Christian Spain and Portugal in the Early Middle Ages: Texts and Societies [1 ed.] 0367345757, 9780367345754

A collection of papers in English by one of the foremost historians of the social and economic structure of medieval rur

615 117 5MB

English Pages 328 [329] Year 2019

Report DMCA / Copyright


Polecaj historie

Christian Spain and Portugal in the Early Middle Ages: Texts and Societies [1 ed.]
 0367345757, 9780367345754

Table of contents :
Half Title
Series Page
List of figures
List of tables
List of abbreviations
1 The early middle ages and Spanish identity
Part I: Charter writing
2 Local priests and the writing of charters in northern Iberia in the tenth century
3 Local priests in northern Iberia
4 Creating records of judicial disputes in northern Iberia before the year 1000
5 Exchange charters in the kingdom of Asturias-León, 700–1000
Part II: Production and exchange
6 Sale, price and valuation in Galicia and Castile-León in the tenth century
7 When gift is sale: reciprocities and commodities in tenth-century Christian Iberia
8 Countergift in tenth-century northern Iberia
9 Notions of wealth in the charters of ninth- and tenth-century Christian Iberia
10 Water mills and cattle standards: probing the economic comparison between Ireland and Spain in the early middle ages
11 Free peasants and large landowners in the west
12 The management of land-use in Old Castile: the early strands of the Becerro Galicano of San Millán de la Cogolla
13 Gardens and gardening in early medieval Spain and Portugal
Part III: Judicial practice
14 Settling disputes in early medieval Spain and Portugal: a contrast with Wales and Brittany?
15 Judges and judging. Truth and justice in northern Iberia on the eve of the millennium
16 Boni homines in northern Iberia. A particularity that raises some general questions
17 On suretyship in tenth-century northern Iberia

Citation preview

Christian Spain and Portugal in the Early Middle Ages

A collection of papers in English by one of the foremost historians of the social and economic structure of medieval rural communities, who here examines local societies in rural northern Spain and Portugal in the early middle ages. Principal themes are scribal practice and the analysis of charter texts; gift, sale and wealth; justice and judicial procedures. Always with a concern for personal relationships and interactions, for mobility, for decision-making and for practice, a sense of land and landscape runs throughout. The Spanish and Portuguese experience has seemed irrelevant to the great debates of early medieval European history that occupy historians. But Spain and Portugal shared the late Roman heritage which influenced much of western Europe in the early middle ages, and by the tenth century records and practice in Christian Iberia still shared features with the Carolingian world. This book offers a substantial corpus of Iberian evidence to set beside Frankish, Italian, English and Scandinavian material and thereby makes it possible for northern Iberia to play a part in these great debates of medieval European history. Wendy Davies taught medieval European history and was also successively Head of the History Department, Deans of Arts and of Social and Historical Sciences and Pro-Provost (Europe) at UCL until she retired in 2007. She remains active in research: while her personal research initially centred on Wales and then Brittany, in the last twenty years northern Iberia in the early middle ages has been the focus. She is an associate member of the History faculty, University of Oxford.

Also in the Variorum collected studies series:

PEREGRINE HORDEN and NICHOLAS PURCELL The Boundless Sea: Writing Mediterranean History (CS1083)

MOHAMED EL MANSOUR The Power of Islam in Morocco: Historical and Anthropological Perspectives (CS1082)

JOHN MARSHALL, edited by PHILIP BUTTERWORTH Early English Performance: Medieval Plays and Robin Hood Games Shifting Paradigms in Early English Drama Studies (CS1081)

JENNIFER O’REILLY, edited by CAROL A. FARR and ELIZABETH MULLINS Early Medieval Text and Image II: The Codex Amiatinus, the Book of Kells and Anglo-Saxon Art (CS1080)

JENNIFER O’REILLY, edited by CAROL A. FARR and ELIZABETH MULLINS Early Medieval Text and Image I: The Insular Gospel Books (CS1079)

JENNIFER O’REILLY, edited by MÁIRÍN MACCARRON and DIARMUID SCULLY History, Hagiography and Biblical Exegesis: Essays on Bede, Adomnán and Thomas Becket (CS1078)

MICHAEL BRETT The Fatimids and Egypt (CS1077)

HIROSHI TAKAYAMA Sicily and the Mediterranean in the Middle Ages (CS1076)

STEPHEN KATZ Holocaust Studies Critical Reflections (CS1075)

JOHN W. WATT The Aristotelian Tradition in Syriac (CS1074)

PEREGRINE HORDEN Cultures of Healing: Medieval and After (CS1073)

DAVID LUSCOMBE Peter Abelard and Heloise: Collected Studies (CS1072) For more information about this series, please visit: series/VARIORUMCS


Christian Spain and Portugal in the Early Middle Ages

Wendy Davies

Christian Spain and Portugal in the Early Middle Ages

Texts and Societies

First published 2020 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN and by Routledge 52 Vanderbilt Avenue, New York, NY 10017 Routledge is an imprint of the Taylor & Francis Group, an informa business This edition © 2020 by Wendy Davies The right of Wendy Davies to be identified as author of this work has been asserted by her in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data A catalog record for this title has been requested ISBN: 978-0-367-34575-4 (hbk) ISBN: 978-0-429-32665-3 (ebk) Typeset in Times New Roman by Apex CoVantage, LLC VARIORUM COLLECTED STUDIES SERIES CS1084


List of figures List of tables Conventions Preface Acknowledgements List of abbreviations

ix x xi xii xiv xvii

1 The early middle ages and Spanish identity


Part I: Charter writing 2 Local priests and the writing of charters in northern Iberia in the tenth century


3 Local priests in northern Iberia


4 Creating records of judicial disputes in northern Iberia before the year 1000


5 Exchange charters in the kingdom of Asturias-León, 700–1000


Part II: Production and exchange 6 Sale, price and valuation in Galicia and Castile-León in the tenth century © 2002 Blackwell Publishers Ltd


7 When gift is sale: reciprocities and commodities in tenth-century Christian Iberia © Cambridge University Press 2010


8 Countergift in tenth-century northern Iberia




9 Notions of wealth in the charters of ninth- and tenth-century Christian Iberia


10 Water mills and cattle standards: probing the economic comparison between Ireland and Spain in the early middle ages


11 Free peasants and large landowners in the west


12 The management of land-use in Old Castile: the early strands of the Becerro Galicano of San Millán de la Cogolla



13 Gardens and gardening in early medieval Spain and Portugal © 2019 John Wiley & Sons Ltd


Part III: Judicial practice 14 Settling disputes in early medieval Spain and Portugal: a contrast with Wales and Brittany?


15 Judges and judging. Truth and justice in northern Iberia on the eve of the millennium © 2010 Elsevier Ltd


16 Boni homines in northern Iberia. A particularity that raises some general questions © Oxford University Press 2018


17 On suretyship in tenth-century northern Iberia






0.1 Iberia, c. 950 2.1a, 2.1b Amorino: ACL 104 and ACL 110 2.2 Durabiles: ACL 818A 6.1 The incidence of all transactions, sales and donations, 910–99, in the six collections 7.1 Location of sources of the main charter collections used 12.1 San Millán and principle sites discussed 12.2 Hiniestra’s core estate (based on 1:25,000 maps, Instituto Geográfico Nacional) 17.1 Local courts in eastern Brittany


xx 28 30 98 121 212 213 292


5.1 Number of charters used from the kingdom of Asturias-León 701–1000 5.2 Percentage of tenth-century gift, sale and exchange charters in collections from Asturias-León 5.3 Lay and ecclesiastical parties to ninth- and tenth-century exchanges 5.4 Parties to ninth- and tenth-century exchanges, by status 6.1 Sales as a percentage of all transactions, by collection, by decade 6.2 Percentage of prices cited as miscellaneous objects, or silver ‘currency’, or both, in six collections, by decade 6.3 Percentage of ‘currency’-only expressions in all prices in the Celanova, Cardeña and Sahagún collections, by decade 6.4 Percentage of all citations of objects as price, 910–99, per monastery, by object 7.1 Gift and sale, by region of 10th-century northern Iberia, as percentage of all transactions per region 8.1 Percentage of charters with countergift, per collection, with presence of countergifts in 980–99 12.1 Hiniestra’s acquisitions in the villages immediately to the north (BG382) 12.2 Donations to Hiniestra in both the Montes de Oca and Briviesca (BG382) 12.3 Hiniestra’s acquisitions in and around Briviesca (BG382) 12.4 Turns in the Arlanzón mill acquired by Hiniestra (BG384, c. 951)


77 79 88 89 99 101 102 104 123 141 214 215 216 217


In quoting from Latin texts I follow the punctuation and capitalization of the editions quoted, unless indicated otherwise. I have kept name forms as spelled therein when quoting texts but have used modern forms in discussion and translations where there is a commonly used form. For those unfamiliar with Latin of this period and area, note that orthography is extremely unstable and that grammar does not follow classical norms; I have not attempted any corrections or ‘improvements’. My translations are very free; they are not intended to be literal (unless indicated); rather, they are intended to make sense in modern English. By ‘peasant’ I intend to indicate someone who laboured on the land, whether richer or poorer, tenant or small-scale proprietor. Of course, we cannot always be sure of the social status of individuals in these texts. However, I have taken to be peasants those who were associated with very small units of property, of low value, within a small area (usually that of a single community), who did not themselves travel far, and whose disputes – if they were involved in them – were of a petty kind. They contrast strongly with aristocrats, who had large, scattered estates (sometimes many) and who can be seen to have travelled considerable distances. It is conventional to refer to charters as originals, copies, pseudo-originals and falsos or fabrications. It has very reasonably been pointed out that, in dealing with early medieval texts, we can rarely know with complete certainty if a charter on a single sheet is an original or not; indeed, the idea that there would always have been a single original is in itself questionable. However, the modern editors of many excellent editions have used the conventional terms, and I have followed their usage, unless indicated otherwise. By ‘original’, therefore, I mean a charter on a single sheet of parchment which has no obvious anachronisms in its physical characteristics or in its wording, and which is consistent with a record produced at the time stated, in the name of the principal actor, by the named scribe, if a scribe be named. By ‘pseudo-original’ I mean a charter on a single sheet whose physical characteristics were intended to suggest production at the time stated, but which was created at a later time. By falso, ‘fabrication’ or forgery I mean a charter designed to deceive, which does not necessarily attempt reproduction of appropriate physical characteristics. xi


The papers in this collection derive from work of the last twenty years. They are organized in accordance with three fields of study which have occupied much of my professional life: how charters came to be formulated in the way that we now see them; the social and economic functioning of rural communities in the early middle ages; and justice and judicial process in that same period. Chapters 3 and 5 could as easily have joined the social and economic section and chapter 12 could have joined charter writing – the fields overlap. As a collection they are a mixture of exploratory papers, where I was working things out for the first time (like cc. 2, 6 and 14), papers of detailed primary research (like cc. 6 and 15) and those of more mature reflection (like cc. 4, 7 and 16). They should complement my two earlier Variorum collections (Welsh History in the Early Middle Ages: Texts and Societies and Brittany in the Early Middle Ages: Texts and Societies, both published by Ashgate in 2009) and my two Spanish monographs (Acts of Giving. Individual, Community, and Church in Tenth-Century Christian Spain, Oxford: Oxford University Press, 2007, and Windows on Justice in Northern Iberia, 800–1000, Abingdon: Routledge, 2016). This collection does not include all my Spanish papers – there are another fifteen at the moment of writing this. These do not feature here for different reasons: they are still in press; I failed to get permission to reprint; the fee for reprinting was too high (as in the case of Boydell & Brewer, although I did pay Brepols for reprinting its two papers); much of the content features in one or other Spanish monograph. There would have been no point reprinting the latter but I regret the fact that I have been unable to include ‘Holding court. Judicial presidency in Brittany, Wales and northern Iberia in the early middle ages’, in F. Edmonds and P. Russell (eds), Tome. Studies in Medieval Celtic History and Law in Honour of Thomas Charles-Edwards, Woodbridge: Boydell & Brewer, 2011; ‘The incidence of princeps in the ninth- and tenth-century charters of northern Spain’, in J.-M. Picard, H. Oudart, J. Quaghebeur (eds), Le prince, son peuple et le bien commun, Rennes: Presses Universitaires de Rennes, 2013; ‘Regions and micro-regions of scribal practice’, in J. Escalona Monge, O. Vésteinsson, S. Brookes (eds), Polity and Neighbourhood in Early Medieval Europe, Turnhout: Brepols, 2019; ‘Counts in ninth- and tenth-century Iberia’, in S. Barton and R. Portass (eds), Beyond the xii


Reconquista: Constructing and Representing Social Power in Medieval Iberia, c.700–1300, Leiden: Brill, forthcoming. I decided to include ch. 8, although some of the initial content also features in ch. 7, because the points made in the final pages do not occur elsewhere; readers should therefore focus on the latter half of this paper. I have included ch. 13, although it is easily available for those in English-speaking universities, because I hope there will be other readers, because it made me engage with al-Andalus and because it looks towards the future. I have made minor revisions to some of the texts and have added some recent bibliography to the footnotes. I have not changed my statements of numbers of charters available, although these vary widely across the papers, because they reflect the numbers I used in respect of each paper. I have not changed the following points, since they reflect my thinking at the time of writing, but readers should take note: for ch. 1 one would say more about regional difference and one would note more questioning of feudalismo nowadays; for ch. 14, concluding pages, I would now attribute less to ruler power. In ch. 6 my original paper attracted considerable criticsm because I translated res, in lists of objects used for payment, as ‘things’, whereas res clearly means ‘animals, cattle’ in Modern Spanish; while it is reasonable to express a doubt, I still prefer my translation, given that tenth-century scribes did not hesitate to use bos and vacca when they meant ‘cattle’, and in Galicia in particular a cattle-standard, using those words (valente duos boves), was sometimes used. In addition to the specific additional bibliography, readers should be particularly aware of the recent and forthcoming work of Graham Barrett, Álvaro Carvajal Castro, Ainoa Castro, Julio Escalona, André Marques, Iñaki Martín Viso, David Peterson, Robert Portass, Juan-Antonio Quirós and José Carlos Sánchez Pardo, and of the online database, under the direction of Isabel Alfonso, Procesos Judiciales en las sociedades medievales del norte peninsular at WOOLSTONE 28 August 2019



I owe considerable thanks, as acknowledged in individual papers, to many people, amongst whom Julio Escalona has been a consistent source of constructive criticism and advice. For this volume as a whole thanks are especially due to Michael Greenwood for encouragement and assistance through the press and to Graham Barrett for his perceptive comments. I am very grateful to the British Academy and to the Leverhulme Trust for grants which enabled me to do the archive work which underpins some of the papers and to the Archivo de la catedral de León and the cathedral chapter for permission to print my own photographs for 2.1a, 2.1b, 2.2. I am extremely grateful to the following publishers for permission to reproduce papers from the works cited. Böhlau Verlag for ch. 5, from W. Davies, ‘Exchange charters in the kingdom of Asturias-León, 700–1000’, in I. Fees and Ph. Depreux (eds), Tauschgeschäft und Tauschurkunde vom 8. bis zum 12. Jahrhundert/L’acte d’échange du VIII au XII siècle, Archiv für Diplomatik, Beiheft 13, 2013, pp. 471–89. Brepols Publishers for ch. 9, from W. Davies, ‘Notions of wealth in the charters of ninth- and tenth-century Christian Iberia’, in J.-P. Devroey, L. Feller, R. Le Jan (eds), Les élites et la richesse au haut moyen âge, Collection Haut Moyen Âge 10, Turnhout: Brepols, 2010, pp. 265–84, and ch. 17, from W. Davies, ‘On suretyship in tenth-century northern Iberia’, in J. Escalona and A. Reynolds (eds), Scale and Scale Change: Western Europe in the First Millennium, The Medieval Countryside 6, Turnhout: Brepols, 2011, pp. 133–52. Brill for ch. 4, from W. Davies, ‘Creating records of judicial disputes in northern Iberia before the year 1000’, in I. Alfonso, J.-M. Andrade, A. E. Marques (eds), Judicial Processes in Early Medieval Societies: Iberia and Its European Context, Leiden: Brill, 2020. Cambridge University Press for ch. 7, reproduced with permission of the Licensor through PLSClear, from W. Davies, ‘When gift is sale: reciprocities and commodities in tenth-century Christian Iberia’, in W. Davies and xiv


P. Fouracre (eds), The Languages of Gift, Cambridge: Cambridge University Press, 2010, pp. 217–37. The Department of Anglo-Saxon, Norse and Celtic, University of Cambridge, for ch. 10, from W. Davies, Water mills and cattle standards: probing the economic comparison between Ireland and Spain in the early middle ages, H. M. Chadwick Memorial Lectures 21, Cambridge: Department of Anglo-Saxon, Norse and Celtic, 2012. The Department of Modern Languages and Cultures, Queen Mary University of London, for ch. 8, from W. Davies, ‘Countergift in tenth-century northern Iberia’, in A. Deyermond and M. J. Ryan (eds), Early Medieval Spain: A Symposium, Papers of the Medieval Hispanic Research Seminar 63, London: Department of Hispanic Studies, Queen Mary, University of London, 2010, pp. 79–96. De Gruyter for ch. 3, from W. Davies, ‘Local priests in northern Iberia’; in S. Patzold and A. C. van Rhijn (eds), Men in the Middle: Local Priests in Early Medieval Europe, Berlin: De Gruyter, 2016, pp. 125–44. Editions de la Sorbonne for ch. 12, from W. Davies and D. Peterson, ‘The management of land-use in Old Castile: the early strands of the Becerro Galicano of San Millán de la Cogolla’, in A. Dierkens, N. Schroeder, A. Wilkin (eds), Penser la paysannerie médiévale, un défi impossible?, Paris: Editions de la Sorbonne, 2017, pp. 47–68. Oxford Publishing Ltd for two chapters, both reproduced with permission of the Licensor through PLSClear, ch. 1, from W. Davies, ‘The early middle ages and Spanish identity’, in H. Pryce and J. Watts (eds), Power and Identity in the Middle Ages: Essays in Memory of Rees Davies, Oxford: Oxford University Press, 2007, pp. 68–84, and ch. 16, from W. Davies, ‘Boni homines in northern Iberia. A particularity that raises some general questions’, in R. Balzaretti, J. Barrow, P. Skinner (eds), Italy and Early Medieval Europe: Papers for Chris Wickham, Oxford: Oxford University Press, 2018, pp. 60–72. Presses Méridiennes and Editorial, Consejo Superior de Investigaciones Científicas, for ch. 2, reproduced from W. Davies (2013): ‘Local priests and the writing of charters in northern Iberia in the tenth century’, in J. Escalona and H. Sirantoine (eds), Chartes et cartulaires comme instruments de pouvoir. Espagne et Occident chrétien (VIIIe–XIIe siècles), Toulouse: Presses Méridiennes de Toulouse/CSIC, pp. 29–43. Revue belge de philologie et d’histoire, General Editor, for ch. 11, from W. Davies, ‘Free peasants and large landowners in the West’, in J.-P. Devroey and A. Wilkin (eds), Autour de Yoshiki Morimoto. Les Structures agricoles en dehors du monde carolingien, formes et genèse, Revue belge de philologie et d’histoire, 90 (2012), pp. 361–80. Taylor & Francis for ch. 15, from W. Davies, ‘Judges and judging. Truth and justice in northern Iberia on the eve of the millennium’, Journal of Medieval History, 36:3 (2010), pp. 193–203. xv


University of Wales Press for ch. 14, from W. Davies, ‘Settling disputes in early medieval Spain and Portugal: a contrast with Wales and Brittany?’, in R. Griffiths and P. Schofield (eds), Wales and the Welsh in the Middle Ages, Cardiff: University of Wales Press, 2011, pp. 89–107. John Wiley & Sons, Inc., for ch. 6, from W. Davies, ‘Sale, price and valuation in Galicia and Castile-León in the tenth century’, Early Medieval Europe, 11 (2002), pp. 149–74, and ch. 13, from W. Davies, ‘Gardens and gardening in early medieval Spain and Portugal’, Early Medieval Europe, 27:3 (2019), pp. 327–48.



Charters are cited by number of the edition in which they are published, and in the case of archives by number within a carpeta (folder). A ACL ACO AHDE AHDL AHN Ar Ast BG



Cartulario de Albelda, ed. A. Ubieto Arteta, València: Bautista, 1960. Archivo de la catedral de León. Archivo de la catedral de Oviedo. Anuario de história del derecho español. Archivo histórico diocesano de León. Archivo histórico nacional. Cartulario de San Pedro de Arlanza, antiguo monasterio benedictino, ed. L. Serrano, Madrid: Centro de Estudios Históricos, 1925. Colección documental de la catedral de Astorga (646–1126), vol. 1, ed. G. Cavero Domínguez and E. Martín López, León: Centro de Estudios e Investigación ‘San Isidoro’, 1999. ‘Becerro Galicano de San Millán de la Cogolla’, online edition (2013): Based on F. García Andreva, El Becerro Galicano de San Millán de la Cogolla: Edición y estudio, San Millán de la Cogolla: Cilengua, Fundación San Millán de la Cogolla, 2010. See also SM. Colección documental del monasterio de San Pedro de Cardeña, ed. G. Martínez Díez, Burgos: Caja de Ahorros y Monte de Piedad del Circulo Catolico de Obreros de Burgos, 1998. (Now also El Becerro gótico de Cardeña: el primer gran cartulario hispánico (1086), ed. J. A. Fernández Flórez, S. Serna Serna, 2 vols, Madrid/Burgos: Real Academia Española, Instituto Castellano y Leonés de la Lengua, 2017.) V. Cañizares del Rey, Colección diplomática (569–1463), ed. M. Rodríguez Sánchez and Ó. González Murado, Lugo: Publicaciones Diócesis de Lugo, 2012.





EMAP Floriano

Li, Lii, Liii, Liv

LaC Leire LL

Lor OD


O Tombo de Celanova: estudio introductorio, edición e índices (ss. IX–XII), ed. J. M. Andrade Cernadas with M. Díaz Tie and F. Javier Pérez Rodríguez, 2 vols, Santiago de Compostela: Consello da Cultura Galega, 1995. Cartulaire de l’Abbaye de Redon en Bretagne, ed. A. de Courson (Paris: Imprimerie Impériale, 1863). Consejo Superior de Investigaciones Científicas (Spain). Colección documental del monasterio de San Pedro de Eslonza (912–1300), vol. 1, ed. J. M. Ruiz Asencio and I. Ruiz Albi, León: Centro de Estudios e Investigación ‘San Isidoro’, 2007. Early Medieval Archaeology Project (Ireland). A. C. Floriano Cumbreño, Diplomática española del periodo astur: estudio de las fuentes documentales del reino de Asturias (718–910), 2 vols, Oviedo: Instituto de Estudios Asturianos, 1949–51. Colección documental del archivo de la catedral de León (775–1230), vol. 1 (775–952), ed. E. Sáez; vol. 2 (953–985), ed. E. Sáez and C. Sáez; vol. 3 (986–1031), ed. J. M. Ruiz Asencio; vol. 4 (1032–1109), ed. J. M. Ruiz Asencio, León: Centro de Estudios e Investigación ‘San Isidoro’ (CSICCECEL), Caja de Ahorros y Monte de Piedad, Archivo Histórico Diocesano, 1987, 1990, 1987, 1989. La Coruña. Fondo Antiguo (788–1065), ed. C. Sáez and M. del Val González de la Peña, 2 vols, Alcalá: Universidad de Alcalá, 2003–4. Documentación medieval de Leire (siglos IX a XII), ed. Á. J. Martín Duque, Pamplona: Diputación foral de Navarra, Institución Príncipe de Viana, 1983. The Text of the Book of Llan Dâv, ed. J. G. Evans with J. Rhys (Oxford: [private publication for J. G. Evans], 1893); charters cited by page no. of this edition, differentiating those that begin on the same page by a, b, c. Liber testamentorum coenobii Laurbanensis, 2 vols, León: Centro de Estudios e Investigación ‘San Isidoro’, 2008. Colección documental del monasterio de Santa María de Otero de las Dueñas, ed. J. A. Fernández Flórez and M. Herrero de la Fuente, vol. 1, León: Centro de Estudios e Investigación ‘San Isidoro’, Caja España de Inversiones, Archivo Histórico Diocesano, 1999. Colección diplomática de San Salvador de Oña, ed. J. del Alamo, vol. 1, Madrid: CSIC, Escuela de Estudios Medievales, 1950.






