4529

more information - www.cambridge.
org/9781107025295
Documentary Culture and the Laity in the
Early Middle Ages
Many more documents survive from the early Middle Ages than from
the Roman Empire. Although ecclesiastical archives may account for the
dramatic increase in the number of surviving documents, this new inves-
tigation reveals the scale and spread of documentary culture beyond the
Church. The contributors explore the nature of the surviving docu-
mentation without preconceptions to show that we cannot infer chang-
ing documentary practices from patterns of survival. Throughout Late
Antiquity and the early Middle Ages – from North Africa, Egypt, Italy,
Francia and Spain to Anglo-Saxon England – people at all social levels,
whether laity or clergy, landowners or tenants, farmers or royal func-
tionaries, needed, used and kept documents. The story of documentary
culture in the early medieval world emerges not as one of its capture
by the Church, but rather of a response adopted by those who needed
documents, as they reacted to a changing legal, social and institutional
landscape.
w a r r e n c . b r ow n is Professor of History at the California Institute

of Technology.
m a r i o s c o s t a m b e y s is Senior Lecturer in History at the University
of Liverpool.
m a t t h e w i n n e s is Professor of History at Birkbeck, University of
London.
a d a m j . k o s t o is Professor of History at Columbia University.
Documentary Culture and the
Laity in the Early Middle Ages
Edited by
Warren C. Brown, Marios Costambeys,
Matthew Innes and Adam J. Kosto
c a m b r i d g e u n i ve r s i t y p r e s s
Cambridge, New York, Melbourne, Madrid, Cape Town,
Singapore, São Paulo, Delhi, Mexico City
Cambridge University Press
The Edinburgh Building, Cambridge CB2 8RU, UK
Published in the United States of America by Cambridge University Press,
New York
www.cambridge.org
Information on this title: www.cambridge.org/9781107025295

C Cambridge University Press 2013
This publication 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 2013
Printed and bound in the United Kingdom by the MPG Books Group
A catalogue record for this publication is available from the British Library
Library of Congress Cataloguing in Publication data

Documentary culture and the laity in the early Middle Ages / edited by Warren
C. Brown, Marios Costambeys, Matthew Innes and Adam J. Kosto.
pages cm
Includes bibliographical references and index.
ISBN 978-1-107-02529-5
1. Middle Ages – Sources. 2. History – Sources. 3. Europe – History –
476–1492 – Sources. 4. Civilization, Medieval – Sources. I. Brown, Warren
C. II. Costambeys, Marios. III. Innes, Matthew. IV. Kosto, Adam J.
D113.D627 2012
302.2 2440940902 – dc23 2012017771
ISBN 978-1-107-02529-5 Hardback
Cambridge University Press has no responsibility for the persistence or

accuracy of URLs for external or third-party internet websites referred to
in this publication, and does not guarantee that any content on such
websites is, or will remain, accurate or appropriate.
Contents
List of figures and tables page vii

List of contributors viii
Acknowledgements ix
List of abbreviations xi
1 Introduction 1
2 Lay archives in the Late Antique and Byzantine East: the
implications of the documentary papyri 17
pe t e r s a r r i s
3 Public administration, private individuals and the
written word in Late Antique North Africa, c. 284–700 36
jo n a t h a n p. c o n a n t
4 Lay documents and archives in early medieval Spain
and Italy, c. 400–700 63
n i c h o l a s e ve r e t t
5 The gesta municipalia and the public validation of
documents in Frankish Europe 95
wa r r e n c . b r ow n
6 Laypeople and documents in the Frankish formula
collections 125
wa r r e n c . b r ow n
7 Archives, documents and landowners in Carolingian
Francia 152
matthew innes
8 The production and preservation of documents in
Francia: the evidence of cartularies 189
hans hummer
v
vi Contents
9 The laity, the clergy, the scribes and their archives: the
documentary record of eighth- and ninth-century Italy 231
marios costambeys
10 Sicut mos esse solet: documentary practices in Christian
Iberia, c. 700–1000 259
a d a m j. k o s t o
11 On the material culture of legal documents: charters
and their preservation in the Cluny archive, ninth to
eleventh centuries 283
matthew innes
12 Documentary practices, archives and laypeople in
central Italy, mid ninth to eleventh centuries 321
antonio sennis
13 Archives and lay documentary practice in the
Anglo-Saxon world 336
charles insley
14 Conclusion 363
Index 377
Figures and tables
Figures
5.1 BnF lat. 10756, pp. 63–4. page 110–11
6.1 BnF lat. 2123, fol. 147r, detail. 146
6.2 BnF lat. 2123, fol. 121va, detail. 148
Tables
9.1 Proportions of ‘lay’ documents in some Italian archives. 239
9.2 Proportions of documents of sale in some Italian archives. 241
vii
Contributors
b r ow n , wa r r e n c . , Professor of History, Division of the Humanities

and Social Sciences, California Institute of Technology
c o n a n t , jo n a t h a n p. , Assistant Professor, Department of History,
Brown University
c o s t a m b e y s , m a r i o s , Senior Lecturer, School of History, University
of Liverpool
e ve r e t t , n i c h o l a s , Associate Professor, Department of History,
University of Toronto
h u m m e r , h a n s , Associate Professor, Department of History, Wayne
State University
i n n e s , m a t t h e w , Professor, Department of History, Birkbeck
College, University of London
i n s l e y , c h a r l e s , Senior Lecturer, School of Arts, Languages and
Cultures, University of Manchester
k o s t o , a d a m j. , Professor, Department of History, Columbia
University
s a r r i s , pe t e r , University Senior Lecturer, Faculty of History,
University of Cambridge
s e n n i s , a n t o n i o, Lecturer, Department of History, University
College, University of London
viii
Acknowledgements
This book is less a traditional edited volume than one result of a highly
fruitful process of collaborative research. What began as a discussion over
a pint in Princeton in November 2000 grew into an ever-widening e-mail
exchange spanning eight time zones, and then a panel at Leeds (2002);
formal workshops in London (2003), Pasadena (2004), Vienna (2005),
Canterbury (2006) and Cambridge (2007); a panel at the Medieval
Academy in Vancouver (2008); and finally – for the four editors – a
remarkable series of group writing sessions, from which the introduc-
tion and conclusion emerged and the various parts were woven into a
whole. From the beginning, the ten members of the Lay Archives Work-
ing Group, as we came to be called, were all frustrated by the increas-
ingly outcomes-driven framework of much research in the humanities.
We therefore banned all discussion of a collective publication until several
years into the undertaking; our goals were, rather, undirected exploration
of our common theme and intellectual exchange among a transatlantic
network of younger scholars, most of whom did not know each other
well or at all when we began. Various members did publish work sepa-
rately as we progressed, but only when it became clear that we as a group
had something potentially important to say did we turn to work on the
present volume. This long process (punctuated by the arrival of eight
children!) lies behind what we hope will be seen as a very well-integrated
publication – not a set of papers, but a book with a collective voice and
a common argument.
So as not to succumb to groupthink, from early on we invited other
scholars to sit in on our workshops, and it is their probing questions,
collegial suggestions and encouragement of our collective approach that
we have the pleasure to acknowledge first: Rosamond McKitterick and
Christina Pössel at London; Patrick Geary, Jason Glenn and Claudia
Rapp in Pasadena; Walter Pohl, Peter Erhart and Bernhard Zeller in
Vienna; and Geoff Koziol and Barbara Rosenwein, who served as com-
menters on the Vancouver panel. Koziol went on to serve as an extraor-
dinarily incisive reader of our manuscript for Cambridge University
ix
x Acknowledgements
Press. He pressed us to clarify our thinking and ultimately helped us to

strengthen our collective argument. We also thank the other (anonymous)
reader for that press, as well as readers for what was an unsuccessful grant
application along the way.
Our meetings received institutional support from Birkbeck College,
University of London; the California Institute of Technology; the Hunt-
ington Library; the Institut für Mittelalterforschung (Vienna); the Aus-
trian Academy of Sciences; and Cambridge University – not to mention
the home institutions of the individual authors that paid for much of
the transatlantic travel. Funding from the British Academy (via an Inter-
national Network Award) and the Leverhulme Trust (via some of the
funding awarded as a Philip Leverhulme Prize) proved vital in getting
the project off the ground.
We dedicate this book to our mentors, formal and informal, who taught
us to value collaboration over competition.
Abbreviations
Series
CSEL Corpus scriptorum ecclesiasticorum
Latinorum (Vienna, 1866–)
CCSL / G Corpus Christianorum series Latina
(Turnhout, 1953–); series Graeca (Turnhout,
1977–)
MGH Monumenta Germaniae historica
AA Auctores antiquissimi
Capit. Capitularia
Const. Constitutiones et acta publica imperatorum
et regum
DD Diplomata
Epist. Epistolae
Libri mem. n.s. Libri memoriales et necrologia, nova series
LL Leges (in folio)
LL nat. Germ. Leges nationum Germanicarum
SrG [n.s.] Scriptores rerum Germanicarum [nova
series]
SRL Scriptores rerum Langobardicarum et
Italicarum
SRM Scriptores rerum Merovingicarum
MGH Formulae Formulae Merowingici et Karolini aevi, ed.
K. Zeumer, MGH Legum sectio V (Hannover,
1886)
Form. Andec. Formulae Andecavenses
Form. Arvern. Formulae Arvernenses
Marculf Formulae Marculfi
Marculf [Supp.] Marculfi formulae [Supplementum]
Form. Marc. Kar. Formulae Marculfi aevi Karolini
Form. Tur. Formulae Turonenses vulgo Sirmondicae dictae
Form. Tur. Add. Additatmenta e codicibus formularum
Turonensium
xi
xii List of abbreviations
Form. Bitur. Formulae Bituricenses

Form. Sen. Formulae Senonenses
Cart. Sen. [App.] Cartae Senonicae [Appendix]
Form. Sen. Rec. Formulae Senonenses recentiores
Form. Sal. Big. Formulae Salicae Bignonianae
Form. Sal. Merk. Formulae Salicae Merkelianae
Form. Sal. Lind. Formulae Salicae Lindenbrogianae
Form. Sang. Misc. Formulae Sangallenses miscellaneae
Coll. Sang. Collectio Sangallensis Salomonis III. tempore
conscripta
Form. Salz. Formulae Salzburgenses
Coll. Flav. Collectio Flaviniacensis
Coll. Flav. Add. Collectio Flaviniacensis [Additamenta]
Form. Vis. Formulae Visigothicae
PL Patrologiae cursus completus, series Latina,
ed. J.-P. Migne, 221 vols. (Paris, 1845–6)
PG Patrologiae cursus completus, series Graeca,
ed. J.-P. Migne, 161 vols. in 166 (Paris,
1857–66)
Charters
ChLA Chartae latinae antiquiores, ed. A. Bruckner et al.
(Olten, 1954–)
CDL Codice diplomatico longobardo, ed. L. Schiaparelli
et al., 5 vols. in 7 (Rome, 1929–2003)
Cluny Recueil des chartes de l’abbaye de Cluny, ed. A.
Bernard and A. Bruel, 6 vols. (Paris, 1876–1903)
RF Gregory of Catino, Regestum Farfense, ed. I. Giorgi
and U. Balzani, Il Regesto di Farfa, 5 vols. (Rome,
1879–1914)
Rio, Formularies A. Rio, trans., The Formularies of Angers and
Marculf: Two Merovingian Legal Handbooks
(Liverpool, 2008)
TF Die Traditionen des Hochstifts Freising, ed.
T. Bitterauf, 2 vols. (Munich, 1905–8)
Tjäder J.-O. Tjäder, Die nichtliterarischen lateinischen Papyri
Italiens aus der Zeit 445–700, 3 vols. (Lund,
1954–82) (documents cited volume in Roman,
number in Arabic [I:3]; commentary cited volume
in Arabic: page in Arabic [1:3])
List of abbreviations xiii
TP Die Traditionen des Hochstifts Passau, ed.

M. Heuwieser (Munich, 1930)
TW Traditiones Wizenburgenses: Die Urkunden des
Klosters Weissenburg: 661–864, ed. K. Glöckner and
A. Doll (Darmstadt, 1979)
Wartmann Urkundenbuch der Abtei Sanct Gallen, ed.
H. Wartmann, 6 vols. in 7 (Zürich, 1863–1931)
Law
CE Codex Euricianus, ed. A. d’Ors (Rome, 1960); Liber
iudiciorum sive Lex Visigothorum, ed. K. Zeumer, MGH
LL nat. Germ. 1:3–32
CIC Corpus iuris civilis, 16th/11th/6th edn, ed. P. Krüger
et al., 3 vols. (Berlin, 1954)
CJ Codex Iustinianus, ed. P. Krüger, CIC 2
CTh. Codex Theodosianus, ed. P. Krüger, T. Mommsen and
P. M. Meyer, 2 vols. (Berlin, 1905)
Dig. Iustiniani Digesta, ed. T. Mommsen and P. Krüger,
CIC 1
Epit. Aegid. Epitome Aegidii, ed. G. Haenel, in LRV
ER Edictus Rothari, ed. G. H. Pertz, MGH LL 4:3–90
ETh. Edictum Theodorici regis Italiae, ed. F. Bluhme, MGH
LL 5:145–68
Gaius, Inst. Gaius, Institutiones, ed. F. de Zulueta, 2 vols. (Oxford,
1946–53)
Inst. Iustiniani Institutiones, ed. P. Krüger, CIC 1
J.Edict Iustiniani XIII Edicta quae vocantur, ed. R. Schöll and
W. Kroll, CIC 3
J.Nov. Iustiniani Novellae, ed. R. Schöll and W. Kroll, CIC 3
LV Liber iudiciorum sive Lex Visigothorum, ed. K. Zeumer,
MGH LL nat. Germ. 1:33–456
LRV Lex Romana Visigothorum, ed. G. Haenel (Leipzig,
1847)
Nov. Valentinian III, Marcian, Maiorian: Novellae, ed.
T. Mommsen and P. M. Meyer, in CTh. 2.
Leo III: Leo III: Novellae, ed. C. E. Zachariae a Lingenthal
(Leipzig, 1857)
Peira Eustathios Romaios, Peira, ed. C. E. Zachariae a
Lingenthal (Leipzig, 1857)
PSent. Pauli sententiae, ed. M. Bianchi Fossati Vanzetti (Milan,
1995)
xiv List of abbreviations
Papyri
CPR Corpus papyrorum Raineri (Vienna, 1895–)
P.Bad Veröffentlichungen aus den badischen Papyrus-Sammlungen, 6
vols. (Heidelberg, 1923–38)
P.Ital. = Tjäder
P.Marini I papiri diplomatici, ed. G. Marini (Rome, 1805)
P.Oxy The Oxyrynchus Papyri (London, 1898–)
PSI Papiri greci e latini (Florence, 1912–)
Studies
Brown, ‘Documents’ W. Brown, ‘When Documents Are
Destroyed or Lost: Lay People and
Archives in the Early Middle Ages’,
EME, 11 (2002), 337–66
Classen, ‘Fortleben’ P. Classen, ‘Fortleben und Wandel
spätrömischen Urkundenwesens im
frühen Mittelalter’, in Recht und Schrift,
13–54
Costambeys, Power M. Costambeys, Power and Patronage in
Early Medieval Italy: Local Society, Italian
Politics and the Abbey of Farfa, 700–900
(Cambridge, 2007)
Davies, Acts of Giving W. Davies, Acts of Giving: Individual,
Community and Church in Tenth-Century
Spain (Oxford, 2007)
Geary, Phantoms P. Geary, Phantoms of Remembrance:
Memory and Oblivion at the End of the
First Millennium (Princeton, NJ,
1994)
Hummer, Politics H. Hummer, Politics and Power in Early
Medieval Europe: Alsace and the Frankish
Realm, 600–1000 (Cambridge,
2005)
Innes, State and Society M. Innes, State and Society in the Early
Middle Ages: The Middle Rhine Valley,
400–1000 (Cambridge, 2000)
Kosto, ‘Laymen’ A. J. Kosto, ‘Laymen, Clerics, and
Documentary Practices in the Early
Middle Ages: The Example of Catalonia’,
Speculum, 80 (2005), 44–74
List of abbreviations xv
Les cartulaires O. Guyotjeannin, L. Morelle and

M. Parisse (eds.), Les cartulaires: Actes de
la table ronde organisée par l’École nationale
des chartes et le G.D.R. 121 du C.N.R.S.
(Paris, 1993)
Les tranferts ‘Les transferts patrimoniaux en Europe
occidentale, VIIIe –Xe siècle (I): Actes de
la table ronde de Rome, 6, 7, et 8 mai
1999’, Mélanges de l’École française de
Rome: Moyen âge, 111 (1999), 487–972
McKitterick, Carolingians R. McKitterick, The Carolingians and the
Written Word (Cambridge, 1989)
Recht und Schrift P. Classen (ed.), Recht und Schrift im
Mittelalter (Sigmaringen, 1977)
Rio, Legal Practice A. Rio, Legal Practice and the Written Word
in the Early Middle Ages (Cambridge,
2009)
Rosenwein, Saint Peter B. H. Rosenwein, To Be the Neighbor of
Saint Peter: The Social Meaning of Cluny’s
Property, 909–1049 (Ithaca, NY, 1989)
Settlement of Disputes W. Davies and P. Fouracre (eds.), The
Settlement of Disputes in Early Medieval
Europe (Cambridge, 1986)
Uses of Literacy R. McKitterick (ed.), The Uses of Literacy
in Early Medieval Europe (Cambridge,
1990)
Wickham, Framing C. Wickham, Framing the Early Middle
Ages: Europe and the Mediterranean,
400–800 (Oxford, 2005)
Journals
EME Early Medieval Europe (Harlow, 1992–)
TRHS Transactions of the Royal Historical Society (London, 1872–)
Varia
Amm. Marc. Ammianus Marcellinus, Rerum gestarum libri,
ed. W. Seyfarth, 2 vols. (Leipzig, 1978)
BnF Paris, Bibliothèque nationale de France
CIL Corpus inscriptionum Latinarum, 18 vols.
(Berlin, 1862–1989)
xvi List of abbreviations
CLA Codices Latini antiquiores, ed. E. A. Lowe, 11

vols. (Oxford, 1934–71)
Niermeyer, Lexicon J. F. Niermeyer, Mediae Latinitatis lexicon minus
(Leiden, 1976)
Settimane Settimane di studio del Centro italiano di studi
sull’alto medioevo (Spoleto, 1953–)
1 Introduction
In 1928, archaeologists working along the Tunisian-Algerian border

uncovered an earthenware jar containing forty-five wooden panels. When
deciphered, the scraps were found to contain the texts of thirty-four Latin
documents from the last decade of the fifth century: tables of calcula-
tions, a record of a dowry and several documents of sale. Most of the
latter concern small plots of olive, fig and nut trees; one deals with a slave.
The properties and individuals involved appear to have been associated
with a single rural estate, the Fundus Tuletianos.1
The episcopal archive in Pisa preserves a simple list of eighty-eight
documents given by Teuspert to Ghittia and her daughters. No reason
can be given for its composition, but it can be dated to between 768
and 774. It inventories different kinds of documents, many of them
rare among the large numbers of complete documents that survive in
Italian ecclesiastical archives, but here catalogued and classified with care.
In over half of the various charters (cartae), notes (breves), judgements
(iudicati), letters (epistolae) and precepts (praecepta), neither party to the
transaction was expressly associated with the Church. Ghittia herself was
a nun, but her collection includes documents from a wide variety of
her kin and contacts in which she is not implicated; forty-four of the
documents in her list, for example, concern the layman Alahis.2
A formula collection from Gaul compiled towards the end of the sev-
enth century contains a model letter from a king to an unfortunate who
has had his house destroyed by fire. Among the things he lost were
all of his documents, particularly those that recorded his title to prop-
erty. According to the text, the man provides written testimony to the
king verifying his property claims; the king then issues this letter con-
firming his property rights. Late in the ninth century – two centuries
1 Tablettes Albertini: Actes privés de l’époque vandale (fin du Ve siècle), ed. C. Courtois et al.,
2 vols. (Paris, 1952); see below, 40.
2 ChLA XXVI 808; CDL II 295; see below, 239.
1
2 Introduction
later – a Bavarian formula collection includes the same text word-for-

word, save that the house in which the documents had been stored is
now a monastery.3
A monk at the monastery of Sahagún in León in the early twelfth
century, transcribing charters into a codex for his institution, copied a
document that listed in abbreviated form twelve sales of land concerning
a property called Melgar de Foracasas. These transactions, which did not
involve the monastery, dated from between 945 and 954. In 959 Melgar
de Foracasas was donated to the monastery by its peasant proprietors,
Iscam and Filaura, along with the document. The total value of the twelve
parcels was 133 solidi, 22 measures of wheat, a small plot and a red cow.4
For much of the modern period, a deeply embedded commonplace held

that in Europe and the Mediterranean world in the early Middle Ages
(AD 400–1000) documents were produced, used and preserved primarily
by, and according to the needs and interests of, churchmen and monks. A
corollary belief maintained that the surviving evidence for early medieval
society so completely depended on the Church for its production and
preservation that the investigation of such things as secular practices of
landowning and conflict resolution, as opposed to the strategies used
by the Church to pursue its interests, was impossible. Any effort to
look beyond the confines of ecclesiastical institutions and reconstruct
the workings of society as a whole would necessarily be indirect because
of the nature of the surviving evidence.5
Yet, each of the examples given above shows laypeople using and stor-
ing documents. Many other examples like these survive, mostly – though
not exclusively – scattered among and embedded in ecclesiastical and
monastic archives. In the last several decades, early medievalists inter-
ested in how the written word was used in their period have begun to
explore the implications of this evidence, namely that the early medieval
laity may in fact have used and relied on writing to carry on the busi-
ness of their lives to a considerable degree. Their efforts have begun
to transform dramatically our understanding of writing practices in the
period.6
3 Marculf 1.33; Form. Marc. Kar. 22. See the discussion by Brown, ‘Documents’.
4 Colección diplomática del monasterio de Sahagún (siglos IX y X), ed. J. M. Mı́nguez
Fernández (León, 1976), nos. 94, 162, 164; see below, 279–80.
5 See Kosto, ‘Laymen’, 47nn10–12; M. Clanchy, From Memory to Written Record: England,
1066–1307, 2nd edn (Oxford, 1993), 32.
6 In addition to the work of the scholars represented in this volume, see two decades of
publication on the subject by Cambridge University Press, from McKitterick, Carolin-
gians to Rio, Legal Practice. French and Italian projects centre on the École française de
Introduction 3
This transformation has not, however, taken hold in the non-specialist

literature, either in the work of early medievalists not themselves directly
concerned with writing practices or in the assumptions about the early
Middle Ages perpetuated by those working on the high and late medieval
periods. The reason, we think, has to do with scale. Each specialist work-
ing on document use among the early medieval laity has tended to focus
on the examples with which he or she is familiar. Their insights have,
therefore, been restricted to particular times and particular regions. This
may well have created or reinforced an impression in the wider scholarly
world that their examples were exceptional rather than typical. Though as
a group the specialists now have a sense that the laity used writing more
than was earlier believed, they have not yet put all of their individual
pieces together to form a broad picture of how early medieval laypeople
used documents or how their use of documents developed over time from
the end of the Roman period through the turn of the first millennium.
Nor have they been able to describe in anything like a comprehensive
fashion exactly how laypeople in the period used writing, how we can
discover this from sources whose survival in the main has been medi-
ated by ecclesiastical and monastic archives and interests, or how this
information might change our understanding of early medieval history
in general.
This book represents an effort by a group of scholars who work on
late antique and early medieval documentary culture to put their pieces
of the puzzle of lay documentary use together and to see what picture
emerges. We examine, directly and across a broad range of space and
time, the scattered but nevertheless significant evidence for the use of
documents by laypeople – that is, men and women who were not officials
or functionaries of a church or members of a religious community, in
transactions that did not involve ecclesiastical institutions – as well as
the factors determining the survival or not of lay documents. We also
explore the similar evidence for the existence of what can only be called
‘lay archives’; that is, collections or dossiers of documents kept and used
Rome: Les Transferts; F. Bougard, L. Feller and R. Le Jan (eds.), Dots et douaires dans le
haut moyen âge (Rome, 2002); F. Bougard, C. La Rocca and R. Le Jan (eds.), Sauver
son âme et se perpétuer: La transmission du patrimoine au haut moyen âge (Rome, 2005);
L. Feller and C. Wickham (eds.), Le marché de la terre au moyen âge (Rome, 2005).
The German and Dutch projects are the Sonderforschungsbereich 231: Träger, Felder,
Formen pragmatischer Schriftlichkeit im Mittelalter (Universität Münster, 1986–99),
reported on in Frühmittelalterliche Studien from 1988 on; and the Pionier Project Ver-
schriftelijking (Universiteit Utrecht), published in the series Utrecht Studies in Medieval
Literacy (Turnhout, 1999–).
4 Introduction
by laypeople.7 In other words, we address the degree to which laypeo-

ple were not just using documents, but regarding them as important
enough to store and make retrievable. Some of the evidence for these
lay practices survives more or less independently of ecclesiastical insti-
tutions. However, much of our discussion uses indirect evidence that,
for example, ecclesiastical archives took over or incorporated material
which originated outside the Church, or that ecclesiastical institutions
served as ‘safe deposits’ for the secular records of lay landowners. We
thus directly question the premise that patterns of current document
survival reflect patterns of early medieval access to and control of written
documentation, and that clerics writing charters and ecclesiastical insti-
tutions storing them did so only to serve narrow sectional interests which
were distinct from those of the wider society.8 Putting all of our pieces
together, we will show how laypeople together with clerics and monks
participated in what was a common documentary culture, and how we
can begin to integrate the experiences they recorded in their documents
into our histories of the early Middle Ages.
The laity
To begin with our main category of analysis: the laity, as opposed to
Christian clergy and monks, is not a group whose existence or charac-
teristics we can take for granted or whose unchanging distinctiveness we
can automatically assume. Lay is a relational category that can exist only
in dialogue with categories of the clerical. This relationship shifted rad-
ically in the course of the first millennium. It is only in the late second
and early third centuries, for example, that sources begin to mention
laity (laikos/laicus) at all. Laypeople begin then to be identified as such
by clerics, even while clerics admitted their own origins in the laity.
Fourth- and fifth-century councils insisted on the distinctiveness of the
priesthood, to the extent that ascetics who took a monastic vow were
classified as laymen.9 In the eighth and ninth centuries, as the Carolin-
gians tried to delineate more sharply the distinctions between lay and
clerical, monks came to join the ranks of the ‘clergy’ – the classical tri-
partite theory of social functions, so very different from earlier patristic
models, merged what had previously been two distinct orders into ‘those
7 For the issue of definition, see below, 12–15.

8 See already H. Fichtenau, Das Urkundenwesen in Österreich vom 8. bis zum frühen 13.
Jahrhundert (Vienna, 1971), 79.
9 P. Siniscalco, Laici e laicità: Un profilo storico (Rome, 1986); Siniscalco, ‘I laici nei primi
secoli’, in Il laicato nella Bibbia e nella storia (Rome, 1987), 91–105.
Introduction 5
who pray’.10 By the twelfth century, the reciprocity implicit in such the-
orizing was contested, as clerics attempted to define themselves as a new
and self-conscious power elite; it was precisely in this period that they
also attempted to define themselves as literati, the lettered class, thus
encouraging the identification of the remainder of society as unlettered,
illiterate.11
As our work will make clear, for much of our period the distinction
between lay and clerical does not help us understand how documentary
culture developed. Nevertheless, it remains necessary for our discussion.
The terms ‘lay’ and ‘clerical’, of course, form the frame of reference
within which modern discussions of early medieval writing have taken
place, and they thus perforce form the framework for our enquiry. More
important, however, is the fact that they have real, though changing, ref-
erents; our sources use them. In many times and places, people identified
by our sources as lay, or not explicitly identified as clerics and monks,
can be seen using documents in ways similar to those called clerics or
monks and for similar purposes. However, these same sources tell us that
laypeople were not the same as clergy. The interests of both were often
shaped by the differing contexts in which they lived and acted and by the
interests of those for whom they worked; these interests in turn helped
determine what documents they used and how and where they stored
them. Furthermore, there comes a point towards the end of our period
when, as a result of a particular combination of cultural and political cir-
cumstances, the distinction between clerical and lay does begin to affect
documentary culture in Europe on a large scale.
Laypeople, documents and the Church

Direct evidence that laypeople in Late Antiquity and the early Middle
Ages used documents survives in significant amounts: for example, in
10 On clerical identity in the Carolingian centuries, see, e.g., L. Coon, ‘What Is the Word
If Not Semen? Priestly Bodies in Carolingian Exegesis’, in L. Brubaker and J. Smith
(eds.), Gender in the Early Medieval World: East and West, 300–900 (Cambridge, 2004),
278–300; M. de Jong, In Samuel’s Image: Child Oblation in the Early Medieval West
(Leiden, 1996); F. Prinz, Klerus und Kreig im früheren Mittelalter (Stuttgart, 1971). On
early medieval/Carolingian ‘three orders’ theory and its mutation, see O. G. Oexle, ‘Tria
genera hominum: Zur Geschichte eines Deutungsschemas der sozialen Wirklichkeit in
Antike und Mittelalter’, in L. Fenske et al. (eds.), Institutionen, Gesellschaft und Kultur im
Mittelalter: Festschrift für Josef Fleckenstein zu seinem 65. Geburtstag (Sigmaringen, 1984),
483–99, and D. Iogna-Prat, ‘Le “baptême” du schema des trois ordres functionnels:
L’apport de l’école d’Auxerre’, Annales ESC, 41 (1986), 101–26.
11 R. I. Moore, The First European Revolution (London, 2004); H. Grundmann, ‘Litteratus–
illitteratus: Der Wandel einer Bildungsnorm vom Altertum zum Mittelalter’, Archiv für
Kulturgeschichte, 40 (1958), 1–65.
6 Introduction
the 45 per cent of documents from a sample of some 2,000 charters from
tenth-century Catalonia that record transactions involving only laymen
and laywomen; in discrete groups of lay documents from the eighth- and
ninth-century Frankish and Lombard worlds that later found their way
into monastic archives; and in collections of documents that have sur-
vived on unusual (to us) material, such as the wooden tablets from fifth-
century North Africa, or slates from seventh-century Iberia.12 Because
this evidence is scattered in time and place and often embedded in com-
plex ways in the ecclesiastical sources, its extent and importance has been
easy to overlook.
It is important because it allows us to confront pressing issues in the
history of medieval writing from a new perspective. Chief among these
is the question of institutional change: the effect of the end of Roman
rule on ways of structuring and recording landholding; the fate of late
Roman strategies of record-keeping in the successor states; the devel-
opment of post-Roman institutions and their demands; and the meth-
ods whereby ecclesiastical institutions in particular attempted to control
time and space through documents and archives. These changes in the
institutional landscape and in documentary practices prompted broader
transformations in the relationships between those who produced doc-
uments, the places in which documents were preserved and the uses to
which documents were put.
One aspect of this process of change turns on the development of the
early medieval Church and in particular of ecclesiastical archives. Specif-
ically ecclesiastical archives cannot be taken as given. It is necessary to
ask when and why churches and monasteries felt it necessary to store and
organize their documents in a manner distinct and separate from other
social actors and institutions. The balance between churches and other
loci of documentary production, preservation and archival activities is
historically contingent and tied to specific political contexts as well as
to long-term developments. There may also be a problem of historical
perspective. Did churches and monasteries ever in general come to dom-
inate the preservation and production of documents for everyone, and,
if so, when, where and why? Or, rather, have we privileged certain eccle-
siastical institutions whose impact on document preservation has been
especially durable? How did the practices of churches relate to other
forms of record-keeping, which in some areas predated the dominance
of the Church but in others piggybacked on ecclesiastical institutions?
When, where and why did ecclesiastical documentary practice become
12 See, e.g., Kosto, ‘Laymen’; Tablettes Albertini; Documentos de época visigoda escrita en
pizarra, siglos VI–VIII, ed. I. Velázquez Soriano (Turnhout, 2000); below, Chapters 3
and 4.
Introduction 7
qualitatively different from those of other social institutions and groups,

and what implications did this have for cultural and social practice?
The documentary and archival practices of churches and monasteries
were influenced by those of the laypeople with whom they were intimately
connected. As a rich seam of recent work has emphasized, the connec-
tion between churches and monasteries and the families that founded
and patronized them was a two-way street.13 The needs and interests of
patron families shaped the formation and organization of churches and
monasteries, and vice versa. Documentary and archival practice was as
crucial to the formation of both institutional and family identity as was the
movement of property. How did the inclusion of a core set of documents
concerning the family of a monastery’s founder affect the formation and
future development of that monastery’s archive? Why do so many of the
documents contained in these embedded lay archives structure transfers
of property within families? What role did women play in the ownership
and transmission of property beyond making the donations to churches
and monasteries that dominate the ecclesiastical records in which they
appear as primary actors?
Asking these questions forces us to recognize that documents them-
selves played a role in social and institutional development rather than
simply bearing witness to them. Early medieval archival documents are
not always routine and descriptive; they can themselves be narratives
subject to the same instrumental or strategic manipulation as texts more
traditionally called ‘narrative’.14 The archives in which they were stored
also reflect the differing interests and needs of the individuals, families or
institutions that created them. Documents and archives, therefore, must
themselves be treated as subjects of theoretical and historical enquiry,
and must be understood in their specific historical contexts. A proper
understanding of the ways in which the documentary record has been
filtered, by the creation of the documents themselves and by histori-
cally contingent processes of archive formation, is necessary to under-
stand the relationship of the surviving source material to the society it
records.
When seen in this full context, our material shows that laypeople were
full and conscious participants in early medieval societies that regarded
13 Classically, Rosenwein, Saint Peter; see now, e.g., A.-J. Bijsterveld, Do ut des: Gift Giving,
Memoria and Conflict Management in the Medieval Low Countries (Hilversum, 2005);
Davies, Acts of Giving; Hummer, Politics; Costambeys, Power.
14 E.g. W. Brown, ‘Charters as Weapons: On the Role Played by Early Medieval Dispute
Records in the Disputes They Record’, Journal of Medieval History, 28 (2002), 222–48;
P. Geary, ‘Monastic Memory and the Mutation of the Year Thousand’, in S. Farmer
and B. H. Rosenwein (eds.), Monks and Nuns, Saints and Outcasts (Ithaca, NY, 2000),
19–36.
8 Introduction
documents as important. As a rule, we have been unable to identify

consistent formal differences – in script, layout or formula – between
documents used by laypeople and those written by and for clerics in any
of our samples, although patterns of preservation are in places distinct.
The comparisons we draw between the patterns of document use and
storage by laypeople and ecclesiastical institutions dissolve the picture of
an impermeable divide between a literate clerical and an illiterate/oral lay
culture.15 Breaking down this divide opens up a new line of questions that
can only be answered by analysing ‘ecclesiastical’ and ‘lay’ documents
together, in the context of what was a common documentary culture.
Literacy and the use of documents

Our sources consist principally of documentary records; that is, writ-
ten texts generated primarily for pragmatic purposes.16 These might be
records of legal transactions: conveyances of property for the most part,
but also testaments, or records of dispute settlements. But they also
include a small but significant number of letters and formulaic models
for drafting letters, and accounting documents. Alternative sources for
our questions in our period scarcely exist. Scholars of the high and late
Middle Ages have a number of resources beyond documentary records
or letters for studying the use of writing among a broad swath of the
laity, such as, for example, texts written for entertainment, texts writ-
ten in the vernacular, university texts or religious texts written explic-
itly for a lay market. The very small number of such non-documentary
texts from the early Middle Ages that can help illuminate written culture
outside the Church (such as, for Carolingian Francia, the letter of the
aristocratic woman Dhuoda to her son William, or the Histories of the
equally aristocratic Nithard) reflect the activities, needs and interests of
the very highest members of the elite; moreover, they have already been
thoroughly explored.17 Documentary texts thus provide challenging but
potentially fruitful material with which to expand our picture of early
medieval documentary culture beyond the elites.
15 The ‘impermeable divide’ model has recently been critiqued from other materials, most
notably as regards intellectual and religious culture: see, e.g., J. M. H. Smith, Europe
after Rome: A New Cultural History 500–1000 (Oxford, 2005). Cf. T. Reuter, ‘Gifts and
Simony’, in E. Cohen and M. de Jong (eds.), Medieval Transformations: Texts, Power and
Gifts in Context (Leiden, 2005), 157–68, for eleventh-century change.
16 As in the title of the Münster project (above, n. 6), ‘pragmatische Schriftlichkeit’.
17 E.g. the essays collected in P. Wormald and J. L. Nelson (eds.), Lay Intellectuals in the
Carolingian World (Cambridge, 2007).
Introduction 9
As noted above, the number of lay documents that survive from the
early Middle Ages, while significant, appears very small when compared
to the tens of thousands of documents that concern the business of
churchmen and monks.18 After the twelfth century, the archival and
documentary practices of the laity become much more visible, and they
have therefore been studied directly.19 One important product of the
scholarship on this later period has been a new willingness to push past the
constraints of the concept of ‘literacy’ – defined as the ability to read and
to write – by allowing for the study of practices beyond these abilities and
therefore beyond the clerical elites most likely to possess them. Drawing
on a new historiography that has emphasized the collective contexts –
whether sociable or pragmatic – of premodern reading, these medievalists
have felt more confident in presenting the peasant listening to the reading
of a biblical text20 and the landowner who hires a scribe to redact a charter
as nevertheless ‘using literacy’.21
When this work has looked back to the early Middle Ages, it has
assumed – quite naturally given the views held until recently by many
early medievalists – that the vast majority of early medieval people relied
exclusively on oral communication.22 Both Michael Clanchy and Brian
18 See Kosto, ‘Laymen’; M. Garrison, ‘“Send More Socks”: On Mentality and the Preser-
vation Contexts of Medieval Letters’, in M. Mostert (ed.), New Approaches to Medieval
Communication (Turnhout, 1999), 49–99.
19 The path-breaking work is Clanchy, From Memory to Written Record. For case studies,
see, e.g., R. Berkhofer, Day of Reckoning: Power and Accountability in Medieval France
(Philadelphia, 2004), and, for aristocratic as opposed to ecclesiastical record-keeping:
J. Freed, The Counts of Falkenstein: Aristocratic Consciousness in Twelfth-Century Bavaria
(Philadelphia, 1983); Littere baronum: The Earliest Cartulary of the Counts of Champagne,
ed. T. Evergates (Toronto, 2003); Feudal Society in Medieval France: Documents from the
County of Champagne, ed. T. Evergates (Philadelphia, 1993); A. J. Kosto, ‘The Liber
feudorum maior of the Counts of Barcelona: The Cartulary as an Expression of Power’,
Journal of Medieval History, 27 (2001), 1–22; and K. Heidecker (ed.), Charters and the
Use of the Written Word in Medieval Society (Turnhout, 2000).
20 In any case, as Geoffrey Koziol pointed out to us in his review of an early draft of
this volume, the stereotype of the ‘illiterate’ peasant is in many ways, at least in the
later Middle Ages, an ideological or constructed trope. See S. Teuscher, ‘Textualizing
Peasant Enquiries: German Weistümer between Orality and Literacy’, in Heidecker
(ed.), Charters, 239–53; P. Biller and A. Hudson (eds.), Heresy and Literacy, 1000–1530
(Cambridge, 1994).
21 Smith, Europe after Rome, 13–50, for a synthesis. For the historiography, see L. Melve,
‘Literacy-Aurality-Orality: A Survey of Recent Research into the Orality-Literacy Com-
plex of the Latin Middle Ages’, Symbolae Osloenses, 78 (2003), 141–97; M. Clanchy,
‘Introduction’ and M. Mostert, ‘New Approaches to Medieval Communication’, in
Mostert (ed.), New Approaches, 3–14, 15–40; and work cited above, n. 6.
22 A view still defended by Michael Richter: The Formation of the Medieval West: Studies in
the Oral Culture of the Barbarians (Dublin, 1994); ‘“Quisquis scit scribere, nullum potat
abere labore”: Zur Laienschriftlichkeit im 8. Jahrhundert’, in J. Jarnut and U. Nonn
(eds.), Karl Martell in seiner Zeit (Sigmaringen, 1994), 393–404.
10 Introduction
Stock, for example, argue that the eleventh century witnessed a funda-
mental transition from orality to literacy.23 However, Rosamond McKitt-
erick has extended the study of literate practices back before the eleventh
century, documenting their extent in law and administration, property
transactions and manuscript culture in eighth- and ninth-century Fran-
cia, and describing a society ‘largely dependent on the written word for
its religion, law, government and learning’. She argues specifically for the
substantial involvement of laypeople in this culture of writing.24 McKit-
terick and others have also hypothesized that the use of writing among
the laity extended well below the level of the elites.
We recognize the dangers of reifying as ‘literacy’ a variety of activities
which may have been distinct in their medieval manifestations and which
may not fit with modern assumptions about what ‘literacy’ entails –
especially the assumption that it is tied to the ability to read and write.
We will therefore use the phrases such as ‘documentary practices’ or
‘documentary use’ rather than ‘literacy’, in order to capture the full
range of ways that early medieval people might have interacted with
documentary culture.25
Documents produced and used by early medieval laypeople survive in
a very limited number of contexts. Some texts, like Ghittia’s, exist as sin-
gle sheets of parchment; others, like those from the Fundus Tuletianos,
were written on papyrus, wood or slate. In the case of wood and slate,
they are only preserved when and where they were deposited under con-
ditions that have allowed for their physical survival and rediscovery. Still
other lay documents were copied in the early Middle Ages as formulas,
such as those concerning burned archives. Letters survive in a few letter
collections, but more often because they were likewise incorporated as
model texts into formula collections. Most stand-alone lay documents,
however, whether on papyrus or parchment, have been transmitted to us
along with surviving parts of the ecclesiastical or monastic archives into
which they were later incorporated. Furthermore, a large proportion of
our sources survives not individually, but because they were copied in the
ninth to thirteenth centuries, along with other contents of ecclesiastical
23 Clanchy, From Memory to Written Record; B. Stock, The Implications of Literacy: Written
Language and Models of Interpretation in the Eleventh and Twelfth Centuries (Princeton,
NJ, 1983).
24 See McKitterick, Carolingians (quotation at p. 2); Uses of Literacy; and, for a sense of
some of the directions of subsequent debate, Walter Pohl and Paul Herold (eds.),
Vom Nutzen des Schreibens: Soziales Gedächtnis, Herrschaft und Besitz im Mittelalter
(Vienna, 2002).
25 Cf. D. Ganz, ‘Temptabat et scribere: Vom Schreiben in der Karolingerzeit’, in R. Schieffer
(ed.), Schriftkultur und Reichsverwaltung in der Karolingerzeit (Opladen, 1996), 13–33.
Introduction 11
or monastic archives, into books (codices) called cartularies; this technol-

ogy, first used by the churches of the newly conquered and Christianized
regions of the Carolingian Empire east of the Rhine in the ninth century,
was adopted by major ecclesiastical institutions right across Europe in the
course of the tenth and eleventh centuries, and by the twelfth century
the first aristocratic landowners were likewise compiling cartularies.26
The survival of this class of text therefore directly depends on choices
made by medieval archivists (and, as we shall see, early modern
antiquarians).
Classical diplomatics would classify the sorts of documents that we
are interested in as ‘private charters’ – technically, legal documents not
issued by a public authority (emperor, king, pope).27 However, funda-
mental analytic categories such as ‘public charter’ and ‘private char-
ter’ have been to a significant degree conditioned by nineteenth-century
(and earlier) historiographical concerns with their characteristically rei-
fied notion of legimate public authority and, more to the point, the
then current demands of archival organization and editing. These cat-
egories have had – and continue to have – huge value for particular
kinds of scholarship, but there is nonetheless a danger of their form-
ing a barrier between modern historians and the evidence, and of their
importing anachronistic assumptions. We do not deny that there may be
real generic differences – evident, for example, in the categories used
in Ghittia’s inventory, or in the script and layout of surviving origi-
nals – between a document issued by a king and one prepared for an
otherwise undistinguished landowner. But we prefer to examine these dif-
ferences by considering how documents were actually used – as physical
objects and legal proofs – rather than judging their meaning uncritically
according to the categories of traditional diplomatics. And their use and
meaning includes more than their role as carriers of information. To illus-
trate the distinction: the study of charters in recent decades has taken
two overlapping but distinguishable paths. One focuses on the human
26 On cartularies, see Geary, Phantoms, 81–114; G. Declercq, ‘Originals and Cartularies:

The Organization of Archival Memory (Ninth–Eleventh Centuries)’, and L. Morelle,
‘The Metamorphosis of Three Monastic Charter Collections in the Eleventh Century
(Saint-Amand, Saint-Riquier, Montier-en-Der)’, both in Heidecker (ed.), Charters,
147–70, 171–204; B. Resl, ‘Nutzen des Abschreibens: Überlegungen zu mittelalter-
lichen Chartularen’, in Pohl and Herold (eds.), Vom Nutzen des Schreibens, 205–22; A.
J. Kosto and A. Winroth (eds.), Charters, Cartularies and Archives: The Transmission and
Preservation of Documents in the Medieval West (Toronto, 1999); Les cartulaires.
27 H. Bresslau, Handbuch der Urkundenlehre für Deutschland und Italien, 2nd edn, 2 vols.
(Leipzig, 1912–31), I, 3–4; O. Redlich, Die Privaturkunden des Mittelalters (Munich,
1911), vi; A. de Boüard, Manuel de diplomatique française et pontificale, II, L’acte privé
(Paris, 1948), 7–11; A. Giry, Manuel de diplomatique (Paris, 1894), 823.
12 Introduction
networks and interactions revealed in the charter evidence – for example,

those that determine the mechanisms of dispute settlement or patterns
of witnessing.28 This path necessarily treats documentary evidence prin-
cipally as repositories of information about these networks. In contrast,
the second path explores the way that charters and especially cartularies
reflect and even constitute processes of identity and memory formation.
Thus many early cartularies have been shown to be documents aimed
consciously at commemorating pious donors as well as their estates.29 It
is on insights of this kind in particular that we draw as we stress the his-
toricity of documents themselves as parts of the processes they purport
to describe.
Archives
The study of documentary culture and practice is inseparable from the
study of how documents are stored and transmitted to us; that is, from
the study of archives and their history.30 Archives are not unproblematic
presences; they are not sites from which theoretically minded scholars can
mine facts before subjecting them to a preferred model of interpretation.
Nor are they places where empiricists can simply lock themselves away,
safely quarantined from the distractions of anthropology, sociology or
critical theory, and find out ‘what happened’. If archives are fundamen-
tally the conduit whereby the past is transmitted to the present, then their
history and historiography determine the range of possible histories and
historiographies. They therefore deserve a place at the heart of attempts
to understand historical writing. Moreover, the actual contents of an
archive – the raw materials of history – are themselves products of histor-
ical processes that shaped what was produced, how it was transmitted,
and how it was presented to modern scholars. Archives and documents
28 Settlement of Disputes; W. Brown and P. Górecki (eds.), Conflict in Medieval Europe:

Changing Perspectives on Culture and Society (Aldershot, 2004).
29 Geary, Phantoms; Geary, ‘Entre gestion et gesta: Aux origines des cartularies’, in Les
cartularies, 13–26.
30 A. Blair and J. Milligan, ‘Introduction’ [to the special edition entitled ‘Towards a Cul-
tural History of Archives’ of], Archival Science, 7 (2007), 289–96, as well as the articles
contained in the rest of the issue; F. X. Blouin and W. G. Rosenberg (eds.), Archives,
Documentation, and Institutions of Social Memory: Essays from the Sawyer Seminar (Ann
Arbor, MI, 2006); A. Burton (ed.), Archive Stories: Facts, Fictions, and the Writing of
History (Durham, NC, 2005); A. J. Behne, ‘Geschichte aufbewahren: Zur Theorie der
Archivgeschichte und zur mittelalterlichen Archivpraxis in Deutschland und Italien’,
in P. Rück (ed.), Mabillons Spur: Zweiundzwanzig Miszellen aus dem Fachgebiet für his-
torische Hilfswissenschaften der Philipps-Universität Marburg (Marburg an der Lahn, 1992),
277–97; and below, Chapter 11.
Introduction 13
are shaped by power relations in both past and present and between past
and present, and by the institutions through which power is mediated.
If archives are first and foremost artefacts of power, then their history is
fundamentally a history of changing experiences and practices of power.
For critical theorists inspired by Jacques Derrida, this has become an arti-
cle of faith. Using an oddly early medieval etymological logic, Derrida
thought that the derivation of a name reflected the nature of the phe-
nomenon. The word ‘archive’, he argued, was derived from the Greek
archon (‘magistrate’), and so archives have at their heart an ‘archonitic
principle’: they are places where documents are subject to the physical
and the interpretative control of a ruling elite.31 By this logic, archives are
thus, on the one hand, fundamentally public institutions whose visibil-
ity is essential for them to bestow legitimacy on their holders, while, on
the other hand, they simultaneously subject their contents to processes
of classification and organization accessible only to those with specialist
knowledge and so embed the particular interests of a power elite. This
model has allowed scholars to write a story of ‘modernization’, in which
the action moves seamlessly from the city states of classical Greece to
the rising nation states of classical modernity. The nineteenth century,
unsurprisingly, stands as a turning point in this narrative, when states
both organized ‘public archives’ in the form of cultural institutions that
propagated an agreed national story, and developed more robust systems
to classify and store the information necessary to govern society, an entity
newly defined as a separate and separable subject upon which the state
could act.32
This model of the archive and its development is fundamentally bound
up with discussions of the modern state and its development. It does
not necessarily work well in late antique and early medieval contexts in
which the imperatives driving document production and storage were
diverse, in which patterns and loci of record-keeping were changing, and
in which the existence or nature of ‘states’ themselves is controversial.33
The modern expectation that archives are primarily tools with which
to classify and record official data is arguably misplaced here, even for
the most heavily bureaucratized polities; concerns about commemoration
and notions of identity are more relevant than modern assumptions about
reference literacy. For example, late antique states were indeed direct
heirs of the rich archival heritage of the ancient Near East and eastern
31 J. Derrida, Archive Fever: A Freudian Impression, trans. E. Prenowitz (Chicago, 1996),

1–3.
32 See the articles in History of the Human Sciences, 12 (1999).
33 S. Airlie, W. Pohl and H. Reimitz (eds.), Staat im frühen Mittelalter (Vienna, 2006).
14 Introduction
Mediterranean; hence, in the central bureaus of Byzantium – and the

Islamic Caliphate and its provinces – records sprawled in their own ded-
icated spaces. But the arcane lore of record-keeping appears to have
served to define the group identity of particular sections of the palace
officialdom, as John Lydus’ sixth-century tract on the magistracies of
the Byzantine state indicates. Lydus’ mandarin mindset reminds us that
archives in this context might have had limited use as a practical tool of
government.34 In the post-imperial West, for all the fundamental influ-
ence of Roman precedents on social and political practice, the scale
and character of such central bureaus altered. Kings, inspired by Roman
advisers and expectations, issued documents and laws whose content and
form demonstrated a continuous evolution from established practice.35
But the institutions responsible for producing and preserving documen-
tary material underwent significant changes: by Carolingian times, for
example, the royal writing office had been subsumed into the king’s
household chapel and was manned by clerics.36 Recent scholarship has
underlined the effort and resources expended by Carolingian kings in
making available copies of royal directives and instructions to bishops and
counts across the realm, and there are references to dedicated archives,
for example, at the palace in Aachen in Charlemagne’s last years.37 But
to expect these archives to fulfil identical functions to those described
by John Lydus, or to judge them against the benchmark of the nation
states of nineteenth- and twentieth-century Europe, is to take them out
of context.38
34 On the ancient background, see M. Brosius (ed.), Ancient Archives and Archival Tradi-
tions: Concepts of Record-Keeping in the Ancient World (Oxford, 2003). On John Lydus,
see: De magistratibus populi Romani libri tres (ed. A. C. Bandy [Philadelphia, 1983]); M.
Maas, John Lydus and the Roman Past: Antiquarianism and Politics in the Age of Justinian
(Oxford, 1992); and C. Kelly, ‘Later Roman Bureaucracy: Going Through the Files’, in
A. K. Bowman and G. Woolf (eds.), Literacy and Power in the Ancient World (Cambridge,
1990), 161–76.
35 See; classically, Classen, ‘Fortleben’, and P. Wormald, ‘Lex scripta and verbum regis:
Legislation and Germanic Kingship from Euric to Cnut’, in P. Sawyer and I. Wood
(eds.), Early Medieval Kingship (Leeds, 1977), 105–38; more recently, e.g., P. Heather,
‘Literacy and Power in the Migration Period’, in Bowman and Woolf (eds.), Literacy
and Power, 177–97; M. Innes, ‘Land, Freedom and the Making of the Early Medieval
West’, TRHS, 6th ser., 16 (2006), 39–74; and below, Chapter 4.
36 D. Ganz and W. Goffart, ‘Charters Earlier than 800 from French Collections’, Speculum,
65 (1990), 306–32; D. Ganz, ‘Bureaucratic Shorthand and Merovingian Learning’, in
P. Wormald et al. (eds.), Ideal and Reality in Frankish and Anglo-Saxon Society: Essays
Presented to J. M. Wallace-Hadrill (Oxford, 1983), 58–75; J. Fleckenstein, Die Hofkapelle
der deutschen Könige, I, Die Karolinger (Stuttgart, 1966).
37 E.g. McKitterick, Carolingians, 23–75; Schieffer (ed.), Schriftkultur.
38 C. Pössel, ‘Authors and Recipients of Carolingian Capitularies (779–829)’, in R. Cor-
radini et al. (eds.), Texts and Identities in the Early Middle Ages (Vienna, 2006), 253–74;
M. Innes, ‘Charlemagne, Justice and Written Law’, in A. Rio (ed.), Law and Custom in
Late Antiquity and the Early Middle Ages (London, 2011), 155–203.
Introduction 15
We need to understand archival practice in terms of its own logic

rather than to view record-keeping as an inherently rationalizing ten-
dency lacking any specific cultural and political context. This is true of
central record-keeping, whether at Constantinople or Aachen; it is vital
in making sense of the documentary practices of landowners that are the
primary focus of this book. Does a collection of documents kept in a
chest by an individual or a family even qualify as an archive if it lacks
organization and public accessibility? To what degree is such a collection
fundamentally any different from the church and monastic archives of
ninth-century Francia, which – though often extremely well organized
and geared for retrieval – were equally ‘private’ and primarily concerned
with property rights and privileges? Does it even make sense for us to
tie our definition of ‘archive’ to a distinction between public and private
functions? If so, when and under what circumstances?
The essays in this volume collectively engage these questions. How-
ever, they will place their understandings of the term ‘archive’ in partic-
ular contexts that will produce related but nevertheless different under-
standings of what the word means. These understandings will reflect not
only the different contexts in which document production and storage
took place but also the different needs and interests that documents and
archives satisfied in each time and place.
We aim to revise our understanding of how documentary practices devel-

oped across Latin Europe as a whole, with evidence drawn from the late
antique and early medieval Mediterranean to expand our picture of the
late Roman backdrop. Our effort has been shaped to a large degree
by imperatives imposed upon us by our evidence. We have found major
regional and chronological differences in documentary and archival prac-
tices. Some areas, notably Italy after the eighth century and Iberia in
the tenth, possess voluminous stocks of original, single-sheet parchment
documents, many offering direct evidence of lay practices. Egypt enjoys
a continuous documentary tradition from the Roman period (and ear-
lier) onwards; a similar pattern of continuity can be implied for other
regions of the East Roman world. In still other areas, such as Francia,
the early medieval evidence for lay practices has been transmitted prin-
cipally through later copies in ecclesiastical cartularies, or indirectly, as,
for example, in formularies; moreover, much of it consists of references
to lost documents (deperdita), relegated to the appendices and footnotes
of nineteenth-century editions but in many ways invaluable in widening
our perspective on document use.
We thus begin with the practices of the Roman and immediately sub-
Roman world, examined here in North Africa (the Roman provinces of
Egypt and ‘Africa’), Italy and Iberia. We then study the transformation
16 Introduction
of these practices in the eastern Mediterranean and the emergence of

what are often seen as new documentary cultures in Francia, Iberia and
Italy. We interrogate the regional differences and chronological trans-
formations underlying these contrasts. Many areas of the broader early
medieval world are left out, but their stories are very different. At present,
northern and eastern Europe appear to lack a documentary culture prior
to the eleventh century, while the impact of Islam renders moot the
entire question of lay documentary practice and its legal and institu-
tional underpinnings.39 Nevertheless, the connections between the North
African wooden panels, Ghittia’s inventory, the Frankish formulas and
the sales of Spanish peasants, tell us a great deal about the culture of
documentary practice in the early Middle Ages.
39 For northern, eastern and central Europe, see, e.g., A. Adamska, ‘The Introduction of
Writing in Central Europe’, in Mostert (ed.), New Approaches, 165–92; A. Adamska,
‘“From Memory to Written Record” in the Periphery of Medieval Latinitas: The Case
of Poland in the Eleventh and Twelfth Centuries’, and I. Hlavacek, ‘The Use of Char-
ters and Other Documents in Premyslid Bohemia’, both in Heidecker (ed.), Charters,
83–100, 133–46; L. Wolverton, Hastening Towards Prague: Power and Society in Medieval
Czech Lands (Philadelphia, 2001); A. Nedkvitne, The Social Consequences of Literacy in
Medieval Scandinavia (Turnhout, 2004). Islamic documentary papyri in former Roman
provinces (above all, Egypt), for long an overlooked or ignored source, are now prov-
ing to be an invaluable and voluminous source for the late antique and early medieval
transformation: see particularly the work of P. Sijpesteijn: ‘Landholding Patterns in
Early Islamic Egypt’, Journal of Agrarian Change, 9 (2009), 120–33; ‘The Archival
Mind in Early Islamic Egypt: Two Arabic Papyri’, in Sijpesteijn et al. (eds.), From al-
Andalus to Khurasan: Documents from the Medieval Muslim World (Leiden, 2006), 163–
87; and Shaping a Muslim State: The World of a Mid-Eighth-Century Official (Oxford,
forthcoming).
2 Lay archives in the Late Antique and
Byzantine East: the implications of the
documentary papyri
Peter Sarris
A survey of documentary practices in the eastern Mediterranean raises many

of the key themes that we will be following throughout the book. Among these
are issues of definition, such as of archive, or lay and clerical, or public and
private; issues of the institutional contexts for document production and preser-
vation, such as the shifting nature of the late Roman state and its organs, the
early institutional Church, or structures of property ownership; and issues of
documentary culture, such as the phenomena of destruction and dossierization,
the symbolic function of documents, and intentionally ephemeral documents.
The ancient archive from Babylon to Byzantium

According to the Book of Ezra, in the fifth century BC, the governor of
the Achaemenid province of Abr-Nahrain wrote a letter to the Persian
Great King, Darius I. The purpose of the letter was to check whether,
as they claimed, the empire’s Jewish subjects had formerly been granted
leave by Darius’ predecessor, Cyrus, to rebuild the Temple in Jerusalem.
The letter requested:
Now therefore, if it seem good to the king, let there be search made in the king’s
treasure house, which is there at Babylon, whether it be so, that a decree was
made of Cyrus the king to build this house of God at Jerusalem, and let the king
send his pleasure to us concerning this matter. (Ezra 5:17)
The account goes on to relate how:

Then Darius the king made a decree and search was made in the house of
the rolls, where the treasures were laid up in Babylon. And there was found at
Achmetha, in the palace that is in the province of the Medes, a roll. And therein
was a record thus written: In the first year of Cyrus the king the same Cyrus the
king made a decree concerning the house of God at Jerusalem, Let the house be
builded, the place where they offered sacrifices, and let the foundations thereof
be strongly laid. (Ezra 6:1–3)
17
18 Peter Sarris
Two points of considerable interest emerge from the passage above. First,
it alerts us to how deeply rooted techniques of archival record-keeping
were in the world of the Near Eastern Fertile Crescent. By Darius’ day
they had already been in existence for thousands of years, dating back
to the third millennium BC.1 Second, these ancient archival techniques
were not necessarily quite as efficient as some have imagined: it is worth
noting that the initial search in Babylon for what a civil lawyer would call
Cyrus’ rescript proved fruitless. Eventually, the document was found,
but it was found in the wrong place.2
Just as the Roman Empire of the fifth century AD was heir to the
‘palace culture’ of the civilizations of the ancient Near East,3 so, too,
was it heir to their archival culture, or rather, cultures. The voluminous
collections of documentary papyri from Egypt demonstrate that the late
Roman world was awash with paperwork, and that the late Roman state in
particular relied heavily on written instruments, especially written legal
instruments, for its administration and cohesion.4 The significance of
the written word was both practical and symbolic: imperial pronounce-
ments and laws issued to the provinces of the Roman Empire were, as
we might imagine, read out by court officials (typically the quaestor) and
circulated to provincial governors, but they were also displayed in public
before the eyes of the emperor’s subjects, often in the form of monu-
mental inscriptions.5 As Patrick Wormald reminded us, the association
between Romanitas and the written word as a symbol of authority was one
of which post-Roman duces et reges in the West remained keenly aware.6
Yet, as in Achaemenid Persia, no one archive, not even a central state
archive, could be regarded as entirely reliable. When, for example, in the
420s, a law commission was set up in Constantinople to codify imperial
constitutions in the form of the Codex Theodosianus, we see the com-
missioners adopting two distinct strategies in relation to the legal texts
which they needed. It appears to have been decided that, with respect
1 M. Brosius, ‘Ancient Archives and Concepts of Record-Keeping: An Introduction’, in

Brosius (ed.), Ancient Archives and Archival Traditions: Concepts of Record-Keeping in the
Ancient World (Oxford, 2003), 1–16, at 1.
2 A. Millard, ‘Aramaic Documents of the Assyrian and Achaemenid Periods’, in Brosius
(ed.), Ancient Archives, 230–40, at 237.
3 F. Dvornik, Early Christian and Byzantine Political Philosophy: Origins and Background, 2
vols. (Washington, DC, 1966).
4 See most recently F. Millar, A Greek Roman Empire (Berkeley, CA, 2006).
5 Classical Greek origins: J. K. Davies, ‘Greek Archives: From Record to Monument’,
in Brosius (ed.), Ancient Archives, 323–43. Sixth-century Byzantine practice: R. Scott,
‘Malalas and Justinian’s Codification’, in E. and M. Jeffreys and A. Moffatt (eds.),
Byzantine Papers (Canberra, 1981), 12–31.
6 P. Wormald, ‘Lex scripta and verbum regis: Legislation and Germanic Kingship from Euric
to Cnut’, in P. Sawyer and I. N. Wood (eds.), Early Medieval Kingship (Leeds, 1977),
105–38.
The Late Antique and Byzantine East 19
to laws issued after 398, the imperial archive in Constantinople could be

relied upon. For laws issued before that date, however, much greater use
was to be made of the archives of provincial governors, either because
the laws the commissioners sought were not preserved in the imperial
capital, or because they wished to check the text of laws recorded there
for authenticity.7
Such unreliability on the part even of state archives could be the result
of a number of factors. First, archives and archival documents could
be lost or destroyed by accident – the cause of numerous legal disputes
recorded in the papyri.8 Second, archives were vulnerable to deliberate
vandalism. It is instructive that during the so-called ‘Nika’ riots that
erupted in Constantinople in the year 532, the praetorium of the prae-
torian prefect of the East, along with its associated archive containing
taxation and other financial records, was purposefully destroyed by the
rampaging mob.9 Such disturbances were far from uncommon in late
Roman cities such as Constantinople, Alexandria and Antioch.10 Given
the enormous practical and symbolic importance of the written word to
the imperial system, it should come as no surprise that uprisings on the
part of the urban mob were often associated with the wanton destruction
of governmental and gubernatorial archives.
Indeed, it is worth noting that the deliberate destruction of archival
and documentary materials can also be seen to have been a feature
of periods of heightened social tensions and political conflict in both
the post-Roman West and the post-Roman East. In 579, for example,
Gregory of Tours records how the citizens of Limoges ‘rose up against
the new taxation of King Chilperic and burnt the libri discriptionum’.11
Likewise, the papyrological record of early Islamic Egypt preserves ‘an
eighth-century papyrus listing prisoners among whom some who had set
fire to a fiscal register’.12 In the opening years of the tenth century, as
the Tulunid regime in Egypt teetered on the brink of collapse before
the advancing Ikshidids, ‘Tulunid chancellery officials . . . made sure
7 A. M. Honoré, Law in the Crisis of Empire 379–455 A.D.: The Theodosian Dynasty and
Its Quaestors (Oxford, 1998), 128.
8 W. Clarysse, ‘Tomoi Synkollesimoi’, in Brosius (ed.), Ancient Archives, 344–59, at
344–5.
9 G. Greatrex, ‘The Nika Riot: A Reappraisal’, Journal of Hellenic Studies, 117 (1997),
60–87, at 75.
10 See Greatrex, ‘The Nika Riot’, for discussion and bibliography.
11 Wickham, Framing, 10, discussing Gregory of Tours, Libri historiarum 5.28 (MGH
SRM 1.1:240).
12 P. M. Sijpesteijn, ‘The Archival Mind in Early Islamic Egypt’, in Sijpesteijn et al. (eds.),
From al-Andalus to Khurasan: Documents from the Medieval Muslim World (Leiden, 2007),
163–86, at 165, referring to text no. 1.13 in F. Morelli, ‘Dalle prigioni dell’Arsinoite’,
in H. Harrauer and R. Pintaudi (eds.), Gedenkschrift Ulrike Horak (Florence, 2004),
185–95.
20 Peter Sarris
[as they fled] . . . to take their tax registers with them, since they knew
how crucial these documents were to the effective running of Egypt. But
in doing so they thereby ensured, unfortunately, their loss.’13 In short,
the ‘logocentric’ always risked inciting the ‘logoclastic’.
The typology of the ancient archive

What do we mean by ‘archive’? In English the term signifies two distinct
but related subjects. As Maria Brosius has written:
Archives are first a physical space within a public space (palace or temple complex,
public archive) or within a private building or private complex of buildings, and
second a collection of stored documents . . . Since the point of an archive was to
provide access to documents, the material must have been stored in such a way
as to allow those consulting the archive to retrieve the information according to
some practical rationale.14
It is this imposition of order on documentary materials so as to render

their contents more readily retrievable that sets or set apart the ‘archive’
properly understood from a mere storeroom or assemblage of documents.
From our earliest records we see documents being organized archivally
into dossiers, typically by theme and often according to a chronological
framework. Thus, in the surviving ‘economic’ documentation, in the
form of inscribed tablets, from the third millennium BC (our earliest
extant archival materials),
distinctions were made by subject matter . . . such as grain, wool, fruit, cattle,
metal, labour payments . . . The contents of individual receipts were eventually
transferred to weekly, monthly, and annual accounts . . . Clearly, tablets were
stored according to different types and labelled containers were marked to denote
the ‘dossiers’ they contained.15
These techniques of ‘dossierization’ discernible in what we might choose

to term the ancient Near Eastern ‘tablet cultures’ (so dubbed by virtue of
the medium on which they wrote) were continued in the papyrus docu-
mentation of the Greco-Roman period down to the fourth century AD in
the form of the tomos synkollesimos, by which papyrologists mean a series
of separate documents or sheets (kollemata) pasted together to form a sin-
gle roll (tomos) – a form of documentation used at every level of imperial
government and administration and emulated in private documentary
practice too.16 The use of the tomos synkollesimos, however, appears to
have died out in the Roman Empire some time around the middle of
13 Sijpesteijn, ‘The Archival Mind’, 164. 14Brosius, ‘Ancient Archives’, 10–12.

15 Brosius, ‘Ancient Archives’, 14–15. 16 Clarysse, ‘Tomoi Synkollesimoi’, 346.
the fourth century, a phenomenon that has baffled those papyrologists

who have sought to address the subject.17 Yet the explanation for the
disappearance of the ‘pasted roll’ is perhaps disarmingly simple. Papyrus
in antiquity was never cheap, and many of the tomoi synkollesimoi were
evidently both massive and unwieldy. A normal, shop-bought papyrus
roll ordinarily consisted of about twenty sheets, pasted together in the
factory, and would have been about 4 m in length from end to end.18 By
virtue of the system of ‘sheet numbering’ we encounter on certain of the
papyrological fragments, we know that many tomoi were over a hundred
sheets in length. We even hear of one, recording a series of libelli of the
emperor Decius, that contained no fewer than 433 sheets.19 We should
not be surprised to discover, therefore, that the disappearance of the
tomos synkollesimos from the papyrological records coincides very closely
with the more widespread dissemination of the codex.20 Certainly, in the
sixth century, we find the codex being used in imperial contexts where, in
the third century, we might have expected a pasted roll.21
Lay archives in the early Byzantine period:

problems of definition
Attempting to distinguish private lay archival materials from other
archival types encountered in the documentary record of the early Byzan-
tine period of, say, the fourth to the seventh centuries AD, is not entirely
straightforward. The only region of the late antique or early Byzantine
world for which large numbers of private documentary sources survive
is Egypt, where the dry sands have preserved hundreds of papyrological
fragments and texts ranging from the theological and the literary, to pri-
vate letters, estate accounts and receipts.22 Many of the documents that
we encounter are inevitably difficult to classify. As shall be seen shortly,
from the fifth to the seventh century AD, for example, much of the land
around the Middle Egyptian city of Oxyrhynchus was owned by members
of a single local family, the Flavii Apiones.23 A large number of docu-
ments from this family’s archive survive. Yet much of the documentation
that may be associated with this archive looks distinctly official and public
17 Clarysse, ‘Tomoi Synkollesimoi’, 356–8.

18 Clarysse, ‘Tomoi Synkollesimoi’, 352. 19 Clarysse, ‘Tomoi Synkollesimoi’, 349–50.
20 E. Turner, The Typology of the Early Codex (Philadelphia, 1977).
21 Note the imperial monobiblon referred to by Malalas (Scott, ‘Malalas and Justinian’s
Codification’, 21) and the demosios ptyx (emendavit Rea) referred to in J.Edict 13.15.
22 Although see now L. Koenen, ‘The Carbonized Archive from Petra’, Journal of Roman
Archaeology, 9 (1996), 177–88.
23 R. Mazza, L’archivio degli Apioni (Bari, 2001); P. Sarris, Economy and Society in the Age
of Justinian (Cambridge, 2006).
22 Peter Sarris
in character. By the sixth century, members of the Apion household were

clearly obliged to assist in the collection of imperial taxes for the region
around their hometown, by virtue of the head of the family’s possession
of the title of pagarch, and we have documentation concerned with this:
letters to and from imperial officials, accounts of payment of arrears and
so on.24 Are these documents ‘public’ or ‘private’? Does their existence
within the archive affect its character? Likewise, many of those whom
the Apion household employed as estate managers, or from whom the
household demanded sureties with respect to its workforce, held eccle-
siastical rank (and from the reign of Constantine the Christian clergy
had acquired a distinct legal status signified by their exemption from
civic munera).25 Is a contract made between an Apion-employed artisan
and an Apion-employed deacon a purely ‘lay’ document? What of the
documentation concerned with Apion payments to local churches?
This picture is further complicated when we start looking beyond
Oxyrhynchus. For the mid-to-late sixth century (as we shall see), we pos-
sess a number of documents and texts belonging to the private archive of
the Psimanobet family, a sort of yeoman farmer/gentry family from the
village of Aphrodite (more commonly known in English as ‘Aphrodito’ –
the name given it after the Arab conquest of the seventh century).26 Prob-
ably in the 530s, one member of this family, a certain Apollos, founded
and retired to a monastery, in which the family thereafter maintained an
active interest. How are we to classify documents contained in the Psi-
manobet family archive concerned with the administration of Apollos’
monastery (known, after its founder, as Apa Apollous)? A parallel prob-
lem emerges with respect to the (mostly Coptic) papyri and ostraca that
survive from the sixth to the eighth centuries from the Egyptian village
of Jeme. While many of the documents from Jeme were found in the
houses of those who had owned them, it is clear that many lay members
of local society entrusted their personal documentary archives to the safe
keeping of the Church, which probably began to fulfil this function as
soon as it acquired ‘legal personality’ under Constantine in the fourth
century.27
24 Sarris, Economy and Society, 149–76.

25 C. Rapp, Holy Bishops in Late Antiquity (Berkeley, CA, 2005), 238.
26 H. I. Bell, ‘An Egyptian Village in the Age of Justinian’, Journal of Hellenic Studies, 64
(1944), 21–36; L. MacCoull, Dioscorus of Aphrodito: His Work and World (Berkeley, CA,
1988); T. Gagos and P. van Minnen, Settling a Dispute: Toward a Legal Anthropology of
Late Antique Egypt (Ann Arbor, MI, 1994).
27 T. G. Wilfong, Women of Jeme: Lives in a Coptic Town in Late Antique Egypt (Ann Arbor,
MI, 2002).
This phenomenon requires a certain degree of explanation. In ear-

lier periods of the history of Roman Egypt (and most probably into the
sixth century), there almost certainly continued the traditional Greek
and Hellenistic practice whereby private papers, especially those per-
taining to relations between individuals and the state, could be stored
in an officially run governmental or gubernatorial archive (a grapheion
or suchlike) in the local city.28 However, the fifth and sixth centuries
in Egypt, as elsewhere in the East Roman Empire, witnessed a gradual
emasculation of the structures and institutions of municipal government
as local society came increasingly to be dominated by aristocratic mag-
nate households, such as that of the Flavii Apiones.29 While the state
sought to harness the social authority and clout of these magnate house-
holds through investing them with public obligations that had hitherto
been delegated to civic curiae, the net result was inevitably a certain mea-
sure of ‘privatization’ of functions that had formerly been the preserve of
municipal institutions. As part of this process, the use and maintenance
of ‘public’ archives seemingly went into decline, even with respect to doc-
uments pertaining to public responsibilities and affairs. In a law of 535,
for example, the emperor Justinian noted that ‘the public registration of
property transactions was no longer taking place because cities lacked
archives where the documents might be kept’.30 Increasingly, it would
appear, the private archives of great estates and especially the archival
resources of the Church came to replace the archivum publicum referred
to in the Justinianic constitution.
Perhaps the best way to get around the problems of definition occa-
sioned by these evolving circumstances is simply not to worry about
them too much and instead to fall back upon ‘the primacy of practical
reason’. That a family’s private papers happened to be stored in a church
or monastery, for example, as seems to have occurred at Jeme, reveals a
great deal about the role of the Church in local society, but it does not
really affect the lay character of the papers themselves. Likewise, that the
Apion archive may have contained papyri of an official provenance, or
papyri concerned with dealings between the family and local ecclesiasti-
cal institutions, does not of itself alter the fact that the family archive was
privately run for the Apion household, by the household and its secre-
tariat, irrespective of what other documents were acquired and preserved
along the way.
28 Brosius, ‘Ancient Archives’. 29 Sarris, Economy and Society, 149–99.

30 J. H. W. G. Liebeschuetz, The Decline and Fall of the Roman City (Oxford, 2001), citing
J.Nov. 15 praef.: ‘cum nullum habeant archivum, in quo gesta apud se reponant’.
24 Peter Sarris
Problems of language
The identification of private lay archives in the early Byzantine period
is further complicated, especially with regard to the public/private dis-
tinction, by a linguistic feature of those papyri (the majority) written in
Greek. Here, once again, the problem may be most clearly exemplified
with respect to the Apion archive.
In the Greek documentary papyri that make up the Apion archive,
we often find employed a terminology and vocabulary that bears very
close resemblance to that used by the imperial authorities in public and
administrative contexts. As a result, documents that are entirely ‘private’
in character, may be read as if they are ‘public’ or official. This lin-
guistic feature of the Apion papyri is, however, entirely explicable both
lexically and sociologically in a way that may also have implications for
how we read post-Roman Western Latin documentation written within
or emergent from the late Roman tradition. Members of early Byzantine
aristocratic families, such as the Apiones, typically enjoyed high-ranking
careers in the imperial bureaucracy and army. The same was also true of
many of those employed as their estate administrators and overseers.31 It
would only have been natural for such individuals to deploy, with regard
to their private economic transactions, techniques and terminological
forms derived from their experience of imperial administration, thereby
(to modern eyes) giving them something of an official appearance or
quality. We should also remember that many of the documents we pos-
sess were written by men whose native tongue is likely to have been
Coptic, individuals who may have had a natural tendency, when writ-
ing Greek, to fall back upon well-known phrases, clichés and formulas.
Accordingly, many of the documents we possess from the Apion archive,
such as petitions to the head of the family drafted by individuals bearing
Coptic names, for example, have a distinctly stilted quality about them.
In many ways their linguistic register resembles the formulaic tone at
times evident in modern sub-continental (Indian) English, itself a hang-
over from the verbosity of Victorian and Edwardian ‘officialese’. As a
result, we must be careful not to be misled by the linguistic register of
a document into misidentifying its provenance or character, or even the
‘private’, as opposed to ‘official’, nature of its subject matter.
To a great extent, just such a misapprehension underlies the hypothesis
of the distinguished French papyrologist Jean Gascou, who has argued,
in a highly influential and stimulating work, that the large estates of
31 Thus the estates recorded in P.Bad IV 95 were administered by estate employees bearing
the official imperial title of comites or ‘counts’.
early Byzantine Egypt, such as that of the Apiones, were primarily ‘semi-
public’ bodies created by imperial fiat to meet the needs of the impe-
rial government in Constantinople.32 In its étatisme, this is a splendidly
Parisian stance, albeit one that is not entirely supported by a closer read-
ing of the evidence. At the core of Gascou’s argument, for example, is
the assertion that the ‘rents’ (Greek phoroi) collected by the managers
of the Apion estates around Oxyrhynchus from the inhabitants of the
estates’ labour settlements (epoikia) represented not the private income
of the Apion family, but rather a ‘rent-tax’ much of which was destined
for imperial coffers. Yet the papyrological support for this assertion is
very thin. The Apion estate accounts are careful to distinguish between
rents (phoroi) and taxes (demosia or the synteleia kephales) collected from
estate employees.33 While the latter were handed over to the imperial
authorities, the former tended to be spent on the internal requirements
of the household, from funding the extension of cultivated land to the
hiring of dancing girls. The important point to note from our perspec-
tive, however, which may be of significance to the reading of the Western
Latin sources, is that tone and style of language are not always clear
indicators of documentary or archival ‘type’. In the West, in short, an
‘official’ Latin vocabulary derived from the workings of the Roman state
could have continued to be used in private contexts long after the fiscal
and administrative structures of empire had faded away – a point with
considerable implications for the work of both Jean Durliat and Walter
Goffart.34
Documentation, ‘archivalization’, and social and

economic relations in the early Byzantine world
The earliest archival evidence we have, dating from the third millen-
nium BC, is represented by the ‘economic documentation’ of inscribed
tablets referred to earlier. The economic focus of this documentation is
important for alerting us to the fact that much of the impetus to cre-
ate, store and systematize the collection of documentary evidence was,
from the very beginning, generated by relations of economic dependence
32 J. Gascou, ‘Les grands domaines, la cité, et l’état en Égypte byzantine’, Travaux et

mémoires, 9 (1985), 1–90. For a detailed critique, see Sarris, Economy and Society, 149–
76.
33 E.g. P.Oxy LV 3804 line 93, distinguishing between payments made for the poll tax and
for rent on sheep, or the pasturing thereof, by the inhabitants of the settlement of Luciu.
34 E.g. W. Goffart, Barbarian Tides: The Migration Age and the Later Roman Empire
(Philadelphia, 2006); J. Durliat, Les finances publiques de Dioclétien aux Carolingiens (284–
889) (Sigmaringen, 1990).
26 Peter Sarris
and inequality and by the urge to enforce or contest rights over property
and person. By ‘clarifying the potentially contestable’, documents and
archives served to frame social and economic relations and tensions.35
As a result, it should not surprise us that, just as archivalized documen-
tation continued to be central to the workings of the late Roman state
in the early Byzantine East, so, too, did it remain central to the opera-
tion and workings of society beyond the state, especially the concerns of
landowners and those who interacted with them. So, for example, almost
all of the papyrological evidence we have comes ‘from the viewpoint of
the propertied classes of the cities of Egypt, the metropoleis and nomes’,
and even the village archives we possess (from Karanis, Theadelphia and
Aphrodito) can be seen to have largely been concerned with and deter-
mined by the demands of the Roman state and relations with members
of the locally dominant landed elite, who also tended to reside in the
region’s cities.36
For the period from the third to the sixth centuries, the main extant
private archives recorded in the documentary papyri (excluding those
held by churches) can be summarized as follows. For the third century,
we possess the ‘Heroninos’ archive from Theadelphia in the Arsinoite
nome. For the fourth century, there exist the archive of Apollonios from
the Hermopolite nome, the archives of Theophanes and Aurelia Charite
and her son from the city of Hermopolis itself, and that of Ammon
scholasticus from Panopolis, alongside the archive of Leonides, and that
of Papnoutis and Dorotheos from Oxyrhynchus. For the fifth century,
the main non-ecclesiastical archive is that of Taurinos from Hermopo-
lis, while the sixth century preserves two related archives (of Dioscorus
and Phoibammon) from the town of Aphrodito, and that of the Api-
ones from Oxyrhynchus.37 Of these, by far the most important is that
of the Apion family, whose members we can trace in imperial service
from the early fifth century through to the early seventh.38 The Apion
archive is of inestimable value, for the papyri provide illuminating insights
into the day-to-day realities of life in what was, economically, the most
important, productive and highly developed region of the entire Eastern
Empire. The documents permit us to begin to come to terms with the
35 Sijpesteijn, ‘The Archival Mind’, 163.

36 R. Bagnall, Egypt in Late Antiquity (Princeton, NJ, 1993); I have not yet had an oppor-
tunity to consult Bagnall, Everyday Writing in the Graeco-Roman East (Berkeley, CA,
2011).
37 P. van Minnen, ‘The Other Cities in Later Roman Egypt’, in R. Bagnall (ed.), Egypt in
the Byzantine World 300–700 (Cambridge, 2007), 207–25, at 215.
38 Although the stemma is now obsolete, for a general sense of the family’s history and that
of the archive, see Sarris, Economy and Society, 10–28.
extent and character of the economic resources that underpinned the

social authority and political role of members of the late antique imperial
aristocracy. Crucially, they also enable us to see a late antique archive
‘at work’.39
The economic details concerning the day-to-day management of the
Apion estates, the condition and organization of labour, and the chains
of command and responsibility within the Apion household that we can
reconstruct on the basis of the papyri are not the primary concern here
and have been set out in detail elsewhere.40 What is more significant is
that each and every one of these aspects generated a great deal of paper-
work, which was gathered in at a central Apion bureau in Oxyrhynchus:
there the documentation was often filed in its own right in a retriev-
able form. This is particularly evident with respect to documents of
receipt and other legally enforceable contracts: so, for example, P.Oxy
LXXI 4835 records a loan of money on the part of an employee of the
Apion household to the inhabitant of a labour settlement or epoikion.
The receipt for the loan was then stored in the archive with the inscrip-
tion on its verso ‘receipt of Phoibammon from the epoikion of Peneb’.41
Likewise, in P.Oxy I 134, 135 and 136, we have respectively a contract of
employment between the Apion household and a stonemason, a contract
of surety guaranteeing the residence on an estate labour settlement of
an agricultural labourer, and the work contract of an estate overseer or
pronoetes. On the verso of each of these, we find a label describing the
nature of the document, indicative of its having been filed away within
the archive in case of any need for future reference.
Alternatively, the details required from a document could be extracted
and archivalized, the original document then being reused as scrap: such,
for example, was evidently the fate of the annual accounts of receipt and
expenditure submitted by the estate overseers or pronoetai employed by
the Apiones’ Oxyrhynchite managers to supervise the collection of rev-
enues in kind and coin from the various estate epoikia. These documents
essentially conformed to a pattern of ‘charge–discharge’ accounting, per-
mitting the ready identification through historical comparison of unex-
pected or unexplained shortfalls in estate income from one year to the
next (on which see more below). Accordingly, it would appear that once
in the hands of the managers of the Oxyrhynchite bureau, the documents’
contents were entered into some sort of central ledger or codex, a copy
of which was then probably sent to a higher level of estate management
39 For a detailed account of the contents of the archive and the uses to which they can be
put, see Sarris, Economy and Society, 29–89.
40 Sarris, Economy and Society, 29–80. 41 P.Oxy LXXI 4835 verso.
28 Peter Sarris
in Alexandria, with the verso of the original annual accounts then being
used for jottings, mathematical calculations and the like.42 On the verso
of the annual estate accounts of a pronoetes published as P.Oxy LV 3804,
for example, we find (published as P.Oxy LV 3805) details and sums
pertaining to various financial transactions from throughout the region
of the Oxyrhynchite, totally unrelated to the contents of the accounts
themselves: rents received, for example, on an urban warehouse and a
synagogue, and a payment made by agricultural workers with respect to
hunting rights associated with an estate demesne or autourgia.43 Simi-
larly, on the verso of the estate accounts published as P.Oxy XIX 2243
(a), we find – published as P.Oxy XIX 2243 (b) – a list of contracts
of employment again unconnected to the contents of the accounts. In
addition to estate accounts, receipts for loans in coin or kind and other
legal contracts such as contracts of employment or leases, the archive
also included petitions addressed to the head of the Apion family and
letters concerning estate business or official affairs, although here the
evidence for archivalization as such is more meagre and we may simply
be encountering chance survivals.44
Documentation, credit arrangements and control

One of the most conspicuous features of the Apion papyri is the ‘paper
trail’ created by credit arrangements and associated economic transac-
tions and by attempts to subject to careful scrutiny and control the activ-
ities of estate employees. So, for example, when collecting payments, be
they fiscal or rental, from members of the estate workforce, the Apion
pronoetai were meant to operate according to an annual schedule of
demands or apaitesimon. When collecting such revenues, the overseers
were supposed to issue the estate workforce with receipts styled entagia,
which the peasant could, if need be, bring forth as proof that he had met
his personal obligations. The sums collected and dispensed by the over-
seer were then meant to be worked up into the annual set of accounts sent
off to the Oxyrhynchite office.45 Any act of fraudulence or theft engaged
in by the pronoetes would thus have been readily identified by the estate
accountants or chartoularioi in Oxyrhynchus by means of checking the
42 For the existence of an Alexandrian office, see Sarris, Economy and Society, 46.
43 E.g. P.Oxy LV 3804 and 3805 lines 16–17.
44 Sarris, Economy and Society, for details, examples and discussion of each of these cate-
gories.
45 Each one of these features can be seen in the overseer’s contract of employment: P.Oxy
I 136.
revenues collected against the schedule of demands issued and, if neces-

sary, the receipts held by the estate’s workforce.46 Moreover, the estate
accounts and contracts of employment record that wages were frequently
issued to estate employees in the form of credit notes or chitties termed
pittacia, which could then in turn be cashed in with estate managers and
cashiers styled enoikologoi and trapezitai.47 Certain of these survive from
other contemporary estates in a papyrologically identifiable form.48
Further evidence concerning credit arrangements emerges from the
deeds of receipt contained within the archive, which generally bear the
designation of cheirographiai. These documents often consist of state-
ments made by workers bearing the legal status of enapographoi georgoi
or coloni adscripticii acknowledging that they had received items of irri-
gational machinery from the household. As such, the documents reveal
that while the georgoi of the Apion labour settlements frequently owned
their own agricultural tools and livestock, the household provided major
capital requirements for agriculture, especially for mechanically irrigated
cultivation.
P.Oxy XXXVI 2779 may be taken as a simple example of the
cheirographia-type.49 The deed is dated to 530, and was made on behalf
of a certain Aurelius Epimachus, ‘also called Apima’, from the epoikion
of Panguleeion.50 Apima declares:
Having now had occasion to require one axle for the landowner’s water-wheel
under my charge which is called ‘five arouras’ and supplies water to vine-land
and arable land, I went up to the city and asked your magnificence to order the
said axle to be provided for me. And at once your magnificence, having regard
to the state of your property, gave me the said axle through Moison, oiketes of
the said most magnificent man, a new one, serviceable, suitable for irrigation
and satisfactory. I received it as completion of all the irrigation implements this
very day . . . The old axle was brought in and given to the great household of the
landowner. In declaration of the receipt I have made this deed (cheirographia)
which is valid written as a single copy and in reply to the formal question I gave
my assent.51
Among the more fragmentary deeds found within the Apion archive are
three – P.Oxy I 137, XVI 1988 and XVI 1989 – which reveal how the
parallel system of credit notes in the form of pittakia was then interleaved
into this practice.52 P.Oxy I 137, dating from 584, consists of an acknowl-
edgement given by a georgos named Aurelius Ptollion that he had received
an axle for a waterwheel. The document stands out from that detailed
46 Sarris, Economy and Society, 52. 47 Sarris, Economy and Society, 56, 67, 75, 77, 88.
48 Especially CPR VII 23. 49 P.Oxy XXXVI 2779.
50 P.Oxy XXXVI 2779 lines 1–9. 51 P.Oxy XXXVI 2779 lines 9–24.
52 P.Oxy I 137; P.Oxy XVI 1988–9. Note also P.Oxy XVI 1990.
30 Peter Sarris
above, however, in terms of the information it provides as to the proce-

dure whereby the new axle was issued, for Ptollion states that the price
of the wheel was ‘credited to me in the cheques (pittacia)’.53 Ptollion was
thus provided with the wherewithal to buy the axle by the Apion house-
hold. The pittakion was presumably handed over by Ptollion to whoever
was responsible for issuing or selling the machinery. P.Oxy XVI 1988,
dating from 587, also acknowledging the receipt of a water-wheel axle,
records that the georgos in question had purchased a replacement and
that ‘its price was credited to me in my pittakia’.54 The same procedure
is recorded in P.Oxy XVI 1989, dating from 590×1.55 The use of estate
cheques or credit notes evidently extended well beyond the mere issuing
of wages.
The contractual papyri thus reveal the highly formalized character
of dealings between the Apion household and its employees, and the
household’s extraordinary dependence upon written legal instruments,
and written documentation in general (accounts, entagia, apaitesima, pit-
takia, etc.). Certainly, this feature is indicative of the highly standardized
and professional character of household estate management. On another
level, however, as with relations between state and subject, the evidence
would suggest that the written word also played an important symbolic,
and – it might be inferred, intimidatory – role in relations between master
and servant, especially given how often, in the contractual papyri, the lat-
ter is described as unlettered.56 Even in this highly documentary society
the use of archivalized knowledge and information was ‘functional’ in a
variety of ways.
The other archival collections that survive from the sixth and sev-
enth centuries from elsewhere in early Byzantine Egypt reveal a similar
concentration on economic transactions, credit arrangements, loans and
documents establishing or associated with legal obligations or rights. Two
particularly important thematically and chronologically distinct papy-
rological dossiers survive from early medieval Aphrodito, for example.
The later of these, dating from the Umayyad period, is derived from
the bureau of a local Arab governor, Qurra b. Sharik (r. 709–14), and
it reveals much of the character of early Islamic rule as the Islamic
53 P.Oxy I 137 line 19. 54 P.Oxy XVI 1988 lines 23–5.

55 P.Oxy XVI 1989 lines 17–18 – only here the word used for cheques is entakia. For the
interchangeability of these words, see E. R. Hardy, The Large Estates of Byzantine Egypt
(New York, 1931), 98. That this practice was not uniform, however, is established by
PSI I 60 (a. 595).
56 For literacy rates and levels in late Roman Egypt, see the discussion in Bagnall, Egypt
in Late Antiquity, 230–60.
government began to lay for itself firmer administrative foundations.57

For the earlier period, encompassing much of the sixth century, as noted
above, the extant papyrological record preserves part of the archive of
one of Aphrodito’s most prominent families, the descendants of a certain
Psimanobet – ‘the goose-herd’s son’ – more commonly known as the
‘Dioscorus archive’.58 Although, as this name would suggest, the origins
of the family would appear to have been relatively humble, by the early
sixth century, its male members had risen into the elite of the local village
headmen or protokometai. As a result, the Dioscoran materials preserve
not only accounts and leases concerned with the family’s own direct eco-
nomic interests and dealings, but also documents concerning the village’s
relations with the ‘outside world’. Chief among these is a series of peti-
tions and verse encomia by which the village lawyer and man-of-letters,
Dioscorus, sought to protect the settlement’s independent fiscal status
(autopragia) against the encroachments of local landowners and imperial
officials.59 A focus on economic interests, legal rights and especially credit
arrangements is also evident from the seventh- and eighth-century papyri
that survive from the town of Jeme.60 The documents safeguarded in the
church in Jeme were overwhelmingly legal in character, detailing dispute
resolutions and suchlike. The primary extant archive beyond that held by
the church, interestingly, consists of pottery sherds or ostraca associated
with the business activities of the local moneylender, a woman by the
name of Koloje.61
The seventh century and beyond

In the year 616, the conquering armies of the Sassanian Shah Khusro
II entered imperial territory in Egypt, and by 619 Alexandria had been
wrested from East Roman control. Although the Persian occupation was
eventually reversed, Egypt was not to be restored to Roman rule for long.
In 639, the Muslim general Amr Ibn al-As pursued retreating Byzantine
forces out of Palestine and initiated the Arab conquest of the region.
Accordingly, we lose sight of the one area of the Byzantine world from
which extensive lay archival documentation survives. That a culture of
archival record-keeping remained a feature of lay society in what sur-
vived of the East Roman Empire, however, seems clear. For irrespective
57 See especially Greek Papyri in the British Museum, IV, The Aphrodito Papyri, ed. H. I.
Bell (London, 1910). I am indebted to Dr Petra Sijpesteijn for discussion of the Arabic
papyri.
58 Gagos and Van Minnen, Settling a Dispute, 19, 131 (family tree).
59 Sarris, Economy and Society, 96–114.
60 Wilfong, Women of Jeme. 61 Wilfong, Women of Jeme, 117–50.
32 Peter Sarris
of how much territory was lost to the Arabs in the seventh century, and
no matter how much damage was done to the urban infrastructure of
Anatolia, Asia Minor and Greece, a bureaucratic political culture and
a bureaucratic class survived in Constantinople.62 Moreover, although
the Byzantine aristocracy beyond Constantinople became increasingly
militarized, it remained profoundly literate. Thus, in the eleventh cen-
tury, we see the retired general John Cecaumenos advising a young man,
‘When you are free and not busy with a commander’s duties, read books,
both histories and church writings. Do not say, “What benefit is there
for a soldier from ecclesiastical books?”, for you will profit greatly from
them.’63 There is no reason to suppose that middle Byzantine impe-
rial government was any less reliant on the written word than was late
antique imperial government. Nor is there any reason to suppose that
middle Byzantine landowners made any less use of written instruments.
In the eleventh-century accounts of legal proceedings known as the Peira
of Eustathius the Roman, for example, we find both great landowners
and independent peasant farmers producing written documents, agree-
ments, settlements of claims and contracts at court, so as to advance or
protect their interests against one another.64 That so little documentary
material survives is not the point. As ever, absence of evidence should
not be confused for evidence of absence.
Moreover, it is interesting to note that when documentary evidence
does begin to become more abundant in the Byzantine world, especially
from the eleventh century onwards, the documents that surface bear
striking terminological similarity to their late antique antecedents in a
way that cannot be explained entirely in terms of the survival of legal
and literary traditions from Late Antiquity through to the middle Byzan-
tine period. Rather, the same terms are used to describe the same –
often quite technical – features of agrarian social relations and relations
of production. For example, the directly administered ‘demesne’ of the
aristocratically owned bipartite estates recorded in the early Byzantine
papyri such as those found in the Apion archive was, as we have seen,
commonly known as the autourgia. The same term is used to designate
the same category of land in our late Byzantine sources.65 It is strik-
ing that the contents both of the eighth-century Farmer’s Law, detailing
regulations concerning village life, and the eleventh-century Peira make
62 M. Whittow, The Making of Orthodox Byzantium (London, 1996).

63 Cited in C. Mango, Byzantium: The Empire of New Rome (London, 1980), 239.
64 E.g. Peira 7.6, concerning a private settlement of claims or dialysis.
65 P. Sarris, ‘Economics, Trade, and Feudalism’, in L. James (ed.), A Companion to Byzan-
tium (Oxford, 2010), 25–42.
sense only when placed in the semantic context of our late antique Greek
documentary sources – the Egyptian papyri. This apparent continuity of
both managerial tradition (with respect to the autourgia) and vocabulary
(with respect to the Farmer’s Law and Peira) is perhaps suggestive of a
high degree of survival of specific documentary traditions.
Conclusions and implications

The early Byzantine evidence, seen in the context of the history of archival
culture as a whole in the Near East back to the world of the third mil-
lennium BC, has a number of potential implications for students of the
early medieval West. Certainly, the survival of archival and documen-
tary cultures in Byzantium is likely to have been much facilitated by
the survival of a bureaucratic class in Constantinople and of a strong
tradition of elite literacy. In this respect, Byzantium arguably stands in
marked – some would insist very marked – contrast to the world of the
post-Roman and early medieval West, although only in very few instances
can state structures and elite literacy be seen to have disappeared abso-
lutely. But perhaps the evidence of archival cultures in the round would
suggest that too narrow a concentration on bureaucracies and elites rather
misses the point. As this chapter has hopefully indicated, and as those by
Conant and Everett certainly do, the late Roman world was a world in
which the use and archivalized storage of written instruments were deeply
embedded and are discernible at a range of social levels. The documents
on slate from early medieval Spain clearly indicate that such traditions
were sufficiently embedded to have been more than capable of survival
at the level of village society and within the context of agrarian social
relations.
We are thus dealing with post-Roman societies in the West where peo-
ple were used to dealing with the written word and were perfectly aware
of its utility. Moreover, as we have seen, the impetus first to document
and then to ‘archivalize’ information was to a very great extent gener-
ated by economic transactions and concerns – in particular, relations of
dependence, actual or anticipated coercion, and structures of extraction
and exploitation. Credit relations, in particular, were especially inclined
to leave behind them an extensive paper trail. This is important, in that
systems of credit would appear to have been central to the operation of
rural patronage in the late Roman and very early medieval (that is, fifth-
century) West, where, as in the East, well-connected patrons would offer
credit to their potential peasant clients in return for their labour or in
34 Peter Sarris
return for acquiring ownership of their plots of land.66 It may well have
been similarly generative of archivalized documentation.
Most crucially of all, however, hardly any (and perhaps none at all) of
the societies that emerged from late Roman conditions in the fifth- and
sixth-century West were egalitarian or classless societies. Anglo-Saxon
England has perhaps the strongest claim to have been the exception that
proves the rule, yet even here such a characterization of early medieval
society is highly contestable.67 James Campbell, for example, has argued
that the England of Bede in the seventh to eighth centuries was ‘a rich
place with a booming export trade in cloth’.68 Likewise as Mark Whittow
has noted:
We know almost nothing of an Anglo-Saxon leather industry, but since half

a million people wore shoes and belts it must have existed. Monkwearmouth-
Jarrow was a lavishly-furnished monastery . . . the resources required to sustain
it must have been substantial. We now know that Monkwearmouth-Jarrow was
not alone: Bede’s England was full of monasteries. Multiplied out the resources
required to sustain this culture will have been immense.69
Certainly on the continent, and perhaps even to some extent in Britain,

the Roman Empire bequeathed what Jairus Banaji has described as a
‘legacy of oppression’ that continued to weigh heavily on elements of
the peasantry, most obviously in terms of an inherited tradition of direct
management of estates.70 In Gaul and Italy, such continuity is clearly
discernible.71 Even in lowland Britain, where such continuity is more
unlikely, one should note the striking and highly suggestive correlation
between late Roman villa sites and places of high-status Anglo-Saxon
66 B. Sirks, ‘The Farmer, the Landlord, and the Law in the Fifth Century’, in R. W.
Mathisen (ed.), Law, Society, and Authority in Late Antiquity (Oxford, 2001), 256–61;
see also the classic account of fifth-century rural patronage in Salvian of Marseilles, De
gubernatione Dei (MGH AA 1.1).
67 See esp. M. Whittow, ‘Beyond the Cultural Turn: Economic History Revisited?’, Journal
of Roman Archaeology, 20 (2007), 697–704, esp. 701–3 (review of Wickham, Framing).
68 J. Campbell, ‘Production and Distribution in Early and Middle Anglo-Saxon England’,
in T. Pestell and K. Ulmschneider (eds.), Markets in Early Medieval Europe: Trading and
Productive Sites, 650–850 (Macclesfield, 2003), 12–19.
69 Whittow, ‘Beyond the Cultural Turn’, 703.
70 J. Banaji, ‘Aristocracies, Peasantries and the Framing of the Early Middle Ages’, Journal
of Agrarian Change, 9 (2009), 59–92.
71 As argued in P. Sarris, ‘The Origins of the Manorial Economy’, English Historical
Review, 119 (2004), 279–311. The survival of actual patterns of great landowning
is also acknowledged, but with very different emphasis, in Wickham, Framing, 195 and
299; see P. Sarris, ‘Continuity and Discontinuity in the Post-Roman Economy’, Journal
of Agrarian Change, 6 (2006), 400–13.
settlement.72 The survival and (in certain instances) continuity of tra-

dition of relations of dependence, coercion and exploitation in the early
medieval West would have given both the powerful and the vulnerable
a strong incentive to record and to store evidence pertaining to their
respective rights, obligations and claims, and is likely to have favoured
the survival of documentary and archival cultures. Any such tendency
would have been further intensified by what Wickham has character-
ized as the legacy from the Roman Empire to much of the post-Roman
West of the concept of there being ‘a public sphere of power’ in which
rights could be asserted, challenged and either vindicated or dismissed.73
Social and economic relations of inequality perhaps provide the key to
the existence and perpetuation of archivalized documentation in the early
Middle Ages, not the broader cultural contexts and categories to which
attention can too easily be drawn.
72 Private communication with Dr R. Faith. 73 Wickham, Framing, 57.

3 Public administration, private individuals
and the written word in Late Antique
North Africa, c. 284–700
Jonathan P. Conant
North Africa was also a place where the civil administration drove documentary
use, even through changes in regime. Here, however, we also see the particular
importance of the military. In addition, the nature of the surviving documents
draws attention to the deep penetration of documentary culture well below the
level of the elite, as well as to the importance of documents that were not intended
to be preserved for long periods. This evidence also indicates how crucial the
interests of individuals, or of collectives outside the apparatus of the state, were
to the persistence of documentary culture. The distinction between laypeople and
clergy is visible, but did not determine how documents were used or preserved.
Egyptians were not the only inhabitants of the late antique Mediter-
ranean to make extensive use of written documents. To be sure, outside
Egypt, comparatively few sources have been preserved through which to
study documentary practice in Late Antiquity; but what evidence we do
have seems to suggest that the scattered documents that survive from the
western Mediterranean barely scratch the surface of what was once pro-
duced there. Together with occasional references to documents that once
existed but have subsequently been lost or destroyed, these fragments
raise important questions as to the nature and extent of documentary
practice throughout the late Roman and early medieval West.
Such questions are particularly acute in the case of the Roman Empire’s
African provinces, because they have never been explored synthetically in
this region. Yet, Africa has preserved a remarkable range of materials that
allow us to assess the role of documentation in local society. By the late
antique period, an agricultural regime based on the cultivation of olive
trees and cereals had extended even into marginal regions of the North
African countryside, where arid conditions are favourable to the preser-
vation of the organic materials with and on which ancient documents
were written. As a result, a small but significant amount of written doc-
umentation survives from Africa: at least three assemblages of wooden
36
Late Antique North Africa 37
tablets and – though otherwise rare – over twenty collections of Latin

ostraca, ranging from single finds to a large group of over 140 sherds.
Not all of this material has yet been published, but these scattered rem-
nants combine with references to other similar texts in the historical,
legal, homiletic and other literary sources to suggest that Africa was once
characterized by a vibrant culture of record-keeping, in both the private
and the public spheres. Indeed, as in Egypt, much of the documenta-
tion generated in Africa was produced to satisfy the demands of the late
Roman fiscal, judicial and military administration (and its successors in
the Vandal and Byzantine periods); but a reliance on the written word to
facilitate the practical aspects of property ownership, estate management
and trade also seems to have percolated quite deeply into secular society
in general and to have survived there until at least the mid sixth century,
and quite possibly beyond.
Private documentation
There can be little question that African society relied heavily on written
documentation in the late antique period. To judge from the texts that
survive to us, estate management – or what Chris Wickham has referred
to as ‘accounting’ – seems to have been a major aspect of that documen-
tary culture.1 Extant tables of calculation indicate that this could involve
the computation of income and expenditures, reckoned either in coin
or in real terms. Thus, for example, in the late fifth century on a fun-
dus in the hill country of eastern Algeria, someone tallied up the money
that members of a local family had spent buying various small parcels of
land.2 Three ostraca from the same general area list a series of names,
each followed by a sum; the largest sherd adds a total and the phrase
partis dominice, suggesting that the numbers represent the income that a
local landowner received either in rents or from labour services on his or
her ‘demesne’.3 Such income also seems to be recorded in a collection of
five Vandal-era ostraca discovered in an earthenware jar near Bir Trouch,
again along the edges of the eastern Algerian highlands.4 At least three
1 Wickham, Framing, 266.

2 Tablettes Albertini: Actes privés de l’époque vandale (fin du Ve siècle), ed. C. Courtois et al.,
2 vols. (Paris, 1952) (henceforth TA, with act no. in Roman, volume and page in Arabic
numerals), XXXIII, 1:299.
3 E. Albertini, ‘Ostrakon byzantin de Négrine (Numidie)’, in Cinquantenaire de la Faculté
des lettres d’Alger (1881–1931) (Algiers, 1932), 53–62, at 62; see also R. Cagnat, in Bulletin
archéologique du Comité des travaux historiques et scientifiques (henceforth BCTH) (1908),
ccxlvii–ccxlix. On the ‘demesne’ in Africa and the late Roman world, see Wickham,
Framing, 273–80.
4 J.-P. Bonnal and P.-A. Février, ‘Ostraka de la région de Bir Trouch’, Bulletin d’archéologie
algérienne, 2 (1966–7), 239–49. See also the discussion of these documents in Wickham,
Framing, 266.
38 Jonathan P. Conant
of these documents involve the same individual, a certain Massiese, and

all of them appear to concern two different holdings under simultaneous
cultivation, one in a place called Viginti Boguli and another at Adeudaci.
The five texts follow a standard formula; the fullest of them reads:
annu nonu cartaginis / domni nostri regis Gunta/mundi accessit Massiese / in xx boguli
parte dominica / ordei m(ensuras) iii tres tam/tum.5
In the ninth year of Carthage of our lord king Gunthamund [AD 492×3]
Massiese accepted 3 three measures of barley only on the lord’s portion at Viginti
Boguli.
Though this is the only example to specify that the measures in question
were of barley, the others also involve mensurae, presumably of some
agricultural product. Taken together, these five ostraca likely represent
accounts of rents or labour services generated as a product of the internal
management of a local estate. Yet another document, this one undated,
survives from Ksar Koutine in southern Tunisia and is addressed to a
certain Severus conductor, probably the chief tenant of a local estate, but
perhaps a collector of customs duties.6
Other documents from North Africa suggest that the issuing of receipts
was probably also fairly widespread in Late Antiquity. For example, a col-
lection of eight ostraca survives from Maknassy in the highlands of south-
ern Tunisia (on the road between Sfax and Gafsa); five contain texts that
are legible in varying degrees and which for the most part appear to be
receipts for payments in kind of sheep meat and wine. Fragmentary ref-
erences to the Roman calendar –Kal(endas) mai(i), id(us) iulias, august –
indicate that these records were originally dated, but none can now be
assigned to a precise year. However, one of the ostraca seems to be
dated according to a king’s reign, and therefore probably belongs to the
period of Vandal or Moorish ascendancy in this region. This document is
addressed by a certain Sylvester to one Africanus Tu . . . .rus, while two or
perhaps three of the other texts were written by or addressed to a woman
named Urbanilla. All five indicate that they were once subscribed, though
the names of the subscribers have been lost.7
Other excavations, surveys and chance finds in Tunisia and Algeria
have turned up a handful of additional Latin ostraca from Late Antiquity,
and though they are even more difficult to interpret, many may also
plausibly be connected with the economic life of rural estates. One of
5 Bonnal and Février, ‘Ostraka de Bir Trouch’, 241.

6 A. Merlin, in BCTH (1913), ccxxxii, no. 2. On the term conductor, see A. H. M. Jones,
The Later Roman Empire, 284–602: A Social, Economic and Administrative Survey, 2 vols.
(Oxford, 1964), 417–19 and 788–92, but also 430.
7 P. Delattre and M. Héron de Villefosse, in BCTH (1912), cclviii–cclx.
the ostraca found at Henchir bou Gornine in southern Tunisia contains a

highly abbreviated text dating to 359 that relates to a farm known as the
Fundus Villa Magna. The text on a second sherd from the same site is
not fully intelligible, but seems to concern the repayment of a debt.8 At
least two of the four late antique ostraca from Henchir el Abiod, Algeria,
are associated with a place called Casae Nigrensium, and, of these, one
also refers to a Fundus Puteos. A third renders an account of goats and
sheep.9 One of the eleven ostraca found at Henchir el Maı̈z, Algeria,
makes reference to a Thrasamund (presumably the sixth-century Vandal
king of that name) and seems to have to do with oil production on a local
estate.10
It is worth observing, though, that estate managers were not the
only individuals to employ written documentation in late antique North
Africa. Evidence of its commercial use in the late fourth or early fifth cen-
tury is provided by thirty-two recently published ostraca from the island
of Jerba, many of which are in Greek but some of which are in Latin.
The texts concern places as far distant as Rhodes, the Troad and per-
haps Thrace; they include a letter and records that appear to be orders,
receipts, loans or notes; and they mention a slave trader, murex dye
and olive oil, as well as some unspecified commodity or commodities
measured in discrete units.11
North Africans also appear regularly to have recorded sales and other
transfers of property in written documents. On 24 January 494, for exam-
ple, a certain Julius Restitutus and his wife Donata sold a scrap of land
with a single almond tree on it – together with the water rights associated
with the property – to another local landholder by the name of Geminius
Felix. The whole package sold for 100 folles in cash. The parties involved
secured the services of a local writer familiar with the formulas neces-
sary to draw up a legal deed of sale (probably the local schoolteacher
Quadratianus), and he enumerated, among other details, the owners of
the properties that bounded this small holding in three directions.12
8 A. Merlin, in BCTH (1915), cxcii–cxciii. On the use of ostraca by Koloje the Money-
lender and her family to record loan agreements in late antique Egypt, see T. G. Wilfong,
Women of Jeme: Lives in a Coptic Town in Late Antique Egypt (Ann Arbor, MI, 2002),
117–28.
9 Cagnat, BCTH (1908), ccxlvii–ccxlix.
10 CIL VIII 22646.20. Wickham, Framing, 266n10, advocates reading the abbreviation pd
in this text as expandable once again to pars dominica.
11 Z. Várhelyi and R. S. Bagnall, ‘Ostraka’, in E. Fentress, A. Drine and R. Holod (eds.),
An Island Through Time: Jerba Studies. I. The Punic and Roman Periods (Portsmouth, RI,
2009), 334–44.
12 TA XIX, 1:274; for Quadratianus as the scribe, see 1:54. On the value of these trans-
actions, see P. Grierson, ‘The Tablettes Albertini and the Value of the solidus in the Fifth
and Sixth Centuries A.D.’, Journal of Roman Studies, 49 (1959), 73–80.
The instrument that records this transaction is one of the Albertini

Tablets, a remarkable collection of thirty-four documents preserved on
forty-five wooden panels that were discovered in 1928 along the Tunisian-
Algerian border in the region of Djebel Mrata. The panels include one
of the tables of calculations mentioned above, as well as a dowry and
various deeds of sale concerning a slave, an olive press and numerous
small plots of olive, fig and nut trees, most of which were associated
with a rural estate known as the Fundus Tuletianos.13 These tablets
reveal something of the significance of written documentation on the pre-
desert frontier of post-Roman Africa. The texts that they contain were
written by and for the rural smallholders that made up the Tuletianos
community, and in selecting friends, neighbours, patrons and clients to
witness their purchases and sales of property, local landholders seem to
have shown a preference for individuals who could read and write. The
fact that the wood on which the instruments were written appears to have
been imported into the Djebel Mrata region shows the importance that
members of this community attached to having a written document to
record their property transactions. So, too, the fact that landholders like
Restitutus, Donata and Felix were willing to pay scribes to draw up deeds
for relatively small-scale transactions like the sale of a single almond tree
and its water rights – though, to be sure, a selling price of 100 folles was
not an insignificant sum in the Fundus Tuletianos community: the same
amount could buy six olive trees, and for somewhat less one could buy
two fig trees.14 As important as it appears to have been to document a
sale on the Fundus Tuletianos, tablets that (for whatever reason) had
outlived their usefulness were scraped down and reused to hold the texts
of new instruments. This seems to have happened with some regularity.
Though not all of the documents in the assemblage can be precisely
dated, those that can were all written between 493 and 496, in the final
years of the reign of the Vandal king Gunthamund. The production
of so substantial an amount of written documentation in a restricted
region in such a short period of time may even suggest frequent archival
culling.15
13 See above, n. 2.
14 Six olive trees: TA, XVIII, 1:271; see also TA IX, 1:246 (90 folles). Two fig trees: TA
XVI, 1:269.
15 J. P. Conant, ‘Literacy and Private Documentation in Vandal North Africa: The Case
of the Albertini Tablets’, in A. H. Merrills (ed.), Vandals, Romans, and Berbers: New
Perspectives on Late Antique North Africa (Aldershot, 2004), 199–224; note, however,
that the sermon attributed there (p. 1) to Augustine (Sermo 303 [PL 39:2324–8]) is in
fact pseudo-Augustinian: L. Dossey, ‘Christians and Romans: Aspiration, Assimilation,
and Conflict in the North African Countryside’, unpublished Ph.D. thesis, Harvard
University (1998), 371.
The examples that we have considered so far all come from the high-
land or pre-desert fringes of the North African countryside, but they are
probably a guide to the practices of the region’s agricultural and urban
centre as well. At least, the documentary culture of late fifth-century
western Byzacena that is revealed in the Albertini Tablets seems to mesh
seamlessly with that of the diocese of Hippo Regius evoked by Augustine
nearly a century earlier. Here, too, the use of written deeds (instrumenta
or tabulae) to record sales and gifts of property appears to have been
a familiar practice. In one of his sermons, Augustine tells the story of a
man named Barnabas who received a piece of land as a gift from a certain
Eleusinus. Barnabas was later ordained a priest and a monastery was built
on the land, whereupon ‘he changed the instrumenta so that [the land] was
held in the name of the monastery’.16 The implication, though, is that at
the time of the original gift Barnabas received a written document trans-
ferring ownership of the property. It is also worth emphasizing that at
that point Barnabas was still a layman, and that he had retained the deed
until he needed to alter it once again to reflect the new change in owner-
ship entailed by his subsequent donation of the land to the monastery. In
another sermon, Augustine indicates that in his day deeds of title typically
specified the neighbours whose landholdings bordered on a property.17
As we have seen, in the Vandal period, too, the writers of the Fundus
Tuletianos documents continued to list the members of the local com-
munity whose properties bordered on a landed sale to the north, south,
east, west, north-west, south-west and toward the sea.18 As in Augus-
tine’s day, in the closing years of the fifth century the use of instruments
was not restricted to landed property.19 Indeed, the highest-value sale
in the Tuletianos dossier concerns a young slave boy named Fortinis.20
Another tablet records the movable goods with which a certain Januarius
endowed his future bride, Geminia Januarilla.21 Januarilla’s dowry is also
evocative of – though not a direct parallel to – the tabulae matrimonialae
that Augustine tells us were read aloud to brides at their weddings.22 In
short, then, the Albertini Tablets seem to have functioned in a world of
16 Augustine, Sermo 356.15 (PL 39:1581): ‘mutavit instrumenta, ut nomine monasterii

possideatur’.
17 Augustine, Sermo 358.2 (PL 39:1587): ‘quomodo solent instrumentis quaeri posses-
sores, inter quos sint affines’.
18 TA III–XXIV, 1:218–83.
19 Augustine, Confessiones 9.9.19 (ed. J. J. O’Donnell, 3 vols. [Oxford, 1992], I:112) and
Sermo 37.7 (CCSL 41:454).
20 TA II, 1:216–17. 21 TA I, 1:215.
22 Augustine, Confessiones 9.9.19 (ed. O’Donnell, I:111–12); see also Augustine, De Genesi
ad litteram 11.41 (CSEL 28.1:376), and E. Meyer, Legitimacy and Law in the Roman
World: Tabulae in Roman Belief and Practice (Cambridge, 2004), 277n76.
documentary practice that was recognizably the same as it had been in

Augustine’s day. Unfortunately, no other documents from late antique
North Africa have yet been published that unambiguously record dowries
or private sales. Still-unedited wooden tablets, such as those found at an
ancient well in Bir Trouch or those from Guert, Algeria, may contain
such documents; the same may be true of some of the Latin ostraca from
North Africa which have so far proven too difficult either to read or to
interpret.23 Even so, private individuals in late antique Africa seem to
have demanded, produced and made use of a considerable amount of
practical written documentation.
The late Roman administration and its successors

The culture of private documentation in late antique North Africa was
underpinned and informed by the practices of the late Roman state and
its successors, in both the civil and the military spheres. Indeed, in seek-
ing to control the empire, the later Roman bureaucracy demanded the
production of a truly extensive amount of written documentation. In the
records that survive to us, this official reliance on the written word is
particularly visible at the highest levels of the late Roman administra-
tion in Africa. The praetorian prefects, vicars and proconsuls of Roman
Africa of course received frequent rescripts from the imperial centre con-
cerning the region’s civil, ecclesiastical and military affairs.24 The vicar
of Africa, in turn, had written instructions from the emperor to send
along reports from subordinate provincial officials that required imperial
23 Unpublished wooden tablets: Bonnal and Février, ‘Ostraka de Bir Trouch’, 239; Mar-
seille, Archives Départementales des Bouches-du-Rhône, 100 J 19 (Archives de P.-A.
Février: Tablettes du Guert), with photos in Y. Duval and P.-A. Février, ‘Procès-verbal
de déposition de reliques de la région de Telergma (VIIe s.)’, Mélanges d’archéologie et
d’histoire (École française de Rome), 81 (1969), 257–320, Figures 44–5 (my thanks to
Mark Handley for this reference).
24 Praetorian prefect: e.g. CTh. 1.5.12; 1.15.10, 14, 17; 7.4.2, 3, 33; 7.6.1; 7.13.21, 22;
11.1.31; 11.7.8; 11.13.1; 11.28.8; 12.6.4; 12.12.1; 13.4.1; 13.6.10; 15.14.13; 16.2.31;
16.5.37. Vicar: e.g. CTh. 1.12.6; 7.8.9; 7.15.1; 9.26.3; 11.1.30; 16.2.29, 34; 16.5.35.
Proconsul: e.g. CTh. 1.3.1; 1.12.6, 8; 4.22.6; 6.28.7; 6.29.9; 7.8.7; 8.4.23; 8.5.64;
8.10.3, 4; 9.26.2; 9.39.3; 9.40.19; 10.1.13; 10.9.3; 10.10.27–8; 11.1.24, 28, 32, 34;
11.5.1; 11.7.19–21; 11.8.2; 11.17.2; 11.28.6; 11.30.53, 60, 62, 64–5; 12.1.95, 141–
5, 149, 166, 174, 176, 185–6; 12.5.3; 12.6.27, 28, 29, 31; 13.1.8, 18; 13.5.25, 30;
16.5.39, 41, 44, 54–5; 16.10.17–18; 16.11.1, 2. Rescript system: T. Honoré, Emperors
and Lawyers, 2nd edn (Oxford, 1994), 1–70. Composition of the CTh.: J. Harries and
I. Wood (eds.), The Theodosian Code (Ithaca, NY, 1993), esp. 17–67; T. Honoré, Law
in the Crisis of Empire, 379–455 A.D.: The Theodosian Dynasty and Its Quaestors (Oxford,
1998), 123–53; J. Matthews, Laying Down the Law: A Study of the Theodosian Code (New
Haven, CT, 2000); A. J. B. Sirks, The Theodosian Code: A Study (Friedrichsdorf, 2007).
Responsibilities of these officials: P. S. Barnwell, Emperor, Prefects, and Kings: The Roman
West, 395–565 (Chapel Hill, NC, 1992), 58–67.
attention.25 The proconsul of Africa – who was outside the vicar’s juris-
diction – similarly forwarded such cases to the emperor, and furthermore
submitted his own reports.26 At least on occasion, we also hear of less
exalted African officials sending written appeals to the emperor for his
consideration. Thus, for example, the leading citizens of Tripolitania
sent envoys to Valentinian I (364–75) in the late fourth century who
carried written decreta intended to inform the emperor of the severe and
widespread devastation that their province had experienced at the hands
of Moorish raiders.27 Similarly, though his missives were intercepted
before they reached the emperor, the Mauretanian general and future
rebel Firmus sent written letters to the imperial court to defend him-
self against the spurious accusations of one Romanus, who was then the
supreme commander of the Roman forces in Africa, or comes Africae.28
However, legislation of Constantius II (337–61) sought to ensure that
such direct exchange of letters between lower-ranking officials and the
emperor was the exception rather than the rule. When the governor of
Numidia wrote to Constantius to present a matter for the emperor’s
consideration, he was rebuked and informed that he should have first
submitted the issue to the vicar of Africa.29
The role that the vicar and proconsul of Africa played in relaying impor-
tant cases to the emperor also hints at the volume of written exchange
that probably took place laterally between the province’s civil and mili-
tary officials. The nature of our sources is such that we are not often able
to see such communications, though some light is shed on them in the
context of the coordination of military action. Thus, for example, if the
hostility of the local population made it impossible for the civilian author-
ities to apprehend those who had committed ‘outrages’ against priests, a
provincial governor was permitted to request military support from the
comes Africae by sending him a letter that included a copy of an imperial
rescript approving the use of troops for such purposes.30 Similarly, when
Constantius II sent a certain Gaudentius to Africa to prepare the province
for war against the emperor’s upstart nephew and junior colleague Julian,
the imperial representative discharged his task by writing letters both to
the comes Africae and to the region’s other military commanders.31 The
Mauretanian rebel Firmus likewise sent a letter that sought to explain
and justify the reasons for his revolt to the Roman military commander
charged with the pacification of Africa.32 When the commanding officer
of the recalcitrant Christian centurion Marcellus sent the soldier to be
25 CTh. 1.15.2. 26 CTh. 9.19.3. 27 Amm. Marc. 28.6.9.

28 Amm. Marc. 29.5.2. 29 CTh. 1.15.3; see also 1.15.2; contrast CJ 1.33.1.
30 CTh. 16.2.31. 31 Amm. Marc. 21.7.4. 32 Amm. Marc. 29.5.8–9.
tried for refusing to continue to serve in the imperial army, he, too, sent
a letter to the court officials detailing the man’s actions.33 Letter-writing
of this sort appears to have remained important to the functioning of the
army in Africa into the Byzantine period: we have at least fourteen lead
seals struck in the names of sixth- or seventh-century duces and magistri
militum from Africa that would originally have been attached to written
documents issued by these military officials, and in the late sixth century
Gregory the Great wrote a letter recommending the Suevic comman-
der Droctulf for military service in Africa under the Byzantine exarch
Gennadius.34
The use of documentation in the civil and military administration of
Africa was not confined to the writing and circulation of letters. A sense
of the notarial sophistication of which the imperial administration in
Africa was capable can be gleaned from the documentation generated
over the course of the conference held in Carthage in 411 to resolve
the long-running rivalry between the region’s contentious Catholic and
Donatist Churches.35 The conference was convened on the order of the
Western emperor Honorius (395–423) and presided over by the imperial
legate Flavius Marcellinus; the secretarial team that oversaw the record-
ing of the conference’s proceedings was headed by two scribes (scribae)
and four stenographers (exceptores) attached to Marcellinus’ staff and to
those of the proconsul, the vicar of Africa and the curator of Carthage.
The official stenographers recorded the proceedings of the conference
in shorthand, as did two teams of ecclesiastical notaries, one from each
of the rival African Churches. Under the supervision of one of the two
exceptores from the proconsul’s office, the opposing parties checked the
resulting transcripts against each other and resolved any discrepancies
before a finalized text was written out in longhand. The rival Catholic and
33 Passio S. Marcelli Tingitani 4 (ed. H. Delehaye, in ‘Les Actes de S. Marcel le centurion’,

Analecta Bollandiana 41 (1923), 257–87, at 262 (M) and 265 (N)).
34 Seals: P. Delattre and P. Monceaux, in Bulletin de la Société nationale des antiquaires de
France (henceforth cited as BSNAF ) (1908), 164 (no. 1) (= Revue archéologique, 4th
ser., 12 [1908], 449 [no. 184]); (1912), 331–2 (no. 1); (1913), 316 (no. 1); (1914),
284 (nos. 1 and 2); P. Monceaux, ‘Enquête sur l’épigraphie chrétienne d’Afrique’,
Revue archéologique, 4th ser., 2 (1903), 59–90, at 75–6 (nos. 15–16); J. Nesbitt and N.
Oikonomides (eds.), Catalogue of Byzantine Seals at Dumbarton Oaks and in the Fogg
Museum of Art. I. Italy, North of the Balkans, North of the Black Sea (Washington, DC,
1991), no. 6.1; G. Zacos and A. Veglery, Byzantine Lead Seals, I/3 (Basel, 1972), nos.
2885, 2898a–b; F. Icard, ‘Sceaux et médailles de plomb trouvés à Carthage’, BCTH
(1917), 1–18, no. 33; F. Icard, ‘Sceaux et plombs marqués découverts à Carthage’,
BCTH (1927), 479–87, no. 1; and A. Merlin and P. Monceaux, in BSNAF (1915), 300
(no. 1); V. Laurent, Les sceaux byzantins du médaillier vatican (Vatican City, 1962), no.
92. Letter of recommendation: Gregory I, Registrum epistularum 9.9 (CCSL 140A:570);
on this Droctulf’s career, see also Paul the Deacon, Historia Langobardorum 3.18–19 (ed.
L. Capo [Milan, 1992], 148–50), and CIL XI 319.
35 Gesta conlationis Carthaginiensis, anno 411 (CCSL 149A:1–257).
Donatist speakers were asked to confirm the accuracy of their recorded

statements by writing in their own hands the word recognovi (‘I have
recognized’ or ‘I have recalled’) after the words attributed to them, and
only rarely does the surviving account indicate dissent on the part of a
participant.36
The conference of Carthage, of course, was something of an excep-
tional case; but even so, down to at least the early fifth century it was
standard procedure in Africa for court recorders to keep transcripts of
municipal sessions and legal proceedings held before local officials, even-
tually including Christian bishops.37 On at least one occasion we hear
of a proconsul asking that these records be read back to him before he
passed sentence on the accused.38 Moreover, sentences were themselves
typically read aloud from writing tablets.39 For Augustine, too, the gesta
proconsularia and gesta municipalia were not only the tribunals where indi-
viduals could be attacked or interrogated, and where the traditores had
surrendered the scriptures and liturgical instruments of the Church; they
were also the written records left behind by such legal actions, records that
could be accessed and read aloud to prove or disprove what the bishop of
Hippo perceived as the Donatists’ partisan allegations.40 Record-keeping
of the sort employed by state officials at the conference of 411 seems to
36 Composition of the acta: E. Tengström, Die Protokollierung der Collatio Carthaginensis:

Beiträge zur Kenntnis der römischen Kurzschrift nebst einem Exkurs über das Wort scheda
(schedula) (Göteborg, 1962), 7–30; Actes de la conférence de Carthage en 411, ed. S.
Lancel, 4 vols. (Paris, 1972–91), I, 337–53. Oral qualities of the surviving text: Actes, I,
289–327, esp. I, 309–27. Objections: e.g. Gesta conlationis Carthaginiensis 1.67 (CCSL
149A:90).
37 Passio sanctorum Scilitanorum (ed. F. Ruggiero, Atti dei martiri Scilitani: Introduzione,
testo, traduzione, testimonianze e commento [Rome, 1991], 71–4); Acta proconsularia sancti
Cypriani (ed. R. Reitzenstein, Die Nachrichten über den Tod Cyprians: Ein philologischer
Beitrag zur Geschichte der Märtyrerliteratur [Heidelberg, 1913], 12–17); Passio S. Maxim-
iliani (ed. P. Siniscalco [Turin, 1974], 159–61); Passio s. Marcelli Tingitani (ed. Delehaye,
260–4 [M] and 264–7 [N]); Passio sanctae Crispinae (ed. P. Franchi de’ Cavalieri, Nuove
note agiografiche [Rome, 1902], 32–5); Gesta apud Zenophilum (CSEL 26:185–97); Acta
purgationis Felicis episcopi Autumnitani (CSEL 26:197–204); Augustine, Contra Fortuna-
tum disputatio (CSEL 25.1:81–112); Augustine, Contra Felicem (CSEL 25.2:799–852);
C. Humfress, Orthodoxy and the Courts in Late Antiquity (Oxford, 2007), 248–52; and,
in general, G. Lanata, Gli atti dei martiri come documenti processuali (Milan, 1973).
38 Passio s. Crispinae 2 (ed. Franchi de’ Cavalieri, 35).
39 Passio ss. Scillitanorum 14 (ed. Ruggiero, 73–4); Acta proconsularia s. Cypriani 2 (ed.
Reitzenstein, 16); Passio S. Maximiliani 3 (ed. Siniscalco, 160); Passio s. Crispinae 2 (ed.
Franchi de’ Cavalieri, 35).
40 Tribunal: Augustine, Epistulae 76.2–3, 108.5.14, 114 (CSEL 34:327–8, 627, 660); Con-
tra litteras Petiliani 1.21.23, 2.15.35 (CSEL 52:18, 40); De unitate ecclesiae (= Epistula ad
Catholicos de secta Donatistarum) 18.46, 19.50 (CSEL 52:291, 297). Documents: Augus-
tine, Epistulae 88.11 (CSEL 34:418); Contra litteras Petiliani 2.20.45, 2.58.132 (CSEL
52:45, 93); De unitate ecclesiae 12.31 (CSEL 52:270); Contra Cresconium grammaticum
et Donatistam 3.47.51, 3.48.53, 3.56.62, 3.72.84 (CSEL 52:458, 460–1, 468–9, 489);
De unico baptismo 17.31 (CSEL 53:32); see also Epistulae 108.5.14 (CSEL 34:627) and
De unitate ecclesiae 12.31 (CSEL 52:270) (‘gesta forensia’).
have been very much a part of the legal and judicial culture of late Roman
Africa.
Writing in general and letter-writing in particular appears to have been
so central to the conduct of the late Roman and Byzantine administra-
tion of Africa that its misuse was the focus of a certain amount of official
anxiety. In part this surrounded the treasonous thoughts or sentiments
which state officers or their correspondents might commit to writing.
In the fourth century, for example, the introduction as evidence at his
trial of the ‘more private papers’ (secretiores charti) of Hymetius, a for-
mer proconsul of Africa, revealed an embarrassing note written in the
man’s own hand inveighing against the avarice and ferocity of the reign-
ing emperor.41 The papers of Romanus, a contemporary of Hymetius
and the commanding general of Africa, included a letter that hinted at
some of the plots in which the officer was implicated.42 Much later, in the
seventh century, imperial officials would accuse Maximus the Confessor
of having written a similarly treasonous letter to the Byzantine military
commander of Numidia, urging him not to take part in the defence of
Egypt against the Islamic invasion.43 In part, however, such anxieties sur-
rounded the dangers of forgery. By 367 the script employed by the writing
office of the proconsul of Africa had come to imitate that of the imperial
chancery so successfully that the emperor Valentinian I began to worry
that provincial officials might be tempted to forge imperial commands.
The emperor ordered that thenceforth communications from the procon-
sul’s office be written in ‘common letters’ (litterae communes).44 Similarly,
in 641, George, the praetorian prefect of Byzantine Africa, received a let-
ter in support of some Monophysite nuns that had been written in the
name of the empress Martina and which (for his own reasons) the prefect
rejected as a forgery.45 But it was not only letters that aroused such sus-
picions. Early in the fourth century, Constantine had fretted that African
provincials might try to dupe local fiscal officials by faking tax receipts.46
Forgeries of the latter sort could potentially have far-reaching ram-
ifications. Indeed, in all probability the most prominent way in which
the bureaucratic apparatus of the late Roman state made itself felt in
the lives of the vast majority of Africans would have been through the
41 Amm. Marc. 28.1.20. 42 Amm. Marc. 28.6.26.

