingian kin-group could often override the wishes of the individual. According to the second version of the Vita Waldradae (c. 2), when Waldrada's father Leutherius granted additional estates to his daughter's abbey of St-Pierre-aux-Nonnains in the early seventh century, he was murdered by his kin. The story may well be apocryphal but this fear remained into the Carolingian period, as shown by the repetition of clauses safeguarding donations of land against attempts by the donor's family to regain them. The fact that the worth of an individual was very often subordinate to his or her value to the kin-group is further underlined by examination of the vertical and diagonal barriers delineated by gender and age. Gender19

The nature of the social barriers created by gender can only be studied in geographical, chronological and, above all, social context. Much work has been reductive, making gender into a straightforward binary polarity, applicable not only across huge geographical zones and widely differing regions but also at all levels of society. The construction of gender differed, as we shall see in chapters 3-4, across comparably short distances and from one century to the next. The constraints on male and female power and action were differently traced among different hierarchical classes, especially when these became separated by clear horizontal barriers. To speak of 'class 19

The literature on the subject is, though patchy in coverage and quality, voluminous. Affeldt et al. (ed.) 1990, an annotated bibliography, is invaluable. Wemple 1981 is a useful general survey which, without eradicating them, suffers less from the problems of diachronicity than many other works, and wisely restricts the discussion to the Frankish realms.

61

SOCIAL ORGANIZATION

transcending gender' (as Nelson 1990:63 does) is to assume that gender construction is uniform across society (and in this case time too!). For now it is important to sketch out the nature of the 'gender barrier', how it constrained action and how it delineated the space within which women and men could act to further their own aims. As is common, I have used the condition of women to highlight the nature of the gender barrier. Most of these features have been extensively dealt with elsewhere, so their treatment here can be summary. Analysis of sixth-century women must, as ever, proceed from the laws.20 The legal condition of sixth-century women is by no means easy to establish. Past work on this topic has suffered from the usual aggregating tendencies. The legacy of the Rechtsschule is the treatment of all 'barbarian' law as a single corpus. Historians have only recently rebelled against this, and the study of women is still in its thrall. Separating the Frankish codes from those of the other post-Roman kingdoms allows us to see that many of the commonplaces of early medieval women's condition are not supported by the evidence in this region. As everywhere in this work, legal material is only treated as a general back-drop against which we can try to compare more strictly contextualized regional surveys of society. The Pactus Legis Salicae is surprisingly tacit on the legal condition of women, and this is a point which must be constantly borne in mind. Its silence is difficult to interpret, but we should not do this by extrapolating from the Burgundian or other Germanic law-codes. We must keep our investigation rigidly within context. The legal role of women remins unspecified. The Pactus (PLS 44) outlines how someone who wishes to marry a widow must pay a betrothal fee (reipus; ring-money) to her late husband's relatives (Murray 1983:ch.ll). This might imply that a woman lived under the legal guardianship of a man, but only PLS 100, from a later capitulary, makes this clear, when, again in relation to a widow's remarriage, it talks of the payment of achasius (the price of mundium, guardianship, over the woman) to the husband's relatives. The laws are silent about women in legal procedure. Pactus Legis Salicae is almost entirely masculine in its provisions. There is no mention of female oath-helpers, but three clauses {PLS 19.iv, 55.iv and 58.iii) imply that women could pay fines. Women are specified as criminals in only a handful of instances. The only case of violence is PLS 19.iv, where a woman may be accused of causing another's infertility through witchcraft. It is significant that the violence is magical rather than physical and inflicted by a woman upon a woman, and that the crime relates to child-bearing. That apart, female crimes concern illicit sex and marriage, either by elopement or with the unfree. In both cases attitudes changed through the sixth century, and in one, under the influence of Roman law, they undeniably hardened. The csixty-five-title text' of the Pactus underlines its exclusion of women from the law by making the man alone responsible in cases of fornication (PLS 15.iii) and adultery/bigamy (15.i). By the end of the century, though a later capitulary repeated male responsibility for elopement with another's wife (PLS 133), Childebert II 20

Ganshof 1962 is the classic study; see, more recently, von Olberg 1989 and Schmidt-Wiegand 1989.

62

SOCIAL ORGANIZATION: DOCUMENTARY EVIDENCE

decreed in 595 (Childebert's Decrees 2.ii) that a woman who consented to elope with a man should suffer death or outlawry with him, or exile if they sought sanctuary in a church. Initially the Pactus (PLS 25.iv) simply sentenced a freewoman who married a slave to slavery, in the same way as it condemned a freeman who married a slave woman. However, by the time that the third capitulary was added, the freewoman who married a slave was outlawed, her possessions confiscated by the fisc and her relatives given permission to kill her without penalty. That apart, most laws, as mentioned, begin si quis; most criminal activity is not gender specific. That the law's provisions are almost entirely couched in the masculine might be taken along with the Pactus' refusal to accept female culpability in cases of fornication or adultery/bigamy to suggest a conscious exclusion of women from the law; even if a woman does commit a crime, the man under whose mundium she lives will be responsible. Nevertheless, this remains only an assumption. There is no outright statement that women are legally incapable of certain crimes (cp., in Lombard Law, Rothari's Edict 278 and 378, and Liutprand's Twenty-Second Year Laws, Women are not mentioned in the resolution of guilt, but we should nevertheless hesitate before assuming absolute legal incompetence. The Pactus never refers to oath-swearing cwith right hand armed3, which, on comparison with other evidence, might suggest a restriction to males, and never states outright that women cannot swear (again, compare Rothari's Edict 204). That women could pay fines might suggest otherwise. Slave women were treated by the law in the same way as slave men in means of proof and punishment (PLS 40). This might imply that amongst the free guilt was similarly established in the same way for men and women. It might alternatively indicate a different legal construction of gender at different ranks of society. However, the main point of PLS 40 is that, unlike the free, slaves are subject to torture and physical punishment, and (at PLS 4O.xi) its concern is with establishing a female equivalent of castration. The legal condition of women might also be examined through the system of wergilds. A freewoman's wergild is the same as a freeman's: 200 solidi (PLS 24.ix, 41.xv and xvii). When of child-bearing age, this increased to 600 solidi (PLS 24.viii, 41.xvi), three times a freeman's and the same as a free count's (PLS 54.i). Whilst this has been interpreted as indicative of a concern to protect the vulnerable (Irsigler 1969), this argument is easily countered. It is difficult to interpret the laws thus when similar high wergilds are not assigned to old people, widows or female children; the Pactus does give a high wergild to boys before their first hair-cutting, at about twelve years of age (PLS 24.i, 24.iv, 41.xviii). The wergild in any case was paid to the family of the deceased, and therefore represents, in our terms, institutionalized value. The victim obviously derives no active benefit from a high wergild payable upon his or her murder! Grahn-Hoek (1976:30-1) also refutes the claim that the high wergilds for child-bearing women represent a concern for the vulnerable, arguing that the concern was to protect the family line. The 600-solidus wergild thus indicates only a high value to the woman's family based upon her child-bearing capacity. Nevertheless, that apart, it is 63

SOCIAL ORGANIZATION

significant that a woman was regarded as of equal value to her family as a man, which contrasts clearly with the Visigothic legal situation (cp. the table of wergilds given in Bonnassie 1991c:75). The bases of a woman's status can be further illuminated by consideration of crimes against women and their punishment. The Pactus includes a clause about rape (PLS 15.ii); the fine is 62| solidi, compared with 45 for fornication (PLS 15.iii, where, nevertheless, the woman is not considered responsible for her actions). This implies that in matters of sex a woman's wishes were taken into account. Needless to say, this applied only to the free and semi-free (PLS 13O.i-ii); in clauses about fornication with slave women, there is no distinction between cases where the woman consented and those where she did not. Most of the other specified crimes against women can be interpreted, like fornication, as a crime against the family rather than against the woman herself. This is certainly the case with abduction, bigamy and adultery, as already discussed. Other laws deal mainly with the effect on a woman's child-bearing capability (PLS 19.iv; 104.ivff). Clearly the ability to have children is often central to a woman's own desires for her life; here, in conjunction with the high wergild for child-bearing women, it might again be argued that the law has the family's rather than the woman's concerns in mind. This difficulty in establishing whether the laws are dealing with crimes against a woman or against her value to her family applies equally to the clauses about hair-cutting (PLS 24.iii, 97), touching the arm or breast (20), calling a woman a witch (64.ii) or a whore (3O.iii),21 or pulling her hair (104.iii). A woman will not wish to be publicly dishonoured or humiliated verbally or physically but again we might suspect that it is the dishonour brought upon the family which is being penalized. This is difficult to resolve, but in some ways perhaps it does not need to be. The alternatives are not mutually exclusive. Even if the laws do envisage the crime as an insult to the family rather than to the woman, the latter can still use the law in just the same way. Similarly, a woman of child-bearing age could use the protection afforded her by her 600-solidus wergild to derive some active benefit from life. Thus, although the laws ascribe female status in terms of value, we can see how a woman might nevertheless acquire worth. The laws make it clear that a woman could own property. References to women paying fines have already been noted. Other laws show that women received a share of a wergild paid for dead relatives (PLS 68, where the mother of a victim receives one quarter of his wergild). The laws show that a woman's dos remained her property. A woman was thus able to manage an independent property base, upon which she was to live in her widowhood. However, she was not able to dispose of it if she had children (PLS lOO.ii). If she predeceased her husband, he was to manage it on behalf of their children (PLS lOl.i); if there were no children, her family received back two-thirds of the dos on provision of two covered beds, two benches and two chairs, otherwise one third. If the woman remarried, she could take one or twothirds of the dos with her, similarly depending upon the provision of covered bed, bench and chair, and, in this case, the payment of the achasius. It is the 21

Note how verbal insults again only concern witchcraft and extra-marital sex.

64

SOCIAL ORGANIZATION: DOCUMENTARY EVIDENCE

widow who pays the achasius, not the second husband (who, in PLS 44, pays the reipus, albeit to a significantly different group of relatives; Murray 1983:ch.ll). Laws relating to property are complex. PLS 67 suggests that an equal share be given to daughters and sons upon a daughter's betrothal and a son's first hair-cutting. This probably relates only to movable property. In inheritance the general rule was that male heirs succeeded to land in preference to female, but that daughters inherited before their uncles (PLS 108; Murray 1983). Most of the debate concerning female inheritance has centred upon the celebrated c. 59 (de alodis) of the Pactus Legis Salicae. The most disputed clause is that (PLS 59.vi) which reads: cBut concerning Salic Land (terra salica) no portion of the inheritance is for [or 'shall pass to'] a woman, but all the land belongs to members of the male sex who are brothers [or 'but the male sex acquires it, that is the sons succeed to the inheritance'].' Murray (1983:ch.l5 and refs.) effectively excluded the sixth century from the debate over the meaning of the word salica by showing that this word does not appear in the earliest recension of the law. In sixth-century law, quite simply, the deceased's property was divided into movable and landed property. The former was divided equally between male and female heirs; sons postponed daughters in the inheritance of land. Early Frankish law did not define particularly clearly the boundaries within which women could act. We might note the absence of lengthy consideration of betrothal, marriage, divorce and other sexual concerns (which contrasts greatly with Burgundian and Lombard law), and the absence, already mentioned, of explicit statements about female legal competence or incompetence. The gender divide, in terms of possible action, is not explicitly delineated in the Pactus, though it is made clear that a free male, once he has come of age, is entirely responsible for his own actions. Women in sixthcentury northern Gaul may have had rather more space within which to acquire social status than is often assumed. On the other hand, as suggested above, the Pactus' silences might be very deliberate, refusing to countenance active female involvement; they may imply that the gender divide was very rigid, but lay within the jurisdiction of family (in the same way as a master's dealings with his own slaves are never dealt with in the Pactus). There is little within the legal evidence itself to allow us to choose between these alternatives. It will suffice to repeat the point made earlier, that many of the commonplaces about female status among the Franks are based upon filling in the Pactus' silences from Burgundian or other Germanic codes (cp. Wemple 1981:esp. ch.2), and this will not do. However, even the more optimistic reading of the laws suggests that women could acquire prestige in the community most easily within the contexts of marriage, child-bearing and the family ascribed to them by men. Other written evidence from sixth-century northern Gaul is sparse, especially from the region of Metz. Like aristocratic men, aristocratic women could be educated (Riche 1976:218-30), to help mark their distinction from other people, as is shown by Venantius' epitaph for Vilithuta, a 'barbarian' who lived in Paris (Carm. IV.26; George 1992:93-4). Aristocratic women could acquire significant wealth and power through their status as powerful 65

SOCIAL ORGANIZATION

men's wives, as shown by Gregory's vignette {LH IX.9) of Duke Rauching's wife riding through Soissons bedecked in jewellery and accompanied by two troops of pueri. More light is shed upon female status by a story from Paris {LH V.32). A woman had left her husband and was rumoured to be living with another man; her father was challenged by the husband's relatives to prove her innocence or let her die, clest shame of this kind strike down our honour'. He swore to his daughter's innocence on the relics of St Denis, his in-laws refused to accept this, a fight broke out in the church itself, and some were killed. Eventually both sides paid fines and were readmitted into the communion of the church by Bishop Ragnemod. The woman was then summoned to trial, but strangled herself before this could take place. The story has a number of interesting implications. First, it was not the act of leaving the husband (put into an ablative absolute sub-clause) which shamed the family, but the accusation of living illicitly with another man (this would tally with the interpretation of sixth-century law put forward above, p. 60). If this was generally the case, it would contrast starkly with the situation in Burgundy {Liber Constitutionum 45.i); killing adulteresses out of hand was common throughout Gaul {ibid. 68; L//VI.36, VIII.19; GM 69). Secondly, it was her father, who, in these circumstances, was responsible for her legally and swore the oath to her innocence. Thirdly, the stress on honour amongst these powerful families should be noted. This is made explicit by the husband's relatives; the woman herself committed suicide rather than suffer disgrace.22 This tale would support the gloomier alternative interpretations of the Pactus' provisions and its silence with regard to female legal competence. However, it is worth stressing that there is nothing in the evidence from Francia which suggests that men there had any more right to divorce a wife. A feud in Tournai {LH X.27) was sparked by a husband's ill-treatment of his wife, and alleged infidelities. Most statements about gender difference in rights of divorce among the Franks are based upon southern Gaulish or Burgundian evidence, or upon the practices of the Merovingian kings, whose marital customs were hardly typical (see above, p. 37, for reasons). Generally, the only sixth-century north Gallic women visible to us are royal women, queens and princesses. These acquired an impressive degree of power, wealth and even sexual freedom {LH IX.34) but, as has been mentioned, the Merovingian royal family was not typical, and we cannot generalize from it. Gender was clearly constructed differently at this altogether exceptional level of society.23 In seventh-century legal material, there are some subtle changes in the treatment of female status. Perhaps most noticeable is the absence in Lex Ribvaria of a clause on rape {pace McNamara & Wemple 1976:101). There is now one clause covering all forms of extra-marital sex with a free-born girl {Lex Ribv. 39.ii), punishable by a fifty-solidus fine, whether or not the 22

23

It might be significant that Boniface (Ep. 73; Whitelock (trans.) 1979:819) says that the Old Saxons forced adulteresses to hang themselves, but this example is well removed from sixth-century Paris in time and space. I am grateful to Ross Balzaretti for this possible parallel. Female royalty is well covered by the literature. See, for example, Nelson 1978 on Merovingian queens, and Stafford 1983 more generally.

66

SOCIAL ORGANIZATION: DOCUMENTARY EVIDENCE

'girl' consented. This may well represent a change in the laws' attitude to female status; the woman's wishes are ignored; the crime is clearly against her family. Lex Ribvaria introduces clauses requiring oath-swearers to swear c with right hand armed' (Lex Ribv. 37, 69.i) and judicial duels (ibid. 36.iv, 6O.ii-iii, 84), which might imply a restriction to males. Lex Ribvaria 84, however, implies that a woman may bring a suit and answer in court after the age of fifteen, and the choice of csword or spindle' opens the use of judicial violence to women in at least one specified situation. That apart, Lex Ribvaria's provisions concerning women's wergilds (Lex Ribv. 12-15), attacks upon women (ibid. 39.iii, 43), and bigamy/adultery (ibid. 39.i) remain largely as in the Pactus, but there is even less treatment of women in the later code. Gone are the lengthy discussions of bodily mistreatment and hair-cutting, of accusations of witchcraft and prostitution, of magical violence, of widowhood and remarriage, and so on. There are changes in the area of mixed marriages (ibid. 61.ix-xviii); as mentioned, the laws now give greater attention to the status of the descendants of such marriages. The laws, following the later sixth-century decrees, view a freewoman's marriage with a slave in a harsh light (ibid. 61.xviii), allowing the woman's family to give her the choice of csword or spindle'. Note, however, that the family has no right to try to stop a freeman's marriage. A final area of change concerns the woman's dower. Lex Ribvaria 41 distinguishes dower given by charter from other kinds. The former is irrevocably the woman's. Where nothing was bestowed by charter the woman receives a dower of fifty solidi, and may ask, upon her husband's death, for one third of all the things they acquired together. If, however, all the things given to the woman were consumed jointly by her and her husband, she may ask for nothing. How this was determined is not especially clear (but see below). The charter gives the husband the right to limit (or increase) the amount of property alienable, upon his death, from his family to his widow, by specifying her dower in writing. The Pactus' lengthy treatments of a woman's dos are absent from Lex Ribvaria. In questions of inheritance, Lex Ribvaria again largely repeats the Pactus' formulae (ibid. 57.i-iii). Lex Ribvaria (ibid. 62.ix) specifies, however, that a man may not leave any one of his children (son or daughter) over twelve solidi more than he left any of the others. This looks like a similar provision to PLS 67, and may refer to movable property only. The specification that daughters may not inherit 'ancestral landed inheritance' (hereditas aviatica; ibid. 57.iv) represents a new departure. Inherited property was now distinguished from acquests. Sons were given clear priority in regard to the former, whereas the latter was treated together with the movable property. This development is probably more or less contemporary with the insertion of the word salica into clause 59.vi of the Pactus Legis Salicae. Bergengruen (1958) argued that terra salica was land around the household (sola) or estate centre, rather than land belonging to the Salic Franks. Again, there seems to be an increased concern to keep 'ancestral' land within the male line, though it should be stressed that females and maternal kin are still by no means excluded. There seemed to be redefinitions of the gender divide in later sixth-century capitularies, which restricted, or punished more severely, female actions. 67

SOCIAL ORGANIZATION

Certain aspects of Lex Ribvaria appear to continue this trend, but, on the whole, the legal status of women remains puzzling and often contradictory. Though Marculf's formulary originated elsewhere, the Wissembourg charters suggest that it (or something very similar) was in use in the region of Metz by c. 700. Its documents allow us to examine further the legal status of women. They show that, as intimated by Lex Ribvaria, charters and deeds could be used to modify custom. Critical is Marculf 11.12, which leaves an equal share of a landed inheritance, 'whether paternal alod, purchased land, mancipia or chattels or whatever' to a daughter. The formula states that although it is a 'long-standing custom' that daughters should not inherit paternal land (paralleling Lex Ribv. 57.iv and the later recensions of the PLS) the donor regards this as an impiety, and gives his daughter a share equal to that of her brothers. Marculf illuminates marriage too, including a nuptial gift (Marculf 11.15), and reciprocal gifts between husband and wife (ibid. II.7-8). The latter are particularly interesting. They apply in the case of childless marriages and detail how the husband and wife give each other all of their property in case of one predeceasing the other. The deed is entirely equal, and the man and woman are on an equal footing. Moreover, the man gives the woman his alodial land as well as that acquired by other means. These documents shed some light upon the practicalities of Lex Ribvaria 41 (see above). There are also documents about raptus (Marculf 11.16) and illegal intermarriage (ibid. 11.29). In the former, the husband, who has carried off his 'beloved' wife without her family's consent, gives her landed and movable property, which henceforth shall entirely be her property, as though it were a normal dower. The document stresses the crime against the family, and mentions that the penalty for raptus is normally death. This is underlined in Marculf 11.29, which states that the slave who ran off with the freewoman should have been executed and that, if the woman had followed him willingly, she should suffer enslavement and become the property of the slave's owner. Nevertheless, in this document, the slave's owner grants freedom to any children produced by this marriage. Finally, Marculf 11.30 is a deed of divorce, wherein a man and wife, 'discord reigning between them so that they can barely talk to each other', are allowed to separate, with their belongings, and neither's relatives may claim any of the other party's property. This certainly argues that divorce was possible in some circumstances, at least among certain classes of seventh-century society (those with knowledge of, and recourse to, this kind of legal deed). Marculf's documents, if they were in common use around Metz, and so can reflect social realities in this region, imply that the institutionalized aspects of the gender divide (and other social barriers) could be moderated by the use of written documents. Women could own and control property. They appear as parties to wills (Marculf 11.17) and precaria (ibid. II.5), and, according to the deeds of reciprocal gift (ibid. II.7-8), they could alienate such property too. The documents imply a greater freedom of action, including divorce. But, again, it must be repeated that these documents may relate only to a restricted social class, hence the repeated concern with land- and slave-owning. 68

SOCIAL ORGANIZATION: DOCUMENTARY EVIDENCE

In the seventh century, evidence allows us to home in more specifically on the region of Metz. Here the discussion will be kept rigidly in geographical context. Much work has been done on the women of neighbouring, northern regions, especially in the Carolingian period, using the Carolingian family saints' lives for the earlier period, and the polyptychs (see Kuchenbuch 1978 and 1986) for the later. As noted, the areas to the north were geographically different, the appearance there of large, influential aristocratic monasteries contrasts sharply with the region of Metz and suggests a dissimilar socio-political structure. The Merovingian cemeteries of the Trier region also suggest a different social organization (above, p. 7). Moreover, society had changed by the ninth century, when the polyptychs appear. It is dangerous to extrapolate from these studies southwards into the civitas of Metz, and backwards in time from the Carolingian period. Of the seventh-century saints' lives, the Vita Glodessindis is the most interesting. Like the Vita Waldradae, it dates from the Carolingian period; its subject is difficult to date exactly24 but a vague seventh-century floruit is certain. Her vita shows subjection to her family. Marriages were arranged for her, and, after her first fiance, named as Obolenus, was executed by a King Childebert (II or III), she refused to go through with a second marriage. Fleeing to Metz cathedral, she was besieged there by her irate father and his followers, until, after a miracle, they saw God's will at work, and allowed her to become a nun, founding and endowing a monastery for her. The story is not unusual, and illustrates a common Merovingian family conflict (at least among the aristocracy). When subjection to a man, and the conventional opportunities for power only within the confines of marriage and child-bearing become a reality, the daughter rebels and becomes a religious, where she may achieve prestige and privilege based upon her own worth, rather than her value to a family. However, we should not assume that this conflict was entirely restricted to daughters. Certain male saints' lives (VP 6.i, 2O.i) show a similar course of events, though the family's response is invariably more muted and less violent. As the laws on free-slave marriage also indicate, a man had more freedom to choose the path which his life would take. By the late seventh century we begin to have 'documents of practice', charters, and these allow us to examine certain aspects of female status in some detail. First, we may consider land ownership. Whatever the laws say about this, there are very few instances of women alienating land. There are thirty-nine documents in the Wissembourg cartulary relating to the region of Metz, but in only seven of these (18%) are women named as donors (Wiss. 202, 225, 228-9, 261-2, 265). Two of these women issued the charter 24

McNamara et al. (trans.) 1992 place her not very convincingly in the very early seventh century, at the same time as Waldrada; Gauthier dates her more convincingly to the end of that century (again, in her interpretation, at the same time as Waldrada), but, in doing so, clashes with certain explicit statements of her vita, notably Vita Glodessindis 14, which states that, at the time of Glodessindis' burial, Arnulf 's body did not yet rest at Holy Apostles, thus dating her burial to before c. 640. Given the late date of the life it is possible that its author, like later editors, supposed Glodessindis' burial to have predated Arnulf's purely because of her father's name, Wintrio, possibly erroneously equated with the duke of Champagne of that name mentioned by Gregory of Tours (LH VIII.18, X.3). Gauthier's arguments for a connection with the Arnulfings and a late seventh-century date seem more solid. 69

SOCIAL ORGANIZATION

jointly with their husbands, and another with her son (her brother also appears in the witness-list). Two of the remaining four charters are issued by Gundoin's wife Wolfgund (Wiss. 228-9) without any male occurring in the text of the charter as co-donor. Nonetheless, it is significant that in Wiss. 228 her son Ermbert appears as the first witness, and in Wiss. 229 her two sons (Ermbert and Otto) head the witness-list. Wiss. 261 is issued by a Geretrude-Gaila, but on behalf, not only of herself but also of her three sons, one at least of whom, Buccellin, was old enough to be the first witness. This leaves only one Merovingian charter out of thirty-nine (2.56%) in which a woman grants land to Wissembourg Abbey other than in partnership with, or with the consent of, an adult male. All of the other Merovingian charters relating to the region of Metz are issued by men (though Pippin 'II's' wife, Plectrude, is co-donor in Pertz, Mayoral 2). However, charters represent only the last, formal stage of a transaction. How much informal, 'behind the scenes' influence existed is unclear. We must remember, nevertheless, that these documents relate to fairly powerful classes of the free, as has been mentioned. It is difficult to know how far down the social scale we can push any conclusions suggested by the charters about female alienation of property. Except where they are donors, women never appear in witness-lists. This might suggest legal incompetence, and support one reading of the laws. It might also suggest that women did not attend the public gatherings where such documents were issued in any active capacity. Again, however, this applies to a comparatively restricted social group, and, perhaps, only in restricted circumstances (land-giving). Another means of examining female status is through name-giving practices, a method used by Herlihy (1976), who suggested that the early middle ages saw a growth of matronymics to identify men. We may divide matronymics and patronymics into 'primary' (the child takes exactly the same name as his/her father or mother; obviously this is only possible where child and parent are of the same sex), 'secondary' (the child is identified as 'X, son (or daughter) of Y (father's or mother's name)') or 'tertiary' (the child takes an element, prefix or suffix, of his/her father's or mother's name). The last may well be more significant than 'secondary' naming practice, since the name makes a direct link with that of one of the parents. However, the Germanic name-pool, or name-element-pool, of the early medieval world was so small that deliberate choice is rarely susceptible of proof. Because of this chance or coincidence factor, this practice must be relegated to tertiary status. A common theory is that early medieval names were frequently of a hybrid 'tertiary' formation; that is, the child took an element of both his father's and mother's names (James 1982:76). There is hardly any indisputable evidence to support this. Even leaving aside his rather dubious use of statistics, Herlihy's thesis finds singularly little support in the region of Metz. Obviously, there are no men with the same names as their mothers, but there is similarly no evidence of what we might term maternal patronymics (taking the maternal grandfather's name), as far as this can be detected. Nor are there any examples of a charter's donor being identified as the son/daughter of his/her mother 70

SOCIAL ORGANIZATION: DOCUMENTARY EVIDENCE

(secondary matronymics). This makes it very nearly impossible to look at primary matronymics for women or at tertiary matronymics, but wherever a mother-son relationship can be found, there is no evidence that an element of the mother's name was used in that of her son. Conversely, tertiary patronymics are sometimes indicated for women (cp. Wolfgund, the daughter of Wulfoald (Wiss. 228-9)), and secondary patronymics are also known for women (Wolfgund again, but see also Geretrude-Gaila, daughter of Theudo (Wiss. 261)). The only times when men are identified by their relationship to a woman are in the witness-lists to charters issued by their female relatives, and here the reason is obvious. On the other hand, primary patronymics are evidenced in the charters. Duke Theotchar's son was also called Theotchar (Wiss. 213), and one hundred years later we meet a Gebold son of Gebold. Primary patronymics at various removes (equating with the practice of 'leading names') are known in some families. The Merovingians themselves are the best-known example, but in our region the repetition of the names [ChJRodoin, Gebehart and Lantfrid amongst the descendants of Petrus is noteworthy. Secondary patronymics are also common (Chrodoin son of Petrus, Weroald son of Audoin, Erlafrid son of Wolfrid, etc.). Tertiary patronymics are not perhaps as common as might be expected but they are attested (Erlafrid's case again; Gundoin's son Otto seems to have been called Audoin as we have noted they appear as Gundun and Ottun in Wiss. 226; Sigoin son of Haroin). This use of the father's suffix is particularly common in the Gorze cartulary between 770 and 790, when almost every father-son relationship bears witness to this name-giving custom. After this short period the names return to the same, more or less random-looking patterns as were attested earlier. We should note, however, that there are many possible name-giving customs which would simply be invisible in the charter evidence, such as naming after godparents or even, in most cases, after uncles or grandparents. It is, given the Merovingian practice of not naming children in lists of mancipia^ impossible to look at patronymics and matronymics amongst that element of the population. Later, Carolingian charters more often name children, but with the exception of Reginulf, son of Reginhar (Wiss. 193, of 19 Nov. 764), an example of tertiary patronymics, no patterns are suggested. We may examine the mancipia of the charters to look at other aspects of female status at this level of society. In contrast to the Carolingian charters of the region, where there are significantly fewer women listed (for whatever reason),25 there are almost exactly equal numbers of men and women among the Merovingian named mancipia (twenty-nine men to thirty-one women). The only other point which might be suggested concerns sale away from families. In this region, this only occurs in Wiss. 210 (Wissembourg, 24 Nov. 783), where four female mancipia are given to the abbey, without being attached to any land-grant. Other such gifts of unfree women are well 25

This is not uncommon in the Carolingian period; Wemple 1981:72. Coleman 1976 suggested the existence of selected female infanticide. Her argument attracted the powerful criticism of J.-P. Devroey (1981). The existence of female infanticide may be unpalatable to the modern mind, but it is well attested in pre-industrial societies, and it is dangerous to dismiss it, as Goetz (1993:45) does, as 'improbable'.

71

SOCIAL ORGANIZATION

attested in other regions. Whatever the purpose of the gift, it might be significant that we have no gifts of male mancipia away from land and families. Might this suggest a difference in attitude to unfree female children, that they might be taken away from their families? By the eighth century, in Pippin Fs decrees on marriage and divorce, a freewoman largely has the same rights as a man. By this period, with its new and more clearly defined horizontal social barriers, the roots of female power may have been traced differently at different levels of society. Age2'

There is little that can be said from the documentary records about the role of age and position in the life-cycle in the creation and acquisition of power. In the Pactus (PLS 24; 41.xviii), male children were valued at a higher wergild (600 solidi). A later capitulary (PLS 104.viii), however, punishes the killing of a female foetus with an astonishing 2,400-solidusfine(compared with 600 for a male foetus). A boy is, moreover, not held to be responsible for any of his actions (PLS 24.vii). None of this appears in Lex Ribvaria. By the time we have charter evidence, children in the communities of mancipia are usually left unnamed (they are named more frequently in early Carolingian Wissembourg charters). Children, whether male or female, came of age at the same time, at twelve (PLS 24) or fifteen (Lex Ribv. 84). However, the narratives suggest that, at least for the royal family, fifteen was the age of majority even in the sixth century. Ceremonies included the cutting of hair, especially for males (PLS 24). There is little evidence from the north about male age at betrothal. PLS 67 implies that women were betrothed at about the time when they came of age, and when boys had their hair cut, that is at about twelve. This age for female betrothal is supported by Venantius' epitaph for the 'barbarian' Vilithuta (Carm. IV.26), who was married at thirteen. Age of marriage may have increased, with legal majority, to fifteen in the seventh century. At this point we might return to the subject of the pueri. As mentioned, on most occasions when such people appear in the narrative sources, they do not seem to be in any way unfree. PLS 54.ii, moreover, mentions that a puer regis might become a count. It was noted that the word puer usually simply means 'boy' when it is used in the laws. It might be proposed that pueri are retainers, who might be of free birth, but who have yet to settle down as heads of families, and thus are an equivalent of the young landless Anglo-Saxon warriors, discussed, for example, by Charles-Edwards (1976). The word thus relates directly to at least 'structural' age (for which, see Bernardi 1985). The significance of old age is difficult to establish from the written sources. Once a freewoman ceased to be able to bear children her wergild fell back to the usual 200 solidi. There are no other legal provisions concerning age. Anecdotal evidence suggests that in some cases (such as that, probably exceptional, of Queen Brunechildis), people might hold on to their power into old age, but the implication of the narratives is clearly that the tenure of 26

There is so far very little on this topic. Minois 1989, though problematic, as ground-breaking studies often are, opens a way into the study of the old.