Si, Sii

Sam San Pedro

SantA SJP SM Sob


Colección diplomática del monasterio de San Vicente de Oviedo (años 781–1200), ed. P. Floriano Llorente, Oviedo: Diputación de Asturias, Instituto de Estudios Asturianos del Patronato José María Quadrado (CSIC),1968. Colección de documentos de la catedral de Oviedo, ed. S. García Larragueta, Oviedo: Diputación de Asturias, Instituto de Estudios Asturianos del Patronato José María Quadrado, 1962. Portugaliae Monumenta Historica a saeculo octavo post Christum usque ad quintumdecimum, Diplomata et Chartae, vol. 1, ed. A. Herculano de Carvalho e Araujo and J. J. da Silva Mendes Leal, Lisbon: Academia Scientiarum Olisponensis, 1867–73. Colección diplomática del monasterio de Sahagún (857– 1230), I (siglos IX y X), ed. J. M. Mínguez Fernández; II (1000–1073), ed. M. Herrero de la Fuente, León: Centro de Estudios e Investigación ‘San Isidoro’ (CSIC-CECEL), Archivo Histórico Diocesano, Caja de Ahorros y Monte de Piedad de León, 1976, 1988. El Tumbo de San Julián de Samos (siglos VIII–XII), ed. M. Lucas Álvarez, Santiago de Compostela: Caixa Galicia, 1986. Tumbo Viejo de San Pedro de Montes, ed. A. Quintana Prieto, León: Centro de Estudios e Investigación ‘San Isidoro’, Archivo Histórico Diocesano, Caja de Ahorros y Monte de Piedad de León, 1971. La documentación del Tumbo A de la catedral de Santiago de Compostela: estudio y edición, ed. M. Lucas Álvarez, León: Centro de Estudios e Investigación ‘San Isidoro’, 1997. Cartulario de San Juan de la Peña, ed. A. Ubieto Arteta, 2 vols, València: Bautista, 1962–3. Cartulario de San Millán de la Cogolla (759–1076), ed. A. Ubieto Arteta, València: Anubar, 1976; see also BG. Tumbos del monasterio de Sobrado de los Monjes, ed. P. Loscertales de García de Valdeavellano, 2 vols, Madrid: Dirección General del Patrimonio Artístico y Cultural, Archivo Histórico Nacional, 1976. Cartulario de Santo Toribio de Liébana, ed. L. Sánchez Belda, Madrid: Archivo Histórico Nacional, 1948. Los becerros gótico y galicano de Valpuesta, ed. J. M. Ruiz Asencio, I. Ruiz Albi, M. Herrero Jiménez, 2 vols, Madrid/ Burgos: Real Academia Española, Instituto Castellano y Leonés de la Lengua, 2010.


Figure 0.1 Iberia, c. 950


The middle ages have for a long time been seen as a critical period in the formation of modern Spain.1 The creation, expansion, and consolidation of the Christian kingdoms of the north and of the county of Castile, movements characteristic of the ninth, tenth, and eleventh centuries, are held to lay the basis for the unification of Spain in the sixteenth century, the Spanish Empire and the transmission of Spanish culture overseas. The images of territorial expansion, Christian self-confidence and the settlement of a desert have seemed to be essential to understanding what makes Spain Spain. Indeed, the distinctiveness of Spain was a recurrent preoccupation of Spanish historians in the twentieth century: what made Spain different from the rest of Europe?2 Different it obviously seemed to be, given the strength of continuing cultural contacts with Spanish America and given its civil war, its long period of dictatorship, its particular brand of Catholicism, and its very strong regional identities. For most historians, what made Spain different was the distinctiveness of its medieval past, a past of overt cultural and political conflict between Christians and Moslems, and a past in which the almost total Moslem conquest of 711 was gradually, across many centuries, overturned by the ‘centuries-long frontier struggle between Christians and Moslems’: ‘Reconquest [is the] key to the history of Spain’.3 Indeed, the view is not confined to historians

1 First published in H. Pryce and J. Watts (eds), Power and Identity in the Middle Ages (Oxford: Oxford University Press, 2007), pp. 68–84. 2 A. Castro, The Structure of Spanish History, trans. E. L. King (Princeton: Princeton University Press, 1954), p. 3; C. Sánchez-Albornoz, España; un enigma histórico, 2 vols (Buenos Aires: Editorial Sudamericana, 1956); J. M. Jover, ‘Panorama of current Spanish historiography’, Journal of World History, 6 (1960–61), pp. 1023–38; T. F. Glick, Islamic and Christian Spain in the Early Middle Ages (Princeton: Princeton University Press, 1979), pp. 10–12; P. Linehan, History and the Historians of Medieval Spain (Oxford: Oxford University Press, 1993), pp. 4, 192. 3 C. Sánchez-Albornoz, ‘The frontier and Castilian liberties’, in A. R. Lewis and T. F. McGann (eds), The New World Looks at Its History (Austin: University of Texas Press, 1963), pp. 27–46, at p. 29; id., España; un enigma, vol. 2, p. 9. Cf. A. Ubieto Arteta, Atlas Histórico. Como se formó España (València: Anubar, 2nd edn, 1970).



and still colours the way many modern Spaniards think of themselves, even if it is to maintain that Spain is no longer different.4 In what follows I shall firstly consider the major issues that have absorbed medieval historians in the twentieth century; I shall then take a look at twentieth-century political developments, at the use of the medieval past by twentieth-century politicians, and at the interaction of both with history writing; and I shall finally consider the fate of historians’ preoccupations of the twentieth century and the issues that remain to be resolved. The interplay between contemporary politics and history writing is subtle – and unusually important.

Writing medieval history Medieval historians had different approaches to the past, and twentieth-century history writing is marked by sustained debates about it.5 The most prominent debate, and the most influential, is that between Américo Castro and Claudio Sánchez-Albornoz, largely conducted in mid-century. Castro argued that the peculiar identity of Spain derived from the integration of Christian, Moslem, and Jewish cultures, which had interacted with each other to produce a new culture during the course of the reconquest across the ninth to thirteenth centuries.6 This interaction was creative, he maintained, and determined the future quality of the Spanish, for the earlier Roman and Visigothic traditions had been wiped out by the very fact of the Arab incursion and settlement. Sánchez-Albornoz, by contrast, argued that the identity of Spain was marked by different characteristics from those of other western European states and derived from a Spanishness that had developed in antiquity. The palaeolithic had been ‘more unified ethnically and culturally than people supposed’ and the Roman period saw further manifestations of the temperament of homo hispanus; despite the apparently dramatic political changes, there were cultural continuities from that time and a quintessentially Hispanic culture survived the Arab onslaught.7 Both historians agreed, however, that the medieval development of Spain had been different from that of western Europe as a whole, for Spain did not share Europe’s feudal characteristics. For Sánchez-Albornoz pre-feudal development was beginning much as it did elsewhere in Europe but was cut off by the invasion of 711 and its consequences; for Castro the ideology of feudalism was completely inappropriate and Spain showed

4 J. Hooper, The New Spaniards (London: Penguin, 1995), pp. 374, 445–6; C. Closa and P. M. Heywood, Spain and the European Union (Basingstoke: Palgrave Macmillan, 2004), p. 31. 5 What follows focusses on the historiography of northern Spain; the Arab historiography raises equally important but different issues, which merit separate treatment. 6 A. Castro, España en su historia. Cristianos, moros y judíos (Buenos Aires: Losada, 1948); id., The Spaniards: An Introduction to Their History, trans. W. F. King and S. Margaretten (Berkeley: University of California Press, 1971). 7 Sánchez-Albornoz, España; un enigma, vol. 2, p. 349 (Spanish pride surviving in the little-Romanized Asturias and Basque country, ibid., vol. 1, p. 617); id., ‘The frontier and Castilian liberties’, p. 30.



no signs of the class conflict that he held accompanied feudalism elsewhere.8 Here then was another critical difference in Spanish development – it had no feudalism. One sub-theme of the debate, although it was much more strongly associated with Sánchez-Albornoz than with Castro, was that of depopulation and repopulation. In response to the initial Moslem incursion, King Alfonso I of Asturias (739–57) adopted a policy of strategic depopulation in Galicia and in the large expanse of the Duero basin, he argued. Moslem conquest provoked further Christian migration, particularly from the Duero valley (see figure 0.1), the Christian population of Hispania fleeing north to the inhospitable Cantabrian mountains and north-west to the farthest corner of Galicia, leaving vast expanses of unoccupied and depopulated land. As Christian kings slowly won back territory in the central middle ages, the Hispanic population gradually returned from the north to the deserted lands, reinforced by Mozarabic settlers from southern Spain, and together they restored settlement and agriculture.9 This was colonization of the frontiers in a truly American sense and Sánchez-Albornoz exploited the parallels. Like the American pioneers, the colonizers brought a spirit of freedom to central Spain and their settlements became free communities of small-scale proprietors.10 This free peasant spirit did much to determine the distinctive qualities of Spanish society in the central middle ages and even beyond: ‘an island of freemen in a Europe in the process of feudalization’; ‘a major factor in forging the Spaniard and his differentiation from the man beyond the Pyrenees’.11 Belief in depopulation and repopulation was highly influential, at scholarly and popular levels, until very recently; the approach has affected most aspects of the interpretation of social and political change in the early middle ages. Even in the late twentieth century distinguished scholars wrote about depopulation beginning

8 Sánchez-Albornoz, España; un enigma, vol. 1, p. 680 (‘Ni feudalismo ni burguesía’, ibid., vol. 1, pp. 678, 679); Castro, Structure of Spanish History, pp. 607–15; J. Beverley, ‘Class or caste: A critique of the Castro thesis’, in R. E. Surtz, J. Ferrán, D. P. Testa (eds), Américo Castro: The Impact of His Thought (Madison: Hispanic Seminary of Medieval Studies, 1988), pp. 141–9, at p. 143. Cf. P. Linehan, ‘History in a changing world: The case of medieval Spain’, in P. Linehan, Past and Present in Medieval Spain (Aldershot: Ashgate, 1992), ch. 1, pp. 1–22. 9 Mozarabic: Arabized Hispanics, often Christian. 10 C. Sánchez-Albornoz, Despoblación y repoblación del valle del Duero (Buenos Aires: Instituto de Historia de España, 1966); id., ‘Repoblación del Reino Asturleonés’, in his Viejos y nuevos estudios sobre las instituciones medievales españolas, 3 vols (Madrid: Espasa Calpe, 1976–80), vol. 2, pp. 581–790 (first published 1971, but see his comments already in id., ‘Las behetrías: la encomendación en Asturias, León y Castilla’, Anuario de Historia del Derecho Español, 1 (1924), pp. 158–336, at pp. 198–203); id., ‘Pequeños propietarios libres en el reino Asturleonés. Su realidad histórica’, Settimane di studio del centro italiano di studi sull’alto medioevo, 13 (1966), pp. 183–222; Ubieto Arteta, Atlas Histórico, pp. 38–49, 59, for visual representations of maximum depopulation in the eighth century and gradual repopulation thereafter. J. Vicens Vives, Approaches to the History of Spain, trans. J. C. Ullman (Berkeley: University of California Press, 2nd edn, 1970), pp. 36–9 (translation of update of his Aproximación a la historia de España [Barcelona: Vicens Vives, 1952]). 11 Sánchez-Albornoz, ‘Repoblación’, p. 790.



already in the later seventh century (that is before the Moslem incursion), about desertion caused by the Arabs, about families who colonized ‘deserted lands’ in the ninth and tenth centuries and about repopulation.12 School textbooks of the twenty-first century detail the abandonment of lands occasioned by the Moslem conquest and the subsequent repopulation of extensive territory in the ninth and tenth centuries, especially in the Duero basin.13 History writing of the past one or two generations has been immensely influenced by the reaction to Sánchez-Albornoz, often self-consciously so.14 To begin with, feudalism was admitted and instated.15 In the mid-1960s Barbero and Vigil began to publish their work on the formation of Spanish feudalism, arguing that although in part it evolved directly from late and post-Roman society, that is from the fifth and sixth centuries, those parts of northern Spain which were little touched by Romanization sustained a tribal society whose transformation from collective property-owning kin groups to villages of individual owners constituted a different and later route to feudal society.16 There followed more than a generation of lively discussion of the precise character of Spanish feudalism and the extent to which it diverged from a north Frankish model. Paying relatively little attention to Barbero and Vigil’s ‘Visigothic feudalism’, most historians identified the central middle ages as the key period of feudalization, charting the development at different rates within a tenth- to twelfth-century bracket. Bonnassie’s

12 J. A. García de Cortázar, La sociedad rural en la España medieval (Madrid: Siglo XXI, 1988), pp. 19–27; R. Pastor, ‘Sur la genèse du féodalisme en Castille et dans le León, Xe–XIIe siècles. Point de départ pour une histoire comparative’, in H. Atsma and A. Burguière (eds), Marc Bloch aujourd’hui. Histoire comparée et sciences sociales (Paris: EHESS, 1990), pp. 259–70, at pp. 263, 266; J. A. García de Cortázar and E. Peña Bocos, ‘Poder condal ¿y “mutación feudal”? en la Castilla del año mil’, in M. I. Loring García (ed.), Historia social, pensiamento historiográfico y edad media (Madrid: Ediciones del Orto, 1997), pp. 273–98, at p. 297; R. Collins, ‘The Spanish kingdoms’, in T. Reuter (ed.), The New Cambridge Medieval History, III, c.900–c.1024 (Cambridge: Cambridge University Press, 1999), pp. 670–91, at pp. 673, 691; G. Martínez Díez, El condado de Castilla (711–1038). La historia frente a la leyenda, 2 vols (Madrid: Junta de Castilla y León, 2005). 13 M. Burgos Alonso, J. Calvo Poyato, M. Jaramillo Cervilla, S. Martín Guerrero, Historia. 1er ciclo, Serie Aula Abierta (Madrid: Anaya, 2001), pp. 128–9. Cf. G. Martín Muñoz, ‘Entre cliché et préjugé. L’Islam, le monde arabe et Al-Andalus dans le système éducatif espagnol’, in J. A. Alcantud and F. Zabbal (eds), Histoire de l’Andalousie. Mémoire et enjeux (Montpellier: L’Archange Minotaure, 2003), pp. 133–44. 14 Cf. R. Pastor, Resistencias y luchas campesinas en la época del crecimiento y consolidación de la formación feudal. Castilla y León, siglos X–XIII (Madrid: Siglo XXI, 2nd edn, 1990), pp. 9, 45 (first published 1980); R. Pastor, C. Estepa Díez, J. A. García de Cortázar, J. L. Abellán, J. L. Martín, Sánchez Albornoz a debate (Valladolid: Universidad de Valladolid, 1993). 15 Note the comments of Linehan, History and the Historians, pp. 191–9. 16 M. Vigil and A. Barbero, ‘Sobre los orígenes sociales de la Reconquista. Cántabros y vascones desde fines del Imperio romano hasta la invasión musulmana’, Boletín de la Real Academia de la Historia, 156 (1965), pp. 72–339; their main work is A. Barbero and M. Vigil, La formación del feudalismo en la Península Ibérica (Barcelona: Crítica, 1978), especially pp. 155–200, 354–404, and particularly pp. 370–1, 401, in this respect.



distinctive contribution, for example, was the speed of the change; he argued for a sharp break with the past and the very rapid development of feudal relationships in Catalonia in the mid-eleventh century (but for the slower evolution of essentially the same processes in Galicia, León and Castile).17 When Reyna Pastor spoke at the Marc Bloch conference of 1986, she adopted essentially the same chronology for León and Castile but pointed to some differences in the process: in particular, the development of the seigneurie banale was a direct consequence of desertion and subsequent colonization, and not of the development of anything like the classical Frankish manor; and vassalage developed after the feudal relations of production rather than before. Earlier she had stressed that León and Castile shared in western European trends, notwithstanding differences like the absence of religious overtones to peasant conflicts.18 Difference was also a key aspect of Carlos Estepa’s influential paper of 1989, with its much quoted identification of the three key elements of feudal property – propiedad dominical, dominio señorial, señorio jurisdiccional (lordly property rights [cf. domanial lordship], seigneurial lordship, and the more established territorial seigneurie); he stressed both the different forms of feudalism to be found in Castile and the several different factors leading to its formation. The complexity of the changes is a point he has emphasized recently, while insisting that Barbero and Vigil’s work remains the starting point, a view that is widely shared.19 Salrach has also recently suggested the greater power of the king in the feudalism of León and Castile.20 Discussion still continues today – witness the major conference on lords, serfs, and vassals in 200121 – although nowadays it focusses as much on the nature of lordly power as on feudalism as such. The presence of feudalism, however, is now embedded in

17 P. Bonnassie, La Catalogne du milieu du Xe à la fin du XIe siècle: croissance et mutations d’une société, 2 vols (Toulouse: Association des Publications de l’Université de Toulouse-Le-Mirail, 1976), especially vol. 2, pp. 539–680; id., ‘From the Rhône to Galicia: Origins and modalities of the feudal order’, in P. Bonnassie, From Slavery to Feudalism in South-Western Europe, trans. J. Birrell (Cambridge: Cambridge University Press, 1991), pp. 104–31 (first published 1980); P. Bonnassie, La Catalogne au tournant de l’an mil. Croissance et mutations d’une société (Paris: A. Michel, 1990), especially pp. 263–358. 18 Pastor, ‘Sur la genèse du féodalisme’, pp. 266–8; ead., Resistencias, e.g. pp. 14, 249–50. 19 C. Estepa Díez, ‘Formación y consolidación del feudalismo en Castilla y León’, in En torno al feudalismo hispánico. I congreso de estudios medievales (Ávila: Fundación Sánchez-Albornoz, 1989), pp. 157–256, at pp. 161–2, 196, 255; id., ‘Comunidades de aldea y formación del feudalismo. Revisión, estado de la cuestión y perspectivas’, in M. J. Hidalgo, D. Pérez, M. J. R. Gervás (eds), “Romanización” y “Reconquista” en la península ibérica: nuevas perspectivas (Salamanca: Universidad de Salamanca, 1998), pp. 271–82, especially p. 282. The view may be widely shared, but there remain some opponents; see C. J. Wickham, ‘Le forme del feudalesimo’, in Il feudalesimo nell’alto medioevo, Settimane di studio del centro italiano di studi sull’alto medioevo, 47 (2000), pp. 15–46, at pp. 24–5. 20 J. M. Salrach, ‘Les féodalités méridionales: des Alpes à la Galice’, in E. Bournazel and J.-P. Poly (eds), Les féodalités (Paris: Presses Universitaires de France, 1998), pp. 313–88, at p. 373. 21 Señores, siervos, vasallos en la alta edad media. XXVIII Semana de estudios medievales, Estella, 16 a 20 de julio de 2001 (Pamplona: Gobierno de Navarra, 2002).



Spanish historiography, without the ‘measure of excitability’ which Rees Davies noted usually attended such talk.22 Alongside the acceptance of feudalism, other themes emerged in recent decades. In contrast to the single unique identities of Castro and Sánchez-Albornoz, regional history became prominent in all its variety; close attention to local development brought systematic study of, for example, monastic seigneuries – like the hugely influential works of García de Cortázar on San Millán de la Cogolla and of Mínguez on Sahagún; in the 1970s and 1980s, as Reyna Pastor put it, there developed a ‘sea of specificities’.23 Other issues emerged in the context of the feudal discussion, taking on a life of their own: the hierarchization of flat social structures; the collapse of community collectivities in the face of growing seigneurial power; the strength, or otherwise, of public power.24 Most recently a new trend has emerged, a trend that questions the fact of depopulation itself and therefore of repopulation and colonization, especially in the Duero valley. Moslem invasion and campaigning may not after all have sent the Hispanic population of the meseta fleeing north into the mountains; most stayed where they were, continuing to farm; and their settlements were in some parts connected through networks of supra-local units. One heard it said in Madrid in May 2005 that no-one can possibly believe the old idea of repopulation any more. The continuity of population in – for example – Castile has been developed by Julio Escalona and for parts of León by others such as Iñaki Martín.25

22 R. R. Davies, The First English Empire: Power and Identities in the British Isles 1093–1343 (Oxford: Oxford University Press, 2000), p. 103. For a useful survey of approaches, J. A. García de Cortázar, ‘La formación de la sociedad feudal en el cuadrante noroccidental de la península ibérica en los siglos viii a xii’, Initium, 4 (1999), pp. 57–121, especially at pp. 69–75. 23 J. A. García de Cortázar y Ruiz de Aguirre, El dominio del monasterio de San Millán de la Cogolla (siglos X a XIII). Introducción a la historia rural de Castilla altomedieval (Salamanca: Universidad de Salamanca, 1969); J. M. Mínguez Fernández, El dominio del monasterio de Sahagún en el siglo X (Salamanca: Universidad de Salamanca, 1980); Pastor, ‘Sur la genèse du féodalisme’, p. 261. 24 See for example I. Álvarez Borge, ‘Sobre la formación de la gran propiedad y las relaciones de dependencia en Hampshire (Wessex) y Castilla en la alta edad media’, in id. (co-ord.), Comunidades locales y poderes feudales en la edad media (Logroño: Universidad de la Rioja, 2001), pp. 21–63, at pp. 43–7; J. Escalona Monge, ‘De “señores y campesinos” a “poderes feudales y comunidades”. Elementos para definir la articulación entre territorio y clases sociales en la alta edad media castellana’, in Álvarez Borge (co-ord.), Comunidades locales, pp. 115–55; Estepa Díez, ‘Formación y consolidación del feudalismo’, pp. 164–74; Salrach, ‘Féodalités méridionales’, pp. 320–1. 25 J. Escalona Monge, Sociedad y territorio en la alta edad media castellana. La formación del alfoz de Lara, International Series No. 1079 (Oxford: BAR, 2002) and id., ‘Unidades territoriales supralocales: una propuesta sobre los orígenes del señorío de behetría’, in C. Estepa Díez and C. Jular Pérez-Alfaro (co-ord.), Los señoríos de behetría (Madrid: CSIC, 2001), pp. 21–46; M. C. Rodríguez González and M. Durany Castrillo, ‘Ocupación y organización del espacio en el Bierzo Bajo entre los siglos V al X’, and J. M. Mínguez, ‘Continuidad y ruptura en los orígenes de la sociedad asturleonesa. De la villa a la comunidad campesina’, both in Studia Historica. Historia Medieval, 16 (1998), pp. 45–87, 89–127; I. Martín Viso, Fragmentos del Leviatán. La articulación



Modern Spain and the past What was made of the past was, as ever, conditioned by contemporary contexts. By western European standards Spain was economically underdeveloped at the start of the twentieth century; the late nineteenth and early twentieth centuries were characterized by the presence of exceptionally unstable governments, not helped by crushing naval and military defeats in Cuba (1898) and Morocco (1921)26 – this was a country only to be explained by the ‘tensions caused by the imposition of “advanced” liberal institutions on an economically and socially “backward” and conservative society’ (Raymond Carr).27 The days of Spanish Empire were clearly over. Against a background of poverty, and of agricultural and industrial revolutions that had barely started in 1900,28 republican, socialist, and anarchist parties began to compete for the commitment of workers.29 Organized labour, however, was highly fragmented.30 At the same time, the strong regional identities apparent in northern Spain began to be politicized, as in Catalonia in the 1890s. The upshot of this political instability, and of the violent protests which sometimes accompanied it, was military dictatorship – at first the dictatorship of Primo de Rivera in 1923, which signalled the end of parliamentary monarchy for over fifty years, followed (after the interlude of the Second Republic) by Civil War and the dictatorship of Franco (1939–75).31 Franco’s regime brought economic transformation, industrialization, consumer goods, and massive migration from country to town, but it was politically very repressive, lacking adult suffrage and freedom of political association.32 The system was showing signs of change

26 27 28

29 30 31 32

política del espacio zamorano en la alta edad media (Zamora: Instituto de Estudios Zamoranos, 2002); S. Castellanos and I. Martín Viso, ‘The local articulation of central power in the north of the Iberian peninsula (500–1000)’, Early Medieval Europe, 13 (2005), pp. 1–42, especially pp. 21–41; cf. J. J. Larrea, La Navarre du IVe au XIIe siècle (Paris/Brussels: De Boeck, 1998), pp. 165–6, 183–211. For a helpful survey in English see J. Escalona Monge, ‘Mapping scale change: Hierarchization and fission in Castilian rural communities during the tenth and eleventh centuries’, in W. Davies, G. Halsall, A. J. Reynolds (eds), People and Space in the Middle Ages, 300–1300 (Turnhout: Brepols, 2006), pp. 143–66, at pp. 143–6. Vicens Vives, Approaches, p. 178; J. Pan-Montojo (ed.), Más se perdió en Cuba. España, 1898 y la crisis de fin de siglo (Madrid: Alianza, 1998), especially pp. 151–260. R. Carr, Modern Spain 1875–1980 (Oxford: Oxford University Press, 1980), p. 1. For the background, see N. Sánchez-Albornoz, España hace un siglo: una economía dual (Madrid: Alianza, 2nd edn, 1977), pp. 69–80; for detailed analysis, J. Vicens Vives (ed.), Historia social y económica de España y América, 5 vols (Barcelona: Vicens Vives, rev. edn, 1974 [first published 1957]), vol. 5, pp. 3–283; R. Carr, Spain 1808–1975 (Oxford: Oxford University Press, 2nd edn, 1982), especially pp. 398–411. Vicens Vives (ed.), Historia, vol. 5, pp. 315–49. See Carr, Spain 1808–1975, pp. 439–55. Vicens Vives (ed.), Historia, vol. 5, pp. 368–83. See Carr, Spain 1808–1975, especially pp. 705–9, 750–2; E. Malefakis, ‘Spain and its Francoist heritage’, in J. H. Herz (ed.), From Dictatorship to Democracy: Coping with the Legacies of Authoritarianism and Totalitarianism (Westport, CT: Greenwood Press, 1982), pp. 215–30, at pp. 217–19.



from the late 1960s, as student radicals began to protest, reformists argued for liberalization, the support of the Church weakened, and the earlier movements of regional autonomy revived.33 Constitutional monarchy and democracy were restored rapidly after Franco’s death: King Juan Carlos immediately appointed ministers and the first free legislative election since 1936 established the new Cortes. The constitution was agreed and approved in 1978 and Catalan regional government (the Generalitat) re-established in that year.34 There have been massive changes in Spanish society and politics in the postFranco era, given the end of censorship and of language suppression, the arrival of a free press and freedom to worship, the great increase in university enrolment, the development of a welfare state and the establishment of a constitution that provides for regional government.35 Already opening up to wider European influence and contact in the 1960s, through tourism and emigration, accommodation to European standards and practices reached a new stage with accession to the European Union in 1986 and another with adoption of the Euro in 1999. As John Hooper commented in 1995, no longer conquistadores, large sections of the Spanish public have become hostile to war;36 church attendance rapidly declined; immigrants brought new cultures to Spain; and seventeen regions achieved autonomous status within the Spanish state. The development of strong regional political identities is of considerable significance for the understanding of modern Spain. The movement was already apparent by 1900, when the Catalans were leading the argument for political devolution and the end of ‘Castilian hegemony’.37 The movement for regional autonomy was initially a characteristic of the north (Galicia, the Basque country, Catalonia): three separate languages (Galician, Basque, and Catalan), each with its own range of dialects, symbolized northern identities, in opposition to the Castilian that centralizing governments adopted as Spanish – a dominance in language that was echoed in politics and culture, castellanocentrismo.38 Late in the Franco period, a movement for greater regionalism became evident.39 The constitution of 1978 provided mechanisms for other regions to achieve autonomous status and they quickly seized the opportunity as autonomanía spread.40 Hence, nowadays, for many people the region comes before the state.41 33 A. López Pina and E. L. Aranguren, La cultura política de España de Franco (Madrid: Taurus, 1976), pp. 191–7. 34 Malefakis, ‘Spain’, pp. 226–7. 35 J. García Morillo, La democracia en España (Madrid: Alianza, 1996), pp. 294–5, 421. 36 Hooper, New Spaniards, p. 105. 37 Vicens Vives, Approaches, pp. 141–50; Carr, Spain 1808–1975, pp. 547–56. 38 Escalona Monge, Sociedad y territorio, p. 8. 39 Cf. regional attitudes polled in the 1960s and early 1970s, López Pina and Aranguren, Cultura política, pp. 96–7, 148–54. 40 Malefakis, ‘Spain’, p. 228. See M. García Ferrando, Regionalismo y autonomias en España 1976– 1979 (Madrid: Centro de Investigaciones Sociológicas, 1982). 41 Hooper, New Spaniards, p. 371.