43 Relatio factae motionis inter domnum Maximum monachum et socium eius coram principibus
in secretario (CCSG 39:15, lines 30–7).
44 CTh. 9.19.3.
45 Maximus Confessor, Epistulae 12 (PG 91:460–509, esp. col. 460); on this George,
see also Epistulae 1, 16, 18, 44–5 (PG 91:364–92, 576–80, 584–8, 641–9), and J. R.
Martindale, The Prosopography of the Later Roman Empire, III, A.D. 527–641 (Cambridge,
1992), s.n. ‘Georgius 50’.
46 CTh. 11.1.2.
extraction of revenue in the form of taxes. The sources that survive to

us suggest that this process must originally have generated an enormous
amount of documentation, most of which has subsequently been lost
or destroyed. At a minimum this would have included the census lists
and tax assessments generated by imperial officials, and the tax receipts
that landholders were expected to deposit with the public registrars.47
On occasion, however, the process would have generated far more in the
way of paperwork. In 365, for example, the new emperor Valentinian I
instituted an elaborate system to ensure the payment of delinquent taxes
by absentee landowners resident in Rome but with property holdings in
Africa. The staff of the vicar of Africa was to draft annual briefs (breves)
indicating the amounts that such landholders owed; representatives of
the urban prefect of Rome and the prefect of the annona were to travel to
Africa to record how much these proprietors had paid; and copies of the
resulting documentation were to be filed with both the praetorian prefect
and the emperor himself.48
Whatever else changed in Africa with the coming of the Vandals, the
regular extraction of revenue certainly continued, at least for a time. To be
sure, the Vandal ruling class was said to have been granted hereditary tax-
free allotments; but Procopius tells us that the property of the Vandals’
Romano-African subjects continued to be heavily taxed.49 Before his
conversion to the monastic life the future bishop of Ruspe, Fulgentius,
had enjoyed a successful career in the Vandal civil service as a procurator
or tax collector.50 Victor of Vita indicates that during the persecution of
484, provisions and wages (annonae et stipendia) were denied to Catholics
serving in the Vandal administration; this would seem to imply that under
normal circumstances African office-holders continued to receive both,
and that the system for procuring them was still in place.51 The fiscus
47 CTh. 11.1.2, 28, 32; 13.5.12. Tax collection in late Roman Africa: C. Saumagne, ‘Un
tarif fiscal au quartième siècle de notre ère’, Karthago, 1 (1950), 105–206; J. T. Peña,
‘The Mobilization of State Olive Oil in Roman Africa: The Evidence of Late 4th-c.
Ostraca from Carthage’, in J. H. Humphrey (ed.), Carthage Papers: The Early Colony’s
Economy, Water Supply, a Public Bath, and the Mobilization of State Olive Oil (Portsmouth,
RI, 1998), 117–238, at 153–65.
48 CTh. 11.1.13.
49 Victor of Vita, Historia persecutionis Africanae provinciae 1.13 (CSEL 7:7), with Pro-
copius, De bello Vandalico 1.5.12–15 (ed. J. Haury [Leipzig, 1962], 333–4). Nature of
these allotments: Y. Modéran, ‘L’établissement territorial des Vandales en Afrique’,
Antiquité tardive, 10 (2002), 87–122; A. Schwarcz, ‘The Settlement of the Vandals in
North Africa’, in Merrills (ed.), Vandals, Romans, and Berbers, 49–57. Transformative
effect on early medieval society: M. Innes, ‘Land, Freedom and the Making of the
Medieval West’, TRHS, 6th ser., 16 (2006), 39–74.
50 Ferrandus, Vita S. Fulgentii episcopi Ruspensis 1 (ed. G.-G. Lapeyre [Paris, 1929], 13).
51 Victor of Vita, Historia persecutionis 2.10 (CSEL 7:27).
also seems to have continued to function under the fifth-century Vandal

regime, and one of the few surviving edicts of the Vandal king Huneric
(477–84) refers to the existence of a domus regia and a decurial order as
well.52 This edict is modelled on earlier Roman anti-Donatist legislation,
and therefore cannot necessarily be taken to reflect contemporary social
reality; but taken as a whole the constellation of evidence would seem to
suggest both that many of the fundamental elements of the late Roman tax
system survived for at least the first sixty or so years of Vandal occupation,
and that they were redirected to serving the interests of Africa’s new
rulers.53
After his reconquest of the Vandal kingdom, the Byzantine emperor
Justinian (527–65) similarly arranged for a new census to be taken in
Africa.54 The region’s much-lauded wealth had been a major factor in
motivating the emperor’s African venture, and down to at least the reign
of Constantine IV (668–85), Justinian’s successors continued to take
a close financial interest in the province. An overwhelming number of
the lead seals that survive from Byzantine Africa concern the region’s
financial administration. Many of them date to the seventh century.
These include the seal of a certain Stephen, honorary consul and dio-
cetes provinciarum, as well as those of three sacellarii and perhaps sixteen
commerciarii.55 In addition, a sixth- or seventh-century funerary inscrip-
tion from Carthage records the burial of one Redemptus ‘arcarius of the
fifth region’ (ar[car]iu[s r]egi/onis qui[ntae]).56 The combination of a cen-
sus, taxes and tax officials empowered to issue sealed documents implies
tax records as well.
The fiscal impositions of the late Roman state probably also provide
the context in which we should understand a Latin ostracon discovered
at Ksar Koutine, north-west of Médenine in southern Tunisia. The text
describes itself as a chitty (pittacium) issued on 17 July 419 in connec-
tion with the assessment of some 1,800 folles on olive oil that was to be
52 Victor of Vita, Historia persecutionis 2.23, 3.11 (CSEL 7:33, 77); J. H. W. G.

Liebeschuetz, ‘Gens into regnum: The Vandals’, in H.-W. Goetz, J. Jarnut and W. Pohl
(eds.), Regna and gentes: The Relationship Between Late Antique and Early Medieval Peoples
and Kingdoms in the Transformation of the Roman World (Leiden, 2003), 55–83, at 75;
but see also Wickham, Framing, 636–7.
53 See, however, Wickham, Framing, 89–92, who makes a strong case for the deterioration
of the classical city-based tax system in the later Vandal period.
54 Procopius, De bello Vandalico 2.8.25 (ed. Haury, 455).
55 P. Delattre and A. Merlin, in BCTH (1924), ccxx–ccxxi (no. 7); C. Morrisson and W.
Seibt, ‘Sceaux de commerciaires byzantins du VIIe siècle trouvés à Carthage’, Revue
numismatique, 6th ser., 24 (1982), 222–40; CIL VIII 22656.25; and Monceaux, ‘Enquête
sur l’épigraphie chrétienne’, 73 (no. 11). We hear in addition of John, sacellarius of the
general of Numidia in Relatio factae motionis . . . (CCSG 39:15, lines 28–9).
56 L’année épigraphique (1986), 259 (no. 717).
sold.57 In all probability, the document is a receipt for the payment of

customs duties at a local town or rural market.58 However, our best evi-
dence for what tax records looked like in the late Roman period comes
from a group of thirty-two Latin ostraca discovered on the Admiralty
Island in Carthage’s circular harbour, recently the subject of a magnifi-
cent edition and commentary by J. T. Peña. The ostraca all seem to date
to the year 373 and form two distinct groups, one concerned with the
receipt and inspection of olive oil shipments sent to Carthage, and the
other concerned with the weighing and storage of oil to be exported from
Carthage as part of the region’s tax levies. The ostraca were made from
locally produced amphorae and most were cut to a fairly standard size
and shape, with the result both that the sherds could be held firmly in
the palm of one hand and that they could be easily archived.59 These
ostraca provide a vivid glimpse of the mundane bureaucratic functioning
of the late Roman state. The group concerned with the receipt of oil
shipments for the most part give the date, the name of the shipmaster or
the owner of the vessel delivering the oil, and the amount of oil accepted
and rejected, all on one side of the sherd; on the other side, much the
same information is provided in more summary form, though with the
addition of the name and title of the official receiving the oil shipment.
The group of ostraca concerned with the weighing and storage of oil to
be exported similarly provide (with some variation) the date, the names
of oil-weighers and amounts of weighed oil, the aggregate sum of these
figures, plus a ‘to do’ (fieri) list with estimates of further quantities of
oil. As with the records of delinquent taxes that Valentinian I ordered
to be drawn up, copies of these records (or a summary of them) were
presumably sent to the appropriate imperial authorities.60
Shaping African documentary practice

The heavy reliance of the late Roman administration and its successors
on written records probably shaped African documentary practice in a
57 Merlin, in BCTH (1913), ccxxxi (no. 1): ‘pc dd nn honorio xii / et theudosio octies /
xvi kal aug aurel . . . / et iu..ntes sasucthani / constitui estinasse ex pre/tium olei quod
uendes / perienslume folles / mille octingentas tnt. es (?) / estum pitacium / emisi’.
58 Jones, Later Roman Empire, 429–30, 825–6. Rural markets in North Africa: B. Shaw,
‘Rural Markets in North Africa and the Political Economy of the Roman Empire’,
Antiquités africaines, 17 (1981), 37–83; L. de Ligt, Fairs and Markets in the Roman Empire:
Economic and Social Aspects of Periodic Trade in a Pre-Industrial Society (Amsterdam,
1993), 119–22, 134–5, 155–96; D. Kehoe, Investment, Profit, and Tenancy: The Jurists
and the Roman Agrarian Economy (Ann Arbor, MI, 1997), 209–13.
59 Peña, ‘Mobilization of State Oil’, 120–2, 128, 130.
60 Peña, ‘Mobilization of State Oil’, 122–52.
number of important ways. Local African aristocrats pursued careers in

service to the imperial and, later, Vandal regimes; and though we have less
evidence for the Byzantine period, it seems likely that at least the lower-
level functionaries who staffed the administration were generally of local
origins.61 Thus, as Peter Sarris has argued for Egypt, the language and
techniques of the bureaucracy will also have provided high-status Africans
with their most natural means of expression. As a result, in Africa, too,
it can be difficult to distinguish in the surviving sources between ‘public’
and ‘private’ documents. For example, an early Byzantine-era ostracon
from Négrine, Algeria, reads:
anno XVI domni / nostri iustiniani in/peratori extima/tus fuit laudeti in / portione
dominica / oliariu arcariu / unu tantum.62
In the year 16 of our lord emperor Justinian [AD 542×3] one oil arcarius only
was assessed at Laudeti on the lord’s portion (in portione dominica).
The text thus records the assessment of oil on the portio dominica of a local
estate. This reference to assessment might suggest that the document
is a tax receipt, but to judge from the ostracon’s dating formula it is
probably not an official document. The Négrine text is dated according
to Justinian’s regnal year only; but in 537, five years before this record
was composed, the emperor had mandated that documents also be dated
according to the consular year, indiction, month and day.63 To be sure,
the choice of a small (but carefully manufactured) 5.5-cm-wide ostracon
and the decision to write the text of the Négrine document without
abbreviations made the inclusion of the other imperially mandated dating
elements impractical; but these same decisions probably also imply that
this document was private in character.
It is worth emphasizing, however, that the production and use of doc-
uments was hardly the exclusive preserve of the administrative elite.
Indeed, the physical act of writing was not a high-class pursuit in the
later Roman world.64 The scribes who drew up the surviving African
documents engaged in a diverse array of professions, and though some
of them probably figured among the notables of their local commu-
nities, none appear to have been of particularly elite status. Between
nine and eleven writers produced the Albertini Tablets. They included
the local priest, Saturninus, as well as two magistri, probably local
schoolteachers – both of them humble occupations in the later Roman
61 M. Overbeck, Untersuchungen zum afrikanischen Senatsadel in der Spätantike (Kallmunz,

1973), 23–40, 55–74, 83–8.
62 Albertini, ‘Ostrakon byzantin de Négrine’, 54, with commentary at 54–8.
63 J.Nov. 47.1. 64 Jones, Later Roman Empire, 515.
countryside.65 A fourth scribe, named Donatianus, identified himself

simply as a ‘citizen’ (cibis) of Cappraria, probably another estate neigh-
bouring the Fundus Tuletianos. Together with a certain Saturninus, this
Donatianus was the co-owner of a 6-year-old slave boy, who was sold
to Geminius Felix, ‘citizen of Tuletianos’ (cibis tuletianensis), on 5 June
494.66 Yet another of the scribes, Paul, was at least a moderately well-
off smallholder whom we meet selling a plot of fifteen olive and fifteen
fig trees.67 The remaining writers are, at best, little more than names
to us. The scribes who wrote up the contemporary Bir Trouch ostraca
are not even that. The Massiese that they mention seems not to have
drawn up these documents himself, though, or at least not all of them.
Indeed, these five ostraca appear to have been written by five different
scribes: though the writing on all of the sherds is similar, no two texts
seem to have been written by the same hand.68 Significantly, at least
two of these scribes were active in the same year.69 The responsibility
for drawing up texts of this sort would thus appear to have been shared
between numerous writers, presumably estate officials of some sort. The
Jerba ostraca, too, seem to have been written up by numerous different
scribes, though these texts otherwise betray little about those who wrote
them up.70 Soldiers also show up as writers in a number of cases. In the
third century Roman military recruiters in Africa appear to have kept
written lists of recruits, and the fourth-century Admiralty Island ostraca
seem to have been drawn up by individuals with military training.71 The
same, of course, was true of the voluminous daily reports, lists and letters
produced at local garrisons like that at Bu Njem – despite some of the
Punic-speaking recruits’ apparently low levels of literacy and imperfect
command of the Latin language.72
65 See Conant, ‘Literacy and Private Documentation’, 203–4, 223–4, but also L. Dossey,
Peasant and Empire in Christian North Africa (Berkeley, CA, 2010), 260n153 (on magister
as ‘schoolmaster’). Status of rural priests and schoolteachers: Jones, Later Roman Empire,
908–9, 997.
66 TA II, 1:217; the name is common in Africa, but he could conceivably be the same
Donatianus who had a field on the Fundus Tuletianos: TA XXIV, 1:283.
67 TA XXVI, 1:288. 68 Bonnal and Février, ‘Ostraka de Bir Trouch’, 246.
69 Bonnal and Février, ‘Ostraka de Bir Trouch’, 239–42 (nos. 1–2).
70 Várhelyi and Bagnall, ‘Ostraka’, 341, 343.
71 Passio S. Maximiliani 3 (ed. Siniscalco, 160): ‘Sterne nomen eius’; J. Godfrey, ‘Who
Wrote the Ostraka from the Ilôt de l’Amirauté, Carthage?’, in Merrills (ed.), Vandals,
Romans, and Berbers, 181–98.
72 R. Marichal, Les ostraca de Bu Njem (Tripoli, 1992), with J. N. Adams, ‘Latin and Punic
in Contact? The Case of the Bu Njem Ostraca’, Journal of Roman Studies, 84 (1994), 87–
112, and, in general, Adams, Bilingualism and the Latin Language (Cambridge, 2003),
213–30; however, see also Adams, ‘The Poets of Bu Njem: Language, Culture, and the
Centurionate’, Journal of Roman Studies, 89 (1999), 109–34.
The individuals who used the documents that these scribes produced
also appear to have represented a wide range of social statuses. Our best
evidence naturally concerns the more prosperous elements of late antique
society. These would include the merchants for whom the Jerba ostraca
were produced. The Massiese from the Bir Trouch ostraca may well him-
self have been an estate owner, and, as we have seen, the Severus to
whom a southern Tunisian ostracon was addressed was a conductor.73 The
Albertini Tablets similarly seem to preserve what remains either of an
estate archive or of the private dossier of Geminius Felix and his family;
the Geminii may have run the Fundus Tuletianos, and in any case they
were likely kinsmen of the estate’s owner, the civic priest Flavius Geminus
Catulinus.74 However, the use of documents also seems to have extended
deeper into society than these relatively elevated examples would suggest.
The peasant cultivators of the Fundus Tuletianos certainly appear to have
been familiar, even comfortable, with the culture of written documenta-
tion that surrounded them. In addition to the nine or more scribes that
the community produced, at least four of the twenty-five men who sold
land or other property in this community were able to sign a document
with an autograph formula indicating that a sale had taken place and
that the selling price had exchanged hands. This is not a high number,
in either absolute or proportional terms; but in the same area, twenty-
four local witnesses could be found to authenticate such a document in
their own hands.75 Moreover, if the Albertini Tablets constituted a family
rather than an estate archive, then there may also be some slight evidence
to suggest that both buyers and sellers of property received a copy of the
deed of sale: a number of these instrumenta are palimpsests, and in one
or two cases documents in which one of the Geminius brothers was a
seller were scraped down and reused to draw up acts in which another
of the brothers was the buyer.76 In this case, the Albertini Tablets may
simply have been the most extensive (and best preserved) of dozens of
similar dossiers maintained in the Djebel Mrata region by estate owners,
managers and tenants alike.
Thus, service in the imperial (or royal) civil administration was only
one of the many factors that will have shaped documentary practices
73 Massiese: Wickham, Framing, 266. Severus: see above, n. 6.

74 D. J. Mattingly, ‘Olive Cultivation and the Albertini Tablets’, in Attilio Mastino (ed.),
L’Africa romana: Atti del VI convegno di studio (Sassari, 1989), 403–15, at 404.
75 Conant, ‘Literacy and Private Documentation’, 203–10.
76 TA, tablets 1a and 18a, 1:215, 245, where the position of the names near the beginning
of the act suggests that they were sellers, not buyers. Brothers: Cresconius and Felix
were both sons of Fortunus: TA, XV, 1:266 (Cresconius); TA II, III, XXIX, XXXI,
1:217, 218, 293, 295 (Felix).
in late antique North Africa. Military service was another; interaction

with officialdom and its expectations of written documentation was a
third. As Wickham has pointed out with respect to the late Roman West
in general, it seems likely that ‘the main stimulus, and model, for pri-
vate accounting was always the record-keeping that taxation depended
upon’.77 It is worth pointing out, though, that African officials imposed
other documentary expectations on their region’s inhabitants as well. As
part of the Vandal kings’ efforts to secure the conversion to Arianism of
their kingdom’s Catholic subjects, for example, those Africans who con-
sented to be rebaptized were issued with written certificates attesting to
their actions. These documents functioned as passports that allowed their
bearers to travel without molestation within the territory of the Vandal
kingdom. By contrast, those who refused rebaptism and thus lacked the
appropriate certificate were by necessity forced to avoid the public roads
in their movements; if they were discovered, they could be subjected to
beatings and other indignities.78
Finally, the production, keeping and use of documents in Africa were
probably encouraged by the region’s legal culture. An inscription from
Timgad (mod. Thamugadi, Algeria) dating to the reign of the emperor
Julian (360–3) sets out the schedule of fees that local litigants could
expect to pay to court stenographers, among other officials, in pursuit
of their cases. The fees included 5 modii of wheat (or the equivalent) to
draw up a formal legal claim (postulatio simplex); 12 for the presentation
of the litigant’s case in court (contradictio); and 20 for a ‘finished case’
(definita causa), which presumably included a transcript of the argument
as well as the final judgement. Litigants also had to pay for the papyrus
(carta) used to write up these documents, and in order to limit expenses
the quantities were restricted to one large roll for the postulatio simplex,
four for the condradictio and six for the definita causa.79 Moreover, as
was increasingly the case throughout the later Roman world, written
documentation could also be entered into evidence in the course of a trial
or official hearing in Africa.80 At the centurion Marcellus’ trial during the
reign of Diocletian, for example, a letter from his regimental commander
was read out detailing the officer’s refusal (as a Christian) to continue
77 Wickham, Framing, 268.

78 Victor of Vita, Historia persecutionis 3.47 (CSEL 7:95); Ferrandus, Vita Fulgentii 6 (ed.
Lapeyre, 35–9).
79 CIL VIII 17896; Jones, Later Roman Empire, 497; A. Chastagnol, L’album municipal de
Timgad (Bonn, 1978); J. Harries, Law and Empire in Late Antiquity (Cambridge, 1999),
100; C. Kelly, Ruling the Later Roman Empire (Cambridge, MA, 2004), 138–42.
80 Harries, Law and Empire, 108–9.
serving in the Roman army.81 At a hearing in 320 before an official named

Zenophilus, no fewer than seven letters were read out as evidence, as were
the formal records (acta) of an earlier enquiry held before the curator
of Cirta (mod. Constantine, Algeria).82 The acta of another case, this
one tried in Carthage in 314, were read out at the legal proceedings that
vindicated Bishop Felix of Aptunga of the charge of having been a traditor
during the Great Persecution.83 Of course, such written documents could
also be doctored, as – under threat of torture – a certain clerk by the
name of Ingentius admitted to having done to these very acta.84 The
comment of a Catholic advocate named Apronianus, who was present
for Ingentius’ confession, suggests that altering official documents was a
familiar problem in early fourth-century Africa. ‘It is not new to do this,’
he remarked – adding, as a barb, ‘for them,’ by which he meant Donatists
and their supporters.85 Apronianus was scoring a rhetorical point in what
was both a legal and a sectarian debate; but given the persistent fear of
forgery that we encounter in Late Antiquity, his allegation may not have
been entirely unfounded.86
Record-keeping and archives

If public records could not entirely be trusted, Africans may have con-
sidered it all the more essential to retain their own copies of documents
that they deemed to be particularly valuable. It is worth stressing that
these never seem to have been the only kinds of documents generated
in the region: even as late as the mid seventh century, for example, we
hear of an African who ‘wrote down on a leaf of parchment’ (‘egrapsen en
chartēi’) the directions that people gave him to a particular church that
he wanted to visit in Constantinople.87 However, the ostraca and wooden
tablets that survive to us from late antique Africa seem for the most part
to involve either substantial amounts of money or property of some sig-
nificance. Though individual acts in the Albertini Tablets could involve
fairly small-scale transactions, the surviving assemblage as a whole estab-
lishes title to 178 olive, 61 fig, 4 almond and 2 pistachio trees; an olive
press; a dowry worth at least 750 olive trees and a slave worth another
81 Passio s. Marcelli Tingitani 4–6 (ed. Delehaye, 262 [M] and 265 [N]).
82 Gesta apud Zenophilum (CSEL 26:186–8, 189–92).
83 Acta purgationis Felicis episcopi Autumnitani (CSEL 26:199–200).
84 Acta purgationis Felicis episcopi Autumnitani (CSEL 26:203).
85 Acta purgationis Felicis episcopi Autumnitani (CSEL 26:203): ‘nec nouum est illis hoc
facere’.
86 See above, nn. 44–6, and also CJ 4.21.20.
87 Miracles of St Artemios 4 (ed. and trans. V. S. Crisafulli and J. W. Nesbitt [Leiden, 1997],
82–4).
65.88 To judge from a table of calculations jotted down on the back of

one of the extant acts, other now-lost tablets once existed that recorded
purchases of another 3 solidi, 3,155 folles worth of property, or, in real
terms, the equivalent of between 262 and 290 additional olive trees.89 By
contrast, D. J. Mattingly notes that ‘the vast majority of modern Tunisian
peasant families specialising in olive cultivation has owned between only
20 and 100 olive trees’.90 It is harder to assess the value of the barley
mentioned in the Bir Trouch ostraca, though the quantities involved var-
ied widely from year to year: in 485×6, for example, the holding at Viginti
Boguli yielded 26 measures, but seven years later it produced only 3.91 At
1,800 folles, however, the Ksar Koutine ostracon (which probably records
a customs payment on olive oil to be sold at an early fifth-century town
or rural market) also involved a considerable sum. Though the value of
the follis was severely undermined by the inflation of the bronze coinage
over the course of the fourth and fifth centuries, in early fifth-century
Africa 1,800 folles would perhaps have been the equivalent of a little over
1 gold solidus.92 If the text does represent a customs payment, and if that
payment was levied at the canonical rate of 2 to 21/2 per cent, then the
total value of the oil to be sold may have been as much as 90,000 folles, or
upwards of 50 solidi.93 In this case, the figures involved would doubtless
have made the chitty valuable enough for the vendor to retain, at least
for a time.
Of course, valuable documents – both private and public – needed to
be stored somewhere. Most institutions in Africa appear to have main-
tained their own files in archives of one sort or another. The African
countryside was peppered with municipalities of widely varying sizes,
and, as we have seen, major ones (at least) like Carthage, Cirta and
Timgad maintained records of their official acts. Augustine clearly had
access to such gesta municipalia, and he assumed that his adversaries
did as well.94 The bishop of Hippo also makes numerous references
to public archives (archiva publica) in general, and to the proconsular
archive (archivum proconsulis or archiva proconsularia) in particular. Such
88 In the Tuletianos community, olive trees normally sold for about 15–16 folles each:
Grierson, ‘The Tablettes Albertini’, 74. The slave Fortinus was sold for 1 solidus, 700
folles (TA II, 1:217); Januarilla’s dowry was valued at 12,000 folles (TA I, 1:215).
89 TA, XXXIII (tablet 29b), 1:299 written on the back of TA XV (tablet 29a), 1:267. The
sum of 3 solidi, 3,155 folles does not include the sums from TA XXXIII lines 5–7, 10–11,
15, 19–20, or 22, which appear to record transactions recorded in the extant acts.
90 Mattingly, ‘Olive Cultivation’, 407n19.
91 Bonnal and Février, ‘Ostraka de Bir Trouch’, 239–41, 244–5 (nos. 1, 4).
92 A. H. M. Jones, ‘The Origin and Early History of the Follis’, Journal of Roman Studies,
49 (1959), 34–8, esp. 37 for early fifth-century Africa.
93 Jones, Later Roman Empire, 825. 94 See above, n. 40.
repositories, he wrote, were full of Catholic protests against Donatists’

acts of violence, and of course they also contained records of the accu-
sations levied and court cases tried before the proconsul.95 In addition,
the public archive preserved Constantine’s judgement in favour of the
controversial appointment of Caecilian as bishop of Carthage.96 At the
conference at Carthage in 411, the Catholic bishops cited a report (rela-
tio) from the proconsul Anulinus to the emperor Constantine dating to
313; when the Donatist bishop Petilianus asked whether it came from a
public or private cache of documents, Augustine first tried to avoid the
question, and then claimed that this text, too, could be found in the pro-
consular archive.97 Elsewhere, when quoting part of the fourth-century
decision of the proconsul Aelianus exonerating Felix of Aptunga of the
charge of traditio, Augustine told his readers, ‘If you want to read the
whole gesta, get it from the archive of the proconsul.’98 Thus, in addi-
tion to the province’s tax accounts and the imperial edicts and rescripts
sent to its leading officials, the state archives in late Roman Africa would
appear to have contained copies of important communications sent to
the emperor, records of court trials held before the proconsul and other
legally significant documents. On the orders of Constantine, the latter
were to include samples of illicit defamatory writings circulated in Africa,
which the offices of the proconsul and the vicar were to keep on file.99
In 414, three years after the disputation between the leading Catholic
and Donatist bishops at Carthage, the emperor Honorius indicated in a
rescript to the proconsul of Africa that the proceedings of that council
were deemed to have permanent validity and that they, too, had been
transferred to the public records (publica monumenta).100
In the centuries following the Vandal conquest, the impression that
such public record-keeping leaves in our sources becomes fainter. In the
midst of the Vandal king Huneric’s persecution of the Catholics in 484,
for example, a local bishop by the name of Habetdeum is said to have
told his Arian tormenters that he had drawn up a gesta against them in his
heart, which would seem to suggest that Africans continued to be famil-
iar with the concept of transcribing judicial procedures.101 Whether the
95 Augustine, Epistulae 129.4 (CSEL 44:36–7); Augustine, Contra Cresconium 3.45.49,

3.61.67 (CSEL 52:455–6, 473–4).
96 Augustine, Contra partem Donati post gesta 16.20 (CSEL 53:118–20).
97 Gesta conlationis Carthaginiensis 3.216–20 (CCSL 149A:232–3); Augustine, Breviculus
conlationis cum Donatistis 3.7.8 (CCSL 149A:276). Augustine cited the same relatio
earlier in his letter to the Donatist bishop Januarius: Epistulae 88.2 (CSEL 34:408).
98 Augustine, Contra Cresconium 3.70.80 (CSEL 52:485): ‘si tota gesta uis legere, ex
archiuo proconsulis accipe’.
99 CTh. 9.34.2. 100 CTh. 16.5.55.
101 Victor of Vita, Historia persecutionis 3.46 (CSEL 7:94–5).
practice continued to be characteristic of Vandal-era realities is less clear,

though it seems plausible. To be sure, Habetdeum’s claim, ‘I have drawn
up a gesta in the headquarters of my heart’ (in praetorio cordis . . . gesta
confeci), is likely to be a play on the widespread concept that one could
bear beliefs or memories in that same organ (in corde gestare); but, per-
haps significantly, the distinctive turn of phrase that the bishop was said
to have employed does not appear to have been proverbial in late ancient
Latin.102 In either case, the old state archives likely continued to function
for as long as the Vandals extracted taxes from their Roman subjects. Vic-
tor of Vita certainly had access to the texts of the only three Vandal royal
laws that survive to us, though, of course, copies of these could just as
well have been preserved in the archives of the Carthaginian Church.103
More telling is the fact that the specific provisions of Huneric’s edict of
persecution were modelled on the western emperor Honorius’ rescript of
412 ordering the suppression of Donatism. This strongly suggests that,
over seventy years on, the bureaucrats who staffed the Vandal royal writ-
ing office not only continued to maintain old Roman imperial laws on file,
but also were still able to access them when necessary.104 In the Byzantine
period, Justinian deplored the state of provincial archives in general, but
at a minimum the newly reimposed fiscal regime would probably have
required a functioning method of storing and retrieving tax records in
Africa, and it is difficult to believe that the new flow of imperial rescripts
to the province would not also have been preserved locally.105
The need to access important documents probably also meant that
it was not unusual for private individuals to maintain personal or fam-
ily dossiers in late antique North Africa. We have already seen that the
proconsul Hymetius and the general Romanus kept private papers that
102 In corde gestare: see, e.g., Augustine, Enarrationes in psalmos 34.2.8, 34.2.11, 51.11,
80.10, 88.1.7, 90.2.9 (CCSL 38:318, 319; CCSL 39:631, 1125, 1225, 1276); Augus-
tine, Sermo 24.3, 199.1.2, 288.4 (CCSL 41:327; PL 38:1027, 1306). See also (corde
gestare), e.g. Augustine, De doctrina christiana 1.13.12, 2.18.28 (CCSL 32:13, 53);
Augustine, De sermone domini in monte 2.2.5 (CCSL 35:95); Augustine, In Iohannis
evangelium tractatus 3.2 (CCSL 36:20); Augustine, Sermo 37.7, 62.1.1, 169.13.16,
187.3 (CCSL 41:454; PL 38:415, 925, 1002).
103 Laws: Victor of Vita, Historia persecutionis 2.3–4, 2.39, 3.3–14 (CSEL 7:25, 39, 72–8).
On African ecclesiastical archives: Augustine, Breviculus collationis 3.17.31–2 (CCSL
149A:296–7) and Augustine, Contra partem Donati post gesta 15.19 (CSEL 53:117). A
single, much-abraded ostracon written in a fourth-century cursive found in the basilica
at Skhira, Tunisia, may be the only document to survive from such an archive: M.
Fendri, Basiliques chrétiennes de la Skhira (Paris, 1961), 56–7.
104 Victor of Vita, Historia persecutionis 3.3–14 (CSEL 7:72–8); CTh. 16.5.52. On the
symbolic importance of written law to early Germanic kingship, see P. Wormald, ‘Lex
scripta and verbum regis: Legislation and Germanic Kingship, from Euric to Cnut’, in
P. H. Sawyer and I. N. Wood (eds.), Early Medieval Kingship (Leeds, 1977), 105–38.
105 J.Nov. 15 praef.; see also 36–7 and App. 2, 3, 6, 9 (rescripts to Africa).
included their correspondence; the legislation of Valentinian III (425–

55) assumed that fifth-century African landowners kept records detail-
ing their property holdings, though the emperor also imagined that such
documentation might have been lost or destroyed in the disorder of the
Vandal invasion.106 Nearly a century later, Justinian thought it entirely
plausible that the descendants of African refugees living in the East might
still have access to the documents they would need to prove their title to
lands expropriated by the Vandals.107 In Gigthis, four unedited ostraca
have been found in a private house neighbouring on the forum, and they
may well hint at a private archive that was once much larger.108 As we
have seen, the Albertini Tablets and the Bir Trouch ostraca would seem to
represent portions, at least, of similar assemblages. The latter two collec-
tions also provide valuable evidence that, in times of disorder, Africans
did what they could to protect their documents: both the ostraca and
the tablets were found in earthenware jars, where they had presumably
been stashed for safe keeping.109 At least, the presence of dorsal notes
on the Albertini Tablets, summarizing the acts contained on the polyp-
tychs’ interior faces, would seem to suggest that these documents were
not normally kept in a pot.110
Barring disorder, documents could be maintained for wildly variable
periods of time. Even when the later Roman emperors remitted tax
arrears, they continued to lay claim to what was owed the state from
on average the previous five or six years, suggesting that the fiscal author-
ities generally maintained good records for at least that long.111 These
general numbers correspond well with the specifically African evidence:
in 410 the western emperor Honorius remitted all delinquent taxes for
the region down to 405, and in 414 he forgave African navicularii what
they owed down to 409.112 As was often the case in instances of this sort,
Honorius’ rescript of 410 also ordered the destruction of old chartae ren-
dered obsolete by his remission.113 Similarly, neither of our two datable
collections of private records from late antique Africa suggests that such
documents were preserved for more than a decade. As we have seen,
some of the Albertini Tablets were scraped down and reused in as little
as two years, and the extant assemblage as a whole spans only the three
106 Valentinian III, Nov. 12, esp. 12.2.

107 J.Nov. 36.1. 108 P. Gauckler, in BCTH (1903), ccvi.
109 Bonnal and Février, ‘Ostraka de Bir Trouch’, 239.
110 Conant, ‘Literacy and Private Documentation’, 215.
111 CTh. 11.28.3, 9, 16; Marcian, Nov. 2; Valentinian III, Nov. 1.1, 3; Maiorian, Nov. 2;
J.Nov. 147.1, 148.1; and Jones, Later Roman Empire, 467.
112 CTh. 11.28.6, 8. 113 CTh. 11.28.6; see also (in general) CTh. 11.28.2, 3, 14.
years between 13 March 493 and 21 April 496.114 The Bir Trouch ostraca
collectively cover only eight years.115 On the other hand, it would seem
that at least some privately held documents could enjoy a considerably
longer life: a rescript of Honorius addressed to the praetorian prefect
of Africa tried to put an end to lawsuits against shipmasters based on
50-year-old documents.116
Finally, in late antique Africa, writing of a different sort was also often
on what was probably intended to be undying public display. In the
late Roman period, the Christian emperors imagined that their written
edicts – especially those on the unity of the faith – would be
posted in prominent, heavily frequented places throughout the African
provinces.117 In the immediate wake of the Byzantine reconquest of
Africa, new or restored fortifications were dedicated with inscriptions
in both Latin and Greek. Both such uses of writing served to project
imperial power; the bilingual military inscriptions also made a potent
statement about the permanence of Justinian’s reintegration of Africa
into the Roman respublica.118
Even before the Byzantine reconquest, though, late antique Africa had
been enormously productive of inscriptions. After Rome itself, more
inscriptions survive from Carthage than from any other city in the
West; indeed, the African metropolis alone has preserved nearly as many
inscriptions as has the whole of Gaul. Provincial towns such as Hadrume-
tum (mod. Sousse, Tunisia), Ammaedara (mod. Haı̈dra, Tunisia) and
Altava (mod. Ouled Mimoun, ex-Lamoricière, Algeria) also produced
epigraphic corpora whose surviving elements run into the hundreds.119
114 Scraped down: see above, n. 76, and in general Conant, ‘Literacy and Private Docu-
mentation’, 212–13. Earliest act: TA XVI, 1:269 (13 March 493). Latest act: TA XXIV,
1:283 (21 April 496).
115 The earliest dates to Gunthamund’s second regnal year; the latest two, to his ninth:
Bonnal and Février, ‘Ostraka de Bir Trouch’, 239–42, 244–5 (nos. 1–2, 4).
116 CTh. 13.6.10.
117 CTh. 16.5.37, 16.11.2. That few survive has to do with the fact that texts of this sort
were generally inscribed in bronze, much of which was subsequently melted down:
e.g. S. Hornblower and A. Spawforth (eds.), The Oxford Classical Dictionary, 3rd edn
(Oxford, 1996), s.n. ‘epigraphy, Latin’ (at p. 544); G. Woolf, ‘Literacy’, in A. K.
Bowman, P. Garnsey and D. Rathbone (eds.), The Cambridge Ancient History. XI. The
High Empire, A.D. 70–192 (Cambridge, 2000), 875–97, at 886–7.
118 J. Durliat, Les dédicaces d’ouvrages de défense dans l’Afrique byzantine (Rome, 1981),
7–91; D. Pringle, The Defence of Byzantine Africa from Justinian to the Arab Conquest:
An Account of the Military History and Archaeology of the African Provinces in the Sixth
and Seventh Centuries (Oxford, 1981; repr. 2001), 315–32.
119 M. A. Handley, Death, Society, and Culture: Inscriptions and Epitaphs from Gaul and
Spain, AD 300–750 (Oxford, 2003), 18. See, however, R. Duncan-Jones, The Economy
of the Roman Empire: Quantitative Studies, 2nd edn (Cambridge, 1982), 360–2, on the
rate of inscription survival.
Many of these inscriptions are epitaphs commemorating secular mem-

bers of the regional elite: men and women of senatorial status, lawyers,
doctors, goldsmiths, ships’ captains and other local notables and their
families.120 Such inscriptions are revealing of the documentary culture
of Africa in two particular ways. First, occasional mistakes – like that in
an epitaph from Hippo Regius which reads, ‘here lies the body of the boy
(to be named)’ Hic iacet corpus pueri nominandi – would seem to indicate
that funerary inscriptions were produced by workshops working from
model books.121 Second, such inscriptions indicate the importance with
which prominent secular families continued to regard the commemo-
ration of their beloved departed in writing. This public display does
not necessarily imply anything about the ability of African provincials,
either individually or collectively, to read or write; but it does say some-
thing meaningful about the ways in which the written word was used
to express power, authority, social prominence and a sense of imperial
and family permanence in Africa, and indeed throughout the late Roman
world.
Conclusions
Of course, Africa, like the eastern Mediterranean, is something of a spe-
cial case. Here, too – in marked contrast to the situation in much of the
early medieval West – a bureaucratic class survived down to c. 700, and
on the face of it, it seems likely that the strength of Africa’s documentary
culture owed something to this survival. Yet the very contrast between
Africa and the north-western Mediterranean (let alone north-western
Europe) raises two closely related questions about developments in doc-
umentary practice in the early medieval West in general. First, Africans
made use of the written word to record a remarkable array of social activ-
ities. These included estate management and bookkeeping; the sale, gift
and endowment of property; the entering into and discharging of debt;
paying taxes; appealing to the emperor; contracting marriages, recording
rebaptisms and choosing epitaphs; pursuing legal claims; recommending
protégés to official appointments; and getting directions. A Byzantine-
era inscription from the military installation at Thabraca (mod. Tabarka,
Tunisia) even reserves a horse for the use of a certain Michael, a courier
of the cursus publicus.122 What is more, Punic appears to have been used
120 Handley, Death, 37. Lawyers: G. Charles-Picard, in BCTH (1950), 74–89, at 88 (=

Revue archéologique, 6th ser. 37 [1951], 173–4 [no. 45]); CIL VIII 25837 (= Inscriptiones
Latinae selectae, ed. H. Dessau, 3 vols. [Berlin, 1892–1916], 2.1:414 (no. 5731]).
121 Revue archéologique, 4th ser., 34 (1931), 367 (no. 112); Handley, Death, 26.
122 Pringle, Defence, 338 (no. 56), with discussion at 230–1.
alongside Latin in the third-century Bu Njem ostraca, and the language

was also employed in ostraca from Silin and Wadi el-Amud in Libya as
well as from Henchir Khanefi, on the borderlands south-west of Gabès,
Tunisia. Punic was also used as an epigraphic language in late antique
Tripolitania, as, for example, in the fourth-century Christian catacombs
at Sirte.123 This, in turn, would seem to suggest that a culture of writ-
ten documentation had penetrated African society to such an extent
that it had even transcended the most ‘Romanized’, Latin-speaking seg-
ments of the population in Late Antiquity. Indeed, recourse to written
instruments was so deeply embedded in local society that, as late as the
490s, it is still discernible at a range of social levels, from the wealthy
elite to the prosperous rural peasantry. The use of documents, in other
words, appears to have been all-pervasive in late antique North Africa –
which raises the question of whether the same was true elsewhere in
the West.
Second, Africa’s documentary culture was underlain and probably
strengthened by the practices of the late Roman state and its succes-
sors, which required or assumed the use of the written word not only
at the highest levels of the civil administration, but also in the army,
in the courts and – critically – in the extraction of taxes. This culture
was thus built on complex foundations whose component parts rein-
forced one another, with the result that the edifice as a whole could
probably survive the weakening of one or another individual element,
at least for a time. Just how long is not completely clear, but references
to the indiction in the Négrine and Henchir el Abiod ostraca show that
on some level the use of documents in estate management survived in
Africa into the Byzantine period.124 The fact that the imperial collection
of taxes continued in this region well into the seventh century – and in
fact seems to have intensified late in the reign of the emperor Constans II
(641–68) – probably provided African landowners with an incentive to
document how much produce they got out of their estates all the way
123 Marichal, Ostraca de Bu Njem, no. 146, with Adams, ‘Latin and Punic’, 88–9;
P. Berger and P. Gauckler, in BCTH (1902), clxxvi; R. Bartoccini, ‘Scavi e rinven-
imenti in Tripolitania negli anni 1926–1927’, Africa italiana, 2 (1928), 187–200, with
K. Jongeling and R. M. Kerr (eds.), Late Punic Epigraphy: An Introduction to the Study
of Neo-Punic and Latino-Punic Inscriptions (Tübingen, 2005), 71–4 (Sirte) and, in gen-
eral, 105–6. The most important collections of late Punic inscriptions include R. G.
Goodchild, ‘La necropoli Romano-Libica di Bir ed-Dréder’, Quaderni di archeologia
della Libia, 3 (1954), 91–107; F. Vattioni, ‘Glosse Puniche’, Augustinianum, 16 (1976),
505–55, at 536–55; and the material cited in J. M. Reynolds and J. B. Ward-Perkins,
Inscriptions of Roman Tripolitania (Rome, 1952), 10–13.
124 Albertini, ‘Ostrakon byzantin de Négrine’, 54; Cagnat, BCTH (1908), ccxlviii.
down to the Islamic conquest.125 Yet, without the fiscal, administrative,

military and legal imperatives of the late Roman state, a culture of written
documentation is likely to have survived in the West only for as long as
merchants, property owners, soldiers, litigants and possibly even rural
tenants found it to be useful. How long that was will be the subject of
the rest of this book.
125 Collection of taxes in seventh-century Africa: see above n. 55; M. McCormick, Origins
of the European Economy: Communications and Commerce, A.D. 300–900 (Cambridge,
2001), 102.
4 Lay documents and archives in early
medieval Spain and Italy, c. 400–700
Nicholas Everett
Late Roman governmental institutions and imperatives had a profound impact

on documentary culture in the West, as well as on the nature of our evidence.
Changes in these institutions and imperatives – driven to a large degree by fiscal
needs but also by changes in the nature of the empire itself – made documents
more important and encouraged their use by individuals both lay and clerical.
The traces left by these developments can be seen particularly well in late and
post-Roman Italy and Spain, where the legal and documentary inheritance
of the late Roman Empire, adapted to meet local conditions, remains visible
at all levels of society. Visigothic Spain demonstrates how late Roman legal
imperatives concerning documents were adopted and developed in a barbarian
kingdom, as well as providing a model for reckoning with our loss of surviving
evidence elsewhere. Surviving papyri from Ravenna show how barbarians
and Romans cultivated documentary norms and rituals, such as registration
in municipal archives, quite beyond the requirements of Roman law; they
also reflect, however, immediate political circumstances and the growth of a
powerful church whose interests had become intertwined with those of secular
government.
From late Roman diocese to barbarian kingdom

The peninsulas of Spain and Italy, under Roman rule since the Sec-
ond Punic War, shared the similar fate in Late Antiquity of two highly
urbanized Roman dioceses, once central to Rome’s imperial power, that
fell to rule by barbarian peoples in the fifth century. Undeniably, the
demise of the Roman Empire in these regions was a blow to governmen-
tal use of documents, and to the literate traditions that supported it: a
liberal arts education focused on rhetoric, and a bureaucracy to facili-
tate communication across a vast stretch of territory. But at the level of
individual laypeople, developments in late Roman law had increased the
importance and frequency of documents for private transactions. The
barbarian governments of Spain and Italy were direct heirs of legal tradi-
tions that privileged writing as a form of proof, and made it mandatory to
63
64 Nicholas Everett
exercise juridical authority. That ‘barbarian’ governments in Spain and

Italy embraced the legal traditions of the late empire, and developed
them to suit their more regionally compact kingdoms, certainly upsets
our conventional picture of the ‘fall of Rome’ and ‘barbarian invasions’.
But neither the history of literacy, nor of technology (in this case, writing),
need subscribe to periodizations based on value-laden views of political
change. Instead, the rich evidence of the use of documents in barbarian
Spain and Italy demonstrates that documents continued to be valued as
a superior form of proof for claims to title, and as a source of authority,
whatever their format: whether written out by a professional scribe on
papyrus, by a friend on parchment, or by a neighbour scratched onto
slate.
In examining the use of documents at this level of private transactions,
this chapter focuses on certain issues raised by the evidence, rather than
attempting to survey it. For these issues go to the heart of understanding
how the end of the Western Roman Empire affected documentation in
post-Roman Europe, and may serve as historical background for many
of the later chapters in this volume. The word ‘may’ is operative here, as
the arguments put forward here are not ones with which all historians of
the period would agree, and stem from research on the issue of literacy,
which often provides a different perspective from, say, legal or economic
history. These arguments are best signalled here and now, and reflect the
order of the discussion.
Firstly, I argue that changes in late Roman law increased the impor-
tance of writing, as proof to claims, and as part of court procedure, a
development buoyed by political and cultural developments, such as the
expansion of late Roman bureaucracy and the conversion to a religion of
the book. Our two peninsulas under barbarian rule were direct heirs of
late Roman legal culture, and subsequently inherited the emphasis upon
written proofs and the use of documents in courts of law.
Secondly, the evidence of the papyri of Ravenna is discussed with a
particular focus on the registration of documents in municipal archives
(known as gesta municipalia), for Ravenna provides us with the only sur-
viving examples of this practice. It is argued that gesta municipalia were a
short-lived creation of the late Roman state for fiscal purposes, and their
survival in Ravenna reflects not only that city’s role as an administra-
tive capital but also the influence of a powerful metropolitan church that
effectively functioned as an arm of the state. The gesta municipalia, and
the registration of documents, are therefore something of a distraction
for understanding wider currents of document production and preserva-
tion. If the Ravennate gesta provide any sort of model for understanding
developments outside Ravenna, it is that of how churches engaged in
Spain and Italy, c. 400–700 65
registration as they merged with local curiae and assumed their functions,
echoes of which are found in formula collections. Moreover, many of the
gesta documents from Ravenna demonstrate peculiarities that reflect a
particular historical context – that of a political capital, for both imperial
and barbarian rulers, in a period of fundamental change and regional
instability.
Thirdly, the discussion turns to the legislation of Visigothic Spain,
and measures its copious references to documentary practices against
the evidence of the Formulae Visigothicae, the few surviving scraps of
parchment, and above all the extraordinary documents written on slate. It
is argued that not only does such evidence dispel reasonable doubts about
the applicability of Visigothic legislation on documents, but also that
Visigothic legislation typifies the development of Roman law mentioned
in the first section, and as such provides a model for how documents
continued to be valued in post-Roman Europe.
Finally, to respect the chronology posted in the title and as a fitting
conclusion, I shall briefly point to the main issues concerning the use of
documents in seventh-century Lombard Italy.
In concentrating on private transactions between lay individuals, I leave
out some subjects that might well be considered pertinent. State docu-
ments and archives are considered where relevant to issues raised by the
papyri of Ravenna and Visigothic law; Cassiodorus’ Variae, although a
rich source for the production of state documents and chancery tradi-
tions in Ostrogothic Italy, is referred to only on occasion, as uncertainty
surrounds the date of its contents and its publication. Similarly, the close
alliance between the Church and the Visigothic monarchy in the seventh
century meant that legislation in Church councils was also considered
law, but I have not included that legislation here: it would not change
the argument anyway. Finally, an aspect of ecclesiastical influence which
affects all documentation in this period is the emphasis on the oath as a
legal instrument and its effect on legal procedure, both in Roman and
in barbarian law, but the subject is too large to tackle here, and awaits
thorough treatment.1
1 On the Variae (MGH AA 12:1–385), see: G. Vidéu, The Roman Chancery Tradition:
Studies in the Language of the Codex Theodosianus and Cassiodorus’ Variae (Gothenburg,
1984) (chancery traditions); R. Macpherson, Rome in Involution: Cassiodorus’ Variae and
Their Literary and Historical Setting (Poznań, 1989) (historical setting); A. Gillett, ‘The
Purposes of Cassiodorus’ Variae’, in A. C. Murray (ed.), After Rome’s Fall: Narrators
and Sources of Early Medieval History: Essays Presented to Walter Goffart (Toronto, 1998),
37–50 (publication). Visigothic councils: R. Stocking, Bishops, Councils, and Consensus
in the Visigothic Kingdom 589–633 (Ann Arbor, MI, 2000). Oaths: Settlement of Disputes;
N. Everett, ‘Diritto tardo romano e alfabetismo giuridico nell’Europa alto-medievale’,
in Scrivere e leggere nell’alto medioevo, Settimane 59 (Spoleto, 2012), 213–46; Everett,
66 Nicholas Everett
From tablets to charters and late Roman ‘popular law’

The ‘late antique’ period (traditionally AD 285–632, from the emperor
Diocletian to the death of the prophet Mohammad) witnessed a revolu-
tion in the use and importance of the written word which greatly affected
legal documentation. The momentous changes behind this revolution
encompassed ideological and practical developments which overlapped:
a more centralized and more bureaucratic Roman Empire; its highly
rhetorical education system that prized literary knowledge and convo-
luted literary expression as evidence of the right to rule; the dominance
of a religion of the book; the shift from roll to codex, and from papyrus to
parchment; and new forms of script and page layout that facilitated read-
ing for a wider audience of less experienced readers.2 Legal developments
were an integral aspect of this revolution. The first, large collections of
Roman law were compiled in this period, from the two privately commis-
sioned codes under Diocletian (c. 290s), to the Theodosian Code (438)
and Justinianic Code (529), plus a host of legal compilations comparing
Roman, Greek and even Mosaic law. Born from aspirations of imperial
unity, these collections primarily served the practical need to supply con-
densed and organized repositories of legal precepts in an accessible and
‘Paulinus of Aquileia’s Sponsio episcoporum: Written Oaths and Ecclesiastical Discipline

in Carolingian Italy’, in W. Robbins (ed.), Textual Cultures of Medieval Italy (Toronto,
2011), 167–216. There is still much in H. Brunner, Zur Rechtsgeschichte der römischen
und germanischen Urkunde.I. Die Privaturkunden Italiens; das angelsächsische Landbuch; die
fränkische Privaturkunde (Berlin, 1880).
2 For a survey of these developments, see N. Everett, ‘Literacy from Late Antiquity to
the Early Middle Ages’, in D. Olson and N. Torrance (eds.), Cambridge Handbook of
Literacy (Cambridge, 2009), 362–85. Education: R. A. Kaster, Guardians of the Language:
The Grammarians and Society in Late Antiquity (Berkeley, CA, 1988); M. Irvine, The
Making of a Textual Culture: ‘Grammatica’ and Literary Theory, 350–1100 (Cambridge,
1994); R. Browning, ‘Education in the Roman Empire’, in A. Cameron, B. Ward-Perkins
and M. Whitby (eds.), The Cambridge Ancient History, XIV, Late Antiquity: Empire and
Successors, A.D. 425–600 (Cambridge, 2000), 855–83; R. Cribiore, Gymnastics of the
Mind: Greek Education in Hellenistic and Roman Egypt (Princeton, NJ, 2001); Cribiore,
The School of Libanius in Late Antique Antioch (Princeton, NJ, 2007). Bureaucracy: H.
C. Teitler, Notarii and Exceptores: An Inquiry into the Role and Significance of Shorthand
Writers in the Imperial and Ecclesiastical Bureaucracy of the Roman Empire from the Early
Principate to c. 450 A.D. (Amsterdam, 1985); C. M. Kelly, ‘Later Roman Bureaucracy:
Going Through the Files’, in A. K. Bowman and G. Woolf (eds.), Literacy and Power
in the Ancient World (Cambridge, 1994); C. M. Kelly, ‘Emperors, Government and
Bureaucracy’, in A. Cameron and P. Garnsey (eds.), The Cambridge Ancient History, XIII,
The Late Empire, A.D. 337–425 (Cambridge, 1998), 138–88; C. M. Kelly, Ruling the Later
Roman Empire (Cambridge, MA, 2004). Rhetoric: P. Brown, Power and Persuasion in Late
Antiquity: Towards a Christian Empire (Madison, WI, 1992). Books: L. Hurtado, The
Earliest Christian Artifacts: Manuscripts and Christian Origins (Grand Rapids, MI, 2006);
A. Grafton and M. Williams, Christianity and the Transformation of the Book (Cambridge,
MA, 2006); W. Klingshirn and L. Safran (eds.), The Early Christian Book (Washington,
DC, 2007).
reference-ready format: parallel developments can be witnessed in other

fields of literature, including history, medicine and, of course, religion,
wherein careful scrutiny of texts and debate over their meaning formed
the basis for formulating and establishing orthodoxy.3
But another motive for the legal compilation was the need to define
exactly what was Roman law in the aftermath of Caracalla’s Antonine
Constitution of 212, which extended Roman citizenship and, as a conse-
quence, the application of Roman law to all inhabitants of the empire.4
The explosion of juristic activity that resulted grappled with an influx of
new ideas and concepts from provincial legal practices, which often could
not be reconciled with the subtleties and distinctions of classical Roman
law, nor with its requisite formal and ceremonial niceties. This process
has been distastefully labelled the ‘vulgarization’ of Roman law, partic-
ularly in regard to property law, though ‘popularization’ is perhaps the
better term.5 But it was precisely this ‘vulgar’/‘popular’ Roman property
3 Codification: J. Matthews, Laying Down the Law: A Study of the Theodosian Code (New
Haven, CT, 2000); A. D. Lee, ‘Decoding Late Roman Law’, Journal of Roman Stud-
ies, 92 (2002), 185–93; C. Humfress, ‘Judging by the Book: Christian Codices and
Late Antique Legal Culture’, in Klingshirn and Safran (eds.), The Early Christian Book,
141–58. Religious texts: B. D. Ehrmann, The Orthodox Corruption of Scripture: The
Effect of Early Christological Controversies on the Text of the New Testament (New York,
1993); K. Haines-Eitzen, Guardians of Letters: Literacy, Power and the Transmitters of
Early Christian Literature (Oxford, 2000). History: G. B. Greatrex, ‘Lawyers and Histo-
rians in Late Antiquity’, in R. Mathisen (ed.), Law, Society and Authority in Late Antiq-
uity (Oxford, 2001), 148–61. Medicine: V. Nutton, Ancient Medicine (London, 2004),
292–309; N. Everett, The Alphabet of Galen: Pharmacy from Antiquity to the Middle Ages
(Toronto, 2011).
4 On the (fragmentary) text: A. Lukaszewicz, ‘Zum Papyrus Gissensis 40 I 9 (Constitu-
ito Antoniana)’, Journal of Juristic Papyrology, 20 (1990), 93–101. Effect: P. Garnsey,
‘Roman Citizenship and Roman Law in the Later Empire’, in S. Swain and M. Edwards
(eds.), Approaching Late Antiquity: The Transformation from Early to Later Empire (Oxford,
2004), 133–55, and, for a revisionist account, R. Mathisen, ‘Peregrini, Barbari, and Cives
Romani: Concepts of Citizenship and the Legal Identity of Barbarians in the Later Roman
Empire’, American Historical Review, 111 (2006), 1011–40.
5 The term ‘vulgar law’ was first coined by Brunner, Zur Rechtsgeschichte, 113–19, but the
concept was fully articulated by E. Levy: West Roman Vulgar Law: The Law of Property
(Philadelphia, 1951); Pauli sententiae: A Palingenesia of the Opening Titles as a Specimen
of Research in West Roman Vulgar Law (Ithaca, NY, 1945); ‘Vulgarization of Roman
Law in the Early Middle Ages’, Mediaevalia et Humanistica, 1 (1963), 14–40. A dif-
ferent interpretation was offered by F. Wieacker, Vulgarismus und Klassizismus im Recht
der Spätantike (Heidelberg, 1955). The concept of ‘vulgar law’ has been heavily criti-
cized by Tony Honoré (‘Conveyances of Land and Professional Standards in the Later
Empire’, in P. Birks (ed.), New Perspectives in the Roman Law of Property: Essays for Barry
Nicholas (Oxford, 1989), 137–52, at 151; ‘Roman Law 200–400: From Cosmopolis to
Rechtsstaat?’, in Swain and Edwards (eds.), Approaching Late Antiquity, 109–32), but
it is still a useful concept. Overview of debates and implications for post-Roman law:
J. B. Sirks, ‘Shifting Frontiers in the Law: Romans, Provincials and Barbarians’, in R.
Mathisen and H. Sivan (eds.), Shifting Frontiers in Late Antiquity (Aldershot, 1996),
146–57; cf. T. Charles-Edwards, ‘Law in the Western Kingdoms Between the Fifth and
68 Nicholas Everett
law that was transmitted to the barbarian kingdoms of Spain and Italy
as the basis of their own property law, and in particular the treatment
of documents as legal instruments. Leaving aside the complicated shifts
in legal theory associated with this vulgarized Roman law, one practical
aspect of the change was clear: writing – and documentation – became
more important than it had been in classical law – as proof, as testimony
to correct procedure followed, and as a source of authority. This is par-
ticularly true for the courts, as is clear from Constantine’s decree (320)
that all accusations now had to be in writing, and a later decree of 374
which mandated that all judges ‘in the provinces’ must read their judicial
decisions from a written statement: any oral pronouncement that was not
a recitatio scripti was invalid, a precept we hear echoed in our barbarian
codes and law courts of Spain and Italy.6
Moreover, a wider range of forms and formats of documents became
acceptable. Romans and Roman law had privileged tabulae, inscribed
wooden tablets embodying ‘unitary’ acts in steeply formalized language,
as the authoritative legal instrument until the third century AD But
the Antonine Constitution, and an increasingly eastern-oriented empire,
demanded that other forms of documentation be recognized, particularly
Greek and Egyptian practices; thesse used papyrus, were much less for-
mal in style and were more epistolary.7 Caracalla’s chief legal adviser, the
jurist Ulpian (d. 228), noted the change when he remarked that a will,
that most traditional of Roman documents and customarily written upon
tablets amidst solemn rituals and formalized speech, should now be valid
‘when written on any kind of material, whether it be on bound tablets, or
papyrus, or parchment, or on the skin of any kind of animal’.8 A century
later, the emperor Constantine, who created a new capital in the East
and introduced many eastern ‘vulgarisms’ into Roman law, opened up
the format of the will to include simplified language:
in the execution of last wills, the requirement of formal speech is hereby removed,
and those who desire to dispose of their own property can write their wills upon
any kind of material whatsoever, and are freely permitted to use any words which
the Seventh Centuries’, in Cameron, Ward-Perkins and Whitby (eds.), The Cambridge
Ancient History. XIV. Late Antiquity, 260–87.
6 CTh. 9.1.5; 4.17.1. Also CTh. 11.3.40. Repeated in ETh. 64 (see below, n. 39) (cf.
Cassiodorus, Variae 11.38 (MGH AA 12:351–2; according to Variae 6.3.3 (MGH
AA 12:176), praetorian prefects were exempt), and LV VIII.1.1.5, IV.1.2, II.3.4
(see below, n.74). On recitatio in the imperial period: E. Meyer, Legitimacy and Law
in the Roman World: Tabulae in Roman Belief and Practice (Cambridge, 2004), 88–99. See
also Classen, ‘Fortleben’, 14–15, 51–2.
7 A full account is given by Meyer, Legitimacy and Law, with references to earlier literature.
8 Dig. 37.11.1 (Ulpian).
they may desire . . . and it makes no difference whatever what grammatical forms
of the verbs indicate in his will, or what way of speaking pours out.9
Specific forms, language or format did not matter, so long as the doc-
ument was clear in its intent and was rooted in a social context (in this
case, the use of witnesses) that could be examined if contested. No mat-
ter where we look in late Roman law – whether in the realm of sales,
donations, wills or different types of agreements between consenting
parties – the victory of documentation over other forms of proof is a
recurring theme, even if this meant simplification of legal concepts, such
as a reduction in the number of different types of donation, or the col-
lapse of contract and conveyance into a single, dispositive act for sales.10
By 472, Emperor Leo declared that any document recording any trans-
action (sales, loans, private agreements), even if not subscribed by the
interested parties or even witnessed, was to have the full force of those
that had enacted these measures.11 Justinian, particularly in his Novels
(largely unknown in the West), was forced to concede to similar precepts,
despite his efforts in the Codex and Digest to turn back the clock and rein-
troduce classical concepts and formalities centred on oral procedures.12
The supremacy of written documents as proof, therefore, was one of
the most important aspects of late Roman law that fundamentally shaped
the legal landscape of barbarian Spain and Italy, and the role of docu-
ments within those kingdoms. Both the Visigothic and the Ostrogothic
kingdoms were direct heirs of Roman legal traditions, and barbarian kings
relied directly on Roman legal advisers to repackage Roman law with a
royal imprimatur: the interpretationes to the Visigothic king Alaric’s com-
pilation of Roman law, the Breviarium (published 506), are directly par-
alleled in the precepts of the near-contemporary Edict of King Theoderic
the Ostrogoth.13 The Lombards took a different legislative route, but they
9 CJ 6.23.15 + 6.37.21 + 6.9.9.

10 Levy, West Roman Vulgar Law; R. Zimmerman, The Law of Obligations: Roman Founda-
tions of the Civilian Tradition (Oxford, 1996).
11 CJ 8.17.11.
12 Everett, ‘Diritto tardo romano’, 223–4 (documentary practices in the Novels); S. Puliatti
and A. Sanguinetti (eds.), Legislazione, cultura giuridica, prassi dell’impero d’Oriente in età
giustinianea tra passato e futuro: Atti del Convegno, Modena, 21–22 maggio 1998 (Milan,
2000); C. Humfress, ‘Law and Legal Practice in the Age of Justinian’, in M. Maas (ed.),
The Cambridge Companion to the Age of Justinian (Cambridge, 2005), 161–84; Levy, West
Roman Vulgar Law.
13 Interpretationes to the Breviarium: Matthews, Laying Down the Law and ‘Interpreting the
interpretationes of the Breviarium’, in Mathisen (ed.), Law, Society, and Authority, 11–32.
Similarity to provision in the ETh.: S. Lafferty, ‘The Edictum Theoderici: A Study of a
Roman Legal Document from Ostrogothic Italy’, unpublished Ph.D. thesis, University
of Toronto (2010), ch. 2.
70 Nicholas Everett
were also heavily reliant on late Roman property law and its documentary
norms for security of title and the administration of justice.
Before we turn to examine the evidence, three points should made
concerning the late Roman legal heritage and its effect on documentation
in barbarian Spain and Italy:
(1) In late Roman law as in the barbarian laws that followed, who actually
wrote the document did not matter. The validity of the document
resided in its concordance with the law and, where relevant, the
subscriptions of witnesses who attested to the document’s contents
as the intent of the person who requested its redaction. One could
certainly hire a professional notary or tabellio, as was the case in
nearly all of our documents from Ravenna, and as is hinted at in one
Visigothic law mentioning ‘public notaries’, but this had no effect
on the document’s legal validity, which was a (much) later medieval
development.14
(2) Under classical Roman law, tampering with documents or forgery
constituted a less serious delictal act of fraud (dolus). In late Roman
law and barbarian law, forgery is elevated to the crime of falsum,
forgery proper, on a par with counterfeiting money and other crimes
considered to be lèse majesté or against the state itself.15 The far
heavier penalties for forgery, including amputation and capital pun-
ishment, further illustrate the increased importance of documents in
the eyes of both barbarian and Byzantine successor states.
(3) With the questionable exception of the gesta municipalia (see below),
people were responsible for preserving their own documents, and
there was no insistence on archival practice of any sort for the vali-
dation of documents or for determining the superiority of one kind
of document over another. Let us now turn to the Italian evidence,
which forces us to confront the issue of archives head on.
The papyri of Ravenna and gesta municipalia

The small corpus of fifty papyrus documents dating to the period AD
455–700 and collected in an edition by the Swedish scholar Jan-Olof
Tjäder16 constitutes the only surviving documentary evidence from Italy
14 N. Everett, ‘Scribes and Charters in Lombard Italy’, Studi Medievali, 41 (2000), 39–83,
at 42–5. Ravennate scribes: Everett, ‘Scribes and Charters’, 56–9. Visigothic law: LV
VII.5.8, and below, 85.
15 Delict: Dig. 4.3.1 (Ulpian), based on the Cornelian law: cf. PSent. 5.7.2. Constantine
extended the definition and added capital punishment: CTh. 9.19.1, 2; 9.21.9; 9.22.1.
16 Hereafter cited as P.Ital (papyri), and Tjäder (will be cited by volume and page numbers
[vol.:pp.]) (commentary). Tjäder’s edition can be difficult to obtain: the papyri can also
be consulted in the relevant volumes of ChLA, many of which Tjäder also edited; the
in this period, which witnessed the end of the Roman Empire and the
establishment of three successive barbarian kingdoms in Italy (that of
King Odovacar 476–93, the Ostrogoths 493–535 and the Lombards 568–
774). Most are associated with the city or church of Ravenna in some way,
in terms of their production or preservation, but not all. The documents
include sales (eleven documents), donations (sixteen), wills (seven), a few
ecclesiastical leases (from the seventh century), a number of inventories of
Church property, part of an inventory of the praetorian prefect’s archive
(c. 510–50) and a few court proceedings. They are fascinating artefacts in
a number of ways: full of convoluted legal formulas in rough Latin written
by professional scribes in cursive scripts, and containing transactions
and subscriptions of laity – bankers, soldiers, silk-makers, soap-makers,
trouser-makers, shipbuilders, grammarians and so on; some used Greek
characters for their Latin, while some Gothic clergy wrote in Gothic,
and a barbarian king scrawled ‘hello, keep well’ on a charter. A dozen of
these papyri comprise the only surviving examples of gesta municipalia, or
proceedings of the city council (curia), in which documents are presented
by citizens to the council for inspection, recitation, confirmation and
registration in the city archives, with copies of the transcript of these
proceedings given out to the citizens.
These gesta documents have exercised a considerable amount of influ-
ence on our understanding of late Roman and post-Roman documen-
tary practices, for some scholars saw them as representing a dividing
line between the ‘public’ documentary practices of Late Antiquity and
the ‘private’ documentary world of the early Middle Ages.17 Hence it is
thought that the eclipse of municipal curiae sometime in the sixth cen-
tury (no one is sure when, and it depends on where) explains why our
private charters only seem to survive in decent numbers from the sev-
enth century onwards, and likewise records of court cases (placita), as
though these somehow filled a void left by the collapse of public archives.
There is no need to rake over how and why such theories appealed to
scholars in the Rechtsschule tradition, or how a few templates for such
documents in Merovingian and Visigothic formularies,18 combined with
largest number (sixteen) are preserved today in the Vatican (ChLA XX–XXII), with
other libraries (Paris, Padua, Naples, Florence, Vienna, Venice) possessing only a few
each. Cross-references to ChLA are given when individual papyri are discussed.
17 B. Hirschfeld, Die Gesta municipalia in römischer und frühgermanischer Zeit (Hamburg,
1904); Classen, ‘Fortleben’, 48–9; W. Bergmann, ‘Untersuchungen zu den Gericht-
surkunden der Merowingerzeit’, Archiv für Diplomatik, 22 (1976), 1–186, at 105–6.
18 Merovingian evidence for gesta: Rio, Formularies, 255–8; Rio, Legal Practice; A. Murray,
‘Review Article: The New MGH Edition of the Charters of the Merovingian Kings’,
Journal of Medieval Latin, 15 (2005), 246–58. The rich array of documents: Brown,
‘Documents’. Formulae Visigothicae: below, n.54 and n.73. Rechtsschule scholarship: Set-
tlement of Disputes, 2–4; P. Wormald, The Making of English Law: King Alfred to the
72 Nicholas Everett
the Ravennate examples and references in late antique legislation, have

helped contribute to a false picture of thriving registration throughout
the late empire.
As a creation of the late Roman state, the gesta municipalia are indeed
another reflection of the increased importance of writing in Late Antiq-
uity, particularly for disputes in court.19 But the primary purpose of
registration of common transactions between laypeople, such as sales
and donations in the gesta municipalia, was not, as is often suggested, the
creation and storage of an ‘authoritative’ document, but rather a means
by which the late Roman government kept track of property ownership
for taxation; as such, it was largely restricted to wealthy, landowning
elites, particularly the senatorial and decurion classes.20 That registra-
tions of sales and donations only survive from Ravenna, and not, say,
Egypt, which provides an abundance of evidence for official documents
and formal notifications of transfer of title (and tax obligations) to the
local council,21 suggests that Ravenna may in some ways be an exception.
But Ravenna was no ordinary late antique city. From AD 400 onwards,
it was the imperial, administrative capital of the West, and home of a
powerful metropolitan church that became inextricably entwined with
secular administration upon the resumption of imperial control from
540 onwards, the period from which most of gesta papyri date (up to
557, then a gap until the last in 625), nearly all of which concern the
church.22 In fact, there are only three purely ‘secular’ documents in
Twelfth Century. I. Legislation and Its Limits (Oxford, 2001), 11–12. The few formulaic
references to gesta in the Merovingian material invite theories of institutional continuity,
most recently in J. Barbier, ‘Testaments et pratique testamentaire dans le royaume franc
(VIe –VIIIe siècle)’, in F. Bougard, C. La Rocca and R. Le Jan (eds.), Sauver son âme et se
perpétuer: Transmission du patrimoine et mémoire au haut moyen âge (Rome, 2005), 7–79,
and Barbier, ‘Pouvoirs et élites dans le monde franc (VIe –XIe siècle): Matériaux pour
servir à l’histoire des élites des cités (VIe –IXe siècle): Le dossier des gesta municipalia,
étude no 2’, unpublished mémoire d’HDR, Université de Paris I–Panthéon-Sorbonne
(2009) (though only parts have been made available); Barbier argues that gesta con-
tinued in Francia into the ninth century. Such theories, to my mind, misunderstand
the function of gesta in the late Roman world (as described here), and the ecclesi-
astical nature of the Merovingian evidence – see the explanation offered below, at
nn. 33–5.
19 R. Lim, Public Disputation, Power and Social Order in Late Antiquity (Berkeley, CA, 1995);
J. Harries, Law and Empire in Late Antiquity (Cambridge, 1999), 70–6 and passim;
C. Humfress, Orthodoxy and the Courts in Late Antiquity (Oxford, 2007).
20 As is clear from CTh. 2.29.2; 5.15.20.1; 6.2.18; 8.1.9; 8.2.2; 11.1.3; 11.28.13; 12.1.173;
12.6.20; 13.10.8; 13.11.2, 13 – and the references cited below.
21 E.g. P.Oxy LIV 3754, 3756, and esp. 3758 (a compilation), containing petitions
addressed to or proceedings before a logistes (curator), c. 320s; also P.Oxy. I 126. See
above, Chapter 2, and R. Bagnall, ‘Papyrus Documents in Egypt from Constantine
to Justianian’, in R. Pintaudi (ed.), Miscellenea papyrologica (Florence, 1980), 12–23;
R. Bagnall, Egypt in Late Antiquity (Princeton, NJ, 1993).
22 Only three are pre-540: P.Ital 10–11 (a. 489, Odovacar), 12 (a. 491, church donation),
29 (a. 504, cleric). Last: P.Ital 21 (a. 625), donation of cleric to the church. Church
gesta format among the Ravennate collection, and two of these were
originally registered elsewhere: one is a complicated donation made by
King Odovacar in 489 on the eve of his demise, which was registered
in Syracuse, where the property was located; the other records a court
case concerning a Gothic widow in the town of Rieti in 557, where it was
registered (but our copy is from elsewhere), and which reflects social ten-
sions caused by the Gothic wars in Italy.23 The document registered in
Ravenna itself also reflects political change: a purchase by the Ostrogothic
king Witigis’ personal notary in the inauspicious year of 540, on the eve
of Ravenna’s reconquest by imperial forces.24 Several other (non-gesta)
documents also reflect ‘this barbaric time’ (as declared in a donation by
yet another Gothic widow – as we shall see, there are several) of conflict
in sixth-century Italy and troubles for Goths, which raises suspicions of
selection behind what survives – an aspect of this corpus that has largely
gone unnoticed.
This is not to say that registration was unimportant – in fact, it seems
to have been more important to Ravenna’s ruling classes, barbarian or
otherwise, than it was to contemporary emperors, for we find Ravenna’s
citizens registering transactions beyond the requirements of imperial law.
There is no need to review in detail the imperial legislation that maps
the rise and decline of gesta municipalia, but a few salient points should
be made to explain Ravenna’s unique position, and the character of its
surviving papyri. Constantine introduced the requirement (in 316) that
all donations of immovable property be registered ‘in the acts (acta)
before a iudex (that is, a provincial governor) or a magistrate’ (of the
municipality), but legislation towards the end of the century seems to
restrict such measures to senatorial patrimonies, and to forced sales of
land by town councillors (decurions). Moreover, although magistrates
were given the right to create their own records in 366, thirty years later
it was declared that they could do so only with three members of the curia
and one public secretary (exceptor publicus) present, ‘so no opportunity
for fraud should arise’.25
and state administration: T. S. Brown, ‘The Church of Ravenna and the Imperial
Administration in the Seventh Century’, English Historical Review, 94 (1979), 1–28,
and T. S. Brown, Gentlemen and Officers: Imperial Administration and Aristocratic Power
in Byzantine Italy, A.D. 554–800 (Rome, 1984). The city as capital: D. Deliyannis,
Ravenna in Late Antiquity (Cambridge, 2010).
23 See below, 80–82.
24 P.Ital 10–11; 7; 31. ‘tempore hoc barbarici’: P.Ital 13. See Everett, ‘Scribes and Charters’,
68.
25 Donations, confirmed in 341: CTh. 8.12.1, 6. Senatorial: CTh. 2.29.2. Councillors:
CTh. 12.3.1. ius actorum conficiendorum: CJ 1.56.2. Three curiales and exceptor: CTh.
12.1.151 = CJ 10.32.47. This last was addressed to the vicar of Spain, where no
evidence of registration survives, save for templates in the Formulae Visigothicae (Form.
74 Nicholas Everett
It is worth reflecting that this type of registration before the curia,

which we find in the Ravenna papyri, was decreed on the eve of the
Western empire’s destruction. Over the next thirty years the territories
of Gaul, Spain and Africa fell to barbarian rule. True, city councils con-
tinued to collect taxes in many places, even rapaciously – hence the quip
of Salvian of Marseilles (writing c. 439–50), ‘What cities are there, and
not only cities but even towns and villages, in which the councillors
(curiales) are not so many tyrants?’26 Barbarian governments maintained
the existing tax system where they could, but it became increasingly
marginal, and revenues were collected through royal officials, military
officers and local bishops. By 444, a desperate Valentinian III introduced
the requirement that all sales of land be registered in the gesta munici-
palia: the purpose was to collect a new sales tax, the siliquaticum (that
is, 1 siliqua or 1/24 per solidus) from both buyer and seller, a tax he
also extended to all sales of movable property – an impossible scenario,
even for late Roman legislative ideals.27 Most later compilations or com-
mentaries on Roman law, including Valentinian’s own later edicts, fail
to mention it. In any case, Valentinian’s law could aim only at what was
left of the Western empire, namely Italy; hence, it is not surprising that
it is only in Ostrogothic Italy that we find evidence for the survival of
the siliquaticum, a tax which ‘antiquity ordained’, according to Theoderic
(via Cassiodorus), though he struggled to collect it, and granted several
exemptions.28
Yet everywhere else we see a retreat from government archives at the
local level. In 479, the emperor Zeno had declared that donations without
Vis. 21, 25): see below, 82–93. Registration laws: Everett, ‘Scribes and Charters’, 73–
9. Postclassical legislation on registration is collected in S. Tarozzi, Ricerche in tema di
registrazione e certificazione del documento nel periodo postclassico (Bologna, 2006). Fees
for documents: Harries, Law and Empire, 122–4; Kelly, Ruling the Later Roman Empire,
74–6, 153, 163; Teitler, Notarii and Exceptores, 27–37.
26 Salvian, De gurbernatione Dei 5.4(18) (MGH AA 1.1:58), cited in A. H. M. Jones, The
Later Roman Empire, 284–602: A Social, Economic and Administrative Survey, 2 vols.
(Baltimore, MD, 1986), II, 756, who surveys the evidence for councils, II, 737–63;
updated by Wickham, Framing, 596–602. Arguing for continuity in Spain, at least in
terms of local prestige and identity, is M. Kulikowski, Late Roman Spain and Its Cities
(Baltimore, MD, 2004), 42–9.
27 Valentinian III, Nov. 15.1.3. As desperate taxation: Jones, Later Roman Empire, 205,
435, 826. In the context of Valentinian III’s legislation: T. Honoré, Law in the Crisis of
Empire, 379–425 A.D.: The Theodosian Empire and Its Quaestors (Oxford, 1998), 259–74.
The use of gesta municipalia is part of a jumble of measures to collect the tax: special
praepositi were appointed in each province to attend personally to each transaction,
collect the tax and issue a receipt; this required that city councils designate specific
market days for different classes of goods.
28 Cassiodorus, Variae 4.19 (MGH AA 12:122–3), exempting sales of corn, wine and oil
‘to stimulate trade’. Its collection was paired with another, monopolium: Variae 2.4, 12,
26, 30; 3.25–6; 5.31 (MGH AA 12:49, 52, 61, 63, 92–3, 160).
registration, or even made without any documents, were valid, and we

can trace similar sentiments in legislation of fifty years earlier.29 Justinian
limited registration to donations and gifts over 500 solidi in value, did
not countenance registration for sales anywhere in his legislation, and
even allowed sales without any document.30 Moreover, his attempt to
restore the dignity of the office of defensor civitatis, another late antique
bureaucratic invention (c. 368), which he bemoaned as having become
‘an insult rather than distinction’, reveals the poor state of city archives
across the empire. In 535, Justinian conceded that:
When documents are drawn up by them [defensores] in the first place, they only do
this for money; and then, as there are no archives in which these documents can
be deposited, they are lost; and no monuments of former times are ever found in
the possession of those who receive them.31
Such comments suggest that Ravenna, and its functioning curial archives,
may have been exceptional. A glimpse at the fragments of Ravenna’s prae-
torian prefect’s archive – or more precisely, an archival index – preserved
among our papyri shows an effective system at work. Two fragments
provide forty-four lines (entries) that list documents preserved with the
arcarius (treasurer) sometime between 510 and 550, but most likely in
the Ostrogothic period. I quote a few entries to give something of its
flavour:
Notice [brevis] between Stefan, Peter and Iacob from the collection in Liguria
[de colligatione Liguria].
Guarantee [cautio] of Macedonus made out to Peter for 1,250 solidi in the year
after the consulship of the uir clarissimus Paulinus [499].
Guarantee, again of Macedonus, and in Greek, made out to Peter for 1,100
solidi.
Guarantee of Verissimus, assistant in the accounting office [numerariorum], made
out to Peter for 4 solidi in the year after the consulship of Paulinus [499].
Promissory note [pittacium] of Stefan, vir spectabilis, made out to Peter, for 4 solidi
and 2 tremisses.
29 CJ 8.53.31.1, looking back to CJ 8.53.29 (a. 428). In contrast, note CJ 8.53.31pr.,

that there is no need for witnesses ‘cum publica monumenta sufficiant’. Note also CJ
8.53.32 (Anastasius, a. 496), dismissing CJ 8.53.30.2 (a. 459).
30 Inst. 3.23. Note also the casual process of redacting a document. Registration of dona-
tions was abolished altogether by Leo III (717–41), Nov. 50.
31 J.Nov. 15.5.1–2; see M. Frakes, ‘Late Roman Social Justice and the Origins of the
Defensor Civitatis’, The Classical Journal, 89 (1994), 337–48, and A. Dietl, Defensor
civitatis: Der Stadtpatron in romanischen Reliefzyklen Oberitaliens (Munich, 1998).
76 Nicholas Everett
Letter of notice [epistula praecaturia] made by Nonno, Asclepius and Aelius con-
cerning 4,000 solidi.
Likewise [by Nonno?], a document folder made of soft leather which was found
here bound with a [ . . . ] [possibly br(evis)]
The archive is organized by name rather than dates and the amounts vary
greatly, from 2 to 4,000 solidi (a considerable sum). The list also refers
to other collections, such as the soft leather folder mentioned above
but also a ‘fascicle . . . of promissory notes’, an ‘accounting document
of the envoy’s office’ of Picenum (Ancona), a ‘chest in which there are
found different documents concerning 152 solidi’ and – notably using a
Latinized Gothic word – ‘twelve little lists’ (listula duodecim).32
I should like to suggest here that changes to the office of defensor civi-
tatis and the city council in the early fifth century may explain our gesta
evidence from elsewhere, and much of that from Ravenna. Legislation
of 409 and 413 granted the local bishop and his clergy important roles
in councils, including the election of the defensor civitatis and the right of
bishops and defensores ecclesiae to sit beside a judge during a trial, offer
him advice and enter his private chambers – that is, direct access to and
involvement with council procedures.33 We may suspect that our evidence
for registration from Merovingian and Visigothic formula-books, all dis-
playing ecclesiastical origin, stems from such changes and the increasing
power of the Church in local governance. For we see the same in Ravenna:
a gesta document dating between 552 and 575 records how Ravennate
clerics (two defensores ecclesiae and two chief notaries) presented before
the ‘praetorian prefect of Italy’ accounts of the opening of six wills which
had taken place before the curia, the earliest dating to 480, the latest
to 552, held before a defensor civitatis. That is, these clerics were re-
registering gesta documents which had already been registered.34 The
32 P.Ital 47–9 (aa. 510–50) (= ChLA XXV 792, XXIX 870). Ostrogothic period: the
‘Peter’ mentioned in twenty-two documents (out of around sixty), and apparently titled
arcarius in no. 18, is possibly the arcarius of Cassiodorus, Variae 12.20 (c. 536) (MGH
AA 12:376–7); Venantius conrictor Lucaniae et Brittiorum of no. 22 appears in Variae
3.7–8 and 46 (all a. 507×11) (MGH AA 12:83–4, 101–2). Leather: adlatum chartarium
alutacium. Picenum accounting document: ratiocinia leagturae. Prefecture: secured by
‘de suffragio praefecturae’ and titles in the list (cursor, strator scolaris, all inlustris potestatis,
plus two adiutores numeriorum, and a consularis Flaminiae).
33 CJ 1.55.8; CTh. 6.26.16. Defensores ecclesiae and the ius sedendi: C. Humfress, ‘A New
Legal Cosmos: Late Roman Lawyers and the Early Medieval Church’, in P. Linehan
and J. Nelson (eds.), The Medieval World (London, 2001), 557–75.
34 P.Ital 4–5 (= ChLA XVII 653 + XXIX 878). Notaries: primicerius and secundocirius
notariorum. Dates: Tjäder 1:200–3. The defensor civitatis appears in the last will of 552.
Other donations registered by defensores ecclesiae: P.Ital 14–15 (a. 572), 21 (a. 625, the
last known gesta) (= ChLA XXIX 889, XXII 720). Note another will benefiting the
church of Ravenna in 575: P.Ital 6 (= ChLA XXI 714).
wills, of laity and clergy, all benefited the church in some way; hence, we
can only assume that the church was engaged in some sort of archival
reorganization – or perhaps we witness here a glimpse of how the city and
church archives merged. One lay will among these, that of Georgius, a
silk-maker (olosiricoprata), shows that the testator dictated his will from
his deathbed to his ‘known friend’ the scribe (forensis) Deusdedit, ‘in
accordance with civil or praetorian law or even the new laws recently
given’, a phrase that resonates with the Visigothic material, just as the
term defensores constantly appears in the Merovingian evidence, the bulk
of which concerns the registration of wills.35 Georgius adds a plea ‘that
any charters [of his] written on parchment or any other material’ remain
inviolable, suggesting that he does not ordinarily register documents, and
his last words in the will reveal why it was registered: ‘You, holy mother
Catholic church of Ravenna, in which the entire Christian people wor-
ship for the forgiveness of sins, I decree to be heir for a tenth of my
property.’
The Ravennate church encouraged registration to secure its gains, but
the laity of Ravenna faced legal and cultural imperatives to register due to
Ravenna’s role as the administrative capital of Italy. The legal imperatives
stemmed from the maintenance of a robust system of taxation and fiscal
surveillance. Nearly all of our surviving gesta papyri record the vendor’s
request to the curia to ‘remove my name from the public tax list (de
polypticis publicis) and write the name of the abovementioned buyer in its
place’.36 Many gesta documents included (or alluded to including – most
are fragmentary) copies of a letter of conveyance (epistula traditionis),
an instrument designed to inform the council of transfer of title. Yet,
these ‘letters of conveyance’, too, took the opportunity to relieve tax
obligations, as is clear from one letter that survives in its original format
(as opposed to being a gesta copy), in which the brothers Milanius and
Gerontius inform the curia of nearby Faenza in 540 (again that auspicious
date!) that they have sold a property to Laurentius, uir strenuus (a soldier),
in Ravenna:
but because the land is in your territory, and still corporal conveyance has not yet
taken place, therefore lords and senior cultivators of the law, may Your Praisewor-
thiness accept this letter, so that when the abovementioned buyer and his men
arrive to take up solemn conveyance as is customary, you have the letter read out
and inserted into your records (acta), and that you also hasten to advise the tax
35 Form. Vis. 21 (‘ad ius praetorium et urbanum’), 25, both concerning wills. Cf. Marculf
2.37, Form. Arvern. 1, Form. Andec. 1, Form. Tur. 3, etc. See Rio, Formularies, and Rio,
Legal Practice, with references cited.
36 P.Ital 31 (= ChLA XX 707). Others use vasaria publica, from which the tributa fiscalia
are known (P.Ital 10–11 [= ChLA XX 703]), or similar terms: Tjäder 2:26–7.
78 Nicholas Everett
collector of your city (tabularius civitatis) concerning the amount of taxable land
(cespitis iugationem) for this property, which has been recorded in the documents
mentioned above, and likewise that you order the buyer and his agents to be
designated [for the liabilities], so that he might be aware of everything owed from
the third indiction to the time of his ownership.37
Despite the fact that the letter was written in the year Belisarius captured
Ravenna, the revenue system was still fully operational, even rapacious,
as it was certainly reputed to have become under Justinian.38
While emperors had largely revoked or simply given up on registra-
tion in city archives by the end of the fifth century, the barbarian king
Theoderic, who lavished patronage upon Ravenna after he assumed rule
there in 493, and who maintained the trappings of late Roman civilian
bureaucracy, maintained and even extended registration, if the Edictum
Theoderici was really his.39 The Edict stipulated that it was necessary
to register donations of ‘rural or urban estates’ before three councillors
(curiales) and either a magistrate or defensor civitatis. But a following pro-
vision admits the difficulty of finding a magistrate or defensor and so offers
the alternative of three councillors ‘provided that corporal conveyance is
carried out with neighbours knowing about it (uicinis scientibus)’, a much
more casual process. Indeed, the papyri show a more casual curia, in
terms of personnel, as the sixth century progresses.40 Following a Novel
of Valentinian (451), the Edict required that a free man who defiled a vir-
gin slave and wished to marry her must acknowledge her owner’s consent
and his own loss of freedom in the gesta. A provision against forgery of
37 P.Ital 32 (a. 540) (= ChLA XX 708). The letter was dictated to a notarius, an unusual
title for Ravenna’s scribes, suggesting it may have been written elsewhere, and was
signed by three witnesses.
38 See references in Brown, Gentlemen and Officers; P. Sarris, Economy and Society in the Age
of Justinian (Cambridge, 2006); and J. F. Haldon, ‘Economy and Administration: How
Did the Empire Work?’, in Maas (ed.), The Cambridge Companion to the Age of Justinian,
28–59.
39 On the ETh., see now the work of Sean Lafferty: ‘The Edictum Theodorici’ (above, n. 13);
‘Law and Society in Ostrogothic Italy: Evidence from the Edictum Theodorici’, Journal
of Late Antiquity, 3 (2010), 377–404; ‘Italy in the Twilight of Empire: The Decline of
Roman Law and Culture Under Theodoric the Great, c. 493–526’, Canadian Journal of
History, 45 (2010), 457–84. Lafferty proves beyond doubt that the ETh. derives from
Italy, and argues convincingly for Theodoric’s paternity; I reserve the possibility that
it was drawn up under Odovacar. Others, particularly G. Vismara (Edictum Theoderici
[Milan, 1967]), have argued for King Theuderic II of the Visigoths; this has been per-
suasive, though rejected early: H. Nehlsen, ‘Edictum Theoderici’, Zeitschrift der Savigny-
Stiftung für Rechtsgeschichte, Germanistische Abteilung, 86 (1969), 246–60. Others have
dismissed it as forgery (e.g. B. Paradisi, ‘Critica e mito dell’editto teodericano’, Bollettino
dell’Istituto di diritto romano, 68 [1965], 1–47).
40 ETh. 52, modelled on CTh. 8.12.8; ETh. 53, modelled on CTh 8.12.8.2, is likewise
concerned about location, but does not water down official presence. Fewer curiales:
Brown, Gentlemen and Officers, 16–19.
documents includes tampering with gesta, an addition to the passage in

the Pauli Sententiae upon which it was based. And anyone who receives
a man claiming to be freeborn (ingenuus) must lead him before the gesta,
where he may publicly declare his status, that way protecting the receiver
in case the self-proclaimed freeborn turns out to be a fugitive slave; there
seems to be no precedent in Roman law for this.41
We also see a penchant for registration in some of the Ravenna papyri
that went beyond Roman laws. In 491, while Theoderic was besieg-
ing the city (490–3), the defensores of the Ravennate church registered
before the curia a charter and an epistula donationis recording the donation
of the widow Maria, femina spectabilis (suggesting that her deceased hus-
band was a military officer), to the church. Maria’s epistula mentions that
while she and her husband had drawn up a charter, his death meant
‘that we could not register it (eam allegare), so now it has been necessary
for me to write this letter in my own name’ to reiterate the donation.42
But it was not necessary: the only purpose for Maria’s letter should have
been to advise the curia to remove her name from the tax register (it
may have done so, as the letter is fragmentary). Registration was the
defensores’ responsibility as the new owners, as Maria’s charter rightfully
indicates by containing the necessary ‘permission to register’ (licentia
allegandi) formula that appears in all donations. That this same formula
appears in sales after 540 further suggests Ravenna’s enthusiasm for reg-
istration, which Tjäder noted.43 The political uncertainty of the times,
or the fact that the donation included the condition that Maria and her
husband be buried inside the church of S. Lorenzo, may both have played
a role. Either way, registration for Maria was a cultural imperative, not
a legal one, just as it was not legally required for sales in Justinianic
law.44
41 Marrying a slave: ETh. 64; cf. Valentinian III, Nov. 31.5, where the man keeps his free
status but cannot leave. Forgery: ETh. 90, ‘testamentum, codicillum, tabulas, rationes,
gesta, libellos, cautiones, epistolas’; cf. PSent. 5.25.1–5, ‘instrumenta, epistolae, cau-
tiones, chirographa’, though both laws also concern counterfeit money; ETh. 80, ‘ut
eum ducat ad gesta, et se profiteatur ingenuum’.
42 P.Ital 12 (a. 491) (= ChLA XXV 791). Theoderic’s siege: J. Moorhead, Theoderic in Italy
(Oxford, 1992), 24–5. Spectabilis: Jones, Later Roman Empire, 143, 282, 528–9.
43 Tjäder 2:26–7 (for the receiver’s benefit, not for the state). On changes in the formula
c. 600, see Everett, ‘Scribes and Charters’, 75n118.
44 Instances of the licentia allegendi formula aside, three sales were registered: P.Ital 29 (a.
504), for only 18 solidi (the purchaser was an acolyte of the Roman church, and the land
in question bordered that of the Ravennate church (lines 11–12); this explains why it
was registered); P.Ital 31 (a. 540), for 40 solidi (Montanus vir clarissimus, Witigis’ notary,
purchases land in Faenza from a local layman, Domnicus); P.Ital 33 (a. 541), for 20
solidi (Minulus, a Gothic cleric; see below, n.46).
80 Nicholas Everett
It should not be surprising that the citizens and church of Ravenna

were steeped in civic rituals and practices that proclaimed their Roman-
ness and the capital status of their city. But is it a coincidence that these
rituals keep appearing in papyri that date to periods of conflict, and
include Goths? A year before Belisarius took the city, a Gothic widow,
Thulgilo and her children sold land in Faenza to a soldier (vir strenuus)
named Pelegrinus for 110 solidi: the witnesses were all officials or nobles
associated with the palatial administration, including Julian the banker,
and more official names were dropped in the description of the land.
While the charter was not registered, Thulgilo states that she has properly
drawn up a letter of conveyance (dyploma vacuole), and that the sale was
‘performed with the customary imperial nummus and with Serapio vir
strenuus acting as scale-holder [libripens] and Opilio vir strenuus as the
special witness [antetestator], both of whom are present to sign below
as witnesses’. Perhaps this Gothic widow really did perform the ancient
Roman emancipation ceremony, striking a set of scales with a bronze coin
and announcing, according to Gaius, ‘These things I declare to be mine
by Quirinary right.’45 We have already mentioned King Witigis’ personal
notary registering his purchase in 540: surely he was aware of the political
winds that could sweep away his property rights. Another gesta from the
next year, after the conquest, records a sale of land by a Gothic cleric,
Minnulus, to Isaac, a soap-maker in Classe; Isaac promptly brought the
charters to the curia of Ravenna and asks for their registration ‘for the
protection of my ownership’ (ad munimen dominii mei).46 That the words
munimen (protection, defence) and monumen(tum) (monument, record,
document) become semantically fused together in the Ravennate papyri
(and continue so in Lombard Italy) further highlights the association
of documents with security of title. And it seems significant that this
phrase appears only in 540 and after, replacing the older, pre-Justinianic
formula that first requested the judge to change the tax records, and then
requested registration.47
45 P.Ital 30 (a. 539) (= ChLA XX 706). Witnesses: Serapio v.s. de scrinio cursorum; Opilio
strator inlustris potestatis; Candidianus vir laudabilis; Petrus collectarius, etc. Neighbours:
dromonarius, scutarius, praepositi dromonariorum, etc. Gaius, Inst. 1.119.
46 P.Ital 33 (a. 541) (= ChLA XXV 973). A fragmentary text (P.Marini 118) that used
the same formula may date a year earlier: see Tjäder 2:40–5. In P.Ital 20 (c. 590×602),
another Gothic widow cites ‘de sexu femineo Belliianus senatusconsultus’; that is, the
senatusconsultus Velleianum, probably dating from the mid first century AD; see H. Vogt,
Studien zum Senatus Consultum Velleianum (Bonn, 1952).
47 P.Ital 14–15 (a. 572), 31 (a. 540): missed by Niermeyer, Lexicon, 771, who cites Lombard
charters, the Liber diurnus and other slightly later Italian documents for the fusion of
meaning. Pre-Justinianic formula: P.Ital 10–11 (a. 489), 12 (a. 491); see Tjäder 2:67.
Finally, our longest and most complete gesta papyrus by far (107 lines,
four pages of printed text) also seems to reflect Byzantine–Gothic con-
flict. It records a hearing in 557 before the curia in Rieti (290 km south
of Ravenna), in which the Gothic widow Gundihild (inlustris femina)
requests the appointment of a legal guardian (tutor) for her two sons,
Lendarit and Landarit, who were still minors and unable to defend their
property against three Gothic soldiers (Adiud inlustris uir, Rosemud and
Gundirit uir magnificus) who had begun to encroach upon it. Gundihild’s
husband, Gudehals (uir inlustris), had died only days before, but had pre-
viously been defending the property in court. Gundhild herself was not
present at court; her well-drafted and somewhat moving request (petitio)
is read out before the curiales, who order it entered into the gesta. What
follows is the record of a protracted process, in very formulaic speech,
of interrogating the guardian, named Flavianus, and then his guarantor
(fideiussor), Liberatus, about whether they fully understand and accept
their obligations.48 Gundihild’s choice of a tutor and guarantor with
Roman names was probably wise: sometime between that year (557) and
565, Justinian ordered that all property belonging to the Goths, ‘not only
in the cities, but also in suburban villas and hamlets’, be confiscated and
given to the Ravennate church. Another of our surviving papyri docu-
ments this process: the fiscales of the state presented to the archbishop
(Agnellus, 557–70) a list of rents owed from the properties and the taxes
due from these.49 As it happens, yet another court case survives from 557
(not in gesta format): the Goth Gundila tried to regain property in Nepi
(near Viterbo, north of Rome), lost in the war with Byzantine forces in the
530s, by converting to Catholicism. Pope Vigilius and a Gothic bishop
intervened and ordered that Gundila’s property be restored to him, but
the Gothic resurgence under Totila (540s) meant Gundila lost it again
to a Gothic count, who gave it to his sons.50 And a papyrus charter of
551 records how the clergy of the Arian church of St Anastasius sold
their neighbouring swamplands to the defensor of the Ravennate church,
48 P.Ital 7 (a. 557) (= ChLA XX 712). It seems the copy we have is that of the tutor,
which he obtained from a Constantine, ‘vir devotus and comitiacus’, and which Con-
stantine obtained ‘from originals’ (ex autentico edidi). Comitiacus seems to have been a
military office: Cassiodorus, Variae 2.10, 5.6, 6.13, 7.31 (MGH AA 12:51, 147, 186–7,
218); Jones, Later Roman Empire, 253–5. The copy was written crosswise, rather than
lengthwise, and without columns, and hence not in regular gesta format.
49 P.Ital 2 (a. 565×700); Brown, ‘The Church of Ravenna’, 7. Confiscation: Agnellus,
Liber pontificalis ecclesiae Ravennatis 85–6 (MGH SRL 334–5). Taxes: Jones, Later Roman
Empire, 820–1; Wickham, Framing, 64–6.
50 P.Ital 49 (a. 557) (= ChLA XXIX 885). Date and reconstruction of events: P. Amory,
People and Identity in Ostrogothic Italy, 489–554 (Cambridge, 1997), 321–5, 382–3.
82 Nicholas Everett
Peter, in part to pay back a loan they had taken from Peter sixteen years
earlier.51
That so many of the Ravennate papyri record the demise of the Goths
in Italy says more about their preservation than about their produc-
tion: if Goths or Byzantines registered or simply recorded transactions
in times of trouble it is further proof that such practices were common
and considered secure. But the regional and chronological arc of these
papyri is narrow. It makes sense that the Ravennate church would pre-
serve together (perhaps in a soft-leather folder) documents for ex-Gothic
property, particularly since it was the direct beneficiary of Justinian’s
seizure of Gothic lands between 557 and 565. Their dispersal might be
blamed on the fire that burned the church archives around 700, in which
‘the flames consumed many documents, while many were seized and
hidden by wicked men’, who subsequently faced angry clerics hurling
anathemas at them in the local tavern to give the documents back.52 For
clearly Ravennati valued documents, and the traditions which produced
them, as is clear from the almost exclusive use of public scribes (tabel-
liones/forenses), who always qualified themselves as ‘of the city of Ravenna’
and maintained a ‘college of scribes’ (schola forensium).53 The use of gesta
municipalia to register transactions was another such tradition, but their
use in Ravenna was limited to lay and ecclesiastical elites of a capital
city that clung to Roman identity during a period of change and conflict,
as barbarian kings and the Byzantine generals who replaced them pro-
moted that identity to provide a veneer of continuity. The gesta among
the Ravennate papryi should not be used as evidence for such practices
elsewhere: the non-gesta documents in that collection provide the better
guide.
Visigothic Spain: legislation, formularies and slate

Visigothic Spain provides a startling array of evidence for lay uses of
documents in the legislation of Visigothic kings and in a formula-book,
but almost nothing of it survives in terms of the documents them-
selves, with the exception of five fragmentary documents on parch-
ment and, extraordinarily, about 160 texts written on slate from the
central western regions of the peninsula (mainly Avila, Segovia and
51 P.Ital 34 (a. 551) (= ChLA XX 704).

52 Agnellus, Liber pontificalis 134 (MGH SRL 365–6). Ravennate archepiscopal archive:
Tjäder 1:17–27.
53 P.Ital 24 (c. 650) (= ChLA XXIX 865): ‘primicerius schola forensium civitatis Ravennae
seo Classis’; Everett, ‘Scribes and Charters’, 57–9.
Salamanca).54 The following seeks to highlight some of the more illu-

minating references to the production and preservation of documents in
the law, and to correlate those with surviving examples in the Formu-
lae Visigothicae (thought to date c. 615–45), on parchment and on slate.
Topics addressed (in order) are forgery, wills, court documents, sales,
family law and finally estate management with reference to the slates.
The discussion is framed by two wider points being made: firstly, the
reliance of Visigothic law on late Roman legal tradition shows how the
latter’s preference for written proofs continued to be developed in a post-
Roman political order, and as such provides a model for understanding
the importance of documents elsewhere in early medieval Europe; sec-
ondly, the explosion in the number of surviving documents in Spain from
the ninth century onwards,55 which match the templates of the Formu-
lae Visigothicae and cite precepts from Visigothic law, reflects improved
methods of preservation, rather than greater appreciation of the use
and value of documents, which have secure precedent in this earlier
period.
Caution is needed in using the Visigothic law code as evidence, for
the political or ideological element of royal legislation always demands
consideration, and this is particularly true for the seventh-century Visig-
othic kings. Each promulgation of the law code known as the Book of
Judgements (Liber iudiciorum), in 654, 681 and 692, coincided with a
period of crisis for the monarchy, and there are large chronological gaps
(up to 75 years) between legislative efforts. Moreover, Karl Zeumer’s
critical edition of the laws published in the MGH (1902) was estab-
lished on principles that no longer seem secure against the advances of
modern research on the manuscript traditions.56 Nonetheless, Visigothic
54 Charters: ChLA XLVI 1398–1402 (referred to individually below). A contemporary

copy of Visigothic law survives: Vatican City, Bibliotheca Apostolica Vaticana, Reg.
lat. 1024 (= CLA 1.111; s. VII). Slate texts: Las pizarras visigodas: Edición cŕitica y
estudio, ed. I. Velázquez Soriano (Murcia, 1989); reissued with same numbering and
better commentary as Documentos de época visigoda escritos en pizarra (siglos VI–VIII),
2 vols. (Turnhout, 2000); reprinted as Las pizarras visigodas entre el latiń y su disgre-
gación: La lengua hablada en Hispania, siglos VI–VIII (Burgos, 2004) (hereafter Pizarra
# [no.]). Formulary: Form. Vis.; also in J. Gil, Miscellanea Wisigothica (Seville, 1972).
A tenth-century Catalonian collection may contain earlier material: M. Zimmermann,
‘Un formulaire du Xème siècle conservé à Ripoll’, Faventia, 4 (1982), 25–86. Following
convention, and because of the problems mentioned below, I refer to ‘Visigoth law’ (LV)
rather than Liber iudiciorum.
55 R. Collins, ‘Literacy and the Laity in Early Medieval Spain’, in Uses of Literacy, 109–33;
A. J. Kosto, Making Agreements in Medieval Catalonia: Power, Order, and the Written
Word, 1000–1200 (Cambridge, 2001); Kosto, ‘Laymen’.
56 Y. Garcı́a López, Estudios crı́ticos y literarios de la ‘Lex Wisigothorum’ (Alcalá de Hernares,
1996), who also provides a new edition of the laws of Egica and Wittiza; R. Collins,
Visigothic Spain, 409–711 (Oxford, 2004), 223–39.
84 Nicholas Everett
Spain produced the most voluminous and comprehensive legislation of

all the barbarian successor kingdoms, and it is replete with references
to the use of documents (usually termed scriptura or instrumenta) among
the laity. As one law makes clear, ‘Whatever is set out in writing in
contracts or agreements shall have full validity (plena firmitas)’, thus echo-
ing the sentiment of Emperor Leo cited earlier, and bestowing author-
ity on writing well beyond the Roman laws on which this precept was
based.57
The late Roman roots of Visigothic legislation are hidden by the
‘nationalist’ sentiment of seventh-century kings, often taken at face value
by modern scholarship with nationalist preoccupations of its own. The
Book of Judgements promulgated by King Recceswinth in 654 claimed
to have abrogated all previous versions of the law, but in reality it built
on a long tradition of royal legislation extending back to King Euric
(466–84) and possibly his father, King Theoderic (419–51), and contin-
uing through the kings Leovigild (569–86), Reccared (586–601), Sisebut
(611–20) and Chindaswinth (643–53). Their legislation (with the excep-
tion of fragments of Euric’s code) survives only in Recceswinth’s Book,
which attributes laws to only four kings (Recceswinth 88 laws, Chin-
daswinth 99, Reccared 3, Sisebut 2), while the greater part (some 315
laws) are simply called ‘ancient’ (antiqua), so their date is often impos-
sible to determine. Legislation continued under later kings, so that the
Book was updated by Ervig (681), Egica (692) and Wittiza (702–10).
Most of the Book of Judgements resembles late Roman law because Visig-
othic kings adopted it as their own quite early, as is clear from a surviving
section of Euric’s Code (about one-sixth of the original text) preserved
in a sixth-century manuscript.58 Admittedly, this section covers areas
in which Roman law excelled (entitled ‘On deposits and loans’, ‘On
sales’, ‘On donations’, ‘On inheritances’), but this has not stopped schol-
arly imagination concerning the missing ‘Germanic’ elements of Euric’s
Code, based on the comment of Isidore of Seville (writing a century and
half later) in his Historia Gothoroum that under Euric the Goths ‘began to
have ordinances of the laws in writing, for before this they were bound by
customs and habit’. Evidence within Euric’s Code and elsewhere would
deny that, and given what we now know about the accommodation of
(small numbers of) barbarians within the military framework of the late
empire, the fluidity of both Roman and barbarian ethnic identity, and
the stereotypes of late Roman ethnography, there is no need to look for
57 LV II.5.5, based on LRV II.9.1, and in turn CTh. 9.19.

58 BnF lat. 12161 (= CLA 5.626).
‘Germanic’ traits anywhere in Visigothic Spain, let alone the laws: like-
wise for Ostrogothic Italy.59
Moreover, Roman law was officially embraced by King Alaric II (485–
507) when he published in 506 his Lex Romana Visigothorum (or Breviar-
ium, as it was dubbed in the sixteenth century), an abridgement of the
Theodosian Code with later fifth-century imperial rescripts and excerpts
from earlier juristic texts such as the Pauli Sententiae and Gaius’ Institutes.
Alaric ordered that copies be given to all counts (comites, presumably
Gothic) and that no other law be used in their courts: if Goths did use
a separate law, they were a small percentage of the population, which at
this point included most of southern Gaul. As mentioned earlier, Alaric
II’s use of Roman legal advisers also to draw up a commentary on the
laws, in the form of clarifications and modifications called interpretationes,
continued the tradition of Roman praetorian prefects adapting law for
their provinces, as we saw in the Edict of Theoderic.60 Reliance upon
Roman models of governance and legal traditions is evident in the Book
of Judgements also, neatly organized into twelve books (probably imitating
Justinian’s Code; the Twelve Tables have also been suggested as a model),
which are subdivided into chapters and numbered laws (called eras), and
containing almost no element of identifiable ‘Gothic’ tradition or custom.
One (undated) law against forgery of royal documents reveals a society
steeped in traditions of written law, as we shall see in other spheres:
many not only write out royal orders themselves, but also promulgate them, and
publish documents confirmed by the signatures of notaries, whereby many things
have been introduced into the laws of our kingdom, and many provisions have
been written, or attempted to be added, which have not received the sanction of
our authority.61
The law goes on to warn royal, official and, surprisingly, ‘public notaries’
(notarii publici), who are not mentioned anywhere else in our sources, not
59 Isidore of Seville, Historia Gothorum 35 (MGH AA 11:281). Evidence of previous

legislation in CE 277, and Sidonius: Collins, Visigothic Spain, 227–9. The tenacity of
the Roman–German dichotomy is evident in P. D. King, Law and Society in the Visigothic
Kingdom (Cambridge, 1972), 8–9, who, despite defending the territoriality of Euric’s
Code, believes there were parts ‘exclusively Germanic’, citing Levy, West Roman Vulgar
Law, 125–6, but these traits are found in PSent. and similar sources. Ethnography:
Amory, People and Identity. ‘Germanic’: Collins, Visigothic Spain, 239–46 and passim.
60 Matthews, ‘Interpreting the interpretationes’; R. Collins, Early Medieval Spain: Unity in
Diversity, 400–1000, 2nd edn (London, 1995), 24–32; M. Rouche and B. Dumézil, Le
Bréviaire d’Alaric: Aux origines du Code civil (Paris, 2008).
61 LV VII.5.9. Undated: LV, p. 308n2 (neither antiqua nor Recceswinth). An antiqua
law (LV VII.5.1) on the same subject, which quotes PSent. 5.25 on Cornelian law,
distinguishes between honestiores, who lose half their property, and humiliores, who lose
their hand; cf. Lombard King Rothari, ER 243.
86 Nicholas Everett
to write decrees without royal approval. One of our (five) surviving parch-
ment fragments from this period appears to be a royal decree, revealing
traces of an otherwise unattested Visigothic chancery script, unless it is
clever forgery of the sort complained about above.62 King Chindaswinth
complained that new laws on forgery were needed because forgers
kept inventing new means of deception. Interestingly, one fragmentary
slate text contains two separate inscriptions, one mentioning that king’s
name (‘Cindasuindus dominus’) and possibly imitating a royal decree,
immediately followed by a writing exercise that addresses ‘assembled
citizens, senators and prefects’, before mixing phrases from the Disticha
Catonis and the Bible in some type of moralizing letter. The alliance
between Roman rhetoric and Visigothic kingship, so evident in the laws,
extended to scratching on slate in rural Diego Álvaro (Ávila), 120 km
from royal Toledo.63
The issue of forgery reveals how valued, and how common, documen-
tation was, for the Visigoths not only punished forgery severely but also
employed palaeographical analysis as part of court procedure to detect
forgery, even coining a new legal term for this practice (contropatio).
Forgery of royal documents (as above) incurred the penalty of 200 lashes,
a scalping and amputation of the thumb, whatever the social rank of the
forger. Forging, suppressing or tampering with any (non-royal) docu-
ments automatically incurred 100 lashes, and additional punishments
were inflicted according to social rank: the upper ranks (potentiores) lost
a quarter of their property; lower and poorer classes (humiliores sane vil-
ioresque) had to ‘sign a profession to the effect . . . that they shall forever
be the slaves of those they deceived’.64 The stakes were high, for as we
saw earlier, written contracts were considered to have plena firmitas. This
even extended to doctors and their patients, who, after the initial diag-
nosis was given, signed an agreement on the method of healing and the
price to be paid – or not paid, if the patient did not survive (fortunately no
such cases do).65 Doubts about the authenticity of documents could be
resolved in court by palaeographical analysis (contropatio), best described
in a law concerning documents of deceased persons:
All documents (omnes scripturae), where he who made them and the witnesses to
the same are dead, and in which the signature of the former and the attestation of
the witnesses appear, when brought into court to be verified, may be proved by
comparison of their seals and signatures with those of other documents; and the
proof shall be sufficient in this investigation, if the seals and signatures of three
62 ChLA XLVI 1399, s. VIIin . The script is similar to that of Merovingian royal charters.
63 LV VII.5.8 (Chindaswinth). Pizarra #59, and Las pizarras (2004), 105–10.
64 LV VII.5.1, 9 (royal); VII.5.2.
65 LV XI.1.3–4. Note also the specified fees for training apprentices (5) and removing
cataracts (7).
or four other documents, when introduced, shall be evidently those of the parties
in question.
Recceswinth clarified further that documents (instrumenta chartarum)

for the comparison are to be found in the household archives (in scriniis
domesticis) of the deceased; if these proved insufficient, other examples of
similitudo scripturarum were to be found, or the case fell.66 This is the only
mention of lay archives in the laws, but their existence is clearly assumed,
and we know now that their format extended to chunks of slate.67 The
importance of signatures is confirmed by the spectacle of King Wamba
(c. 679) forcing rebels to confirm their handwriting on oaths of loyalty
signed at his election, and then showing the written oaths the rebels
later signed to support the usurper Paul.68 Soon after (681), King Ervig
mandated that every converted Jew sign a written oath or ‘profession’
abjuring his past faith, the text for which was provided in the laws, and
then place it upon an altar with his own hands before the local bishop,
who preserved it the archives of the church.69
The laws against forgery were particularly concerned with wills,
another sphere in which the Visigoths borrowed heavily from late Roman
legislation. Persons over 14 years of age could dispose of their property
as they wished, though there were some limits and reservations, and the
law envisaged four types of written will that reflect different degrees of
literacy or ability among the testator and/or witnesses: one with subscrip-
tions by the testator and witnesses; one with only the mark (signum) of
the testator, guaranteed by the subscription of a witness, who later was
forced to swear to its authenticity; one used when (presumably because
of illness) the testator could not even make a mark, requiring a spe-
cial representative and witnesses to swear before a judge; and finally, a
holograph will, written and subscribed by the testator in the absence of
witnesses, probate of which required comparison of the handwriting in
the will with at least three specimens of the alleged testator’s script before
66 Quotation: LV II.5.15 (Chindaswinth). Recceswinth: II.5.17. See also I.2.23; II.4.3,

5.15. Contropatio (from contrapactio), can also mean compensation: cf. III.1.5, IV.5.3,
VI.1.5; K. Zeumer, ‘Zum westgotischen Urkundenwesen. I: Subscriptio und signum.
II: Die Schriftvergleichung contropatio’, Neues Archiv der Gesellschaft für ältere deutsche
Geschichtskunde, 24 (1899), 13–38.
67 The only other archives mentioned are an archivum ecclesiae (for Jewish professions to a
priest: LV XII.3.28; for examples given in the code, see below, n. 69), and arca publica
(V.4.19) for the state treasury.
68 See J. Martı́nez Pizarro (trans.), The Story of Wamba: Julian of Toledo’s Historia Wambae
regis (Washington, DC, 2005), 237–8. Oaths in Visigothic law: King, Law and Society,
113–15.
69 LV XII.3.14–15. On this despicable ‘twelfth book’, see Collins, Early Medieval Spain,
132–4; King, Law and Society, 130–45.
88 Nicholas Everett
a judge or a bishop, who then validated the will by adding his subscrip-
tion. The casual prescription for the will’s contents reads very much like
that of Constantine mentioned above: ‘Let the testator write out how he
wishes to dispose of things, and let it be especially noted who he wishes
to act as executor, and what of his property he chose to give to whom.’
The only formal requirement for a will of whatever type was that it record
the exact date of production (day and year). Even an oral will needed to
be confirmed in court by witnesses, who swore to its veracity and then
subscribed their oaths before a judge.70
Visigothic courts operated on the assumption that documents would be
provided and produced, not only to resolve cases but also to introduce
them. Summons to court required a letter impressed with a judicial
seal and delivered to the defendant in the presence of a witness.71 In
a case where torture (quaestio) might be employed, the Visigoths took
over the late Roman practice of inscriptio, whereby an accuser presented
to the judge a subscribed statement in which he promised to suffer the
same penalties as his opponent if he failed to prove his case. Moreover,
the accuser had to submit secretly to the judge a written account of
the events of the crime, which the judge would then compare to the
confession obtained under torture.72 As with Roman law, there were
considerable disincentives for an accuser or a judge to turn to torture.
Nonetheless, these safeguards revolved around the use of documents, and
the acquitted left the court armed with documents issued by the judge
to prevent the case from being reopened. In important cases the judge
himself was compelled by law to write out and subscribe two identical
accounts of the proceedings and his verdict, one for each party. In fact, an
example of this survives in the Formulae Visigothicae, which also records
that one party cited Visigothic law by book, chapter and verse (eras).73
In all cases, however, the judge’s decision was required to be written out,
and the judge was to keep a copy for himself; otherwise, the decision
was considered invalid – a late Roman law we find also in the Edict of
Theoderic.74
In lesser cases, the parties received a sworn protocol of the evidence of
the witnesses of the victorious side; indeed, one of our longest slate texts
70 Wills: LV II.5.1–17; also IV.2.20; II.5.10, 12 (oral); King, Law and Society, 244–8.
71 Summons: LV II.1.19–20; King, Law and Society, 93–6.
72 Inscriptio: LV II.3.4, IV.1.2, VIII.1.1.5. Late Roman inscriptio: Harries, Law and Empire,
101, 120–2.
73 LV II.1.24; Form. Vis. 40: ‘sententias legis libri ill[e] protulit, legam illam, qui est sub
titulo illo, era illa’. Finish: ‘ad singula decernentes in hanc iudicii paginam inseruimus,
quatenus futuris temporibus iustitiam habens congaudeat’.
74 LV II.1.25; CTh. 4.17, 11.3.40; ETh. 64.
appears to be such a document. It is entitled Conditiones sacramentorum,

and its format is replicated in a formulary template with the same title.
The slate records a dispute over an exchange of horses in which a cer-
tain Lolus was ordered by four judges and two vicarii to swear an oath
accepting the terms of their decision. The oath itself is directly mirrored
in the formulary template and is in strict accordance with Visigothic law:
in the absence of evidence or witnesses, the judge was to administer an
evidentiary oath, which seems to be exactly what happened in this case.
In taking the oath, Lolus emerged as the victor and departed from the
court with a written copy of it as insurance against future prosecution.75
In another slate recording a court case (placitum) held before at least two
vicarii, a certain Gisadus took an oath in which he pledged a pig as a secu-
rity for the debt he had incurred to Aninaus and Teudoteus for 10 solidi.
Again the terms used and the general framework of the case reflect two
provisions on securities in Visigothic law, and other fragmentary slates
recording court cases also reflect the legal language of Visigothic law or
the Roman sources on which they are based.76 Two parchment fragments
further reflect court activity: one seems to be the order of a count called
Medema, in handsomely executed cursive script (unfortunately, his seal
was removed), instructing judges to obtain an oath in court (c. 697); the
other is an oath concerning agricultural goods (c. 704).77
While it is true that for sales, exchanges and donations writing was not
absolutely essential and witnesses were considered equally valid, it is often
stated throughout the laws that writing offers greater security, couched
in expressions such as firmitas/dignatas/veritas scripturae, veridica/stabilis
scriptura and so on.78 Recceswinth even legislated against people using
documents to camouflage conditions that were illegal.79 Euric’s Code
had already determined that ‘a sale contracted through writing has full
75 Pizarra #39, Form. Vis. 39; cf. LV II.1.12–13.