72

SOCIAL ORGANIZATION: DOCUMENTARY EVIDENCE

prestige and power was more likely to be challenged as one grew older (cp. Chramn's rebellion against his ageing father, Chlothar I; the removal from power of Queens Brunechildis and Baltechildis). Widowhood is mentioned in the laws, largely in relationship to remarriage. PLS 44 and 100 make it clear that a widow is at least partly under the control of her husband's relatives, and that she must consult with her children before remarriage. PLS lOO.i states that a woman for whom there are no immediately recognizable responsible kin is in the guardianship of the king. What this meant in practice is unclear. It may have implied that she lived under the protection of the community. There is little such evidence for the status of old men. Obviously, according to the law, they were still entirely responsible for their actions and legally competent. Any loss of status can only have been the result of the everyday demands of the habitus, rather than being institutionalized (this contrasts, again, with Visigothic law, where an old man was valued at half, or less, of the wergild of a man between twenty and sixty-five; Bonnassie 1991c:75). It is worth wondering whether the term senior which occasionally appears in Gregory's work (cp. L//VIII.31 for a northern Frankish context) might mean, more literally, 'elder', rather than 'senior' in the broader, ranking sense. Marculf's collection contains two documents (Marculf 11.11, 11.13) which shed a grim light on old age. In both, an old man is seen to be no longer able to fend for himself, and so adopts an heir. In one document (Marculf 11.11) this is one of his grandchildren; in the other, the heir is not a member of the family. In both cases the heir undertakes to feed, clothe and protect the donor. Fictive kinship27

A final element in social organization concerns groupings of people, the membership of which is not based upon either consanguinal (blood) or affinal (marriage) relationships. Though it is not without its problems, the term c fictive kinship' can be used to describe such links. In the Merovingian period there were a number of ways in which fictive kinship bonds could be created. Spiritual kinship, godparenthood (Lynch 1986:ch.6), is perhaps the most obvious, but we should also note the existence of adoption (Lynch 1986:179-80). Both Frankish law-codes allow for the adoption of an heir when there were no children (PLS 46; Lex Ribv. 50). Marculf includes a deed of adoption of an heir (11.13). The ceremonies involved in the first cutting of a boy's hair, traditionally carried out by someone other than the boy's father (Lynch 1986:180; alluded to by PL5 24.H), and of the first bestowing of weapons on a young male, were also used to create bonds between families. A comparable bond could be created by giving a child to another family to bring up. Arnulf of Metz was brought up by Gregory of Tours' great-uncle Gundulf. Clearly, relationships like this served important secular as well as spiritual purposes. Bonds between godparents and godchildren could, of course, bridge other barriers such as class or rank. Thus Sigivald could be the godson 27

This topic is thoroughly covered by Lynch 1986.

73

SOCIAL ORGANIZATION

of the Merovingian Theudebert (although there was some blood-relationship between the two; LH III. 13, 111.23). We may suppose that godparenthood and spiritual kinship were mechanisms used in improving the status and prestige of the individual, either the godchild or his or her parents. Brunechildis' sponsorship of Berthefried's daughter {LH IX.9) may have been a ploy used to win his support during Childebert IPs minority. It is worth noting how female as well as male children could be used equally in the creation of these ties. The obligation to treat a godchild as a blood-relative could be used to protect a child. Godparents might be chosen from amongst those who could otherwise threaten a child's inheritance. When Fredegundis and the Neustrian aristocrats placed the infant Chlothar II in Guntramn's care and asked Guntramn to stand sponsor to him at the font this was a deliberate, and very necessary, policy to protect the child-king's inheritance {LH VII.5, VIII.9). Similarly, Theuderic II chose Chlothar II to be the godfather of his son, Merovech (Fredegar, Chron. IV.42). Bringing up or educating another's child could also bridge rank barriers, and (most noticeably in the case of the nutritor of a royal child) be of great political use to the family of a guardian of lower social rank. Guardianship, like godparenthood, also helped to remove the risks involved in the transfer of prestige from one generation to the next. Should a boy's father die, then his guardian would, theoretically, help to secure his inheritance. Association with a guardian would (where he was of the same, or higher, standing as the child's father) also enhance the transmission of prestige. Of course, godparenthood, adoption and guardianship all bridged the barriers created by age difference. A final form of fictive kinship would be constituted by membership of the sworn bands or gangs of freemen mentioned in the Pactus, the contubernia, which may have been a form of age group. Lex Ribvaria's hariraida may or may not have been similar; it has already been seen that Lex Ribvaria is more concerned with Vertical' ties between a leader and his followers {satellites, ingenui in obsequio). Ties of lordship, vassalage and dependence can be viewed as yet another form of fictive kinship, bridging and blurring the horizontal barriers of class and rank. This brings us back to the first section of the chapter, which described the creation of such links in the establishment of a more formal aristocratic caste, a nobility and the creation of more rigid social hierarchy.

74

Creating a model: cemeteries of the Merovingian civitas of Metz

Introduction In 1990, 305 possible Merovingian sites could be catalogued in the civitas of Metz (fig. 3.2).* The civitas' Merovingian burial customs differed from

•

•Florange \ Bouzonvillet

Wittersheim #Altheim ••Walsheim

Berthelming Bettborn St-Ulrich#

Fig. 3.1

Places mentioned in chapter 3

Halsall 1990, app.2.a, is the most recent and complete catalogue of the sites of the whole civitas, to which must be added the newly discovered site at Fontoy {Arch. Med. 1991:386) and the latest excavations at Florange-Daspich {Arch. Med. 1991:385-6). For other general surveys of archaeological data, see Clermont-Joly 1978; Forrer 1934; Guillaume et al. 1988; Hurstel 1984, vol.11; LB; LFS; LS; LT; Lutz 1952; 1956; 1978; Maisant 1971; Salin 1937; Simmer 1987; Stein 1989b; Toussaint 1938a; 1938b; 1938c; 1938d; 1943-5. These draw on older repertoires, such as Beaupre 1897; Benoit 1862; Boulange

75

SOCIAL ORGANIZATION

Fig. 3.2

Merovingian cemeteries in the civitas of Metz (including all possible sites)

those of the late Roman era in several ways. Late Roman burial in the civitas, as most clearly shown at the sadly unpublished site at Scarponne,2 took the form of inhumation with a decreasing number of grave-goods, usually ceramics. The dead were buried in simple trenches or in stone- or tile-lined cists. Orientation was irregular, as was the lay-out of the cemeteries. The custom of stone-lining the grave can be traced back to early Roman times when cremations were similarly placed in stone-lined pits (Baudoux et al. 1990:66-80 for a recent discussion of the region's Roman cemeteries). Around 400, changes occurred. In place of the largely unfurnished and standardized late Roman rites, burial with weaponry and jewellery began to reappear (weapon-burials had been known in the highlands of the Vosges until the third century (Lutz 1964), but in the civitas' lowlands such graves had been unknown in the Roman period). By the beginning of the sixth century, burial with numerous and diverse grave-goods had become common, though the inhumation of the dead in simple trenches or stonelined cists remained. At about this time many new cemeteries were estab-

2

1853-4; Hoffmann 1893; Huber 1889-90; Joly 1870; Kraus 1889; Lepage 1843; REL; Simon 1859; Verronais 1844; Viville 1817; Wiener 1895. For other sites of the civitas, not discussed here, see Ancelon 1874; Barbier 1867; Barthelemy 1842-3; Bellard 1933; Chatelain 1881-2; Dufresne 1854-5; Guillaume 1864; Merciol 1886; Moineaux 1966; Morlet 1863-6; des Robert 1878; de Sailly 1869; Sauer 1958; Schlosser 1881-4; 1895; Schmit 1879; Simon 1843-4; 1850-1; Welter 1902; Wichmann 1894. Gallia 1968:374-6; 1970:281-3; 1972:349; 1974:338-9; 1976:356; 1978:335-6; Baudoux et al. 1990:67. Masson 1975 contains the only accessible inventory.

76

CEMETERIES IN THE CIVITAS OF METZ

lished. As elsewhere in the post-Roman West, these were organized in rows, more neatly than in late Roman cemeteries, though the extent to which such organization was adhered to varied from site to site. The nature of the gravegoods and their patterns of association will be discussed further below, as will alleged instances of deviation from burial norms. The publication of all the sites discussed here (except Bettborn, which alas contained only eleven graves) suffers from certain problems. None has really been published to the standard expected in the 1980s and 1990s, even Audunle-Tiche, the most recent (see ch.4). Only selected grave-plans are given, leaving us with vague written descriptions of skeletal lay-out, artefact position and grave construction, and no information on grave-shape. Further, the areas around the graves are not planned in detail, and where pits and similar features are found, these are not properly described or illustrated. The only features deemed noteworthy are earlier Roman buildings and occasionally later medieval structures; these too are not published to the standard expected today. We may conclude that these sites were excavated and published according to some very outmoded precepts: 1 That the grave is the only thing worth recording on a cemetery site (because burial is assumed to be the only thing which went on there), and 2 that within the grave the only important variable is the nature of the grave-goods. These assumptions were reflected in the excavation methods, which sometimes involved digging parallel trenches a metre or so wide and a metre or so apart. This method is relatively quick and almost certainly discovers all the graves on a site, but it is not particularly useful if the possibility of other structures on a site is accepted. Nevertheless, this method, used by Salin in the 1920s to 1940s, was used at Altheim in the 1970s (Hackler 1989:fig. 3). It can be said that the things which interest the excavator of a Merovingian cemetery, the things which he or she records and the way in which he or she does so, have hardly changed since the Voinots excavated Chaouilley in 1902. These limitations in the quality of observations reflect the limited analyses performed by Merovingian archaeologists upon their data. The data used are, as will be seen, often very old, and frequently problematic. The age of the data is not a problem in itself, as Huggett and Richards (1990) have shown. As will be seen, the larger sites discussed here were excavated and recorded (if not always published) competently. Other problems remain, however, notably concerning sample size and the ageing of skeletal data, and it is far from impossible that the thesis proposed here might be shown to be fundamentally mistaken, because based upon statistical chance. Nevertheless, the picture to emerge from these analyses seems remarkably consistent. We cannot ignore this apparent consistency and refuse to attempt its explanation for the simple reason that there are, in theory, problems with the evidence; to do so is to fall into the most convenient positivist 'cop-out'. There are no (and will never be) perfect, valueneutral data. It must be the duty of the early medievalist to put forward a model explaining the evidence as it is seen at a particular point in time, and 77

SOCIAL ORGANIZATION

which can be tested, modified or rejected as and when better quality data become available. That proposed here rests only upon those aspects of the sites' recording and excavation which can be taken as reliable. It is at least as likely that more plentiful and higher quality evidence will allow greater subtlety to be extracted, in support of this model, as that it will force its rejection.

Ennery cLes Trois Arbres' (fig. 3.3) Analysis of the archaeological cemetery evidence from the civitas of Metz began with the site of Ennery (Moselle), the only sizeable cemetery of the region yet published. The site was excavated in 1941, with, for its day, unusual rigour (Reusch 1941-2:40-52 and Taf.I-VII), and extensive records were kept (Alain Simmer is currently preparing a definitive publication). It was not, unfortunately, published as well, though a detailed inventory can be built up from careful study of the original publication (Delort 1947), the anthropological report (Heuertz 1957) and the catalogue of the Musees de

&S&i£s&2

wi

a N • 5

10

M

Fig. 3.3

78

Ennery 'Les Trois Arbres': site-plan (after Delort 1947 and Clermont-Joly 1978:18)

CEMETERIES IN THE CIVITAS OF METZ

Metz (Clermont-Joly 1978), to which most of the finds were sent (the remainder largely going to the Musee du Pays de Sarrebourg). Eighty-two graves were excavated, fifty-eight of which were intact and suitable for analysis. The excavated section of the site seems to have been occupied during the last three-quarters of the sixth century, but with one or two graves dating from the very early seventh century. The bulk of the sample is then to be placed in Bohner-Perin period III (c 525-c. 600; Bohner 1958; Perin 1980). That the excavated section of the cemetery covers only two or three generations means that chronological distortion of mortuary practices is unlikely to have affected the data to any great degree. Associations between artefacts and sex The analyses performed on the Ennery data began with an attempt to examine the artefactual construction of gender. Rather than using Heuertz's (1957) physical anthropological study from the start, this was a valuable opportunity for testing a methodology of determining the deceased's sex from grave-goods. If such a method was found to be accurate it could then be used with the region's other cemeteries which had not been studied by a physical anthropologist. The method would, above all, test the common assumption that weapons are always an indicator of a male burial, whilst jewellery is always the mark of a female grave. The method involved dividing the grave-goods into a series of 'artefacttypes' (or object-types), types of artefact found in at least two intact graves containing at least two such artefact-types. The term 'grave-goods' has been used to refer, more generally, to any and all objects placed with the dead. Thus a grave containing three spears contained three grave-goods but only one artefact-type. Artefact-types were given a numerical code (fig. 3.4) which, it was hoped, would reduce preconceptions about the expected strength of the gender-related messages of particular artefacts. Graves disturbed by later burials, building or farming were excluded from the analyses, as the integrity and completeness of their assemblages was doubtful. A diagram was then drawn wherein the artefact-types were plotted in identical order on X- and Y-axes. Weaponry was placed at the end of the axes nearest the origin, and jewellery at the other, according to the preconceived notions about their gender associations mentioned above. Other items were placed in between. In some cases, their position was influenced by the results of analogous work on Anglo-Saxon cemeteries (e.g. Shephard 1979). When this was done, any association between object-type A and object-type B was plotted at the intersection of the two types. After the first test, which showed no clear patterns, the order of the object-types was adjusted to bring together groups of associated items. After seven tests the results shown in figure 3.5 were produced, revealing three groups of artefact-types, two of which were never associated with each other. In the Ennery analyses, however, weapons were kept grouped together. This was not done in later tests on other sites for the very good reason that some artefact-types seemed to be at least as strongly male-specific as, and occasionally more so than, certain weapon-types. The three groups of artefact-types could be described as mas79

SOCIAL ORGANIZATION

Code 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24.

Artefact-type Bronze hairpin. Earring(s). Brooch. Bronze ring. Bracelet. Necklace. Disc brooch (Lavoye phase 2 only). (Not used). Spindlewhorl. Iron buckle. Flint. Bronze, silvered bronze or silver buckle. Comb. Shoe or garter buckle. Shears. Bone or antler medallion. Knife. Grey or grey/black ceramics. Red or red/orange ceramics. Glassware. Intaglio. Coin(s) (except where mounted as part of a necklace). Bronze needle. Bronze tweezers. Fig. 3.4

Code Artefact-type 25. Awl. 26. Spear (any type except ango) or arrowhead. 27. Purse-fastener or purse-buckle. 28. Single bead. 39. Axe (including francisca). 30. Belt-fittings (counter-plaques, dorsal plaques, strap-ends, &c). 31. Scramasax. 32. Sword. 33. Plaque-buckle. 34. Strike-a-light. 35. Chatelaine. 36. Key. 37. Bone pin. 38. Truss. 39. (Not used). 40. Fork. 41. Shells. 42. (Not used). 43. Dress-pin. 44. (Not used). 45. Shield-boss. Ango. 46. 47. Bronze vessel. 48. Bucket.

Numerical coding of artefact-types used in analyses described in chapters 3-4

culine, feminine and cneutral\ The word 'neutral' is here used guardedly simply to mean those artefact-types which, in the cemetery evidence, revealed no exclusive association with either males or females. Reservations must be expressed, nevertheless. The contents, for example, of otherwise identical pots may have been vitally important in gender construction; a brooch may be feminine simply because people were not buried in cloaks and women alone wore brooches to fasten other clothing; in the deliberate act of furnishing a grave, objects may have been given specific meanings which were treated loosely in everyday life; symbolisms may even have been inverted in the transition from a 'living' to a mortuary context. The terms 'neutral' and 'neutrality' must thus be used with care, and note taken of the foregoing riders. Three objections might be raised here. The first is that the reduction to a binary opposition of two groups of objects oversimplifies the data. The second objects that the small sample size leads to results which are too sample-sensitive. The third concerns the small size of the samples themselves. These are valid objections, but the results of the tests do seem to show that this binary opposition was in fact vitally important within the customs of grave-good deposition. As for the second objection, the 80

CEMETERIES IN THE CIVITAS OF METZ

Artefact-types (see fig. 3.4 for key) O 6 9 3 11 3 2 2 3 3 3 L 1 2 3 1 1 2 2 2 2> 1 1 >2 4 4 7 3 9 5 62 2 6 9 1 3 0 "7 1 7 6 8 0 3 0 5 2 32 26 29 31 33 30 17

I

I

(J 11 27 36 18 10 23 20 25 22 12

1 1 1

1

1

I

II 1 1 I 1 1

I 11 I 1 11 I 11 1 1 1 I 11

I

1

I 1

1 1I 1 1 11

I 11 I I I 11 11 I 11 I 11

I 1 1 1 I

1 I 1 1

I1 I 11 1

I

I

I I 1

I I

° 14 4 37 13 19 5

9

6 2

I 1

11 I 1 11 1 I

I I

I I 1 I I I 11 I I

III •

I I 1 I 1 I 1

1

11

I

1

I

| = artefact-type X is associated with artefact-type Y in one or more intact graves containing two or more artefact-types. Fig. 3.5

Ennery (Les Trois Arbres3: artefact associations

requirement that an artefact-type be found in two or more intact burials containing two or more such artefact-types was based upon the notion that the material record was meaningfully constituted. An object-type must have had symbolic significance if deliberately (and publicly) placed with other objects in two or more burials (between 1% and 4% of individual samples). In practice, cases where the results were affected by small sample size (as at Audun-le-Tiche; see ch.4) showed up fairly clearly. The third objection is the most important. Separating the Ennery graves into gender and age groups did result in tiny samples. The picture to emerge from these analyses cannot therefore be regarded as, and is not presented as, more than a sketch. Its outlines, nevertheless, are very interesting and deserve attention. After the artefact-types were arranged into the series shown in figure 3.5, it was felt that one ought to be able, if the two unrelated groups did refer to gender, to predict the sex of the deceased. Consequently, a new diagram (3.6) was produced in which the sorted order of artefact-types was placed 81

SOCIAL ORGANIZATION

Vrtefact-types (see fig. 3.4 for key )

6 •j j

1

"\

Z J J 2 6 9 13

Grave

Z

o •1

III 1 / 11 I

70 71 54 6 5 4 39 22

1

|

1

11 •I• I i••

1 37 65

36 3 55 35

•1

•

11 1. 1,

28 63 61 75 47 7 11 67 42

•

81 8 82

52 58 60 40

9

o 1 3 1 1 z zo z ABC 0 '7 1 7 6 8 0 3 2 4 4 7 3 9 5 6 2 0 5 2 6 6I 6 66 I I 6 6 4 6 6 4 6 64 . ' ; a a 6 6 5 6 64 | 6 6 5

"X " J

M•

.1'1

11 I

•

I 1

1'

1 1

2 57

•

48

49 79 51 73

•

34

6 6 6 6 6 6 6

1 i I

•

1

I

6 6 66 AB.2 6 5 6 5 O2 O5 O2

6 96

1 1

32

1

6 cJ3 o o4 o o? B 9 o6 o o? o o l 9 95 9 93 9 03 9 o1 BC 9 03 9 6*5 9 93 9 9 4 9 9 4 C

1 1 1 i

c=

ag e according to Heuertz = 7-14 yrs; 3 = 14-22 yrs; 4 = 22-40 yrs; 5 = 40-60 yrs; 6 = 60+ yrs) .

A = F )redicte m \ J^Metrich ^ Florange • \ Koenigsmacker V

Homecourt

) _ • .„ >^v_ % Bouzonville ^>^^-^ g Saarbrucken # V F Giidingen Amneville (fEnnery LeHeraple#

v

L Verdun S»Belleray

#

V»Dieue-sur-Meuse

Lavoye

Klt^r, ^MCIZ

M Jouy-aux-Arches ! : >

II \

I

Scarponne

A

Bar-le-k Due ^ ^

• Imling

J^Nancy

V

^ * ^ Varangeville Morley *

• N

\ I

Chaouilley

I

V

S.

0

10

I

Km

)

Fig. 8.1 Places mentioned in chapter 8 James 1989 is a recent summary with good bibliography. See also Halsall forthcoming ch.6; Steuer 1982 and especially Pader 1980 and 1981. Young (1984 and 1986a) studies some of the sites studied here, but the analyses are flawed and ultimately, particularly in Young 1984, lead nowhere.

245

20

CONCLUSIONS

artefacts with the dead. In general terms, these are attested on all the sixthcentury sites examined. It appears that there were clear mores about what types of grave-goods were appropriate, and in what numbers, in the burials of particular people. But within these rules there is significant diversity. At Lavoye, it seems that males above about twenty, who had not yet reached old age, were deemed worthy of burial with numerous grave-goods, particularly weapons and other masculine artefacts. There is, nevertheless, a wide variety of practice within this general rule. This might suggest competitive display of resources. On the other hand, it might simply be that, within this general rule, the distribution of grave-goods reflected accurately a family's rank or wealth-based class. If the argument against variability in grave-goods (within age and gender groups) representing social competition is to be maintained, we must ask why grave-goods were buried at all. Why should there have been a 'passive' (reflective) rather than an 'active' (competitive) gravegoods custom? In the former case we are forced to accept that the burial custom was governed by the exigencies of religious belief or Germanic tradition. The latter possibility, that grave-goods manifest Frankish as opposed to Gallo-Roman ethnicity was, in its usual crude sense, rejected in chapter 1. The religious interpretation of grave-goods has long been questioned (cp. Ucko 1969). The many flaws in the idea that burial with grave-goods was pagan and burial without them Christian have long been pointed out, and need not be repeated (Young 1975; 1977). One more recent attempt to plot Christians within the region's cemeteries was made by Alain Simmer (1987:388; 1988:130-5), who argues that sarcophagus burials are Christian because they are, allegedly, conscious imitations of things Roman and urban, like Christianity itself. This is clearly a flawed argument, though more interesting than previous efforts to identify Christian burials. The distribution of sarcophagi is uneven and governed, obviously, by the availability of suitable stone or the ability to transport such stone to the cemetery (Cuvelier &c Guillaume 1989). If we accept Simmer's hypothesis, we must also accept that this manifestation of Christianity is only to be found in geologically and geographically suitable areas, and thus that no argument can be made from the absence of sarcophagi. Except where a local quarry made sarcophagus production easier and the use of stone coffins a viable proposition for the bulk of the community, sarcophagi were expensive items and are usually rare within cemeteries. They are therefore more likely to be manifestations of the wealth and prestige of the deceased's family than a sign that within a cemetery of 200 graves, five or six were those of Christians. Simmer (1988) does pay lip-service to this point but does not allow it to interfere with his suggestion (discussed above) that one area of the site at Audun-le-Tiche was reserved for Christians. The religious element in a funeral is likely to be that which leaves no archaeological trace, such as what was said over a grave, as part of the funerary feast, or at other stages of mourning and burial. It is surely worth asking why religion appears to have played so little part in the archaeologically revealed aspects of the disposal of the dead. The lack of a religious element in the material remains of Merovingian burials, or the fact that such an element remains ambiguous, is worthy of note, not least because it coincides 246

TOWN AND COUNTRY, c. 350 - c. 600

so well with the testimony of the written sources, which likewise attest the perpetuation by Christians of apparently non-Christian beliefs, the separation of what seem superficially to be religious acts from religious creeds, and finally the slowness of the church to take interest in these particular c rites of passage' in this period (Paxton 1990). Religious and ethnic explanations cannot then be adduced for the existence of the furnished burial custom, which, in turn, makes it difficult to see the latter as a passive reflection of society. A case has been made for the reasons for the appearance of the rite in the late fourth and early fifth centuries to be sought in the social instability of the period (Halsall 1992a). The deposition of grave-goods becomes a strategy in the competition for prestige within the local community. For such strategies to make sense, burials must be regarded, as set out in chapter 1, as deliberate, actively and meaningfully constituted 'texts'. This should be pursued further. If, following Shanks and Tilley (1987:75), cno social practices exist without signification and without being situated in an overall symbolic field', funerals and burials, as religious and ritual events, are clearly particularly important arenas for symbolism. The funeral and the resources expended on the construction and furnishing of the grave were important means of transmitting to an audience information about the deceased and their families. We can argue that burial practices are an especially important and dynamic field of social expression; in our period they underwent fundamental change more frequently than spoken language. Let us consider the detectable elements of Merovingian burial in the region of Metz. Even the patchy data from the region have suggested that the shape, size and construction of the tomb were deliberately chosen, and thus important variables in the funerary process. Then the deceased was clothed for his or her burial. It has been noted that certain elements of dress or bodily adornment, such as jewellery and plaque-buckles, had important symbolic value, as did other items such as knives, and scramasaxes, which were often worn rather than deposited in the grave separately. Other grave-goods were laid in the grave alongside the corpse: items of weaponry such as spears and, it seems, swords; pottery, glassware and other vessels; even dress adjuncts were sometimes piled neatly beside the deceased rather than worn as in life (and a distinction must be drawn between these two options). In both cases, the grave-goods were governed by general customs. The analyses performed on the cemeteries of the region have suggested that the positioning of the body, and its treatment between death and burial, were further significant variables. Finally, we have evidence that the burial was accompanied by feasting. The pits on cemetery sites at Morley (Meuse; Salin 1946:205), Koenigsmacker (Moselle; above, pp. 107-8), Varangeville (Meurthe-et-Moselle; Salin 1946:205), Bouzonville (Moselle; above, p. 99) and Ennery (Moselle; above, p. 91), which contain ashes, bones and pieces of pottery, are just the kind of evidence one would expect of the sweeping up after a feast. The presence of food and drink offerings, either in the form of animal bones or in the deposit of pottery in the graves, argues further for the existence of funeral feasts. The charcoals, pot sherds and animal bones found in the fill of graves are probably also best explained as the remains of such activities. 247

CONCLUSIONS

We can consider the cemetery as an arena for competitive discourse, with the burials as statements. We need to consider further two related issues; the audience at which the statement was aimed, and the temporal aspect of the message. For how long was it intended to be read (and re-read) and understood? All the elements of sixth-century burial noted above are, it must be stressed, temporary. Outside Metz, there is hardly any evidence for permanent above-ground grave-markers. A display of wealth or of access to resources manifested through the variables noted would be invisible to anyone not actually present during the funeral process. Once the grave was filled in, the display was ended. To be effective, then, such a display needed a large audience. Written evidence supports the indications of the archaeological data, that early medieval funerals were important public displays rather than quiet family affairs. As we saw at Lavoye, sixth-century cemeteries seem to serve a large community. There we suggested that two or three groups used the same cemetery. As shown in chapter 5, it seems reasonable to suppose that the sixth-century cemetery served as a common ritual focus for a large but dispersed community. The potentially large audience at a sixth-century burial in turn probably required fairly clear rules determining the appropriate symbolism of the grave-goods. The often implicit belief that grave-goods and burials are a static mirror of a fixed social structure can thus be countered. It is not infrequently cited as an anomaly, as an example of the complexity and 'difficulty' of reading archaeological evidence (cp. Miller 1985:203), that whilst, as in sixth-century Anglo-Saxon and Frankish society, settlements show no evidence of hierarchical distinctions, with all houses or dwelling units being of more or less the same order of size, cemeteries show huge social differentiation. This seems to miss the point. Even leaving aside the fact that the 'huge social differentiation' can, as here, be structured around age and gender differences rather than hierarchical rank or class distinctions, it is easy to understand that if one family acquires sufficiently assured control over resources to build itself a much bigger and better dwelling unit than its neighbours then the local cemetery is not likely to be an arena where that family's neighbours engage with it in serious competitive displays of wealth and power. The deposition of grave-goods should thus be seen primarily in terms of competitive display. Rather than being a passive mirror of social organization, it is an active strategy in the creation of social reality, an act of 'social theatre'. This conclusion is central to this interpretation of Merovingian society in the region of Metz.

An Indian summer: late Roman society in the region of Metz Some overall conclusions about late Roman society in Belgica Prima are a further necessary preliminary to the understanding of early Merovingian social structure. Fourth-century society in the region of Metz was fairly stable. The general, gradual decrease in grave-goods in the fourth century, 248

TOWN AND COUNTRY, c. 350 - c. 600

after the reintroduction of inhumation in the third, and the decline of larger stone monuments slightly earlier (these peaked in the third century), suggests that the funeral was not an arena for fierce local competition. Study of the region's settlements has suggested (above, pp. 176-8) that the late thirdcentury upheavals were survived fairly well. The fourth century did not see, as has been suggested in other regions, an expansion of large villas at the expense of smaller establishments. In this region, as Burnand (1990:233-4) argues, the fourth century is an era of smaller establishments. The traditional picture of the creation of large latifundia finds no support here. In the civitas-czpital, despite some contraction and dereliction, urban life continued to thrive (above, p. 224). Metz's public buildings continued in use until at least the mid-century and there were major new civic works such as St-Pierre-aux-Nonnains, the city walls and possibly the small amphitheatre, as well as new religious edifices, at St-Pierre-aux-Arenes, St-Felix and Holy Apostles. The aqueduct was refurbished as late as Valens' reign (one of his coins was found cemented into it at Jouy-aux-Arches; Toussaint 1950:100). The vici and castra continued to be occupied through the century with, again, evidence of wall-building, resurfacing of roads and rebuilding of houses. This suggests an organized industrial context for such building activity (stone-quarrying and transportation, and tile-manufacture). This is supported by the continuing use of the Yutz kilns, and the supply of the region with coinage and manufactured goods, such as Argonne ware. The economy seems to have been provincially based, after the break-up of the general, imperial economic system in the third century (Reece 1980; 1981). All the evidence points to the persistence of complex social organization. The stability of this social structure was built upon the presence of the Roman state. The provincial administration and its network of offices and titles, founded upon the civitas unit, provided the social order's legitimation. The status and power of individual landholders was maintained by the acquisition of titles and their attendant privileges. Stability in the countryside permitted the continuing sponsorship of urban life, even if, as is well known, the villa became the centre of fourth-century aristocratic culture (Matthews 1975). The landowning elite still came to town to participate in civitas politics and to carry out their administrative functions. That Metz and its region remained prosperous into the later fourth century, whilst other, particularly north-western, areas of northern Gaul suffered more acute decline, may have been related to its strategic position on the lifeline between Trier and the Mediterranean.

Collapse The Alamannic wars of the mid-fourth century hit the civitas of Metz badly. In the 350s, the Moselle valley was something of a frontier war-zone and the aristocracy's rural base was seriously affected (Ammianus, Res Gestae XVI.2.ix-x, XXVII.2.iff). At this point there appears to have been some dereliction and abandonment in the villas and in Metz, but the region seems to have recovered. Building of walls is attested at Saarbriicken (above, p. 203) 249

CONCLUSIONS

and, possibly, Metz (above, p. 223), whilst at least part of the St-Pierre-auxNonnains complex dates to Valentinian I's reign. Coinage continued to find its way to Metz (above, p. 227). This was a short-lived recovery. In the region of Metz, the Tall of Rome' came quickly and it came early. The coin supply to Metz dried up quickly after c. 388. Leaving aside the thorny problem of the St-Pierre-aux-Nonnains basilica, it is difficult to see any occupation in the town by c. 400. By the time that the Notitia Dignitatum was written, the procurator gynaecii rei privatae at Metz had moved to Aries {Notitia Dignitatum occ.l2.xxvii, though the entry is corrupt). The same picture is found in vici, castra and villas across the region. The coin supply persisted in the countryside until the early fifth century, when it too fizzled out. At about the same time, the last evidence of the pottery industry ceases. On the other hand, around 400, we have the earliest of the new, furnished inhumations.2 The cemetery has become an arena for local displays of power, perhaps for the first time since the late third-century crises. Further west, in Picardie and what is now Normandy, these burials appear somewhat earlier in the late fourth century. This may again relate to an earlier collapse of the Roman order in those regions, with that in Metz persisting longer because of its strategic and political importance. Explanation of such changes must be sought in the removal of effective imperial presence from the region. In chapter 1 it was argued that there is little evidence of central governmental activity in the region after the suppression of Magnus Maximus in 388. The imperial state's withdrawal had catastrophic results. Metz lost most of its administrative functions. This cannot, even were imperial authority still effective, have been helped by what Goffart (1972:186) termed ca remarkable devolution of tax obligations' from the city to the individual in this period. That the main function of late Roman Metz was administrative is attested by the rapid end of coin supply to the town after 388. The end of effective governmental presence furthermore removed the organized network of offices and titles. Efficiently regulated patronage, and thus the legitimation of the social order, was withdrawn. As we have seen (above, p. 34), many of the most powerful people of the region fled south with it. Thus, in the appearance of weapon burials and other lavish graves, we see a new form of social competition, making use of what, hitherto, were the prerogatives of imperial authority: weapons, demonstrating the control of armed force, and official belt-sets and brooches, badges of rank (Halsall 1992a). Yet it must be pointed out that, as yet, these burials are few in number. They appear to represent an old elite devising new ways of demonstrating local power. It is this instability and struggle for power which Salvian described. Following and developing the ideas of Van Dam (1985, chs. 2-3) and Drinkwater (1989; 1992), I would interpret Salvian's complaints as evidence of a struggle for local power, and the legitimation thereof, in the vacuum left by the Roman state's disintegration in northern Gaul (above, ch.l). This manifested itself as an opposition between those who 2

Scarponne, Le Heraple and Florange (see above, ch.6); Dieue-sur-Meuse grave 101; Belleray (Bohme 1974:308-9 and Taf.115; Guillaume et al. 1988:70).