Debate about the peoples of Spain surfaced during much of the twentieth century: was there one Spanish people or were there several, reflecting those strongly demarcated and developing regional identities? The impact of the Franco regime here was marked: the Nationalists were strongly on one side of the peoples debate. When they fought for power in 1936, they stood for pan-Spanish nationalism – the whole country, one state – as against the different potential nationalisms of the regions or the supra-national ideology of communism.42 Their very insistence on the one-ness of Spain in itself enhanced regional identities. Moreover, the Nationalists used the past in the argument for unity: Spain had a long history, its modern political identity established by the middle ages, particularly in the context of the long war fought by the Christians against the Moslems. For them, the ultimate outcome of all that fighting, and of all the small kingdoms involved, was the Spanish state. Indeed, Castile was the core of the state, and the rise of Castile could be traced back to the tenth century. Nationalist Spain was also Catholic Spain, as some historians had argued even before the Franco era:43 it was Christianity that had brought unity to Spain; Christianity had in the end triumphed.44 And just as Spain had been the bulwark of western Christendom in the early middle ages, the last stand against the Moslem onrush, so the Nationalists – having won against republicans and against the supra-national alliances of the labour movement – could pursue a new reconquest against communism. A veritable crusade.45 The image enhanced the political message and in its turn was reinforced as an interpretative model. Historians, however, turned in another direction. Castro reacted against these twentieth-century attempts to ‘de-Islamize and de-Judaize the Spanish past’;46 he worked to demonstrate the extensive interaction of Islamic and Jewish populations with the Christian population in Spain, reversing an anti-semitic trend that had been present for centuries. Disturbed by the political regime, he left Spain for the Americas, holding chairs in the United States after a brief spell in Buenos Aires. (Sánchez-Albornoz also fled Spain, although his different views oddly reflected those of the regime he so disliked.) Thus did two of the most prominent historians of the mid-twentieth century both reflect and oppose the dominant political ideology. Historical attitudes began to change further in the 1960s, even before Franco’s death in November 1975, as scepticism about the determining effects of

42 Linehan, ‘History in a changing world’, pp. 10–11; cf. Jover, ‘Panorama’, p. 1026. 43 Cf. Barbero and Vigil’s caricature: ‘Spanish nationality already existed in the Visigothic kingdom of Toledo; disappearing, but then reviving with Pelayo in Covadonga, it was continued by the Astur-Leonese kings, who transmitted the ancient Gothic national heritage to the Castilian monarchs, who completed the national Reconquest with the fall of Granada’, Formación del feudalismo, p. 234. 44 Linehan, History and the Historians, pp. 79–80, 192, 206. 45 Linehan, ‘History in a changing world’, pp. 12–14. 46 Castro, Spaniards, p. ix.



Christianity began to grow and the power of the nationalist, Catholic ideology began to weaken. Historians dealt more with social and economic dimensions and less with the religious, in which context came the arrival of feudalism and the concern with solid local studies. Catalan history writing, in particular, tended to stress the difference between Catalan development and that of the rest of Spain. As the pace of political change increased in the late twentieth century, the contexts in which the historian operated not only widened; they multiplied. When Spain joined the European Community, it was no longer ‘different’; it became part of the European family, although a part that retained strong cultural links with Spanish America.47 Parliamentary monarchy within the European Union provides the context for Spain to share in European initiatives and European concerns. In the world of the twenty-first century, political similarities across Europe are underlined (and shared global threats – as the Madrid bombing of March 2004 [itself reflected in the London bombing of July 2005] reflects New York on 9/11 [2001] rather than more local problems). Cultural fusion has now become a matter of acute political sensitivity: Christian/Moslem integration within civil society is the message of liberal political leaders, although neo-conservative and extremist factions argue for polarization as they do in many parts of the world. Indeed, a conservative stream in Spanish politics still rehearses the Christian versus Moslem paradigm – witness ex–Prime Minister Aznar’s speech at Georgetown University, when he commented that the Spanish had had problems with Al-Qaeda and Islamic terrorism since the early eighth century: at that time, he said, ‘a Spain recently invaded by the Moors refused to become just another piece in the Islamic world and began a long battle to recover its identity. This Reconquista process was very long’ (21/9/2004). Now, therefore, in politics there is a complex interplay between the autonomous regions, the state, the wider political association that already encompasses twenty-five European countries and global imperatives.48 The current generation of medieval historians reflects this complexity – well informed about historical development beyond Spain, open to new ideas, thinking comparatively, embracing cultural diversity; while it tends to reflect the liberal position on integration, it is also strongly characterized by the strength of the reaction to old orthodoxies, particularly to those of the early and mid-twentieth century.

The continuing medieval issues What then has happened to the issues that occupied medieval historians of the twentieth century? Does Spain continue to be different? No; it now seems to be accommodated within the wider European world, although there are clearly differences of development and emphasis. Are there some unchanging continuities?

47 Closa and Heywood, Spain and the European Union, pp. 31, 34, 245 (citing El País 11/6/95). 48 Cf. García Morillo, La democracia, pp. 178–91.



Thomas Glick, writing in 1979, thought that the real underlying issues were to do with cultural substrates and their continuity, as opposed to cultural fusion: ‘to what extent do enough cultural elements persist over very long periods of time?’49 His answers were subtle but the question is not really answerable from early medieval material. In an intellectual environment which had expectations of a politically correct cultural integration, historians’ preoccupations moved away from issues of cultural fusion or continuity in the 1990s, although the reaction against depopulation struck a blow for continuities of a kind. The overt preoccupations were different. Was Spain feudal? Although Spanish historians had now adopted ‘feudalism’ as a Spanish characteristic, European feudalism had moved on (and some would say it had disappeared). Gone were the institutional invariables of fief and vassal; in had come the complex of fluctuating feudo-vassallic relationships. In too were a couple of decades of debate on feudal mutation – a shift in the entire socio-economic system of, for some, astonishing rapidity; for others a change so slow it was barely noticeable.50 Spanish historians quickly accommodated the change, characterizing the feudality of much of northern Spain as a social system rather than a complex of institutions and noting seigneurial pressure on peasants there as elsewhere.51 In the past five years or so seigneurialization (that is, embedding, and often territorializing, the political power of private lords) has largely been accepted as the dominant process of the tenth to twelfth centuries in northern Spain and there is something approaching consensus on the fact of it.52 Where there are remaining issues, they are on the speed and timing of the development rather than on the fact – did it start in the tenth century? was the real change in the eleventh? Other issues under current discussion are to do with fiscality, with public power (presence, absence, growth, and its relationship with comital power in Castile) and with seigneurial relationships with peasant communities.53

49 Glick, Islamic and Christian Spain, p. 8. 50 For changing approaches, see especially Structures féodales et féodalisme dans l’Occident méditerranéen (Xe–XIIIe siècles). Bilan et perspectives de recherches (Rome: Ecole française de Rome, 1980); J.-P. Poly and E. Bournazel, La mutation féodale: xe–xiie siècles (Paris: Presses Universitaires de France, 1980); Bournazel and Poly (eds), Les féodalités. For the sceptical approach, S. Reynolds, Fiefs and Vassals: The Medieval Evidence Reinterpreted (Oxford: Oxford University Press, 1994). 51 Salrach, ‘Féodalités méridionales’, p. 369; A. Isla Frez, La alta edad media. Siglos VIII–XI (Madrid: Síntesis, 2002), pp. 291–2. 52 See, for example, Larrea, La Navarre, pp. 295–6, 406; Escalona, ‘Unidades territoriales supralocales’, p. 26; I. Álvarez Borge, ‘Estructuras de poder en Castilla en la alta edad media: señores, siervos, vasallos’, in Señores, siervos, vasallos, pp. 269–308, at p. 308. 53 Fiscality: Larrea, La Navarre, pp. 251–3; Isla Frez, La alta edad media, p. 284; C. Estepa Díez and C. Jular Pérez-Alfaro, ‘Prólogo’, in iid. (co-ord.), Los señoríos de behetría, pp. 9–18, at p. 15. Public power: Salrach, ‘Féodalités méridionales’, pp. 320, 362–3; García de Cortázar and Peña Bocos, ‘Poder condal ¿y “mutación feudal”?’; Álvarez Borge, ‘Estructuras de poder en Castilla’, pp. 300–4. Peasant communities: ibid., pp. 281–83.



These are largely new issues. What of the legacy of depopulation and repopulation? Depopulation was in fact central to the interpretative models that were dominant for most of the twentieth century. The free spirit of the colonizers can no longer be called on to explain the ‘delay’ in the development of aristocratic power, the ‘late’ onset of feudality, the quirks of the Spanish experience (like feudo-vassallic institutions following feudality). Without it we need different explanatory models. Despite the strength of the idea of depopulation, doubts had been expressed as long ago as the early twentieth century and qualifications to the stark model became very evident in the later twentieth century.54 Even Ubieto Arteta’s 1958 maps have areas ‘without organized population’, rather than areas with no population.55 And Menéndez Pidal was at the same time making the point that the word populare in early medieval texts could refer to the imposition of new political and administrative structures rather than to new settlements.56 Some (more modest) decline in population now tends to be invoked in its place, as is new colonization arising out of sheer peasant dynamism in a context of demographic growth;57 ‘depopulation’ is understood as loss of structure rather than whole-scale desertion;58 and the current debate tends to focus on the relative significance in colonization of indigenous as against immigrant population. But the textual references to deserts and to journeys to populate new lands are still present in tenth-century and later texts – like the peasants of Braga who claimed that their ancestors had come from Oviedo.59 Following recent analyses, some of the texts can be reasonably interpreted as political claims: repopulation is a means of legitimizing the claims of ninth-century and later kings.60 This works particularly well for the chronicles and for royal and aristocratic charters. But the language of repopulation runs through all levels

54 R. Menéndez Pidal, ‘Repoblación y tradición en la cuenca del Duero’, in M. Alvar, A. Badía, R. de Balbín, L. F. Lindley Cintra (eds), Enciclopedia Lingüística Hispánica, 2 vols (Madrid: CSIC, 1960–7), vol. 1, pp. xxix–lvii; cf. Barbero and Vigil, Formación del feudalismo, pp. 226–7. Even Sánchez-Albornoz allowed for continuing isolated settlements south of the Duero, ‘The frontier and Castilian liberties’, p. 37. 55 Ubieto Arteta, Alas Histórico, p. 40; cf. Pastor, ‘Sur la genèse du féodalisme’, p. 264. 56 Menéndez Pidal, ‘Repoblación y tradición’, p. xxx; cf. García de Cortázar, Sociedad rural en la España medieval, pp. 21–2; Castellanos and Martín Viso, ‘Local articulation’, p. 23. 57 Isla Frez, La alta edad media, pp. 293–4; Larrea, La Navarre, pp. 196, 589–91; Martín Viso, Fragmentos del Leviatán, especially pp. 43–59; peasant colonization with comital control, Bonnassie, Catalogne, pp. 39–57. 58 J. J. García González in the ‘Mesa Redonda’, in Estepa Díez and Jular Pérez-Alfaro (co-ord.), Los señoríos de behetría, pp. 261–309, at pp. 263–4. 59 See Sánchez-Albornoz, ‘Repoblación’, especially pp. 591–620, for a review of occurrences; id., Despoblación y repoblación, pp. 215–343, depends heavily on evidence of the existence of small proprietors for its repopulation argument; Pastor, Resistencias, p. 33. 60 Barbero and Vigil, Formación del feudalismo, pp. 216–28, 232–49; Linehan, History and the Historians, pp. 88–127; J. Escalona, ‘Family Memories. Inventing Alfonso I of Asturias’, in I. Alfonso, H. Kennedy, J. Escalona (eds), Building Legitimacy: Political Discourses and



of charter, from royal to peasant issuers; it is embedded in tenth-century perspectives. It cannot all be explained in terms of political claims. We could draw some lessons from Rees Davies on colonization. In the context of Britain and Ireland of the central middle ages, he demonstrated how ‘colonizing’ may at one level have been about kings staking claims to new lands but often in practice it was about aristocrats establishing their physical presence. He drew attention to the complexities of frontier life and the multiplicity of frontier lands – ‘their boundaries changing over time and cutting across the national and state divisions within which we have imprisoned so much of our historiography’.61 Chris Wickham’s work on the cultivation of ‘deserts’ and ‘clearance’ of forests is also relevant. He demonstrated that such textual references have more to do with the alienation of royal rights than with any previous economic marginality.62 In effect ‘clearance’ often meant the establishment of new kinds of proprietary right, new powers over inhabitants and new expectations of regular and increasing returns to owners.63 There is much in this model that can help to explain the repeated references to desertion and repopulation in Spanish texts: while, of course, there might have been some genuine clearance, just as there must have been some genuine migration, the presura (cf. ‘assart’ in English) did not have to be a freshly cleared plot; it might be a plot with freshly determined liabilities. For the future, some more field survey and settlement archaeology would help: we need to know more about the chronology of rural settlement, the changing size of rural settlements over time and the physical frequency of new settlements.64 Cultural fusion may have disappeared as an historical issue, but identity is an aspect of it that very much remains. Tenth-century material is critical to any assessment of identity issues because the tenth century offers the first opportunity to address them with the help of a reasonable quantity of written material. How did people express identity in the tenth century, when substantial (and varied) records begin? Did a consciousness of shared religious or ethnic culture influence their sense of group identity? Were they conscious of being Christian or Jew or Moslem, or did they define themselves in other ways? Recent historians have pointed out that chroniclers of medieval Spain only gradually came to use labels with a religious connotation for different parties, because political alliances were

61 62



Forms of Legitimation in Medieval Societies (Leiden: Brill, 2004), pp. 223–62, especially p. 257 for depopulation/repopulation as a ‘legitimating device’. Davies, First English Empire, p. 92. C. J. Wickham, ‘European forests in the early middle ages: Landscape and land clearance’, in L’ambiente vegetale nell’Alto Medioevo, Settimane di studio del centro italiano di studi sull’alto medioevo, 37 (1989), pp. 479–548. See also P. Fouracre, ‘Marmoutier: Familia versus family: The relations between monastery and serfs in eleventh-century north-west France’, in Davies, Halsall, Reynolds (eds), People and Space, pp. 255–73. Cf. García de Cortázar, San Millán, p. 101.



determined by other factors.65 References to Christians do occur but they are rare before the late eleventh and early twelfth centuries – at which point religious difference began to colour the portrayal of military conflict. For the Moslems of southern Spain, tenth-century chronicles used terms borrowed from the Old Testament – Ishmaelites and Chaldeans – but also occasionally the words Arab and Saracen. The charters of tenth-century northern Spain number thousands and are full of localizable and personal detail; as such they have considerable potential as a source of information on naming practice. Usage of the word ‘Christian’ in these texts essentially reflects that of the chronicles: it is surprisingly rare before the eleventh century and when it does occur it is nearly always in a formulaic expression of piety, not specific to individuals or groups (for example, ‘I, the least of Christians, make this gift’ or ‘whoever denies this gift shall be excluded from the company of Christians’).66 Occasionally in the late tenth century the term occurs as a political identifier: in 988 we hear that war against the Christians was started by the barbarians and in 998 that rebels went to the king of the Moslems (muzlemitarum), who was intending to destroy the kingdom of the Christians.67 However, such references are extremely rare, though much more frequent in the charters of the following centuries. In tenth-century charters the enemy from the south is more usually Saracen (sarraceni, sarrazeni) than any of the more literary terms – Saracens invaded, rebels joined the Saracens – although ‘barbarians’ feature occasionally too.68 Perhaps surprisingly, the word Moor occurs as well (strictly, Berbers from north Africa): Moors were put to flight from León c. 958; Moorish slaves were handed over in gifts to wives and to monasteries earlier in the century; and a rather different but suggestive dimension is indicated by the Moorish blankets (genabe maurisca) sometimes handed over as part of the price paid in commercial transactions, as happened twice in León in 959.69 Jews are occasionally identified in the charters: the purchaser of a vineyard is identified as ebreo in 984, as is a purchaser of a farm in the Porma valley in 997, while the bounds of another vineyard in the suburbs of León touch that of Abzecri iudeo – ‘bounded on one side by land of the sisters of Savarico, on the second by that of the priest Stephen, on the next by that of Abzecri iudeo as far as the road to Covellas, and on the fourth by the road to the River Torío’.70 The references suggest a picture of Jewish members of the rural farming community, mingling side by side with Hispanic farmers in rural society, but differentiated. (Some of

65 A. Christys, ‘Crossing the frontier of ninth-century Hispania’, in D. Abulafia and N. Berend (eds), Medieval Frontiers: Concepts and Practices (Aldershot: Ashgate, 2002), pp. 35–53, at pp. 43–9. 66 For example, S128; Liii515. 67 S340, Liii581. 68 S340; Cel266; S9. 69 S159, Cel576, S114; Lii318, Lii322. 70 Lii495, Liii579, S290; but note Li19: ‘quondam iudeus, postea vero christianus’.



the properties are small-scale – interspersed with the holdings of peasant society.) The term is rare, but the point might perhaps be expanded by the quite frequent references to people called David and Abraham – who were also integrated with the local farming communities, and were not ethnically differentiated – arguably forerunners of Castro’s cultural fusion. What is much more striking than the specific terms, however, is the fact that all such terms are rare. Most people, and there are thousands, have no such label. If they are identified they are much more commonly identified by region or by local group – the person from Galician or Leonese or Zamoran territory or the woman from Villa X or Place Y. Alternatively, they might be members of the council (concilium, sometimes collacium/-tium as a direct synonym) of this village or that – a limited occurrence but one whose contexts suggest a stronger sense of the identity of a collectivity than the religious labels do.71 Concilium is used of single village councils in León and Castile from the 930s. It explicitly refers to a collectivity which made gifts, or made agreements about rights, or confirmed or witnessed deals; named individuals might speak for the whole concilium.72 It was clearly the assembly of that part of the community that went to meetings (not necessarily the whole community) and was the group that took decisions for the community. In some cases it looks as if a community developed a collective identity where it had to meet frequently, for example to negotiate with new lords (rather than, as is sometimes maintained, communal identity collapsed in the face of new lordships). Clearly people in their daily lives thought much more in terms of place and of very local collectivities than of religious or ethnic differentiation. So, four related issues dominated twentieth-century medieval history writing (cultural fusion, continuities, reconquest and repopulation, feudalism) – all with a bearing on identity and on the distinctiveness of Spain. All are very easily understandable in the context of twentieth-century politics. Changes from those preoccupations came in the 1970s; again the changes are very understandable in the context of twentieth-century political change. But issues remain. They remain because of lacunae in textual provision for the eighth and ninth centuries and because as yet there is not enough archaeology. But there are also aspects to address that have not yet been properly addressed. The construction of group identity is important; it needs much more thought in a tenth-century context both because it underlies the much-discussed developments of the central middle ages and because it is central to our understanding of social relationships, especially at local level. The political issues of the moment are both global and regional: how to build a successful multi-cultural society; how to deal with terrorism.73 Will

71 For example, Liii515, Liii577; S268, S298, S338; C70, C89, C192. 72 S44, S335, C89, Lii466, for example. 73 García Morillo, La democracia, pp. 415–25, ‘El mañana no está escrito’, on terrorism and on twenty years of experiment at breakneck speed.



‘Europeanization’ become deeper, the associated political bonds tighter? Will the regions remain satisfied with present powers or press towards a looser federation or even independence? And how will EU membership impact on the relationship between state and region, especially in the context of enlargement?74 It will be interesting to see how the next generation of historians will reflect those issues: a concern with the microlevel in the context of the macro is already apparent, as is a concern systematically to draw out the parallels between Spain and other parts of Europe.75 What will happen to identity and the relationship of identity to microand macro-polities? Rees Davies could say that in the late eleventh century ‘to outsiders the distinctiveness of Wales was as impressive as its fragmentation’.76 Different as it is in so many respects, perhaps Spain too was as distinctive as it was fragmented. It is the interplay between the fragments – the local – and the various levels of the supra-local that emerges as perhaps the most interesting issue for the next decade.77

74 Closa and Heywood, Spain and the European Union, pp. 84–9, 192–201. 75 For example, I. Álvarez Borge, Comunidades locales y transformaciones sociales en la alta edad media. Hampshire (Wessex) y el sur de Castilla, un estudio comparativo (Logroño: Universidad de la Rioja, 1999). 76 R. R. Davies, The Age of Conquest: Wales 1063–1415 (Oxford: Oxford University Press, 1991 [first published as Conquest, Co-existence, and Change, 1987]), p. 15. 77 I am very grateful to Chris Wickham, Ann Christys and Julio Escalona for their comments on a draft of this paper and to my colleagues Christopher Abel and Axel Körner for advice on modern issues.


Part I

Charter writing

2 LOCAL PRIESTS AND THE W R I T I N G O F C H A RT E R S I N N O RT H E R N I B E R I A I N T H E T E N T H C E N T U RY There is already plenty of excellent discussion in print of the wide range of scribes and notaries who composed and wrote charters in Iberia in the tenth century.1 There were scribes who wrote for kings in royal households; scribes who wrote for bishops; scribes attached to great monasteries like Albelda, San Millán, Cardeña, Sahagún and Celanova, some of whom also wrote for kings; scribes who wrote for smaller monasteries and lay religious households, sometimes in a very personal style. All that is very well discussed, in the many excellent recent editions of charters, as well as in dedicated books and papers.2 What I want to do is explore some of the less obvious scribes, at a less elevated social level, so that we can extend our consideration of the power that comes with writing to a wide social spectrum. My focus is on northern Iberia, excluding Catalonia, in the late ninth and tenth centuries – that is, the period when we first have extensive documentation that comes from a wide social context. Although I have considered material from across this entire area, that is from Galicia to Aragón, most of the examples I shall cite will come from León, Otero de las Dueñas and Sahagún material; that is because of the number of charters on single sheets that still survive in those collections.

1 First published in J. Escalona and H. Sirantoine (eds), Chartes et cartulaires comme instruments de pouvoir. Espagne et Occident chrétien (VIIIe–XIIe siècles) (Toulouse: Presses Méridiennes de Toulouse/CSIC, 2013), pp. 29–43. 2 Notable editions are: Emilio Sáez in L, vol. 1 (775–952); Fernández Flórez and Herrero de la Fuente in OD; Carlos Sáez and Ma del Val González de la Peña in LaC. Notable books and papers include: the list of royal scribes and other analyses of royal charters in M. Lucas Álvarez, El reino de León en la alta edad media. VIII. La Documentación real astur-leonesa (718–1072) (León: Centro de Estudios e Investigación ‘San Isidoro’ [CSIC-CECEL], Caja de Ahorros y Monte de Piedad, Archivo Histórico Diocesano, 1995), pp. 219–33 especially; C. Sáez and A. Castillo Gómez (eds), Libros y Documentos en la alta edad media. Los libros de derecho. Los archivos familiares. Actas del VI Congreso Internacional de Historia de la Cultura Escrita, 2 vols (Madrid: Calambur, 2002); J. A. Fernández Flórez, ‘Los documentos y sus scriptores’, in Monarquía y sociedad en el reino de León de Alfonso III a Alfonso VII, 2 vols (León: Centro de Estudios e Investigación ‘San Isidoro’, Caja España de Inversiones, Archivo Histórico Diocesano, 2007), vol. 2, pp. 97–139.