76 Gisadus: Pizarra #92; cf. LV V.6.3–4. The property to be pledged should be decided by
a judge or three respectable persons (trium honestorum uirorum); the slate is fragmentary
but may refer to three or four vicarii present. Other placita: Pizarra #12–14 (CE 285,
PSent. 1.12.3; LV III.4.10, V.4.14, etc.), 18, 43 (Form. Vis. 2, 6, 38).
77 ChLA XLVI 1398, at the petition of a deacon and priest; ChLA XLVI 1400 (i.e. the
superior script of the royal decree mentioned above), reread and signed by the actor
(Cixa) and several witnesses.
78 I omit here (for reasons of space) discussion of manumission, which likewise could
be performed per scripturam or among witnesses with equal validity (LV IV.2.18, 5.12;
V.1.13–14, 6.6–7, 7.1–20, etc.), though V.7.14 addresses the use of a document to
stipulate the terms of the freedman’s peculium. The templates for manumission in the
formulary (Form. Vis. 1–6) show ecclesiastical influence, and no examples survive in the
slates (though some fragments might be such) or parchments.
79 LV II.5.10, ‘scripturas conficit amplius quam lex iubet’; cf. III.1.5, IV.5.1. Firmitas: e.g.
LV V.4.3. Sales with scriptura have plena firmitas, dating back to CE 286. Written and
oral evidence: King, Law and Society, 105–11.
90 Nicholas Everett
validity’, and we see similar statements privileging writing in the Formu-

lae Visigothicae: one template quotes and then updates a precept of the
Pauli Sententiae: ‘Granted that “written instruments of sale are not essen-
tial in contracts of sale and purchase defined in good faith” (= PSent.
2.18.10 interp.), however, it confers security for the buyer if the specific
terms (definitio ipsa) are confirmed with the solidity of writing.’ Other
templates for sales make similar statements.80 Moreover, many examples
of sales can be found among the slate texts, and these repeat the legal lan-
guage for such transactions found in the formulary and in Visigothic law.
Unfortunately, most of these are very fragmentary, but the small sums of
money and property they record show that even minor transactions were
considered worthy of scratching into slate and preserving.81
One slate in particular shows how they were used for archival purposes.
On one side the slate preserves a charter recording the sale of land by
Gregorios to his ‘lord and nephew’ Desiderius; the other side preserves
a document entitled Professio de seruitudine, in which a certain Unigildus
testifies that he arrived in the area to initiate a lawsuit relating to fraud,
and had adduced Froila as a witness at the house of Desiderius. Froila
then, we are told, led Unigildus to another house, where he was set to
work ‘in the strawberry fields’ (in fragis), and received in return certain
objects from Desiderius.82 It seems that Desiderius, by recording the
testimony of Unigildus, could prove that he paid for the work that Froila
had performed by proxy, thereby avoiding legal problems that might arise
once Unigildus had left the area. Here one slate has functioned as the
record of two different transactions that could be used by Desiderius in
court to verify his claims.
Before turning to the issue of documents for estate management raised
by these slates, it is worth noting that Visigothic law excelled other bar-
barian legislation in the use of documents in family law, particularly for
marriage and divorce. Donations, either causa mortis or inter vivos, could
be made with or without documentation, but documents were absolutely
80 CE 286 (plena firmitas); Form. Vis. 13. Sales and writing: Form. Vis. 12 (‘ut scripturae
firmitas emittatur’), 27 (cartula commutationis, ‘per scripturae conscribere tramitem’),
28 (‘licet mutuo . . . consensu . . . sed ad posteritatis memoriam reservandam adicitur
testimonium literarum’); cf. Form. Vis. 11, only ‘iuratione confirmo’.
81 Sales: Pizarra #8 (land for 3 solidi), 19, 30 (wine), 41, 66 (wine), 107 (wine), 121,
etc.; too fragmentary are #9, 15–17, 22, 26, 37–8, 44, 60–1, 64–5, 70, 72B, 73, 80,
94, 98, 112, 122, 126, 130, 136–8, 142, 144, 150. A fragmentary sale or exchange
of cows is also preserved on parchment (ChLA XLVI 1402, s. VIIex ), as is a car-
tola commutationis vel resepenationis subscribed by no fewer than six different hands,
none giving any qualification, hence they are presumably laymen (ChLA XLVI 1401,
post-687).
82 Pizarra #40; the charter uses language similar to Form. Vis. 11. My thanks to Graham
Barrett for sharing with me his observations on this text and the slates in general.
necessary for donations between husband and wife.83 A formal betrothal

(disponsatio), the second stage of marriage proceedings after the suit (peti-
tio) and before the wedding (nuptiae), involved legal obligations of estab-
lishing the ‘dowry’ – not the traditional Roman type, but rather one
conveying property from the groom’s side to the bride, common to late
Roman society (donatio ante nuptias) and Germanic traditions (‘bride-
price’, ‘morning gift’). This needed to be witnessed, either in writing
or by the delivery of a ring as pledge, but the benefits of a dowry con-
tract (dotale scriptura, or dotale tabula) ensured ‘future conjugal dignity’.84
Indeed, the templates for several such contracts survive in the Formulae
Visigothicae (and Merovingian formularies also), including one in verse
form for a Goth marrying a girl ‘from senatorial stock’ and another that
cites ‘Papean, Popean and Julian law about the statutes of marriage’,
which date back to Augustus.85 While divorce was only acceptable on
certain grounds (mainly adultery), the law notes that ‘reckless men’ try
to obtain divorce by producing false documents or coercing their wives
into signing them.86
We have discussed how slate texts were used in the legal contexts
mentioned above, but a final word needs to be said about their role
as documentation for estate management and their place in the his-
tory of documentation.87 The use of slate for texts is known in a few
other places (such as tenth-century Ireland and fifteenth-century Eng-
land), but can be compared to the use of birch bark at Novgorod or
wood and ostraca in Africa as constituting a locally convenient medium
for preserving information. Nearly all of our slate texts come from the
northern Meseta region, from around forty-two different sites located
along the Salamanca-Ávila border, most in contexts – apparently from
villas or walled castros – without any obvious ecclesiastical presence, such
83 Husband–wife: LV V.2.7. Visigothic donations: Levy, West Roman Vulgar Law, 138–42,
164–6.
84 LV III.1.9 (possibly Ervig); also III.1–8.
85 Form. Vis. 14–20.14: ‘ex lege Papeam Popeam et ex legem Iulianum quae de maritandis
ordinibus’. Lex Papia Poppaea (c. AD 9), was a supplement to the Leges Juliae of 18 and
19 BC concerning marriage and adultery: S. Treggiari, Roman Marriage: Iusti Coniuges
from the Time of Cicero to the Time of Ulpian (Oxford, 1993); the same laws are alluded
to (though not named) in Marculf 2.15 (see MGH Formulae 85n1). Form. Vis. 20:
‘ex stirpe senatus’, also suspiciously mentioning a morgincap. Cantabria maintained a
council called a ‘senate’, and the term appears in other places: King, Law and Society,
46n5; Wickham, Framing, 94, 223. Pizarra #67 may be a marriage contract.
86 LV III.6.1–3. King, Law and Society, 235, suggests divorce by mutual consent, but
III.6.2 denies it.
87 As the editions of Velázquez Soriano replaced previous attempts, so does her accom-
panying commentary on the slates. A succinct overview is Collins, Visigothic Spain,
170–3.
92 Nicholas Everett
as that of Diego Álvaro, which has yielded the largest number (fifty-three
slates).88 Despite the different archaeological contexts, the slates display
striking similarities of form and content, and nearly all date from the
period 560–700, pointing to a shared literate culture of legal norms and
estate management removed from any major urban or ecclesiastical cen-
tre. Why they cease with the Visigothic kingdom is difficult to answer:
the Arab conquest (711), which turned the Meseta into a frontier zone
for two centuries, and which reconnected Spain to a wider world in
which papyrus and eventually paper could be imported, may be part of
the explanation. The argument that they were primarily taxation records
that had no purpose after the eclipse of the Visigothic kingdom has no
weight.89
As is known, the slates comprise a range of different purposes and
texts besides those mentioned above, including educational and religious
texts (alphabets and writing exercises; prayers and talismans), and, of
course, many slates have little or no writing at all, merely recording num-
bers alongside names or types of produce, such as a ‘record of cheeses’
(notitia de casios) or ‘record of sheep’ (notitia de ueruices). These probably
concerned rents, as suggested by other slates with phrases such as ‘he has
paid’, ‘I keep back’, ‘I have given firewood’, and so on, or other estate
business that needed to be recorded (‘John paid 33 modii for the horses’),
or lists of animals recording their age and sex.90 Contemporary confir-
mation that these practices were routine for lay landowners is found in
the seventh-century Life of St Fructuosus, which records how the saint’s
father visited his estates in Bierzo (200 km from the slate finds) to receive
reports on his flocks, reports that probably looked like our numerical
slates, whatever material was used.91
While the spelling and grammar of the slates is far from classical and
well on its way to proto-Romance (even more so than the Ravennate
88 Five find sites (yielding twenty-four slates altogether) contain archaeological traces of a
chapel (e.g. La Aceña de la Fuente and Navahombela), but they may be villas with a
private chapel, as was common in late antique Spain, or villas later reused as churches:
Las pizarras visigodas (2004), 23–41. The archaeology and architecture of Visigothic
Spain is notoriously difficult and in need of serious revision: Collins, Visigothic Spain,
174–222.
89 I. Martı́n Viso, ‘La sociedad rural en el suroeste de la Meseta del Duero (siglos VI–VII)’,
in Martı́n Viso and G. del Ser Quijano (eds.), Espacios de poder y formas sociales en la
edad media (Salamanca. 2007), 171–88. Spain and trade: O. R. Constable, Trade and
Traders in Muslim Spain: The Commercial Realignment of the Iberian Peninsula, 900–1500
(Cambridge, 1994), 30–2.
90 Cheese: Pizarra #11. Sheep: #97. Exprendit, recondo and horses: #5. Dedi licias: #125.
Animals: #53–4, 75–7.
91 Vita sancti Fructuosi 2 (ed. M. C. Dı́az y Dı́az [Braga, 1974], 82; trans. A. T. Fear, Lives
of the Visigothic Fathers [Liverpool, 1997], 124).
papyri and Lombard charters), they reveal a degree of practical literacy in

the remote countryside that was difficult to imagine before their discovery
(in the 1940s and 1950s) and subsequent publication. We have seen this
in the legal texts mentioned above, but it is also apparent in a handful of
letters on slate relating to estate management. By far the most complete
letter (fourteen lines of text on both sides of a slate c. 89×168 mm) is
that of Faustinus, who writes to a ‘Lord Paul’ to request that he gather
the olive harvest in the customary manner (ut comodo consuetudinem),
warn him of potential fraud by his slaves or tenants (mancipii), ask him to
verify the receipt of certain goods with his seal-ring (sigilla de tuo anula),
and transfer certain dependants (all named) to different locations. Even
more striking than the mention of a seal-ring (referred to several times
in Visigothic law) is the impression of an administrator telling a lord
(dominus) what to do and how to do it.92 Another letter, unfortunately
very fragmentary, also appears to be a report to a landowner, perhaps
one who resides in or holds office in Toledo, for it mentions the city (the
only location named among all the slates), money (tremisses) and even a
Gothic loanword for profit (gannatione), along with a list of names and
quantities in the manner of the notitiae mentioned above.93
Conclusion: towards Lombard Italy

Lombard Italy presents problems of evidence that are similar to those of
Visigothic Spain: legislation and other evidence offer strong indications
that there was a considerable amount of documentary activity going on,
but hardly a scrap of it survives.94 Our surviving charters begin in the
eighth century, which tally over 370, and the numbers double, even triple
in some places, in the ninth century. Yet, twenty years after their initial
conquest of north and central Italy, and once monarchic rule was secured
(568–90), Lombard kings adopted late Roman trappings of court culture,
issued royal charters in chancery style, and engaged in diplomatic corre-
spondence with Rome, Ravenna and Constantinople. The Edict of King
Rothari, published in 643, on the eve of an aggressive (and successful)
campaign against the Byzantines, contained far fewer references to doc-
uments than Recceswinth’s Liber iudiciorum (654), but it, too, adopted
much of late Roman property law, and was clearly intended to be used
92 Pizarra #103. Seal-rings: LV VI.6.2 (royal); II.1.19, 20, etc.; King, Law and Society,
94n3. Other letters: Pizarra #61, 68, 111–12, 134–5.
93 Pizarra #75.
94 What follows is fully documented in N. Everett, Literacy in Lombard Italy (Cambridge,
2003); Everett, ‘Scribes and Charters’; Everett, ‘Literacy and the Law in Lombard
Government’, EME, 9 (2000), 93–127.
94 Nicholas Everett
in courts, as it certainly was in our eighth-century evidence, along with

later legislation, most notably that of King Liutprand (712–44), which
contains many references to the use of documents. There is no doubt that
Rothari’s Edict is full of Lombard customs that derive from a more Ger-
manic world – wergelds, compositions, oath-helpers (a type of character
witness) – and glosses of crimes, practices and legal categories with Old
High German words (plotraub, fulcfree, gairethinx, etc.). There is plenty
here for the strain of historiography that highlights Lombard difference,
or argues that their kingdom represents a complete break from the Roman
past, unlike Odovacar or the Ostrogoths. True, Rothari envisages oral
procedures (pledges, oaths, sureties) for most transactions, but his few
references to documents are telling: two laws, one insisting on documen-
tation for leased land (for later use in court if needed), and another pre-
scribing amputation of the hand for forging charters, book-end a section
concerning property, and immediately before this Rothari recommends
the use of a charter to secure manumissions performed with the ‘four
roads’ ceremony, a ritual involving a whip and arrow, and described with
six Old High German words.95 The use of such terms and practices was
undoubtedly in part ideological: Liutprand introduced Germanic words
where Rothari was content to use Latin, yet his legislation was far more
Roman in character. In any case, the eighth-century evidence, both leg-
islation and documents, shows that Lombard elites adapted to written
forms of administration and legal instruments quickly. To illustrate, let
us end with an archive: a list of documents belonging to a Lombard offi-
cer, Alahis vir magnificus, active in Tuscany c. 720–30, mentions about
fifty charters, a dozen of which were royal diplomas of Liutprand.96 That
we only know about this archive because it was later preserved by a nun
(and her daughter), and then by the bishop of Pisa, is the same pattern of
preservation we have seen in Ravenna, and see everywhere else in early
medieval Europe, as the chapters in this volume amply show.
95 ER 227 (libellus for leases), 243 (amputation), 224 (manumission). Royal charters: CDL
III 197. Private charters: CDL I–II.
96 Alahis and Pisan list: CDL II 295 (= ChLA XXV 808); on the date, Everett, ‘Scribes
and Charters’, 39–40, 80.
5 The gesta municipalia and the public
validation of documents in Frankish Europe
Warren C. Brown
The degree of continuity or change in documentary practices after the disap-

pearance of the western Roman Empire is a question raised with particular
force by continued references to the gesta municipalia. Although their core fis-
cal function had vanished along with the Roman state, references to the gesta
continue to be found in Frankish formularies – that is, collections of models
for different kinds of legal document – dating from the late eighth and ninth
centuries, and referring to procedures undertaken between the late sixth and the
ninth. The deliberate inclusion, and sometimes adaptation, of gesta formulas
in manuscripts of formula-books, and references in charters, show that even
though now used by different institutions in different contexts, the gesta’s estab-
lished language and procedures continued to be adopted, as one among several
options for the pubic validation of transactions and their documents.
For students of documentary practices in the early medieval West, the

gesta municipalia have long been an important touchstone.1 As Everett
notes, the disappearance of these civic document registers has been
linked in particular to a putative transition from a public to a private
documentary culture as Roman Late Antiquity gave way to the early
Middle Ages.2 Pushing against this narrative, Everett shows us that the
gesta reflected a very particular context. They were never intended to
1 H. Cancik and H. Schneider (eds.), Brill’s New Pauly: Encyclopaedia of the Ancient World,
16 vols. (Leiden, 2002–10), s.v. gesta (V, 827–8); P. Johanek, ‘Gesta municipalia’, in
Lexikon des Mittelalters, 10 vols. (Munich, 1980–99), IV, 1408; B. Hirschfeld, Die Gesta
municipalia in römischer und frühgermanischer Zeit (Marburg, 1904), 8–15; O. Redlich,
Die Privaturkunden des Mittelalters (Munich, 1911), 8–10; P. Classen, Kaiserreskript und
Königsurkunde: Diplomatische Studien zum Problem der Kontinuität zwischen Altertum und
Mittelalter (Thessaloniki, 1977), 33, and Classen, ‘Fortleben’, 42; I. Wood, The Merovin-
gian Kingdoms 450–751 (London, 1994), 204, and Wood, ‘Disputes in Late Fifth- and
Sixth-Century Gaul: Some Problems’, in Settlement of Disputes, 7–22, at 12–14. Of J.
Barbier, ‘Pouvoirs et élites dans le monde franc (VIe –XIe siècle). Matériaux pour servir
à l’histoire des élites des cités (VIe –IXe siècle): Le Dossier des gesta municipalia, étude
no 2’, unpublished mémoire d’HDR, Université de Paris I–Panthéon-Sorbonne (2009),
only Part I, ch. 3, and Part II were available to me at the time this volume went to press.
2 Above, 71.
95
96 Warren C. Brown
serve, nor did they serve, as public document repositories, on which

document users depended to safeguard their documents and through
which, therefore, document use was mediated. Their raison d’être was
fiscal; they represented, above all, efforts by successive imperial govern-
ments to keep track of properties that could be taxed and of those who
were responsible for the tax burdens.3 As the late Roman state dissolved,
the gesta lost their reason for being. They remain visible in Ravenna
into the early seventh century only because Ravenna was a centre of late
Roman and Romanized barbarian administration and power in a time
when uncertainty drove both Roman and Gothic elites to cling to Roman
traditions, and because Ravenna’s correspondingly powerful church pre-
served records of gesta protocols in which it had been involved.4
The gesta municipalia have nevertheless left us with a puzzle that needs
to be solved in order to understand why our evidence for documentary
culture in the post-Roman West looks the way that it does. The gesta left
a Nachklang that resonated in the West for centuries, especially in the
Frankish kingdoms north of the Alps. References to the gesta in Frankish
sources are abundant and surprisingly persistent; they stretch from the
Merovingian period deep into the Carolingian period and beyond. These
references appear in a variety of sources, including wills and charters.5
The most numerous and detailed, however, are to be found in the Frank-
ish formula collections.6 These are albums of document forms; that is,
collections of texts that could serve as complete templates for documents
and letters, or as sources for language. Some of them were obviously
created as models for typical or frequently needed documents. Others
were apparently drawn from actual documents that had some interesting
or important characters or features; they were turned into formulas by
having some or all of their case-specific information removed. The for-
mula collections survive in manuscripts ranging in date from the eighth
through the tenth centuries that span the entire Frankish world both east
and west of the Rhine. The document forms in them for the most part
stem from roughly the same period, but many can be traced back to the
seventh and even the sixth centuries.7
3 Above, 64, 72. See also Everett, ‘Scribes and Charters in Lombard Italy’, Studi Medievali,
41 (2000), 39–83, at 43–4, and Wickham, Framing, 70.
4 Above, 64–5, 72–82. 5 Barbier, ‘Pouvoirs et élites’, Part II, pp. 114–97.
6 Edited by K. Zeumer as MGH Formulae.
7 Rio, Legal Practice; R. Buchner, Deutschlands Geschichtsquellen im Mittelalter, Beiheft, Die
Rechtsquellen (Weimar, 1953), 49–55; Classen, ‘Fortleben’, 15; Settlement of Disputes,
271 (glossary, s.v. ‘Formula and Formulary’); McKitterick, Carolingians, 25; I. Wood,
‘Administration, Law and Culture in Merovingian Gaul’, in Uses of Literacy, 63–81,
esp. 64–5; C. Lauranson-Rosaz and A. Jeannin, ‘La résolution des litiges en justice
durant le haut moyen âge: L’exemple de l’apennis à travers les formules, notamment celles
d’Auvergne et d’Angers’, in Le règlement des conflits au moyen âge: XXXIe Congrès de la
S.H.M.E.S. (Angers, juin 2000) (Paris, 2001), 21–33, esp. 23–5; Brown, ‘Documents’.
Francia: gesta municipalia 97
Some of the formulas that refer to the gesta do so only in short, formu-
laic phrases, as, for example, when a formula for a transaction mandates
that the record be submitted to the gesta.8 Others, however, represent
complete gesta protocols that describe the entire process of submitting
a document to the gesta in detail. These latter formulas give a very late
antique impression; that is, they refer specifically to such things as a city’s
municipal assembly or curia, to officials bearing late Roman titles and so
forth. Most of them describe transactions by laypeople (or at least people
who are not explicitly identified as clerics) for the benefit of laypeople.
Only a few concern transactions in which a layperson gave, or might have
given, something to a church or monastery.9 On the surface, then, they
would indeed seem to reflect tradition-minded Frankish scribes carrying
forward relics of late Roman documentary culture into a world to which
they no longer applied.
And yet, these formulas kept getting copied, and recopied, for a very
long time. Though many of them were originally Merovingian, they all
survive in manuscripts from the Carolingian period: from the late eighth
century in a few cases, in most cases from the ninth. If the gesta municipalia
themselves were by this point dead, and if the context that had produced
them and the culture of lay document use of which they were a part had
disappeared, why did anyone go to the (often significant) effort to copy
complete gesta protocols into the surviving manuscripts? Tradition might
explain the persistence of formulaic phrases, and the formulas in general
are indeed rife with phrases that clearly represent vestigial memories of
long-vanished practices.10 But for Frankish scribes to go to the effort
to copy and recopy complete formulas for this particular document type
over such a span of time suggests to me that they were motivated by more
than tradition.
Their efforts can be explained by going back to late antique Italy and
looking more closely at how the gesta served the needs and interests not
only of the late Roman and Gothic authorities but also of the people
who brought their documents to the gesta for registration. If the author-
ities used the gesta to keep track of tax obligations as property changed
hands, those who came to the gesta achieved something equally valuable:
they had their legal transactions validated and secured by having the
8 See, inter alia, Form. Andec. 41: ‘gestis municipalis sit oblegatum’; Form. Tur. 20: ‘et
adhuc mihi inserere placuit, ut hoc mandatum civitate illa cum curia publica, ut mos
est, gestis municipalibus facias alligari’.
9 What I mean by ‘might have given’: three formulas give churches or monasteries as
options; that is, a person N gave property to the church N, or to his nephew N, or to
his bride N.
10 For example, the tag stipulatione subnixa at the end of charter formulas, or the references
to manumitted persons enjoying the freedoms of Roman citizens, both of which are
discussed below, 133, 135.
98 Warren C. Brown
documents in which those transactions were recorded publicly ratified

by the civic authorities and by receiving from the authorities a document
saying that the necessary steps had been carried out.
The Frankish gesta formulas, and especially the evidence of their
manuscript transmission, suggest strongly that a desire to secure trans-
actions in this way persisted in the Frankish kingdoms even after the late
Roman state that had created and maintained the gesta had dissolved.
Accordingly, some of the procedures that had been associated with the
gesta persisted there as well. Many people in both Merovingian and Car-
olingian Europe continued to resort to what alongside the civic document
registers had always formed an essential part of the gesta, namely, pub-
lic, ceremonial procedures, carried out in an authority-bearing forum,
that validated documents recording legal transactions. That these proce-
dures were descended from those associated with the gesta municipalia is
revealed by the fact that in some places in Frankish Europe, even in the
ninth century, formulas describing the gesta process were still considered
the right and proper way to capture these legal rituals in writing.
Among those who chose to validate and secure their documents this
way were laypeople as well as monks and clergy. This tells us that mem-
bers of both groups still saw written documents as important. When we
compare the gesta formulas to actual surviving transaction records from
both the Merovingian and the Carolingian periods, it becomes clearer
how and why documents were important. A legacy of late Roman doc-
umentary culture remains clearly visible in the Frankish West: many
people, both lay and clerical, still believed that recording their transac-
tions in writing and having the resulting documents publicly validated in
an authority-bearing forum was a powerful way to secure them.
The gesta municipalia in the Italian papyri

Despite substantial variation in their subjects and in the formulaic lan-
guage that they use, the papyri from fifth-, sixth- and early seventh-
century Italy give a fairly consistent picture of the gesta municipalia.11
The picture begins with someone who wants to carry out some sort of
legal transaction, such as making a new will or verifying the validity of
an old one, giving or selling property, acknowledging receipt of a debt
11 The relevant texts are Tjäder I 4–5, 7, 8, 10–11, 14–15, 18–19, 21; II 29, 31, 32,
33. Following Tjäder’s lead, I have chosen to label the papyri under discussion here
as ‘Italian’ rather than ‘Ravennate’ because, although most of them were undeniably
produced in Ravenna or mediated through the archives of the city’s cathedral church,
there is some uncertainty about whether all of them were. See Tjäder 1:21–3.
payment, or setting up a ward for a minor.12 This person draws up the

appropriate document, or has it drawn up by a scribe, in the presence
of witnesses who then add their signatures to it. Occasionally, the docu-
ment will include a clause asking that it be submitted to the gesta (gestis
municipalibus allegare).13 Next, either the originator of the transaction or
a representative, or the intended beneficiary or a representative,14 goes
with the document to the city whose authorities had jurisdiction over
the transaction by virtue either of where the principals lived or where
the property at issue was located. There they bring the document to an
assembly headed by one or more authority figures. In most cases the
assembly can be identified as the city’s municipal council.15 The assem-
bly heads are generally called ‘magistrate’ (magistratus), or occasionally
‘advocate’ (defensor);16 an exalted exception is the praetorian prefect who
headed an assembly in Ravenna in the middle of the sixth century.17 The
papyri also mention officials or functionaries, such as notaries or record
keepers (exceptor, notarius, chartarius).18
Once the petitioners come before the assembly, the magistrates ask
them what they wish. The petitioners reply that they would like their
document taken up, read out loud and entered into the records (acta,
gesta). The heads of the assembly then order the document to be taken
up, displayed openly and read out loud. In some cases, the original
witnesses are present and testify to their role in the document’s creation.19
If the principal responsible for drafting the document is not present, the
assembly heads accept the document from his or her representatives and,
at their request, order a delegation sent to the principal to verify orally
that he or she was responsible for creating it, that it said what he or she
wanted it to say, and that he or she had gathered the witnesses whose
signatures were on it; in this situation the witnesses also sometimes testify
12 In the sequence given, Tjäder I 4–5, 14–15; II 29; I 8, 7.

13 Tjäder I 8. See Everett, ‘Scribes and Charters’, p. 75.
14 Originator/representative: the embedded wills in Tjäder I 4–5, 7; beneficiary/
representative: Tjäder I 8, 10–11.
15 The members of the assembly are referred to variously as curiales, principales (or prin-
cipales viros) or municipes. See G. W. Bowersock, P. Brown and O. Grabar (eds.), Late
Antiquity: A Guide to the Postclassical World (Cambridge, MA, 1999), s.vv. ‘councils’ and
‘curiales’.
16 Magistratus: throughout the collection; the occasional modifier quinquennalis simply
means a magistrate serving for a 5-year term. Defensor: e.g. Tjäder II 31 (section II,
line 6); see C. T. Lewis and C. Short, A Latin Dictionary (Oxford, 1879), s.v. defensor;
Cancik and Schneider (eds.), Brill’s New Pauly, s.v. defensor.
17 Tjäder I 4–5.
18 Exceptor: e.g. Tjäder I 10–11 (section II, line 8). Notarius: e.g. Tjäder I 10–11 (section II,
line 11). See Bowersock, Brown and Grabar (eds.), Late Antiquity, s.v. notarii. Chartarius:
see, for example, Tjäder I 10–11 (section III, line 6); Niermeyer, Lexicon, s.v. Chartarius.
19 See the will validations embedded in Tjäder I 4–5.
100 Warren C. Brown
in person.20 When the delegation has returned with positive answers, the
document is taken up by the assembly, read out and entered into the
record.
Once all this is done, the assembly heads ask the petitioners what more
they wish. The petitioners ask for an official record of the proceedings
(gesta) ‘as is the custom’ (ex more). The assembly heads order this done.
Though it is not always possible to tell for certain, it appears that in most
if not all cases it is these gesta records given to the petitioners, or later
copies, that we have in the papyri. In several examples, the text then
concludes with validating statements by the members of the assembly
and/or the officials involved.21
This broad template covers various kinds of transactions, and various
kinds of principals, both lay and ecclesiastic. A sub-deacon gives property
to the church at Ravenna.22 A lay couple gives half of their property to the
church at Ravenna.23 A cleric from the Gothic (that is, Arian) church in
Ravenna sells property to a soap-maker from a nearby town;24 a banker
sells an acolyte from Rome some property in Ravenna.25 The ward for
a minor who has inherited property from a man who has died gives a
security for the receipt of the property to the dead man’s widow.26 In
Rieti, a Gothic woman appoints a ward for her two children to handle
the lawsuits in which their dead father was involved.27
The gesta records include the texts of the original transaction docu-
ments they were intended to validate, as well as the texts of any attach-
ments, such as a property list.28 They can also contain the records of
other gesta processes. The most spectacular example, from 489, comes
from Syracuse. This document represents the last stage in a gift of prop-
erty in the territory of Syracuse, from the king Odovacar, who had in 476
seized power in Italy from the western emperor Romulus Augustulus, to
his faithful follower Pierus. Odovacar’s original gift document is embed-
ded within the record of that document’s validation before an assembly
in Ravenna and its entry into the Ravenna gesta. The texts of both of
these documents are embedded in the text written on the actual surviv-
ing papyrus, which is the copy (gesta) issued to Pierus’ representatives of
the record written up by the curia in Syracuse recording the gift’s entry
into the gesta of Syracuse, the formal transfer of the property (traditio) and
the change of the owner of record in the tax rolls (polypthicis publicis).29
This last example makes explicit the fact that the late Roman state
used the gesta municipalia to keep track of people’s tax obligations. The
20 See Tjäder I 14–15. 21 See, for example, Tjäder I 7, 14–15, or 31.

22 Tjäder I 21. 23 Tjäder I 14–15. 24 Tjäder II 33. 25 Tjäder II 29.
26 Tjäder I 8. 27 Tjäder I 7. 28 E.g. Tjäder I 8. 29 Tjäder I 10–11.
papyri accordingly highlight the entry of documents and the record of

the proceedings into a central, official register. But the gesta also served
their petitioners; those who submitted their documents to the gesta had
reasons for doing so that went beyond simply responding to the state’s
imperatives. First, they obtained the legal validation (firmitas) of their
transactions. Second, they had an entry in the city’s official records (acta
or gesta) made that contained the text of their original transaction doc-
uments and stated that the actions necessary to validate them and thus
secure the transaction had taken place. Third, they received an official
copy of this record to take with them.30 They would presumably have
kept this copy with the original transaction document to protect their
rights should their transaction ever be challenged. These ends are cap-
tured by a statement appearing towards the end of several of the papyri,
made by the petitioners to the officiating magistrate(s) at the end of the
process, that runs more or less as in the following example: ‘Since every-
thing that pertains to the validation of these documents of sale and letter
of transfer has been properly carried out in sequence, I therefore ask your
Excellency, highest magistrate, that you order the gesta to be issued to
me, for my defence, by the responsible official according to custom.’31
The gesta records demonstrably did what they were supposed to do.
Sometime between 552 and 575, some notaries from the church of
Ravenna wanted for some unknown reason to re-establish their church’s
claims to properties that had been willed to it.32 So they came before
an assembly in Ravenna headed by the praetorian prefect of Italy. There
the gesta documents recording the relevant wills’ validation were read out
and entered anew into the city’s records.33 The result is another set of
nested documents; the original gesta records containing the texts of the
original wills (albeit truncated), and recording their opening, verifica-
tion, validation and entry into the city records, were copied into a new
gesta document that in turn recorded their revalidation and entry into the
record.
30 That this latter copy had an official character emerges from a comment at the end of one
of the papyri, from 557, which notes that it was produced from an ‘authentic’ original
(‘ex authentico editum’): Tjäder I 7.
31 ‘Quoniam omnia ordine suo, que ad firmitatem ipsorum instrumentorum venditionum
vel epistule traditionis rubor pertinebat, rite ademplita sunt . . . ut gesta mihi propter
monimen meum a conpetenti officio edi iubeatis ex more’: Tjäder II 31. Cf. Everett,
‘Scribes and Charters’, 43–4, and above, 80 (on monimen).
32 Tjäder I 4–5. Cf. above, 76–7.
33 Since the beginning of this text is missing, it is impossible to say for sure whether these
gesta records were brought from the church, that is, that they were the church’s copies, or
were pulled from the city’s archives. Tjäder (1:200) assumes the latter, though without
giving a reason.
102 Warren C. Brown
The papyri place officially sanctioned written records at centre stage

in the making and securing of legal transactions. They also, however,
tell us that much more was involved. The transactions described in the
papyri were first set down in writing, and then they were witnessed and
signed (and in some cases sealed). The documents were then taken to
a public, authority-bearing forum, openly displayed and read aloud by
an official attached to that forum. The creator of the document orally
verified that its contents matched his or her wishes and that the witnesses
whose marks were on it were those that he or she had gathered. In
some cases the witnesses themselves also orally verified the contents of
the document and their participation in its creation. The document was
then copied into an official record along with a written description of
the ritual acts surrounding its verification. The final act in the process
had the petitioners, in the last of a series of ritual actions, asking for and
receiving an official written description of all of the steps, written, oral
and ritual, that had been taken to validate their transaction. In short,
to submit documents to the gesta municipalia meant to go through this
entire process, in which written records, speech, memory and public
ceremony were combined and interdependent.34 The process produced
two records, one in the city’s official register, and one that went home
with the petitioner(s), presumably to be kept with the original transaction
documents.
The Frankish formulas describing the submission of

documents to gesta municipalia
Eight surviving manuscripts from Frankish Europe contain formulas
describing the submission of documents to gesta municipalia.35 As noted
above, these manuscripts stem from the Carolingian period; that is, the
eighth (mostly the late eighth) and ninth centuries. Their geography is
a bit harder to pin down because not all of them have been localized
precisely. They all, however, come from the regions west of the Rhine.
Those whose place of origin has been determined more precisely stem
from areas on the Loire.36 The formula texts themselves appear to cover
34 On the tradition of publicly performing transactions, visible in late Roman legislation

from Constantine on, see Everett, ‘Scribes and Charters’, 52 and 60.
35 Fulda (Hessische Landesbibliothek) D1; Leiden (Universiteitsbibliothek) BPL 114;
BnF lat. 2123, 4409, 4627, 4697, 10756; Warsaw (Biblioteka Uniwersytecka) 1.
36 Fulda D1: s. VIIIex , Angers or Tours; Leiden BPL 114: Bourges?, s. VIIIex or s. IXin ;
BnF lat. 2123: W. Francia, possibly Burgundy, s. VIIIex or s. IXin ; BnF lat. 4409:
N. Francia, s. IXex ; BnF lat. 4627: Tours?, s. IX (possibly soon after 818); BnF lat.
4697: Francia, s. IX; BnF lat. 10756: the relevant part is Burgundian, s. VIII (727?);
a much broader span of time.37 They have been dated from the late sixth
century to the early ninth century. They too are West Frankish. Most
stem from cities located on the Loire and southwards, except for one
from eastern Neustria and two from Sens.38
The gesta formulas all describe more or less the same sequence of
events. First, someone carries out a transaction of some kind (such as
a property gift, a dowry grant, an inheritance arrangement, etc.) and
has a charter written that describes it; this charter is witnessed. Then,
the person responsible for the transaction gives another person a writ-
ten mandate to take the charter to a nearby city and submit it to the
gesta municipalia (gestis municipalibus allegare or prosequere). The mandate
bearer (from this point on often called the prosecutor) duly appears before
the curia of the city, which is headed by a defensor. He asks the defensor
and curia to have the public books (codices publici) opened to him because
he has something that ought to be submitted to the gesta municipalia. The
defensor replies that the public books lie open to him and that he should
do what he wishes. In response, the prosecutor offers first his written man-
date from the original donor, then the original charter to the assembly,
and asks that they be read out loud. The documents are accepted by
the assembly and read out, most often by a notary (amanuensis).39 The
documents having been read, the curia recognizes them as valid. At the
prosecutor’s request, the curia then issues a written document (that is, this
formula) to the prosecutor saying that all this has been done. Finally, in
some cases we get a note from the prosecutor back to the original donor
saying that he had done what he was supposed to do.
Although the gesta formulas all outline the same procedure, and in
some places use virtually identical language, they are not all the same.
They are not derived from a single source or even much related to each
other beyond their essential content. For one thing, each deals with a
different kind of transaction.40 Most of the transactions involve only
Warsaw 1: Tours, s. IX1/2 . See Rio, Legal Practice, 241–71 (‘Appendix: A Handlist of
Manuscripts’), and the literature she cites in the entry on each manuscript.
37 Using Zeumer’s names and numbering, the formulas are as follows: Form. Andec. 1abc;
Form. Arvern. 2ab; Form. Bitur. 6 and 15abcd; Marculf 2.37–8; Cart. Sen. 39–40; Cart.
Sen. App. 1abcd; Form. Tur. 2–3; Form. Tur. Add. 4–5 = Coll. Flav. 9–10.
38 Form. Andec. (Angers, s. VIex ); Form. Arvern. (Clermont, s. VIIImed ); Form. Bitur.
(Bourges, c. 800); Marculf (St-Denis or Meaux?, s. VII2/2 ); Cart. Sen. (Sens, 768×74);
Cart. Sen. App. (Sens, s. VIII?); Form. Tur. (Tours, s. VIIImed ); Form. Tur. Add. (Tours
or Flavigny, s. VIIIex or s. IXin ?); Coll. Flav. (Flavigny, s. VIIIex or s. IXin ). See Rio,
Legal Practice, 67–101, 111–26.
39 Bowersock, Brown and Grabar (eds.), Late Antiquity, s.v. notarii; Niermeyer, Lexicon,
s.v. amanuensis.
40 Very little of the formula evidence for the gesta, Merovingian or Carolingian, deals with
the registration of wills. See also J. Barbier, ‘Testaments et pratique testamentaire dans
104 Warren C. Brown
laypeople: husbands make dotal gifts to their brides,41 a mother gives

power of attorney to her son,42 a man donates property.43 One case
deals with a husband who never quite got around to giving his wife
her morning gift or dos, so that their children were legally illegitimate;
the transaction has him passing property to his children anyway.44 Three
formulas include churches as optional parties: one man makes a donation
to a church, a monastery, or an illustrious man;45 another either gives
property to a church or makes a dotal gift to a woman;46 a third gives
property either to a church or to his grandson.47 The formulas’ structure
also varies. Some present each element of the procedure in distinct parts,
in the following order: the original donation charter, the mandate, the
procedure before the curia, and the mandate bearer’s notification to the
original donor. Others present the pieces in the order in which they would
have taken place before the curia. That is, we start with the prosecutor
appearing before the curia; then, as the mandate and original transaction
charter are read out, their texts are given; then we close with the ending
of the gesta procedure. Still others compress some or all of these parts
together, or stop halfway through with the procedure incomplete. Finally,
the length and detail vary from formula to formula; some present all
of the elements I have described; others are truncated to a greater or
lesser degree. In other words, these formulas appear to reflect a set of
assumptions about the gesta municipalia procedure that was common to
several areas in the central regions of Gaul. Nevertheless, they represent
independent formula traditions; they were not all derived from a single
source.
Many of these formulas do share one other feature in common, how-
ever: it is very hard to understand how a central document register fits
into the process they describe. As noted above, all of the formulas start
out by referring to a set of public books (codices publici). The prosecutor
asks the assembly and its heads to have the public books opened because
he has something that he wishes to submit to them;48 the defensor then
le royaume franc (VIe –VIIIe siècle)’, in F. Bougard, C. La Rocca and R. Le Jan (eds.),
Sauver son âme et se perpétuer: Transmission du patrimoine et mémoire au haut moyen âge
(Rome, 2005), 7–79, at 31n129. Many of the references to the gesta outside the formulas,
however, do come in the context of testaments; see Barbier, ‘Pouvoirs et élites’, Part I,
Chapter 3, passim; compare above, 77 and n.35.
41 Form. And. 1abc; Form. Bitur. 15abcd. 42 Form. Arvern. 2ab.
43 Form. Tur. 2–3; Form. Tur. Add. 4–5 = Coll. Flav. 9–10. 44 Cart. Sen. App. 1abcd.
45 Marculf 2.37–8. 46 Cart. Sen. 39–40. 47 Form. Bitur. 6.
48 E.g. Form. Andec. 1a: ‘utique coticis puplici patere iobeatis, qua habeo, quid apud acta
prosequere debiam’; Marculf 2.37: ‘ut mihi codices publicus patere iubeatis, quia habeo
aliquid, que gestis prosequere debeam’; Form. Tur. 3: ‘ut mihi codices publicos patere
iubeatis, quia inluster vir ille per hunc mandatum ad me speravit, ut donationem illam,
replies that they are open. But in most of the formulas, these public books
then drop out of sight; they take no further part in the proceedings and
leave behind no hint of how they were used. Only at the end of one exam-
ple from Tours do they make another appearance; the prosecutor, after his
documents have been read out before the assembly, asks that the public
records (publica momenta = monumenta) be taken up, and that once the
books have been opened (patefactis codicibus) he be given a copy of the
gesta.49 This vague request implies that the copy was to be produced from
the public books.
In only two formulas does a central, written register maintained by
the curia play a clear and coherent part in the process. Both derive from
the same source, namely the formula collection of the monk Marculf,
compiled in eastern Neustria sometime in the second half of the sev-
enth century. One appears in Marculf’s collection itself; it is titled ‘Gesta
according to the custom of the Romans, how donations or testaments
should be read’. The other was drafted either in Tours or in Flavigny,
but it lifts the relevant language directly from Marculf.50 In both, we
start as usual with the opening dialogue about the public books. As the
process unfolds, the defensor orders that the original donation charter
be read out and then inserted into the public gesta (gestis publicis). At
the end, he directs that a written copy of the proceedings (gesta) be
given to the prosecutor and that the gesta be preserved in the public
archives to be remembered (ut in arcipibus publicis memoranda servetur). It
appears that Marculf drew on a documentary tradition, which he quite
consciously thought of as Roman, in which central records and archives
were important. The Tours or Flavigny copyist picked this tradition up
from Marculf.
These two formulas are, however, exceptions. In the other gesta formu-
las, it is very difficult to tell what part centrally maintained written records
quem de rebus suis proprietariis de loca nuncupantia illa, sitas in pago illo, partibus
illius per sua legittima strumenta confirmavit, gestis municipalibus cum curia publica et
defensore prosequere et alligare deberem’; Form. Bitur. 15c: ‘uti mihi codicis publicae
paterae praecipiatis, que abeo, gestorum alegatione cupio roborarae’.
49 Form. Tur. 3: ‘rogo, ut publica momenta suscipiat, et, patefactis codicibus, gesta, cum a
vobis fuerit subscripta, mihi nobilitas vestra, ut mos est, tradi precipiat’.
50 Marculf 2.37–8; Form. Tur. Add. 4–5. These last two formulas appear only in the two
surviving manuscripts of a formula collection from Flavigny, called by Zeumer the
Collectio Flaviniacensis; they are Coll. Flav. 9 and 10. Because they blend material from
Marculf with formulas that clearly come from a Tours formula collection, and because
in one of the manuscripts they were incorporated en bloc with the rest of the Tours
collection into the Flavigny collection, Zeumer apparently concluded that they had
come from Tours and published them as Form. Tur. Add. 4 and 5. However, there is no
compelling reason to conclude that they come from Tours; they may just as easily have
been created and copied in Flavigny. Cf. Rio, Legal Practice, 117–21.
106 Warren C. Brown
are playing.51 The opening references to ‘public books’ suggest that cen-
tral records did once form an important element of the gesta municipalia
tradition in the areas where these formulas were produced. However, as
the formulas present them, the references appear to be vestigial. The
public books are disconnected from events; they remain peripheral to the
action and their purpose is unclear.
One standard for judging whether or not a piece of formulaic language
in a formula text had any real connection to the world of the scribe who
copied it might be whether it makes any sense within the formula’s inter-
nal logic. In the gesta formulas that stand outside the Marculf tradition,
the references to a central written record do not make this kind of sense;
they have no essential connection, or at best a vague one, to the process
being described.
Yet the gesta formulas do describe a process with a recognizable logic
whose purpose we are able to grasp. In this process, a petitioner secured
a transaction by having the original transaction document ratified by a
public assembly that had the authority to ratify it. The document was
ratified by having it read out loud before the assembly and by having
the assembly publicly acknowledge that it was valid. The assembly then
produced for the petitioner a document saying that all this had been
done. This comprises only part of what we saw in the Italian papyri.
There the gesta process had two elements: on the one hand, the central
written register, and, on the other, the public display of documents,
oral performance, human memory and final issuing of a document that
supported the central register and validated the petitioners’ transactions.
In the Frankish formulas that fall outside the Marculf tradition, the part
of the process having to do with the central written register has become
incoherent. The other part, however, remains in focus.
Within the picture created by these formulas, then, it appears that to
submit a document to the gesta municipalia essentially meant to secure
a transaction (gesta) by having the original charter with which it was
carried out (gesta) read out in a public, authority-bearing forum (gesta
51 In the oldest of the gesta formulas, from sixth-century Angers, the prosecutor, at the end
of the process, says, ‘Gracias agere magnitudine vestrae quod dotem sua scripta quem
prosequio gestis municipalibus ut abuit karitas vestra alegassetis.’ It is very hard to tell
what exactly this means; that is, whether the prosecutor is thanking the curia for having
taken up the dotal gift that he (i.e. the prosecutor) had submitted to the gesta, for having
taken up the dotal gift before the gesta (i.e. the assembly) which they had held, or for
having inserted it into the gesta as a central written register, i.e. the coticis puplici (sic)
with which the formula sequence opens. Rio, Formularies, 49–50, indeed translates the
passage in the latter sense. The point to take away here, however, is that it is not at all
clear.
municipalia) and recognized as valid; the security thus given the transac-
tion by this procedure was embodied in the document (gesta) produced
at the forum saying that this had been done. In other words, what the
original donor cared about – what he wanted to have done and have a
record saying had been done – was the public reading and ratification of
his charter in an authority-bearing forum. The most important thing for
his representative the mandate-bearer/prosecutor to do was not to insert
a document into a gesta municipalia as a central, written record, but rather
to take away a document from a gesta municipalia as an authority-bearing
public assembly describing what that assembly had done.
Supporting this argument is some language that appears in examples
from Bourges and from Sens. Here the original donor writes to the
mandate-bearer: ‘do not delay to write me back about the proceeding that
was celebrated’ (de caelebrata prosecutione mihi rescriberae non tardaris).52
In other words, for the donor it was the public ceremonies, the public
declaration and confirmation of the transaction, that lay at the heart of
the matter.
The manuscripts
There is no way to tell by looking at these formulas by themselves that
they were not, for all of their individual variation, just late antique fossils
carried forward in time by conservative and tradition-bound (or uncom-
prehending) compilers and copyists. To say something firm about how
these formulas might have connected to some early medieval reality, we
have to look at the manuscripts.
Some manuscripts reveal little or nothing about the history of the for-
mula collection they contain. For example, the Angers formulas, which
have been dated to the late sixth century, survive only in one late eighth-
century manuscript from the Angers/Tours region.53 There is no way to
tell what processes of choice or selection (or lack thereof) might have led
this collection to include a gesta municipalia sequence,54 or when these
processes might have taken place. It is possible that the eighth-century
copyist simply transcribed a group of late antique formulas whole,
52 Form. Bitur. 15b (see also Form. Bitur. 15d; Cart. Sen. App. 1b and d). Note the earlier
use of celebrari in Tjäder I 7.
53 Fulda D1; Rio, Legal Practice, 67–80, and Rio, Formularies, 37–101, 248–54;
Bergmann, ‘Formulae Andecavenses’; Bergmann, ‘Verlorene Urkunden’; B. Bischoff,
Die südostdeutschen Schreibschulen und Bibliotheken in der Karolingerzeit. I. Die bayrischen
Diözesen, 2nd edn (Wiesbaden, 1960), 258.
54 Form. Andec. 1abc.
108 Warren C. Brown
without concerning himself too much about what was in it or about

anachronistic language.
Other formula manuscripts, however, show clearly that the formula
collections they contain had histories and that formula collections could
evolve over time. The people who compiled and copied out these
manuscripts constructed their formula collections by deliberately select-
ing, adding or dropping individual formulas or groups of formulas from
older sources, sources which themselves sometimes reflect similar pro-
cesses. The most direct example that applies to the gesta municipalia is
a codex, now kept in Paris, which consists of pieces from several other
codices sewn together.55 The interesting piece for our purposes is a frag-
ment of an eighth-century manuscript containing some formulas from
Bourges. The formula collection represented by this fragment must have
originally contained at least fifteen formulas, but the first ten formulas
were somehow lost; the quire begins with a formula numbered 11 and
continues through number 15. These formulas formed part of a unified
collection; they were written out in one go, with the same page layout, in
a single eighth-century hand.56 They stop at the top of the recto side of
a folium. Then comes a short hymn text in Tironian notes, in a different
hand and in a different ink. The remaining half of this recto side orig-
inally contained some sort of drawing involving circles that is now too
faint to make out. Written over the drawings, squeezed tightly into the
available half-page, and in yet a different (but still eighth-century) hand,
is a gesta municipalia formula from Bourges (see Figure 5.1).57 Specifi-
cally, it is the text of a gesta procedure whereby a man, acting on behalf
of (and bearing the mandate of) his brother, comes before a civic curia
and its defensor to submit a property gift made by his brother to the gesta
municipalia.58 On the top of the following page is something completely
different: a calendar. In short, this Bourges gesta formula was not part of
the original formula collection of which this quire formed a part. It was
later added deliberately by someone who thought it important enough
to squeeze it into the space left at the end of this page and on top of the
original drawing.
The gesta formulas could not only be selected; they could also be modi-
fied and adapted. The best evidence for this comes from two manuscripts
containing a formula collection from the abbey of St Praeiectus in
55 BnF lat. 10756; Rio, Legal Practice, 259–60.

56 BnF lat. 10756 fols. 62–4; Rio, Legal Practice, 111; MGH Formulae 166 (notes to Form.
Bitur.); H. Mordek, Kirchenrecht und Reform im Frankenreich: Die Collectio Vetus Gallica:
Die älteste systematische Kanonessammlung des Fränkischen Gallien: Studien und Edition
(Berlin, 1975), 107n43.
57 BnF lat. 10756 fol. 64r. 58 Form. Bitur. 6.
Flavigny.59 Sometime in the late eighth century or early ninth, monks

at the abbey mixed formulas from Marculf together with a formula col-
lection from Tours and with formulas derived from documents and texts
in their own archive and library to create a new formula collection. The
original manuscript of this collection has been lost, but it must once
have existed, for its two surviving manuscripts derive from a common
exemplar.60
Early in the Flavigny collection comes a large group of formulas about
property gifts, sales, inheritances, benefices, etc., that reproduces the
Tours formula collection in almost the same order that it appears in other
manuscripts.61 The exceptions come at the beginning. Right after Tours
1 are inserted two formulas from Flavigny.62 Then, after the two Flavi-
gny formulas, come Tours 2 and 3. Except that they are not really Tours
2 and 3. The original Tours 2 and 3 are a gesta municipalia sequence.
Tours 2 represents a mandate given by a property donor to another
man to go to the city, submit the donation to the gesta muncipalia and
return word that he had done so. Tours 3 describes the actual gesta
process, held in the city of Tours, in which the mandate bearer carries
out his mandate before the defensor and curia of the city. In the Flavi-
gny collection, what we find in place of Tours 2 and 3 are two similar
formulas, – likewise, a mandate formula and a gesta formula. But they
have been constructed by blending sections lifted from Tours 2 and 3
with pieces of the Marculf gesta formulas; I mentioned them above when
noting the language about central document registers and archives that
they borrowed from Marculf.63 The blending has a substantive impact.
For example, the revised mandate leaves out a line from the Tours origi-
nal in which the property donor asks the mandate bearer to report back
about what he has done. The revised gesta formula leaves out the orig-
inal’s specific reference to the city of Tours. The revised gesta is also
more detailed than its Tours counterpart; the extra detail comes from
Marculf.
In short, the compiler of the Flavigny formula collection wanted to
have a gesta municipalia sequence in the spot where this group of Tours
formulas would have had one. However, he did not like what the Tours
59 Coll. Flav.: Copenhagen, Kongelige Bibliotek, Coll. Fabr. 84; BnF lat. 2123.
60 Rio, Legal Practice, 96–9; P. Depreux, ‘La tradition manuscrite des “Formules de Tours”
et la diffusion des modèles d’actes aux VIIIe et IXe siècles’, Annales de Bretagne et des
pays de l’Ouest, 111 (2004), 55–71, at 61–3. See the more extensive discussion of this
collection in the next chapter.
61 I.e. Zeumer’s Form. Tur. 1–32. 62 Zeumer’s Coll. Flav. 7 and 8.
63 Coll. Flav. 9–10. Cf. Rio, Legal Practice, 179–80.
110 Warren C. Brown
Figure 5.1 BnF lat. 10756, fol. 63v: fragmentary formula collection
from Bourges, no. 15.
Figure 5.1 (cont.) BnF lat. 10756, fol. 64r: the Bourges gesta sequence
appears on the lower half of the page.
112 Warren C. Brown
collection offered, so he modified the Tours formulas by blending them

with the corresponding Marculf formulas. Later on, the Flavigny col-
lection reproduces a great deal of Marculf Book II, albeit in a differ-
ent order from the original and with several formulas left out. Among
the formulas left out, of course, are the Marculf gesta formulas, from
which the compiler had borrowed to make his replacements for Tours 2
and 3.
Finally, the text of at least one of the surviving gesta formulas indicates
that it was actually used in a document at about the time the manuscript
containing it was put together. A ninth-century manuscript, which was
most likely put together in Bourges, contains in its last quire one of the
most detailed and complete of all the gesta sequences in the formulas.64 It
starts with a very long and ornate charter by which a man gives property
in the pagus of Bourges to his betrothed. Then comes the mandate, in
which the donor asks a friend to take the charter to Bourges and have
it submitted and affirmed. Then comes the gesta, in which the friend
submits the charter to the city’s defensor and to the public curia. The gesta
starts with the following dating clause: ‘in such-and-such a month, on
such-and-such a day, in the thirty-fourth year of the reign of our lord
Charles and five years from when, with the aid of Christ, he took up the
imperial rule’. This dating clause can refer only to Charlemagne. It is,
therefore, virtually certain that this formula was derived from an actual
document; the copyist removed the names of the people involved, but
failed for some reason to remove the specific date. So, somewhere in the
area around Bourges, in 805 (that is, five years after Charlemagne was
crowned emperor), someone used this series of gesta formulas to record
an actual property transaction. The record was then used as the basis for
the Bourges gesta formulas.
Evidence for the gesta outside the formulas

This last example connects immediately to some corresponding evidence
from outside the formula collections, namely a sequence of charters pre-
served in the Liber Aureus of the monastery at Prüm. This codex was put
together starting in the last decade of the ninth century to record the
monastery’s property holdings in the wake of Viking attacks in the 880s.
It contains a property catalogue (Urbar) and a series of charter transcrip-
tions. A first group of transcriptions was complete by 919. The codex
64 Leiden BPL 114; Form. Bitur. 15abcd; Rio, Legal Practice, 245–6.
remained in this form until the twelfth century, when more transcriptions
were added.65
Transmitted in this twelfth-century section of the Liber Aureus is a
fully articulated gesta municipalia sequence.66 First comes a property gift,
which according to the text was carried out in the city of Angers in
April of the fourth year of Charlemagne’s imperial rule: a man named
Harwic gave property in the territory (pagus) of Angers to the monastery
of the Holy Saviour at Prüm. After the donation, comes a traditio; that
is, the formal legal handover of the property, carried out according to
the dictates of Roman law, which repeats the terms of the donation. The
traditio was likewise carried out in Angers and bears the same date as the
donation. Then comes the mandate, dated to the same place and day:
the donor gives one Aganbert a mandate to go to Angers and submit the
gift charter to the gesta municipalia. Then comes the gesta itself: the by
now familiar blow-by-blow account of the procedure by which Aganbert
carried out his mandate before the defensor and curia of Angers. Here, as in
the Marculf formulas, a civic document register plays a comprehensible
role in the proceedings. The amanuensis reads out the mandate, and
then he takes up the donation that is to be inserted into the gesta (gesta
inserendi) and reads it out. Once this is done, the defensor orders that the
donation be inserted into the gesta (gestum [sic] inserere iubeatur). Finally,
Aganbert asks that once the proceedings (gesta) have been written up and
subscribed by the curiales, he be given a copy.
It seems safe to assume that the Prüm copyists of the twelfth century
copied this text into the Liber Aureus because they wanted to preserve the
record of this property gift. Why in this case they included the entire gesta
procedure rather than just the original donation charter remains unclear.
The scribe copied the sequence as a whole rather than in distinct parts; it
is possible, therefore, that he thought it was all the donation charter (the
complete text of the entire sequence in the Liber Aureus ends with the
comment in large letters: Carta quam Hariuuicus Fecit). It is also possible
that he thought all of the pieces were necessary for this particular charter
to be valid, or that they would somehow strengthen it.
I have no reason to think that the text is not a verbatim transcript of the
original or that it is a forgery, so I will take it at face value. It tells us that,
in 804, someone in Angers used formulas for the gesta sequence to record
a gift of property to Prüm. The place where the gift was carried out is
65 Das ‘Goldene Buch’ von Prüm (Liber aureus Prumiensis): Faksimile, Übersetzung der Urkun-
den, Einband, ed. R. Nolden (Prüm, 1997), 8–9.
66 Das ‘Goldene Buch’ von Prüm, fols. 97b–99b, trans. pp. 276–80. See also Urkundenbuch
zur Geschichte der, jetzt die Preussischen Regierungsbezirke Coblenz und Trier bildenden
mittelrheinischen Territorien, vol. I, ed. H. Beyer (Coblenz, 1860), nos. 41–2.
114 Warren C. Brown
important; while this is the only gesta sequence preserved in the Prüm
Liber Aureus, it is also the only donation recorded in the Liber that was
carried out in Angers. So perhaps the tradition of using these document
forms had survived in Angers where it had not in the other areas from
which Prüm received property. The sequence does not, however, match
the gesta sequence that survives in the Angers formula collection, which
is much more vague and deals with a dotal gift. Nor does it follow any of
the other extant gesta formulas; while it shares the basic elements of the
gesta sequence and procedure with them, in its details it is unique (most
noticeably in its inclusion of a separate traditio, which does not appear in
any other gesta sequences).67 So, evidently, a tradition for writing these
documents survived in Angers in 804 that was independent both of the
Angers formula collection and of all the other formula traditions that we
know about.
Earlier writers on the gesta municipalia have already noted this sequence
in the Prüm Liber Aureus. These scholars have disagreed with each other,
sometimes sharply, about its implications. Some have taken an optimistic
view, to the point of arguing that, in Angers in 804, a gesta municipalia on
late antique lines was still functioning.68 Others argue that it represents
an archaic charter tradition and that it says nothing about what really
took place.69
In my opinion, the Prüm gesta sequence does connect in an important
way to the reality of early ninth-century Angers, though not in the way
that the argument for absolute continuity would have it. My reason for
thinking so has to do with the fact that there were people involved in
the transaction, and that these people had names. The witnesses to the
donation, the traditio and the mandate all overlap (more or less – one
or two drop out here and there); it was essentially the same group of
people who witnessed all three.70 But the people involved in the gesta
are entirely different, indicating that it was a distinct assembly.71 The
67 See Classen, ‘Fortleben’, 44–5, for a discussion of the formulas, or pieces of formulas,
that this text echoes.
68 Staab, Untersuchungen, 137–8.
69 Hirschfeld, Die Gesta municipalia, 86–7; H. Aubin, ‘Vom Absterben antiken Lebens im
Frühmittelalter’, Antike und Abendland, 3 (1948), 88–119, at 94; Classen, ‘Fortleben’,
44–5.
70 Donation: vicarius Godald, Dilidran, Gedulf, Agisbert, Ramirtran, Gaginhard, Adial-
hard. Traditio: Godald, Dilidranus, Tedulf, Agisbert, Ramirtan, Cacinhard, Adalhard.
Mandate: Godald, Dilidran, Thedulf, Agisbert, Ramirtan, Cacinhard.
71 Count Nono, curator Risclen, vicedomus Uuigfred, Hermedran, Gendrad, centenar-
ius Letbaud, Gerald, Saidris, Srodald, David, Lethard, centenarius Stabul. Classen,
‘Fortleben’, 45, argues that Protokollierung before the gesta has become part of the gift
gesta itself presents us with a familiar cast of characters, namely the curia
of the city of Angers, the defensor, the court notary (amanuensis) and
the mandate bearer. What is striking is that the last three are named:
the defensor, Uulfred; the notary, Leodegar; and the mandate bearer,
Aganbert, as noted above.72 At the end we get the list of witnesses:
Count Nononus; a municipal official of some sort (curator)73 named
Riscleno; an administrative assistant, possibly to the count (vicedomus),
named Uuigfred;74 then Hermedranus and Gendradus; then a regional
officer under the count’s authority (centenarius)75 named Letbaud; then
a Gerald, a Saidris, a Srodaldus, a David and a Lethard; and, finally,
another centenarius named Stabulus. This list starts off with a count; the
titles for the other office-holders it names are unequivocally Carolingian.
One might be tempted to say, therefore, that this was not in fact a ‘real’
gesta municipalia, but rather a typical Carolingian comital court of the
early ninth century wrapped up in the archaic language of the gesta.
However, the defensor Uulfred, who heads the assembly, and the notary
Leodegar are not on the witness list; they occupy their own separate
position in the proceedings.76 Count Nononus appears only in the witness
list. If it had been a comital court that a tradition-bound scribe had tried
to capture with an archaic formula, I would have expected him to name
the count, rather than Uulfred, as the person heading the assembly.
The point is that the scribe who drafted this text was able, apparently
without any strain, to fit named people into each of the slots required
by the gesta formula. This does not mean that a gesta process took place
exactly as described by the text.77 We do not need to argue that the men
charter itself, that there was no longer a curia in Angers, and that we are therefore faced
here with an attempt to convey some sort of judicial legitimacy by couching the trans-
action in old forms. While I agree with the third statement, for reasons that will become
clear below, the difference in the witnesses indicates that the parts of the transaction
were distinct and that some sort of separate legitimating assembly had in fact gathered
in Angers.
72 Staab, Untersuchungen, 137n569 assumes from the address in the mandate (‘domno
in christo fratri Aganberto’) that the mandate bearer Aganbert was a monk of Prüm.
However, frater appears frequently throughout the formulas in contexts indicating that
it can also serve as a general form of address.
73 Niermeyer, Lexicon, s.v. curator.
74 F. L. Ganshof, Frankish Institutions under Charlemagne, trans. B. and M. Lyon (Provi-
dence, RI, 1968), 33.
75 Ganshof, Frankish Institutions, 32–3.
76 Here I disagree sharply with Hirschfeld, Die Gesta municipalia, 87, who, in order to see
the assembly as a Carolingian court, arbitrarily decided that the defensor Uulfred and
the vicedomus Uuigfred were the same person.
77 Here I differ with Staab, Untersuchungen, 138, who apparently thought that it meant just
this.
116 Warren C. Brown
listed at the end of the text thought of themselves as curiales, although the
text refers to them as such, to think that they most likely represent the
members of the assembly; we do not need to think that Uulfred saw him-
self acting as a defensor in the late antique mould as he headed the assem-
bly, if he were even aware that this is what the text called him. Nor can
we say whether the documents involved were actually entered into a civic
document register. But it does seem that something took place that was
more or less equivalent to what the text says took place. I would suggest
the following: the original donor Harwic had the donation, the traditio
and the mandate drawn up (or drew them up himself? In the witness lists
to these three, we read Haruuicus hanc donatione/traditionem/mandatum a
me facta) and witnessed. He then gave them to his mandate-bearer Agan-
bert. Aganbert took them to an assembly that had the authority to ratify
them, that was composed of the signatories listed at the end, and that
was presided over by Uulfred. The notary for the assembly, Leodegar,
used the gesta formula to capture the ratification process, and copied the
other associated documents along with it. According to our text, a signed
copy of the gesta written by Leodegar was given to the mandate bearer
Aganbert. Count Nononus, as the city’s leading secular authority, would
naturally have been present at such an assembly, but it does not appear
that he presided over it.
Beyond the gesta municipalia

I would argue that the compilers of our formula manuscripts copied
gesta municipalia formulas into their collections because such formulas
were actually used in the Carolingian period. This is emphatically not to
suggest that real, live gesta municipalia, exactly as the formulas describe
them, were still operating in the same way and in the same form. How-
ever, I do think it is safe to say that there was a perceived need in the
late eighth and ninth centuries, on the part of at least some people, to
secure transaction charters by having them read out at and ratified by
a public assembly that was understood to have some authority to ratify
them, and to have the assembly issue a document saying that this had
been done. This process was sometimes, or even often, carried out by a
surrogate – hence the mandate. That the need for this procedure and the
procedure itself derived from the tradition of the Roman gesta municipalia
is indicated by the fact that the gesta municipalia formulas were used to
record what had happened. The details of what these formulas said, or
the terminology that they used, may not have been important, but it is
evident from the ways that they were selected, copied and used that they
were considered to be the right and proper way to record the ratification
process.78
The belief that documents ought to be ratified in this way evidently
covered transactions among laypeople only, as well as transactions involv-
ing clerics and churches. Hence, these formulas provide evidence from
the eighth and ninth centuries for a public (in the sense covered by the
German öffentlich rather than staatlich) system of validating documents in
which laypeople as well as clerics participated, a system that was, more-
over, not anchored in churches (as was, say, the practice of validating a
charter by placing it on a church altar). The formulas reflect, therefore, a
general world of document use comprising both lay and clerical members
of the upper crust of Carolingian society (by which I mean those people
propertied enough to be involved in the kind of transactions covered by
the formulas).
The gesta formulas were used to record the public validation of docu-
ments in the western parts of the Frankish world. The practice appears
to have been most at home in cities on the Loire and southwards,79 but
it is also visible elsewhere in West Francia. This is, of course, precisely
where one would expect to find Roman-style documentary practices and
a Roman institutional legacy surviving the longest. No formulas like this
stem from East Francia, despite the fact that important formula collec-
tions have survived from Bavaria and Alemannia.
However, the documentary practices that lay at the heart of the gesta
municipalia formulas are visible over a much wider area.80 I have argued
that from the point of view of those carrying out transactions, what was
most important about the gesta municipalia was public ceremony; people
thought it important to secure their transactions by having their trans-
action documents read out loud before an assembly or forum that had
the authority to ratify them and that could produce another document
saying that it had done so. The two formula sets from the Marculf tra-
dition, as well as the actual charters from Poitiers and Angers/Prüm, do
78 Cf. Rio, Legal Practice, 180–2; Rio, Formularies, 257–8. Cf. Barbier, ‘Pouvoirs et élites’,
Part I, ch. 3, esp. pp. 106–8. Barbier infers from references to the gesta in other kinds
of sources (i.e. diplomas, testaments and charters, but also Marculf 2.3) that real gesta
municipalia, with central document registers, persisted in some places up to the begin-
ning of the ninth century (though she does suggest that the gesta evolved in the eighth
century away from central archiving and increasingly stressed public reading and valida-
tion of documents). The surviving gesta, she argues, reflected an ongoing interest on the
part of civic elites in controlling local property arrangements and preventing or resolving
property disputes. Her argument, however – at least as it appears in the chapter that
was available to me – is speculative and depends on meagre and ambiguous evidence
that could be (and has been) interpreted differently.
79 Cf. Wickham, Framing, 110–11.
80 For the framework of what follows, see below, Chapters 8–9.
118 Warren C. Brown
integrate the civic document register intelligibly into the gesta process.
They survive, however, alongside a greater number of formulas in which
references to such registers appear to be vestigial. What remains in focus
in all of the formulas is what consumers of the gesta had been getting out
of the process ever since Late Antiquity: public ceremonial validation of
their transaction documents and a written record of that validation.
To achieve this essential result, one did not have to go to a gesta munic-
ipalia. One simply needed to find an authority-bearing and public venue
where one could present one’s documents for validation and which could
issue a document saying that the necessary public steps had been taken.
One could, for example, go to a king and his court. In a formula from the
so-called Marculf Supplements, a king writes to a count to say that a man
(or the man’s missus) came before him and reported that he had bought
property from another.81 The buyer had apparently brought with him
the original document of sale, for next the seller stepped forward, con-
firmed the document’s contents, and acknowledged that he had received
payment. The king then confirmed the transfer of ownership with this
document (per presentem . . . preceptum). This scenario also appears in
several Merovingian diplomas. In an example from 596, a priest vir inlus-
ter and a nun founded and endowed a chapel dedicated to St Martin in
Le Mans per eorum strumenta. They then went to see King Theuderic
II and asked him to confirm the transaction. After the epistola donationis
had been viewed, the king did so.82
The same end result was likewise achieved in the so-called Schein-
prozesse, or ‘staged disputes’ from the Merovingian period. These were
cases brought up before a royal judicial hearing, or placitum, in which one
party charged another with unjustly holding its property, the defendant
conceded without a defence, and the plaintiff received a document con-
firming his rights.83 At a placitum headed by King Childeric III in 710,
the inluster vir Ragnesind declared that he had purchased property from
one Sicland and his wife Dinan, that they had accepted the purchase
price from him, and that a document of sale had been drawn up. Rag-
nesind held up the charter to be read, and it was read.84 Sicland, who was
present, confirmed Ragnesind’s account. Accordingly, the king together
with his leading men ordered that, the document of sale having been
81 Marculf [Supp.] 2 = Coll. Flav. 105.

82 MGH DD Mer. 1, no. 25; see also nos. 22, 28, 32, 75.
83 See P. Fouracre, ‘Disputes in Later Merovingian Francia’, in Settlement of Disputes, 26
and n. 13.
84 Presumably out loud: ‘et ipsa vindicione in presente ostendedit relegienda relicta [sic =
relecta?] ipsa vindicione’.
examined, Ragnesind be confirmed in his possession of the property; the

order was written up in the form of a royal judgement.85
People similarly resorted to kings to validate their documents in the
Carolingian period. In Pavia in 774, for example, Charlemagne con-
firmed an exchange of property between Bishop Merol of Le Mans
and Abbot Rabegaud of St-Calais. The parties produced the original
exchange document to be read; from the detail given, it appears that it
was substantially copied into the royal diploma confirming the exchange,
of which the king ordered two identical copies made.86
This kind of document validation process also shows up in the East.
A Salzburg manuscript of the late ninth century contains an only super-
ficially altered version of the Marculf Supplement formula discussed
above.87 Other examples appear in the charters of St Gall. According
to one, in Worms in 780, Charlemagne confirmed an agreement made
between Bishop Sidonius of Constance and Abbot John of St Gall about
the dues owed by the monastery to the bishopric.88 The two men had
reached the agreement at some earlier point, and had fixed it in writing in
two identical copies. When they came before the king, they displayed the
documents for reading.89 Thereupon Charlemagne ordered this charter
produced, confirming their contents. In Frankfurt in April 875, Charle-
magne’s grandson Louis the German similarly issued a diploma con-
firming an exchange between Abbot Hartmot of St Gall and the royal
vassal Widpert. The exchange had originally been carried out by means
of documents affirmed ‘by the hands of good men’; the pair displayed
these documents before the king to be read out.90
Kings were not, however, the only authorities to which one could
resort. Some people in Bavaria and Alemannia carried out their transac-
tions before a comital court, or mallus. This achieves the same result as
the gesta procedure, but compresses it. The transaction and the ceremo-
nial ratification before an authority-bearing assembly are all wrapped up
in one public process; at the end, the principal or principals come away
with a document saying that the necessary public steps had been taken.
A formula from the same Salzburg formula manuscript discussed above,
for example, has a man and his wife appear at a public mallus and give
85 MGH DD Mer. 1, no. 158; see also nos. 136, 143, 153, 155, 187.
86 MGH DD Kar. 1, no. 79; see also nos. 130, 136, 161, 206.
87 Form. Marc. Kar. 28 = Munich, Bayerische Staatsbibliothek, Clm 4650 fols. 57v–58v
(no. 51).
88 Wartmann I 92.
89 ‘eorum manus roboratus uno tenore conscriptas nobis ostenderunt relegendas’.
90 Wartmann II 586. For the original exchange charters, see Wartmann II 576; see also
454, 479, 587, 588, 694, 716.
120 Warren C. Brown
property to God and Sts Peter and Rupert (the patrons of the monastery
of St Peter in Salzburg); the gift would take effect after both had died.
The transaction is witnessed and confirmed by people who have been
given the power to witness by the typical Bavarian practice of ear-pulling
(testes per aurem tracti).91 Actual charters that describe more or less the
same thing survive in the St Gall archive. Among them are three record-
ing transactions carried out on the same day before a mallus held in Egg,
just south-east of Zurich, sometime between 841 and 872.92
Examples such as these reveal a continuum of procedures for vali-
dating and securing transactions. All of these procedures involve public
enactment;93 in all of them ritual, oral and written forms of communi-
cation and memory mutually reinforce each other, and in all of them
documents were produced at the end saying that the necessary steps had
been carried out. Venues range from a central place on the property that
was being donated or sold (such as a villa or a church) to courts and to
churches (especially the atria of churches) that were not themselves the
subjects of transactions. In these latter cases, it would seem that God
himself and/or the patron saint of the church or monastery was serving
as the necessary authority figure.94 In Bavaria before the Carolingian
conquest, several transactions were carried out before Agilolfing dukes
(Odilo or Tassilo III) and/or in a location called a castrum, implying
that it embodied ducal authority.95 In cases that do not explicitly assign
any special authority to the venue, the witness lists sometimes contain
office-holders, such as counts, tribunes or centenarii, suggesting that the
assembly may have had some authority-bearing character.96
In her well-known 1989 work on the St Gall charters, Rosamond
McKitterick identified a loose hierarchy of such venues. Some transac-
tions recorded in the charters were carried out locally, on the site of
the property concerned, in front of local witnesses; in these cases, noth-
ing suggests that anything more than simple publicity was involved in
carrying out and securing them. However, McKitterick also identified
91 Form. Salz. 4 = Clm 4650 fols. 64v–65v (no. 58).

92 Wartmann II 565–7. See also Wartmann I 11, 297, 321; II 643.
93 On publicity in the sense of the German Öffentlichkeit in this context, see Fichtenau,
Das Urkundenwesen in Österreich vom 8. bis zum frühen 13. Jahrhundert (Vienna, 1971),
80.
94 Cf. Fichtenau, Urkundenwesen, 77–8, and Barbier, ‘Pouvoirs et élites’, Part I, ch. 3,
pp. 94–8.
95 TF 1, 2, 7, 12.
96 But one does need to be careful; sometimes office-holders were present because they
were related to the principals in the transactions. In assemblies explicitly called a mallus,
the witness lists often do not mention any counts; conversely, the presence of a count in
the witness list does not mean that the transaction was carried out at a mallus.
places to which people went regularly and repeatedly to take care of legal
and judicial matters, places that they evidently perceived as having the
necessary authority and offering the necessary public visibility to secure
transactions, and at which documents could be produced saying that the
necessary steps had been taken. The assemblies held at these places are
not always labelled as courts (that is, as a mallus or placitum), but their
locations were independent of the properties concerned in the transac-
tions that came before them. McKitterick has also documented what she
calls, following Rolf Sprandel, ‘witness leaders’. These were landowners
who show up in the charters heading witness lists at several locations;
they appear to have acted as local authority figures who headed assem-
blies before which transactions were carried out. Following McKitterick,
one can imagine that, depending on the wealth and importance of the
parties involved, and the size or symbolic importance of the property
involved or the legal matter at hand, there was a corresponding loose
hierarchy of public venues and authorities to which people went to carry
out their business, and from which they took home a document saying
that they had done so.97
Even at this local level, some Eastern records describe two-stage pro-
cesses that closely resemble those of the gesta municipalia. In the winter
of 804, for example, a man named Starcholf and his son Hitolf came
to Bishop Atto of Freising in Bavaria.98 They asked him for a benefice,
namely a church in Assling, in exchange for other property in Assling
as well as in Steinkirchen, and annual payments in kind and in silver.
The father and son then pledged or enacted somehow their end of the
bargain.99 Next, a Freising archpriest named Ellanod, acting as Bishop
Atto’s missus, travelled together with Starcholf and Hitolf to a public mal-
lus headed by Count Cundhart in Steinkirchen. There Ellanod, coram
omnibus, pledged to give Starcholf and his son the benefice that they
had asked for. Starcholf and Hitolf in turn pledged again to give what
they had promised. At the end, the document describing the process was
subscribed by witnesses. In 884, one Moyses gave property to St Gall.
In the charter recording the gift, just before the witness list, comes the
line: ‘Carried out publicly at this monastery, in the presence of those
whose names are contained here.’ Then we read: ‘Afterwards this char-
ter was lifted up in the public placitum before a multitude of the people
and once more affirmed, with those agreeing whose signatures are noted
97 McKitterick, Carolingians, 98–115. 98 TF 195.

99 ‘pro qua (i.e. the benefice) spoponderunt se daturos usum beneficii. Quod ita et
fecerunt.’
122 Warren C. Brown
below.’100 In other words, the transaction was first carried out by the
two parties involved at St Gall; then, the charter was taken to a judicial
assembly and publicly reaffirmed. Another St Gall charter, recording
an exchange of unfree dependants between St Gall and one Chadaloh,
suggests the same two-step process with the line, ‘Carried out and lifted
up in the pagus of Munteriheshuntere in the villa of Dieterskirch, affirmed
and accomplished in the pagus of Eritgau, in the place called Bussen,
publicly in the atrium of [the church of] St Laudegarius’.101 Still other
St Gall charters include two separate lists of witnesses given as present in
different locations (and sometimes segregated into clerical and lay), or a
statement that a transaction was carried out first before one office-holder
and then another.102
Conclusions
The gesta municipalia did not simply disappear from the Frankish world.
Instead, they were gradually transformed, at different rates and in dif-
ferent ways in different places, into a set of varied but related practices,
as the decline of late Roman civic and fiscal institutions made municipal
document registers largely irrelevant. The gesta papyri from late antique
Italy still emphasize storing copies of records in central registers or record
books because, at the time they were written, a state still existed in Italy
that had an interest in keeping track of its own fiscal rights and its sub-
jects’ tax obligations. Nevertheless, they also record a series of public
ceremonies, including the very formal public ceremony that took place
when a transaction was secured in the gesta. What the gesta papyri pro-
vided for those who received copies of them was a written guarantee
that the public ceremonies necessary to validate and thus secure their
transaction documents had been properly carried out. In Frankish Gaul
from the sixth century on, it appears that many people who bought, sold,
exchanged or gifted property continued to believe that it was essential
to securing transactions to record them in writing, to have the records
publicly ratified by an authority-bearing body and to get another writ-
ten record saying that this had been done. This belief is reflected in the
survival, copying and continued use of the gesta formulas. Where gesta
100 Wartmann II 639: ‘Postea vero in publico placito sub frequentia populi levata atque
iterum firmata est haec eadem carta, astipulantibus his, quorum hic signulacula sub-
notatur.’
101 Wartmann II 684: ‘Acta et levata in pago Munteriheshuntere in villa Diethereskiriha,
firmata et perpetrata in pago Eritgeuve, in loco, qui dicitur Pusso, in atrio sancti
Laudegarii puplice.’
102 Wartmann I 87, 92, 115, 144; II 486.
formulas were copied or used as the basis for real documents, they were
copied or used because the people involved thought that they were the
proper way, or at least one valid way, of recording how documents were
validated and that they thus were, or might be, useful. In other words,
the formulas still had some perceived relationship to reality. But in many
places the language in them that had to do with central civic registers
or archives dissolved, along with the institutions that had inspired that
language, until it became incoherent. Outside Frankish Gaul, in areas
where Roman institutional and municipal traditions were weak or non-
existent, people nevertheless handled and ratified documents in a way
consonant with the gesta tradition. They simply found different venues
and authority figures to fulfil the same purpose.
It has been suggested that, as the eighth century progressed, churches
came to take over the role of the gesta municipalia. The evidence behind
this suggestion lies in explicit references to storing documents in church
archives, such as that in the Flavigny will formula discussed above, as
well as in the undeniable number of transactions that were carried out
at churches and recorded by church scribes.103 To assign to churches
the role formerly played by the gesta, however, is to assume that the
gesta served primarily as archives where both the civic authorities and
those who carried out transactions could store copies of transaction doc-
uments for safe keeping and later reference. Our evidence indicates that
even in Late Antiquity this was only part of the gesta’s purpose, and
that as time passed its importance waned. The gesta municipalia served
to record transactions in a central register for the purposes of the late
Roman state; for their users, they guaranteed that their transactions were
valid through public ceremonies, document readings and the issuance of
an official document. What was important to the users were the official
documents that they took with them, with their lists of authority figures
and witnesses, that said that they had gone through the necessary public
steps to validate their transactions. They kept these records with their
others at home, or sometimes (over time and in some places) in churches
or monasteries because churches and monasteries offered the security of
better buildings and divine protection.104 Churches kept their own docu-
ments for the same reason everyone else did. By the Carolingian period,
no one any longer had any institutional or fiscal interest in preserving
registers of property transactions that recorded everyone’s transactions,
or better, the transactions of those wealthy enough to matter, in the way
103 Rio, Legal Practice, 181; McKitterick, Carolingians, 89; Fichtenau, Urkundenwesen, 70.
Cf. above, 76–7.
104 Brown, ‘Documents’, 351–2; below, 371.
124 Warren C. Brown
that the gesta municipalia in late antique Italy had.105 The official docu-
ment registers compiled and kept as part of the late antique gesta process
served the interests of the late Roman state; they served the interests of
users only in so far as they protected them from getting in trouble for not
registering, and provided a backup source of documentation.
If we choose to see the gesta muncipalia as primarily central docu-
ment registers or archives, we can say that they died as the institutional
infrastructure of the late Roman state died, at different rates in different
regions. But if we focus on the part of the gesta that had most interested
its users, that is, on the public, ceremonial validation of documents, then
they lived on, at least in spirit. The vigour of the gesta tradition is attested
in the gesta formulas that continued to be copied. It is equally strongly
attested in the other venues that people found for accomplishing the
same purposes. Churches did number among those venues; they were,
however, but one such among several possible.
As the early Middle Ages progressed, institutional change or transfor-
mation altered the landscape of available venues to which people could
resort to have their transactions and documents validated. Where and
while the fiscal infrastructure of the Roman state persisted, the gesta
municipalia in something like its old form endured. But when that infras-
tructure withered, or when documentary practices spread to (or were
revived in) areas where they had never or had hardly existed, other venues
were found that combined the publicity, the authority and the ability to
produce written documents that had characterized the gesta.
105 Cf. the suggestion by Barbier (above, n. 78) that local civic elites may have had such
an interest.
6 Laypeople and documents in the Frankish
formula collections
Warren C. Brown
The surviving Frankish collections of legal formulas demonstrate in the round

how documents were used at the time when the manuscripts were written,
at the (usually earlier) time when the identifiable collections were originally
compiled, and at the time when individual formulas were first composed; and
they therefore also show how documentary practices changed over time. The
three collections examined here – that of Marculf, and those from Tours and
Flavigny – were first put together in, respectively, the late seventh, mid eighth,
and late eighth centuries. While the majority of their formulas are for laypeople’s
documents, all three collections were made in ecclesiastical milieux, and some
formulas responded to the distinct needs of the clergy. These collections therefore
suggest that in the Carolingian period the clergy were taking responsibility for
producing documents for both themselves and the laity.
The Frankish formula collections addressed in the previous chapter are

not just a useful source for the history of the gesta municipalia. They are
perhaps the richest source available for studying the ways that laypeo-
ple used documents in Frankish Europe.1 They contain a great many
examples of documents or document types that do not survive in extant
charter or letter collections. Heavily represented among these are doc-
uments in which none of the actors involved, including the issuer and
recipient, are clerics or monks, and in which there is no apparent role
played by ecclesiastical institutions.
These formulas presumably represent document or letter types that
were useful in some way, or at least comprehensible, to the people who
copied them. There is, as we saw before and will see again in the specific
case study below, simply too much evidence that they were deliberately
selected and crafted by their compilers to think otherwise.2 If they were
1 Edition: MGH Formulae. Basic literature: R. Buchner, Deutschlands Geschichtsquellen im

Mittelalter, Beiheft, Die Rechtsquellen (Weimar, 1953), 49–55; Rio, Legal Practice.
2 See also Rio, Legal Practice, 63–6, 238–40; W. Brown, ‘Die Karolingischen Formel-
sammlungen – warum existieren sie?’, in P. Erhart, K. Heidecker and B. Zeller (eds.),
125
126 Warren C. Brown
intended to serve as templates for complete documents, then they rep-

resented documents that could conceivably have been needed and that
therefore captured transactions or behaviours that could conceivably have
happened. If they were intended to be sources of language, then they con-
tained language or phrases that might conceivably have been useful. If
they were intended to be practice models for student scribes, or cases for
them to study, then they represented document types or procedures that
students might need to know, or at least profit from knowing about. The
formulas for lay documents, therefore, can tell us something about the
situations in and purposes for which laypeople might have used docu-
ments, or at least about the range of possibilities for which their compil-
ers felt they needed to be prepared or for some reason wanted to have a
record of.
The formulas present a somewhat fractured or blurred picture of lay
documentary culture, however, because their transmission histories are
complicated. As we saw previously, individual formula collections often
include documents or formulaic passages that are older, even centuries
older, than the manuscripts in which they survive. Sometimes these older
formulas were updated to match contemporary Latin norms or to render
the incomprehensible understandable; sometimes they were not.3 Later
formula collections not infrequently took formulas from older collections,
mixed them together, altered them, and added new texts to produce a
new collection that differed in content and emphasis from its sources.
As a consequence, it can be difficult or impossible to pin down a given
formula in time and space.
Nevertheless, by untangling the transmission history of given formula
sets, particularly those that draw on older formula collections, we can use
the formulas to say some things about the ways that laypeople interacted
with documents at several points along a continuum: at the time the
earliest formulas or language in a given collection must have originated;
at the time(s) earlier formula collections from which that collection drew
were compiled; at the time the collection itself was put together; and at the
time when the extant manuscript was copied out. Moreover, by looking at
Die Privaturkunden der Karolingerzeit (Dietikon, 2009), 95–101. An exception might be

the Angers formula collection, whose texts probably stem from the late sixth century but
which is preserved in the late eighth-century manuscript Fulda, Hessische Landesbib-
liothek, D1; since this is the only copy of this group of very old documents, we can only
guess at possible processes of selection or adaptation involved in its creation. It remains
possible, therefore, especially since it is copied together with a (poor) copy of the LRV,
that it represents an antiquarian compilation. But cf. Rio, Legal Practice, 77–80.
3 See, for example, Munich, Bayerische Staatsbibliothek, Clm 4650, which contains a
version of some of Marculf’s formulas that were updated to reflect Carolingian political
realities (Form. Marc. Kar.); Rio, Legal Practice, 104–5.
Francia: formula collections 127
the formulas in the context of the entire manuscript in which they survive,
we can say some things about the needs and purposes behind their final
copying out, and thus about the cultural and institutional context in
which lay documents were being copied, studied and produced.
To draw one such picture of lay documentary practices in Frankish
Europe, I will discuss one manuscript: BnF lat. 2123. This codex encom-
passes a continuum covering Neustria, the Touraine and Burgundy from
the late sixth through the early ninth centuries. The formula collec-
tion it contains offers rich evidence for a culture of document use in
which laypeople actively participated; in this envelope of time and space,
laypeople, both men and women, both high status and low, regarded
documents as important and used them. They interacted through the
medium of documents not only with churches and monasteries but also
directly with each other. Although clerics and monks were institution-
ally, legally and vocationally distinct enough from laypeople that they
sometimes needed their own versions of common documents, in general
laypeople and clergy used similar kinds of documents in similar ways for
similar things.
The relationship between lay documentary culture and ecclesiasti-
cal institutions along this continuum, however, was not static. Our
manuscript points to a second process of institutional change that fol-
lowed the disappearance of the late Roman civic institutions that had
originally lain behind the gesta municipalia. Over the course of the late
Merovingian period, but especially in the Carolingian period, churches
and monasteries began to fit into the societies around them in new ways
such that they became focal points for the production of documents, to
the point that they compiled collections of formulas allowing them to
study and produce documents for laypeople alongside those they needed
for their own purposes. By the ninth century, at least some ecclesiastical
institutions had developed if not a monopoly on, then a significant and
perhaps dominant role in, the production and use of documents in their
regions.
BnF lat. 2123 was most likely copied out in Burgundy at the end of the
eighth century or the beginning of the ninth.4 It is primarily an ecclesias-
tical reference book; its contents include church council acts, sermons,
an epitome of the Liber Pontificalis, and religious writings and letters by
such figures as Gennadius of Marseilles, Leo the Great and Theodore of
4 Rio, Legal Practice, 252; P. Depreux, ‘La tradition manuscrite des “Formules de Tours” et
la diffusion des modèles d’actes aux VIIIe et IXe siècles’, Annales de Bretagne et des pays de
l’Ouest, 111 (2004), 55–71, at 61; L. Kéry, Canonical Collections of the Early Middle Ages
(ca. 400–1140): A Bibliographical Guide to the Manuscripts and Literature (Washington,
DC, 1999), 54–5.
128 Warren C. Brown
Canterbury. Towards the end it contains an extensive collection of for-

mulas, briefly discussed in the previous chapter, that can be traced, on
internal evidence, to the abbey of St Praiectus at Flavigny. This collection
was edited by Karl Zeumer for the MGH as the Collectio Flaviniacensis.5
The Flavigny formula collection predates the manuscript. It survives in
one other manuscript, now in Copenhagen,6 that differs from our Paris
manuscript in ways indicating that they both stemmed from a common
model.7 This common model can be dated on internal evidence to the
second half of the eighth century.8
The Flavigny formula collection is a hybrid.9 It blends some formulas
from Flavigny itself with formulas from two earlier collections: a mid-
eighth-century collection from Tours, and the collection of Marculf from
the latter part of the seventh century.10 It thus holds embedded within it
layers of time, layers that contain interlocking images of lay documentary
practices. Not only does it offer its own image but also its two major
sources (that is, the Tours and Marculf collections) each provide theirs.
These sources in turn contain fragmentary images inherited from a still
deeper past. I will begin by working through the images of lay docu-
mentary practice offered by the Marculf and Tours collections. Then I
will explore how those images were combined into the image projected
by the Flavigny collection. Finally, I will examine the Paris manuscript
itself. The information offered by the manuscript does not change the
picture of documentary culture that the formulas provide. However, it
does tell us something about how and why the Flavigny collection was
put together in the form in which it survives.
Marculf
Marculf’s formula collection comprises ninety-two formulas,11 divided
into two sections: Book I, containing praeceptiones regales or royal docu-
ments, and Book II, containing cartas pagenses or non-royal documents.
The collection is unusual in that it leads off with a prologue that says
5 Coll. Flav.
6 Copenhagen, Kongelige Bibliothek, Coll. Fabr. 84 (Francia, s. IXmed ); Rio, Legal Prac-
tice, 243–4.
7 See below, n. 87. 8 See below, n. 84. 9 See above, 108–9.
10 Form. Tur.; Marculf. On the former, see Rio, Legal Practice, 112–17; Depreux, ‘La tra-
dition manuscrite’; on the latter, see Rio, Legal Practice, 81–101. A. Jeannin, ‘Formules
et formulaires: Marculf et les praticiens du droit au premier moyen âge (Ve –Xe siècles)’,
unpublished Ph.D. thesis, Université Jean Moulin–Lyon 3 (2007), was not available to
me.
11 Not counting the Supplements.
who its author was and how that author understood its purpose.12 The
monk Marculf, by his own testimony 70 years old or more, states that
he had compiled the collection at the behest of a certain Bishop Lan-
deric as a resource to begin the training of young boys (ad exercenda initia
puerorum). The jury is still out on the precise time and place in which
Marculf put the collection together; the various arguments depend on
different identifications of Bishop Landeric, on different interpretations
of the kinds of documents Marculf chose to include, and on the ways that
his formulaic language compares, or does not, with that of extant docu-
ments. The two chief possibilities are somewhere near Paris around the
650s, or in the area of Meaux closer to 700. None of the arguments are
conclusive; the safest statement is that Marculf compiled the collection
somewhere near Paris or in eastern Neustria, sometime in the second
half of the seventh century, though some significant evidence would tend
to push the date towards the end of the century.13
Even leaving aside the testimony of the preface, there is no question
that Marculf ’s collection once existed; it is not, unlike others of the for-
mula collections published in the MGH, an editorial fiction.14 It survives
in more or less identical form in two manuscripts, and in varying stages
of completeness and alteration in five more.15 The documents that Mar-
culf selected reflect what the old monk thought were important examples
or possibilities. They were, therefore, comprehensible and relevant to his
world and presumably to that of his imagined students as well. Neverthe-
less, his collection captures impressions of Frankish life over a span rather
than at a single point in time. Given his age, Marculf drew on experience
that may have covered up to half a century. Furthermore, some of the
documents he chose contain echoes of language and procedures that can
be traced back to Late Antiquity.
Book I contains the royal documents. Fully half of this book’s forty
formulas do not in any way involve clerics or monks. They therefore
project writing as an instrument of late Merovingian royal power that
embraced and worked through laypeople as well as ecclesiastics.16 We see
kings confirming property grants or property arrangements made by their
predecessors;17 we see kings making property arrangements themselves
(including granting benefices and taking property from rebels and hand-
ing it out to others).18 We see kings helping people to divide property;19
we see them settling disputes and overseeing placita.20 We also see them
12 Marculf, prologue (MGH Formulae 36–7). 13 Rio, Legal Practice, 85–8.

14 Rio, Legal Practice, passim. 15 Rio, Legal Practice, 92–101.
16 Marculf 1.8, 9/10, 13, 17, 18, 20–1, 25, 28–32, 33/34, 37–40.
17 Marculf 1.17, 31. 18 Marculf 1.13, 30, 32.
19 Marculf 1.20. 20 Marculf 1.25, 38.
130 Warren C. Brown
appointing people to office,21 assigning advocates to people,22 issuing

orders and summoning people to show up at court.23 An example of the
latter, specifically titled Indecolum [sic] ad laicum, has a fidelis complain-
ing to a king that someone had attacked him on the road, beaten and
wounded him, and stolen his property. The king in this letter therefore
commands the accused either to make amends, or, if he wished to contest
the charge, appear at court to give a legal response.24 Some royal commu-
nications were, however, evidently too detailed or important to entrust to
writing, for two formulas represent letters between kings authenticating
oral messages carried by their bearers.25
Marculf’s Book I projects an image of a world in which, as far as
the royal use of documents was concerned, clergy and laity hardly dif-
fered from one another; kings used the same sorts of documents in the
same way for the same things regardless of whether they were deal-
ing with clergy (or their institutions), laypeople or both. This impres-
sion is strengthened by the formulas that provide options allowing users
to choose either ecclesiastical or lay actors. A king confirms a prop-
erty arrangement whereby a childless lay couple could let property pass
either to a monastery or to their relatives as they wished.26 A formula
for a royal property gift comes with different prologues that couch the
gift as going either to a layperson, as a reward for loyal service, or to a
monastery.27
These ‘two-way’ formulas, however, also tell us that in Marculf ’s world
there was a difference between clergy and laity; one sometimes needed to
have separate versions of the same document for each. A king’s confirma-
tion of a gift to a bishop’s church by one of his predecessors is followed
by language for the case that the king was confirming a gift made to an
inluster vir.28 Two formulas titled indecolum ad episcopo regarding charges
of unjust property seizure are followed by the indecolum ad laicum that
deals with essentially the same situation.29 In short, a cleric was not a
layperson, and in some situations this distinction was important enough
that Marculf provided for it.
While Book I of Marculf’s formulas is focused on kings, it does occa-
sionally indicate that the use, and storage, of documents extended beyond
the royal courts. A letter to a king from a group of pagenses documents
the fact that one of their lay neighbours had lost all of his documents in
a fire; it is followed by a formula that has the king confirming the layman
21 Marculf 1.8. 22 Marculf 1.21. 23 Marculf 1.28–9, 37, 39–40.

24 Marculf 1.29. 25 Marculf 1.9–10. 26 Marculf 1.12.
27 Marculf 1.14a–c. See also 1.23, a royal precept crafted to fit either a layman or an
ecclesiastic.
28 Marculf 1.16–17. 29 Marculf 1.26–7, 29.
in his property rights.30 Another royal formula states that some people
were reported to have given property to a church or monastery ‘by means
of their charters’ (per eorum instrumenta).31 A group of laypeople demon-
strate their rights before the king by displaying charters.32 A king gives a
man a precept in which he makes the man one of his military retainers
(antrustion) and raises his wergeld accordingly; the recipient would surely
want to keep such a document.33 So, too, would two disputing parties
whom a king orders be given copies of a precept regulating how their
dispute was to be handled.34
However, by far the best evidence in Marculf ’s collection for lay doc-
ument use outside the arena of the royal courts comes in Book II, con-
taining the cartas pagenses. More than half of these – thirty-two out of
fifty-two – qualify as ‘lay’ formulas according to the criteria used above.35
A good number of them deal with inheritance, marriage or other intrafa-
milial property arrangements.36 Others concern dispute settlements or
judicial procedures that required documents.37 We find here securities for
debts and receipts for debts paid,38 and manumission charters and other
documents relating to serfs and to free or unfree status.39 We find formu-
las for letters of commendation and for letters asking for intercession.40
Some particularly striking examples speak volumes about the society
that produced them. A father declares that the usual custom, according
to which only sons should inherit, is impious and that he will not follow
it; he therefore gives his daughter a share in the family inheritance equal
to that of her brothers.41 A man has forcibly married a girl without her
parents’ consent. The formula gives two options for the girl herself: the
marriage occurred either with her consent or without it. In either case,
the man is in danger of his life until, with the intercession of priests
and other good men, he agrees to make the required prenuptial property
gift.42 Members of a kindred write to a man who had killed their brother,
to say that the man had paid the agreed-upon wergeld and should hence-
forth remain free from trouble about the matter.43 A man got in some
sort of trouble for which he was unable to pay compensation. His lord
stepped forward and paid the compensation for him; in return, the man
agrees to become the lord’s servus.44 A marriage has been marked not
by God’s charity but by discord to the point that the man and wife can
30 Marculf 1.33–4. 31 Marculf 1.36. 32 Marculf 1.17.

33 Marculf 1.18. 34 Marculf 1.38.
35 Marculf 2.9–19, 21–2, 24–38, 41, 50–2 (a companion to 1.39). Note that in 2.16 and
18 priests are mentioned as among the intercessors in a dispute among laypeople.
36 Marculf 2.9–17. 37 Marculf 2.16?, 30–1, 41. 38 Marculf 2.18, 25–7, 35.
39 Marculf 2.28–9, 32–4. 40 Marculf 2.50–1. 41 Marculf 2.12.
42 Marculf 2.16. 43 Marculf 2.18. 44 Marculf 2.28.
132 Warren C. Brown
no longer talk to each other; as a consequence, the couple agree to get

divorced.45
None of these lay formulas in Book II represent documents that a
church or monastery would have needed because they dealt with trans-
actions among laypeople in which a church or monastery had an interest,
for example, because some of the property involved belonged to it. The
collection does cover such a case, but with its own specific formula: two
men exchange villas, but one has to get a bishop’s permission to do
so, presumably because the bishop’s property interests were somehow
involved.46 In short, these lay formulas represent documents for which a
church or monastery would have had no direct use.
Like Book I, Marculf’s Book II tells us that there was no difference
between laypeople and ecclesiastics as far as getting involved in matters
that could produce or require documents was concerned. Another ‘two-
way’ formula secures property arrangements between a childless couple
in case one survives the other. One version allows the surviving spouse to
include a church among his or her heirs, while another specifies that the
property pass undiminished to the couple’s legitimate heirs.47 Two of the
formulas summarized above, the stolen bride formula and the receipt for
a wergeld payment, have priests as well as laymen involved in bringing
about the settlements concerned.48 If Marculf is any guide, ecclesiastics
wrote more letters, though laypeople did on occasion as well.49
Nevertheless, by presenting clergy and laity as options, Marculf once
more recognizes the distinction between lay and ecclesiastical. Clerics
and monks, monasteries and churches, were in his world qualitatively dif-
ferent from laypeople and lay kindreds, and therefore sometimes needed
their own documents, even if the transactions involved were of the same
sort that laypeople carried out by themselves. For their part, laypeople
did some things with documents that clerics did not, like get married or
divorced, pass property to children, pay wergeld for homicide, etc. The
formulas covering these situations strongly support the conclusion that
laypeople in Marculf’s world sometimes used writing in spheres and for
purposes different and separate from those of the churches and monas-
teries.
While I have argued so far that our picture of lay document use reflects
Marculf ’s world of the late seventh century (or at least the world of
possibilities that Marculf imagined for students), his formulas contain
echoes of lay documentary culture that go back much further in time.
As noted in the previous chapter, Marculf included in his collection a
45 Marculf 2.30. 46 Marculf 2.23. 47 Marculf 2.7–8.

48 Marculf 2.16, 18. 49 See Marculf 2.52.
pair of formulas that carry forward the language and procedures asso-
ciated with the late antique gesta municipalia, presumably because these
formulas were still useful for recording processes by which documents
were publicly validated.50 A formula for a testament similarly evokes a
procedure visible in the Italian papyri for validating a will. The cou-
ple concerned have a notary draw up their testament; a specified num-
ber of days after their deaths, specified legates were to take the will
before the gesta municipalia and – after the seal had been recognized and
the cord binding the document roll had been cut as the authority of
Roman law declared – have it validated.51 Several of the Marculf for-
mulas end with the tag stipulatione subnexa. This is a fossilized vestige
of the Roman stipulatio Aquiliana, a clause attached to transaction doc-
uments that declared both parties to be free from claims by one party
against the other.52 None of this is to say that in Marculf ’s world Roman
procedures and institutions were still functioning as they had in Late
Antiquity. But people were apparently still doing things similar enough
to Roman precedents that Marculf considered formulas using the old lan-
guage and describing the old procedures to be the proper way to record
them.53
Tours
The formula collection from Tours is another that survives in several
manuscripts and whose manuscript tradition points to the existence of
an original ‘ur-collection’. Although some manuscripts contain more
formulas, and some less, the core of the collection comprises thirty-
three document texts, some of which refer directly to the monastery of
St Martin in Tours54 or to the city of Tours itself.55 It was most likely
assembled in the eighth century, possibly during the reign of the sons of
Charles Martel as mayors of the Frankish palace.56
The first thing to note about the Tours collection is that very little
of it could have been of direct use to the monks of St Martin – only
three formulas, to be specific: one for a donation to a church, one for
a precarial grant, and one in which the poor on the alms rolls of the
50 Marculf 2.37–8; see above, 105. 51 Marculf 2.17; see above, 98–102.
52 Marculf 2.1, 3–4, 6–7, 9–11, 14, 19, 22–4, 29, 32, 36, 39, 52. Stipulatio: R. Zim-
mermann, The Law of Obligations: Roman Foundations of the Civilian Tradition (Cape
Town, 1990; repr. Oxford, 1996), 757; F. Sturm, Stipulatio Aquiliana: Textgestalt und
Tragweite der Aquilianischen Ausgleischsquittung im klassischen Römischen Recht (Munich,
1972).
53 See also above, Chapter 5. 54 Form. Tur. 1b, 11. 55 Form. Tur. 3, 28–9.
56 Rio, Legal Practice, 113, 116–17; Depreux, ‘La tradition manuscrite’, 55–9, 67–70.
134 Warren C. Brown
monastery (matricularii) find an abandoned infant and locate someone

to adopt it.57 A couple of others record transactions that tangentially
involve a monastery or a church: a layperson sells to another layperson
land that lay within the boundaries (termini) of a monastery, and the
manumission of a servus takes place before the altar and clergy of a
church.58
The overwhelming majority of the Tours formulas do not involve a
church or monastery, or clerics or monks, at all. We see laypeople giving
each other power of attorney to handle their affairs59 and having docu-
ments validated by the civic authorities.60 We see them giving or receiving
property or benefices to/from each other and exchanging property;61 we
see them selling property to each other62 (including selling themselves
into someone else’s service)63 and freeing servi.64 We see them making
property arrangements in advance of a marriage,65 kidnapping (or elop-
ing) with brides and settling the resulting disputes,66 regulating inheri-
tance arrangements and legal powers within their families and adopting
heirs,67 and getting divorced.68 We see judges giving orphans to people
to be raised and educated;69 we see people having their property rights
confirmed by a king after their documents had been destroyed70 and set-
tling conflicts over everything from seizure of property to homicide and
assault on the road.71 One of these latter dispute formulas has a man
coming before a judicial assembly in Tours and charging another with
having unjustly seized his property. The formula has an optional variant
prologue in case the suit concerned homicide rather than property.72
Like their counterparts in Marculf, some of the Tours formulas suggest
strongly that these were documents that laypeople would take home and
keep. A letter by which a judge assigns a guardian for an orphan specifies
that two copies of the document be made, one for the guardian and one
for the person who would serve as guarantor for the arrangement.73 A
man who had lost his archive because he had feared the approach of a
marauding army and had buried it, only to find later that the documents
had rotted and were ruined, receives a copy of the charter confirming
his property rights from the civic authorities.74 The servus freed by his
master would plausibly want to keep the document recording his man-
umission, even if he could not read it; more likely, the master would
57 Form. Tur. 1a, 7, 11.

58 Form. Tur. 8, 12; see Depreux, ‘La tradition manuscrite’, 58.
59 Form. Tur. 2. 60 Form. Tur. 2–3. 61 Form. Tur. 4, 6, 13, 26.
62 Form. Tur. 5, 8, 9. 63 Form. Tur. 10. 64 Form. Tur. 12. 65 Form. Tur. 14–15.
66 Form. Tur. 16, 32. 67 Form. Tur. 17–18, 20–3, 25. 68 Form. Tur. 19.
69 Form. Tur. 24. 70 Form. Tur. 27–8. 71 Form. Tur. 29, 30–1, 33.
72 Form. Tur. 29. 73 Form. Tur. 24. 74 Form. Tur. 27–8.
want to keep it as evidence for his position as the newly freed person’s
benefactor.75
Other Tours formulas indicate that documents were ‘out there’ filtering
through society in the world beyond the collection. The lost archive case
just mentioned lists a number of possible kinds of documents that the
victim could have had in his possession: documents recording property
gifts from kings, as well as property sales, gifts, grants and exchanges.76
A post obitum gift of property to the monastery of St Martin explicitly
blocks anyone from trying to undermine the arrangement by showing up
with a charter that the donor had not personally approved.77
Also like Marculf, the language used by the Tours collection echoes
a culture of document use going back to Late Antiquity. Many of the
Tours formulas refer to Roman law, some citing specific passages from
the Theodosian Code or the Lex Romana Visigothorum.78 The tag stipu-
latio subnixa likewise shows up frequently. One formula even recalls the
tag’s origins; in a formula describing a property arrangement between a
husband and wife, the penalty clause actually says that the two copies
of the document to be produced should remain firm cum stipulatione
Aquiliana.79 As discussed in the previous chapter, several of the Tours
formulas assume that the transactions they record would be validated by
the local gesta municipalia;80 others have their transactions carried out in
the presence of the civic curia.81 The manumission charter mentioned
above guarantees the newly freed servus the right ‘to lead a free life just like
other Roman citizens’ (et sicut alii cives Romani vitam ducat ingenuam).82
As with the similar language in the Marculf collection, it is impossible
to tell to what degree, if at all, this language might have been relevant or
comprehensible to those who might have used the Tours formulas as a
basis for real documents in the course of real transactions. Nevertheless,
as far as the compilers of the collection were concerned, it apparently
still formed part of the process of drafting these documents in the proper
way. The monks may have understood some of the references to Roman
law or procedures that they copied; others may have been traditional but
less comprehensible. Nevertheless, the Tours monks regarded the old
75 Form. Tur. 12; compare above, 130–1. 76 Form. Tur. 27. 77 Form. Tur. 1b.
78 Cf. Form. Tur. 11 (CTh. 5), 14 (LRV), 15, 16 (LRV, Epit. Aeg.), 17 (LRV), 19–21 (all
LRV), 22, 24, 25, 29–30 (both LRV), 32. Those that do not specifically quote the law
contain general phrases such as ‘secundum legem Romanam’ or ‘lex Romana constrin-
git’. On the continued use of Roman law in the region, especially among churchmen,
see P. Stein, Roman Law in European History (Cambridge 1999), 38–41.
79 Form. Tur. 17.
80 Form. Tur. 2–3, 17, 20, 23; see above, 105, 109. 81 Form. Tur. 23.
82 Form. Tur. 12. See also the (albeit transformed) echoes of the late antique appennis
procedure in Form. Tur. 27–8.
136 Warren C. Brown
language as necessary to the document models that they wanted to have

on hand.
The fact that the Tours collection refers directly to the monks of St
Martin, but that most of the formulas it contains would have been of little
direct use to the monks, suggests that in mid-eighth-century Tours the
scribes of St Martin were generating – or at least wanted to know about –
documents needed by the laypeople around them, of all social stripes.83
This too, then, was a world in which laypeople of all kinds were actively
using and keeping documents, and producing them for examination when
necessary. Like Marculf, the compilers of the Tours collection dragged
with them a great deal of far older language and references to older
procedures. It appears, therefore, that in the older world or worlds from
which the Tours monks drew their documentary traditions, laypeople
had actively used documents as well.
Flavigny
Sometime in the second half of the eighth century, and possibly towards
the very end of the century, monks at the Burgundian abbey of St
Praeiectus at Flavigny compiled a formula collection of their own. They
took much, though not all, of Marculf’s collection, added substantial
pieces from the Tours collection, and filled in the blended collection that
resulted with a small number of unique formulas that came directly from
their own archive and library.84 There is no doubt that the collection was
put together at Flavigny; among the documents that the monks pulled
out of their abbey’s archive to make formulas were two, a property gift
and a testament, written by the abbey’s founder, Widerad. How the Flav-
igny monks got hold of the Marculf and Tours formula collections is not
clear. Philippe Depreux has speculated quite plausibly that the Tours
collection made its way to Flavigny through the good offices of, or the
connections inspired by, Alcuin, who was abbot of Tours and, after 797,
simultaneously abbot of Flavigny.85
83 But it is possible that the collection was not put together by the monks of St Martin
at all, but rather by someone working outside the monastery, but who included some
potentially useful documents from the monastery.
84 Rio, Legal Practice, 117–21. The collection includes two formulas for testaments that
its scribes derived from testaments by Flavigny’s founder, Widerad; these testaments
were most likely written in 717 and 719. However, the incorporation of the mid-eighth-
century formula collection from Tours pushes the terminus post quem for the collection up
to that point. If the Tours collection reached Flavigny through connections inspired by
Alcuin (see below, at n. 85), then the Flavigny collection would have been put together
some time during Alcuin’s simultaneous tenure as abbot of Tours and of Flavigny, i.e.
after 797.
85 Depreux, ‘La tradition manuscrite’, 61–2.
The original manuscript of this Flavigny collection has been lost. It

must have once existed, however, because the collection survives in
two manuscripts, one now in Copenhagen86 and the other our Paris
manuscript BnF lat. 2123, that differ enough from each other to indi-
cate that both stemmed from a common model rather than one from
the other. Though the Copenhagen manuscript reproduces the formu-
las in a different order from the Paris manuscript, the Paris manuscript
is assumed to reproduce the ordering of their common source because
it gives its versions of the Marculf and Tours formulas in more or less
their original order. Some of the formula texts given in the Paris table of
contents are missing from the Paris manuscript; nevertheless, they can
be reconstructed with the aid of the Copenhagen version.87
The Flavigny collection is numbered through from 1 to 117.88 Its
compilers cast a wide net in their search for models. The collection
starts off with five late antique texts, four letters and a speech, drawn
from the Acts of S. Sylvester and from the Latin translation of the sixth-
century Greek Historia tripartita of Theodorus Lector. These include
letters between the Roman emperor Constantine and his mother Helena,
a letter from the emperor Constantius to Bishop Athanasius, part of a
letter from Pope Julius to the clergy and people of Alexandria, and a
short fragment of a speech purportedly given by Emperor Valentinian
II on the election of Bishop Ambrose of Milan. All of the names and
specific details were left in the texts and/or titles, making them easy
to identify. Nevertheless, it does appear that they were intended to be
models or sources for language. None of the five are set off or in any way
distinguished from the rest of the formula collection. The last of the five
consists only of three sentences. Though we can identify it as part of an
imperial speech said to have been given in a specific context,89 this brief
excerpt makes no mention of that context. It consists only of general
exhortations and is titled in a way – de ordinando episcopo. Valentinianus
imperator ad episcopos ait – that generalizes it into something that could be
generically useful as an example of an imperial letter on the ordination
86 Copenhagen 84 (Francia, s. IXmed ); Rio, Legal Practice, 243–4.

87 The scholarship now assumes a stemma that starts with the original and proceeds
to a copy from which both the Copenhagen and Paris manuscripts stem; another,
reordered exemplar must have lain between this copy and the Copenhagen manuscript.
See Rio, Legal Practice, 118; Depreux, ‘La tradition manuscrite’, 63; K. Zeumer, ‘Die
Lindenbruch’sche Handschrift der Formelsammlung von Flavigny’, Neues Archiv der
Gesellschaft für ältere deutsche Geschichtskunde, 14 (1889), 589–603, at 593–603.
88 Thus the table of contents in BnF lat. 2123. The manuscript includes five more formulas
after no. 117, some of which may have been part of the original Flavigny collection; see
below, n. 123.
89 The speech is reported by Theodoret, as transmitted in Cassiodorus-Epiphanius, His-
toria ecclesiastica tripartita 7.8 (CSEL 71:394).
138 Warren C. Brown
of bishops.90 The impression that the compilers of the collection were

treating these antique texts as formulaic models is strengthened by some
letter fragments from a similar source that appear at the end of the
collection. The table of contents titles the very last entry, Flavigny 117,
as ‘letters made in different ways’ (Indicolos diversos modos factos). This
entry consists of twelve openings for letters addressed by bishops to a
variety of recipients, the first few of which are simply headed item alio.
The last two of the twelve, titled Dionisius episcopus ad Romanos and Item
ad imperatorem respectively, have been identified as coming from Rufinus’
translation of Eusebius’ Ecclesiastical History.91 The first of these two
retains in the title the name of the person who wrote it. Nevertheless,
the texts themselves are very short excerpts – two and three sentences
respectively – that mention no names; they are completely subsumed
within Flavigny 117 as examples of Indicolos diversos modos factos.92
The bulk of the Flavigny collection consists of large pieces of the Mar-
culf and Tours collections interspersed with formula texts taken from the
Flavigny archives that all involve clerics, monks or ecclesiastical affairs.
The Tours and Marculf sections are for the most part copied in their
original order and with very few changes (though most of the specific
references to St Martin’s in Tours have been dropped);93 they therefore
pick up and reproduce the priorities of their original compilers. These
copies of the Tours and Marculf formulas were, however, by no means
blind; the Flavigny scribe or scribes did not simply copy them out of
antiquarian interest or a respect for the old that was disconnected from
the realities of their world. On the contrary, they copied with a visible
intent to put together a collection that fitted their needs.
The first evidence for this comes right away with the sixth formula.
This formula represents a post obitum gift to a church; it is taken from
Tours 1.94 The scribe did not move immediately on to Tours 2, how-
ever. Instead, he followed Tours 1 with a Flavigny formula, taken neither
90 Rio suggests that the names were left in these texts because the protagonists were
prestigious; she also suggests that esp. no. 5 ‘is very much the sort of thing one would
expect to find in such a [formula] collection’; Rio, Legal Practice, 119.
91 MGH Formulae 498–9 (Coll. Flav.), marginal notes to 117l and 117m.
92 It is possible to question whether Flavigny 1–5 and 117 belonged to the original Flavigny
collection; they appear only in the Paris manuscript. Because they are listed in the Paris
table of contents, however, which is presumed to follow the exemplar being copied by
the compilers of the Paris manuscript, it seems justified to assume that they did form
part of the original collection. If they did not, then the comments here about their use
as formulas apply to their inclusion in the Paris manuscript, where they are presented
physically, and treated in every way, as organic parts of the larger formula collection.
93 The exceptions are Coll. Flav. 35–6 = Form. Tur. 28–9, which preserve the specific
references to the civitas of Tours.
94 Coll. Flav. 6ab.
from Tours or Marculf, that represents a gift to a church or monastery a

die presente.95 He thus created a matching pair of gifts, one post obitum,
the other a die presente. Next comes a testament, in which a cleric leaves
property to a church.96 The text was adapted from the testament of Flav-
igny’s founder, Abbot Widerad. However, the beginning of the formula
was taken from the corresponding formula for a will in Marculf.97 The
Flavigny copyist took with the Marculf excerpt its late antique language
mandating the will’s validation before the gesta municipalia. He modified
the Marculf text, however, so that a copy of the will would also be stored
in a church archive (in archivis basilice sancti illius). In other words, he
stitched together exactly the text he wanted.
Just as telling are the next two formulas. These are the two Flavingy
gesta municipalia formulas discussed in the previous chapter.98 First is the
mandate from a property donor to a magnificus frater, asking him to go in
his place to the city, have the donation validated by the gesta municipalia,
and return with confirmation that he had done so. The second is the
blow-by-blow description of the mandate bearer carrying out his task
before the gesta assembly.99 Both the Tours and the Marculf collections
contain formulas for such documents: Tours 2–3 and Marculf 2.37–8.
One would expect that, having led off this group of formulas with Tours
1, the Flavigny copyist would have simply used Tours 2–3. He did not.
What he did was to make up new formulas by mixing pieces from Tours
2–3 with pieces of Marculf 2.37–8.100 In short, despite their late antique
language and descriptions of procedure, the compiler of the Flavigny
collection understood these formulas. He also knew what he wanted,
and he knew how to blend pre-existing but apparently not completely
right material together to get what he wanted. He must, therefore, have
considered these formulas to be not only relevant to his world but also
important enough to be worth putting in the effort to get them right.
The twenty-nine formulas that follow, Flavigny 11–39, are copies, in
order, of Tours 4–32. They therefore reproduce the order, contents and
priorities of the Tours collection, including its view of lay participation in
documentary culture. Flavigny 39, however, contains an interesting wrin-
kle; someone working on the text changed it to fit his preconceptions.
In the original Tours version of this formula, a couple eloped without
permission from the girl’s parents but also without objection or com-
plaint from either the parents or the girl. The crucial sentence explaining
the situation reads as follows: ‘accusabat aliquo homine nomine illo, eo
95 Coll. Flav. 7. 96 Coll. Flav. 8. 97 Marculf 2.17.

98 Above, 109. 99 Coll. Flav. 9–10. 100Into Zeumer’s Form. Tur. Add. 4–5.
140 Warren C. Brown
quod aliqua femina nomine illa iam anno expleto sine diffinicione par-
entum vel sine eius clamore aut vociferatione eam volentem rapuisset
atque in coniugio sibi malo ordine contra legem et iustitiam sociasset’.
In both the Copenhagen and the Paris manuscripts of the Flavigny col-
lection, the meaning of this sentence has been altered. Instead of ‘eam
volentem rapuisset’, we read ‘eam violenter rapuisset’.101 It is impossible
to tell exactly when this change took place. It may have been present
already in the exemplar from which the Flavigny copyist was working;
it may have happened when the Flavigny version was compiled, when
the exemplar that served as the common source for the Copenhagen
and Paris manuscripts was copied out, or at some point in between. It
is also impossible to say whether the change was conscious or uncon-
scious. Whatever the case, the original Tours language did not fit the
assumptions of whoever made the change. It made more sense to him,
or he automatically assumed as he was writing, that any raptus, that is, in
this context any carrying off of a girl for the purpose of marriage, would
have been carried out forcefully or impetuously, and would have been a
violation of right, regardless of what the girl or her parents thought about
the matter.102
The last formula in this Tours section, Flavigny 40, brings evidence
of unequivocally deliberate change. It has the title of Tours 33, iudicium
evindicatum, but it has different content. The Tours original was a direc-
tive from a king to a count; a plaintiff had come to the king saying that
a man of the count’s pagus had beaten him up on the road, taken his
property and seriously injured him. The accused had been summoned
to a placitum but had failed to show up. The king therefore ordered the
count to constrain the accused and force him to make adequate amends.
The Flavigny copyist103 evidently felt that he needed to include a for-
mula like this in his collection, but that the most likely judicial scenario
would be different. So he altered the text to turn the charge into one of
unjust property seizure rather than assault and robbery; the king ordered
101 Turning eam from part of an accusative absolute into a direct object. See Copenhagen
84 fol. 28v, and BnF lat. 2123 fol. 122v.
102 The word violenter in this period does not necessarily mean ‘violently’ in the modern
sense; it has the more specific connotations of acts carried out with forcefulness or
impetuosity, or of a violation of boundaries. See W. C. Brown, Violence in Medieval
Europe (London, 2010), 6–7 and nn. 25–6.
103 It is possible that this change was introduced not by the Flavigny copyist, but by
whoever wrote the exemplar from which he was working. Given that this alteration
is not present in any other extant manuscript of the Tours collection, however, and
given the other evidence for intervention in these texts by the Flavigny copyists, I have
concluded that this change was most likely made when the Flavigny collection was
compiled.
that since the defendant had failed to attend the placitum and swear an
oath to his rights, the plaintiff’s possession of the disputed property be
confirmed.104
From here we move into a section of the Flavigny collection dominated
by formulas from Marculf’s Book I. As with the group taken from Tours,
at first we get some additions. Three unique formulas are inserted right
after Marculf 1.1. Then come Marculf 1.2–4, then another formula is
inserted and, after Marculf 1.5–7, yet another.105 The last two inserted
formulas appear in other Marculf manuscripts.106 They must have come
from the manuscript of Marculf that the Flavigny copyist had in front of
him. Seen as a whole, this newly constructed group of formulas is not
random, but rather has an identifiable logic; it deals with episcopal rights
and privileges, especially immunity. Its last three formulas cover all the
steps involved in appointing a new bishop, from the initial royal command
that the bishop take up his post, to the royal request to another bishop
to consecrate him, to the letter from the citizens of the city formally
requesting that the new bishop be installed.
What follows is the rest of Marculf Book I, complete. The Flavigny
compiler(s) thus reproduced the preoccupations, and assumptions about
lay document use, of this part of Marculf’s collection. Next, we move
into the material drawn from Marculf Book II; here the Flavigny com-
piler(s) did a great deal more selection and reordering. The selection and
reordering were again not random. A number of formulas from Marculf
Book II were left out, apparently because they were redundant.107 They
are either covered by formulas from Tours or Flavigny that appear earlier
in the collection, or they offer variations on other material from Marculf,
variations that the Flavigny compiler(s) must have thought unimportant.
The formulas that the Flavigny compiler(s) did take from Marculf Book
104 Note that a similar Marculf formula, 1.29, is also in the Flavigny collection, as no.
74. It is not a duplicate of Form. Tur. 33 = Coll. Flav. 40, however. It represents a
different stage in the proceedings; rather than a royal order to constrain someone who
had ignored a summons to a placitum, this formula represents the original command
to the accused that he attend the placitum.
105 Coll. Flav. 41–52. 106 Marculf Supp. 1, 6.
107 It is possible that these formulas were already missing from the Marculf manuscript
from which the Flavigny copyists were working. However, the fact that they would have
been redundant specifically in the context of the Flavigny collection, where what was
copied was arranged in coherent groups (see below), and that their content was covered
in part by formulas from Flavigny, suggests to me that they were dropped as the Flavigny
collection was compiled. An alternate scenario would have the Flavigny copyists, faced
with a Marculf manuscript in which these formulas were missing, working with what
they had to fill in the gaps and fill out their groupings. If this were the case, then
someone earlier, compiling a now lost recension of Marculf, engaged in the process of
evaluation and decision-making that led to these formulas being dropped.
142 Warren C. Brown
II were gathered together and reordered to make new groups of formulas.

For example, Flavigny 86–90 takes Marculf 2.1, 3, 5, 39–40 and 41 and
groups them into a set of formulas dealing with donations of property and
with benefice arrangements. The missing formulas Marculf 2.2, 4 and
6 are covered by formulas from Tours or Flavigny, or by other Marculf
formulas.108 Then comes a group of formulas for property arrangements
within families that was streamlined compared to the similar section
in Marculf.109 Then, a security for a wergeld payment is followed by
property sales, more securities and debt payments, and one power of
attorney.110 A set of formulas dealing with unfree status111 is followed
by a mixed bag: a royal confirmation of a sale, a letter from a mayor of
the palace to a bishop, and a letter from a bishop to his mother,112 all of
which segue easily into a group of letters to bishops, kings and queens
to be sent at Christmas and Easter.113 The section ends with letters of
supplication and commendation.114
To sum up: the compiler(s) of the Flavigny formula collection selected
their formulas to meet their abbey’s needs for model documents as they
understood those needs. They drew their texts from the Marculf and
Tours collections, and from their own archives and library, and delib-
erately selected and arranged them. Although the Flavigny additions
all concern clerical/monastic affairs, the Flavigny collection as a whole
retains a large number of forms for documents that would only be useful
to laypeople. The way these lay formulas are arranged reveals the same
kind of evidence of conscious selection, arrangement and modification
as do the clerical formulas, indicating that the compiler or compilers
understood what they were doing when they copied them.
The collection that resulted projects assumptions about lay partici-
pation in documentary culture in the area around Flavigny in the later
eighth century that are essentially the same as those projected by the
earlier Tours and Marculf collections. The fact that the Flavigny com-
pilers chose to copy the formulas that they did indicates either that they
felt a need to know about and study, or to teach young scribes about,
lay documents in addition to those their abbey would need for its own
108 Marculf 2.2 is a variant prologue for the same kind of donation to a church given by
Marculf 2.3; Marculf 2.4 is roughly equivalent to Coll. Flav. 7; Marculf 2.6 is roughly
equivalent to Form. Tur. 1b.
109 Coll. Flav. 91, 92–5 = Marculf 2.11–12, 14–15.
110 Coll. Flav. 96–101 = Marculf 2.18, 19–22, 25–7, 35, 28, 31.
111 Coll. Flav. 102–4 = Marculf 2.29, 32–4, 36.
112 Coll. Flav. 105–7 = Marculf Supp. 2–4.
113 Coll. Flav. 108–10 = Marculf 2.42–4. 114 Coll. Flav. 111–16 = Marculf 2.46–51.
purposes, or that they needed the lay document formulas because they
would be producing documents for laypeople.
BnF lat. 2123

All of the threads that we have been following come together in a sur-
viving physical object: our manuscript. This manuscript reveals some
things about how formula collections were created. Equally important, it
embeds the Flavigny formulas in a particular context that tells us more
about why our formulas for lay documents might have been copied and
kept.
BnF lat. 2123 has, with varying degrees of precision, been dated to the
end of the eighth century or the beginning of the ninth, to 795×816, or
even to 814×16.115 It most likely originated somewhere in Burgundy.116
The codex now comprises 156 folia; a few of its original folia are missing.
Its texts are written out in two columns, in a pre-Caroline minuscule,
which might indicate a date of the very end of the eighth century rather
than the ninth.117 We even have a named scribe; on fol. 91 is the comment
Walefredus me fecit.118
The codex appears to be an ecclesiastical reference book. It contains, in
addition to the Flavigny formulas, excerpts from Church council acts and
papal letters, an incomplete copy of Gennadius of Marseilles’s De eccle-
siasticis dogmatibus, some pseudo-Augustinian sermons, the Cononian
epitome of the Liber pontificalis, Polemius Silvius’ Laterculus, the Dicta of
Theodore of Canterbury, the canon law collection known as the Heroval-
liana, a computus of the ages of the world up to the reign of Charlemagne,
and the sections from Isidore of Seville’s Etymologies on weights and
measures.119 It was conceived as a whole and executed with some care;
it is written out in the same way throughout, with the same kinds of
(often rather nice) rubrics and capitals (and occasional drawings in the
margins), by if not the same scribe then scribes with very similar hands,
with no visible breaks across the various texts.
115 See the literature cited above, n. 4.

116 Attributions to Flavigny itself appear to rest too much on its copy of the Flavigny
formulas.
117 Rio, Legal Practice, 252.
118 At the bottom of fol. 75r, the name Hathulfus is written next to the name Walefredus; at
the bottom of fol. 65v, the name Hathulfus appears by itself. This suggests that there
were perhaps two scribes involved in writing the codex. If so, their hands are very
similar. On fol. 52r, a later hand (s. X–XI?) continues the list of popes, which had
originally ended with Hadrian I, from Leo III to John IX.
144 Warren C. Brown
The Flavigny formula collection starts on fol. 105v. The scribe began
with a spectacular initial and capitals for the title Incipit praefacio libri
huius. Following the title comes the prologue to Marculf ’s formula col-
lection; it has been adapted here as a preface for the Flavigny collection.
The prologue preserves Marculf’s self-attribution and his explanation for
the collection’s purpose. There is one change, however: Bishop Landeric
has become Bishop Aeglidulf.120 It has proven impossible to identify
Aeglidulf. It is also impossible to say whether this change was made
when this manuscript was compiled, or at some earlier stage in the his-
tory of the Flavigny collection.121 Nevertheless, someone – either our
scribe or an earlier one – used Marculf’s prologue essentially as a for-
mula in its own right, by simply plugging in the name of the appropriate
bishop.122
After the prologue comes a lengthy table of contents. The table of
contents does not entirely match the actual contents of the formula col-
lection. Five formulas are written on fols. 151–153v that are not listed
in the table of contents; these concern property donations, precarial and
prestarial grants to and from a monastery, and a monk who had fled his
monastery.123 Moreover, as noted earlier, several of the formulas given in
the table of contents are missing from the manuscript. That they formed
part of the original Flavigny collection is ensured by the other (Copen-
hagen) manuscript of the collection.124 Their absence here is not due to
missing folia. It appears to have been due to simple scribal error; the text
moves seamlessly from one formula to the next but skips one or more
formulas listed in the table of contents, including their numbers. For
example, on fol. 148r, we continue smoothly, in the middle of a column,
from Flavigny 96 to Flavigny 98 with no visible change of hands.125
Nevertheless, the manuscript offers considerable evidence that care
was taken with the formula collection, that at least some level of oversight
went into its production, and that the person or people doing the work
understood what he or they were doing and made some effort to get
it right. This applies to the lay formulas as well as to those in which
clergy were involved. Words found to be missing were inserted along with
120 BnF lat. 2123 fol. 105v.