250

TOWN AND COUNTRY, c. 350 - c. 600

still retained, and exploited (cp. Governance V.8.xxxivff), 'official5 imperial backing, those who wielded 'unauthorized' local power (the so-called bacaudae\ ibid. V.6.xxivff) and those who turned to the incoming barbarian rulers to maintain their local authority (ibid. V.5.xxiff). It is not surprising that Salvian records that the people of the Trier region craved a return of imperial patronage (ibid. VI.15.1xxxvff). With the insecurity of local power, aristocratic attentions had to be turned to their rural land bases. Large estates broke up (Wickham 1981:182 for a similar process in post-Carolingian north Italy). In this context the rapid collapse of urbanism is not unexpected. The end of social stability meant that few people had any assured control of surplus; the necessary social structure for organized industry and manufacture ceased. The elites no longer had cause to come to town. Much of Metz became derelict. As subsistence became increasingly important in the settlement's economic basis, the population drifted away from the now dangerous walled area, parts of which seem to have been turned over to horticulture. Similar processes of abandonment can be seen on the via and castra sites as well as on villas, and similar dereliction, following the break-down of organized stone and tile industries, may be adduced. An explanation for the rapid demise of urbanism in Metz can be proposed which is not radically different from Esmonde Cleary's (1989) for the end of Roman British towns. This drift was given an added focus by the shifts in mentalities of the time, so brilliantly evoked in the ceuvre of Peter Brown (1981; 1982). In the collapse of secular power networks, alternatives, as stated, needed to be found. One alternative was spiritual, the patronage of the saints and their representatives. Hence it is not difficult to see settlement moving away from the old secular centres of power (the forum, basilicas, and so on) to new spiritual extra-mural power bases (the churches) as a physical manifestation of this mental shift. In this context, the translation of St Stephen's relics and the construction of his church near the presumed site of the forum, perhaps with the bishop taking over former public buildings there, take on added significance.3 Nevertheless, the bishop appears to have resided in the Great Amphitheatre, which perhaps served as a refuge, until after the sack of 451.

Early Merovingian society in the region of Metz The fifth century is a dark period in almost all areas of evidence. This is only to be expected given the collapse and confusion outlined in the previous section and in chapter 1. The invasions, such as of 406-7 and 451, will have fuelled this anarchy. In the conditions just described, Attila's Huns could easily ravage without check until they reached central Gaul and the limits of Aetius' effective power. 3

St-Stephen is the only certainly fifth-century church and probably the only fifth-century piece of monumental architecture in the region. This does not contradict the model proposed above. The construction materials and labour can have been provided within the context of Voluntary supply', and from quarrying nearby buildings; cp. the construction of the mid-fifth-century church of St Martin, Tours (Van Dam 1993:18). For St-Stephen's construction on the site of an old public building, see Jolin 1975.

251

CONCLUSIONS

In the late fifth century the region of Metz was, like the remainder of northern Gaul, incorporated into the Frankish kingdoms, becoming part of the Merovingian realm by the beginning of the sixth. This appears to have brought with it, as might have been expected, some social reorganization. This is difficult to examine in detail but it is associated with the burst back into popularity of furnished burial (sparse in the century's middle decades, perhaps suggesting that Aetius did manage to restore some stability), the creation of new cemeteries, and changes in the forms and design of artefacts (marking the beginning of Bohner-Perin period II; Ament's AM I). These developments appear to have taken place in the final quarter of the fifth century and at the very beginning of the sixth. In the region of Metz, the new burials do not appear to be common until c. 520 (the beginning of Bohner-Perin III, or Ament AM II). Sometimes cemeteries appear to be 'founded5 by a particularly lavish burial. In our region, Lavoye 319, dated to the 470s or 480s, is the classic example. Another noteworthy instance is the lavish grave of c. 520 at Kirschnaumen (Simmer 1985), of which the material is now lost. The relationship between these changes and the creation of the Frankish kingdoms is unclear. Clovis' promulgation of the Pactus Legis Salicae is probably itself an indication of social change, and of the king's desire to legitimize his rule. The new political order was responsible for these transformations according to most commentators. However, it is equally possible that the stress within these regions manifested by these changes, possibly associated with the final collapse of the old Roman elites (with the end of the possibility of any central Roman legitimization in the 470s), enabled the Franks to take over. As discussed at the beginning of this chapter, the purpose of furnished burial was public, competitive display, and should be seen in terms of the instability of 'horizontal barriers'. Of course, the act of establishing a new burial ground or custom, and of encouraging others to use it, required an exceptional display. This makes it a mistake to compare these elaborate 'founder graves' with later burials, and to derive significance from the lack of exactly comparable graves in subsequent phases (cp. Young 1986a; 1986b; Stein 1992). The ability to make such a successful display may have stemmed from involvement with the Franks, either as an incoming warrior or as a local leader accorded royal sanction. Such people may have been given fiscal or administrative responsibility for a region. In chapter 2, it was demonstrated that such power was rarely permanent, could be withdrawn as easily as it was granted, and was only with difficulty transmitted from one generation to the next. This uncertainty is the context within which we must place these founder graves. It must be mistaken to see people like the 'chiefs' of Lavoye 319, Giidingen 4 or Chaouilley 20c, or the subjects of the lavish burials at Montenach (Guillaume & Gambs 1989), as in any conventional sense 'lords' of the local lands or community. They had been people of undoubted power, prestige and even, perhaps, temporary privilege, but the effort expended on their funeral is eloquent testimony to the difficulty which their heirs had in making an effective claim to inherit that local standing. From c. 525 we can, largely through the medium of the new furnished 252

TOWN AND COUNTRY, c. 350 - c. 600

cemeteries, begin to study local society in the region, and its dynamics, in detail. The sixth-century population lived in small, dispersed settlements, either hamlets (as at Amneville) or individual homesteads, but were apparently grouped into communities by their shared use of large cemeteries. The large size of sixth-century cemeteries in the region is amply demonstrated at Lavoye, Cutry (cp. Legoux & Lieger 1989), Homecourt,4 Imling (Linckenheld 1930-1; LS:68; Toussaint 1938d:17-18), Koenigsmacker (above, pp. 107-9), Metrich (above, ch. 5, n. 12) and Puxieux.5 These communities may well have been organized within villae, possibly older, fragmented estates (with the old estate centre standing near the new cemetery, and possibly a focal hamlet). The villae probably now served as fiscal units of assessment. Thus there may have been continuity in the divisions of the landscape, but discontinuity in their internal organization, management and articulation (contra Morimoto 1988:99-101). These large but dispersed communities provide the setting for the sixth century's furnished burial rites, which, I argued earlier, required a large audience. In this society, a family's standing required underpinning through regular participation in acts of social theatre. Thus, sixth-century people who cannot yet have had important inheritable social positions are often buried with the most attention: teen-age women and young adult males. The family had to bury all its dead with the appropriate grave-goods, to the best of its ability. Furnished burial was not, as some observers (Samson 1987a) have argued, simply a matter of showing ability to inherit status. There is perhaps a correlation between the public nature of the funeral and the fact that it is frequently those who may have been most visible within sixth-century society (see below) who were interred with the greatest displays of grave-goods: marriageable women, young warriors and active family heads. In addition to the displays of grave-goods, the funerals were the occasion for feasting. The distribution of food, as gifts, served to maintain relationships between people. The use of food within such gift relationships is well attested in the literature of the period. As just one example, when Chilperic I wanted to win over Gregory of Tours, he asked him to partake of some broth which he had had prepared; Gregory, however, saw through the strategy (LH V.I8). This interpretation of the sixth-century cemetery as a field for competitive discourse leads to a particular conclusion about horizontal barriers in that century; local power which could be transmitted from one generation to the next was a rare commodity in northern Gaul. The burial was an act of maintaining status, but it could also be used to enhance or create social standing. We might perhaps see the lavish burials within cemeteries which are not surrounded by similar graves as failures to enhance prestige within the community, whilst those which are followed by similarly distinguished burials (here we might adduce the sequence of special graves at Lavoye; above, fig. 4.30) are more successful attempts. The evidence of sixth-century 4

5

JSAL 1882:188; 1889:120; 1907:240; Wiener 1888; Salin 1937:214; Toussaint 1938a:39; Hurstel 1984, vol.II:130; Halsall 1990:465-7. de Bouteiller 1866-7; BSAHM 1867:96-7; 1868:12-14; JSAL 1889:120; Hoffmann 1893; Beaupre 1897: 119; Wiener 1895:54, 59; Salin 1937:207; Toussaint 1938a:40; Clermont-Joly 1978:44-5; Hurstel 1984, vol.II:267.

253

CONCLUSIONS

cemeteries allows us to conclude that power in the locality was built expensively and precariously. This does not mean that a family might not dominate a community for several generations; simply that such domination was difficult and costly. This interpretation tallies well with the written sources, which similarly attest the insecurity and transience of power. It has been argued that at the end of the century the royal antrustiones attempted to create some group identity (Grahn-Hoek 1976:51). On the whole, however, effective social barriers were not raised by birth and descent; one could not adopt a powerful identity based upon these factors. Family identity was firmly rooted in the present, in current de facto status, wealth and alliances within the community. To create a more powerful personal or more narrowly familial identity, titles or other royal legitimation were needed. Sixth-century community dynamics can be further examined by consideration of the life-cycle, and gender differences at its various stages (for the funerary, artefactual elements of this, see chapters 3 and 4). Children do not appear to have inherited social status from their parents. Child burials are usually poorly furnished with a similar array of generally 'neutral' gravegoods. Some children, however, were buried with feminine artefacts but, as discussed in chapters 3 and 4, this might mean that they were considered as extensions of their mothers rather than that they were female. The construction of small children's gender does not seem to have been strong. A person appears to have acquired a gender identity essentially upon coming of age. Though the laws place majority at the same age for males and females, cemetery archaeology suggests something different. Women from puberty onwards receive feminine artefact types, above all jewellery, but adolescent males remain artefactually 'neutral'. I would suggest that women came of age, effectively, earlier than men. Their acquisition of the artefactual and costume attributes of female identity in the early teens tallies with the evidence adduced (above, p. 72) of betrothal at that age. If their comparatively lavish burial indicated status it was a status based upon institutionalized value to the family (it is important to remember that the community rules of grave-goods deposition reflect institutionalized, 'theoretical' social organization, every bit as much as the law-codes). As demonstrated (above, p. 60), in sixth-century community politics, marriage was exceptionally important, and marriageable, child-bearing women were important as the lynch-pins in such alliances. As counters in such politics, young marriageable women were doubtless much seen in public; the 'publicity' of their lives is reflected in the lavish display involved in their burial, and possibly by the fact that items of jewellery are the artefacts which reveal the greatest sixth-century decorative craftsmanship. Here, an anecdote of Gregory of Tours (LH VIII.21) is suggestive. Gregory tells us of the interment of a relative of Guntramn Boso's wife, in a church near Metz. Before describing her burial with 'much gold and a profusion of ornaments' he mentioned that she had died childless (sine filiis). We do not know whether or not she was still of child-bearing age, or how her lack of children related to the wealth placed in her grave, but Gregory nevertheless implies in this story that whether or not one had had children was a factor of some importance in determining female status. 254

TOWN AND COUNTRY, c. 350 - c. 600

It is also at this age that women were assigned the high, 600-solidus wergild, based upon similar factors. Just as it was argued that a woman could use institutionalized legal status based upon 'passive' value actively to further her own aims, similarly, the goods placed in her burial to demonstrate Value' were nevertheless regarded as her possessions. However, the social theory of the times, as indicated both by the law-codes and by the grave-goods customs, marked out the space within which a woman could act very clearly, in terms of marriage and child-bearing. Males seem to have undergone a long process of maturity. Hair and beard may have been ritually cut in the early teens (the ceremony serving the purpose of cementing alliances between families). Some masculine artefacts are occasionally found in their graves, but never weaponry. The latter was acquired, it seems, at about the age of twenty. The formal bestowal of weaponry (again used as a ceremony to ally families) marked the final entry into manhood. Weaponry seems especially important in the burials of young males and, if the Ennery data may be used as a basis for generalization, there was comparatively little distinction between the graves of men of this age group. These young men, with their weaponry, may have made up the gangs of equals, the contubernia of the PLS. They may also have been the pueri found in the service of others. Many of these 'boys' are armed, ready to engage in violence at the drop of a hat, and are often dispatched upon violent missions (to murder or kidnap) by their masters. The behaviour of these, and the contubernia, is not dissimilar to that of other c y o u n g warrior' age groups in, for example, East Africa. Their attendance upon the person of a count, duke, bishop, member of the royal family or other notable may have led to the term pueri being used for certain household slaves. This interpretation is by no means certain, but we should bear in mind the possibility that the term puer did actually relate to one's age. It might also be that it was the adolescents, between legal maturity and weapon-bearing, who formed these groups. For men, prestige appears to have been acquired whilst of this 'young warrior' age. For those who survived the violence of the times, success in comital, ducal or royal service may have brought rewards in gifts, offices titles and, perhaps, land. A warrior would then be able to settle down and marry. This career path is similar to that proposed by Charles-Edwards (1976) for early Anglo-Saxon noblemen. Marital age is difficult to establish. The Ennery evidence suggests a change in burial at around forty years of age. This may, however, simply reflect the anthropologist's use of twentyyear age groupings. Marriage may have taken place at about thirty, as suggested by other evidence (the poetry of Venantius; Alamannic cemeteries of Baden- Wurttemberg).6 Late marriage for men is suggested further by the demography of the cemeteries studied. At Ennery, Chaouilley and Lavoye, there seemed to be strikingly low numbers of women. At Chaouilley and, to a lesser extent, Ennery, this may be due to the incompleteness of the sample. At Ennery, however, chance finds from around the excavated section still suggested a 6

These age grades may have been clearly structured. I hope to return to this subject in a future article, drawing upon evidence from other regions of the Merovingian world.

255

CONCLUSIONS

preponderance of masculine graves. At Lavoye, the sixth-century phase is surrounded and 'sealed' by later, seventh-century burials so this explanation will not suffice. Separate burial of women is unlikely (no such female cemeteries have been found; a few buckles would be enough to date one), and there is no evidence of differential burial rites for women (cremation without urn; exposure in trees or on platforms) which might make their burials less archaeologically visible, as perhaps in Viking Age Scandinavia. The sixthcentury cemeteries of the region of Metz appear to be community cemeteries, with all sections of the community interred there. When, as at Lavoye, we have a seemingly complete sample of such a community and it includes a surprisingly low number of females, we must accept the possibility of female infanticide.7 Reducing the numbers of female children may have been done simply by differential care invested in the infants, rather than by deliberate acts of killing. It would here be coupled with the cults of weaponry and violence encouraged in young men, and the endemic warfare of the age, as a population-control mechanism (Harris 1978:41-55 for general anthropological discussion). With men marrying at, say, thirty, and women at about fifteen, the sex ratios would by then be more even. Female infanticide would demand early marriage for those who remained, and, of course, increase their value to their families. In young adulthood (defined as between twenty and forty), women were buried less often with jewellery but commonly with the most numerous grave-goods, including pottery and other items. We might detect a change in the roots of female status, from the marriageable teenager dressed up as a visible counter in family marriage strategies, to a wife and mother surrounded by the symbols of her domain, the household. These furnishings are more indicative of status based upon active worth. The evidence of the burials of people older than forty suggests further changes by this stage of people's lives. Male graves remain well furnished. Weaponry is joined by other masculine artefact-types, such as awls, strike-alights, flints, bronze needles and knives, perhaps more symbolic of household life. This is not unexpected. By this stage, a married man would have children (if he married at thirty, his eldest male children would not come of age until he was about fifty) and thus a household of dependants under his mundium. It is perhaps these married men, 'older warriors', who are the ingenui of the laws. The legal concern with violent crimes against them and crimes against their freedom represents the fact that they are legally responsible for a number of other people. Hence their legal identity is particularly important. At Ennery, at this age, male burials begin to be separated into well-furnished and more poorly equipped graves, with some entirely without grave-goods. The well-furnished burials find their extremes in the graves of the forty- to fifty-year-olds in Lavoye 319 and Giidingen 4. As stated, it is men of this age who held the most important social positions to be inherited. Their death, before their eldest sons had yet established their identities as family heads, created tension. Hence these people are buried with the most lavish displays. 7

Leaving aside the vexed issue of the Carolingian polyptychs, Burnell (1988) has reached a similar conclusion from Alamannic and Bavarian cemeteries. The Vita Liudgeri (Lebecq 1990) tells of an instance of attempted female infanticide. Such practices are well attested in rural, pre-industrial societies.

256

TOWN AND COUNTRY, c. 350 - c. 600

Women of this age, however, are buried with far fewer grave-goods. Feminine artefacts are increasingly rare, with very little jewellery, and graves of women of this age are often artefactually 'neutral' or unfurnished. By this age (if she married at about fifteen, like Vilithuta) a woman's daughters will themselves have married, and even her sons will have joined the 'young warriors'. Her child-bearing days and her responsibility for young children are past. She has probably given her female finery to her daughters, as attested by the presence of 'old' jewellery in the graves of young women (cp. Lavoye 114) and by literary anecdotes from southern Gaul (cp. Caesarius of Aries, Sermons, 78.4; LH IX.34). It was at the age of forty, according to seventh-century Lex Ribvaria (12.i), that a woman's wergild fell to 200 solidi. PLSy (65e) ascription of sixty years of age as the end of child-bearing surely lies ultimately in a scribal error (LX for XL)! The lawyers and the people of local communities placed the end of a woman's institutional status based upon child-bearing at about forty: a neat and rare correlation and one which perhaps shows that the social theory of the law-makers was interestingly similar to that of sixth-century Frankish rural communities. In old age, people of both sexes received little attention in their burials. Graves of the elderly are frequently 'neutral' or unfurnished, and often small too. Male graves further continued their separation into well- and poorlyfurnished graves. At this age, men lost their most important social roles as family heads, seeing their sons now reach the 'active family head' or 'older warrior' age. Their deaths created less tension and required less lavish marking. Nevertheless, some men retained items of weaponry and occasional other masculine items. Possibly a familial social role was replaced by a more 'communal' one, perhaps as the mysterious 'rachimbergs' or as the seniores mentioned in the literary sources. Women had long left the arena marked out for them by social institutions, that of marriage, child-bearing and childrearing. Most would be in their widowhood by now, perhaps living under the mundium of their sons, or, effectively, of the general community (as we should understand the laws' references to the king's mundium). Some, however, as we have seen, had a property base to fall back on and manage, and may have had the space to create their own social standing. Their lack of continuing, institutional status within the family may have meant that their death did not cause great tension, and was not marked by lavish funerary display, but we should not assume from that that old women had little standing in the community. Familial social roles, again, may have been replaced by communal or even religious ones. In many societies, returning to the liminal zone, also inhabited by young children, between this world and the next, the old replace worldly status with supernatural. The graves of children and old people (with the exception of male graves with weapons) are very similar. Hence, though grave-goods may no longer have been appropriate, other aspects of the funerary procedure could have been used to show respect: the size and construction of the grave, for example. This survey of the sixth-century life-cycle demonstrates something of rural social dynamics. In what Ian Wood (1986:11) has termed the 'face-toface' communities of Merovingian Gaul, prestige was acquired, not ascribed; families had to work to maintain local standing; the stages of the life-cycle were clearly marked and ceremonies accompanying them were used to ally 257

CONCLUSIONS

different families. When status within a family was lost, as with women older than forty, institutional status appears to have been lost too, though other forms remained. This point should be expanded. In the sixth century, identity was built more around one's position within a family, and fulfilment of the roles related to that position, than around membership of a particular family. As argued above, the power attached to family identity was based around the achievements of current family members, their acquired wealth, connections and perhaps titles. The temporary nature of grave-goods display is analogous to the legal procedures of the Pactus Legis Salicae. Both are concerned with the 'here and now', the use of public, ritualized, essentially temporary displays. This seems to represent a particularly transient view of time. In this society, where status could be acquired and lost within a lifetime, this failure to show a concern with projection into the future is not surprising. Where do the half-free and the slaves fit into the schema above? In chapter 2 it was concluded that slaves and laeti were probably quite few in number, and that there may have been a weakening of the barriers separating them from the free. The cemeteries appear to attest active participation in the 'discourse' of furnished burial across the community. The realization that the lavishness of burial relates largely to age and gender rather than hierarchical rank removes the old, glib assumptions that burials without gravegoods are those of slaves. The absence of separate unfurnished cemeteries (as mentioned, a buckle here and there would suffice to date one such) makes it difficult to perceive distinct slave cemeteries. If some of the poorly furnished graves of young men and women are those of the half-free, or even slaves, though this is never testable (better palaeopathology might one day show correlations between poorly furnished burial and heavier physical work), this would nevertheless show that they were allowed to participate in the life of the community, and that a peculium could be disposed of as grave-goods. Further testimony to the fluidity of sixth-century horizontal barriers would be provided. Similar points apply to the question of ethnicity. 'Romans' and 'Franks' all participated in the theatre of the cemetery. The functional difference which the adoption of sixth-century ethnic identities created (above, p. 29) might be revealed in the use of weaponry. It is just possible that 'Franks' symbolized their ethnicity and their military role by the deposit of weapons, whilst 'Romans' used other gender-specific items; there are some fairly well-furnished masculine graves without weapons (Ennery 28; Lavoye 255), though not many. Perhaps this indicates how the 'Franks' were acquiring consistently higher social standing than Romans. This is, of course, untestable, but, among younger adult males, there may be a correlation between AB.l- and AB.2-graves, and 'Franks' and 'Romans' respectively. If prestige was most often acquired through service in the army, and Romans did not often serve, then another reason for the gradual decrease of Roman social status becomes apparent. The adoption of Frankish ethnic identity, the acquisition of Frankish status, would thus be a valuable strategy in local politics, especially if it brought at least partial tax exemption. It is not difficult to see why Frankish identity should have become almost universal by c. 600. 258

TOWN AND COUNTRY, c. 350 - c. 600

Another aspect of sixth-century community dynamics concerns the gender struggle. The ways in which community mores mapped out a woman's path to prestige have been discussed. Another side to this might be suggested by the way in which, by the end of the sixth century, women had adopted certain artefact-types which were exclusively masculine in the century's early decades (the plaque-buckle at Lavoye, for example). This led to a redefinition of the masculinity of such objects, through the differentiation between decorated and undecorated, between bronze and iron, and so on. We should further consider economics. In sixth-century north-eastern Gallic society, with its stress on gift-giving, particularly in the processes of marriage, and the obsession with partible inheritance, land-units will have changed hands regularly. Though boundaries, and the general arenas within which such exchanges took place, remained fixed, the patterns of land ownership were a perpetually changing kaleidoscope, with continual build-up and fragmentation of family estates (Bois 1992:42ff for a similar picture in tenth-century Burgundy). Permanent possession of large landed estates was difficult in the extreme. The region's economy appears to have been nonmonetary, though the old Roman coins found in graves may have been used as bullion; the villae communities possibly used their shared ritual focus, their cemetery, as the location for exchange, the site's nature guarding against violation of the 'peace of the market' (Polanyi's 'magic guarantee'; Polanyi 1978). Exchange appears to have been local, and artefacts of the period seem to attest little craft specialization. Often, as with the plain belt buckles, it may have been the moulds rather than the objects which changed hands. The relationships between local society and the royal government now require attention. Those successful in creating local prestige might then, as ingenuiy attend the assemblies of the Frankish army, bringing them the opportunity to come to the notice of the kings and their officers. This could also be done at an earlier age, as stated, by serving among the notables' pueri. The young warriors who settled down on the proceeds of previous service might still attend the army and look for further rewards. The acquisition of royal titles and offices, or status as an antrustio, was invaluable in local politics, as outlined in chapter 2. Fiscal lands could be exploited; legal privilege was accorded and a particular role in surplus-extraction and the administration of justice granted. The king might also bestow upon the successful subject particular estates or villae. In the sixth century, this may well have meant the rights to royal dues from these areas, following late Roman administrative practice (Goffart 1972:172-3). Even if it were not, the apparently temporary nature of such grants would make a similar collection of customary royal dues the only practical means of surplus extraction. This thesis requires more detailed treatment than is possible here,8 but would certainly explain the ease with which 'estates' were granted from and reincorporated into thefisc.We can also understand the passage in LH VI.46 where Gregory claims that Chilperic I tore up wills which bequeathed lands to the church. According to this interpretation, Chilperic's ire derived from the fact that recipients of temporary grants of the revenue from certain fiscal lands had 8

I hope to deal with the issue more fully in a future article. 259

CONCLUSIONS

no right to dispose of these lands by testament. Interpreting royal grants of Villas' in this way also explains the nature of the villa as it emerges in the later seventh-century charters. The recourse to central authority in the resolution of local power struggles holds the key to the interpretation of figures 3.23 and 3.24. Within a zone extending twenty to thirty kilometres from Metz, conflict could be resolved by appeal to the king, duke, bishop and other magnates at Metz. Further away, local authority was more open to competition, and so more lavish funerary displays become common. These use symbols of violent power, or access to military resources (angones, shield-bosses, horse-harness), of the distribution of food and drink (buckets, bronze bowls), and balances, which might suggest control over exchange (Halsall 1992b:275), or possibly a role in the conversion of local produce into taxes. Physical geography is important. Note how the zone free of such burials extends further to the east, where communication from Metz into the Plateau Lorrain is easier, than to the west, where the Cote de Moselle presents a more serious physical barrier. This is further underlined by comparison with the region of Trier. Here lavish burials come much closer to the civitas-czpital> but in the hilly Triererland, much more difficult terrain than the region of Metz, communication would be harder, and society more fragmented into local units. Note too how these displays appear to respect Metz far more than the lesser centres of Toul and Verdun. Herein lies what I have elsewhere (Halsall 1992b:275-7) called the twoedged sixth-century social struggle. On the one hand we can see a struggle between aristocrats and their neighbours and rivals, striving to create and maintain local prestige. In this struggle, as just outlined, royal favour and support was an extremely important weapon. On the other hand we may see a conflict between these aristocrats and the kings. The kings needed local 'big men' to administer their kingdom, but worked hard to prevent the creation of an established class or noble rank between themselves and the remainder of the free population (above, pp. 36-9). The aristocrats, though needing royal support, struggled against the kings to establish more permanent, assured control of resources, and used their rural, local power bases as an instrument in this. In this society, it is not difficult to explain the nature of the settlement at Metz. The insecurity of power and the attendant importance of gift-giving and display meant that few people had any assured control of surplus or an established, landed base. This in turn will have made it difficult for the aristocracy to sponsor a revival of industries, such as stone-quarrying and transport, or craft specialization. Unless they possessed significant royally bestowed authority, they will not have been able to remove themselves for long periods of time from their localities without leaving their power open to usurpation. Hence Metz remained largely derelict, with a fluctuating, heterogeneous population. Its economic base was divided between subsistence, especially for its probably fairly dispersed secular population, and supply, for the ecclesiastical and administrative presence. However, the royal residence there will have meant that the periods when the population rose will have been more frequent, and it does not seem unreasonable to attribute the 260

TOWN AND COUNTRY, c. 350 - c. 600

fact that Metz began to recover somewhat earlier than other northern towns to its status as a royal residence. The kingdom's aristocracy, including its most powerful figures, attended court there, and competed with each other within the arena of the royal 'capital'. Hence there are traces of lavish burial in Metz (seefigs.3.23 and 3.24), attested, at Holy Apostles, by an interesting passage in the History ofSt-Arnulf, Metz (Waitz (ed.) 1879:533). Builders extending the church choir in 1239 discovered the remains of lavish graves, assumed by the History's author to be those of kings and emperors. Religious processions similarly served as fields for such competition. The episcopal policy of instituting such occasions successfully maintained some periodic high status for the site at Metz. Sixth-century society used burial and, apparently, costume as its most important areas of display of social status. This might help to explain why the settlements of the period remain so difficult to detect. Dwelling units were not yet the locus of status display, and, as stated, the social context for the construction of large stone-built or similarly elaborate houses did not yet exist. The importance of age groups alluded to earlier may have led to a physical separation of various groups within a community, as in East Africa, with younger adult males living separately in seasonal or periodic settlements, and to a somewhat fluctuating, if homogenous, population for the other settlements. Society in the sixth-century region of Metz was structured around identities based upon ethnicity, gender and place in the life-cycle, and status related strongly to the achievements made whilst carrying out the functions associated with these. Power, however, was transient, and it required skill and expense to pass it from one generation to the next. To be successful, a man had to transmit his power to his heirs during his lifetime. Death before this was done created stress and was likely to lead to his social position being usurped by others. This is attested by written and archaeological evidence.

261

The later Merovingian period

Change around AD 600: society in the age of Dagobert A long list of transformations can be loosely dated to the period c. 580 c. 620 (Halsall 1992b is a preliminary outline). The chronology is not always exact, however, and some may have occurred later in the seventh-century (see below). Description and explanation of these changes is intricately bound up with an account of the social structure which emerged from these critical decades. The laws issued at the end of the period of change, Chlothar IFs Edict of Paris and Lex Ribvaria, suggest changes in the nature of horizontal barriers between the aristocracy and the remainder of the free; in the relationships

Audun-le-Tiche Hayange •

Saarbriicken Verdun

Altheim

Gorze

FrouardV.