In considering ‘scribes’ it is well to note at the start that there are some serious issues concerning the differentiation of authors and writers, that is, there can be difficulty in distinguishing between the person who composed the words and he or she who wrote them down (where they were different), as well as that in distinguishing between the person in charge of the operation of charter production (if such there was) and the actual writer of a particular document.3 Occasionally there are indications in a text, like the record of Bishop Oveco’s endowment of San Juan de Vega in 950, composed by Abbot Gondesindo and written by the deacon Adulfo, or like the deacon and notary (notarius) Aloyto, who composed and wrote a record of Bishop Rosendo’s endowment of Celanova in 942.4 However, more often either we simply have an indication that one person wrote (scripsit) the charter, with no precise indication of which role he performed, or we have no indication at all. We cannot even rely on the proposition that the verb notuit indicates the composer of the text and scripsit the writer, since the terms can be used haphazardly and inconsistently.5 There is good reason to think that some composers, like Aloyto, wrote their own charters, both choosing the words and writing them down, without the supervision of a master scribe, so we certainly cannot assume that the drafter was always different from the writer. In what follows I shall use the word ‘redactor’ for the person who chose the words where he can be distinguished from the person who wrote them down, ‘writer’ for the person who did the writing where this was clearly the case and ‘scribe’ for ambiguous cases; I shall return to the issues of differentiating between them in due course.

Local priests Just as there were those who wrote for kings and bishops and great monasteries, there were some who wrote for peasants.6 While some of these scribes came from small (or sometimes larger) monasteries, many were priests, and these often seem

3 See H. Atsma and J. Vezin, ‘Autour des actes privés du chartrier de Cluny (Xe–XIe siècles)’, Bibliothèque de l’école des chartes, 155 (1997), pp. 45–60, for a succinct statement of the problem; H. Atsma and J. 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. Poulla, M. H. Smith (eds), Le statut du scripteur au moyen âge (Paris: École des chartes, 2000), pp. 9–20; B.-M. Tock, Scribes, souscripteurs et témoins dans les actes privés en France (VIIe–début du XIIe siècle) (Turnhout: Brepols, 2005), pp. 110–14, for an eloquent discussion of the fact that scribere meant many more things than ‘to write’. See further for scribes in charge of others. 4 Li220; Cel2: qui et notarius dictavi et scribsi, cf. Cel48 (941). 5 Cf. the comments of Fernández Flórez, ‘Los documentos y sus scriptores’, pp. 106–10, noting his caution that we should not expect too precise a distinction between writers and redactors (‘notaries’); also L. Casado de Otaola, ‘Escribir y leer en la alta edad media’, in A. Castillo Gómez (co-ord.), Historia de la cultura escrita: Del próximo oriente antiguo a la sociedad informatizada (Gijón: Trea, 2002), pp. 113–77, at p. 132. 6 See above, p. xi, for my use of ‘peasant’.



to have been local priests.7 We do not have to wait for the development of full parish structures in the later eleventh and twelfth centuries to find priests in rural localities, some of whom served rural communities. There is good evidence of priests placed by an absent owner (be that owner a monastery, bishop or secular aristocrat) to serve a single church – just as the monastery of Samos ‘gave’ the priest Cagito the church of Santo Tomé to live in, a church which he gave back to Samos at the end of his life (c. 973), together with all the gifts he and the church had accumulated.8 There is plenty of evidence of lay religious households (which are often in practice difficult to distinguish from small monasteries) with priests attached9 and a host of evidence of individual priests who owned their own churches, often living there with family and dependents. The Cardeña cartulary, for example, includes many cases of priest proprietors handing their churches over to the major monastery of Cardeña, especially from the 940s onwards,10 but this happened in many other places too: the priest Argivolo gave his church, which his ancestors ‘had built’, and the houses he himself ‘had built’, to the monastery of Sobrado in 964;11 the priest Ero had to forfeit his church in Trobajo, near León, for the homicide he had committed;12 the priest Julián gave his church of Peñacorada, and appurtenances, to Sahagún in 996.13 That there were plenty of local rural priests is clear enough; what we do not know is the density of their distribution, although we should expect it to have been patchy, the result of uncoordinated personal histories, endowments and chance. If we look at records of transactions between lay parties (sale transactions, especially, but also some gifts), a significant proportion concerns the transfer of small properties – plots of arable rather than whole estates, side by side with others in a community’s arable fields, or patches of vines, or fractions of orchards; these are peasant transactions, dealing in peasant properties, like the sale of a small orchard by one lay couple to another in 948 or the sale of half a vineyard in Ribaseca in 949 between two more lay couples.14 That such people could want records is explicit in an early record of c. 850, from Aragón, in which it is noted that three pious lay donors had asked a brother to find them a scribe (scriptor).15 Such records were produced by individuals who often identified themselves as priests; many of these priests were people who appeared in that location only, or in that and a neighbouring locality, and whose records were witnessed by the same 7 That some texts were written by local priests is a point clearly made by Fernández Flórez, ‘Los documentos y sus scriptores’, pp. 104, 107, 126. 8 Sam217. See further W. Davies, Acts of Giving. Individual, Community, and Church in TenthCentury Christian Spain (Oxford: Oxford University Press, 2007), pp. 44–50. 9 E.g. Sam61 (976). 10 E.g. C48 (945), C49 (945), C57 (946), C91 (957). 11 Sob8. 12 Lii507 (985). 13 S350. 14 Li200, Li205. 15 SJP5.



pool of witnesses, that is by people of the locality. They were therefore priests who provided the service of writing, in quasi-notarial fashion, for transactions in which they themselves had no apparent interest, for local people (there were also deacons, and others to whom no office is ascribed, who performed similar functions); they contrast strongly with those who wrote for great monasteries, or for bishops, or for kings and high aristocrats, people whose acts tended to relate to a wide range of places, some distance apart, and which were characteristically witnessed by a different group of witnesses for each different locality. The fact that local priests were indeed local is emphasized by cases in which they had personal property close by that transacted: the priest Stephano, who recorded exchanges between lay parties in Marialba, lying between the rivers to the south of León, in the 920s, himself exchanged plots in Marialba with a peasant couple in 926;16 in 949 one Vicente recorded lay sales to Sahagún in and near Cofiñal, on the southern slopes of the Cantabrian Mountains (75 km north-north-west of Sahagún), and himself sold plots he had there to the monastery a month later. Quite apart from his own local property, he clearly was not attached to Sahagún because three brothers came from the monastery to hand over the price for the lands bought, on behalf of Sahagún.17 The priest Ermegildo, who recorded sales between lay parties in Monte Aurio, just to the east of León, in 985, had a plot adjacent to one of those sold.18 Local priests made records for peasants. There is nothing intrinsically unlikely about this practice: we know of local priests and others writing for minor aristocrats, medium landowners and peasants elsewhere in western Europe in the early middle ages, where there was no apparent church interest in the transactions at the time of the transfer – in Germany, Italy, Switzerland, Brittany, Burgundy, for example.19 That being so, can we point to specific charters and say that they were composed or written by local scribes, who are identifiable as individuals? While we could rarely be completely sure, given the considerable instabilities in the format of the texts and ambiguities in relation to names, nevertheless some individuals are indeed visible, regardless of whether the text survives as a cartulary copy or single sheet copy or original, simply on the

16 17 18 19

Li70; Li69, Li78. S117; S115, S116. Lii502. Cf. Lii449 (977). R. McKitterick, The Carolingians and the Written Word (Cambridge: Cambridge University Press, 1989), especially pp. 89, 111, 118–33; P. Erhart and J. Kleindinst, Urkundenlandschaft Rätien (Vienna: Österreichischen Akademie der Wissenschaften, 2004), pp. 83–105; Tock, Scribes, souscripteurs et témoins, pp. 301–3; P. Erhart, ‘Der Rotulus des Valerius: das Schreiben von Urkunden im frühmittelalterlichen Churrätien’, Geschichte und Region/Storia e regione, 15 (2006), pp. 38–61, especially pp. 50, 58–9; W. Davies, ‘People and places in dispute in ninth-century Brittany’, in W. Davies and P. Fouracre (eds), The Settlement of Disputes in Early Medieval Europe (Cambridge: Cambridge University Press, 1986), pp. 65–84, at pp. 68–70; M. Innes, ‘Archives, documents and landowners in Carolingian Francia’, in W. Brown, M. Costambeys, M. Innes, A. Kosto (eds), Documentary Culture and the Laity in the Early Middle Ages (Cambridge: Cambridge University Press, 2013), pp. 152–88.



basis of their locations and personal associates: the place(s) where they occurred, the individuals for whom they wrote and the people who witnessed the transactions they recorded. To take some precise examples of named local scribes. The priest Durabiles in 939 recorded lay sales of land in Villobera on the Bustillo Brook (a tributary of the Valderaduey) to the monastery of Abellar – some 75 km to the north-west.20 Fredenando, not called priest, recorded lay sales to the rich priest Vincimalo in Villagoya in or near the Valderaduey valley in 943, writing for transactions in this location but not for Vincimalo’s other purchases.21 The priest Iohannes recorded sales between lay parties in Los Oteros, an area to the south-east of León, in 950–51;22 the priest Piniolo recorded sales between lay parties in Tendal (east of León) in 959–60;23 the priest Amorino recorded sales between lay parties in Ujo, in the far north, on the route to Oviedo, in 959 and 962.24 And the priest Vermudo recorded sales between lay parties in Montefrio, on the road to León, in 996 and 999.25 Taking a different dimension of the local, there were also some who wrote in a dedicated fashion for aristocratic households and it is relevant to consider these alongside local rural priests. The priest Ferronius wrote for the aristocrat Assur in the 950s and 960s; Assur bought property here and there, in León, outside the city, and in Tendal, and he was given property in Villa Alva. These transactions relate to scattered places and were witnessed by different sets of witnesses, but Ferronius remained the named scribe for all.26 Fredemundo Liveriz (a priest, though not always labelled such) wrote for the powerful Munio Flaínez from the late 940s to 960s. Munio bought here and there – in the Cea valley not far from Sahagún and on the Esla, especially in and near Aleje (just south of Valdoré, in sight of the Picos) – with different sets of witnesses and Fredemundo also wrote for his wife (perhaps by then widow) on the river Porma; these places are 50 km apart.27 Then 20 Li132, Li135. For the location, see P. Martínez Sopena and M. J. Carbajo Serrano, ‘Notas sobre la colonización de Tierra de Campos en el siglo X: Villobera’, in El pasado histórico de Castilla y León. Actas del I Congreso de Historia de Castilla y León, 3 vols (Burgos: Junta de Castilla y León, 1983), vol. 1, pp. 113–25, at p. 114. Durabiles seems to have been local to Villobera because he only made records there and did not make them for other Abellar purchases, of which there were many, and because he is not noted in the many other Abellar charters; see the following further for this individual. 21 Li172, Li173, Li174; perhaps also the scribe of the lay sale to the monastery of Sahagún of lands in Ataula, near the Valderaduey valley, earlier in the same year, S85. 22 Li226, Li237; cf. also Li234, Villacedré. 23 S169, S182. 24 Lii310, Lii357. 25 Liii572, Liii583. 26 Three sales between lay parties and one gift: S368 (956), S199 (962), S210 (963), S214 (963). 27 Five purchases from lay parties and one gift: OD6 (947), S107 (947), S108 (948), S124 (950), S189 (961); S205 (962); cf. S110 (948), an unlocated sale to Munio witnessed by the priest Fredmundo, but written by the priest Sebastian. We last see Munio in March 962 (S198); his wife’s charter (S205) is dated October 962. For the location of Otero Morisco (OD6) and Valle de Lorenzo (S107,



there is Braolio, who wrote for Munio’s son, Count Flaino Muñoz, in the 980s and 990s; he recorded judicial proceedings at which Flaino presided and confirmation of a gift (on which occasion Flaino also seems to have presided), nearly all at Valdoré, on the upper waters of the Esla.28 The priest Sendino wrote for another count, Munio Fernández, recording gifts and sales to him and an exchange with him variously in Valdevimbre, Toral and La Bañeza, within a zone 20–40 km south of León.29 The priest Domengo wrote for Fruela Vimáraz, in the Astorga region in the 990s.30

Scribal practice Is there any trace of individual identity in the scribal practice of any of these records? There are several ways in which these and other scribes could be thought to demonstrate their individuality: in the grammar and orthography of the Latin they wrote, in their use of standard formulas and in their physical habits in writing, especially with regard to hands, marks and abbreviations. In many cases we only have cartulary copies of tenth-century texts and can do little of the latter, but where we have charters on single sheets there is plenty of scope for investigating the materiality of these texts. Some scribes for rural communities, and some members of aristocratic households, wrote a Latin which has some very unclassical forms, both in spelling and in grammar. This is certainly not true of all, and even where it is true it is not easy to analyse since both spelling and grammar tend to be unstable: scribes use now one convention, now another, and inconsistency is common. However, it is broadly true to say that whereas scribes from great monasteries tended to write a Latin which is closer to classical (though frequently with orthographic variation), scribes who wrote for peasants and for aristocratic households wrote a Latin which (in being farther away from the classical) reflected local patterns of speech and vocabulary,

S108, S124), see J. J. Sánchez Badiola, El Territorio de León en la edad media: poblamiento, organización del espacio y estructura social (siglos IX–XIII), 2 vols (León: Universidad de León, 2004), vol. 1, pp. 208, 205–6. The writing of the records of Munio’s many other transactions was attributed to at least nine other scribes, in different locations. 28 OD26 (986), confirmation of a gift to a monastery, before Flaino; OD31 (991) and OD34 (993), judicial proceedings, with fines to Flaino; OD43 (997), court-case before Flaino about intrusion into the property confirmed in OD26. At least one other named scribe wrote for Flaino (Sancho, OD44), but several hands wrote records for him (OD27, OD32, OD33, for example). 29 Liii556 (993), Liii561 (994), Liii562 (994), Liii578 (997). Perhaps he should be identified with the priest scribe Sendino responsible for OD46 (999), a record of a sale to Fruela Vimáraz, Munio’s father-in-law (see A. García Leal, ‘Los condes Fruela Muñoz y Pedro Flaínez: la formación de un patrimonio señorial’, Anuario de Estudios Medievales, 36:1 (2006), pp. 1–110, at pp. 7–13, for their relationship); the scribal signa of Liii562 and OD46, which both survive on single sheets, are very different and there are other differences in scribal habit, but Sendino could have had responsibility for the production of both. 30 OD40, OD41 (both 995). See also n. 29.



that is local vernacular Romance, and may have to some extent reflected pronunciation.31 So, the local priest Iohannes, writing to the south east of León c. 950, used very variant orthography (with much interchangeability of v and b, d and t, n and m) and had a strong tendency to lose or substitute case endings of nouns and adjectives: ‘Ego Todolfo et uxor mea Maria uobis Agube et uxor tua Auria. Uindimus bouis tera nostra propria que abemus in teridorio Leionens’; ‘Ego Gisando et uxor mea Argilo bouis Agodin et uxor Bonilde. Uindimus bouis uila nostra propria que abemus in teridorio Leionens’.32 Local priest Durabiles, writing in the Bustillo valley in the 930s, shows a similar orthographic instability and an even greater tendency to lose or substitute endings. Fredemundo, from the household of Munio Flaínez, wrote with slightly less unstable orthography (apart from interchangeability of t and d ) but more of a tendency to loss of endings: ‘Et si aliquis ommo ad erumpendum uenerit, anc ego, anc filis meis . . . comodo parie uobis ipsa terra duplada uel quantum ad uos fuerit meliorada’; ‘Et si aliquis ommo ad erumpendu ueneri uel uenerimus comodo pariemus uobis ipsas terras dupladas quantum ad uos fuerit meliorada’.33 While Braolio, writing for his son’s household a generation later, in the 990s, wrote with a strong (and unusual) tendency to substitute g for c and with cases, tenses and possessive pronouns that are distinctly unexpected: ‘Ego Fredino et uxor suam Leobina ad uobis Fllayno Munizi et uxorit uestram Iusta Pepizii . . . Et iurabit, indem, Ermegildo gonditionem quet non fecera Ermegildo ipsa plagam’; ‘Et qui de nos . . . unc nostro factum et isto scrito ad erupere quesieri, aut per nos, au per alika pesona, aut per quemlibem homine, pariemus solidos D’.34 Without wanting to overstate the point, there seem to be traces here of individual usage. Charter writing is – of course – highly formulaic. This is particularly valuable because it allows us to see how individual authors treated standard formulas. For example, the local priest Iohannes characteristically introduced a personal element into his dating clauses: the day of the week and in two cases the time of day of the transaction; both are highly unusual in the tenth-century corpus: ‘die dominico, ora III, xi kalendas septembres’; ‘die dominico, v kalendas feuera’; ‘disabado, ora erad quasi uesperum, viiii kalendas aprilis’ (Sunday at the third hour on the eleventh of the kalends of September, Sunday on the fifth of the kalends of February, on the Sabbath [Saturday] at the time of Vespers on the ninth of the kalends of April).35 Or, to take a different approach, there is one standard formula which 31 See R. Wright, Late Latin and Early Romance in Spain and Carolingian France (Liverpool: F. Cairns, 1982), especially pp. 165–75; id., ‘The non-existence of “Leonese Vulgar Latin”’, in R. Wright, Early Ibero-Romance (Newark: Juan de la Cuesta, 1994), pp. 127–34; also my longer discussion in Acts of Giving, pp. 98–106. 32 Li226 (950); Li234 (951). 33 OD6 (947); S205 (962). Both appear to have been written by the same hand (AHDL Otero 3, AHN Sahagún carpeta 875 no. 3). 34 OD31 (991); OD43 (997). 35 Li226 (22/8/950); Li234 ([26]/1/951) – there is a problem with the date of this charter since the fifth of the kalends in 951 was on a Tuesday, not Sunday, but I follow the editor’s emendation (and see his note on p. 325 of Li); Li237 (22/3/951).



is striking in the number of its variant forms, and is thereby particularly useful. The formula is ‘placuit nobis atque convenit nullius cogentis imperio sed propria nostra voluntate accessit nobis ut vinderemus . . .’. It is especially characteristic of sale records, although it can also be found in some records of gift and of exchange, and it is usually the opening formula after invocation and address; it announces that the principal actor has agreed of his or her own free will to take the action recorded. It is extremely common indeed, during the ninth and tenth centuries, occurring more sporadically in the eleventh century but still cited in the early twelfth. It normally appears in one of two versions: that cited above, which has both Frankish analogues and a Catalan analogue in the formulary from Ripoll, and in a variant second version.36 The second version is so common, however, it must have been learnt in many places, and across a long period: ‘placuit nobis atque convenit, nullius quoque gentis imperio neque suadentis articulo, set propria nobis accessit voluntas ut venderemus . . .’; in other words, nullius cogentis imperio is replaced by nullius quoque gentis imperio. Each version tends to occur in micro-regions, regions which are much smaller than the autonomous regions or the provinces of modern Spain: the area around León has different practice from that round Sahagún, both on the meseta, as those around Sobrado, Samos and Celanova in Galicia are also different from each other; these micro-regions reflect the several kinds of local practice around the monasteries and not simply the practices of the monasteries themselves.37 Micro-regional variation is important as a contribution to understanding the training of scribes, but we should note that there is individual variation too: I am currently aware of 38 variants of the phrase nullius cogentis, some of which can be related to individual practice.38 The priest Ferronius, writing for the aristocrat Assur, uses the form nullius . . . egentis: ‘placuit mici adque conuenuenit (sic) nullius quoque egentis imperium neque suadentis articulo set propria mici accessit bone pacis uolumtas’; ‘placuit nobis adque conuenit nullius quoque egentis imperium neque suadentis articulo set propria nobis accessit uolumtas’.39 Fredemundo, writing for the household of Munio Flaínez, uses the spellings nunlis . . . ientis: ‘placuit nobis aque couenit nunlis quoque ientis inperio neque suadentis articulo set propria nobis acit uoluntas’; ‘placuit nobis

36 Formulae Merowingici et Karolini Aevi, ed. K. Zeumer, MGH Leges 5 (Hanover: Hahn, 1886), pp. 10, 90, 186, 229; M. Zimmermann, ‘Un formulaire du Xe siècle conservé à Ripoll’, Faventia, 4:2 (1982), pp. 25–86, at p. 77. 37 For the notion of regional diplomatic – Urkundenterritorien – see, classically, H. Fichtenau, Das Urkundenwesen in Österreich vom 8. bis zum frühen 13. Jahrhundert (Vienna: Hermann Böhlaus, 1971), pp. 46–55, 110–14, 255–6, and more recently P. Erhart, ‘Erratische Blöcke am Alpennordrand? Die rätischen Urkunden und ihre Überlieferung’, in P. Erhart, K. Heidecker, B. Zeller (eds), Die Privaturkunden der Karolingerzeit (Zürich: Urs Graf, 2009), pp. 161–71, at p. 164, and K. Heidecker, ‘Urkunden schreiben im alemannischen Umfeld des Klosters St. Gallen’, ibid., pp. 183–91, at pp. 187–8, for Rhaetian and Alemannic practice, respectively. 38 For updated figures and further discussion of this formula, see W. Davies, Windows on Justice in Northern Iberia, 800–1000 (Abingdon: Routledge, 2016), pp. 97–106. 39 S368 (956) – a cartulary copy; S214 (963).



adque conuenit nunlis quoque ientis imperio neque suadentis articulo set propria nobis accesit uoluntas’.40 The priest Durabiles, in Villobera, uses the forms nullus coque (e)gentis and introduces a local phrase into the formula – prumtu animo (ready in spirit); this phrase can undoubtedly be found in other formulas from time to time but is rare in this formula in the tenth century, although there is another example from Villobera, written a couple of years before, by a different scribe;41 so, Durabiles has: ‘placuit nouis prumtus animus adque conuenit nullus coque gentis imperio nec suadentis artigulo set propria nouis azesit boluntas’; ‘placuit nouis adque concesimus prumtu animo et bolumtate nullus coque egentis imperio nec suadentis articulo set propria nouis azessit bolumtas’.42 And someone, unnamed, in the Flaínez household used the phrase excluding drunkenness from the possible pressures limiting voluntary capacity: I did this of my own free will, without pressure, not being drunk, and so on. It occurs both in a record of sale to Flaino’s brother in 973 and in a record of a gift to Flaino in 987, and thereafter several times for that particular household in the early eleventh century and occasionally elsewhere: ‘placuit nobis aduenit uolutas ut nec per metum nec per eberiadate set expontana mea uolutate’; ‘sanus et clara mente, bono amimo et expontania nostra uoluntatem nec per medum nec per ebriadatem set probria nobis et fuit uoluntas’.43 While I leave full analysis to expert palaeographers, there are also aspects of the materiality of the texts on single sheets which merit some comment. There are from time to time physical characteristics that point to the individuality of the writers, or the practice of a household, some of whom are the local and aristocratic household priests cited above. Abbreviations and contractions, for example, tend towards consistency within a single text although overall there are many inconsistencies in practice; however, in two charters local priest Iohannes used a contraction mark over omo (i.e. homo) which indicates that he was thinking of the spelling ommo; this is not unique but the combination of the contraction and his distinctive contraction mark (rather like a figure ‘2’ with an extended horizontal) is unusual.44 Again, while most chrismons are extremely similar some scribes used distinctive chrismons at the start of their charters and beside the columns of witness names: local priest Durabiles again and another local Vermudo, writing in the Cea valley 40 OD6 (947); S205 (962). I am very grateful to Roger Wright for pointing out to me that the spelling nunlis is an understandable error: given that the letter n at the end of a syllable had become silent in several words, the idea arose that sometimes n ought to be added in apparently comparable positions, as in Herculens for the name Hercules. 41 Li116 (937), a record of a gift to Abellar, noted by Vellite. 42 Li132 (939); Li135 (939). 43 OD17; OD27. Cf. OD74 (1007) – ‘placuit nobis aduenit uolutas karo animo et spotania mea uolutate ut nec per metum nec per ebriadate set expotiia nostra uolutate ut’ – and OD85 (1010), OD104 (1016), OD146 (1022); Liii742 (1016), Liv1090 (1053), Liv1323 (1107). 44 Li234, Li237 (ACL 98, ACL 97). Other contraction marks for this spelling do sometimes occur, but there are not many; for example, Fredmundo, of OD6 (AHDL Otero 3), uses a short straight suprascript line.



in mid-century, as well as ‘aristocratic’ Braolio.45 It is also common, though by no means invariable, for a scribe’s name to be followed by a signum, often florid and sometimes very large; some of these (like the ruche) occur in texts from other parts of Europe in the early middle ages and some do not.46 Though these can vary too, some writers have personal signs and also signatures that seem to proclaim their individuality; local Amorino, from Ujo in the north, is a good case of the coincidence of hand, signature and signum.47