121 The Marculf preface is not in the Copenhagen 84 version of the Flavigny collection.
123 These formulas were edited by Zeumer as Coll. Flav. Add. 1–5; Rio, Legal Practice,
118–19. The beginning of Add. 1 has been lost along with the folium on which it was
written.
124 See Rio, Legal Practice, 118, 243.
125 See also fol. 140r, skip from Coll. Flav. 80 to 86; fols. 147v–148r, skip from Coll. Flav.
94 to 96; fol. 148v, skip from last formula of Coll. Flav. 98 group to Coll. Flav. 100;
fol. 150v, skip from Coll. Flav. 104 to 111.
insertion marks to show where they belonged. For example, on fol. 122v,
when copying out Flavigny 39 = Tours 32 (the formula discussed above
in which a man ran off with a girl violenter instead of eam volentem), the
scribe, when he reached the sentence explaining the situation, skipped
all the way from sine diffinicione to eam violenter rapuisset. Catching his
error, he placed a sign at the end of diffinicione and a corresponding sign
in the bottom margin of the page, where he wrote the missing text: vel
sine eius clamore aut vociferacione (though he still left out the parentum that
should have followed diffinicione).126 Mistakes were marked with rows of
dots underneath or above the affected lines, which told readers to ignore
them. For example, on fol. 138r, in the middle of the second column, in
the text of Flavigny 77 = Marculf 1.32 (in which a king gives property
confiscated from rebels to some viri inlustri), the words mixti fuerunt are
supposed to appear twice; the first time in the phrase, ‘qui cum eodem
mixti fuerunt, ex hoc conprehensum adduxerunt’, and the second time in
the phrase, ‘qui cum illo mixti fuerunt, nec ab heredes eorum’. After the
scribe wrote the first mixti fuerunt, he let his attention wander and skipped
to the nec ad heredes that followed the second. Realizing his mistake after
writing these three words, he marked them as an error with a row of
dots underneath, and then continued correctly with ex hoc conprehensum
adduxerunt.127
The most spectacular mistake tells us something about how the for-
mula collection in this manuscript was put together; the scribe had in
front of him separate leaves or quires from his sources, leaves or quires
that could get mixed up. On fol. 147r, in the middle of the second col-
umn, Flavigny 91 = Marculf 2.9 ends (see Figure 6.1). Then comes
the title of Flavigny 92 = Marculf 2.11. Instead of the opening line of
this formula, however, we get, complete with a nicely decorated capital,
126 See also fol. 137rb middle (in Coll. Flav. 73 = Marculf 1.28 the missing ‘mo’ in ‘cum
omnibus modis’ is written in above the line with an insertion mark); fol. 115rb, two-
thirds of the way down (a missing ‘quis suum’ inserted in the text of Coll. Flav. 18 =
Form. Tur. 11); fol. 144va, bottom (in the opening lines of Coll. Flav. 117a, the missing
‘moda’ inserted for the phrase ‘salutem vobis multimoda’).
127 See also: fol. 135v, in Coll. Flav. 67 = Marculf 1.22: dots mark a missing ‘t’ in ‘aut
sua’. Fol. 136r: in the same formula text, dots mark mistaken letters copied between
‘titulum’ and ‘a iugo’. Fol. 145v: at the beginning of Coll. Flav. 117e, the scribe skipped
part of a line and wrote a few words that should have come later; he made a mark,
blocked off the incorrect text with two dots over and under the offending words at
either end, made a second mark, and then started again where he made the mistake
but this time wrote correctly. Fol. 146v: in Coll. Flav. 117i, the scribe wrote ‘dign&tis’
instead of ‘dignastis’, so he crossed out the ‘&’, marked it with dots above and below,
and wrote the letters ‘as’ above it. Fol. 147v, at the beginning of Coll. Flav. 93 =
Marculf 2.12, the scribe wrote, instead of ‘ut de terra paterna’, ‘ut d& erna’, so he put
dots above and below between the ‘e’ and ‘r’ of ‘erna’, and simply placed the missing
letters ‘ra pate’ above, to create ‘ut d& er[ra pate]rna’.
146 Warren C. Brown
Figure 6.1 BnF lat. 2123, fol. 147r detail: the opening line of Marculf
2.15 written by mistake between the title and first line of Flavigny 92,
with cross-marks to guide the reader around it.
the line, ‘Quod bonum filex prosperumque veniat disponsandis maritan-

disque ordinibus ac procreatione liberorum causis’. At the beginning of
this line, the scribe placed a cross-mark; at the end, he did so again.
Next comes the proper opening line of Flavigny 92, which represents a
property gift from a grandfather to a dutiful grandson. The line in

between the title and the correct text turns out to be the beginning of
Marculf 2.15, in which a man about to be a father-in-law makes a dotal
gift to his son’s betrothed; according to the table of contents, this formula
was supposed to appear in the collection farther down as Flavingy 95. The
scribe apparently had leaves or quires of his exemplar of the Flavigny col-
lection in front of him but got them mixed up; after he copied the begin-
ning of Flavigny 95 = Marculf 2.15, he realized his mistake and made a
note for readers to skip over the abortive copy to the correct beginning
of Flavigny 92. As it happens, Flavigny 95 is one of the formulas that got
left out of the manuscript; when he got there, the scribe skipped straight
from 94 to 96. When he came to copy it, he must have thought he had
copied it already, or had misplaced his exemplar, or had simply forgotten
about it.
The care taken to show readers how to get around mistakes indicates
that this was indeed a codex intended to be used as a practical reference.
More evidence to this effect is provided by scattered efforts to demar-
cate formula texts proper from instructions for using formulaic options
and from shorthand references to common formulaic phrases. While
the scribe concerned did not do this consistently, on a few occasions
he marked such words or passages by suddenly shifting from the pre-
Caroline minuscule that dominates the codex into a flowery chancery-
like script. On fol. 121va, eight lines from the bottom, for example (see
Figure 6.2), we come to the place in the judicial formula Flavigny 36 =
Tours 29 where an optional prologue is provided for the case of homi-
cide; the direction, ‘Et si de homicidio accusacio facta fuerit’, stands out
clearly from the rest of the text by virtue of its suddenly more flowery
script.128
Then there is some internal cross-referencing. On fol. 120r, the scribe
copied Flavigny 34 = Tours 27, which is a royal confirmation of the
property rights covered by a lay archive lost to a fire. Next to the title,
he wrote, ‘alibi require sub hera lxxviii’; that is, ‘for another one see
number 78’. Sure enough, Flavigny 78 = Marculf 1.33 is another royal
precept confirming the contents of a lost lay archive. On fol. 136v, at the
bottom of the first column, the end of the royal placitum prologue Flavigny
70 = Marculf 1.25, we find inserted the line, ‘require in hera lxxxiii’; that
is, ‘see number 83’. Flavigny 83 is another one of the formula texts that
got left out, but, according to the table of contents, it was to have been the
royal judicial precept Marculf 1.38. Since Flavigny 70 is just a prologue, it
looks like the scribe wanted to refer searchers to another royal judicial text
that had more content (his note also, by the way, gives further evidence
128 See also fol. 114rb, line 4: ‘facta loco’; fol. 149rb: ‘et si voluntaria servo accipit dicis’.
148 Warren C. Brown
Figure 6.2 BnF lat. 2123, fol. 121va, detail: directions for use of an
optional prologue are written in a different script.
that the missing Flavigny 83 was supposed to be included but was left
out by mistake).
The choice of the word hera (or hira) for ‘chapter’ or ‘number’ in this
context is interesting. It is used three times in this copy of the Flavigny
collection in the same way to make cross-references.129 In general, this
word is rare in the formula collections, but it is used in the Herovalliana
129 See also fol. 142r, at the end of Coll. Flav. 88a = Marculf 2.5 (a precarial grant): ‘et
reliqua sicut in hira’. There is no number given for this one. Zeumer (MGH Formulae
484) seems to indicate that it refers back to Coll. Flav. 87 = Marculf 2.3, which is the
prologue for a donation to a church. However, it makes little sense for the ending of a
precarial grant to be completed by referring upwards to the beginning of a donation.
collection of canons, which appears earlier in this manuscript. Appar-

ently, the scribe who used the word in the formulas was influenced by its
use in the canon law collection – another sign that the same person or
same team was involved in producing other parts of the manuscript.130
Conclusions
In the end, this manuscript offers us evidence across an envelope of space
and time for a culture of document use in the heart of the Frankish world
that included laypeople. The envelope extends from the mid or late sev-
enth through the early ninth century and covers the Touraine, Neustria
and Burgundy. Echoes from formulaic language take the evidence for lay
documentary culture back even earlier.
At each stage of the timeline that culminated in our manuscript, the
formulas indicate that laypeople as well as clergy were using documents
for a variety of transactions and legal processes that involved personal
rights and privileges as well as property. The people involved ranged
from the powerful (kings and counts, bishops and archbishops), to the
powerless (that is, the unfree or those falling into unfreedom). Some
of these transactions and processes connected laypeople and clergy but
many did not; they involved only laypeople, or only clergy.
At least in the worlds of possibility imagined by those who compiled
the formula collections, laypeople and clerics or monks did similar things
with similar kinds of documents; both thought documents were impor-
tant and when necessary went to some effort to get them and to keep
them. The only distinction that we can really draw is this: clerics and
monks sometimes needed their own versions of common documents,
or engaged in kinds of transactions that were unique to them and their
institutions; to that degree, the formula collections treat laypeople and
clergy differently. The fact that formula scribes felt compelled at times
to include specifically clerical versions of common documents, however,
is good evidence that the specifically lay versions were important in their
own right, rather than simply sources or models for documents that clergy
needed. Moreover, the care with which our scribes collected, selected,
arranged and sometimes altered or combined formulas indicates that
they were all, lay as well as clerical, comprehensible and important to
them.
However, it is very important to note that monasteries played a funda-
mental role in collecting and preserving the model documents contained
130 H. Mordek, Kirchenrecht und Reform im Frankenreich: Die Collectio Vetus Gallica: Die
älteste systematische Kanonessammlung des Fränkischen Gallien (Berlin, 1975), 122n66.
150 Warren C. Brown
in our manuscript. Marculf was a monk who compiled his collection

for students at the behest of a bishop. The monastery of St Martin in
Tours compiled its own formula collection. Copies of both the Mar-
culf and Tours collections made their way to the abbey at Flavigny. If
Philippe Depreux is right, it was no less a man than Alcuin as lay abbot
of both Tours and Flavigny who was at least indirectly responsible for
the Tours collection coming to Flavigny. The monks at Flavigny took
these collections, added material from their own library and archives,
and made their own collection. It is impossible to say for sure where the
manuscript BnF lat. 2123 itself was put together. However, its contents
point unmistakably to an ecclesiastic or monastic milieu.
Clearly, some monasteries from the end of the seventh, but especially
from the middle of the eighth through the first decades of the ninth
century, were strongly interested in preserving and maintaining access
to models for documents that were potentially of use to a wide range of
people, lay as well as clerical, high born as well as low. By implication,
then, they felt they needed to be able to generate such documents, or
at least that their scribes and students needed to know about them, and
probably both. The question is, why?
The answer to this question lies in looking at all of the formula
manuscripts, not just at BnF lat. 2123, as well as at the other kinds of
evidence for early medieval documentary culture that is discussed else-
where in this volume. Here I will make some preliminary suggestions. All
of the extant manuscripts containing collections of formulas date from
the eighth century through the early tenth; the overwhelming majority of
them come from the ninth century.131 Some of the formula collections
contained in these manuscripts are indeed older; Marculf wrote in the
late seventh century, while the content of the Angers collection strongly
suggests that at least the documents in it stem from the late sixth. Even
so, however, most of the ur-collections that have been reconstructed from
later manuscripts, or that exist only in single copies, date from after the
middle of the eighth century.132 The majority of these formula collec-
tions project an essentially similar picture of a common documentary
culture, shared and participated in by both clergy and laity, to what we
have seen here; moreover, they extend the picture to cover not only Fran-
cia but also Alemannia and Bavaria.133 They appear in a wide variety of
131 Rio, Legal Practice, 241–71; Brown, ‘Die Karolingischen Formelsammlungen’.

132 Rio has pointed out that all seven Marculf manuscripts date from the Carolingian
period and could thus be called ‘Carolingian versions’ of Marculf: Rio, Legal Practice,
59.
133 See the geographical range of the collections given in Rio, Legal Practice, 67–164.
manuscript contexts: in collections of canon law and secular law (that

is, Roman and/or Germanic law, Carolingian capitularies, etc.), school-
books and miscellanies and by themselves in their own codices. We can
directly trace the origins of only some; we can infer the possible origins of
others from their contents. Many of those whose origins we can trace or
at least surmise point to monasteries or cathedral churches. Some betray
in addition evidence of direct involvement by the Carolingian court.
The Carolingian world of the late eighth and early ninth centuries was
one in which secular authority and ecclesiastical authority are difficult if
not impossible to separate. It was a world in which very powerful peo-
ple wore both secular and ecclesiastical or monastic hats, and in which
churches and monasteries were being used as instruments of government
and power to an unprecedented degree. It was a world in which writing
itself was also being used as an instrument of government and power
to an unprecedented degree, and in which churches and monasteries, as
well as those overseeing the royal fisc, were being encouraged by the court
to make and keep written records of their properties and incomes. The
formula evidence suggests that, in this context, churches and especially
monasteries were, consciously or unconsciously, taking responsibility for
producing documents for everyone. They therefore needed to create and
to keep collections of models that covered the range of possible docu-
ments that people might need, and that their scribes might need to draw
on or study.
7 Archives, documents and landowners
in Carolingian Francia
Matthew Innes
We have just seen how in the Frankish world the needs of ecclesiastical institu-
tions generated a set of sources – formula-books – that give us an indirect picture
of the broader documentary culture in the period following the atrophy of late
antique institutions. In this chapter, we see directly how these same needs play
out with real documents and real archives. In particular, we see how church and
monastic archives filtered and reshaped the documentary record in response to
their own needs, warping our view of the broader documentary culture. Peering
through this filter we can see traces of the wider world of document use and how
some of these traces got into churches and monasteries in the first place. Clerical
and monastic documentary practices were not sui generis, but rather reflect
assumptions about and practices surrounding records and record-keeping in the
local societies of which they were a part.
Evidential horizons and their

historiographical repercussions
Histories of written law and legal documentation in the early medieval
world are inevitably framed in terms of the Roman inheritance to the
western Empire’s successor states.1 Assessing the nature of that inher-
itance is difficult, however, on account of the different profiles of the
evidence for the Roman and immediately post-Roman centuries on one
1 Law: P. Wormald, ‘Lex scripta and verbum regis: Legislation and Kingship from Euric
to Cnut’, in P. Sawyer and I. Wood (eds.), Early Medieval Kingship (Leeds, 1977), 95–
138. Charters: Classen, ‘Fortleben’; McKitterick, Carolingians, esp. 77–134; essays in
Settlement of Disputes, esp. Nelson; and, in the wider context of written law, P. Wormald,
The Making of English Law. I. Legislation and Its Limits (Oxford, 2001), 30–108, all
now supplemented by a range of regional studies; for a fresh appraisal, see Rio, Legal
Practice, esp. 9–27, and G. Koziol, The Politics of Memory and Identity in Carolingian
Royal Diplomas: The West Frankish Kingdom (840–987) (Turnhout, 2011), ch. 1. I am
grateful to the Lay Archives Working Group; to Peter Erhart, Karl Heidecker, Geoffrey
Koziol, Wendy Davies and Charles West for their comments; and to Jon Jarrett for work
undertaken as a research assistant paid by Birkbeck College and the Leverhulme Trust,
in which capacity he identified ‘lay’ documents embedded within cartularies and also
undertook analysis of surviving dispute-settlement (placita) texts.
152
Francia: archives, documents and landowners 153
hand, and the eighth to eleventh centuries on the other. In Francia, our
understanding of the documentary and legal culture of the fifth and sixth
centuries – and to a significant degree the seventh century also – rests
on normative sources such as law codes, with some contextualization
from hagiographical and historiographical narratives.2 It is only as we
progress towards the end of the seventh century that we begin to find
a steady trickle of documentary texts, mostly transmitted through later
copies, from which to work; the compilation and copying of collections of
model documents in formularies relates to a similar time horizon and in
important ways was a response to the same processes that determined the
vectors of documentary transmission.3 Once we move into the eighth and
ninth centuries, the new or refounded ecclesiastical institutions closely
associated with the ‘Carolingianization’ of the Frankish world, particu-
larly its eastern regions, offer a much fuller supply of documents, mostly
in the ‘books of traditions’ compiled by these newly founded and heavily
patronized churches; and from a handful of institutions we get a steadier
trickle of originals, particularly of royal diplomas.4
The changing nature of the documentation thus immediately sets up
major interpretative problems in handling change over time. For we view
the successive phases of Frankish documentary culture through differ-
ent sources accessed via contrasting historiographies. In dealing with
Merovingian documentary culture, we adopt historiographical strategies
and interpretative techniques familiar to scholars of Roman legal prac-
tice, nuancing an idealized normative picture against isolated and indirect
hints at a more varied practice in narrative and normative texts. Such a
historiography poses questions about the extent to which the agendas
2 E.g. I. N. Wood, ‘Dispute Settlement in Late Fifth- and Sixth-Century Gaul: Some
Problems’, in Settlement of Disputes, 7–22, and Wood, ‘Administration, Law and Culture
in Merovingian Gaul’, in Uses of Literacy, 63–81.
3 Surviving original documents (overwhelmingly transmitted from a single archive, St-
Denis, for reasons discussed below, 317–19, with a significant stream from the Car-
olingian period from St Gall, discussed below, 155–64) are in ChLA; D. Ganz and W.
Goffart, ‘Charters Earlier than 800 from French Collections’, Speculum, 65 (1990),
906–32. Royal documents are edited by T. Kölzer (MGH DD Mer.); see A. C. Murray,
‘The New MGH Edition of the Charters of the Merovingian Kings’, Journal of Medieval
Latin, 15 (2005), 246–78. There is no systematic collection or study of the late seventh-
and early eighth-century documents transmitted in cartulary copies. Formularies: above,
Chapter 5; Rio, Legal Practice.
4 Carolingian cartularies: below, Chapter 8. Transmission via high medieval cartularies
and compilations: below, Chapter 11. ChLA is now expanding its scope into the ninth
century; for original documents from the ninth and tenth centuries, below, 313–20,
with references to the ARTEM database. Even in the ninth century, transmission in the
original is dominated by three archives (St-Denis, St Gall and Cluny), with a handful
of others that preserve some important royal and founding documents as originals (e.g.
Hersfeld, Murbach).
154 Matthew Innes
of rulers and their agents to regulate society could be put into prac-
tice, but it does not allow the presentation of alternative visions of social
practice from those implicit in the normative discourses of the time. For
the historian of the Carolingian and post-Carolingian periods, the inter-
pretative disjuncture is more severe, with the appearance of a sizeable
body of evidence directly relating to legal practice on the ground. The
brute fact of the dominance of ecclesiastical archives in our evidence for
legal practice is difficult to avoid, and skews our perspective on prac-
tice; and the fact that we know so much more about practice in the
later period makes us all the more aware of the problems of interpreting
normative sources. In fact, it produces an interpretative paradox. Schol-
ars dealing with the immediately post-Roman period of Merovingian
Gaul, because they have little or no surviving practical documentation
but a range of hints at social practice, do not rub against concerns about
ecclesiastical monopolization on the documentary record and so feel
confident in drawing inferences about wider documentary cultures. In
contrast, students of the Carolingian world, acutely aware of the wealth
of traditions transmitted through ecclesiastical archives in copy and orig-
inal, and the overwhelming dominance of documents directly establish-
ing the interests of the Church in those traditions, feel apologetic or
defensive in attempting to recreate the legal and tenurial practices of
lay landowners.
This chapter deals directly with the issues raised by the changing pro-
file of the evidence and the interpretative problems these create. Explain-
ing changes in the nature of the sources, and the relationship of these
changes to social processes, is the first job of the historian. What follows
argues that the ecclesiastical archives of the Carolingian world not only
define our evidential horizons but also were themselves an important
product of cultural, legal and political change. These processes are elu-
cidated through a detailed case study of the biggest surviving collection
of original documents from the early Middle Ages, those from St Gall.
Because the St Gall material is preserved as originals with notations and
markings, it offers a priceless opportunity to see inside the archive of an
early medieval ecclesiastical institution and reconstruct the concerns and
techniques underpinning documentary classification and preservation;
it therefore sheds light on the parallel processes feeding into the pro-
duction of ninth-century ‘books of traditions’ elsewhere in the eastern
regions of the Carolingian world. At St Gall, it is clear that the institu-
tional archival practices which produce such a dramatic upsurge in doc-
umentary texts do not represent the totality of Carolingian documentary
culture: they need situating amidst a wider documentary culture, many of
whose features can be reconstructed. Moving beyond St Gall, its unique

transmission of originals and its particular political context, we can trace
hints of a similar variety of practices elsewhere; the estate of Perrecy in
Burgundy provides an example, for a region with a very different history,
topography and society.
Charters and their transmission at Carolingian St Gall

Archives – even ecclesiastical archives – are notoriously elusive in Car-
olingian sources. We have a handful of uses of the Latin archivum to refer
to the place where documents were stored, amidst a more varied and less
specific terminology (‘treasury’, ‘store’). The variety of words used sug-
gests that document storage may not have been defined by a dedicated
space with its own specialized rules. It certainly is striking that in their
voluminous writings the likes of Einhard and Hincmar nowhere describe
archives as a source of authority, nor did the practices of document
keeping define the identities of any of the many servants of the Carolin-
gian rulers, in marked contrast to the later Roman and early Byzantine
world of John Lydus.5 To search for a set of institutional spaces with
defined public profiles is probably fruitless; we would do better to focus
on archival practices – the pragmatics of how and when documents were
utilized after their initial production – rather than institutional history.
If we do so, can we read archival practices as a means of acquiring
physical and interpretative control of the documentary record, as some
recent theorists have suggested?6 The case of the monastery of St Gall in
modern Switzerland allows us to reconstruct the mechanisms whereby a
major church, as its patrimony reached a certain scale and it acquired a
certain degree of institutional self-consciousness, might develop a fixed
archival memory of a new kind. The evidence from St Gall also sug-
gests that, prior to reaching this ‘critical mass’, such an institution might
‘piggyback’ on pre-existing archival and documentary practices and man-
age its muniments in a manner familiar to the landowning classes around
it.
5 H. Fichtenau, ‘Archive der Karolingerzeit’, in Fichtenau, Beiträge zur Mediäevistik,

3 vols. (Stuttgart, 1986), III, 115–25; for Lydus, above, 14. I hope to discuss Carolingian
royal archives elsewhere. Recent arguments for the systematic central registration of the
contents of royal diplomata do not take on board the findings of previous scholarship:
cf. D. Bachrach, ‘The Exercise of Royal Power in Early Medieval Europe: The Case of
Otto the Great 936–973’, EME, 17 (2009), 389–419; B. and D. Bachrach, ‘Continuity
of Written Administration in the Late Carolingian East c. 887–911: The Royal Fisc’,
Frühmittelalterliche Studien, 42 (2008), 109–46.
6 See above, 13.
156 Matthew Innes
St Gall, famously, provides a potentially unique insight into Car-

olingian record-keeping, thanks to the preservation of its charters in
the original rather than via a cartulary.7 However, although the abbey
has had a continuous history since its refoundation c. 720, its archival
holdings – for all their transmission in the original – have not. In partic-
ular, the early modern period saw the disruption of the archive and the
removal of a significant part of its contents following the plundering of
the abbey in 1531 – a reminder that systematic post-medieval disruption
is a constant in the history of medieval archives. The centuries follow-
ing 1531 saw recurrent campaigns of restitution, a process which went
hand-in-hand with the editing of collections of documents. As a result,
although the last known group of surviving documents removed from St
Gall in 1531 was only returned to the abbey in 1946, the shape of the
surviving charter collection of over 700 original Carolingian documents
has been well known since the nineteenth century.8 While this trans-
mitted material seems to provide a relatively full picture of the abbey’s
property acquisitions, it is salutary to remember that we have no way
of knowing what was destroyed in 1531, or subsequently lost, nor how
the processes of dispersal and restitution altered the pattern of surviving
documentation.
For all the need to consider the possibility of gaps, the survival of
over 700 originals allows us unrivalled insight into a major Carolingian
archive, and the work of Peter Erhart and others has used this material to
elucidate St Gall’s archival practice.9 By the ninth century, an archiving
process seems to have been well established. Scribes often made exten-
sive notes (the so-called Vorakte) on the parchments, either prior to a
transaction when they were ‘commissioned’ or as a contemporaneous
note at a public meeting; these then informed the working up of a formal
7 The St Gall charters are published as Wartmann I–VI; those prior to 800 have been
edited, with facsimile, in ChLA I–II; the ninth-century material is in ChLA C–CIII. See
also P. Erhart and L. Hollenstein (eds.), Mensch und Schrift im frühen Mittelalter (St Gall,
2006).
8 P. Erhart, ‘Dem Gedächtnis auf der Spur: Das frühmittelalterliche Archiv des Klosters
St. Gallen’, in Erhart and Hollenstein (eds.), Mensch und Schrift, 59–66, with references;
P. Erhart and J. Kleindinst, Urkundenlandscaft Rätien (Vienna, 2004), 24–6.
9 Erhart, ‘Dem Gedächtnis auf der Spur’; Erhart, ‘Carta ista amalfitana est et nescitur legere:
The Charters of Cava dei Tirreni and St Gall and Their Evidence for Early Medieval
Archival Practice’, Gazette du livre médiéval, 50 (2007), 27–39; K. Heidecker, ‘Les actes
privés de la période carolingienne dans les archives de Saint-Gall’, Annuaire de l’École
pratique des hautes études (EPHE), Section des sciences historiques et philologiques, 139 (2008):
ashp.revues.org/index570.html.
charter text.10 The charter was then brought to St Gall, where the indi-
vidual parchment was folded into a long, finger-like, packet, and a note
on the contents (normally the name of the donor and land) was made on
the outside. Sometimes this note was written by the actual charter scribe
(who when he was a monk seems to have been responsible for archiving
his own documents), but sometimes in another hand (presumably that
of the monk then responsible for the charter collection). The physical
location in which the documents were then stored remains obscure – it
is not indicated in the ideal monastery on the so-called Plan of St Gall –
but we do know that charter scribes doubled as book scribes and worked
in close proximity to the scriptorium. In all likelihood, the scriptorium or
the sacristy doubled as the storage place for documents; by the eleventh
century the armarium, close to the treasury (thesaurus), served as the
charter store.11
Archiving practice – in particular, the annotation of documents – seems
to have been motivated by immediate concerns from its very beginnings.
The earliest dorsal notices began c. 770 and were made by the monk
and later abbot, Waldo; the practice of making annotations for reference
was subsequently adopted by other monastic scribes and became general
by the ninth century. Waldo’s moves were in all probability a response
to urgent need; he referenced not only documents he wrote himself but
also older muniments, including several of the earliest surviving St Gall
documents.12 These documents predated the Carolingian takeover and
recorded the gifts to the abbey made by local Alemannian landowners
prior to the battle of Cannstadt in 746. They had contemporary signif-
icance precisely because so many of these gifts had been subsequently
seized by Ruthard and Warin, the Frankish aristocrats who ruled Ale-
mannia with an iron rod in the middle decades of the eighth century.
Not only was the establishment of Carolingian rule through Warin and
10 E.g. McKitterick, Carolingians, 94–8, and Heideicker, ‘Les actes privés’. It seems to me
rather odd that some (most famously, M. Richter, ‘“Quisquis scit scibere, nullam potat
abere labore”: Zur Laienschriftlichkeit im 8. Jht’, in J. Jarnut and U. Nonn [eds.], Karl
Martell in seiner Zeit [Sigmaringen, 1994], 393–404), have seen the Vorakte as indicating
a disjuncture between the legal moment of disposition and the written record.
11 On the elusiveness of archival space, see Fichtenau, ‘Archive der Karolingerzeit’, 117–
19; for St Gall, Erhart and Kleindinst, Urkundenlandschaft Rätien, 22–3, citing Ekkehard
IV, Casus sancti Galli 112 (ed. H. F. Haefele [Darmstadt, 2002], 220); also S. Barret,
La mémoire et l’écrit: L’abbaye de Cluny et ses archives (Xe –XVIIIe siècle) (Münster, 2004),
91–6.
12 The dorsal notices were edited by O. Clavadetscher and P. Stärkle, Die Dorsualnotizen
der älteren St Galler Urkunden (St Gall, 1970); they are distinct from the Vorakte; a good
example of the early annotation of significant donations to Otmar comes in ChLA I 40,
44 = Wartmann I 8, 9.
158 Matthew Innes
Ruthard effected by systematic confiscations and expropriations whose

documentary footprint structures the record for the region through the
ninth century; it also involved the subjection of St Gall to the bishop
of Constance, and the trial, deposition, exile and subsequent death of
Abbot Otmar (717–59).13
St Gall tradition not only remembered the misfortunes visited by
Warin and Ruthard; it also fostered active memories of lost lands that
informed long-running attempts at reclamation. In the house histories in
which these memories found written reworking, these carefully nursed
grievances were often articulated through stories about documents. St
Gall’s subjection to Constance, for example, was presented as an act of
arbitrary seizure encapsulated in a complex tale about the removal of
charters and their subsequent incarceration in the treasury (scrinium)
of Constance cathedral. St Gall’s claims here need taking with a pinch
of salt. With the benefit of hindsight, the monks claimed that since the
time of Pippin they had held privileges that were in fact identical to those
systematically granted to royal monasteries in the reign of the current
Carolingian, Louis the Pious; indeed, they were successful in mobiliz-
ing this version of the past to procure new privileges from Louis. But
documents mattered both as symbols of identity – the theft of royal privi-
leges by the bishop of Constance was almost a ritual demonstration of St
Gall’s humiliating subjection – and practically in eliciting Louis’ grant of
royal protection and associated rights.14 Documents such as those Waldo
indexed were likewise crucial points of memory, for the lands lost under
Warin and Ruthard were the subject of a St Gall campaign for restora-
tion spanning a century and more, a campaign that resulted in a slow but
steady trickle of restitutions from the very last years of the eighth century
through to the end of the ninth. As the house histories make clear, issues
of identity loomed large in the memories of such estates, their donors
and their complex histories. But studying and referencing the abbey’s
documentary resources was also, simultaneously, a practical action with
instrumental intentions, not least as confiscated estates might have been
subsequently transferred by charter themselves, creating a rival parch-
ment trail to St Gall’s. A remarkable pair of royal charters from 790, for
example, show Charlemagne dealing with the case of Alemannian estates
13 For confiscation, expropriation and Carolingian rule in Alemannia, see M. Innes, ‘Prop-
erty, Politics and the Problem of the Carolingian State’, in W. Pohl and V. Wieser
(eds.), Das frühmittelalterlichen Staat: Europäische Perspektiven (Vienna, 2009), 299–
314.
14 Ratpert, Casus sancti Galli, esp. cc. 2–3 (MGH SRG 75:151–65); P. Depreux, ‘La
plainte des moines de Saint-Gall auprès de l’empereur Louis le Pieux (815)’, Zeitschrift
für schweizerische Kirchengeschichte, 94 (2000), 7–16.
that had been confiscated in the aftermath of the Carolingian conquest,

and then passed on by Ruthard to private landowners by charter of sale.
In both cases, Charlemagne overturned these (now lost) deeds on the
grounds that the initial confiscation had been unjust, and returned the
land to the abbeys of St-Denis and St Martin at Tours.15 It would there-
fore be a mistake to take the relatively laconic placitum tradition from this
region – which tends not to give detailed accounts of argumentation in
court, simply listing witnesses without normally mentioning documents –
at face value, and argue that these documents were not used in court:
their referencing by St Gall scribes implies otherwise.16
The notices show not only continual reference and use but also at
least two systematic efforts at archival organization. The first, which saw
charters annotated with names of donors and the location of the land, was
completed by 815. This may be linked to the compilation of a listing of
the monastery’s monks, headed by their saintly patron, Otmar, c. 802: the
same hand is evident at work in both projects.17 This would confirm the
commemorative aspect of such archival activity, evident also in cartulary
compilation elsewhere. But this codification of institutional memory was
also linked to fundamental changes in the governance of Alemannia and
its relationship to the royal court, which made the cult of Otmar, and
complaints about the actions of Ruthard, Warin and the first generation
of Carolingian agents, politically live issues.18
A further systematic reorganization c. 840 underlines the relationship
between internal archival organization and external social and political
changes. This second campaign of reorganization involved the reordering
and numbering of the charters, region by region, with each region given –
in true Carolingian fashion – its own capitulum, with the relevant docu-
ments probably stored together in a single drawer. Simultaneously, royal
diplomas were differentiated from private charters, creating a contrast
which Notker was to play on in appealing to Charles the Fat for further
15 MGH DD Kar. 1, nos. 166–7. Charlemagne’s simultaneous resolution of both cases

probably relates to a wider moment in Alemannia’s history, and the royal charters likely
obscure local compromises: Innes, ‘Property, Politics’, 307–8.
16 Cf. T. Reuter, ‘Property Transactions and Social Relationships Between Rulers, Bishops
and Nobles in Early Eleventh-Century Saxony: The Evidence of the Vita Meinwerki’,
in W. Davies and P. Fouracre (eds.), Property and Power in the Early Middle Ages (Cam-
bridge, 1995), 165–8, at 171–2; or for the classic denial of any dispositive or probative
value for this material, Richter, ‘“Quisquis sit scribere”’.
17 Erhart, ‘Dem Gedächtnis auf der Spur’; Erhart, ‘Carta ista’; Heidecker, ‘Les actes
privés’, with references to the foundational work of Bruckner and Stärkle.
18 Innes, ‘Property, Politics’, with M. Borgolte, Geschichte der Grafschaften Alemanniens in
karolingischer Zeit (Sigmaringen, 1983) for administrative change as reflected in scribal
formulas.
160 Matthew Innes
privileges.19 This distinction between royal praecepta and cartae pagenses

dates back at least as far as Marculf’s formula-book in the late seventh
century, but as a principle of archival organization it is characteristic
of Carolingian charter compilations in East Francia, the prototypes for
later cartularies right across the post-Carolingian world; they normally
list royal documents separately at the beginning, and private charters by
geographical region.20 The St Gall evidence perhaps suggests that these
principles of organization reflect an archival system originating in the
early ninth century. At St Gall, these archival practices can probably also
be linked to the final establishment of a regular network of counties in
Alemannia – evident from changing charter formulas, and reflected in the
regional organization of the material – and perhaps to the contestation of
political power in the area.21
Because St Gall’s campaigns of archival organization took place
through the physical reordering of original documents, rather than the
compilation of a ‘book of traditions’ of copied texts, they can be viewed
in an unusual degree of detail. The St Gall experience, indeed, raises
issues about the impulses underpinning the copying of collections of
deeds in other major churches east of the Rhine. At Freising, where we
have the original manuscript of the charter compilation made by the
priest Cozroh, we can perhaps sense the practicalities of archival classi-
fication interacting with twenty-plus years spent copying deeds into the
cartulary.22 Cozroh’s codex consists of three parts bound together, the
first two written contemporaneously with each other by Cozroh him-
self, and the third compiled c. 848 under his supervision. The first con-
tains the documents of the four bishops of the Agilolfing period and
follows neither a chronological nor topographical order. It does con-
clude by emphasizing the records of judicial proceedings in the episco-
pate of Atto (783–811), in which the titles of gifts received prior to the
Carolingian takeover were publicly affirmed in court and synod before the
agents of the new regime. The second part consists of both documents
of gift and sale, and records of litigation, written under Bishop Hitto
19 Notker, Gesta Karoli 2.10 (MGH SRG n.s. 12:66–7); for his success in so winning
privileges from Charles the Fat, see S. Maclean, Kingship and Politics in the Late Ninth
Century: Charles the Fat and the End of the Carolingian Empire (Cambridge, 2004), 201–2.
20 Marculf’s formula-book is divided into two books on these lines; see Rio, Formularies.
It is a moot point whether Marculf’s distinction was based on an archival scheme.
21 Borgolte, Grafschaften Alemanniens, for this context.
22 TF. The Cozroh codex is now available in a magnificent interactive facsim-
ile with hyperlinks to Bitterauf ’s edition and Adelheid Krah’s notes at www.
bayerische-landesbibliothek-online.de/cozroh. On its compilation, see A. Krah, ‘Die
Handschrift des Cozroh: Einblicke in die kopiale Überlieferung der verlorenen ältesten
Archivbestände des Hochstifts Freising’, Archivalische Zeitschrift, 89 (2007), 407–31.
(811–35), at whose behest Cozroh made the cartulary; these are orga-
nized by imperial year. The third records transactions made under the
next bishop, Erchanbert (835–58), up to the year 848. While the order-
ing of the documents exemplifies the commemorative concerns spelled
out in Cozroh’s preface, the codex is also an eminently practical artefact;
witness the register of property acquisitions under Hitto with which the
first part ends and the systematic use of headings and reference marks
through the text. This is no surprise, given Cozroh’s own extensive work
as a scribe and his responsibility for charter production under Hitto. His
role most likely extended to include acting as what we would term the
‘archivist’ for the charter collection and presumably involved assembling
written proofs for the recurrent litigation evident through the codex. As
Adelheid Krah has shown, there is clearly a complex set of interactions
at play between Cozroh’s organizational scheme in the cartulary and the
classification and physical archiving of the original charters in the Freis-
ing cathedral, which was perhaps inspired by Cozroh’s own experience
of visiting Aachen, noted in the cartulary codex.23
Cozroh sought to preserve the memory of benefactors and bishops
and order their traditions so as to establish Freising’s title to its estates,
but he was also aware that the archive from which he worked was far
from static. His copies, indeed, were in part motivated by a recognition
that many documents had fallen into oblivion, whether wilfully destroyed
through malice or lost through carelessness, and a fear that more might
be removed in future by those who wished to contest their contents.24
These fears underline that, as at St Gall so at Freising, the Carolingian
takeover fundamentally altered the distribution of property. The Church
was left dependent on claiming royal patronage to safeguard its interests,
using written proofs to mobilize official support and popular opinion in
local assemblies and setting off a cycle of land litigation that reverberated
right through the first half of the ninth century.25 The Cozroh codex,
still more than the dorsal notes made by Waldo and his successors at
St Gall, encourages us to see an archivist marshalling his charters and
awaiting his moment to assert title to this or that property as the political
configuration of both the empire and the local landowning community
changed. While we cannot see quite so clearly these concerns feeding
23 Krah, ‘Die Cozroh Handschrift’.

24 TF, pp. 1–2. Cozroh as an author: P. Geary, ‘Medieval Archivists as Authors: Archival
Memory and Social Memory’, in F. Blouin and W. Rosenberg (eds.), Archives, Docu-
mentation and Institutions of Social Memory: Essays from the Sawyer Seminar (Ann Arbor,
MI, 2006), 106–13.
25 Impact of Carolingian takeover: W. Brown, Unjust Seizure: Conflict, Interest and Authority
in an Early Medieval Society (Ithaca, NY, 2001).
162 Matthew Innes
through into the other East Frankish cartularies, we can be clear that
what united them was the tenurial and political impact of Carolingian
conquest, and its interaction with the rapidly growing landholdings of
the newly institutionalized Church in the region.26
There can be little doubt that the kind of activity we can see at St Gall or
at Cozroh’s Freising reflects an archival consciousness, and a concern to
structure documentary holdings so as to facilitate their use. But the ‘com-
pact and efficient storage system’27 created at St Gall was only viable in
an archive of a thousand or so documents when related to known estates
managed by the monks and tied to the commemoration of patrons; one
had to know in advance where to look. This was not reference literacy
of a modern kind, but the structuring of an archive of a particular size
according to a topographical and geographical mnemonic: those con-
cerned with a particular estate would remember the carefully cultivated
traditions about its history, and look in the appropriate drawer for the
appropriate region to find the documentary record. It is this style of doc-
ument keeping – in the context of a living tradition of estates and their
erstwhile owners – that should form the baseline against which we assess
Carolingian archives. It relied on a living memory of the monastery’s
episodic relationships with its patrons and neighbours.
The surviving originals overwhelmingly consist of property acquisi-
tions by St Gall. Nonetheless, even a cursory look underlines the fact
that the monastery’s archive was not hermetically sealed or separate
from the lay society around it. A handful of documents record trans-
actions in which the monastery was not directly implicated, detailing the
prehistory of property later acquired by St Gall.28 The run of land acqui-
sitions immediately prior to Cannstadt in 746, for example, includes
one now-mutilated charter in which Daghilinda sold property at Geb-
hardswil for 30 solidi. A gap in the parchment means that the identity
of the buyer is now unknowable, but the document is preserved because
the scribe Odo wrote it on the same parchment as the gift of property,
also at Gerbhardswil, made by Gauzoin to Abbot Otmar; it was anno-
tated by a ninth-century St Gall hand as ‘Agilind’s charter concerning
Gebhardswil’.29 Similarly, a fascinating charter from the 850s involves
26 E.g. for Salzburg, H. Wolfram, ‘Die Notitia Arnonis und ähnliche Formen der
Rechtssicherung im nachagilolfischen Bayern’, in Recht und Schrift, 115–30; for Wis-
sembourg, see Hummer, Politics; for Fulda, below, Chapter 8.
27 Erhart, ‘Carta ista’.
28 Erhart and Kleindinst, Urkundenlandschaft Rätien, 21–2; they cite Wartmann I 70 (a.
772), a sale to a local church dedicated to St Gall, which the monks may have misread
as a gift to the monastery.
29 ChLA I 40, 44 = Wartmann I 8, 9. This was filed with gifts and purchases by St Gall
by the ninth century, with the dorsal note ‘car[t]a Agilind de Gaebaratuilari’; if Otmar
Balfred and his wife Evalia selling land to Wacharen and his daughter
Odalswind, with the proviso that Odalswind should not sell the land to
anyone, Roman or Aleman, other than to Priectus or his offspring, or if
she wished to make a gift for her soul to St Gall. Although we have no fur-
ther documentation concerning Odalswind, a dorsal notation in a St Gall
hand from the ninth century labels this the ‘charter of Priectus’, perhaps
indicating that this document was passed on as a title-deed with a later
donation.30 The dorsal notices provide further priceless evidence for the
interaction of the monastic archive with a wider documentary culture
among the landowning classes. They demonstrate, for example, the exis-
tence of pairs of charters, one within and one without the monastery, and
they also record the removal of charters from the monastery to a secular
household where they might be consumed by fire along with other items
of an individual’s possessions.31 We know that the granting of life inter-
ests (precaria) to donors and their descendants was widespread at St Gall
as elsewhere in this period, and that such transactions must have gener-
ated a ready flow of parchments between the monks and their patrons.
It is also clear that – other than where a donor reserves a life interest
in an initial donation charter – such transactions are underrepresented
in the surviving charter collection, presumably as written acts recording
precaria had little value once they had expired.32 Further indirect evi-
dence for lay documentary practices has been assembled in Rosamond
was the beneficiary, as has been suggested, this would make sense, but there is no real
reason to suppose so other than the charter’s ending up at St Gall, and it is equally
likely that the charter was filed simply because it was attached to a genuine donation
at the same place on the same day, and later interpreted by the St Gall archivist as a
further acquisition. Note that, bizarrely, the Gauzoin document was also noted as ‘carta
Aegilinda’. Did the pair of acts have Dagalinda selling land to Gauzoin for him to gift
immediately to Otmar? The extreme rarity of sales to St Gall may be relevant here.
30 Erhart and Kleindinst, Urkundenlandschaft Rätien, no. 42. The crucial clause is slightly
garbled: ‘ipse pretius fuit Odolsenda; in tale vero rationem, quod <si> ipsa iure ven-
didere voluerit, non habeant licentiam nec ad Romanos nec ad Alaemannos, set Priecto
vel ad suos infantes; exceptum si pro remedium animae sue voluerint donare ad monas-
terium sancti Galli licentiam habeant’. I would interpret this as limiting any further
transfer of this property other than to Priectus or his offspring, unless Odalswind wished
to make a gift to St Gall. In the event, it looks as if the land came to Priectus, who then
made a gift to St Gall, with this charter passing on as a title-deed, and doubtless would
be of value if the process whereby the land came to St Gall were questioned.
31 See Erhart, ‘Carta ista’, citing Clavadetscher and Stärkle (eds.), Die Dorsualnotizen, nos.
609, 610, 626 (alluding to paired documents in exchanges, one for the monastery, one
for its lay partner), 783 (‘priorem cartam illorum cum ceteris rebus suis ignis consump-
sit’). Cf. below, Chapter 8, for other East Frankish examples; Brown, ‘Documents’, for
destruction and replacement; and for the Freising and Marseille evidence, above, n. 24,
and below, n. 64.
32 P. Depreux, ‘L’apparition de la précaire à Saint-Gall’ and L. Morelle, ‘Les “actes de
précaire”, instruments de transferts patrimoniaux (France du nord et de l’est, VIIIe –XIe
siècle), both in Les transferts, 649–73, 607–47; Hummer, Politics, esp. 76–129.
164 Matthew Innes
McKitterick’s pathbreaking work, which demonstrated the existence of

an infrastructure of local scribes, some lay and some priests, supporting
a documentary culture among the regional landowning class prior to the
foundation of St Gall.33
Given this evidence for St Gall’s involvement in a wider documen-
tary culture, it would be a mistake to see the relative paucity of surviv-
ing lay documents as indicative of original patterns of production and
use, rather than the result of the complexities of transmission. Recent
research on cartularies has emphasized that they were not copies of an
archive’s entire contents at a particular moment, so much as a carefully
organized selection of documents designed to commemorate benefac-
tors and establish title, which therefore potentially excluded significant
tranches of documents.34 While St Gall had no cartulary, the ninth-
century reorganization of the charters into neatly labelled parcels each
allocated to the correct capitulum may have left little space for documents
that now had limited utility. For example, one late ninth-century charter
that documents a group of over a dozen landowners giving their shares
of a mountain to the church of St Saviour at Roncalam only survives
because it was used in the thirteenth century for jottings relating to the
monastery’s accounts.35 In any case, any secular documents held by the
abbey through the Middle Ages are less likely to have been transmitted
across the disruptions of the early modern period, not least as after 1531
the abbey had little reason to seek their restoration. Such a conclusion
would be supported by the contents of several legal compilations and
formula collections made at St Gall, many of which outline documents
for which there are no obvious analogues in the surviving charters, and
imply the existence of now lost secular documentation. One such doc-
ument template, for example, records the reconciliation of Otulf and
Undolf following a quarrel over a wood which broke into open conflict
and was settled by a gift of property. Even if we see the case as fictitious, it
still has important implications: that such events needed a documentary
record, and that St Gall scribes might provide a legal template for such
an eventuality, without leaving a trace in the monastery’s archive.36
33 McKitterick, Carolingians, 77–134; and now K. Heidecker, ‘Urkundenschreibern im

alemannischen Umfeld des Klosters St. Gallen’, in Erhart, Heidecker and Zeller (eds.),
Die Privaturkunden der Karolingerzeit, 183–91.
34 See further below, Chapters 8 and 11.
35 Wartmann III A10 = Erhart and Kleindinst, Urkundenlandschaft Rätien, no. 54. Like the
other Rhaetian material, this was excluded from the capitulum system – perhaps material
not filed in this way was particularly vulnerable to loss or reuse.
36 The famous formula-book of Salomon (Coll. Sang.) needs treating carefully on account
of its ideological motivation; more practical in orientation is the material edited as
A locality, its documents and their survival:

Rankweil and St Gall
For one locale, it is possible to move beyond these hints of a wider doc-
umentary culture and gain direct insight into documentary practices,
thanks to a run of fifty-three surviving documents from Rhaetia, the
overwhelming majority of which are concerned with Rankweil and sur-
rounding villages high on one of the main passes over the Alps. Of these
fifty-three ninth- and tenth-century Rhaetian documents transmitted via
the St Gall archive, the abbey is involved in only seven.37 This collection
itself provides clear evidence for the post-Carolingian loss of documents
from the St Gall archive, for several of the Rankweil documents came to
light in early modern book-bindings, some at St Gall but most recently
the fifty-third document at Zurich.38 Once again, the dorsal notices sup-
ply priceless evidence as to the archival and transmission history of the
Rhaetian material. In a handful of Rhaetian documents gifting land to
St Gall, the dorsal notices indicate that the normal processes of entry
into the monastery archive were followed, and even though Rhaetia was
excluded from the capitulum system (the monastery’s holdings there were
scarce), some were entered into another ‘chapter’ on grounds that remain
obscure.39 The bulk of the Rankweil documents, however, came to St
Gall not via this standard process of archival entry but as a group at a
far later date than their redaction. The dorsal notices on the Rhaetian
material from the 800s, 810s and 820s – in which St Gall was not a direct
actor – were made in the first instance by the local scribes who wrote the
documents. Around half of the Rhaetian material dates from the later
part of the ninth century, but a significant number of documents have
dorsal notation which does not conform to St Gall practice. Quite when
these charters came to St Gall remains unclear: there is a rich seam of
Rhaetian charters in which St Gall is not implicated through the 880s and
Form. Sang. Misc., which is actually drawn from four distinct ninth- and tenth-century
manuscripts. Otulf and Unolf are Form. Sang. Misc. 5.
37 See now for analysis, texts and facsimiles in Erhart and Kleindinst, Urkundenlandschaft
Rätien, with P. Erhart, ‘Erratische Blöcke am Alpennordrand? Die rätischen Urkunden
und ihre Überlieferung’, in Erhart, Heidecker and Zeller (eds.), Die Privaturkunden der
Karolingerzeit, 161–72, and the classic H. Fichtenau, Das Urkundenwesen im Osterreich
vom 8. bis zum frühen 13. Jahrhundert (Vienna, 1971), 38–53.
38 ChLA CI 178 = Erhart and Kleindinst, Urkundenlandschaft Rätien, no. 26.
39 The following documents bear traces of normal archiving as evident in the other St Gall
documents: Erhart and Kleindinst, Urkundenlandschaft Rätien, nos. 1 and 2 (as a pair),
40, 42. No. 39 is a donation to St Gall, made at Gams, which lacks dorsal notices or
other archiving marks; similarly no. 46, made at Rankweil; no. 53, a swap; and no. 57,
a gift; nos. 45 and 51 both have dorsal notices that cannot be definitively linked with St
Gall. See further below, 168–71.
166 Matthew Innes
890s, with further surviving documents from 920, three from 933 and a
final document from 975.40 What is clear is that at some point, almost
certainly in the tenth century, these charters were physically moved from
Rankweil – an important judicial and legal centre – to St Gall. Here they
remained through the Middle Ages, within the monastery’s archive but
never integrated into the capitulum system, an anomalous group separate
from the main classification system documenting St Gall’s holdings.41
For modern scholars, these documents are associated above all with
the career of Folkwine, a minor official active c. 820; twenty-seven of the
surviving charters document his wheeling and dealing between 817 and
825.42 Only two documents predate Folkwine’s, a pair of placita from the
first years of the ninth century, when the hold of the prince-bishop of
Chur was weakened by Charlemagne and a Frankish count dispatched
to wield secular power in Rhaetia.43 Folkwine was a follower of the new
Frankish counts and in all probability an outsider, appointed to act as
a local official; in his first surviving charter, he is styled escultaizo, a
Latinization of the Germanic term Schultheiss, a generic vernacular label
for a subcomital official, comparable to the catch-all terms centenarius or
vicarius of the normative sources. He was based at Rankweil, a strate-
gically crucial point on the Alpine pass, and one where there was an
important royal villa that doubled as a base for the new counts.44 Refer-
ences abound in his documents to his ‘good offices’, and to the network
of favours and obligations he was able to build up in the local community
as he fulfilled his fiscal and judicial roles. Folkwine’s personal acquisition
of a sizeable portfolio of landed property – he is the buyer or recipient
of land in all twenty-seven surviving documents in which he features –
thus cannot be divorced from his public role, which in any case proba-
bly supplied the wherewithal for him to make his purchases. Given the
40 Fichtenau suggests a possible political context in the tenth century when the county of
Rhaetia was combined with the duchy of Swabia; any hypothesis would need to account
for the five surviving tenth-century documents from Rhaetia finding their way to St
Gall. See below, 168–71.
41 The exclusion of the Rankweil material from the capitulum system does raise the question
of whether only documents recording monastic acquisitions were so classified, with
material less directly related to the monastery’s interests excluded.
42 In addition to Erhart and Kleindinst, Urkundenlandschaft Rätien, esp. 83–90, and Fichte-
nau, Urkundenwesen, 38–53, see K. Bullimore, ‘Folqwine of Rankweil: The Local World
of a Carolingian Official’, EME, 13 (2005), 3–17, and Innes, ‘Practices of Property’,
256–62.
43 Erhart and Kleindinst, Urkundenlandschaft Rätien, nos. 9 and 10; for the politics, Bor-
golte, Grafschaften Alemanniens, 219–29.
44 See the inventory of royal rights made in 842×3: Bündner Urkundenbuch, vol. I, ed. F.
Perret and E. Meyer-Marthaler (Chur, 1955), 375–96; Lothar issued a royal charter in
Count Hunfrid’s villa at Rankweil (no. 44).
political context of Folkwine’s activities – the evident tensions caused

by the presence of his masters and his arrival in a tight-knit mountain
community, and the campaign of appeal and complaint to successive
Carolingians waged by the bishops of Chur against the new disposition
imposed by Charlemagne – the careful preservation of these documents
was a necessity. After all, there was a real possibility of the investigation
and overturning of the new order, and at least some of Folkwine’s transac-
tions were potentially contestable.45 Continued interest in the Folkwine
charters is indicated by the addition of a further layer of dorsal notices to
this material, most made in a single hand of the second half of the ninth
century, which undertook the further classification of the carta Folquini.
As this is not a known St Gall hand, the classification must have taken
place prior to the monastery’s acquisition of these documents.46
What was the agency for the transmission and continued scrutiny of
Folkwine’s dossier? The key lies in the scribes who sustained the doc-
umentary culture that produced it. The texts of Folkwine’s documents
show a priest, Andreas, bearing official responsibility for their drafting,
while occasionally also noting that another clerical scribe – Valerius or
Vigilius – actually wrote the charter. A group of four charters from Schlins
are drafted under another priest, Bauco. Palaeographical analysis indi-
cates a variety of hands even in those charters where Andreas alone is
named as scribe, a phenomenon that has been paralleled in other contexts
where a significant body of originals survives. Andreas’ role therefore
has a public element, as he was an individual bearing responsibility for
the documents drafted on his authority, but with a team of other scribes
working with him: an important reminder that ultimately charters needed
to be publicly trusted.47 In fact, a series of priest-scribes is responsible
for more or less the totality of the Rankweil documentation. Indeed,
they appear as actors, neighbours and witnesses throughout it, in some
cases perpetuating priestly and scribal dynasties. Andreas and company
are almost certainly to be seen as priests serving the church associated
45 We have no record of any appeal against Folkwine, nor do we know what happened to
his land; he is commemorated in the St Gall necrology, suggesting some link with the
monks, but the political context, regional and regnal, was fraught, and we know that
when an aristocrat fell from grace in this period his property acquisitions and those of
his subordinates might well be subjected to scrutiny, as in the case of Matfrid of Orleans.
Note also that in the context of the civil strife of the 830s the bishop of Chur was able
to move on long-standing grievances and reclaim land seized by the counts of Rhaetia.
46 Erhart and Kleindinst, Urkundenlandschaft Rätien, 26–32, drawing on Fichtenau, Urkun-
denwesen, 45.
47 Erhart and Kleindinst, Urkundenlandschaft Rätien, esp. 54–70; P. Erhart, ‘Der Rotulus
des Valerius: Das Schreiben von Urkunden im frühmittelalterlichen Rätien’, Geschichte
und Region / Storia e regione, 15 (2006), 38–61; for the Cluny comparison, see
below, Chapter 11.
168 Matthew Innes
with the comital residence at Rankweil, an important administrative and

judicial centre.48 Tied into the community of local landowners whom
they served as priests and scribes, they were allied with and writing for
the rulers of the region, not least Folkwine, the local face of a newly
intrusive Frankish power. It is their activities that connect the cluster of
Folkwine-related documents with the steady trickle of local documen-
tation through the remaining decades of the ninth century and into the
tenth. The Rankweil documents, therefore, are best seen as the remnants
of a local set of records kept at the church in Rankweil, and connected
with the work of the priests as scribes closely tied to the royal officials
and the royal estate and villa. The first set of documents, from the earlier
part of the ninth century, reflects a need to document contested and con-
testable activities of the earliest Frankish officials in a context of repeated
challenges led by the bishop of Chur. Thereafter, we have priests main-
taining their own documents, and holding records of some charters they
produced for local landowners, and indeed for other local churches such
as St Saviour ad Roncalem, of which the priest-scribes were important
patrons.
Looking at the interests of the priest-scribes, it may indeed be possible
to identify the mechanisms by which their charter collection came to St
Gall. The monastery had little property in the area before the middle
decades of the ninth century: just the one gift of land at Gebhardswil,
from 745, survives prior to this date.49 By the middle decades of the
ninth century, however, parts of the community around Rankweil were
becoming implicated in St Gall’s patronage network. In 835, Berengar
and his wife Imma made a sizeable gift to the monks for the health of
their souls, with the right to buy back against a lump sum and annual
payment.50 In 844 or 851, Job and his wife Andustria made a gift to
the monastery, where their son Drucio had evidently been oblated; if he
became a monk or left the monastery of his own volition, the gift was
to stand, but if the monks forced him to leave it was to fall.51 In 851 or
858, Balfred and his wife Evalia sold land to Wacharen and his daughter
Odalswind, who had permission to resell the land only to Priectus, or
for the good of her soul to St Gall; a dorsal note recording the archiving
of the document at St Gall indicates that the land had been passed on
48 Erhart and Kleindinst, Urkundenlandschaft Rätien, 114–15; the 842×3 inventory lists
three royal churches, one of them attached to the curtis dominica, which was the comital
residence.
49 Erhart and Kleindinst, Urkundenlandschaft Rätien, nos. 1–2, on which see above, 162.
50 Erhart and Kleindinst, Urkundenlandschaft Rätien, no. 39.
51 Erhart and Kleindinst, Urkundenlandschaft Rätien, no. 40.
to St Gall by Priectus.52 The gift connected with Drucio’s oblation was

similarly entered into the St Gall archive with the normal dorsal note,
and an earlier non-St Gall dorsal note was erased.53
The next half century saw four further acquisitions by St Gall: in 864
Valerius made a sale on behalf of Magna to the advocate Onoratus and
his maior Abraham, both apparently acting for St Gall; in 881 Abraham
gave his inherited land to the monks for the good of his and his ancestors’
souls; in 884 Hisuanus and his son Isuano travelled to St Gall to make
a gift for the soul of another son, Nordolo; in 893 the priest Valerius
swapped land with the monks in a charter redacted at Rankweil, giving
in return his church of St Victor.54 None of these transactions bears
any sign of St Gall archiving protocols: only the 864 sale and the 884
gift bear any dorsal notices and these read, respectively, ‘purchase [by] St
Gall from Valerius’ and ‘charter of Isuano to the benefit of St Gall’.55 The
explicit mention of St Gall as beneficiary in both cases is striking, because
dorsal notices associated with the monastery archive rarely felt the need
to specify the institution’s interest; it was taken as read. While the script
of both notices is ninth-century, neither has yet been identified with any
known St Gall hand. It therefore appears likely that in both cases the
dorsal notices were made not at St Gall, but at Rankweil: other Rankweil
charters in which St Gall was not implicated have similar notices, while
the charter associated with Drucio’s oblation likewise had a non-St Gall
dorsal notice erased when the charter was entered into the monastery
archive. The explicit noting of St Gall as beneficiary would make more
sense in an extra-monastic context. Valerius’ gift of 893 is particularly
intriguing, for Valerius was a frequent charter scribe and his family and
contacts dominate the surviving charters, and in return for land he gave
St Gall the proprietary church of St Victor in which he was a priest: was
this where the charters were kept?
The startling conclusion, then, is that of the seven ninth-century char-
ters from the Rankweil area in which St Gall directly or indirectly ben-
efited, only two can be definitely shown to have been entered into the
St Gall archive in the ninth century. Moreover, in both of these cases
(Drucio’s oblation and Priectus’ passing on of Odalswind’s purchase),
the route was indirect, with the charter being preserved outside the
52 Erhart and Kleindinst, Urkundenlandschaft Rätien no. 42, and see above, n. 30.
53 Erhart and Kleindinst, Urkundenlandschaft Rätien, no. 40, with the erased dorsal notice
‘traditio de Iobones’ (perhaps a local entry prior to the gift being confirmed when Drucio
became a monk), replaced with the St Gall hand noting ‘traditio Iob de Vinumna’ and
then ‘cap. XV’.
54 Erhart and Kleindinst, Urkundenlandschaft Rätien, nos. 45–6, 51, 53.
55 ‘Conparacio sancti Galli de Val[erio]’; ‘Carta de Isuano ad parte sancti Galli’.
170 Matthew Innes
monastery for a period until the conditions attached to the prospec-

tive gift were met and the charter passed over. While three remaining
documents have no dorsal notice – which would in itself be remarkable
had they been placed at St Gall immediately on their redaction, so rou-
tine were the protocols of archival entry – another two, including the
only one of the series actually written at St Gall, have dorsal notices
that have more in common with the remainder of the run of non-St
Gall Rankweil charters than with the context of the monastic archive. It
must be possible, then, that they were not entered into the monastery’s
charter collection on redaction, but rather stored locally with the other
charters of the priest-scribes, perhaps themselves functioning as some-
thing approaching local ‘agents’ for the monastery. Could this ongoing
relationship – and the presence at Rankweil of some charters in which St
Gall was the direct beneficiary – explain why, at some point in the tenth
century, at least part of the Rankweil charter collection was brought to
St Gall? Or was it advantageous – and indeed expected – for the monks
to piggyback on local archival practices and store documents concerning
Rankweil and its hinterland on site?
Whatever the date and reason for their removal from Rankweil to St
Gall, there is no doubt that these charters are the relict of a once inde-
pendent document collection later transferred to St Gall. This collection
has a relatively long chronological horizon, reaching back over an arc
of a century or more to the immediate aftermath of the imposition of
direct Frankish rule in Rhaetia. Although the twenty-seven documents
connected with Folkwine’s activities dominate and were annotated as a
coherent subset, this document collection cannot be explained in terms
of the interests of any single individual, institution or family. Could we
classify the Rankweil collection as a lay archive? It served a range of
actors and has a chronological depth reaching back several generations,
while dorsal notices and the folding of the documents attest to some
level of organization. Just as at St Gall charter scribes normally indexed
and archived their own documents, so at Rankweil the agency of local
scribes may have been vital in the storage as well as the production of this
material. The fact that these scribes were priests does not mean that they
were acting solely as agents of their church, nor that they were divorced
from the local community of which they were members and which they
served; the fact that storage probably took place in a church does not
negate the fact that it was serving as, in effect, a public store for the
secular documents of the local community. Just as lay landowners who
were patrons might use the monastic archive at St Gall and elsewhere, so
even more at Rankweil we should not envisage the local document store
as a clearly defined institution. Perhaps its functions were multiple, and
its boundaries permeable and porous, with landowners able to deposit

and remove their documents as they needed them – hence the ad hoc and
piecemeal nature of much of its contents.
The contents of the Rankweil material give further precious glimpses
of the ways in which landowners kept their documents. Most have short
notes, of a similar type to the St Gall dorsal notices and comparable to
the headings used to introduce a document in a cartulary, jotted by the
original scribe on the reverse and apparently used as a reference tool. In
several cases, it is also clear that a single, long piece of parchment was
used for several transactions; sometimes the individual ‘charters’ have
been subsequently cut apart, but in a number of cases groups of docu-
ments survive on a single parchment. In most identified cases, this seems
to be a case of several legal acts that took place on the same day, presum-
ably at the same public meeting. Thus, among Folkwine’s records, we
have a series of transactions over a six-month period in 820, now cut into
several separate ‘charters’ but from the same parchment, and a roll of four
transactions from 821 on a single parchment, used in an early modern
bookbinding. From 882 we have a single parchment with two versions
(one rough, one fair) of the text of the same purchase by Otulf, his wife
and his son; in 883 Otulf, his wife and son similarly made two related
purchases from two different vendors and had them recorded on a single
parchment. A now fragmentary and undated document from the last two
decades of the ninth century takes the form of a roll where at least thirteen
different landowners gave their shares of mountain land to the church
of St Saviour ad Roncalem, while in 933 Magnus and his wife Quintella
made their last will and testament and recorded a land purchase on the
same parchment.56 These kinds of linked documents are only known,
within the St Gall archive, in the Rankweil documents. Presumably, such
practices had little use for the monks, whose archive was organized into
geographical ‘chapters’ and who had developed defined archiving proto-
cols. At Rankweil, however, an individual landowner might keep a series
of parchments, uncut and if necessary rolled. The documentary practices
of lay landowners remind us of the materiality of the written word as a
crucial aspect of its practical use, an aspect all too easily ignored if we
approach early medieval documents solely through the expectations of
high medieval cartularists, and classical diplomatic and modern printed
editions.
The Rhaetian material immediately stands out within the St Gall
archive as a regional block. In part, as we saw, this is because of its
56 Erhart and Kleindinst, Urkundenlandschaft Rätien, nos. 16–22, 27–30, 47–50, 54, 58–9;
see also nos. 74–9.
172 Matthew Innes
different provenance and transmission from the mass of material in the

monastery’s archive. But Rhaetia itself was a distinct region in the early
medieval period: nestled in the Alps, politically and socially relatively
self-contained, and linguistically Romance rather than Germanic, it was
the last of the great independent bishoprics to be firmly integrated into
the Carolingian firmament, by Charlemagne at the beginning of the ninth
century.57 As is demonstrated by the early ninth-century local compila-
tion entitled Lex Romana Curiensis, this region enjoyed a continuous legal
tradition that had developed from Roman times. Both the Lex and the
surviving documents indicate that Rhaetian legal practice differed from
that of St Gall’s Alemannian heartland.58 It is therefore tempting to see
this as an island of Romanitas whose exceptional footprint within the St
Gall archive is hardly surprising. It certainly is the case that the closest
comparisons for the Rankweil material come from south of the Alps,
in the charter collections associated with the family of Totone of Cam-
pione d’Italia, or (still more) the documents of Peter of Niviano, who
was, like Folkwine, a local official (sculdassius). These parallels may be no
accident given the similarities in post-Roman experience and the close
cultural links over the Alps in the pre-Carolingian centuries. Indeed, on a
macrohistorical level, the Rankweil charters might be linked to a broader
geopolitical process whereby Rhaetia’s cultural and political alignment
shifted from northern Italy to Alemannia.59
As it came into contact with an area like Rhaetia, with its rich docu-
mentary traditions, the St Gall archive could not help but suck in some
of the charter collections which sustained those traditions; because they
formed an exceptional block in an area where St Gall otherwise had min-
imal interests, it was possible for the Rankweil material thus acquired to
remain physically within the St Gall archive but not fully integrated into
its systems of classification or internal organization. Presumably, if St
Gall had made a cartulary, or the originals had been lost, the Rankweil
57 R. Kaiser, Churrätien im Frühmittelalter (Basel, 1998) is the best study of the region.
58 Lex Romana Curiensis (ed. E. Meyer-Marthaler, Die Rechtsquellen des Kantons
Graubünden [Aarau, 1959]) exerted influence on the formulas and legal concepts evident
in the Rankweil charters, and the sole surviving manuscript (St Gall, Stiftsbibliothek,
Cod. Sang. 722) is the copy of Ursicinus, one of our charter scribes: McKitterick,
Carolingians, 110. Note that Lex Romana Curiensis 12.1 (ed. Meyer-Marthaler, p. 341)
equated gesta with all charters witnessed by four curiales; see Erhart and Kleindinst,
Urkundenlandschaft Rätien, 33, perhaps linking to a similar post-Roman evolution of
public ‘registration’ as proposed above, Chapter 5.
59 S. Gasparri and C. La Rocca (eds.), Carte di famiglia: Strategie, rappresentazione e memoria
del gruppo familiare di Totone di Campione (Rome, 2005); F. Bougard, ‘Pierre di Niviano,
dit le Spolétin, et le gouvernement du comté de Plaisance à l’époque carolingienne’,
Journal des Savants (1996), 291–337; see below, Chapters 9, 12 and (for the Iberian
evidence) 10.
collection and with it Folkwine’s career would have disappeared without

trace. As it was, the Rankweil material was too peripheral to the main
interests of the monks to be integrated into their archive, and so para-
doxically relatively immune from the periodic pruning and reorganization
that shaped it to present needs. The abbey’s Alemannian hinterland may
not have produced secular collections as rich as that from Rankweil, but,
in any case, charters held at St Gall were strictly organized in chapters to
document and support the monastery’s holdings, meaning, as we have
seen, that such secular documents as existed have left only the faintest
of footprints. In Alemannia, St Gall clearly did not bring documentary
practices into a previously oral society. In fact, the evidence of placita
and judgements suggests that written records were particularly impor-
tant precisely because confiscated or seized property might be quickly
passed on by charter of sale to establish a ‘clean’ title and an alternative
parchment trail to that held in the monastery.60 Nonetheless, St Gall
rapidly acquired a dominant position as a provider of documentary ser-
vices, and as a result the traces of lay documentary culture are so much
harder to find.
An estate and its documents: Perrecy

in the ninth century
The St Gall evidence may be rich, but it pertains to a very particular
regional context of conquest, Christianization and confiscation in the
peripheral principalities east of the Rhine in the eighth century, followed
by the ninth-century institutionalization of new aristocratic and eccle-
siastical hierarchies. These processes were the prime drivers of archive
formation and changes in documentary culture in the eastern half of
the Carolingian world. The time lag of one or two generations between
the eighth-century moment of Carolingian conquest and the horizons
of archival reorganization and cartulary compilation underline that the
latter were not always an immediate and direct response to the former.
But the conquest led to the confiscation and redistribution of the estates
of erstwhile opponents of the Caroligians and a major investment in a
select handful of ecclesiastical institutions that were definitively aligned
60 MGH DD Kar. 1, nos. 166–7, discussed above, 158–9; the parallel with Folkwine’s
careful archiving of written proofs of the ‘hot’ property he acquired by hook or crook
is clear. Carolingian legislation insisting on written judgements was responding exactly
to these kinds of escalating and competing written claims: see, e.g., M. Innes, ‘Charle-
magne, Justice and Written Law’, in A. Rio (ed.), Law, Custom, and Justice in Late
Antiquity and the Early Middle Ages (London, 2011), 155–203.
174 Matthew Innes
(and sometimes forcibly realigned) to the Carolingian dynasty. In soci-

eties where landholding was characteristically diffuse and scattered, with
even aristocrats owning dozens of tiny parcels instead of agglomerated
estates, these changes created a veritable parchment trail in ecclesiasti-
cal archives as the tensions they set in motion were worked through in
succeeding generations.61
In the old Merovingian heartlands of western Francia and southern
Gaul, the Carolingian impact took a different form. The crucial phase
focused on the renegotiation of tenurial arrangements around Church
land, but in a context of a publicly acknowledged regalian prerogative to
call on and make temporary grants of ‘excess’ Church land, not of ‘raw’
conquest and expropriation.62 These developments took place in a region
where the Church enjoyed a continuous and vital post-Roman history
and thus had long-established rights and prerogatives, and where land-
holding at least partially consisted of larger and more compact blocks,
especially where the Church and king had been able to consolidate their
rights by building on public prerogatives inherited from Roman times.
Ecclesiastical institutions had for centuries been interacting with docu-
mentary cultures that had evolved from those of the Roman world to
safeguard its interests. In the pages of Gregory of Tours, for example,
we see a range of individuals and institutions manipulating documents
for good or ill, but when Gregory thinks of archives he thinks first and
foremost of gesta episcopalia recording the prerogatives of churches and
the commemorating the activities of bishops.63 East of the Rhine, where
Church property had been rapidly acquired over a few generations and
where documentary traditions were thinner in the first place, new archival
technologies had been necessary to allow ninth-century prelates to deal
with the sharp imperative of working through the consequences of eighth-
century tenurial upheaval. In western Francia and southern Gaul, on the
other hand, well-established expectations about document collection and
use could cope with a tenurial changes that worked to a different rhythm,
just as at Rankweil existing techniques of record-keeping were adapted
to cope with the dramatic imposition of Carolingian rule and the arrival
61 For models, see, e.g., M. Innes, ‘Kings, Monks and Patrons: Political Identities at the
Abbey of Lorsch,’ in R. Le Jan (ed.), La Royauté et les élites dans l’Europe carolingienne
(Lille, 1998), 301–24; Innes, ‘Peoples, Places and Power in Carolingian Society’, in
M. de Jong and F. Theuws (eds.), Topographies of Power from Late Antiquity to the Early
Middle Ages (London, 2001), 397–437; Innes, ‘Property, Politics’.
62 E.g. G. Constable, ‘Nona et decima: An Aspect of the Carolingian Economy’, Specu-
lum, 35 (1960), 224–50. Discussion has focused on Charles Martel’s attitudes towards
Church property rather than the systematic appropriation of rights over ‘excess’ Church
land under Pippin.
63 I hope to return to discuss this at greater length in future.
of its agents. Thus, when two royal justices (missi dominici) visited Mar-
seilles in 780, they heard an incredible story concerning the activities of
a patricius of Provence earlier in the eighth century, who had seized the
charter chest (arca) of a local church and burned its contents to facilitate
the seizure of property and its redistribution to his followers; the deed
recording the gift of the widow Adaltruda and its public validation, so
they heard, had survived because a local prelate had hidden it under his
robe and secreted it under an altar; witness testimony was then necessary
to unravel the complex and contested transactions involving ecclesiasti-
cal claims, royal grants, and the activities, licit and illicit, of successive
patricians of Provence.64
In western Francia and southern Gaul, cartularies were a phenomenon
of the eleventh century and beyond. As a result, our access to Carolingian
documentary culture is largely obscured by a later archival horizon, and
we can only trace the working through of Carolingian rule on the ground
indirectly.65 If we look for documentary responses to this process, we can
point to the production first, in the middle decades of the eighth century,
of inventories of Church land and then, in the ninth century, of polyp-
tychs documenting the Church’s rights over its property and tenants in
a wealth of practical detail.66 Even these were not novel: judging from
several of the gesta abbatum and gesta episcoporum that were such a popu-
lar new genre of the Carolingian period, ecclesiastical archives (likewise,
potentially, gesta) had long been based around inventories of estates and
their dues alongside royal privileges and the testamentary depositions
recording the heroic piety of past prelates and illustrious benefactors.67
64 Cartulaire de St-Victor de Marseille, vol. I, ed. B. Guérard (Paris, 1857), no. 31; this case is
also noted by P. Geary, ‘Umgang mit Urkunden in frühen Mittelalter’, in P. Erhart and
L. Hollenstein (eds.), Mensch und Schrift im frühen Mittelalter (St Gall, 2006), 11–24,
and it deserves full analysis elsewhere; for political context, see Geary, ‘Die Provence
zur Zeit Karl Martells’, in Jarnut and Nonn (eds.), Karl Martell in seiner Zeit, 382–92,
and Geary, Aristocracy in Provence: The Rhone Basin at the Dawn of the Carolingian Era
(Stuttgart, 1985), which gives further indications of the rich documentary culture of
the region.
65 For these vectors of transmission, see below, Chapter 11.
66 E.g. I. Heidrich, ‘Das Breve der Bischofskirche von Macon aus der Zeit König Pippins
(751–768): Mit Textedition’, Francia, 24 (1997), 17–37; I. Wood, ‘Teutswind, Witlaic
and the History of the Merovingian precaria’, in W. Davies and P. Fouracre (eds.),
Property and Power in the Early Middle Ages (Cambridge, 1995), 31–52; for polyptychs,
see below, nn. 68–9.
67 M. Sot, Gesta episcoporum, gesta abbatum (Turnhout, 1981). For a model analysis and
edition of the instrumental use of Merovingian archival muniments in such material, see
M. Weidemann, Geschichte des Bistums Le Mans von der Spätantike bis zur Karolingerzeit:
Actus pontificum Cenomannis in urbe degentium und Gesta Aldrici (Mainz, 2002); the
Carolingian context of controversies over Church property that shaped this material is
controversially analysed by W. Goffart, The Le Mans Forgeries: A Chapter from the History
176 Matthew Innes
Indeed, the nomenclature and conception of the Carolingian polyptychs,

if not their function, had its roots in these ecclesiastical collections. In the
tenth century, Flodoard of Reims found a continuous tradition of such
estate inventories, styled polyptychs, alongside the wills of Merovingian
bishops and the voluminous correspondence of their ninth-century suc-
cessors, while fragmentary survivals from Tours suggest that this was not
exceptional.68 Against such a backdrop, polyptychs could serve as proof
of title and of obligation in Carolingian courtrooms.69
To understand how inherited documentary cultures might cope with
the challenges thrown up by eighth- and ninth-century changes, let us
turn to a western case study as a counterpoint to St Gall. The texts of
a dozen ninth-century documents concerned with the estate (villa) of
Perrecy in Burgundy are transmitted via the monastery of Fleury.70 As at
Rankweil, the assembly of this dossier predates the acquisition of these
documents by a major church. Perrecy had been willed by Count Eccard
of Mâcon to the monks of Fleury in 876, and although his kin had con-
tested this bequest, Perrecy subsequently became an important holding
of the monastery and the site of a major priory.71 A cartulary was com-
piled at Perrecy c. 1014; the collection contained material relating to the
of Church Property in the Ninth Century (Cambridge, MA, 1966); see also P. Janin, ‘Heiric
d’Auxerre et les “Gesta pontificum Autissiodorensium”’, Francia, 4 (1976), 89–105; I.
Wood, ‘St Wandrille and Its Hagiography’, in G. Loud and I. Wood (eds.), Church and
Chronicle in the Middle Ages: Essays Presented to John Taylor (London, 1991), 1–16; M.
Sot, Un historien et son église au Xe siècle: Flodoard de Reims (Paris, 1993).
68 W. Goffart, ‘Merovingian Polyptychs: Reflections on Two Recent Publications’, in Gof-
fart, Rome’s Fall and After (London, 1989), 233–53 (orig. Francia, 9 [1982], 55–77); for
the Tours documents (ChLA XVIII 659), see S. Sato, ‘The Merovingian Accounting
Documents from Tours: Form and Function’, EME, 9 (2000), 1–19; for Rheims, J.-P.
Devroey, ‘Les premiers polyptyches rémois’, in A. Verhulst (ed.), La grande domaine
aux époques mérovingienne et carolingienne (Ghent, 1985), 78–97.
69 E.g. J. L. Nelson, ‘Dispute Settlement in Carolingian West Francia’, in Settlement of Dis-
putes, 45–64, esp. 48–51; J.-P. Devroey, ‘Libres et non-libres sur les terres de Saint-Remi
de Reims: La notice judiciaire de Courtisols (13 Mai 847) et le polyptyque d’Hincmar’,
Journal des Savants (2006), 65–103; Devroey, ‘Elaboration et usage des polyptyques:
Quelques éléments de réflexion à partir de l’exemple des descriptions de l’église de
Marseille (VIIIe –IXe siècles), in W. Haubrichs, J. Jarnut and D. Hagermann (eds.),
Akkulturation: Probleme einer germanisch-romanischen Kultursynthese (Berlin, 2004), 436–
72. Note also the reference to a polyptych in the Marseille document, cited above
n. 64, which predates the surviving descriptio of 814.
70 Recueil des chartes de l’abbaye de St-Benoit-sur-Loire, ed. M. Prou and A. Viader (Paris,
1903), esp. pp. lviii–lxvii for transmission. For analysis, see B. Kasten, ‘Erbrechtliche
Verfügungen des 8. und 9. Jahrhunderts: Zugleich ein Beitrag zur Organisation und zur
Schriftlichkeit bei der Verwaltung adeliger Grundherrschaften am Beispiel des Grafen
Heccard aus Burgund’, Zeitschrift der Savigny-Stiftung für Rechtsgeschichte, Germanistis-
che Abteilung, 107 (1990), 236–338; Wormald, The Making of English Law, 76–91; and
Innes, ‘Practices of Property’, 249–56.
71 Recueil . . . St-Benoı̂t-sur-Loire, no. 25 (Eccard’s will), with related nos. 26–8; for his kin
contesting this bequest, see no. 30.
priory’s complex relationship with Fleury, alongside royal privileges and

other documents concerning both Fleury and Perrecy such as Eccard’s
will, but it began with a run of ten documents from the years 796–876
that predated Eccard’s gift of the Perrecy estate to Fleury. Although the
Perrecy cartulary is now lost, it was via these copies that the texts of these
documents were known at Fleury and so preserved in further cartulary
copies.72
The documents split into two chronologically and thematically distinct
sets. The first, earlier, set consists of seven documents from 796 to 821
which concern the interests of a number of powerful aristocrats holding
Perrecy and other nearby parcels of royal land as benefices. The driving
force behind this documentation is the activities of the local agents of
these aristocratic benefice-holders as they take a succession of named
individuals to court and establish their unfree status and therefore their
possession as part and parcel of a particular estate. These notices give
a rare insight into the use of the public courts to buttress the rights
of landlords over the peasantry, especially by adjudging issues of free-
dom and servitude. In two of these cases, an initial hearing is deferred
and a document drawn up recording that deferral, prior to the produc-
tion of witnesses and the writing up of their testimony, thus creating a
paired document. Not all of these cases concern Perrecy itself, although
there are clear links explaining their association. For example, the earli-
est case, from 796, concerns the status of a peasant at Le Jeu, held as a
benefice by the Count Hildebrand who appears to hold Perrecy in the
remainder of the series.73 A second case, concerning property at Baugy
unjustly held by Amelius and restored to Count Nibelung, who held it as
a benefice, relates to an estate that was both earlier and later associated
with Perrecy; it may have been a local estate centre whose relationship
to the main Perrecy holding changed.74 Throughout the series, it is local
agents rather than the aristocrats who held these benefices who are the
main protagonists; often described as ‘advocates’ of their aristocratic
72 For the context and technique of cartulary compilation in this period, see below, Chapter
11.
73 Recueil . . . St-Benoı̂t-sur-Loire, no. 9.
74 Recueil . . . St-Benoı̂t-sur-Loire, no. 13. Nibelung was in any case almost certainly a kins-
man of Hildebrand. Eccard’s will (Recueil . . . St-Benoı̂t-sur-Loire, no. 25) tells us that
Hildebrand – Eccard’s father who had held Perrecy as a benefice – had held Baugy as
benefice, but makes Baugy – which Eccard had granted as a benefice to Leutold – an
exception to his gifting of Perrecy to Fleury. That is, it acknowledges that it might be
seen as an appendage of the larger estate (presumably Eccard’s rights at Baugy came
from the grants of 836 and 839), but one that was not integrated or integral to that estate;
thus Baugy is willed differently from the wider Perrecy estate, going to the monastery
of St-Antoine.
178 Matthew Innes
masters, these figures should be seen as something like ‘estate-managers’

recruited locally. Fredelus, for example, who asserts Hildebrand’s rights
at Perrecy in most of the series, had appeared as a witness in the first
document, from 796, where Moses acted as Hildebrand’s advocate. Even
in the outlier, the reclamation of Baugy, it is one Fulchard who acts as
advocate of the benefice-holder Count Nibelung. The parchment trails
assembled by Fredelus, Moses and Fulchard as they pursued dependent
peasants through the local courts vividly demonstrate the vital impor-
tance of procuring publicly affirmed and validated documents to safe-
guard rights. As they are transmitted via cartulary traditions, not in
the original, we can only speculate as to their original appearance and
preservation – perhaps in bundles of linked parchments, with contempo-
raneous or related cases copied beneath each other on a single sheet, as
we saw at Rankweil?
The second half of the Perrecy ‘dossier’ concerns the contestation
of control of the estate among the aristocracy in the half-century prior
to Count Eccard’s gifting it to Fleury. It consists of three documents
concerning Count Eccard, which form an obvious prologue to the will
and its associated documents in which he gave Perrecy to Fleury. But
these documents also imply that from the rebellions of the 830s through
the recurrent intradynastic familial conflicts that structured Carolingian
politics for the next generation, Perrecy was a plum prize that aristocrats
coveted from their royal masters. They comprise the outright gift, in
full property, of Perrecy to Eccard by King Pippin of Aquitaine in 836;
the subsequent (re)gift of Perrecy, in full property, to Eccard by the
emperor Louis the Pious in 839 (after Pippin’s death and so presumably
reflecting a date where this erstwhile follower of a rebellious royal son
had come to terms with the imperial father seeking rapprochment); and
finally the incomplete record of a court hearing held between 866 and
877 before two royal justices (missi dominici) of King Charles the Bald
at the villa of Mont, at which Eccard’s title to Perrecy was contested by
the bishop of Bourges and their cases referred to the testimony of local
witnesses.75
The Mont notice is a complex and fascinating document, and of partic-
ular interest here because within it are references to two further tranches
of documents concerning Perrecy that were produced in court but do
not survive. Bishop Wulfad of Bourges brought three documents, which
75 Recueil . . . St-Benoı̂t-sur-Loire, nos. 20–1, 24. The Mont case is justly famous thanks to
its discussion by Nelson, ‘Dispute Settlement’, at 53–5. She suggests a date of c. 875; the
absolute limits remain the ordination of Wulfad to Bourges in 866 and Count Eccard’s
will in 877.
were read aloud: two Merovingian charters related to Perrecy, and a

document from the time of the first Carolingian king in which a previ-
ous bishop granted a life interest (precaria) in the Perrecy estate to one
Nibelung in return for an annual rent of three pounds on the feast of St
Mary.76 The count, Eccard, produced two more recent documents, like-
wise read aloud: an imperial charter issued by Louis the Pious, and the
notice recording a hearing held in a general assembly of the current king,
Charles the Bald, in which Eccard recovered property granted to him in
the earlier document, against the claims of an otherwise unknown John.
While the first of Eccard’s documents (Louis the Pious’ grant) survives
in cartulary copy, the second (the judgement made under Charles the
Bald) does not. The use of documents in court was evidently a matter of
careful strategic choice: Eccard had first been granted Perrecy by Pippin
of Aquitaine in 836, but he chose to base his claims on Louis the Pious’
subsequent grant; evidently, a single, ‘clean’, grant by an emperor was
preferable to one made by a rebellious son whose right to make the grant
might have been seen as questionable, indeed a troubling precedent given
Charles the Bald’s recurrent problems with rebellious sons in Aquitaine.
Conversely, the court case whereby Eccard was able to prevail before
Charles the Bald and thus (re)establish his control of a part of Perrecy,
probably after the civil war between Louis the Pious’ sons, is omitted
from the cartulary dossier, yet it was vital at Mont in showing that the
imperial grant by Louis had been previously accepted in a public court
and was so ‘accredited’. This case shows charters being used as weapons,
but they were wielded with a degree of sophistication and precision that
is easily overlooked. The surviving notice of the Mont meeting is unfin-
ished, but after relating the documents produced it recites the carefully
worded sworn testimonies given by summoned witnesses. It is quite clear
from this alone that the hearing was heading in the direction of Count
Eccard.
Well-established interpretative traditions have informed the analysis of
records of dispute settlement such as the notice of the Mont hearing.77
76 The Merovingian diplomata among the Bourges documents are from the reigns of
otherwise unspecified Childebrand and Chilperic; I take the ‘precaria ex tempore Pippini
regis’ to have been issued under the first Carolingian king (751–68), rather than one of
the later Pippins who acted as subkings in Aquitaine, given that we know a Nibelung was
a crucial backer of the new regime in Burgundy, and that the granting of excess Church
land in precaria was a feature of the reign: following Nelson, ‘Dispute settlement’,
54n38, citing L. Levillain, ‘Les Nibelungen historiques et leurs alliances de famille’,
Annales du Midi, 49 (1937), 337–407, at 354.
77 See W. Brown and P. Górecki, ‘What Conflict Means: The Making of Medieval Conflict
Studies in the United States, 1970–2000’, in Brown and Górecki (eds.), Conflict in
Medieval Europe: Changing Perspectives on Culture and Society (Aldershot, 2004), 1–36.
180 Matthew Innes
Through the nineteenth and much of the twentieth centuries, historians

interrogated sources of this kind with a view to weighing the relative
value placed on written and oral proof. In doing so, they were influenced
by a legal-historical method designed to adduce hierarchies of argument
and proof and measure progress towards a ‘rational’ legal mentality; they
thus drew a strong contrast between the use of a written record and
sworn testimony as proof. More recent scholarship, on the other hand,
has stressed the interaction between the written and the oral in a world
where documents were read aloud and their contents tested against the
sworn testimony of locals.78
In fact, a careful reading of the surviving notice of the Mont hearing
underlines the extent to which the documents produced by the two par-
ties framed the questions put and the responses given. Perrecy, it seems
clear, had been Bourges’ property in the Merovingian period, by royal
grant, but in the context of Carolingian takeover in Burgundy, it was
deemed ‘excess’ to Bourges’ needs and so granted out for a lifetime to
a loyal follower of the Carolingians,79 with the church retaining ultimate
title. By the ninth century, however, Bourges’ claims to ultimate own-
ership were forgotten, and in the context of the struggle between rival
Carolingians in the civil strife of the 830s the young Eccard was rewarded
with the gift of what was now seen as royal land ‘in full property’.80
When the various witnesses who came to Mont gave their testimony to
the royal missi who had called the meeting, their words were carefully
chosen; they swore not only as to whom they had seen ‘holding’ the
property, but also as to what they had seen and heard said about ultimate
78 See Nelson, ‘Dispute Settlement’, 54, on Perrecy: ‘oral testimonial proofs . . . were the
more necessary where documents conflicted’. For the chronological and geographical
variations in the explicit citation of earlier documents in records of dispute settlement,
see K. Heidecker, ‘Communication by Written Texts in Court Cases: Some Charter
Evidence (ca. 800–ca. 1100)’, in M. Mostert (ed.), New Approaches to Medieval Commu-
nication (Turnhout, 1999), 101–26; Heidecker, ‘Emploi de l’écrit dans les actes judicia-
ries: Trois sondages en profondeur: Bourgogne, Souabe et Franconie (VIIIe –début XIIe
siècle)’, in M.-J. Gasse-Grandjean and B.-M. Tock (eds.), Les actes comme expression du
pouvoir au haut moyen âge (Turnhout, 2003), 125–37. See also below, Chapter 11, for the
ways in which documents might structure court hearings without their being explicitly
noted by a placitum scribe – another reminder that we should avoid reading placita as
stenographic transcripts rather than interested representations of dealings in court, as in
the salutary warning of W. Brown, ‘Charters as Weapons: On the Role Played by Early
Medieval Dispute Records in the Disputes They Record’, Journal of Medieval History,
28 (2002), 227–48.
79 This Nibelung was none other than the kinsman of the Carolingians who was a key sup-
porter in Burgundy and commissioner of the first set of continuations to the Chronicle
of Fredegar, recounting Pippin’s reign. On the family (with important information on
Perrecy), see Levillain, ‘Les Nibelungen historiques’.
80 Kasten, ‘Erbrechtliche Verfügungen’, is convincing on the estate’s history.
rights of ownership. As it was common ground that Eccard ‘held’ the

estate, particular care was given to the question of his payment of any
dues to another, his performance of legal rituals denoting obligations to
another, or any other evidence that might constitute a ‘mark of dominium’
over Perrecy. Ultimately, even those whose claims about Bourges’ right-
ful title had precipitated the case could summon no tangible evidence
to support the bishop’s charters; their claims had been intermittently
repeated in the locale, but had never been publicly acknowledged by the
local community.81 The case thus did not turn on any underlying conflict
between written and oral proof, or any rule as to the ultimate value of
charters of different pedigrees, but rather on a pragmatic approach that
tested two competing legal arguments constructed around documents
against local public opinion.82
Reading the testimony in the context of the charter proofs produced
by both sides therefore reveals a structured interrogation of a specific
set of legal questions left implicit in the surviving notice. At Mont, both
parties produced not only royal diplomas granting them possession of the
Perrecy estate but also subsequent documents showing the public affir-
mation and implicit validation of the claims in those documents. The sur-
viving notice stresses that Bourges’ life-grant to Nibelung with its explicit
acknowledgement of the bishopric’s ultimate ownership was witnessed
by many boni homines, while Eccard felt the need to supplement his impe-
rial diploma, for all its undeniable charisma, with a notice recording his
successful defence of this grant in court.83 In short, documents needed
affirmation, and a demonstration that a document – however ancient or
charismatic – had been tested in the public courts and not found want-
ing added to its value. The Mont notice thus gives us priceless insight
into how documents – all bar one now lost – were received, used and
understood. Although deperdita such as those mentioned in the Mont
document are characteristically treated as phantoms and excluded from
editorial and historiographical canons, it might be argued that footprints
such as these – even where we have no surviving text – have a particular
value for some forms of historical analysis, precisely because they reveal
use and reception.
81 Following Innes, ‘Practices of Property’, arguing that the words chosen to record the
testimony were carefully selected to distinguish claims of ‘holding’ from ‘ownership’;
Nelson’s translation tends to see the Latin as less precise in its meaning.
82 This model of ‘substantive legalism’ ultimately draws on the work of Stephen D. White;
e.g. ‘Inheritances and Legal Arguments in Western France, 1050–1150’, Traditio, 43
(1987), 55–103.
83 See below, 249–50, for ostensio cartae procedures in Italy, and above, Chapter 5, for
the move from formal validation to public recognition in earlier Frankish validation
procedures.
182 Matthew Innes
It has been suggested that the survival of the Perrecy documents

‘implies the existence of a family archive transferred to Fleury on Eccard’s
death’.84 In fact, the pattern of surviving documentation cannot easily
be reduced to such an institution, and the transmission suggests that it
was stored at Perrecy, where it remained. True, the documents establish-
ing Eccard’s title to Perrecy might be seen as part of Eccard’s personal
archive, handed over with his will and associated material to the monks
of Fleury. But it is not at all clear that Eccard’s documents should be seen
as a part of a coherent whole with the documents that preceded Eccard’s
acquisition of the estate. Although Count Hildebrand, who had held
Perrecy to the 820s, was Eccard’s father, we do not know that the estate
was passed on in direct or seamless succession from them, for Hildebrand
had died in 827 and Eccard received his father’s former benefice as full
property in the scramble for honores in the 830s, and in rivalry with other
aristocrats. The earlier documents might be seen as a useful set of test
cases for the unfree status of certain sections of the local peasantry tied to
the Perrecy estates, and thus understood as part of the apparatus of estate
management held on site by those locals entrusted with running Perrecy
and defending the rights of its owners: hence their preservation at Perrecy
right through to 1014 and the compilation of the priory’s cartulary.85 At
Perrecy, then, our window on the ninth century is only possible because
the Carolingian estate records could be used to underpin the claims of
the eleventh-century seigneurie, and because they gave Perrecy a history
independent of Fleury, and so aided the cartulary compiler in his aims of
providing a documentary template for the complex relationship between
priory and mother house.86
From Roman public records to Frankish

charter collections
How does the documentary culture revealed at Rankweil and Perrecy
relate to our inherited models of long-term change? In view of the
84 Nelson, ‘Dispute Settlement’, 54, and later Nelson, ‘Literacy in Carolingian Govern-
ment’, in Uses of Literacy, 258–96, at 276.
85 Landownership and estate structures at Perrecy: O. Bruand, ‘La Villa carolingienne,
une seigneurie? Réflexions sur les cas des villas d’Hammelburg, Perrecy-les-Forges et
Courçay’, in D. Barthélemy and J.-M. Martin (eds.), Liber largitorius: Études d’histoire
médiévale offertes à Pierre Toubert par ses élèves (Geneva, 2003), 349–73, although there is a
teleology implicit in the argument that Perrecy was therefore ‘already’ a seigneurie. I owe
this reference to Charles West. Bruand’s discussion does provide a clear context for the
collection and preservation of documents classifying and controlling unfree dependants.
It also makes sense of the Baugy reclamation and the lost case in which Eccard defeated
the claims of John, cited at Mont; some of the testimony at Mont also seems to imply
the possibility of levels of agency and subjection within the villa.
86 See above, at n. 70.
problems of organizing complex and changing bodies of source mate-

rial into a compelling account of change, it is no surprise that historians
of documentary practice have developed grand narratives that are fun-
damentally influenced by and intertwined with familiar historiographi-
cal meta-narratives of post-Roman transformation. True, those favoured
meta-narratives have freed themselves from notions of cataclysm and
decline, in large part because the distant Roman roots of early medieval
documentation as it emerges from the end of the seventh century are
incontrovertible. But meta-narratives about the shift from public to pri-
vate power, and from the institutions of the late Roman state to the early
medieval Christian Church, have been powerful but often hidden orga-
nizing principles in much scholarship. They have allowed historians to pit
the inevitable messiness that emerges when we can glean detailed insights
into the practical workings of legal documentation after c. 700 against
the norms that are all that remain from the fifth, sixth and early seventh
centuries. The result, unsurprisingly, has been a contrast between the
public institutions of Late Antiquity and the variety of interests and insti-
tutions involved with written documentation in the Carolingian period.
Late antique regulations on the drafting and registration of documents
are read as indicating a clearly defined and strictly demarcated set of insti-
tutions – public notaries and public archives – whereas surviving charters
from c. 700 onwards seem to show a variety of scribes and mechanisms
for the validation and publication of documents. Similarly, whereas our
normative sources from the immediately post-Roman period envisage
a wide range of contexts in which documents might be used by private
individuals, the actual documentation from later centuries has a narrower
range, and is dominated by property transfers to the Church. One impor-
tant strand of scholarship has seen these shifts as ultimately indicative of
a series of wider changes in documentary culture: late antique systems of
public notarization and registration gave way to the private production
and preservation of documents, with the Church becoming the domi-
nant institution of written record; ecclesiastical scribes and ecclesiastical
archives, therefore, both aligned to the particular interests of the Church,
marginalized the landowning classes from documentary practice.87
Care needs to be taken in moving from the changing profile of our
sources to institutional and social practice. After all, where we do have a
meaningful volume of documents to elucidate late antique archival and
scribal practice, the clearly defined and strictly demarcated public insti-
tutions of the normative sources are immediately shot through with the
interests of powerful individuals and specific interests, as the Egyptian
87 Classen, ‘Fortleben and Wandel’; Ganz and Goffart, ‘Charters Prior to 800’.
184 Matthew Innes
and Ravennate evidence demonstrates.88 The run of over fifty original

papyrus and parchment documents from the late Merovingian St-Denis
archive – without which we would be clutching at fragments in evalu-
ating seventh-century Frankish documentation – fits only with difficulty
into inherited grand narratives, and raises issues paralleled in the run
of sixth- and seventh-century papyrus documents of diverse provenance
preserved in the archives of the church of Ravenna.89 It is not a narrowly
ecclesiastical chartrier as the historiography would have us expect, and it
includes a priceless run of royal documents and of private wills of which
St-Denis was not a beneficiary. While St-Denis’ close links with the late
Merovingian court might go some way towards explaining the royal doc-
umentation, the range of material makes it difficult not to imagine some
wider social function; unlike the Ravenna material, however, none of that
from St-Denis uses the template of Roman public registration in the gesta
municipalia.
In any case, a careful look at Carolingian scribal practices shows some
continuing framework of public authority within which a variety of indi-
viduals operated, some churchmen who might work for private individu-
als as well as their own institutions, and some local landowners who might
serve patrons, clients and neighbours.90 The evidence discussed above for
practical charter use likewise underlines the importance of formal records
of public recognition and validation in ‘accrediting’ a document; claims,
even when recorded in documents, were stronger when they had been
affirmed in court and subjected to public scrutiny. In other words, the
horizons of our evidence may suggest a clean cut and complete transfor-
mation from state registration and regulation to ecclesiastical monopoly,
but if we take the care to deal with the surviving documents on their own
terms it becomes clear that in fact we are dealing with a shift in under-
standings of ‘the public’, no longer so closely tied to the constraints of
formal law and regulated state institutions.91
In reconstructing these vectors of change, we need to take full account
of regional difference within the Frankish world, lest we end up com-
paring the last gasp of Roman urban civilization south of the Loire in
the fifth and sixth centuries with the new ecclesiastical frontier of the
provinces east of the Rhine in the eighth and ninth, and ripping our
88 See above, Chapters 2 and 4.

89 I differ here from Ganz and Goffart, ‘Charters Earlier Than 800’. On Ravenna, see
above, Chapter 4.
90 For the situation as revealed by the Cluny and St Gall originals, see above, 298, 156–7,
167; see also McKitterick, Carolingians, 77–134; Innes, State and Society, 111–18; and
below, Chapter 8.
91 Cf. above, Chapter 5.
evidence from context. When it comes to the history of written law

and legal documentation, regional comparisons are normally couched in
terms of the differential shading of the Roman inheritance. The pioneer of
the study of Roman institutional legacies in Frankish Gaul, Eugen Ewig,
even imagined an internal division stretching diagonally, north-west to
south-east, across the Merovingian world, starting on the Channel coast,
skirting the Loire valley and Burgundy, and reaching the Alps. South of
this line, the civic and episcopal institutions preserved Roman practices
on a local level into the seventh century and beyond. To the north of
this line, Roman institutional survival was at best fragmentary and hap-
hazard; internal acculturation between north and south was vital in the
effective revival of the Church in the sixth and seventh centuries and with
it a whole spectrum of practices, written documentation not least among
them.92 Inevitably, the details of Ewig’s picture need careful nuancing,
with a range of other comparisons and variables adding colour to the
black-and-white of the original picture. Nonetheless, it is striking that
the regional comparison sketched by Ewig seems to map onto the differ-
ences in documentary profile evident once we begin to get a meaningful
transmission of charters, in the eighth to tenth centuries. Direct continu-
ity of documentary practices through the Merovingian centuries is most
marked in those regions south and west of Ewig’s line. Rhaetia’s history is
almost emblematic of the continued evolution of Roman forms on a local
level, while the rich documentary traditions evident in Burgundy likewise
belong to a region where the documentary habit clearly enjoyed a contin-
uous post-Roman history. Paradoxically, though, in these regions where
legal writing enjoyed a continuous post-Roman evolution, the archival
horizons are late, with documentary traditions surviving overwhelmingly
in post-Carolingian compilations, above all cartularies of the eleventh
and twelfth centuries, which cannot be read as providing full or unmedi-
ated access into Carolingian documentary culture. One result is a certain
tentativeness in scholarly discussion of the extent to which document use
extended beyond court and cloister in precisely those regions where doc-
umentary traditions enjoyed most continuity through the early medieval
centuries.
In Francia north of the Paris basin, and in the East Frankish world
beyond the Meuse and Moselle, our archival horizon differs. It is deter-
mined by a succession of Carolingian compilations of copied documents
92 E. Ewig, Spätantikes und fränkisches Gallien: Gesammelte Schriften, 3 vols. (Munich,

1976–2009), esp. I, 409–34 (‘Das Fortleben römischer Institutionen in Gallien und
Germanien’, orig. Relazioni del X Congresso internazionale di scienze storiche, 6 vols.
[Florence, 1955], VI, 561–98).
186 Matthew Innes
made by major ecclesiastical institutions. As we have seen, these com-

pilations related to the specific situation of newly conquered and Chris-
tianized regions; even at St Gall where – exceptionally – no cartulary
was compiled either in the Carolingian period or later, the complex-
ities of the abbey’s relationship with central authority led to a coher-
ent system of archival organization being developed very early in the
institution’s history, a system reminiscent in its underlying principles of
contemporary cartularies and one that effectively determined the con-
tours of institutional memory. As a result, for eastern Francia we have
far thicker documentary traditions, with a succession of major institu-
tional compilations of hundreds – and sometimes thousands – of docu-
ments, and a sense of a far fuller social picture than is possible for much
of western Francia.93 But the sheer weight of this material produces
a sense of ecclesiastical dominance – even in regions like the Moselle,
the middle Rhineland or the area around Lake Constance, where odd
lay documents transmitted within ecclesiastical traditions hint at local
documentary traditions prior to the foundation of the great churches of
the Carolingian period. Moreover, because our Carolingian cartularies
do not appear to have produced a continuous tradition of updating and
recopying, in most cases the surviving profile of document transmission
tails off dramatically, with the early ‘peak’ captured in our ninth-century
cartularies followed by a thinner tradition dependent on high medieval
cartularists. The resulting impression has further encouraged some his-
torians to think in terms of a ‘crisis’ in the practical use of writing,
which gells with received narratives of Carolingian ‘failure’ and a tenth-
century society organized on ‘archaic’ and ‘oral’ lines around political
ritual.94
Ultimately, it is the East Frankish cartularies that dominate our image
of Carolingian documentary practice, because of the sheer wealth of
evidence; the close association of their producers with the patronage of
Carolingian kings; and the degree of economic, social and political detail
they offer. Moreover, although our West Frankish cartulary traditions are
later and thinner, they develop the basic forms set down by their ninth-
century predecessors, thus supporting our tendency to see the cartulary
as a default image of the ecclesiastical archive of the Carolingian period.
Yet, as we saw in our case studies, cartulary compilation was a product
of tenurial and political upheaval as new forms of institutional landhold-
ing familiar from the Frankish heartlands spread to the peripheries of
93 With obvious exceptions such as eastern Brittany, on which see below, 187 at n.95, 320.
94 For further historiographical discussion, see below, Chapter 11.
the Frankish world.95 In organizing their records, churches reacted to

the patterns of extra-ecclesiastical document use in their hinterlands and
to immediate institutional imperatives, and so there were marked dif-
ferences between regions and between institutions. Crucial factors here
were the differing impacts of the Carolingian state, and the degree to
which viable archival practices that allowed institutions to cope with the
tensions implicit in Carolingian state-formation were already present.
How, then, should we think of the charter collections of eighth-,
ninth- and tenth-century landowners in terms of the questions that
frame this book: post-Roman continuity and change, the role of eccle-
siastical institutions in archival and documentary culture, and the rela-
tionship between record-keeping and power? We should, unsurprisingly,
allow for regional difference, not only in our period but also within the
Roman inheritance. The documentary cultures behind and before the
East Frankish cartularies may have developed through internal accul-
turation, and may increasingly have been sustained through the inter-
play between aristocratic landowners and the churches they patronized,
but they were not wholly confined to the Church, and they need judg-
ing against the frontier literacy of the Roman army and its subsequent
legacy. The earliest document from early medieval Bavaria, the ‘Rottach-
gau fragment’ (written some time between the end of the sixth and the
early eighth century), suggests that such practices were remembered even
on the peripheries of the Frankish world, while the rich traditions of title-
deeds and judgement notices passed on to the Breton abbey of Redon
as it acquired property in the ninth century likewise show the adaption
of documentary practices by landowning communities on the fringes of
the late Roman world in the centuries that followed.96 We should also
acknowledge that, where it is possible to assess the archival and docu-
mentary practices that preceded the formation of ecclesiastical archives
and the compilation of cartularies, those practices witness a complex
post-Roman evolution that cannot be easily contained within the grand
narrative of a shift from public to ecclesiastical production or preserva-
tion. It is true that at Rankweil, in a context where the sixth to eighth
centuries had seen the establishment of an episcopal principality, local
95 Hence the anomaly of the Redon charters within the West Frankish tradition. See further
below, 320.
96 The ‘Rottachgau fragment’, transmitted as the first document in the ninth-century
Traditionsbuch of Passau: TP 1, and see now F.-R. Erkens, ‘“Actum in vico fonaluae
die consule”: Das Rottachgau-Fragment und die romanische Kontinuität am Unterlauf
des Inns’, in U. Ludwig and T. Schilp (eds.), Nomen et Fraternitas: Festschrift für Dieter
Geuenich zum 65. Geburtstag (New York, 2008), 491–510. For Redon, see above all W.
Davies, Small Worlds: The Village Community in Early Medieval Brittany (London, 1988).
188 Matthew Innes
priests in the service of royal officials played a crucial role in documentary

culture, and a royal church is perhaps the most likely physical location for
document storage. But, conceptually, these activities seem to have been
understood as part of a wider public culture of document use that tran-
scended particular institutional interests, to the extent that even St Gall
seems to have operated through these documentary mechanisms in this
locale. Meanwhile, in Burgundy, the Perrecy evidence shows a rich doc-
umentary culture in which priests and churches doubtless participated,
but in which the Church as an institution had no special role in the pro-
duction or preservation of documents. Had we the space to expand our
range of examples, the range of possibilities would be expanded. But in
all of these cases, we see a society in which documents were used and
stored by landowners without defined public institutions of record, and
without the Church enjoying the kind of hegemony as a provider of doc-
umentary services that it may have been establishing east of the Rhine in
the eighth and ninth centuries. In other words, we need to acknowledge
the existence of a complex set of practices that cannot be understood
as a mere staging post between the lost late antique world of public
archives and the great ecclesiastical charter compilations of the eleventh
and twelfth centuries, but needs evoking in its own terms. The notion of
‘Roman survival’ hardly does justice to the richness of the practices like
those we have glimpsed at Rankweil or at Perrecy, which, after all, were
managed not by first-order aristocrats but by their local agents. After all,
while a military officer or major landowner in late antique Egypt might
have left a comparable documentary footprint to Fredelus or Folkwine,
we have no evidence that the same is true of their predecessors in the
Empire’s western provinces. If we are to understand the documentary
transformations of the post-Roman world, we must put the endeavours
of estate-managers, local officials and small-time landowners to secure
their interests centre stage.
8 The production and preservation of
documents in Francia: the evidence
of cartularies
Hans Hummer
Other cartularies from the East Frankish regions show processes of archival
reorganization in response to Carolingian expansion similar to those visible at
St Gall, but with very different material outcomes. Carolingian expansion also
led to the centralization of scribal practice in churches and monasteries; though
the Carolingian state was a very different kind of polity from that of Roman
Late Antiquity, here, too, the state and its imperatives shaped documentary
culture. Most importantly, however, monastic archives in this region appear
once again not as timeless givens, but rather institutional creations that emerged
in particular contexts for particular reasons – reasons tied to the ways that the
Carolingians changed the relationships between ecclesiastical institutions and
the people around them.
In the Frankish realms of the early Middle Ages there existed no lay
method or system of archival management to rival that of monasteries
and churches. At least, we have no firm evidence of one. Nonetheless,
strong circumstantial and vestigial evidence testifies that many individu-
als routinely kept documentation. Narrative accounts of disputes some-
times refer to the destruction of documents held by laypersons; charter
collections compiled by monks include within them on occasion trans-
actions among the laity; donations to a monastery or church might refer
to lay records that once lay behind the transaction; and there are several
stunning examples of dossiers of records held by lay individuals and taken
up into ecclesiastical archives when they had donated their properties.1
Some of these are quite large, such as the Folkwine dossier, but without
exception they cover a few generations at most. They lack the chronolog-
ical depth that one generally associates with ecclesiastical and monastic
archives, whose holdings span centuries. How widespread these dossiers
1 Brown, ‘Documents’; Kosto, ‘Laymen’, esp. 60–2; H. Fichtenau, Das Urkundenwesen in

Österreich vom 8. bis zum frühen 13. Jahrhundert (Vienna, 1971), 38–45.
189
190 Hans Hummer
might have been is impossible to say. The surviving examples are limited
mostly to deeply Romanized areas, such as Rhaetia and Catalonia, and
perhaps speak to the greater importance attached to documentation in
particular regions.2
Or maybe not. The unusual survival of originals from St Gall and
Catalonia might have seen to the preservation of dossiers that otherwise
were present, but not included, in cartularies, the codices of property
transactions assembled by monks who were preoccupied mostly with
donations. We need not jump to the conclusion from this that cartularies
were mere reflections of ecclesiastical interests and therefore preclude
any insights into the participation of the laity in written culture. For
one, the abbots and monks came from many of the same kindreds who
patronized the monasteries, so that the ecclesiastical records encompass
within them the interests of their patrons. Indeed, donation charters
were overwhelmingly represented, not for material reasons per se, but
because they bore material witness to an individual’s or a family’s spiritual
relationship to the monastic institution. Consequently, the second most
common records in cartularies after donations are precarial documents,
which document the patrons’ requests for the continued use of donated
property. In some cases, the donation charter was not included at all, the
precarial record sufficing to bear witness to the gift.
Although the monastery was responsible for assembling cartularies,
many of the charters therein – even donation charters – were not authored
by monks. These documents often originated from beyond the cloister,
only later to be gathered up and organized by monastic copyists. More-
over, the notaries responsible for drawing up many of the transactions
between patrons and a monastery might have had no formal connection
to the monastery at all, an indication of a more dynamic, complex world
of document production than one might casually assume. Cartularies,
rather, should be thought of as something like a fossil, a static assem-
blage of bones which, upon forensic examination, reveals clues that allow
us to flesh out the processes responsible for the patterns before us. When
examined in this light, monastic charters provide tantalizing evidence for
the regular participation of the laity in the use, production and preserva-
tion of documents.
Crucial to recognize for the present analysis is the distinction between
the original production of a document and its subsequent preservation.
Cartularies represent acts of preservation. They are made up of copies
of charters: a team of monastic scribes made copies of loosely kept
charters, either original drafts of transactions, or copies of originals, at a
2 See above, Chapter 7, and below, Chapter 10.

Francia: cartularies 191
particular moment in time. This was in some cases only the beginning of
a process, the cartulary having been supplemented thereafter for several
decades. Cartularies must be distinguished from the records produced at
the moment of a transaction. While a cartulary might preserve only one
record of a transaction, sometimes the charters reveal that copies were
produced for the donor(s) too. As we shall soon see, cartularies in fact do
include multiple drafts of the same transaction, including copies which
by all accounts had once belonged to the lay patrons. These personal
records produced for the donors also represent acts of preservation in
that the patrons presumably took these copies away with them from the
transaction and kept them. This mode of preservation, however, was
mostly ephemeral. Deprived of the institutional structures which might
see to their preservation in a coherent way over time, these documents
vanished, unlike those kept by the more stable, ecclesiastical foundations.
This chapter will examine the intersections between lay and ecclesi-
astical record-keeping. It first will attempt to deduce from the evidence
the documents that laypersons might have possessed during the early
Middle Ages in Francia, and how their records might have come into the
possession of a monastery. It then will offer a chronological analysis that
will illuminate the formation of enduring ecclesiastical archives out of
the multipolar world of late Merovingian documentary culture. The pro-
found and sweeping transformation of monastic archival consciousness
in the late eighth century was responsible for the preservation and the
shape of much of the evidence that we have before us. Before we launch
this investigation, however, it will be helpful to say a few words about the
shape and scope of the evidence.
All of the extant early medieval cartularies stem from the eastern
regions of Carolingian Francia, having been compiled at the Rhineland
monasteries of Wissembourg in Alsace and Fulda in the Rhine–Main
region, and at the Bavarian church of Freising.3 In addition, vestigial
collections remain from the Danube-valley bishoprics of Passau, Regens-
burg and Salzburg, and the monastery of Mondsee.4 Why these collec-
tions should have arisen, by all appearances, exclusively in the east, has
3 TF; TW; Codex diplomaticus Fuldensis, ed. E. J. F. Dronke (1850; repr. Aalen, 1962)
(hereafter CDF, with document number); Urkundenbuch des Klosters Fulda, ed. E. E.
Stengel (Marburg, 1958) (hereafter UF, with document number).
4 TP; Die Traditionen des Hochstifts Regensburg und des Klosters S. Emmeram, ed. Josef Wide-
mann (Munich, 1943; repr. Aalen, 1988); F. Lošek, ‘Notitia Arnonis und Breves notitiae:
Die Salzburger Güterverzeichnisse aus der Zeit um 800: Sprachlich-historische Ein-
leitung, Text und Übersetzung’, Mitteilungen der Gesellschaft für Salzburger Landeskunde,
130 (1990), 5–193; Das älteste Traditionsbuch des Klosters Mondsee, ed. G. Rath and E.
Reiter (Linz, 1989).
192 Hans Hummer
remained a vexed issue. Briefly, they might represent efforts to consol-

idate proprietary interests in the wake of Carolingian conquest,5 but, if
so, the chronology does not always align well with the surviving patterns.
Bavaria, the argument’s favourite example, was conquered by Charle-
magne in 788, yet the cartulary of Freising, our most complete collection,
was assembled in the 820s, over three decades later, and those of Mond-
see, Passau and Regensburg even later in the ninth century.6 Alsace never
was ‘conquered’, and even if one allows that the area was ‘taken over’
by Carolingian kings, its incorporation took place a good century before
the compilation of the cartulary at Wissembourg around 860.7 While
the areas east of the Rhine were conquered in the late eighth century,
the cartulary of Fulda again was compiled well afterwards in the 820s.8
Moreover, we have areas in West Francia subjected to grinding conquest,
notably Aquitaine, none of whose monasteries produced a cartulary, so
far as is known. On the other hand, we can at least admit a connection
to Carolingian reform, itself an expression of Frankish consolidation,
which does provide a broader – if more diffuse – explanatory framework
for the appearance of cartularies.9 Indeed, the monasteries that produced
cartularies were bound up within the memorial network of monasteries
beginning in the latter eighth century and elaborated in the ninth,10 a
manifestation of the Carolingian ideals of Augustinian order, permanence
and social stability. This network also included western establishments,
none of which, however, bequeathed to posterity a cartulary before the
tenth century, although they did produce abbatial or episcopal chroni-
cles which memorialized proprietary holdings within a wider institutional
history.11 Thus, if, early on, cartularies were unknown in West Fran-
cia, property records did exist there, suggesting that if the assembling
of cartularies was a regional phenomenon, the patterns of lay partici-
pation in documentary culture that can be excavated from their con-
tents nonetheless illuminate practices and developments beyond eastern
Francia.
If we consider the greater cartularies from Fulda, Freising and Wis-
sembourg, we have about 1,450 extant charters from the period before
900. Six hundred of these hail from Fulda, but represent just two of
5 Geary, Phantoms, 88–92.

6 TF, pp. xx–xxi; Traditionsbuch . . . Mondsee, 36–8; TP, pp. xix–xx; Traditionen . . .
Regensburg.
7 Hummer, Politics, 56–75. 8 UF, pp. xviii–xxi. 9 Geary, Phantoms, 92–3.
10 Karl Schmid, ‘Wege zur Erschließung des Verbrüderungsbuches’, in Das
Verbrüderungsbuches der Abtei Reichenau (MGH Libri mem. n.s. 1:lx–ci, at lx–lxv).
11 Michel Parisse, ‘Les cartulaires: Copies ou sources originales’, in Les cartulaires, 503–11,
at 506.
the ten books of charters that made up Fulda’s original cartulary.12 A

digest of the Fulda cartulary produced in the high Middle Ages reveals
that these other eight books contained about 1,200 charters, bringing
the original total at Fulda to nearly 1,800.13 Similarly, the cartulary of
Wissembourg possesses 273 charters dedicated to property in the dis-
tricts to the south and west of the monastery in Alsace, the Saargau and
the Saulnois.14 There are good reasons to believe that Wissembourg pro-
duced a cartulary covering only these districts; nonetheless, other records
and surveys from the tenth and the thirteenth centuries reveal that dur-
ing the early Middle Ages Wissembourg lorded over a concentration of
mid-Rhine properties to the north and east which probably exceeded
those recounted in the extant cartulary.15 How numerous these might
have been is impossible to say, but we know from the testimony of a
seventeenth-century Jesuit who examined its now-lost contents that the
cartulary of the nearby monastery at Honau contained over a thousand
Carolingian-era charters.16 Thus, the charters in the cartularies of Fulda,
Wissembourg and Honau probably numbered at least around 4,000. The
cartulary of Freising contributes another 577 charters, but this figure rep-
resents only the records in the cartulary compiled by the priest Cozroh
between 824 and 848.17 The scribes of Freising were remarkably indus-
trious, compiling yet other collections of transactions down to 900; these
include about another 500 charters, many of them property exchanges,
bringing the total at Freising to over 1,000, and our grand total so far to
roughly 5,000.
This number can be augmented with the fragmentary or partial col-
lections from Murbach in Alsace, Mondsee, Passau, Regensburg and
Salzburg, which run to almost 470 extant charters and whose original
scale presumably matched those of the surviving greater collections.18
And we have not even considered the vestigial records of numerous
12 UF, p. xix, and E. Stengel, Abhandlungen und Untersuchungen zur hessischen Geschichte
(Marburg, 1960), in particular the articles, ‘Über die karlingischen Cartulare des
Klosters Fulda (Fuldensia II)’ and ‘Untersuchungen zur Frühgeschichte des Fuldaer
Klosterarchivs (Fuldensia IV)’ (147–93, 203–65).
13 Der Codex Eberhardi des Klosters Fulda, ed. H. Mayer zu Ermgassen, 2 vols. (Marburg,
1995–6).
14 TW, pp. 40–4.
15 Hummer, Politics, 181–90; W. Metz, ‘Das Kloster Wissembourg und der Vertrag von
Verdun’, in C. Bauer, L. Boehm and M. Müller (eds.), Speculum Historiale: Geschichte
im Spiegel von Geschichtsschreibung und Geschichtsdeutung (Freiburg, 1965), 458–68.
16 C. Wilsdorf, ‘Le monasterium scottorum de Honau et la famille des ducs d’Alsace au
VIIIe siècle: Vestiges d’un cartulaire perdu’, Francia, 3 (1975), 1–87, at 11.
17 TF, pp. xx–xxi.
18 See above, n. 4, and for Murbach, A. Bruckner, Regesta Alsatiae aevi Merovingici et
Karolini (496–918), vol. I (Zurich, 1949).
194 Hans Hummer
monasteries, which surely amounted to more than now exist; or the

roughly 3,000 abbreviated charters from Lorsch, which, unfortunately,
because the copyists of the high Middle Ages left out the wider social
circle of witnesses and notaries as they abridged the documents, are less
useful for the current purposes.19 Nor have we considered the western
establishments, whose records in that more document-rich area prob-
ably were more voluminous than now exist, but whose holdings we
know of from a smattering of charters included in later medieval col-
lections and from aggregate summaries of property in res gesta, rather
than cartularies.20 Let it not be said that early medieval Europe had ever
suffered a poverty of records.
These figures, naturally, reflect what monasteries held, and they also
reflect particular institutional imperatives; that is to say that collections
of charters were not intended to keep track of just any old record having
to do with the cloister. Cartularies were devised to memorialize the gifts
made to monastic foundations for salvific purposes, and therefore they
necessarily bear witness to the network of patrons bound to the eternal
‘family of God’.21 In addition to their deliberate obliviousness to records
unrelated to the cloister, cartularies seldom even bother with property
records that the monastery surely possessed in some abundance, but
which have to be guessed at from notarial formulas or glimpsed in stray
entries in cartularies: the monastery’s own more purely economic wheel-
ing and dealing for property, the prestarial grants of property in response
to patrons’ precarial requests (which are more likely to be preserved,
but still in numbers much reduced by comparison with donations), the
drafts of donation and precarial charters drawn up for patrons, and the
sales of property between laypersons. These records of sales between
laypersons probably were transferred to the monastery upon the dona-
tion of property previously purchased by the donor. Indeed, the donation
of property is usually called a ‘handing over’ (traditio) of property, a pro-
cess which might include ritual actions or the transfer of objects (though
these are only intermittently attested in formulas and even less frequently
in ordinary charters),22 and perhaps the simultaneous conveyance of any
19 Codex Laureshamensis, ed. K. Glöckner, 3 vols. (Darmstadt, 1929–36).