Tarquimpol

Nancy Toul

Fig. 9.1 262

Places mentioned in chapter 9

THE LATER MEROVINGIAN PERIOD

between the kings and the aristocracy; in the nature of ethnic identities, with Romans descending formally to effectively dependent status; in the form of the barriers between free and slave, which seem (in some respects) to have weakened with the presence of more numerous semi-free classes; in the discussion of the family; and so on. The creation of the new code, Lex Ribvaria, might itself be seen as indicative of the social transformations of this period, as was argued for the Pactus Legis Salicae. Royal power remained strong after Chlothar IFs conquest of Austrasia, his reunification of Gaul and his issue of the Edict of Paris, an attempt to resolve the difficulties of the previous decades. However, the relationship between king and aristocracy had changed. The latter were a very different body from their sixth-century predecessors. Aristocratic power was more established, with legal recognition of ties of dependence between them and lesser freemen. Their land-based power was also greater. The beginnings of the survival of charters suggest important changes here, as noted in chapter 2. Royal grants in the sixth century seem to have been temporary contracts, revocable at will by the king, and may have been gifts of the taxes due from a villa, rather than of outright ownership of the land. In the seventh century, gifts of land by charter made these lands into a part of the aristocrat's heritable patrimony, held in perpetuity and of which he was in turn free to dispose. Hence the charters which begin to survive from c. 600 are often aristocratic; the few surviving sixth-century north Gallic deeds (none of which is of unimpeachable authenticity) are exclusively royal. The basis of aristocratic preeminence in the localities had changed. From being built upon constant expensive renegotiation through gift-giving and display in the acts of social theatre, it now stemmed from outright ownership of more of the region's surplus-producing assets. Lands acquired in this way would sometimes be worked by slaves, but it is easy to see how free occupants would be reduced in status, becoming the aristocrat's tenants. Even if there was little change in the dues collected, this was a relation of dependence which would not and did not occur where the aristocracy had simply acquired temporary rights to collect certain royal taxes from a villa's inhabitants; hence it does not appear in the Pactus. In the latter case, the aristocrat was effectively a royal servant, but all remained freemen together. Nonetheless, the aristocracy was not yet entirely removed from the remainder of free society. Other strategies were necessary to maintain its position. One was to foster a greater class identity. This, it seems, was done by employing, within the group, the systems of alliance which had been used within local communities in the sixth century. Marriage was one strategy (see Weidemann 1987 for north-western Gaul). The marriage of Pippin cthe Old's' daughter, Begga, to Ansegisel, the son of Arnulf of Metz1 is perhaps 1

The relationship is first mentioned in Pertz, Mayoral 2 (villa Nielsio, 20 Feb. 685, though the charter does not survive as an original), issued at about the time of the death and burial in St-Arnulf of Arnulf's putative elder son Chlodulf. Here Pippin 'IF refers to Arnulf as avus noster. That apart, beyond problematic charters such as Pertz, Spuria 7, the relationship is not explicitly attested until Paul the Deacon's late eighth-century Liber de Episcopis Mettensibus. I am not convinced that this is reason to doubt its historicity. The juxtaposition of Chlodulf and AnsegisePs names in the list of advisers in Pertz, Royal 29, suggests that Paul was correct in making them brothers, and that his knowledge of the Arnulfing family tree was not as suspect as has recently been proposed. Paul (pace Wood 1994:259 n.35) never

263

CONCLUSIONS

the most famous example. Other aspects of fictive kinship were employed; Arnulf was sent to the subregulus, Gundulf, to be brought up. The age groupings which may have been so important in sixth-century community organization were also used to bind aristocratic children. Thus the young aristocrats sent in the early seventh century to the schola palatina, including Goeric-Abbo and Chlodulf of Metz, Eligius of Noyon and Desiderius of Cahors, formed a bond which united them through their subsequent years, as Desiderius' correspondence demonstrates. The new aristocracy's identity was further fostered by its adoption of new forms of monasticism, particularly that associated with Columbanus. Arnulf ended his days as a hermit in the Vosges; his friend Romaric founded the similarly Vosgian monastery of Remiremont (VRAH); Arnulf's kinsman Bertulf became abbot of Columbanus' monastery of Bobbio (Vita Columbani 11.23). The outpouring of literature venerating these 'noble saints', stressing their high birth, was also important. The control of lands which 'family monasteries' gave to aristocratic factions is well known (Gerberding 1987:96-105), and their removal of exclusive control of sanctity from the urban, episcopal arena is noteworthy. However, it is again worth stressing that, unlike the neighbouring regions to the north, north-west and east, the region of Metz saw no foundation of aristocratic rural monasteries. This is important. The see of Metz remained the single most important ecclesiastical landholder in the region, and was, as argued in chapter 1, kept under royal control. As for the remainder of the region, this may have remained largely under the control of the royal fisc. The aristocratic estates attested in later Merovingian and early Carolingian charters make little headway into the Plateau Lorrain until the ninth century. If so, this might account for the lack of aristocratic rural monasteries, and also, perhaps, explain why some late Merovingian social developments, especially concerning the aristocracy, may have appeared later in the region of Metz than elsewhere. Cemetery evidence allows us to examine further the relationships between the aristocracy and the remainder of the free. There is a decrease in the numbers and variety of the grave-goods chosen to furnish burials, and on newly founded sites pottery is scarce. There also appears to be a greater correlation between grave-goods and straightforward wealth (possibly explaining the 'normal curve' in fig. 4.28, c-f). The grave-goods custom still appears to result from display, but the element of competition seems reduced. Despite some adherence to sixth-century customs, lavish groups of burials stand out by including subjects of all ages and both genders; they were less subtle than before. Sixth-century lavish graves, on the whole, played within the general community mores governing grave-goods deposits. Those which broke mores (e.g. Ennery 70 and 71) were nevertheless quite subtle. Those which were not (Lavoye 319 and 189) are almost invariably 'founder graves' of some sort. Seventh-century graves, on the other hand, demonstrated wealth with all of the family's dead, even, as at Audun-lementions AnsegiseFs marriage to Begga, though the Liber de virtutibus sanctae Geretrudis (c. 2) suggests some connection between Chlodulf and Pippin 'the Old's* family. As outlined in chapter 1, Arnulf may not have been especially popular in Metz, and his cult was slow to develop. Given Metz's political importance, he seems an odd choice as a fictitious ancestor.

264

THE LATER MEROVINGIAN PERIOD

Tiche, old women and adolescent males. At Lavoye, we see lavishly furnished children's graves from this period. As noted earlier, this represents a change, and seems to stress the family's local standing further, suggesting that the status brought by family identity was inherited and retained throughout the life-cycle. The increased permanence of family status shown by the grave-goods and by the acquisition of more stable possession of land enabled early seventhcentury aristocrats to project their place in the local landscape into the future in a way which had simply not been practical for their sixth-century predecessors. The latter had seen personal and family power built up and lost within a lifetime; it is small wonder that their concerns were, on the whole, with the present. As noted in chapter 2, it is from this period that our earliest surviving documents date. When coupled with the other changes in this period, this begins to look less like the result of chance or coincidence. This was a period of change across Merovingian Gaul. Goffart (1982a) dated to this period the collapse of certain royal rights of taxation. Earlier (1972) he noted how the earliest documents of immunity which Flodoard of Reims could find in that see's excellent archive dated to about the 590s, whereas the first recorded attempts by the bishops of Reims to organize their landed estates were datable to the early seventh century (ibid.:374; see also Goffart 1982b:74). This supports the general thesis proposed here, that it was in these decades that documents began to be retained permanently by north Gallic individuals and ecclesiastical institutions. The counter-argument noted in chapter 2, which sees the survival of documents as a function of the collapse of the gesta municipalia can be weakened by consideration of a wider array of evidence. As has been described, Metz remained largely derelict for most of the fifth and sixth centuries but began to revive in the late sixth century and underwent a clearer resurgence in the early seventh. We must therefore accept that gesta municipalia survived 200 years of stagnation when it is difficult to demonstrate the continuity of any late Roman urban institutions in northern Gaul, only to die out just as the town revived. If this were the case, it would only make the change from communally organized to private archives even more significant. The survival of Merovingian records (or copies thereof) in northern Gaul from around AD 600 represents a real change in practice associated with a change in the concepts of time. In chapter 2 the correlation was observed between the date of our oldest north Gallic deeds and the beginning, in Lex Ribvaria, of a legal concern with written documents. There was now a purpose in keeping documents, which could be used as instruments of proof, and to modify the institutionalized barriers on the csocial map'. A change can be seen in the laws from an exclusive concern with the present, with public ritual and with the use of living witnesses as the sole means of proof, to an increased projection of situations into the future.2 A charter proved 2

Gregory's discussion of the trial of Egidius of Reims is important for the light it sheds upon royal and episcopal archives, and shows the legal use of documents in sixth-century Metz (Z.//X19). The documents in themselves do not prove the case, having, as in Roman law, to be examined and authenticated. In this trial, though, the desire of the Austrasian court to rewrite the history of Childebert II's minority outweighed any considerations of the actual evidential value of these documents (Wood 1993).

265

CONCLUSIONS

ownership forever, unless superseded by a later charter. The conception of the future is also attested by Lex Ribvaria (c 63). Where documents are not used to mark landownership, young boys are brought in to witness the transaction and boxed on the ears to make them remember it. This is an important statement about concern with the future, also demonstrated by Lex Ribvaria 's interest in the status of the descendants of mixed marriages {Lex Ribv. 61) and its statement (ibid. 12.ii) that, where a 600-solidus fine could not be paid, the guilty man's family could pay over three generations. Both compare interestingly with the Pactus Legis Salicae which, on the one hand, shows no concern with mixed marriages' offspring, and on the other, spreads liability for heavy fines across an ever widening circle of living kin. Seventh-century aristocratic awareness of the future may also have resulted in an increase in the use of documents, leading, in the later years of the century, to the first north Gallic formularies. This change in the views of time is supported by changes in funerary practice. In the seventh century we can see a gradual change from the transient, temporary display of gravegoods, to a concern with more permanent, above-ground markers: sarcophagus lids visible at surface level, gravestones, walls around groups of graves and, above all, churches. This gradual change is best seen at Audun-le-Tiche (above, ch.4). Other aristocratic strategies can be suggested from the cemetery evidence. As shown in chapters 3 and 4, seventh-century graves show a marked standardization, with most males being buried with a plaque-buckle. These new types of buckle, with intricate inlaid patterns, suggest increased craft specialization. If an aristocrat could sponsor the production of these buckles, evidently so important in the construction of seventh-century male identity, he might control their distribution, using them as, in a way, prestige goods, and so increasing other community members' dependence upon him. Nevertheless, it is worth noting that, on the whole, the local elites were still buried with their communities in the region of Metz. In other regions of the Frankish world, such as Alamannia (Christlein 1991:50-62), a removal of aristocratic graves away from the community cemetery took place in the early seventh century, and was often associated with extreme displays of prestige; exceptionally lavish funerary assemblages, horse-burials and the construction of barrows and even churches (the process is not dissimilar to that found in Anglo-Saxon England at the same time). Apart from those who built family churches in Metz (see below), and these may have come from across the kingdom, there is little yet to suggest aristocratic separierung in funerary matters. We can see the growing importance of family identity amongst the remainder of the community in the increasing organization of cemeteries into what appear to be familial plots (as at Audun-le-Tiche, Lavoye, Altheim, etc.). At Hayange, one family may have created a very distinct burial plot, with separate orientation (above, p. 102). This, as mentioned, tallies well with the changes in the concerns of the law-makers towards the status of descendants. The greater standardization of deposits also suggests reduced competition, Nonetheless, to my knowledge, in Gregory's recordings of land disputes, in no instance does possession of a deed decide the issue. This requires further research.

266

THE LATER MEROVINGIAN PERIOD

and that standing in the community was more fixed. The break-down of communal cemetery organization, in rows and with apparent rules governing the location of women and children's burials, fits nicely with the weakening of the communal mores governing what grave-goods were appropriate to which kinds of people. Consequently, we might also see a weakening of identities based on communally ascribed social roles, founded upon gender and age, in favour of family identity. Seventh-century grave-goods, like those of the fourth, suggest greater stability. We might also consider briefly changes in the symbolism of certain grave-goods. Weaponry in graves, it was suggested above (ch.8), was, when it appeared around AD 400, symbolic of military power, or access thereto; in the sixth century it was argued that it might have symbolized a right to fight in the army, and thus have had something to do with Frankish ethnic identity. The spread of weapon-burial noted at Lavoye in the seventh century, when most adult males are buried with scramasax or spear, might reflect the spread of Frankish identity, usual in northern Gaul by the earlier seventh century. If so, following the equation of Frankish identity with 'freedom', it might have taken on a new meaning, symbolizing free status as opposed to half-freedom or slavery. Again, this is untestable, but it seems plausible, and invites us to remember that the semiotics of material culture are rarely static. These changes seem to have made class and, at the lower reaches of the social hierarchy, rank identity more important as society became more stable. The increasing existence of freemen in relations of dependence to other freemen, whilst probably weakening the barriers between poor free, half-free and slave in terms of their everyday existence, seems to have led to added importance being given to legal differences between these identities. Social transformations had their effects upon economics and the settlement pattern. In the decades around 600, many new cemeteries were founded (fig. 5.4), often, it seems, smaller than sixth-century sites. Some, like Audun-leTiche (itself unusually large for the seventh century), were founded, like the earlier sites, by lavishly furnished graves. This might suggest a weakening, if not a fragmentation, of the old villae communities, with individual settlements creating their own cemeteries and even locating them within the ruins of earlier habitations. This is associated with the other evidence of a weakening of the community vis-a-vis the family. It was perhaps at this point that the old Roman villa sites were finally abandoned to funerary use, the implantation of cemeteries in ruins possibly destroying vestigial traces of earlier occupation. Associated with this may have been a slight expansion of the settlement pattern, with new settlements, like Frouard Phase I (Henrotay & Lansival 1992:336, 348, 351), being more archaeologically visible. In Metz itself there appears to be an increase in church foundation, with written and archaeological evidence at St-Pierre-aux-Nonnains. Remains of funerary activity in the town increase, as do traces of occupation. Churches were also founded in the castra, as for example at Tarquimpol (above, p. 208). In the economic sphere, increased stability of social structure allowed individuals to remove themselves from subsistence and engage in greater craft specialization and industry. The use of stone sarcophagi resumes, suggesting 267

CONCLUSIONS

a revival of organized stone-quarrying industries. The increase in church building would support this. Though, of course, Merovingian churches could well have been wooden, the upsurge in construction, especially on aristocratic initiative, would still argue for a greater 'industrial3 network and the necessary social context. At the same time new artefact forms and designs appear (marking the beginning of Bohner-Perin IV; Ament JM I). The elaborate new plaque-buckles and inlaid disc brooches suggest increasing craft specialization in the early seventh century. Associated with this is the revival of a money economy. In the early seventh century, or at the very end of the sixth, a number of mints appear throughout the region, striking increasingly debased gold trientes. The triens, a third of a solidus, was of much greater practical value than the royal gold solidi of the sixth century, which had been few in number and probably circulated only among the highest social strata, largely as vehicles for royal propaganda. Seventh-century coins do not bear royal busts or names, however; simply the names of the moneyer and the mint. Durliat (1990:118— 20) argues for royal economic initiative behind these coins, because of their standard weight, but, given the complete lack of evidence for a regular north Gallic money economy between c. 400 and c. 600, he stretches his point when he proposes direct continuity of monetary policy from the late Roman Empire to the seventh century. These economic developments were doubtless sponsored by the more securely established aristocracy, which used them to further their own ends. The use of specialized metalwork has been described. Stone-quarrying and working was used to demonstrate family prestige both through urban church building, and in the provision of stone funerary monuments, especially sarcophagi, in town and country. We can also see the revival of coinage as part of this process. It certainly bears witness to a livelier system of exchanges, notably around the salt-producing areas. In this context there might have been an upturn in the activity in the Saulnois, with salt being produced for exchanges further afield (though Gregory, VSM IV.29, mentions a salt-trader in Metz in the late sixth century). The use of coinage enables exchange over longer distances, with people over a wider area accepting the guarantee of the moneyer's name and the mint-mark that a coin is of a standard weight and value. Even so, the moneyer's names and the spread of the coins still suggests that exchange was generally intra-regional. Outside the region of Metz, these coins only seem to have spread north, down the Rhine and ultimately to Anglo-Saxon England, probably as bullion (Stahl 1982:139). More permanent control of rural land bases and of agricultural surplus had enabled aristocrats to sponsor craftsmanship, manufacturing and industry. This in turn is to be associated with the upsurge of church foundation in Metz, and perhaps in the via and castra. The aristocracy's more assured position also meant that some could remove themselves from the rural communities more often to spend time and money in towns, competing, at court and in monumental building activities, no longer with the local community, but within their peer group. Change in social organization, towards more socially and economically distinct groupings, was probably reflected in

268

THE LATER MEROVINGIAN PERIOD

changes in the settlement pattern, towards greater diversification of settlements through their economic basis, population and status. Seventh-century society in the region of Metz seems thus to have differed greatly from its sixth-century predecessor. The hierarchy seems to have been more stable. Horizontal barriers were more rigid and class and rank identity more powerful than before, as, in obvious connection with this, was family identity. The construction of gender appears to have changed, too. The increasing importance of descent seems to have switched the primary locus of lavish bodily adornment from young marriageable women, the lynch-pins of marriage strategies, to adult males, potential fathers and ancestors. Changes occurred in the nature of female inheritance, with an increased desire to keep ancestral land in male hands. The aristocracy in particular had greatly increased stability and power. New relationships were being forged to bridge these horizontal barriers, based upon links of dependence and helping to break up the old large rural communities. Nevertheless, the king remained powerful, and most aristocrats in this region were still intimately part of their local communities. What had brought about these changes? The catalyst is surely to be sought in the series of minorities between Sigibert Fs assassination in 575 and the invasion of Austrasia by Chlothar II in 613. Of these thirty-eight years, Austrasia was ruled by a minor for about seventeen (575-85, the minority of Childebert II; c. 596 - c. 602, that of Theudebert II; and 613, the short-lived minority of Sigibert II). To this we might add that it is difficult to know exactly how effective a fifteen-year-old king was, even if he had reached legal majority, and that Theudebert II (who ruled until 612) was, according to Fredegar {Chron. IV.35) 'simple-minded5. In this period, rival factions fought for control of the royal court. These doubtless rewarded themselves and bought the support, or acquiescence, of others with royal privileges. Queen Brunechildis herself, who maintained a striking grip on Austrasian politics for much of the period between 575 and 613 (Nelson 1978) similarly had to purchase loyalty. Her sponsorship of Berthefried's daughter at the font may have been an attempt to win support by bestowing afictivekinship alliance with the royal family. The importance of aristocratic loyalty is, of course, amply demonstrated by Brunichildis' ultimate fate in 613, when Arnulf of Metz and the Austrasian magnates invited Chlothar II to take over the kingdom. Both Chlothar II (in 621-2) and Dagobert I (in 632) had to accede to the Austrasian aristocracy's demands that they be given their own king (in the latter case one who had not reached his majority, and in the former one who had been brought up by the leaders of the Austrasian aristocracy). The process of social change was doubtless influenced by contacts between Austrasia and the south, where, in the Auvergne, Aquitaine and Provence, there had been greater continuity of the Roman social system and its nobility. In Burgundy, there existed a Germanic nobility, the Burgundaefarones, alongside the old Roman senatorial class. Austrasian aristocrats like Guntramn Boso served in central and southern Gaul; some were rewarded with estates there; the Austrasian churches certainly held estates in the

269

CONCLUSIONS

Auvergne and the south. The spread of Auvergnat, Aquitainian and Burgundian saints' cults north to Metz, especially in the late sixth and seventh centuries, also demonstrates these links with the south. Lex Ribvaria, as long noted, shows the influence of Burgundian law. It may be that the appearance of documents, many of which are traceable to Roman originals (Classen 1955; 1956), and which served new aristocratic needs, also results from southern influence, rather than coming via gesta municipalia, surviving by an obscure process through the dark years of the fifth and sixth centuries. By c. 620, the aristocracy had won both sides of the sixth century's twoedged social struggle. They had acquired a degree of independence of royal authority, and more assured local preeminence. We can justly think of the changes which were wrought between c. 580 and c. 620 as a revolution. They radically altered the nature of social structure and the bases of power. Like the 'feudal revolution' of late tenth-century France they involved an increase in unrestrained violence (the period of change saw wars of unusual ferocity at its beginning and end, in the bloodbaths of 574-5 and 612-13), a change of royal dynasty (in 613, from the descendants of Sigibert I to those of Chilperic I) and an ideological element, incorporating new religious observance (the literary output praising the new 'noble saints', and the association with Columbanus' monastic reforms).

Later Merovingian society in the region of Metz The period between c. 650 and c. 750 was one of constant change, which is difficult to separate into discrete stages. The changes of the later seventh century, though not as dramatic as those of the turn of the century, were in many ways very important indeed. As outlined in chapter 5, the landscape which emerges in the charters of this period was a patchwork of free and dependent farms (the latter worked either by free or half-free tenants or by slaves), organized into villae, sometimes called fines or marca, small regions within which several people might hold land. If villa could denote a single community, then it may have been used to describe some of the newer, smaller communities with their own cemeteries founded around 600. This might have led to the use of fines and marca to denote wider areas, and may be behind the rather fuzzy overlaps between the different terms. Some aristocrats owned numerous farms in widespread villae, their holdings being scattered amongst the farms of independent free smallholders. If some villae were single estates, this may have been because they had been transmitted intact direct from the fisc to the aristocrat in question. Aristocrats' control of surplus may still largely have been in the form of what had originally been royal taxes, even if their 'eminent' ownership of the land meant that they could reorganize it as they wished. In the 'patchwork' of landholding suggested above, in which, it seems, very large contiguous estates were rarely built up, it is probable that the levying of formerly royal dues from scattered lands, using the mansus as the basic unit of assessment, was the only practical way of extracting surplus. Aristocratic control of salt production in the Pagus Salinensis also 270

THE LATER MEROVINGIAN PERIOD

falls into this framework. There may nevertheless have been some reorganization of the rural settlement pattern. Some elements of the 'Grand Domaine' may be traced to the late seventh century, and Eply-Raucourt cHaut-deVillers' seems to date to this period. From about the death of Dagobert I (638), the term nobilis begins to be used for Frankish aristocrats, and in the later century magnates are identified as the son of their father (above, p. 46). Isidore's criteria for nobility are beginning to be fulfilled. It might be significant that it was only in this period that rural monasteries began to attract donations of land from the region's aristocracy. Duke Boniface, active in Alsace and in the Pagus Salinensis during the reign of Childebert 'the Adopted5 (Gerberding 1987:54-5), was associated with Wissembourg, founded at this time. By 700, Wissembourg received growing support, especially from the Gundoinings and their allies. Urban monasteries may have been founded and endowed with rural estates earlier, but this seems to be the first attempt to break away from the episcopally dominated urban arena. As noted, it took place here slightly later than elsewhere in northern Gaul, and probably for the reasons mentioned above, a continuing royal and episcopal control of land. The lack of rural family monasteries is significant in other ways. Ifprecaria were used to consolidate family estates against divided inheritance (above, p. 52), an interpretation which explains Eleutherius' murder by his kin in the second version of the Vita Waldradae (above, p. 61), then this option was rarely open to the aristocracy of the region of Metz. The see of Metz was under royal control, and urban monasteries remained under episcopal supervision. The only means of preventing the constant fragmentation of landholdings was by using written documents to modify the demands of custom, as discussed in chapter 2, hence the further increase in surviving documents. By the mid-seventh century, the new mints and coinage discussed earlier died out, to be replaced by silver denarii struck at a single mint at Metz. This would appear to mark economic change (cp. Rouche 1986). Instead of higher-denomination currency struck at numerous local centres, and probably used in the relatively confined areas within which the minters were recognized, coinage was issued which was evidently valuable across far wider regions, hence the reduction in the numbers of mints. This and the still lower denomination of the currency testifies to an increasing monetization of the region's economy. This development would have opened up the landmarket, as witnessed by sales of land recorded in the charters. Aristocrats could now, even more than before, purchase lands to replace those lost through division or gift. In the area of funerary behaviour, we can see the continuing decline of grave-goods. By this time there are very few significant displays of funerary furnishings. Those which remain are extremely standardized, as can be seen at Moyeuvre-Grande and elsewhere. By this time, weaponry and jewellery are hardly ever buried. Instead, the movement to above-ground monuments increased, as demonstrated clearly at Audun-le-Tiche, area 2. Here we see possible family plots marked out by stone walls, the reuse of Roman sculpture as grave-stones, as well as the construction of new grave-markers such as the stone cross, and decorated sarcophagus lids. 271

CONCLUSIONS

It is in the second half of the seventh century that we might see the removal of the local aristocrats from the rest of the local community in matters of burial. It seems that it was in this period that the rural elites built churches in the countryside. The Wissembourg charters show early eighth-century aristocrats giving the abbey churches built by their grandfathers (Wiss. 234 and 237). The church at Mousson seems to have been built in the second half of the seventh century. The earliest possible grave there dates to c. 650 (Cuvelier 1990:165). Cuvelier (ibid.:lG5-G) dates the church to 580-650 but I see no reason to make the building precede the first burial by so much. As he argues (ibid.:l67), it is unlikely that the church was a parish church. Though he postulates a monastic function, the range of possibilities discussed (a parish church, a garrison church (?), or a monastic church) is not wide, and his argument is not compelling. An aristocratic funerary church would appear more likely. Audun-le-Tiche 103 is a graphic illustration of the tensions involved in aristocratic funerary separation. The burial of two young men, with heads removed, might be interpreted thus: their family, clearly of local standing, still paid lip-service to burial with the rest of the community, thus disposing of the bodies with, for the time, a lavish display of grave-goods, in an elaborate tomb marked at surface level by a stone surround; however, their heads were removed before burial and interred separately, in a new aristocratic funerary church. Thus it is only in the second half of the seventh century that aristocrats in the region of Metz commonly felt able to distance themselves from their local communities. This dating is 100-150 years later than the conventional chronology for such processes (cp. Young 1986b; Stein 1992), but this, I would argue, is because the latter is based upon a misunderstanding of the nature of grave-goods, crude comparisons of 'founder-graves' with later burials, too unsubtle searches for prestigious graves and a misconception of the nature of the sixth-century Frankish aristocracy. At a local level, aristocratic dominance was built on assured control of surplus and economic independence, in the form of rents in money and produce. These rents may have served to maintain forces of household retainers. However, aristocratic lands were still likely to be broken up by partible inheritance, and there were few means of combating this. Furthermore, support was bought with, and service rewarded by, outright gifts of land. Estates thus continued to build up and fall apart, though the component lands circulated within a narrower section of society than before. This helped to prevent the establishment of large concentrated estates on the classic, 'bipartite' model. The consolidation of large holdings was thus still arduous, and aristocrats were only slowly, and with difficulty, able to remove themselves from their local communities in this period. Consequently, the power of families like the Gundoinings was, at least in part, built upon their acquisition of titles such as count, still bestowed by the court, even if only as a formality, and often under the effective control of the mayor of the palace. Control of such offices helped to bring support together. The Gundoinings appear to have been at the centre of an Austrasian landholding faction (above, pp. 50-1), but this seems fairly fluid. Their leadership was probably based around their comital and ducal titles and 272

THE LATER MEROVINGIAN PERIOD

privileges. Gerberding noted that aristocrats still needed Merovingian kings as figureheads to unite their factions: 'without the aura of royal leadership, few would rally to the cause and the movement would gain no scale' (1987:83). One might say the same about royally bestowed titles. These titles and the resources of which they brought control enabled aristocrats to recoup the losses of land through gift and partition. Noble power may have been more assured, but aristocratic factions remained insecurely bound together. The increase in aristocratic power from the middle of the seventh century might be associated with the rapid collapse into royal minorities and political confusion after the death of Sigibert III (c. 651); the usurpation of Grimoald, the murders of Childeric II and Dagobert II, the civil wars between Ebroin and his opponents, and so on. In this period, royal weakness resulted in further strengthening of aristocratic control in the countryside. Again, the royal court could be dominated, and support bought by continuing grants of privileges and titles. Whereas royal power had recovered after the problems of the period c. 580 - c. 620, the contestants in the royal-aristocratic struggle were rather different in c. 650. The aristocracy would not this time relinquish the political power it obtained. Nevertheless, the Merovingian court retained much power (above, pp. 49-50), and an energetic king like Childebert III might still, with tact, wield considerable authority. Further changes in the nature of social organization came about in the region of Metz at about the time of Childebert IIPs death. Some are difficult to date exactly; they may have spread over a number of decades, and might therefore continue the changes of the later seventh century. Around 700, most of the old cemeteries were finally abandoned. Some which had churches built upon them continued. Otherwise burial moved to new graveyards around churches. This might represent a culmination of the processes which began in c 600. First, the elites demonstrate their power by exceptionally lavish funerary displays to the local audience, disregarding community rules. The next stage is represented by physical removal of aristocratic burial away from that of the rest of the community to separate cemeteries, especially in and around churches. Finally, when elite power is even more entrenched, the community grave-yards' removal to the aristocratic churches is allowed. The aristocracy controlled these churches, and the churches controlled the rites of burial from around this time. In this context it is perhaps not unreasonable to see some significance in the appearance in this region of the terms precaria and, especially, beneficium. The former, resulting from association with rural monasteries like Wissembourg, permitted the stabilization of aristocratic estates. The latter, beneficium, as noted, heralded a new kind of landholding relationship. More powerful landowners granted land to followers on a temporary basis, for a fixed number of lifetimes, perhaps the ultimate expression of aristocratic projection into the future of status in the landscape. This way, a large landed base might be kept intact whilst support could still be bought. The ties of dependence so created would give more cohesion to aristocratic factions. Some reorganization of the landscape might be suggested by the fact that Phase II at Frouard dates to about 700 (Henrotay & Lansival 1992:348). 273

CONCLUSIONS

Here a slight digression is necessary. The developments in the nature of horizontal barriers and in the security and permanence of family prestige are clearly interdependent, and surely affected other identities too. As a more clearly defined social hierarchy began to emerge, and as fulfilment of generally ascribed roles within families became less important than membership of particular families in the construction of personal identity, then it is likely that the construction of gender began to differ at the various levels of society (as noted briefly above, p. 72). This is unfortunately difficult to examine in any detail or with any certainty, but the charters provide one (albeit extremely problematic) starting point. Study of the Merovingian charters of Wissembourg Abbey reveals a superficially odd reversal of the norms of early medieval history: the slave women of Merovingian Lorraine are more historically visible than their free and aristocratic counterparts. These documents name almost five times as many female slaves as free or aristocratic women. Is this significant? Of course, dependent and unfree women were recorded because of the value of their labour to their masters and owners, whereas freewomen did not attend the assemblies where the charters were witnessed in an active capacity. A single explanation, based upon the patriarchal nature of society, can be adduced for both the recording of the slave women and the non-appearance of the free. Though this explanation in no way contradicts what we know of later Merovingian society, it is not, I think, entirely satisfactory. The role of work as a determinant of female status may hold the key. Amongst the dependent, peasant population, as in similar societies, women may well have been valued for their work, at home and in the fields. Unmarried daughters especially formed a vital element of the peasant domestic labour force, but it is unlikely that work would play an identical role in gender construction at the social level of families like the Gundoinings, with their diverse estates, rents and resources, control over land and men, and indeed ability to give away whole families of dependants. Amongst the aristocracy, women doubtless played a crucial part in the control of labour, and the management of the household. There are general analogies, but it is mistaken to equate this directly with the indispensable role of women's physical labour in subsistence agriculture. Amongst the aristocracy it may be that, in the formal construction of gender, a woman's work was less important than her role within the family, as wife and mother and as the lynch-pin of marriage alliances between families. The roots of female power would be similar, at this level, to those of sixth-century society, when archaeological evidence shows that the undoubtedly important work of women was subordinated to their roles as marriage counters and mothers in the institutionalization of gender roles. This, for aristocratic women, may have been the end result of the increased security of family identity. Thus any part they played in the alienation even of their own property was generally confined to undocumented 'behind the scenes' influence. On the other hand, in this period and in this context, to the authors of the charters and their aristocratic informants, unfree women appear to have been as valuable, and as worthy of record by name, as their male counterparts. To these men, in the array of identities possessed by any individual, an unfree woman was a slave first 274

THE LATER MEROVINGIAN PERIOD

and a woman second, whereas an aristocratic woman was a woman first and an aristocrat second. The significance of this is further suggested by contrast with slightly later evidence (see below). Nonetheless, the many problems involved in dealing with charter evidence render this no more than a suggestion. The changes around 700 might be associated with the civil wars of the period, which established the power of Charles Martel after the death of Pippin CIF. Royal power was almost non-existent after Childebert Ill's death, which signalled the break-away from the Merovingian kingdom of areas like Provence. These wars resulted in the final Arnulfing triumph over their rivals but, not surprisingly, this was a period of great stress in the region, with anti-Arnulfings (or at least opponents of Charles Martel and his sons) such as Count Wulfoald fortifying their own strongholds. In chapter 1, it was noted that an upsurge of donations of land by the Gundoinings and their allies to Wissembourg Abbey took place in the second decade of the eighth century, the period of Charles' wars. In this situation new sources of power were sought. According to Marculf's formulary, the terms precaria and beneficium may have appeared earlier in other regions of northern Gaul. The absence of precaria can be explained by the peculiarities of the region, with its lack of aristocratic Eigenklosters. The chronology of the appearance of benefices in the region of Metz, around 710-20, is not untypical. The decision to adopt new vocabulary is itself significant. If benefices did appear slightly later than elsewhere in the region of Metz, this might again stem from the differences in aristocratic power there. It has been suggested that the fisc and the probably anti-Arnulfing (or at least anti-Charles Martel; above, pp. 16-17) see of Metz retained control of much of the region. Metz, as argued by Gerberding, was late in joining Charles Martel.