(a) Figure 2.1a, 2.1b Amorino: ACL 104 and ACL 110

Thinking about the physical characteristics of a charter raises plenty of questions too. A few signatures and signa from apparently identifiable individuals actually look very different, although this does not happen often. Take the cases of Braolio, from the Flaínez household, and of local Durabiles, both of whom have some indications of individual identity: while the signatures of Braolio in several charters could be read as the same, those of Durabiles do not look the same at all. Three of Braolio’s four signa look very different from each other, although the hands of Braolio’s charters are similar, especially those of OD31 and OD43, and they all have a distinctive spelling habit.48 For Braolio, the signum of OD43 is bold and clear, with three very clear letter Bs within it; those of OD31 and OD34 are similar to each other but different from OD43; that of OD26 is quite different. Of course, Braolio may simply have used different signa at different stages of his 45 Durabiles: Li132, Li135 (ACL 818A & 818B); Vermudo: Li197, Li221 (ACL 804 and ACL 805); Braolio: OD43 and OD31 (AHDL Otero 16v and Otero 159) have identical, unusual chrismons but for one short stroke; cf. OD34 (ACL Fondo Raimondo Rodríguez 4), also a Braolio record, which is almost the same. 46 Other witnesses and confirmers, especially priests, can have elaborate signa after their names too and some charter witness lists have many elaborate signa following witness names; while the latter tend to be associated with records of high-status transactions, overall the inclusion of signa is inconsistent. See Tock, Scribes, souscripteurs et témoins, pp. 145–90, for a thorough discussion of the range of graphic signs in use in France in the early middle ages, several of which are also found in this Iberian corpus. 47 Amorino: Lii310, Lii357 (ACL 104 and ACL 110) – there are slight differences in the hand and signature, but they nevertheless appear to be the same. 48 Braolio: OD26 (AHDL Otero 16rB), OD31 (AHDL Otero 159), OD34 (ACL Fondo Raimondo Rodríguez 4), OD43 (AHDL Otero 16v). OD26 is particularly difficult to read because it is squashed around the witness list of another charter (OD21, Otero 16rA); I am especially grateful to David Ganz for his opinion of these hands.



life: OD26 is the earliest (986) and is a standard form; OD31 and OD34 are close in time (991, 993) and could be read as a kind of three times B; OD43 is four years later and has the mature sign with the clear Bs. Alternatively, did someone (i.e. Braolio) take charge of production of the text, but someone else do some of the writing, as we know happened in royal households and great monasteries but we might not have expected in a rural context? It certainly looks as if Braolio was the person responsible for the charters but other scribes could have written at least some of them.49 Since Braolio wrote for a count, Flaino Muñoz, it is worth considering if there was a group of scribes working at the comital centre of Valdoré. Indeed, such a circumstance cannot help but remind us of the well documented and well published magister Andreas, assisted by junior scribes, at the comital centre of Rankweil in Rhaetia (Switzerland) in the early ninth century.50 Both Durabiles charters (Li132, Li135) are written on a single large sheet (ACL 818A and 818B). The name Durabiles is written in cursive script in the first record and spelled Durabiles, followed by two signa; the name in capitals (spelled Durables), follows the second charter, together with another signum, which shares an element with the first of the first record but is overall very different. This pair of charters is also striking because the two sale records, dated three weeks apart in 939, look as if they were started by one person and completed by someone else, in a more clumsy hand. The first seven lines of Li132, the first charter, are well spaced out; the rest of Li132 and also Li135 are more cramped and use a different pen, and are in the same hand as each other. The two signa after the scribal name on the first record perhaps indicate the two scribes. This parchment raises all kinds of question. Was Durabiles the redactor, and in control of the whole operation, or one of the writers? Was he (or one of the parties involved in construction of the records) truly local or an agent of the distant monastery? As pointed out above, the name Durabiles does not feature in other Abellar charters and it is striking that the local prumtu animo phrase occurs in both hands, and that the chrismons of both charters are distinctive and unlike those of Abellar charters. Now, the monastery of Abellar had had interests in Villobera since a major gift in 927 and it was involved in many transactions there, largely buying additional lands. It must be likely that Abellar representatives were installed to look after the properties, although a formal decanía does not seem to have been established.51 49 There are plenty of parallels beyond Iberia for one scribe’s name attached to a charter written by another; see Fichtenau, Das Urkundenwesen, p. 43; McKitterick, Carolingians and the Written Word, pp. 96–7; Atsma and Vezin, ‘Autour des actes privés’, pp. 51–6; Atsma and Vezin, ‘Les responsables’; Erhart and Kleindinst, Urkundenlandschaft Rätien, p. 68; Tock, Scribes, souscripteurs et témoins, pp. 111–13; K. Heidecker, ‘Les actes privés de la période carolingienne dans les archives de SaintGall’, Annuaire de l’école pratique des hautes études, Section des sciences historiques et philologiques, 139 (2008), pp. 372–7 (‘aussi que derrière deux noms pouvait se trouver une seule main’). 50 Cf. Erhart and Kleindinst, Urkundenlandschaft Rätien, pp. 59–64; Erhart, ‘Erratische Blöcke am Alpennordrand?’, pp. 167–8. 51 See M. J. Carbajo Serrano, El monasterio de los santos Cosme y Damián de Abellar. Monacato y sociedad en la epoca astur-leonesa (León: Centro de Estudios e Investigación ‘San Isidoro’



Figure 2.2 Durabiles: ACL 818A

That being so, the first writer of Li132 could originally have been Abellar trained, though the distinct local associations of both records suggest local acculturation. It could also be significant that Durabiles ‘noted’ (notavit) both documents rather than ‘wrote’ (scripsit) them. If the first writer of Li132 were also the redactor of both texts, and in charge of their production, then it could make sense to suggest that Durabiles was the redactor and started off the writing of the first record but that that record and the second were finished off by someone else, whom he had perhaps trained to make the signa and the chrismons. That would suggest the presence of a ‘chief scribe’ or ‘master’ and others (other clergy and/or boys being trained), as is in any case suggested by the names of the scribes of other Villobera charters.52 It remains odd that Durabiles does not feature in other Villobera, or other Abellar, charters, however, given the numbers of both; and it is odd that the name Durabiles is written by the same writer in two different ways, the version in [CSIC-CECEL], Archivo Histórico Diocesano, 1988), pp. 79, 158–60, 197–202; and Martínez Sopena and Carbajo Serrano, ‘Notas sobre la colonización de Tierra de Campos’. 52 Cf. the two priests who identified the priest Bera as their magister, S279 (975). For the magister Andreas and his trainees Valerius and Vigilius, who wrote charters in Rhaetia in his name, see Erhart and Kleindinst, Urkundenlandschaft Rätien, p. 68, and the especially clear charters, nos 13, 14, 15, 16, 17, 20; Erhart, ‘Der Rotulus des Valerius’, especially pp. 46–9.



capitals being particularly awkward.53 One can think of other scenarios to explain the production of Li132 and Li135 but on balance, given the similarities of formulation between both records and of the chrismons, a local group is perhaps the best explanation – in this case a local group serving a monastic interest rather than that of a count.

Tools of power? There are three useful points to make. Firstly, the fact that local priests provided writing services for local people – peasants and others, richer and poorer – in itself says something about the dynamic of power relations in local communities. Even allowing that they must have been paid, it gave the priests power: to write or not, to massage the contents or to join the negotiation between the parties, as they must have done in sales and in court-cases – local priests often played a part in court, on inquest panels, as expert witnesses, as advocates, as sureties, as judges and as saiones (legal officers).54 The fact that some local priests (not necessarily all) had a relationship with the local community is emphasized by the, admittedly few, rural concilia of which we have knowledge in the tenth century.55 Some of the local priests who recorded lay transactions, like Vermudo of the 990s, seem to have been associated with a rural concilium: in this latter case witnesses of Vermudo’s records were identified as members of the concilium of Sancti Marcelli, a concilium clearly defined by the name of a saint.56 The implication is that Sancti Marcelli was Vermudo’s church and that the local community that attended the church (or some of them) formed its concilium; Vermudo made records for them, witnessed by members of the council. Secondly, it is notable that when aristocrats, lay or ecclesiastical, were buying properties from peasants or minor aristocrats, although the transactions were often recorded by the household scribe, sometimes it was the local priest who made the record. Munio Flaínez did not always use his household priest, Fredemundo; sometimes he used a local, as he used the priest Principio in 960, and

53 Again, I am especially grateful to David Ganz for his opinion. 54 E.g. inquest: Li128 (938); expert witnesses: Liii556 (993); advocates: Li192 (946); surety: OD4 (946); judge: Li191 (946); saio: Sob109 (990s). 55 In these contexts the concilium was the assembled group relating to a given place, a collectivity; see P. Martínez Sopena, La tierra de campos occidental. Poblamiento, poder y comunidad del siglo X al XIII (Valladolid: Institución Cultural Simancas, 1985), pp. 289–90 (where he points out that some concilia still owned churches in the eleventh and twelfth centuries), pp. 506–13; Davies, Acts of Giving, pp. 201–7, and references there cited. The word concilium also signified many other things in the tenth century. 56 Liii572 (996); Vermudo is not called priest here, but he is in Liii583 (999), also a Montefrio transaction, as Liii572 was. Cf. Lii396 (966), sale of a vineyard in Valdesogo, recorded by the priest Filanum and witnessed by members of the concilium of Sancte Eufemie.



there are several others.57 The same is true of Assur and his scribe Ferronius: sometimes he used the local Piniolo.58 Records of the powerful priest Melic’s many purchases nearly all have different scribes named, who were presumably local. The powerful priest Vincimalo used the local Fredenando for his Villagoya purchases (all recorded on one sheet) but other scribes for other purchases. Local Amorino wrote for the aristocrat Taurello. And local Vicente wrote for purchases by the monastery of Sahagún in Cofiñal. This is another aspect of the power of the local priest, perhaps negotiating, certainly an intermediary, in part controlling the interaction between peasant vendor and aristocratic outsider. Thirdly, household priests who wrote for aristocrats suggest a different kind of relationship. Where an aristocrat used a household scribe then he, the aristocrat, was making a clear statement about his power – expressing a capacity to move across a wide area with his personal entourage and thereby expressing dominance. It reinforced his control of the transactions. The priest Domengo, writing for Fruela Vimáraz, recorded two ‘gifts’ to Fruela in lieu of debts on the same parchment on the same day in the same place in 995.59 This text invokes a picture of a dominant aristocrat calling in debts, a day of summoning peasant debtors to his presence, receiving compensatory grants and getting a record made. We need to think of both aspects of the aristocratic interaction with the local: the dominant, using the household scribe, and the more hesitant, using local priests, seeking a foothold in a functioning community. Both aspects are a part of the changing dynamic of the latter half of the tenth century. So, there seem to have been plenty of local priests, and others, who wrote for local people of all kinds. There were also priests, and others, who wrote for aristocratic households and who are best considered as members of those households. One can find traces of individual practice in their use of Latin, their treatment of standard formulas, their abbreviation and contraction marks, their use of graphic signs and their hands. The search for these indicators certainly throws up some problems, but it looks as if there were occasions, as elsewhere in western Europe, on which some scribes wrote under the names of others; that means that there were some groups (or at least pairs) of scribes in rural localities as well as individuals. In the two cases considered above, which are unusual, it is worth suggesting that one such group served a comital centre and the other a locally embedded monastic outpost. We need, then, to think of a variety of ways of providing writing services – local scribes, household priests, monastic outposts, comital centres, as well as the monasteries we might be inclined to expect – though there remain many questions. As for charter writing as a tool of power, this material clearly 57 S184. Cf. the choice of scribes by private actors in tenth-century Burgundy, B.-M. Tock, ‘Les actes entre particuliers en Bourgogne méridionale (IX–XI siècles)’, in Erhart, Heidecker, Zeller (eds), Die Privaturkunden der Karolingerzeit, pp. 121–34, at pp. 125–7. 58 S182 (960). 59 OD40, OD41 (AHDL Otero 31 A and B).



shows the local scribe influencing local relationships and mediating between locals and outsiders, with the aristocratic household scribe a more overt expression of aristocratic power. That there may have been groups of comital scribes suggests a degree of organization of comital affairs that would have reinforced, and indeed extended, the power of the count.60

60 I am extremely grateful to Roger Wright, Bernhard Zeller and the conference organizers for their comments on drafts of this paper; to David Ganz for long conversations on the palaeography; and to the archivists of the cathedral and the diocesan archives in León for their very considerable assistance.



This is about the nature and activities of local priests in northern Iberia up to the year 1000 – northern Iberia excluding Catalonia (because of its Frankish background and tendency to look eastwards; and the north of the peninsula because the centre and south were Muslim-controlled).1 The political and religious context in which local priests operated in this period was one of very limited governmental and administrative activity. There were two kingdoms: the larger Asturias-León in the north west and a very small Navarre in centre north. There were no archbishops; bishops appear now and then (particularly the bishop of León, whose activities are well documented); but there was little of the apparatus of ecclesiastical administration that we come to expect in the central middle ages. However, there were plenty of monasteries, large and small, there were some lay religious households and there appear to have been a good number of local priests.2 Our information comes overwhelmingly from charters, of which there are about 2,750 for the area covered, about two thirds surviving in cartulary copies and the remaining third on single sheets, most of which are originals. The availability of charters determines the date focus: there is a tiny number from the eighth century, a couple of hundred from the ninth century and the rest (that is over 90%) come from the tenth century. There are many more eleventh-century texts but this survey stops at the year 1000 because different kinds of behaviour, relationships and institutions developed in the eleventh century. There are some other kinds of text available too, but very few have material relevant to local priests: the few 1 First published in S. Patzold and A. C. van Rhijn (eds), Men in the Middle: Local Priests in Early Medieval Europe (Berlin: De Gruyter, 2016), pp. 125–44. 2 For general surveys of political history, see for example in English, H. Kennedy, Muslim Spain and Portugal: A Political History of al-Andalus (London: Routledge, 1996); R. Collins, ‘The Spanish Kingdoms’, in T. Reuter (ed.), The New Cambridge Medieval History III c.900–c.1024 (Cambridge: Cambridge University Press, 1999), pp. 670–91, 716–17; in Spanish: A. Isla Frez, La alta edad media. Siglos VIII–XI (Madrid: Editorial Síntesis, 2002). For ecclesiastical structures, A. Isla Frez, La sociedad gallega en la alta edad media (Madrid: CSIC, 1992), pp. 71–103, is focussed on Galicia but has wider significance; and see my longer comments in W. Davies, Acts of Giving: Individual, Community, and Church in Tenth-Century Christian Spain (Oxford: Oxford University Press, 2007), pp. 44–50.



chronicles that cover the ninth and tenth centuries are primarily concerned with the deeds of kings;3 there is very little ninth- or tenth-century hagiography, and there are very few tracts or letters of northern Iberian origin;4 but there is a large corpus of liturgical material, on which I will comment in due course.5 Although there were not so many texts other than charters composed in the tenth century, this was nevertheless a period when the scriptoria of the great monasteries of northern Iberia were energetically copying biblical, patristic, theological and legal texts of earlier date, especially of the fourth to seventh centuries. There is a background of Visigothic texts from before 700 which are potentially relevant to the subject of local priests in the ninth and tenth centuries. The proceedings of Visigothic church councils indicate concern about the quality and training of local priests at that time: the second council of Braga (572) made provision for inspection of the clergy; and the fourth council of Toledo (633) provided for a libellum to be given by the bishop to local clergy so that they knew how to baptize and celebrate the office (although, while there are tenth-century copies of the corpus of conciliar provisions, I am not aware of any ninth- or tenth-century citations of these clauses).6 It is also worth remembering that Martin of Braga wrote a tract called ‘De Correctione Rusticorum’ in the late sixth century, which was copied at least once in the later tenth century (in a monastic context) and so was clearly known in some parts.7

3 The principal chronicle, with notes and critical study, is published in J. Pérez de Urbel, Sampiro. Su cronica y la monarquía leonesa en el siglo X (Madrid: CSIC, 1952), at pp. 273–346; it is of very late tenth- or early eleventh-century date. 4 The Hispanic Passionary is a compilation of the passiones of Roman, Visigothic and Mozarab martyrs in Spain, a compilation which began in the seventh century and reached its definitive form in the eleventh; the few items which could be of tenth-century origin (most are much earlier) probably derive from Mozarabic contexts, in southern Iberia; for detailed description of the contents, and a survey of arguments for dating, see M. A. Andrés Sanz, C. Codoñer, S. Iranzo Abellán, J. C. Martín, D. Paniagua, La Hispania visigótica y mozarabe. Dos épocas en su literatura (Salamanca: Universidad de Salamanca, 2010), pp. 336–59. There are two short tenth-century Vitae: the exceptionally brief so-called Life of Abbot Salvo of Albelda, of c. 976, is published by C. J. Bishko, ‘Salvus of Albelda and frontier monasticism in tenth-century Navarre’, Speculum, 23 (1948), pp. 559–90, at pp. 575–6; see also Andrés Sanz et al., La Hispania, p. 364. For the Vita Froilanis, see the following, n. 34. See also Andrés Sanz et al., La Hispania, pp. 269–96, for the corpus of ninth-century Córdoban letters, from southern Iberia. 5 See what follows for liturgy and M. C. Díaz y Díaz, Index Scriptorum Latinorum Medii Aevi Hispanorum (Madrid: CSIC, 1959), pp. 141–64, for other works, comments and notes written in the tenth century. 6 Concilios visigóticos e hispano-romanos, ed. J. Vives, with T. Marín Martínez, G. Martínez Díez (Barcelona: CSIC, 1963): II Braga cc. 1, 2, 3, at pp. 81–2, IV Toledo ch. 26, at p. 202, ‘Quando presbyteres in parrochiis ordinantur, libellum officiale a sacerdote suo accipiant, ut ad ecclesias sibi deputatas instructi succedant, ne per ignorantiam etiam in ipsis divinis sacramentis offendant, ita ut quando ad letanias vel ad concilium venerint, rationem episcopo suo reddant qualiter susceptum officium celebrant, vel babtizant’. 7 Paris MS BN lat. n.a. 235, fos 93v–108v, a Silos manuscript; A. Millares Carlo, Corpus de códices visigóticos, prep. M. C. Díaz y Díaz et al., 2 vols (Las Palmas de Gran Canaria: Universidad de



The limited portfolio of texts available means that all kinds of important question about local priests in the ninth and tenth centuries cannot be answered. What can be discussed, however, is as follows: what was a local priest and how did he get to be a priest? What were his resources and how, if at all, was he controlled? How far can we see a local priest interacting with the local community? What did he do as priest? What else, if anything, did he do? I shall deal with these questions in that order.

What was a local priest? In identifying the local priest I am interested in priests who were not members of an established monastery, nor an episcopal household or cathedral chapter, nor a royal household. That therefore means priests of the countryside, associated with a named church, operating alone or – more probably – as a member of a small group of priests or minor clergy. This was not the ‘parish priest’ and there was as yet none of the apparatus of parish territory or tithe: the implementation of a parish system did not begin until the changes of the later eleventh and twelfth centuries in northern Iberia.8 There are over 400 ninth- and tenth-century individuals in charter texts who could well be local priests; in some cases it is easy to demonstrate that they were local; in many cases it is not, because there is insufficient information to tie them to a local context. Meaningful statistics are therefore impossible and some of what follows will inevitably remain arguable. Despite that, it is perfectly clear that there were local priests and that there were different kinds of local priest: the single priest (or small group), serving a church and local community; priests who were members of lay religious households, where the leader of the community was often a lay person – most of these were clearly rural establishments, operating on a small scale, concerned with their own business and perhaps not providing any kind of service for the local community; and there were also priests in aristocratic households, again rural, certainly dealing with local business and local networks, but beyond that probably again not providing any service for the local community. The first and second of these groups are in practice difficult to differentiate from very small monasteries, that is monasteries with no more than five or six members. In fact the words monasterium and ecclesia can be used interchangeably in the same text of the same establishment and it looks as if contemporaries saw no great distinction – these were all small groups of committed religious persons,

Educación a Distancia, Centro Asociado de Las Palmas de Gran Canaria, Gobierno de Canarias, 1999), no. 253. 8 See J. A. García de Cortázar, La sociedad rural en la España medieval (Madrid: Siglo XXI, 1988), pp. 90–5; J. Orlandis, ‘La elección de sepultura en la España medieval’, reprinted in J. Orlandis, La Iglesia en la España visigótica y medieval (Pamplona: Universidad de Navarra, 1976), pp. 257–306, at pp. 277–84; J. Orlandis, ‘Reforma eclesiástica en los siglos XI y XII’, in Orlandis, La Iglesia, pp. 307–48. Cf. Isla Frez, La alta edad media, pp. 248–51. On tithe see further in what follows.



working, and often living, together. Small monasteries, and other communities, of this kind should be differentiated from the medium-sized and great monasteries of northern Iberia – communities with 30 or 60 or 200 members, living a regulated day, with formalized commitment, following a known Rule. In what follows, when I use the word ‘monastery’ I am referring to the latter; when I am referring to the former I will use the word ‘church’ or ‘small group’.9 That some of the 400 individuals were local priests is demonstrable for the following kinds of reason: their personal property is mentioned in the boundaries of another person’s small plot, like that of the priest Ihoannes, which lay on one side of the plot sold (to the great monastery of Cardeña in Castile) by a man called Donno in this description of 964: ‘a plot beside the plot of Gomiz Bellaza, and on another side [that of] the priest Ihoannes, on the third side the plot of Tellu Feles, and on the fourth side the hill of the Cardeña monks’.10 Priests also bought, sold and exchanged small plots in a single locality, where they may also have acted as scribe, like the priest Stephano in Marialba to the south of León, recording exchanges between lay parties in 926 and 928 and himself exchanging a plot in Marialba in 926.11 Or priests were identified as priest of a named place, as in the following witness list of a sale to the important monastery of Eslonza (east of León) in 946; part of this long list (from an original charter) is organized by location and some of the locations included priests: ‘De Legina Maruan presbiter testis; De Uilla de Auriolo [four lay names], Recemondo presbiter testis . . . De Reuellinos [four lay names], Raysendo presbiter testis . . . Uincemalo presbiter’.12 Or they could be involved in small-scale joint transactions with peasants, like the priest Daniel who, along with a local family, sold a mill to the monastery of Lorvão in central Portugal in 977; the priest had a two thirds share, the others one third.13 Priests like these are ‘local’ because they operated within a restricted locality, had limited personal property and worked with their peasant neighbours. We need to attempt some estimate of the scale of the sphere in which such priests operated, although there is no kind of contemporary survey and it is impossible to make any generalization. In the Spanish historiography the dominant model is drawn from the so-called ‘Parochiale Suevum’ (list of Sueve parishes), a list of thirteen bishoprics and dependent ecclesiastical territories from north-west Iberia, which appears ultimately to have been of late sixth-century origin. The list exists in several later versions, over which there has been much

9 See further Davies, Acts of Giving, pp. 46–8. 10 ‘agro iuxta agro de Gomiz Bellaza, et de alia parte Ihoannes presbiter, et de tercia parte agro de Tellu Feles, et de IIII pars monte de fratres de Caradigna’, C119. 11 Li69, Li78; Li70. 12 E21 (the second priest of Revellinos could well be the powerful Vincemalo, who had interests in many different locations; see following, n. 38); cf. Li147 (941), T41 (932), SJP14 (928), which has sixteen localized priests. 13 Lor76.



debate, although the text established by Pierre David is central to most discussions.14 The reconstructed ecclesiastical territories of the original list vary considerably in size but are within the range 300–1600 sq km. The detailed work that has been done on the later diocese of Santiago de Compostela demonstrates that many names of ecclesiastical territories have changed since the sixth century, sometimes several times, and suggests that by the late ninth century there were thirty church territories dependent on the bishopric of Santiago, each of whose size was of the order of 200–300 sq km, each with a principal church but often with other churches within the territory.15 That pattern – of a principal church with other churches within its territory – certainly occurs in other regions in the ninth and tenth centuries, as recently demonstrated for the presura of Tobillas in Álava.16 By the twelfth and thirteenth centuries, when the territorialization of the parish was much clearer (a process that gathered momentum from the later eleventh century), the size of many ecclesiastical territories was much reduced, to as little as 8 sq km in some cases.17 Of course, we do not need to think of the territory of the local priest as necessarily consisting of a single solid block of land and for many areas we probably ought to have in mind a network of interests and perhaps a series of links with scattered communities. But we still need to have some idea about how far a local priest’s interests might reach. The number of churches mentioned on the meseta (the central plateau of Spain) in tenth-century texts might suggest rather smaller spheres in the tenth century than those of either the sixth- or the ninth-century Galician model.18 The core of the Tobillas presura was about 4 km across, stretching north of Tobillas to Quintanilla and north east to Basabe.19 To this one might add that Rabal in southern Galicia was referred to as a concilium (unit of civil association, often the basis for later parishes) in the early eleventh century;20 it has plausibly been reconstructed to have been 3–4 km across and 5.2 sq km, rather a small area (compare 14 P. David, Études historiques sur la Galice et le Portugal du VIe au XIIe siècle (Lisbon/Paris: Livraria Portugália, 1947), pp. 1–82, at pp. 30–44. 15 F. López Alsina, La ciudad de Santiago de Compostela en la alta edad media (Santiago de Compostela: Ayuntamiento de Santiago de Compostela, Centro de Estudios Jacobeos, Museo Nacional de las Peregrinaciones, 1988), pp. 155–70, 309–10. 16 J. J. Larrea Conde, ‘Construir iglesias, construir territorio: las dos fases altomedievales de San Román de Tobillas (Álava)’, in J. López Quiroga, A. M. Martínez Tejera, J. Morín de Pablos (eds), Monasteria et territoria. Elites, edilicia y territorio en el Mediterráneo medieval (siglos V–XI), BAR International Series S1720 (Oxford, 2007), pp. 321–36, at pp. 322–3, 332–3. 17 García de Cortázar, La sociedad rural, pp. 91–2. 18 Cf. the process of fission in tenth- and eleventh-century southern Castile, which produced ‘the building blocks of the new parish system’, in which one village normally corresponds to one parish; J. Escalona, ‘Mapping scale change: Hierarchization and fission in Castilian rural communities during the tenth and eleventh centuries’, in W. Davies, G. Halsall, A. Reynolds (eds), People and Space in the Middle Ages, 300–1300 (Turnhout: Brepols, 2006), pp. 143–66, at pp. 149–59. 19 I am extremely grateful to Juan José Larrea for providing me with maps of the extent of both the presura and its core. 20 For concilia, see further in what follows.



the comparable unit of the Breton plebs in the ninth century – upwards of 6 km across).21 Farther south in Galicia, a group of villas round Verín, each one documented in tenth-century texts, corresponds precisely with present-day parishes; these parishes range between 2 km and 7 km across.22 For the moment, then, we might think of a range of 2 km to 20 km across for the geographical scale of a local priest’s activity in the ninth and tenth centuries, a block or network including anything from 5 sq km to 400 sq km.23 These are clearly inadequate calculations but they have the virtue of suggesting that there was considerable variation in the range of the territory a local priest might serve, as indeed must have been the case.