20 Cf. M. Sot, Gesta episcoporum, gesta abbatum (Turnhout, 1981).
21 Cf. Geary, Phantoms, 81–114.
22 These would be the festuca (rod), wadium (gage) or andelangus (handshake?, see F.
Staab, Untersuchungen zur Gesellschaft am Mittelrhein in der Karolingerzeit (Wiesbaden,
1975), 259n494). They are attested to in Merovingian and Carolingian-era formularies.
Festuca: Marculf 1.13, 21; 2.14; Form. Sen. 7, 29, 34; Form. Sen. Rec. 2; Form. Sal. Merk.
21, 27–8; Form. Sal. Lind. 1. Festuca and wadium: Marculf 2.18. Wadium: Form. Tur.
6, 32; Form. Sal. Merk. 29; Form. Sal. Lind. 19. Andelagnus and wadium: Form. Sen.
50. Festuca, wadium and andelangus: Form. Sal. Lind. 2, 6–8, 12, 14, 16, 18. However,
prior records. For example, in 754 at Thulbach in Bavaria, a certain

Timo received the epistula of his donation and at the altar handed it over
into the hands of Bishop Joseph of Freising.23 Glimpses like this might
explain why the documentary form itself, the carta donationis, could
also be called a traditio.24 A transfer of documents would, at any rate,
explain the otherwise odd appearance of records between laypersons in
cartularies.
Let us now turn to particular examples to get a better grasp of the circu-
lation of documentation among laypeople, as we can perceive it through
ecclesiastical collections. The few purely lay transactions included in car-
tularies involve mostly sales. A remarkable example in the Wissembourg
corpus involves two drafts of a sale in 736×7 between two laymen, the
purchase of property by a certain Rantwig, son of Chrodwig, from the
Etichonid duke Liutfrid.25 These properties, which had been acquired
by Liutfrid from several sources – from his father, from a certain Liudulf
and by uendicionis titulum (‘title of sale’) from a woman named Ingina –
had been granted in benefice by Liutfrid to Rantwig’s father. Both copies
were made from two different drafts of the sale, which presumably found
their way into Wissembourg’s monastic archives with Rantwig’s impres-
sive donation of property in twenty-nine locations in 742,26 a gift which
included the previously purchased properties. The two drafts of his ear-
lier purchase were then copied into the cartulary when it was compiled
around 860.
We might assume that the two drafts of this sale came to Wissembourg
within the context of Rantwig’s donation in 742, and presumably were
copied into the cartulary because of the wealth of the donation, the
prominence of its actors, and the knot of feudal, familial and personal
claims previously bound to the property. But to whom had these drafts
the andelangus appears just twice in the Wissembourg charters: TW 233, 237; a wadium
once in the charters of Passau, TP 78, and twenty-nine times out of about 700 charters
of Freising, TF 91, 185, 195, 197, 227, 235, 251, 258, 288, 299, 324, 328, 338, 345,
351, 358, 364, 378, 390, 400, 412, 435, 470, 502, 507, 509, 543, 549, 704. The festuca
appears in none of these collections, though it does appear later in the charters of Cluny
(see below, 308–9), as does the andelangus, albeit rarely; e.g. Cluny 738. Neither the
festuca, the wadium nor the andelangus appears among the charters in the Fulda cartulary.
23 TF 7: ‘Acta haec epistula accepit eam ipse Timo propriis manibus coram testibus seu
ceteris adstantibus atque haec consentientibus involuto palleo ipsam epistulam super
altare posita tradidit in manus supradicti venerabilis Joseph episcopi.’
24 Notaries among the Wissembourg charters, for example, use both terms interchange-
ably. Guntbert says that he ‘wrote this charter of donation’ (‘hanc carta donationis
scripsi’) in TW 171. In TW 175, Guntbert substituted traditio for carta donationis: ‘hanc
traditionem scripsi’. For good measure, just above this notarial subscription, the donor
Erhart says he had ‘caused this carta donationis to be done and affirmed’, thus under-
scoring that traditio in this context was the semantic equivalent of carta donationis.
25 TW 35, 162. 26 TW 52.
196 Hans Hummer
once belonged? Some of the variation in the two copies of Liutfrid’s

sale to Rantwig can be attributed to the copies made when the cartulary
was assembled, but more substantial deviations indicate that each was
descended from separate documents made at the time of the sale between
Rantwig and Liutfrid.27 Presumably, one of these had been Rantwig’s
record of the transaction, which was transferred to the monks upon his
impressive donation of 742. But what of the other? Was that Liutfrid’s? If
so, how had his draft made its way into the monastic archive? And were
these records removed from these individuals to the institution, or did
the institution have copies made from documents which the individuals
continued to keep?
Moreover, what had been the fate of the records of the earlier sale,
of the uendicionis titulum, between Liutfrid and Ingina, alluded to in
the sale between Liutfrid and Rantwig? Presumably because they did
not directly impinge on Rantwig’s donation of 742, these records never
were handed over to the monastery when Rantwig made his donation.
Copies apparently would have been held by Liutfrid and Ingina, neither
of whom, nor their descendants, would have felt compelled to preserve
their uendicionis titulum once the properties had been sold to Rantwig.
That is, the case points to the one-time existence of deperdita, of now-lost
records behind the documents we possess, as well as the likely reasons
for their ephemerality.
Rantwig’s donation, and the details of the earlier proprietary history
of some of his lands, which exist only because the monks happened to
include Liutfrid’s sale to Rantwig, hint at a much busier world of doc-
umentary give and take than the surviving materials normally indicate.
Rumour of this activity is not limited to the records of the entitled such
as Liutfrid; it also reverberates in the donations of less exalted figures
who also refer to the earlier purchase of the property from another
layperson by ‘title of sale’.28 In addition to sales, a donation char-
ter might exclude properties granted previously to the donor’s wife in
libeldote, an apparent reference to the libellum dotis of Frankish dowry
27 For example, both copies refer to Liutfrid in Merovingian-era usage as vir inluster, but
TW 35, produced by copyist B, who otherwise has a tendency to compress the text,
included the title ‘dux (Fratri Rantuuigo, uir inluster Liutfridus dux ui[n]de)’, whereas
the draft produced by copyist A, TW 162, left out the ducal title, an unlikely omission
considering the fuller style of the document from which the scribe had copied (‘Fratre
Rantuuigo emptore, vir inluster Liutfridus uenditor’). This kind of deviation indicates
that the copies had descended from two separate documents, cf. TW, pp. 213–14 and
nn. 162a, 35b.
28 TW 232, 233.
formulas.29 These records have not survived, either because they were
present in the monastic archive but not preserved for posterity in the
cartulary, or because they remained in the possession of the layperson
who lacked the means (or the intention?) to preserve them beyond his or
her own lifetime, or the lives of their immediate heirs. That they could
survive in abundance can be perceived at the monastery of Cluny, which
preserved – independent of its eleventh-century cartulary – a trove of
transactions, many of them sales between laypersons, that took place in
the century before its foundation in 909.30
We can augment these clues about the documents that laypersons must
have once possessed with several dozen instances from the codices of Wis-
sembourg, Fulda and Freising, where two drafts, and in two cases three,
of particular transactions were preserved.31 These offer tantalizing clues
about documentary practices because, as we have seen with the drafts
of Rantwig’s purchase, the deviations among pairs or triplicates seem to
spring from the separate copies issued to both the monastery and the
patron, whose version sometimes made its way back to the monastery
and thence into a cartulary. That is, the monks had copied into the car-
tulary a record which had once belonged to the lay actor. Many of these
double drafts from Wissembourg have to do with precarial transactions,
whose patron copies perhaps had been returned to the monastery upon
the expiration of the lease.32 If so, this would have required the precarist’s
heirs to produce the documentation, in some cases many decades later,
a process that assumes the keeping of clutches of charters by patrons.
However, these lifelong leases were usually – where the evidence permits
us to see it – assumed by the precarists’ heirs, who presumably would
have kept copies for themselves as evidence of their rights, rather than
returning the record to the institution. If the monastery did cull precarial
records directly from families, we might then assume the persistence of
copies of at least precarial documents among the patrons.
We do encounter multiple drafts of donations in the Wissembourg and
Freising codices that were made up to a century before their cartularies
were compiled. This leaves dangling the question of how the patron’s
29 TW 128, 53 = 178; cf. Marculf 2.15. In another case, TW 222, a certain Sigoin gave
property inherited from his mother, who had in turn received the property ‘in dotes
[titulum]’ from her husband.
30 See below, Chapter 11.
31 Wissembourg: TW 8/47, 17/159, 26/105, 35/162, 53/178, 110/154, 198/251, 204/254,
205/223/252, 194/224, 218/239, 245/250. Freising: TF 45, 76, 121, 127, 184, 193,
232, 251, 418, 464, 568, 680. Fulda: UF 24, 86, 163–4, 187, 190, 232, 237, 248, 264,
268, 279, 285.
32 Or so the editors of the cartulary, Glöckner and Doll, consistently assume; see TW 8/47,
17/159, 26/105, 35/162, 53/178, 110/154, 204/254, 205/223/252, 218/239, 245/250.
198 Hans Hummer
copy of the donation had been acquired by the monastery. That copies
of transactions were handed down within families is possible, though the
surviving lay dossiers suggest that a century-long persistence of records,
at least as a coherent group, is unlikely. Moreover, the well-established
cognaticism and partible inheritance practices of early medieval fami-
lies presumably would have seen to the dispersal of family records from
generation to generation.33 Exceptions might have been records of trans-
actions with ecclesiastical institutions, particularly, as we have seen, those
held in precaria. These tenures, rendered indivisible by the very nature of
the grant, allowed individuals to hand down and receive blocks of prop-
erty from generation to generation.34 Even so, the extant lay dossiers do
not show the handing down even of precarial documents over such long
periods of time.
It may be that patron copies of charters made their way indirectly
into monastic archives through estate or village churches, many of which
had been donated, and continued to be controlled as precariae, by patron
groups. The extant donations make clear enough that many of these local
churches had been founded by aristocratic kindreds, were supported with
their donations, and then eventually were handed over as gifts to powerful
monasteries. Might these parish churches have served as local centres of
estate records? The monastery of Schlehdorf in Bavaria is perhaps the
most outstanding example. Founded and administered by members of
the Huosi family, its records were apparently transferred en bloc when the
foundation was turned over to the see of Freising.35
The experience of the Huosi was replicated by humbler, middling
patrons in village churches elsewhere. In 714, a certain Nordolf gave
part of his property to the basilica of St Martin in the mid-Rhine vil-
lage of Edesheim.36 He then gave half of the basilica to the monastery
of Wissembourg and had an ‘epistle of donation’ made out for both the
monastery and the local church (‘propterea tibi hanc epistolam dona-
tionis fieri rogaui vel ad partem ipsius basilicae’), which he says had
been built by his ancestors (‘apud antecessores meos constr[u]cta esse
33 On Frankish kinship, see K. Schmid, ‘Zur Problematik von Familie, Sippe und
Geschlecht, Haus und Dynastie beim mittelalterlichen Adel’, Zeitschrift für die Geschichte
des Oberrheins, 105 (1957), 1–62, and A. C. Murray, Germanic Kinship Structure: Studies
in Law and Society in Antiquity and the Early Middle Ages (Toronto, 1983).
34 See the remarkable case of the Rodoins in Hummer, Politics, 71–5, 115–27.
35 J. Jahn, ‘Virgil, Arbeo und Cozroh: Verfassungsgeschichtliche Beobachtungen an
bairischen Quellen des 8. und 9. Jahrhunderts’, Mitteilungen der Gesellschaft für Salzburger
Landeskunde, 130 (1990), 201–91, at 251–7. On the Huosi, see W. Störmer, Adelsgrup-
pen im früh- und hochmittelalterlichen Bayern (Munich, 1972), 91–112. For the phe-
nomenon, compare below, Chapter 10.
36 TW 41.
uidetur’). The cartulary of Wissembourg also contains several donations

made not to the monastery, but directly to one of the monastery’s village
churches by members of the Ratbald-Wicbald kin group, which pos-
sessed an array of properties in northern Alsace and in the mid-Rhine
region, as well as connections to Wissembourg’s abbot, Ermbert, and the
founders of the powerful monastery at Lorsch.37 In 766, a woman named
Richsvind gave a contiguous field marked off by stones ‘to the most holy
church of St Peter which was built in the mark Uhlweiler, where at this
time in the name of God Abbot Ermbert [of Wissembourg] is seen to
preside’. Likewise, five years later, her kinsman Sigibald donated a vine-
yard directly ‘to the most holy church that was constructed in honour of
St Peter the apostle in the place that is called Uhlweiler’.38
A donation to the church of St Mary in the Alsatian village of Dauen-
dorf offers a particularly striking example. In 774, its presiding lord and
priest Ado, who was closely connected to the Ratbald-Wicbalds, gave his
share of the church to Wissembourg.39 At some point on either side of
Ado’s donation, a group of twelve petty patrons made an undated dona-
tion to Ado and the church of St Mary.40 Its editors date the charter to
anywhere between 760 and 790, reasoning that Ado had retained control
of the church after his donation and so could have been the recipient
of any gifts made after 774. As the presiding cleric, he no doubt did,
but it seems much more reasonable to assume that any gifts made after
774 would have been addressed to Wissembourg’s abbot (as was the case
with Richsvind’s donation to St Peter’s at Uhlweiler), rather than to Ado,
which would place the donation prior to 774. Whatever the case, the
charter was directed to the parish church and its pastor: ‘sacrosancta
ecclesia sancta Maria quod est constructus in marcha Dachunheim ubi
in dei nomine Ado pres[b]iter preesse uidetur’. This formula looks very
much like those we encounter in the gifts to monasteries which dominate
the extant record, except that the local church was the recipient of the gift
and presumably the charter too, an indication that the records had once
been housed in Dauendorf and then later transferred to Wissembourg,
presumably after Ado had donated the church.
The donation of property at Berg and Waldhambach on the other
side of the Vosges yielded a cluster of documents whose patterns can be
explained most easily if one assumes their intermediate preservation at
37 W. Alter, ‘Studien zur mittelalterlichen Siedlungs- und Volksgeschichte der mittleren

Vorderplatz. II. Teil: Die in den Klosterkodizes genannten Personen, insbesondere die
Angehörigen der Familie Ratbald-Wicbald’, Mitteilungen des historischen Vereins der Pfalz,
57 (1959), 39–135; cf. Hummer, Politics, 111–13.
38 TW 108, 189. 39 TW 71. 40 TW 181.
200 Hans Hummer
an estate church. On 27 June 717, a certain Chrodoin, son of Peter, con-

firmed an earlier donation of property to Wissembourg.41 This donation
had included an estate in Waldhambach and the basilica of St Martin at
Berg, as well as the presiding priests, who were listed among the serfs
as if they were servants. Chrodoin received the properties in precaria, as
did his descendants for the next three generations down to at least 830,
and perhaps longer.42 This precaria was reworked three more times over
the next year as Chrodoin elaborated his donation into a more compre-
hensive will.43 These revisions reveal that Chrodoin’s ancestors (plural)
had previously donated property to the church of St Martin at Berg, an
indication that the church had been under family control for at least two
generations prior to Chrodoin’s donation. When we combine Chrodoin’s
testimony with the trail of records attached to his descendants, we can
perceive that the properties had been held for at least six generations,
from the late seventh to the mid ninth century.
The drafts of Chrodoin’s will suggest that the transaction had gener-
ated records for both the family and the monastery. Chrodoin’s precariae
and wills were drawn up by a scribe named Leudoin, who identified
himself variously as an emanuensis and a cancellarius emanuensis. More-
over, Chrodoin’s final will exists in two drafts, one made by Leudoin
and a second by yet another cancellarius emanuensis, Guntbert, neither of
whom was a monk of Wissembourg.44 These two drafts, each produced
by a different notary, suggest copies issued separately to Chrodoin and
to the monastery. How had a version made for Chrodoin made its way
to Wissembourg? That it was returned upon the expiration of the lease,
as the editors routinely assume in the cases of other doubles,45 seems
doubtful when we consider the long persistence of this family’s precar-
ial control of the properties well into the ninth century. Moreover, the
copy in question applies to the will, which was separate and distinct from
the precaria, which was drawn up several months later. We perhaps need
look no further than the church of St Martin at Berg as the repository of
Chrodoin’s copy of his will, and perhaps some of the earlier workings of
the will, especially when one ponders, as we shall soon see, the relative
underdevelopment of monastic record-keeping prior to the late eighth
century.
In one case, we possess three copies of a transaction: one belonging to
the monastery, and the other two to the joint donors. In 699, the brothers
Ermbert and Count Otto made an extensive donation of properties to
Wissembourg in the Saulnois, in the upper Moselle region just south
41 TW 196. 42 Hummer, Politics, 118–20. 43 TW 194/224, 195, 227.

44 TW 194/224; on the two notaries, see TW, pp. 126–8. 45 See above, n. 32.
of Metz.46 Three exemplaria of this transaction were drawn up by a lay

notary named Chroccus,47 one of which was issued to the monastery of
Wissembourg. The other two were dubbed, by the scribes who produced
the cartulary over a century and a half later, the ‘exemplaria Ottoni atque
Eremberti’, the copies belonging to Otto and Ermbert. The gift included
several basilicas established by their ancestors and dedicated to St Martin,
St Hilarius and other saints. These copies might have been acquired
by Wissembourg from these ‘basilicas’, or perhaps from the church at
Biberkirch, where some of the donated properties lay and where twenty
years later they were received in precaria by a Count Adalhard.48 Again,
we encounter a node of documents that cluster around a local church.
Taken together, the examples of the Huosi, Nordolf, Richsvind,
Sigibald, Ado, Chrodoin, and Otto and Ermbert strongly hint that char-
ters could be regularly deposited and stored in local churches, where
they might be culled both by the locally powerful, who by all appear-
ances had organized themselves around estate churches, and by the par-
ent monastery. I consider it significant that almost all of the charters that
come down to us in multiple copies in the codices from Freising and
Wissembourg deal with donations of property in villages known to have
had a suffragan church.49 The vestiges of evidence teased out here find
vivid elaboration in the astonishing evidence of the Folkwine dossier,
which, prior to its transfer to St Gall, seems to have been housed in the
local church at Rankweil.50
At the monastery of Fulda east of the Rhine, we are confronted with
contrasting circumstances. Among the Fulda charters are twenty multiple
drafts, all but one of which appear to descend from separate copies
produced at the time of transaction.51 Of these twenty, eight date to
46 TW 223/205/252. 47 TW, pp. 118–19. 48 TW 267.

49 Nine of twelve from Wissembourg and eleven of twelve from Freising (cf. above, n.
31); but for the churches associated with properties in TW 26/105, 35/162, 204/254,
218/239, see TW 52–3, 151, 156, 200. The exceptions are TW 8/47, 110/154, 245/250,
although the first one was transacted in Durstel, where Wissembourg had a suffragan
church (TW 198/251), and dealt with properties located near local churches in nearby
Modern and Kirrwiller (TW 60, 151, 156, 172). The second deals with property in
Lembach, where Wissembourg held property from a dense cluster of patrons, which
must have had a parish church, if the number of donors is a fair guide (TW 29, 49,
71, 76–7, 79–80, 82, 89, 93–6, 98–101, 107, 109–10, 122, 136–7, 144, 148, 154–5,
157, 164, 170). With respect to Freising, a local church does not appear to have been
involved in the transmission of TF 232, though the two drafts of this transaction record
a plea to Charlemagne for the return of property seized by a count, a case that was
delegated to the royal missi, whose court records might have been the source of a second
draft.
50 See above, 165–73.
51 UF 24, 86, 163/164, 187, 190, 232, 237, 248, 264, 268, 279, 285. For the char-
ters after 802, Dronke’s edition must be consulted (CDF). However, Dronke did not
202 Hans Hummer
between 824 and 876. These later transactions deal with property well
east of the Rhine in the Saale basin and were conducted during the
time when the cartulary was conceived in the 820s and then periodically
augmented throughout the rest of the ninth century. Some of the putative
patrons’ drafts of these eight transactions could have been derived from
local churches, since a local monastery or church was implicated in half of
them.52 On the other hand, because these transactions transpired as the
cartulary was being compiled, and because half of them bear no evidence
of the involvement of a village church, it could be that these patron drafts
were culled immediately upon the transaction and rapidly entered into
the codex. That is, the second drafts of these latest eight transactions
might have been added directly to the cartulary without the mediation of
any local archive.
The remaining twelve cases where we have multiple drafts, however,
transpired between 754 and 802, and so date to the period well before
the cartulary was conceived. In these cases, the patrons’ drafts presum-
ably made their way into the monastic archives indirectly. Two involve
property from the Saale basin, one of which records the gift in 795 of
property in several locations, among them Nordheim.53 Although the
charter does not mention a church, we know from later records copied
into the cartulary that Nordheim served as the locus of several transac-
tions and that its church eventually was handed over in 836 to Fulda.54
The other involves the property of the suffragan monastery Milz.55 This
nunnery was given to Fulda at Milz in 799 by Abbess Emhilt, and then
returned to her as a precarial grant. The donation/precaria was recorded
in the Fulda cartulary, as was a second draft by the monk Eberhard in the
twelfth-century Codex Eberhardi. Eberhard had devised his draft from a
record of the transaction, whose formulas suggest derivation from a copy
edit duplicates, which still have to be examined in the ‘Traditionum Fuldensium Libri
Tres’, published by Johann Pistorius, Rerum familiarumque Belgicarum chronicon mag-
num (Frankfurt, 1654), 445–588. In the following, charter headings are from Dronke’s
edition, the numbers in brackets refer to the order in Dronke’s edition, and the page
numbers are to the location of the duplicates in Pistorius, ‘Traditionum Fuldensium’:
Traditio Rihbrahti [562], pp. 495, 496; Traditio Germunti [455], pp. 495, 542; Traditio
Egilharti [598], pp. 495, 587; Traditio Gotehrammi [461], pp. 534, 545; Traditio Egilmari
et Meginolti [462], pp. 534, 545; Traditio Uuitgeri [411], p. 543; Traditio Reginhartes
[588], pp. 566, 489; Traditio Cunihiltae [611], pp. 569, 574. Of these, only the Traditio
Rihbrahti is a verbatim copy.
52 Traditio Rihbrahti [562], pp. 495, 496; Traditio Germunti [455], pp. 495, 542; Tradi-
tio Egilharti [598], pp. 495, 587; Traditio Uuitgeri [411], p. 543. The first three were
transacted at the suffragan monastery of Rore. For the church at Sundheim, see CDF
323.
53 UF 232.
54 CDF 423–5. On the donation of the church at Nordheim, see CDF 493. 55 UF 264.
separate and distinct from that contained in the cartulary,56 presumably

a version that once had belonged to Milz. Emhilt’s precarial request turns
out to have been derived from an earlier precarial arrangement she had
contracted with Milz in 784, when the nunnery was still an autonomous
cloister.57 As we have seen in cases from Freising and Wissembourg, that
this earlier precarial request was directed to Milz indicates that the record
had once lain at this modest cloister. Presumably, a copy of Emhilt’s pre-
caria of 784 was transferred to Fulda from Milz upon Emhilt’s donation
fifteen years later, or shortly thereafter. Although the case of Milz does
not involve a layperson, the fate of the documentation underscores the
importance of humbler, local institutions in the production, transfer and
preservation of documents.
Of the remaining ten cases where we have multiple drafts of a trans-
action from Fulda, eight are traceable to the Wormsgau and one to the
diocese of Strasbourg; the last deals with property in Alsace but was
preserved in a section of the cartulary devoted to transactions in the dio-
cese of Strasbourg.58 Five of the ten cases appear to have been drawn
up by non-monastic notaries: one each at Strasbourg, Mainz and a royal
assembly in Paderborn, and two whose records do not reveal the place of
transaction.59 However, because these last two record gifts of property
in Mainz in 797 and 802, and because both were drawn up by non-
monastic scribes, Mainz would appear, as we shall soon see, to be the
likely place of conveyance. The other five were drawn up by monastic
notaries: three were conducted at Fulda, while the other two do not
reveal a place of transaction, though Fulda would be a likely location,
as we shall see later, given their drafting by monastic scribes.60 Of the
three explicitly done at Fulda, two deal with property in Dienheim, near
Mainz, where Fulda held an impressive concentration of property, while
the other recounted the gift of property located just beyond the walls of
Mainz in 791.61 The remaining two, which do not mention a place of
transaction, deal with property in villages near Mainz. In sum, in each
of these ten cases, whether the transactions were drafted by monastic or
non-monastic notaries, we are dealing with multiple drafts that share a
connection to Strasburg or to Mainz and its environs.
56 UF 375. 57 UF 154.
58 UF 24, 86, 163/164 (Strasbourg), 187 (Strasbourg), 190, 237, 248, 268, 279, 285.
59 UF 24 (Mainz), 187 (Strasbourg), 163/164 (Paderborn), 248, 285. On the non-
monastic origins of the notaries, see UF, pp. liii–lix.
60 UF 86, 190, 237, 268, 279.
61 UF 190, 237, 279; on Dienheim, see U. Weidinger, Untersuchungen zur Wirtschaftsstruk-
tur des Klosters Fulda in der Karolingerzeit (Stuttgart, 1991), 124–5.
204 Hans Hummer
The records issued to the monastic party at the time of the transaction
account for one of the two drafts, but where had the other drafts, appar-
ently those taken away from the transaction by the patrons, lain before
they were acquired by Fulda? In none of these ten cases can we identify a
local church, as in the cases of Nordheim and Milz. The prominence of
non-monastic scribes in half of these cases, and the location of donated
properties in Strasbourg and Mainz, or in villages very near them, sug-
gests the mediation of diocesan archives, which certainly existed. This
supposition finds support in the close connections between Fulda and
Mainz stretching back to the days of Boniface and his disciples Lull and
Sturm, a relationship materially attested by the repose of Archbishop
Boniface’s relics at the monastery. One can imagine any number of sce-
narios whereby documents could have passed between Fulda and the
episcopal church of Mainz. More specifically, the very organization of
the cartulary suggests the influence of the diocesan notariate on many
of the charters. Edmund Stengel, who edited Fulda’s charters, observed
that, at least where it pertained to the monastery’s Rhineland posses-
sions, the organization of properties according to Gau, or district, actu-
ally breaks along the underlying diocesan boundaries of Mainz, Worms
and Strasbourg. He noticed that the notaries responsible for many of the
earlier transactions were not monks and their field of activity was cir-
cumscribed by diocesan boundaries.62 Stengel’s observations have since
been taken up by Franz Staab, whose more thorough examination of
the charters from the mid-Rhine region has identified a Mainz notariate,
which operated under the auspices of the archbishop and authored many
of Fulda’s earliest charters.63
We might reasonably posit, then, the agency of diocesan archives –
by which I mean either the cathedral archive or the plethora of smaller
archives in parish churches subject to diocesan supervision – in the preser-
vation of some of the multiple drafts that once had been issued to patrons.
But other possibilities remain. Some of the drafts are within three decades
of when the cartulary was begun and could have been acquired directly
from families. Yet others might have been housed in villages sophisti-
cated enough to have developed an institutional consciousness, and a
local archive to go with it. Dienheim, for example, appears variously in
the Fulda charters as an oppidum, a trading centre along the Rhine, a
vicus publicus, the location of a via publica, the site of a complex of royal
estates, and a place for the performance of transactions.64
62 UF, p. liv.
63 Staab, Untersuchungen, 137–53, esp. 144–52; cf. Innes, State and Society, 114–17.
64 UF 235, 246; CDF 217, 250–1, 264, 281, 328.
One case from Fulda hints at the production of documents to meet a

range of constituencies. In 785, an Alsatian lord named Hugo made a
donation to Fulda of properties near Ribeauvillé. The gift was transacted
at the royal assembly in Paderborn, drawn up by a public notary, and
entered into the Fulda codices in three versions.65 One was labelled an
exemplaria, in this case meaning a copy of an original charter, presumably
the one held by Fulda. Although most copies differ in wording, pointing
to the existence of multiple drafts, the other two versions of Hugo’s trans-
action deviate in their substance in ways that speak to different purposes
and audiences. By contrast with the exemplaria, absent in the two other
versions, is a list of the mancipia attached to the properties and, among
the witnesses, a certain Eberhard, by all appearances Hugo’s son, ‘who
consented and subscribed’. The comprehensive listing of mancipia and
the assent of Eberhard, the person most likely to challenge the gift, make
sense in a copy destined for Fulda. The second and third versions, rather,
include provisions for Hugo’s man Badurich, which are missing in the
first. The third version twice mentions the property Badurich held ‘in my
benefice’. Was this Hugo’s copy? The second, which includes a provision
allowing the successors of ‘my man Baturic’ to retain the benefice, appar-
ently was devised to spell out the conditions for Badurich. Had Badurich
retained this copy? Ostensibly, all of these were drawn up at Paderborn,
though it could be that the latter two were drafted shortly thereafter to
spell out the conditions for Badurich, whose interests perhaps had been
overlooked in the original transaction. It might also be that all three of
these copies were issued to the monastery as the conditions were fine-
tuned, since some notarial formulas claim that duplicates were made so
that each party would have a copy, though even that would suggest the
production of documents for up to three parties, including perhaps a
lowlier henchman.
Whatever the case, these three charters point to a busier and more
sedimented process of documentation than normally is assumed from
the extant record, dominated as it is by the cartularies. In 820, another
Alsatian lord named Hugo, Count Hugo of Tours, exchanged his proper-
ties in Niederbronn, Preuschdorf, Walf, Barr and Froscheim for property
held by Wissembourg in Dettweiler.66 The exchange was made in front
of Louis the Pious at the imperial assembly in Quierzy and, as befitted
a transaction of one of the most powerful men in the empire, was wit-
nessed by a formidable assemblage of counts, abbots, bishops and other
notables, presumably members of Hugo’s retinue. After this witness list,
however, were entered groups of figures from each of the communities
65 UF 163–4. 66 TW 69.
206 Hans Hummer
who had borne witness to the on-the-spot transfer of possession: ‘Now

the witnesses who saw the investment of this traditio, first in the villa that is
called Niederbronn . . . Then in Preuschdorf . . . Then the witnesses con-
cerning Walf . . . At Barr as well as at Froscheim . . . ’ At Niederbronn,
the document explains that Ratbraht and Uodilo, two figures who appear
among the untitled notables at Quierzy, ‘gave that investment’; that is,
apparently, they had been dispatched by Hugo to supervise the transfer
of his property to Wissembourg.67 It might be that the charter of the
transaction conducted at Quierzy had been carted around to these vil-
lages by Hugo’s agents and supplemented with additional witness lists
as the transfer was affirmed by the locals, whose assent apparently was
crucial for validating the swap. On the other hand, it seems more likely
that brief records of local validation had been drawn up, which were then
combined with the record of the transaction at Quierzy and fused into
the composite document now visible in the cartulary.
When augmented with the other examples from the charters of Fulda,
Freising and Wissembourg, the records generated by the two Hugos
of Alsace attest to a much more dynamic documentary culture than the
static and binary impression of patron giving and ecclesiastical documen-
tation and preservation given by the extant records. The forensic evidence
suggests, rather, a regular flow of documentation among individuals, and
between patrons, on the one hand, and local and regional institutions on
the other. Let us now consider how the relationship between monasteries
and this wider documentary culture unfolded over time.
Record-keeping habits in Francia prior to the eighth century, when
monastic scriptoria became dominant, presumably were bound up with
the processes of public authentication and documentation mediated by
the gesta municipalia, at least in some regions. This has to be inferred
from admittedly sparse documentation, but the chapters in this volume
by Everett and Brown suggest that the documents of both individu-
als and ecclesiastical institutions were registered through these obscure
civic institutions for the purposes of taxation and public validation in
Italy and West Francia down to the early seventh century.68 Brown
examines in addition the puzzling references to the gesta municipalia in
West Frankish formulas from the eighth and ninth centuries; he con-
cludes that these late attestations of gesta municipalia probably refer to
67 Ratbraht appears in the Quierzy witness list as the linguistic equivalent Ratbert. The
element -braht appears variously in early medieval sources as -bert, -berht, -breht and
-brecht.
68 See above, Chapters 4 and 5; Barbier, ‘Pouvoirs et élites’, Part I, Chapter 3, 77–80;
Wickham, Framing, 110–11.
some sort of public validation ceremonies, if not to the municipal institu-

tion of Late Antiquity.69 This conclusion is strengthened by a surprising
reference to the gesta municipalia in an actual transaction to the east in
the Vosges region. In his will of 735×7 to the monastery of Murbach,
Eberhard, the Etichonid count in Alsace, claims to have ‘decided by all
means and taken care not at all with disregard to enter the present dona-
tion into the gesta muncipalia, lest anyone at any time should reclaim it
on account of this case’ (‘presentem vero donationem nequaquam vil-
itate gestis minicipalibus alligare curavimus et omnino decernimus, ne
aliquando in eam ob hunc casum quisquam valeat repetere’).70 Horst
Ebling, perplexed, surmised that ‘the phrase of allegation was taken
over only as a formula, which in any case only appears meaningful in
Burgundy’, where the Etichonid family had originated.71 Ebling appar-
ently was thinking of the formula from Marculf 2.3: ‘We took care and
wholly decided by no means to have the present donation entered into
the public archive by contemptible decurions, so that no one should
be able to gain access to it at any point as a result of this’ (‘presentem
vero donationem nequaquam a curalium vilitate gestis municipalibus
alligare curavimus, ne aliquando in eam ob hoc casu quisquam valeat
repperire’).72
This does approximate the phrasing in Eberhard’s will, but if the author
of Eberhard’s charter was familiar with Marculf’s formula, he adapted it
to his own purposes and times, transforming its meaning so that Eberhard
spoke not of avoiding ‘contemptible curials’, but rather of having taken
great care to have the document validated by the gesta municipalia. Nor
does Eberhard speak of restricting access (repperire) to the document,
but rather of the threat that someone might attempt to relitigate (repetere)
the donation. The absence of any reference to curials certainly makes
sense in that by the eighth century curials had probably become extinct
and land most certainly was no longer subject to formal registration
for the purposes of taxation. More positively, donations, which exist
69 Above, Chapter 5. See also Rio, Formularies, 255–8; Rio, Legal Practice, 177–82.
70 Bruckner, Regesta Alsatiae, no. 127. On the Merovingian provenance of Eberhard’s will,
which survives in a copy made in the fifteenth century, see Wilsdorf, ‘Le monasterium
scottorum de Honau’, 63–4, and C. Wilsdorf, ‘Remiremont et Murbach à l’époque
carolingienne’, in M. Parisse (ed.), Remiremont, l’abbaye et la ville (Nancy, 1980), 49–
57.
71 H. Ebling, Prosopographie der Amtsträger des Merowingerreiches von Chlothar II. (613) bis
Karl Martell (741) (Munich, 1974), 131n15.
72 Marculf 2.3; trans. Rio, Formularies, 185; J. Barbier, ‘Pouvoirs et élites dans le monde
franc (VIe –XIe siècle): Matériaux pour servir à l’histoire des élites des cités (VIe –IXe
siècle): Le dossier des gesta municipalia, étude no 2’, unpublished mémoire d’HDR,
Université de Paris I–Panthéon-Sorbonne (2009), Part I, Chapter 3, 88–94.
208 Hans Hummer
in abundance only from the late seventh century on, are replete with
anxious clauses about meddlesome relatives. This might explain why,
whereas the formula of Marculf goes on to warn both heirs (heres) and
greedy judges not to interfere with the gift, the author of Eberhard’s will
expands the circle of potentially troublesome heirs (heres aut proheres),
elaborates Marculf into a more ferocious warning to those heirs (and to
himself), and makes no mention of tax-gouging judges.73
The formula for the gesta municipalia in Count Eberhard’s will, shorn
as it was of the fiscal and functional trappings of the putative late antique
institution, might allude, albeit with linguistic pretension, to the perfor-
mance of a high-profile transaction, in the assembly that had gathered
to witness the assignment (literally) of this count’s impressive chartered
endowment – as we discover at the end of the document – ‘publice at
the castrum Habendum and the monastery of Remiremont’; that is, at
a place that was the functional equivalent of a Frankish vicus publicus.74
Thus, if the phrase was mobilized for the purposes of affectation, that
affectation was appropriate to the solemnity of the occasion: the fabulous
endowment of Murbach made publicly at the fortified, comital hub at
Remiremont.
Whatever the particulars of its development might have been, the dis-
appearance or transformation of the gesta municipalia meant that over the
course of the Merovingian period, especially during the seventh century,
the institutions central to the generation and preservation of documen-
tation in Late Antiquity had become obsolete. Consequently, many of
the records dependent upon it – ecclesiastical records included – passed
with it. This, more than anything else, probably explains the meagre
documentation surviving from the seventh century. The phenomenon
of ephemerality should be distinguished, however, from the production
of documentation at that time. Formulas, wills, allusions to records in
narrative and hagiographical accounts, and the smattering of monastic
charters that do survive point to the wide and deep circulation of docu-
mentation in the seventh century.
The wills of two seventh-century aristocrats of the Paris basin, Ermen-
trude and the son of Idda, will illuminate the contention. Ermentrude’s
will directed property to her son, her nepotes, and several ecclesiastical
73 Marculf 2.3, on which see above, 207; Bruckner, Regesta Alsatiae, no. 127.
74 The initial foundation of Remiremont around 620 was probably supported with royal
grants (H. Büttner, Geschichte des Elsaß I . . . , ed. T. Endemann [Sigmaringen, 1991],
57–8), making it a probable hub of comital authority as early as the seventh century. By
the ninth century, it was a preferred stop on the royal hunting itinerary (Annales regni
Francorum, s.aa. 805, 817, 821, 825 [MGH SRG 6:120, 147, 155, 167]). By 849, a
royal palace at Remiremont finally comes into view: MGH DD Kar. 3, nos. 105–6.
institutions.75 Although the document survives because it was preserved

by one of the interested churches, the document itself was addressed in
the second person to Ermentrude’s son, an indication that a copy of the
transaction had been made out for him. We might also infer the one-time
existence of copies for the other churches, and probably for her nepotes.
Similarly, the son of Idda made provisions for the dispensation of prop-
erty to his wife, Chramnethrude, and several churches.76 Some of the
properties were directed to churches the son of Idda had already donated
by now lost epistolae donacionis. At the end of the will, the subscribing wit-
nesses verified this exemplar autenticum, ‘authentic copy’, of the charter,
an indication that at least one other version had been drafted.
The echoes of erstwhile documentation also reverberate in hagiograph-
ical accounts. The Life of Germanus of Grandval, composed around 680,
informs its readers that sometime in the 630s Abbot Waldebert of Luxeuil
persuaded Duke Gundoin to ‘make firm those places with his hands and
those of good men’ for the foundation of a monastery in Grandval, a
phrase that finds parallels in Salic donation formulas.77 Likewise, the
Life (c. 615–20) of Gundoin’s daughter, Sadalberga, reveals that after
she had founded a monastery at Laon, she became embroiled in a dis-
pute with her brother, who withheld from the foundation ‘some villas
that she had granted through a series of charters’.78 After harmony was
restored, ‘the inviolability of the charters was mutually confirmed’ by
Sadalberga and her brother. Sadalberga’s Life also tells us that she earlier
had founded a monastery, probably by 640, near Langres ‘on her pater-
nal inheritance and succession, which she enriched with revenues from
the estates of her paternal, hereditary succession’.79 In light of the wills
above, we might infer a transfer of these family properties to Sadalberga
by formal testament, and then grants by her to the foundation by for-
mal epistolae donacionis. Moreover, the detailed listing of property in the
wills of Ermentrude and the son of Idda, and the reference in the Life of
Sadalberga to Sadalberga’s revenues, would seem to rest on the one-time
existence of estate records.
75 ChLA XIV 592. 76 ChLA XIII 569.

77 Bobolenus, Vita Germani abbatis Grandivallensis 7 (MGH SRM 5:36): ‘Tunc coepit
Waldebertus verbis lenibus mulcere animum eius, ut, si vellet, pro Dei intuitu vel
pro remedio animae sue vel absolutione peccaminum suorum firmitatem de ipsis locis
manibus suis seu bonorum hominum roboratam manibus exhibeat.’ Cf. Form. Sal. Big.
10: ‘ut per hanc epistolam donacione vel deliberatione nostra, manu mea vel bonorum
firmata’; Form. Sal. Lind. 6: ‘sed magis praesens haec cartola tam a me quam ab aliis
bonorum hominum manibus roborata, omni tempore firma et stabilis permaneat’.
78 Vita Sadalbergae Abbatissae Laudunensis 29 (MGH SRM 5:66).
79 Vita Sadalbergae Abbatissae Laudunensis 12 (MGH SRM 5:56–7).
210 Hans Hummer
In any event, these wills and saints’ lives testify to clusters of collateral
documentation that no longer exist but that, if the casual references to
them are indicative, were assumed to be integral to the give and take
of property settlements. If we could parachute into the period at any
moment, we probably would see that documents were routinely gener-
ated and pragmatically preserved by institutions and individuals alike.
However, while ecclesiastical institutions obtained records of their trans-
actions, monasteries and churches were not yet decisive for the preser-
vation of documentation because they had developed neither the means
nor the will to perpetuate, by themselves, their holdings over long peri-
ods of time. And this should not be surprising, since there is nothing
self-evident about the collection and preservation of documents. Not
even the late Roman state pretended to preserve records of its business
for eternity; nor were clerics – despite stereotypes – somehow geneti-
cally programmed to hoard charters. From the perspective of millennia,
property records – barring a change in the consciousness of the institu-
tions to which they were attached – were bound to pass out of existence.
Ultimately, a system centred on ecclesiastical institutions would be elab-
orated and refined only in the Carolingian period, when monasteries
became conduits of central authority and assumed a greater public role.
Crucially, this public duty was infused with the radical Carolingian ethos
of memoria, which for our purposes placed upon monks the obligation to
preserve forever the records of donations as signs of the eternal, heavenly
order on earth.80
When Carolingian monks created cartularies in the ninth century, they
were at a loss to compile a coherent record of their oldest donations. Wis-
sembourg was founded in the seventh century, although exactly when and
by whom is obscure because the parchment trail is so meagre. During the
high Middle Ages, when their foundation by a great king or lord became
an obsession of monasteries (a defining preoccupation, we might add,
of many later cartularies), the monks of Wissembourg manufactured an
original endowment in 633 by Dagobert I.81 In reality, Wissembourg
probably was founded around 660 by Bishop Dragobodo of Speyer and
Duke Bonifatius, though the charters offer no conclusive proof.82 The
80 O. G. Oexle, ‘Memoria und Memorialüberlieferung im früheren Mittelalter’,

Frühmittelalterliche Studien, 10 (1976), 70–95; Geary, Phantoms, 16–19. The literature
on memoria is now vast; see Oexle, ‘Memoria, Memorialüberlieferung’, in Lexikon des
Mittelalters, 10 vols. (Munich, 1980–99), VI, 510–13.
81 Bruckner, Regesta Alsatiae, no. 32.
82 F. Staab, ‘Noch einmal zur Diplomatik der Weißenburger Traditionen’, Archiv für mit-
telrheinische Kirchengeschichte, 44 (1992), 311–22; A. Doll, ‘Ist die Diplomatik der Wis-
sembourger Urkunden geklärt?’, Archiv für mittelrheinische Kirchengeschichte, 45 (1993),
439–47; and Doll, ‘Kloster Wissembourg, seine Gründung und deren Zeugen’, Archiv
für mittelrheinische Kirchengeschichte, 44 (1992), 287–309.
earliest charter in the cartulary, Bonifatius’ donation of 661, is not a

foundation charter per se: it grants the property and mill of the duke’s late
son Gundebald at Görlingen, yet it does make reference to ‘the monastery
of lord Peter at Wissembourg, which was built by that bishop’ (i.e. Drago-
bodo). This strongly implies that Dragobodo had made earlier grants for
the foundation of the monastery. No records of that act survive, though
if the previous examples of Germanus and Sadalberga are an indication,
they surely had once existed. Nor do we possess seventh-century charters
from the Rodoins, a kin group believed to have been among the nexus of
founding families: their putative ancestors, Chrodohari and Chrodoald,
stood as witnesses to Bonifatius’ donation, and the first recorded abbot of
Wissembourg, a certain Chrodoin, apparently another ancestor, received
the second oldest donation in 682 from Duke Theotchar.83 Records of
the Rodoins’ donations begin only in 717, but, as we have seen, the will
of Chrodoin, son of Peter, refers to the prior donation of a church at
Berg by his ‘ancestors’, perhaps an allusion to these otherwise obscure
figures. Bonifatius’ and Theotchar’s donations, and a third charter dated
to 693×4, represent the only surviving records of the monastery’s first
three decades of business.84 Only from 695 on are transactions more or
less continuously represented, an impressive memory to be sure when
compared to the other collections, but a paltry account of what must
have been wider holdings in the seventh century.
The cartularies of Fulda and Freising were compiled three decades ear-
lier than Wissembourg’s, during the 820s.85 Although Fulda was founded
in 744, and thus had less than a century of transactions to remember,
the oldest charter in the extant portions of the cartulary dates to 751.86
A cluster of transactions from the 750s ensues, but very little remains
from the 760s. Only from the 770s and 780s does the flow of charters
become continuous. At Freising, a see established in the 730s when the
Bavarian Church was reformed by Boniface, only one charter from the
tenure of Bishop Erembert (73?–48) was remembered, and that was mis-
takenly listed in the register of the charters transacted during the period
of Joseph (748–64), his successor, itself a sparse record.87 Only from 769
does the cartulary impart a steady stream of records.
83 K. Glöckner, ‘Die Anfänge des Klosters Weißenburg’, Elsaß-Lothringisches Jahrbuch, 19

(1939), 1–46, at 13–20. See TW 213 for Theotchar’s donation.
84 TW 203, 213, 38.
85 TF, p. xxi; UF, p. xx; cf. H. Hummer, ‘A Family Cartulary of Hrabanus Maurus?
Hessisches Staatsarchiv, Marburg, Ms. K 424, folios 75–82v’, in T. Schilp and U.
Ludwig (eds.), Nomen et Fraternitas: Festschrift für Dieter Geuenich zum 65. Geburtstag
(New York, 2008), 645–64, at 646–7.
86 UF 11. 87 TF 1, and p. 3.
212 Hans Hummer
As seen in the previous chapter, the monks of St Gall in Alemannia

never did produce, so far as is known, a cartulary; they bequeathed
instead an impressive trove of unbound, original charters. It is perhaps
significant then that this collection exhibits many of the same patterns
evident in the extant cartularies. Although St Gall was, like Wissembourg,
founded in the seventh century by a cluster of aristocratic kindreds,88 it
could recall absolutely nothing of transactions made at that time. Aside
from a lone charter in 716, and a few from the 740s, the monks failed to
remember much of anything prior to the 760s.89
Thus we have a similar pattern in multiple regions: a fragmentary
record of monasteries’ earliest donations, which becomes continuously
represented – with the notable early exception of Wissembourg – only
after 770. Positive evidence suggests that the memory of the Merovin-
gian and early Carolingian periods was hazy, not because monasteries did
not use or possess documents in these periods, but because they devel-
oped a self-sustaining system of archival management and preservation
only in the late eighth century. Much of what was recorded before the
consolidation of this sharpened archival consciousness passed into obliv-
ion, and the vestiges we do possess we owe to Carolingian-era efforts to
reconstruct documentation from this earlier time.
Illuminating is a plea from the monastery of Honau, located on an
island in the Rhine and founded in 722 by the Etichonids, the ducal
family in Alsace. In 775, its abbot Beatus asked Charlemagne to confirm
all the properties that had been given to the monastery by kings and
God-fearing men because the charters (instrumenta chartarum) had been
lost due to neglect.90 That is a startling statement when one considers
that Honau had been in operation a mere fifty years. The abbot’s com-
plaint, however, gives voice to patterns visible elsewhere: the rapid loss
of charters apparently due to poor records management, as well as to
the attempts in the late eighth century to reconstitute them. Testimony
emanating from the early modern period indicates that the monks indeed
appear to have reconstructed their lost records. A seventeenth-century
Jesuit claims to have seen a cartulary from Honau, apparently lost sub-
sequently during the Thirty Years’ War, which, he says, contained over a
88 Or so one can infer from Notker’s later complaint, albeit exaggerated, that St Gall was
poor because of its traditional dependence upon small-scale donations (‘ex privatorum
tradiciunculis’), by contrast with the royally endowed foundations, Gesta Karoli Magni
imperatoris 2.10 (MGH SRG n.s. 12:66–7).
89 Wartmann I 17–28.
90 MGH DD Kar. 1, no. 10; for a fuller discussion of the context of this plea, see Hummer,
Politics, 63–5.
thousand charters spanning the period from Honau’s foundation through

the reign of Charlemagne (768–814).91
So what exactly would this late Merovingian world of document pro-
duction have looked like? Fortunately, the region of Alsace and neigh-
bouring southern Lotharingia has handed down a fairly tight body of
charters from the late seventh and early eighth centuries, permitting us
to elucidate the patterns of the production and preservation of docu-
mentation in the later Merovingian period, as well as the formation of
a new archival consciousness within ecclesiastical institutions by the late
eighth century. All told, this region yields nearly a hundred Merovingian-
era charters, a veritable flood by the standards of other collections. The
mother lode of charters springs from the Wissembourg codex, which
mediates eighty-three late Merovingian charters from the period 661–
747.92 As we have seen, the two earliest charters date to 661 and 682,
and a steady flow of charters commences only around 695. Of these
late Merovingian charters, an impressive total of thirty-nine emanate
from the period before 720; that is, they belong to the period before
Charles Martel had consolidated Carolingian power, an indication that
the phenomenon of their existence, or their survival, cannot adequately
be explained by the deus ex machina of Carolingian takeover.93
Revealing within these charters is the notarial evidence. Generally
speaking, only during the latter half of the eighth century did the writing
of charters come to be centred on notaries from Wissembourg. Prior
to that, most operated as free agents outside the monastery’s sphere
of authority. Here we must rely on the toilsome work of editors, who
have developed diplomatic techniques to distinguish monastic from non-
monastic notaries. Briefly, the first clue is whether the notary identifies
himself as an emanuensis, a notarius or a cancellarius, often an indication
of his personal professional identity. Of these, the term emanuensis is held
91 Wilsdorf, ‘Le Monasterium scottorum de Honau’, 11. Jodocus Coccius, the Jesuit
scholar, claims the cartulary was assembled in 1079 by a monk of Honau named Leon.
According to Coccius, the charters were arranged in chronological order, perhaps by
the abbot/bishop, and ranged from the reign of Theuderic IV to that of Charlemagne.
It is impossible to say more from such slim testimony, but the dates of the charters, and
their organizational scheme, suggest perhaps that the cartulary was embedded within a
larger survey compiled by Leon, in the way that Cozroh’s codex, also a chronological
survey, was absorbed into the codex produced at Freising in 1187 by Conrad Sacristan.
92 TW 1–7, 8 = 47, 9–16, 17 = 159, 18, 35 = 162, 36–41, 43–6, 52, 136, 141–3, 146–8,
150, 169, 186–8, 192, 194 = 224, 195–6, 202–3, 205 = 223= 252, 213, 218 = 239,
225–6, 228–9, 231–5, 237, 240–4, 247–8, 256–7, 261–2, 265–7. A smattering of other
charters, dating from 722×3 to 749, mostly from Murbach and Honau, can be found
in Bruckner, Regesta Alsatiae, nos. 100–3, 113, 125, 127, 128, 163, 165, 167.
93 Cf. Hummer, Politics, 66, 68.
214 Hans Hummer
to be the most discriminating, since it was rarely used by monks. How-

ever, a scribe might not identify himself professionally, or might do so in
some cases, but not others. Nor can we assume that the absence of self-
identification as a monk or priest necessarily means that the scribe is a
layperson because vocational statuses too were not uniformly expressed.
That is, the consistency of expression that we might expect, or desire, is
frustratingly lacking.
Fortunately, the charters yield other clues that can be brought to bear
on the problem. In the absence of any firm declaration of professional
and vocational status, one can identify the notary’s field of activity –
does he ever write at the monastery, or exclusively in locales in a defined
region? – as well as his name, which might betray a prosopographical
connection to a known kin group, or appear in the lists of monks kept by
monasteries, against which his membership can be checked. Additional
clues can be found in the formulas employed by the scribe (his Fassung),
which allude to a notarial training acquired from within the monastery
or from somewhere else. Sometimes, we can even find a notary who
consistently appears in the retinue of a count or a duke, and who does
not write for a particular institution, or in a particular region, but for his
lord. At other times we encounter a notary who wrote for more than one
institution, revealing his status as an autonomous notary, independent of
any single institution.
From all of this, researchers have identified two main classes of
notaries, ecclesiastical scribes and those who are traditionally, if perhaps
misleadingly, called offentliche Schreiber, or public notaries. The former
can be divided into monastic and episcopal notaries; the latter, into the
so-called Gauschreiber, or notaries who operate in particular districts,
often with a close connection to particular groups; court notaries, who
write for entitled lay lords; and free scribes, who might write for multiple
monasteries or groups, yet remain independent of them. These cate-
gories need not be overstressed, since a Gau or a court notary might
be a priest, working on behalf of the kin groups or powerful lords with
which he was associated, while a local scribe might be someone whose
activity suggests work on behalf of a cloister, yet who was not himself a
monk. This investigation will stress the distinction between monastic and
non-monastic notaries, whoever they might have been, and how the ratio
between the two changed over time so that monastic competency came to
dominate.
We begin with a closer look at the patterns of notarial evidence in
the Wissembourg codex. When Duke Bonifatius made the first recorded
donation to Wissembourg in 661, his transaction was drawn up by the
duke’s own son, Teodoald, who reveals that he was asked to write ‘this
epistola by my father Bonifatius’.94 Teodoald’s statement occurs before

the list of witnesses, after which follows the date and place of the trans-
action. The dating clause and the witnesses, however, appear to have
been added by the scribe Leodobert, who claims at the end of the char-
ter to ‘have written at the request of those written above’ (‘rogatus a
suprascripta scripsi’), presumably referring to the parties to the transac-
tion. Leodebert, otherwise unknown, might have been related to one of
the witnesses, Ledoald, and to the aforementioned early eighth-century
emanuensis Leodoin.95 The hybrid configuration of this charter suggests
that a record had been drawn up for the duke by his own son, and a
second by a local scribe for Bishop Dragobodo, to whom the charter is
addressed.
Duke Theotchar’s donation of a lucrative salt pan in 682 was drawn
up in the Saulnois by a local scribe, Babo emanuensis, who seems to have
been one of the duke’s neighbours.96 The third oldest transaction in the
cartulary was a donation made in 693 or 694 by a certain Hildifrid on
behalf of himself and his two siblings, Managold and Waldswind. It was
drawn up by the otherwise unknown notarius Uadalgari, who ‘wrote, hav-
ing been asked by those written above’ (‘rogatus a suprascriptis scripsi’),
and who, judging by the elements of his name, might have been a relative
of the donors and some of the witnesses.97 In any event, his use of older
formulas and his autonomy vis-à-vis the transacting parties suggest that
he was no monk.98
Around 700, when the number of transactions greatly increases, we
encounter a series of charters down to the 760s written up by thirty
notaries, only two of whom can with any sureness be identified as
monks.99 A third, Theutgar, might eventually have entered the chapter
at Wissembourg, but his style betrays a previous notarial education.100
Ten others identified themselves as clerics – one as a bishop, and the
94 TW 203.
95 Glöckner and Doll speculate that Leodebert probably was a non-monastic notary and
perhaps related to the amanuensis Leodoin, TW, pp. 416, 417n2.
96 TW 213. The charter records the people living around the property, one of whom
was a man named Bobo, a name which, when we take into account the irregularity of
medieval orthography, is virtually indistinguishable from Babo.
97 TW 38. The names of one of the donors, Waldswind, and two of the witnesses, Walthari
and Willihari, may point to kinship with Uadalgari. So far as can be inferred from other
charters, the witnesses mostly were laymen.
98 TW, p. 218.
99 TW, pp. 115–51, 629–39 (Table 76). The two monks were Hildifrid (TW, p. 128) and
Erlabold (TW 257).
100 Theutgar appears first in 712. Glöckner and Doll infer from evidence in the witness
lists to other charters that he probably became active on behalf of the monastery in the
late 730s (TW, p. 125).
216 Hans Hummer
rest as either deacons or priests; but of these ten, four also formally
identified themselves as emanuenses, and one possibly was a monk at
Wissembourg.101 That most of these ten ‘ecclesiastical’ notaries were
priests and deacons, rather than monks, presumably points to origins
within the diocesan ranks. Even still, they seem to have behaved not as
agents of episcopal authority as such, but rather as free scribes or as
notaries in the service of powerful patron groups. The notaries Ansgar
and Heimo, for example, both self-described presbyters, appear only as
notaries to the Etichonids, the ducal family in Alsace.102 The priest and
emanuensis Geroin was active in the Speyergau for the Ratbald-Wicbalds
and appears in the Fulda charters as the notary to a Count Baugulf.103
And the deacon Guntbert was busy in the Saargau on behalf of the
Rodoins and Wolfoald-Gundoins.104
The remaining seventeen, however, were not professional clerics (at
least no formal affiliations are mentioned). Their activities are traceable
rather to the districts within which identifiable patron groups operated.
Of these seventeen scribes, three were (lay) patrons of Wissembourg,
each of whom also worked as a notarius or an emanuensis for the Wolfoald-
Gundoin and Rodoin kin groups;105 another self-identified notarius and
emanuensis wrote on behalf of the same two groups;106 two identified
themselves as cancellarii and appear on the basis of their styles to have
been local scribes;107 one was in the service of a count;108 one bears no
self-designation as a notary, but appears as a scribe in both the Wis-
sembourg and the Honau charters, thus pointing to the activities of a
free notary;109 and the rest are otherwise unidentifiable with respect to
profession.110 So far as one can tell from their self-identification (or
101 Ratker, deacon/emanuensis (TW, pp. 119–21); Desiderius, presbiter (p. 221); Gund-
bert, deacon/cancellarius/emanuensis (p. 128); Heimo, presbiter (pp. 128–9); Sindicho,
deacon/emanuensis (pp. 130–1); Ansgar, presbiter (p. 131); Otricus, presbiter (p. 489);
Basinus-Baldfrid, deacon (p. 132); Geroin, presbiter/emanuensis (pp. 133–5); and Bishop
Hartbert (p. 346). Basinus might have been, on the basis of his field of activity, a monk,
though he bears the same name as the bishop of Speyer. If he was the same bishop, he
could have operated out of the cathedral; cf. TW, p. 132.
102 Ansgar’s formulas are similar to those used in other Etichonid donations to Honau
(TW, p. 131); and Heimo wrote up Etichonid transactions with both Wissembourg
and Honau (cf. Bruckner, Regesta Alsatiae, nos. 103, 110, 126).
103 TW, pp. 133–5; UF 53. 104 TW, p. 128.
105 Chrodoin, notarius (TW, pp. 115–18); Hahicho, notarius (pp. 121–2); Chroccus,
emanuensis (pp. 118–19). All wrote for the Wolfoald-Gundoins, and Hahicho also
for the Rodoins.
106 Leudoin (TW, pp. 126–8).
107 Lantfrid (TW, p. 505); Wibald (p. 507). 108 Williulf (TW, p. 182).
109 Wanulf (TW, p. 371; Bruckner, Regesta Alsatiae, no. 167).
110 Heribert (TW 43), Gerhari (TW, p. 389), Faramund (pp. 129–30), Chudo (p. 390),
Humbert (TW 14), Harbert (TW, p. 348), Grimoald (p. 345), Hildifrid (p. 342),
Fruatolf (pp. 135–6).
lack thereof), in combination with their fields of activity, the location

of the transactions, the location of the donated properties, or notarial
styles, none of these yield any evidence of a formal affiliation with the
monastery.
This is not to say that none of these lay notaries, or the priestly notaries,
had no connection to the monastery at all. Some of them appear to have
written either exclusively, or mostly, for Wissembourg, thus betraying a
cordial, if informal, relationship. The ‘ecclesiastical’ emanuenses Ratker
and Sindicho, for example, wrote up transactions involving properties
donated in multiple districts almost exclusively at Wissembourg.111 Two
of the lay notaries, Chrodoin and Hahicho, when they were not both
working on behalf of local groups, wrote on behalf of the monastery.112
What this tells us, then, is that, early on, monks did not monopolize the
production of charters. When the monks, or lay actors for that matter,
wanted documents drawn up, they called upon a notariate, some of whom
might have been in their service, but most of whom were independent of
either the cloister or a secular patron.
In fact, local scribes dominated notarial activity in the late Merovin-
gian period: of the seventy-six charters that date to the period before
751, thirty-six were drawn up by just five notaries from the Saargau.113
By contrast, only twenty-two charters are traceable to the work of two
monks and two putative monks, and seventeen of those were drawn up
by a single scribe, the aforementioned Theutgar, who, as we have seen,
seems to have become a monk only later.114 The most intriguing of the
five Saargau notaries is the self-described notarius Chrodoin, a prolific
scribe among Wissembourg’s late Merovingian charters and the same
Chrodoin, son of Petrus, who was one of the progenitors of the Rodoins,
the group responsible for many of the transactions in the Saargau section
of the cartulary.115 Between 695 and 718, Chrodoin drafted at least nine
documents, appeared as donor or testator in five others, and appeared as
a witness in six more.116 In his capacity as a notary and a witness, he wrote
and testified for a circle of people, the most prominent of whom were
111 TW, pp. 120–1, 132. 112 See above, n. 105.

113 Chrodoin (10), Chroccus (6), Hahicho (5), Leudoin (10), Faramund (5).
114 See above, nn. 99–101.
115 On the Rodoins, see Glöckner, ‘Anfänge’, 18–21; and F. Langenbeck, ‘Probleme der
elsässischen Geschichte in fränkischer Zeit’, Alemannisches Jahrbuch (1957), 1–132,
at 33–40. On the Rodoins and the cartulary of Wissembourg, see Hummer, Politics,
65–75, 190–207.
116 TW, pp. 115–18. The editors cautiously credit him with a tenth charter whose copy is
missing its notary yet whose formulas are evocative of Chrodoin’s formulas. The donor
in question was a certain ‘priest Chrodoin’, who the editors speculate is one and the
same person. On the other hand, they also present compelling evidence that points to
two separate individuals (TW, pp. 117, 215–16).
218 Hans Hummer
individuals of the Wolfoald-Gundoin group, the former mayoral family

of Austrasia. Chrodoin’s own transactions were drawn up not by himself,
but by another prolific notary, the aforementioned Leudoin, the self-
described emanuensis, notarius and cancellarius, and by all appearances a
layperson, who was responsible for nine charters, among them transac-
tions of the Wolfoald-Gundoins.117 As the intertwined careers of these
two notaries illustrate, in the early eighth century an impressive propor-
tion of documentation arose from associative networks within localities.
More than one notary might be summoned to draft copies of char-
ters for both parties to the transaction. In the purchase of property by
Wissembourg in 715, the notarius Chrodoin claims that he ‘wrote this
sale and subscribed’.118 A second draft reveals that ‘Chrodoin wrote the
charter, and I Gundbert also wrote and made a copy of the charter’,
presumably either for the seller Ermbert or for the monastery, although
Gundbert himself seems not to have been a monk, but yet another local
notary formally unconnected to Wissembourg.119 In a donation of 705 or
706, and in the subsequent precaria of 706×7, two local notaries, Chroc-
cus and Hahicho, indicate that they had each written a version of the
same transaction. According to the donation charter, ‘I Chroccus having
been asked wrote [and] I Hahicho issued the original version (‘auten-
ticum relegi’), [wrote] the copy (exemplaria), and subscribed.’120 In the
precaria drafted the following year, the roles were reversed: ‘I Hahicho
wrote [and] I Chroccus issued the original version, wrote the copy and
subscribed.’121 The wording in these two documents again suggests the
production of copies for the two parties, the patron and the institution,
each presumably employing their own notaries.
By the end of the eighth century, the drafting of property transactions
became centred emphatically on Wissembourg. From the 770s on, most
of the extant documents were drafted by monks who, when they referred
to themselves as cancellarii, clearly meant by that that they were monastic
scribes.122 The overlapping careers of two notaries responsible for draft-
ing the lion’s share of transactions between 760 and 785, Geroin (757,
764–84) and Kadwal (767, 774–82), neatly encapsulate the transition.123
We have already met Geroin, who was both a priest and an emanuensis
in the Speyergau and northern Alsace, and, judging from the documents
he drafted, was closely tied to the prominent Ratbald-Wicbald group
of donors.124 I note here that the patterns of Geroin’s activities evoke
117 TW, pp. 126–8. 118 TW 239. 119 TW 218 and p. 470; cf. p. 128.
120 TW 228. 121 TW 229. 122 TW, p. 137.
123 Of the sixty-seven transactions written up during this time, Geroin produced charters
for twenty-seven and Kadwal fourteen (TW, pp. 538–46).
124 TW, pp. 133–5.
prevailing late Merovingian practices when notaries were frequently tied

to patron groups. On the other hand, although Geroin appears not to
have been a monk, he was also literally close to Wissembourg, perhaps
having resided near or at the monastery: of the twenty-eight transactions
he wrote up, twenty were conducted at Wissembourg, and two of those
were said to have been done ‘in the monastery’ (in ipso monasterio).125
Kadwal, by contrast, was a monk of Wissembourg, was the first to
identify himself as the monastic cancellarius, and stands at the beginning
of an unbroken line of monastic cancellarii at Wissembourg. Although
some of his style bears the influence of Geroin’s, he introduced a num-
ber of changes to the formulas that shaped Geroin’s later charters; these
changes turned out to have had a lasting influence on his successors at
Wissembourg, an indication that the monastery’s cancellarius was har-
nessing the charters to a distinctive Wissembourg style.126 Moreover, if
Geroin’s charters sometimes might refer to transactions conducted ‘in
the monastery’, Kadwal’s are the first charters to distinguish the act of
transaction (actum) from the drafting of the charter itself and, beginning
in 774, to indicate that the charter had been produced (facta cartola)
at the monastic scriptorium.127 Over the next several decades, between
782 and 811, three monks, Adalland, the dominant figure in Wissem-
bourg’s archive in the late eighth century;128 Hildibodo and Wolfhart
were responsible for drafting most of the charters in this, the most robust
period for transactions.129 The acts of donation recorded in these char-
ters, when indicated, occurred overwhelmingly at Wissembourg: of the
128 transactions made between 770 and 811, only a dozen stipulate that
they had been done elsewhere.130 By all accounts, the more diverse notar-
ial culture of the late Merovingian period had decisively come under the
domination of Wissembourg and its notaries.
This reorganization of Wissembourg’s scriptorium went hand in hand
with an effort to systematize and preserve its records. Although Wissem-
bourg’s cartulary was not assembled until 860, it was preceded by other
125 TW, pp. 135, 633. The number of transactions done at Wissembourg might have been
higher since five do not reveal the location of the transaction. That is, in only three did
he write somewhere other than Wissembourg. However, these three transactions, his
activity on behalf of the Ratbald-Wicbalds and his serving as notary to Count Baugulf
in the Fulda codices indicate that he worked independently of the monastery too.
126 TW, pp. 136–7. 127 TW 54, 73, 90, 93, 95; cf. p. 137.
128 H. Butzmann, ‘Die Weißenburger Handschriften: Einleitung zum Katalog’, in H. Butz-
mann, Kleine Schriften: Festgabe zum 70. Geburtstag, ed. Wolfgang Milde (Graz, 1973),
48–103, esp. 82–8.
129 Of the ninety-four documents dated to this period, seventy-two were drafted by the
three, and all but one of the seventy-two were conducted at the monastery (TW,
pp. 546–55). On Adalland, Hildibodo and Wolfhart, see TW, 138–41, 142–4, 145–6.
130 Cf. TW, pp. 540–57.
220 Hans Hummer
compilations. The oldest sections of the high medieval Liber possessionum,

which convey aggregate figures of manorial holdings in the mid-Rhine
region, were drawn up in the ninth century, perhaps in the 840s.131 The
Brevium exempla, a royal capitulary that dates to between 810 and 817
and advised monasteries on the organization of their records, issued as
examples lists of those who held precariae and benefices in the Worms-
gau from Wissembourg, an indication that the monastery had begun to
formalize its record-keeping habits at least by the early ninth century.132
If charter production had come under Wissembourg’s administration
by the late eighth century, this is not to say that the monastery’s role
remained unchanged, or that all notarial activity had been monopo-
lized by monastic scribes. After Wolfhart’s tenure ended in 811, subse-
quent ninth-century transactions were more likely to be conducted away
from Wissembourg at particular villages. Of the dozen charters drawn
up down to 830 by Guntbert, the chief monastic scribe after Wolfhart,
only five were conducted at Wissembourg.133 And this seems not to have
been a phenomenon particular to Guntbert: of the thirty-six transactions
between 811 and 864, fewer than half (fifteen) are known to have been
made at Wissembourg itself, a striking change from the last quarter of
the eighth century.134
Indeed, a separate, albeit subordinate, line of notaries seems to have
persisted for a time in the Saargau and the Saulnois. As we have seen,
this trans-Vosges region was serviced in the late Merovingian period
by a number of notaries, the most prominent of whom appear to have
arisen from, or to have written for, the Rodoin and Wolfoald-Gundoin
groups. Moreover, by contrast with transactions of property in Alsace,
which were conducted almost always at Wissembourg even in the late
Merovingian period,135 those involving property in southern Lotharingia
were just as likely to be conducted in local villages.136 The Saargau’s
distinctiveness in this respect persisted into the Carolingian period. In
763 and 789, a certain cancellarius Lantfrid and a certain emanuensis
Chrodoin, respectively, wrote up transactions in the Saargau.137 In the
latter case, Chrodoin redrafted a donation – originally made by a certain
131 Liber possessionum Wizenburgensis, ed. C. Dette (Mainz, 1987); cf. A. Doll,
‘Die Possessiones Wizenburgenses und ihre Neuedition’, Archiv für mittelrheinische
Kirchengeschichte, 41 (1989), 437–63, and M. Gockel, ‘Kritische Bemerkungen zu
einer Neuausgabe des Liber possessionum Wizenburgensis’, Hessisches Jahrbuch für
Landesgeschichte, 39 (1989), 353–80.
132 MGH Capit. 1, no. 128.
133 Gundbert: TW, pp. 147–9. 134 Cf. TW, pp. 557–63.
135 The exceptions were Etichonid transactions, several of which were conducted at Stras-
bourg, an important centre of Etichonid lordship, TW 10–11, and 35/162.
136 Cf. TW, pp. 629–33. 137 TW 263, 259/260.
Gunthart at Wissembourg and drawn up by cancellarius Adalland – when

a certain Albrich wanted to join the donation. The fact that Chrodoin
identified himself as an emanuensis, and that the transaction was made at
Rieding near Saarbourg rather than at the monastery, strongly suggests
that the joint transaction was drawn up, perhaps at Albrich’s insistence,
by a local notary. When we consider further that Lantfrid and Chrodoin
bear names typical of those belonging to known members of the Rodoin
group, which itself produced the earlier prolific lay notary Chrodoin, it
appears that notaries continued to arise from within these patron groups
throughout the eighth century.
Although there is no evidence that Rodoin notaries persisted into the
ninth century, independent notaries did continue to operate exclusively
in the Saargau, albeit under the general auspices of the monastery. Fore-
most among them was a certain John, who first appears in a charter
of 807 as an emanuensis in the Saargau, and who proceeded to write
up another six transactions down to 847.138 Yet, the elaboration of the
monastic notarial administration was not without consequences even
here. In 846, on behalf of the monastic cancellarius Baldram (‘ad vicem
Baltramni cancellarii’), John wrote up three transactions made at Wis-
sembourg by individuals of the Rodoin group.139 John’s responsibilities
as a local notary were depicted as if they had been assigned to him by
Wissembourg’s cancellarius.
The phenomenon of monastic aggrandizement detectable at Wissem-
bourg exhibited contrasting if complementary patterns elsewhere. The
donation patterns at Fulda reveal that the monastery was supported
first by patron groups in the better-developed mid-Rhine regions around
Mainz and Worms, and then subsequently by patrons residing nearer
to the monastery in the Grapfeld and the Saale basin as these regions
were harnessed more thoroughly to Frankish rule during and after the
Carolingian subjugation of Saxony.140 As we have seen, most of the ear-
liest charters were produced by non-monastic notaries operating within
the dioceses of Strasbourg, Worms and especially Mainz. As at Wissem-
bourg, hardly any of these early notaries were formally attached to Fulda.
This is detectable in their fields of activity, which were limited to trans-
actions performed mostly in the episcopal cities (e.g. actum Mogontiae
civitatis publice) or at villages within the diocesan boundaries, but never
138 TW 201, 230, 242, 268–71. 139 TW 268–71; cf. p. 147.

140 For the period before 770, we have twenty charters from the mid-Rhine region begin-
ning in 751 (UF 11, 18, 22–31, 33, 37, 40–2, 44, 48–9), a mere two from the Grapfeld
beginning in 760 (UF 32, 39), and none from the Saalegau; cf. Stengel, ‘Über die
karlingischen Cartulare’, 181–4.
222 Hans Hummer
at Fulda.141 When they do invoke their professional status, they refer to

themselves as emanuenses, and they often assert that that they had been
asked (rogitus) to write. Half of these ten notaries identified themselves
as priests; the other half by all appearances were laymen.142 This figure,
however, underplays the significance of the lay notaries, because the five
self-identified priests authored just six of the seventy-four charters pro-
duced by the group of non-monastic notaries. To look at it another way,
of the surviving fifty-six charters produced during the monastery’s first
four decades of existence down to 784, forty-one were drawn up by non-
monastic scribes, thirty-four of which were produced by four notaries
working out of Mainz, only one of whom was a priest, and he pro-
duced just one charter.143 The most prolific of these was the emanuensis
141 On the non-monastic notaries, see Stengel, UF, pp. liii–lix, and Innes, State and Society,
114–17. There were ten non-monastic notaries; I have reproduced in parentheses their
fields of activity, the numbers of charters they authored and their date(s) of service, after
which I have listed the charters attributable to them. Anonymously composed charters
believed on stylistic grounds to have been written by the notary are in [brackets]:
Erlefrid, priest and emanuensis, Mainz, one, a. 751 (UF 11); Wolfram, emanuensis,
Mainz, twenty-two, aa. 752–74 (UF 18, 22–4, 26, 28–33, 37, 40, 41, 44, 52, 54–
5, 59, 61, 63, [64], 66); Winibald, Mainz, two, aa. 754–68 (UF 48–9); Weliman,
Mainz, sixteen, aa. 775–88 (UF [70], 71–2, 76, 80–2, 87–8, 160–2, 177, [178], 179,
[180]); Hiltibald, Mainz, fifteen, aa. 790–9 (UF 195–6, 231, 245, 248–53, 255–7,
260); Theotrich, emanuensis, Mainz, seven, aa. 800–15 (UF 263, 266–7, 270, 283–
5); Hiaebo, priest and emanuensis, Worms, two, aa. 765–70 (UF 42, 50); Notbald,
priest, Lobdengau/Worms, one, a. 763 (UF 38); Geroin, priest, Speyer, one, aa. 757–
84 (UF 53); Erhard, southern Alsace/Basel, one, a. 785 (UF 163); Asaph, northern
Alsace/Strasbourg, five, aa. 778–805 (UF 84, [176], 187, 254, 281); Uodalrich, priest,
Alsace/Basel, one, a. 793 (UF 197); Sadrebald, Grabfeld, one, a. 762×3 (UF 39).
Note: I have worked up here only the charters down to 802, where Stengel’s edition
ends. Asaph authored four more Alsatian charters after 802 (CDF 178–9, 208, 225);
Theotrich authored six more in the region of Mainz (CDF 218, 222, 224, 244, 246,
270). The dates for Geroin’s activity have been elaborated on the basis of his more
extensive work at Wissembourg.
142 If any of the other five were priests, they neglected to say so, despite numerous oppor-
tunities. The Mainz scribe Hiltibald, for example, authored fifteen charters, served as
witness in seven others, and was referred to as a neighbour by another donor, for a
total of twenty-three chances to announce any clerical status (UF, p. lv).
143 Erlefrid, Wolfram, Winibald, Weliman; above, n. 141. The location of their trans-
actions, and those of many other notaries, monastic or non-monastic, poses some
problems because many of the charters in the Fulda codex do not reveal where the
transaction had been performed. With respect to the non-monastic notaries, for exam-
ple, Erlefrid’s and Wolfram’s activity in Mainz is well attested: the former’s one, and
eight of the latter’s twenty-three, were done at Mainz. None of Wolfram’s other fifteen
charters divulge a place of transaction, but in the absence of any contrary information,
we assume Mainz. None of Weliman’s charters reveal a place of transaction, but his
style is virtually identical to Wolfram’s, he took a pen in 775 only after Wolfram’s failed
in 774, and a charter he drew up for the monastery at Honau was done at Mainz
(Bruckner, Regesta Alsatiae, no. 275). Most of Theotrich’s charters lack a place of
transaction, but several of his later charters were transacted either at Mainz or in places
nearby (CDF 218, 222, 224, 246). None of Hiltifrid’s charters mentions a place of
Wolfram, who authored twenty-three charters between 752 and 774, and
single-handedly was responsible for fourteen of the first seventeen sur-
viving charters written up during the monastery’s first two decades.144
Fulda’s earliest charters were generated overwhelmingly by a lay notariate
based at Mainz.
Fulda itself became a centre for transactions and document production
only after 775: property donated in 776 by a certain Ruduch signals the
first recorded transaction made at Fulda (actum in monasterio Fulda).145
This charter was written up by the priest and monk Asger (776–96),
who is the first named monastic scribe to appear among the Fulda char-
ters, as well as the most prolific of the monastery’s late eighth-century
scribes, having drafted charters for thirty-one transactions, and proba-
bly ten others that lack a notarial subscription.146 The Fulda codices
do convey six earlier charters written in a similar, monastic style, which
were composed between 770 and 775, but these do not divulge the
name of the notary and only one identifies the place of transaction.147
These would appear to be the work of a priest named Rihelm, whose
lone attested charter bears a similar construction, but his relationship
to Fulda is unknown, and the one of the six unsigned charters that does
mention the location of transaction purports to have been done at Mainz,
not Fulda.148
Monastic notaries gained the upper hand among Fulda’s charters only
after 785 when Asger became most productive.149 Of the 110 charters
transacted between 785 and 802, seventy-four were put out by eight
monastic scribes.150 Fulda now became the major centre of activity:
forty-six transactions were carried out there, thus overtaking Mainz as the
transaction, but again his style mimics that of the other Mainz notaries, and both he
and Theotrich appear as witnesses for people donating property in the region around
Mainz. Thus, all available evidence points without contradiction to Mainz as the field
of activity for these scribes; or, to put it another way, not a scrap of positive evidence
places them anywhere else. See UF, pp. lv–lvi.
144 Above, n. 141. 145 UF 75.
146 Cf. UF, pp. lx–lxiii. 147 UF 51, 56–8, 65, 69. 148 UF 86; cf. p. lx.
149 Asger put out just three charters before 785 (UF 75, 85, 144), and thirty-eight after-
wards (UF 156, [167], 172, [181], 182–4, 186, 188, 190, 192, 200–10, [211], 212–13,
[214], 215–18, [224–7], 229, 241–3). Charters in brackets refer to those lacking a
notary but believed on stylistic grounds to have been drafted by Asger.
150 In addition to Asger (see previous note), these were Abraham, a. 785 (UF 157–8);
Einhart, aa. 788–96 (UF 175, 189, 191, [219–20, 232–3], 234–5, 240); Hracholf,
aa. 795–812 (UF [230, 233, 236–9, 244, 261, 265b, 269b], 275, [276–7], 279, [280,
282, 526–8]); Engilger, a. 798 (UF 258); Reccheo, aa. 800–19 (UF 264); Ramuolt,
aa. 800–1 (UF 269, 278); and one anonymous notary (UF 259). Bracketed charters
refer to those lacking a notary but believed on stylistic grounds to have been drafted by
known notaries. On the monastic notaries, see UF, pp. lx–lxi.
224 Hans Hummer
most frequently mentioned location of transaction.151 This business in

turn necessitated a centre for the production of charters at the monastery.
It perhaps is significant then that this period yields the first explicit state-
ment that a transaction – the donation of a certain Gundacar in 796 –
had been written up in the monastery: ‘scripta haec [k]arta traditionis in
monasterio Fulda’.152
The bulk of the work carried out by Fulda’s notaries, however, was
dependent upon transactions involving property east of the Rhine. If we
single out the books of the cartulary devoted to transactions in the mid-
Rhine region and Alsace, and examine their contents down to 802, we
discover that non-monastic notaries, most of whom as we have seen oper-
ated out of Mainz, accounted for seventy-five transactions, and Fulda’s
notaries for just twenty-seven.153 When Fulda’s notaries did handle trans-
actions of mid-Rhine properties, they did so overwhelming at Fulda,
rather than at locations along the mid Rhine.154 Conversely, the non-
monastic notaries never appeared at Fulda, but always held forth in
locations along the mid Rhine, mostly around Mainz.155
Now, if we examine the books of the cartulary dedicated to charters
in Grapfeld and Saalegau, fifty-six of the fifty-eight charters down to
802 were drawn up by Fulda’s notaries.156 That is, just two transactions
were drafted by non-monastic notaries in the right-Rhine regions. One
of these was done by the aforementioned Wolfram, though he appears to
have handled this donation of Grapfeld property at Mainz;157 the other
151 Those at Fulda were uniformly done by monastic scribes (UF 156–8, 172, 174–5,
181–4, 189–91, 200–4, 206–10, 212–13, 215–20, 230, 234–8, 240, 261, 265b, 269b,
276–7, 279–80, 282). Twenty-nine appear to have been done at Mainz; all but one,
which was done by the monastic scribe Asger (UF 167), were drafted by three non-
monastic notaries of the Mainz region, Weliman, Hiltigar and Theotrich (UF 161–2,
177–80, 185, 195–6, 231, 245, 248–53, 255–7, 260, 263, 266–7, 270, 283–5). None
of these mentions the place of transaction; however, on the probable location of these
transactions at Mainz or nearby, see above, n. 143.
152 UF 230.
153 These books correspond to the extant ninth-century portion of the cartulary. The non-
monastic notaries were responsible for UF 11, 18, 22–4, 26, 28–31, 33, 37–8, 40–2, 44,
48–50, 52–5, 59–61, 63–4, 66, 70–2, 76, 80–2, 84, 87–9, 160–3, 177–80, 185, 187,
195–7, 231, 245, 248–57, 260, 263, 266–7, 270, 281, 283–5; and Fulda’s notaries for
UF 58, 69, 86, 89, 157–8, 167, 182, 184, 190–1, 210, 213, 215–20, 236–7, 259, 261,
268, 277–9.
154 Mainz: UF 69, 167. Fulda: UF 157–8, 182, 184, 190–1, 210, 213, 215–20, 236–7,
261, 277, 279. No place of transaction is mentioned for UF 58, 86, 89, 259, 268, 278.
155 See above, nn. 143, 151.
156 These books correspond to the portions of the cartulary published by Pistorius in 1607.
Fulda’s notaries drafted UF 51, 56, 65, 75, 85, 156, 172, 175, 181, 183, 186, 188–9,
192, 200–9, 211–12, 224–7, 229–30, 232–5, 238–44, 258, 264, 265b, 269b, 275–6,
280, 282, 286–7, 526–8.
157 UF 32. On Mainz as the probable locus of transaction, see above, n. 143.
by a certain Sadrebald in villa publicus Geldersheim, the centre of a royal

estate in the Grapfeld, who was presumably a notary attached to the
royals.158
This geographical bipolarity strongly suggests that the growth in
Fulda’s chancery was stimulated in large part by efforts to organize the
underdeveloped and unruly lands east of the Rhine that, by contrast with
the mid-Rhine region, required the importation of a documentary tradi-
tion, a tradition which would have to be supplied by the monastery itself.
Yet these exertions had consequences for the development of Fulda’s
institutional consciousness as it continued to deal in mid-Rhine proper-
ties. If we compare the transactions undertaken in this region during the
period before 785 to those undertaken between 785 and 802, we find
that the monastery gradually began to assign its own notaries to dona-
tions of mid-Rhine properties, instead of relying so heavily on the Mainz
notariate. Whereas before 785, monastic notaries accounted for just four
out of forty-four mid-Rhine transactions,159 between 785 and 802 they
handled twenty-three out of fifty-eight.160 Although the Mainz notariat
remained a vibrant presence, and would continue to be so beyond 802,161
an ever greater proportion of business gradually was taken up at Fulda
by its own notaries. Thus, as at Wissembourg, we can detect patterns
that testify to the elaboration of a distinctive and autonomous monastic
archival consciousness in the late eighth century.
At first glance, the practices of the church of Freising appear to con-
trast sharply with those at Fulda and Wissembourg. One does not observe
at Freising a transition from lay and diocesan notaries to a reliance on
monks. From the earliest preserved charters, priests dominated the notar-
ial culture of Bavaria and represent a significant proportion of the donors
as well. This might be held up as evidence that the written word in Bavaria
had been the exclusive preserve of the clergy. While these priests necessar-
ily would have been canonically subordinate to the bishops of Freising, it
is doubtful that they operated on behalf of the diocesan administration per
se. Although Theodor Bitterauf, who edited Freising’s charters, assumed
158 UF 39. On Geldersheim, see Weidinger, Wirtschaftsstruktur, 147.

159 UF 11, 18, 22, 23, 24, 26, 28–31, 33, 37, 38, 40–2, 44, 48–50, 52–5, 58, 59–61, 63–4,
66, 69, 70–2, 76, 80–2, 84, 86, 87–8, 89. Those undertaken by monastic notaries are
in italics.
160 In the portions of the cartulary dedicated to the mid-Rhine region, the monastic
notaries were responsible for these transactions made between 785 and 802: UF 157–
8, 167, 182, 184, 190–1, 210, 213, 215–20, 236–7, 259, 261, 268, 277–9. The non-
monastic notaries were responsible for UF 160–3, 177–80, 185, 187, 195–7, 231, 245,
248–57, 260, 263, 266–7, 270, 281, 283–5.
161 The Mainz notary Theotrich, for example, remained a prominent notary until 815
(CDF 218, 222, 224, 244, 246, 270).
226 Hans Hummer
that the earliest notaries were tied closely to Freising,162 their behaviour
points to a more subtle relationship between Freising, its patrons and the
clerical writers.
The two earliest charters preserved in the Freising cartulary were com-
posed in 744 and 748 by the priest Benignus, who appears to have oper-
ated as a free notary of sorts for the donors.163 He tells us in the charter of
a certain Moatbert that the donor ‘asked this donation to be done’, with
‘his wife Totana consenting and affirming’, and that he wrote ‘with Duke
Otilo confirming’. In the second, Benignus claims that he ‘wrote, hav-
ing been asked and sought out’, presumably by the donor, Amilo, ‘who
asked this traditio to be done’, though ‘after this duke Tassilo confirmed
this epistula with his own hand’. Arbeo, the future bishop of Freising,
behaved similarly. He was ‘asked and sought for’ four times between 754
and 757, and in one of those, his notarial statement was followed by
Duke Tassilo’s confirmation of ‘this epistula’.164 Two years later in 759,
Arbeo claims to have written at the ‘command of the glorious duke’.165
Who ‘asked and sought for’ his expertise is not always spelled out, but
in yet another charter, which was transacted in 769 after he had become
bishop, Arbeo reveals that he ‘commanded this charter to be made from
the petition’ of the donor.166 Arbeo’s drafting of the donation of a certain
Haholt in 758 underscores the joint agency of bishop and donors behind
the production of Freising’s charters: ‘Arbeo priest wrote this charter at
the command of Bishop Joseph and at the request of the donors Haholt
and his son Arno.’167
The foregoing examples illustrate that early on the Bavarian dukes
frequently were integral to the documentation of property transactions,
sometimes literally so. Duke Tassilo strongly implies that he wrote and
signed one of his own charters (‘ego Tassilo propria manu mea scripsi et
confirmavi’), and possibly a second.168 It was more usual for the duke
to appear in the document to bolster the transaction with his confir-
mation, in one case apparently by lending his own scribe. Sometime
between 769 and 777, the priest Fater ostentatiously presented himself
as a ducal scribe, ‘the capellanus of lord Tassilo duke of the Bavarians’,
in the donation charter of a certain Rihhart. ‘Having been asked by the
162 TF, pp. xxxviii–xxxix. 163 TF 1–2. 164 TF 7–10.

165 TF 14. 166 TF 31. 167 TF 11.
168 TF 35. See also TF 3: ‘Imprimis Tassilo propria manu signum fecit’. Although neither
charter bears any other indication of a notary, Bitterauf asserts that Tassilo did not
write these charters, that the duke’s statements mean only that he affirmed them. This
might be the case for TF 3, but compare Tassilo’s statement in TF 35 (‘ego Tassilo
propria manu mea scripsi et confirmavi’), with Alprich’s, whom Bitterauf classifies as a
Freising notary (TF, p. xxxix), in TF 6: ‘Ego indignus presbiter rogatus et petitus hanc
donationem scripsi manu mea’ (emphasis mine).
priest Isanhart and the layman Rihhart, he wrote this donation, so that it
might remain stable and unshakable’, a transaction that had been ‘done
in the home of St Pancratius’; that is, in the local church in the village of
Steinhart.169 According to the text of the charter, Rihhart gave the prop-
erties for the support of the church, so it might be that the priest Isanhart
was his brother – as the last element of their names, -hart, hints – and
possibly the pastor of St Pancratius. In any event, Rihhart’s use of Fater
as scribe signals his closeness to Duke Tassilo. On the other hand, Fater
appears elsewhere among the retinue of the bishops of Freising,170 thus
underscoring the ways in which the dukes, the bishops and the patrons all
might appeal to the services of clerics, whose roles could be redefined as
the circumstances warranted. Indeed, the deployment of notaries could
be startlingly ad hoc, as in the episode of a certain David, who made a
donation sometime between 758 and 763 to a church in the Bavarian vil-
lage of Puch. A priestly witness, Reginperht, was pressed into service as
scribe, ostensibly because he was available and could write: ‘Ego Regin-
perht rogatus fui ad scribendun. Non scripsi quomodo volui, sed sicut
potui.’171
Other priestly notaries arose from the surrounding kin groups, pre-
sumably pastored local churches, and wrote up their own donations.
The priest Willahelm, who was part of a nexus of groups tied to one
another through the church of St Zeno at Isen, made a donation in
769 to St Zeno’s and then wrote up the charter.172 Sometime between
769 and 776, the brothers Alprich and the deacon Ascrich made a joint
donation that was to pass to Freising’s control only upon the deaths of
Alprich; Ascrich; their two sisters, Marchraat and Uualtraat; and their
three nepotes, Amiloni, Uuisurihhi and Angilrata. The transaction was
then written up by Ascrich himself, who dated the charter not to the year
of the ruler, but to the eternal reign of ‘our Lord Jesus Christ’.173 Simi-
larly, the priest Oadalger made a donation in 770 of property inherited
from his father Cello, a gift made, as with so many of the early charters
of Freising, with the duke’s consent. Oadalger drafted the charter and
affirmed the witnesses (‘ego Oadalger donator scripsi et subscripsi pro-
pria manu et testes firmavit’), though by all appearances a second draft
was composed by his father Cello (‘ego Cello manu propria scripsi et
169 TF 37. 170 TF 7, 14a. 171 TF 12.

172 TF 33. Willahelm stood as witness at the church of St Zeno when Haholt consecrated
his son Arn, the future Carolingian missus and archbishop of Salzburg, to the church of
Freising (TF 11). Willahelm’s own donation to St Zeno’s was witnessed by the priest
Alpolt, who was among the donors to St Zeno’s and a scribe for a donation to Freising
transacted in the church of St Zeno (TF 4, 32).
173 TF 36.
228 Hans Hummer
ipse tradidi’).174 And finally, the priest Uuatto seems to have been in the
retinue, not of the bishop, but of a local lord, ‘having been commanded
by my lord Machelm’ to write up his charter in 776.175 These examples
indicate that if the diocese was the probable source of scribal education,
a demand for scribal talent arose from within patron groups too.
This more diverse group of notaries began to dwindle beginning in
the 770s during Arbeo’s episcopacy (764–83). The notaries henceforth
almost always appear to be closely tied either to Freising, or its suffragan
monastery, Schlehdorf.176 This change can also be detected in notarial
formulas now deployed to underscore episcopal agency. The notaries’
claims that they had been ‘asked and sought’ (‘rogatus et petitus’) went
extinct at the end of Bishop Joseph’s tenure in 763.177 That is, the usage
was phased out by Joseph’s successor, Arbeo. Moreover, although in
charters up to 770 it was more common for the patrons to claim that
they had ‘asked this [charter of] donation to be made’ (‘hanc dona-
tionem/traditionem fieri rogavit’), than for the notaries to claim that the
bishops had ordered them to write, nearly all references to patron requests
vanish after 776.178 Conversely, only after 772 – that is, at roughly the
same time – do notaries consistently claim that they wrote at the com-
mand (jussus) or ‘from the mouth’ (ex ore) of the bishop.179 Whether
this development means that patrons were less assertive is impossible
to say, though unlikely. The ‘charter’ of a certain Sigifried and his son
Erchanfrid, for example, was ‘ordered to be done’ in 769 by Arbeo, but in
response to the ‘petition of the aforesaid’ donor, an indication that donor
agency probably persisted beneath the formulas.180 But even this episode
places the accent on the authority of the bishop, who is the recipient of a
plea, after which he then orders the notary to act, thus underscoring the
174 TF 39. 175 TF 74. 176 Cf. TF, pp. xxxix–xl.

177 Or similar formulations, such as rogitus and rogatione. Patron rogation appears in TF 2,
6–8, 9b, 10–12, 19.
178 In the forty charters up to 770, patron requests appear in twelve transactions (TF 1–2,
6, 11–12, 13b, 15, 17, 25, 28, 30, 37), and episcopal commands in eight cases (TF
11, 19, 21, 23, 24c, 31–2, 38). A ninth episode occurs in TF 16; but see the following
note.
179 The first of these occurred in 763 towards the end of Joseph’s abbacy (748–64): TF
21. A second appeared in 760 (TF 16), but the record of this transaction is a breve
commemoratorii and thus would appear to have been composed sometime later. In a
third case, Joseph commanded a charter to be made, but at the request of the donor
Haholt and his son Arn in 758 (TF 11). The phenomenon of episcopal command
is much more indicative of charters transacted during Arbeo’s tenure (764–83), but
especially after 772. Beginning with TF 42 (a. 772), the episcopal command appears at
a rate of every other transaction. The incidence is probably greater because many of the
charters lacking the episcopal command were copied into the cartulary without their
witness lists, immediately after which is where the episcopal command often appears.
180 TF 31.
transformation of the church of Freising’s institutional consciousness in

the late eighth century.
The examples of Wissembourg, Fulda and Freising illuminate lines of
evidence detectable elsewhere and which also intersect in the late eighth
and early ninth centuries. Some of the earliest attempts to book charters
are traceable to the late eighth century in Bavaria and Rhaetia.181 The
pattern is also evident at St Gall where we do not encounter the orga-
nization of charters into codices, but whose records nonetheless reveal
an initiative beginning around 770 to organize charters with dorsal nota-
tions and by 840 to group them by district.182 The materials we have
before us, therefore, are not accidents, documents that have arbitrarily
survived the ravages of time and that we are somehow fortunate to have,
but rather records present to us because of profound archival innovations
that saw to their preservation.
Frankish cartularies weave a vivid tapestry of lay and ecclesiastical
interests. They testify both to the production of charters for individuals,
local churches and larger ecclesiastical institutions, and to the develop-
ment of a radical archival consciousness within monasteries during the
Carolingian era that ensured their preservation. If the charters bear wit-
ness to the eventual dominance of document production and preserva-
tion by monasteries beginning in the late eighth century, they also permit
a tantalizing glimpse into the transitional era between the post-Roman
period and the Carolingian period when multiple centres of notaries –
local, court, diocesan and monastic – operated.
The observation that monasteries became crucial for the preservation
of records is to be distinguished from the documentation that once existed
beyond the cloister. Vestiges of that evidence in the cartularies demon-
strate the movement of documents between patrons, village churches
and monasteries, and between laypeople themselves, the incidence of
whose sales and copies of transactions remained fairly uniform through-
out the early medieval period. Moreover, courts and episcopal authorities
certainly continued to generate documentation. The difference was the
memorial ethos which compelled monks to preserve charters as memo-
rial records forever, and which, more broadly, went hand in hand with the
construction of the Carolingian political order.
This pivotal development in European history might explain the phe-
nomenon of Cluny and its unmatched archive. If monasteries in the
181 Geary, Phantoms, 88; P. Erhart and J. Kleindinst, Urkundenlandschaft Rätien (Vienna,
2004), 21–30.
182 Above, Chapter 7, and Peter Erhart, ‘Carta ista amalfitana est et nescitur legere: The
Charters of Cava dei Tirreni and St Gall and Their Evidence for Early Medieval
Archival Practice’, Gazette du livre médiéval, 50 (2007), 27–39.
230 Hans Hummer
Carolingian period had difficulty reconstructing their earliest holdings,

the monastery of Cluny had little trouble at all. Looking back to the
foundation in 910 from the eleventh century when the Cluniacs began
to assemble their cartularies, the monks easily located founding doc-
uments and could assemble a dense record of donations from the
foundation’s first century, a stark contrast with, say, Fulda, Freising
and Wissembourg.183 Moreover, because monasteries had by then long
become crucial to the perpetuation of aristocratic power, Cluny found
itself the repository of a trove of sales between laypersons that had been
transacted in the century before its founding.184 The difference was that
Cluny was founded at a time when the parameters of monastic records
management had become firmly entrenched so that Cluny in effect was
born with a memorial consciousness.
Perhaps it should not be a surprise then that the first extant indepen-
dent lay record in the European tradition, the Codex Falkensteinensis,185
emerged in the twelfth century after the Church reforms that distin-
guished clerical from lay authority, and bears witness not to a qualitatively
distinct lay method for organizing property somehow submerged for cen-
turies beneath an alien clerical culture. Rather, it was organized after the
manner of a monastic cartulary, which had established the paradigm not
merely for an enduring, albeit ever-evolving form of record-keeping, but
also for the memorializing of property rights within an institutionalized
family consciousness.
183 See S. Barret, La mémoire et l’écrit: L’abbaye de Cluny et ses archives (Xe –XVIIIe siècle)
(Münster, 2004), 27–144.
185 E. Noichl, Codex Falkensteinensis: Die Rechtsaufzeichnungen der Grafen von Falkenstein
(Munich, 1978); cf. J. Freed, The Counts of Falkenstein: Noble Self-Consciousness in
Twelfth-Century Germany (Philadelphia, 1984).
9 The laity, the clergy, the scribes and their
archives: the documentary record of eighth-
and ninth-century Italy
Marios Costambeys
Ecclesiastical institutions dominate the transmission of documents in the Lom-

bard and Carolingian kingdoms of northern and central Italy, just as in the rest
of the post-Roman West. While the Italian evidence shows straightforwardly
that it was increasingly professional lay charter scribes who produced docu-
ments, analysis of surviving collections also uncovers other laypeople involved
in the preservation of documents. Lay dossiers ended up in the hands of churches
and monasteries in various ways, and not only because of those institutions’
desire to reinforce their property claims; where such a desire was the object of
preserving documents, other means were also available, such as the confirma-
tion of property rights by a superior authority, or the validation of documents
– and the production of new documents – by a law court. The Italian evi-
dence shows how those who used charters, in a shifting legal landscape in which
rights were never fixed or secure, provoked the development of these and other
ways of managing documents on the part of both ecclesiastical and secular
institutions.
Every document that survives to us from the Lombard and Carolingian

kingdoms of Italy has spent most of its life in a church or monastery.1 This
simple fact certainly sets apart those institutions from other entities, such
as lay families, in one respect: they proved more successful at preserving
documents through eras subsequent to when those documents were writ-
ten. Whether they were also distinctive institutions in that initial, eighth-
and ninth-century period is a more complicated question. The practical
1 Additional abbreviations: CDA: Codex diplomaticus Amiatinus: Urkundenbuch der Abtei

S. Salvatore am Montamiata von den Anfängen bis zum Regierungsantritt Papst Innozenz
III. (736–1198), ed. W. Kurze, 4 vols. (Tübingen, 1974–82); CDV: Codice diplomatico
veronese, ed. V. Fainelli, 2 vols. (Venice, 1940–63); CF: Il Chronicon Farfense di Gregorio
di Catino, ed. U. Balzani, 2 vols. (Rome, 1903); Manaresi: I placiti del Regnum Italiae, ed.
C. Manaresi, 3 vols. (Rome, 1955–60).
231
232 Marios Costambeys
existence of the Church as a single institution in Lombard and Carolin-

gian Italy can very easily be questioned,2 but the clergy continued to be
what they had been for some time: an identifiable group with their own
status, privileges and preoccupations. It was not simply the unique ability
of consecrated priests to bring salvation closer through communion that
automatically imbued them with this common charisma: every clerical
grade – including monks – is generally held to have enjoyed the same,
discrete, cachet.3 The ninth century saw strenuous efforts to reinforce
the lay–clerical distinction,4 and while these may indicate contradictory
trends – either that the barrier between the two was being eroded, or that
its basic permeability was increasingly being recognized – they certainly
affected the historical record that survives to us. As some communities of
clerics – individual ecclesiastical institutions such as some monasteries –
developed more robustly distinct characters, and clerical identities
became increasingly solid, so improved both their capacity to preserve
written records and their tendency to preserve those that best affirmed
their identity. This is most evident in the beginnings in the ninth cen-
tury of the ‘cartulary tradition’, but it also greatly affected archives of
originals: ultimately, a significant role of both was that of repository of
the institution’s proofs of title.5 In short, northern and central Italy in
the eighth and ninth centuries saw decisive developments in the pro-
duction and preservation of documents, developments that can best be
understood by focusing on the lay–clerical divide.
In the same way that most manuscript codices have ecclesiastical prove-
nances, single-sheet documents also survive because they were ultimately
kept by ecclesiastical institutions. This is as true for Italy as for anywhere
else in western Europe, as is clear from a brief perusal of the provenances
in a printed collection such as the Codice diplomatico longobardo.6 Just
2 E.g. M. de Jong, ‘Charlemagne’s Church’, in J. Story (ed.), Charlemagne: Empire and

Society (Manchester, 2005), 103–35, esp. 106.
3 Explicit treatment of clerical difference: R. Reynolds, ‘The Organisation, Law and
Liturgy of the Western Church, 700–900’, in R. McKitterick (ed.), New Cambridge
Medieval History, II, c.700–c.900 (Cambridge, 1995), 589–622, at 605–13.
4 Most obviously evident in the Pseudo-Isidorian forgeries: see now the comprehen-
sive texts and studies at www.pseudoisidor.mgh.de. See also K. Zechiel-Eckes, ‘Pseu-
doisidorische Dekretalen’, in Lexikon der Kirchengeschichte, 2 vols. (Freiburg, 2001), II,
1345–6, which modifies the classic survey of H. Fuhrmann, Einfluss und Verbreitung der
pseudoisidorischen Fälschungen, 3 vols. (Munich, 1972–4). For succinct treatments unin-
formed by the most recent research, see H. Fuhrmann and D. Jasper, Papal Letters in the
Early Middle Ages (Washington, DC, 2001), 135–95, and Reynolds, ‘The Organisation,
Law and Liturgy’, 616–17.
5 Proof of title was of course not the only reason for the preservation or copying of charters:
Geary, Phantoms, 81–114.
6 CDL I and II for non-royal documents from the Lombard kingdom proper down to 774;
CDL III for royal diplomas to 774; CDL IV/1 for diplomas of the dukes of Spoleto to
Eighth- and ninth-century Italy 233
as much as Francia, Italy exhibits a growing clerical self-consciousness,

including a concern to shape institutional memory that is evident, for
instance, in the canon law collection known as the Collectio Anselmo ded-
icata, compiled (if that is the right word) in Milan in 882×96, but that
also affected archival practices.7 Italian medievalists have not always been
alive to the significance of this essential fact. Thus Armando Petrucci,
while acknowledging the competence of numbers of laymen in subscrib-
ing charters in their own hand (which he equated with ‘literacy’),8 still
saw literacy dominated by the clerical subscribers who are a distinct
majority in those charters he examined, without recognizing that this
predominance may be due to the charters’ subject matter and places of
preservation.9 It is hardly surprising, to take the most egregious example,
that the witness lists of charters concerning the cathedral of Lucca should
be dominated by its clergy,10 and this impression is confirmed if we look
at subscriptions to the eighteen ‘private’ charters in our most easily iden-
tifiable ‘lay’ archive, the Campione dossier discussed below, in which
786×7; CDL IV/2 for diplomas of the dukes of Benevento; CDL V for private charters
from the duchies of Spoleto and Benevento down to 787.
7 On the Collectio Anselmo dedicata, see L. Kéry, Canonical Collections of the Early Middle
Ages (ca. 400–1140): A Bibliographical Guide to the Manuscripts and Literature (Wash-
ington, DC, 1999), 124–8, and K. Zechiel-Eckes, ‘Quellenkritische Anmerkungen zur
“Collectio Anselmo dedicata”’, in W. Hartmann and A. Grabowsky (eds.), Recht und
Gericht in Kirche und Welt um 900 (Oldenburg, 2007), 49–65.
8 Petrucci’s equation of literacy with the ability to sign one’s name in this way has of
course given way to an appreciation of the diversity of literate capability: for the point
in relation specifically to late antique and early medieval Italy, see G. Cavallo, ‘Dal
segno incompiuto al segno negato: Linee per una ricerca su alfabetismo, produzione e
circolazione di cultura scritta in Italia nei primi secoli dell’impero’, in A. Bartoli Langeli
and A. Petrucci (eds.), Alfabetismo e cultura scritta nella storia della società italiana: Atti
del Seminario tenutosi a Perguia il 29–30 marzo, 1977 (Perugia, 1978), 119–45, at 126–8;
in general, McKitterick, Carolingians, 271–3.
9 A. Petrucci, ‘Libro, scrittura e scuola’, in La scuola nell’Occidente latino dell’alto medioevo,
Settimane 19 (Spoleto, 1972), 313–37, at 325–6 (trans. as ‘Book, Handwriting and
School’, in A. Petrucci, Writers and Readers in Medieval Italy, trans. C. M. Radding [New
Haven, CT, 1995], 59–76, at 67–8) examined 180 original charters from the Lombard
kingdom dating between 720 and 774, containing 988 subscriptions, of which 326
are autograph (32.7%). Of the 988 subscriptions, 355 (35.9%) include clerical titles;
assuming for the moment that all those who could designate themselves as clerics did
so, this leaves 633 lay subscriptions. Of the latter, 93 are autograph: 9.4% of the total
and 14.7% of the lay subscriptions; of the 355 clerical subscriptions, on the other hand,
233 are autograph: 65.6%, and 23.5% of the total. Of the 326 autograph subscriptions,
71.4% are by clerics, 28.6% by laymen. It is important to remember that the number of
subscriptions is not the same as the number of subscribers, since the latter could, and
in a few cases demonstrably did, witness more than one charter: see, e.g., P. Supino
Martini, ‘Le sottoscrizioni testimoniali al documento italiano del secolo VIII: Le carte
di Lucca’, Bullettino dell’Istituto storico italiano per il medio evo e archivio muratoriano, 98
(1992), 87–108.
10 E.g. (among many) CDL I 99 (= ChLA XXXI 931).
only eleven of the seventy-one charter witnesses (15.5%, compared with

Petrucci’s average of 35.9%) give themselves or are given clerical titles
(in a collection in which all but five witnesses have some sort of designa-
tion, mostly secular epithets like vir devotus or vir honestus).11 The latter
collection does show, however, that although the ecclesiastical colouring
of our survivals exaggerates the absolute number of literate clergymen
visible to us, the rate of clerical literacy may still have been somewhat
higher than that among laymen: the twenty-eight autograph subscribers
in the Campione charters include ten clerics; of the remaining forty-
three, subscribing with signa manus, only one bears a clerical title. This
should not surprise us, of course, given the straightforward requirements
on the priesthood and monasticism as occupations, and it bears on the
question of lay literacy that I shall consider shortly, but equally it should
not detract from our recognition of the distorting effect that the circum-
stances of preservation exert on the documentary record. Its evidence
makes society look more ecclesiastical than it really was.
A striking paradox appears when the clerical monopoly on preservation
is set alongside another feature of the Italian documentary record: the
growing dominance of laymen over the creation of documents. Increas-
ingly through the ninth century, charter scribes were laymen. This is evi-
dent because they, along with the majority of others who set their hands to
charters, took care to give themselves some kind of title, and in particular,
it seems, to distinguish themselves as clerics when appropriate.12 This
becomes palpable in instances where we can trace promotion through the
clerical hierarchy in the self-designations given by issuers and witnesses.
Thus we can see Teudo graduate from archipresbiter of the see of Rieti to
its bishop, or Audo progress from subdiaconus to diaconus to archidiaconus
to bishop in Verona.13 This attention to clerical identity extends to the
11 A full breakdown is given in A. Bartoli Langeli, ‘I documenti’, in S. Gasparri and C. La

Rocca (eds.), Carte di famiglia: Strategie, rappresentazione e memoria del gruppo familiare
di Totone di Campione (721–877) (Rome, 2005), 237–64, at 256–61. I take a more
optimistic view of the regularity and reliability of indications of clerical status than does
N. Giovè, ‘In margine al dossier di Totone di Campione’, in Gasparri and La Rocca
(eds.), Carte di famiglia, 265–82, at 275.
12 For scribes’ self-designations in the charters to 774, see N. Everett, Literacy in Lombard
Italy (Cambridge, 2000), 200–1, who finds scribes giving themselves some kind of
designation in 175 out of around 260 cases.
13 Teuto: CDL V 3 (= RF V 1220, a.739) (archipresbiter); CDL V 20, 22 and CDL IV/I 14
(= RF II 34, 37, 46; a.753, 755, 761) (episcopus). Audo: CDV 131 (a.829: subdiaconus);
CDV 153, 172 (a.839, 849: diaconus); CDV 197, 201 (a.855, 856: archidiaconus); CDV
217, 219, 226 (a.860, 860, 862: episcopus). Audo’s episcopate lasted 860–9: see M.
Miller, The Formation of a Medieval Church: Ecclesiastical Change in Verona, 950–1150
(Ithaca, NY, 1993), 151 with n. 45.
writers of charters.14 In the Farfa archive, for instance, scribes who were
clerics all designated themselves as such, though sometimes alongside the
term notarius, and variations in these designations are easily explicable by
their progress from one grade to another (and/or their monachization).15
The term notarius, unqualified, is the norm for a lay scribe in the Farfa
archive, pointing towards the generally less regular use of designations
by lay charter scribes.16 Across northern and central Italy as a whole,
there is enormous variety of titles for lay scribes: Costamagna’s survey
revealed twenty-three different designations.17 It has nevertheless been
generally accepted that in the absence of a qualifying clerical designation
such titles do refer to laymen. The rising incidence of such terms through
the course of the ninth century has been seen as the corollary of the
widely observed decline of the clerical charter scribe in most parts of
Italy in that period.18
At Lucca this change has been linked directly to the Frankish takeover:
charter scribes who had belonged to the Lombard episcopal clergy gave
way, it is suggested, to notaries serving the Frankish counts.19 But in
many parts of Francia itself the task of charter-writing was actually mov-
ing in the opposite direction – from secular to ecclesiastical scribes –
in precisely this period.20 In Lucca, meanwhile, the shift was still not
complete more than two generations after the conquest, and in any case
the association between the city’s notaries and the bishop before 774 was
unusually close. Comparison with the charters preserved at the abbey of
S. Salvatore on Monte Amiata, fortunately all also extant in the original,
reveals a gradual and likely quite complex development, in which the
14 H. Zielinski, Studien zu den spoletinischen ‘Privaturkunden’ des 8. Jahrhunderts und ihrer

Überlieferung im Regestum farfense (Tübingen, 1972), 137–41; H. Keller, ‘Der Gerichtsort
in oberitalienischen und toskanischen Städten’, Quellen und Forschungen aus italienischen
Archiven und Bibliotheken, 49 (1969), 1–72, esp. the table on 13.
15 E.g. Palumbus in CDL V 5 (a.745: diaconus), 7 (a.746: diaconus), 25 (a.757: monachus),
27 (a.757: presbiter et monachus); or Theuferius in CDL V 52 (a.768: presbiter notarius),
63 (a.773, presbiter notarius), 75 (a.777: presbiter) – here he wrote his own charter offering
himself to Farfa – which might explain 81 (a.778: presbiter et monachus) and 87 (a.778:
presbiter et monachus); in the latter, though, he also calls himself scriptor. See Zielinski,
Studien, 130, 135.
16 Zielinski, Studien, 137–41.
17 G. Costamagna, ‘L’altomedioevo’, in M. Amelotti and G. Costamagna, Alle origini del
notariato italiano (Rome, 1975), 147–314, at 157, and, for the Lombard period, Everett,
Literacy in Lombard Italy, 199–207.
18 Zielinski, Studien, 140–1; Keller, ‘Der Gerichtsort’, 9–19, 31. Piacenza at least retained
some kind of ecclesiastical notariate: F. Bougard, La justice dans le royaume d’Italie de la
fin du VIIIe siècle au début du XIe siècle (Rome, 1995), 69 and n. 11.
19 A. Mailloux, ‘Modalités de constitution du patrimoine épiscopal de Lucques, VIIIe –Xe
siècle’, in Les transferts, 701–23, esp. 710–12.
20 See above, Chapter 8.
turning point comes around 840. The 166 genuine acts datable from
the collection’s inception in 736 down to 900 were written by eighty-five
scribes. Of these, twenty (23.5%) gave themselves clerical designations.
But of the twenty-four scribes active in the period 840×900, only one
did so (4.2%).21 By comparison, three of the fifteen scribes (20%) of the
Campione dossier, dating from 721 to 863, were clerics.22 The laicization
of the Italian charter scribe was uneven and drawn out, then, because of
a number of factors: the deliberate encouragement of Frankish counts,
certainly, but also a more general proliferation and professionalization of
legal writing of this kind.
While there is little difference in terms of graphic ability between lay
and clerical charter scribes, Petrucci and Romeo identified two basic lev-
els of script among charter subscribers, characterized by competence in
the new Italian cursive on the one hand, and in a more basic elemen-
tary minuscule on the other.23 Part of their case was an interpretation
of the subscriptions in the Campione dossier, though a recent detailed
treatment of the same documents has shown that variations in graphic
competence did not neatly follow the lay–clerical distinction.24 Never-
theless, both studies show conclusively that education in a level of reading
and writing necessary to engage in documentary culture was widespread
and did not vary in its basics across those parts of Italy from which our
evidence survives. Functional literacy in documents observed no obvious
lay–clerical division.
The fact that endowments of monasteries could be recorded in quite
humble documents – mere scribbled scraps, donating equally paltry
pieces of land25 – has been rightly interpreted as revealing the impor-
tant social and economic roles of churches and monasteries in Italy, as
in Francia.26 It also incidentally reveals that written documents in gen-
eral were deeply familiar and often very mundane. Just how familiar and
mundane needs underlining. From northern and central Italy (down to
and including the modern regioni of Lazio and Abruzzo), there survive
to us dating from the years between c. 680 (when the run of documents
21 The data are recoverable from the lists and tables in CDA I, pp. xxiv–xxv, and CDA IV,
pp. 178–86.
22 Bartoli Langeli, ‘I documenti’, 239.
23 A. Petrucci and C. Romeo, ‘Scriptores in urbibus’: Alfabetismo e cultura scritta nell’Italia
altomedievale (Bologna, 1992), 23.
24 Petrucci and Romeo, ‘Scriptores in urbibus’, 57–63, and see now Giovè, ‘In margine al
dossier di Totone’, 272–9.
25 E.g. CDA 36.
26 E.g. Mailloux, ‘Modalités’, and B. H. Rosenwein, ‘Property Transfers and the
Church, Eighth to Eleventh Centuries’, in Les transferts, 563–75; see also Costambeys,
Power.
really begins)27 and 1000 some 7,500 documents that were not issued
by a ‘public’ authority.28 Two-thirds of these are single-sheet documents,
90 per cent of them originals, while the other third comes from a handful
of monastic cartularies, principally those of Farfa and Casauria. Exclud-
ing the latter, plus the smattering of registers and extracts, the figure for
survivals is broadly comparable with that for Catalonia.29 There are two
major differences, however: northern and central Italy covers a rather
larger area, of course, but it nevertheless also preserves no significant
archives of lay public figures to set alongside the comital records of
Catalonia. The Italian survivals come from a minimum of fifty-three
archives, all of them ecclesiastical. This fact is perhaps itself sufficient to
explain why it has often been assumed that lay documents survive only
rarely from Italy before the millennium. Moreover, we cannot benefit,
as we can north of the Alps, from formula-books that indicate a preva-
lence of purely ‘lay’ documentary forms (at least in theory, as Brown
argues). The earliest formula-book from north-central Italy, the Cartu-
larium Langobardicum, dates from the late eleventh century.30 But we
do have a number of archives that are sufficiently substantial to include
a diverse range of material, offering the possibility of discovering some
of the documentary practices of laypeople.31 The largest single reposi-
tory by far is the archiepiscopal archive in Lucca, which contains over
1,800 pre-1000 documents, nearly all originals. There are also over 900
pre-1000 documents in the Casauria collection, and nearly that number
in Farfa’s cartularies. The abbey of S. Salvatore on Monte Amiata pre-
served about 180 pieces (now mostly in the Archivio di Stato in Siena).
These mean that central Italy – Tuscany and Spoleto – is by far the
best-represented area. Next best is Emilia-Romagna, with about 650
documents extant in Piacenza and about 300 in Ravenna. There are
27 The only original Italian documents dating from before 680 are the Ravenna papyri,
edited in Tjäder and in ChLA: see the concordance in ChLA XXIX, p. 10. The earliest
extant charters from Lombard Italy are copies: CDL I 4 (Siena vs. Arezzo); CDL III
1–3, 5 (Bobbio); 4, 6 (Piacenza vs. Parma).
28 The figure is that of F. Bougard, ‘Actes privés et transferts patrimoniaux en Italie centro-
septentrionale (VIIIe –Xe siècle)’, in Les transferts, 539–62, at 539. Definitions here are
of course quite difficult, given the multiplicity of powers who could and did claim some
sort of ‘public’ authority in Italy over this long period. So any number must remain
approximate.
29 As Bougard points out (‘Actes privés et transferts patrimoniaux’, 540), this is even more
the case if we exclude cartulary copies, regests and the like. For Catalan figures, see
Kosto, ‘Laymen’, 52.
30 Above, Chapter 5; MGH LL 4:595–602. For the date, I follow the convincing arguments
noted by Bougard, La justice, 308 with n. 4.
31 For what follows, full publication details down to 1995 are given in Bougard, La justice,
76–108. The principal change since then is the continuing publication of ChLA.
about another 600 from the whole of Lombardy (principally from Milan,
with fewer in Brescia, Bergamo and Cremona), and about 200 from the
Veneto (mostly from Verona).
‘Lay’ documents and their preservation

in Italian archives
We do not have to penetrate far into this material to find abundant traces
of lay documentary practices. I adopt here a definition of a ‘lay docu-
ment’ as one in which neither party carries an explicit clerical designation.
Thus I have not tried to separate documents involving secular officials
because, while clerics are generally straightforward to identify, it is not
always easy, in early medieval Italy’s amorphous and shifting institutional
context, to define where the official bureaucracy ended and ‘private’ lay
families began. A similar problem exists with archives. There has been
an assumption that where lay documents (as just defined) do exist, they
survive because they were of interest to the ecclesiastical institution that
absorbed them, with the corollary assumption that the landed property
to which they generally relate followed the same trajectory. We shall see
below that this can sometimes be demonstrated, but should underline
here that the question does always need to be addressed, that we should
be wary of drawing conclusions about the extent of the kind of compe-
tence with documents that was sufficient to commission, comprehend or
preserve them simply from the places where those documents ended up
(which in turn tends to have dictated how they are presented to us), and
that a single explanation does not cover all instances of lay documents.
The need for complex and properly informed explanation becomes
apparent as soon as we start to enumerate the number of documents in
our archives that satisfy our basic definition of ‘lay’ status (both parties
being non-clerical). Table 9.1 below shows the incidence of such lay
documents in four archives of originals – Lucca, Monte Amiata, Verona
and Bergamo – and one cartulary, that of Farfa, between the earliest
survivals (c. 680) and 850.
The prominent exception of the Farfa cartulary simply emphasizes
how important the process of selection was in the formation not just of
cartularies but also of the ecclesiastical archives from which they were
drawn: a point to which I will return. It should already be clear, though,
that some surviving archives include lay documents in sufficient numbers
to allow us to draw comparative conclusions. We can point not only to
reasons why some lay documents were transmitted through ecclesiastical
archives but also to differences in archival practices between the laity and
ecclesiastical institutions.
Table 9.1 Proportions of ‘lay’ documents in some Italian archives
Non-royal, ‘Lay’ ‘Lay’ Percentage

non-publica Originals documents originals ‘lay’
Lucca 625 625 31 31 5.0

Monte Amiata 112 111 31b 31 27.7
Verona 44 24 8 7c 18.2
Bergamo 13 11 5 4 38.5
Farfa 298 0 0 0 0
a I.e. not issued by rulers, and not court notices.

b One of these (CDA 41) includes as one party a primicerius, who, following Roman
precedent, could have been either clerical or lay.
c One is an immediately contemporary copy.
A fortunate survival from Lombard-period Italy is particularly reveal-

ing in this regard. Some time in Desiderius’ reign, the nun Ghittia was
the recipient of a collection of over a hundred documents, only the list of
which survives, giving, in most cases, the names of the issuers and recipi-
ents of each.32 Some of these were wholly ‘lay’ documents; others related
directly to the church of S. Pietro ai Sette Pini near Pisa; forty-four of
them concerned a single individual, Alahis, possibly a relative of Ghittia.
This is certainly a record of an ‘archive’, therefore, but one that was nei-
ther wholly lay nor entirely ecclesiastical. It may have come down to us
because the references to the Pisan church led to its preservation in the
episcopal archive of that city. But the number of types of legal document
listed is much larger than the average ecclesiastical archive preserved.
The largest number of types of transactions in the tables in François
Bougard’s survey of the private documents from the early medieval Ital-
ian kingdom is six (admittedly he runs a few types together);33 among
the seventy-two private documents in Ghittia’s list, on the other hand,
there are twenty different legal terms for transactions. The largest cat-
egory by far (thirty-four) is the sale, but we also have acquisitions of
mundium, deeds of thingatio (public confirmation of a legal contract) and
declarations of fidelity, none of which are at all common among surviving
documents.
The impression that collections of lay documents generally included
different kinds of transaction from ecclesiastical ones is reinforced by
the best-known and best-studied identifiable corpus associated with a lay
32 CDL II 295. See Everett, ‘Scribes and Charters’, 39.

33 F. Bougard, La justice, 65–108.
family, the dossier of the Totoni of Campione d’Italia.34 These twenty-

one charters, fourteen of which survive as originals, include ten different
types of transaction. Even if we exclude the royal or imperial diplomas,
we are left with records of the acquisition of free women’s mundia (three),
of the acquisition of slaves (three), of the loan of a slave (one), of a loan
of money (one), of a betrothal and dowry (one), and of compensation
for a killed slave (one), as well as the sales of land (three) and pro anima
and post obitum donations (two each) that are familiar from ecclesiastical
archives. What is perhaps most striking here is the emphasis on people
rather than on landed property. Not only do they record the transfer of
mundium (legal personality) of both free women and slaves (male and
female), but they also enact the sale of the latter.
As has already been hinted, sales are one type of transaction that pro-
vides an index of difference between lay and ecclesiastical archives. Sales
are consistently under-represented in ecclesiastical archives, apart from
exceptional cases where an institution acquired much of its property
through purchase, as did S. Clemente in Casauria, as examined below.35
While a few ecclesiastical institutions did engage extensively in purchase
and sale in building up and managing their patrimony, on the whole they
stuck to the letter of the prohibition on the alienation of Church land,36
utilizing legal devices including ad tempus transfers such as precaria, and
often very unbalanced exchanges, as their principal means of parting
with property.37 But when we have records of archives of laymen’s doc-
uments, they include large numbers of sales. A tabular arrangement of
the incidence of documents of sale in some of the main collections from
the period c. 680–850 reveals some striking patterns(Table 9.2).
Several features of Table 9.2 deserve notice. The first is simply that,
taken together, these data reveal that sales and purchases of land, goods
and people were numerous, and were activated by documents. We are
coming to recognize the importance of an economy of sale and purchase
in early medieval Europe, while at the same time appreciating that land,
and other goods, can often not be regarded simply as commodities in a
34 Fully examined in Gasparri and La Rocca (eds.), Carte di famiglia.