The change of dynasty: revolution or evolution? The developments of the early eighth century had, as their eventual result, the change of dynasty from the Merovingians to the Carolingians. The new system of benefices enabled aristocratic factions to be bound by ties of dependence, independent of royal legitimation. For the first time the king and the royal court became irrelevant and so, when Theuderic IV died, in 737, Charles managed without a king at all, something which his father had never managed to do. Wiss. 241 is dated 'in the first year after the death of King Theuderic, Charles being Mayor of the Palace'. This, needless to say, is practically unique in early medieval history, and was certainly without precedent in the eighth century. Though Pippin the Short and Carloman put a Merovingian, Childeric III, on the throne after Charles' death, presumably to make sure of their position, the experiment was short-lived; Pippin had Childeric deposed and himself proclaimed king, with papal blessing, in 751. This interpretation, placing Merovingian obsolescence only in the early eighth century, is, these days, hardly novel. Nevertheless, the change of dynasty was not achieved without stress. In chapter 2 it was noted that in a short period in the 770s 275

CONCLUSIONS

to 790s almost all of the father-son relationships attested in the Gorze cartulary bear witness to the use of tertiary patronymics, in this case employing the suffix of the father's name. This burst of patronymics is unusual. Is it significant that these sons, whose names make a direct link with those of their fathers, will have been born about the time of Pippin's usurpation? In part of the region of Metz the change of dynasty produced a concern to demonstrate paternal descent in naming practices. Like that in 613, in the region of Metz the change of dynasty in 751 was accompanied by other important ideological changes, in this case the founding of Gorze by Chrodegang, the Carolingians' loyal supporter, his introduction of Roman cults and liturgical practices, and his reform of the see of Metz. Further dimensions may have been given by the fact that, whereas the new monasteries of the early seventh century had been founded and endowed by an aristocracy in search of a new identity, Gorze rapidly attracted the support of much lesser landholders, perhaps keen to preserve their lands against an encroaching aristocracy. This interpretation might be supported by the fact that the benefices of the Gorze documents are much more like earlier precaria than the early eighth-century benefices of the Wissembourg cartulary. The donor gave land to the abbey and then received it back as benefice in a companion charter. At the same time, we see, as in 580-620, a burst of legislative activity in Pippin's edicts. Pippin also reformed the coinage and the economy. The supplanting of the Merovingians by the Carolingians was the culmination of social changes which, at least in the region of Metz, only stretched back a generation or so. It produced social tension and required considerable ideological readjustment. As is well known, the literary output legitimizing the Carolingian usurpation continued well into the ninth century.

Postscript: transformations in the late eighth century Social organization in the region of Metz did not long remain stable after the usurpation. A series of major transformations took place at the end of the eighth century. In Metz, an important phase of urban renewal appears to have begun at this time. In the north of the walled area, which seems to have stood derelict since the late fourth century, there are numerous traces of levelling associated with Carolingian pottery. The ceramics found at the unpublished Rue Boucherie-St-Georges site dates this activity to c. 800 (Henrotay & Lansival 1992:348). From about the same time, occupation begins to be attested at sites across the town and in the northern extra-mural zone (above, p. 238). The upsurge in urbanism in Metz is doubtless associated with widespread restructuring in the countryside. As shown in chapter 5, in the 770s and 780s, the charters of both Gorze and Wissembourg Abbeys manifest new ways of describing settlements. In Wissembourg, the coupling of villa with other more regional terms, as in in villa vel in marca, suddenly becomes the most common formula used for general land-grants. In the Gorze documents the change is even more interesting, to the formula cin X villa (or curtis) or 276

THE LATER MEROVINGIAN PERIOD

in the fines of Y [the same stem as X but with the regional suffixes -ensis or -acum]\ Both cartularies in their way seem to show an increased concern to distinguish the villa from the surrounding marca or fines. At the same time, mansus appears more frequently as a small, measured unit, using formulae common to land-grants in towns. The translation of mansus as tenement seems to be more usually justifiable. This suggests a reorganization of settlement with, perhaps, nucleation into hamlets. That this might well be the case is supported by archaeological evidence. At Nomeny, the seventh- and early eighth-century site is replaced by a new, apparently larger settlement 100m or so to the east. To this period, too, belongs Frouard phase III, yielding the same kinds of pottery as Metz cRue Boucherie St-Georges' (Henrotay & Lansival 1992:348, 351). Frouard phase III is interesting in that the wooden sunken-featured and post-built structures of phases I and II are replaced by a square stone structure. The imposition of stone buildings into the countryside is indicative of a desire to make a more permanent mark upon the landscape, as well as showing the control and use of considerable resources for vernacular buildings. For the first time since the fourth century, the dwelling place was increasingly the focus of social display. The greater organization of rural surplus may be further indicated by the sudden upsurge in the use of the terms jornalis and carradas, in the 780s in the Wissembourg charters, and the attendant increased interest in the measurement of land. All this might suggest regular management of arable and meadow, and the produce it generated. As suggested in chapter 2, this might have been associated with more systematic organization of labour services. It is not difficult to see how this change in the management of land and the control of resources could have resulted in increased surplus in the hands of the regional aristocracy and nobility, enabling them to construct more elaborate and imposing buildings in the countryside. It also allowed an even greater upsurge in urbanism in Metz, probably stimulating reorganization and rebuilding projects and new markets for longer distance exchange. If the premises upon which this study is based are sound, this reorganization of the settlement pattern should be reflected in changes in social organization. As has been noted (above, p. 59), this period saw changes in the terminology of the dependent classes. By c. 800, the multiplicity of technical terms for free, semi-free and unfree dependants was replaced, in the charters, by references to mancipia alone. By the end of the eighth century, legally defined identities based on the horizontal barriers within society had become less important than those founded upon the relations of dependence which bridged these barriers, only to create new 'diagonal' ones. In a sense, this is an extreme development of the changes of the century's first decades. At this time, as noted above, unfree women are less often recorded by name, instead simply being referred to as the wife of their husband. This suggests some change in attitudes towards gender, though quite what is difficult to say. It certainly adds significance to the earlier recording of women's names. These changes in society, especially those in the organization and management of estates, find their ultimate expression in the great Carolingian 277

CONCLUSIONS

monastic polyptychs (Morimoto 1988) which begin to appear at about this time. In north-east Gaul, this process eventually led to the creation of the Prum polyptych, so thoroughly, and often brilliantly, studied (Perrin 1935; Kuchenbuch 1978). In c. 810, Wissembourg drew up, under royal supervision, a brevium exempla of its precaria and benefices (Boretius (ed.) 1883: 252-4), and by 818 it had its own polyptych (Dette (ed.) 1987). Although, as Morimoto and others have shown, these documents were composite and rarely static documents, their appearance marks a significant stage in the organization of estates (Morimoto 1988:122-3), probably the final emergence of the 'Grand Domaine' from origins which, as outlined earlier, perhaps go back to the mid-seventh century. The organization of the countryside of the region of Metz into consolidated bipartite estates, with nucleation of settlement into hamlets, instead of scattered and fragmented holdings of mansi, meant a greater concentration of local people in dependence on the same lord. In these circumstances the old divisions into various legally defined free, semi-free and unfree ranks broke down. The era of the polyptychs, arising out of the changes of the eighth century's final quarter, mark a new age of social organization. Discussion of these transformations and their importance must, however, show how dangerous it is to argue back from the social situations outlined in these remarkable documents. Detailed explanation of the late eighth-century changes lies outside the scope of this book. It might not be unreasonable, however, to see them at least partly as associated with the expansion of the early years of Charlemagne's reign. In this 'decisive and fertile period' (Folz 1974:37), new, aggressive and successful campaigns were waged in Italy, Saxony and Bavaria, bringing varied rewards to the Austrasian aristocracy and church. Charlemagne's extension and manipulation of the systems of vassalage and benefice provided a spur to changes in landholding and management, and the king instituted further economic and monetary reforms.

Conclusions The picture just outlined of eighth-century society in the region of Metz is by no means revolutionary, either in its details or in the explanations proposed. However, this study has aimed to show that the social organization which existed around 750 was only a generation or so old, the product of social change in the early eighth century, and not of 300 years of gradual evolution. Furthermore, it did not long survive. By 800, it had itself been submerged by far more widespread and far reaching transformations. The primary interest of this survey has been in sixth- and early seventh-century society, and in the transformations around AD 600. As stated in chapter 1, however, I have hoped to show that society had not remained fixed in a particular 'mode' before the sixth century, and did not remain so after c. 650. A series of, admittedly rather blurred, snapshots may be taken of society in the region of Metz, as has been done in the last two chapters. In each we see a social structure which differed in important respects from its predecessors, sometimes fundamentally. The nature of the aristocracy and the 278

THE LATER MEROVINGIAN PERIOD

roots of its power was significantly different in 350, 450, 550, 650 and 750, and at various points in between. Similar changes can be noted at equally regular intervals in the conditions of the free population or slaves, or in the construction of gender. Jean Dunbabin has recently argued that the case of the '"gradualists" (those who spread the evolution of medieval society over seven centuries)' (Dunbabin 1993:698) cannot be undermined by Bois' (1992) study. Both sides in this debate seem equally misled. Bois, as mentioned (above, p. 2), saw society as remaining in some more or less static 'slave mode of production' from the ancient world until the end of the tenth century, when, in a dramatic period of change, it was transformed into the 'feudal' mode. This is demonstrably false. Bois' argument that a 'slave mode' persisted to the tenth century, based ultimately upon Bonnassie's (1991b) work, has been roundly attacked (Verhulst 1991). The idea of a static social formation before the tenth century finds no support in the analyses performed here and, given the perpetual renegotiation of identities and social barriers in 'customary' society, the very concept is inherently unlikely. Nevertheless, in detail, Bois' account of fundamental structural changes taking place over a short period in late tenth-century Burgundy is persuasive, plausible and not especially radical. We need to consider how unusual such a period of change was in the early middle ages. Is the 'gradualists" position any more tenable than Bois'? I think not. As I have argued throughout this book, the picture of slowly evolving postRoman society is based largely upon the aggregation of evidence. When the fragmentary written data of post-Roman Gaul are separated out by date and geographical origin, numerous interesting and important differences emerge. This becomes clear when, instead of grouping all Frankish law together as a single 'corpus', Lex Ribvaria is separated from the Pactus Legis Salicae, and the various royal edicts kept separate from both. The same is true of charters. There were changes through time in the ways in which their writers talked about land, dependants and so on. Keeping evidence rigidly within geographical and chronological context reduces our samples often to statistical insignificance, but at the same time opens up before us exciting new vistas of diversity and dynamism in the early medieval world. But if there are differences in our documents, be they in the picture they paint, in the forms they take or in their very existence, can we deduce from this real social difference and change? At one level we might insist that the choice of words to express thoughts, the choice of the documentary form within which to put that expression, and the choice of whether or not to keep that document, are all conscious, deliberate and significant. Even where new phrases and formulae, or new types of deed, can be shown to have been copied or introduced from elswehere, the acceptance of the new form, or the decision to copy, still resulted from particular circumstances and choices. Nevertheless, the case can only really be proved by consideration of a broader range of evidence. Decisions to abandon cemeteries used for 100 years or more, to change radically the methods of disposal of the dead, or to introduce completely new methods, are not made lightly. Nor are decisions to abandon settlements or to introduce wholly new organizations 279

CONCLUSIONS

of dwelling units and living spaces, and nor are changes in the material constructions of gender, age and other identities. Neither these material changes nor those in the methods of written expression may be dismissed as 'fashion'. Fashion is a description, not an explanation. When we find transformations in a number of these areas occurring at about the same time, we can plausibly argue that real social change was taking place. The comparison of different forms of evidence thus does not take the form of simplistic one-to-one correlations but stems from adducing common factors which might explain the developments in different areas of evidence. These factors often revolve around mentalities; it has been argued that the increased survival of written documents, and the changes in funerary practices in the seventh century, all hinge on developments in the concepts of time. In other areas we can plausibly see certain aspects of both written and material culture as resulting from particular human relationships. The question remains, however. How significant are the changes revealed? Do they qualify simply as phases in long-term processes of gradual evolution? Certain periods of change studied here, like that analysed by Bois, are clearly fundamental. The transformations of the end of the fourth century, of c. 600 and of the late eighth century all qualify as revolutionary, affecting all areas of life and effecting major structural changes in society. That nevertheless leaves 150-200 years between these revolutionary periods but, as demonstrated, society did not remain static in the intervening years. If we had better evidence, the social changes around 500 would doubtless appear still more important. The transformations between c. 650 and c. 750 can be outlined in more detail. As argued earlier, some of these developments involved important transformations in social structure and in people's perceptions. These changes were difficult to separate into entirely discrete phases; change was dynamic in the later Merovingian period. It thus seems difficult to argue that change during the 300 years of the Merovingian period represents 'gradual evolution'. These changes were neither more nor less 'gradual' than those of any other period before or since. An old man alive around AD 700 would look back upon his childhood as upon another world. The problem with envisaging 'the gradual evolution of medieval society over seven centuries' is that it begins with a particular definition of medieval society. Here, we return to the problems with which this book opened. This approach defines what is and what is not medieval society; it then finds points at which these criteria are satisfied; and declares the intervening period to be 'transitional'. To take an example, we might decide that a 'medieval' society is one with some system of royal vassalage, a rural landscape organized into Great Estates, with demesne and dependencies, an undifferentiated dependent peasantry, and so on, a society, in short, much as existed in c. 800. On the other hand, we might see society around 300 as 'ancient', with its continuing slavery, parasitic 'Roman' urbanism, taxation and symbiotic relationship between state and elites. It does not help to define the intervening period as one of transition. Similar problems arise with the Marxist problematic of progression from one 'mode of production' to another, in our case from a 'slave' or 'ancient mode' to a 'feudal mode of production'. I have largely avoided the term 280

THE LATER MEROVINGIAN PERIOD

feudal in this work. It is now difficult to discuss feudalism without defining one's terms. Traditional definitions have proved increasingly unworkable (Stray er's 1968 narrow governmental formulation seems to be the best) and recent Marxist attempts to reduce feudalism to power based upon landholding (Wickham 1984; Hilton 1992:9-10) are not very useful. The problem with the approach is that once terms are defined, history reduces to a search for the point where these predetermined criteria are satisfied, and then backwards to look for the origins of the component elements. Late eighth-century society may again stand as an example. We can trace the origins of the 'Grand DomaineY components, of the rural framework of the late eighth century, 130 years or so earlier; we can see changes to the bonds between aristocrats towards basing these upon landholding relationships, eighty years earlier. Even to say that this is a shorter time-scale than is often envisaged still misses the point. Late eighth-century social structure was not inevitable once certain components were in place. In the late seventh century, more than one outcome of the day's social struggles was possible. The early eighth-century changes were closely related to political events. We need to consider further the nature of social change. It has been discussed here in terms of the interplay of identities. Changes in the nature of certain identities affected the others so that changes in the strength and permanence of family identity brought about changes in gender, age and class identities. Within society, there are groups with conflicting interests, making use of the various identities which it is possible to adopt, and consequently social organization was in perpetual renegotiation. Hence change can be seen as constant. Nonetheless, the outcomes of particular struggles can be viewed as heavily influenced by political events. The latter are very important, not least in the way in which they shape the precise nature of the identities adopted. Historical development is often seen in terms of an opposition between two schools of thought. On the one hand, we see longterm, gradual change evolving over the clongue duree', often with inevitable results expressed (with varying degrees of sophistication) either as a Marxist progression of modes of production, or in similar terms of 'social evolution', or determined by environmental (climatic, geographical) or economic factors. On the other hand, there is the short-term, 'Great Men and Battles' approach, with history being made by a few 'great' individuals or by the outcome of particular events. Both approaches, in their own way, are 'safe', removing the individual from any role in affecting societal change. In the first, social change is inevitable, removed from the hands of individual mortals, especially in environmentally determinist histories. In the other, change results from the actions of certain chosen individuals, and 'fate' plays a large part in determining when and where they arise, and the results of their actions: 'history's lucky breaks'. Again most people have no role in the shaping of history. Both approaches, admittedly reduced to extremes, must be rejected. The structures of society define the framework within which people can act, and thus the specific nature of political events. Nevertheless, individuals may all act to change the frameworks within their communities, helping to bring about the circumstances for political events to effect major redefinitions of social 281

CONCLUSIONS

structure. It will no longer do, therefore, to reject chistoire evenementielle' as 'surface disturbances, crests of foam which the tides of History carry on their strong backs' (Braudel 1972:21). Bloch, Braudel and the other great Annalistes were undoubtedly the finest historians of this century, and struck a vital blow against the political-institutional history of their day. They did so not least by forcing us to consider a more sophisticated and broader range of human activity and interaction, and the interplay between man and the environment. Nonetheless, it is time to reinstate political, and even military, events in a significant place in the analysis of social change. The other side of the coin from dynamic change through time is geographical diversity. This study has attempted to describe and explain social development in one region, and to propose a chronology of the principal social changes which took place. It is nowhere argued that this picture is of general application. As has been repeatedly stressed, even the neighbouring Triererland differs significantly from northern Lorraine. Both the physical geography of the region of Metz and, in particular, its place in the political structures of the late Roman and Merovingian worlds contribute to the area's individuality. This is not to say that the major transformations recorded here did not take place in other regions, but that they may have taken place at different times, and for differing reasons, as similar responses to different situations. In, for example, Bonnassie's discussion of early medieval slavery, the chronological variations are noted between one area and another in the end of particular legal forms of unfreedom (Bonnassie 1991b:55-9); historians of the Carolingian and later periods have long accorded particular importance to regional surveys. They are similarly readier than historians of the immediately post-Roman centuries to accept the existence of dynamic change in customary societies. In the fundamentally rural societies of Merovingian northern Gaul, the construction of social reality differed over very short distances; one need only compare the nearby cemeteries of Lavoye and Dieue-sur-Meuse to see that. Brief study of cemeteries in the Rhineland, the Triererland and further into the Paris Basin brings home the point more forcefully. Merovingian society was essentially local in its outlook, and is therefore best studied on a regional basis. Thus, in the same way as I make no claims that the results of this study are applicable outside northern Lorraine, they cannot be simply 'disproved' by studies of other parts of the Merovingian world, let alone of later periods. Critique of this study must be carried on within its own terms, that is either by taking issue with the theoretical premises upon which it is based, or by alternative interpretations of this region's data. More archaeological data from settlement and cemetery sites have been accumulated and are in the process of publication. It will be interesting to see how these modify and correct the models proposed here. It will be interesting, too, to see how this region compares with others within the Merovingian realms. How decisive is the fact of its central location within Austrasia, for example? We still need more contextualized regional surveys of Merovingian Gaul, for a more sophisticated picture of postRoman Gallic society. Only then can we restore to the Merovingian world some of its undoubted dynamism and diversity. 282

Bibliography

Primary sources Location of sources (see also abbreviations) Ammianus Marcellinus, Res Gestae: Rolfe (ed. & trans.) 1935-52. Annales Mettenses Priores: von Simson (ed.) 1905. Bede, Ecclesiastical History: Colgrave & Mynors (ed. & trans.) 1969. Caesarius of Aries, Sermons: Mueller (trans.) 1956. Chlothar II, Edict of Paris: de Clercq (ed.) 1963:283-5; Hillgarth (trans.) 1986:113-16. Decretio Childeberti (Childebert's Decrees): Eckhardt (ed.) 1962:267-9; Eckhardt 1967; Rivers (trans.) 1987:144-8; Drew (trans.) 1991:156-9. Fredegar, Chronicle: Krusch (ed.) 1888:1-193; Wallace-Hadrill (ed. & trans.) 1960. History of St-Arnulf Metz (Ex Historia S. Arnulfi Mettensis; 'Petit Cartulaire de St-Arnoul'): Waitz (ed.) 1879:529-49. Hydatius, Chronicle: Burgess (ed. & trans.) 1993:1-172. Isidore of Seville, Etymologies: Lindsay (ed.) 1911. Jonas, Vita Columbani: Krusch (ed.) 1902:1-152. Law of Seven Categories: Eckhardt (ed.) 1962:269-73; Rivers (trans.) 1986:148-52; Drew (trans.) 1991:159-63. Leges Alamannorum: Lehmann (ed.) 1888:55-159. Lex Baiouariorum: de Schwind (ed.) 1926:261-473. Liber Constitutionum (Burgundian Law): Drew (trans.) 1949. Liber de virtutibus sanctae Geretrudis: Krusch (ed.) 1888:464-71; McNamara et al. (trans.) 1992:229-34. Lombard Law (Rothari's Edict and Liutprand's Laws): Drew (trans.) 1973. Marculf's Formulary: Uddholm (ed. & trans.) 1962. Notitia Dignitatum: Seeck (ed.) 1876. Procopius, Wars: Dewing (ed. & trans.) 1914-28. Revised Annals: Kurze (ed.) 1895; King (trans.) 1987:108-31. Royal Frankish Annals: Kurze (ed.) 1895; King (trans.) 1987:74-107. Salvian, Governance of God: Halm (ed.) 1877; Sullivan (trans.) 1962 (numeration after Halm). Sidonius Appollinaris, Epistulae: Anderson (ed. & trans.) 1936-65. Stational List of Churches: Klauser & Bour 1929. Venantius Fortunatus, Carmina: Leo (ed.) 1881; Nisard & Rittier (ed. & trans.) 1887. Vita Radegundis: Krusch (ed.) 1888:358-77; McNamara et al (trans.) 1992:70-86. Vita Eligii: Krusch (ed.) 1902:634-761. Vita Goaris Confessoris Rhenani: Krusch (ed.) 1902:411-23. Vita, translations, miracula S. Glodessindis: Acta Sanctorum Iul VI, pp.203-10; McNamara et al. (trans.) 1992:137-54. Vita Waldradae (First Version): Mabillon (ed.) 1733:58-9. Vita Waldradae (Second Version): Acta Sanctorum, Mai II, pp.51-2. Wettinus, vita Galli: Krusch (ed.) 1902:256-80. 283

BIBLIOGRAPHY Editions and translations Anderson, W.B. (ed. & trans.), 1936-65. Sidonius. Poems and Letters, 2 vols. (London). Bachrach, B.S. (trans.), 1973. The Liber Historiae Francorum (Laurance, Kans.). Beyerle, F., & Buchner, R. (eds.), 1951. MGH Legum, Sect. 1, vol. 3, Lex Ribvaria (Hannover). Boretius, A. (ed.), 1883. MGH Legum, Sect. II, vol.1, Capitularia Regnum Francorum (Hannover). Burgess, R. (ed. & trans.), 1993. The Chronicle of Hydatius and the Consularia Constantinopolitana (Oxford). Busson, G., & Ledru, A. (eds.), 1901. Archives Historiques du Mans II. Actus Pontificum Cenomannis in Urbe Degentium (Le Mans). de Clercq, C. (ed.), 1963. Corpus Christianorum 148A. Concilia Galliae (Turnhout). Colgrave, B., & Mynors, R.A.B. (ed. & trans.), 1969. Bede's Ecclesiastical History of the English People (Oxford). Dette, C. (ed.), 1987. Liber Possessionum Wizenburgensis. Neu herausgegeben und kommentiert (Mainz). Dewing, H.B. (ed. & trans.), 1914-28. Procopius, vols. I-V (London). Douglas, D.C., & Greenaway, G.W. (trans.), 1981. English Historical Documents, II: 10421189, 2nd edition (London). Drew, K. F. (trans.), 1949. The Burgundian Code (Philadelphia)(reprinted 1972). (trans.), 1973. The Lombard Laws (Philadelphia), (trans.), 1991. The Laws of the Salian Franks (Philadelphia). Eckhardt, K.-A. (ed.), 1962. MGH Legum Sect. I, vol.4, Part 1, Pactus Legis Salicae (Hannover). Glockner, K., & Doll, A. (ed.), 1979. Traditiones Wizenburgenses: Die Urkunden des Klosters Weissenburgy 661-864 (Darmstadt). Gundlach, W. (ed.), 1957. Corpus Christianorum 117 (Turnhout). Halm, C. (ed.), 1877. MGH Auctores Antiquissimi 1 (Berlin). d'Herbomez, A. (ed.), 1898. Cartulaire de VAbbaye de Gorze. (Mettensia 2) (Paris; introduction dated 1902) Hillgarth, J.N. (trans.), 1986. Christianity and Paganism, 350-750. The Conversion of Western Europe (Philadelphia). James, E. (trans.), 1991. Gregory of Tours. The Life of the Fathers, 2nd edition (Liverpool) (1st edition, Liverpool, 1985). King, P.D. (trans.), 1987. Charlemagne. Translated Sources (Lancaster). Klauser, T., & Bour, R.-S., 1929. 'Note sur la liturgie de Metz et sur ses eglises anterieures a Tan mil' ASHAL 38, pp.497-639. Krusch, B. (ed.), 1888. MGH Scriptores Rerum Merovingicarum 2 (Hannover), (ed.), 1896. MGH Scriptores Rerum Merovingicarum 3 (Hannover), (ed.), 1902. MGH Scriptores Rerum Merovingicarum 4 (Hannover). Krusch, B., & Levison, W. (eds.), 1951. MGH Scriptores Rerum Merovingicarum 1, Part 1 (Hannover). 1969. MGH Scriptores Rerum Merovingicarum 1, Part 2 (Hannover). Kurze, F. (ed.), 1895. MGH Scriptores Rerum Germanicarum 6 (Hannover). Lehmann, K. (ed.), 1888. MGH Legum Sect. I, vol. 5, Part 1, Leges Alemannorum (Hannover). Leo, L. (ed.), 1881. MGH Auctores Antiquissimi 4, Part 1 (Berlin). Lesort, A. (ed.), 1909-12. Chronique et chartes de VAbbaye de St Mihiel (Mettensia 6) (Paris). Lindsay, W.M. (ed.), 1911. Isidori Hispalensis Episcopis Etymologiarum sive Originum libri XX, 2 vols. (Oxford). Mabillon, J. (ed.), 1733. Acta Sanctorum Ordinis Sancti Benedicti in Saeculorum Classes Distributa. Saeculum Secundum quod est ab anno Christi DC ad DCC, ed. L. D'Achery (Venice). McNamara, J.-A., Halborg, J.E., & Whatley, E.G. (trans.), 1992. Sainted Women of the Dark Ages (Durham, N.C., and London). 284

BIBLIOGRAPHY

Mueller, Sister M.M. (trans.), 1956. Caesarius ofAries. Sermons I Sermons 1-80 (New York). Miihlbacher, E. (ed.), 1906. MGH Diplomata Karolinorum 1 Pippini, Carolomanni, Caroli Magni Diplomata (Hannover). Nisard, C , & Rittier, E. (eds. & trans.), 1887. Venance Fortunat: Poesies Melees Traduites en Franqais (Paris). Pardessus, J.-M. (ed.), 1843-9. Diplomata, Chartae, Epistolae, Leges aliaque Instrumenta ad res Gallo-Francicas Spectantia, 2 vols. (Paris). Pelt, J.-B. (ed.), 1937. Etudes sur la Cathedrale de Metz: La Liturgie L V-VIII Siecle (Metz). Pertz, K.A. (ed.), 1872. MGH Diplomata Imperii 1 (Hannover). Rivers, TJ. (trans.), 1986. The Laws of the Salian and Ripuarian Franks (New York). Rolfe, J.C. (ed. & trans.), 1935-52. Ammianus Marcellinus, 3 vols. (London). Rouche, M. (trans.), 1986. 'Vie de St Gery, ecrite par un clerc de la basilique de Cambrai entre 650 et 700', Revue du Nord 68, no.269, pp.281-8. de Schwind, E. (ed.), 1926. MGH Legum, Sect. I, vol. 5, Part 2, Lex Baiwariorum (Hannover). Seeck, O. (ed.), 1876. Notitia Dignitatum accedunt Notitia Urbis Constantinopolitanae et Latercula Provinciarum (Frankfurt am Main) (repr. Frankfurt, 1962). von Simson, B. (ed.), 1905. MGH Scriptores Rerum Germanicarum 10 (Hannover and Leipzig). Sullivan, J.F. (trans.), 1962. The Writings of Salvian the Presbyter (Washington). Thorpe, L. (trans.), 1974. Gregory of Tours. The History of the Franks (Harmondsworth). Uddholm, A. (ed. & trans.), 1962. Marculfi Formularium Libri Duo (Uppsala). Van Dam, R. (trans.), 1988a. Gregory of Tours. The Glory of the Martyrs (Liverpool). 1988b. Gregory of Tours. The Glory of the Confessors (Liverpool). Waitz, G., 1879. MGH Scriptores 24 (Hannover). Waitz, G. (ed.), 1878. MGH Scriptores 2 (Hannover). Wallace-Hadrill, J.M. (ed. & trans.), 1960. The Fourth Book of the Chronicle of Fredegar with its Continuations (London). Whitelock, D. (trans.), 1979. English Historical Documents, I: c.550-1042, 2nd edition (London).