How did a person get to be a priest? There are two aspects to this question: how did a priest acquire the incumbency of a particular church; and how did he reach the order of priesthood? I will deal with each in turn. Often there is no way of knowing how a priest came to be the incumbent of a church, but where there is some indication of the process we can see that priests took up their churches because they were appointed by the proprietor of the church or because they themselves had a hereditary right to it or because they acquired a right by purchase, gift or exchange. There are some unambiguous cases of appointment by bishops, great monasteries and founders (who could be lay persons); so, for example, one Cendamiro and his brother made a written agreement with the bishop of Lugo to serve him from the church of Santa Columba, on the river Flamoso, in Galicia, in 974.24 In fact, this kind of arrangement is very rarely documented; the parchment on which the agreement is recorded is small and scrappy; it is an exceptional record. Much more common are the many cases of priests who held churches because they acquired them from a family member. This happened across the period, in all areas, although there was some tendency for it to decrease in the later tenth century. There were certainly some married priests, and priests and clerics with offspring;25 but recorded cases of family succession characteristically specify a church passing from priest uncle

21 M. del C. Pallares Méndez, Ilduara, una aristócrata del siglo X (La Coruña: Do Castro, 1998), pp. 38–47; M. C. Pallares and E. Portela, ‘El lugar de los campesinos. De repobladores a repoblados’, in A. Rodríguez (ed.), El lugar del campesino. En torno a la obra de Reyna Pastor (València/ Madrid: Universidad de València, CSIC, 2007), pp. 61–87, at p. 69. For Breton plebes, W. Davies, Small Worlds: The Village Community in Early Medieval Brittany (London: Duckworth, 1988), pp. 63–7. 22 Pallares and Portela, ‘El lugar’, p. 66. 23 Cf. the Catalan estimate of 300 parishes in the bishopric of La Seu de Urgell, of 7,000 sq km; G. Ripoll and I. Velázquez, ‘Origen y desarrollo de las parrochiae en la Hispania de la antigüedad tardía’, in Ph. Pergola and P. M. Barbini (eds), Alle origini della parrocchia rurale (iv–viii sec.) (Vatican City: Pontificio istituto di archeologia cristiana, 1999), pp. 101–65, at p. 156. 24 AHN Clero Lugo, carpeta 1325A no. 11. 25 For example, OvC12 (889), Sam61 (976), OD36 (993).



to priest nephew, or priest cousin to priest cousin.26 And there is good evidence of priests giving, selling and exchanging their churches with other priests and with lay parties: the priest Jimeno sold his church in Burgos to the priest Ariolfo in 914 and the priest Belisario sold his church of São Miguel in Portugal to the priest Gondesindo in 924; the priest Adulfo gave his church to a lay couple as the result of a court-case; and the abbot of Eslonza – in the extreme circumstance of loss of food and stock to Muslim raiders – sold a church to the aristocrat Oveco Téllez and his wife.27 It is explicit that wealthier priests could own several churches and, as in the case of Artemio in Asturias, that a rural church could have dependent churches.28 But how did a man become a priest in the first place? We must assume that most priests were ordained (although one might expect there to have been some uncanonical ordination and, sometimes, one has to wonder if some were ordained at all, especially since there are those who are referred to as quasi-presbiter and since the bishop seems to have been a very distant presence in some parts).29 I know of no references to the ordination of priests in ninth- or tenth-century texts, although the surviving eleventh-century liturgical ordines do include such a rite.30 (This liturgical material is Mozarabic, a pre-Roman rite, and is ultimately of Visigothic origin; doubtless it underwent some changes between sixth and seventh centuries and the eleventh century, but much of what we find in eleventh-century manuscripts must have been known in some parts of the north in the ninth and tenth centuries.)31 There are also occasional references to the fact that the order of priesthood was distinctive – like a reference to a man to whom churches on the outskirts of Coimbra were bequeathed for use during his lifetime if he were to be of the priestly grade (gradum sacerdotalem) but not if he had left it; and to he who reached the clerical order (clerigatum hordine) through the study of the fathers and the holy rule.32 But we know nothing of the process of ordination nor of how far a priest might travel to be ordained – there is little suggestion that bishops 26 For example, nephew: Sam153 (982), or C35 (941), in which the priest nephew had expected to inherit, although his uncle diverted the church to Cardeña; great nephew: AHN Clero Lugo, carpeta 1325A no. 4, dorse (987) (to return to the family after his death); cousin: S25 (921), Lor17 (957); multiple relationships: Ov24 (990), in which Priest Artemio inherited a church from his relation Priest Dulcidio, who had received it from his grandfather, Artemio then passing the church on to Priest Modesto, except for a dependent church which went to Artemio’s son and daughter. 27 C7, PMH DC 29, PMH DC 53 (943), E30 (988). 28 For example, Ov24 (990), Lor45 (946), T76 (990); see n. 26 above for Artemio. 29 Even in the seventh century the ordination of priests and deacons, and the consecration of churches, by priests alone had to be prohibited, as it was by the second council of Seville in 619; Ripoll and Velázquez, ‘Origen y desarrollo’, p. 149. Quasi-priests: Ast24 (923), S166 (959), PMH DC 93 (967), PMH DC 167 dorse (993), all of whom were scribes; S336 (987), a donor. None of these places is remote – S166 deals with property in León, beside the bishop’s gate. 30 Liber Ordinum Episcopal (Cod. Silos, Arch. Monástico, 4), ed. J. Janini (Silos: Abadía de Silos, 1991), pp. 88–97. There is a charter recording ordination of an abbot by a bishop: San Pedro 4 (896). 31 See further in what follows. 32 Lor17 (957), SamS-9 (995).



undertook regular visitation and performed necessary rites in the course of those visits.33 As for training: priests were unquestionably associated with literacy and all must have experienced some kind of training. We know very little about it, but we can deduce from variations in standard formulas that there could be an aural element in learning to write the charters that many local priests wrote. The tenth-century Vita of an early tenth-century bishop offers no clues, stressing only that Bishop Froilán of León was learned from infancy and from the age of eighteen wanted both eremitic solitude and the ‘office of preaching’, in order to teach others (like his companion Atilano, who became bishop of Zamora); people flocked to hear their words.34 On the other hand there are isolated but widespread references to magistri, and occasionally to scolae, in charters occurring from Castile to Portugal.35 It must be likely that some of those who ended up as local priests had been to monastic or episcopal schools, and that some, as earlier hagiographic models suggest, had gone to noted holy men for training, but – because some of the magistri appear in very rural locations (and one was married and one gave away a church) – it is reasonable to suggest that some local churches were training places;36 it is also reasonable to suggest that the adults in lay religious houses trained their children, and priest fathers and uncles (and why not mothers and aunts) trained their sons and nephews. If we were to think of many of them as members of a hereditary learned class, passing their skill, learning and experience from generation to generation – carrying on the tradition in some isolated locations – it might help us understand how there came to be so many who were sustaining what was in effect an ancient culture.

Resources The local priest’s principal resource was income from landed property. In this period this would have been produce (for no coin was minted until the late 33 Cf. even in the ninth-century Breton church, which was much more dominated by bishops, the request for ordination in and by a monastery, because of the dangers of travel to the bishop’s see; Davies, Small Worlds, p. 22. 34 ‘predicationis officium ad docendum alios’; J. C. Martín, ‘La Vita Froilanis episcopi Legionensis (BHL 3180) (s.x): introducción, edición crítica y particularidades lingüísticas’, in M. Goullet (ed.), Parva pro magnis munera. Études de littérature tardo-antique et médiévale offertes à François Dolbeau par ses élèves (Turnhout: Brepols, 2009), pp. 561–84, at pp. 579, 581. There is no reference to Froilán’s ordination in this text. For Atilano, see F. Luis Corral, ‘En busca de hombres santos: Atila, Ildefonso y el obispado de Zamora’, in I. Martín Viso (ed.), ¿Tiempos oscuros? Territorios y sociedad en el centro de la Península Ibérica (Madrid/Salamanca: Silex, Universidad de Salamanca, 2009), pp. 203–27. 35 Scola: Lii404 (967), the place where a charter recording a small-scale exchange between lay parties was written. Magistri: S279 (975), SM77 (= BG2) (929) – a fabrication of the mid-twelfth century, Lor45 (946), Ast77 (954), for example. 36 Seeking holy men for training: for example, the disciples attracted by the seventh-century Valerio of El Bierzo, M. C. Díaz y Díaz, Valerio del Bierzo. Su persona. Su obra (León: Centro de Estudios e Investigación San Isidoro, 2006), Replicatio 12, 23, at pp. 292–4, 302.



eleventh century) and it is quite likely that many priests would have spent a proportion of their time farming. Much of the property was family property; priests transacted with their fathers, sisters, mothers, brothers and cousins, and inherited from them too.37 Some property was also attached to churches, irrespective of the family associations of the incumbent. Priests also acquired land, partly by receiving gifts but particularly by buying property. Some priests acquired a considerable volume of property, although the individual plots were usually of small scale – an arable plot, an orchard, a garden; two priests of the meseta, Melic and Vincemalo, made very substantial acquisitions and clearly reached a status well beyond that of the purely local priest.38 Priests also feature as litigants, fighting hard to keep their land, water rights and churches, or to acquire more: they contested with monks, bishops, other priests and lay people.39 It is difficult to get a sense of the scale of a local priest’s property, and it is again impossible to generalize, but there are a few individual examples that are useful. When the priest Juliano endowed the church of Santa Juliana in 974 with what he had received from his father he gave arable plots, two meadows, a quarter of an orchard, two days and a night of mill rights, several vineyards, three farmhouses, barns with grain, three barrels, two beds and bedding, a chalice and patten, horses, oxen, cows, sheep and pigs. Twenty years later, when he made his last testament and gave the church of Santa Juliana to the great monastery of Sahagún, he had acquired more, but apparently not so much – another six days of mill rights, three days in yet another mill and six vineyards.40 When the priest Beato endowed the church of San Salvador in Galicia in 889, he gave barrels, jugs, dishes, a wine press, beds and bedclothes, as well as land, liturgical vessels and books.41 It is clear that local priests had goods – movables that were surplus to their subsistence needs. When they bought land, for the most part they used farm produce to pay for it. If we look at the list of purchases made by local priests in León charters of 939 to 985, the prices paid include: a horse and two measures of wheat, silver bullion, two measures of barley, a rabbit fur, six measures of grain, a cow and a quarter of wine, a black cow and grain and silver, a horse, more grain. Elsewhere the prices paid are very similar, although there were also lengths of cloth in Galicia, pigs and

37 There are dozens of examples of this: for example, fathers: Lii316 (959), Lii396 (966); mother: Lii428 (974) – with a deacon, T76 (990); sister: C84 (954), Liii569 (995); brothers: Lii278 (954), Sob57 (962); cousin: Cel354 (989). 38 Melic: e.g. S69 (938), S75 (940), S81 (942), S90 (943). Vincemalo: e.g. S71 (938), S83 (942), S86 (943), Li172 (943); cf. P. Martínez Sopena, La Tierra de Campos occidental (Valladolid: Institución Cultural Simancas de la Diputación Provincial de Valladolid, 1985), p. 427; F. Luis Corral, Villavicencio en la edad media. Propiedad y jurisdicción en los Valles del Cea y del Valderaduey (Valladolid: Diputación Provincial de Valladolid, 2003), pp. 107–13. 39 For example: S21 (920), C22 (932), Li144 (941), OD27 (987), PMH DC 163 (991), Liii578 (997). 40 S274; S350 (996). Of course, it is possible that he had acquired other property that he did not wish to include in the bequest, but the items still remain of small scale. 41 Cel36.



goats and cider in Asturias, and a load of wood in Castile. Local priests really did have surplus goods. What we do not find, except very occasionally, are references to tithes. It would be conventional to say that evidence of institutionalized tithe collection did not start until the beginnings of parish formation in the mid- to late eleventh century.42 I see no reason to dissent from this conventional view, for nothing suggests that tithes were a significant source of income in the ninth or tenth century. There are a few references to tithe, however, in tenth-century charters. Most of these cases are very questionable, since they occur in late or suspect copies which could well have been interpolated. Even if we were to allow that all were credible, the number of them is tiny and most come from areas well to the east, in Navarre and Aragón (two from Leire, one from Albelda and one from San Juan de la Peña).43 However, one such reference is Portuguese, and occurs in a charter on a single sheet, recording a dispute in 991 between two local priests over control of the local church of Guilhabreu, about 15 km north of Porto.44 This is a credible reference to tithe, a contribution from which was part of the settlement made on the defeated priest;45 he went away with some of the church’s income, explicitly in respect of the services he had performed. This must mean that by the late tenth century the concept of tithe was familiar, at least in northern Portugal, which is hardly surprising given its Old Testament occurrences. However, overall, it remains the case that credible references to tithe are exceptionally rare in these texts and that tithe was not a significant source of income for local priests at this time; collecting it from those who might benefit from pastoral care does not appear to have been systematic or regular. There are no explicit references to the payment of fees either, but it must be possible that fees were paid for baptism and burial: the priest Cagito passed on to the monastery of Samos what he had acquired while holding the church of Santo Tomé de Cancelada and the priest Leodulfo handed over what he had got from the people; in both cases the contexts suggest that the acquisitions were primarily material goods.46 The best case that can be made that a request for burial was directed to a local priest, a Castilian case of 964, was however accompanied by a gift of land, presumably a de facto fee.47

42 See Martínez Sopena, La Tierra de Campos occidental, p. 296; García de Cortázar, La sociedad rural, p. 90. 43 Leire 6 (c. 918), Leire 7 (938), SJP17 (947); A28 (983), an agreement between the bishop of Nájera and the abbot of Albelda about tithes, is a credible record. 44 PMH DC 163. 45 ‘dederunt inde de ilo decimu inter pane et a beuere lx modios’. 46 Sam217 (973); ‘quod ex populo’, LaC 19 (914). 47 C117. Cf. the recent suggestion that some of the shorter donation charters of Becerro Gótico of Valpuesta are in effect notices of burial and hence that what was donated were in practice fees for burial, V pp. 122–6; I am grateful to Julio Escalona for this reference.



Control Control of local churches, and thereby of priests, was overwhelmingly proprietary, as one would expect given the strength of proprietary interests in churches. There were lay, monastic, episcopal and other clerical owners of churches, in many cases with families associated with the church. Succession to an incumbency was therefore often explicitly through the family. There are plenty of well recorded examples of this, including some from the eighth and ninth centuries: as indicated above, priests were succeeded by their cousins, nephews and probably occasionally by sons; and they could be given charge of churches by a brother.48 Families which owned churches might also appoint a priest who was not one of their number, and monasteries which had acquired ownership of local churches made appointments too. The latter was a marked and growing trend across the tenth century as some of the great monasteries acquired significant numbers of proprietary churches.49 For example, the priest Obeco gave his own church to the monastery of Cardeña some time before 957; Cardeña then put in two priests and a confessor to serve the church – a pattern repeated many times elsewhere.50 Monastic control of local churches is an increasing fact of the tenth century. Bishops can be seen controlling some churches too (just as some bishops took a particular interest in selected monasteries, like the bishop of Astorga with San Pedro de Montes and related houses).51 Centuries earlier, the Visigothic church had expected a third of the income of local churches to go to the bishop (another third for the clergy and a third for buildings); although competition for the third surfaced in the later eleventh century, especially in the context of claims that monasteries which owned local churches were exempt from such payments to a bishop, it is very rare to find any kind of reference to this in ninth- or tenth-century texts and it looks as if, for the most part, the practice had lapsed. However, the 947 San Juan de la Peña charter which mentions tithe also mentions an opus episcopatus, perhaps ‘bishop’s portion’ (which the bishop gave away); and a San Millán charter of 984 records a court-case between Bishop Muño of Álava and the abbot of Ocoizta/Acosta over payment of the third (tercias), a case which the abbot won.52 There is one other reference to obligations to a bishop which is quite clear: an early case relating to the church of Moreda, in Galicia. There are three texts about this church on a single parchment: in order of writing, the first is the priest Toresario’s confession, in court, before Count Fruela, that after his

48 49 50 51

See above. Cf. Orlandis, ‘Reforma eclesiástica’, pp. 322–5. C90. San Pedro 1 (892), 2 (892), 3 (895), 4 (896); Ast12 (915). See also I. Martín, ‘The memory of the “holy men” in Hispanic monasticism: The case of the Bierzo region’, Imago Temporis. Medium Aevum, 6 (2012), pp. 165–90, at pp. 179–80. 52 SM98 [and SM213 – a single document] (= BG221). For the issues surrounding payment to bishops, see Isla Frez, La sociedad gallega, pp. 96–100.



grandfather’s and mother’s deaths he had stopped making payments (obsequium, tributum) to the bishop of Braga and had put in a stranger to serve the church, so that it dropped out of the jurisdiction (iure) of Braga. The second is Toresario’s agreement on the same day in 861 to present himself at the church, before the legal officer, the saio, to restore its status as it had been in the past. The third is another agreement about the same church, with the same bishop, by another priest (Siseguto) to hold the church as the faithful servant of the bishop and pay him the obsequium and the due portion of the thirds (tertias).53 It looks as if at least one bishop expected, and took action to restore, control. It is equally clear that, where any such control existed, receiving the income from the church was the principal expression of control. One might note that the personal bequests of bishops identified churches in the same way as other properties, like those listed (geographically) by Sisnando, bishop of Santiago, when he endowed the monastery of Sobrado with lavish gifts in 966.54 His list of gifts includes a great mixture of properties, including churches and farms, in very haphazard order, and he often indicates how he had acquired them in the first place – some gifts, some purchases, some exchanges; this is the mixed portfolio of a very wealthy proprietor. Charters like this suggest that churches were as much a proprietary interest for bishops as they were for wealthy laymen. Episcopal control of churches clearly did occur, but when it did it was largely within the framework of the bishop’s proprietary interests, rather than within the framework of anything like diocesan, let alone metropolitan, organization. It was another aspect of the dominant pattern of lay and ecclesiastical proprietary control. The converse is equally true: many churches were quite beyond the control of bishops, just as Toresario’s clearly was for a time. As noted above, there is nothing to suggest regular or irregular visitation, although there are a few cases of episcopal consecration of churches.55 Nor are there anything like the tests of competence we would expect in the Carolingian world.56 The bishop of Astorga gave books to some churches and monasteries in the early tenth century, which might be taken to indicate episcopal concern for proper provision at local level.57 However, in

53 Floriano 74, 75, 73. There is something odd about the last charter (which Floriano dated as the first) – it was recorded last, is a copy, has gaps and has an unclear date; it is presumably a copy of the agreement by the bishop with the priest who took over from Toresario once the church had been recovered and would therefore be the third document in time. These three texts have at times been considered forgeries; see Isla Frez, La sociedad gallega, pp. 58–9. The most recent palaeographical assessment treats them as original (A. Castro, Colección diplomática altomedieval de Galicia I: documentación editada en escritura visigótica (662–1234) (La Coruña: Toxosoutos, 2011), no. 8, pp. 7–8), but see Castro’s further thoughts in W. Davies, Windows on Justice (Abingdon: Routledge, 2016), p. 85, n. 82. No. 73, which has the reference to tertias, is problematic. 54 Sob6. 55 Sam128 (849); PMH DC 5 (870); cf. Sam226 (947). 56 See S. Patzold and C. van Rhijn (eds), Men in the Middle: Local Priests in Early Medieval Europe (Berlin/Boston: De Gruyter, 2016). 57 See what follows below.



practice the behaviour and standards of local priests do not appear to have been a concern of bishops in this area at this time.

How far did the local priest interact with the local community? Although it was still some time before they became institutionalized, there clearly were some local communities with a sense of identity in northern Iberia: on occasion groups of peasants took collective action and we can see the local lay collectivity being referred to as a concilium in some areas, assembling – especially – for the performance of transactions. These concilia are regionally and temporarily limited: they occur in the second half of the tenth century and not before; they occur on the meseta, stretching into Castile and the Rioja, but apparently not in other regions in this period. There are many more references to such groups in the eleventh and twelfth centuries, and they became widespread, forming the concejos of the later middle ages, the smallest unit of civil association.58 To summarize a complex argument briefly, these collectivities began to institutionalize and acquire a recognized and consistent spatial expression round about the year 1000. Of course this does not mean that there were no groups before that but they are much harder to spot in the texts that survive; there were, however, some groups characterized as plebes in Galicia and on the Pardomino Mountain, north of León.59 As for the interaction of priests with these and other local groups, some tenth-century concilia (by no means all) were defined by the name of a church – San Marcello, San Juan, San Julián, San Andrés. Bearing in mind the fact that these concilia had lay members, this convention in naming them seems to imply that the group that constituted the concilium was defined as those people who went to or related to the named church. Characteristically, the members are named as witnesses of transactions, and sometimes – explicitly – a priest recorded the transactions, as the priest Fortunio recorded a sale between lay parties in the concilium of San Esteban in 965; and the priest Filano recorded a sale from a layman to his priest son in the concilium of Santa Eufemia in 966.60 These priest scribes are best understood as the incumbents of the churches, performing services of several kinds for the rural community. In such cases the priests had a very clear relationship with the local group, as a group, and not just with the individuals who participated in a particular transaction.

58 For a longer comment see Davies, Acts of Giving, pp. 193–207. 59 Sam128 (849), Floriano 74 and 75 (861), SantA39 (932), SantA43 (952). The term can have an ecclesiastical connotation (as one would expect from Italian analogues) but does not appear to do so in Li184 (944) and perhaps A28 (983) (as it does not in ninth-century Breton contexts – see W. Davies, ‘Priests and rural communities in East Brittany in the ninth century’, Études Celtiques, 20 (1983), pp. 177–97). 60 Lii391, Lii396 (arguably a sale by a laywoman not a layman).



There are also times when priests were named among the members of a local group taking collective action: the priest Abolmondar acted with twenty-two locals in a dispute with the abbot of San Torquato over water to power mills.61 There were five priests, along with thirty-six laymen, among the men of Los Ausines who were held to have given land to the count of Castile to build castles.62 There were also occasions when a priest appears to have acted as representative of the group, like the priest Nemorelli who led locals against the abbot of Valdevimbre in another dispute over water and mills in 941.63 The priests who had shares in local property, as mentioned above, were also locally embedded, as were those identified by place, together with a group of lay persons, in witness lists.64 And there were the priests who were associated with locals in other kinds of shared action, like Cardello who joined with locals to guarantee burial rites for a poor monk; or the sick priest Pedro whose affairs were sorted out by local boni homines.65 Local priests were clearly members of their local communities, joining in their actions and sometimes providing leadership.