35 Bougard, ‘Actes privés et transferts patrimoniaux’, 543.
36 On this very venerable and problematic prohibition, see now S. Wood, The Proprietary
Church in the Medieval West (Oxford, 2006), 9–11, 19–20.
37 On precaria, see L. Morelle, ‘Les “actes de précaire”, instruments de transferts patrimo-
niaux (France du Nord et de l’Est, VIIIe –XIe siècle)’, and P. Depreux, ‘L’apparition de
la précaire à Saint-Gall’, and for the peculiarities of Italian precarial contracts, L. Feller,
‘Précaires et livelli: Les transferts patrimoniaux ad tempus en Italie’, all in Les transferts,
607–47, 649–73, 725–46; in general, see D. Ganz, ‘The Ideology of Sharing: Apostolic
Community and Ecclesiastical Property in the Early Middle Ages’, in W. Davies and P.
Fouracre (eds.), Property and Power in the Early Middle Ages (Cambridge, 1995), 17–30.
Table 9.2 Proportions of documents of sale in some Italian archives
Non-royal,
non-public, Percentage lay lay Percentage
documents Sales sales documents sales lay sales
Lucca 625 74 11.9 31 21 67.7

Monte Amiata 106 47 44.3 31 8 25.8
Verona 44 15 34.1 8 4 50.0
Bergamo 13 3 23.1 5 3 60.0
Farfa 191 28 14.7 0 0 0.0
Campione dossier 21 6 28.6 11 6 54.5
Ghittia lista 78 34 43.6 – – –
a Ghittia’s list does not give sufficient information for us to be able to distinguish ‘lay’
documents. Nevertheless, the incidence of sales within it offers a useful comparator.
profit-driven market, but changed hands according to a social at least as

much as an economic logic, in a society in which generosity, mutuality
and reciprocal obligation were important determinants of wealth.38 A
second feature is that Table 9.2 gives a reason for historians’ relative
lack of attention to sales until recently: their smaller proportion in the
larger collections of documents. That the larger proportion in smaller
collections is not simply a statistical quirk of those numbers is shown
by the percentage in Ghittia’s list and, above all, in the Monte Amiata
collection (on which, see below). The final notable feature is of course
the huge proportion of lay documents that are sales. This simple analysis
therefore reveals that sale and purchase, perhaps especially of land, needs
to be woven into our view of economic relations that have often been
dominated by a culture of gift-giving.
The imprint of land sales and purchases in ecclesiastical archives is
arguably strongest where that archive absorbed groups of documents
relating to properties that were eventually acquired by the institution.
We can see this quite clearly in cases from both the Lucca and Monte
Amiata collections. From Lucca we have charters revealing that between
742 and 752 a certain Crispinus bought land in several places, including
twice from laymen at Pescia near Lucca.39 With these he established
the church of S. Martino in Lunata, the future of which he provided
38 See now M. Innes, ‘Framing the Carolingian Economy’, Journal of Agrarian Change, 9
(2009), 42–58; Innes, ‘Practices of Property in the Carolingian Empire’, in J. Davies
and M. McCormick (eds.), The Long Morning of Medieval Europe: New Directions in
Early Medieval Studies (Aldershot, 2008), 243–69; Wickham, Framing, 216–18 and, in
general, 694–708, with references.
39 CDL I 80, 88, 102, 106 (the latter two not ‘lay’ documents).
for in 764: after the deaths of himself and his immediate heirs, it was
to pass to the bishop of Lucca.40 Similarly, the Monte Amiata archive
includes charters of 765 and 791 by which a certain Walcari obtained
from other laymen lands at Marano (near Sovana, southern Tuscany).41
These, evidently, he attached to the church that he had founded there,
half of the substance of which in 793 he gave to Monte Amiata.42 In both
these cases, the ‘lay’ documents passed through a limited series of fairly
clear contexts, from production in an entirely lay environment, through
an Eigenkirche, to a monastery. The creation of a dossier of documents
was here, then, part of the process of consolidating property around a
single focus, an Eigenkirche.
It is interesting to compare these cases with another evident in docu-
ments in the Monte Amiata collection. These relate to property at ‘Agello’
(mod. Gello, in the valley of the Orcia, quite close to Monte Amiata).
Four charters record sales of land at ‘Agello’ between different sets of
laymen in the period 750 to 786.43 In 814 and 827, Monte Amiata itself
bought land from laymen in the same place. Only one of these seems
to have been related to the previous purchasers.44 In other words, the
abbey seems to have received not a single dossier of documents but a
haphazard scatter of charters, some of which may have passed through
several hands before reaching it. The time period here is also interest-
ing, as Monte Amiata did not receive property at ‘Agello’ until much
more than thirty years after some of the documents relating to it were
issued, and therefore long after they would have had any legal force.45
It is difficult to determine straightforward reasons for the retention of
such documents. Initially, it is surely likely that they were brought into
Monte Amiata’s archive because the proprietary context was such that
they might still be called upon to bolster rights of ownership. But they
must soon have become obsolete, and their further retention is an inter-
esting testimony to the archival culture at institutions like Monte Amiata.
Perhaps it was simply that any document relating to important estates
(and ‘Agello’ certainly became that) was regarded as worth keeping.
Both the complex forces driving the retention of documents and the
importance of the traffic in landed properties emerge most clearly in the
40 CDL II 179. 41 CDA 12, 39: the former was a purchase, the latter an exchange.
42 CDA 42. CDA 91 reveals that this gift was evidently not unproblematic.
43 CDA 7, 9, 10, 33. 44 CDA 74, 101.
45 The ‘thirty-year rule’ in Lombard Italy: C. Wickham, ‘Land Disputes and Their Social
Framework in Lombard and Carolingian Italy, 700–900’, in Wickham, Land and Power:
Studies in Italian and European Social History, 400–1200 (London, 1994), 229–56, with
further references at n. 7 (orig. in Settlement of Disputes, 105–24); in general, E. Levy,
West Roman Vulgar Law: The Law of Property (Philadelphia, 1951), 176–90.
records preserved by the monastery of Casauria.46 Founded in 873 by the

emperor Louis II near the frontier between the Carolingian domain and
the southern, Lombard, principality of Benevento, Casauria preserved
an extremely rich record of its fund of documents, latterly through the
work of its exceptional twelfth-century archivist Giovanni di Berardo.47
In the course of the 1170s, Giovanni transcribed into a single manuscript
more than 2,000 documents, arranged into two sections.48 First were the
munimina: often abbreviated copies of documents that traced the pro-
prietorship of landed properties before their acquisition by Casauria and
were contracts mostly between two lay parties. They were arranged topo-
graphically, grouped according to the settlement boundaries of the ninth
century (that is, before the dramatic settlement shifts of incastellamento);
many evidently represent the remains of dossiers of documents belonging
to a variety of lay families. In the second section were copied, usually in
their entirety, the instrumenta: documents of which the monastery was
either author or beneficiary, arranged in rigorously chronological order.49
The munimina in particular therefore show signs of selection, edition and
abbreviation, some of which can be ascribed to the initiative of Giovanni
di Berardo; but we can also be fairly certain that there had been preceding
collections and redactions of the original documentary material, perhaps
in the form generally known as Liber traditionum.50 A number of people
intervened in the shaping of Casauria’s archive; it is unlikely that all were
pursuing the same motives or programme.
Layers of archiving serve to obscure the reasons why lay families’
dossiers came into the abbey’s hands, or why they were preserved in the
form they were, abbreviated and often shorn of dating evidence. Laurent
Feller’s work has focussed on the ninety-seven documents, spanning the
period c. 830 to 879, concerning the region of Vico Teatino, dominated
by the dossier associated with the family of Karol, son of Liutprand.51
46 L. Feller, Les Abruzzes médiévales (Rome, 1998), 63–84; L. Feller, A. Gramain and
F. Weber, La fortune de Karol: Marché de la terre et liens personnels dans les Abruzzes au
haut moyen âge (Rome, 2005).
47 L. Feller, Les Abruzzes, 69–74; Feller, ‘Le cartulaire-chronique de San Clemente a
Casauria’, in Les cartulaires, 261–77.
48 BnF lat. 5411. Facsimile edition: Liber instrumentorum seu chronicorum monasterii
Casauriense, ed. A. Pratesi (L’Aquila, 1982). On the structure of the codex, see C.
Manaresi, ‘Liber instrumentorum seu chronicorum monasterii Casauriensis della
Nazionale di Parigi’, Rendiconti dell’Istituto lombardo di scienze e lettere, Classe di let-
tere e scienze morali e storiche, 80 (3rd. ser., 11) (1947), 29–62.
49 For a brief history of the ebb and flow of Casauria’s landed wealth, see Feller, Gramain
and Weber, La fortune de Karol, 19–24.
50 On the distinction between cartularies and libri traditionum, see Geary, Phantoms, 81–2.
51 On additions to and exclusions from the Karol dossier, see Feller, Gramain and Weber,
La fortune de Karol, 149.
They document Karol’s burgeoning landed estate, built up through pur-

chases between about 850 and 870, made easier in at least one instance
by a loan, secured against his land and lent at what seems to have been an
exorbitant rate of interest. It may in part have been this acquisitiveness
that forced his sons to sell to the monastery some, at least, of their estate,
which they then leased back for an annual rent.52 What is in doubt is
the extent to which Casauria’s Karol dossier represents the totality of the
family’s documents – and therefore documents the totality of its landed
property – or constitutes simply the residue of an ordinary level of trans-
action between significant proprietors in the area. Put another way, did
the presence of the monastery have a distorting effect either on the land
market or on the documentary culture of the region?
The foundation of the monastery almost certainly affected the nature
of transactions – we can only speculate as to what may have become
of Karol’s sons if they did not have the option of ‘sale and re-lease’
provided by Casauria – but had far less effect on the overall level of
property transfers. ‘Land market’ seems an inadequate description of
these, since it carries the baggage of modern economic definitions. While
those linked by personal, familial and social ties in the ‘Vico Teatino’
could transact property between themselves through deeds of sale and
purchase without significantly affecting their personal relationships, those
relationships were often influential in determining the agreed price.53 The
‘Vico Teatino’ dossier includes ‘market exchanges’ determined solely by
the size and condition of the goods and a monetary evaluation of the
compensating payment, but it also contains transaction records that do
not state the quantity of the goods (the surface area of land) and/or its
quality (the nature of its exploitation), and those that detail goods in kind,
rather than money, as payment for land. Even a monetary payment could
be subject to social pressures, but where it was absent, or where goods are
not presented according to quantifiable criteria, there is plenty of scope
for unwritten, informal, social elements to affect the transaction.54
Since our estimation of socio-economic relations depends as often on
omissions as on inclusions of certain elements (like specified sums of
money) in the transcribed records, the process of archiving becomes
crucial. The regularity of the pattern of documentation in the ‘Vico
Teatino’ dossier – those giving a surface area for transferred lands are
52 For differing interpretations of Karol’s family’s position, compare Feller, Gramain and
Weber, La fortune de Karol, 93–115, with P. Cammarosano, ‘Marché de la terre et
mobilitié sociale dans les Abruzzes aux IXe –XIe siècles: À propos d’un livre récent’,
Revue historique, 646 (2008), 369–82.
53 Feller, Gramain and Weber, La fortune de Karol, 6.
54 Feller, Gramain and Weber, La fortune de Karol, 90–1.
a group distinguishable also by other criteria – already suggests that

both Giovanni di Berardo and previous generations of Casauria archivists
retained the key substance of the dispositive clauses of the charters they
were copying or abbreviating, even where they failed to copy the date
clauses. The reliability of their work can be confirmed by comparison
with originals. Those from Monte Amiata and Lucca have already been
mentioned, but especially pertinent here is a small corpus of original
charters in the cathedral archive of Piacenza documenting the activities
of Peter ‘de Niviano’: twenty-six acts in which Peter and his wife built
up their patrimony mostly through purchase.55 They also demonstrate
that Peter was able to liquidate his landed assets, both in order to make
cash purchases and to offer loans to those in need. The latter look very
similar to the ‘help’ delivered to the sons of Karol by Abbot Romanus of
Casauria.
Lucca, Monte Amiata, Casauria and Piacenza all show that an ecclesi-
astical archive could become the repository of batches of lay documents.
Often we have reason to assume that this was because the properties
concerned ended up in these institutions’ possession. A Veronese charter
of 841 gives especially vivid testimony of this, because, although it is
a lay document, by which a lay landowner sold land to a sculdahis, the
dorse of the charter records the transfer of the properties described in the
document, twenty or more years later and by an unknown donor, to the
monastery of S. Maria in Organo.56 Since the documents that we read
in standard editions or in later cartularies are very often actually tran-
scriptions of the recto side, the question arises how many more might
in fact include such stipulations. The transfer of documentation to a
monastery because of a transfer of land is also the obvious explanation in
the case of Casauria, founded after Karol’s major transactions. However,
the examples of the estate dossier of ‘Agello’ at Monte Amiata and the
personal dossier of Peter ‘de Niviano’ at Piacenza suggest an alternative:
that these institutions functioned as convenient places of safe keeping for
documents that, during their practical lifetime, had nothing to do with
them.57
55 F. Bougard, ‘Pierre de Niviano, sculdassius, dit le Spolétin, et le gouvernement du comté

de Plaisance à l’époque carolingienne’, Journal des savants (1996), 291–337, including
editions of all documents in the dossier.
56 CDV 166; it names Abbot Romuald, abbot from August 860 (CDV 219) until 865×81
(CDV 232, 280).
57 For a list of documents from north of the Alps that included properties no longer in the
abbey’s hands, see the discussion of the earliest Passau cartulary by Geary, Phantoms,
90–3.
An alternative to the archive: public

validation of documents
The argument has been made that the increasing predominance of lay
charter scribes indicates not only their professionalization but also the
implication of the state in the process of charter redaction. According to
the standard narrative, this resulted from the quest to give greater legal
validity to documents. It envisages a fairly linear development in which
a growing concern for the secure validation of activities and/or transac-
tions led to the incorporation of a number of formal, public elements in
the process of redaction, and ultimately to the formalization and public
validation of the occupation of scribe (or, more narrowly, of notary –
notarius) itself. This linear progression has been taken to imply that at its
starting point in the Lombard period recourse to writing was relatively
ineffective. Thus Paolo Delogu: ‘in Lombard Italy . . . private acts were
not publicly valid, even when written in the presence of witnesses’.58 It
is not clear what kind of public–private distinction is meant here. Everett
has shown that the means of bestowing legal validity on charters in the
Lombard period were identical to – and certainly inherited from – late
and vulgar Roman practice.59 Means of validation such as authentication
by the correct number of witnesses reflect the emphasis on public perfor-
mance – on the charter as capturing the moment of redaction – that was
transmitted to, and even enhanced by, charter scribes of the Lombard
period. In any case, if documents had no legal validity, in particular in
court, then there would have been little point in drawing them up in
the first place. In fact, however, there is substantial evidence that even
documents commissioned by ‘private’ individuals had sufficient validity
to be invoked as means of proof in formal, ‘public’, judicial hearings.60
The breve lacked even those elements of formal enactment that the carta
had61 – including witness subscriptions – yet the term encompasses doc-
uments that were regularly advanced as proofs in court, not least the
abbreviated records of court hearings themselves.62 As elsewhere in the
post-Roman West, Lombard Italy allowed legal validity to all sorts of
documents.
58 P. Delogu, An Introduction to Medieval History, trans. M. Moran (London, 2002 [1994]),

174–5 (quotation at 175).
59 Everett, Literacy in Lombard Italy, 202–3 with references.
60 For examples, see Wickham, ‘Land Disputes’, and Costambeys, ‘Disputes and Courts
in Lombard and Carolingian Central Italy’, EME, 15 (2007), 265–89.
61 The terms carta, charta, cartula and chartula used for single-sheet legal documents are
interchangeable in this context.
62 E.g. Manaresi I 8, 16, 27, etc.
But to say this is not to deny the traditional narrative’s central point that
the Carolingian period witnessed a shift towards the authentication and
validation of documents by the state. It is really this, surely, that Delogu
had in mind when he wrote the words quoted above, and there is plenty
of evidence to support such a view.63 It falls into two categories: royal
pronouncements about the appropriate activity of scribes, which taken
together have been seen to have brought about their professionalization
at the behest of the state; and the confirmation or revalidation of docu-
ments through public means. To take the former first, ninth-century royal
acts focused on notaries in several noteworthy ways.64 Capitularies about
charter scribes involve the combination of exhortation and ‘best practice’
that has recently been identified as typical of Carolingian capitulary acts
in general: it is in this sense that we should read the requirement that
notarii be ‘registered’ by the royal missi, for instance.65 The Italian capit-
ularies, on the other hand, include stipulations that have been thought
to reveal a more specific concern with procedure: notaries’ redaction of
cartae publicae was to be supervised by the count and his scabini and
vicarii, they were to be required to forswear forgery or colludium, and
maximum and minimum sums were set for their payment.66 Yet, only
the last, I suggest, is sufficiently precise to speak of any kind of system
or verification, and it in fact looks to be quite anomalous.67 The other
clauses can be seen as establishing the general environment of notarial
practice, rather than regulating specific activities. These therefore need
to be set apart from the evidence for the laicization of charter scribes and
cannot be seen as themselves aimed at or creating a notarial profession. It
used to be argued that it was in the ninth century that notaries acquired
fides publica: an authority bestowed by the state, through the law, to give
validity to documents. But while it is true that by the late ninth century
63 In what follows, I am drawing principally on Bougard, La justice, 66–76; see also the
summary in Delogu, An Introduction to Medieval History, 175–6.
64 MGH Capit. 1, no. 40, c. 3 (a. 803) ruled that missi should register the names of notarii;
no. 43, c. 4 (a. 805) that each bishop, abbot and count should have ‘his’ notarius, a
regulation that follows on from no. 90 (a. 781), specifically for Italy, which envisaged
that the count would have a personal notarius. See Bougard, La justice, 66–7.
65 C. Pössel, ‘Authors and Recipients of Carolingian Capitularies, 779–829’, in R. Corra-
dini et al. (eds.), Texts and Identities in the Early Middle Ages (Vienna, 2006), 253–74; and
M. Costambeys, M. Innes and S. Maclean, The Carolingian World (Cambridge, 2009),
ch. 4. Interpretation of the nomination of notarii differs here from Bougard, La justice,
68.
66 MGH Capit. 1, no. 158, cc. 12, 15 (a. 822×3) (‘ut cancellarii [= notarii] . . . veraces
cartas publicas conscribant ante comitem et scabinis et vicariis eius’); MGH Capit. 2,
no. 201, c. 13 (a. 832); no. 202, c. 5 (a. 832). For cancellarius as a synonym for notarius,
see Bougard, La justice, 68–9.
67 Bougard, La justice, 68 with n9.
some notaries were also judges, and therefore validated documents pre-
sented to them in court that they themselves, or their fellow notaries,
had written, the notion that in doing so they possessed some recognized
authority qua notaries is very anachronistic. When it came to validation,
we can go along with François Bougard’s observation that in the Carolin-
gian period ‘nothing has really changed in relation to the Lombard era:
what gives the act force and validity is above all its redaction “in pub-
lic”. The adjective publicus, when it qualifies the scribe or the charter,
most often has no other meaning than “not secret”.’68 True fides publica,
and therefore the redundancy of the formal witness subscription, had to
wait until the twelfth century.69 The ninth-century notarius belonged to
a profession only in an informal sense, his job description being set by
practice and custom.
The use of ‘public’ instruments or mechanisms to confirm or vali-
date documents points rather more directly to state involvement. The
most obvious were the so-called pancarte confirmations: diplomas peri-
odically issued by kings and emperors confirming, often, whole lists of
‘private’ documents.70 Their preservation seems decisively to reinforce
the notion that only Church institutions had the capacity to preserve title
through generations. The fact that the surviving confirmation diplomas
all concern the property acquisitions of monasteries raises an important
question: was it only monasteries, or at any rate ecclesiastical institutions,
that had the archival capacity that such documents required? We know
that ecclesiastical institutions were not alone in receiving, and keeping,
documents confirming the provisions of other documents: several are
mentioned among Alahis’ documents in the Ghittia list.71 But the fact
that these do not survive makes it hard to judge their shelf life, whereas
ecclesiastical institutions’ better archival capacity opened up different
possibilities to them. Farfa, for instance, kept the confirmation diplomas
of successive emperors and kings of Italy, even though each successive
ruler’s diploma only reiterated, and therefore in theory made redundant,
that of his predecessor. What was the point of keeping a confirmation
charter of Charlemagne if you had one, referring to the same properties,
68 Bougard, La justice, 69: ‘rien n’a vraiment changé par rapport à l’époque lombarde: ce
qui donne à l’acte force et validité, c’est avant tout sa rédaction “en public”. L’adjectif
publicus, lorsqu’il qualifie le scribe ou la charte, n’a le plus souvent pas d’autre signifi-
cation que “non occulte”.’
69 Bougard, La justice, 281–95; see also, e.g., C. La Rocca, Pacifico di Verona: Il passato
carolingio nella costruzione della memoria urbana (Rome, 1995), 9.
70 Pancartes that included provisions for immunity must be read in the light of B. H. Rosen-
wein, Negotiating Space: Power, Restraint and Privileges of Immunity in Early Medieval
Europe (Ithaca, NY, 1999), 99–134; in general, see Geary, Phantoms, 88.
71 CDL II 295.
from Otto III?72 Was it simply that rights improved through sheer weight
of documentation? The answer must in part be that Otto was presented
with this pile of diplomas precisely in order that their provisions could be
renewed: they were kept so that they could be shown to him. This speaks
of a consciousness that the efficacy of such documents diminished with
time, and that, in particular, their provisions needed confirmation by the
reigning king or emperor. It reminds us that institutions as well as individ-
uals had personal bonds with their lords, protectors, patrons, superiors
and/or rulers that had to be renewed with each new incumbent.73 In this
respect, it must have impressed Otto that Farfa could demonstrate an
association with every single one of his predecessors. Publicly display-
ing this part of its archive produced diplomas which in their turn could
be, and were, displayed before judicial tribunals as proofs of title.74 An
archival capacity of this magnitude was something lay families struggled
to match, and it allowed such institutions to strengthen their identity –
and their socio-political clout – by selectively constructing their past.75
Farfa’s sheer longevity was one of the features that convinced rulers like
Otto III that it was worth favouring (though we should not lose sight of
others: the location, as well as the extent, of its properties, claimed or
actual; its liturgical activity; its saintly associations).76
Lay families did, however, have access to the other form of public
validation that ecclesiastical institutions used: the law courts. The ninth
century saw the development of documentary forms apparently delib-
erately designed to validate existing charters through judicial proceed-
ings. One such form was the ostensio cartae, first appearing in 880×1,
in which a plaintiff’s charter was copied out, other parties were asked
if they wished to contradict it, and they declined to do so.77 It was not
the only such innovation. In the 870s, we get the first examples of a pro-
cedure known as finis intentionis (finis intentionis terrae or finis intentionis
72 RF III 413, issued in 996, not coincidentally the year of Otto’s imperial coronation.
73 It would be inappropriate here to pursue further the crucial issue of vertical ties of
dependence: see, e.g., A. Rio, ‘Freedom and Unfreedom in Early Medieval Francia:
The Evidence of the Legal Formulae’, Past and Present, 193 (2006), 7–40. See also
above, Chapter 5, and below, Chapter 10.
74 E.g. Manaresi I 38, 50 (= RF II 270, 286).
75 Geary, Phantoms, 81–114; W. Pohl, Werkstätte der Erinnerung: Montecassino und die lan-
gobardische Vergangenheit (Vienna, 2001).
76 Properties: Costambeys, Power; M. Stroll, Medieval Abbey of Farfa: Target of Papal and
Imperial Ambitions (Leiden, 1997). Liturgy: S. Boynton, Shaping a Monastic Identity:
Liturgy and History at the Imperial Abbey of Farfa 1000–1125 (Ithaca, NY, 2006). Saints:
E. Susi, ‘Agiografia e territorio’, in I Longobardi dei ducati di Spoleto e Benevento: Atti
del XVI congresso internazionale di studi sull’alto medioevo: Spoleto, 20–23 ottobre 2002,
Benevento, 24–27 ottobre 2002 (Spoleto, 2003), 317–56.
77 Manaresi I 91; further examples until 900 are nos. 99, 105, 106, 107.
status).78 Here, the plaintiff made a statement of his rights in a thing,

declared himself ready to confront an opponent, and requested a formal
recognition of his rights from a ‘defendant’ (often, this latter party would
simply state that he had, or could make, no claim to rights in the thing).
The main difference from the ostensio cartae was that the text of a charter
was not included in extenso in the notice.79 Precisely this method was
used in 873 by Casauria to validate its ownership of the lands of Karol,
son of Liutprand.80 The third procedure, appearing first in 874 (but
not again then until 896), was the investitura salva querela, in which the
plaintiff stated his rights, no defendant appeared, and the court formally
invested the plaintiff with possession, while leaving open the possibility
that a defendant could press a claim to the property in the future, within
a certain period.81
These forms give the impression of having been deliberately designed
to provide plaintiffs with a new document to preserve or strengthen their
title. The fact that not one of the earlier documents thus confirmed
survives suggests quite strongly that these notitiae were in fact meant to
replace those earlier documents. This in turn suggests that documents had
shelf lives, and raises several questions. Were these time limits a function
of the legal transactions that the documents contained (which may, for
example, have been subject to the thirty-year rule),82 of the materials
from which they were made,83 or of the limited archival capacity of the
plaintiffs? Where did the initiative for these new forms lie: with those
who employed charter scribes, or with the scribes themselves? It used to
be thought that all three had their origins in the formula-book known
as the Cartularium Langobardicum, in which they all appear. But, as we
have seen, this text has now been firmly redated from the late ninth to
the late eleventh century, so it must be seen as recording practice, not
78 Manaresi I 74–5, 79–80, 82–6 (all from Casauria). Compare the various Frankish
processes for testing and validating a document discussed above, Chapter 5, and below,
Chapter 11.
79 Bougard, La justice, 312–14, argues that this procedure was developed as a necessary
response to the practice in which properties were initially occupied illegally and then
successively alienated, or their usufruct granted and regranted, until the original owner’s
rights were lost sight of. The finis intentionis offered the opportunity to ‘wipe the pro-
prietorial slate clean’, as it were: either the original owner could revindicate his rights,
or he would lose them, and a new owner would be invested.
80 Manaresi I 75.
81 First in Manaresi I 77, then in 101–2. Bougard, La justice, 314–19, argues that this was
simply the logical progression from a series of previous norms and procedures aimed at
combating contumacy.
82 Note, though, that the rule made redundant more recent documents, not older ones.
83 Though note that by the ninth century almost everyone outside Rome seems to have
been writing their charters on parchment.
setting it. The question of origins is a debate between ad hoc development

and centralized direction. Chris Wickham has pointed to an increasing
tendency in the Italian notitiae through the earlier decades of the ninth
century for written documents to be taken uncritically as proof, and for
their production in court to end the hearing. This would seem to be a
first step on the road to the non-confrontational format of the ostensio
cartae. On the other hand, the remarkably rapid spread of these three
forms, once arrived at, seems to point to a controlling hand guiding their
use. By the last decade of the tenth century, the vast majority of dispute
notitiae were written in one of these three forms,84 an increase so rapid as
to suggest a very effective bureaucracy, centred on the legists attached to
the royal court at Pavia.85 The pressure to which they were responding
is generally thought to have been generated by courts themselves, and
in particular by judges’ growing tendency to prefer charters as means
of proof. This, it is argued, led them to develop a new set of practices,
focused exclusively on the charter, that took over relatively quickly.
This goes some way to recognizing that the judicial system appears
to us, through the notitiae, as concerned not with disputes at all, but
with documents. Documents, that is, were not simply tools with which
to argue, but very often appear to be reified as the thing in dispute: the
dispute was either about the document or was intended to produce a
document. Disputes could also lead to the destruction of documents:
the (unlawful) destruction of documents is recorded in notitiae from the
eighth and ninth centuries,86 and their destruction as a result of court
judgements in those from the eleventh.87 The ostensio cartae form in par-
ticular seems to indicate that the court hearing had become entirely a
matter of documents: the brandishing of one, and the writing of another
(the notitia). Time and space were both important here. What has been
said above suggests that even properly witnessed charters were not con-
sidered secure guarantors of title: their efficacy was limited by time (not
84 Wickham, ‘Land Disputes’, 237: ‘The [north Italian] legists worked out formularies for
the recording of placita in the late ninth century that caught on so fast that they covered
nearly every formal placitum by 920, except in non-Carolingian southern Italy. (When
the formulaic placitum became dominant, in fact, the usefulness of the formal court case
for social historians drops dramatically.)’ See also pp. 245–6.
85 While it is true that the finis intentionis, while common in Spoleto, really only got going
in the north from the 960s, it still constitutes about a third of surviving cases from the
last decade of the tenth century; the investitura salva querela makes up about a sixth of
surviving notitiae from the 890s to the 1020s; and the ostensio cartae accounts for another
third of survivals in the first half of the tenth century, rising to a half after about 950.
86 See below, Chapter 12, and A. Sennis, ‘“Omnia tollit aetas et cuncta tollit oblivio”:
Ricordi smarriti e memorie costruite nei monasteri altomedievali’, Bullettino dell’Istituto
storico italiano per il medio evo, 106 (2004), 93–135.
87 Manaresi II 294, 314bis; RF IV 658.
least by the lifetime of the witnesses) and by the narrow range of those
they involved. As the civic institutions that had performed the function
of validating titles (that is, gesta municipalia) atrophied between the sev-
enth and ninth centuries, new mechanisms came to be needed.88 It has
often been noted that the decline of the gesta left a void, but sugges-
tions for how it was filled have generally invoked vaguely individual or
family enterprise.89 The notitia obtained from the court might therefore
either replace a document that was regarded as past its ‘use by’ date or
strengthen one by broadening knowledge of it. Archiving was therefore
not simply a question of once-and-for-all preservation. Archives had to
be managed, and had to be accessible. Their management must there-
fore often have involved an element of turnover: documents had to be
preserved in order to be brandished, but once the new court notice had
been produced, the old document could be discarded (or the parchment
reused).90
The development of the notitia through the ninth century therefore
suggests that the judicial system was not simply a forum for settling
disputes; it was also a mechanism for validating old documents and
producing new ones. This may have been done in anticipation of dis-
putes, rather than in response to them. But it may also have become
necessary because other forms of validation and renewal were no longer
available. Wickham, as we have seen, sees the appearance of the ostensio
cartae and similar procedures as an essentially organic development then
set in stone by some on-the-ball Pavese legists. For François Bougard,
they were intended to confirm transactions already entered into (and
in all cases already enacted by ‘private’ charters); and he notices that
the majority seem to be especially important transactions, or those of a

89 E.g. J. M. H. Smith, ‘Aedificatio sancti loci: The Making of a Ninth-Century Holy
Place’, in M. de Jong and F. Theuws (eds.), Topographies of Power in the Early Middle
Ages (Leiden, 2001), 361–96, at 371: ‘although the late Roman procedure of registering
title in municipal archives, the gesta municipalia, lingered on here and there into the early
seventh century, the onus for redacting and preserving texts gradually passed everywhere
to individual or institutional proprietors’, citing P. Johanek, ‘Zur rechtlichen Funktion
von Traditionsnotiz, Traditionsbuch und früher Sigelurkunde’, in Recht und Schrift,
131–62, esp. 136–45.
90 What we can see of their preservation indicates that it was their immediate beneficiaries
who benefited from, and so preserved, notitiae employing ostensio cartae and similar
forms: of the twenty-four surviving examples of these forms from the first thirty years of
their existence (873–903), nineteen are preserved by the institution (always a church or
monastery) that benefited from the document or by that institution’s ultimate repository:
Manaresi I 74–5, 77, 79–80, 82–6, 88, 91–2, 95, 99, 101–2, 105–7, 109, 111–13; of
which nos. 77, 80, 91, 107 and 113 benefit parties other than those who preserved the
document. By ‘ultimate repository’, I mean that, for instance, documents benefiting the
monastery of S. Ambrogio in Milan are now in the Archivio di Stato there.
certain complexity. Some of these may have had real disputes behind
them, and some not. What is equally important for Bougard is the
simple fact that in switching to this more formulaic style, the notaries
were setting the documentary record, the notitia, at one remove from
the actual court case (if there was one).91 For him, the ostensio cartae
was the notaries’ deliberate invention, prompted by a desire to make
unnecessary the ‘jeux d’écritures’ that both lay and ecclesiastical parties
had been engaging in previously: that is, the creative use of standard
‘private’ charter forms that were not individually sufficiently flexible to
cover, for instance, the restructuring of a loan.92 As he observes, ‘A dona-
tion could hide a sale, a sale a loan.’93 Recourse to the court apparatus
also added a sought-after element to temporally limited or contingent
transactions: donations that reserved usufruct for the donor’s lifetime,
or acts like manumissions that may have been initially provisional, could
be confirmed by application to the court. One study indicates that the
majority of notitiae employing new models like the ostensio cartae were
produced only a short time after the documents they confirmed,94 per-
haps because they enabled the current possessor of a property to validate
his right in a context in which traditional norms like the thirty-year rule
tended to favour the residual rights of earlier possessors. This may have
become especially important when time-limited leases were running out:
it is surely significant, then, that the first such one- and two-generation
leases appear in our records in the first half of the ninth century, and the
first of these new forms some sixty or so years later.95 These observa-
tions therefore confirm that the pressure for change may have come from
the ‘consumer’ (that is, those who commissioned charters), and that it
may have been a response to the growing inadequacy (rather, perhaps,
than a fading away) of older means of preserving, validating, renewing or
copying documents. The state looks to have been less of a driving force
behind this change than is often argued; its only importance, in fact,
91 Hence, if the term placitum refers first and foremost to the court hearing itself, it
becomes ever less helpful to use it to refer to the document that that hearing eventually
produced.
92 See the example of A. R. Natale, Il museo diplomatico dell’Archivio di Stato di Milano, vol.
I (Milan, 1968), nos. 82–3, in which S. Ambrogio in Milan had to perform all sorts of
tergiversations of standard diplomatic simply in order to renew the loan of one of its
debtors.
93 Bougard, La justice, 328.
94 S. H. Brunsch, ‘The Authority of Documents in Early Medieval Italian Pleas’, in
B. Bolton and C. Meek (eds.), Aspects of Power and Authority in the Middle Ages (Turn-
hout, 2007), 277–87, at 285–6.
95 See Bougard, ‘Actes privés et transferts patrimoniaux’, 551–5, on the chronological
patterning of property concessions.
was that it provided the connections between notaries that enabled new
documentary forms to be legitimized and disseminated.
Acknowledging the role of the courts in validating documents has a
number of implications. First, it suggests that we should not restrict
our definition of lay documents to what have traditionally been classed
as ‘private’ acts. The notitiae of court cases bridged the divide between
the traditional categories of ‘public’ and ‘private’, since they were very
often simply a means of giving extra legal weight to transactions between
‘private’ individuals, families or institutions by having them confirmed by
a recognized authority and, perhaps most importantly, by having them
exposed to the scrutiny of the actors’ peers: in other words, being no
longer secret, they were now ‘public’. Secondly, it reinforces the point
that archiving was not a once-and-for-all procedure. A document could
be periodically renewed, rendering redundant the previous document.
But, of course, periodic renewal would only take place for as long as it
was necessary and relevant to the parties concerned. Once the pattern
of family property changed through death, inheritance and marriage – in
other words generally after just a few generations – all trace of the title that
had previously been renewed would be lost. Thirdly, as the notitia became
more formalized, its function as a document of validation separated from
its role as record of dispute, judgement and/or agreement. There had
always been a necessity for breves recording negotiations not recorded in
the court notitia (which can hardly be regarded as extra-judicial); that
need only increased with the growing formality of the latter.96
Ecclesiastical institutions and the preservation

of documents
Was there, then, something special – something peculiarly ecclesiasti-
cal – about those ecclesiastical institutions that preserved documents?
One reason above all why documents sometimes survived – and occa-
sionally still survive – in the hands of ecclesiastical institutions is that
those institutions did not undergo the kinds of changes that led to the
dispersal and/or eventual disappearance of lay patrimonies – and the
documents associated with them. They were entities that did not have
to partition their patrimonies with every new generation, did not trans-
fer property as part of marriage settlements, and, crucially, did not die.
They have been subject to different acts of history. These may have led to
96 E.g. RF II 205, 250, both breves (rather than official notitiae) recording dispute settle-
ments; further, Bougard, La justice, 74–6.
the dispersal or destruction of their fund of documents and even extin-

guished the institutions themselves – how often have medievalists cursed
Cromwell’s commissioners, or the incendiary mobs of the French Rev-
olution? – but such historical accidents were nowhere near as frequent,
as routine – indeed, as inevitable – as those suffered by lay families. This
contrast with the natural fragility of lay families makes it easy to assume
that ecclesiastical institutions’ superior durability is a peculiarly ecclesi-
astical quality, and to take it for granted. It is this, after all, that makes
so many of them so much the obvious foci for local or regional studies
across the entire early-to-high medieval period.97 However, it needs to
be underlined that institutions that managed to conserve their archives
for posterity are ipso facto exceptional among the full complement of
ecclesiastical institutions: most ecclesiastical archives did, in fact, have
limited shelf lives. There are five broad reasons why ecclesiastical insti-
tutions’ records might not remain in situ: natural accident (fire, decay);
the disappearance or dramatic upheaval of the society that supported the
institution (as with the bishoprics of northern Gaul and much of Britain
in the fifth century); the absorption of the institution by another, with
consequent transfer, and streamlining, of the record; the extinction of
the institution with consequent dispersal of its archive; and (related to
the latter) the extinction of the family that supported the institution. The
latter three are those most relevant to the documents that survive from
Italy, and the last reminds us that, among the total number of ecclesiasti-
cal institutions, those not dominated by a single family or interest group
were the exception.98 In particular, we can distinguish most (arguably not
all) bishoprics in this respect, since their existence depended on a rather
different basis (though they could, of course, be amalgamated and/or
extinguished). Where monasteries are concerned, the provenances of the
surviving documents reveal which of them had the most durable archives
and suggest that their longevity depended on the ability of the institution
to distinguish itself from the laity – that is, usually, from particular lay
interest groups – to a sufficient extent to survive changes in those groups’
fortunes.
97 E.g. K. Schmid, Die Klostergemeinschaft von Fulda im früheren Mittelalter, 3 vols. in 5

(Münster, 1978); for Lucca, C. Wickham, The Mountains and the City: The Tuscan
Apennines in the Early Middle Ages (Oxford, 1988), and now M. Stoffella, ‘Aristocracy
and Rural Churches in the Territory of Lucca Between Lombards and Carolingians:
A Case Study’, in S. Gasparri (ed.), 774: Ipotesi su una transizione (Turnhout, 2008),
289–311.
98 Wood, The Proprietary Church, esp. 221–30.
Broadly, therefore, the ecclesiastical institutions that were best at pre-

serving documents were those that were most successful in defining them-
selves clearly and distinguishing themselves from the non-ecclesiastical.
Indeed, the documents that they kept played a crucial part in this self-
definition, because this was not simply a question of preservation for its
own sake. These institutions had archival strategies involving selecting
as well as retaining documents – strategies, moreover, that themselves
shifted over time and according to circumstance. Perhaps the best exam-
ple of this in Italy is Verona. Cristina La Rocca has made a strong case
for the careful attention paid to shaping their past by Verona’s episco-
pal clergy in the late eleventh and twelfth centuries.99 The most strik-
ing aspect of this, she argues, is their selection of an obscure ninth-
century archdeacon, Pacificus, whom they transformed into a Carolin-
gian Renaissance man, the equal or superior of any such figure at the
court of Charlemagne, making him responsible for all sorts of preco-
cious initiatives that established a pioneering programme of document-
ing and record-keeping as the single most important distinguishing trait
of Verona’s clergy. La Rocca shows that this was achieved principally
by the fabrication of a series of key documents that all argued for the
independence of the cathedral chapter (Pacificus’ institution) from the
bishop. Elsewhere, cathedral chapters did not acquire such autonomy for
another two centuries. She argues that the mid ninth century emerged
as the point at which cathedral chapter and bishop could be portrayed as
having separated because Verona tradition remembered genuine friction
at that time between the cathedral clergy and the then bishop Ratold,
an Aleman introduced into the diocese by the Carolingians. If she is
right, then it is very likely that later generations of Veronese clergy not
only added to their institutional tradition by forging charters but also
shaped it through selecting what charters to keep. The question then
arises of whether we can see some trace of archival selection (as well as
augmentation through forgery) in the profile of the charters that have
been preserved.
Although Verona’s charter collection is not the easiest to study, because
of the rather indiscriminate approach of its modern editor, La Rocca’s
work allows us to identify which of the later copies of charters are reliable
and can therefore be added to the surviving originals. Of the resulting
total of fifty-seven charters from the period 751–875, eight are ‘lay’ docu-
ments. All of these belong to dossiers associated with particular churches
99 La Rocca, Pacifico di Verona, and La Rocca, ‘A Man for All Seasons: Pacificus of Verona
and the Creation of a Local Carolingian Past’, in Y. Hen and M. Innes (eds.), The Uses
of the Past in the Early Middle Ages (Cambridge, 2000), 250–77.
or estates, the most prominent among them being the monastery of S.

Maria in Organo. This was jurisdictionally independent of the bishop of
Verona throughout the period that concerns us here, falling under the
purview of the patriarch of Aquileia. Only in the twelfth century, after the
Concordat of Worms ensured that the bishopric was no longer part of the
imperial power network in northern Italy, did the episcopate of Tebaldus
(1135–57) see collaboration with the patriarch of Aquileia and a com-
ing together of bishop and monastery.100 It is to that period – precisely
the time when the legend of Pacificus was in the making – that we can
date an attempt to reshape and sift Verona’s cathedral archive to exclude
references to Bishop Ratold. Nearly all the documents preserved from
the period of his episcopate (c. 814–40) are those concerning either the
cathedral clergy or S. Maria in Organo. Only a handful of the bishop’s
own documents survive, and examination of their witness lists reveals the
complete absence of the cathedral clergy; in the same way, the bishop
is absent from the latter’s charters. Crucial here is the unusual survival
of a grant by Ratold, in which, in conjunction with the new Carolingian
count of Verona, Hucpald, the bishop gave the estate of the late count,
Ademar, to the church of S. Pietro in Castro.101 Not only did this latter
church belong to the episcopal and ‘Carolingian’ side of Verona’s political
division – by 871 it had come into the hands of the imperial monastery of
S. Zeno, which from the start was antagonistic towards the much earlier
foundation of S. Maria in Organo102 – but also, here as elsewhere, the
witness list shows Ratold having to turn to laymen, quite possibly Frank-
ish immigrants, to help authenticate his act.103 The cathedral clergy, on
the other hand, often feature in the ‘private’ transactions that took place
in Ratold’s episcopate, often involving S. Maria in Organo.104 Verona
therefore shows quite palpably that in the ninth century documents were
arranged and stored for reasons quite apart from any structural distinc-
tion between the ecclesiastical and the lay. In Verona, charter use followed
political affiliation.
While ecclesiastical institutions’ dominance of our documentary
record for Italy looks at first sight to discourage any notion that we
may be able to touch lay documentary culture, we have seen that this
masks a variety of practices that cast doubt on the usefulness of the
lay–clerical distinction in this context. Those who structured ecclesias-
tical archives may have preferred to preserve lay documents from early
100 Miller, The Formation of a Medieval Church, 163–74.

101 CDV 89. 102 CDV 241.
103 La Rocca, Pacifico di Verona, 240, gives the example of Manaresi I 31 (CDV 121).
104 E.g. CDV 124–5, 128, 131–2, 134–5, 144–6.
in – or before – the history of their institution because it reinforced

its longevity. But this also implies that they often selected for omission
or destruction later lay documents. Where these survive, as at Monte
Amiata and Piacenza, they provide a precious indication that sometimes
monasteries and episcopal centres offered the most practical repository
for leaves of parchment that were valuable, at least to their owners. That
those owners did not last down the generations may explain why many
may later have been discarded. Family longevity affected the shelf life
of a document. On the other hand, the need to preserve a document
was felt less where mechanisms existed for reaffirming the provisions of
a legal act, and often for thereby producing a new document. This was
clearly one of the main functions of the law courts in eighth- and ninth-
century Italy, and it pointed to a future in which the state – or at least
those claiming its authority – would play a stronger role in documentary
culture.
10 Sicut mos esse solet: documentary practices in
Christian Iberia, c. 700–1000
Adam J. Kosto
In Catalonia, an extraordinary number of documents survive: clear evidence

of the continued vitality of the documentary culture that we have seen in
the sources from the Visigothic kingdom. The evidence from the rest of Iberia
shows that Catalonia was not exceptional, although here documents survive in
ecclesiastical cartularies rather than originals. We can nonetheless see laypeople,
even servi at the bottom of the social hierarchy, using documents for petty
transactions, and not just land transfers. We can also spot signs of the earlier
existence of lay family archives, in some cases of several stages of preservation
by different lay families, and of dossiers of family documents being stored in
churches. Overall, the Iberian evidence attests how the ways of producing and
keeping documents, which had been as diverse as the patterns of landholding
and transfer they recorded, tended to become more institutionalized from the
tenth century, at the same time as the aristocracy consolidated their landed
estates.
In the mid tenth century, in the county of Portugal, Vermudo Pepiz

granted to his wife Guntroda a villa called Fano. The grant was recorded
in a document (scriptura benefactis), possibly along with other properties
in a charter recording Guntroda’s endowment (dos). Soon thereafter,
Guntroda exchanged the villa for another in Salinas with Didaco, son of
Menendo. This transaction, too, was recorded in a document, a ‘charter
of exchange’ (cartula commutacionis). Neither of these two documents
survives; both are mentioned in the prologue (narratio) of a grant made
in 959 by Didaco and his wife, Ildonca, of the villa Fano, to the recently
founded monastery of Guimarães. The couple granted Fano ‘just as I
exchanged it with lady Guntroda’. The presence of this document in
the cartulary of Guimarães, the Livro de Mumadona, so called after the
founder of the house, is no surprise, for one of the principal functions
of a cartulary was to collect evidence of donations of property. The
grant’s mentions of these two earlier, lost, documents – deperdita – are
259
260 Adam J. Kosto
more intriguing, for they offer evidence of transactions not involving an

ecclesiastical institution, transactions solely between laypeople. Further-
more, Didaco’s grant suggests the existence not just of two charters, but
of many more, for it notes that his earlier exchange with Guntroda was
recorded in duplicate cartulae commutacionis and that, furthermore, such
record-making was the reigning custom: ‘ex quibus utrorumque cartulas
sibi commutacionis uicissim fecerunt sicut mos esse solet’.1
I have made the case elsewhere for the routine use of written records
by laypeople in tenth-century Catalonia, as well as for the existence of
lay archives, all in the context of a common (lay and clerical) culture of
documentary practice.2 The present chapter extends that investigation
to the rest of the Iberian peninsula. Despite the fact that the eastern
Pyrenean counties shared a common Visigothic heritage with the Chris-
tian kingdoms to the west, the much stronger and long-lasting Frankish
influence on Catalonia led to cultural differences, notably (for present
purposes) in areas such as linguistic development, palaeography and
diplomatic practice. Original early medieval documents from Catalonia
feature a version of caroline minuscule, while scribes from further west
were using a Visigothic cursive. Similarly, scribes in Catalonia continued
to date documents by the regnal year of French kings until 1180, while
their Iberian counterparts elsewhere dated according to the Spanish era.3
A hypothesis of the documentary exceptionalism of Catalonia would be
reasonable. It proves, however, not to be valid. The evidence suggests
that the regular use of documents by laypeople in the rest of Christian
Iberia was very much the custom.
The dossier of documentary evidence for the Christian Iberian polities
outside Catalonia – Aragon, Pamplona/Navarre, Castile, León, Asturias,
Galicia and Portugal – is considerably more resistant to comprehensive
analysis than the material from the Catalan counties, where scholars can
1 PMH 77 (additional abbreviations given at end of chapter). There is a facsimile edition

of the Guimarães cartulary: Livro de Mumadona: Cartulário medievo existente no Arquivo
Nacional da Torre do Tumbo, vol. I (only), Reprodução facsimilada do códice, ed. I. A. Pinto
Ferreira (Lisbon, 1973). Other deperdita are indicated, for example, in Sobrado I 12
(‘Ipsum uillare iam supradictum uobis uendimus atque concedimus sicut illum habuit
mea auola iam supradicta per textus scripture incartatum de domna Unisco’) and PMH
88 (‘ego illa habui incartata cum uiro meo diue memorie Ranimirus Menendi secundum
illa habuimus incartata de Rex domno ordonio’). My thanks to Wendy Davies and Roger
Collins for comments on earlier versions of this chapter.
2 Kosto, ‘Laymen’.
3 A. J. Kosto, Making Agreements in Medieval Catalonia: Power, Order, and the Written Word,
1000–1200 (Cambridge, 2001), 5–7, and works cited there. See also the judgement of
R. Collins, Early Medieval Spain: Unity in Diversity, 400–1000, 2nd edn (New York,
1995), 236: ‘the Leonese and Asturian charters are less varied and more restricted in
their style and contents than those of the contemporary counties of Catalonia’.
Christian Iberia, c. 700–1000 261
benefit from a single modern editorial project.4 The Iberian material,

surviving as original single-sheet documents as well as cartulary and
antiquarian copies, is scattered among over two dozen modern, over-
lapping (and often competing) editions, all with different standards and
practices with respect to dating, identification of likely forgeries, and,
most importantly, judgements about what to label a single record. Only
the most impressionistic arguments based on simple numerical analysis
of the corpus as a whole are possible. Furthermore, the geographical
range of the documents alone, with the concomitant differences in polit-
ical, economic and social environments which gave rise to them, argues
against an attempt at comprehensive study, although for the purposes of
a comparison with Catalonia and other European regions, aggregation of
the evidence is necessary and, it is hoped, instructive. For, taken together,
the evidence is substantial: over 2,000 records from 1000 and earlier, with
over 90 per cent from the tenth century. The single largest source is the
composite archive of the cathedral of León (599 records), but there are
other major collections, such as those of Celanova (241 records), San
Millán de la Cogolla (119 records), Sahagún (360 records), Sobrado
de los Monjes (119 records), the ‘Fondo Antiguo’ of La Coruña (100
records) and the remarkably under-utilized assemblage of Portuguese
sources published by Herculano in the nineteenth century, in addition
to many smaller dossiers.5 This total is still only one-third the size of the
complete corpus from Catalonia, although it is twice the size of the cor-
pus from, for example, Anglo-Saxon England – a similarly heterogeneous
area of similar size.6
The first part of this study discusses aspects of the evidence for the
use of documents by laypeople to record their transactions. The second
part examines the evidence for lay archival practices: the preservation
4 Only the earliest documents have been edited as a coherent group (although even this
has substantial gaps): DEPA. The project Regestalia, based at the Universidad de Alcalá
de Henares, remains incomplete, and its website, previously at www2.uah.es/historia1/
carlossaez/Regestos/default.htm, has been taken down.
5 On the material generally, see Davies, Acts of Giving, esp. 22–6; E. Pastor Dı́az de Garayo,
‘Los testimonios escritos del sector meridional de Castilla (siglos X–XI): Ensayo de crı́tica
documental’, Historia. Instituciones. Documentos, 24 (1997), 355–79; L. Sierra Macarrón,
‘Producción y conservación de la documentación altomedieval: Del Cantábrico al Duero
(siglos IX–XI)’, Signo: Revista de historia de la cultura escrita, 13 (2004), 99–120.
J. Escalona Monge, P. Azcárate Aguliar-Amat and M. Larrañaga Zulueta, ‘A la ide-
ologı́a polı́tica: Los diplomas fundacionales de San Pedro de Arlanza y la construcción
de una identidad para la Castilla medieval’, in Actas del VI Congreso internacional de his-
toria de la cultura escrita. I. Libros y documentos en la Alta Edad Media: Los libros de derecho.
Los archivos familiares (Madrid, 2002), 159–206, is a model of the penetrating diplomatic
scholarship that these sources demand.
262 Adam J. Kosto
and organization of documents outside and independent of ecclesiastical

institutions. The arguments presented here, based as they are on an initial
survey and entirely on printed sources, are necessarily preliminary; they
are meant to provide a foundation for future work and a comparison
to the material from Catalonia, but above all a basis for the inclusion
of Christian Iberia as a whole in this broader survey of early medieval
documentary practices across Europe and the Mediterranean.
Lay documents
In the substantial corpus of evidence from the region of Osona-Manresa
in Catalonia, the proportion of documents with no evident link to an
ecclesiastical institution, cleric or monk was 39% (45% if counts and
viscounts were included). The figure for the rest of Iberia is significantly
lower, but still notable, around 20%, or about 400 of the roughly 2,000
documents.7
As other chapters in this volume suggest, the gap between Catalonia
and the rest of the peninsula likely has less to do with a difference in early
medieval documentary practices than with later decisions about preserva-
tion. While original single-sheet records constitute over half of the Osona-
Manresa corpus, cartulary copies dominate the holdings elsewhere. The
significance of the means of preservation can be demonstrated by an
analysis of the one coherent edition that cuts across archival sources.
The classic collection of Antonio C. Floriano, Diplomática española del
perı́odo astur (718–910), contains 225 documents (including fragments
and deperdita), the bulk of the early medieval Iberian records from outside
the Frankish March for those years.8 Of the 225 documents, 20 per cent
(44) are transactions between laypeople (excluding counts and kings) –
a figure that matches the overall estimate above.9 The proportion of
original10 documents that are transactions between laypeople, however,
7 The incomplete database of documents being used currently includes 2,068 records,
of which between 391 and 417 are lay documents according to the criteria used in the
Osona-Manresa sample (Kosto, ‘Laymen’, 54–5).
8 DEPA. Note that the main database includes only 163 records for this period. I include
in the total the appendices of fragments and deperdita, as well as documents reputed false.
I exclude the document at II, 398, which dates from after 910, as Floriano suspected;
see Sahagún 152. For the Visigothic period, see Á. Canellas López, Diplomática hispano-
visigoda (Zaragoza, 1979), and above, Chapter 4.
9 DEPA 17, 26, 34, 43–4, 47, 66–8, 70, 72, 76, 81, 83, 94, 96, 111, 114, 117, 119, 121,
129–30, 139–40, 154–5, 163, 168–9, 172, 177, 183, 188, 190–1, 193–5, 199, 201, 203;
also vol. I, appendix, 5, and vol. II, appendix, 9 and 12.
10 I use the term as shorthand for ‘document on a single sheet of parchment patently
written around the time it claims to be written’. Some of these may be contemporary
copies or forgeries, although it is worth exploring whether forgery was more common
in documents related to ecclesiastical institutions than in documents generally.
is much higher: of the 225 documents, 39 are originals, and of these,

44 per cent (17) concern transactions between laypeople – a figure in
line with the overall estimate for Osona-Manresa.11 These figures under-
line the importance of transmission for our understanding of documen-
tary practices in a given region,12 and this type of analysis still needs
to be applied to the Iberian corpus as a whole, and indeed beyond
Iberia.13
If the surviving record understates the importance of the produc-
tion of documents recording transactions between laypeople in numer-
ical terms, early medieval formula-books, such as those surveyed by
Brown, and sources such as the Ghittia list, discussed by Costambeys,
lead us to suspect that the evidence is also skewed in terms of sub-
stance – that there were many more types of documents produced than
have survived and that many of those were documents particularly rel-
evant to transactions between laypeople.14 The little-studied Formulae
Visigothicae support such an argument.15 The forty-five surviving for-
mulas from this collection, which probably dates from the early sev-
enth century, address the following topics: manumission (nos. 1–6 [of
slaves] and 7 [of a son]), donations (7–10, 29–31), sales (11–13), nup-
tial endowments (14–20), testaments (21–6), exchanges (27–8), com-
mendation (32), pact of division of property (33), judicial proceedings
(36, 39–43), precariae (36–7), loans (38, 44) and monastic pacts (45–
6). Of these, only fourteen explicitly concern churchmen or Church
11 I follow Floriano’s determination of the originality of various documents, although in

most cases it is not a problem, as he is dealing with cartulary copies. The original
documents (with lay documents in italics) are as follows: DEPA 9, 18, 24, 66, 69, 72,
73–5, 76, 81, 94, 96, 97, 103, 108, 112, 121, 138, 141, 149, 154–5, 159, 169, 172,
173–4, 176, 178, 183, 191, 193, 195, 203; vol. II, appendix, 4–5, 8, 9. Floriano does
not seem to have confirmed the existence of an original for DEPA 97 and 103.
12 The problem of transmission is, of course, complex. The surviving original documents
may offer a more accurate reflection of the original documentary production than suviv-
ing documents of all types (originals and later copies, in and outside cartularies). But
if, as Patrick Geary has suggested, documents relating directly to the interests of eccle-
siastical institutions are less likely to have survived as originals once cartularies were
created – if they were systematically destroyed, or, for institutional reasons, more likely
to be lost – then the presence of lay documents among original charters is an exaggera-
tion of the original documentary production. See Geary, Phantoms, 82.
13 See esp. Kosto, ‘Laymen’, 73.
14 Above, Chapters 6 and 9; see also Brown, ‘Documents’, 339.
15 Form. Vis. Dedicated studies are rare: J. G. Bidermann, Commentatio ad formulas Visig-
othicas novissime repertas (Berlin, 1856); B. Martı́n Mı́nguez, Las fórmulas tenidas por
visigodas (Madrid, 1920); K. von Schwerin, ‘Sobre las relaciones entre las Fórmulas
Visigóticas y las Andecavenses’, Anuario de historia del derecho español, 9 (1932), 177–
89; R. Mentxaka, ‘En torno a Formulae Visigothicae 44’, in S. Crespo Ortiz de Zárate
and Á. Alonso Ávila (eds.), Scripta antiqva in honorem Ángel Montenego Duque et José
Marı́a Blázquez Martı́nez (Valladolid, 2002), 827–40. See also Pastor Dı́az de Garayo,
‘Los testimonios’, and above, Chapter 4.
264 Adam J. Kosto
business (6–10, 25,16 31[?],17 , 45–6). The surviving Iberian documents

themselves are dominated by donations of various types and sales; in
her corpus of 1,960 tenth-century transactions, for example, Wendy
Davies identified 980 gifts (50%) and 850 sales (43%), although, as she
demonstrates well, the difference between the two is not always clear.18
Exchanges and various records of judgement, types also found in the
formula list, account for many of the remaining documents, although
only a few of the types are completely absent from the corpus of early
medieval Iberian charters in one form or another: the manumissions (1–
7), a gesta municipalia protocol (25) and a series of judicial mandates
(41–3).
Surviving sales, gifts, exchanges and even judgements record the dis-
position of land, which provides a clear incentive for their long-term
preservation; the same is true of some of the less commonly attested
documents, such as nuptial endowments.19 Documents have survived,
however, that are not even mentioned in the formula-book, and that offer
tantalizing hints of a much broader world of document production. Con-
sider profiliation, the process of transfer by adoption of an heir, usually
(but not always) from outside the family. About half of the surviving
cartae profiliacionis and other records referring to the process concern
transfers between laypeople. In 875, for example, the couple Licerio and
Fakilo made Adeit their heir (in loco filio) with respect to a villa; and in
905 the couple Severo and Recesinda made a second couple, Gundefredo
and Basilia, their heirs (in loco filios) with respect to three villas.20 In some
cases, profiliation represented the submission of the humble to the more
powerful, as when in 870, the servus Flaciano made his lord Nuño a co-
heir of two villas.21 This interpretation has borne much historiographical
weight. But Davies has shown that profiliation was not just a means of
domination, but in fact a very flexible instrument for the manipulation
of social ties.22 The next step is to examine profiliation in the context of
intrafamilial transfers of land, which are not as rare for the early period
and at apparently low social levels as has been assumed; again, this is not
a problem limited to transactions between laypeople, but it is a problem
dominated by them.23
16 Form. Vis. 25, a model transcript of proceedings of a gesta municipalia, includes the
passage ‘per quam ecclesiabus sanctarum Dei aliqua concessit’.
17 The address of this document hints at an ecclesiastical beneficiary.
18 Davies, Acts of Giving, esp. 25; W. Davies, ‘Sale, Price and Valuation in Galicia and
Castile-León in the Tenth Century’, EME, 11 (2002), 149–74.
19 E.g. Sobrado I 119; Otero 50. 20 DEPA 114, 183.
21 León 4; below, at n. 38. 22 Davies, Acts of Giving, 83–4, 160–1.
23 Examples: Liébana 38, 69; Corias pp. 71, 106; PMH 14, 16, 61, 97; Cardeña 1; Celanova
11, 12, 27, 40, 158; La Coruña 9; León 22, 53, 97, 139, 204, 316, 396, 488, 525; Sahagún
Documents have also survived, however, that concern transactions not

involving land. These sorts of transactions would again have been equally
open to laymen, clerics, monks and ecclesiastical institutions alike, but
it is the possibility of their use by laypeople that must be considered in
reimagining the scope of early medieval document production.
Two texts reveal a documented trade in servi, a subject already seen in
the chapters by Conant and Costambeys, above.24 In 914, Silo granted
to Muño and his wife, in a document that reads like a typical carta dona-
tionis, a mancipella by the name of Facquina, the daughter of Moderic.25
Similarly, in 995, Rodrigo Guimirez and his wife Basselissa handed over,
by means of a scriptura donationis, a 3-year-old mancipia by the name of
Adosinda, the daughter of Ero and Munia, whom Rodrigo and Basselissa
had inherited; they also granted, to another party, but by the same docu-
ment, Adosinda’s 15-day-old sister, Gossenda. For Adosinda (but not for
Gossenda), they received a countergift of real property, which provides
an explanation for the preservation of the document. But in the drafting
of the document, the reference to real property is limited to the part of
the text where one would expect to find the price or countergift. The
penalty clauses, for example, refer simply to the mancipiae: ‘Ita ut de odi
[sc. hodie] tempore sic siant ipsas mancipias de iuri nostro abrasa et in
uestro iuri uel dominio sint traditas adque confirmatas, et si aliquis omo
proinde vos calomniare uoluerit de progenie nostre que nos in iudicio
deuindicare non potuerimus aut autorigare noluerimus aut uos in uoce
nostra tunc pariemus illas uobis duplatas uel quantum ad uos fuerint
melioratas et uobis perpetim auiturum.’ The carta protects the interests
of the recipients of the mancipiae, not the the title to the lands.26 Both of
these are wholly lay transactions.
Even more surprising are two texts that record entry into servile status.
The first is a placitus (sic) of 956, subscribed before eight witnesses, in
which Astrulfo and Nomina undertook service in the house of the priest
Zamario and his companion Farega: ‘We promise that we will remain
and live in your house and at your side and in your villa, and we will
do there service (servitio) in the manner of good men (homines bonos).
And if we should commit any fraud in your house or your villa, either
with respect to your work or ours, or if we should attempt to go else-
where (‘in alia parte transire’) without your permission or without your
blessing, as it has been written we will remain as your servi, given over
to your benefit and that of the church of St Martin, and furthermore we
1, 206; Sobrado I 82, 83. See Davies, Acts of Giving, 44; L. Feller and C. Wickham (eds.),
Le Marché de la terre au moyen âge (Rome, 2005), e.g. 68–9, 169, 204–5, 328.
24 See above, Chapters 3 and 9. 25 Liébana 19. 26 PMH 174.
266 Adam J. Kosto
will compensate you with ten oxen without delay and be held by you in
perpetuity.’27 This record, found in the cartulary of Guimarães, was per-
haps initially preserved along with another from four years later in which
the same couple (‘Astrulfo et uxori mee Teodili congomento Nomina’)
sold to the same pair (‘Zamarius et Farega deovota’) a series of proper-
ties that ultimately passed to the monastery.28 It is not unique. In 991,
Argimirus and his wife Ermilla promised lady Trastala (in a document
also called a placitum) that if he dared to transfer his services to some-
one else (‘si ausatum fuerit me ad alia podestate proclamare’) or left
her villa without her permission or knowledge or orders, he would com-
pensate her without delay with 5 solidi plus double damages (‘de quo
agitor dublato’) . . . and he would lose his rights in his cows (‘caread mea
rationem de illas vacas que aveo’).
This is something more than the entry into patronage networks evident
in many contemporary charters,29 but something distinctly less than the
sort of submission-with-reciprocal-obligations evident in commendation
contracts such as the much-studied example from the Tours formula-
book or its analogue in the Formulae Visigothicae.30 They are acknowl-
edgements of dependence, with statements of fines for violation of that
dependence. The function of such records is a bit of a puzzle. But the
fact that Astrulfo and Nomina were simultaneously bound servants and
capable of a substantial donation of land reminds us how poorly we
understand the nature of personal service in the early Middle Ages.
Like the records of sales of servi discussed, and like a fidiatura from
Valpuesta that simply records the penalty for wrongs committed against
the monastery, these documents are of only very temporary use.31 They
are not ephemera, like the famous tenth-century monastic cheese list
from León,32 but proof of uses of documents for more than the preser-
vation of title to land. To the extent that this is one of the reasons given
for an ecclesiastical monopoly on documentary practices, the Iberian
evidence renders that argument weaker.
27 PMH 70: ‘conpromittimus quomodo sedeamus uel habitemus in uestra casa et apud
uos et in uestra villa et faciamus ibidem seruitio sicut facent homines bonos. et si in
uestra casa fraudem fecerimus aut in uestra villa aut in uestro labore aut de nostro
aut si in alia parte transire uoluerimus sine uestro mando aut sine uestra beneditione
sicut in scriptura resona que sedeamus uestros seruus traditus post parte uestra et parte
ecclesia sancti martini et insuper pariemus uobis x boues extra aliqua dilatatione et
uobis perpetim [habituri?]’. Presumably, Zamario served the church of St Martin.
28 PMH 78. 29 Davies, Acts of Giving, 128–9, 149–54.
30 Form. Tur. 47; Form. Vis. 32.
31 Valpuesta 45. In this sense, they are analogous the lone carta securitate from Osona-
Manresa (Kosto, ‘Laymen’, 44).
32 León 480; cf. above, Chapter 4 at n.90.
A third way to think about the use of documents in early medieval

Christian Iberia is to consider the sorts of people who used them. The
low average value of sales and the absolute number of very low-value
sales in a decade-long sample of documents in the Osona-Manresa cor-
pus from Catalonia suggest that the majority of the actors were of modest
social standing.33 A comparable analysis is not possible for the remaining
Iberian material for two reasons: first, we are dealing with data from a
much wider geographical and chronological range, meaning that valu-
ations are likely to be inconsistent; more importantly, price is not con-
sistently recorded, if it is recorded at all, appearing variously in solidi,
modii or (most commonly) in-kind units.34 Nonetheless, Davies, in her
study of these sources, came to the similar conclusion that many of the
actors were not aristocrats: ‘Some will have been relatively rich and some
will have had relatively high status in their local communities; others will
have been relatively poor.’ She adopted the following convention: ‘where
properties of small scale were conveyed, where prices were low, where
their donors and vendors were only associated with properties in one
community, where these donors and vendors did not themselves appear
across a wide area, and/or where witness lists do not include identifi-
able aristocrats, then it is a reasonable assumption that these actors were
peasants’.35 This is perhaps overly restrictive, but given the impossibility
of numerical analyses and the difficulty of identifying individuals (a task
complicated by a limited range of personal names), it is a sensible one,
and one that allows us to consider a portion of the four hundred or so
lay transactions as evidence of peasant document use. The argument is
not a new one: Sánchez-Albornoz described the importance of small-
time property holders in this region (although property holders under
considerable seigneurial pressure), prefiguring the argument made by
Bonnassie for Catalonia that an analysis of the Osona-Manresa corpus
confirmed.36 Scholars have now begun to discuss the extent to which the
circulation of properties controlled by such individuals was or was not a
‘market’.37
33 Kosto, ‘Laymen’, 56. 34 For these problems, see Davies, ‘Sale, Price and Valuation’.
35 Davies, Acts of Giving, 16–22.
36 C. Sánchez-Albornoz, ‘Serie de documentos inéditos del Reino de Asturias’, Cuadernos
de historia de España, 1–2 (1944), 298–351, at 320–1, and Sánchez-Albornoz, ‘Pequeños
proprietarios libres en el reino asturleonés: Su realidad histórica’, in Sánchez-Albornoz,
Investigaciones y documentos sobre las instituciones hispanas (Santiago [Chile], 1970), 178–
201; E. Sáez, ‘Documentos gallegos inéditos del perı́odo asturiano’, Anuario de historia
del derecho español, 18 (1947), 399–431, at 404–6; P. Bonnassie, La Catalogne du milieu du
Xe à la fin du XIe siècle: Croissance et mutations d’une société, 2 vols. (Toulouse, 1975–6),
I, 224–36, 307–43; Kosto, ‘Laymen’, 55–6.
37 Feller and Wickham (eds.), Le marché de la terre; R. Pastor et al., Transacciones sin mercado:
Instituciones, propiedad y redes sociales en la Galicia monástica, 1200–1300 (Madrid, 1999;
trans. as Beyond the Market: Transactions, Property and Social Networks in Monastic Galicia,
268 Adam J. Kosto
It is possible, however, to go further. In two early documents, the

juridical actors are identified as servi. In one, from 864, the servus Auso-
nio granted his lord (domno), Nuño, some orchard land. In 870, the
servus Flaciano made this same Nuño a co-heir of two villas.38 Sánchez-
Albornoz asked the key question: ‘What did men of the kingdom of
Asturias mean by servus?’ – although he went on to assume rather than
argue his answer, that a servus was a slave.39 The existence of servitude
in Iberia in the eighth to tenth centuries is generally accepted, even if
there is debate over the pace of its disapperance. Furthermore, in the
case of domestic servants, servi could accumulate substantial property,
enough even to purchase their own freedom.40 The couples described
in the previous section – Astrulfo and Nomina, and Argimirus and
Ermilla – were not identified as servi, but they were certainly servile.41 It
is not implausible, therefore, that the documents of Ausonio and Flaciano
are the documents of slaves. Indeed, slaves – or rather former slaves –
would have been the parties most interested in the six manumission for-
mulas that begin the Formulae Visigothicae, and the term ‘chartermen’
(cartularii, cartulati) that appears elsewhere in Europe as late as the tenth
century may very well refer to freedmen.42 The case remains unproven,
but, combined with the acknowledgements of service, these two docu-
ments provide strong evidence for the use of documents by laymen at
or very near the bottom of the social hierarchy – in the interest of their
superiors, to be sure, in one case a cleric – but given the assumed dif-
ference in power between a servant and a master, the existence of such
documents is a striking indicator of the routine nature of documentary
practices.
1200–1300 [Leiden, 2002]); L. Feller, A. Gramain and F. Weber, La fortune de Karol:

Marché de la terre et liens personnels dans les Abruzzes au haut moyen âge (Rome, 2005).
Much of the discussion focuses on later evidence, but as the latter volume – and many
of the chapters in the present book – shows, there is an early medieval aspect to the
story.
38 León 3, 4.
39 Sánchez-Albornoz, ‘Serie de documentos’, 321–2. He raises the possibility that this is
a simple humility topos, starts discussing in the conditional the idea that servus means
‘slave’ (‘Si servus equivalı́a a siervo . . . ’), lists various points of evidence, and then asks,
‘who, in view of these texts, will dare deny the advances made by siervos?’ . . . without
ever proving that a siervo is a slave. Cf. León, I, xlix–l.
40 Collins, Early Medieval Spain, 228; Bonnassie, La Catalogne, I, 299–300; Claudio
Sánchez-Albornoz, ‘Los siervos en el Noroeste hispano hace un milenio’, in Sánchez-
Albornoz, Viejos y nuevos estudios sobre instituciones medievales, 2nd edn, 3 vols. (Madrid,
1976–80), III, 1523–611; Davies, Acts of Giving, 18–19.
41 Above, at nn. 27–8.
42 J. L. Nelson, ‘Literacy in Carolingian Government’, in Uses of Literacy, 258–96, at 262;
Niermeyer, Lexicon, s.v. chartulatus; but compare above, 134–5.
Lay archives
Virtually all of the documentary evidence from the early Middle Ages
comes down to us from collections that were for most of the period
between the eleventh and eighteenth centuries, if not longer, in cathedral
and monastic archives, whether in cartulary or single-sheet form. Still,
many of the institutions that preserve these records were themselves
founded only after the year 1000, or in the years just before. Where were
the documents beforehand? The case of Sant Benet de Bages in Cat-
alonia shows how the earliest records, in large part transactions between
laypeople, documented the formation of the (lay) patrimony at the core
of the institution’s initial endowment. Monastic archives, in other words,
could originate in the lay archives of their founders – a phenomenon
seen above in the chapters by Hummer and Costambeys.43 Individuals
wealthy enough to found monasteries are clearly members of at minimum
a local aristocracy. But their archives might contain obvious groupings of
documents – dossiers within dossiers – that reflect the merger, or absorp-
tion, of archives well before the act of foundation, dossiers that might well
have been assembled by less socially prominent individuals, even well-
off peasants.44 Analysis of the early content of several foundations from
elsewhere on the Iberian Peninsula supports these findings.
One opportunity for such documentary archaeology is presented by
the thirteenth-century cartulary of Sobrado de los Monjes. The forma-
tion and organization of the work have not been well studied, but it seems
to contain traces of earlier efforts at cartularization.45 The first volume,
of 196 folios, includes some 645 entries, with documents dating from
803 through the thirteenth century; the second, with 176 folios, con-
tains 541 entries, with documents from 787 through 1252.46 The overall
organization of the volumes is geographical, but of the 119 documents
from the first millennium, all but three are found in the first volume. In
that first volume, the 116 documents from the first millennium appear
among the first 194 entries. If the last entry is discounted, the early
grouping is confined to the first 137 entries, running to folio 58. Despite
the overall geographical organization, then, 97 per cent (115/119) of the
pre-1000 documents in the cartulary are confined to the first 58 folios
of the first volume. In addition, the volume begins with a list of capitula
corresponding to the Roman numerals in the rubrics of the first 106
43 Above, Chapters 8 and 9. 44 Kosto, ‘Laymen’, 60–2.

45 I am working simply from the printed edition. Direct examination of the manuscript
might modify these hypotheses.
46 Sobrado I 220 (a. 1355) and II 334 (a. 1506) are clearly later additions.
270 Adam J. Kosto
documents,47 although in the text the numerals continue through entry

144 on folio 59v, just before a rubric reading, ‘Incipit Liber de Kartulis
de Hereditatibus et possessionibus Graugie Guardianes’.48 A break in
the cartulary is thus evident around folio 59. Of the 115 documents
from 1000 and before in this proposed early section, fully half record
transactions involving only laymen.
The source of many of these documents is not hard to discern. The core
of the monastery’s endowment originated in a series of grants between
952 and 966 by Ermengildo and Paterna and their sons (Bishop) Sis-
nando and (Count) Roderic, along with the latter’s wife Geloira.49 All
of the lay transactions date from 953 and earlier. From between 916
and 953, there are twenty-eight documents recording alienations by lay-
people to the founding couple, plus three to Ermengildo alone and four
to Paterna alone.50 These include sales, gifts, exchanges, profiliations
and fines, the latter category attesting to the power of the couple in the
region. To these can be added two alienations by priests to the couple.51
Most of the documents record the acquisition of land reflected in the ini-
tial endowment. The same is true of two grants by laypeople in 946 and
951 to Roderic and Geloira.52 These thirty-nine documents are likely
to have been part of the archive of a local aristocratic family that was
turned over complete to the new institution when the estate became the
monastery.
An inventory included in the archive may give us an even more pre-
cise glimpse of the moment of transformation.53 The date with which
the document starts (‘In era [Ma VIIIIa ]’; that is, 971) is evidently a
later addition. The text proper begins, ‘Noticia uel inuentario de uillas
et hereditates que in testamentos de monasterio Superato resonat, que
47 ‘Incipiunt capitula primi libri de hereditatibus et possessionibus que fuerunt antiquo

tempore de monasterio Superaddi in diebus regis domini Ordonii et regis domini Uere-
mundi et regis domini Ranemiri et regine domine Tasarie et regis domini Adefonsi.’
48 The exceptions are Sobrado II 46 (s. X), 172 (a. 787), 278 (a. 996).
49 Sobrado I 1–6.
50 Couple: Sobrado I 10–12, 14, 16, 21–7, 29, 31, 60, 66–7, 70–1, 78–9, 95, 98–9, 101–
3, 111. Ermengildo alone: Sobrado I 28, 53. Paterna alone: Sobrado I 15, 91–3. The
documents to Paterna alone may instead be to the Paterna married to a Menendo,
attested as a couple between 930 and 953 in the five lay alienations Sobrado I 54, 85,
87, 89, 100 (see also no. 105, a. 921, a grant from a priest) – unless Menendo is the
same person as Ermengildo. Sobrado I 63 (Ermengildo and Paterna, a. 877) and 75
(Ermengildo alone, a. 858) are probably either different individuals or misdated.
51 Sobrado I 32, 61; see also 105, a sale by a priest to Menendo and Paterna.
52 Sobrado I 69, 90.
53 Sobrado I 112. Cf. below, at n. 72. Other tenth-century inventories include Celanova
177, 183–4, 239; La Coruña 87; León 505; Sobrado I 194; PMH 27.
fuerunt de Heremgildo et Paterna, et postea de filiis eorumdem, Ses-

nandi episcopi et germani sui Ruderici ducis et uxoris sue Giloire, et
posuerunt illas insimul in ipso prefato monasterio.’ The document lists
over one hundred villas and churches, organized geographically. The
grammatical antecedent of illas could be villas et hereditates, but the sense
of the sentence suggests that it instead refers to testamentos: not all of
the lands of Ermengildo and Paterna passed first to their children (‘que
fuerunt . . . postea de filiis eorumdem’), and it makes more sense to place
documents than villas ‘in’ a monastery. Thus this may be as much an
inventory of documents – of the family archive – as it is a record of the
lands passed on to the Church. It might be argued that it is simply a
collation of the properties named in the six founding documents, but
although many of the place names in the inventory appear there, not
all of them do (e.g. Abolim, Anobre, Auteiro, Aurioles, Avianco, Balio,
Bavegio . . . ). If this hypothesis is correct, the geographical organization
of the inventory may reflect a geographically organized archive of the
sort seen at St Gall (described by Innes, above54 ), along the lines of what
would later become standard monastic practice (shown in the predom-
inantly geographical contemporary dorsal notations on surviving early
medieval original charters).
Whether or not the contents of the Ermengildo/Paterna archive were
inventoried in this manner, it is likely to have contained not just docu-
ments recording alienations to the family but also documents related to
the earlier histories of those acquisitions. The villa Salamir, mentioned in
the inventory and the founding documents, is the subject of a document
recording a sale in 858; likewise, the villa Mundini, also in the inventory
and the founding documents, is the subject of a document recording a sale
in 905.55 These are not yet direct links: documents recording the transfer
to the person who ultimately transferred the property (and presumably
the original document) to Ermengildo and Paterna. We come closer with
the villa Marciani. A document of 927 records that Ermengildo and
Paterna purchased the property there of one Lebora and her son Gisulfo.
Gisulfo is said to have inherited the land from his father Aloito, who is in
turn attested in a document of 895 purchasing lands at the villa Marciani
(and elsewhere) from the priest Gundesindus.56 The lands may in fact
be different – Gisulfo claims to be selling land that had been in the family
already with his grandfather – but the coincidence of person and place
suggests a documentary link.
54 Above, Chapter 7.
55 Salamir: Sobrado I 72; cf. I 4, 6, 112. Mundini: Sobrado I 65; cf. I 3, 112.
56 Sobrado I 79, 81.
272 Adam J. Kosto
In addition to working backward from the documents of Ermengildo

and Paterna, one can work forward from the earliest documents pre-
served in the cartulary. Four of the ten oldest documents in the collec-
tion – indeed, some of the oldest surviving post-Visigothic documents
from anywhere on the peninsula – are associated with the family of a cer-
tain Pompeyano. In 803, his children Ysilo and Fafila exchanged various
properties. In 817, he granted to Fafila and Fafila’s wife, Penetruda, a
group of five villas. In 838, he sold to the couple Herfonsus and Hermildi
a plot of land. Then, in 860, four of his grandchildren (Fafila’s children)
sold a plot of land to a certain Cenebrida.57 The lands mentioned in the
documents of 803, 817 and 860 overlap; the 838 sale seems to be sep-
arate. Many of the place names mentioned in the documents appear in
the first endowment charter of 952, granted by Ermengildo and Paterna,
and in the inventory from around 971,58 but none of the names of the
beneficiaries of these documents outside the family – that is, Herfonsus,
Hermildi and Cenebrida – appear later in the Sobrado cartulary. There is
no evidence, therefore, that any of these documents entered the Sobrado
archive in connection with a family other than that of Pompeyano, fur-
ther supporting the idea of these documents as a coherent group. The
transactions of 803 and 817 were both documents that Fafila and his
descendants would have kept to demonstrate title, but the other two are
alienations from the Pompeyano family, and for a private individual to
keep a duplicate parchment copy of an alienation – as proof, for example,
that he or she was no longer subject to legal obligations attached to the
property – would represent a level of document use not usually imagined
for this period.
If the Pompeyano documents are the remnants of an archive – for they
may just be a random collection of documents that happen to be inter-
related – it is likely that it is not the archive of Pompeyano, but rather
of a later individual who consolidated the lands in question. One can-
didate is Sisnando, the author of an endowment (cartula dotis) in 887:
the thirty villas he granted to his wife in 887 (thus a generation after the
last Pompeyano document) include at least three mentioned in the Pom-
peyano documents: Laureda, Mandeo and Codesosso.59 Furthermore,
Sisnando was likely related to Ermengildo or Paterna.60 But some lands
57 Sobrado I 82, 77, 18, 76. See also Sáez, ‘Documentos gallegos inéditos’.
58 Sobrado I 1, 112. 59 Sobrado I 119.
60 The genealogy at M. del C. Pallares Méndez, El monasterio de Sobrado: Un ejemplo de
protagonismo monástico en la Galicia medieval (La Coruña, 1979), 60, gives an Eldondia
as a cousin of Ermengildo, and one of the children of the counts is named Sisnando.
The dual coincidence of names suggests a family connection.
and documents of Pompeyano may have been granted directly to the lat-
ter couple: in 927, Segeredo, grandson of Fafila, granted them the villa
‘que dicunt Pompeiani’, with similar boundaries to the lands described in
Pompeyano’s grant of 817.61 Again, the coincidence of names, places and
documents is a strong indicator of a real documentary tradition behind
the one represented in the cartulary.
The monastery of Santa Maria de Otero de las Dueñas offers a similar
opportunity for this sort of approach, but with two important differences:
in this case, the documents are preserved in large part as originals, but
the monastery itself was not founded until the thirteenth century. Two
of the great-great-grandfathers of the founder, Marı́a Núñez, were the
eleventh-century counts Fruela Muñoz and Pedro Flaı́nez, whose fami-
lies’ acquisitions dominate the earliest holdings of the monastic archive.
The lineages were joined by the marriage of the daughter of the former
to the son of the latter, probably in the 1040s. Alfonso Garcı́a Leal has
identified 243 documents from the Otero collection presumably in the
archives of these two counts when they were united shortly after the
death of Pedro Flaı́nez in 1048 (Fruela Muñoz died in 1046), 72 record-
ing acquisitions of Fruela Muñoz between 993×1003 and 1046, and 68
of Pedro Flaı́nez between 996 and 1047.62 In the period of interest here,
however, the two patrimonies, and thus their putative archives, would
have existed separately. In fact, as the dates of the earliest documents
of these two counts suggest, the tenth-century and earlier records offer
instead a picture of the archives of their ancestors.
Of the fifty-one documents from 1000 and earlier in the Otero archive,
thirty-one concern transactions between laypeople; in addition, there is
one royal grant to a layperson and one sale from a priest to a layperson.
Of these, three are the earliest acts of Pedro Flaı́nez and Fruela Muñoz,
but most relate instead to the activity of their parents and grandparents:
two sales and one grant to Muño Flaı́nez and Froileba (Pedro’s paternal
grandparents, discussed below in the context of Sahagún);63 seven acqui-
sitions by Flaı́n Muñoz, with and without his wife Justa (Pedro’s parents),
61 Sobrado I 27: ‘iuxta riuulo Mando et discurrit ipsa villa per terminos de Roadi et inde
in terminos de Colimbrianos et inde in terminos de Sancto Iuliano, et sunt ipsos terminos
antiquos’ (my emphasis). Cf. Sobrado I 77: ‘in riba de Mandeo . . . villa que est usque in
terminos de Colimbrianos, et inde usque in terminos de Roadi’. See Pallares Méndez,
El monasterio de Sobrado, 6n8.
62 A. Garcı́a Leal, ‘El archivo de los condes Fruela Muñoz y Pedro Flaı́nez (854–1048):
Una visión nueva de viejos documentos’, Signo: Revista de historia de la cultura escrita, 13
(2004), 121–47; Garcı́a Leal, ‘Los condes Fruela Muñoz y Pedro Flaı́nez: La formación
de un patrimonio señorial’, Anuario de estudios medievales, 36 (2006), 1–110. Marı́a
Fruélaz was born after 1029 (Leal, ‘Los condes’, 10).
63 Otero 6, 11–12; below, 278–9.
274 Adam J. Kosto
six of which were directly linked to his judicial activity;64 and five acquis-
tions by Fruela Vimaraz, with and without his wife Adosinda (Fruela’s
maternal grandparents), three of which are linked to judicial activity, and
one of which was a royal grant from Vermudo II.65 The volume of records
linked to judicial activity, which only increases under Pedro (19 records)
and Fruela (17 records), might be seen as giving the collection a public
or institutional character,66 but these were very much family archives.
They contain not only acts benefiting the counts themselves but also the
endowment of Fruela’s half-sister, Urraca (daughter of the first marriage
of his mother, Jimena),67 and an inventory of documents assembled by
the parents of the second wife of Pedro Flaı́nez’s father.68 Most of the
remaining lay transactions concern properties easily associated with one
of the other families; several, for example, refer to Valdoré and the Esla
River valley, one of the principal bases of the Flaı́nez, while two refer to
Cirujales, in the territory dominated by the Fruélaz.69 As in the Sobrado
archive, direct links between early documents and later transmission are
difficult to discern, but there are hints. Of the earliest records, three are
sales to the couple Adriano and Eremesinda. One (a. 936) refers to Val-
doré, and the other two (both a. 949) to El Montecillo; inherited property
at the latter location was granted by Pedro’s son to his own children in
1057.70 A sale in 999 of property at Corniero is echoed in a transfer
(for judicially ordered compensation) in 1048, and then again in Pedro’s
son’s grant of 1057.71
The inventory merits further analysis.72 It is a list of nine documents
prepared by Fernándo Vermúdez and Elvira, likely the parents of the
second wife of Pedro Flaı́nez’s father.73 The date with which the doc-
uments begin (‘Era millesima XIIII’, or 976) is plausible, but also evi-
dently a later addition. It is labelled an ‘Agutione [recte agnitione?] vel
inventarium . . . de cartas de Villa Naptaulio’. The place in question has
been identified as El Natahoyo, which also appeared in the 1057 grant
by Pedro Flaı́nez. The documents themselves are not preserved in the
64 Otero 27, 33–4, 36, 38–9, 44. 65 Otero 30 (royal), 37, 40–1, 46.
66 Davies, Acts of Giving, 143–6; A. Prieto Prieto, ‘Documentos referentes al orden judicial
del monasterio de Otero de las Dueñas’, Anuario de historia del derecho español, 44
(1974), 619–74; C. Estepa Dı́ez, ‘Poder y propiedad feudales en el perı́odo astur:
Las mandaciones de los Flaı́nez en la montaña leonesa’, in Miscel.lània en homenatge
al P. Agustı́ Altisent (Tarragona, 1991), 285–327. For the difficulties of making such
distinctions, above, 11, 15, 21–3, 50, etc.
67 Otero 50; Garcı́a Leal, ‘Condes’, 8–9. 68 Otero 22; below, at n. 72.
69 Otero 3–4, 14, 24–5, 49 (Valdoré, Esla); 22, 28 (Cirujales).
70 Otero 3, 7–8; Garcı́a Leal, ‘Condes’, 86–7.
71 Otero 45; Garcı́a Leal, ‘Condes’, 86–7, 103.
72 Otero 22. 73 Garcı́a Leal, ‘Los condes’, 66.
Otero archive; indeed, none of the individuals or place names appear

earlier in the corpus. Of course, the grandchildren to whom the lands
passed were not the direct ancestors of Marı́a Nuñez; if the charters
passed with the land, they would have left the family archive. Around the
year 1000, however, the land (and thus the documents) would have been
in the possession of Pedro Flaı́nez.
Nine documents are listed and follow a general form: the carta of so-
and-so, concerning such-and-such a type of property, located in such-
and-such a place, such-and-such a portion. Thus the first is the char-
ter of Donno, concerning an orchard (pumare) in Ramma, next to the
house of Negella, for the whole of the enclosure (clusa). The second is
the charter of Graciano, again concerning an orchard, in La Quintana,
where Serpento lived, for a one-third portion. And so on through the
ninth, the charter of Ranimiro Froilaniz, in Las Linares, bordering the
lands of the sons of Abbatinus, for half of the lands. Documents 4 and 5
are charters of a Queen Uracca, one along with her son Ordono, another
with her sons Ordono and Ramiro.74 These descriptions reveal the author
of the inventory working directly from the original documents, not both-
ering to clarify pronominal references: ‘one property borders on the strata
kavallari, and the second is above the house of Armentarius, and the third
borders your lands’; ‘the field that is next to your house’.
The contrast of this inventory with the Ermengildo/Paterna inventory
is striking. Here there is no ambiguity: it is unquestionably a list of
documents. And the scribe who drew it up was clearly interested in
being able not just to identify the documents but also to digest their
content; if the beneficiaries are assumed to be Fernándo Vermúdez and
Elvira, who are after all named at the outset of the document, then the
only key details missing from the point of view of a modern diplomatist
are dates and (if sales) prices. In offering information about boundaries
and the precise portions of property transferred, the scribe gives more
information than would be required if the record were merely a finding list
for the documents (compare, for example, the Ghittia list from Lombard
Italy, which distinguishes various types of documents in phrases that
recall dorsal notations).75 What the Fernándo Vermúdez/Elvira list offers
is the details of the legal acts; in that sense it is more of a miniature
cartulary or Traditionsbuch, with a narrow geographical focus on a single
74 This could be either Urraca, second wife of Fruela II of Asturias, who had sons by him
of with those names between 917 and 925, or Urraca Fernández, who had a son named
Ordoño by Ordoño III of León between 951 and 956, and a son named Ramiro by
Sancho II of Pamplona after 970. Alternatively, the charters could be from two different
Urracas. The first scenario seems most likely.
75 ChLA XXVI 808; CDL II 295; above, 239.
276 Adam J. Kosto
villa. It is impossible to know how difficult it was to assemble these nine

documents (How many others were with them in the scrinium? Were they
already organized in labelled bundles, or was it a messy pile?), but the
author of the list was clearly engaged in the preservation, organization and
transmission of documentary records. The author was, in other words,
working not with a mere collection of documents, but with an archive,
and was doing so for laypeople.
Sobrado and Otero are examples of what might be called the ‘founder
phenomenon’: monastic archives with their origins in the family archives
of an aristocratic founder. San Toribio de Liébana offers a variation on
this theme: the lay transactions that dominate the earliest holdings of the
archive are not those of the founder, but rather of the family of an early
abbot, Opila. The earliest forty-two documents in the archive (to 932)
include twenty lay transactions, of which fourteen record the acquisitions
of Opila’s parents, Bacaudano and Faquilona.76 This example of the
integration of a lay archive into a monastic archive – rather than the
wholesale transformation of one into the other – is fairly straightforward.
Traces of lay archives in early institutional holdings are not always so
obvious, as is shown by the case of Sahagún.
At least seven substantial dossiers survive among the ninth- and tenth-
century documents of Sahagún, all of which look different from an
archival perspective. Four of the five oldest documents are sales by
laypeople to laypeople of properties at Piasca.77 Of these, the three ear-
liest are sales to the couple Argemundo and Recoire, one of a vineyard
from Recoire’s brother in 857, and two of various properties from the
couple Daildi and David in 861. The daughter of one of the vendors
in the fourth document, from 904, is named Daildi; given the rarity of
the name, which appears nowhere else in the Sahagún corpus, a family
connection is possible.78 The monastery of Santa Marı́a de Piasca, which
was incorporated into Sahagún in the twelfth century, is first recorded in
930,79 well after all of these documents. A plausible hypothesis, then, is
that the lands of Argemundo and Recoire – along with these documents
– were at some point transferred to the monastery of Piasca, and then
to Sahagún in the twelfth century.80 Yet the coincidence of the place
76 Liébana 18, 21, 23, 25–8, 32, 35–6, 39–42. The other lay transactions are Liébana
13–14, 19, 22, 31, 38.
77 Sahagún 1–3, 5. The fifth document, DEPA 94 (a. 869), has no evident connections to
any others in the corpus.
78 See, however, DEPA 117 (a. 877), in which the donor’s grandfather is named Daildi.
79 Sahagún 39.
80 Other documents from a hypothetical ‘Santa Marı́a de Piasca dossier’ include Sahagún
79, 96, 134–6, 153–4, 242, 288, 304–5.
name and the presence of these documents in the Sahagún archive are
their only connections to the monastery of Piasca (although a grant to
the community in 945 includes lands iusta [domum] Require),81 so three
alternatives remain: first, that there is no connection between the lands in
question and the monastery of Piasca, and at some point the lands were
transferred, along with the documents, directly to Sahagún; second, that
the land was never transferred either to Piasca or to Sahagún, but that
documents were stored in the archive of Piasca after its founding and
later incorporated into the Sahagún archive; and third, that the land was
never transferred either to Piasca or to Sahagún, but that the documents
were stored at Sahagún itself from the beginning.
The survival of these four documents as originals has some relevance
here. Most of the early charters of Sahagún survive not as originals, but
in the early twelfth-century Becerro gótico.82 Two-thirds of the originals
before 950 (twenty-two of thirty-four) are royal or episcopal documents
or transfers directly to Sahagún;83 of the remaining third (twelve), seven
relate to Piasca,84 and two others are part of other groups of docu-
ments that probably represent archival dossiers.85 Both of these dossiers
offer instructive comparisons. The first concerns the monastery of San-
tiago de Valdávida. This monastery was endowed by Vermudo Nuñez,
count of Cea, in 949, on confiscated lands that had been granted to
him by Ramiro II in 943. The ninth- and tenth-century Sahagún cor-
pus includes ten additional documents concerning Valdávida (without
mentioning Sahagún). Of all these documents, only two, the endowment
charter and a sale of 969, survive as originals, and all of the documents
(including the two just mentioned) appear in the Becerro gótico.86 Thus
the Valdávida documents must have been incorporated into the Sahagún
archive when the monastery was absorbed sometime before the compo-
sition of the cartulary in 1110. The second dossier concerns the priest
81 Sahagún 96.
82 Of Sahagún 1–129 (that is, the documents from 950 and earlier), ninety-three docu-
ments list the Becerro as the only or the earliest exemplar. This includes a number of
document groups listed under a single number (Sahagún 34 [four acts], 36 [seven acts],
94 [twelve acts]).
83 Royal diplomas for Sahagún: Sahagún 6–8, 61, 70, 93, 98–9, 129. Royal diploma for
another monastery: Sahagún 21. Transfers directly to Sahagún: Sahagún 18, 25, 29, 56,
80, 87, 120, 126, 128. Episcopal documents not mentioning Sahagún: Sahagún 33, 78.
Episcopal confirmation for Sahagún: Sahagún 101.
84 Sahagún 1–3, 5, 39, 79, 96.
85 Sahagún 114 (Valdávida), 125 (Melic). The remaining three include two transfers
between laypeople that I have not yet traced (Sahagún 4, 52) and a grant to the church
of San Félix de Cisneros, which again I have not yet traced (Sahagún 105).
86 Sahagún 84 (grant to Vermudo), 113, 114 (endowment), 142, 157, 172–3, 208, 217,
252 (sale of 969), 254, 257.
278 Adam J. Kosto
Melic, whose documents include twelve sales and a profiliation (from 932
to 951) before his own grant to Sahagún and a testament of 959; only
one of the thirteen earlier documents survives as an original, and it, too,
is included in the Becerro.87 The presence among the earliest Sahagún
originals of two charters clearly linked to dossiers with demonstrable
links to Sahagún might be used to support an argument that the Piasca
documents are part of a similar dossier; that is, that the Pisaca originals,
too, were preserved because they documented holdings of Sahagún. The
fact that Piasca was not absorbed until after the composition of the
Becerro, however, means that we cannot similarly test whether the com-
piler of the Becerro gótico judged those documents to be relevant to a
picture of Sahagún’s holdings. It remains possible, then, that these docu-
ments were stored in the archive despite having no formal connection to
Sahagún.
A fourth substantial dossier (in addition to Piasca, Valdávida and
Melic) appears to be a true lay archive, and from members of a family
we have already seen to have archival habits: Muño Flaı́nez and Froileba,
the grandparents of Pedro Flaı́nez of the Otero archive. Muño and his
wife, and then Froileba alone after Muño’s death, systematically acquired
property at several locations: Valle de Laurenzo (by purchase in 947, 948
and 950, all drawn up by the priest Fredemundo),88 Aleje (by purchase
in 958, 959 and 961),89 Noanca (by profiliation in 961; by purchase in
961, 962 and 963; by default on a loan in 962)90 and Caso (by grant in
960 and by the same profiliation of 961).91 Three other places appear
in only one document each (Loides, profiliation, 960; Corniero, profilia-
tion, 961;Vilasecos, purchase, 962); one sale, from 948, does not specify
a location.92 One output of this spree of acquisition was a set of fifteen
documents produced over a fifteen-year period, all with Muño and/or his
wife as beneficiaries. In 985, their daughter Jimena made a large donation
to Sahagún, which mentions lands at Noanca and Valle de Laurenzo.93
The correspondence is not perfect: the land at Noanca is said to have
been purchased from the ‘omes de Colle’ rather than inherited from her
parents; the money for the purchase of the land at Valle de Laurenzo,
rather than the land itself, is said to have come from her parents; and the
87 Sahagún 45 (profiliation), 69, 75, 81, 90, 103, 109, 111–12, 121, 125 (original), 131,
133, 167 (grant to Sahagún; original, not in Becerro), 168 (testament; original and in
Becerro). Sahagún 183, a royal confirmation of the testament of Melic, which survives
as a single sheet and in the Becerro, appears to be a forgery. An additional sale to Melic,
in 951, survives in the Tumbo of the cathedral of León: León 232.
88 Sahagún 107–8, 124. 89 Sahagún 161, 163, 189.
90 Sahagún 191, 193, 198, 201, 213. 91 Sahagún 178, 191.
92 Sahagún 110, 184, 191, 205. 93 Sahagún 328.
other place names do not appear. A subsequent grant by Jimena of lands

at Valle de Laurenzo does mention her parents as previous owners, how-
ever, so we should not rule out still other donations of family lands that
would cover the missing properties.94 In contrast to the earliest dossiers,
only one of the fifteen documents survives as an original;95 the rest are
preserved in the Becerro gótico of 1110.
A similarly sized dossier is more difficult to connect to Sahagún.
Between 958 and 965, Ansur and his wife Marı́a made nine purchases of
land at Tindal, Villanueva, León, Otereulo, Lampreana and Trejeco;96
they also acquired by profiliation land at Villa Alva.97 They granted the
land at León, a field (ferragine) acquired in 961, to Sahagún in 963.98
At some point between 965 and 970, Ansur acquired a new wife, Ild-
uara, and the new couple resumed the old couple’s campaign, making six
more purchases between 970 and 976 at Trobajo, Rebolare, León, Villa
de Petro, Villa de Belone and Quintila.99 The lands at Villa de Petro and
Villa de Belone were purchased from the monasteries of San Martı́n de
Valdepopulo and Santa Marı́a de León (?) and then immediately granted
to San Salvador de Palazuelo; they then appear in a royal confirmation
to Sahagún of 976 and in a dispute in 998.100 The only grant directly to
Sahagún other than of the field at León was made by Ilduara, now the
widow of Ansur, of two villas not named in earlier documents.101 In all,
then, there are sixteen documents recording transactions in which Ansur
and his wives are the beneficiaries that ultimately made their way into
the Sahagún archive. All of them are preserved in the Becerro, while only
three also survive as originals102 – a good indication that even though few
of the lands can be traced to the monastery (in the early documents, at
any rate), they eventually became its property.
The dossier of Iscam and Filaura is slightly different, in that all but one
of the surviving transactions concern a single place, Melgar de Foracasas,
and the collection is preserved in a curious manner.103 Twelve sales from
between 945 and 954, totalling over 133 solidi,104 are listed in abbreviated
94 Sahagún 345.
95 Sahagún 205. Jimena’s grant of 985 also survives as an original, although the later grant
does not.
96 Sahagún 158, 169, 182, 187, 194, 199, 210, 223, 237.
97 Sahagún 214. 98 Sahagún 194, 215. 99 Sahagún 260, 268, 278, 282.
100 Sahagún 269, 272–3. The sales are recorded in confirmations dated after the grant, and
the second grant refers to the first – I have not yet sorted out this puzzle.
101 Sahagún 308.
102 Sahagún 214, 268, 272. The grant to San Salvador, Ansur’s testament and Ildonza’s
grant also survive as originals (Sahagún 269, 284, 308).
103 Sahagún 94.
104 107 solidi in coin or kind, 26 arenzos (solidi in coin), 22 modios of wheat, a plot of land
(terra placabile) and a cow (vacca ruvia); for one of the transactions, the price is missing.
280 Adam J. Kosto
form in a single document in the Becerro. Each item lists just the seller,
the property, the price and the date. A sanction clause and subscriptions
(‘nos omnes supra scripti hoc confirmamus’) follow the last document.
Whether a multiple document was copied whole into the Becerro or the
individual documents were synthesized for the purpose of the cartulary
(the common subscription, absent in two other multiple documents in
the corpus, may indicate the former solution105 ), at some point there was
a collection of twelve charters relating to Melgar, which would have been
transferred to the monastery at the time of two grants by Iscam, Filaura
and Iscam’s sister in 959, the second of which survives as an original.106
The dossier of Iscam and Filaura is striking in that it brings us down to
a quite modest social level: that of well-off peasants. The couple were
landowners, to be sure, but they dealt with property in only a single
place, and with small, low-value plots. If the twelve documents represent
an archive or part of an archive, it was a peasant archive.
Muño Flaı́nez and Froileba were not, by contrast, just any laypeople,
nor were Fernándo Vermúdez and Elvira, or Ermengildo and Paterna.
They were local aristocrats – if not always counts themselves then moving
in comital circles. Given the relative flatness and fragmented power struc-
ture of Iberian society in this period, comital designations and connec-
tions reveal more about their relative than their absolute power. Nonethe-
less, these were the families with the inclination and resources to amass
the amount of property that could endow a monastic institution stable
enough to last and thus to preserve their documents for posterity. But
the Iberian dossiers just examined reveal that the archival habits of such
families – accumulation, organization and transmission of documents –
preceded their ties with ecclesiastical institutions, and furthermore drew
on documentary habits of their social inferiors. Early medieval archives,
like early medieval documentary practices generally, were not an ecclesi-
astical monopoly.
Uncovering the rhythms and logic – economic, social and cultural – of the
massive transfer of property out of lay hands and into the hands of eccle-
siastical institutions has been one of the great achivements of medieval
historical scholarship. François Menant offers, however, a caution in the
Italian context:
There is a risk that the dominant impression is one of a permanent shift from lay
property to Church property, an impression that is surely false, or at least very
incomplete. On the other hand, the intermediate stages of concentration of small
105 Sahagún 34, 36.

106 Sahagún 162, 164. The first grant also includes lands in Valderaduey and Boñar.
peasant property, equally evident in the documentation, must take on greater

importance: the concentration of small peasant property in the hands of more
substantial landowners, rural but especially urban, whose holdings could pass in
turn to the Church.107
The lay archives discussed in the second part of this chapter illustrate
the validity of this model: the archival dossiers of lay monastic founders
record clearly a process of concentration of smaller individual properties
in the hands of substantial landholders. But this was merely the penul-
timate step, as even older records, the ones that record the transfers
previous to the absorption of a property into a great estate, hint at the
situation before: a world of modest aristocrats, small-time landowners
and even humble servants, clerical and lay, buying and selling land, peo-
ple and goods; paying fines; entering into agreements – and writing it all
down frequently enough that to do so could be considered local custom.
Additional abbreviations:
Cardeña Colección documental del monasterio de San Pedro de

Cardeña, ed. G. Martı́nez Dı́ez (Burgos, 1998)
Celanova Colección diplomática del monasterio de Celanova
(842–1230), ed. E. Sáez and C. Sáez, vols. I–III (Alcalá
de Henares, 1996–2006)
Corias El registro de Corias, ed. A. Garcı́a Leal (Oviedo, 2000)
La Coruña La Coruña: Fondo antiguo (788–1065), ed. C. Sáez and
M. del V. González de la Peña, vols. I–II (Alcalá de
Henares, 2003–4)
DEPA Diplomática española del perı́odo astur: Estudio de las
fuentes documentales del Reino de Asturias (718–910), ed.
A. C. Floriano, 2 vols. (Oviedo, 1949–51)
León Colección documental del archivo de la catedral de León
(775–1230), ed. E. Sáez et al., vols. I–III (León,
1987–90)
Liébana Cartulario de Santo Toribio de Liébana, ed. L. Sánchez
Belda (Madrid, 1948)
Otero Colección documental del monasterio de Otero de las Dueñas,
ed. J. A. Fernández Flórez and M. Herrero de la Fuente,
vol. I (León, 1999)
107 F. Menant, ‘Les transactions foncières dans le Royaume d’Italie du Xe à la fin du

XIIe siècle’, in Feller and Wickham (eds.), Le marché de la terre, 147–60, at 151. In
Iberia, Davies finds the increase in ‘volume of peasant engagement with ecclesiastical
landlords’ – replacing relationships with lay lords – to be the main lesson of the tenth-
century charters (Davies, Acts of Giving, 221).
282 Adam J. Kosto
PMH Portugaliae monumenta historica . . . , ed. H. de Carvalho e

Araujo and J. J. Da Silva Mendes Leal, part 3, Diplomata
et Chartae, vol. I (Lisbon, 1868)
Sahagún Colección diplomática del monasterio de Sahagún (siglos IX
y X), ed. J. M. Mı́nguez Fernández (León, 1976)
Sobrado Tumbos del monasterio de Sobrado de los Monjes, ed.
P. Loscertales de Garcı́a de Valdeavellano, 2 vols.
(Madrid, 1976)
Valpuesta Cartulario de Valpuesta, ed. M. Desamparados Pérez
Soler (Valenica, 1970)
11 On the material culture of legal documents:
charters and their preservation in the Cluny
archive, ninth to eleventh centuries
Matthew Innes
Cluny has been central to the previous scholarship on lay documentary practices.
In this chapter, however, and in the context of the rest of this volume, it appears
now as a late tenth- and early eleventh-century manifestation of precisely the
same processes we have seen elsewhere that generate our incomplete image
of a common documentary culture. We note again the importance of intra-
familial transactions, and the ongoing nature of the relationship between found-
ing families and the institutions that eventually absorbed their documents. The
case of Cluny also underlines that documents served purposes beyond storing
and transmitting their content, in the same way that archives served purposes
beyond storing and transmitting their documents.
It may seem optimistic to devote an entire chapter of a book on the lay use
of documents to the Cluny charter collection. After all, Cluny is probably
the best studied monastery of the earlier medieval period: surely any char-
ters preserved there can do no more than simply confirm the well-known
maxim that our sources survive because of the institutional continuity of
ecclesiastical archives? This chapter demonstrates that the contents of the
Cluny archive, properly analysed, categorically refute this idea. Precisely
because the Cluny charter collection was not only voluminous but also
the subject of intensive early modern and modern interest, its remnants
allow a detailed consideration of lay documentary and archival practices,
and their relationship to the development of the institutional practices of
the Church. Building on the other contributions to this book, this chap-
ter sets out to sift through the contents of this institutional repository
and to identify earlier layers of archival and documentary practice within
the Cluny material. It goes on to argue that we can not only reconstruct
archival and documentary practices that preceded and were subsumed
by ecclesiastical institutions, but we can also reach a deeper understand-
ing of the ways in which these practices were eventually pushed beneath
the historiographical horizon by post-Carolingian archival reorganization
283
284 Matthew Innes
and institutional memorialization. With the Cluny material it is possible

not only to surmise the existence of a submerged iceberg of lay documen-
tation but also to map the contours of the frozen matter hidden beneath
the waves. The findings of this chapter also speak to two other crucial
debates, the first epistemological and the second historiographical.
First of all, considering the transmission of the Cluny charters exposes
the need to integrate classical diplomatic scholarship, with its empha-
sis on textual typologies, with a more contextualizing approach that
embraces the material culture of legal documents: the physical traces,
annotations and alterations that are the best evidence we have of how
a document was received, used and understood.1 After all, Jean Mabil-
lon and the other early modern founders of diplomatic scholarship were
concerned with establishing rules to ascertain the validity or otherwise
of documents as a means to advance live legal and political arguments.
The subsequent reception of their methods in the nineteenth and twen-
tieth centuries and their adoption as the bedrock of national archives
and the newly professionalized historical scholarship of the university
curriculum owed much to the imperative of establishing an objecti-
fied narrative of identity shaped by the needs of the modern nation
state.2 Nonetheless, the criteria they established to test the reliability and
date of particular texts have been indispensable in building an agreed
evidential corpus and a set of methodological rules on which histor-
ical study can be based. The editions made possible by this work have
brought the fruits of medieval archives to a wider constituency of modern
1 Cf. G. Koziol, The Politics of Memory and Identity in Carolingian Royal Diplomas: The
West Frankish Kingdom (840–987) (Turnhout, 2011). Like Koziol, I treat documents as
material objects implicated in living social relationships that they helped to define, but
I differ in my specific readings of the relationships between their embedded and ritual
meanings, on the one hand, and their practical intent as written record on the other. I
have benefited from the comments and observations of my colleagues in the Lay Archives
Working Group, in particular Adam Kosto and Hans Hummer, whose own observations
on Cluny sparked my own; from discussion with Jon Jarrett of work he undertook as a
research assistant databasing placita and cartularies in work funded by Birkbeck College
and the Leverhulme Trust; and from the comments and generosity of Wendy Davies,
Barbara Rosenwein, Julian Swann, and Charles West. It is more than usually necessary,
therefore, to emphasize that all conclusions and interpretations, and, above all, all errors
and infelicities, in what follows are my own.
2 Mabillon: D. Knowles, ‘Jean Mabillon’, in Knowles, The Historian and Character (Cam-
bridge, 1964), 213–39, and Knowles, Great Historical Enterprises (Cambridge, 1962),
2–62. Functions of professionalized medievalism in the nation states of classical moder-
nity: P. Geary, The Myth of Nations: The Medieval Origins of Europe (Princeton, NJ,
2003), 15–40; M. Innes, ‘A Fatal Disjuncture? Medievalism and Medieval History in
Britain’, in J. Jarnut and H.-W. Goetz (eds.), Mediävistik im 21 Jahrhundert: Stand und
Perspektiven der internationalen und interdisziplinären Mittelalterforschung (Munich, 2003),
73–100; R. J. W. Evans and G. P. Marchal (eds.), The Uses of the Middle Ages in Modern
European States: History, Nationhood and the Search for Origins (London, 2010).
The Cluny archive, ninth to eleventh centuries 285
historians. Thus, documents as initially experienced by many modern

scholars have undergone an ontological transformation, from material
remains into standardized texts, divorced from the context of their storage
and transmission. The establishment of canonical source-criticism has
embedded a set of unwritten assumptions – some empirical and some
epistemological – about what might constitute historical evidence and
thus limited the perspectives of possible history.
The material culture of documents – their physical appearance,
production and performance – was central to their reception by
contemporaries, yet it is rarely considered in modern analysis.3 Partly,
this may be because many texts are transmitted through cartularies, but
where original charters have been discussed as physical objects, this has
tended to be framed by a ‘strong thesis’ positing a fundamental opposi-
tion between orality and literacy; the material presence of writing can be
relegated to the status of an ‘oral’ hangover redolent of ‘archaic’ men-
talities and so somehow alien and embarrassing to the progress of liter-
acy; yet, the fact that writing is simultaneously both a physically present
mnemomic and a means of communication, with the capacity to tran-
scend the immediacies of time and place, is fundamental to its practi-
cal use in most pre-modern and indeed many modern social contexts.4
Although there is a rich stream of recent scholarship on charter collec-
tions and their commemorative and legal functions, there has been next to
no sustained study of the kinds of abbreviations, annotations and physical
traces of preparation, storage and retrieval that – as we saw in the case of
the St Gall documents discussed above – are crucial in reaching defini-
tive conclusions about document use.5 Moreover, by focusing atten-
tion on documents that can be reconstructed as ‘first-order’ historical
phenomena – that is, as self-standing texts – documentary scholarship
has diverted attention from a wider range of evidence that has been effec-
tively if inadvertently excluded from serious historical analysis: references
to documents – whether now lost or still surviving – embedded in surviv-
ing texts, for example, or evidence for the use of a surviving document
3 It is particularly striking that even in facsimile reproductions of documents, the accom-

panying discussion focuses on the questions privileged by the traditional concern of
diplomatic scholarship to establish an ‘authoritative’ text: see, e.g., D. Ganz and W. Gof-
fart, ‘Charters Earlier Than 800 from French Collections’, Speculum, 65 (1990), 906–32.
G. Koziol, ‘A Father, His Son, Memory, and Hope: The Joint Diploma of Lothar and
Louis V (Pentecost Monday, 979) and the Limits of Performativity’, in J. Martschukat
and S. Patzold (eds.), Ritual, Inszenierung und Performanz vom Mittlelater bis zur Neuzeit
(Cologne, 2003), 83–103, takes materiality seriously.
4 For a range of approaches to charters as evidence for literacy, see K. Heidecker (ed.),
Charters and the Use of the Written Word in Medieval Society (Turnhout, 2000).
5 See below, nn. 11, 87–8, for studies of charter collections and their functions.
286 Matthew Innes
after its initial production in the form of later additions or annotations.6

In other words, the foundational agenda of modern historical scholarship
has assumed a specific epistemology that has limited our engagement with
medieval documents. While we respect the achievements of centuries of
documentary scholarship, the need now is to treat documents as objects
that, although they may now look like passive records of past social struc-
tures, were once material actors within the societies they purported to
define. This will allow us not only to reframe some of the great set-piece
questions of medieval scholarship but also to expand dramatically the
evidential base from which they can be answered.
The second set of wider issues addressed by this analysis of the Cluny
charters is historiographical. In recent scholarship, debate about the pace
and nature of social transformation in post-Carolingian West and East
Francia has been helpfully linked to a series of questions about archival
and documentary transformations in the tenth and eleventh centuries.
This connection between historiographies of social development and
documentary change has achieved a high profile largely because of
the revisionism championed by Dominique Barthélemy.7 Questioning
a dominant methodological presupposition of post-war Francophone
medieval scholarship – the post-Carolingian ‘feudal revolution’ as first
essayed by Marc Bloch and subsequently elaborated by Georges Duby’s
study of the Mâconnais and a host of other regional thèses – Barthélemy
drew attention to the ways in which the texture of the documentary
evidence changed in the tenth and eleventh centuries.8 Documents –
and in particular notices recording the progress of disputes – grew less
laconic and tightly constrained by expectations about documentary
form, and more likely to include the anecdotal and incidental details
necessary to sustain an episodic narrative. So Barthélemy pointed out
that they inevitably granted modern historians a fuller understanding
of social bonds, in all their complexity and messiness, than did earlier,
more formalized, documents. There were thus dangers in assuming that
the kind of motivations and obligations made apparent by this new doc-
umentary narrativité directly reflected new forms of social relationship;
rather, they might simply attest changing scribal conventions. From a
complementary angle, other scholars have followed Patrick Geary’s lead
6 A ‘forensic’ as opposed to editorial approach has important epistemological and histori-

ographical implications for traditional positivist methodologies, an issue which I hope to
discuss elsewhere.
7 D. Barthélemy, La société dans le comté de Vendôme: De l’an mil au XIVe siècle (Paris, 1993).
8 M. Bloch, Feudal Society, trans. L. Manyon (London, 1961 [1937]); G. Duby, La société
au XIe et XIIe siècles dans la région mâconnaise (Paris, 1953). Of course, Bloch’s agenda
was not the same as Duby’s, but it was via Duby’s example that annaliste regional history
became dominant. See T. N. Bisson, ‘La terre et les hommes: A Programme Fulfilled?’,
French History, 14 (2000), 322–45.
in emphasizing the agency and interests of ecclesiastical institutions, and

in particular newly founded and reformed monasteries, Cluny foremost
among them. For Geary, the establishment of these institutions as
the primary repositories of social memory wrought significant cultural
changes, not least in the ways aristocratic families remembered their
pasts and so represented their presents: the picture of social transforma-
tion confidently identified by Duby half a century earlier might therefore
be more apparent than real, first and foremost a result of changes in
the realm of cultural representation.9 Close examination of the narrative
strategies of specific dispute notices, indeed, showed how these were not
disinterested legal records but rather carefully constructed accounts that
reflected the self-representation of ecclesiastical institutions as the stable
central points of this society, and therefore presented aristocratic actions
as wilfully and spontaneously violent, when the reality might have been
rather more complex.10
Both Barthélemy and Geary drew on an older tradition of diplomatic
scholarship that posed significant questions about the nature of the
documentary record for post-Carolingian West Francia. A long tradition
of studies of cartularies – compilations into which monastic institutions
began to copy an edited and reordered version of their documentary hold-
ings from the eleventh century onwards – reaches back through the cen-
tral institutions of modern French medieval scholarship and the builders
of national archives and collections in the nineteenth century to the anti-
quarians of the pre-Revolutionary era. Cartulary scholarship has increas-
ingly demonstrated the danger of assuming that cartulary compilers of
the eleventh century and later sought to present a direct and undistorted
view of the contents of the archives from which they worked. They evi-
dently made active choices about which documents to include and which
to exclude, as well as improving texts and imposing their own editorial
and organizational schemes, motivated by the interlocking imperatives of
commemoration and the establishment of title.11 As a result, students of
post-Carolingian documentation have become aware of the ways in which
cartularies effectively ‘froze’ a potentially more fluid and ambivalent
archival situation, fixing a particular image of the documentary practices
9 Geary, Phantoms, which in part built on the cultural turn in Duby’s own work. For
subsequent debate, see, e.g., L. Morelle, ‘Histoire et archives vers l’an mil: Une nouvelle
“mutation”?’, Histoire et archives, 3 (1998), 119–41.
10 P. Geary, ‘Monastic Memory and the Mutation of the Year Thousand’, in B. H. Rosen-
wein and S. Farmer (eds.), Monks and Nuns, Saints and Outcasts: Essays in Honor of
Lester K. Little (Ithaca, NY, 2000), 19–36.
11 Les cartulaires; A. Kosto and A. Winroth (eds.), Charters, Cartularies and Archives: The
Preservation and Transmission of Documents in the Medieval West (Toronto, 2002); ‘Pra-
tiques de l’écrit documentaire au XIe siècle’, published as Bibliothèque de l’École des
chartes, 155 (1997), 5–349.
288 Matthew Innes
of the earlier period. The stark judgements of previous generations, as

passed down through handbooks on diplomatic, thus need wholesale
revision.12 Cartularies characteristically include only the selected high-
lights of available Merovingian and Carolingian documentation, solemn
acts recording the merciful munificence of a charismatic and now
legendary royal authority or commemorating the reckless piety of
founders and celebrated benefactors. The comparison between the
stately formality of such material and the more scrappy and often
formless notices in which contemporary dealings were done need not,
therefore, be indicative of a decline in documentary standards: where
later cartularists chose to include more workaday material from the
earlier material it is similarly pragmatic and eclectic.13
Documentary practices before the monastic archive:

the Cluny evidence
These interlocking questions frame our discussion of the Cluny archive.
Cluny’s is probably the best-known charter collection of the entire
medieval period, one that has provided the empirical underpinnings
for a series of paradigm-setting works on the social history of the post-
Carolingian centuries.14 The monumental six-volume edition of the
surviving material prior to the thirteenth century, published by Bernard
and Bruel over the last quarter of the nineteenth century, contains over
5,500 documents15 ; this has been supplemented by a range of subsequent
studies correcting details (for example, of chronology)16 ; by fundamental
12 O. Guyotjeannin, ‘Penuria scriptorum: Le mythe de l’anarchie documentaire dans la

France du Nord (Xe –première moitié du XIe siècle)’, Bibliothèque de l’École des chartes,
155 (1997), 11–44.
13 D. Barthélemy, ‘Une crise de l’écrit? Observations sur des actes de St-Aubin d’Angers
(XIe siècle)’, Bibliothèque de l’École des chartes, 155 (1997), 95–117.
14 Most obviously, Duby, La societé, and Rosenwein, Saint Peter.
15 Cluny, and for their sources, methods and subsequent scholarship, see S. Bar-
ret, ‘Cluny, note sur le Recueil des chartes de l’abbaye de Cluny d’Auguste Bernard
et Alexandre Bruel’, Bulletin du Centre d’études médiévales d’Auxerre, 13 (2009):
cem.revues.org/index11017.html. An indispensable research tool, comprising the
searchable texts of the Cluny charters hyperlinked to a scanned Cluny, is the Car-
tae Cluniacensis Electronicae database hosted by the University of Münster: www.
uni-muenster.de/Fruehmittelalter/Projekte/Cluny/CCE
16 Most importantly, M. Chaume (and J. Marilier), ‘Observations sur la chronologie des
chartes de l’abbaye de Cluny’, Revue Mabillon, 16 (1926), 44–8; 29 (1939), 81–9,
133–42; 31 (1941), 14–19, 42–5, 69–82; 32 (1942), 15–20, 133–6; 38 (1948), 1–6;
39 (1949), 41–3; 42 (1952), 1–4; and unpublished work by Maria Hillenbrandt. I am
most grateful to the kindness of Barbara Rosenwein for sharing with me her collation of
various corrections to Bernard and Bruel’s datings suggested in subsequent scholarship:
I have tried to integrate these into any quantifications in what follows. Obviously, any
errors and anomalies that remain are my responsibility, but in any case the numbers
provided in the discussion that follows are used indicatively, to demonstrate trends, and
are not subjected to statistical analysis.
studies of the processes of archive formation and cartulary compilation

by Atsma and Vezin, Barret and Iogna-Prat17 ; and by a series of studies
on the palaeography of the surviving original charters, accompanied by
their publication in facsimile by Atsma, Barret and Vezin,18 while the
electronically literate can enjoy a wealth of online resources hosted by
the Institut für Frühmittelalterforschung at the University of Münster.19
It might be thought that this seam of documentation is so thoroughly
worked as to be historiographically exhausted. However, we have the
texts of 106 charters that predate Cluny’s foundation in 909 or 910.20
These pre-foundation documents do not consist of the familiar fare
of donations to the Church, but instead are almost without exception
secular exchanges of property. Six of these documents are royal diplo-
mas, of which five give royal possessions (sometimes estates, sometimes
churches) to various recipients, while the sixth confirms the swap of
the Cluny estate between the bishop and count of Mâcon: the kind of
material we commonly find in ecclesiastical cartularies of both Carolin-
gian and post-Carolingian vintage. But the overwhelming majority of
these documents are not royal, nor are they types of material routinely
found elsewhere. Land sales involving laypeople account for well over
half. Gifts from one layperson to another – evenly split between those
where a kinship tie between vendor and buyer is explicitly recorded, and
those between apparently unrelated parties – account for over a quar-
ter, and include a significant number of transfers of dowry lands and
morning-gifts from husband to wife (the vagaries of documentary form
17 H. Atsma and J. Vezin, ‘Gestion de la mémoire à l’époque de saint Hugues (1049–

1109): La génèse paléographique et codicologique du plus ancien cartulaire de l’abbaye
de Cluny’, Histoire et archives, 7 (2000), 5–29; S. Barret, La mémoire et l’écrit: L’abbaye
de Cluny et ses archives, Xe –XVIIIe siècles (Münster, 2004); D. Iogna-Prat, ‘La confection
des cartulaires et la historiographie à Cluny (XIe –XIIe siècles)’, in Les cartulaires, 27–44.
18 Les plus anciens documents originaux de l’abbaye de Cluny, ed. H. Atsma, J. Vezin and S.
Barret, 3 vols. to date (Paris, 1997–).
19 www.uni-muenster.de/Fruehmittelalter/Projekte/Cluny/. Other invaluable online resou-
rces are also being developed: see, e.g., the Chartae Burgundiae Medii Aevii project
hosted by the Centre d’études médiévales at the University of Auxerre: cem.revues.org/
index2372.html.
20 I have counted the documents chronologically prior to the foundation charter’s stated
date of 910, and excluded one outlying gift to Cluny from 909, which is problematic.
For the foundation, see now D. Iogna-Prat, ‘Cluny, 909–910, ou l’instrumentalisation
de la mémoire des origines’, Revue Mabillon, 11 (2000), 161–85. As explained above,
n. 16, I have also endeavoured to take account of modern scholarship revising Bernard
and Bruel’s chronology. I have also excluded material transmitted via other institutions
(e.g. the cartulary of the cathedral of St-Vincent at Mâcon), some of which is intimately
intertwined with Cluny and was included by Bernard and Bruel. I thus count the
following charters: Cluny 3, 5, 7–16, 18–28, 30–77, 79–111 (with 89bis), 219, 1095;
no. 90, a single document recording both a sale and a placitum, is counted twice.
Classification inevitably involves some degree of interpretation, but I have tried to
follow the implicit typologies in scribal language and formulas as closely as possible.
290 Matthew Innes
making it difficult to make a hard and fast division between gifts and mar-
riage agreements). In addition we have four records of dispute settlement
(three heard before counts and one before imperial missi), two swaps of
land, two swaps of unfree dependants and one poignant document in
which an individual places himself in the servitude of another. Where
churches do appear, they do so as the possessions of individual landown-
ers, with the exception of two precarial grants made by bishops, another
episcopal document confirming a grant to an abbot, one donation by a
priest to his church and one transaction centred on a chapel. Hence, while
there is the odd clerical actor or witness and a mixture of clerical and
non-clerical scribes, the Church as an institution is far from hegemonic
in determining these social practices or their documentary record.
This field of secular transactions and the documentary culture that it
witnesses, moreover, continues to loom large among the surviving char-
ters even after Cluny’s foundation, through much of the tenth century.
From the abbacy of the first abbot, Berno (910–27), almost 40 per cent
of the transmitted documents – over sixty charters – are such secular
transactions in which Cluny itself was not directly implicated, and under
his successor Odo (927–42), in spite of a much thicker transmission of
documents in which Cluny was the recipient as gift-giving took off, secu-
lar transactions in which neither the abbey nor the monks were involved
still make up around 20 per cent of surviving documents – another haul
of over fifty charters. In fact, through to at least the abbacy of Odilo (994–
1049), we have a fairly steady rate of transmission of secular documents,
with an average of three or four such charters surviving per year. The
take-off of gifts to Cluny means that this constitutes a decreasing propor-
tion of an increasing volume of surviving material, but in sum it gives us a
sample of several hundred transactions in which local landowners bought
and sold land from each other, contracted dowries, took each other to
court and bought out each other’s claims. Cluny’s foundation does not
mark a break in the nature, preservation or, apparently, the transmission
of this material.21
These documents provide a priceless window onto the workings of a
provincial society prior to the advent of a major monastery or aristocratic
network, with all the changes that integration into a wider system of
patronage brought. In fact, this rich seam of charters offers the possibil-
ity of complementing and testing the series of other regional studies that
21 Here it has simply been used as a clean, convenient and unambiguous cut-off, prior to
which a period of non-monastic preservation is a given and needs no further demon-
stration. Future research, indeed, might usefully separate tenth-century material that
had an independent transmission prior to coming to the Cluny archive from that whose
transmission was wholly monastic.
have had to work from the perspective of networks created around great
monasteries, and adopt an indirect approach to reconstructing society
before the advent of imperial aristocrats and the imperial Church; these
documents directly record social phenomena that are normally inferred
rather than demonstrated.22 Given the formative influence on the his-
toriography of post-Carolingian change of works whose empirical basis
lay in the Mâconnais, it is striking that these late Carolingian documents
have never been properly used to analyse social process on the ground;
they have been oddly neglected by modern scholars.23
The preservation of lay documents and the history

of the Cluny archive
The vast majority of these pre-910 charters must therefore be categorized
as ‘lay documents’ by any meaningful definition. With respect to archival
and documentary practices, they therefore raise important issues.24 The
date of this material means that it must have been not only produced
but also preserved, independently of Cluny, for a number of years; as we
shall see, on occasion we can surmise how a document may have come
to Cluny, but even for the mass of documents where this is not possible
the chronological profile of the surviving sample indicates that it was
possible for a secular document to survive without the archival agency of
the Church for several decades.
The sheer volume of lay documentation transmitted via Cluny, and its
presence so far into the abbey’s history, is a brute fact that raises impor-
tant questions of provenance and transmission. How and why did it come
to be preserved? The keys to answering this question lie in the history
of the Cluny archive. We are lucky that that history has recently been
uncovered by Sébastien Barret, and that Barret’s work itself was possible
thanks to the work of the eighteenth-century antiquarian Louis-Henri
Lambert de Barive.25 Lambert worked in the Cluny archive in a series of
visits over the course of the two decades before the French Revolution;
he provided a priceless description of the thirteenth-century ‘tower of the
22 E.g. W. Brown, Unjust Seizure: Conflict and Authority in an Early Medieval Society (Ithaca,
NY, 2001); Costambeys, Power; Hummer, Politics; Innes, State and Society.
23 The Carolingian backdrop to Duby’s Mâconnais was based on normative sources, and
then interpreted as evidence for state institutions. For a powerful recent critique on both
empirical and theoretical grounds, see S. D. White, ‘Tenth-Century Courts at Mâcon
and the Perils of Structuralist History’, in W. Brown and P. Górecki (eds.), Conflict in
Medieval Europe: Changing Perspectives on Society and Culture (Aldershot, 2003), 37–68.
24 They have already been noted by several commentators, from Duby, La societé, xii, to
Kosto, ‘Laymen’, 72, but I know of no extended discussion.
25 Barret, La mémoire et l’écrit.
292 Matthew Innes
archives’ on one side of the gatehouse, as well as many hundreds of tran-

scripts of original documents he found within that tower.26 Lambert’s
antiquarianism is clear from the care with which he made not merely
transcriptions but almost mini-facsimiles, which attempted to reproduce
aspects of the physical appearance and script of the documents.27
This antiquarianism, however, had very clear connections with wider
currents in ancien régime cultural and political life. Lambert’s activity
at Cluny in the 1770s and 1780s coincided with a series of other ini-
tiatives in which archives were used for an immediate purpose. The
records of the Cluniac order at the mother house, for example, had been
extensively drawn on earlier in the century by scholars at St-Martin-des-
Champs eager to assemble proofs for the monastery’s rigorous obser-
vance and to defend the order against the laxness encouraged by the
royal appointees who served as commendatory abbots of Cluny through
this period; indeed, the vocation became a particularly live issue after
the General Assembly of the Gallican Clergy in 1765, and the setting
up of the Commission des Réguliers, which investigated the monastic
orders right across the kingdom, reporting on standards of observance
and winding up some houses.28 Meanwhile, from 1774, a team of other
antiquarians, like Lambert professional avocats from Autun, was hired
by the local bishop and president of the Estates of Burgundy, Yves Mar-
beuf, to organize the records of the estates – this in the context of the
continuing financial demands placed on the estates by the Crown, those
same demands that made the fiscal privileges of the clergy as a whole, and
the monastic orders in particular, such an issue of debate.29 Lambert’s
own activity, however, was funded by the Cabinet des Chartes, founded
in 1765 by the historiographer royal Jacob-Nicolas Moreau to create
a national archival depository to underpin an unashamedly antiquar-
ian and royalist history of France rich in useful precedent for his master.
26 S. Barret, ‘La Tour des Archives de l’abbaye de Cluny’, Livraisons de l’histoire de

l’architecture, 10 (2005), 9–17.
27 On Lambert’s antiquarianism, see above all S. Barret, ‘Un avocat au service du Cabinet
des chartes: Les travaux de Louis-Henri Lambert de Barive dans les archives de Cluny
(v. 1770–v. 1790)’, Histoire et archives, 15 (2004), 29–64.
28 See J. McManners, Church and Society in Eighteenth-Century France. I. The Clerical
Establishment and Its Social Ramifications (Oxford, 1998), 141–73 (for 1765 and the
issue of clerical taxation), 571–614 (for the commission and the monastic orders on
the eve of Revolution); on Cluny at this date, see G. Chevron, ‘L’abbaye et l’ordre de
Cluny à la fin du XVIIIe siècle’, Revue Mabillon, 39 (1949), 44–58; 40 (1950), 1–28;
and on the impetus for the history of the order, G. Charvin, ‘À propos d’un projet d’une
histoire de l’ordre de Cluny au XVIIIe siècle’, in À Cluny: Congrès scientifique (Dijon,
1950), 181–5. A revised constitution for the Cluniac order was finally approved in 1789.
29 See J. Swann, Provincial Power and Absolute Monarchy: The Estates General of Burgundy,
1661–1790 (Cambridge, 2004).
Moreau’s plan drew on the extensive labours undertaken by the Maurists

over the previous century in collecting materials for unashamedly patri-
otic provincial and national histories, and built on the Maurist endeavour
to institute a network of paid antiquarian helpers, each charged with ran-
sacking the archives of a given province; Lambert’s activity at Cluny
needs placing in that context.30
The cultural, fiscal and political issues that framed Lambert’s anti-
quarianism, in any case, fed through the period of the Revolution and
the secularization of Cluny’s estate, the dispersal of the archive and the
destruction of the buildings.31 It is normal to attribute the destruction of
the archive to an episode of plundering and burning in 1793, and thus
to invoke that random actor ‘chance’ in attributing survivals. In fact, the
subsequent fate and shape of the archive suggests that as the abbey and
its estates were dismembered, documents were falling into lay hands by
various mechanisms. Historians who invoke ‘chance’, as so often, are
actually adopting a shorthand for an unknown or uninvestigated process:
mathematical and philosophical treatments of probability and truth may
offer a better way to think about these problems.32
The Revolution and its aftermath, indeed, saw continued interest in
the type of project masterminded by Moreau, alongside concern and
controversy regarding control over initiatives that could not but be politi-
cally loaded, with the fate of royal and governmental archives and record
collections the subject of vigorous debate. With the establishment of sta-
bility of a sort, however, it became increasingly clear that the concerns
about precedent that had made archival control such a loaded issue in
the unravelling of the ancien régime were no longer so pressing. Instead, a
less immediately partisan political concern for the fate of documents and
records disturbed in the Revolutionary and post-Revolutionary periods,
and the need to preserve the monuments of the French past, emerged,
evident in the foundation of the École des Chartes to create a cen-
trally trained cadre of archivists who could classify and publish the
30 K. Baker, ‘Controlling French History: The Ideological Arsenal of Jacob-Nicolas

Moreau’, in Baker, Inventing the French Revolution: Essays on French Political Culture
in the Eighteenth Century (Cambridge, 1990), 59–85. Maurists: McManners, Church
and Society, 594–600; Knowles, Great Historical Enterprises, 33–62; M. Lecomte, ‘Les
Bénédictins et l’histoire des provinces aux XVIIe et XVIIIe siècles’, Revue Mabillon, 17
(1927), 237–46; 18 (1928), 39–58, 110–33, 302–23 (important for any early medieval-
ist using Bouquet’s Recueil des historiens des Gaules et de la France or Devic and Vaisette’s
Histoire générale de Languedoc).
31 B. Marguery-Melin, La destruction de l’abbaye de Cluny, 1789–1823 (Cluny, 1985).
32 The best discussion of these issues, in the context of charter transmission in Lucca, is
A. Esch, ‘Uberlieferungs-Chance und Überlieferungs-Zufall als methodisches Problem
des Historikers’, Historische Zeitschrift, 240 (1985), 529–70.
294 Matthew Innes
building blocks of the nation’s history. Archives and archivalism were

thus ‘medievalized’: it is striking that the various transcriptions and col-
lections of documents made in the post-Revolutionary period contain
only early and high medieval pieces. As the nation’s historical imagi-
nation was fuelled by romantic essayists flirting with a glamorous and
distant past, a gap opened between this endeavour and current contro-
versy. At the same time, as the emerging French nation state devoted
more and more energy to the cultivation of official records enabling it
to rule its subjects, the ordering of the medieval past through public
archives became a legitimating tool that expressed a grand narrative of
French national identity rooted in the state.
The fate of cartularies and related documents thus became a major
concern of archivists in Paris and beyond through the middle and late
nineteenth century.33 As regards Cluny in particular, it was the tireless
efforts of Leopold Delisle at the Bibliothèque Nationale in the first half
of the nineteenth century that led to so much of the material relating
to the archive being reassembled in Paris. Delisle was able to purchase
several sizeable collections of original charters from private individuals
in Burgundy. These were to complement the abbey’s surviving cartu-
laries, which had been brought to Paris in the Revolutionary period,
and Lambert’s papers, which like most of the other material concerned
with Moreau’s endeavours had been seized to underpin the new national
archives.34 Bernard and Bruel’s edition was really the culmination of
Delisle’s work in pulling back together as much material as could be
identified; they also built heavily upon Lambert’s work as they realized
that his transcriptions were often closer to the original documents than
the cartulary copies.35
The archival history of the Cluny charters thus in many ways exem-
plifies the transformation of archival cultures as the ancien régime was
transformed into the nation state of classical modernity.36 Similarly, the
emergence of an archival culture less dependent on exigencies of access
and place is clear in the late twentieth and early twenty-first centuries,
with the growth of an international scholarly community devoted to the
study of the records of Cluny, at first through conferences and papers
but more recently through shared tools disseminated electronically on
33 See I. Vérité, ‘Les entreprises françaises de recensement des cartulaires (XVIIIe –XXe
siècles)’, in Les cartulaires, 178–213.
34 See L. Delisle, Inventaire des manuscrits de la Bibliothèque nationale, fonds de Cluny (Paris,
1884).
35 J. Richard, ‘La publication des chartes de Cluny’, in À Cluny, 155–60.
36 Cf. above, 13.
the internet.37 More immediately important for our purposes, however,

is the fact that the disjuncture of the crucial Revolutionary and post-
Revolutionary decades meant that the surviving material, as published
by Bernard and Bruel, rests on three distinct vectors of transmission
through the period of the abbey’s suppression: via the abbey’s cartu-
laries, via those originals taken from Cluny in the Revolution but then
sought out and bought up by Delisle, and via Lambert’s papers.
Each of these distinct vectors of transmission privileges different kinds
of material, creating a different image of the abbey’s early medieval
archive. The process of cartulary compilation, for example, had its origins
in the later part of the abbacy of Odilo (994–1049) but reached fruition
under Abbot Hugh (1049–1109), with the compilation of three inde-
pendent but complementary collections, known to modern historians as
cartularies A, B and C, but in fact best conceived as a single venture
comprising three physically separate but conceptually interdependent
parts.38 The cartularies need to be contextualized alongside a range of
other initiatives to provide written memorials of the abbey in its heyday:
custumals and hagiographies designed to fix institutional memory and to
preserve institutional identity drew on similar impulses to those evident
in the cartularies.39 The A cartulary organizes the abbey’s documents
in the form of a gesta abbatum of the first four abbots, beginning with a
chronologia of the abbots and a dossier of foundational documents, and
then evoking the history of the church and the memory of its benefac-
tors, before giving the texts of the donations, purchases and privileges for
Berno, Odo, Aymard and Maiolus in turn, with texts often embellished
or improved for style, and context – including in many cases the date –
elided. The B cartulary completes the process for the abbots under whom
the cartularies were compiled, Hugh and his predecessor Odilo, and in so
doing gives some clues as to the various stages of cartulary organization.
The C cartulary, published on the occasion of the visit to Cluny of Pope
Urban II – a former Cluny monk – in 1095–6, complements the two
volumes of gesta abbatum in the A and B cartularies. It begins with the
same evocation of the church and donors used in the A cartulary, but
then organizes documents by a different principle: foundational material
37 These developments differ from the ‘postmodern’ ‘democratization of the archive’ dis-
cussed by Derrida and other theorists; above, 13.
38 H. Atsma and J. Vezin, ‘Gestion de la mémoire à l’époque de saint Hugues (1049–
1109): La génèse paléographique et codicologique du plus ancien cartulaire de l’abbaye
de Cluny’, Histoire et archives, 7 (2000), 5–29, with full references to earlier bibiliography.
39 Iogna-Prat, ‘La confection des cartulaires’; D. Iogna-Prat, ‘La geste des origines dans
l’historiographie clunisienne des XIe –XIIe siècles’, Revue bénédictine, 102 (1992), 135–
91.
296 Matthew Innes
is followed by papal and then royal diplomata, and then private charters
grouped according to a geographical rationale. The momentous visit of
the abbey’s most illustrious ‘old boy’ was marked not only by the launch
of a new kind of Christian warfare aimed at recapturing the Holy Places,
but also by the consecration of the altars in the abbey’s magnificent new
church, and Urban’s declaration of a ‘sacred ban’ – a zone of protected
space – around this particular Holy Place. The publication of the C car-
tulary therefore both marked the culmination of decades of work aimed
at establishing a definitive institutional memory and underpinned a pro-
gramme of territorial sacralization.40
Comparison of the cartularized documents with surviving originals
shows quite clearly that the process of cartulary compilation effected a
series of changes to the charters included. Their texts were ‘improved’ by
abridgement, correction and embellishment to fit the linguistic and liter-
ary expectations of this grand memorial project.41 Inevitably, the act of
recopying removed cues and markings that might guide the interpretation
or aid the use of an original document. As we shall see below, such cues
might be textual – dorsal notes summarizing a document, or annotations
recording subsequent confirmations or contestations of a transaction – or
non-textual – the fixing together of one document to another, earlier doc-
ument relating to the same people or place, or the attachment of a legal
staff or festuca whose exchange had symbolized the handover of prop-
erty. Above all, cartularization involved the reorganization of the docu-
ments in terms of categories and classifications determined by the abbey’s
institutional needs: the commemoration of abbots and their patrons in
the A and B cartularies, the geography of the estate and the need to
preserve the abbey’s liberties and privileges over that estate in the C
cartulary.
The rich secular documentation of the late ninth and tenth centuries
had no utility to the cartulary project, and is therefore almost wholly
excluded from the cartulary tradition. Documents involving the circu-
lation of estates within the family of the founder, Count William, were
copied as necessary prologues to the foundational history of the abbey,
but even these were not included uncritically or en masse, as the cartu-
larists’ exclusion of an exchange of unfree dependants between William
40 E.g. B. H. Rosenwein, ‘Visualising a Dispute Resolution: Peter of Albano’s Protected