Archaeological sources for the region of Metz This bibliography, by no means exhaustive, includes those works which are primarily publications of data (site reports, regional inventories, catalogues etc.). Works which are essentially discussions and interpretations of archaeological evidence are included under 'Secondary References'. Most of the articles in BLAFAM 12 (1988) were subsequently published, and occasionally developed, in Actes . . . . To save space, I have usually cited only the Actes . . . publications. See also p. xx. above. Altmayer, 1828-9. 'Sur les mines du Hieraple, la voie romaine qui y aboutit et les traditions fabuleuses du pays', MAM, pp.395-65. Ancelon, H., 1874. 'Ou ont succombe les deux legions romaines de Julien, surprises par les Germains? - Pres de Tarquimpol', JSAL, pp. 183-7. Barbier, Abbe, 1867. 'Anciennes decouvertes a Mittelbronn et a Romelfing', JSAL, pp. 10914. Barthelemy, L'Abbe, 1842-3. 'Note des antiquites recueillees sur le territoire de la commune de Bettelainville et offertes a l'acadamie', MAM, pp.393-4. Beaupre, J., 1897. Repertoire Archeologique pour le departement de Meurthe-et-Moselle (Nancy). Bellard, A., 1933. 'En construisant une voie ferree entre Mondelange et Richemont, on decouvre de sepultures franques', Cahiers Lorrains, pp.9-10. 1957. 'Introduction a une histoire de Noveant', ASHAL 57, pp. 153-8. 1964. 'Sur les "Cardo Maximus" et "Decumanus Maximus" de la Metz gallo-romaine', ASHAL 64, pp.21-4. Benoit, L., 1862. 'Repertoire archeologique de l'arrondissement de Sarrebourg', MSAL 12, pp.220ff (article has special pagination). 285

BIBLIOGRAPHY Biehler, E., & Francois, M.-F., 1977. 'Une officine de ceramique romain decouverte a Metz dans l'enceinte de la caserne De-Lattre-de-Tassigny', Cahiers Lorrains, pp.99103. Biehler, E., & Schlemaire, G., 1974. 'Vestiges gallo-romains decouverts au 4-6 Rue du Chanoine Collin a Metz', ASHAL 74, pp.9-18. Boulange, G., 1853-4. 'Notes pour servir a la statistique monumentale de la Moselle', MAM, pp.295-327. de Bouteiller, E., 1866-7. 'Notice sur des sepultures franques', MAM, pp.257-65. Brunella, P., Dautremont, N., Thion, P., Wagner, P.E., Coudrot, J.-L., Heber-Suffrin, F., & Lefebvre, C., 1992a. Metz. Document du Patrimoine Archeologique Urbain (Tours). 1992b. 'Metz du IVe au milieu du VHP siecle', in Brunella et al 1992a, pp.25-32. Burnand, Y., & Collot, G., 1980. 'Metz', in Archeologie Urbaine. Actes du Colloque Internationale (Tours), pp.559-63. CAHCG (Cercle Archeologique et Historique de Chatel-St-Germain), 1978. Dix Ans de Travaux au Mont-St.-Germain (?Metz). 1980. Fouilles Archeologiques au Mont-St-Germain. Rapport 1979 (?Metz). 1982. Fouilles Archeologiques au Mont-St-Germain. Rapport 1981 (?Metz). 1989. Archeologie a Chdtel-St-Germain (Moulins-les-Metz). Chatelain, L'Abbe, 1881-2. 'Notice sur le chateau et les sires de Warsberg', MAM pp.51170. Clement, R., 1936. 'Trouvailles archeologiques a Metz et dans les environs', ASHAL 45, pp. 157-72. Clercx, J., Dufresne, A., & D'Huart, E., 1843-4. 'Daspich, annexe de Florange, canton et arrondissement de Thionville, dept. de la Moselle', MAM, pp.267-80. Clermont-Joly, M., 1978 (with Wagner, P.E.) UEpoque Merovingienne (Catalogues des Collections des Musees de Metz 1) (Metz). 1980. 'Metz a l'epoque merovingienne: etat des decouvertes archeologiques', in Patrimoine et Culture en Lorraine, ed. F.-Y. Le Moigne (Metz), pp.67-81. Collot, G., 1961a. 'Decouverte d'un mur romain chez Mayer Freres, 17 rue des Clercs', ASHAL 61, pp.9-11. 1961b. 'Decouverte de deux autels gallo-romains, Rue des Clercs', ASHAL 61, pp.12-14. 1963. 'Nouvelles decouvertes archeologiques rue des Clercs', ASHAL 63, pp. 15-19. 1964a. 'Fouilles archeologiques sur le chantier du parking souterrain de l'Esplanade et Rue Poncelet', ASHAL 64, pp.41-70, 77-8. 1964b. 'Fouilles et degagement des thermes du Carmel en 1963 et 1964', ASHAL 64, pp.71-6. 1967-8. 'Fouilles archeologiques sur le chantier de l'Ecole des Arts Appliques', ASHAL 67-8, pp.5-32. 1980. La Sculpture du Haut Moyen Age (Catalogues des Collections des Musees de Metz 2) (Metz). Collot, G., Clermont-Joly, M., & Wagner, P.E., 1979. De la Fin du Monde Antique a VAn Mil. Musee Archeologique de Metz Guide 4 (Metz). Cuvelier, P., 1988a. 'Necropoles et habitats du haut moyen-age en val de Seille (Meurthe-et-Moselle)', BLAFAM 12, pp.65-73. 1988b. 'Eglise et sepultures du VIP siecle a Mousson', BLAFAM 12, pp.74-6. 1988c. 'Le paysage rural en val de Seille: structures et evolution de l'antiquite au moyenage', in Delestre (ed.) 1988, pp.53-8. 1988d. 'Les premiers edifices chretiens de Mousson', Cahiers Lorrains (part 2, June), pp.205-16. 1990. 'Mousson, de la prehistoire aux temps modernes. Premiere synthese archeologique', in Lotharingia, II (Nancy), pp. 149-92. Cuvelier, P., & Delestre, X., 1986. 'L'agglomeration rurale de Nomeny et son environnement', Cahiers Lorrains (part 4, Oct.), pp.325-36. Cuvelier, P., Delestre, X., & Heber-Suffrin, F., 1988. 'Le paysage monumentale', in Delestre (ed.) 1988, pp.41-52. 286

BIBLIOGRAPHY unpub. 'Peuplement et habitat en Lorraine centrale (de l'antiquite au moyen-age)', resume for 39e Symposium Saxon, Caen, 12-16 Sept., 1988. Dautremont, N., unpub. Tontiffroy 1987: Salle de Conseil-St.-Clement', unpublished 'Rapport de fouille', DAHL, Metz. Delestre, X., 1988a. St.-Pierre-aux-Nonnains (Metz, Moselle). De VEpoque Romaine a VEpoque Gothique (Guides archeologiques de la France 15) (Paris). 1988b. 'St.Pierre-aux-Nonnains: une nouvelle datation par l'archeomagnetisme', Cahiers Lorrains (part 2, June), pp. 195-8. Delort, E., 1935. 'Le cimetiere merovingienne de Koenigsmacker, pres de Thionville', Revue Archeologique, pp.271-2. 1947. 'Le cimetiere franc d'Ennery', Gallia 5, pp.351-403. Delort, E., & Lutz, M., 1948. 'Les decouvertes de Berthelming', ASHAL 48, pp.95-108. Dufresne, A., 1854-5. 'Notice sur des sepultures gallo-franques trouvees en 1854 a Farebersviller (Moselle, canton de St-Avold)', MAM, pp.539-43. Fisenne, F. von, 1896. 'Das Mithraeum zu Saarburg in Lothringen', JGLGA 8, pp.119-75. Forrer, R., 1934. 'Varia merovingiennes et cimetieres merovingiens inedits de Bettwiller, Behlenheim, Gambsheim, Schiltigheim, Friesenheim etc.', CAHA 99-100, pp.221-58. Gauthier, N., 1975. Receuil des Inscriptions Chretiennes de la Gaule anterieures a la Renaissance Carolingienne. 1. Premiere Belgique (Paris). Guillaume, L'Abbe P.E., 1864. 'Decouverte d'un cimetiere merovingien a Maizieres-lesVic', JSAL, pp.60-5. Guillaume, J., 1974-5. 'Les necropoles merovingiennes de Dieue/Meuse (France)', Acta Praehistorica et Archaeologica 5-6, pp.211-349. Guillaume, J., Pargny, D., Quiry, G., Quiry, G., Sibert, P. & Wieczorek, A., 1982. 'Un couvercle de sarcophage decore a Chatel-St-Germain (Moselle)', BLAFAM 6, pp-7982. Guillaume, J., Bertaux, J.-P., Clermont-Joly, M., Cuvelier, P., Delestre, X., Gambs, A., Heckbenner, D., Lefebvre, C., Legoux, R., Lieger, A., Lemant, J.-P., Marguet, R., Rouillon, M., Simmer, A., & Wieczorek, A., 1988. 'Les necropoles' in Delestre (ed.) 1988, pp.61-109. Guillaume, J., & Gambs, A., 1989. 'La necropole de 'Colmette' a Montenach (Moselle)', Actes..., pp.137^1. Guillaume, J., Lefebvre, C., & Wieczorek, A., 1989. 'La necropole de Chatel-St-Germain (Moselle)', Actes..., pp. 107-10. Hackler, C., 1989. 'La necropole de Altheim (R.F.A.)', Actes..., pp.179-91. Hatt, J.-J., 1958. 'Fouilles stratigraphiques a Metz (Juin-Novembre 1957). Rapport provisoire', ASHAL 58, pp.35-44. 1959. 'Fouilles de Metz. Campagne 1958', ASHAL 59, pp.5-16 (plates between pp.60-1). 1961. 'Les fouilles de la basilique Saint-Pierre-aux-Nonnains de Metz en 1961', ASHAL 61, pp.15-26. Hatt, J.-J., & Lutz, M., 1960. 'Fouilles a Sarrebourg', ASHAL 60, pp.41-64. Heckenbenner, D., & Thion, P., 1986. 'Fouille archeologique de 1'Arsenal Ney a Metz', Cahiers Lorrains (part 4, Oct.), pp.337-52. Henrotay, D., & Lansival, R., 1992. 'Un habitat rural du haut moyen-age a Frouard (Meurthe-et-Moselle)', RAE 43, pp.329-52. Heuertz, M., 1957. 'Etudes des squelettes du cimetiere franc d'Ennery', Bulletin et Memoire de la Societe dyAnthropologie de Paris 8, pp.81-147. Hoerner, B., 1986. 'Grostenquin: vicus gallo-romaine?', Cahiers Lorrains (part 4, Oct.), pp.317-24. Hoffmann, O.-A., 1893. 'Die Kleinaltertumer des romisch-mittelalterlichen Museums der Stadt Metz',/GLG.4 5 (part 2), pp.172-87. Huber, E., 1894. 'Excursion archeologique au Herapel', JGLGA 6, pp.296-304. 1889-90. 'Notice sur les collections et les collectioneurs de Lorraine', MAM, pp. 109-17. 1907. Le Herapel Les Fouilles de 1881 a 1904. Description des Planches (Strasbourg). Hurstel, J., 1984. 'Sepultures et techniques artisanales dans la cite des Mediomatriques du Ve au VHP siecle', These de doctorat de Hie cycle, Universite de Nancy II (2 vols.). Joffroy, R., 1974. Le Cimetiere de Lavoye (Meuse) (Paris). 287

BIBLIOGRAPHY Jolin, R., 1975. 'Implantation des vestiges romains aux environs de la cathedrale', ASHAL 75, pp.38-41. 1977. 'Vestiges romains entre la Rue des Clercs et Nexirue a Metz', ASHAL 77, pp.1726. 1982. 'Les thermes St-Jacques a Metz', Cabiers Lorrains (part 3), pp.266-76. Joly, A., 1870. 'Repertoire archeologique des cantons nord et sud de Luneville', MSAL, pp.69-104. Keune, J.-B., 1901. 'Spat-merovingischer Friedhof bei Gross Moyeuvre (Kr. Diedenhofen)', JGLGA 13, pp.355-60, 480. 1902. 'Gross-Moyeuvre', KW2GK 21 (nos.1-2, Jan.-Feb.), cols.1-3. 1903. Triedhof der friihen Volkerwanderungszeit aus dem Bann von Metrich', JGLGA 15, pp.480-1. 1905. 'Metz [Altertumsfunde zu Metz und Sablon]', KWZGK 24 (nos.3-4, March-April), cols.40-1. 1907-9. 'Die Flur Sablon in romischer Zek\JVEM 26, pp.1-98. Keune, J.-B., Schramm, R., & Wolfram. G., 1892. 'Das grosse Amphitheater zu Metz', JGLGA 14, pp.340-430. Klein, 1932. 'Neue Funde der Merowingerzeit im Saarland', Ber.SDS 4, pp.85-6 (+Taf. 17 and 18). Knitterschied, E., 1901. 'Graberfunde bei Metz', JGLGA 13, pp.363-6. Kolling, A., 1965. 'Der Name des romischen Saarbriicken', Ber. SDS 12, pp.61-6. Kraus, F.-X., 1889. Kunst und Alterthum in Elsass Lothringen. Beschreibende Statistik (Strasbourg). Laumon, A., 1977. 'La necropole merovingienne de Bettborn', ASHAL 77, pp.51-71. Lefebvre, C., & Colardelle, M., 1984. 'L'Exemple de Metz', Revue des Monuments Historiques (Dec, special issue: 'Archeologie et Projet Urbain'), pp.49-54. Lefebvre, C., & Pargny, D., 1981. 'Le prieure et la necropole du Mont Saint-Germain a Chatel-Saint-Germain', Cahiers Lorrains (part 3), pp. 149-55. Legendre, J.-P., & Krausz, S., 1989. 'Deux fosses merovingiennes d'habitat avec traces d'artisanat a Florange (Moselle)', Actes..., pp. 125-31. Legoux, R., & Lieger, A., 1989. 'La necropole gallo-romaine et merovingienne de Cutry (Meurthe-et-Moselle)', Actes..., pp.111-23. Lepage, H., 1843. Le Departement de la Meurthe. Statistique, Historique et Administrative (Nancy), vol.11. Lieger, A., & Marguet, R., 1992. 'Le cimetiere merovingien de Royaumeix/Menil-la-Tour (Meurthe-et-Moselle)', RAE 43, pp.99-149. Lieger, A., Marguet, R., & Guillaume, J., 1984. 'Sepultures merovingiennes de l'abbaye de St-Evre a Toul (Meurthe-et-Moselle)', RAE 35, pp.301-17. Linckenheld, E., 1930-1. 'La necropole barbare d'Imling-Xouaxange et ses rites funeraires', Bulletin Archeologique, pp.529-40. 1931. Repertoire Archeologique de VArrondissement de Sarrehourg (Moselle) (Sarrebourg: 1929 on title page). 1932a. Archdologisches Repertorium der Kreise Forbach und Saargemiind (Forbach). 1932b. 'Une villa romaine avec tombes de Pepoque barbare a Gondrexange (Moselle)', Revue Historique de la Lorraine 78, no.4, pp. 130-40. 1932-3. 'Villa romaine avec tombes de l'epoque des invasions a Gondrexange (Moselle)', Bulletin Archeologique, pp.685-90. 1933. Archdologisches Repertorium des Kreises Bolchen (Forbach). 1934. Repertoire Archeologique des Arrondissements de Thionville-Est et Ouest (Metz). Lutz, M., 1950. 'La villa gallo-romaine et la necropole merovingienne de Berthelming', RAE 1, pp. 180-4. 1952. 'Decouvertes archeologiques dans la region de Sarrebourg', ASHAL 52, pp.59-93. 1956. 'Decouvertes merovingiennes recentes ou inedites en Moselle', ASHAL 56, pp.1730. 1960. 'Sarrebourg. Le Chantier Blum, Rue de la Paix', ASHAL 60, pp.65-86. 1972. 'Le domaine gallo-romaine de Saint-Ulrich (II)', Gallia 30, pp.41-82. 1978. 'Archeologie des rives de la Sarre', ASHAL 78, pp.11-39. 288

BIBLIOGRAPHY 1991. La Moselle Gallo-Romaine (Drulingen). Maisant, H., 1971. Der Kreise Saarlouis in vor- und friihgeschichtlicher Zeit (Saarlouis). Masson, J.-P., 1975. 'Etude anthropologique de la necropole gallo-romaine de Scarpona', These pour obtenir le grade de Docteur en Medecine, Universite de Nancy I. Merciol, L'Abbe, 1886. 'Decouverte faite pres de Morville-les-Vic', JSAL, pp.38-9. Miron, A., & Reinhard, W., 1990. 'Zur Problematik archaologischer Quellenerschliessung und Interpretation der Quellenlage. Ein Beitrage aus saarlandischer Sicht', Kolner Jahrbuch fur Vor- und Fruhgeschichte 23, pp.689-701. Miron, A., Reinhard, W., Petit, J.-B., & Schaub, J., 1986. Archeologie de la Vallee de la Blies. Exposition Franco-Allemande, Musee de Sarreguemines (Sarreguemines). Moineaux, P., 1966 'Des nouvelles de la necropole "merovingienne" de Blies-Ebersing', Cahiers Sarregueminois 4, p. 174. Morhain, E., 1953. 'Decouvertes archeologiques dans l'eglise de Cheminot', ASHAL 53, pp.87-101. Morlet, Le Col. de, 1863-6. 'Notice sur quelques decouvertes effectues dans les cantons de Saar-Union et de Drulingen (arr. de Saverne)', BSCMHA, 2nd ser., 2, pp. 1-6. Nicolay, P.-X., 1937. Histoire d'Hayange. Tome 1. La Seigneurie d'Hayange (Hayange). Oswald, F., 1967. 'Romische Basilike und ottonische Kirche St Peter auf der Zitadelle in Metz', Fruhmittelalterliche Studien 1, pp. 156-69. Piechaud, S., & Poinsignon, V., 1988. 'Les thermes gallo-romaines de Ste.-Ruffine (Moselle)', Cahiers Lorrains (part 2, June), pp. 153-7. REL 1901-3. Reichsland Elsass Lothringen, 3 vols. (Strasbourg). Reusch, W., 1941-2. Trankische Funde aus lothringischem Boden; Bericht iiber den gegenwartigen Stand der fruhgeschichtlichen Spatenforschung in Lothringen', Westmarkische Abhandlungen 5, pp.39-58. 1943. 'Die St Peter-Basilika auf der Zitadelle in Metz', Germania 27, pp.79-92. des Robert, F., 1878. 'Notice sur quelques sepultures decouvertes pres de Dommartin par un habitant de Dampvitoux', JSAL, p.22ff and p.45. de Sailly, H., 1869. 'Premiere excursion dans le Barrois Mosellan5, MSAHM., pp.161-206. Salin, E., 1937. 'Caracteres generaux, nomenclature et carte archeologique des cimetieres du haut moyen-age du departement de Meurthe-et-Moselle', Bulletin Monumental 1, pp.191-218. 1939. Rhin et Orient, I: Le Haut Moyen Age en Lorraine dyapres le Mobilier Funeraire (Paris). 1946. 'Le cimetiere de Varangeville (Meurthe-et-Moselle)', Gallia 4, pp. 199-245. Sauer, C , 1958. 'Trouvailles romaines et merovingiennes d'Oermingen', CAHA 38, pp.834. de Saulcy, C , 1831-2. 'Notes sur quelques antiquites trouves a Dieulouard', MAM, pp.186-97. Schahle, W., 1961. 'Merowingerzeitliche Frauengraber aus Giidingen, Kr. Saarbriicken', Ber.SDS 8, pp.11-22 (+Taf. 3-4). 1965. 'Die Reihengraber von Walsheim', Ber.SDS 12, pp.107-20 (+Taf. 17-18). Schaub, J., & Petit, J.-B., 1984. Bliesbrucken. Gallo-rb'mischen Siedlung in Lothringen (Sarreguemines) (exactly parallel French edition also exists). Schaudel, L., 1894. 'Decempagi-TarquimpoP, JSAL, pp.156-8. Scheneker, P., 1899. 'Das alamannisch-frankisch Grabfeld bei Busendorf', JGLGA 11, pp.367-72. Schindler, R., 1962. 'Neues vom Kastell und Vicus Saarbriicken', Ber.SDS 9, pp.l2ff. Schlemaire, G., 1974. 'Fouilles de Sauvetage au Pontiffroy a Metz en 1973. Cave S.l', ASHAL 74, pp.19-28. 1976. 'Fouilles de sauvetage au Pontiffroy a Metz. Batiments S.2 et S.3', ASHAL 76, pp.37-59. 1978. 'Fouilles de sauvetage au Pontiffroy a Metz. Sites S.4 a S.ll', ASHAL 78, pp.4163. Schlosser, 1881-4. 'Notice sur un sarcophage decouvert dans l'ancienne eglise de Diedendorf', BSCMHA 2nd ser., 12, pp. 100-29. 289

BIBLIOGRAPHY 1895. 'Bericht iiber zwei bei Rimsdorf aufgefundene Sarkophage', BSCMHA, 2nd ser., 17, pp.49-56. Schmit, J.A., 1879. 'Promenades aux alentours de Chateau-Salins', MSAL, pp.51-80. Schroter, F., 1846-59. Mittheilungen des Historisch-antiquarischen Vereins fiir die Stddte Saarbriicken und St-Johann und dem Umgegend. Uber die romischen Niederlassungen und die Romerstrasse in den Saargegend, 3 vols. (vol.1 1846; vol.11 1852; vol.III 1859) (Saarbriicken). Simmer, A., 1985. 'Une redecouverte d'epoque merovingienne. La riche sepulture de Kirschnaumen (Moselle)', RAE 36, pp.311-16. 1987. 'Le nord du departement de la Moselle a l'epoque merovingienne', RAE 38, pp.333-96. 1988. Le Cimetiere Merovingien d'Audun-le-Tiche: Archeologie d'Aujourd'hui No.l (Memoire de VAFAM no.l) (Paris). Simon, V., 1834-5a. 'Note sur les instruments en fer trouves pres de Grosyeux, commune d'Augny, pres Metz', MAM, pp.425-6. 1834-5b. 'Note sur quelques antiquites trouvees a Metz', MAM, pp.427-31. 1842-3. 'Notice archeologique sur Metz et ses environs', MAM, pp.337-55. 1843-4. 'Notice sur les sepultures des anciens', MAM, pp.245-66 1848-9. 'Notice sur le Sablon, pres Metz, et sur les sepultures qui y ont ete decouvertes', MAM, pp.46-60. 1850-1. 'Observations sur des sepultures antiques decouvertes dans diverses contrees des Gaules, et sur l'origine qui parait devoir leur etre attribue', MAM, pp.143-58. 1859. 'Documents archeologiques sur le departement de la Moselle', MSAHM., pp.57-75. Stein, F., 1989a. 'Les tombes d'un chef franc et de sa famille a Giidingen: considerations sur le role de l'aristocratie dans l'implantation franque entre la Meuse et la Sarre', Actes..., pp. 153-69. 1992. 'Les tombes d'un chef franc et de sa famille a Giidingen: considerations sur le role de l'aristocratie dans l'implantation franque entre la Meuse et la Sarre', Saarbrucker Studien und Materialen zur Altertumskunde 1, pp.117-44. Toussaint, M., 1938a. 'Essai sur la question franque en Lorraine. Deuxieme partie: repertoire des necropoles et sepultures isolees de l'epoque franque decouvertes en Lorraine. I. Dep. Meurthe-et-Moselle', Revue des Questions Historiques, no.4, pp.27-57. 1938b. 'Essai sur la question franque en Lorraine. Deuxieme partie: repertoire des necropoles et sepultures isolees de l'epoque franque decouvertes en Lorraine. II. Dep. Meuse', Revue des Questions Historiques, no.l, pp.26-51. 1938c. 'Essai sur la question franque en Lorraine. Deuxieme partie: repertoire des necropoles et sepultures isolees de l'epoque franque decouvertes en Lorraine. III. Dep. Vosges', Revue des Questions Historiques, no.l, pp.52-74. 1938d. 'Essai sur la question franque en Lorraine. Deuxieme partie: repertoire des necropoles et sepultures isolees de l'epoque franque decouvertes en Lorraine. IV. Dep. Moselle', Revue des Questions Historiques, no.6, pp.3-33. 1943-5. 'L'Origine du peuplement dans la vallee moyenne de la Seille, d'apres les noms de lieux et les decouvertes archeologiques', Bulletin Archeologique, pp.503-45. 1948. Metz a VEpoque Gallo-Romaine (= ASHAL 49) (Metz). 1950. Repertoire Archeologique du Departement de la Moselle (Nancy). Verdel, E., 1986. 'Fouille de sauvetage de l'Espace Serpenoise a Metz', Cahiers Lorrains (part 4, Oct.), pp.353-61. unpub. 'Archeologie urbaine a Metz. Fouille de sauvetage urgent de l'Espace Serpenoise, 1984' (typescript of unpublished article, DAHL, Metz). Verronais, M.G., 1844. Statistique Historique, Industrielle et Commerciale du Departement de la Moselle (Metz). Viville, M., 1817. Dictionnaire de Departement de la Moselle. . ., 2 vols. (Metz). Voinot, J., 1904. 'Les fouilles de Chaouilley. Cimetiere merovingien', MSAL 54, pp.5-80. Waton, M.-D., 1986. 'Metz: Pontiffroy (Moselle). Sauvetage 1983-1985', RAE 37, pp.75-97. Welter, T., 1902. 'Das frankische Grabfeld 'Haut-Zabes' bei Fraquelfing-Loerchingen', JGLGA 14, pp.474-5. Wichmann, 1892. 'Decempagi-TarquimpoF, JGLGA 4 (part 2), pp.116-66. 290

BIBLIOGRAPHY 1894. 'Ausgrabungen und Funde bei Saarburg, i.U,JGLGA 6, pp.313-23 ('Frankische Graber bei St-Ulrich', pp.316-17). Wiener, L., 1888. 'Cimetiere merovingien d'Homecourt', JSAL, pp.198-200. 1895. Catalogue des Objets d'Art et d'Antiquite. Musee Historique Lorrain (Nancy).

Secondary references This book touches upon many areas of research with voluminous secondary literature; to cite all of it would be prohibitive in terms of space, even if I could claim to be familiar with it all. Consequently, I have, wherever appropriate, limited general references to recent works, preferably in English, with good bibliographies. I apologize to the authors of those works which I have studied and learnt from, but which do not appear here. Affeldt, W. (ed.), 1989. Frauen in Spdtantike und Fruhmittelalter. Lebensbedingungen Lebensnormen - Lebensformen (Sigmaringen). Affeldt, W., Nolte, C, Reiter, S., & Vorwerk, U. (eds.), 1990. Frauen im Frubmittelalter. Fine ausgewdhlte kommentierte Bibliographie (Frankfurt-am-Main). Ament, H., 1977. 'Zur archaologischen Periodisierung der Merowingerzeit', Germania 55, pp.133-40. Amory, P., 1993. 'The meaning and purpose of ethnic terminology in the Burgundian laws', Early Medieval Europe 2.1, pp. 1-28. Anderson, P., 1974. Passages from Antiquity to Feudalism (London). Arnold, C.J., 1980. 'Wealth and social structure: a matter of life and death', in Rahtz, Dickinson & Watts (eds.) 1980, pp.81-142. Baker, D. (ed.), 1978. Studies in Church History. Subsidia 1: Medieval Women (paperback edition: Medieval Women, Oxford 1978). Bange, F., 1984. 'L'Ager et la villa: structures du paysage et du peuplement dans la region Maconnaise a la fin du Haut Moyen Age (IXe-XIe siecles)', Annales ESC 41, pp.52969. Barley, M.W. (ed.), 1977. European Towns. Their Archaeology and Early History (London). Barnish, S., 1992. 'Old Kaspars: Attila's invasion of Gaul in the literary sources', in Drinkwater & Elton (eds.) 1992, pp.38-47. Barnwell, P.S., 1992. Emperor, Prefects and Kings. The Roman West, 395-565 (London). Baudoux, J., et al, 1990. La Lorraine Antique. Villes et Villages (Metz). Beaujard, B., 1976. 'Les vici de Mediomatriques au Bas-Empire', Caesarodunum 11 (numero special: actes du colloque 'le vicus gallo-romain'), pp.296-308. Bergengruen, A., 1958. Adel und Grundherrschaft im Merowingerreich (Wiesbaden). Bergthol, E., 1955. cLe Christianisme au Mont Herapel', ASHAL 55, pp.5-11. Bernardi, B., 1985. Age Class Systems. Social Institutions and Polities Based on Age (Cambridge). Bertemes, F., Miron, A., Petit, J.-P., Reinhard, W., Schaub, J., Scmitt, G., Schumacher, F., & Strahl, E., 1988. Saar-Pfalz-Kreis (Fuhrer zu archaologischen Denkmdlern in Deutschland 18) (Stuttgart). Bloch, M., 1947. 'Comment et pourquoi finit Pesclavage antique?', Annales ESC 2, pp.3044, 161-70. 1962. Feudal Society, 2 vols., 2nd English edition (trans. L.A. Manyon) (London). Blouet, V., Brunella, P., Heckenbenner, D., Lefebvre, C, Legendre, J.-P., Ollivier, L., & Waton, M.-D., 1983. 'La Lorraine', in Architectures de Terre et de Bois: VHabitat Prive des Provinces Occidentales du Monde Romain. Antecedents et Prolongements: Protohistoire, Moyen Age et Quelques Experiences Contemporaines. (Documents d'Archeologie Franqaise 2), ed. J. Lasfragues (Paris), pp. 103-12. Bohme, H.-W., 1974. Germanische Grabfunde des 4 bis 5 Jahrhunderts zwischen untere Elbe und Loire. Studien zur Chronologie und Bevb'lkerungsgeschichte (Munich). Bohner, K., 1958. Die Frdnkischen Altertiimer des Trierer Landes. (Germanische Denkmdler der Volkerwanderungszeit, Serie B, Band 1), 2 vols. (Berlin). 1977. 'Urban and rural settlement in the Frankish kingdom', in Barley (ed.) 1977, pp.185-201. 291

BIBLIOGRAPHY Bois, G., 1992. The Transformations of the Year 1000: The Village of Lournand from Antiquity to Feudalism, trans. J. Birrell (Manchester). Bonnassie, P., 1991a. From Slavery to Feudalism in South-Western Europe, trans. J. Birrell (Cambridge). 1991b. 'The survival and extinction of the slave system in the early medieval west (fourth to eleventh centuries)', in idem. 1991a, pp. 1-59. 1991c. 'Society and mentalities in Visigothic Spain', in idem. 1991a, pp.60-103. Bour, R.-S., 1930. 'Observations sur la plus ancienne cathedrale de Metz', ASHAL 39, pp. 155-62. 1938. 'Un dernier mot sur la plus ancienne cathedrale de Metz', ASHAL 47, pp.7-15. Bourdieu, P., 1977. Outline of a Theory of Practice, trans. R. Nice (Cambridge). Bourgeat, G., & Dorvaux, N., 1907. Atlas Historique du Diocese de Metz (Metz). Braudel, F., 1972. The Mediterranean and the Mediterranean World in the Age of Philip II, English translation of 2nd edition (London). Brenot, C , 1991. 'Les monnaies', in Duval et al. (ed.) 1991, pp.164-83. Brown, P.R.L., 1981. The Cult of the Saints. Its Rise and Function in Latin Christianity (Chicago). 1982. Society and the Holy in Late Antiquity (London). Brunella, P., Heckenbenner, D., Lefebvre, C , & Thion, P., 1988. Metz. Cinq Annies de Recherches Archeologiques, 1982-1987 (Metz). Burnand, Y., 1990. Encyclopedic Illustre de la Lorraine. Histoire de la Lorraine. Les Temps Anciens 2, de Cesar a Clovis (Metz and Nancy). Burnell, S., 1988. 'Merovingian to early Carolingian churches and their founder-graves in southern Germany and Switzerland: the impact of Christianity on the Alamans and Bavarians', 3 vols., Univ. of Oxford DPhil thesis. Burnouf, J., 1988. 'Postface: Bilan et Perspectives', in Delestre (ed.) 1988, pp.111-14. Campbell, J., 1979. 'Bede's words for places', in Sawyer (ed.) 1979, pp.34-54. 1989. 'The sale of land and the economics of power in early England: problems and potentials', Haskins Society Journal 1, pp.23-37. Cardot, F., 1987. VEspace et le Pouvoir. Etude sur VAustrasie Merovingienne (Paris). Carver, M.O.H. (ed.), 1992. The Age of Sutton Hoo. The Seventh Century in NorthWestern Europe (Woodbridge). Charles-Edwards, T.M., 1976. 'The distinction between land and moveable wealth in Anglo-Saxon England', in Medieval Settlement. Continuity and Change, ed. P.H. Sawyer (London), pp. 180-7. Chazan, M., 1992. 'L'Erudition et conscience urbaine dans Teloge de la ville de Metz' de Sigebert de Gembloux', Cahiers Lorrains (part 3-4, Oct.), pp.441-53. Cheyette, F.L. (ed. & trans.), 1968. Lordship and Community in the Middle Ages. Selected Readings (New York). Christlein, R., 1991. Die Alamannen. Archdologie eines lebendigen Volkes, 3rd edition (Stuttgart and Aalen; 1st edition Stuttgart, 1978). Clarke, D.L., 1972. 'Glastonbury lake village. A provisional model of an Iron Age society and its settlement pattern', in idem, (ed.) Models in Archaeology (London), pp.801-69. Clarke, H.B., & Brennan, M. (eds.), 1981. Columbanus and Merovingian Monasticism, BAR (S) 13 (Oxford). Classen, P., 1955. 'Kaiserreskript und Konigsurkunde: Diplomatische Studien zum romisch-germanischen Kontinuitatsproblem. Teil 1', Archiv fiir Diplomatik 1, pp. 1-87. 1956. 'Kaiserreskript und Konigsurkunde: Diplomatische Studien zum romischgermanischen Kontinuitatsproblem. Teil 2', Archiv fiir Diplomatik 2, pp.1-115. Coleman, E.R., 1976. 'Infanticide in the early middle ages' in Stuard (ed.) 1976, pp. 47-70. Collins, R., 1983. Early Medieval Spain. Unity in Diversity, 400-1000 (London). Cracco Ruggini, L., 1992. 'The crisis of the noble saint: The "Vita Arnulfi'", in The Seventh Century. Change and Continuity, ed. J. Fontaine & J.N. Hillgarth (London), pp.116-53. Cuvelier, P., & Guillaume, J., 1989. 'Inventaire et typologie des sarcophages en Lorraine', Actes..., pp.87-96. 292