What did he do as priest? Given this, it is not surprising that we do not know much about pastoral care in northern Iberia at this time, although there are some occasional hints. There is a Galician charter of 930 which mentions in passing that a local community had asked for someone to come and preach to them and there is the Castilian charter of 964 which records the layman’s request for a local priest to arrange burial; in the Portuguese dispute about the church of Guilhabreu, the defeated priest Sagulfu received a payment explicitly because he had ‘said the hours’, which must mean some regular performance of the liturgy.66 By the second half of the tenth century there were plenty of pro anima gifts, for self or family, although most of these were gifts to sizable monasteries rather than to local priests or local churches. That being so, it is theoretically conceivable that much burial was in monastic cemeteries, although this would seem to be highly unlikely, given the acquisitions of local priests from the people and the surviving physical existence of cemeteries which were not associated with monasteries, as at sites in the Basque country, western meseta and Portugal.67 It must be possible, then, that some local priests 61 C22 (932); the twenty-two came from five different settlements; it is clear that here and in other cases peasants from different settlements combined to take collective action. 62 C153 (972); as it stands some of the text of this cartulary copy is suspect, but the listing of names is likely to derive from a credible list of the period. 63 Li144. 64 Lor76, cf. Lor40 (927); E21, Li147, T41, SJP14. 65 V17 (939), Lor17 (957). 66 Sam170, C117; PMH DC 163: ‘per ipsas oras que disera In ipsa eglesia’. 67 See J. A. Quirós Castillo and B. Bengoetxea Rementeria, Arqueología (III) (Arqueología Postclásica) (Madrid: Universidad Nacional de Educación a Distancia, 2006), pp. 190–239; I. Martín Viso, ‘Enterramientos, memoria social y paisaje en la alta edad media: propuestas para un análisis



conducted burial rites for members of the local community, although we cannot assume that all did. The surviving liturgy includes ordines for burial, as also for baptism, although there is nothing about baptism in the charter texts. This liturgical material, although surviving in eleventh-century texts, nevertheless has some relevance for ninth- and tenth-century local situations. There are several eleventh-century liturgical manuscripts, including Silos MS 4 (from Albelda, dated to 1052) and Madrid MS A56 (from San Millán, undated but perhaps even of the late tenth century); both are called ‘Episcopal’ by their most recent editor, although they were clearly produced in the scriptoria of major monasteries.68 They are substantial books. There is also Silos MS 3, finished in 1039, of unknown provenance, which is much shorter. Of course, we cannot know how much, if any, of this material was known at local level in the tenth century but there are some relevant points to make. The short Liber Ordinum (Silos MS 3) was believed by Janini to have been based on a tenth-century model, a book he thought had been compiled to include the things indispensable for local use.69 It does not have, for example, rites for the ordination of a priest, but it does have baptism, visitation of the sick, burial, marriage, blessings for wine, the patten, the chalice and so on.70 Janini cites a reference to books in Sobrado charter 2 (one of the charters recording large gifts from Bishop Sisnando to the monastery of Sobrado, of 955); these included a psalter, a hymn book, prayer book and two ordines, unum episcopalem et alterum minorem (one episcopal and the other smaller); he proposed that the archetype of Silos MS 3 was just such a minor set of ordines, in other words a Liber Ordinum for local priests. He also argued that this book must have been very like the libellum mentioned in the proceedings of the fourth council of Toledo in 633, the essential handbook for priests.71 And, indeed, the ordination rite in the larger, episcopal, Liber Ordinum explicitly provides that the priest be given stole, chasuble and manuale at ordination.72

68 69 70 71 72

de las tumbas excavadas en roca en el centro-oeste de la península ibérica’, Zephyrus, 69 (2012), pp. 165–87; I. Martín Viso, ‘El espacio del más acá: las geografías funerarias entre la alta y la plena edad media’, in E. López Ojeda (ed.), De la tierra al cielo. Ubi sunt qui ante nos in hoc mundo fuere. XXIV Semana de Estudios medievales, Nájera 29 de julio al 2 de agosto de 2013 (Logroño: Instituto de Estudios Riojanos, 2014), pp. 75–140. Recent archaeological analysis suggests that some cemeteries were not initially associated with Christian centres of any kind and that some rural churches were constructed in the early middle ages beside already existing community cemeteries; see also I. Martín Viso, ‘Paisajes sagrados, paisajes eclesiásticos: de la necrópolis a la parroquia en el centro de la península ibérica’, Reti Medievali Rivista, 13:2 (2012), pp. 1–32. By contrast, one charter recording the consecration of a church in this period certainly does specify an associated burial place (PMH DC 5); see above, n. 55. Liber Ordinum Episcopal, ed. Janini. Liber Ordinum Sacerdotal (Cod. Silos, Arch. Monástico, 3), ed. J. Janini (Silos: Abadía de Silos, 1981), pp. 21–3. Ibid., sections I, II, III–VIII, XXVIII–XXXI, IX–X and XIII–XXVII respectively. For IV Toledo, see above, n. 6. Liber Ordinum Episcopal, ed. Janini, p. 96.



One could dismiss all this as wishful thinking, or as provisions made on high that had little implementation at local level, but we do know that many local churches had books included in their endowment in the tenth century, as well as liturgical vessels, and here our charters are precise, localizable and very useful. So, for example, a deacon gave his proprietary church near Oviedo to a priest in 889, together with land, stock, furniture, liturgical vessels and five books: Manual, Passionary, Psalter, Antiphonary and Ordinum.73 The founder of a church in Galicia, in 947, built it, endowed it with family property, acquired some relics and gave it houses, a silver chalice, cross and reliquary, two chasubles and a dalmatic, four altar cloths, a bell, a censor and four books: Manual, Psalter, Comicum (that is, readings from the Bible) and Ordinum.74 In Castile, a priest gave his two proprietary churches to the monastery of Cardeña and joined its community in 942, giving property and five books: Antiphonary, Manual of prayers, Ordinum, Psalter and Hymn book.75 And in 915 Bishop Genadio of Astorga made gifts to four churches and monasteries he had founded or restored, at the same time leaving a substantial library for common use; these endowments included Psalter, Canonicum, Antiphonary, Manual of prayers, Ordinum, Passionary and Book of Hours for the important house of San Pedro de Montes; Psalter, Antiphonary, Manual of prayers, Ordinum and Passionary for San Andrés de Montes; Psalter, Comicum, Antiphonary, Manual of prayers, Ordinum and Passionary for Santiago de Peñalba; and just a Psalter for Santo Tomás.76 We do not have to suppose that the various Manuals that are listed were indeed the Toledo successor (and many of them are explicitly manuals of prayers rather than service books), but most churches had a Liber Ordinum, a book specifying the appropriate ritual for particular occasions. It is clear that local churches did have books, and that there was something like a standard pack of four or five books, usually including Psalter and Antiphonary, a service book (Ordinum) and a prayer book.77 Liber Comicum only occurs twice in the examples here but it was also very commonly cited.78 Given all this material, and there is much more, it is reasonable to conclude that many, if not most, local priests were in a position to provide basic rites, prayers and biblical readings for their local communities. How far this translated into pastoral care is an open question. Notably, neither penitentials nor homiliaries are a common element among the books listed at these locations and very few 73 74 75 76 77

OvC12. Sam226. C37. Ast12. Cf. M. Zimmermann, Écrire et lire en Catalogne (IXe–XIIe siècle), 2 vols (Madrid: Casa Velázquez, 2003), vol. 1, p. 527, on provision of books for Catalan local churches; he identified a standard pack of four (Psalter, Antiphonary, Missal and Lectionary), noting (p. 531) the probable adoption of new books from Francia. The Missal in this Catalan list clearly performs the functions of Liber Ordinum and Prayer Book farther west, as the Lectionary does that of Liber Comicum. I am grateful to Jonathan Jarrett for this reference. 78 See Díaz y Díaz, Index, no. 640, pp. 156–7, for surviving manuscripts of Liber Comicum.



penitentials and homiliaries of this period survive from any location.79 There was also very little pious donation from the agricultural community at this date (though it increased later); small gifts were certainly made to local churches by peasants but when we know the intentions behind such gifts they are more often for reasons of practicality than piety – paying off debts, buying a service and so on.80 The provision of pastoral care may well have been uneven.

What else did local priests do for local communities? As already indicated, local priests acted as de facto notaries for peasant transactions: if a lay person sold a plot of arable or an orchard or water rights to another lay person, then it was usually a local priest who made the record, just as the priest Amorino recorded the sale of half a garden and fruit trees way up in the Cantabrian Mountains from one lay party to another, for a price paid in cider.81 This was an important aspect of the functioning of local communities, although it looks as if there were more transactions in some regions than in others and a greater need to have such records. The second kind of priestly activity to mention is an extension of the recording function. Local priests must often have assisted in negotiating price when they recorded sale transactions, because they prepared the record in advance, leaving a gap for the price which was filled in later. Given the fact that they controlled the record, and the addition of the price, once negotiated, one can imagine they did not stand idly by while the deal was done. There are examples of these additions from Asturias and Galicia as well as the meseta.82 Priests were also especially active in court-cases. Some were participants as a natural extension of their rural quasi-notarial functions and because so many hearings of court-cases were heard in local churches; they made records of fines, for example.83 While there are cases of priests acting as judges, expert witnesses, special witnesses of oath or ordeal, members of inquest panels and the innocent going to the ordeal, most of these priests seem to have come from monasteries or high-status households. But there were also priests who acted as the saiones who took people to court, administered oaths, and supervised some recording. Some of these seem to have been local, holding the office for a short term as a kind of community representative, like the priest saio in the witness list of a record of a Liébana dispute over a church between two laymen and a monastery, in 885; or

79 For a listing of surviving tenth-century and other books written in visigothic cursive and their contents, see Millares Carlo, Corpus de códices visigóticos. Only one tenth-century penitential from northern Iberia is listed, part of no. 117, pp. 88–9. 80 See further, Davies, Acts of Giving, pp. 113–30. 81 Lii310 (959). On this function, see further ch. 2 above, ‘Local priests and the writing of charters in northern Iberia in the tenth century’. 82 Li221 (950), Ov15 (962), S135 (951), for example. 83 Li138 (940).



the priest saio Fulgentio who managed the oaths and ordeal in a northern Galician case of the 990s; or, slightly differently, the two priests who accompanied a lay saio when he went to give notice to García Refugano to appear in the count’s court to answer a charge about the disputed ownership of a church.84 Some of the groups of supporters that took an oath on behalf of one party included people who seem to have been local priests.85 A priest might also act as surety: when the son of a peasant committed theft near Cumbraos in the Tambre valley in northern Galicia, in 931, six sureties were appointed that he would not run away and two of them were priests; the priest Ermegildo was surety for two lay parties in the agreement following a dispute before a saio and a judge in Valdoré, in the Upper Esla valley, in 946; and in Castile the priest Cardello was surety for his neighbours that their gifts to provide for burial would not be diverted.86 We also hear of criminal priests in court, some of whom look local: the priest Adulfo committed homicide in northern Portugal in 943 and asked a local aristocrat to help him out (and hence the subsequent gift of his church to the aristocrat); the priest Oderigo stole from a monk, some time before 988, and had to pay half his church as a fine; the priest Ero committed both fornication and homicide, his church of San Cristóbal being confiscated by the king in 985.87

Concluding thoughts From this survey of the questions that can be answered about the nature and role of local priests in northern Iberia in the ninth and tenth centuries, the things that emerge as important are: that there clearly were local priests; that their system of support was based on personal and family property rather than on any institutionalized levy such as tithe; that control was overwhelmingly exercised through proprietary interests (be they the interests of lay person, cleric, monastery or bishop), and not much through episcopal disciplinary procedures; that they probably delivered more pastoral care for the local community than is immediately obvious; and that they also provided all sorts of other practical functions – notarial, business and legal – functions which were important for community cohesion. And their churches were clearly useful locations for the conduct of ordinary secular business and for the swearing of oaths in regular judicial procedure. For the local community priests were landlords too. And for some communities they must have been the gateway to spiritual power, although it is impossible to estimate the level of public participation in the ritual that local priests were evidently equipped to perform – in some parts the ritual may have been the private activity of a restricted

84 85 86 87

T17, Sob109, C90 (957). E.g. SJP14 (928), Lii410 (968); C151 (972) notes a false oath sworn by a priest. Sob21, OD4, V17 (939). PMH DC 53, LaC 102, Lii507.



group, the preserve of a hereditary learned class maintaining in some fashion a system inherited from a distant Late Antique world. Overall, the character of the ecclesiastical landscape of northern Iberia was extremely mixed, with pockets and networks of episcopal or monastic interest, sites of lay proprietorship, the accumulations of excessively proactive priests, local churches with dependent churches and isolated institutions.88 The notion of ‘microchristendoms’, of varying modes of organization, proposed by Iñaki Martín for El Bierzo in the seventh century is a helpful approach to understanding how such a complex landscape may have come about.89 There were probably some strong regional differences in the ninth and tenth centuries: there were more identifiable communities, more transactions in churches, more notarial services, more proactive purchasing by priests on the meseta. From Galicia, by contrast, we have fewer indications of the priest as community representative and a stronger sense of the family church going on from generation to generation (even late in the tenth century). But in many areas, despite the differences, the absorption of private churches into monastic networks, as they expanded, was a characteristic trend of the tenth century, and a major trend; the end of this period was clearly a time of change. A century later, a system of regular ecclesiastical administration was under development and local and other churches were beginning to look more like their counterparts in other western European regions.90

88 Cf. the comments of Larrea, ‘Construir iglesias’, pp. 329–30. 89 Martín, ‘The memory of the “holy men”’, pp. 167–9; he was drawing on Peter Brown’s notion of microchristendom. 90 I am extremely grateful to Bernhard Zeller and other members of the ‘Local Identities and Social Cohesion’ project for their comments on earlier drafts of this paper, as also to Walter Pohl and the Institut für Mittelalterforschung, Vienna, for encouragement and support for the project; to Steffen Patzold and Carine van Rhijn for their energy and industry in bringing a volume on local priests to publication; to Álvaro Carvajal and Julio Escalona for their invaluable comments on my final draft; and to the Leverhulme Trust for the award of an Emeritus Fellowship which allowed me to undertake some crucial archive work.


4 C R E AT I N G R E C O R D S O F JUDICIAL DISPUTES IN N O RT H E R N I B E R I A B E F O R E THE YEAR 1000 This paper is about making records of judicial disputes, throughout ninth- and tenth-century northern Iberia, excluding Cataluña; Cataluña is excluded because of its different socio-political contexts and strong Frankish background.1 Records of disputes that were handled in judicial assemblies survive in many forms, as is clear from the ‘Processos Judiciales en los reinos del norte peninsular’ (PRJ) database, and have different kinds of content. I have used a ten-fold classification of the content of these records. This is different from the PRJ classification system, because it was devised independently and started long before my awareness of PRJ. Since the scheme deals with records up to the year 1000 only, this means that recording practice is less wide-ranging than that detailed by PRJ, although there is plenty of overlap.2 I originally identified 289 cases in 257 records, in an analysis based on numbers of events rather than on numbers of documents, although it is now clear that 8 cases in 8 additional records should be added.3 The essential characteristics of each category are as follows: Accounts are retrospective narratives, constructed after the events; Incidental records are just that – brief incidental references to judicial proceedings in a record whose purpose was to record something else, such as a sale or gift; Fines are records of payment of penalties consequent on judicial proceedings; Confessions and Oaths appear to be verbatim statements made during court proceedings, although their origin is more complicated than that; Agreements cover several categories of agreement, some immediately consequent on confessions and some subsequently constructed (with the benefit of hindsight, this category would probably have been better split); Mixed and Implied are, I hope, obvious, and in any case are tiny groups; Debts and Confiscations are for the most part not consequent upon court-cases, but need taking into account as indicators of non-judicial mechanisms of dispute

1 First published in I. Alfonso, J.-M. Andrade, A. E. Marques (eds), Judicial Processes in Early Medieval Societies: Iberia and Its European Context (Leiden: Brill, 2020). 2 For detailed discussion, see W. Davies, Windows on Justice in Northern Iberia, 800–1000 (Abingdon: Routledge, 2016), ch. 2, for full exposition of the classification system. 3 I am very grateful to André Marques and Bernardo Cañon for this update.



resolution. Some types are very common (Accounts, Incidental, Fines) and others rare (Oaths especially, Confessions). Understanding how these texts were constructed is important if we want to use them to illuminate events, relationships and processes in the ninth and tenth centuries and if we want to assess their levels of credibility. This is not as simple as supposing that there was a court reporter recording what happened in court, or a public notary constructing an ‘official’ document, or even a scribe in a monastery producing a record for monastic consumption, because the processes behind the creation of these records were much more complicated. It is the purpose of this paper to unravel some of these processes. I will begin by drawing attention to four different elements of format, across the whole corpus; and then indicate some regular stages in the construction process, before deconstructing some of the more complicated texts.

Significant elements of format Firstly, formulaic framework. As it was everywhere in early medieval Europe, charter writing was highly formulaic. Favourite formulas recur over and over again, like ‘ut de odierno die uel tempore ipsa uinea de nostro iure abrasa et in uestro iure sit traditam uel confirmata’, an example of a standard formula transferring title in regular charters of gift and sale.4 Dispute records are much less formulaic than gift and sale charters, but there are nevertheless recurrent formulas in most categories. Hence, as many people have pointed out, ‘orta fuit intentio/ contentio’ is a standard way to begin an Account;5 ‘cognosco/cognoui in ueritate’ is a standard way to summarize a confession, although ‘que manifesto uerum esse fateor’ is the way to introduce an actual Confession; ‘conditiones sacramentorum’ is the standard introduction of an Oath and ‘nostrum placitum tibi compromittimus’ the standard introduction of one kind of Agreement; ‘presentemus nostras personas in presentia N’ is the standard formula for expressing a commitment to appear in court;6 while the sanctions of Incidental records, Fines and Debts, as also of gifts and sales on the meseta, frequently include the formula ‘Si quis contra hanc cartulam donationis aliquis homo ad disrumpendum uenerit uel uenero an 4 Li143 (941). See W. Davies, ‘Regions and micro-regions of scribal practice’, in J. Escalona, O. Vesteinsson, S. Brookes (eds), Polity and Neighbourhood in Early Medieval Europe (Turnhout: Brepols, 2019), pp. 305–23, for more detailed discussion. 5 See A. Prieto Morera, ‘El proceso en el reino de León a la luz de los diplomas’, in El reino de León en la alta edad media, II: Ordenamiento jurídico del reino (León: Centro de Estudios e Investigación San Isidoro, 1992), pp. 381–518, at p. 390; I. Alfonso, ‘El formato de la información judicial en la alta edad media peninsular’, in J. Escalona and H. Sirantoine (eds), Chartes et cartulaires comme instruments de pouvoir. Espagne et Occident chrétien (VIIIe–XIIe siècles) (Toulouse: Presses Méridiennes de Toulouse/CSIC, 2013), pp. 191–218, at p. 205. 6 I did not include commitments as one of my ten classes of dispute text, but see Davies, Windows on Justice, pp. 135–7. With hindsight, I should have done so; in what follows I treat them as a distinctive type of text.



ego uel quelibet subrogata persona’. Those who wrote dispute records, of whatever kind, wrote within an established formulaic framework. That means that the redactors will have drawn on a learned tradition, applying words deriving from much earlier practice, handed down through skilled families, as well as in some cases using formularies and model charters as well as the varied resources of the scriptoria of major monasteries and cathedrals. While some dispute records will have been produced in the context of a monastic or episcopal household, others will have been produced by people working in the inherited learned legal tradition; and others seem to have come from scribes working for aristocratic households (Braolio for Flaino Muñoz, Sendino for Munio Fernández, for example).7 The second element is narrative. Many records include an element of narrative, although this can be very brief. Some records include lengthy, detailed free-flow narrative of events leading up to judicial proceedings, then detail of the hearings themselves and then the outcome. Lengthy narrative is especially characteristic of the type I call Accounts, which use the third person and past tense and are explicitly retrospective. The third element is that of ritual words. In contrast with the narratives, some records are cast in the first person and present tense. These records tend to be much briefer and they represent a single aspect of judicial proceedings, not a whole story – for example, the words of evidence from witnesses or the moment of a confession. Such records include formulas as well as circumstantial detail. These formulas are so distinctive, their repetition so precise (with variants sometimes explicable by oral transmission) and they are used at such widely distant locations, hundreds of kilometres apart, that I think it likely that these were ritual words uttered in court, words that everyone knew, characteristically invoked when a given part of the proceedings was reached – just as ritual words are used to introduce the oath in a modern English court. The fourth element is verbatim reporting. It is extremely rare to encounter anything which looks like verbatim reporting. However, there are one or two records which suggest negotiated Agreements, that is, negotiated between two parties with assistance from reputable members of the court. These records seem to contain an attempt to represent what each party said. The short sentences read like speech: a series of events recounted by the party who initiated the case, sometimes repetitively, without formulas; then the voice shifts to the other party, outlining what was agreed.8 This really is rare. If such records were common, they were not normally kept. But it does show that it is possible to conceive of a kind of court reporting, not necessarily written during court session, for records could be constructed immediately afterwards, in conversation with one or both parties. But these must be records that were made close to the meeting of the court.

7 For the latter, see ch. 2 above, ‘Local priests and the writing of charters in northern Iberia in the tenth century’. 8 OD4 (946), OD43 (997), Liii597 (1000).



Stages of construction Looking at the structures and contents of the entire corpus of dispute records makes it clear that the records we now have were created at different stages in relation to court-cases. In the Agreements just mentioned, as in Oaths and Confessions and commitments, there is material that derives directly from the activity of the court – this is what might be termed a primary record. Here is an example, surviving as a separate record on a single sheet: In iudicio Flacenti, Carcia Ordonizi uel cetero concilio, manifestum sum, ego Flaino, ad peticione Elias, qui asere uoce de comite Flaino Monizi, uerum est, cot necare non ualeo, quia peccato inpidiente et diauolo inmitente sic toliuit, ego Flaino, muliere aliena ad suo marito et fecit qum ea adulterio et post ec sic fecit qum sua filia, qum matre et filia. Et que manifesto uerum est factore et manifestu nicili abeo que opona. Facto manifesto XI kalendas madias, era XXXIII. Flaino in hoc manifesto rouorabit.9 Secondly, there are cases in which such material, and/or more, has been worked up into a charter of conveyance: a raw confession becomes the reason for a transfer of property and so the transfer is recorded, with the usual formulas of conveyance and an allusion to the fact of confession, but not a quotation of the words of confession. This might be termed a secondary record. Here is an example, surviving in a cartulary copy: Nos homines qui sumus habitantes in uilla Zacoys. Id sumus prenominati, Hordonio, Ueremudo, Habze, Abdella et Timimi, uobis domni Manillani abbati et domni Didaci et fratrum uestrorum monasterio Cellenoue. Plerisque manet cognitum quod obtinuimus quandam partem uillule sancti Felicis iuri nostro de dato pontificis domni Rudesindi episcopi beate memorie quod nobis dederat ad stipendium usufructuario. Et peccato nobis inpediente misimus illam in contemptionem ad fratrem uestrum frater Oduario qui ipsam deganeam sancti Felicis obtinebat pro que peruenimus in ipsam uillam in presentia iudicum Froyla Monniiz, Abraham abba, Pelagius Aspasandiz, Oduario Tetoniz, Gemondo Froylaiz, Alfidio Gulderes, Pelagius Tructesendiz, Fagildus confessus, Hoduario confessus, Uimara Froylaiz et aliorum multorum filii benenatorum turba non modica qui ipsos terminos preuiderunt secundum in uestra diuisione resonat. Et cuncti dixerunt quia uestra est ueritas.10 Proinde causatus fuit nobis mandator uester nomine Saluator et roborauimus manifestum

9 OD38 (21 April 995). 10 A typical second-stage formula.



et placitum ut fuissemus ad legem, et quod nobis ordinasset adimplessemus.11 Obinde agnouimus nos in ueritate et adsignauimus uobis in uestra uilla per textum scripture placiti et per saionem Gudesteo,12 et fecimus uobis ligali placiti ut si in quacumque tempus, an nos an filii, neptis uel bisneptis nostris uobis aut fratribus uestris qui ipsam deganeam contenuerint aliquam calumpniam in ipsam uillam uobis inferre uoluerint pro ipsis terminis sicut in istam diuisionem resonat, quomodo pariemus uel tam nos quam qui hoc ausus fuerit infringere ipsam uillam duplatam et insuper per singula capita XX XX boues. Notum die V ydus Magii. Era XXV post Millesima. Hordonius in hoc placitum manu mea (signum). Ueremudo manu mea (signum).Habze manu mea (signum). Abdella manu mea (signum). Timimi manu mea (signum). Qui presentes fuerunt: Didacus presbiter. Crizila presbiter. Timimi presbiter. Melhe presbiter. Gudesteo presbiter. Nouidius presbiter. Teodemirus diaconus. Geodesindus Pepizi. Froyani Gaudio. Berosindus Iohanniz. Ero Audinizi. Azo Uimarizi. Fonso Sunilla et alii plures.13 Thirdly, there are the narrative Accounts, retrospective constructions which might refer to confessions, oaths and agreements but rarely quote them. The purpose here is to provide some explanatory background, preferably involving a legitimizing figure like a king, as also to focus on outcome and consequent confirmation of rights and/or conveyance of property – this might be termed a tertiary record, for an example of which, see the following S356. There were, then, at least three distinct stages of creating records and three different types of context for their creation: brief documents written in or shortly after court sessions, arising from court procedure; longer documents of conveyance arising from the consequences of a court session, written subsequently; narrative Accounts of background, court proceedings and outcomes, written at some distance from proceedings. The best demonstration of first and second stages is, as Roger Collins pointed out thirty years ago, the Confession in the Celanova cartulary quoted above, which has four separate texts attached (not quoted).14 Since I have discussed this text elsewhere, I will not do so at length here. However,

11 Reference to confession but not the actual words of confession, followed by another typical second-stage formula. 12 Reference to the actual handover of the property, but not the commitment to consign, as in a primary record. 13 Cel94 (11 May 987); fo. 38v. 14 See R. Collins, ‘Visigothic law and regional custom in disputes in early medieval Spain’, in W. Davies and P. Fouracre (eds), The Settlement of Disputes in Early Medieval Europe (Cambridge: Cambridge University Press, 1986), pp. 85–104, at pp. 89–90. My interpretation of detail differs, but his main point that a single case might generate ‘a plethora of documents’ must be right. See Davies, Windows on Justice, pp. 44–8.



the essentials are that the initial Confession is a retrospectively constructed text, referring to the fact of confession and recording the conveyance that was a consequence. The appended documents are a raw Confession, as made in court, a commitment by the representatives of both parties to appear in court, another commitment by the parties to appear in court with their evidence before a different judge and a final commitment to hand over the disputed property. All four of these appended documents have the signa of one or both parties but no witnesses. Now, since there is another copy of the initial charter of conveyance later in the cartulary, which is almost identical but does not have the four additional appended documents,15 it looks as if the retrospective record was the one for keeping and the four additional documents were the short-term outcome from the proceedings of the court-case, not intended for the long-term archive. This set of texts is extremely important in indicating the existence of brief texts (which not surprisingly do not survive in large numbers) generated in the course of a court-case. Recognition of these stages of construction allows us to differentiate the evidential value of ‘complete’ texts; it also allows us to disentangle some composite texts which – at the distance of some eleven hundred years – are difficult to comprehend. In what follows I will demonstrate what can be understood about the construction of some of these difficult texts by following the principles outlined above.