Zone’, in Brown and Górecki (eds.), Conflict in Medieval Europe, 85–108.
41 A. Bruel, ‘Note sur le transcription des actes privés dans les cartulaires antérieurement
au XIIe siècle’, Bibliothèque de l’École des chartes, 6 (1875), 445–56, and, more broadly,
Geary, Phantoms, 103–7; also now H. Atsma and J. Vezin, ‘Originaux et copies: La
reproduction des éléments graphiques des actes des Xe et XIe siècles dans le cartulaire
de Cluny’, in Kosto and Winroth (eds.), Charters, Cartularies and Archives, 113–26.
and his sister Ava copied by Lambert shows.42 The absorption of the
family documents of aristocratic founders into monastic archives is a
phenomenon widely paralleled, and it suggests a continuing give and
take between founders and the institution which defined their family
identity and memory and so was an obvious place to store family deeds;
the phenomenon is well attested precisely because cartulary compilers
right across the Carolingian and post-Carolingian world were anxious to
commemorate pious founders, and indeed characteristically where secu-
lar documentation is copied into cartularies, it seems to consist of ancient
muniments concerned with the founder’s family.43 Other than that, where
Cluny received once royal estates or churches that had previously been
granted out to favoured aristocrats, the royal documents recording those
prior gifts – presumably stored at the estates or churches in question
and passed over with them, often long after the abbey’s foundation –
were similarly favoured by cartulary compilers; again, this phenomenon
is widely paralleled in cartularies right across the Frankish world.44 But
while those compiling these collections were interested in documents of
charismatic antiquity that might commemorate near-legendary founders
or add regalian lustre to their codices, they had no interest in records of
sales or dowries between private individuals whose relationship to the
abbey and its patrimony must by then, in most cases, have been long
forgotten.
Had we the cartularies alone, any attempt to understand document
use in ninth- and tenth-century Burgundy would run up against familiar
evidentiary problems. We would be confident in demonstrating that the
greatest aristocratic families – those of the abbey’s founders, or the aris-
tocratic fideles who received royal gifts – preserved documentary proofs
relating to some of their patrimonies, particularly when dealing with the
Church or with kings. We would have some tiny scraps of evidence for
written documentation at a lower social level and end up asking whether
this was the tip of the iceberg or the exception that proves the rule. The
existence of a vigorous tradition of secular documentation by lesser local
landowners in their dealings with each other would remain an unknow-
able unknown, as would the preservation of a significant volume of secular
material within Cluny’s archives. In and of itself, this serves as a sobering
illustration of the dangers of taking a cartulary as a snapshot of a monas-
tic archive at the point of compilation, rather than as a carefully crafted
42 Cluny 74; 89bis, a charter of William for one of his fideles, which was later passed with
the property it concerned to Cluny, was similarly excluded; see below, n. 53.
43 As has been underlined by Jonathan Jarrett in an as yet unpublished work, which I am
grateful to him for discussing with me; see also above, Chapters 8–10.
44 Though not all such documents were included in cartularies.
298 Matthew Innes
and consciously organized attempt to use archival materials for present

purposes, to shape institutional memory.
The transmission of a significant number of originals, most retrieved or
bought up by Delisle, has made clear the extent of reworking involved in
copying original documents into cartularized form and proved invaluable
in illuminating scribal practice and the processes of charter production.
The surviving originals from Cluny provide further indirect evidence for
the continuity of a broadly based documentary culture through the tenth
century in this region. Palaeographical analysis has shown that charters
subscribed by a single scribe are not necessarily all in the same hand,
and suggests that the actual scribal subscription named the individual
under whose authority a document was written, not the individual who
physically wrote it. In other words, even as Cluny’s monks were becoming
important providers of literate services for the local community, they did
so in the context of a complex scribal culture, a culture that rested on
the notion that scribal office involved public trust, with the named scribe
vouching for documents written up under his authority.45
Does the pattern of surviving originals hint at the scale of exclusion
involved in cartulary compilation, or serve as a meaningful control on
the cartularies?46 Very few of our secular documents are preserved in
original, rather than in antiquarian, copies: of the 106 pre-foundation
documents, just thirteen. Of these thirteen originals, three were royal
documents concerning property that later came to the abbey, including
Louis the Pious’ confirmation of the bishop of Mâcon’s swapping of the
villa of Cluny with Count Warin in 825, whence the estate came into
the hands of Count William’s family; a further charter records the gift
of Cluny to William by his sister, Ava, abbess of St-Pierre-les-Nonnains.
Of the remaining nine originals, no fewer than seven survive thanks to
a remarkable, enigmatic and anonymous collection of more than fifty
45 H. Atsma and J. Vezin, ‘Autour des actes privés du chartier de Cluny (Xe –XIVe siècles)’,
Bibliothèque de l’École des chartes, 155 (1997), 45–60; Atsma and Vezin, ‘Les responsables
de la transcription des actes juridiques et les services de l’écriture au Xe siècle: L’exemple
de Cluny’, in M.-C. Hubert, E. Poulle and M. Smith (eds.), Le statut du scripteur au moyen
âge (Paris 2000), 9–20; B.-M. Tock, Scribes, souscripteurs et témoins dans les actes privés
en France (VIIe –début XIIe siècle) (Turnhout, 2005); Tock, ‘Les actes entre particuliers
en Bourgogne méridionale (IXe –XIe siècles)’, in P. Erhart, K. Heidecker and B. Zeller
(eds.), Die Privaturkunden der Karolingerzeit (Zurich, 2009), 121–34.
46 In analysing the distribution of originals, I have made use of the ARTEM database
of original documents prior to 1121 in modern France: see B.-M. Tock et al., La
diplomatique française du haut moyen âge: Inventaire des chartes originales antérieures à 1121
conservées en France, 2 vols (Turnhout, 2001). The numbers in the following paragraph
are based on the searchable online database at www.cn-telma.fr/originaux/index/, used
in conjunction with the Münster Cartae Cluniacenses Electronicae database (above, n.
19), and with chronological adjustments made as outlined above, n. 16.
original charters made after Cluny’s secularization and passed on to the

Bibliothèque Nationale, where it still resides; the two remaining docu-
ments are the record of a judicial hearing at which Count Eblo of Poitiers
found in favour of the abbey of Nouaillé, and a gift by Count William
to one of his retainers, which was passed to Cluny with the estates it
concerns.47
With Cluny’s foundation, we have a far higher volume of surviving
originals, as a significant number of donations to the abbey survive in
the original as well as in cartulary copy: from the first four decades of
the abbey’s existence, in the period prior to 950, we have thirty-three
original documents of which Cluny is the beneficiary, and of which just
two are royal diplomata and the remainder ‘private’ donations covering
the range of the social spectrum from Count William to small-scale local
landowners. Through the tenth century, we continue to have a steady
stream of originals in which Cluny is not directly implicated, preserved
at a similar rate to that evident in the decades immediately before the
abbey’s foundation: from the period 910–50, fifteen originals in which
Cluny is not involved, of which three are royal documents and eight are
non-royal documents preserved in the same post-1789 collection that
dominates the pre-910 originals.48 Strikingly, the breakdown of trans-
mitted originals of around one-third ‘secular’ transactions to two-thirds
benefactions to Cluny is very similar to the overall ratio of all transmit-
ted documents, including cartulary copies and early-modern antiquarian
transcripts as well as originals. The similarity in the ratio of monastic to
secular documentation in both original and all other vectors of transmis-
sion must confirm that the abbey’s documentary traditions ‘piggybacked’
on and expanded an already rich documentary culture in the region. But,
just as the activities of Lambert were crucial in preserving traces of the
secular documents kept in the Cluny archive prior to its dispersal, so the
seam of original lay documents would have vanished were it not for
the antiquarian efforts evident in the charter collections now housed in
the Bibliothèque Nationale in Paris: without the single collection in BnF
lat. 17715, we would have just half a dozen non-royal secular documents
prior to 950, two prior to the abbey’s foundation and four from the
47 On the charter collection BnF lat. 17715, note the comments of Barret, ‘Cluny, note’,
n. 17: this dossier eloquently shows the role of chance in transmission. For the trans-
mission of Cluny 89bis, see below, n. 53; I am at a loss to explain the transmission of
the Poitiers placitum, no. 81.
48 The non-Cluny documents from 910–50 are ARTEM nos. 1582–3, 1594–5, 1597,
1599, 1815–19, 1822, 1827, 1831, 2504. Of these, 1582–3 and 1599 are royal; the non-
royal 1815–19, 1822, 1827 and 1831 are collected in BnF lat. 17715. Of the documents
directly benefiting Cluny, four (ARTEM nos. 1821, 1824–5, 1830) are from BnF lat.
17715.
300 Matthew Innes
period 910–50. It is only if we strip out the efforts of antiquarian copyists

like Lambert, and the collectors of original documents from depart-
mental archives and the black market whose collections now grace the
Bibliothèque Nationale, that the pattern of transmitted documents from
Cluny matches that more familiar from other ecclesiastical foundations.
At Cluny, however, we can move beyond an argument of implication
and suggestion and directly demonstrate a significant documentary cul-
ture among lay landowners great and small. This is, above all, thanks
to the work of Lambert and other eighteenth- and nineteenth-century
antiquarians; without them, there would be no real clue to the volume of
what Cluny once held, and the submerged part of this particular archival
iceberg would remain invisible to the modern eye.
Landowners and dossiers: documentary practices before

and within the Cluny charter collection
How did this rich seam of secular material come to reach Cluny, and in
what form was it ‘archived’ both before and after its acquisition by the
monastery? Where secular documents are transmitted via cartulary tra-
ditions, it is normally in the form of ‘dossiers’ of title-deeds passed over
to the monastery when it acquired an estate.49 The effort to trace the
contours of such groups of documents has been profoundly influenced
by the Cluny material, and in particular by André Déléage’s identifica-
tion of a dozen such dossiers relating to the prehistory of an individual’s
gift to Cluny.50 These dossiers normally contain between half a dozen
and a dozen charters, documenting an individual’s acquisition of prop-
erties later given to Cluny; they are not usually chronologically deep,
characteristically dealing only with the activities of one conjugal couple
rather than reaching back across the generations, and therefore poten-
tially covering transactions spread over several decades but not more.51
49 E.g. Kosto, ‘Laymen’, 72: ‘The phenomenon of the integration of “dossiers” into eccle-
siastical archives at the time of acquisition of a piece of land is perhaps best known
from the Cluny chartrier.’ The phenomenon has also been noted by Wendy Davies
for Brittany and northern Iberia (W. Davies, Small Worlds: The Village Community in
Early Medieval Brittany [London, 1989]; Davies, Acts of Giving), and Adam Kosto for
Catalonia (Kosto, ‘Laymen’, which also includes a discussion of wider comparanda at
71–3).
50 A. Déléage, La vie rurale en Bourgogne jusqu’au début du onzième siècle, 2 vols. (Macon,
1941), I, 234–40; Gratsianskii’s work, noted by Kosto, ‘Laymen’, 72n134, is beyond
my linguistic capacities.
51 Compare, on this chronological-generational structure, above, 198, for the East Frank-
ish evidence, and 269–81 for Iberia; on the Italian evidence and the possible influence
of the ‘thirty-year rule’ based on the post-Roman development of vulgar law, see above,
Chapter 9.
These parameters – dossiers of a relatively small number of linked docu-

ments typically reaching back three decades or so – are also suggested by
the distribution curve of secular documents in the Cluny archive, which
seems to mirror the pattern of growing gift-giving to the abbey, except
three decades earlier. In other words, as the monks acquired land, they
frequently acquired with it small collections of documents connected
with that land and reaching several decades into the past.
Of the twelve identified dossiers, most are tenth-century in origin, with
a couple spilling over into the eleventh; the earliest is that associated with
Lilia. Lilia’s dossier is larger than most of those identified, comprising up
to a dozen documents that were passed on to Cluny when in 926 Lilia
herself granted the abbey ‘all that I have acquired with my husband in
the county of Mâcon’, naming three particular villages (villae).52 Only
one of these documents is transmitted in the original, a charter from 905
recording the gift of three fields and a clearance to Hugh and Lilia by
Count William, who was to found Cluny just a few years later, and his
sister Ava. This seems to have been preserved in the original on account
of the lustre added by William’s name: a dorsal note styles it as a record of
the generosity of ‘our lord William’, suggesting that at some date after its
passing into the Cluny archive with Lilia’s other title-deeds, this charter
was misunderstood as a record of William’s munificence and so came
to be treated differently from its fellows.53 The only two documents
transmitted via the cartulary are two charters recording Lilia’s gifts to
Cluny, the first from 916 and the second from 925; the former of these
was also copied by Lambert, and the remaining nine texts survive thanks
to the antiquarian endeavours of Lambert in the last decades of the abbey
prior to its secularization in the Revolution.54
52 Cluny 275; Lilia had made a more limited, gift in 916 (no. 202) of holdings in a specific
area within a single villa, but it makes more sense to see the totality of her title-deeds
passing over with the totality of her estates following 926. Villa: F. Bange, ‘L’ager et la
villa: Structures du paysage et du peuplement dans la région mâconnaise à la fin du haut
moyen âge (IXe –XIe siècles)’, Annales ESC, 39 (1984), 529–69; and in Carolingian
sources generally, M. Costambeys, M. Innes and S. Maclean, The Carolingian World
(Cambridge, 2011), 229–41, with references.
53 Cluny 89bis (V, 833–4). It is easy to see how on a superficial reading, in the context of
the Cluny archive, this charter could have been interpreted as a pious gift to the abbey
rather than to Hugh and Lilia – it was, however, not copied in the cartulary, presumably
as its date preceded the foundation of the abbey, so it could not be linked to the gesta
of any of the abbots. Its separation from its fellows might be implied by the fact that it
was not used by Lambert.
54 Cluny 202, 275. The possibility must arise, given the transmission of the latter document
via the cartulary alone, that the 925 charter represents a post hoc working-up of an earlier
agreement negotiated at the time of the smaller 916 gift; the stress in no. 275 on William’s
relationship with Lilia might fit with this being the ‘official’ Cluny viewpoint.
302 Matthew Innes
These nine texts are all charters of sale in which Lilia and her husband
Hugh are named as joint purchasers; they reach back three decades, to
897, and record a cluster of purchases of closely related – often con-
tiguous – fields, meadows, vineyards and farmsteads over the period
897–908. Lilia and Hugh’s activities were focused in Bierre, northwest
of Cluny, and a handful of neighbouring settlements, some newly emerg-
ing and others longer established, but most notably Chateau, where the
couple also seem, appropriately, to have had a residence. They underline
the exceptional richness of this kind of material for the social historian.
Purchase prices are given in solidi, or, on three occasions, other movables
whose value is expressed in solidi and, once, libri; the prices vary from 1 to
50 solidi, with livestock being used instead of coin in the smaller transac-
tions and ‘silver and other goods’ in the largest. There is the unmistakable
bustle of sociability within this land market: these are local landowners
dealing with the exigencies of the family fortunes by buying and selling
smallholdings, and using documents to record those dealings, and Hugh
and Lilia bought from both kin and neighbours. Hugh and Lilia, indeed,
expended a large amount of movable wealth in both cash and kind over a
decade, and for well-connected locals like them land appears to have been
easily available for purchase – facts that sit uneasily with some received
views of early medieval landed society.55
As Hugh and Lilia bought up property, they seem to have prospered
and acquired a certain social cachet. Hugh in particular enjoyed impor-
tant links with the high echelons of the aristocracy: charter scribes felt
able to give Hugh the honorific title ‘magnificent lord’, and in the last
donation in the series – the largest, from 908, in which Hugh and Lilia
bought out another couple for 50 solidi’s worth of movables, acquiring the
only demesne farm in the series – Count William, Cluny’s founder, was
among the witnesses. The couple’s relationship with Count William was
such that in 905 he and his sister had given Hugh and Lilia three fields that
they owned in one quarter (fine) within the villa of Bierre, and a further
smallholding in a clearance nearby. This charter – the only lay gift in the
series and the only document surviving in the original – is unusual in the
series in that there is a note of conscious self-representation that is lacking
from the pragmatic records of Hugh and Lilia’s ventures in the land mar-
ket. Not only are both Hugh (‘our beloved faithful [man]’) and William
55 Cf., in different contexts: Davies, Acts of Giving; W. Davies, ‘Sale, Price and Valuation
in Castile and Galicia-Leon in the Tenth Century’, EME, 11 (2002), 149–74; L. Feller,
A. Gramain and A. Weber, La fortune de Karol: Marché de la terre et liens personnels dans
les Abruzzes au haut moyen âge (Paris, 2005).
(‘count, count of the palace56 and margrave’) given grand titulature, but
also William, addressing Hugh, publicly states his motivation in giving
‘on account of the love and benevolence we have towards yourselves’. As
Barbara Rosenwein has recently shown, such expressions of love – present
in around a fifth of the ninth- and tenth-century charters from Cluny –
were consciously used by scribes to express the social and emotional
ties that bound together the landowning class of the late Carolingian
Mâconnais.57 The dry pragmatism of Hugh and Lilia’s other documents
is thus dramatically juxtaposed with the more expressive note sounded
to articulate the affective but also hierarchical relationship between the
couple and their patron, who was simultaneously friend, benefactor and
representative of a charismatic authority charged with ensuring divinely
ordained regalian order. This charter indicates that the ideologies devel-
oped by Cluny’s monks in the course of the tenth and eleventh centuries
‘piggybacked’ on shared ideas and expressions already current in the
landowning communities out of which the abbey itself grew – and that
the economy of gift-giving in land that grew up so rapidly around the
tenth-century abbey did so precisely because gifts of land were already
a crucial mechanism for creating and prolonging social relationships of
patronage among the landowning classes.58 Indeed, the existing relation-
ship between William on the one hand, and Hugh and Lilia on the other,
56 Comes, conspalatius et marchio, a titulature not untypical of the emerging ‘territorial

princes’ of the decades around 900: generally, K. Brunner, ‘Der fränkische Fürstentitel
im neunten und zehnten Jahrhundert’, in H. Wolfram (ed.), Intitulatio. II. Lateinische
Herrscher- und Fürstentitel im neunten und zehnten Jahrhundert (Vienna, 1973), 179–340,
and on William, J.-P. Brunterc’h, ‘Naissance et formation des principautés au temps
de roi Eudes: L’exemple de l’Aquitaine’, in O. Guillot and R. Favreau (eds.), Pays de
Loire et Aquitaine de Robert le Fort aux premiers Capétiens (Poitiers, 1997), 69–116, a
reference I owe to Steve Robbie. On conspalatius, Adam Kosto has reminded me that
Niermeyer, Lexicon, 257, has a handful of tenth-century charter references (from, e.g.,
Gorze and Montier-en-Der) for this title as a conflation of comes palatii, all of which
postdate the Hugh gift; given the ease with which the language of the private documents
in the Cluny archive slips from Latin into proto-Romance, it would make sense to see
this as a vernacularization of the formal Latin title and so an indication of how William’s
standing was represented on a village level. See also conpalatinus: J. van de Straeten, ‘Vie
inédite de S. Hughes évêque de Rouen’, Analecta Bollandiana, 87 (1969), 215–60, at
238–9 (a reference I owe to Stuart Airlie).
57 B. H. Rosenwein, ‘Circles of Affection in Cluniac Charters’, in D. Bousseuil et al. (eds.),
Écritures de l’espace social: Mélanges d’histoire médiévale offerts à Monique Bourin par ses
élèves et amis (Paris, 2009), 397–415, esp. 408–10 (affection for fideles); also Rosenwein,
‘Political Uses of an Emotional Community: Cluny and Its Neighbors, 833–965’, in
D. Boquet and P. Nagy (eds.), Politiques des émotions au moyen âge (Florence, 2010),
205–24.
58 Rosenwein, Saint Peter, is fundamental; for gifts and Carolingian officials, see M. Innes,
‘Practices of Property in the Carolingian Empire’, in J. Davies and M. McCormick
(eds.), The Long Morning of Medieval Europe: New Directions in Early Medieval Studies
(Aldershot, 2008), 243–69.
304 Matthew Innes
so evident in 905, must raise the question of whether Hugh and Lilia’s
ability to buy up land so quickly was dependent on the material wealth
and social support of their patron: was their building up of a local network
part of the micropolitics by which aristocratic power was exercised in the
localities, and was Cluny itself able to prosper precisely because it could
take over these carefully built and painstakingly maintained networks on
behalf of its founder?59 Certainly, Lilia played on these links after her
husband’s death: in her gift of 926, she invoked the now dead Count
William as her former lord (senior) and asked that both his soul and that
of her current lord, his eponymous nepos and successor, benefit from her
patronage of Cluny.
As Hugh and Lilia acquired property, their relationship with local
churches appears to have grown closer: whereas their earlier transactions
were written up by apparently lay scribes with no ecclesiastical affiliation,
and publicly enacted in local meetings, five of their last six were written
up by priests (three by the ‘archpriest Josbert’), and one was staged at a
local church. Other notable features of the documentation are the status
enjoyed by wives as equal participants with their husbands in buying and
selling land – though not in witnessing others’ transactions – and their
independent control over their acquisitions, as in Lilia’s own donation to
Cluny after her husband’s death and several sales in the series in which
women acted independently or alongside their sons. Lilia’s activity on the
land market appears to have stopped in 908; perhaps the last and largest
acquisition, of a demesne farm in that year, had consolidated her status.
Her first gift to Cluny of 916 appears to allude to a second husband,
Walter, now apparently deceased, and this remarriage, along with the
continued patronage and protection of successive counts claimed in the
926 donation, must have allowed her to secure her position.
The first of Lilia’s gifts to Cluny, in 916, consisted of a smallholding
(curtilem) within the villa of Bierre; it was made nearby at Chateau,
where Lilia had acquired substantial interests in 908.60 In contrast,
Lilia’s second gift to Cluny, from 926, was far less specific and more
59 The couple were probably local to the villages in which they operated, in that they
appear only in the charters in which they acquired interests in those places, never as
neighbours or witnesses elsewhere. The Hugh with interests in the villa Caucilla in the
county of Vienne (Cluny 125, 137), and the landowner(s) of this name in Cluny 136,
145, 208, are different men with different interests and kin, as is the count Hugh in nos.
241, 256, active after our Hugh’s death.
60 Cluny 202 is the 916 gift; was this gift to William’s foundation based on William’s gift
to Hugh and Lilia, no. 89bis from 905? The two charters stipulate land in the same ager
within the villa, but most of the couple’s other acquisitions in Bierre were also in this
ager, so the 916 gift was likely a composite estate of which William’s fields formed just
part.
open-handed in its terms, simply stipulating the granting to Cluny of all

of her and her husband’s acquisitions in the Mâconnais, and naming the
three villae of Chateau, Bierre and Rufiaco,61 which had been the focus
of her activities. We cannot therefore be sure that the surviving dossier of
sales documents the totality of her holdings – the 926 charter cannot be
taken as an exhaustive list, and in any case is transmitted only through
the cartulary, whose contents are known to have been ‘touched up’ so
as to meet the dictates of institutional memory – or whether there were
more acquisitions for which we now lack a charter record. Perhaps it was
possible for a donation charter like that of 926 to be drafted in relatively
generic terms, and to focus on pious motives and the identities of those
who should benefit from the gift’s salvific powers, precisely because its
public witness of Lilia’s wishes could be supported by title-deeds out-
lining her holdings; this would explain why the monks held onto Lilia’s
dossier. Equally, it does seem clear that some property covered in Lilia’s
dossier was not directly mentioned in either of the gifts to Cluny. The
earliest document in the series, for example, concerns a smallholding in
a perhaps embryonic settlement in the Autunois which – if it was still
in Lilia’s possession some thirty years later – is not explicitly alluded to
in the text of her 926 gift to Cluny, which mentions only the three named
places in the Mâconnais.62
While Lilia preserved the documents recording her acquisitions, and
the Cluny monks felt that they needed keeping as title to their estates,
within the surviving dossier there are embedded references to still more
documents that do not survive: land bought by Hugh and Lilia in 900, for
example, included one parcel that had been ‘chartered’ (the scribe used
the verb incartere from the noun carta, ‘charter’) from husband to wife.63
In other words, in this social circle it was assumed that the charter was
the routine and usual mechanism for transferring dowry land or selling
land, even to kin – so routine, indeed, that only those documents with
61 The dossier vividly shows Hugh and Lilia building up a substantial estate in Bierre, and
also acquiring land and residing at Chateau; Rosiliago, however, does not appear in the
documents, and even if the land here were inherited or dowry land, we would expect
documentation, given the profound penetration of written deeds evident in this society.
The uncertainty underlines that even in a vivid case like this we must in all probability
envisage lost documents that did not survive the dislocations of the Revolution.
62 The villa de Kasale Bertruda in Cluny 60 would most likely be an emerging settlement,
but it is in the Autunois, not the Mâconnais; if Lilia had sold or otherwise lost her holding
here in the meantime, it is striking that she still had the charter. Cluny 84 concerns a
small clearing, and a further small clearing is included with Bierre land in no. 89bis;
nos. 64 and 100 involve land at Rosiliago (and no. 100 also identifies property in two
other villae in addition to Chateau and Bierre): none of these places are mentioned in
no. 275, although the clearings at least might easily be appendages of Bierre.
63 Cluny 67; cf. above, 260n1.
306 Matthew Innes
immediate utility in terms of establishing title were passed on when land

passed to a new owner. The evidently regular circulation of title to land
among Lilia’s class must have eased the rapid acquisition of property
by Cluny after its foundation, not least because – as Lilia’s own final gift
reminds us – Cluny fitted snugly into the systems of patronage and power
in tenth-century provincial society.
Of course, not all the deeds later piled up at Cluny were transmitted in
this way. Within the documents, there are some clear subsets – the royal
and episcopal documents granting local churches to individual aristocrats
or churchmen, for example – for which we can posit alternative vectors
of transmission, for such pieces presumably sat in the churches they
concerned, possibly with related documents or those pertaining to a
church’s owners or patrons, until that church was passed on to Cluny
and its documents absorbed into the monastery’s archive; those royal
diplomata preserved or copied precisely because of their illustrious glitter
overwhelmingly fall into this class of document. But even on the most
optimistic reading, such mechanisms can account for only a tiny share of
the secular documentation preserved at Cluny: fewer than a dozen of the
pre-foundation documents – that is, under 10 per cent or, perhaps more
pertinently, a smaller share than that connected with Lilia. If we are to
understand the nature of this material, then, it is on cases like Lilia’s that
we must base our model.
Dossiers like Lilia’s ultimately suggest that Cluny ended up holding
a large volume of secular deeds precisely because as it acquired land
through the tenth century it also acquired the collections of title-deeds
that went with that land. When Cluny began to acquire property and with
it deeds to that property, it was drawing on well-established documen-
tary practices rooted in the expectations of the landowning class. The
monastery swept into its archive the folded packets of pierced parchment
that patrons kept to document their holdings. This effect – the pull on the
rich documentary traditions of its host society of an institution rapidly
building up its holdings – probably also explains the variety of material
acquired by the abbey’s archives. For while it is possible to identify neat
dossiers like Lilia’s, directly related to property later handed over to the
monks, it is also clear that a very significant proportion of the transmitted
secular documents do not relate in a direct or obvious way to any subse-
quent gift.64 Take the nine or so secular documents – three surviving in
the original – from the late ninth century relating to the villa of Caucilla:
these are not dominated by any single landowner in the way that Hugh
and Lilia monopolize our secular documentation for Bierre, but show
64 Cf. above, 269–81, on the Iberian evidence.

the circulation of property within a small but loosely connected group of

landowners within the villa.65 Cluny acquired land within Caucilla in the
time of the first abbot, Berno, and its interests there were consolidated
in 945, thanks to a gift by the viscount Ratburn that included a church;
by 976, land parcels here were being let back as precarial grants, and by
the end of the century, Caucilla was listed as one among the dozens of
villae where Cluny’s acquisitions were confirmed by King Raoul.66 The
connection between these tenth-century interests and the run of ninth-
century charters is elusive – there is no explicit or direct link with the
grant made by Ratburn half a century after the last transmitted secular
document – and this encourages us at least to consider indirect mecha-
nisms, such as the storage of local title-deeds whose utility was long since
lost in a local church subsequently taken over by Cluny: a parallel state
of affairs to those documented at Perrecy and Rankweil earlier in this
book.
Of course, asserting a negative – in this case, demonstrating a lack of
correlation between ancient title-deeds and current property interests –
is always tricky for a historian, but the advent of electronically search-
able databases of the Cluny material makes it possible to do so with
reasonable confidence. In this bustling regional society, indirect and tan-
gential connections of course abound, for, by the abbacy of Odilo at the
latest, Cluny had a presence in most rural communities and a connec-
tion with most segments of the landowning class in the region. Even if
we allow for some ‘missing links’, lost documents that would have tied
apparently unconnected transactions together, we must admit the pres-
ence of a varied and multivalent documentary culture, and the possibility
that as bundles were handed over to new owners, miscellaneous, stray
or now redundant pieces of parchment were passed over also, rather like
Lilia’s 897 purchase, which ended up with Cluny three decades later even
though the property itself seems to have passed into other hands.
From dossier to archive: the social life of legal documents

How was an individual landowner’s dossier organized, and what does
the existence of such dossiers suggest about archival and documentary
cultures? The surviving originals from Cluny give priceless clues to these
otherwise intractable questions. Barret has drawn attention to the ways –
evident in the originals – in which dossiers connected individual doc-
umentary texts relating to specific people or places: by the addition of
subsequent notices, by the inclusion of documents on the same piece of
65 Cluny 8, 23, 26, 32, 37, 42, 49, 79, 86. 66 Cluny 125, 137, 671, 1423, 2465–6.
308 Matthew Innes
parchment as an existing transaction, by the physical connection of sepa-

rate parchments, and by the piercing through of one or more documents
by the staff or festuca with which legal rituals of property transfer were
performed.67
Such confections could be complex in structure, particularly when they
related to an initial bequest to Cluny to which monks appended subse-
quent episodes relating to the same people or property. In the abbacy
of Odilo, for example, Odalrich and Gauzfred made a donation to the
monastery for the soul of their father Wicfred, which included some
lands that had been previously been the object of a dispute with Cluny.
The document records Odalrich and Gauzfred’s subscription, and then,
in a different hand, the consent of their family; a note of the gift of an
unfree female dependant (ancilla) appears to have been inserted between
the two groups of subscriptions, and a festuca driven through the lower
corner of the parchment; a one-line notice recording a supplementary
donation for the soul of Wichard is also recorded. A second parchment is
sewn onto this mélange of documentary episodes. On the second parch-
ment is an account of the events that took place in 1047, when Wichard,
close to death and unable to talk, was taken to Cluny for burial: his rel-
atives, after consultation, decided to make a gift and have a document
drawn up, but contested the earlier gift of the ancilla, whose husband
demanded to speak to Odilo for satisfaction, prior to the original charter
being produced by the monks and the episode being concluded in high
ritual involving the solemn reconfirmation of the first document and the
paying off of the kin with 23 solidi.68
In Wichard’s case, it is clear that ‘dossierization’ was undertaken by
the monks, with the careful assemblage of the various material relating to
his gifts, and its production when his kin at his deathbed contested earlier
bequests. But there is good reason to think that as a strategy dossierization
originated among the local landowning classes prior to the monastery’s
foundation. In 905, for example, Azo and Hildeard sold lands in the
Lyonnais to Stephen and Gergergana, before witnesses. Lambert, thanks
to whom we have the text of this act, also transcribed the notice of a quit-
claim made by Azo and Hildeard in favour of Stephen and Gergergana,
at Mâcon before the count and the boni homines, where witnesses attested
to having seen the initial investiture (vestitura), the legal ritual whereby
67 Barret, La mémoire et l’écrit, 96–107.

68 Les plus anciens documents III 81–2 = Cluny 2008–9; see also Barret, La mémoire et l’écrit,
98–102, and K. Heidecker, ‘30 June 1047: The End of Charters as Legal Evidence in
France?’, in P. Schulte, M. Mostert and I. van Renswoude (eds.), Strategies of Writing:
Studies on Text and Trust in the Middle Ages (Turnhout, 2008), 87–94.
property was symbolically passed from one individual to another nor-

mally through the handing over of a proxy such as a festuca.69 The care
taken in recording the quitclaim (werpetitio) here outweighs that given to
documenting the original sale: the ritual performance of a quitclaim and
its subsequent recording in writing were crucial strategies for the public
demonstration of title, and the care taken by the abbey to maintain such
records of quitclaim and to ‘dossierize’ them with original donations
shows that it drew on and exploited these long-standing mechanisms for
asserting title.70 Because we are dependent on Lambert’s transcription
and the cartulary for the 905 case, we do not know exactly what this pair
of documentary texts looked like on parchment, nor, indeed, whether
there was a physical mnemonic of the vestitura that was the point of the
second episode; what we can see clearly, however, is the creation of a
mini-dossier to establish and publicly validate title.
This 905 case is suggestive of the pairing and grouping of documents in
conflict-led dossiers, a phenomenon common in cartularies right across
the Frankish world.71 It also invites us to think of the other material
traces of storage and use – dorsal notes, folding, uncut documents on a
single piece of parchment – that we found in the St Gall archive. Many of
the surviving originals from Cluny have dorsal notes, but we have less of
a critical mass of such documents than at St Gall, so the reconstruction
of monastic archiving systems and pre-monastic annotations has not yet
been possible for Cluny. Nonetheless, because of Lambert’s interest in
recording the texts of dorsal notices where he found them, we do get a
sense that some of our secular dossiers were organized and labelled with
practical ends in mind. We have transcriptions of dorsal notes for around
half of Lilia’s dossier, for example, and these texts would tend to indicate
that the documents were not classified as a single set of cartae Ugonis et
Lilianis or titles to Bierre after their passing over to Cluny. The script of
the only original dorsal note is roughly contemporary with that of the
charter scribe, but we do not have dense enough evidence to draw firm
conclusions about the when and why of dorsing, except to conclude that
the dorsing was clearly not made on the entry of the charters into the
Cluny archive – if it were, they would be consistently linked to Lilia.72
69 Cluny 90. 70 Cf. above, Chapter 9, for a parallel position in Italy.

71 See above, Chapters 7–8.
72 Cluny 60 (‘the sale of Deodatus in Kasal Bertrudo), 67 (‘Hugh’s sale in Bierre’), 68
(‘the sale of Raimbod’), 69 (‘Hugh’s sale in Bierre’), 82 (‘Hugh’s sale in Château’), 84
(‘Ageda in Bierre’), 89bis (‘The lord Count William in Bierre’), 93 (‘Hugh’s sale in
Bierre’), 94 (‘Theotrada’s sale in Bierre’), 202 (‘Lilia’s donation in Bierre’). Cluny 89bis
is an original; no. 275 is transmitted only by cartulary, so we have no such record.
310 Matthew Innes
These phenomena underline the extent to which we need to consider

the materiality of the written word on parchment, and the interpenetra-
tion of written and oral in the legal rituals performed with and around
charters. To older generations of historians, weaned on a strong con-
trast between literacy as a ‘rationalizing’ force and ‘irrational’ forms of
legal ritual, the use of the charter as an object might be seen to mili-
tate against its ‘common-sense’ function as a written instrument, even to
indicate its debasement; similarly, the existence of legal rituals of trans-
fer led to questions about whether written documents were the ‘real’
dispositive mechanisms by which property ownership was transferred, or
whether they served as written mnemonics of a dispositive oral ceremony;
such worries were often linked in scholarship of a certain vintage to wor-
ries about ‘Germanic’ influence undermining ‘Roman’ legal principles.73
These concerns indicate how, while it remains indispensable, traditional
diplomatic scholarship’s treatment of documents as disembodied texts,
and its links to an equally weighty legal history rooted in strong opposi-
tions between custom and written law, helped to shape an entire research
agenda here. Yet, does it make any sense to compare the relative signif-
icance of oral testimony, performed legal ritual and written documen-
tation in a context in which law and justice were conducted face to
face, argued according to shared norms and reliant on local public opin-
ion? The absence of any evidence indicating that early medieval people
worried about issues of diplomatic or formal legalism must be telling.
If our questions made sense to them, we would have cases turning on
whether a written document or a legal ritual was the dispositive moment
in a legal transaction; we do not. It must make more sense, in the con-
text of the Cluny evidence, to stress the way in which a parchment was
simultaneously a written document with probative value – one to which
other documents or related notices might be added – and also an object
that might be attached to other objects, with symbolic as well as legal
functions.
As objects whose transfers between different members of a network
defined that network, charters might be seen as acquiring a ‘social life’
of their own, playing a more active role in defining relationships than
as mere passive records.74 These annotations and attachments illustrate
73 The bibliography on these issues is huge. For a convincing demonstration of the inter-
penetration of writing and ritual in a comparable legal culture, see J. Bowman, Shifting
Landmarks: Property, Proof and Dispute in Catalonia Around the Year 1000 (Ithaca, NY,
2004), and Bowman, ‘Did Neo-Romans Curse? Law, Land and Ritual in the Midi
(900–1100)’, Viator, 28 (1997), 1–43.
74 I owe this point to Charles West; the key reference is A. Appadurai, The Social Life of
Things (Cambridge, 1988).
vividly that documents were used, interacting with the repertoire of legal
rituals and oral testimony that were central to public meetings, even
when the notices of those meetings do not mention the production of a
charter: the 905 notice of quitclaim discussed above, for example, does
not mention the existence of an earlier deed of sale, but its transmission
context clearly indicates that an earlier deed informed the quitclaim. In
working from cartulary copies, as we normally must, these elements are
missing, for the documents have been ripped from the context of the
network in which they had initially operated and recopied in a different
register. In assessing the role of written documents in early medieval
legal culture, then, we should take care before assuming that documents
played a role only in cases where they are explicitly cited in records of
court proceedings; we sould acknowledge the textuality of those records,
especially where we are dependent on later cartulary copies that remove
texts from their original contexts and prevent us from seeing the kind of
linkages and usages clear from the processes of addition, annotation and
dossierization apparent at Cluny.75
In and of itself, this must mean that we cannot attempt to examine the
penetration of written documents in court proceedings simply by count-
ing when and where they are explicitly mentioned in records of cases.76
While documents such as the second parchment concerning Wichard’s
donation to Cluny can tell us much about how cited documents were
used and received, early medieval scribal cultures were too fluid and the
mechanisms by which documents might inform legal action too varied
for us to assume that documents were routinely or regularly mentioned
when used. In fact, in the case of Wichard’s donation, considering a
single document out of its archival context obscures a richer hinterland
of archival and documentary practices.77 Locating them in their original
contexts shows us how documents acquired value and validity above all
through their public authentication. The institutions and processes of the
75 For placita texts as textual artefacts, see now W. Brown, ‘Charters as Weapons: On the
Role Played by Early Medieval Dispute Records in the Disputes They Record’, Journal
of Medieval History, 28 (2002), 227–48.
76 Which does not, of course, mean that such endeavour is meaningless, just that it cannot
be assumed to provide a definite or objective picture of document use. Cf. K. Heidecker,
‘Communication by Written Texts in Court Cases: Some Charter Evidence (ca. 800–
ca. 1100)’, in M. Mostert (ed.), New Approaches to Medieval Communication (Turnhout,
1999), 101–26; Heidecker, ‘Emploi de l’écrit dans les actes judiciaires: Trois sondages
en profondeur: Bourgogne, Souabe et Franconie (VIIIe –début XIIe siècle)’, in M.-J.
Gasse-Grandjean and B.-M. Tock (eds.), Les actes comme expression du pouvoir au haut
moyen âge (Turnhout, 2003), 125–37.
77 A point resoundingly made about the Wichard document – which in traditional diplo-
matic manuals had been seen as marking the end of the charter as a legal proof – by
Heidecker, ‘30 June 1047’.
312 Matthew Innes
Roman state had provided precise contexts for the publication and vali-
dation of documents, but by our period these possibilities had receded.
Instead, landowners sought to demonstrate that their documents had
enjoyed a more generic form of public acceptance, one not bound to any
institutional context or informed by firm legal rules so much as respond-
ing to norms and expectations shared by the charter-using classes.78 The
stress on scribal subscriptions and the confirmation of witnesses evident
in original documents needs to be read in these terms, as demonstra-
tions of a document’s acceptance by the local ‘public’, not necessarily
as indicating a particular institutional standing on the part of a scribe or
a specific legal relationship to the transaction on the part of a witness.
So, too, the tactics of annotation and linking documents to demonstrate
their acceptability to a local public rested on a similar logic.
Dossierization through annotation, the copying of connected docu-
ments on a single parchment or the attachment of related parchments,
was thus an archival practice that predated the monastery and was
adopted by the monks simply because it was a practical and sensible strat-
egy in this documentary culture. Through the tenth century, it remained
the dominant archival strategy at Cluny, to the extent that we should
probably think of the monastery’s archive as a collection of dossiers each
relating to dealings with a particular patron or kin group, or a particular
estate.79 Barret suggests that dorsal notices on the surviving originals
indicate some attempt to classify and organize these documents to facil-
itate access by the beginning of Odilo’s abbacy, at the end of the tenth
century. It may be that these first stages of archival reorganization fed
directly into, and perhaps anticipated, the process of cartulary compila-
tion that began later in Odilo’s abbacy. Although we have clear cases of
dossierization well into the eleventh century and beyond, especially in
cases like Wichard’s where donations were contentious and contested, it
does seem that cartulary compilation was one of a series of wider changes
in archival organization designed to create an institutional memory that
served both commemorative and practical purposes. In part, this was
surely a simple necessity given the sheer scale of the abbey’s holdings,
in terms of both land and deeds documenting title to that land. In the
tenth century, when the abbey had been in existence only one or two
generations and the stories of the acquisition of specific estates were
tied to the memory of the monks’ dealings with particular benefactors,
dossierization remained a viable – even a sensible – archival practice,
78 Classen, ‘Fortleben’; P. Johanek, ‘Zur rechtlichen Funktion von Traditionsnotiz, Tra-

ditionsbuch und früher Siegelurkunde’, in Recht und Schrift, 131–62; obviously, I read
the post-Roman end differently. Cf. also above, Chapter 5.
79 This is Barret’s conclusion: La mémoire et l’écrit, esp. 103–7 on ‘un “dossier” d’archives’.
facilitating as it did the documentation of episodic and oscillating rela-

tionships with patrons. By the end of the tenth century, however, the
abbey’s archive must have contained perhaps a thousand documents,
and so probably well over a hundred groups of charters and dossiers,
the earliest strata of which concerned the gifts of Lilia and her con-
temporaries, now beyond the horizon of living memory. The episodic
relationships between Cluny and its patrons are so visible in the tenth-
century documentation precisely because they defined archival practices
and so determined the pattern of documentation at the abbey; by the sec-
ond half of the eleventh century, as a different archival system established
itself, the norms governing the ‘gift economy’ binding the monastery to
its benefactors began to alter. The result was a shift from archival prac-
tices immediately identifiable to any lay landowner, which connected
parchment to people to place to make sense of current relationships, to a
new and deeper system of archival memory that underwrote institutional
identity while enabling easy reference for those searching for documents
relating to a particular privilege or specific estate. The case of Cluny
thus shows how a monastery, in its early phases, might ‘piggyback’ on
pre-existing archival and documentary practices. It also shows, however,
how once an ecclesiastical patrimony had reached a certain scale, and
an ecclesiastical institution a certain degree of self-consciousness, a fixed
archival memory of a new kind, more distant from the practices of the
landowning classes, might develop.
Conclusions: charter collections, archival

transformations and social change
Shifts in archival practice at Cluny, encapsulated in the compilation of the
cartularies in the eleventh century, thus had their roots in the changing
relationship between the monastery and its ‘host’ society in the first cen-
tury and a half of its existence. These developments, however, were not
unique to Cluny: they parallel a series of wider archival transformations
right across the post-Carolingian kingdoms in the eleventh and twelfth
centuries. In particular, major monastic houses sifted their charter col-
lections as they had their traditions edited and reordered in cartularies
and pancartes, technologies rapidly imitated by other ecclesiastical institu-
tions and, in the course of the twelfth century, by major lay landowners
too. In monasteries, the process seems to have had important links to
wider currents of intellectual influence and to the spread of ideologies
of ‘reform’ from centres like Cluny.80 Initially motivated by the dual
imperatives of institutional commemoration and tenurial protection, but
80 Geary, Phantoms, 81–113; Les cartulaires.

314 Matthew Innes
increasingly tied up with the crystallization of new forms of seigneurial

power over judicially defined territories, these eleventh-century compi-
lations did have precedents in the ‘books of traditions’ of Carolingian
times, into which some of the newly established churches in the recently
conquered and Christianized provinces beyond the Rhine had copied
records of donations. But ninth-century initiatives had been shaped by
very specific impulses in particular contexts, above all by the tenurial
disruption wrought by the ‘Carolingianization’ of these regions.81
While the records of Freising and Fulda have profoundly influenced
modern historiographical expectations about Carolingian and post-
Carolingian record-keeping, they should not be seen as representing
a new archival template that swept aside the varied practices that had
developed in the post-Roman centuries, nor did they replace the char-
ter collections and dossiers that typified documentary practice within
Frankish society.82 The adoption of new forms of record-keeping right
across eleventh- and twelfth-century western Europe, on the other hand,
took place as landholding institutions – first ecclesiastical landowners and
then the newly institutionalized secular lordships of the twelfth century –
responded to far-reaching economic, social and tenurial changes.83 Fore-
most among those imperatives in the earliest waves of cartulary compi-
lation was a heightened awareness of the special status of ecclesiasti-
cal property and increased concern with identifying the dividing lines
between the sacred and secular in this world – concerns very visible
in the Cluny cartularies and their publication in concert with Urban
II’s promulgation of a ‘sacred ban’ around the heart of the monastery’s
territory. This sense that ecclesiastical possessions formed an inviolable
patrimony interacted with notions of spiritual protection and immunity
as norms about the interaction between churches and lay benefactors
underwent fundamental revision.84 Economic change – the increasing
81 E.g. Hummer, Politics; M. Innes, ‘Property, Politics and the Problem of the Carolin-
gian State’, in W. Pohl and V. Wieser (eds.), Der frühmittelalterliche Staat: Europäische
Perspektiven (Vienna, 2009), 299–314; H. Wolfram, ‘Die Notitia Arnonis und ähnliche
Formen der Rechtssicherung im nachagilolfischen Bayern’, in Recht und Schrift, 115–30;
and above, Chapters 7–8.
82 See above, Chapters 7–8.
83 Chronology of cartulary compilation: I. Rosé, ‘Panorama de l’écrit diplomatique en
Bourgogne: Autour des cartulaires (XI–XVIII siècles)’, Bulletin du Centre d’études
médiévales d’Auxerre 11 (2007): www.cem.revues.org/index2352.html.
84 See Rosenwein, Saint Peter, and Rosenwein, Negotiating Space: Power, Restraint, and
Privileges of Immunity in Early Medieval Europe (Ithaca, NY, 1999), 156–83. On changing
relationships between monks, benefactors and commemoration, see D. Iogna-Prat, ‘The
Dead in the Celestial Bookkeeping of the Cluniac Monks’, in L. Little and B. H.
Rosenwein (eds.), Debating the Middle Ages: Issues and Readings (London, 1998), 340–
62, and on monks and laity, D. Iogna-Prat, Order and Exclusion: Cluny and Christendom
monetization and marketization of rural production, and the linked phe-

nomenon of price inflation – led to the adoption of new techniques
of estate management based on the more direct exploitation of posses-
sions, as opposed to long-term fixed-rent arrangements embedded in
social relationships.85 Similarly, the crystallization of new forms of power
in the countryside, rooted in the territorialized jurisdictional rights of
the seigneurie, created a new layer of formal rule over and above the
flux of sociable exchange.86 For landowners – increasingly the great lay
lords, now able to pass on institutionalized territorial power from gen-
eration to generation, as well as ecclesiastical institutions – new tech-
niques of accountability and record-keeping were vital in facing these
changes.87
As a result of these changes, earlier medieval documentary practices
were effectively lost beneath our historiographical horizon as novel forms
of archival organization reshaped earlier traditions.88 Cartulary compila-
tion involved both the selection and the reorganization of original doc-
uments which had been carefully ordered and preserved, in some cases
for centuries, in institutional charter collections. While dorsal notes on
surviving originals can, as we have seen, give some hint at the classifi-
cation systems of early medieval charter collections, we can only access
them indirectly, via the kind of methodology attempted above. We cannot
know how much or how quickly old charter collections were dispersed or
destroyed in the aftermath of cartulary compilation: at Cluny it is striking
that the bundles of early charters selectively copied into or omitted from
the cartularies seem to have remained, dormant and unused, with the
monastery into the early modern period and the age of Lambert and his
fellow antiquarians, but elsewhere there are hints at dispersal and reuse,
Face Heresy, Judaism and Islam, 1000–1150, trans. G. R. Edwards (Ithaca, NY, 2002
[1998]), 9–96.
85 G. Duby, ‘Économie domaniale et économie monétaire: Le budget de l’abbaye de Cluny
entre 1080 et 1155’, in Duby, Hommes et structures au moyen âge (Paris, 1973), 61–82
(orig. Annales ESC, 7 [1952], 155–71).
86 The literature is huge, and intimately tied up with questions of causation that are hotly
contested as part of the ‘feudal revolution’ debate; for a recent perspective, see T. N.
Bisson, The Crisis of the Twelfth Century: Power, Lordship and the Origins of European
Government (Princeton, NJ, 2008).
87 Monastic accounting: R. Berkhofer, Day of Reckoning: Power and Accountability in
Medieval France (Philadelphia, 2004).
88 L. Morelle, ‘The Metamorphosis of Three Monastic Charter Collections in the Eleventh
Century (St.-Amand, St.-Riquier, Montier-en-Der)’, in Heidecker (ed.), Charters and
the Use of the Written Word, 171–204. Cf. S. Vanderputten, ‘Transformations in Charter
Production and Transmission during the “Iron Age” (Tenth–Early Eleventh Centuries):
Some Evidence from Northern France and the Southern Low Countries’, Jaarboek voor
Middeleeuwse Geschiedenis, 7 (2004), 7–30; this paper is undermined by the fact that it
adopts a quasi-quantitative methodology while failing to consider modes of transmission.
316 Matthew Innes
with some originals surviving precisely because they were recycled.89

Once a donation was a century or more old, the practical utility of the
dossier of title-deeds recording a property’s prehistory and establish-
ing the right of the donor to grant it must have receded. Cartularies
were arguably more suited to documenting historical privileges, claims
to immunity and rights of seigneurie – needs close to the hearts of many
eleventh- and twelfth-century abbots and bishops – than were chests full
of dossierized charters. Meanwhile, parchment remained in high – indeed
increasing – demand, and new uses could always be found for old doc-
uments. More or less ‘organic’ turnover, with ‘active’ dossiers becoming
dormant, gathering dust and perhaps eventually being lost or discarded,
must have been common within early medieval charter collections; car-
tularies, on the other hand, fixed a very different kind of memory, shaped
by the cultural and legal needs of a new social order, and in the long run
that memory came to dominate institutions’ understanding of their own
past.
We certainly should not assume that the shift to cartulary copies was
motivated by a ‘crisis’ that undermined the legal utility of original docu-
ments. The case of Wichard in 1047, discussed above, has been cited as
an indication of the failure of such documents to be used as legal proof,
but, in fact, in this case, negotiation between monks and their bene-
factors was structured around a dossier of documents and appended
notices that was immediately fetched from the monastic archive when
questions about the donation were raised. Transactions such as this,
and the texts of the documents recording them, may be messy, and
not comply to any formalized juristic rules about hierarchies of proof
or any sense of a ‘clean’ legal process whereby rights were absolutely
and irrevocably transferred, but that is precisely because at this date
rights and norms remained embedded in social relationships in a world
of ‘substantive legalism’ rather than subject to an abstracted ‘formal
law’.90 In any case, the whole notion of a ‘crisis’ or ‘decline’ in written
89 The extent of conscious archival destruction has been a major focus of debate following
Geary, Phantoms.
90 Rosenwein, Saint Peter, and, for formal law versus substantive legalism, S. D. White,
Custom, Kinship and Gifts to Saints: The Laudatio Parentum in Western France, 1050–1150
(Chapel Hill, NC, 1988), and White, ‘Inheritance and Legal Arguments in Western
France, 1050–1150’, Traditio, 43 (1987), 55–103; on Wichard, Heidecker, ‘30 June
1047’. For the notion – inherited from diplomatic textbooks – of a decline in the
use of written documents as proof, used to explain the move to cartularies, see G.
Declercq, ‘Originals and Cartularies: The Organisation of Archival Memory (9th –11th
Centuries)’, in Heidecker (ed.), Charters and the Use of the Written Word, 147–70; cf. P.
Geary, ‘Umgang mit Urkunden im frühen Mittelalter’, in P. Erhart and L. Hollenstein
(eds.), Mensch und Schrift im frühen Mittelalter (St Gall, 2006), 11–24, esp. 20–1.
documentation is itself an illusion created by changes in archival hori-

zons. Because those Carolingian and earlier documents copied in the
new cartularies of the eleventh and twelfth centuries overwhelmingly
recorded the munificence of kings and founding benefactors and the
disinterestedness of their justice, they create a backdrop of highly for-
malized acts of carefully formulated Latinity. From a certain perspective,
searching for legal clarity and hierarchy, this backdrop might encour-
age criticism of the more episodic narrativité and breathless practicality
of surviving documents: hence notions of a rising barbarity, even ‘doc-
umentary anarchy’. But if we set aside the assumptions underpinning
these judgements, a more continuous development of pragmatic legal
writing becomes apparent.91
Cluny, because of its unique position in the high Middle Ages and
the remarkable interest in its muniments in the eighteenth, nineteenth
and twentieth centuries, offers a rare possibility of catching tantalizing
glimpses of what lies beyond the evidentiary horizon set in the eleventh
century. It is no accident that the only truly comparable case comes from
St-Denis in Paris. St-Denis’ continuing history through Merovingian and
Carolingian times, and its close relationship with the Frankish Crown
from the seventh century, mean that its early medieval archive must have
looked very different in composition and chronological depth from that of
Cluny. But, like Cluny, it was the subject of significant antiquarian inter-
est through the ancien régime – modern diplomatic scholarship, indeed,
was born from the St-Denis archive and Mabillon’s efforts to prove the
veracity of its traditions – and its contents fed directly into the national
collections of the nineteenth-century French state, hardly surprisingly
given the monastery’s ‘national’ profile since medieval times. Thus the
great cartulary compilations of the abbey – most especially the White
Cartulary of the early thirteenth century – are complemented by a rich
transmission of originals, and by the works of antiquarians, which may
not be systematic efforts at transcription like Lambert’s but nonetheless
give a clear sense of what has been lost.92 What is immediately clear is the
extent to which the high medieval cartulary tradition was determined by
current concerns. Remnants of a stream of donations by benefactors of
the seventh to tenth centuries, ranging from members of the Carolingian
family and the ‘imperial aristocracy’ to more local landowners, survive
91 Guyotjeannin, ‘Penuria scriptorum’; Barthélemy, ‘Une crise de l’écrit?’

92 An indispensable study aid for anyone working on the St-Denis archive is the web-
site directed by O. Guyotjeannin, ‘Chartes de l’abbaye de St-Denis, VII–XIII siècle’,
saint-denis.enc.sorbonne.fr, with full transcriptions of the Cartulaire blanc and the early
modern Inventaire générale; see also D. Sonzogni, Le chartrier de l’abbaye de St-Denis en
France au haut moyen âge: Essai de reconstruction (Paris, 2003).
318 Matthew Innes
in the original, but the cartularies exclude most in a relatively system-

atic manner, with the exception of royal diplomata; our early modern
inventories and antiquarian notes identify significant further numbers
of now lost documents.93 Within these traditions, there are tantalizing
indications of now lost lay traditions similar to those we have analysed at
Cluny: for example, a lone original document from 769 in which Aege-
fredus and Archesidana sold land in the villa Pociollus in the Pincerais to
Nautlindo.94 Famously, these survivals include a run of original papyri
from the seventh century paralleled only at Ravenna in the West, which,
along with a body of original royal diplomata on parchment from the last
quarter of the seventh century, constitute the primary evidential basis for
our knowledge of late Merovingian documentary culture. The fifty-nine
surviving original Merovingian documents were transmitted outside the
mainstream archival tradition, in that they were not copied in the car-
tularies, nor is there any evidence for their use at medieval St-Denis.
Some of the magnificent ancient papyri were used to confect a series of
eleventh-century forgeries: the medium evidently now had more value
than the Merovingian message inscribed on it, as Abbot Suger sought
to buttress the abbey’s position by the production of allegedly ancient
papal privileges. Others simply sat, excluded from the main monastic
tradition rather like the secular material at Cluny. Others still – for exam-
ple, twenty-two of the twenty-nine Merovingian documents used by the
ninth-century author of the Gesta Dagoberti – were lost.95 Significantly,
very many of our surviving Merovingian documents do not directly relate
to the abbey or its interests: some are royal privileges or judgements in
favour of other beneficiaries, and some are the documents of wealthy
private landowners from the Paris region, including an important run of
last wills and testaments.
The close links between St-Denis, the Merovingian court at Paris,
and the highest echelons of the Frankish elite doubtless explain how
much of this came to end up at the abbey: as at Ravenna, there is a
political nexus at work. But this specific context should not obscure
the broader phenomenon of a church perhaps acting as a recognized
store for significant and contestable documents, and so vacuuming up a
significant quantity of papyrus and parchment in which its interests were
not directly implicated. While elsewhere we simply do not have the weight
93 E.g. D. Sonzogni, ‘Un acte de vente inédit du chartrier de Saint-Denis (11 avril 702?)’,
Bibliothèque de l’École des chartes 159 (2002), 609–13.
94 ARTEM no. 4488, ChLA XV 609.
95 Geary, Phantoms, 107–13, with bibliography; H. Atsma and J. Vezin, ‘La faux sur
papyrus de l’abbaye de St-Denis’, in J. Kerherve and A. Rigaudière (eds.), Finances,
pouvoirs et mémoire: Mélanges offerts à Jean Favier (Paris, 1999), 684–99.
of antiquarian scholarship and rescued originals that typifies Cluny and

St-Denis, we do have a constant murmur of hints that their cases were
far from unusual. From Murbach, for example, our surviving originals
focus on the abbey’s founders and major benefactors, but we also have
an original document from 762 in which Sighifrid gave a considerable
portfolio of estates to his son Altmann, some of which later ended up
in Murbach’s hands; from Cambrai, a gift in 875 by Henry to his son
Warmund; from Nouaillé, another gift between private individuals from
886 and a sale from 897 amidst a fascinating set of private dossiers; from
Narbonne, two related sales from 912 and 925; from Poitiers, a sale from
923; from Limoges, a sale from 941; from La Grasse, a sale from 947;
and so on.96 The spread of this ‘background noise’ of chance survivals
is evident from the database of original early medieval charters from
modern France compiled by the ARTEM (Atelier de Recherche sur les
Textes Médiévaux) project hosted by Nancy 2 University. In the ARTEM
database, almost 60 per cent of the 672 surviving originals before 950
were issued by emperors, kings or popes; most of the remainder involve
churches or priests as one of the actors, with only 56 where all parties
were laymen.97 But even if we acknowledge that 22 of these documents
come from Cluny (almost all in the single charter collection in BnF lat.
17715), the scattering of other traditions does indicate that churches right
across the Frankish world had contact with lay documentary traditions
that they did not create or control.
Even cartularies – whose date and function mean that they are least
likely to include lay documents – contain echoes of such traditions. It is
almost normal for such collections to have a document or two concerning
the families of founders or important early benefactors, and in a handful
of cases specific impulses encouraged high medieval cartulary compilers
to copy out whole dossiers of lay documents. At Perrecy – discussed
above – our remarkable run of Carolingian estate documents and title-
deeds survived on site through to the eleventh century, when they were
copied in order to add ballast to a cartulary of the priory now founded
on the Perrecy estate, which was concerned above all with the complex
relationship between that priory and its mother house, the monastery of
96 ARTEM no. 3873; ChLA XIX 676; ARTEM nos. 352, 1068, 1073, 2372–3, 1090, 647,
3785, respectively.
97 The figures here are based on the discussion in Kosto, ‘Laymen’, 73–4, and in Tock,
‘L’acte privé en France, VIIe –milieu du Xe siècle’, in Les transferts, 499–537, at 502–7,
534–5. Note that the majority of the St-Denis documents discussed above are excluded
from these calculations as they are either royal or (in the case of many of the wills
and ‘private’ documents) involve some property going to the Church. In any case the
ARTEM search categories are not watertight when not backed up by a manual search.
320 Matthew Innes
Fleury.98 At Redon in Brittany, an eleventh-century scribe came across

dossiers of sales and exchanges similar to those we found at Cluny, was
unable to fit them easily into his scheme of a cartulary in the guise of
gesta abbatum, but copied them nonetheless.99
In all of these cases, we see a society in which documents were used and
stored by landowners without defined public institutions of record, and
without the Church enjoying hegemony as a provider of documentary
services. In other words, we need to acknowledge the existence of a
complex set of practices that cannot be understood as a mere staging
post between the lost late antique world of public archives and the great
ecclesiastical charter compilations of the eleventh and twelfth centuries,
but needs evoking in its own terms. Crucial to these practices was the
need to demonstrate that a document was a publicly validated record –
hence the importance of scribes, subscriptions and witnesses, and the
care taken to make public acceptance of a document physically tangible
on the document itself, through annotation or the attachment of further
documents or objects. This was a world in which establishing title through
creating a convincing narrative that could mobilize public opinion was
an important driver of document use, as late antique concerns about the
registration of fiscal liabilities were no longer active, and the claims of
the seigneurie had not yet developed. Our fragmentary dossiers show that
documents were crucial episodes in ongoing narratives that linked land
and landowners, and that were periodically debated, openly, in public. As
pieces of parchment or papyrus they may be unremarkable, but their very
scrappy materiality takes us to the heart of the archival and documentary
culture of the early medieval centuries.
98 See above, 176–82.

99 W. Davies, ‘The Composition of the Redon Cartulary’, Francia, 17 (1990), 79–100;
Davies, ‘Intra-family Transfers in Southeastern Brittany: The Dossier from Redon’,
in Les transferts, 881–94; Davies, ‘People and Places in Disputes in Ninth-Century
Brittany’, in Settlement of Disputes, 65–84; Davies, ‘Priests and Rural Communities in
Eastern Brittany in the Ninth Century’, Études celtiques, 20 (1983), 177–97; Davies,
Small Worlds.
12 Documentary practices, archives and
laypeople in central Italy, mid ninth to
eleventh centuries
Antonio Sennis
The common documentary culture of tenth- and eleventh-century Italy is so

highly visible in the sources as to have long been recognized, though we can
now see that it was neither a recent development nor exceptional. Yet this
chapter shows that active questions remain about laypeople’s engagement with
documents. The Italian evidence not only shows that lay families preserved doc-
uments but also hints at how they did so – for instance, through depositing them
in chests and cupboards, sometimes apparently randomly. It also demonstrates
how archives could be filtered by the deliberate destruction of documents that
were considered redundant, irregular or forged. Conversely, when the provisions
in documents needed to be maintained, they could be confirmed through the law
courts. Once more, we see that it is not a distinction between clergy and laity
that shaped documentary practices. Rather, lay families and their patrimonies,
which shifted and fragmented with each generation, lacked the institutional
mindset that ecclesiastical institutions developed, which encouraged them to
preserve records for the long term.
In the spring of 983, the emperor Otto II received a letter from Gerbert,
the newly appointed abbot of the monastery of Bobbio. ‘I would rather
be bringing you good than sad news,’ the monk wrote, ‘but how could I
keep silent, when I see my brethren consumed by hunger and degraded
by nakedness? The situation would perhaps be bearable, had we not been
deprived even of the hope that it might improve. By means of I do not
know what documents, which here they call libelli, this church of God has
been entirely sold. The money that was paid is nowhere to be found; our
cupboards and granaries are empty; there is nothing in our pockets.’1
The emperor died later in the year and never granted any privilege to
Bobbio. We therefore do not know what, if anything, he had to say about
the rather peculiar use that Gerbert’s predecessors and their (mainly lay)
1 Gerbert of Aurillac, Epistolae 2 (ed. P. Riché and J. P. Callu [Paris, 1993], 1:5).
321
322 Antonio Sennis
counterparts seem to have made of a type of contract that was never

meant to transfer permanently rights of property. For his part, Gerbert
was quick to leave the Italian monastery and, in the course of time, was
quite successful in grazing new, if by no means less troubled, pastures:
first the see of Reims, then that of Ravenna and eventually the greenest
of them all, Rome.2 He did not, however, lose interest in the fortunes of
Bobbio. Keeping, at least nominally, his abbatial title, Gerbert evidently
kept an eye on how things were developing. On 1 October 998, Otto III
intervened, at Gerbert’s request, to put an end to the manipulation of
libelli, or charters of lease.3 Although highly contentious, the Capitulare
Ticinense, issued on 20 September of the same year, sanctioned that,
from then on, leases and emphyteuses granted by a church would expire
with the death of the person who had granted them and would therefore
not impinge on the will of his or her successor, who was completely free
to confirm them or not. It is true that the fact that this capitulary was
not included in the Liber Papiensis indicates that, most probably, it was
never properly enforced.4 Nevertheless, this episode and its legal results
clearly remind us that, perhaps surprisingly for some contemporaries, the
inhabitants of most parts of Italy, whether lay or cleric, indeed knew how
important preserving written documents was to establish, claim and even
abuse their rights.5 Moreover, even if often many among them could not
read what was written on the charters they owned, they knew how to use
them well – at times, as in the case of Bobbio’s counterparts, all too well.
When compared to other regions of Europe, Italy seems, at first
glance, strikingly peculiar. If one looks at what are usually considered the
fundamental elements in any written tradition – levels of literacy, social
2 Libelli: P. Toubert, Les structures du Latium médiéval: Le Latium méridional et la Sabine du

IXe siècle à la fin du XIIe siècle (Rome, 1973), 516 ff.; C. Violante, ‘La signoria rurale nel
secolo X’, in Il secolo di ferro: Mito e realtà del secolo X, Settimane 38 (Spoleto, 1991), 349
ff. Gerbert of Aurillac: M. Oldoni, ‘Silvestro II’, in Enciclopedia dei Papi, 3 vols. (Rome,
2000), II, 116–24.
3 MGH DD Otto III, no. 303: ‘iubemus . . . ut ea quae male his temporibus acta sunt sine
abbatis Gerberti auctoritate et detinentur vel in precariis aut commutationibus rerum vel
hominum sive in libellis aut aliquibus scriptis, nemo retineat, nullus ex eis se intromittere
audeat, sed propria nostra auctoritate frustrentur et omnia destruantur, nisi ab eodem
iterum melius ordinentur et restaurari videantur’.
4 MGH Const. 1, no. 23; F. Bougard, La justice dans le royaume d’Italie de la fin du VIIIe
siècle au début du XIe siècle (Rome, 1995), 53; A. Visconti, ‘Una legge feudale di Ottone
III esclusa dal capitolare italico’, Reale istituto lombardo di scienze e lettere: Rendiconti, 2nd
ser., 60 (1927), 198–208. On the potentially destabilizing effects of this capitulary, see
K. Görich, Otto III. Romanus, Saxonicus et Italicus: Kaiserliche Rompolitik und sächische
Historiographie (Sigmaringen, 1993), 269, 277; D. A. Warner, ‘Ideals and Action in the
Reign of Otto III’, Journal of Medieval History, 25 (1999), 1–18, at 11.
5 C. Wickham, ‘Land Disputes and Their Social Framework in Lombard-Carolingian
Italy, 700–900’, in Wickham, Land and Power: Studies in Italian and European Social
History, 400–1200 (London, 1994), 229–56, at 242–3 (orig. in Settlement of Disputes,
104–24).
Central Italy, mid ninth to eleventh centuries 323
uses of writing, the development of a group of professionals who ful-

fil the documentary requirements of society – Italy has traditionally
been considered to be an exception in medieval Europe. The (relatively)
widespread use of writing and of written documents by the laity in early
medieval Italy emerges as a fairly established fact, as Everett and Costam-
beys have already pointed out in their chapters above.6
Clearly, even in Italy, churches did have strong control over the pro-
duction and circulation of written documents, and this fact had relevant
consequences. For example, the relationship between a monastery and
the notaries writing its charters tended to result in very strong, if not
exclusive, control on the part of the institution over local social dynam-
ics. Notaries could not only write acts; they could also, for example,
testify in court in favour of the monastery or prove similarly useful in
arbitrations.7 Moreover, by conveying, often in an aggressive way, ideas
about the right ordering of society, the undeniable liturgical dimension
that any charter had was also used by churches to shape the relation-
ships that they had with the outside world. When charters were placed
beside the eucharistic species on a monastic altar, lay generosity was
turned into a liturgical act, and the identification between the religious
and material treasures of the community was thus reinforced.8 Besides,
6 A. Bartoli Langeli and A. Petrucci (eds.), Alfabetismo e cultura scritta: Con alcuni contributi
su psicologia e storia (Bologna, 1978) (= Quaderni storici, 13 [1978], 437–796); F. H.
Bäuml, ‘Varieties and Consequences of Medieval Literacy and Illiteracy’, Speculum,
55 (1980), 237–65; A. Bartoli Langeli, ‘Storia dell’alfabetismo e storia della scrittura:
Questioni di metodo’, Annali della Facoltà di lettere e filosofia dell’Università degli studi di
Perugia, 26, n.s. 12 (1988–9), 221–37; M. Banniard, Viva voce: Communication écrite et
communication orale du VIe au IXe siècle en Occident latin (Paris, 1992); F. Cardini, ‘Alto e
basso medioevo’, in G. Cavallo, C. Leonardi and E. Menestò (eds.), Lo spazio letterario
del medioevo: Medioevo latino (Rome, 1992–8), 1.1:130–43, with bibliography; above,
Chapters 4 and 9.
7 E.g. the case of a Dominicus tabellio who, in the course of a long career spanning at
least the period between 1068 and 1118, wrote almost all the charters related to the
properties of the monastery of Fonte Avellana in the territory of Cagli (Carte di Fonte
Avellana. I. 975–1139, ed. C. Pierucci and A. Polverari [Rome, 1972], index). Similarly, a
certain Batalla, between 1152 and 1189, wrote 121 charters (practically all the surviving
ones) regarding the properties of the Venetian monastery of S. Zaccaria in Monselice
(Codice diplomatico padovano dall’anno 1101 alla pace di Costanza (25 giugno 1183), ed.
A. Gloria [Venice, 1879–81]; G. Tasini, ‘I notai e le badesse: La gestione delle proprietà
del monastero di San Zaccaria di Venezia in territorio di Monselice (secoli XII–XIII)’,
in Chiese e notai (secoli XII–XV) [Verona, 2004], 245–59). Medieval notaries who, in
spite of a privileged relationship with a church, continued to write documents for other
institutions or individuals: L. Fois, ‘I notai del monastero di Sant’Ambrogio di Milano
nel XIII secolo (una prima indagine)’, in Chiese e notai, 261–84, at 273.
8 F. Bougard, ‘Trésors et mobilia italiens du haut moyen âge’, in J.-P. Caillet (ed.), Les
trésors de sanctuaires, de l’Antiquité à l’époque romane: Communications présentées au Centre
de recherches sur l’antiquité tardive et le haut moyen âge de l’Université de Paris X–Nanterre
(1993–1995) (Paris, 1996), 161–97, at 167 (and nn. 43–5 for a number of examples).
Donations to monasteries performed in front of the altar: M. Ragnow, ‘Ritual before
324 Antonio Sennis
monasteries were communities whose members could suspend the

celebration of sacraments for political gain and use sacred rites to invoke
the eternal damnation of someone who had seized a piece of their land.9
It is therefore not surprising that, for example, in the eleventh century
the comparatively low literacy of laypeople rendered their religious life,
in the eyes of clerics, inferior.10
Lay individuals, however, could prove to be quite at ease with the dif-
ferent stages (and values) of production, circulation and preservation of
a charter. In late eighth-century Lucca, for example, the notary Filip-
pus, on request of the donor Porphoreus, corrected the document from
which it appeared that Porphoreus had given himself, as well as his prop-
erties, to the Tuscan church of S. Pietro di Castiglione.11 Lay awareness
of documentary practices progressed throughout the centuries. In the
early twelfth century, the use of different systems of dating on the part of
notaries writing charters for the abbey of Sassovivo frequently depended
on precise requests of the parties (often lay donors).12 And as for the
liturgical dimension of documentary practices, it has to be said that it
is not always easy to tell how effective all those frightening rituals were
and whether the prospect of being excommunicated and damned was
enough to keep people, especially lay outsiders, at bay. Although the sur-
vival of monastic communities largely depended on the spiritual function
they performed, their members continually alluded to the suffering they
had to endure from outsiders who evidently were not too worried about
divine anger.13 Benefactors could quickly change their minds and take
back their charters of donation, at times cunningly exploiting just the
elaborate liturgical setting that was supposed to bind them. This is, for
the Altar: Legal Satisfaction and Spiritual Reconciliation in Eleventh-Century Anjou’, in

J. Rollo-Koster (ed.), Medieval and Early Modern Ritual: Formalized Behaviour in Europe,
China, and Japan (Leiden, 2002), 57–79, esp. 69–71.
9 G. Constable, The Reformation of the Twelfth Century (Cambridge, 1996), 182; J. F.
Romano, ‘Julian of Vézelay, a Twelfth-Century Critic of His Monastery’s Worldly Suc-
cess’, Medieval Sermon Studies, 50 (2006), 51–69, at 59–61; B. H. Rosenwein, ‘Feudal
War and Monastic Peace: Cluniac Liturgy as Ritual Aggression’, Viator, 2 (1971), 129–
57; L. K. Little, Benedictine Maledictions: Liturgical Cursing in Romanesque France (Ithaca,
NY, 1993).
10 Y. Congar, ‘Clercs et laı̈cs au point de vue de la culture au moyen-âge: “laicus” = “sans
lettres”’, in Studia mediaevalia et mariologica P. Carolo Balić . . . dicata (Rome, 1971),
309–32.
11 CDL II 250, 251: ‘per uoluntate domni Peredei in Dei nomine episcopi et per rogitum
Porphorei, ego Filippus clericus iscriptor huius cartule abstuli de hanc cartulam, persona
ipsius Porphorei non esset offerta in ecclesia Sancti Petri; et iterum iuidem rescripsi,
sicut ambarum partium placuit’.
12 Le carte dell’abbazia di S. Croce di Sassovivo. II. 1116–1165, ed. V. De Donato (Florence,
1975), p. xx, n. 37.
13 J. Nightingale, Monasteries and Patrons in the Gorze Reform: Lotharingia, c. 850–1000
(Oxford, 2001), 261.
example, the case of a certain Gaitelgrima, who, in the first decades of

the twelfth century, argued that the donation with which her deceased
husband had granted a castrum to S. Bartolomeo in Carpineto was not
valid because he had not placed the charter on the altar.14
The most immediate answer to the question of why Italy would be so
different from the rest of Europe is, usually, that the legacy of Roman
traditions must have played some major role in determining this speci-
ficity. When new analyses have added nuance to this image of Italian
exceptionalism, they have done so not by undermining what can be
called the ‘documentary awareness’ of laypeople in Italy but, rather, by
reappraising the spread and social distribution of documentary practices
throughout the rest of Europe.15 However, it remains a fact that until the
eleventh century, even in Italy, documents – whether produced by and
for the laity or the clergy – have been transmitted to us via ecclesiastical
institutions.16 This raises the question of whether laypeople, as individu-
als or as groups, ever felt the need for the intentional, organized, mindful
preservation, over significant periods of time, of those documents we
know they were using. In other words, since the use of documents was
not restricted to clerics, and since laypeople, albeit with different degrees
of intensity, were fully involved in the documentary practices of their
respective social groups, were they also involved in their preservation?
And, if so, in what ways, and to what extent?
To help explain the pre-eminence of churches in the conservation and
transmission of texts and documents, some historians have tended to
appeal vaguely to different lay and clerical attitudes towards writing.
According to this view, the coexistence, not necessarily in conflict, of two
separate spheres of culture, one related to writing (reserved to monks
and clerics) and one to orality (peculiar to the laity), would have induced
laypeople to preserve documents with much less care, if at all. When
confronted with the entirety of documentary practices, lay individuals
would have accessed some of them in ways that were structurally differ-
ent from their ecclesiastical counterparts, ways that depended on their
status, their aims and their education (or, better, lack thereof).17 This
14 Alexander the Monk, Liber chronicorum monasterii sancti Bartholomei de Carpineto 4 (ed.
B. Pio [Rome, 2001], 154). On the problem of hypocritical submissions to churches
by individuals who would then quickly renege, see also G. Koziol, Begging Pardon and
Favor: Ritual and Political Order in Early Medieval France (Ithaca, NY, 1992), 316–21.
15 Uses of Literacy, and the research that collection has inspired.
16 P. Cammarosano, Italia medievale: Struttura e geografia delle fonti scritte (Rome, 1991),
39–111.
17 H. Grundmann, ‘Litteratus-illitteratus: Der Wandel einer Bildungsnorm vom Altertum
zum Mittelalter’, Archiv für Kulturgeschichte, 40 (1958), 1–65; cf. the perceptive critiques
of G. Severino Polica, ‘Cultura ecclesiastica e culture subalterne: Rileggendo alcuni saggi
di H. Grundmann’, Studi storici, 23 (1982), 137–66.
326 Antonio Sennis
would explain, at least in part, why archives that can be attributed to the
activity of lay individuals or institutions before the twelfth century seem
not to have survived in their original form – if they ever existed, some
might be tempted to say. However, as seen in previous chapters, there is
ample evidence that laypeople clearly did keep archives of documents of
some sort; it is the processes of archival transformation that have masked
the ways laypeople preserved and organized those documents.18 Similar
patterns will emerge in the following pages.
In revisiting the question of lay documentary practices in early medieval
Italy, I concentrate here on three specific issues: the ways in which doc-
uments were originally preserved by laypeople and by churches; the
destruction of charters as a sign that laypeople had a surprising famil-
iarity with written documents; and the preservation of documents as a
way of promoting group identity and awareness. Documents, in other
words, were not just preserved; they were used self-consciously by their
possessors.
When and where were documents preserved?

On 1 February 968, in the presence of and with the consent of her hus-
band Gratianus, the nobilissima foemina Theodora donated to Silvester,
abbot of the Roman monastery of SS. Cosma and Damianus in Mica
aurea, a field situated just outside the Porta Portuense, in the south-west
quadrant of the city’s suburbium. In doing so, Theodora and her husband
specified that they could not actually give the abbot the documents,
recent and old (monimina novas et vetustas), that attested their rights over
that field. ‘However,’ they said, ‘they seem to be somewhere in arcibus
nostris, but once we have found them we promise that we, or our heirs, will
give them to you and, if we refuse to do so, those documents will become
void and invalid, and will forever remain deprived of any authority.’19 It is
difficult to understand what Theodora actually meant with the term arx –
chests? a room? simply, ‘the secure place where charters are kept’? – but
this document does constitute explicit evidence of the fact that laypeo-
ple in tenth-century Italy preserved their documents in a specific place,
presumably in their place of residence. It is negative evidence, because
we cannot actually see Theodora retrieve the documents, yet it is quite
revealing.
Moreover, unlike Demetrius and his consortes (see below), Theodora
clearly refers to a complex of documents, some of which could have been
18 See above, Chapters 7, 9–11.

19 Il regesto sublacense del secolo XI, ed. L. Allodi and G. Levi (Rome, 1885), no. 52.
issued quite some time before. We may also notice that Theodora refers
to herself as nobilissima, claiming membership in a very specific and self-
distinctive social group in tenth-century Rome. We can therefore suppose
that her patrimony was quite substantial, as would have been the number
of documents related to its management. It is therefore possible to deduce
that a system of keeping documents could be in place here.20 We should,
however, be careful not to overestimate how systematically those written
records might have been kept and stored, as Theodora seems not to
have been sure where exactly they were. Nevertheless, this is not specific
to complexes of documents preserved by laypeople. At Subiaco in the
tenth century, charters and other documents seem to have been kept, all
together and without particular order, in chests and sacks, and finding a
specific one might prove an impossible endeavour, as a document from
July 993 reveals.21
The case of Theodora is not unique. The same thing happened, for
example, in the context of a sale of some land in the Sublacense region
in December 929, when the seller specified that he would keep all the
documents related to those properties safe (‘unde charte noue et uetuste
aput me meosque heredes reiacent ad conseruandas et saluas faciendas’).
And some years later, in November 935, in selling to the two brothers
John and Peter a fundus just outside the Porta Salaria in Rome, a certain
Peter specified that he would keep all the documents pertaining to that
property safe, and added that he would give them to the buyers whenever
they needed them.22
The practice of delivering the documents related to a piece of land
that was being sold is quite widespread among both laypeople and clerics
in central and southern Italy between the ninth and twelfth centuries.23
However, these transfers of documents, evidence of a mentality predis-
posed to the preservation of documents, did not necessarily happen in a
straightforward way. If the document included other properties excluded
from the sale, then the seller – lay or cleric – could declare that he or she
was not going to transfer the document as well, but promised to care for it
20 Cf. the case of the illustris vir Kaloleus, whose landed patrimony is the only one for
which management can be clearly documented (Il regesto sublacense, nos. 3, 93, 155).
21 Il regesto sublacense, no. 78: ‘post Georgi abbati appare libelli que˛ abuit. infra chartas
monasterii iactauit, et minime eam inuenire potuit’.
22 Il regesto sublacense, nos. 92, 61, respectively.
23 Other examples: BnF lat. 5411 (Liber instrumentorum de possessionibus, rebus, sive digni-
tatibus, quas Casauriense monasterium habuit, habet, vel habere debet), fol. 112r–v (4 June
879) (ed. A. Sennis, Strategie politiche, affermazioni dinastiche, centri di potere nella Mar-
sica medievale [Rome, 2002], appendix 1, doc. 29); Il regesto sublacense, no. 42 (20 June
967); Codex diplomaticus Cavensis, 10 vols. to date (Milan, 1873–93; Cava de’ Tirreni,
1984–), VIII 1321 (July 1061).
328 Antonio Sennis
and show it to the buyer, should it be necessary.24 This practice, in which

the careful keeping of the documents was a legal obligation, can be fur-
ther clarified by some evidence from the Codex Diplomaticus Cavensis. In
Salerno, as early as the mid ninth century and throughout the tenth and
eleventh centuries, documents were read and copied, either integrally or
in an abridged form, even long after their original date. For example, a
preceptum of 846 was read out and copied during a sale of 995; a series of
documents from 1019, 1027, 1036, 1043 and 1045 was copied in a sale
of 1069; in June 1070, twenty documents issued between 1028 and 1054
were read and copied as part of the division of patrimony between two
lay brothers; twenty-eight documents issued between 948 and 1040 were
read out and copied during a dispute between two laymen in 1054.25
The legal obligation, sometimes explicitly stated in a specific clause, to
safeguard the other party’s documents could even be part of the strategies
of litigation in the course of a dispute. This is what happened in 1067,
when Theophylact, abbot of the monastery of S. Nicola in Gallucanta
near Salerno, accused a certain Constantinus and his mother Porpora of
hiding and destroying some of the monastery’s documents. Constanti-
nus and Porpora were prepared to make the same accusation against the
abbot. The judge made both parties promise to return to court with all
the documents they possessed and swear not to have hidden or destroyed
any other document related to their opponents.26
Destroying documents
Destroying documents can indeed be considered a sign of a sophis-
ticated capability to manipulate written culture. Traditionally, we are
inclined to think that the need to safeguard documentary patrimony was
24 The usual formula was (with irrelevant variations): ‘Ipsa suprascripta venditionis charta
apud me meosque heredes reiacunt ad conservandum et salvas faciendum pro alia res
que nobis in eas continere videtur. Et pro hoc eam vobis minime contradere potui, sed
quandoque vobis vestrisque successoribus necesse fuerit pro vestra defensione semper
ostendere et demonstrare spondeo.’ E.g. Il regesto sublacense, nos. 39–40, 43, etc.
25 Codex diplomaticus Cavensis III 476 (a. 995); VII 1195 (a. 1054); IX 72 (a. 1069), 94
(a. 1070). Other examples: Salerno: II 338, 339, 368, 377, 395; V 725, 728, 731, 760,
797, 825; VII 1143, 1164, 1195; VIII 1265, 1281, 1292, 1297, 1316, 1321, 1322,
1325, 1361, 1373; IX 13, 53, 57, 94, 104. Capua: IV 621, 686; V 708, 779, 781, 845;
VI, 1048; VII 1072, 1083, 1087, 1093, 1115–16, 1216, 1221, 1230; VIII 1235, 1241,
1245, 1261, 1271, 1278, 1282, 1323, 1335, 1348, 1352, 1355–6, 1358–9, 1366, 1379,
1381; IX 19, 38, 45, 55, 71, 78, 85, 89, 96–9, 102, 108, 132.
26 Codex Diplomaticus Cavensis IX 36. The terms of the obligation (ten days to deliver
the documents, one hundred days during which the recipient of the documents had to
defend the other’s rights on the basis of those documents, a penalty of 300 golden solidi)
are fully specified, for example, in Codex Diplomaticus Cavensis VIII 1321.
peculiar to ecclesiastical institutions. Effectively, the prospect that the

loss or destruction of documents could have serious effects on a com-
munity’s rights, not only patrimonial, played a major role in the relations
that various churches established with the holders of sovereign authority.
Churches and monasteries could in fact explicitly request that the char-
ters granting or confirming privileges include a specific clause according
to which, in the event that their written titles were lost or destroyed,
they were allowed to attest to the legitimacy of their claims by oath.27
However, churches were not alone in doing so, as this clause was also
included in documents possessed by laypeople. This is confirmed, for
example, by a charter issued in 969 by Otto I in favour of his fidelis,
Ingo, in relation to his patrimony infra Regnum Italicum, and by another
conceded in 1038 by Conrad II to Raimbaldus, comes of Treviso.28 As
happened to churches, the rights of lay landowners were also protected
when their validity could not be established because written evidence
had been destroyed before the concession of a sovereign charter granting
or confirming properties. This is what can be inferred from some exam-
ples included in formula-books.29 At least in the Lombard principality of
Benevento, by the mid ninth century the procedures for the certification
of the status of property when the relevant documents (munimina) had
been destroyed had already been turned into a general norm, applied to
the whole of the society, as a law issued in 866 by Adelchis II indicates.30
Conversely, in Carolingian Italy documents that were considered
worthless or in some way irregular had to be destroyed by law. This
was the case for the cartulae obligationis related to men whose free sta-
tus had in fact been acknowledged, or for goods sold at too high or
too low a price due to necessity, or for donations whose value had been
overestimated.31 In the mid eleventh century, according to some formu-
las of the Cartularium Langobardicum, at the end of a dispute, the incisio
or capsatio chartarum was required if the documents presented had been
27 See, for example, a charter issued by Otto III in 997 for the abbey of Nonantola (MGH
DD Otto III, no. 237), another issued by Conrad II in 1026 for the monastery of Leno
(MGH DD Conrad II, no. 57), and a third conceded by Conrad II, in the same year,
to the monastery of S. Dionigi near Milan (MGH DD Conrad II, no. 58).
28 MGH DD Otto I, no. 371; MGH DD Conrad II, no. 277.
29 Among the numerous cases of intervention by a king or an emperor to remedy the
loss of documents due to fire or flood, whether intentional or through negligence, see,
for example, a charter issued in 967 by Otto II for the monastery of Subiaco (MGH
DD Otto II, no. 336). Needless to say, this could be an opportune chance for the
community to take advantage of the lack of written evidence in order to succeed in
making groundless claims (a parallel for England is analysed in M. T. Clanchy, From
Memory to Written Record: England, 1066–1307, 2nd edn (Oxford, 1993), 148–9; cf.
168); see also Brown, ‘Documents’.
30 MGH LL 4:211–12 (c. 7). 31 MGH Capit. 1, no. 88, cc. 1–3.
330 Antonio Sennis
judged to be false or, more simply, if their content had been superseded
by new agreements or events. Moreover, the traditio promissionis pro debito
required the charta venditionis with which someone had sold a piece of
land to a creditor as warranty on the debt to be capsata et taliata once the
debt had been fully repaid.32 This practice became more and more com-
mon after the beginning of the eleventh century, when the documentary
processes of private transactions started to include temporary acts that
remained valid for a very short period of time.33
Preservation and group dynamics

But how, and to what extent, can the preservation of charters be seen as
one element of lay group dynamics, as a way of promoting group identity
and awareness? On 17 August 942, in Rome, Alberic princeps, assisted
by some Roman judges, presided over a case brought before him by Leo,
abbot of the monastery of SS. Benedict and Scholastica in Subiaco. Leo
complained that a group of inhabitants of Tivoli were making unlawful
claims over the fundus Paternu, located some 20 miles from Rome. A
certain Demetrius, apparently the spokesman of the group, responded,
declaring that their claims were well grounded and that in fact they had a
charta to prove it. Only, he said, they did not have it right there with them
(‘Demetrius cum supradictis litigantibus dixerunt, “habemus chartam
sed non est hic”’). Alberic and the judges allowed Demetrius and his co-
litigants some days to produce that document in court; they failed to do
so and therefore lost the case.34 This example might refer to a procedure
quite common in Italian court cases between the 880s and the 1030s,
a procedure that evokes the old Ravenna gesta municipalia. One of the
litigants, usually the plaintiff, brought in front of the judges a document
attesting to his (or sometimes her) rights that was then validated by
the authority of the court. It is true that in this case it seems that the
defendants, and not the plaintiff, are the ones to produce the charter,
but this is probably due to the fact that the narratio of the document
implies that the men from Tivoli had actually started the dispute.35
What this case anyway clearly reveals is that laypeople could at least
claim that they were able to preserve documents attesting their rights.
The fact that, according to the document copied in the cartulary of
Subiaco, Demetrius came back before the judges empty-handed does not
undermine this impression. Perhaps he was just bluffing and the judges
32 MGH LL 4:597 (c. 9).

33 Bougard, La justice, 71n20, 319–29. 34 Il regesto sublacense, no. 155.
35 Ostensio chartae: Bougard, La justice, 319–30; above, 98–102, 249–54.
successfully called his bluff, but the tone of the document indicates that
it was perfectly credible that those lay litigants had the document stored
in a place from where it could be easily retrieved. Obviously, the use of
a formulaic tone does not mean that the underlying events were, or were
perceived to be, ordinary. However, and this is what interests me here, it
seems to indicate that what was being followed was a quite standardized,
evidently common procedure.
This document, however, also tells us something more. Demetrius was
not alone, but was acting together with his consortes. Consequently this,
and presumably every other, document attesting to the prerogatives of
the group had to be handled in a way that guaranteed the rights of each
member. Apart from their obvious patrimonial value, for consortes such
as Demetrius and his co-litigants, documents could indeed be a way to
define and emphasize their identity and solidarity as a group. This is con-
firmed, for example, albeit for a slightly later period, by a memoratorium
transmitted by Gregory of Catino that relates how in 1066 (or 1077) a
certain Teuto, an inhabitant of the castellum of Bibaro, went to Berar-
dus, abbot of Farfa, begging him to return to him the document with
which he and his kin had donated Bibaro to the monastery. Apparently,
Berardus, moved to pity and acting ‘sine consilio vel assensu congre-
gationis’ and ‘contra legem et contra libertatem huius monasterii’, gave
the cartula to Teuto. Once he was back in Bibaro, Teuto summoned his
kin and showed them the document. At this point, glad to have back
the charter of donation, they tore it into pieces and burned it – but
the fire went out of control and, iudicio Dei, the whole of Bibaro was
destroyed.36 It is clear that, in this case, the community’s narrative of the
events set the malicious destruction of this document in a framework of
actions – the illicit (and preposterous) request to have the document
back; the rash (and unbecoming) submissiveness demonstrated by the
abbot; the sacrilegious (and ritually savage) rejection of the monastery’s
full rights – whose grand finale confirmed the community’s absolute
supremacy over the neighbouring society. Nevertheless, what is revealing
is that, in this piece of propaganda, the adversaries of the monastery are
shown as acting together as a group with respect to the document. Teuto
brought it back and showed it to his kin, and every member of the group
took part in its destruction, each tearing it, holding one piece and setting
it on fire.37
36 RF V 1299; see A. Sennis, ‘“Omnia tollit aetas et cuncta tollit oblivio”: Ricordi smarriti
e memorie costruite nei monasteri altomedievali’, Bullettino dell’Istituto storico italiano
per il medio evo, 106 (2004), 93–135, at 107–8.
37 Rituals and symbolic values of the destruction of charters: A. Sennis, ‘“Omnia tol-
lit . . . ”’, 101–14.
332 Antonio Sennis
On the other hand, the configuration and articulation of these groups

did not prevent each of their members from having patrimonial relations
autonomous from, and external to, the ones maintained by the group
as a whole.38 Therefore, a single layperson could handle, and have to
preserve, documents related to him or her both as an individual and
as a member of a kin group. In other words, a system of preservation
such as the one alluded to by Demetrius had to be structured enough
to ensure the retraceability of each document according to its functions.
This was obviously true when laypeople, as individuals or as members of
a group, were involved in a dispute with a church. However, the use of
documents could clearly take place outside the sphere of influence of a
main monastery, as is demonstrated, for example, by the documentary
dossiers from the cartulary of San Clemente a Casauria, dating between
the ninth and the twelfth centuries and referring to donations, sales and
exchanges in which the monastery did not take part, as discussed in the
chapter by Costambeys.39
If laypeople, as I have argued, used and preserved documents in the
early Middle Ages in the ways I have described, why have the vast majority
subsequently disappeared while documents involving clerics and monks
have survived? The preservation of documents by religious institutions –
even documents that have lost any juridical value – may be explained by
the theoretical ban on the alienation of the ecclesiastical patrimony of a
church, which would have provided an ideological restraint on the alien-
ation (or loss by other means) of documents. The contempt (and it is just
one case among many) with which Gregory of Catino commented upon
Abbot Berardus’ decision to give the document back to the inhabitants
of Bibaro could be a sign of this. Laypeople were not subject to such
restraints. More importantly, however, churches and monasteries would
have had the tendency to preserve documents because they operated
with an institutional mindset. Gregory of Catino’s own documents sur-
vived not because he was a monk per se, but because they were preserved
in the archives of the institution to which he belonged: the monastery of
Farfa.40 An institutional mindset of this sort was largely absent among
the laity in the early Middle Ages. Only in the late eleventh and twelfth
centuries do we see lay institutions become generally present, and this
38 E.g. L. Feller, Les Abruzzes médiévales: Territoire, économie et société en Italie centrale du
IXe au XIe siècle (Rome, 1998), 505–6. Social dimensions of the construction of landed
patrimonies: Rosenwein, Saint Peter (Ithaca, NY, 1989).
39 BnF lat. 5411, fols. 6r–72v; above, Chapter 9.
40 A number of documents related to Gregory’s family were copied by him in the Regestum
(RF IV 948–75). They were probably attached to RF IV 948, which is a sale by Gregory’s
parents to Farfa.
is precisely when the laity regained a certain degree of control over the
process of record-keeping.
Moreover, throughout the early Middle Ages, lay society in central
Italy was characterized by a high degree of mobility. The highest strata
did not constitute a compact order but, rather, a fluid group whose mem-
bers were characterized by their wealth and power. The political fortune
of a family could rapidly rise or be effaced, as could its patrimony. The
patrimonies of lay elites became more stable in the twelfth century, when
the number of prominent families decreased as they were integrated into
larger political entities (namely the papal territories and the Norman,
then Angevin, kingdom). This process had repercussions on lay docu-
mentary complexes while, at the same time, the publica fides that notaries
had by then generally acquired made it possible, indeed necessary, for
the documents of the laity to be preserved in the form of inbreviaturae.41
We might add that it is only in the late eleventh and twelfth centuries
that lay institutions return to general prominence in the region, and
it is only then that the laity regain a certain control over the process
of record-keeping. Thus the relevant comparison is not between clergy
and laity, but between institutionalized and non-institutionalized juridical
persons.
Laypeople in the early Middle Ages were not as successful at preserv-
ing documents; this was not because they were not part of the clergy,
but because they were not grouped into institutions. By the time lay
institutions emerged, however, churches had strongly established the
institutional model of reference for medieval Italy. Besides, one of the
hallmarks of the ideology and self-representation of medieval monaster-
ies, the capability to control the dynamics through which written doc-
uments were produced and preserved by society, became an even more
central element of their rhetoric after the mid eleventh century, when a
new sense of historical development and therefore a different relationship
between the present and the past began to emerge throughout Europe.42
I have argued elsewhere that, towards the end of the Middle Ages,
at least some intellectuals outside the cloister considered monasteries as
the custodians of the memorial patrimony of society. Moreover, monastic
ideals became almost inextricably associated, by outside observers, with
the fundamental function of preserving the material, written traces of the
41 M. Lenzi, La terra e il potere: Gestione della proprietà e rapporti economico-sociali a Roma

tra alto e basso medioevo (secoli X–XII) (Rome, 2000), 103.
42 See above, Chapter 11; G. Constable, ‘Past and Present in the Eleventh and Twelfth
Centuries: Perceptions of Time and Change’, in L’Europa dei secoli XI e XII fra novità
e tradizione: Atti della decima Settimana internazionale di studio (Mendola 25–29 agosto
1986) (Milan, 1989), 135–70.
334 Antonio Sennis
past.43 We can now see that this was true not only for the very specific
type of written memories – that is, the most precious legacies of classical
culture, that I was discussing in that case.
Thus, in the first phase of their existence, Italian communes used
local monasteries and convents to archive their documents. And this
remained the case even after lay status had become a prerequisite for
office holding.44 For example, on 16 October 1244, magister Acto Ben-
tivolii, syndicus of the commune of Cingoli, in the Marches, deposited
four copies of the agreement that the commune had reached with the
lords of Civitella. The copies were deposited as follows: one ‘apud sanc-
tum Franciscum, locum fratrum minorum; the second at the church of
S. Esuperazio’; the third ‘apud hospitale Spreti’, where it was received
by the abbess, Catherine; and the fourth ‘apud dominum Rainaldum
Glodii’.45 Similarly, at Brescia, in 1252, the ‘notae venditionum fac-
tarum sub extimatoribus seu consulibus iustitiae’ were considered valid
only if the corresponding charta ‘inventa fuerit scripta in libris’ which
were given ‘in deposito domino archipresbitero Brixiensi’.46 At Verona,
the originals of imperial charters issued by Frederick I were kept in the
sacristy of the Friars Minor, while copies were kept in the sacristy of the
commune and in that of the capitanei.47
The function of a church as a place where documents, both of an
institution such as the commune and of individuals, could be put ‘in
salvamento’ is also evident from a statute from Vicenza which mentions
ten books of copies of charters: ‘tempore compromissi perfidi Ecelini
de Romano, qui libri sunt positi in salvamento in loco Sancte Corone
in custodia Fratrum Predicatorum’.48 Lay institutions and individuals
in Italy continued to use churches to store their most precious docu-
ments well into the late medieval period. When Florentines had to escape
the pestilence or political turmoil, they often deposited their ‘scritture
d’importanza’ in monasteries.49
43 A. Sennis, ‘The Power of Time: Looking at the Past in Medieval Monasteries’, in A.

Müller and K. Stöber (eds.), Self-Representation of Medieval Religious Communities: The
British Isles in Context (Berlin, 2009), 307–26.
44 P. Torelli, Studi e richerche di diplomatica comunale, II (Mantova, 1915), 37.
45 Cingoli, Archivio del comune, Pergamena 19, II; see also E. Sebastiani, ‘Genesi, con-
cetto e natura giuridica degli archivi di Stato in Italia’, Rivista italiana per le scienze
giuridiche, 37 (1904), 44.
46 Statuti di Brescia, ed. F. Odorici, Monumenta Historiae Patriae, XVI (Turin, 1876), III,
XXIII, 1726–7.
47 A. M. Allen, A History of Verona (London, 1910), appendix I, CCXLVIII, p. 131.
48 Statuti del comune di Vicenza MCCLXIV (Venice, 1886), 86.
49 This is what Michele di Bindaccio Cerchi does in 1363 (Florence, Archivio di Stato,
Dono Canigiani Cerchi, 310, fol. 6v) and Luca da Panzano in 1424 (Florence, Archivio
di Stato, Strozziane II, 9, fols. 26v–27).
However, at such a late date, placing documents in the library or in the

archive of a monastery was clearly an act meant to emphasize the memo-
rialistic dimension of those written records, rather than to safeguard
their immediately practical value. So, at Padua, only the chest containing
‘omnia instrumenta, privilleia et rationes et statuta, que comune padue
non operatur’ was kept by the local Franciscans, and the podestà and the
anziani kept its keys and a register of its contents. Similarly, when, still at
Padua, the statutes were updated and four copies of the new version were
distributed among four offices of the commune, the four outdated copies
of the previous version of the statutes were sent to be kept ‘unum penes
monasterium sancti benedicti, aliud penes monasterium sancti iohannis
de verdaria, tercium penes sanctam mariam de vancio, et quartum penes
monaterium sancte marie de porcilia’, possibly as a sort of documentary
monument.50
It was, then, when new (mainly lay) institutions became mature and
formalized enough to frame the complex of documentary practices in
the society around them, that the role of churches and monasteries as
overseers of the daily elements of those practices began to decline.
50 Libri statutorum communis Padue 4.9/1133 (a. 1265), 4.15/1178 (a. 1263) (ed. A. Gloria
[Padua, 1873], pp. 344, 360).
13 Archives and lay documentary practice
in the Anglo-Saxon world
Charles Insley
As previous chapters have traced continental documentary practices that moved

fairly seamlessly between lay and ecclesiastical settings, it is instructive to
examine the evidence from Anglo-Saxon England from this new perspective.
Although it preserves no lay archive as such, England offers evidence substan-
tial enough to allow us to detect some of the outlines of a cohesive documentary
culture that included the laity. Exceptional survivals like the Fonthill Letter
point to the intensive use of documents, as well as to historical developments
that have shaped the surviving corpus, which is dominated by royal diplomas
preserved by major monastic houses and bishoprics. Traces of the operation of
documents in property disputes, coupled with the evidence for a lively traffic in
land, indicate how common documents must have been; they also attest to a
culture in which laypeople routinely handled and kept them. The elite probably
had quite sophisticated archival methods. The forces shaping long-term preser-
vation, especially after 1066, tended to preserve some types of document – such
as Latin royal diplomas for major monasteries – and discouraged the retention
of others – in particular, vernacular documents like leases – that must once
have proliferated.
In the early tenth century, probably in its second decade, a West Saxon
ealdorman, Ordlaf, sat down and either wrote or dictated an account
of the recent wrangling over an estate at Fonthill, in Wiltshire, and the
crimes of a man called Helmstan.1 The ‘Fonthill Letter’, as it has become
1 Canterbury, Cathedral Archive, Dean and Chapter, Chartae Antiquae C 1282; P. H.

Sawyer, Anglo-Saxon Charters: An Annotated List and Bibliography (London, 1968), since
updated online by Susan Kelly and Sean Miller: www.trin.cam.ac.uk/chartwww/esawyer.
99/eSawyer2.html (hereafter S) 1445. The Fonthill Letter is translated in D. Whitelock,
English Historical Documents. I. c. 500–1042, 2nd edn (London, 1979) (hereafter EHD),
no. 102; much the best modern edition and commentary is that of Simon Keynes (S.
Keynes, ‘The Fonthill Letter’, in M. Korhammer with K. Reichl and H. Sauer (eds.),
Words, Texts and Manuscripts: Studies in Anglo-Saxon Culture Presented to Helmut Gneuss
[Cambridge, 1992], 53–97); a new edition is included in The Charters of Christ Church,
Canterbury, ed. N. P. Brooks and S. E. Kelly (Oxford, forthcoming), no. 104. Keynes
(‘Fonthill’, 56) argues for a date in the second decade of the reign of Edward the Elder;
336
The Anglo-Saxon world 337
known, now preserved in the archives of the Dean and Chapter of Canter-
bury Cathedral, is one of the great documentary treasures of the Anglo-
Saxon period. As one might expect, it has seen the expenditure of a great
deal of scholarly ink, for it is one of the best witnesses to the nuts-and-
bolts functioning of criminal justice in Anglo-Saxon England, and so of
quite exceptional importance in the effort to understand its culture of
document use.2
In the simplest terms, the letter is an account of the recent history of
the Fonthill estate by a third party to a litigation episode, in this case
between a man called Æthelhelm Higa and the bishop of Winchester.
The waters are muddied somewhat by the fact that Ordlaf, the probable
author of the letter, had once owned the Fonthill estate before giving it to
the bishop, and was the godfather of the Helmstan whose kleptomaniac
behaviour had first sparked off the dispute at some point in the late 890s.3
However, interesting as the legal exegesis of the document is, it does not
directly concern us here, except insofar as it hints that documents were
part and parcel of legal processes in late ninth- and tenth-century Eng-
land, even those exclusively involving laymen, and so must have once
been very common indeed.4 The Fonthill Letter encapsulates many of
the issues raised by thinking about archives and lay documentary practice
in Anglo-Saxon England. First, there is the lay role in production, since
the document was probably produced by a layman and almost certainly
in a lay aristocratic household; second, since the interactions and pro-
cesses it describes largely involved lay actors, it suggests that we may be
able to access some of the functions of documents in lay society; third,
it was written in the vernacular and forms part of a raft of vernacu-
lar documentation dealing with legal process and property ownership,
pointing to language as a factor affecting document use and preserva-
tion; fourth, since it links documents, legal process and conveyancing,
it highlights the relationship between documents and the operation of
the law; and fifth, since it was preserved by a church, but one with no
connection to either the property concerned or the events described,
it testifies to the difficulty of discerning how and why documents were
kept.
see also P. Wormald, ‘A Handlist of Anglo-Saxon Lawsuits’, Anglo-Saxon England, 17

(1988), 247–81, nos. 23–26a; P. Wormald, The Making of English Law: King Alfred to
the Twelfth Century. I. Legislation and Its Limits (Oxford, 1999), 144–8; N. P. Brooks,
‘The Fonthill Letter, Ealdorman Ordlaf and Anglo-Saxon Law in Practice’, in S. Baxter
et al. (eds.), Early Medieval Studies in Memory of Patrick Wormald (Aldershot, 2007),
301–17.
2 Keynes, ‘Fonthill’, 54.
3 See below, 346 and n33, for a discussion of the authorship of the letter.
4 As Brooks (‘The Fonthill Letter’, 306) has suggested.
338 Charles Insley
Literacy and lay documentary culture

In surveying and contextualizing the historiographical landscape of lay lit-
eracy in Anglo-Saxon England, Kathryn Lowe has argued for a ‘change
in attitude to the written document that was neither confined to the
church or the state nor, indeed, to the centuries following the Norman
Conquest’. This change, she suggests, is best exemplified by the chiro-
graph, the name given to a form of charter, almost certainly originating
in Anglo-Saxon England, in which two or three copies of the same doc-
ument were made from the same sheet of parchment for safe keeping by
interested parties, and where these counterparts could be easily matched
and checked.5 Lowe suggests that the key to understanding the role of
the chirograph was that it was ‘in form and function so attractive to
the layman’, seeming to respond to an audience that was not partic-
ularly literate but was interested in the possession and deployment of
documents as part of the legal and conveyancing process. This impres-
sion of familiarity with documents is reinforced by Catherine Cubitt’s
reading of the eleventh-century Old English translation of the legend of
the Seven Sleepers as showing that the late Anglo-Saxon judicial system
made frequent reference to written texts. Cubitt is concerned here not
simply to accept a degree of pragmatic literacy in late Anglo-Saxon lay
society but also to argue that that pragmatic literacy might have reached
some way down the social scale.6 These persuasive demonstrations of the
deep social penetration of documents contrast with earlier work that had
regarded Anglo-Saxon England, right up until 1066, as what anthropol-
ogists in the mid twentieth century referred to as a ‘traditional society’,
featuring only restricted literacy.7 Such views had already been effectively
countered by research indicating that literate practices, including the use
of documents, were embedded within the norms and assumptions of
Anglo-Saxon lay society.8 The work of Lowe and Cubitt thus reinforces
5 K. A. Lowe, ‘Lay Literacy in Anglo-Saxon England and the Development of the Chi-
rograph’, in P. Pulsiano and E. M. Treharne (eds.), Anglo-Saxon Manuscripts and Their
Heritage (Basingstoke, 1998), 161–204; quotations at 180.
6 C. Cubitt, ‘“As the Lawbook Teaches”: Reeves, Lawbooks and Urban Life in the Old
English Legend of the Seven Sleepers’, English Historical Review, 124 (2009), 1021–49,
esp. 1048.
7 P. Wormald, ‘The Uses of Literacy in Anglo-Saxon England and Its Neighbours’, TRHS,
5th ser., 27 (1977), 95–114, esp. 96–7; Wormald’s conclusions are amplified in his The
Making of English Law, 480–3; M. T. Clanchy, From Memory to Written Record: England,
1066–1307, 2nd edn (Oxford, 1993).
8 S. E. Kelly, ‘Anglo-Saxon Lay Society and the Written Word’, in Uses of Literacy, 36–62;
see also S. D. Keynes, ‘Royal Government and the Written Word in Late Anglo-Saxon
England’, in Uses of Literacy, 226–57, and H. Collingridge, Anglo-Saxon Literacy at
Worcester: The Evidence of the Oswald Leases, unpublished M.Phil. thesis, University of
York (2002), 3–8, and, on Worcester’s archive, F. Tinti, ‘From Episcopal Conception to
Monastic Compilation: Hemming’s Cartulary in Context’, EME, 11 (2002), 233–61.
the strong impression that laypeople, perhaps even at a relatively humble

level, valued documents, used them in their business and, presumably,
therefore, stored them in ways that meant that they could find them
easily.9 In this context, the Fonthill Letter shows how important infor-
mal lay document use may have been in a judicial process that English
historiography of the early Middle Ages has often seen as underlying the
importance of the state and the institutions of government.10
But the intention in the following discussion is not to reheat further
the arguments over the importance and existence of socially or admin-
istratively significant lay literacy in England, but instead to look at the
ways in which documents were used, stored and preserved by laymen as
part of the relationship between lay society and document use – in other
words, the documentary culture of Anglo-Saxon lay society. This is not to
say that discussions about lay literacy are not important – quite the
reverse. The apparent contradiction between the Fonthill Letter – a
vernacular document written by a layman – and, for example, Patrick
Wormald’s rejection of a direct relationship between vernacular docu-
mentation and lay literacy shows that this letter may be very significant.
Nevertheless, the a priori assumption adopted in this discussion is that a
layman did not necessarily need to be able to read a document to under-
stand its importance and place value on it as a piece of writing as well as
an artefact with symbolic power.
It is perhaps worth qualifying at the outset some of the terms and
concepts that have appeared elsewhere in this volume in the specifically
Anglo-Saxon case. For the purposes of the present discussion, given that
the vast majority, if not the entirety, of the extant corpus of Anglo-Saxon
documentary material survives in an ecclesiastical context, ‘lay archive’
is understood very loosely as referring to a collection of documents that
had been in lay hands at some point during their earlier history. While we
cannot point to surviving ‘lay archives’ as such, it is possible to uncover
their footprints, either in the surviving collections of documents in
ecclesiastical hands, or in the processes in which documents were
9 The use of documents in legal procedure has been especially highlighted, in anglophone
scholarship, by the two important collections of essays edited by Wendy Davies and
Paul Fouracre: Settlement of Disputes, and Property and Power in the Early Middle Ages
(Cambridge, 1995). Patrick Wormald’s contribution to the former (‘Charters, Law and
the Settlement of Disputes in Anglo-Saxon England’, pp. 149–68) is invaluable when
considering the relationship between law, disputes and the deployment of charters; see
also P. Wormald, ‘A Handlist’.
10 J. Campbell, ‘Observations on English Government from the Tenth to the Twelfth
Century’, TRHS, 5th ser., 25 (1975), 39–54; Campbell, ‘Some Agents and Agencies of
the Late Anglo-Saxon State’, in J. C. Holt (ed.), Domesday Studies (Woodbridge, 1987),
201–18; Campbell, ‘The Late Anglo-Saxon State: A Maximum View’, Proceedings of
the British Academy, 87 (1994), 39–65; Keynes, ‘Royal Government and the Written
Word’; Wormald, ‘Uses of Literacy’.
340 Charles Insley
deployed. Similarly, very few surviving Anglo-Saxon documents are gen-

uinely ‘lay documents’, in the sense that there was no clerical involvement
in their production or use – the Fonthill Letter is, as will be argued, per-
haps the only surviving document where we can have some confidence of
lay authorship, but, even then, the document was probably produced for
use in a legal process where one of the parties was a bishop, and ultimately
preserved in an ecclesiastical archive. Costambeys and Kosto adopt a def-
inition whereby neither party to a document is explicitly identified as a
cleric.11 This is perhaps more helpful, since a significant number of our
surviving Anglo-Saxon documents do not explicitly involve the Church
as a party to the transaction or occasion recorded in the document, even
if the Church was the ultimate repository of the relevant documentation.
The evidence
England presents us with a corpus of a little over 1,600 documents that
survive from before 1066, along with references to around 270 now-lost
documents. In overall numbers, it fares badly in comparison with con-
tinental Europe, including Francia, physically and culturally the closest
part of the continent. Even allowing for the depredations inflicted on
ecclesiastical archives by the Henrician reformation of the mid sixteenth
century and the civil wars of the mid seventeenth, the relatively small
size of this corpus of documents may tell us something important about
the extent to which documentary use was or was not a culturally embed-
ded practice in Anglo-Saxon England. Certainly, the assumption of most
scholars is that whatever level of documentary literacy there was in the
eastern and southern parts of the Roman province of Britain it had com-
pletely vanished by the sixth century and was perforce reintroduced by the
Augustinian mission in the 590s. It is possible that elite society in Kent, in
some sense part of the wider cultural hegemony wielded by the Merovin-
gian kings,12 had some awareness of the potential and power of document
use, but the overwhelming likelihood remains that, in what became the
territories of Anglo-Saxon England, late Roman documentary culture,
even in the ossified forms in which we see it in Merovingian Francia,
had ceased to exist. This is an important point: in this respect England
was qualitatively different from the rest of western Europe, where some
bedrock of documentary culture seems to have survived the political and
11 See above, 238, 259–60, 262. Cf. above, 3.

12 I. N. Wood, The Merovingian North Sea (Ålingsas, 1983), 12–18; Kelly, ‘Anglo-Saxon
Lay Society’, 41.
social traumas of the fifth and sixth centuries. This fissure in documen-
tary culture may help to explain why we simply have fewer surviving
documents from England before, say, 1000 than from anywhere else in
the former Western Roman Empire. It may also account for the distinc-
tive form much of that documentary evidence takes. Both Pierre Chaplais
and Susan Kelly make the point that although the Anglo-Saxon solemn
diploma, our most plentiful surviving document type, in its basics con-
forms to wider continental diplomatic norms, in important respects – in
particular the absence of any internal or external validating structure –
it remains highly unusual, a distinctiveness born of the circumstances
of the reintroduction of documentary culture to southern and eastern
England.13
Of our 1,600 or so surviving documents, a large majority, around
1,060, are in the form of Latin royal diplomas. A further 100 are Old
English royal writs, largely dating from the reign of Edward the Confessor
(1042–66). There are 80 documents that Peter Sawyer classified as grants
by the laity, mostly in Latin with some wholly vernacular documents and
many that contain both.14 Another 150 documents are grants by bish-
ops, while the remaining documents are an eclectic mixture of wills,
leases, detached boundary clauses and other ‘miscellaneous’ documents,
as Sawyer called them. These include records of law cases, agreements,
lists of sureties and other notifications. The documentary record, then,
is skewed towards the survival of the most prestigious documents, royal
diplomas. Royal diplomas make up the major part of the vast majority of
surviving Anglo-Saxon archives.15 In an Anglo-Saxon context this pre-
ponderance is to some extent understandable, since diplomas functioned
as title-deeds to property. But the contrast with the surviving bodies of
charters from continental Europe, in which ‘private’ charters consistently
13 P. Chaplais, ‘Who Introduced Charters to England? The Case for Augustine’, in F.

Ranger (ed.), Prisca Munimenta: Studies in Archival and Administrative History Presented
to Dr A. E. J. Hollaender (London, 1973), 88–107, at 100–1; Kelly, ‘Anglo-Saxon
Lay Society’, 40–5; P. Wormald, Bede and the Conversion of England: The Charter Evi-
dence (Newcastle upon Tyne, 1984), 5, 11–15. D. Howlett, Sealed from Within: Self-
Authenticating Insular Charters (Dublin, 1999), 13–59, has controversially argued that
the biblically derived numerical structure he has detected in the word ordering of a hand-
ful of early Kentish diplomas in effect constitutes an authenticating mechanism. For a
robust rebuttal of Howlett’s thesis, see, inter alia, H. McKee and J. McKee, ‘Chance or
Design? David Howlett’s Insular Inscriptions and the Problem of Coincidence’, Cam-
brian Medieval Celtic Studies, 51 (2006), 83–102, and ‘The Problem of Coincidence: A
Brief Final Response’, Cambrian Medieval Celtic Studies, 56 (2008), 97–99.
14 S.
15 For example, of the thirty-two extant pre-Conquest documents that survive as part of
the bishop of Exeter’s archive, twenty-seven are diplomas. In contrast, the surviving
archives of Worcester and Bury St Edmunds are dominated by vernacular material,
mostly leases.
342 Charles Insley
form a higher proportion than royal issues, is worth noting. In neither the
English nor the continental cases are patterns of survival likely to reflect
original levels of production at all closely. Something of what we have
lost from England can be glimpsed in fragments: a layer of documenta-
tion concerning sales, leases and transfers of property and rights, often
in the vernacular, of which only a few examples survive.16 It should not
be forgotten that one of the effects of the Norman Conquest of England
was to eliminate Old English as a language of administration and busi-
ness, and one can imagine a situation as early as the mid twelfth century
in which documentation in Old English may have been perceived, for a
variety of reasons, to be irrelevant. Indeed, at some point in the twelfth or
thirteenth century, a Canterbury Christ Church archivist wrote ‘inutile’
on the dorse of the Fonthill Letter, a fact which makes its survival all the
more remarkable.17 In terms of our discussion about lay documentary
culture and practice, this is an important caveat, since the documents
that have survived are the most ‘ecclesiastical’ in their form; that is, they
are often written in difficult and sometimes downright obscure Latin
and shot through with religious symbolism and imagery, whose force, as
Chaplais noted, is entirely spiritual.18
There are two other points to be made in conjunction with the evidence
and the pattern of its survival. The first is that a significant proportion
of it, somewhere between a third and a half of the surviving diplomas,
is of questionable authenticity. The use of the word ‘authenticity’ is, of
course, fraught with problems for the Anglo-Saxon diplomatist. Since the
vast majority of documents lack the sort of marks of authentication we
take for granted on the continent, such as autograph signa or subscrip-
tions, it is unclear to what extent any Anglo-Saxon document, certainly
any diploma, can be regarded as strictly ‘authentic’.19 The distinction
between an original document and a single-sheet copy then becomes
very hazy and may often hinge on detailed palaeographical analysis and
dating of scripts. The reliability of the information in any given docu-
ment becomes a test of patient textual and archival detective work. The
phrase ‘questionable authenticity’ can therefore cover a multitude of sins,
16 Much of the Worcester material exists in the only surviving pre-Conquest English car-
tulary, London, British Library, Cotton Tiberius A.VIII.
17 It is possible that the uselessness of the letter was related to the fact that it concerned
property in which Christ Church had no interest, rather than the language in which it
was written.
18 P. Chaplais, ‘The Origin and Authenticity of the Royal Anglo-Saxon Diploma’, Journal
of the Society of Archivists, 3 (1965), 48–61, at 56.
19 Wormald, Bede, 5. For the absence of authenticating marks, see Kelly, ‘Anglo-Saxon
Lay Society’, 42–3.
ranging from documents that have been subtly altered but remain basi-
cally as produced, to elaborate but complete forgeries. This is an especial
problem for early Anglo-Saxon diplomas (pre-c. 900), where perhaps half
of the surviving corpus contains information that is hard to corroborate
and might be regarded as untrustworthy, and where there has been a
significant degree of later interference – often clerical – in the surviving
documents.
The second point concerns the chronological distribution of our mate-
rial, in particular the diplomas. The earliest surviving original diploma
dates from 679, although, as Chaplais, Kelly and Wormald have all
argued, given the already idiosyncratic diplomatic of these early charters,
they were unlikely to have been the first to have been written in England,
and it is possible that diplomas were produced much earlier in the sev-
enth century.20 It is tempting to posit a situation not unlike Merovingian
Francia, where some of these early diplomas may have been written on
papyrus, before a switch to the much more robust and readily available –
in northern Europe – parchment.21
Something like 350 diplomas survive dated from the seventh, eighth
and ninth centuries. A further 540 surviving diplomas date from the
tenth century, with 160 from the period from 1000 to 1066. A similar,
if less marked, shift in the pattern of survival can also be detected in
the non-diploma material: writs, wills, bounds, lay charters, agreements
and other memoranda. This jump in the number of surviving diplo-
mas after 900 requires some thinking about, especially in relation to the
notion of a growing dominance of monastic archives among centres of
document preservation.22 The simplest explanation for this great leap in
survival is that many more diplomas were produced after around 925
than before. This is not inherently implausible. The tenth century saw
a massive kingdom-building project launched by the West Saxon kings
and their successors as kings of the English, and it seems likely that this
political and military activity was attended by significant changes in land-
holding and landownership that required a large rise in the production
of diplomas. For example, the political tension between the half-brothers
Eadwig and Edgar in the years 955–7 saw the issuing of a large num-
ber of diplomas in the year 956, of which some eighty-five survive, as
Eadwig presumably sought to use the granting of bookland, or heritable
20 Chaplais, ‘Who Introduced Charters into England?’; Kelly, ‘Anglo-Saxon Lay Society’,
40–3; Wormald, Bede, 11–15.
21 The majority of scholarly opinion is sceptical about the existence of papyrus diplomas
in England in the earlier seventh century, but it remains a possibility: Wormald, Bede,
11–15; Kelly, ‘Anglo-Saxon Lay Society’, 42.
22 See the similar pattern in Iberia: Kosto, ‘Laymen’; Davies, Acts of Giving.
344 Charles Insley
land held by charter, as a means of securing backing from an English

political and clerical establishment who were by no means universally
enthusiastic about him.23 As we also see elsewhere in this volume, polit-
ical conflict here influenced document production and preservation: the
coincidence of a complete lacuna in the run of documents for the years
910–24 with Edward the Elder’s campaigns of conquest in the Danelaw
might be another example.24 The preponderance of royal diplomas in the
Anglo-Saxon corpus of charters makes it inherently likely that changes
in the activity and ambition of royal government might have affected the
overall pattern of survivals.25
The influence of royal action on charter survival can be linked directly
to the principal institutions in which documents survived: the major
monasteries of the period. Since the tenth-century kingdom-building
project in England also included extensive foundation, refoundation and
patronage of monastic communities, the dominance of monastic centres
for the preservation of documents should not be surprising. Increased
activity, though, is not the only explanation for the great increase in
post-900 survival. We might also point towards shifts in tenurial practice
and perhaps towards a more aggressive attitude to ecclesiastical property
fostered by the new monastic institutions founded or refounded in the
tenth century. The notices of land transactions concerning Ely, compiled
as part of the Libellus Æthelwoldi and copied into Book Two of the Liber
Eliensis, are testament to the ruthlessness with which these tenth-century
refoundations husbanded and expanded their properties, especially in
terms of securing their title-deeds.26 This is a question to which we will
return.
Fonthill and beyond: lay archives and legal process

To return to the document introduced at the beginning of this chapter,
it is worth spending a little time with the Fonthill Letter and drawing
23 P. Stafford, Unification and Conquest: A Political and Social History of England in the Tenth
and Eleventh Centuries (London, 1989), 47–50; S. Keynes, The Diplomas of King Æthelred
‘The Unready’ 978–1016 (Cambridge, 1980), 48–69.
24 Compare Everett, above, 80–2, on the context of the gesta municipalia of Ravenna.
For the lacuna in 910–24, see D. Dumville, ‘Between Alfred the Great and Edgar the
Peacemaker: Æthelstan, First King of England’, in Dumville, Wessex and England from
Alfred to Edgar (Woodbridge, 1992), 141–71, at 151–3.
25 C. Insley, ‘Where Did All the Charters Go? Anglo-Saxon Charters and the New Politics
of the Eleventh Century’, Anglo-Norman Studies, 24 (2002), 109–27; R. Fleming, Kings
and Lords in Conquest England (Cambridge, 1991), 21–52.
26 Liber Eliensis, ed. E. O. Blake (London, 1962), p. xxxiv; see also Liber Eliensis 2, esp. cc.
25, 27 (ed. Blake, 63–236; trans. J. Fairweather [Woodbridge, 2005], 84–284).
out some of its points of interest.27 The letter is addressed to King

Edward the Elder (899–924) and describes a dispute over a five-hide
estate at Fonthill Magna in Wiltshire. The main actors in the story are a
light-fingered king’s thegn, Helmstan; another thegn, Æthelhelm Higa;
and Ealdorman Ordlaf, the probable author of the letter.28 The dispute
hinges on the possession of Fonthill, initially claimed by Æthelhelm Higa
against Helmstan. Helmstan, although clearly a man of some property
and well connected – as well as Fonthill, the letter refers to a much larger
estate belonging to Helmstan at neighbouring Tisbury – was also a rather
incorrigible thief. The letter opens with a description of Helmstan’s theft
of a sword belt and the subsequent judgement made against him; the theft
of a belt was not, perhaps, a trivial matter, given the status of sword belts
as marks of rank and status.29 This judgement seemed to have provoked
Æthelhelm Higa’s suit: as Keynes suggests, Helmstan’s law-worthy status
may have been affected by the theft suit, thus making him vulnerable to
further suits.30 Helmstan’s actions as a thief also arguably put him outside
the law, since theft was regarded effectively as a breach of faith.31 At this
point, Helmstan, who presumably needed all the friends he could lay his
hands on, sought the support of his godfather, Ealdorman Ordlaf, who
interceded with King Alfred on Helmstan’s behalf. This intercession,
Keynes suggests, was to secure Helmstan the right to bring an oath
to prove his title to the estate; in a meeting at Wardour, the case was
heard and Helmstan adjudged to be in the right. The price of Ordlaf’s
intercession was that Helmstan then gave the land at Fonthill to Ordlaf,
with a life interest in it for himself.
There matters should have rested, except for Helmstan’s recidivist
tendencies. Within eighteen months or perhaps two years of the initial
lawsuit, Helmstan stole untended cattle at Fonthill and drove them to
Cricklade. Once again, judgement was given against him, outlawry, and
this time his property was forfeit, although Ordlaf secured exemption
for Fonthill, which now belonged to him. By this time, Alfred the Great
had died and the outlawry of Helmstan was lifted by his son, again after
Ordlaf’s intercession. At this point, Ordlaf then exchanged the five hides
27 Wormald, The Making of English Law, 144–8.