BIBLIOGRAPHY Davies, W., 1988. Small Worlds. The Village Community in Early Medieval Brittany (London). Davies, W., & Fouracre, P. (eds.), 1986. The Settlement of Disputes in Early Medieval Europe. (Cambridge). Delestre, X. (ed.), 1988. La Lorraine Merovingienne (Metz). Demolon, P., 1972. Le village Merovingien de Brebieres (VP-WIY siecles) (Arras). Devroey, J.-P., 1981. 'Les methodes d'analyse demographique des polyptyques du Haut Moyen Age', in Histoire et Methode (Acta Historica Bruxellensia 4) ed. M.A. Arnould, C. Billen & C. Dupont (Brussels), pp.71-88. Dockes, P., 1982. Medieval Slavery and Liberation, trans. A. Goldhammer (Chicago). Doehaerdt, R., 1978. The Early Middle Ages in the West. Economy and Society, trans. W.G. Deakin (Amsterdam). Drinkwater, J.F., 1989. 'Patronage in Roman Gaul and the problem of the Bagaudae', in Patronage in Ancient Society, ed. A. Wallace-Hadrill (London), pp.189-203. 1992. 'The Bacaudae of fifth-century GauF, in Drinkwater & Elton (eds.), pp.208-17. Drinkwater, J.F., & Elton, H. (eds.), 1992. Fifth-Century Gaul: A Crisis of Identity? (Cambridge). Dunbabin, J., 1993. Review of Bois 1992. English Historical Review 108, no.428, pp.697-8. Durand, M., 1988. 'Cimetieres et habitats: l'emergence du village au sud-est du departement de l'Oise', Revue Archeologique de Picardie (part 3—4) (Actes des VHP Journees de VAFAM (19-22 Juin 1986)), pp.153-60. Durliat, J., 1990. Les Finances Publiques de Diocletien aux Carolingiens (284-890 (Beiheft der Francia 21) (Sigmaringen). Duval, N., 1991. 'L'Architecture Cultuelle', in idem, et al (ed.) 1991, pp.186-219. Duval, N., Fontaine, J., Fevrier, P.-A., Picard, J.-C, & Barruol, G. (eds.), 1991. Atlas Archeologique de la France. Naissance des Arts Chretiens. Atlas des Monuments Paleochretiens de la France (Paris). Duval, Y., & Picard, J.-C. (eds.), 1986. L3Inhumation Privilegiee du IVe au VHP siecle en Occident (Paris). Ebling, H., 1974. Prosopographie der Amstrdger des Merowingerreiches, von Chlothar II (613) bis Karl Martell (741) (Beiheft der Francia 2) (Munich). Eckhardt, W.A., 1967. 'Die Decretio Childeberti und ihre Uberlieferung', Zeitschrift fur Rechtsgeschichte, Germ. Abt. 84, pp. 1-70. Elziere, J.-B., 1993. 'Sud du Massif Central. Eglises de Rhenanie et royaume des Austrasiens', Bulletin Monumental 153.1, pp.47-72. Ennen, E., 1959. 'Die Entwicklung des Stadtwesens an Rhein und Mosel vom 6 bis 9 Jahrhundert', La Citta Nell3Alto Medioevo. Settimane di Studio 6 (Spoleto), pp.41952. Esmonde Cleary, A.S., 1989. The Ending of Roman Britain (London). Ewig, E., 1952. 'Die frankische Teilreiche (511-613)', Akadamie der Wissenschaften und der Literatur Mainz. Abhandlungen der geistes- und sozialwissenschaftlichen Klasse 9, pp.651-715. 1953. 'Die frankische Teilreiche im 7. Jahrhundert (613-714)', Trierer Zeitschrift 22, pp.85-144. 1954. Trier im Merowingerreich. Civitas, Stadt, Bistum (Trier). 1963. 'Residence et capitale pendant le haut moyen-age', Revue Historique 230, pp.2572. Feller, M., & Poplineau, G., 1988. 'Ateliers de ceramique gallo-romaine d'Argonne: prealables methodologiques', Cahiers Lorrains (part 2, June) pp. 173-80. Filmer-Sankey, W., 1992. 'Snape Anglo-Saxon cemetery: the current state of knowledge', in Carver (ed.) 1992, pp.39-51. Finley, M. 1980. 'The decline of ancient slavery', in idem, Ancient Slavery and Modern Ideology (London), pp. 123-49. Folz, R., 1974. The Coronation of Charlemagne, 25 Dec, 800, trans. J.E. Anderson (London). Fouracre, P., 1986. ' "Placita" and the settlement of disputes in later Merovingian Francia', in Davies & Fouracre (eds.) 1986, pp.23-43. 293

BIBLIOGRAPHY 1990. 'Merovingian history and Merovingian hagiography', Past and Present 127, pp.338. 1992. 'Cultural conformity and social conservatism in early medieval Europe', History Workshop Journal 33, pp.152-61. Franc, ois, J., & Tabouillot, N., 1769-90. Histoire Generate de Metz par des Religieux Benedictins de la Congregation de Saint-Vanne (Metz and Nancy). Galinie, H., 1978. 'Archeologie et topographie historique de Tours lVeme-IXeme siecle', Zeitschrift fiir Archaologie des Mittelalters 6, pp.33-56. 1988. 'Reflections on early medieval Tours1, in Hodges & Hobley (eds.) 1988, pp.57-62. Ganshof, F.L., 1962. 'Le statut de la femme dans la monarchic franque', Receuil de la Societe Jean Bodin 12, pp.5-58. 1964. Feudalism, 3rd edition (London). Gauthier, N., 1980. UEvangelisation des Pays de la Moselle. La Province Romaine de Premiere Belgique entre Antiquite et Moyen-Age (IIIe-VHIe siecles) (Paris). 1986. Topographie Chretienne des Cites de la Gaule des Origines au Milieu du VHP Siecle, I: Province Ecclesiastique de Treves (Belgica Prima) (Paris). 1988. 'La fondation de l'abbaye de Longeville-les-St.-Avoid', Cahiers Lorrains (part 4, Dec), pp.369-78. Geary, P., 1988. Before France and Germany. The Creation and Transformation of the Merovingian World (Oxford). Genicot, L., 1962. 'La noblesse au moyen-age dans Pancienne "Francie": continuite, rupture ou evolution?' Comparative Studies in History and Society 5, pp.52-9 (transl. in Cheyette (ed. & trans.) 1968:128-36). George, J.W., 1992. Venantius Fortunatus. A Poet in Merovingian Gaul (Oxford). Georges, M., 1988. 'Le "vicus" gallo-romain du Herapel a Cocheren (Moselle): chronologie et fonctions', Cahiers Lorrains (part 2, June), pp.163-71. Gerberding, R., 1987. The Liber Historiae Francorum and the Rise of the Carolingians (Oxford). Goetz, H.-W., 1993. 'Serfdom and the beginnings of a "seigneurial system" in the Carolingian period: a survey of the evidence', Early Medieval Europe 2.1, pp.29-51. Goffart, W., 1972. 'From Roman taxation to medieval seigneurie: three notes', Speculum 47, pp. 165-87 and 373-94. Reprinted in Goffart 1989, ch.7. 1982a. 'Old and new in Merovingian taxation', Past and Present 96, pp.3-21. Reprinted in Goffart 1989, ch.8. 1982b. 'Merovingian polyptychs: reflections on two recent publications', Francia 9, pp.55-77. Reprinted in Goffart 1989, ch.9. 1988. The Narrators of Barbarian History, (A.D.5 50-800). Jordanes, Gregory of Tours, Bede, Paul the Deacon (Princeton). 1989. Rome's Fall and After (London). Grahn-Hoek, H., 1976. Die frdnkische Oberschicht im 6 Jahrhundert. Studien zur ihrer rechtlichen und politische Stellung (Sigmaringen). Graus, F., 1965. Volk, Herrscher und Heiliger im Reich der Merowinger (Prague). Hall, R.A., 1988. 'York 700-1050', in Hodges & Hobley (eds.) 1988, pp.125-32. Halsall, G., 1989. 'Anthropology and the study of pre-Conquest warfare and society: the ritual war in Anglo-Saxon England', in Weapons and Warfare in Anglo-Saxon England, ed. S. Chadwick-Hawkes, Oxford University Committee for Archaeology Monograph 21 (Oxford), pp.155-77. 1990. 'Civitas Mediomatricorum: settlement and social organisation in the Merovingian region of Metz', Univ. of York DPhil thesis. 1992a. 'The origins of the Reihengrdberzivilisation: forty years on', in Drinkwater and Elton (eds.) 1992, pp. 196-207. 1992b. 'Social change around 600 A.D.: an Austrasian perspective', in Carver (ed.) 1992, pp.265-78. forthcoming. Early Medieval Cemeteries: An Archaeological Introduction to Burial in the Post-Roman West (Glasgow). Harris, M., 1978. Cannibals and Kings: The Origins of Cultures (London). 294

BIBLIOGRAPHY Haslam, J., 1987. 'Market and fortress in England in the reign of Offa', World Archaeology 19, no.3, pp.76-93. Hedeager, L., 1992. 'Kingdoms, ethnicity and material culture: Denmark in a European perspective', in Carver (ed.) 1992, pp.279-300. Heidrich, I., 1965. 'Titulatur und Urkunden der arnulflngischen Hausmeier', Archiv fiir Diplomatik 10, pp.71-279. 1974. 'Die merowingische Munzpragung im Gebiet von oberer Maas, Mosel und Seille', Rheinische Vierteljahrsbldtter 38, pp.78-91. Heighway, C. (ed.), 1972. The Erosion of History. Archaeology and Planning in Towns (London). Heitz, C , 1967. 'La groupe cathedrale de Metz a Pepoque de St. Chrodegang', in Saint Chrodegang, ed. J. Schneider (Metz), pp.123-32. 1982. 'Metz et son groupe episcopal a Pepoque pre-carolingienne et carolingienne', Eglises de Metz dans le Haut Moyen-Age (Cahier no A du Centre de Recherches sur VAntiquite Tardive et le Haut Moyen Age, Paris X Nanterre) (Paris), pp.5-13. Herlihy, D., 1976. 'Land, family and women in continental Europe (701-1200)', in Stuard (ed.) 1976, pp.13-45. 1985. Medieval Households (Cambridge, Mass.). Hermann, H.W., 1963. 'Zum Stande der Erforschung der fruh- und hochmittelalterlichen Geschichte des Bistums Metz', Rheinische Vierteljahrsbldtter 28, pp. 131-99. Hiegel, C , 1981. 'Le sel en Lorraine du VHP au XIIP siecle', Annales de VEst, 5th ser., 33, no.l, pp.3-48. Hilton, R., 1992. English and French Towns under Feudalism. A Comparative Study (Cambridge). Hingley, R., 1989. Rural Settlement in Roman Britain (London). Hlawitschka, E., 1962. 'Zur landschaftlichen Herkunft der Karolinger', Rheinische Vierteljahrsbldtter 27, pp. 1-17. Hodder, I., 1986. Reading the Past. Current Approaches to Interpretation in Archaeology (Cambridge). Hodges, R., 1982. Dark Age Economics. The Origins of Towns and Trade, AD 600-1000 (London). Hodges, R., & Hobley, B. (eds.), 1988. The Rebirth of Towns in the West, 700-1050 (Council for British Archaeology, London). Hodges, R., & Whitehouse, D., 1983. Mohammed, Charlemagne and the Origins of Europe (London). Hugget, J., & Richards, J., 1990. 'Anglo-Saxon burial: the computer at work', in AngloSaxon Cemeteries. A Reappraisal, ed. Southworth, E. (Stroud), pp.65-85. Hunt, E.D., 1992. 'Gaul and the Holy Land in the early fifth century', in Drinkwater & Elton (eds.) 1992, pp.264-74. Irsigler, F., 1969. Untersuchungen zur Geschichte des fruhfrdnkischen Adels (Bonn), ch.2 translated as 'On the aristocratic character of early Frankish society', in Reuter (ed. & trans.) 1979, pp.105-36. James, E., 1977. The Merovingian Archaeology of South-West Gaul, 2 vols., BAR (S) 25 (Oxford). 1979. 'Cemeteries and the problem of Frankish settlement in Gaul', in Sawyer (ed.) 1979, pp.55-89. 1982. The Origins of France. From Clovis to the Capetians, 500-1000 (London). 1988a. The Franks (Oxford). 1988b. 'Childeric, Syagrius et la disparition du royaume de Soissons', Revue Archeologique de Picardie, pp.9-12. 1989. 'Burial and status in the early medieval West', TRHS 5th ser., 29, pp.23-40. Janssen, W., & Lohrmann, D. (eds.), 1983. Villa - Curtis - Grangia. Landwirtschaft zwischen Loire und Rhein von der Rb'merzeit zum Hochmittelalter (Beiheft der Francia 11) (Munich and Zurich). Kaiser, R., 1973. Untersuchungen zur Geschichte der Civitas und Diozese Soissons in romischer und merowingischer Zeit (Bonn). 295

BIBLIOGRAPHY Kapelle, W.E., 1979. The Norman Conquest of the North. The Region and its transformation, A.D.1000-1135 (London). Kennedy, H., 1985. 'From polis to madina: urban change in late antique and early islamic Syria', Past and Present 106, pp.3-27. Kiefer, T.M., 1972. The Tausug. Violence and Law in a Philippine Moslem Society (New York). Kienast, D., 1962. Die Fundmiinzen der Rb'merzeit in Deutschland. 3 (Saarland) (Berlin). King, C.E., 1992. 'Roman, local and barbarian coinages in fifth-century Gaul', in Drinkwater & Elton (eds.) 1992, pp.184-95. King, P.D., 1980. 'King Chindasvind and the first territorial law-code of the Visigothic kingdom', in Visigothic Spain: New Approaches, ed. E. James (Oxford), pp. 131-57. Kuchenbuch, L., 1978. Bduerliche Gesellschaft und Klosterherrschaft im 9. Jahrhundert. Studien zur Sozialstruktur der 'familia3 der Abtei Pru'm (Vierteljahrschrift fur Sozial- und Wirtschaftsgeschichte Beiheft 66) (Wiesbaden). 1986. Trennung und Verbindung im bauerlichen Werken des 9. Jahrhunderts. Eine Auseinandersetzung mit Ivan Illichs Genus-Konzept', in Frauen in der Geschichte 7: Interdisziplindre Studien zur Geschichte der Frauen im Fruhmittelalter. Methoden-ProblemeErgebnisse, ed. W. Affeldt & A. Kuhn (Dusseldorf), pp.227-42. 1988. cDie Klostergrundherrschaft im Fruhmittlelater. Ein Zwischenbilanz', in Herrschaft und Kirche. Beitrdge zur Entstehung und Wirkungsweise episcopaler und monastischer Organisationsformen (Monographien zur Geschichte des Mittelalters 33), ed. F. Prinz (Stuttgart), pp.297-343. 1991. Die Grundherrschaft im fruhen Mittelalter (Historisches Seminar Neue Folge 1) (Idstein). Lafaurie, J., 1964. 'Monnaie en argent trouvee a Fleury-sur-Orne. Essai sur le monnayage d'argent franc des Ve et VP siecles', Annales de Normandie 14, pp. 197-222. 1987. 'Les frappes monetaires de Metz et sa region aux VIe-IXe siecles', in Riche, Heitz & Heber-Suffrin (ed.) 1987, pp.89-115. Latouche, R., 1967. The Birth of Western Economy, trans. E.M. Wilkinson (London). Laumon, A., 1973. 'Le pays de Sarrebourg a l'epoque merovingienne', Memoire de maitrise, Universite de Nancy II, faculte de lettres et sciences humaines. Lebecq, S., 1990. 'La famille et les apprentissages de Liudger, d'apres les premiers chapitres de la vita Liudgeri d'Altfrid', in Haut Moyen-Age. Culture, Education et Societe. Etudes Offertes a Pierre Riche, ed. M. Sot et al (Paris), pp.283-99. Leclercq, H., 1933. 'Metz', in Dictionnaire d'Archeologie Chretienne et Liturgique, vol.11, cols. 790-866. Lefebvre, C., & Demarolle, J.M., 1986. 'Mettis, cite de l'antiquite tardive', in Histoire de Metz, ed. F.-Y. Le Moigne (Toulouse), pp.38-44. Levison, W., 1948a. Aus rheinischer und frdnkischer Fruhzeit (Dusseldorf). 1948b. 'Das Testament des Diakons Adalgisel-Grimo vom Jahre 634', in idem. 1948a, pp.118-38. 1948c. 'Metz und Sudfrankreich im fruheren Mittelalter', in idem. 1948a, pp.139-63. Lorren, C., 1985. 'Le village de St.-Martin-de-Mondeville, de l'antiquite au haut moyen age', in Perin & Feffer (eds.) 1985, pp.350-61. Loseby, S.T., 1992. 'Bishops and cathedrals: order and diversity in the fifth-century urban landscape of southern Gaul', in Drinkwater & Elton (eds.), 1992, pp. 144-55. Lutz, M., 1964. 'Considerations sur la civilisation dite 'des Sommets Vosgiens' a la lumiere de decouvertes recentes', ASHAL 78, pp.11-39. Lynch, J.H., 1986. Godparents and Kinship in Early Medieval Europe (Princeton). McNamara, J.-A., & Wemple, S.F., 1976. 'Marriage and divorce in the Frankish kingdom', in Stuard (ed.) 1976, pp.95-124. Mann, M., 1986. The Sources of Social Power, I: A History of Power from the Beginning to A.D. 1760 (Cambridge). Mathisen, R.W., 1992. 'Fifth-century visitors to Italy: business or pleasure?', in Drinkwater & Elton (eds.) 1992, pp.228-38. Matthews, J., 1975. Western Aristocracies and Imperial Court, AD 364-425 (Oxford). 296

BIBLIOGRAPHY Mehlis, C , 1877. Studien zur dltesten Geschichte der Rheinlande (hersgeg. von hist. Verein. der Pfalz), vol.III (Leipzig). 1879. Studien zur dltesten Geschichte der Rheinlande (hersgeg. von hist. Verein. der Pfalz), vol.IV (Leipzig). Miller, D., 1985. Artefacts as Categories. A Study of Ceramic Variability in Central India (Cambridge). Minois, G., 1989. History of Old Age from Antiquity to the Renaissance, trans. S.H. Tenison (London). Morhain, E., 1960-1. fLes origines du christianisme a Metz et en Moselle', ASHAL 60, pp.87-129. Morimoto, Y., 1988. 'Etat et perspectives des recherches sur les polyptyques carolingiens', Annales de I'Est, 5th ser., 40, no.2, pp.99-149. Morris, I., 1992. Death-Ritual and Social Structure in Classical Antiquity (Cambridge). Murray, A.C., 1983. Germanic Kinship Structure: Studies in Law and Society in Antiquity and the Early Middle Ages (Toronto). 1986. The position of the Grafio in the constitutional history of Merovingian GauP, Speculum 64, pp.787-805. Naveau, J., 1987. 'L'Archeologie merovingienne en Mayenne: etat des recherches', in Bretagney Pays de Loire-Touraine-Poitou a VEpoque Merovingienne (Paris), pp.87-92. Nelson, J.L., 1978. 'Queens as Jezebels: the careers of Brunhild and Balthild in Merovingian history', in Baker (ed.) 1978, pp.31-77. repr. in Nelson 1986, no.l. 1986. Politics and Ritual in Early Medieval Europe (London). 1990. 'Women and the word in the early middle ages', in Women in the Church (Studies in Church History 27), ed. WJ. Shiels & D. Wood (Oxford), pp.53-78. Notin, V., & Desbordes, J.-M. (eds.), 1990. Augustoritum. Aux Origines de Limoges (Limoges). Oexle, O.G., 1967. 'Die Karolinger und die Stadt des heiligen Arnulf, Fruhmittelalterliche Studien 1, pp.250-364. von Olberg, G., 1989. 'Aspekte der rechtlich-sozialen Stellung der Frauen in den fruhmittelalterlichen Leges', in Affeldt (ed.) 1989, pp.221-35. Pader, E.-J., 1980. 'Material symbolism and social relations in mortuary studies', in Rahtz, Dickinson & Watts (eds.) 1980, pp.143-59. 1981. Symbolism, Social Relations and the Interpretation of Mortuary Remains, BAR(S) 130 (Oxford). Parisse, M., 1978. 'Paroisses etpagi\ in idem, (ed.) 1978, pp.97-9. 1981. La Lorraine Monastique (Nancy). 1987. 'La reine Hildegard et PAbbaye St.-Arnoul de Metz', in Riche, Heitz & HeberSuffrin (ed.) 1987, pp.41-47. 1990. Encyclopedie Illustre de la Lorraine. Histoire de la Lorraine. VEpoque Medievale (Nancy and Metz). Parisse, M. (ed.), 1978. Histoire de la Lorraine (Toulouse). Patterson, O., 1982. Slavery and Social Death (Cambridge, Mass.). Paxton, F.S., 1990. Christianizing Death. The Creation of a Ritual Process in Early Medieval Europe (Ithaca). Percival, J., 1969. 'Seigneurial aspects of late Roman estate management', English Historical Review 82, pp.449-73. 1976. The Roman Villa. An Historical Introduction (London). Perin, P. (with contribution by R. Legoux), 1980. La Datation des Tombes Merovingiennes (Paris and Geneva). Perin, P., & Feffer, L.C., 1987. Les Francs, 2 vols. (Paris). Perin, P., & Feffer, L.-C. (eds.), 1985. La Neustrie. Les Pays au Nord de la Loire, de Dagobert a Charles le Chauve (Creteil). Perrin, C.-E., 1935. Recherches sur la Seigneurie Rurale en Lorraine d'apres les Plus Anciens Censiers, IXe-XIIe Siecle (Paris). Picard, J.C., 1988. 'L'Austrasie: entite geographique ou politique?', in Delestre (ed.) 1988, pp.13-21. 297

BIBLIOGRAPHY Polanyi, K., 1978. 'Trade, markets and money in the European early middle ages', Norwegian Archaeological Review 11, no.2, pp.92-6. Pontal, O., 1989. Histoire des Conciles Merovingiennes (Paris). Prinz, F., 1965. Frubes Monchtum im Frankenreich (Munich and Vienna). 1967. 'Heiligenkult und Adelsherrschaft im Spiegel merowingischer Hagiographie', Historische Zeitschrift 204, pp.529-44. 1981. 'Columbanus, the Frankish nobility and the territories east of the Rhine', in Clarke & Brennan (eds.) 1981, pp.73-87. Rahtz, P.A., Dickinson, T.M., & Watts, L. (eds.), 1980. Anglo-Saxon Cemeteries, 1979, BAR(B) 82 (Oxford). Reece, R., 1980. 'Town and country. The end of Roman Britain', World Archaeology 12, no.l, pp.77-92. 1981. 'The third century: crisis or change?', in The Roman West in the Third Century ed. A. King & M. Hennig, BAR(S) 109 (Oxford), pp.27-38. Reuter, T., 1991. Germany in the Early Middle Ages, 800-1056 (London). Reuter, T. (ed. & trans.), 1979. The Medieval Nobility (Amsterdam). Reynolds, S., 1991. 'Bookland, folkland and fiefs', Anglo-Norman Studies 14, pp.211-27. Riche, P., 1976. Education and Culture in the Barbarian West. Sixth through Eighth Centuries, trans. J. J. Contreni (Columbia). Riche, P., Heitz, C , & Heber-Suffrin, F. (eds.), 1987. Actes du Colloque "Autour d'Hildegardeyi (Cahier no. 5 du Centre de Recherches sur VAntiquite Tardive et le Haut Moyen Age, Univ. de Paris X-Nanterre, et Centre de Recherches d'Histoire de VUniv. de Metz) (Paris). Roberts, C.A., et al. (eds.), 1989. Burial Archaeology. Current Research, Methods and Developments BAR(B) 211 (Oxford). Rolph, T.C., Shaw, J., & Grant, J.E., 1987. 'Geomagnetic field variations as a dating tool: application to Sicilian lavas', Journal of Archaeological Science 14, pp.215-25. Rouche, M., 1979. UAquitaine des Wisigoths aux Arabes, 418-781. Naissance dyune Region (Paris). 1986. 'La crise de l'Europe au cours de la deuxieme moitie du Vile siecle et la naissance des regionalismes', Annales ESC 43, pp.347-60. Runciman, W.G., 1983. A Treatise on Social Theory, I: The Methodology of Social Theory (Cambridge). 1989. A Treatise on Social Theory, II: Substantive Social Theory (Cambridge). Salin, E., 1949. La Civilisation Merovingienne d'apres les Textes les Sepultures et la Laboratoire, I (Paris). 1952. La Civilisation Merovingienne d'apres les Textes les Sepultures et la Laboratoire, II (Paris). Samson, R., 1987a. 'Social structures in Reihengrdber: mirror or mirage?', Scottish Archaeological Review 4, part 2, pp. 116-26. 1987b. 'The Merovingian nobleman's home: castle or villa?' Journal of Medieval History 13, pp.287-315. Sawyer, P.H. (ed.), 1979. Names, Words and Graves. Early Medieval Settlement (Leeds). Sawyer, P.H., & Wood, I.N. (eds.), 1977. Early Medieval Kingship (Leeds). Schlesinger, W., 1976. 'Hufe und Mansus im Liber Donationum des Klosters Weissenburg', Beitrdge zur Wirtschafts und Sozialgeschichte des Mittelalters. Festschrift fur H. Helbig (Cologne and Vienna), pp.33-85. Schmidt-Wiegand, R., 1989. 'Der Lebenskreis der Frau im Spiegel der volkssprachigen Bezeichungen der Leges Barbarorum', in Affeldt (ed.) 1989, pp. 195-209. Schneider, J, 1950. La Ville de Metz au XIIP et XIVe Siecles (Nancy). Shanin, T., 1987. 'Introduction: peasantry as a concept' in Peasants and Peasant Societies, ed. idem., 2nd edition (Harmondsworth). Shanks, M., & Tilley, C, 1987. Social Theory and Archaeology (Cambridge). Shephard, J., 1979. 'The social identity of the individual in isolated barrows and barrow cemeteries in Anglo-Saxon England', in Space, Hierarchy and Society. Interdisciplinary Studies in Social Area Analysis, ed. R. Burnham & J. Kingsbury, BAR(S) 59 (Oxford), pp.47-79. 298

BIBLIOGRAPHY Simmer, A., 1982. 'Le prelevement des cranes dans l'est de la France a l'epoque merovingienne', Arch. Med. 12, pp.35-49. 1983. 'Les tombes doubles a l'epoque merovingienne: l'exemple d'Audun-le-Tiche', RAE 34, pp. 170-2. Sprandel, R., 1961. 'Struktur und Geschichte des merowingischen Adels', Historische Zeitschrift 193, pp.33-71. Stafford, P., 1983. Queens, Concubines and Dowagers. The King's Wife in the Early Middle Ages (London). Stahl, A.M., 1982. The Merovingian Coinage of the Region of Metz (Louvain). Stein, F., 1974. 'Franken und Romanen in Lothringen', in Studien zur vor- und fruhgeschichtlichen Archdologie. Festschrift fur Joachim Werner. (Miinchener Beitr. Vor- und Fruhgesch. Ergbd.l), pp.123-32. 1989b. 'Die Bevolkerung des Saar-Mosel-Raumes am Ubergang von der Antike zum Mittelalter', Archaeologia Mosellana 1, pp.89-195. Steuer, H., 1982. Fruhgeschichtliche Sozialstrukturen in Mitteleuropa: Eine Analyse der Auswertungsmethoden des Archdologischen Quellensmaterials. (Abh. Akad. Wiss. Go'ttingen Phil.-Hist. Kl 3. Folgey 128) (Gottingen). Stieber, A., 1954. 'Decouverte d'un cabane merovingien a Furdenheim (Bas-Rhin)', CAHA 34, pp.97-100. Stiller, G., 1986. Guide du Musee de la Tour aux Puces, Thionville (Thionville). Strayer, J.R., 1968. 'Feudalism in western Europe', in Cheyette (ed.) 1968, pp. 12-21. Stuard, S.M. (ed.), 1976. Women in Medieval Society (Philadelphia). Theis, L., 1976. 'Saints sans famille? Quelques remarques sur la famille dans le monde franc a travers les sources hagiographiques', Revue Historique 225, pp.3-20. Tilley, C. (ed.), 1991. Reading Material Culture (Oxford). Ucko, P., 1969. 'Ethnography and the archaeological interpretation of funerary remains', World Archaeology 1, no.2, pp.262-77. Van Dam, R., 1985. Leadership and Community in Late Antique Gaul (Berkeley). 1993. Saints and their Miracles in Late Antique Gaul (Princeton). Verhulst, A., 1991. 'The decline of slavery and the economic expansion of the early middle ages', Past and Present 133, pp.195-203. Verhulst, A. (ed.), 1985. Le Grand Domaine aux Epoques Merovingienne et Carolingienne (Ghent). Verlinden, C., 1955. UEsclavage dans VEurope Medievale (vol.1) (Ghent). Vieillard-Troiekouroff, M., 1976. Les Monuments Religieux de la Gaule, d'apres les oeuvres de Gregoire de Tours (Paris). Vigneron, B., 1986. Metz Antique (Metz). Vince, A., 1990. Saxon London. An Archaeological Investigation (London). Wagner, P.-E., 1987. 'Le paysage urbain de Metz', in Le Pay sage Monumental de la France autour de VAn Mil, ed. X. Barral I Altet (Paris), pp.510-16. 1992. 'Metz', in Der Kreis Merzig-Wedern und die Mosel zwischen Nennig und Metz (Fuhrer zu Arch. Denkmdlern in Deutschland 24), ed. J. Lichardus & A. Miron (Stuttgart), pp.212-21. Wallace-Hadrill, J.M., 1957. 'Frankish Gaul', in France, Government and Society, ed. idem. & J. McManners (London), repr. in idem. 1962, pp. 1-24. 1962. The Long-Haired Kings (London). 1975. Early Medieval History (London). Weber, M., 1982. 'Max Weber: selections from Economy and Society, vols. 1 and 2, and General Economic History', in Classes, Power and Conflict. Classical and Contemporary Debates, ed. A. Giddens & D. Held (London). Weidemann, K., 1970. 'Zur Topographie von Metz in der Romerzeit und im friihen Mittelalter', Ber.RGZM 17, pp.147-71. Weidemann, M., 1987. 'Adelsfamilien im Chlotharreich. Verwandtschaftliche Beziehungen der frankischen Aristokratie im 1. Drittel des 7. Jahrhunderts', Francia 15, pp.829-51. Wemple, S.F., 1981. Women in Frankish Society: Marriage and the Cloister, 500-900 (Philadelphia). 299

BIBLIOGRAPHY Werner, J., 1950. 'Zur Entstehung der Reihengraberzivilisation', Archaeologia Geographica 1 (April), pp.23-32. Werner, K.F., 1979. 'Noble families in the kingdom of Charlemagne. A prosopographical study of the relationship between king and nobility in the early middle ages', in Reuter (ed. & trans.) 1979, pp.137-202. Wickham, C , 1981. Early Medieval Italy. Central Authority and Local Society, 400-1000 (London). 1984. 'The other transition: from the ancient world to feudalism', Past and Present 103, pp.3-36. 1992. 'Problems of comparing rural societies in early medieval western Europe', TRHS, 6th ser., 2, pp.221-46. Wightman, E.M., 1977. 'The towns of Gaul with special reference to the north-east', in Barley (ed.) 1977, pp.303-13. 1985. Gallia Belgica (London). Willaume, M., 1991. 'Bilan et orientations de la recherche archeologique', BSL, pp.7-16. Wolfram, G., 1892. 'Die alteste Kathedrale zu Metz', JGLGA 4, pp.240-50. Wolfram, G., & Gley, W., 1931. Elsass-lothringischer Atlas (Frankfurt-am-Main). Wood, I.N., 1977. 'Kings, kingdoms and consent', in Sawyer & Wood (eds.) 1977, pp.6-29. 1983. 'The ecclesiastical politics of Merovingian Clermont', in Ideal and Reality in Frankish and Anglo-Saxon Society ed. C.P. Wormald, R.J.H. Collins & D. Bullough (Oxford), pp.34-57. 1986. 'Disputes in fifth- and sixth-century Gaul: some problems', in Davies and Fouracre (eds.) 1986, pp.7-22. 1988. 'Forgery in Merovingian hagiography', in MGH Schriften 33: Fdlschungen im Mittelalter. Internationaler Kongress der M.G.H., Miinchen 16-19 Sept., 1986 (V: Fingierte Briefe Frb'migkeit und Falschung Realienfdlschung) (Hannover), pp.369-84. 1990. 'Administration, law and culture in Merovingian Gaul', in The Uses of Literacy in Early Medieval Europe, ed. R. McKitterick (Cambridge), pp.74-8. 1992a. 'Continuity or calamity? The constraints of literary models', in Drinkwater & Elton (eds.) 1992, pp.9-18. 1992b. 'The Merovingian kingdoms, 481-751', Medieval History 2, no.l, pp.14-27. 1993. 'The secret histories of Gregory of Tours', Revue Beige de Philologie et d'Histoire 71, fasc. 2, pp.253-70. 1994. The Merovingian Kingdoms, 450-751 (London). Wormald, C.P., 1977. 'Lex scripta and verbum regis: Legislation and Germanic kingship from Euric to Canute', in Sawyer & Wood (eds.) 1977, pp.105-38. 1984. Bede and the Conversion of England: The Charter Evidence, Jarrow Lecture 1984. Young, B.K., 1975. Merovingian Funeral Rites and the Evolution of Christianity: A Study in the Historical Interpretation of Archaeological Material, Univ. of Pennsylvania PhD thesis (Ann Arbor). 1977. 'Paganisme, christianisme et rites funeraires merovingiens', Arch. Med. 7, pp.5-81. 1984. Quatre Cimetieres Merovingiens de VEst de la France: Lavoye, Dieue-sur-Meuse, Mezieres-Manchester, Mazerny, BAR(S) 208 (Oxford). 1986a. 'Quelques reflexions sur les sepultures privilegies, leur contexte et leur evolution dans la Gaule de l'Est', in Duval & Picard (eds.) 1986, pp.69-88. 1986b. 'Exemple aristocratique et mode funeraire dans la Gaule merovingienne', Annales ESC 43, pp.379-407.