28 Brooks, ‘Fonthill’, 312–14, although see below, at n. 33.
29 C. Wickham, ‘Rural Society in Carolingian Europe’, in R. McKitterick (ed.), The New
Cambridge Medieval History. II. c.700–c.900 (Cambridge, 1995), 510–37, at 510.
30 Keynes, ‘Fonthill’, 73; Brooks, ‘Fonthill’, 310n16.
31 Wormald, The Making of English Law, 146–8. Wormald argues that Chapter 1 of Alfred’s
law code, which dealt with oaths, turned any criminal activity into a breach of faith.
See Die Gesetze der Angelsachsen, ed. F. Liebermann, 3 vols. (Halle, 1898–1912), I, 29;
Ancient Laws and Institutes of England, ed. B. Thorpe, 2 vols. (London, 1840), I, 60–1;
EHD, no. 33.
346 Charles Insley
at Fonthill for five hides at Lydiard with the bishop of Winchester. The
penultimate act in this narrative was that Æthelhelm Higa renewed his
claim to the estate, this time against the bishop of Winchester, at whose
behest Ordlaf may have written his account. The final act is recorded in
an endorsement of the letter, added not long after it was written, that
records a meeting at Warminster where Æthelhelm Higa withdrew his
suit.
The timescale of the events in the letter is unclear; Helmstan’s ini-
tial theft of the belt happened no earlier than 897, since in the year or
eighteen months that elapsed between this case and the cattle-rustling
episode, King Alfred had died. The subsequent claim by Æthelhelm
Higa against the bishop of Winchester may have happened considerably
later; Keynes suggests a date in the second decade of the tenth century,
a suggestion supported on philological grounds by Gretsch and palaeo-
graphical grounds by Dumville.32 A date further away from belt-thieving
and cattle-rustling is also suggested by the fact that Ordlaf could no
longer remember the exact timescale between the two episodes, simply
recording that he was not sure whether it was eighteen months or two
years. Although Susan Reynolds and Mark Boynton have suggested an
alternative author of the letter, most commentators seem satisfied that
it was Ordlaf, notwithstanding the switch between the first- and third-
person voice which occurs part-way through.33 It has even been argued
that Ordlaf physically wrote the letter; if he did, that would mark it as a
very rare example of lay literacy in both reading and writing. If, on the
other hand, Ordlaf had dictated the letter to a scribe, at whose status as
layman or cleric we can only guess, this would raise the issue of what
‘authorship’ might have meant in this context: the letter might not then
be a reliable indicator of laymen’s possible command of the technology
of writing, but it still shows clearly that many (at least of Ordlaf ’s status)
could be deeply familiar with the use of documents in their everyday
business.
Here we have, then, a letter dictated – if not written – by a layman,
describing a world of legal processes into which the Church intrudes only
at the end. The action takes place in both formal environments, clearly
32 M. Gretsch, ‘The Language of the Fonthill Letter’, Anglo-Saxon England, 23 (1994),

57–102, at 90–2; D. N. Dumville, ‘English Square Minuscule Script: The Background
and Earliest Phases’, Anglo-Saxon England, 16 (1987), 147–79, at 171.
33 M. Boynton and S. Reynolds, ‘The Author of the Fonthill Letter’, Anglo-Saxon England,
25 (1996), 91–5; Brooks, ‘Fonthill’, 313–14, notes that shifts between the first and
third person are not uncommon in other types of vernacular documents, such as wills.
Elsewhere in Europe, the named scribe, while clearly taking overall responsibility for
the production of the document, did not necessarily physically write the script; see
above, 167, 298.
meetings of shire or hundred courts, and informal ones: in the first

claim launched by Æthelhelm Higa, Ordlaf describes an occasion in the
king’s chamber at Wardour, where King Alfred was washing his hands.
The legal processes referred to in the letter involve both the deploy-
ment of documents and non-documentary proofs, such as oath-takings.
Indeed, the Fonthill Letter highlights some of the problems of creating a
polarity between documentary and non-documentary culture in an early
medieval context since the two are inextricably linked in this case.34 The
discussion in Alfred’s chamber at Wardour concerned whether Helmstan
should be allowed to proceed with a judicial oath to prove his title against
Æthelhelm. What secured Helmstan this right was that some form of
charter granting the land to Oswulf, from whom Helmstan had acquired
Fonthill, was read out, along with the attestations of the king and others.
As Keynes’ translation of the letter puts it: ‘then it seemed to all of those
who were at that arbitration that Helmstan was nearer to the oath on that
account’; in other words, possession of the charter improved Helmstan’s
chances of being able to clear his name by judicial oath.35 The oath was
crucial because it was a public declaration of the settlement reached at
court, before the parties’ peers, who were now implicated in making the
settlement secure; only Helmstan’s repeated criminality threw it back
into dispute.36
Therefore, while the explicit object in dispute was the property at
Fonthill, also at issue was Helmstan’s trustworthiness, put in question by
his indictment for theft, twice. Since the first item that he stole, the sword
belt, was a traditional and typical badge of social standing, Helmstan’s
status may also have been a factor in the case: the extent of the property at
issue at Fonthill, five hides, accords with the minimum holding associated
in some sources with the rank of thegn.37
It is clear, though, for all the complex interplay of documentary
and non-documentary proofs, that documents – charters, agreements,
records of sales – were a vital part of this dispute and presumably of the
whole process of transferring and conveyancing property in general.38
34 Keynes, ‘Royal Government’, 256; Wormald, ‘Charters’, 167.

35 Keynes, ‘Fonthill’, 75–6.
36 For the significance of oaths in dispute settlement, see Innes, State and Society, 138–9,
and M. Costambeys, ‘Disputes and Courts in Lombard and Carolingian Central Italy’,
EME, 17 (2007), 265–89.
37 Brooks, ‘Fonthill’, 314–15; for five hides and thegnly status, see the text christened by
Liebermann Geþyncðo (ed. Liebermann, Gesetze, I, 456–69; trans. EHD, no. 51).
38 Wormald, ‘Charters’, 167: ‘what could be called the “oral-formulaic” approach to
Anglo-Saxon litigation is not borne out by the records of actual pleas. Oaths mattered,
but so, to a much greater extent than hitherto appreciated, did what modern justice
would consider evidence, and such evidence was preferably in writing’.
348 Charles Insley
As well as the letter itself, there are at least two other documents – and
maybe more – floating around during the dispute. The first is the charter
that one presumes accompanied the initial grant of Fonthill to a man
called Æthelwulf, or one of his predecessors. Æthelwulf gave the land
to his wife Æthelthryth as part of her dower, and she subsequently sold
it to Oswulf. The second document concerns this purchase, since the
letter refers, as noted above, to King Alfred giving his ‘signature’ (hond-
setene) to Oswulf at this purchase: presumably some form of charter by
Æthelthryth granting Fonthill to Oswulf, attested by Alfred. There may
have been – probably were – more documents underlying Ordlaf ’s let-
ter. When Helmstan gave the land to Ordlaf, he gave him the charter;
whether this was a putative original diploma for the estate or a record
of Oswulf’s purchase from Æthelthryth is unclear. When Ordlaf subse-
quently exchanged the estate with the bishop of Winchester, one imagines
that at least one charter changed hands, and it seems unlikely that the
bishop of Winchester would not also have had some sort of record of the
sale drawn up.
The Fonthill Letter seems, therefore, to stand as representative of a
substantial documentary culture, in which sale agreements, conveyances
and records of property transfers – possibly making extensive use of the
vernacular – circulated more widely than is suggested by most of
the surviving evidence, dominated as it is by documents from the top
of the documentary hierarchy, such as royal diplomas. The events sur-
rounding the composition of the letter also hint at a world where laymen
were adept at manipulating charters and conveyancing, where a sharp
operator like Ordlaf would smoothly offload disputed land in exchange
for unencumbered property. An important point to note here is that the
documents seemed to have circulated readily; in other words, laymen
clearly kept their charters in a way that allowed their easy retrieval for
use in such cases. While the Fonthill Letter is a unique survival, it seems
very unlikely indeed that the sort of episode it describes was not com-
mon, and we should take seriously Brooks’ suggestion that there were
once many more documents like this in circulation.39 Wormald’s hand-
list of Anglo-Saxon lawsuits, although small in comparison to lists of
similar surviving Carolingian material such as placita, provides plenty of
evidence of lay document use in dispute settlement, often interwoven
with other forms of testimony.40 Although Fonthill is not a placitum in
39 Brooks, ‘Fonthill’, 306.

40 Wormald, ‘Charters, 151, makes the point that the records we have of Anglo-Saxon
disputes are not formal placita in a continental sense, but usually descriptions of cases
embedded in other documents.
the sense that a Carolingian aristocrat might understand the term – that
is, as a type of document with distinctive formulation – it and the other
narratives concerning lawsuits are to some extent analogous in function.
The well-known Kentish dispute over Snodland, for instance, was
resolved by testing the boundaries detailed in the relevant charter.41
The Snodland narrative, from the last quarter of the tenth century, also
includes the theft of the relevant landbooks (charters) by priests, and
their sale to the son of the man who had originally granted Snodland to
the cathedral church of Rochester. It is a remarkably complex dispute
and, like Fonthill, took place over a long period of time, in this case over
the course of at least a generation. Although the surviving documentary
record of the dispute is preserved in charters now part of the archive of
Rochester Cathedral, the charters refer to the existence of other docu-
ments in lay hands during the earlier stages of the case. The widow of the
man who had had the title-deeds stolen, Brihtwaru, surrendered them
after being sued by the bishop of Rochester; she then, following King
Edgar’s death, recovered them in a counter-suit in which the bishop
was compelled to hand them over. Again we have frequent document
deployment in legal process and what appears to be a family archive.42
Another of the cases noted by Wormald, the dispute in 824 concern-
ing Westbury between the church at Berkeley, in Gloucestershire, and the
bishop of Worcester, hinged on the title-deeds originally possessed by the
father of the lay donor of Westbury, a thegn called Æthelmund. The lay
donor, Æthelmund’s son Æthelric, had given the estate to Worcester,
with a life interest for his mother, Ceolburh. To forestall any problems
over his dispositions, the charter recording Æthelric’s will notes that he
had also given copies of the Westbury donation to ‘friends’.43 Wormald’s
interest in this dispute was to uncover the mechanism behind its resolu-
tion, but it is clear that whatever the process, it involved the deployment
and preservation of documents by laypeople, including Æthelmund’s
widow Ceolburh, and again what may have been some sort of family
‘archive’.
The consignment by Æthelric of copies of the donation to his friends
points to the value attached to the charter as an object in its own right,
an impression reinforced by further evidence for Æthelric’s use of doc-
uments. He was party to the earliest surviving copy of a chirograph, a
charter in which the same text was copied two or three times on a single
membrane, which was then cut to produce congruent copies of identical
texts; the chirograph in question documents an agreement of around 804
41 Wormald, ‘A Handlist’, no. 46 (S 1457).

42 Wormald, ‘Charters’, 157–60. 43 Wormald, ‘Charters’, 152–7.
350 Charles Insley
between Æthelric and the bishop of Lichfield.44 Counterparts were held

by two of Æthelric’s friends in this case too, explicitly as security, in addi-
tion to the parts held by Æthelric and the bishop.45 The example of the
chirograph suggests that these documents were not just put somewhere
and forgotten, but that they could be retrieved and deployed if necessary:
what would be the point of Æthelric giving copies to his friends (‘amici
necessarii mei et fidelissimi’) if they were likely to lose them? To be of
use, the charter had to be a recoverable object.
That laymen fully understood the value of documents is reinforced by
an account of a transaction, dating from the later 970s or 980s, recorded
in the Libellus Æthelwoldi, a twelfth-century Ely chronicle also used in
the compilation of the better-known Liber Eliensis.46 The account con-
cerns a dispute over an estate at Hauxton and Newton, in which the
relevant charters were in the possession of a man called Ælfwold, brother
of the vendor Eadric.47 So worried was Abbot Byrhtnoth of Ely about
the ‘claims and trickeries’ that might arise from his abbey’s failure to
possess the charters – described in the Libellus as cyrographi – that he
asked first Ealdorman Æthelwold and then Ealdorman Byrhtnoth to pur-
chase the charters from Ælfwold, in return for charters for Ramsey and
Sproughton, charters Ælfwold greatly desired (‘quo cyrographo multum
indiguit’). Again, the actors here all understood the value of the charters,
even if they could not all read them, not just as symbolic of ownership
but also as tools in dispute resolution.
Evidence like that for Snodland, Westbury and Hauxton and Newton
shows a world where different modes of testimony happily coexisted but,
significantly, one where the deployment of documents was commonplace,
and where legal cases could run across a generation or two. It also seems
likely that this document use did not include just diplomas but also
an array of other documentary material: records of sales, conveyance
notices and even memoranda such as the Fonthill Letter. The frequent
use by laypeople of documents in judicial processes, sometimes with the
Church, sometimes with other lay individuals, has implications for how
we approach the question of lay archives. It would seem likely that if
individuals expected to use their documents – title-deeds, notices of sale
or whatever – then they would take steps to store them in places that
were accessible and in a way that allowed for reasonably easy retrieval.
The evidence of Æthelric suggests that laymen did indeed expect to have
to use their documents and did expect their gifts and bequests to be
44 Lowe, ‘Lay Literacy’, 174; S 1187. 45 Lowe, ‘Lay Literacy’, 174.

46 Liber Eliensis 2.27 (ed. Blake, 100–1; trans. Fairweather, 122).
47 Keynes, ‘Royal Government’, 250–1; Lowe, ‘Lay Literacy’, 168–9.
challenged. Why else would Æthelric supply two copies of his agreement
with the bishop of Lichfield to two close friends? Why else was the abbot
of Ely so exercised about the charters held by Ælfwold?
What is also striking about many of the records of transactions in the
Libellus Æthelwoldi is the extent to which purchases and exchanges were
an essential part of the way in which a religious institution built up its
endowments. The Libellus in fact attests to a busy traffic in land in tenth-
century East Anglia, where estates, charters and money changed hands
frequently and where both the abbey and its lay neighbours drove hard
bargains. It seems unlikely that this was a phenomenon confined either to
East Anglia or to the late tenth century. There is a strong argument that
even in early Anglo-Saxon England a good number of estates changed
hands for reasons other than straightforward inheritance and were often
transferred in return for cash, payment in kind or other pieces of land,
the quantities of which may have depended more on social factors than
on a ‘pure’ market mechanism (that is, a payment in cash determined by
supply and demand). It is overwhelmingly likely that the same exchanges
occurred, perhaps even more frequently, in the later period too.48 It
is possible that the foundation or refoundation of land-hungry abbeys
in East Anglia, such as Ely, Ramsey or Thorney, made the traffic in
land in late tenth-century East Anglia particularly lively, but this does
not alter the basic picture; indeed, if we return to Fonthill, the events
described in the letter also describe the same piece of land changing
hands several times. This, too, has implications for how we view the way
in which laymen used and stored their documents. It is clear that diplomas
functioned as title-deeds, and therefore possession of the diploma was
key. Indeed, the Libellus’ account of the acquisition of Bluntisham makes
this quite clear: ‘the person who had the charter was nearer to having
the land than the one who did not have it’.49 In other words, possession
of the document did not guarantee possession of the land, but it helped
enormously. This suggests a world where charters may have changed
hands quite frequently and where additional documentation recording
sales and transfers may have been drawn up to accompany such transfers.
The Libellus episode concerning Hauxton and Newton referred to above
also makes it clear that charters did change hands and that the new owners
of a piece of property were very keen to have the title-deeds, often going
to great lengths to secure them. This, in turn, suggests an environment
where laypeople could organize and retrieve their title-deeds and other
48 J. Campbell, ‘The Sale of Land and the Economics of Power in Early Anglo-Saxon
England’, Haskins Society Journal, 1 (1989), 23–37, at 25–6, 37.
49 Liber Eliensis 2.25 (ed. Blake, 99; trans. Fairweather, 122).
352 Charles Insley
documents, and where the vernacular endorsements made on diplomas

would have aided such organization. If we stray a little further into the
realms of speculation, it is possible to imagine a situation in which a
layman with a significant amount of property held by charter might
have bundles of documents – diplomas, leases and notices of sale or
exchange – stored in some sort of chest.
Family archives
As well as the ready deployment of charters and chirographs in legal
processes, both the Æthelric and Snodland cases also hint at the existence
of family archives that operated across at least two generations. This is an
important consideration; we might postulate an individual layman with
the ninth- or tenth-century equivalent of the box under his bed, but for
these collections to be transmitted across a number of generations of one
family argues for something a little more systematic. Related to this is
the question of whether such family archives or dossiers can be detected
in the surviving record. There may be a number of such collections
embedded in the surviving archives of Old English religious houses, but
the archives of Burton Abbey in Staffordshire offer one of the clearest
examples.
Although later medieval Burton traditions suggest an early commu-
nity associated with St Modwenna on the site, the abbey as it was in
the later Middle Ages was founded, or refounded, in the two decades
before 1004, when its possessions and privileges were confirmed by King
Æthelred II.50 The founder was an influential and well-connected Mid-
lands thegn called Wulfric Spot, whose surviving will shows him to have
had a significant assemblage of estates stretching right across the Mid-
lands, with a concentration in the counties of Staffordshire, Derbyshire,
Warwickshire and Shropshire.51 Wulfric was a ‘king’s thegn’, that is, a
senior member of the king’s court, a connection emphasized by the fact
that his brother Ælfhelm was ealdorman of York between 993 and 1006,
and the fact that Wulfric, in his will, specifically commended his new
foundation of Burton to King Æthelred’s care.52 That will also details
Wulfric’s extensive property bequests, and it is the connection of a num-
ber of these bequests to charters in the archive of Burton that allows us to
identify a possible family archive embedded within the monastic archive.
50 S 906.
51 S 1536; Charters of Burton Abbey, ed. P. H. Sawyer (Oxford, 1979) (hereafter Burton),
pp. xvi–xvii.
52 S 1536; Burton, pp. xxxviii–xliii.
The Anglo-Saxon charters of Burton are preserved in a thirteenth-

century cartulary incorporated into a manuscript now in the National
Library of Wales, Peniarth 390.53 The thirteenth-century copyist almost
certainly had the original charters in front of him, since he went to the
effort of copying into the cartulary some of the chrismons that open a
number of them.54 The ordering of the texts in Peniarth 390 does not
help much in reconstructing the archive, since the cartularist opted to
copy the texts in roughly chronological order, rather than in another way.
Nevertheless, a number of the charters that can be plausibly connected
to Wulfric’s family predate the refoundation of the abbey at the very end
of the tenth century or start of the eleventh, and what we have in Peniarth
390, therefore, is the sort of ‘founder’s archive’ that we see in continental
cartularies.55
Several of the charters in the Burton archive are grants directly to
members of Wulfric’s family, although there has been a degree of tam-
pering with some of them. Two of the charters, dated 942, are grants to
Wulfsige ‘the Black’ (‘prenomine Maurus’), who was probably Wulfric’s
grandfather, covering lands in Derbyshire and Staffordshire.56 A further
grant, dated 951, of land at Marchington (Staffordshire) to the thegn
Wulfhelm may be part of Wulfric’s archive, as might a charter of King
Eadwig for ‘his man’ Mæglsothen.57 The grant to Mæglsothen is for land
subsequently bequeathed by Wulfric to Burton Abbey, and the charter,
therefore, may have been Wulfric’s title-deed to the estate.
Wulfric himself was the beneficiary of several charters in the Bur-
ton archive, covering estates at Pillaton and Abbots Bromley in
Staffordshire.58 Other charters in the archive can be linked more ten-
tatively to Wulfric’s family. The ten-hide grant at Parwich made by King
Edgar in 966 to the thegn Ælfhelm may have been to Wulfric’s brother
of that name.59 More securely, the grant by Edgar to the thegn Wulfgeat
of land at Arley in Warwickshire should also be regarded as part of
Wulfric’s family archive. Arley was among the gifts made by Wulfric’s
mother Wulfrun to her foundation of Wolverhampton, while Wulfgeat
was named in Æthelred’s confirmation of Wulfrun’s bequests to Wolver-
hampton as her kinsman.60 It is unclear how a bequest to one foundation,
53 Aberystwyth, National Library of Wales, Peniarth MS 390, fols. 173–84v; Burton,

pp. xiii–xv.
54 Burton, p. xiv. 55 See above, Chapters 7, 8, 10–11.
56 Burton, p. xl, and nos. 5, 6; S 479, 484. 57 Burton, nos. 11, 15; S 557, 628.
58 Burton, nos. 26–7; S 878–9. S 879 is regarded as spurious, but Pillaton was among the
estates bequeathed by Wulfric to Burton.
59 Burton, no. 21; S 739.
60 Burton, no. 20; S 720, 1380; Anglo-Saxon Wills, ed. D. Whitelock (Cambridge, 1955),
164–5.
354 Charles Insley
Wolverhampton, ended up first being granted again to a member of the

founder’s family and then bequeathed by another member of the same
family to a new foundation. Since at no point are the hidage assessments
given, either for Wulfrun’s initial bequest or in the 963 charter grant-
ing Arley to Wulfgeat, it is possible that the grants are separate, with
Wulfrun bequeathing one portion of the estate and then another portion
being booked to Wulfgeat.
Not all of these links are certain, and there is a degree of supposi-
tion about some of those posited between possible members of Wulfric’s
family.61 It is also likely that the dossier of Wulfric’s charters embedded
in the Burton archive may have originally been larger than it was when
Peniarth 390 was copied in the thirteenth century. Two charters now in
the William Salt Library in Stafford, and not copied into Peniarth 390,
also appear to have been part of Wulfric’s collection.62 What allows us
to think of Wulfric’s charters as a family archive is its multi-generational
aspect. A number of charters in this collection were for beneficiaries
other than Wulfric; one may have been from his brother and three from
his grandfather, and there is also the possibility of grants from his mother.
The evidence for Arley points to another strand of preservation by lay-
men, which may even indicate a different family archive. A grant by
King Æthelred in 1001 of land in Arley and Itchington is extant on a
single sheet that survived through much of the subsequent centuries at
Coventry, but all the ecclesiastical institutions that might have preserved
it there came into existence some time later (probably the earliest was
Coventry abbey, founded in 1043).63 So the charter must have spent at
least the first generation after its creation in other hands, which must
have included the initial beneficiary, the lay magnate Clofi. Whether he
was receiving the estate that had previously belonged to Wulfgeat, or a
different part of Arley, is not clear; but like Wulfric and his family, he
seems to have had the desire and the means to preserve documents.
As already suggested, embedding lay collections in ecclesiastical
archives raised the possibility that diplomas, in particular, could be
amended, copied or interpolated after entering clerical hands. Since
diplomas effectively conferred title, a perennial concern for a church was
holding diplomas in which the church was not the beneficiary; something
of these worries can be seen in the Hauxton and Newton case already
encountered. One way in which a church could quite easily resolve this
was to amend the charter to reflect the church’s possession of the relevant
estate. That this happened is clear; how often it happened, or in what
particular circumstances, is much less clear. One example illustrates the
61 Burton, pp. xxxviii–xliii. 62 S 878, 922. 63 S 898.

problem: Stoke Canon in Devon, a possession of the bishop of Exeter

in 1066. Stoke was one of the estates that eleventh-century Exeter tradi-
tion believed had originally been given to the minster by King Æthelstan
(924–39).64 However, in 1031, Stoke was the subject of a grant by Cnut
to a thegn called Hunuwine.65 At some point between 1031 and 1066,
Stoke passed to Exeter, as did Cnut’s diploma, which is still preserved
among the muniments of Exeter Cathedral. There is, though, another
charter for Stoke preserved in the cathedral archives, one of a group
of not-very-good forgeries produced at some point during the 1050s or
1060s.66 This forgery was clearly based on Hunuwine’s charter: it has the
same hidage assessment – five – and exactly the same vernacular bound-
ary clause. In this case, the desire of Bishop Leofric (bishop of Crediton
and St Germans 1046–50, bishop of Exeter 1050–72) and his scribes
seems to have been to make a recent acquisition from a layman appear
to be a much more ancient grant direct to the minster. Fortunately for
us, the original charter survives in this case, although why is unclear;
had Hunuwine’s charter vanished, we would have no idea that the estate
had been in lay hands. This, of course, raises the question of the preva-
lence of this sort of practice. Where lay charters were copied into the
twelfth-century chronicle-cartulary of Abingdon, they were attached to a
statement that the layman named in the charter had made the church his
heir.67 The modern editor of Abingdon’s charters, Susan Kelly, has sug-
gested that the frequency with which this happens means that we must
exercise great caution before taking such statements at face value.68 As
with the case of Hunuwine, the attachment of the simple statement of
bequest may mask a more complicated mode of acquisition by the abbey.
Such a process would also serve to break up collections of documents as
they entered the abbey’s archival landscape. The reasons for the preser-
vation of embedded family documents in ecclesiastical archives must
ultimately be various, some reflecting acquisitiveness on the part of the
institution and its personnel, and some resulting more directly from the
association of the institution with a lay family, and the position of both
in the skein of local allegiances.
64 Oxford, Bodleian Library, MS Auct. D. 2.16, fols. 8r–14r; Oxford, Bodleian Library,
MS 579, fol. 6r–v (Leofric Missal, ed. N. Orchard, 2 vols. [London, 2002], II, 8–9); see
C. Insley, ‘Remembering Communities Past: Exeter Cathedral in the Eleventh Century’,
in P. Dalton, C. Insley and L. Wilkinson (eds.), Cathedrals, Communities and Conflict in
the Anglo-Norman World (Woodbridge, 2011), 41–60, at 48–54.
65 S 971. 66 S 387.
67 Charters of Abingdon Abbey, ed. S. E. Kelly, 2 vols. (Oxford, 2000–1) (hereafter Abing-
don), pp. cxxii–cxxiii.
68 Abingdon, p. cxiii.
356 Charles Insley
Related to the issue of family archives is the important but difficult

question of whether Anglo-Saxon laymen used local churches or monas-
teries as their record repositories. The evidence here is, perhaps unsur-
prisingly, very unclear. We might start with the example of Ealdorman
Æthelfrith, who, in the early tenth century, petitioned Edward the Elder,
Æthelred and Æthelflæd, rulers of the Mercians, to replace charters he
had lost.69 According to one of these replacement charters, all of his char-
ters of inheritance (hereditarii libri)70 had been consumed by a ravaging
fire. This raises some interesting issues. The first is that the mention of
his ‘books of inheritance’ is clearly a reference to some sort of archive.
Second, the charter states that Edward, Æthelred and Æthelflæd ‘unan-
imously consented’ to other charters being produced ‘as far as he could
relate them from memory’ (in quantum eos memoriter recordari potuisset);
this suggests not only that Æthelfrith had a number of diplomas, but
also that he was familiar with their contents.71 Third, there is the fire
itself: given the date of Æthelfrith’s petition, the first decade of the tenth
century, it is tempting to see his archives as a casualty of the campaigns
of the Vikings in Mercia in the late ninth century. If so, it is possible that
his archives were burnt because they were in a church, often the targets
of Viking attacks. This is, of course, speculation, and there is no reason
why Æthelfrith’s charters could not have been kept in his hall or some
other such location. The history of the Fonthill Letter itself shows how
preservation was often a matter of a series of personal choices. Since
in the end the beneficiaries of the Fonthill dispute were the bishopric
of Winchester, which got the estate at Fonthill, and Ordlaf, who got an
estate in exchange for Fonthill, we would expect the letter initially to have
resided with one of them. By the twelfth century, though, it was where it
remains, in the archive of Christ Church, Canterbury, and where the later
archivist’s designation of it as ‘useless’ was, fortunately, ignored. It has
been hypothesized that it reached Canterbury when one of the bishops of
Winchester transferred to the archiepiscopal see of Canterbury, perhaps
Ælfheah in 1006.72 The other possibility is that Christ Church had a link
to, or even provided a repository for the documents of, Ordlaf or one
of his family; but it is less likely in this case simply because the known
associations of Ordlaf and his son are with Winchester, Malmesbury and
Wilton.73 Nonetheless, both scenarios envisage that the letter spent time
69 S 367; this incident is discussed by Lowe, ‘Lay Literacy’, 170–1.

70 Here liber is a direct translation of Old English boc, ‘book’ or ‘charter’.
71 S 367. 72 Brooks, ‘Fonthill Letter’, 314–15.
73 See S 354, 359, 362, 372–8, 381–3, 1286 (Ordlaf witnessing Winchester charters);
1284 (exchange of Fonthill for Lydiard between Ordlaf and bishop of Winchester);
in an archive associated with a person (ecclesiastical in one case, lay in

the other), rather than an institution.
We are possibly on more concrete ground if we consider those archives
that contain charters for property to which the Church had no claim.
Cases in point are the so-called ‘alien diplomas’ in the Abingdon
archive.74 These are diplomas copied into the two twelfth-century Abing-
don cartularies but concerning lands to which Abingdon seems to have
had no claim at any point. Stenton suggested that these were charters
that had been deposited for safe keeping with the abbey and then not
retrieved (for whatever reason) and that had subsequently been merged
with the abbey’s archives.75 As Kelly notes in her discussion of the Abing-
don archive, this thesis is attractive on the face of it, since it might explain
how Abingdon and, indeed, other houses such as the Old Minster Winch-
ester or Glastonbury, came to be in possession of charters for estates that
seem never to have been part of the monastic endowment. However,
Kelly goes on to argue that most of the charters that apparently have no
connection to the endowment were, in fact, related to estates the abbey
had or still held in the twelfth century. She also raises the point that, since
possession of a title-deed (that is, charter) conferred possession of the
land, it would be a rash layman who would entrust his precious landbooks
to monastic keeping, something that might have been especially pertinent
given the much more rigid attitude to property displayed by many of the
tenth-century foundations or refoundations, such as Abingdon, Ely or
Ramsey.76
Keynes suggests that in some circumstances laymen might have
deposited their diplomas with a monastery such as Abingdon for safe
keeping in times of necessity, but such circumstances must have been
rare, and there is precious little pre-Conquest evidence of such a prac-
tice. However, these reservations really concern diplomas, and it is likely
that other sorts of records, such as agreements or wills, especially those
drawn up in chirograph form, were deposited with churches. The Libellus
Æthelwoldi contains an account of a will-making episode, in which Abbot
Byrhtnoth of Ely had the will of a thegn called Siferth written as a tripar-
tite chirograph before being read out and distributed to Siferth himself,
Abbot Byrhtnoth and Ealdorman Æthelwine.77 Once again we have the
intersection of performance and text, since a key element of the process
1205 (exchange between Ordlaf and Malmesbury); 364, 368 (in the Wilton archive,
granted to or witnessed by Ordlaf); 469 (grant to Wilton by Ordlaf’s son Ordwald).
74 Abingdon, pp. cxxxvii–cxlix.
75 F. M. Stenton, The Early History of the Abbey of Abingdon (Reading, 1913), 43.
76 Abingdon, pp. cxxxviii–cxlv.
77 Liber Eliensis 2.11 (ed. Blake, 87; trans. Fairweather, 110).
358 Charles Insley
was the reading out of the will, but the episode also underlines the close
relationship of many local aristocrats to these foundations and the prob-
ability that they did entrust some of their documents to ecclesiastical
keeping.
There remains the question of the extent to which the late tenth-
century ‘monastic reform’ changed the archival landscape of Anglo-
Saxon England. Much ink has been spilled on the nature of the monastic
reform and the ‘anti-monastic’ reaction that followed the death in 975
of King Edgar, the most active proponent and patron of the reform.78
What is relevant is that the reform appears to have redefined the rela-
tionship between monastery, benefactor and property, inasmuch as the
houses founded or refounded during this period seem to have had a
much more rigid and even aggressive attitude towards their property and
the kin of their benefactors than older churches, which seem to have
enjoyed a rather more cosy relationship with local aristocracies. Many of
the episodes related in the Libellus Æthelwoldi, for instance, concern the
thwarting by Ely of the claims made by relations of their benefactors.79
This attitude, in so far as it can be discerned, is especially important
to keep in mind when we consider the nature of the corpus of surviving
Anglo-Saxon charters. What survives is dominated by the archives of a
handful of great houses: well over half of survivals, especially of diplo-
mas, can be linked to the archives of Worcester, Abingdon, Old Minster
Winchester and Christ Church Canterbury. It is very significant, there-
fore, that all of these institutions were, in some way, participants in the
reform movement. How we are to regard this dominance or its implica-
tions is not at all certain, but a few lines of enquiry can be advanced.
What is clear about the ‘reform’, even before its apogee in the reign
of Edgar, is that it involved a significant transfer of landed wealth into
the hands of the monastic Church.80 This was a matter not simply of
piety but of royal policy; the Regularis Concordia, promulgated under
Edgar’s auspices at a great council in Winchester, specified one rule for
all monasteries, enjoined all monasteries to pray frequently for the king
and the kingdom, and placed all monasteries under the king’s lordship,
while nunneries were subject to the queen.81 These strictures, along with
78 See C. Cubitt, ‘The Tenth-Century Benedictine Reform in England’, EME, 6 (1997),

77–94.
79 See above, 351.
80 C. Cubitt, ‘The Institutional Church’, in P. Stafford (ed.), A Companion to the Early
Middle Ages: Britain and Ireland c. 500–c. 1100 (Oxford, 2009), 376–94, at 386–9.
81 T. Symons, ‘Regularis concordia: History and Derivation’, in D. Parsons (ed.), Tenth-
Century Studies: Essays in Commemoration of the Millennium of the Council of Winchester
and the Regularis Concordia (Chichester, 1975), 37–59.
active promotion by Edgar of lay patronage of these institutions, turned

these foundations, especially the Mercian and East Anglian ones, into a
form of surrogate royal presence. These institutions were very like the
abbeys founded by the Carolingians east of the Rhine, in that they func-
tioned, in part, as centres of royal control in areas outside the West Saxon
heartland of the kings of the English.
This transfer of landed wealth can be detected in the surviving archives,
whether the Abingdon Chronicle or the Liber Eliensis, or even in those
charters belonging to the family of Wulfric Spot that ended up in the
hands of Burton Abbey.82 Such a transfer also undoubtedly changed the
relationship between these religious foundations and local aristocrats and
landholders. It became at once closer, as lay patrons had a much larger
stake in the Church, but also more rigid, as the Church in turn sought to
loosen lay control over land in Church hands. This changing relationship
affected the archival landscape; first, by generating an increase in the
issuing of diplomas, as laymen sought permission to transfer land to
the Church – an increase that can be seen in the pattern of survival of
diplomas through the tenth century. Second, the new relationship also
saw the transfer of diplomas and other documentation into the hands of
the monastic Church, the sort of processes witnessed by the accounts
in the Libellus Æthelwoldi. This in turn had implications for the way in
which documents were preserved, and even for what documents were
preserved.
The one part of England seemingly untouched by these developments
was the North; that is, what had been the kingdom of Northumbria.
Much of the preceding discussion has considered England south of the
river Humber, with Burton Abbey in Staffordshire the northernmost out-
post of the tenth-century expansion in monastic foundation and refoun-
dation. The question of lay documentary and archival practice in the
North is a very tricky one, not least because of the convoluted and violent
history of Northumbria from the eighth century onwards. It is generally
assumed that the archives of most northern houses were badly hit during
the Viking attacks and settlements of the ninth and early tenth centuries,
while the confused political situation that followed, until the extinction
of the Viking kingdom of York in the 950s, was also hardly conducive to
archival stability. Not that we should lay the blame solely at the feet of
the Vikings: King Eadred’s campaign in the North in 948, which saw the
burning of the church at Ripon and the transfer to the South of the relics
82 See above, n. 46; Burton, pp. xliv–xlvii; Abingdon, pp. cxxxi–cxcii.

360 Charles Insley
of Wilfrid, one of the North’s most potent saints, presumably also had
implications for whatever documents were stored there.83
What did laymen in the North do with their charters and other records?
The much poorer survival of pre-Conquest material from northern Eng-
land makes it tempting to posit – even allowing for Viking depredation –
a much lower level of documentary use and culture than in the South.
Tempting, but probably wrong; there is no reason to suggest that the
same imperatives that drove lay document use and preservation in south-
ern England were absent from the North. In southern England laymen
collected and stored documents because they were useful, both in sym-
bolic and practical ways. It is likely that these needs also obtained in
northern England. We can, perhaps, detect glimpses of lay documentary
culture in the twelfth-century corpus of writing from Durham Cathedral
priory. These texts are generally regarded as drawing on earlier collec-
tions of annals, regnal lists and, possibly, documents such as charters.
One episode, in particular, hints at a world of lay property transactions
and collections of charters. Early in the first decade of the eleventh cen-
tury, a prominent York thegn named Styrr Ulfson granted a large num-
ber of estates in North Yorkshire and County Durham to the community
of St Cuthbert.84 This episode is described in the Historia de Sancto
Cuthberto – one of these twelfth-century Durham texts – in terms that
suggest that the author had in front of him either the charter or some sort
of written notice of it.85 The episode is also described as having taken
place under the witness of Archbishop Wulfstan and King Æthelred, so
if the Durham scribe had a document in front of him it could have been
either Styrr’s charter or even a royal charter allowing Styrr to alienate
the land.86 Crucially, though, these estates had not been in Styrr’s family
for very long, since in the text he is described as having purchased them
83 Anglo-Saxon Chronicle (D), s.a. (ed. G. P. Cubbin [Cambridge, 1996], 44; trans. EHD,
no. 1 [p. 223]).
84 The Early Charters of Northern England and the North Midlands, ed. C. R. Hart (Leicester,
1975), no. 130; W. Aird, St Cuthbert and the Normans: The Church of Durham, 1071–
1153 (Woodbridge, 1998), 47–8; H. H. E. Craster, ‘The Red Book of Durham’, English
Historical Review, 40 (1925), 504–35, at 526. The estates concerned were Darlington,
Coniscliffe, Cockerton, Haughton-le-Skerne, Normanby, Ketton and Lumley.
85 Historia de sancto Cuthberto 29 (ed. T. Arnold, Rolls Series 75.1:212–13).
86 Styrr’s grant to the Cuthbert community seems to have been part of the complex
interplay of regional political rivalries in what had been the kingdom of York; see Aird,
St Cuthbert, 48–9; W. Kapelle, The Norman Conquest of the North: The Region and Its
Transformation, 1000–1135 (London, 1979), 17–24; C. Insley, ‘Politics, Conflict and
Kinship in Early Eleventh-Century Mercia’, Midland History, 25 (2000), 28–42, at
35–7, for a discussion of these rivalries.
himself.87 The question is then whether Styrr had charters, or at least

memoranda of sale, for those transactions, charters that he then handed
over to Durham. This seems to have been the sort of situation hinted at
in texts such as the Libellus Æthelwoldi, and there is no reason to believe
that property transaction did not happen in the same way north of the
Humber.
Where the North might have differed, especially in terms of the con-
tours of the archival landscape, is in the absence there of aggressive,
archive-minded monasteries of the sort we see in East Anglia or Wes-
sex; there were no ‘Æthelwoldian’ abbeys in the north. The imperatives
of places such as Abingdon, Ely, the New Minster or Worcester shaped
the archival landscape of southern England on the eve of the Norman
Conquest, not least because they generated a significant shift of land
into monastic hands. Their absence from the North might therefore be
a major determinant of the much more exiguous survival of diplomatic
evidence from northern England.
Conclusions
The evidence surveyed here, perforce briefly and episodically, allows a
number of conclusions to be drawn about lay documentary culture in
England in the three and a half centuries before the Norman Conquest.
The first is that those looking for lay archives in England will be disap-
pointed; no such thing survives. However, it is also clear from what does
survive that laypeople fully understood the importance of documents,
not just as artefacts but also as texts, even if they could not read them;
laypeople also involved documents in all sorts of transactions and social
action, from an early date. The frequency with which documents appear
in the fragmentary accounts that survive of disputes and conveyancing
also suggests that lay individuals had mechanisms for accessing the nec-
essary documentation. For a small-scale landholder, this may not have
been difficult, since he or she may only have had a few charters; for
larger-scale landowners, however, the ease with which documentation
was deployed and the arenas in which it was deployed suggest some sort
of rudimentary record-keeping beyond the proverbial box under the bed.
Second, the likelihood that there was a lively traffic in landed property
from an early date in England suggests a certain amount of documen-
tation in circulation in addition to royal diplomas. This is hinted at by
87 Historia de sancto Cuthberto 29 (ed. Arnold, 212), ‘Et ego emi propria pecunia et dedi
Sancto Cuthberto iiii carucatas terrae.’
362 Charles Insley
such lucky survivals as the Fonthill Letter. Third, the Fonthill Letter
also illustrates how much material, especially in the vernacular, has now
been lost. Only at Worcester and Bury St Edmunds does a significant
body of Old English material – mostly leases – survive. However, it
seems inconceivable that these were the only two institutions to lease
land using vernacular documents.88 Much of this additional documenta-
tion was probably lost in the centuries following the Norman Conquest.
However, it is likely that this process of reshaping the archival landscape
began before 1066, as religious institutions sought to control and orga-
nize their archives from the second half of the tenth century onwards and
to prioritize the preservation, interpolation and copying of royal diplo-
mas. We know what we have, but we can only guess at what we do not
have but that once existed.
88 See, for instance, S 1387, an Old English chirographic lease made by the bishop of
Crediton in around 1018; also Lowe, ‘Lay Literacy’, appendix, 185–203.
14 Conclusion
Many more documents survive from the early Middle Ages than from the
Roman Empire or its Byzantine successor, outside Egypt. Yet the view
that Roman society at all levels was permeated with documents is per-
sistent. Narrative sources assume routine documentation, and the fact
of imperial unity has given ancient historians confidence in generalizing
from scant direct evidence, like the papyri preserved in the unique con-
ditions of Egypt, to provincial society as a whole. We might now be less
likely to be swayed by the literature of the golden age than our prede-
cessors; we might be more likely to stress the increasing prevalence of
documents in third- and fourth-century legal procedure in the context of
ever more ambitious attempts at social regulation on the part of the state;
we might also point to the differing brands of Romanitas on offer in the
provincial societies of the third, fourth and fifth centuries. Nevertheless,
the general statement that Romans routinely resorted to documentation
will be contested by few.
When we deal with the ever-thickening stream of surviving documen-
tation from the last centuries of the first millennium, however, we are
reluctant to generalize. Regional difference, a leitmotif of current early
medieval studies, drives us to draw contrasts even between different
provinces of the same kingdom when it comes to documentary prac-
tice or social structure.1 More importantly, we are provoked to caution
by the nature of our evidence. We have far more actual evidence from
1 C. Wickham, ‘Problems of Comparing Rural Societies in Early Medieval Europe’, in

Wickham, Land and Power: Studies in Italian and Early Mediaeval Social History, 300–1100
(Rome, 1994), 201–26 (orig. TRHS, 6th ser., 2 [1992], 221–46); see, emblematically,
Wickham, Framing, as well as J. M. H. Smith, Europe After Rome: A New Cultural History
(Oxford, 2005). While there have been regional studies of the late Roman Empire,
especially of Egypt, whence the bulk of documents survives for the period (R. S. Bagnall,
Egypt in Late Antiquity [Princeton, NJ, 1996]; C. Adams, Land Transport in Roman
Egypt 30 BC–AD 300: A Study in Administration and Economic History [Oxford, 2007];
P. Sarris, Economy and Society in the Age of Justinian [Cambridge, 2005]), these are taken
as to some extent symptomatic of the imperial system as a whole. Studies of regions after
c. 650, on the other hand, are generally taken to speak of diversity and difference.
363
364 Conclusion
the eighth and following centuries than we do for the Roman period
precisely because the great churches and monasteries of early medieval
Europe enjoyed a continuous archival history into early modern times,
and because the records they held were therefore absorbed to a greater
or lesser extent into the national and regional archives of modern gov-
ernments. From there they entered into the documentary canons of the
professionalized historical academies. We can debate what a Roman tax
list looked like, in the absence of unambiguous direct evidence, but
we know exactly how early medieval churches documented the dues of
their tenants. Nevertheless, we are profoundly suspicious of this evidence
because it represents only one institutional context, that of the Church.
The legacy of nineteenth-century anticlericalism, and the corresponding
assumption of a fundamental and structural antagonism between Church
and State, has exerted a profound influence here, one that makes it far
harder for us to accept ecclesiastical writings as representative of a culture
as a whole than those of a Cicero or a Tacitus. The shadow of Gibbon’s
triumph of superstition and barbarism also looms large,2 for in spite
of now three generations of scholarship wrestling with the assumptions
of cataclysm and decline, alternative characterizations of post-Roman
change remain elusive. There is a striking historiographical comparison
with early medieval China, where historical writing took place in a dedi-
cated office at the imperial court and where the different mechanisms of
transmission mean that practical and provincial documentation is more
or less absent. Yet, in spite of political upheavals, the ideologies essayed
in official historiography and their adaptation by early modern and mod-
ern regimes mean that the assumption of structural continuity remains
unquestioned. Specialists are thus happy to generalize from exceptional
survivals such as the archives of the Buddhist cave monastery of Tun-
huang and assume a society in which the written instruments of imperial
government penetrated deeply. Furthermore, the provenance of these
remarkable but isolated documents is seen as unproblematic as there is
little sense of a cultural divide between religious foundations and the
wider populace.3
In the specific case of writing in early medieval Europe, the changing
distributions of one particular form of writing – the inscribed word –
have been used to shore up received views of a generalized decline of
literacy, when in fact they relate to the shifting cultural, political and
2 Edward Gibbon, The History of the Decline and Fall of the Roman Empire, ed. J. B. Bury,
7 vols. (London, 1898–1901), IV, 150: ‘The progress of superstition and barbarism was
rapid and universal.’
3 D. Twitchett, ‘Chinese Social History from the Seventh to the Tenth Centuries: The
Tunhuang Documents and Their Implications’, Past and Present, 35 (1966), 28–53.
Conclusion 365
ideological meanings of inscriptions.4 In any case, in Byzantium the

distribution of inscriptions likewise changes, and (outside Egypt) doc-
uments of practice from the early medieval period are far, far scarcer
than in western Europe, but specialists have little doubt that documen-
tation continued to play a significant role in the regulation of social
relations.5 In contrast, in western Europe, the thickening evidence for
documents and document use from the eighth century onwards is
generally understood in terms of revival and innovation, driven above
all by the Church and by the imperatives of the Carolingian rulers
who invested so heavily in it; debate turns on whether one takes a
maximalist or minimalist view of the effects of this investment. As a
result, scholars have tended to assume a gap between the documentary
culture of Late Antiquity and that of the Carolingian and post-
Carolingian world, a view underpinned by Pirenne’s seizing on the move
from papyrus to parchment in seventh-century Gaul as emblematic of a
series of wider cultural and political changes.6 A variety of grand narra-
tives have been created to explain this perceived gap; the most familiar
one describes a transition ‘from public to private’ in the shape of a shift
from state to Church as the primary agency for document production
and preservation.
This book has eschewed inherited grand narratives and instead has
investigated the pattern and nature of the surviving documentation. And
we have shown that we cannot infer changing documentary practices
from changing patterns of documentary survival. We have demonstrated
that people at all social levels used documents for a variety of purposes
throughout Late Antiquity and the early Middle Ages. This is as true of
laypeople as it is of the clergy. People with no ecclesiastical affiliation,
ranging from the Apion family in fourth-century Egypt to prosperous
peasants in tenth-century Iberia, deployed documents in ways that were
unrelated to the needs of churches and monasteries. Incidental references
to such documents suggest that their use was thoroughly mundane, while
cases of their intentional destruction nevertheless reflect a sophisticated
appreciation of their value. At the beginning of our story, in the fourth
4 Cf. W. V. Harris, Ancient Literacy (Cambridge, MA, 2001), esp. 285–322, with M.
Handley, Death, Society and Culture: Inscriptions and Epitaphs in Gaul and Spain, 300–700
(Oxford, 2003).
5 See above, Chapter 2, and C. Holmes, ‘Provinces and Capital’, in L. James (ed.), A
Companion to Byzantium (Oxford, 2010), 55–66.
6 H. Pirenne, ‘Le commerce du papyrus dans la Gaule mérovingienne’, Comptes ren-
dus des séances de l’Académie des inscriptions et belles-lettres (1928), 178–91; Pirenne,
Mahomet et Charlemagne (Brussels, 1937), 124–5. Cf. M. McCormick, Origins of the Euro-
pean Economy: Communications and Commerce A.D. 300–900 (Cambridge, MA, 2001),
704–8.
366 Conclusion
century, the distinction between lay and ecclesiastical documentary cul-

tures does not even apply; as we go forward, the distinction remains
problematic. The ecclesiastical aspects of a shared documentary culture
may become more visible, but a systematic comparison with lay practices
indicates a continuum of values, imperatives and technologies. It is the
differential histories of transmission, allied to modern assumptions about
the Church as separate from and opposed to the state, that have created
a historiographical divide.
When we examine this shared culture, we find that documentary prac-
tices – the production and use of documents as well as their preservation
in archives – were shaped by the needs of institutions, but also by the
needs of the people who resorted to them. These needs did not always
overlap. The best-known examples of this phenomenon, and ones that
have loomed large in the preceding chapters, are the lay patrons of the so-
called proprietary monasteries, whose dossiers of title-deeds ended up in
those monasteries’ archives. The late antique gesta municipalia, although
established to respond to the fiscal needs of the state by keeping track of
tax obligations, also happened to meet the needs of local landowners by
authenticating their documents, and therefore legally strengthening the
transactions that those documents enacted.7 Similarly, landowners in
Italy from the Carolingian period onwards could and did take advantage
of the judicial process to obtain documentary support for their property
rights.8 In other words, for us, writing with hindsight and attempting
to tease out and then explain change over time, it is tempting to see
shifting patterns of document production and storage as evidence for
new institutions establishing cultural and social power. If we are theoret-
ically minded, we might even invoke the processes of domiciliation and
consignation, physical and interpretative control.9 But post-Roman soci-
eties were multifaceted, and the wealth of evidence for lay documentary
practice assembled in this book undermines a simplistic and misleading
teleology of ecclesiastical takeover. We must therefore put the agency of
those who needed documents and sought out archives into the picture
and see institutions as responsive to social needs. Control of archives and
control through archives need reading as complex and fundamentally
contestable processes, not as raw facts.
The ultimate disappearance of the late Roman state in the West did
not mean the disappearance of landowners and their needs; landowners
simply found new ways to achieve old purposes. The language of the gesta
municipalia continued to hold an attraction for some scribes in parts of
the medieval West. Ongoing references to the registration of documents
7 See above, Chapters 4 and 5. 8 See above, Chapter 9. 9 See above, 13.
Conclusion 367
in municipal archives in post-Roman contexts suggest that people were

still seeking public validation of their transactions, despite the evident
disappearance of the archives themselves.10 A similar conclusion can be
drawn from seventh-century estate accounts from Tours that still bear
the imprint of Late Antique fiscal administration. The rare survivals of
papyri in the ecclesiastical archives of the Merovingian monastery of
St-Denis and the archbishopric of Ravenna also perpetuate recognizably
late Roman forms.11 While the language used in these documents may be
archaic, the documents themselves are not fossils; they reveal practices
with living roots in late Roman institutions that were evolving to meet
changed circumstances. This evolution in documentary practices would
continue throughout the early Middle Ages.
A core purpose of documents has always been to record and secure
transactions and rights. This applies as much to a bureaucratic state as
to an individual landowner. Late Roman documentary culture encom-
passed both, but it was essentially driven by the state’s requirements both
to register and to enforce its fiscal claims. Neither these requirements,
however, nor the mechanisms through which they were achieved, were
static. The need for documents in the Roman world actually increased
in the third and fourth centuries as Roman government and Roman law
grappled with the consequences of a reshaped empire.12 One outcome
was the abandonment of strictures on the formal aspects of documentary
production, and a recognition that writing itself was a constitutive act,
something that has profound implications for how we read early medieval
documentary practices. In addition, the late Roman army also generated
a significant amount of documentation. The direct needs of the Roman
state, however, were not the only drivers of documentary production and
use. The very existence of great estates promoted an intense circulation
of documents.13
In this world, people used documents to meet their own needs.
Estate records, designed in the first instance to register the income of
landowners, also allowed peasants to document their creditworthiness.14

11 Tours: W. Goffart, ‘Merovingian Polyptychs: Reflections on Two Recent Publications’,
in Goffart, Rome’s Fall and After (London, 1986), 233–53 (orig. Francia, 9 [1982],
55–77); J. Durliat, ‘Qu’est-ce qu’un polyptyque? À propos des documents de Tours
(ChLA 659)’, in M. Sot (ed.), Media in Francia: Recueil de mélanges offerts à Karl-
Ferdinand Werner à l’occasion de son 65e anniversaire par ses amis et collègues français (Paris,
1989), 129–38; S. Sato, ‘The Merovingian Accounting Documents of Tours: Form
and Function’, EME, 9 (2000), 143–61. St-Denis: D. Ganz and W. Goffart, ‘Charters
Earlier Than 800 from French Collections’, Speculum, 65 (1990), 902–36; Ravenna:
above, Chapter 4.
12 See above, Chapters 2–4. 13 See above, Chapters 2–3. 14 See above, 28–31.
368 Conclusion
The Ravenna papyri reveal that at least in Ravenna the gesta municipalia
process, while clearly designed to help the state keep track of tax obli-
gations, also gave owners the ability to validate transactions and secure
their property.15 These sorts of practices could only work in a society
permeated by a more general use of documents at all social levels. The
existence of writing of this kind is demonstrated not only by the Egyptian
papyri but also by ephemeral documents such as receipts on ostraca or
even curses on slate found in Visigothic Spain.16
The entry of Christians into the public sphere in the fourth century
did nothing to change this situation. As already noted, the distinction
between lay and clerical was not manifested on papyrus. Christians (and
their developing institutions) simply participated in this fluid documen-
tary culture like everyone else, whether as owners of and workers on great
estates, commanders and soldiers in the army, or imperial bureaucrats
and municipal functionaries. It would take centuries for Christian institu-
tions to develop in a way that could affect broader documentary practices.
In late sixth-century Gaul, for example, Gregory of Tours could compare
his historiographical record of the gesta of the present to the public entry
of titles and obligations into the gesta of the cities that still structured soci-
ety; the great bishoprics that dominated Gregory’s landscape may have
had their own special archives, gesta episcoporum, but these were part of
a common culture of document keeping that permeated both royal and
local politics.17
As Gregory demonstrates, some aspects of late Roman documentary
culture survived in many parts of the West at least into the sixth century.
But by 700, document use was responding to a very different social order.
In a state that rested on a regular cycle of tax assessment and registration,
local documentation of fiscal liability was vital for both the operation of
government and the security of local landowners; the underlying principle
was well put by the Roman lawyer in the Burgundian kingdom who
explained that no transfer of fields or lands could be effected unless the
recipient was able to meet the tax liability and register his interests. In
post-Roman times, however, assessment – always a potential flashpoint
in local politics – became more and more vexed, as bishops rehearsed
supposedly ancient liberties on behalf of their churches and cities, and
ordinary landowners increasingly came to claim a new form of freedom,
connected with the bearing of arms in the service of their king and
15 See above, 70–82 and Chapter 5.

16 See above, Chapter 3, and Las pizarras visigodas: Edición crı́tica y estudio, ed. I. Velázquez
Soriano (Murcia, 1989), nos. 2–3, and pp. 613–14.
17 Ganz and Goffart, ‘Charters Earlier Than 800’.
Conclusion 369
the adoption of barbarian ethnicity, but also encompassing claims to

freedom from taxation. By the end of the seventh century, Marculf was
able to equate a ‘truly free’ man with one who ‘is not counted in the
public register’. The late Roman fiscal system had atrophied, and while
practices associated with it continued in places, at a local level they did
so as custom, detached from any wider discourse of public obligation or
fiscal law.18
Yet, although documents were no longer produced in direct association
with this fiscal system, the documentary habit persisted. One of the
ways that it persisted, and the best known, was among rulers. Thus,
diplomas of the Merovingian Frankish kings predominate among the
earliest charters in the archive of the abbey of St-Denis, to the extent that
the latter may be seen as in some way a ‘royal archive’. But this habit was
not limited to rulers. The few extant non-royal charters and wills from
this period (for example, the nine non-royal pre-Carolingian documents
from the St-Denis archive that have no connection to St-Denis itself)
show that the highest aristocracy also managed their affairs in writing and
sought out the same forms of security and storage as their royal masters.19
The earliest formula-books, meanwhile, indicate a much wider range of
business carried out with documents: manumissions, powers of attorney,
securities, receipts, quitclaims and so on. Similarly, the slate documents
of Visigothic Iberia suggest that the basic business of running rural estates
was being carried out in writing: the recording of oaths, professions of
servitude and simple accounting.20
There was, then, no fundamental change in how people used docu-
ments. The scarcity of evidence for lay documentary practice, however,
as well as the absence of the imprints of late Roman fiscality and above
all the survival of what little evidence there is in ecclesiastical archives,
has encouraged the view that laypeople in the West became marginal-
ized from documentary culture. Historians have therefore looked to the
Church as the ark in which late Roman documentary practices sur-
vived. Yet monks and clerics had not yet come to take leading roles
in documentary culture; in many places major churches and monasteries
18 This paragraph draws on W. Goffart’s studies, collected in Goffart, Rome’s Fall and After;
also M. Innes, ‘Land, Freedom and the Making of the Early Medieval West’, TRHS,
6th ser., 16 (2006), 39–74, and Innes, University of London inaugural lecture, 2007:
‘Roman Liberty and Gothic Freedom: From Late Antiquity to the Early Middle Ages’
(forthcoming). The quotations are from the Lex Romana Burgundionum 60 (MGH LL
nat. Germ. 2.1:159) and Marculf 1.19 (trans. Rio, Formularies, 153).
19 The St-Denis records are edited in ChLA XIII–XIV; for bibliography and discussion,
see above, 317–19.
20 Formularies: Rio, Legal Practice. Slates: above, Chapter 4.
370 Conclusion
had not even begun routinely and systematically to store their own
documents.21
In the Byzantine East, a bureaucratic, fiscal and military superstruc-
ture endured. The state, therefore, though it underwent profound trans-
formations, continued to provide institutional frameworks within which
documents circulated. The documentary practices and survival pat-
terns visible in the Byzantine world differ correspondingly from those
in the West.22 Here, Marculf’s famous seventh-century formula-book is
a harbinger of the future. It reveals a broadly shared legal and docu-
mentary culture, but it also exemplifies a moment when the institutional
resources available to landowners who wished to secure and preserve
documents were beginning to change.23 In Marculf ’s Neustria, churches
were arguably becoming the institutions of record by default as alter-
natives receded, as the continued deposit of ‘private’ documents in the
archive of St-Denis indicates. The remarkable run of early secular papyri,
complete with gesta municipalia protocols, in the archives of the church
of Ravenna immediately invites comparison. The chronology of change
differed from region to region, depending on the range of local possibil-
ities. Notably, churches were only one option among many for storing
documents. Several early Frankish formulas assume that laypeople were
keeping documents in their own homes.24 As late as the reign of Charles
the Bald, a lay aristocrat was able to resist the documentary claims of
the local bishop by producing documents stored at the disputed estate.25
Churches and monasteries themselves were subject to the same processes
whereby a dominant church in a region became the institution of record.
In Italy, the documents of the local churches in the relatively remote val-
ley of the Garfagnana were deposited in the episcopal archive in Lucca.26
21 See above, Chapter 8, as well as Hummer, Politics, 76–104; McKitterick, Carolingians,

77–126, esp. 122–4.
22 See above, Chapter 2. 23 Marculf; see above, Chapter 6.
24 Brown, ‘Documents’. 25 See above, 176–9.
26 For general comment on the ‘registering’ of charters at Lucca, see C. Wickham, The
Mountains and the City: The Tuscan Apennines in the Early Middle Ages (Oxford, 1988),
9, 18, with nn. 5, 27. Dorsal notes on some of the Garfagnana charters confirm this,
notably ChLA XXXVII 1074 (= CDL II 285), a charter concerning the church of
S. Maria in Campori and written in 773, on which a dorsal note in an eighth-century
hand reads ‘+ dos ecclesie Sancte Marie de Campulo’; in other words, the charter was
already, within a few years of its issue, filed according to the name of the church it
concerned, and therefore almost certainly not in that church. Since this charter in fact
transferred future control of the church to the bishop of Lucca, it seems very likely that it
was stored in the latter’s archive. A similar note appears on the dorse of ChLA XXXVII
1094, equally important for the association of the church with the bishopric. Archiving
according to the name of the estate is suggested by dorsal notes on the Garfagnana
charter ChLA XXXII 951, issued in 759.
Conclusion 371
This phenomenon can also be seen at Wissembourg, where, for exam-

ple, a series of transactions concerning the church at St Martin at Berg
was ultimately deposited among the records of the great monastery.27 In
short, when people and institutions accumulated land and built relation-
ships with each other, they accumulated documents and preserved them
through the most convenient means available. Churches in particular
were attractive for storage not because they controlled documentary cul-
ture, but because they were community centres and because they enjoyed
supernatural and physical protection. In rural archaeological contexts,
churches tend to be the first, and often the only, stone building. Even
wooden churches such as proliferate in the seventh and eighth centuries
were generally more substantial than the surrounding dwellings.
A long-standing paradigm in the historiography of this period con-
nects a particular set of ecclesiastical institutions – bishoprics and large
monasteries – to the rising Carolingian dynasty. As the Carolingians
gained power, they established patronage over bishoprics and certain
privileged monasteries in order to gain access to crucial material and
social resources. Identifying churches and monasteries as especially con-
venient and potentially well-organized centres of documentary culture,
they actively encouraged some of them to record and organize their rela-
tionships and income in writing in order to take advantage of local land
and patronage networks. A number of monasteries drew their particular
attention because they stood in strategically important locations: Nonan-
tola and Farfa, for example, benefited in this way after Charlemagne’s
conquest of Lombard Italy in 774, as did several of their Aquitanian and
East Frankish counterparts in the aftermath of conquest.28 Some eccle-
siastical institutions – for example, at St Gall and Salzburg – responded
to the encroachment of Carolingian power by expanding their use and
storage of documents.29
It is at precisely this point that we can see the great churches and
monasteries in the Frankish heartland trying to locate old documents and
to organize document storage and production in newly useful ways. The
earliest polyptychs and cartularies are manifestations of this impulse, in
part stimulated by the compact begun under Pippin whereby ‘excess’ land
27 See above, 199–200.

28 As is clear from F. Prinz’s map of ‘Schenkungen und Privilegien Karls des Großen’,
in W. Braunfels (ed.), Karl der Große: Lebenswerk und Nachleben, 5 vols. (Düsseldorf,
1965–8), I, 488. Cf. J. Semmler, ‘Karl der Große und das fränkische Mönchtum’, in
Braunfels (ed.), Karl der Große, II, 255–89, here 267–70, 273–8.
29 See M. Innes, ‘Property, Politics and the Problem of the Carolingian State’, in W. Pohl
and V. Wieser (eds.), Der frühmittelalterliche Staat: Europäische Perspektiven (Vienna,
2009), 299–314, and H. Wolfram, ‘Die Notitia Arnonis und ähnliche Formen der
Rechtssicherung im nachagilolfischen Bayern’, in Recht und Schrift, 115–30.
372 Conclusion
was distributed on the king’s word.30 So are ecclesiastical and monastic

archives and staffs of clerical and monastic scribes. And it is only at this
point that we can say with confidence that the overwhelming majority of
scribes in the Frankish kingdoms – even those who served lay consumers –
were clerical or monastic. This is evident from the patterns of charter
production at St Gall and Wissembourg.31 This is also the period in
which churches and monasteries began compiling and copying on a large
scale collections of formulas, many of which were relevant only to lay
users. Gregory of Tours in the preface to his Histories had alluded to
a multi-tiered and multivalent literate culture structured by the regular
record of deeds (gesta) within cities that were increasingly dominated by
bishops; Regino of Prüm, writing two centuries later, compared his work
as a historian to the workaday task of a notary making a legal notice
commemorating praiseworthy actions.32
In Francia, therefore, it appears as though the Church had in fact
taken over documentary practice. The traditional picture of a clerical
and monastic monopoly of documentary culture, however, generalizes to
the entire early Middle Ages a reality that was in fact a consequence of
Carolingian efforts to strengthen their regime in certain crucial regions.
Furthermore, it ignores the continued and lively use of documents by
laypeople both within and outside this ecclesiastical framework, as wit-
nessed by the intensive record-keeping at Perrecy, Folkwine’s anxiety to
maintain documentary proof for his shady dealings, or the dozens of
secular transactions that later found their way into the archives of the
monasteries of Cluny and Redon.33
Our mental map tends to normalize and generalize the Carolingian
story precisely because of the patterns of documentary survival. The
Carolingian situation rests on the fact of large ecclesiastical institutions
that serviced both their own needs and those of the imperial aristoc-
racy and the royal court. As managers of estates, people and memory,
Carolingian ecclesiastical institutions developed characteristic forms that
they bequeathed to later centuries: estate registers, cartularies with a
commemorative function, ecclesiastical notaries serving the wider com-
munity, and archives of ecclesiastical property records that might incor-
porate dossiers of lay documents. As it is the ecclesiastical cartularies and
archives that have tended to survive, it is they that dominate the tractable
30 For the earliest stages of this impetus, see I. Heidrich, ‘Das Breve der Bischofskirche von
Mâcon aus der Zeit König Pippins (751–768): Mit Textedition’, Francia, 24 (1997),
17–37.
31 See above, 155–73, 214–21.
32 Gregory of Tours, Libri historiarum, preface (MGH SRM 1.1:1); Regino of Prüm,
Chronicon, preface (MGH SRG 50:1).
33 See above, Chapters 7 and 11.
Conclusion 373
record with which modern historians grapple. When, from the twelfth
century onwards, secular institutions and aristocratic families did feel the
need to create dedicated long-term archives on the model of those forms
utilized by Carolingian and post-Carolingian ecclesiastical institutions,
our picture of document use changes.34
The power of the traditional model has also tended to obscure both
variation within the Carolingian heartland itself and the existence of dif-
ferent situations on its periphery. In areas with documentary habits con-
tinuing from Merovingian and ultimately late Roman times, lay landown-
ers as well as churches had a variety of documentary resources on which
to draw, even though it is the churches that loom large in the surviving
evidence.35 In those areas that experienced a distinct sub-Roman hiatus,
and those that were never part of the empire, documentary practices had
to be effectively reinvigorated or created ab initio by late Merovingian
and early Carolingian churchmen. The emergent elites of the Carolin-
gian period in these latter regions were induced to use documents in
ways similar to their Neustrian and Burgundian cousins: they met their
needs by piggybacking on the new ecclesiastical institutions to which
they were so closely tied and on the documentary culture that had devel-
oped around them.36 The formula evidence illustrates clearly the pro-
cess of cultural translation that was often involved.37 These differences
are reflected in the contrasting documentary profiles of the regions of
the Frankish world. Ninth-century cartularies were entirely a product of
East Frankish churches that were newly founded in the eighth century
and that rapidly amassed great quantities of landed wealth; although in
many areas it is clear that these churches did not bring documentary
practices into contexts where they were previously wholly lacking, the
resulting picture is nonetheless one of churches and their scribes rapidly
becoming the dominant institutions of record, preserving rich traditions
34 E.g. E. Noichl, Codex Falkensteinensis: Die Rechtsaufzeichnungen der Grafen von Falken-
stein (Munich, 1978), online with facsimile at www.bayerische-landesbibliothek-online.
de/codexfalkensteinensisen, and W. Rösener, ‘Codex Falkensteinensis: Zur Erin-
nerungskultur eines Adelsgeschlechts im Hochmittelalter’, in Rösener (ed.), Adelige und
bürgerliche Erinnerungskulturen des Spätmittelalters und der Frühen Neuzeit (Göttingen,
2000), 35–55; Littere baronum: The Earliest Cartulary of the Counts of Champagne, ed.
T. Evergates (Toronto, 2003); A. J. Kosto, ‘The Liber feudorum maior of the Counts
of Barcelona: The Cartulary as an Expression of Power’, Journal of Medieval History,
27 (2001), 1–22; H. Débax, ‘Le cartulaire des Trencavel (Liber instrumentorum vice-
comitalium)’, in Les cartulaires, 291–9; Liber instrumentorum memorialium ou cartulaire
des Guillems de Montpellier, ed. A. Germaine and C. Chabanneau, 3 vols. (Montpellier,
1884–6); A. Lewis, ‘The Guillems of Montpellier: A Sociological Appraisal’, Viator, 2
(1972), 159–70; also above, Chapter 12.
35 See above, Chapters 7 and 11. 36 See above, Chapter 8.
37 See, for example, the Bavarian formula derived from Marculf, in which a lay archive
destroyed by fire was kept not in the plaintiff’s home but rather in ‘his’ monastery:
Brown, ‘Documents’, 351–2.
374 Conclusion
in which lay documents, while not wholly absent, are elusive. In West
Francia, on the other hand, the Carolingian takeover reshaped the doc-
umentary and ecclesiastical landscape in more subtle ways, encouraging
churches to marshal documents about their landholdings, for sure, but
impinging less dramatically on long-standing documentary practices that
had evolved from Roman times. Here, we have a more indirect insight
into the entirety of Carolingian documentary culture, via the carefully
selected acts favoured by eleventh- and twelfth-century cartulary compil-
ers; only in exceptional contexts of transmission, as at Cluny or St-Denis,
or where cartularists made unusual choices, as at Perrecy or Redon, do
we get a glimpse of all aspects of Carolingian documentation.38
Anglo-Saxon England provides a control.39 In many ways it paral-
lels very well those areas of East Francia where documentary traditions
were weak. In a similar fashion, it absorbed traditions from the conti-
nent already formed. However, even allowing for the distorting impact
of Viking destruction and associated political changes, it is striking that
many of the institutional forms we associate with the Carolingian and
post-Carolingian heartlands were adopted late and intermittently in Eng-
land: no polyptychs or cartularies survive from there until after 1066. In
addition, the documentary forms of Anglo-Saxon charters have their
own peculiarities deriving from the complex circumstances of their ori-
gins, and from the mid ninth century onwards they increasingly drew on
the written vernacular.
In Italy, scribes were normally laymen throughout our period. That
demonstrable fact is part of the evident continuity of institutions in the
civitas centres, which may have been ecclesiasticized in the sense that
they became focused on the cathedral church but in many fundamen-
tal respects remained late Roman. In their diplomatics, in their script,
and in many areas in their use of papyrus, charters evolved directly from
Roman predecessors.40 Aristocracies in both Lombard and Byzantine
Italy therefore had routine access to a variety of documentary resources.
The social and material distance between those aristocrats and the more
substantial landowning cultivators was less than in Francia.41 As a con-
sequence, documentary culture penetrated as far as the peasant class.
38 East Frankish cartularies: see above, Chapters 7, 8 and 11; Geary, Phantoms, 81–113.
40 See above, Chapter 4; Everett, Literacy in Lombard Italy (Cambridge, 2000), 197–234;
A. Pratesi, Genesi e forme del documento medievale (Rome, 1979), 47–61. On the theme
of the continuity of diplomatic, see the foundational L. Schiaparelli, ‘Note diplomatiche
sulle carte longobarde’, Archivio storico italiano, 17 (1932), 3–34; 19 (1933), 3–66; 21
(1934), 3–55.
41 See C. Wickham, ‘Aristocratic Power in Eighth-Century Lombard Italy’, in A. C. Mur-
ray (ed.), Sources and Narrators of Early Medieval History (Toronto, 1998), 153–70, with
qualifications by Costambeys, Power, 165–6.
Conclusion 375
The social distance was even shorter in Iberia, where we have even more
evidence for the use of documents by peasants. The Iberian evidence
from the tenth century shows something of which we already have signs
in eighth-century Italy, as well as in Francia: the formation of lay archives
that were later absorbed into ecclesiastical institutions, sometimes insti-
tutions founded by the very families who assembled those archives. In
Italy, the emergence of communes in the late eleventh century meant that
there were institutions through which the laity might potentially regain
a measure of control over record production and storage. Political and
economic developments elsewhere in Europe around the same period
would bring similar results.42
All of these observations tell us that we should contrast not clergy and
laity, but rather institutionalized and non-institutionalized actors. This
distinction is perhaps a better way to think about the nature of early
medieval records and the patterns of their survival than the traditional
lay–clerical divide, which, as we have consistently noted, is problematic.
A priest who, along with four siblings, sold a plot of land to a lay couple
can hardly be thought of as a representative of the institutional Church;
that is, as a clericus in its twelfth-century sense. Though our evidence sug-
gests that both he and his buyers would have wanted the sale recorded
and validated by a document, that document would most likely not sur-
vive more than a few generations without the involvement of a long-lived,
transpersonal institution that could preserve it. Only from the seventh
century did some ecclesiastical institutions – churches and monasteries –
develop that could perform this role. But the kinds of documents that an
institution will preserve are determined by that institution’s documentary
interests. The primary raison d’être of the late antique gesta municipalia
was to keep track of tax obligations. The documentary interests of early
medieval churches and monasteries, in contrast, were largely driven by
their own property, rights and privileges. Unlike the gesta municipalia,
many early medieval churches and monasteries, and thus their archives,
survived into modern times along with the institutional religion they
represented. Yet, though our document of sale may well have been pro-
duced by a clerical or monastic scribe, it would not enter an ecclesiastical
archive to be passed down to us unless the property involved ended up
in a church’s or monastery’s possession, or unless the seller or the buyer
decided to deposit it there for safe keeping. Even in an Italy with a thriving
secular notariate, churches and monasteries were often the repositories
of first choice for documents, very probably because they had institu-
tional resources and operated with an institutional mindset that before
the year 1000 was still largely absent within the laity. Such documents
42 See also above, Chapters 9–12.

376 Conclusion
nevertheless still risked being thrown out at some future point unless the
information they contained (or their age and suitability as the basis for
forgeries) came to meet a clerical or monastic need.
We have shown that it is possible to study lay documentary practices, and,

indeed, lay archives, in a period in which their very existence has been fre-
quently doubted. We have presented evidence for the use of documents
by people from all social strata, including those with no connection to
churches or monasteries. It is clear that the ecclesiastical sources that
continue to dominate perceptions of early medieval society were just one
segment of a shared documentary culture. We have also offered a new
narrative of institutional change that explains the patterns both of docu-
mentary production and of evidentiary transmission. One of the effects
of vindicating a documentary culture that extended beyond the Church
is to open up a range of new questions. For example, the evident vol-
ume of land sales adds new layers of complication to an economy usually
seen as dominated by the gift. Similarly, the use of documents for the
transfer of property within families, as visible in the lay dossiers preserved
in monastic archives, invites us to reframe fundamental questions about
early medieval social structures.43
Most importantly, however, we need to rethink archives. Our inherited
model is tied up with the process of modern state formation. As we noted
at the outset of this book, we tend to think about archives in terms of
reference literacy and so as tools of governing institutions. Late antique
and early medieval archives were different. Any attempt to assess them
needs to avoid modern expectations. When we think of them, we should
not imagine a giant central repository accessible only to experts, com-
plete with index and classification systems. Instead, we should think of
small collections of folded packets of charters, of related documents tied
together or pierced with ritual sticks, of fragments of connected narrative
linked by people and places. Stuffed into chests or boxes, kept in homes
and local churches, most of these collections would only make sense to
their possessor. As historians, we should respect their very scrappy mate-
riality and their very local logic. This is how written documentation was
experienced by late antique and early medieval people.
43 L. Feller and C. Wickham (eds.), Le marché de la terre au moyen âge (Rome, 2005);
W. Davies, ‘When Gift Is Sale: Reciprocities and Commodities in Tenth-Century Chris-
tian Iberia’, in W. Davies and P. Fouracre (eds.), The Languages of Gift in the Early Middle
Ages (Cambridge, 2010), 217–37; Davies, Acts of Giving, esp. 44; Davies, ‘Intra-Family
Transfers in Southeastern Brittany: The Dossier from Redon’, in Les transferts, 881–94.
Index
Aachen, 14, 15, 161 Algeria, 1, 37, 38, 39, 40, 42, 50, 53, 54,
Abbots Bromley, monastery, 353 59
Abingdon, monastery, 355, 357, 361 Alps, 96, 165, 172, 237
archive. See ecclesiastical archives Alsace, 191, 192, 193, 199, 206, 207, 212,
cartulary. See cartularies 213, 216, 220, 224
Chronicle, 359 Ambrose, bishop of Milan, 137
Achmetha, palace, 17 andelangus, 194, 195
Adalhard, count, 201 Andreas, scribe, 167
Adalland, cancellarius, 219, 221 Angers, 113–15, 117
Adelchis II, 329 formulas. See formulas
Admiralty Island, 49 Anglo-Saxon England. See England
ostraca. See ostraca Ansur, landowner, 279
Ado, priest, 199, 201 dossier. See dossiers
Aegefredus, landowner, 318 Antioch, 19
Aeglidulf, bishop, 144 Antonine Constitution, 67, 68
Ælfwold, layman, 350, 351 Anulinus, proconsul, 56
Aelianus, proconsul, 56 Aphrodito, village, 22, 26, 30
Æthelflæd, ruler of the Mercians, 356 Apion family, 21, 23, 25, 27, 365
Æthelfrith, thegn, 356 archive. See lay archives
Æthelhelm Higa, 337, 345, 346, 347 Apronianus, advocate, 54
Æthelmund, thegn, 349 Aquileia, patriarch of, 257
Æthelred II, king, 352, 353, 354, 360 Aquitaine, 179, 192
Æthelred, ruler of the Mercians, 356 Aragon, 260
Æthelric, son of Æthelmund, 350–1, 352 Arbeo, bishop and scribe, 226, 228
Æthelstan, king, 355 arca, 175, 326
Æthelthryth, landowner, 348 arcarius, 48, 50, 75
Æthelwine, ealdorman, 357 archives, 12–15, 376. See also document
Æthelwulf, landowner, 348 preservation
Æthylric, son of Æthelmund, 349–50 ancient Near East, 17–18, 20, 25,
Aganbert, mandate bearer, 113, 115, 116 33
Agello, estate, 245 archivum, 155
dossier. See dossiers Byzantine, 21–33
Alahis, vir magnificus, 1, 94, 239, 248 Carolingian, 155, 185–6
archive. See lay archives definition of, 13, 20
Alaric II, 69, 85 ecclesiastical. See ecclesiastical archives
Alberic, princeps, 330 Egypt, 26
Albertini Tablets, 40–2, 50, 52, 54, 58 gesta municipalia. See gesta municipalia
Alcuin, 136, 150 history, early modern, 291–4
Alemannia, 117, 119, 150, 157, 159, 160, lay. See lay archives
172, 173 local, 31, 168, 170–1, 182, 198–203,
Alexandria, 19, 31, 137 204, 241–2, 306, 307, 356–7,
Alfred, king, 345, 346, 347 370–1
377
378 Index
archives (cont.) Bierre, villa, 302, 305, 306, 309

North Africa, 55–8 Bir Trouch, Algeria, 37, 42
practice, 155, 156–7, 159–60, 169, 170, ostraca. See ostraca
283–4, 312–20 birch bark, 91
public. See public archives Bitterauf, Theodore, 225
residential, 87, 326–7, 370 Bloch, Marc, 286
Argemundo, landowner, 276 Bobbio, monastery, 321–2
Argimirus, servant, 266, 268 Boniface, bishop of Mainz, 204, 211
Arley, estate, 353–4 Bonifatius, duke, 210, 211, 214, 215
Arno, bishop of Freising, 226 Bonnassie, Pierre, 267
Ascrich, clerical scribe, 227 Book of Ezra, 17
Asger, monastic scribe, 223 Book of Judgements, 83, 84, 85, 93
Astrulfo, servant, 265–6, 268 bookland, 343
Asturias, 260, 268 books of traditions. See cartularies
Atelier de Recherche sur les Textes Bougard, François, 239, 248, 252–3
Médiévaux (ARTEM), 319 Bourges, 107, 112, 178, 180
Athanasius, bishop of Alexandria, 137 church of, 180–1
Atsma, Hartmut, 289 formulas. See formulas
Atto, bishop of Freising, 121, 160 Boynton, Mark, 346
Audo, bishop of Verona, 234 Brescia, 238, 334
Augustine, bishop of Hippo, 41, 45, 55, breve, 1, 47, 75, 254
56 Breviarium. See Lex Romana Visigothorum
Ausonio, servus, 268 Brevium exempla, 220
authentication. See documentary records Brihtwaru, landowner, 349
Autun, 292 Britain, 34, 255, 340
Autunois, 305 Brittany, 320
Ava, abbess of St-Pierre-les-Nonnains, Brooks, Nicholas, 348
297, 298, 301 Brosius, Maria, 20
Aymard, abbot of Cluny, 295 Bruel, Alexandre, 288, 294, 295
Azo, landowner, 308 Burgundian kingdom, 368
Burgundy, 127, 143, 149, 155, 185, 188,
Babylon, 17 207, 292, 294, 297
Baldram, cancellarius, 221 Burton Abbey, 352, 353, 359
Balfred, landowner, 163, 168 archive. See ecclesiastical archives
Banaji, Jairus, 34 cartulary. See cartularies
Barret, Sébastien, 289, 291, 307, 312 Bury St Edmunds, monastery, 362
Barthélemy, Dominique, 286, 287 Byrhtnoth, abbot of Ely, 350, 357
Baugulf, count, 216 Byzantine Empire, 14, 363
Bavaria, 117, 119, 120, 121, 150, 187, archives. See archives
192, 195, 198, 225, 229
Beatus, abbot of Honau, 212 Cabinet des Chartes, 292
Becerro gótico. See cartularies Caecilian, bishop of Carthage, 56
Bede, 34 Cambrai, 319
Belisarius, 78, 80 Campbell, James, 34
Benevento, principality, 243, 329 Cannstadt, 157, 162
Benignus, clerical scribe, 226 Canterbury, 356
Berardus, abbot of Farfa, 331, 332 archive. See ecclesiastical archives
Berg, estate, 199 church, 337, 342
Bergamo, 238 Capitulare Ticinense, 322
archive. See ecclesiastical archives capitulum, 159, 164, 165, 269
Bernard, Auguste, 288, 294, 295 Caracalla, emperor, 67, 68
Berno, abbot of Cluny, 290, 295, 307 cartae. See charters
Bibaro, castellum, 331, 332 Carthage, 38, 44, 45, 48, 49, 54, 55, 56,
Biberkirch, church, 201 59
Bibliothèque Nationale, 294, 299, 300 conference of, 44–5, 56
Index 379
cartulae obligationis, 329 dorsal notes, 157, 161, 163, 165, 167,
cartularies, 11, 12, 15, 153, 154, 160, 169, 170, 229, 296, 301, 309, 312,
162, 185–7, 190–5, 229–30, 232, 315
371 Italy, 236–8
Abingdon, 355 Lucca, 237, 241, 245
Becerro gótico (Sahagún), 2, 277–8, materiality of, 307–10
279–80 Monte Amiata, 237, 241, 242, 245,
Burton, 353 258
Casauria, 237, 242–3 originals, 11, 15, 153, 154, 155, 156,
Cluny, 294–6, 298. See also charters, 162, 167, 184, 190, 212, 232, 237,
Cluny 238, 240, 245, 256, 263, 273,
Farfa, 237, 238 277–8, 279, 295, 296, 299, 307,
Fernándo and Elvira, 275 309, 312, 315–19
Freising, 160–2, 192, 193, 197, 201, Osona-Manresa, 262–3, 267
211 Otero, 273–5
Fulda, 192–3, 197, 201–4, 211 Piacenza, 237, 245, 258
Guimarães, 259, 266 Rankweil, 165–73
Honau, 193, 212 Sahagún, 276–80
lay copies in, 195–205 St-Denis, 184
Mondsee, 191, 192, 193 St Gall, 119, 120–2, 154–5, 156–8,
Murbach, 193 162–4, 212, 285. See also charters
Passau, 191, 192, 193 Rankweil
Perrecy, 176–7 Verona, 238, 256–7
Regensburg, 191, 192, 193 Chateau, villa, 302, 305
Salzburg, 193 cheirographeia, 29
scholarship of, 287–8 Childeric III, king, 118
selectivity of, 164, 172, 194, 296–300, Chilperic I, king, 19
315–16 China, 364
Sobrado, 269–70, 272 Chindaswinth, king, 84, 86
St-Denis, 317 chirograph, 338, 349, 350
Subiaco, 330 Chroccus, scribe, 201, 218
Wissembourg, 192, 193, 195, 197, 199, Chrodoin, abbot of Wissembourg, 211
201, 210–11, 213, 214 Chrodoin, donor and scribe, 200, 201,
Cartularium Langobardicum. See formulas 211, 217–18, 221
Casauria, monastery, 237, 243–5, 250, Chur, 166, 167, 168
332 Cicero, 364
archive. See ecclesiastical archives Cingoli, commune, 334
cartulary. See cartularies Cirta, town, 54, 55
Cassiodorus, 65, 74 Civitella, lords of, 334
Catalonia, 6, 190, 237, 260, 261, 262, Clanchy, Michael, 9
267, 269 clergy, 4–5
Caucilla, villa, 306–7 clerical identity, 232
Chaplais, Pierre, 341, 342, 343 Cluny, monastery, 197, 229–30, 283–319,
Charlemagne, 14, 112, 113, 119, 143, 372, 374
158, 159, 166, 167, 172, 192, 212, archive. See ecclesiastical archives
213, 248, 256, 371 cartulary. See cartularies
Charles Martel, 133, 213 charters. See charters
Charles the Bald, king, 178, 179, 370 historiography, 286–7
Charles the Fat, king, 159 Cnut, king, 355
charters, 11–12, 58. See also cartularies codex, 21, 66
cartae pagenses, 128, 130, 131, 160 Codex Diplomaticus Cavensis, 328
cartae profiliacionis, 264 Codex Eberhardi, 202
cartae publicae, 247 Codex Falkensteinensis, 230
Cluny, 286, 288–91. See also cartularies Codice diplomatico longobardo, 232
diplomatic scholarship of, 284–5, 287–8 Collectio Anselmo dedicata, 233
380 Index
Collectio Flaviniacensis. See formulas. Dioscorus archive. See lay archives

Flavigny Psimanobet family
commemoration. See document use and diplomas, royal, 153, 249, 341, 343–4
practice diplomatics, 11. See also charters
Commission des Réguliers, 292 Disticha Catonis, 86
Concordat of Worms, 257 document preservation, 190–1. See also
Conrad II, king, 329 archives
Constance, 119, 158 early modern, 291–5
Constans II, emperor, 61 ecclesiastical, 2, 82, 94, 150–1, 210,
Constantine, emperor, 22, 46, 56, 68, 73, 231–3, 242–5, 254–8, 295–301,
88, 137 305–7, 318–20, 325, 332–3, 334–5,
Constantine IV, emperor, 48 344, 354–5, 357–9, 364, 370–2,
Constantinople, 15, 19, 32, 54, 93 375–6. See also cartularies
Constantius II, emperor, 43, 137 late Roman, 54–9
Copenhagen, 128, 137, 140, 144 lay, 70, 130–1, 134–5, 191, 197–8, 325,
Corniero, estate, 274, 278 326–8, 330–5. See also dossiers and
Costamagna, Giorgio, 235 lay archives
Coventry Abbey, 354 mediums, 10–11, 307–10
Cozroh, scribe, 160–2 document production, 66–70
Cremona, 238 ecclesiastical, 2, 127, 218–20, 223–5
Crispinus, landowner, 241 local, 167–8, 214–18
Cubitt, Catherine, 338 royal demands, 343–4
curia, 23, 48, 65, 71, 73, 74, 76, 77, 78, state and fiscal demands, 26, 37, 42–9,
79, 80, 81, 97, 100, 103, 104, 105, 52–3, 61–2, 72, 78–9
106, 108, 109, 112, 113, 115, 135 document use and practice
curiales, 73, 74, 78, 81, 113, 116, 207 commemoration, 159, 161, 164, 192,
Cyrus I, king, 17–18 210, 313
ecclesiastical, 6–7, 136–51
Daghilinda, landowner, 162 late Roman, 6, 49–54
Dagobert I, king, 210 lay, 2–4, 7, 126–7, 130–1, 133, 134,
Danelaw, 344 136, 162–4, 171, 267–8, 289–91,
Danube, 191 324–5, 326, 330–2, 337, 346–52
Darius I, king, 17–18 legal proceedings, 53–4, 88–9, 178–81,
Davies, Wendy, 264, 267 249–54, 310–12, 328
Decius, emperor, 21 public validation, 97–8, 116–22, 123–4,
decurions. See curiales 178, 180–1, 184, 246–54, 311–12,
defensores, 75, 77, 79, 81, 99, 103–15, 116 320. See also gesta municipalia
defensor civitatis, 75, 76, 78 documentary culture
defensor ecclesiae, 76 Carolingian, 154–5, 174–6, 184–8, 229
Déléage, André, 300 common, 7–8, 365–6
Delisle, Leopold, 294, 295, 298 early medieval Europe, 364–5
Delogu, Paolo, 246, 247 ecclesiastical, 130, 132
Demetrius, consort, 326, 330–1, 332 England, 348–9, 360–2, 374
deperdita, 15, 181, 196, 259. See also Frankish, 127, 129–30, 151, 153–5,
documentary records, loss 174–6, 182–8, 370–4
Depreux, Philippe, 136, 150 East Francia, 174, 185–7
Derbyshire, 352, 353 West Francia, 174–6, 185
Derrida, Jacques, 13 Iberia, early Middle Ages, 280–1
Desiderius, king, 239 Italy, 236–7, 374–5
Desiderius, landowner, 90 early Middle Ages, 257–8
Dhuoda, laywoman, 8 late Roman, 33–5, 62, 363, 367–8
Didaco, son of Menendo, 260 lay, 126–7, 149
Diego Álvaro, Spain, 86, 92 local, 187–8
Dienheim, vicus publicus, 203, 204 Merovingian, 153, 174, 206–10,
Diocletian, emperor, 53, 66 213–18
Index 381
public vs. private, 183–4 Totone of Campione (Italy), 233–4, 236,

Roman legacy, 63–4, 70, 132–3, 135–6, 239–40
152–3, 182–5, 187–8, 325, 340–1, Valdávida (Iberia), 277
366–7, 368–70 Vico Teatino (Italy), 244
documentary records, 8 Wichard’s (Burgundy), 308, 312, 316
authentication, 86–8, 342–3 Dragobodo, bishop of Speyer, 210, 211,
destruction, 19, 328–30, 331 215
ecclesiastical. See cartularies, charters Drucio, oblate, 168, 169
and ecclesiastical archives Duby, Georges, 287
economic motives, 25–6 Dumville, David, 346
England, 340–2 Durham, church of, 360–1
estate, 37–9, 91–3, 367–8. See also Durliat, Jean, 25
dossiers and lay archives
Iberia, 260–1 Eadred, king, 359
lay, 5–6, 125, 131–2, 164, 195–205, Eadwig, king, 343
230, 238–42, 256, 262–6, 270, 284, East Anglia, 351, 359, 361
289–91, 300–7, 317–20, 340. See Eberhard, count, 207
also dossiers and lay archives will. See wills
loss, 19–20, 36, 47, 130, 134, 147, 164, Ebling, Horst, 207
165, 177, 181, 212, 307, 318, 342, Eblo, count, 299
362. See also deperdita Eccard of Mâcon, 176–7, 178–9, 180–2
and cartularies selectivity ecclesiastical archives, 6–7, 87, 123, 139,
mediums, 20–1, 66, 68–9, 307–11 154, 212, 229, 237–9, 241–5,
ostraca. See ostraca 255–6, 257–8
private, 37–42. See also dossiers and lay Abingdon, 357, 358
archives archivum ecclesiae, 87
public vs private, 21–3, 49–50, 71–2, Bergamo, 238
254 Burton, 352–3, 354
slates. See slates Canterbury, 337, 356–7, 358
state and fiscal demands, 74, Casauria, 242–3, 245
368–9 Cluny, 229–30, 288, 291–5, 297,
vernacular, 337, 342, 362 312–13
donations, 74, 79, 190 diocesan, 204
carta donationis, 195, 265 Farfa, 235, 248–9, 332
epistula donationis, 79, 118 Flavigny, 138, 150
epistula traditionis, 77 formation of, 218–29
scriptura donationis, 265 Freising, 161
traditio, 113, 114, 116, 194–5, 226 Fulda, 225
Donatists, 44, 45, 48, 54, 56 incompleteness of, 210–13
dorsal notes, 58. See also charters lay roots of, 241–5, 269–81, 353–5
dossiers, 20, 58, 189–90, 300–13, 332, Lucca, 238, 245, 370
366, 372 Monte Amiata, 238, 242, 245
Agello (Italy), 242, 245 Otero, 273–4, 275, 276, 278
Ansur’s (Iberia), 279 Piacenza, 245
Flaı́nez (Iberia), 278–9 Piasca, 277
Folkwine (Rhaetia), 166–7, 170, 171, Pisa, 1
189, 201 Ravenna, 76–7, 82, 184
Geminius Felix (North Africa), 52 Sahagún, 276–7
Iscam and Filaura (Iberia), 279–80 San Toribio, 276
Karol’s family (Italy), 243–4 Sobrado, 272, 274, 276
Lilia’s (Burgundy), 301–6, 309 St-Denis, 184, 317, 369, 370
Melic’s (Iberia), 277–8 St Gall, 120, 156–7, 159–60, 162, 165,
Perrecy (Francia), 176–82 169, 172–3, 271
Peter de Niviano (Italy), 245 Verona, 238, 256
Piasca (Iberia), 276–7 Wissembourg, 219–20
382 Index
École des Chartes, 293 Fernándo Vermúdez, landowner, 274, 275,

Edgar, king, 343, 349, 353, 358–9 280
Edward the Confessor, king, 341 archive. See lay archives
Edward the Elder, king, 344, 345, festuca, 194, 195, 296, 308, 309
356 fides publica, 247, 248, 333
Egypt, 15, 19–20, 21–31, 36, 37, 46, 50, Filaura, peasant, 2, 279–80
72, 188, 363, 365 dossier. See dossiers
archives. See archives Filippus, scribe, 324
documentary culture. See documentary Firmus, general, 43
culture late Roman Flaciano, servus, 264, 268
papyri. See papyri Flaı́nez, landowner, 273, 278, 280
Einhard, 155 archive. See lay archives
El Natahoyo, villa, 274 dossier. See dossiers
Elvira, landowner, 274, 275, 280 Flavigny, monastery, 105, 109, 128, 136,
cartulary. See cartularies 138, 142, 150
Ely, monastery, 344, 350, 351, 357, 358, archive. See ecclesiastical archives
361 formulas. See formulas
Emhilt, abbess of Milz, 202–3 Fleury, monastery, 176–7, 182, 320
Emilia-Romagna, region, 237 Flodoard of Rheims, historian, 176
England, 34, 261, 336–62, 374 Floriano, Antonio, 262
documentary culture. See documentary Folkwine, landowner, 167, 168, 170, 173,
culture 188, 372
documentary records. See documentary dossier. See dossiers
records Fonthill, estate, 336, 337, 345, 346, 347,
Erchanbert, bishop of Freising, 161 348, 356
Erembert, bishop of Freising, 211 Letter, 336–7, 339, 340, 342, 344–9,
Erhart, Peter, 156 350, 351, 356, 362
Ermbert, abbot of Wissembourg, 199 forgery, 70, 78, 85–6, 88, 247
Ermengildo, landowner, 270–2, 280 Formulae Visigothicae. See formulas
archive. See lay archives formulas, 15, 65, 164, 263
Ermentrude, testatrix, 208, 209 Angers, 106, 107, 114, 126, 150
Ermilla, servant, 266, 268 Bourges, 108, 112
Ervig, king, 84, 87 Cartularium Langobardicum, 237, 250,
Esla river, 274 329–30
estate records. See documentary records Flavigny, 108–12, 123, 128, 136–49
Etymologies, Isidore of Seville, 143 Frankish, 1–2, 16, 71, 76, 77, 91, 96,
Euric’s Code, 84, 89 125–7, 194, 209, 370
Eusebius, Ecclesiastical History, 138 gesta municipalia. See gesta municipalia
Evalia, landowner, 163, 168 indecolum ad laicum, 130
Ewig, Eugen, 185 manuscripts, 107–12, 137, 140, 150–1
exemplaria, 201, 205, 218 Marculf, 109–12, 113, 117, 118, 119,
Exeter, church, 355 128–33, 134, 135, 136, 137, 138–9,
141–2, 144, 145–7, 150, 160, 207,
Faenza, town, 77, 80 370
Fafila, landowner, 272, 273 Tours, 109–12, 128, 133–6, 137,
Fano, villa, 259 138–41, 142, 145, 147, 150, 266
Farfa, monastery, 235, 237, 248–9, 331, Visigothic, 65, 71, 76, 83, 88, 89, 90,
332, 371 91, 263–4, 266, 268
archive. See ecclesiastical archives Francia, 6, 8, 15, 16, 96–8, 102, 150, 153,
cartulary. See cartularies 191, 206, 230, 235, 236, 340, 343,
Farmer’s Law, 32, 33 374, 375
Fater, capellanus, 226–7 documentary culture. See documentary
Felix of Aptunga, bishop, 56 culture
Felix, bishop of Aptunga, 54 East Francia, 117, 160, 185, 186, 192,
Feller, Laurent, 243 286, 374
Index 383
West Francia, 117, 174, 175, 186, 192, public validation, 101, 106–7, 366–7,
206, 286, 287, 374 368
Franciscans, 334, 335 Ravenna, 64–5, 70–82, 96, 98–102, 122,
Frankfurt, 119 133, 330, 368, 370
Fredelus, local agent, 178, 188 Remiremont, 207–8
Frederick I, emperor, 334 Roman legislation, 73–5
Freising, church of, 121, 160, 161, 191, gesta proconsularia, 45
197, 203, 225–9, 230, 314 Ghittia, nun, 1, 10, 11, 16, 239
archive. See ecclesiastical archives list, 239, 241, 248, 263, 275
cartulary. See cartularies Gibbon, Edward, 364
French Revolution, 255, 291, 293, 301 Giovanni di Berardo, archivist, 243,
Froileba, landowner, 273, 278, 280 245
Fruela Muñoz, count, 273 Gisulfo, landowner, 271
Fruela Vimaraz, landowner, 274 Gloucestershire, 349
Fulda, monastery, 191, 193, 197, 201–5, Goffart, Walter, 25
221–5, 229, 230, 314 Grandval, monastery, 209
archive. See ecclesiastical archives Grapfeld, 221, 224, 225
cartulary. See cartularies Gregory I, pope, 44
Fundus Tuletianos, 1, 10, 40, 41, 51, Gregory of Catino, 331, 332
52 Gregory of Tours, 19, 174, 368, 372
Gretsch, Mechthild, 346
Garcı́a Leal, Alfonso, 273 Guimarães, monastery, 259, 266
Gascou, Jean, 24–5 cartulary. See cartularies
Gaul, 1, 34, 74, 85, 104, 122, 154, 174, Gundihild, widow, 81
175, 185, 255, 365, 368 Gundoin, duke, 209
Gauzoin, donor, 162 Gunthamund, king, 38, 40
Geary, Patrick, 286–7 Guntroda, landowner, 259, 260
Gebhardswil, estate, 162, 168
Geldersheim, villa publicus, 225 Habetdeum, bishop, 56–7
Geloira, countess, 270, 271 Hahicho, scribe, 217, 218
Geminius Felix, 39, 51, 52 Hartmot, abbot of St Gall, 119
dossier. See dossiers Harwic, donor, 113, 116
General Assembly of the Gallican Clergy, Hauxton, estate, 350, 351, 354
292 Helmstan, thegn, 336, 337, 345–6,
Gennadius of Marseilles, 127, 143 347
Georgius, silk maker, 77 Henry, landowner, 319
Gerbert, bishop of Rome, 321–2 Herovalliana collection, 143, 149
Germanus, abbot of Grandval, 211 Hildeard, landowner, 308
Geroin, clerical scribe, 216 Hildebrand, count, 177–8, 182
Geroin, lay scribe, 219 Hincmar, bishop of Reims, 155
gesta, 56–7, 175, 368, 372 Hippo Regius, 41, 60
gesta abbatum, 175, 295, 320 Historia de Sancto Cuthberto, 360
Gesta Dagoberti, 318 Hitto, bishop of Freising, 160–1
gesta episcopalia, 174 Honau, monastery, 193, 213, 216
gesta episcoporum, 175, 368 cartulary. See cartularies
gesta municipalia, 45, 55, 70, 95–124, 125, Honorius, emperor, 44, 56, 57, 58, 59
127, 133, 184, 206–7, 252, 264, Hugh, abbot of Cluny, 295
366, 375 Hugh, husband of Lilia, 301–4, 305, 306
fiscal imperatives, 72, 77–8, 96, 100, Hugo of Tours, count, 205–6
368, 375 Hugo, donor, 205, 206
formulas, 71, 97–112, 116–18, 122–3 Humber river, 359, 361
Frankish, 96–8, 102–17, 122–4, 135, Huneric, king, 48, 56, 57
139 Huniwine, thegn, 355
protocol, 97, 98–100, 102, 103, 113 Huosi family, 198, 201
public books, 103, 104–5, 106 Hymetius, proconsul, 46, 57
384 Index
Iberia, 6, 15, 16, 259–81, 365, 375 Lambert, Henri Louis, 291–3, 294, 295,
documentary culture. See documentary 299, 300, 301, 308, 309, 317
culture land market, 240–1, 244, 267, 302, 304,
documentary records. See documentary 351
records Landeric, bishop, 129, 144
Visigothic, 63–4, 65, 74, 82–93, 368, Langres, 209
369 Laon, 209
Idda, son of, testator, 208, 209 law
Ilduara, landowner, 279 Lombard, 69–70
Ingo, fidelis, 329 Ostrogothic, 69
inscriptions, 59–60 Roman, 63–4, 65, 66–70, 84–5, 135
Institut für Frühmittelalterforschung, Visigothic, 65, 69, 83–91
289 lay archives, 3–4, 189–90, 233, 339–40,
instrumenta, 41, 52, 84, 87, 131, 212, 243 349, 352–9, 375. See also dossiers
Iogna-Prat, Dominique, 289 Alahis’s (Italy), 94
Iscam, peasant, 2, 279–80 Apion family (Egypt), 21–2, 23, 24–5,
dossier. See dossiers 26–30, 32
Isidore of Seville, 84, 143 definition of, 3–4, 21–3
Islamic caliphate, 14, 31 Ermengildo and Paterna (Iberia),
Italy, 15, 16, 34, 172, 206, 231–58, 275, 270–1, 275
374–5 Flaı́nez (Iberia), 275
central Middle Ages, 321–35 Opila’s family (Iberia), 276
charters. See charters Perrecy (Francia), 182
documentary culture. See documentary Pompeyano (Iberia), 272
culture Psimanobet family (Egypt), 22, 31
early Middle Ages, 366, 370 Rankweil (Rhaetia), 170–1
Late Antiquity, 63–5, 70–82 Sisnando (Iberia), 272–3
Lombard, 6, 65, 69, 80, 93–4, 238, 371, Vermúdez (Iberia), 274–6
374 Wulfric’s (England), 352–4
Ostrogothic, 65, 69, 74, 75, 85 lay records. See documentary records
Le Mans, 118, 119
Jeme, village, 22, 23, 31, 34 lead seals, 48
Jerba island, 39 Leo I, emperor, 69, 84
ostraca. See ostraca Leo I, pope, 127
Jerusalem, 17 Leo, abbot of Subiaco, 330
Joseph, bishop of Freising, 195, 211, 226, Leodegar, scribe, 115, 116
228 Leofric, bishop of Exeter, 355
Julian, emperor, 43, 53 León, 2, 260, 266, 279
Julius, pope, 137 church of, 261
Justinian I, emperor, 23, 48, 50, 57, 58, Lex Romana Curiensis, 172
59, 69, 75, 78, 81, 82 Lex Romana Visigothorum, 69, 85,
Code, 66, 69, 85 135
libelli, 21, 321, 322
Kadwal, cancellarius, 218–19 Libellus Æthelwoldi, 344, 350, 351, 357,
Karol, son of Liutprand, 243–4, 245, 250 358, 359, 361
dossier. See dossiers Liber Aureus, 112–14
Kelly, Susan, 341, 343, 355, 357 Liber Eliensis, 344, 350, 359
Kent, 340, 349 Liber Papiensis, 322
Keynes, Simon, 345, 346, 347, 357 Liber Pontificalis, 127, 143
Krah, Adelheid, 161 Liber possessionum Wizenburgensis, 220
Ksar Koutine, Tunisia, 38, 48, 55 Liber traditionum, 243
Lichfield, bishop of, 350, 351
La Rocca, Cristina, 256 Life of Germanus of Grandval, 209
laity, 4–5 Life of Sadalberga, 209
Lake Constance, 186 Life of St Fructuosus, 92
Index 385
Lilia, landowner, 301–6, 307, 313 Meuse, 185

dossier. See dossiers Milan, 233, 238
Limoges, 19, 319 Milz, monastery, 202–3, 204
literacy, 9–10, 233–4, 236, 285, 322–3, missi, 118, 121, 175, 178, 180, 227, 247,
338–9, 346 290
Liutfrid, duke, 195–6 Mondsee, monastery, 191, 192, 193
Liutprand, king, 94 cartulary. See cartularies
Livro de Mumadona, 259 Mont notice, 178–81
Loire, 102, 184 Monte Amiata, monastery, 235, 237,
Lombard kingdom. See Italy, Lombard 241–2, 245, 258
Lorsch, monastery, 194, 199 archive. See ecclesiastical archives
loss of records. See documentary records charters. See charters
Lotharingia, 213 Moreau, Jacob-Nicolas, 292–3, 294
Louis II, emperor, 243 Moselle, 185, 186
Louis the German, king, 119 Mundini, villa, 271
Louis the Pious, 158, 178, 179, 205, munimina, 243, 329
298 Murbach, monastery, 193, 207, 208, 319
Lowe, Kathryn, 338 cartulary. See cartularies
Lucca, 235, 241, 245, 324
archive. See ecclesiastical archives Narbonne, 319
charters. See charters Négrine, Algeria, 50
church of, 233, 237 Neustria, 103, 127, 129, 149, 370
Lydus, John, 14, 155 Newton, estate, 350, 351, 354
Nibelung, count, 177–8, 179, 181
Mabillon, Jean, 284, 317 Nithard, historian, 8
Mâcon, 289, 298, 308 Noanca, villa, 278
Mâconnais, 286, 291, 303, 305 Nomina, servant, 265–6, 268
Mainz, 203–4, 221, 222, 223, 224, 225 Nonantola, monastery, 371
diocese, 204 Nononus, count, 115, 116
notariate, 204, 223, 225 Nordheim, village, 202
Maiolus, abbot of Cluny, 295 church, 202, 204
mallus, 119, 120, 121 Nordolf, donor, 198, 201
Marbeuf, Yves, bishop of Autun, 292 Norman Conquest, 338, 342, 361, 362
Marcellus, centurion, 43, 53 North Africa, 6, 15, 36–62, 74
Marciani, villa, 271 archives. See archives
Marculf, monk, 129, 132, 133, 136, 144, documentary culture. See documentary
150, 369, 370 culture, late Roman
formulas. See formulas Northumbria, 359–60
Maria, femina spectabilis, 79 notaries. See scribes
Marı́a Núñez, founder of Otero, 273, 275 notitia, 249–54
Marseilles, 175 finis intentionis, 249–50
Massiese of Viginti Boguli, 38, 51, 52 investitura salva querela, 250
Mattingly, David J., 55 ostensio cartae, 250, 251, 253
Maurists, 293 Notker of St Gall, 159
McKitterick, Rosamond, 10, 120–1, Nouaillé, monastery, 299, 319
164 Novgorod, 91
Meaux, 129 Numidia, 43, 46
Melgar de Foracasas, estate, 2, 279 Nuño, lord, 264, 268
Melic, priest, 278 nuptial endowments, 90–1, 264
dossier. See dossiers
memoria, 210 Odalswind, landowner, 163, 168, 169
Menant, François, 280 Odilo, abbot of Cluny, 290, 295, 307, 308,
Mercia, 359 312
Merol, bishop of Le Mans, 119 Odo, abbot of Cluny, 290, 295
Meseta, region, 91, 92 Odovacar, 71, 73, 94, 100
386 Index
Opila, abbot of San Toribio de Liébana, dossier. See dossiers

276 Persian Empire, 17–18
archive. See lay archives Peter, arcarius, 75, 76
orality, 9–10, 173, 180, 181, 285 Peter de Niviano, 172, 245
Ordlaf, ealdorman, 336, 337, 345–7, 348, dossier. See dossiers
356 Petrucci, Armando, 233, 234, 236
original charters. See charters Piacenza, 237, 245, 258
Osona-Manresa, region, 262–3 archive. See ecclesiastical archives
charters. See charters charters. See charters
ostraca, 22, 31, 37–9, 42, 49, 50, 51–2, 54, church of, 245
58, 61, 91, 368 Piasca, monastery, 276–8
Admiralty Island, 49, 51 archive. See ecclesiastical archives
Bir Trouch, 37–8, 51, 52, 55, 58, 59 dossier. See dossiers
Jerba, 39, 51, 52 Piasca, villa, 276
Négrine, 50, 61 Pippin I, king, 158, 371
Ostrogothic Italy, 85 Pippin of Aquitaine, king, 178, 179
Oswulf, landowner, 347, 348 Pirenne, Henri, 365
Otero, monastery, 273 Pisa, 239
archive. See ecclesiastical archives archive. See ecclesiastical archives
charters. See charters church of, 1, 94, 239
Otilo, duke, 226 pittakia, 29–30, 48, 55, 75
Otmar, abbot, 158, 159, 162 placita, 71, 89, 118, 121, 129, 140–1, 147,
Otto I, emperor, 329 159, 166, 173, 180, 251, 253, 265,
Otto II, emperor, 321 266, 311, 348
Otto III, emperor, 249, 322 Poitiers, 117, 299, 319
Otto, count, 201 polyptychs, 58, 77, 100, 175–6, 371
Oxyrhynchus, Egypt, 21, 22, 25, 26, 27, 28 Pompeyano, landowner, 272–3
archive. See lay archives
Pacificus, archdeacon, 256, 257 Porphoreus, donor, 324
Paderborn, 203, 205 Portugal, 259, 260
Padua, 335 praecepta, 1, 128, 160, 328
pancartes, 248, 313 precaria, 76, 163, 179, 198, 200, 201, 203,
papyri, 19–21, 343 218, 220, 240, 263, 290
Egyptian, 21–31, 183, 368 preservation of documents. See document
Italian or Ravennate, 64, 65, 70–1, preservation
72–3, 74, 75–82, 98–102, 106, 122, Priectus, landowner, 163, 168–9
133, 184, 318, 367, 368, 370 private records. See documentary records
rolls, 20–1 production of documents. See document
St-Denis, 318, 367 production
Paris, 108, 128, 129, 137, 140, 185, 294, prosecutor, 103, 104, 105, 106, 107
299, 317, 318 Provence, 175
Passau, church of, 191, 192, 193 Prüm, 112, 113, 114, 117
cartulary. See cartularies Psimanobet family, 22
Paterna, landowner, 270–2, 280 archive. See lay archives
archive. See lay archives public archives, 71, 183
Pauli Sententiae, 79, 85, 90 arca publica, 87
Pavia, 119, 251 archivum publicum, 23
Pedro Flaı́nez, count, 273–5, 278 Merovingian, 19
archive. See lay archives municipal. See gesta municipalia
Peira of Eustathius, 32 praetorian prefect, 19, 71, 75–6
Peña, J. Theodore, 49 proconsular, 55–6
Perrecy, estate, 155, 176–82, 188, 307, Roman state, 18–19, 23, 55–7
319, 372, 374
archive. See lay archives Rabegaud, abbot of St-Calais, 119
cartulary. See cartularies Ragnesind, vir inluster, 118–19
Index 387
Raimbaldus, count, 329 Rome, 59, 93, 100, 322, 327, 330
Ramiro II, king, 277 Romeo, Carlo, 236
Ramsey, estate, 350 Romulus Augustulus, 100
Ramsey, monastery, 351, 357 Rosenwein, Barbara, 303
Rankweil, villa, 165, 166, 168, 169, 170, Rothari’s Edict, 93–4
171, 174, 176, 182, 187, 188, 307 Rottachgau fragment, 187
archive. See lay archives Rufinus, historian, 138
charters. See charters Ruthard, count, 157–8, 159
church at, 168, 201
Rantwig, landowner, 195–6 S. Bartolomeo in Carpineto, monastery,
Raoul, king, 307 325
Ratbald-Wicbald family, 199, 216, 218 S. Clemente in Casauria. See Casauria
Ratburn, viscount, 307 S. Esuperazio in Cingoli, church, 334
Ratold, bishop of Verona, 256, 257 S. Maria in Organo, monastery, 245, 257
Ravenna, 64–5, 70–82, 93, 94, 96, 99, S. Martino in Lunata, church, 241
100, 237, 318, 322 S. Nicola in Gallucanta, monastery, 328
archive. See ecclesiastical archives S. Pietro ai Sette Pini, church, 239
church of, 71, 72, 77, 79, 80, 81, 82, 96, S. Pietro di Castiglione, monastery, 324
100, 101, 184, 367 S. Pietro in Castro, church, 257
gesta municipalia. See gesta municipalia S. Salvatore on Monte Amiata, monastery.
papyri. See papyri See Monte Amiata
Recceswinth, king, 84, 87, 89, 93 S. Zeno, monastery, 257
Recoire, landowner, 276 Saale, 202, 221
Redemptus, arcarius, 21–3 Saalegau, 224
Redon, monastery, 187, 320, 372, 374 Saargau, 193, 217, 220, 221
Regensburg, church of, 191, 192, 193 Sadalberga, abbess of Laon, 209, 211
cartulary. See cartularies Sahagún, monastery, 2, 261, 273, 276–9
Regino of Prüm, 372 archive. See ecclesiastical archives
Regularis Concordia, 358 Becerro gótico. See cartularies
Reims, 322 charters. See charters
Remiremont, monastery, 208 Salamir, villa, 271
gesta municipalia. See gesta municipalia Salerno, 328
Reynolds, Susan, 346 sales, 69, 72, 74, 75, 79, 80, 89–90,
Rhaetia, 165, 166, 170, 172, 185, 190, 229 195–6, 240–5, 302–4
Rhine, 96, 102, 160, 173, 174, 184, 188, in ecclesiastical archives, 241–5
192, 193, 199, 201, 202, 212, 221, uendicionis titulum, 195, 196
224, 225, 314, 359 Salvian of Marseilles, 74
Rhineland, 186, 191 Salzburg, 119, 120
Richsvind, donor, 199, 201 cartulary. See cartularies
Rieti, 73, 81, 100 church of, 191, 193, 371
church of, 234 San Martı́n de Valdepopulo, monastery,
Rihhart, donor, 226–7 279
Ripon, church, 359 San Salvador de Palazuelo, monastery, 279
Rochester, church of, 349 San Toribio de Liébana, monastery, 276
Roderic, count, 270, 271 archive. See ecclesiastical archives
Rodoin family, 211, 216, 220 Sánchez-Albornoz, Claudio, 267, 268
Rodrigo Guimirez, 265 Sant Benet de Bages, monastery, 269
Roman Empire, 18, 34, 35, 36, 63, 64, 66, Santa Marı́a de León, monastery, 279
71, 152, 363 Santa Maria de Otero de las Dueñas. See
documentary culture. See documentary Otero, monastery
culture Santa Marı́a de Piasca, monastery. See
documentary records. See papyri, Piasca, monastery
ostraca, tablets and slates Santiago de Valdávida, monastery, 277
Romanus, abbot of Casauria, 245 dossier. See dossiers
Romanus, comes Africae, 43, 46, 57 Sassovivo, monastery, 324
388 Index
Saulnois, 193, 200, 215, 220 Plan, 157

Sawyer, Peter, 341 scriptorium, 157
Saxony, 221 St Martin in Berg, church, 200, 211, 371
Schlehdorf, monastery, 198, 228 St Martin in Edesheim, church, 198–9
scribes, 44, 213–14, 298, 323 St Martin in Tours, monastery, 133, 135,
amanuensis, 103, 115, 200, 213, 215, 136, 138, 150, 159
216, 218, 220, 221, 222 St Martin-des-Champs, priory, 292
cancellarius, 200, 213, 216, 218, 219, St Mary in Dauendorf, church, 199
221, 247 St Pancratius in Steinhart, church, 227
clerical, 167–8, 225–8, 235–6 St Peter in Uhlweiler, church, 199
diocesan, 221–3 St Praiectus, monastery. See Flavigny
ecclesiastical, 44 St Saviour at Roncalem, church, 164, 168,
lay, 50–1, 234–6, 246–8, 304 171
local, 167–8, 170, 214–19, 220–1, St Zeno at Isen, church, 227
227–8 Staab, Franz, 204
monastic, 156–7, 204, 219–20, 223–5 Staffordshire, 352, 353, 359
notarius, 78, 99, 213, 215, 216, 217, Starcholf, 121
235, 247–8 Stengel, Edmund, 204
public, 44, 70, 82, 85–6, 99, 183, 214 Stenton, Frank, 357
Sens, 103, 107 Stock, Brian, 10
Severus, conductor, 38, 52 Stoke Canon in Devon, 355
Siena, 237 Strasbourg, 203, 204, 221
Siferth, thegn, 357 church of, 203
Sighifrid, landowner, 319 Styrr Ulfson, thegn, 360, 361
Silvester, abbot of SS. Cosma and Subiaco, monastery, 327
Damianus, 326 cartulary. See cartularies
Sisnando, bishop, 270, 271 Suger, abbot of St-Denis, 318
Sisnando, landowner, 272 Switzerland, 155
archive. See lay archives Syracuse, 73, 100
slates, 33, 65, 82–3, 88–93, 368
Snodland, estate, 349, 350, 352 tablets, 20, 25
Sobrado de los Monjes, monastery, 261, wood, 1, 16, 42, 68, 91. See also
269 Albertini Tablets
archive. See ecclesiastical archives Tacitus, 364
cartulary. See cartularies Tassilo, duke, 120, 226
Spain. See Iberia Tebaldus, bishop of Verona, 257
Spoleto, 237 Teudo, bishop of Rieti, 234
Sprandel, Rolf, 121 Teuto of Bibaro, 331
SS. Benedict and Scholastica in Subiaco, Theoderic, 69, 74, 78, 79
monastery. See Subiaco Edict, 69, 78–9, 85, 88
SS. Cosma and Damianus, monastery, 326 Theodora, nobilissima femina, 326–7
St-Calais, monastery, 119 Theodore of Canterbury, bishop, 128,
St Cuthbert, monastery, 360 143
St-Denis, monastery, 159, 184, 317–18, Theodosian Code, 18, 66, 85, 135
367, 374 Theophylact, abbot of S. Nicola in
archive. See ecclesiastical archives Gallucanta, 328
cartulary. See cartularies Theotchar, duke, 211, 215
charters. See charters Theuderic II, king, 118
papyri. See papyri Theutgar, scribe, 215, 217
St Gall, monastery, 119, 120, 122, Thorney, monastery, 351
154–60, 161, 162–6, 168–73, 176, Thugilo, widow, 80
186, 188, 190, 212, 229, 271, 285, Timgad, North Africa, 53
309, 371, 372 Tivoli, 330
archive. See ecclesiastical archives Tjäder, Jan-Olof, 70, 79
charters. See charters Toledo, 86, 93
Index 389
tomos synkollesimos. See papyri rolls Walefred, scribe, 143

Totone of Campione, 172, 240 Wardour, royal site, 345, 347
dossier. See dossiers Warin, count, 157–8, 159, 298
Touraine, 127, 149 Warwickshire, 352, 353
Tours, 105, 107, 133, 176 Westbury, estate, 349, 350
Tunhuang, Buddhist monastery, 364 Whittow, Mark, 34
Tunisia, 1, 38, 39, 40, 48, 52, 55, 59, 60, Wichard, donor, 308, 311, 316
61 dossier. See dossiers
Tuscany, 94, 237, 242 Wickham, Chris, 35, 251
Widerad, abbot of Flavigny, 136, 139
Ulpian, jurist, 68 will. See wills
Uracca, queen, 275 Wilfrid, saint, 360
Urban II, pope, 295, 296, 314 William, count, 296, 298, 299, 301,
302–4
Valdávida. See Santiago de Valdávida wills, 68–9, 77, 87–8, 101, 139, 208–9
Valdoré, villa, 274 Eberhard, 207–8
Valentinian I, emperor, 43, 46, 47, 49 Ermentrude, 208–9
Valentinian II, emperor, 137 Idda, son of, 209
Valentinian III, emperor, 58, 74 Widerad, 139
Novel of, 78 Wiltshire, 336, 345
Valerius, scribe, 167, 169 Winchester, church of, 356, 357, 358
validation of documents. See document use bishop of, 337, 346, 348, 356
and practice council, 358
Valle de Laurenzo, villa, 278 Wissembourg, monastery, 191, 192–3,
Valpuesta, monastery, 266 195, 197, 199, 200–1, 203, 205,
Vandal kingdom, 47–8, 56–7 206, 210, 211, 212, 213, 215, 217,
Vermudo II, king, 274 218–21, 225, 229, 230, 371, 372
Vermudo Nuñez, count, 277 archive. See ecclesiastical archives
Verona, 234, 238, 256, 334 cartulary. See cartularies
archive. See ecclesiastical archives Witigis, king, 73, 80
charters. See charters Wolfoald-Gundoin family, 216, 220
church of, 256–7 Wolfram, scribe, 223, 224
Vezin, Jean, 289 Worcester, church, 349, 361, 362
Vicenza, 334 Wormald, Patrick, 18, 339, 343, 348,
Vico Teatino, 244 349
dossier. See dossiers Worms, 119, 204, 221
Victor of Vita, 47, 57 Wormsgau, 203
Vigilius, pope, 81 Wulfad, bishop of Bourges, 178
Viginti Boguli, North Africa, 38, 55 Wulfgeat, thegn, 353–4
Vikings, 356, 359, 374 Wulfric Spot, thegn, 352, 353, 354
Villa de Belone, 279 archive. See lay archives
Villa de Petro, 279 family of, 353–4, 359
Vosges, 199, 207, 220 Wulfstan, archbishop of York, 360
wadium, 194, 195 York, 360

Walcari, landowner, 242
Waldebert, abbot of Luxeuil, 209 Zeno, emperor, 74
Waldhambach, estate, 199, 200 Zeumer, Karl, 83, 128
Waldo, abbot of St Gall, 157, 158, 161 Zurich, 120, 165

4529

Uploaded by

Copyright:

Available Formats

4529

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

4529

Uploaded by

Copyright:

Available Formats

more information - www.cambridge.

w a r r e n c . b r ow n is Professor of History at the California Institute

This publication is in copyright. Subject to statutory exception

First published 2013

Library of Congress Cataloguing in Publication data

ISBN 978-1-107-02529-5 Hardback

Cambridge University Press has no responsibility for the persistence or

List of figures and tables page vii

b r ow n , wa r r e n c . , Professor of History, Division of the Humanities

Press. He pressed us to clarify our thinking and ultimately helped us to

Form. Bitur. Formulae Bituricenses

TP Die Traditionen des Hochstifts Passau, ed.

Les cartulaires O. Guyotjeannin, L. Morelle and

CLA Codices Latini antiquiores, ed. E. A. Lowe, 11

In 1928, archaeologists working along the Tunisian-Algerian border

later – a Bavarian formula collection includes the same text word-for-

For much of the modern period, a deeply embedded commonplace held

This transformation has not, however, taken hold in the non-specialist

by laypeople.7 In other words, we address the degree to which laypeo-

7 For the issue of definition, see below, 12–15.

Laypeople, documents and the Church

qualitatively different from those of other social institutions and groups,

documents as important. As a rule, we have been unable to identify

Literacy and the use of documents

or monastic archives, into books (codices) called cartularies; this technol-

26 On cartularies, see Geary, Phantoms, 81–114; G. Declercq, ‘Originals and Cartularies:

networks and interactions revealed in the charter evidence – for example,

28 Settlement of Disputes; W. Brown and P. Górecki (eds.), Conflict in Medieval Europe:

31 J. Derrida, Archive Fever: A Freudian Impression, trans. E. Prenowitz (Chicago, 1996),

Mediterranean; hence, in the central bureaus of Byzantium – and the

We need to understand archival practice in terms of its own logic

We aim to revise our understanding of how documentary practices devel-

of these practices in the eastern Mediterranean and the emergence of

A survey of documentary practices in the eastern Mediterranean raises many

The ancient archive from Babylon to Byzantium

The account goes on to relate how:

1 M. Brosius, ‘Ancient Archives and Concepts of Record-Keeping: An Introduction’, in

to laws issued after 398, the imperial archive in Constantinople could be

The typology of the ancient archive

It is this imposition of order on documentary materials so as to render

These techniques of ‘dossierization’ discernible in what we might choose

13 Sijpesteijn, ‘The Archival Mind’, 164. 14Brosius, ‘Ancient Archives’, 10–12.

the fourth century, a phenomenon that has baffled those papyrologists

Lay archives in the early Byzantine period:

17 Clarysse, ‘Tomoi Synkollesimoi’, 356–8.

in character. By the sixth century, members of the Apion household were

24 Sarris, Economy and Society, 149–76.

This phenomenon requires a certain degree of explanation. In ear-

28 Brosius, ‘Ancient Archives’. 29 Sarris, Economy and Society, 149–99.

Documentation, ‘archivalization’, and social and

32 J. Gascou, ‘Les grands domaines, la cité, et l’état en Égypte byzantine’, Travaux et

35 Sijpesteijn, ‘The Archival Mind’, 163.

extent and character of the economic resources that underpinned the

Documentation, credit arrangements and control

revenues collected against the schedule of demands issued and, if neces-

above, however, in terms of the information it provides as to the proce-

53 P.Oxy I 137 line 19. 54 P.Oxy XVI 1988 lines 23–5.

government began to lay for itself firmer administrative foundations.57

The seventh century and beyond

62 M. Whittow, The Making of Orthodox Byzantium (London, 1996).