300

Index

French place-names from the study area are followed by their departmental number in brackets: 54 = Meurthe-et-Moselle; 55 = Meuse; 57 = Moselle; 67 = Bas Rhin. Abbo, bishop of Troyes 16 Abbo II, bishop of Metz 16 accolae 54, 57-8 Adalchard, count 51, 53 Adalramn, count 53, 53 n. 15 adoption 73-4 Aega, aristocrat 46 Aegidius, general 8 Aetius, general 8, 251, 252 age 72-3, 162-3, 254-7 Agiulf, bishop of Metz 15, 29, 32, 38 Alamannic Wars 203, 222, 249 Albestroff (57) 186 Altheim (Saar-Pfalz-Kreis), cemetery 77, 94-5, 266 Altrippe (57) 181 n. 3 Amalbert, slave 58 Amelecourt (57) 183 n. 5 Ammianus Marcellinus 230 Amneville (57) 186-7, 253 ancillae 57-9 Andoald, minter 209 angaria 54 Angevillers (57), cemetery 106 Ansegisel, aristocrat 14, 46, 263, 263 n. 1 Ansoald, minter 209 antmstiones 27, 37, 254, 259 Aptatus, bishop of Metz 16, 32 Aravatius, bishop of Tongres see Servatius Arbogast, count in Trier 7, 34 Arcadius, emperor 227 archaeomagnetism 222 archives: episcopal 47, 265, 265 n. 2; municipal (gesta munidpalia) 47, 265, 270: royal 13, 47, 265 n. 2 Aregundis, queen 37 Aries 8, 31, 250 Arnoald, bishop of Metz 15 Arnulf, St, bishop of Metz 15-16, 46, 73, 217, 230, 263, 263 n. 1, 264, 269 Arnulfings, dynasty 13-14, 16-17, 51, 195, 211, 275 Ars-sur-Moselle (57) 181 n. 3, 214 artefacts: association with age 83-6, 96-7, 98, 99100, 106, 113-15, 119-20, 123-5, 134-6, 145, 147, 154-6, 162-3; association with sex 79-83,

95, 96-8, 100, 104, 112-13, 115, 119, 121-3, 132-4, 144-5, 147, 152-5, 156, 162 Attalus 45 Attila, king of the Huns 8, 228, 230, 251 Auctor, bishop of Metz 230 Audo, tax-collector 30, 37 Audoin (Hodo, Otdo), aristocrat 14, 50-1, 53 Audun-le-Tiche (57), cemetery 77, 81, 91, 100, 151-62, 246, 266, 267, 271, 272 Augny (57) 201, 202 Austroald, minter 209 Avitus, bishop of Clermont 12 bacaudae 7, 251 Baldoin, charter-witness 51 Baltechildis, queen 73 Basse-Ham (57), cemetery 106 Bebing (57), cemetery 106 Begga 263, 263 n. 1 Bellange (57) 11, 201-2 Belleray (55), cemetery 250 n. 2 Benedict, aristocrat 32, 51 benefidum (benefice) 51, 52, 54, 273, 275, 278 Berling (57) 183 n. 5 Bertemund, minter 210 Berthechramn, bishop of Bordeaux 37 Berthechramn, bishop of Le Mans 15 Berthefried, duke 36, 39, 74, 269 Berthelming (57), cemetery 96-7, 100, 181; villa 96, 183 Bertulf, abbot of Bobbio 264 betrothal 60-1, 254 Bettborn (57): cemetery 77, 97-8, 100, 181; villa 97, 181 n. 3 Bettviller (57), cemetery 109 Biberkirch (67) 50, 59, 181 n. 3 Blemerey (54), cemetery 108 Bliesbruck (57) 200, 202 Bliesgau 18, 99, 103 Bobo, minter 205, 208 Bobo, Soissonais 39 Boccichild, minter 208 Bockweiler (Saar-Pfalz-Kreis),villa and church 181 Bodegisil, Soissonais 31, 39

301

INDEX Bodigisil, count 36, 39 Bodilo, Frank 46, 50 Boniface, duke 32, 271 Boniface, St 66 n. 22 Boucheporn (57) 183 n. 5 Boulay (57) 181 n. 3 Bourscheid (57) 51 Bouxieres-aux-Dames (54) 182 Bouzonville (57), cemetery 98-9, 100, 113, 145, 247 Brouviller (57) 181 n. 3 Brunechildis, queen 12-13, 15, 37, 72-4, 269 Buccellin, aristocrat 70 Burgundaefarones 46, 269 burial: ethnic interpretation of 10, 246, 258; religious interpretation of 162, 208, 246-7 burial, face-down 97, 108, 161 Cambrai, count of 17 Cappel (57) 183 n. 5 Carignan 30-1 Carloman, mayor 17, 275 carradas 54, 191, 277 Catalaunian Fields, battle of (451) 8 Cautinus, bishop of Clermont 12 cemeteries, Merovingian 75-163 passim, 175, 17881, 183-4, 186, 245-8, 249, 252-3, 255-6, 2589, 264-5, 266-7, 271-2, 273; organization of 88-90, 95, 103, 116-17, 127-8, 138-9, 157-8, 267; pits on sites of 91, 99, 107-8, 160, 247 cemetery data, problems of 77-8, 79-81, 96, 98, 99-100, 103-4, 111-12, 118-19, 142-4, 151-2 Chaouilley (54), cemetery 77, 111-18, 121, 122, 125, 145, 147, 152, 156, 252, 255 Chardoin, centenarius 18, 51 Charlemagne, king/emperor 211, 238, 278 Charles Martel, mayor 14, 16-17, 46, 49, 275 Charoin, landowner 51, 54, 71 charters 46-8, 56, 59, 67, 265-6; see also Gorze cartulary, Wissembourg cartulary Chatel-St-Germain (57), cemetery 152, 152 n. 1, 181, 187-8, 188 n. 7 Chaussy-sur-Nied (57) 195 Cheminot (54) 188 Chenet, G., archaeologist 118, 142 Childebert 'the adopted', king of Austrasia 13, 50, 271 Childebert I 30 Childebert II 13, 39, 46, 50, 62-3, 69, 74, 233, 265 n. 2, 269 Childebert III 69, 273, 275 Childeric I 8 Childeric II 13-14, 50, 273 Childeric III 275 childhood 58, 63, 71, 72, 74, 254, 264 children, burial of 84-5, 86, 95, 98, 103, 106, 113-14, 115, 116, 119-20, 123, 125, 135, 145, 147, 149, 154-5, 162, 254, 257 Chilperic I 30, 38, 253, 259, 270 Chilperic II 195 Chlodulf, bishop of Metz (son of Arnulf) 16, 263 n. 1, 264 Chlothar I 30, 37, 41, 73 Chlothar II 15, 50, 74, 262-3, 269

302

Chlotild, queen 9 Chramn, son of Chlothar I 73 Chroacar, mancipium 58 Chroccus, aristocrat 51 Chrodegang, archbishop of Metz 16-17, 188, 217, 230, 276 Chrodoin, aristocrat 14, 50-1, 54, 71, 189 church freedmen 32, 55, 56 church slaves 56 churches, rural 51, 181-2, 187-8, 192, 272, 273 churchyard cemeteries 163, 273 cists (stone burial chambers) 76, 88, 89, 98, 117, 119, 208 Clemens, bishop of Metz 219, 221 Clemery (54): cemetery 181, 183; Roman villas 180 Clichy, Council of (626) 9 Clovis I 8-9, 15, 38, 252 Clovis II 13 Clovis III 16 Cologne (Koln) 8, 10 Columbanus 264, 270 comites civitatum 17, 34, 36-7 Constantine, charter-witness 51 contubernia 40, 53, 74, 255 convivia regis 28, 35 Corny-sur-Moselle (57) 183 n. 5, 184, 201 Coume (57) 182 n. 4 Cutry (54), cemetery 253 Cutting (57) 191 Dabo (57) 182 n. 4 Dagobert I 13, 15, 16, 45, 49, 50, 269, 271 Dagobert II 13, 273 Dagobert III 49 Daspich (57) see Florange Delme (57) 200 demesne (dominica) 188, 191, 192, 193, 195-6 Desiderius, bishop of Cahors 16, 264 Deuteria, queen 15, 37-8 Dieue-sur-Meuse (55), cemetery 142-50, 151-7 passim, 163, 187, 197, 250 n. 2, 282 Dieulouard (54), cemetery 106, 206; see also Scarponne Dieuze (57) 205, 208 divorce 60, 66, 68 documents, use of 47-8, 265-6, 271 Dodo 194 Dolving (57) see St-Ulrich Domnolen, minter 211 Dondigny 195 Donjeux (57) 182 n. 4 Donnelay (57) 181 n. 3 dos 64, 67 double burials 90-1, 109, 158 dower 47-8, 54, 61, 67, 68 Drogo, mayor (son of Pippin '11') 16 Dructimund, charter-witness 51 Ebroin, mayor 49 economics: sixth-century 259; seventh-century 267-8 Edict of Paris 45, 48, 50, 262-3 Edicts of Childebert II 27, 31, 37, 56, 60-1, 63

INDEX Egidius, bishop of Reims 12, 265 n. 2 Eichel, river (= Aquila) 18, 189, 190, 191 Einville-au-Jard (54) 50, 51, 53 n. 15, 190 Eleutherius, duke (= Leuthar?) 61, 233, 271 Eligius, bishop of Noyon-Tournai 16, 264 Ennery (57), cemetery 78-94, 95-163 passim, 181, 198, 247, 255, 256, 264 Epistulae Austrasiacae 13, 47 Eply (54), cemetery 181 Eply-Raucourt (54), Merovingian settlement 181, 183, 186-7, 197, 198, 271 Erchinoald, mayor 46 Erfweiler (Saar-Pfalz-Kreis), villa 175, 180 Erlafrid 71 Ermbert, aristocrat 51, 54, 70 ethnicity 9-12, 26-32, 49 Evantius, Arlesien 31 Fainulf, minter 207, 211 Falck (57) 182 n. 4 Fati, minter 207, 209 Felix, bishop of Metz 32 fictive kinship 73-4 fines 190, 192, 194, 197, 198, 270, 277 Firminus, Arvernian 31 Flodoard of Reims 265 Florange (57) 201, 202, 203; cemetery 75 n. 1, 106, 201, 201 n. 3, 250 n. 2 Franks 7-8, 26-32, 252, 258; settlement of 8-12, 252 Fraquelfing (57) 183 n. 5 Fredegar 12-14, 45-6, 269 Fredegundis, queen 30, 37, 74 Fresnes-en-Saulnois (57) 182 n. 4 Frouard (54) 187, 267, 273, 277 Frumoald, charter-witness 51 Fulculin, minter 209 funerary feasts 247, 253 Furdenheim (67) 184 Gaila, mancipium 58 Garoald, minter 209, 210 Gaugeric, St 31, 56 Gebehart 71 Gebold 71 Geretrude-Gaila, aristocrat 70-1 Germanus, minter 210 Gersheim (Saar-Pfalz-Kreis) 99 Gisloald, minter 209 Glodessindis, St 69 n. 24 Godo, bishop of Metz 16 godparenthood 73-4 Goeric-Abbo, bishop of Metz (Abbo I) 16, 264 Gogo, nutritor of Childebert II 17, 29 Gondrexange (57), cemetery 109 Gorze (57) 214; abbey 52, 54, 193, 211, 212, 276; cartulary 52, 54-5, 57, 71, 189, 190-1, 192-5, 209, 211, 218, 276 gradualists 279 grafio 17, 35 Grand Domaine 188, 190, 191, 196, 271, 278, 281 Gratian, emperor 227 grave, arrangement of 86-8

grave-goods 76-7, 79, 96-7, 98, 103, 136, 150, 152-3, 245-8, 250, 264-5, 267; definition of 79; positioning of 86-8, 116, 125-7, 136, 138; see also artefacts grave-structures 88, 159-60 grave-types, definition of 83, 86, 89, 111 graves, size 148, 156-7 Gregory, bishop of Tours 7, 9, 12, 15, 17, 29-31, 36-9, 46, 50, 69 n. 24, 73, 195, 195 n. 11, 217, 228, 230, 233, 239, 253, 254, 259, 265 n. 2, 268 Gregory the Great, Pope 15 Gremecey (57), villa 181 Grimoald, mayor 13-14, 16, 50, 273 Grippo, Frank 31 Giidingen (Kr. Saarbriicken), cemetery 99-100, 119, 252, 256 Gueblange (57) 191 Guentrange (57), deserted village 212 Guerstling (57) 183 n. 5 Guillaume, J., excavator of Dieue 142-4 Gundoin, aristocrat 14, 50, 54, 70-1 Gundoinings, family 51-2, 271, 274, 275 Gundulf, bishop of Metz 15, 29 Gundulf, duke 15, 29, 73, 264 Gunsolinus, bishop of Metz 29 Guntramn, king of Burgundy 74 Guntramn Boso, duke 36, 39, 254, 269 Hadonville (55) 193 Hanonville (57) 194 hariraida 54, 74 Haroin see Charoin Hayange (57), cemetery 100-2, 104, 266 headless burials 159, 161 Heriwin see Charoin Hettange (57) 200 History of St-Arnulf, Metz 231, 261 hoba 188, 191, 192 Hoff (57) 183 n. 5 Homecourt (54), cemetery 106, 253 homines 56, 57; see also church freedmen, king's freedmen Honorius, emperor 7, 203, 205 horses, burial of 109 Huns 8, 228, 230-1, 251 Hunsriick 5 Hydatius, bishop 228, 230 Imling (57) 253 immobilized burials 161 immunity 45, 49 infanticide 71 n. 25 ingenui 27-8, 256, 259 ingenui in obsequio 45, 53-4, 56, 74 Ingundis, queen 37 inheritance 60, 65, 67, 68, 74, 269 Isidore of Seville 39, 271 Jallaucourt (57) 183 n. 5 Janot, A., excavator of Dieue 142-3 jornales 54-5, 190-1, 277 Jouy-aux-Arches (57) 214, 249 Jublains 239

303

INDEX Justinian I, emperor 211 Juvelize (57), cemetery 106 Kadual, scribe 190 king's freedmen 32, 55, 56 king's slaves (fiscal slaves) 56 kinship 60-1 Kirch-Berg (67) 51 Kirschnaumen (57), cemetery 252 Koenigsmacker (57), cemetery 107-9, 247, 253 Kriterienbundels 167, 173-4 labour services 54-5 Labry (57) 183, 194 laeti 41-2, 57, 258 Landoald, minter 209 Lantfrid 71 latrones 31 Lavoye (55), cemetery 118-42, 145, 147, 148-9, 152, 153, 156-7, 163, 184, 246, 248, 252, 253, 255-7, 258, 259, 264, 266, 267, 282 Le Donon, mountain 5 Le Heraple (57), castrum 202-3, 206, 213, 250 n. 2 Le Mans 47 Leuci, tribe 5 leudes 38, 46 Leudesius 46 Leuthar, duke of the Alamans 14, 233 Lex Ribvaria see Ripuarian Law Limoges 239 Liutprand's Laws 63 London 238 Louis I 'the Pious', emperor 238 Mackwiller (67), vicus 200, 202 Magnentius, emperor 203, 227 Magnovald, aristocrat 39, 50 Magnus Maximus, emperor 7, 250 Mairy (57) 196 n. 12 majority, age of 72, 254, 255 mallus 196, 207 mancipia 57-9, 68, 71-2, 189, 191, 191 n. 10, 195, 197, 277 mansi ingenuiles 54, 193 mansus 188-9, 191, 192, 193, 194, 197, 209, 210, 211, 270, 277, 278 Mapinius, bishop of Reims 197 marca 190-1, 192, 197, 198, 270, 276-7 Marchfield 37 Marculf's formulary 48-9, 52, 55, 57, 58, 59, 68, 194, 275 marriage 60-1, 67, 72, 254-6 marriage, royal 37-8, 66 Mars-la-Tour (57) 195 Marsal (57); cemetery 108, 209; vicus 191, 193, 200, 208-10, 211, 213 Martin, Austrasian aristocrat 13 matricularii 236, 241 Meaux, count of 17 Mediomatrici, tribe 5, 176, 214 Merovech, son of Theuderic II 74 Metrich (57) 196 n. 12, 253 Metz, civitas of 4-7, 17-18

304

Metz (57), town of 7-8, 12-17, 18, 39, 49, 107, 188, 196, 202, 213, 214-41, 249-51, 254, 260-1, 265, 267, 270, 271, 276, 277; a duke in 17, 49, 188; early Roman 214-17, 227; name of 214; sack of (451) 8, 228-31; walls of 219, 230 cemeteries in: Brasserie Messine 233 n. 16, 236 n. 17; Jardin Botanique 231, 236 n. 17; Les Halles 233 n. 16, 236 n. 17; Lunette d'Argon 236 n. 17; Porte de Thionville 233 n. 16; Porte des Allemands 236 n. 17; Rue de FEveche 236 n. 17; Rue des Allemands 233 n. 16; Rue Paul Diacre 236 n. 17, 236 n. 19 churches in 228, 233, 236; Holy Apostles see St-Arnulf; St-Amantius 233; St-Arnulf (St-Arnoul) (= Holy Apostles) 16, 46-7, 52, 188, 195, 228, 231, 236 n. 17, 239, 249, 261, fair of 241; St-Felix 228, 249; Ste-Glossinde, nunnery 52; St-John the Baptist 228, 231; St-Julian 233; St-Livier 224, 227; St-Martin 241; St-Pierre-aux-Arenes 221, 228, 231, 249; St-Pierre-aux-Nonnains, nunnery 14, 52, 61, 221-2, 223, 233, 236 n. 17, 236 n. 19, 238, 267, Roman public buildings 221-3, 227, 249; St-Sigolena 233; St-Stephen (Metz Cathedral) 46-7, 69, 193, 211, 228, 230-1, 236; 239, 251, 251 n. 3; St-Victor 219, 233; see also St-Remy-de-Scy coin finds in 218-19, 227, 238; inscriptions 228, 228 n. 12, 231, 231 n. 15, 236, 236 n. 18 medieval quarters of: in Sinodochio 236; in viciniolo 236; infra episcopo 236; portus 240; Portus Saliae 240; suburbium 236 urban archaeology in, problems 217-18, 222, 224, 227 urban archaeology, sites: Centre St-Jacques 214; Chantier de la Visitation 224, 224 n. 9; Ecole des Arts Appliques 236 n. 19; En Nexirue 236 n. 19; Espace Serpenoise 228, 231, 238; Great Amphitheatre 214, 219-21, 231, 238, 251; Grenier de Chevremont, 224 n. 9, 236, 238, 238 n. 21; Hauts-de-Ste-Croix 214, 224, 224 n. 9, 227 n. 10; H6pital-St-Nicholas 227, 227 n. 10, 228, 231, 238; Hotel de Police/Moulin des Thermes 224; La Maison Quarree 233; Parking Avenue Ney 223-4; Parking souterrain de l'Esplanade 224, 238 n. 21; Place de la Republique 236 n. 19; Pont de la Prefecture 240; Pont-St-Georges 233, 240; Pontiffroy 224, 227; Rome-Salle 236-8, 238 n. 20; Rue Boucherie-St-Georges 276-7; Rue Chanoine-Collin 224, 224 n. 9, 236, 238, 238 n. 21; Rue des Benedictins 227; Rue des Clercs 224, 231, 236; Rue des Trinitaires 224 n. 9; Rue St-Eucaire 227; Rue Taison 233, 236; small amphitheatre 214, 233; Square Gallieni 236 n. 19; Thermes du Carmel 224, 224 n. 9, 227 Meunier, Dr, excavator of Lavoye 118-19, 122-4, 134 Meuse, river 192 military service 28, 31-2, 37, 41-2, 49 Mittelbach, stream 189, 191

INDEX Montenach (57), cemetery 252 Morley (55) 247 Morsbach (57) 183 Mosellan, dialect 12 Moselle, river 5, 8, 176, 178, 214, 228, 233, 240 Mousson (54) 181, 188, 272 Mouton, M., excavator of Lavoye 118 Moyenvic (57), vicus 209, 210; St-Pient, monastery 210 Moyeuvre-Grande (57), cemetery 102-3, 271 Muldulnus, minter (= Madelinus) 209, 211 mundium 59, 62-3, 256-7 name-giving 70-1, 276 Nennig, villa 180 Nibelungenlied 230 Nicetius, bishop of Trier 17 nobilis, use of word 38-9, 45-6, 271 Noddilo, aristocrat 56 Nomeny (54), cemeteries 180, 183, 200, 205; medieval settlements 180, 184, 200, 277; Roman vicus 180, 200, 202, 205 Norroy-le-Sec (54) 195 Noveant (57) 200 Obolenus 69 old age 72-3, 257 old people, burial of 83-5, 109, 115, 123, 125, 135-6, 156, 162, 163 Oriocourt (57) 182 n. 4 Orleans, Council of (511) 9 Otto see Audoin Otto, aristocrat (son of Gundoin) 51, 70-1 Otto, palatine official 14 Ottonville (57) 182 n. 4 Pactus Legis Salicae 9, 26-9, 35, 40-4, 55-6, 60-5, 67-8, 72-4, 252, 255, 257, 258, 263, 279 Pactus pro tenore pads 36 Pagus Aquilensis 18 Pagus Mettensis 18 Pagus Mosellensis 18 Pagus Salinensis 5, 18, 53, 191, 210-11, 213, 240, 271; see also Saulnois Pagus Sarroensis 18, 53, 205 Pagus Scarponensis 18, 194, 207 Pagus Wabrensis 7, 193, 195 see also Woevre Pange (57) 183 n. 5 Papolus, bishop of Metz 15, 32 Parisot, excavator of Chaouilley 111, 115 Parthenius, tax-collector 30, 36 patronage 33-4, 38, 49-50, 53 Paul, count in Angers 34 Paul the Deacon 15, 37, 206, 208, 217, 219-21, 230, 230 n. 13, 263 n. 1 Petrus, bishop of Metz 17, 29 Pfalzerwald 5 Pippin I, king 16, 49, 61, 72, 275, 276 Pippin T , 'the Old', mayor 263 Pippin 'IF, mayor 13, 16, 46, 49, 196, 263 n. 1, 275 Pippin CIIF, 'the Short', mayor see Pippin I, king

place-names 182, 190, 194, 196; ethnic interpretation of 10-12 Plappeville (57) 181 n. 3 Plateau Lorrain 5, 260, 264 Plectrude, wife of Pippin 'IF 70 Pont-a-Mousson (54) 187 Praetextatus, bishop of Rouen 30 precaria 52, 68, 271, 273, 278 prestige graves 91-4, 99, 102, 109, 117-18, 120-1, 128-32, 139-41, 148-9, 155-6, 157, 158-9, 252, 260, 264-5, 272 Procopius 38 Prum, monastery 59, 278 puellae 57 pueri 42, 55-6, 66, 72, 255, 259 pueri regis 42, 55-6, 72 Puttigny (57) 182 n. 4 Puxieux (54) 253 Quincy (55) 193 rachimbergs 41, 257 Ragnemod, bishop of Paris 66 raptus 60, 68 Rauching, duke 36, 39, 66 Raucourt (54), cemetery 181, 183 Reginhar, mancipium 71 Reginulf, mancipium 71 Reibengraberzivilisation 3 n. 4 Reims 12 Reinheim (Saar-Pfalz-Kreis) 200 Rembercourt (54) 194 Remilly (57) 181 n. 3 Renatus Profuturus Frigeridus, lost History of 7 'Revolt of the Dukes' (587) 39 Rigunth, princess 30 Ripuarian Law (Lex Ribvaria) 9, 31-2, 45, 47-9, 53-7, 58, 61, 66-8, 72-4, 195, 257, 262-3, 265, 266, 270, 279 Robert, bishop of Metz 219 Romans, ethnic group in Merovingian Gaul 2632, 35, 55-6, 258, 263 Romaric, abbot 13, 264 Rothari's Edict 63 Rouen, count of 17 Royaumeix/Menil-la-Tour (54), cemetery 162 Saar, river: see Sarre Saarbriicken castrum 203-4, 213, 249; Halberg cemetery 203-4; Malstatt-Burbach cemetery 203-4 sacerbaro 35 St-Arnual (Saarbrucken) 203-4 St-Avoid (57), monastery 17, 52, 188 St-Baussant (57) 55 St-Jean-de-Bassel (57) 183 n. 5 St-Julien-les-Gorze (57) 55 St-Marcel (54), vicus 202 Ste-Marie-aux-Chenes 182 n. 4 St-Mihiel (55): castrum 14, 16; monastery 51-2, 209

St-Nabor, monastery see St-Avoid St-Paul, monastery see St-Avoid St-Remy-de-Scy (57) 239

305

INDEX Ste-Ruffine (57) 184, 201 St-Ulrich (57), cemetery 106 villa 106, 176 Salic law see Pactus Legis Salicae salt making 5, 209, 210-11 Salvian 8, 3 3 ^ , 250-1 Samuel, aristocrat 51 Sanon, river 50, 189 sarcophagi 103, 157, 159, 162, 188 n. 7, 208, 210, 246,267-8, 271 Sarraltroff (57) 182, 183 n.5 Sarre (Saar), river 5, 11, 14, 50, 96, 97, 98, 176, 186 Sarre-Union (57) 186 Sarrebourg (57) 199, 204-5, 206, 208, 213, 240 Sarreguemines (57) 186, 190, 212 satellites 45, 5 3 ^ , 56, 74 Saulnois 5, 268 scarae 49 Scarponne (54): castrum 199, 203, 205-7, 208, 211, 213, 230; late Roman cemetery 76, 206, 250 n. 2 Schwarzenacker (Saar-Pfalz-Kreis), vicus 202 Seille, river 5, 14, 51, 175, 176, 178, 180, 183, 189, 213, 214, 236, 240 Servatius, bishop of Tongres (= Aravatius) 230 servus 57-9; see also slavery settlement data, problems of 175-6 Severus II, emperor 222 Sevirus 194 Sigibald, bishop of Metz 16-17, 188 Sigibert I 12, 30, 269, 270 Sigibert II 269 Sigibert III 13, 16, 46, 273 Sigibert of Gembloux 13 Sigivald 73-4 Sigoin 71 Simmer, A., excavator of Audun 151-2, 156, 158, 160, 162, 246 skeletal data, ethnic interpretation of 9-10; problems of 77, 83-4 slavery 42-5, 56-60, 258, 270, 274, 277 Speyer 7 Strasbourg 7 Sulpicius Alexander, lost History of 7 Syagrius, count of Soissons ('Rex Romanorum')

Tarquimpol (57): castrum 199, 207-8, 213, 230, 267; cemetery 109, 208 taxation 28-31, 36-7, 41, 48-9, 253, 263 Theodemund, 'leader of the citizens' 29 Theodosius I, emperor 227 Theodosius II, emperor 183 Theotchar, duke 51, 71 Theotchar, son of Duke Theotchar 71 Theudebald I 12 Theudebert I 15, 30, 37, 74 Theudebert II 13, 195, 233, 269 Theudemund, minter 209, 210 Theuderic I 8-9, 30 Theuderic II 13, 15, 74, 195, 233 Theuderic III 233 Theuderic IV 275

306

Theudo, charter-witness 51, 71 Theudoald, mayor 49 Theudulf, minter 209 Thiethar, mancipium 59 Thionville (57) 200, 211-12, 239 Thirty Years War 182 Thuringia 30, 43, 46 Tieffenbach (67) 190 Totto, minter 208, 209 Toul, civitas of 5-7, 207 Toul (54), town of 260; St-Evre, cemetery under 122; St-Stephen (Toul cathedral) 210; battle of (612) 49 Tournai 66; Franks of 30 Tours 238-9; St-Martin 31, 251 n. 3 towns see urbanism Trasoald, minter 211 Trasulf, minter 210 Tribocci, tribe 5 tributarius 28, 31, 40-1, 55 Trier, region of 5, 7, 17, 69, 180, 251, 260 Trier, town of 7, 10, 30, 69, 209, 222, 240, 249; Aula Palatina 222; St-Leodegar 209; St-Maximin 209 trustis regis 28, 35, 37 unfurnished graves 86 urbanism: definition of 173-4, problems of 1678; see also Metz, town of Ursio, duke 36, 39 Valens, emperor 227, 249 Valentinian I, emperor 222-3, 227 Valentinian II, emperor 227 Valentinian III, emperor 8 Vannecourt (57) 182 n. 4 Varangeville (54) 193, 247 Varize (57) 106, 181 n. 3 vasalles 57 Vaxy (57) 181 n. 3, 182 n. 4 Venantius Fortunatus 12, 15, 17, 65, 217, 233, 255 Verdun, civitas of 7, 205-6 Verdun (55), town of 260 Vic-sur-Seille (57), vicus 193, 199, 209, 210-11, 213; St-Marianus 210-11 Vilithuta 65, 72, 257 villas: Merovingian 189-98, 253, 259-60, 263, 270, 276-7; Roman 176-8, 180-2, 183 Villicus, bishop of Metz 197 Vinchy, battle of (717) 14 Vita Glodessindis 69, 69 n. 24, 217 Vita Liudgeri 60 n. 18, 256 n. 7 Vita Waldradae 46, 46 n. 10, 61, 69, 217, 271 Voinot, Dr J., excavator of Chaouilley 77, 11113, 115, 118-19 Vosges 11, 15, 50-1, 52-3, 76, 184, 186, 189, 195, 196 Vouille, battle of (507) 9 Waddo, count 36 Waldhambach (67) 11, 51, 59, 189-90, 191, 191 n. 10 Waldrada, saint 61, 69, 69 n. 24, 233

INDEX Walfchramn, minter (= Luolframn) 209, 210, 211 Walsheim (Saar-Pfalz-Kreis), cemetery 103-4 Waregisel, minter 207 Warmar, Frank 31 Weroald, aristocrat 11, 14, 50-1, 54, 71, 189-90 widowhood 60, 62, 64-5, 73, 257 Wiesviller (57), cemetery 181 Willichar, count 18, 51, 205 Wintrio, duke 69 n. 24 Wissembourg: abbey 11, 14, 50-1, 54, 59, 70, 189, 193, 209, 271, 273, 275, 278; cartulary of 14, 32, 47, 50-3, 54-5, 57-9, 68, 69-70, 72, 182, 186, 189-95, 197-8, 200, 209, 274, 276, 277, 272 Wittersheim (Saar-Pfalz-Kreis), cemetery 104-6

Woevre 7, 39, 195 Woippy (57) 182 Wolfgait, mancipium 59 Wolfgund, aristocrat 70-1 Wolfrid 71 Worms 7, 202, 203 Wuisse (57) 182, 182 n. 4 Wulfoald, count 14, 16, 51-2, 275 Wulfoald, mayor 13-14, 71 Xonville 55 York 238 Yutz (57), vicus 184, 200, 201, 202; Palace 212; Haute-Yutz kilns 222, 249

307