Deconstruction of composite texts Li192 There is a set of related records in the early twelfth-century Tumbo de León, a set published by Emilio Sáez as a single record, and laid out in the cartulary as one record – although separate sections are very clearly differentiated; the fact that the twelfth-century scribe considered all of this to be one record is indicated by the space for a single large initial capital, which has faded or was never added.16 Since the sections are indeed differentiated, and since each section is intrinsically coherent, it looks as if these sections originated at different times and that they or their exemplar may have been copied from a set of separate sheets, perhaps clipped or folded together. The texts relate to a dispute between the monastery of Santos Justo y Pastor (which lies in Ardón, to the south of León) and one Mateo who in 946 – according to the monastery – had attempted to appropriate the church of San Esteban and surrounding land and had attacked the attached dependents; the case was prosecuted on behalf of the monastery by the priest Berulf, who acted as advocate in other cases in this area. The case, and the records, look relatively straightforward, although there was clearly some uncertainty about the 15 Cel150; fo. 54r. 16 Fos 39v–40v.



monastery’s title to the church, since two separate occasions of making the gift are recorded by two different sets of donors, a year apart in 943–944, in near identical charters.17 Most of these original donors appear to have witnessed the transfer of property resulting from the court-case recorded here, as also three (perhaps four) of the witnesses of the second donation.18 1

The first document is a commitment made to the saio (legal officer) Sando that the representatives of both parties (Berulf for the deacon Sisebut and the monastery, Reuelle for Mateo [his relation, perhaps, or a neighbour]) will appear in León, where Berulf will produce evidence of the monastery’s title to the church and of the wounds inflicted by Mateo, and Mateo will receive the evidence; a very high 100-solidi sanction to the saio and court holder, in the event of default, is attached. It is dated 26 March 946 and ends with confirmation and signa of Berulf and Mateo, but no witnesses, as is common in primary records.19 It seems odd that although Reuelle is named as spokesman for Mateo, it is Mateo’s signum that is recorded and Mateo who agrees to appear (compare ACO B.1.11, parts 11–13, below);20 and it is odd that the sanction is extremely high (again, compare ACO B.1.11 below): [B]erulfus presbiter, qui asseret in sua uoce21 et de Siseuuto diacono uel de parte ecclesie uel monasterio, contra isto Riuelli, qui asseret in uoce Matheo, tibi saioni nostro Sando. Per hunc nostrum placitum, tibi conpromittimus qualiter die quod erit in alio die, post intrata dominica, in Legione de Galletia, presentem, ego Berulfus presbiter, mea persona et mea testimonia per quos firmem super isto Matheo, in cuius uoce asseret Reuelle, qualiter tenente in nostro iure uel post parte ecclesie ipsa

17 Li176 (27 August 943), Li186 (24 September 944); these charters are essentially the same, although there are a few variations in the formulas and Li176 has corrupt elements. Both of these texts, like Li192, survive as cartulary copies. Since the texts of Li176 and Li186 are so very close, it looks as if a single record of donation has been duplicated to provide for two sets of donors, perhaps different branches of one family as there is some repetition of personal names. Since Li176 cites Godesteo son of Mazaref as father of the donors, as does the witness evidence of Li192 (see parts 2 and 3 below); and since Li176 has a long witness list which appears to be a representation of the León episcopal household (cf. Li201 [948], Li217 [950], Li220 [950]), whereas Li186 has a short and simple witness list which looks more local, Li176 may have been created in the context of the court-case of Li192, particularly since parts 2 and 3 of Li192 refer to the testamentos (i.e. documents) handed over to the monastery’s representatives by the original donors. 18 The witnesses Vincepti and Braelio had participated in the first donation, sons of Godesteo iben Mazaref; the witnesses Siseuuto, Godesteo and Valerio had participated in the second donation, which had been witnessed by the Li192 witnesses, the priests Kacem, Andreas, and Mantelli, and perhaps the saio’s representative Salomon. 19 See Davies, Windows on Justice, pp. 55, 136. 20 A man called Reuelli appears as witness to the final transfer. 21 ‘qui asseret in sua uoce’, ‘qui asseret in uoce N’, ‘Per hunc nostrum placitum, tibi conpromittimus qualiter die’ are all standard formulas.



ecclesia Sancti Stephani, cum suos dextros, ab integro, per istum testamentum et per auctores qui in ipso testamento resonant, et sic se leuauit22 iste Matheo et suos minores, una et alia uice, et maliarunt nostros minores et fecerunt plagas in solidos LXXII, et promiserunt ipsa ecclesia cum suos dextros de nostro iure. Et ego Matheo, mea persona, pro ipsas testimonias accipere, et qui minime fecerit pariet saioni nostro uel post parte dominica solidos C. Factus placitus23 VII kalendas aprilis, era DCCCCLXXXIIII. Berulfus presbiter et Matheo, in hoc placito manus nostras ++ roborauimus. 2

The second document is a statement of the witness evidence produced in favour of Berulf, Sisebut and the monastery by Gualamir and the priest Flaino, as prefigured in part 1: Verbo testis quod testificauerunt testimonias, de parte Berulfo et Siseuuto, id sunt: Gualamiro et Flaino presbitero, quia occulis uidimus et aures audiuimus et de presente fuimus quando24 habebat ipsa ecclesia Godesteo iben Mazaref, cum suos dextros, ab integro, et sic fecerunt Bani Electa et filii domni Godestei inde testamentos ad isto Berulfo et Siseuuto; et sic se leuauit iste Matheo et expulsauit inde suos minores, de Berullfo et de Siseuuto, et fecerunt plagas in solidos LXXII. Et ec que testificamus et iurare debemus.


The third is the formal Oath of the witnesses, of one of the two standard types, with standard introduction.25 It repeats the evidence cited in parts 1 and 2, and adds the names of the four judges; it concludes with a curse, which is unusual: Conditiones sacramentorum atque ex ordinatione26 Furtunio, Aurelius, Teodela Gutiniz seu et Domnon Uimaraniz. Iuraturi sumus nos testes prolati, qui sunt de parte Berulfo et Siseuuto, id sumus, Gualamiro et Flaino presbitero. Iuramus inprimis per Deum Patrem omnipotentem qui fecit celum et terram, mare et omnia que in eis sunt; iuramus per XII Prophetas et XII Apostolos et sancta quatuor Euangelia, id sunt, Marcus, Matheus, Lucas et Iohannes; iuramus per reliquias Sancte Marie Uirginis et genetricis Domini nostri Ihesu Christi; iuramus per reliquias Sancti Claudii uel per diuina omnia que sunt sancta; iuramus quia occulis uidimus et aures

22 For sic se leuauit, see also ACO B.1.11, parts 2, 4 and 6, and S356, in what follows at nn. 51, 57; it looks as if it was a standard introduction for an accusation of intrusion. 23 Use of the nominative case here is extremely unusual. 24 ‘occulis uidimus et aures audiuimus et de presente fuimus quando’ is a standard formula used in declaring evidence. 25 See Davies, Windows on Justice, pp. 124–6. 26 Cf. adque ex ordinatione in the Oath of ACO B.1.11, part 5, following.



audiuimus et de presente fuimus quando habebat ipsa ecclesia Gudesteo iben Mazarref, cum suos dextros, ab integro, et sic fecerunt Uani Electa et filii domni Godestei inde testamentos ab istos Berulfo et Siseuuto; et sic se leuauit iste Matheo et expulsauit inde ipsos homines de Berulfo et de Siseuuto, et fecerunt plagas in solidos LXXI.27 Et ec que iuramus, recte et fideliter iuramus, et si mentimus et nomen Domini in Filio28 tangimus,29 conprehendat nos iuditius quod sic conprehendit Dathan et Abiron, uiros sceleratissimos, qui propter eorum scelera uiuos et terra obsorbuit. 4

The fourth is not standard and is a separate confirmation of the sworn evidence by the accused, Mateo, with his signum: Quod testes testificauerunt et iudices iudicauerunt, et ego Matheo me in ueritate te agnosco, et iura et pena illis perdimitto et hanc conditionem manu mea confirmo +.


The fifth text is a commitment by Mateo to Berulf and Sisebut to hand the church over to the monastery by a Friday, with another very large sanction in solidi, to be paid to the court holder, completed with Mateo’s signum; it has a non-standard form, for it would be more usual to make the commitment to the saio and to agree to appear on a given day in order to consign the property, rather than express the intention as a condition: Matheo uobis Berulfo et Siseuuto. Facio uobis placitum, ut die VI feria, si non consignauero ipsa ecclesia cum suos dextros, ab integro, post uestra parte uel de monasterio, quomodo pariem post parte dominica solidos C. Matheo in hoc placito manu mea +.


The sixth is a list of confirmers and witnesses of Mateo’s preceding confirmation, including four judges and the saio, set out in two columns; it adds one judge to those named in part 3 but does not include Fortún García among them, as before. Fortún García and Pelayo González were high aristocrats, who attended the royal court, and there is at least one deacon from the cathedral church: In cuius copresentia roborauit ipsa conditione, id sunt: Furtunius Garseani cubiculari conf (monogram). Oueccus Nepotiani ts (signum).30 Felix, diaconus de Sancta Maria, ts. Bonellus diaconus ts. Mirabiles diaconus ts.

27 Note the minor variants from the last citation: LXXII/LXXI, Bani/Uani, Mazaref/Mazarref, minores/ homines. 28 ‘in Filio’ is perhaps an error for ‘in falso’; see n. 29. 29 For ‘et si mentimus et nomen Domini . . . tangimus’, cf. ‘in falso’ of ACO B.1.11, part 6, following; cf. also ‘nomen Domini falsamus’ of Sob109. 30 Here, as elsewhere, I indicate by the word (signum) the signa that were more elaborate than simple crosses, which are represented by +.



Sando, ubi saio fuit, conf. Aurelius quos iudicaui conf.31 Pelagius Gundisaluiz quos iudicaui conf. Teoda Gutiniz quod iudicaui conf. Domnon Uimarani quod iudicaui conf. 7

The seventh document is a commitment by Mateo, made on 7 May to Berulf and Sisebut, that if there are more or worse assaults on the church, property or dependents of San Esteban, a very high sanction of 200 solidi will be payable to the monastery and the court holder: Matheo uel filiis meis tibi Berulfo et Siseuuto, per hunc meum placitum uobis conpromittimus, ut de eodem die uel tempore,32 si ego Matheo uel filiis meis uos Berulfo et Siseuuto in ipsos dextros de Sancti Stephani uel in omnem prestariam eius aliquam inquietationem fecerimus uel ad uestros minores aliquam disturbationem fecerimus, plus quam fecimus, quomodo pariemus uobis uel post parte dominica solidos CC. Factum placitum ipsas nonas maii, era DCCCCLXXXIIII.


The last document might arguably be seen as intrinsic to part 7; however, I have separated it because it relates to a further, subsequent event. This is a list of witnesses of the moment of consigning the church to Berulf by Mateo and his children, through the agency of the court holder’s (Fortún García’s) saio,33 recorded in columns: Qui presentes fuerunt quando Matheo et filii eius consignarunt Sancti Stephani, cum suos dextros, ab integro, ad Berulfus presbiter, per manu saioni de Furtunio Garceiz cubiculari, id sunt: Hacem presbiter. Teouigildus presbiter. Adrias presbiter. Mantelli presbiter. Alamiro presbiter, Mutarraf presbiter. Celso presbiter. Micael presbiter. Abonazar presbiter. Titon presbiter. Gondemarus presbiter.34 Siseuuto ts. Azfon ts. Godesteo ts. Fedzu ts. Vincepti ts. Braelio ts. Baltasari ts. Reuelli ts. Melic ts. Abdelmelchi ts. Peppi ts. Valit ts. Hazmo ts. Valerio ts. Fatta ts. Fofo ts. Besuleo ts. Gaudio ts. Salomon, qui fuit uigarius de Sando, saioni.

There is no reason to suppose that this court-case did not happen. It has many intrinsically credible features – the royal officer as court holder, the actions of his saio and the saio’s representative, the six-week period from accusation to

31 ‘quos/quod iudicaui’ is a standard way of representing the subscriptions of the judges of a court-case. 32 ‘per hunc meum placitum uobis conpromittimus, ut de eodem die uel tempore’ are standard formulas of commitment. 33 Or perhaps, in practice, through his representative, since it is the representative not the saio who is noted in the witness list. 34 Perhaps the protagonist of Li128 (938), recording a dispute in the same area.



final settlement, the use of standard formulas and standard procedures. Mateo and Reuelle appear to have been based in Marialba, just to the south of León and not far from Ardón, and they feature as witnesses to other charters in the 940s and 950s.35 Álvaro Carvajal has pointed out that Fortún García, who was in any case a royal chamberlain, appears to have had considerable personal properties in the Tierra de Campos, one of which territories – Mazules – seems to be the ‘Villa de Mazul’ which included the church of San Esteban.36 It is not surprising, then, that it was his saio who handled the case and that Fortún was the court holder. This text has a number of surprising features. These include the naming of a representative for Mateo, who does not appear to act for him; the background donation of the church to the monastery of Santos Justo y Pastor apparently by two different sets of donors; and the unusually high penalty clauses of 100 and 200 solidi – 5 solidi was the usual penalty for contravention of a commitment.37 This suggests that the penalties, which lie in texts that were probably constructed for the beneficiary, may have been framed by the monastery to serve as a severe warning to other would-be intruders – perhaps an example of methods used by a monastery to warn local people to keep off its property, particularly when the title was recently acquired and somewhat uncertain. Here we see a representation of an intruder’s mark at least three times on different documents, presumably presented to him in public, in an extremely high-level court in which there was an element of royal interest; once his capitulation had been secured, he was threatened with a gigantic sanction in the event of default. Whatever the politics of the situation, this text offers a very clear example of the existence of at least six separate, coherent, credible texts deriving from a single court-case. T66 This record occurs in the early fourteenth-century cartulary of Santo Toribio in the Liébana, although it is clear that its ninth- and tenth-century material was collected earlier at the monastery of San Martín, Turieno, in the valley below Santo Toribio. It is quite difficult to comprehend this text; as it stands, the syntax is extremely misleading and there seem to have been errors and omissions in transmission. There is an occasional emendation of the cartulary text, which suggests that the scribe had some difficulty with his exemplar; and the extant cartulary does not differentiate between the constituent parts of the record; that, and the character of the record, suggests that it was copied from a copy and not from the original texts. However, there is enough to see that this composite document

35 Li187 (944), Li209 (950); see Á. Carvajal Castro, ‘La construcción de la monarquía asturleonesa en la meseta del Duero. Estudio de los procesos de integración territorial (s.ix–xi)’ (Doctoral thesis, Universidad de Salamanca, 2013), p. 603. 36 Lii482 (981); cf. Liii749 (1017). Carvajal Castro, ‘La construcción de la monarquía’, pp. 366–8. 37 Cf. ACO B.1.11, parts 3, 11, 12, 13, following.



details the consequences of a dispute between two lay people, Siseguto and Fredinando, over control of a vineyard at Mieses, close to Turieno, in 962. 1

The first document is a first-person Confession by Siseguto in response to Fredinando’s petition, using the standard formulas of Confession records; it was made on 7 May 962 in a court in the monastery of San Martín, witnessed by three judges, a saio and one other person. Siseguto had asked Fredinando to transfer a vineyard to him by charter. When this had been done, Fredinando was taken to court by Tellu Romanit, questioning his authority to deliver the charter (the text is ambiguous here – Tellu may have acted for the original owner or perhaps Tellu was himself the original owner). As a result Siseguto retracted his part in the transaction and Fredinando paid a fine of 50 solidi (also ambiguous): In iudicio Hopila abbas, Uermudi Allefonsiz, Gotini et Martinus, puerum regis, uel aliorum multorum iudicium (sic) et in conuentum ecclesie Sancti Martini, manifestus sum ego Sesecutus ad peticionem Fredinandi, qui asserit in sua uoce. Uerum est quod negare non ualeo,38 qui fui ad casa de Fredinando cum Froila et rogauimus illum per cartam traslatare de uinia qui est in Mesas cum suos arbores et translatauit uobis39 Fredinandus ipsam cartam et prendibit illum Tellu Romanit ad iudicium, et litabit ipsa carta placitum pro autore, et recesamus nos de atorgatione,40 et pariauit41 Fredinandus quinquaginta sollos; et quod ego manifesto uerum esse factum, et in iudicio nihil abeo que hoponam. Ego Sesecutus in hoc manifesto manus mea + feci et in conuentum ecclesie roboraui. Factus istus manifestus die II feria, ipsas nonas magias, Era M. Ego Hopila abbas, quod iudicaui manus mea + feci. Ueremudus Allefonsiz (signum). Gotinus quod iudicaui testis (signum). Alloritus ubi testis fui +. Sisinandus +.


The second document is a formal commitment by Siseguto and Fredinando, to the saio, that Siseguto will consign the vineyard to Fredinando on 15 May, together with another vineyard and some apple trees near Fredinando’s house, with a sanction of 5 solidi payable to the saio in the event of default. It is not witnessed but has the signa of both parties:

38 ‘Uerum est quod negare non ualeo’ and ‘quod ego manifesto uerum esse factum, et in iudicio nihil abeo que hoponam’ several lines later are the standard Confession formulas. 39 Recte ‘nobis’? This abbreviated word could be read as ‘nobis’, but the minims are unclear on the scan of the microfilm of the manuscript. 40 A technical term; see L. G. de Valdeavellano, ‘Escodriñamiento y otorificación (contribución al estudio de la reivindicación mobiliaria en el derecho español medieval)’, in Centenario de la ley del notariado. Sección primera: estudios históricos, 2 vols (Madrid: Junta de Decanos de Los Colegios Notariales de España, 1964–5), vol. 2, pp. 125–335, at pp. 274–324, who demonstrates that the meaning of autorigare is to provide guarantees of the legitimate possession of a thing purchased. 41 The standard term for ‘pay a fine’ in charters of this period.



Fredinandus et Sesegutus tibi Sanino42 Allorito per hunc nostrum placitum,43 qualiter die II feria quod est ipsos idus magias Era millesima consignem ego Sisecutus tibi Fredinando de uinea que est in Mus, iusta uinia Sancti Iuliani, et alia in Befares, iusta flumen Deua et pomares ad illa Cabella subtus tua casa;44 et ego Fredinandus pro accipere ipsas uinias et ipso pomare per manum sayonis et qui minus facerit pariet tibi sayoni nostro V sollos. Fredinandus in hoc pactum + feci. Sesegutus in hoc pactum + feci. 3

The third text is a guarantor document in which the guarantor Allorito undertakes to Fredinando, in the monastery’s court, that if – three days after the due date – Siseguto should have defaulted on the preceding commitment, then Allorito himself will give Fredinando some comparable properties. There are no witnesses but the document has the signum of Allorito, who is clearly a different person from the saio Allorito since both are mentioned in the final sentence. There are very few surviving guarantor documents, but there are examples from Castile and the meseta:45 Alloritus qui sum fiator de seguro contra Fredinandum usque ad diem V feria ipsus idus magius Era M, si non consignarem ego Sesegutus tibi Fredinando ipsas uineas que in pacto resonat, quomodo dem tibi ego Alloritus similes tales uinias et pomares; ego Alloritus in anc fiatura manus mea + feci in presencia Hopila abbas et sagoni Allorito uel aliorum multorum testibus.


The fourth document is a donation charter, which begins with a short passage of narrative detailing the events which preceded the donation. I will take these two sections separately. a)

A third-person narrative describes how Siseguto defaulted on the agreement, so that Allorito carried out the guarantee to Fredinando, giving him two substitute properties. Allorito then took Siseguto to a different court where he (Siseguto) confirmed the guarantee agreement, which carried a double penalty: mentiuerit Sesegutus placitum et fidiatura que roborauit Alloritus ad Fredinandum ipsa fiatura uinia in Befares et pomare in balle de Bosita et prendebit Alloritus Sesegutus ad iudicium ante

42 Recte ‘sayono’ (i.e. ‘saio’)? 43 ‘per hunc nostrum placitum’ is part of the standard formula of commitment; compare parts 11, 12, 13 of ACO B.1.11 following and parts 1 and 7 of Li192 above. Part 13 also has ‘ego N pro accipere NN per manum sayonis’ and the 5-solidi sanction is characteristic of all four commitments in ACO B.1.11. 44 The form of the placename of the contested vineyard here and in part 4 is Mus, whereas in part 1 it is Mesas. Given the detail of the location here, Mus must refer to Mesas, modern Mieses. 45 C103 (962), V43 (968–75), Lii464 (979).



Monio Uelasco et roborauit pactum pro ipsa fiatura uinia duplare et duplauit illa.46 b) Most of this document is a first-person donation charter, of 26 May, from Siseguto to the abbot of San Martín and to Allorito, through the agency of a different saio, concerning four properties; these properties are the two which had originally been specified in the commitment from Siseguto to Fredinando (that is in part 2), together with two additional properties. It concludes with sanction, dating clause and witness list (including the two saiones and three priests) of standard type; and the disposition includes a transfer formula of standard type: Ego Sesegutus per manum sagoni nostro Albiniano et consignaui uiniam in Toreuno, iusta uiniam regis et de alia parte iusta uiniam in Mus, iusta uinia Sancti Iuliani; et alia in Befares iusta flumen Deua; et pomare in Mus iusta flumen Deua, et alium ad illa Cabella. Consigno uobis ipsas uinias per manum sagonis et roborauit uobis carta et iudicium ut de odie die de meo iure sit abrasum et in iure de abbati nostro Hopila et Allorito sit confirmatum. Si quis tamen, aliquis uos ad disrumpendum pacta uel testamentum, an ego Sesegutus an germanis meis an aliquis subrogita persona, quomodo pariet uobis ipsas uinias duplatas et ipsum dictum uel quantum ad uos fuerit melioratum et hec scriptura ista firmiter abeat firmitatem. Facta pacta uel testamenta VII kalendas iunias Era M, principe sedente Sanzio in Legione. Ego Sesegutus in anc scriptura ista pacta uel testamenta manus mea + feci et testibus47 quos iudicaui +. Gotinus quos iudicaui testis. Alloritus ubi sago fui testis (signum). Albinianus ubi sago fui +. Danus testis (signum). Lemicius presbiter (signum). Egerius presbiter testis (signum). Teodarius presbiter +. Iohannis testis. There are clearly parts of the story that cannot be reconstructed and some parts remain ambiguous, but the process described is credible, especially with its several meetings on different dates in May 962 and with the involvement of two saiones and two different courts. All of the identifiable properties are very close together (within 3 km of each other) and they appear to be of rather small scale. It must be likely that the Allorito who benefitted from the final donation was Allorito the guarantor, since he had suffered a loss in making good his guarantee to Fredinando; the abbot benefitted as original court holder, to whom part of the penalty for default would be due. Whether or not any of the parties were tenants or clients

46 The t of duplauit has been erased. 47 There must be a name missing here.



of the abbot is not clear; it is possible, although this does not have to have been the case. Whether or not the guarantor came from the monastery is not stated but it is also possible. Despite the many difficulties of this text, it is very important for showing explicitly how a guarantor carried out his responsibility and what he might do if the person he guaranteed defaulted. The text also suggests that a 50-solidi fine was at one stage paid – these sanctions were not just unrealisable pious hopes; and, very interestingly, the text explicitly shows the payment of the double penalty, as required by the sanction, in the final settlement. The form of the documents, the use of standard formulas and the circumstantial detail of the text make it perfectly clear that this record was constructed from four or more related but independent texts. ACO B.1.1148 This very large, and very long, document, which survives on a single sheet in Oviedo cathedral archive, has the appearance of being an account of a court-case between Pedro and Victino, heard in Gordón, ending up in favour of Pedro, confirmed in the royal court of the kings of León in 953, and witnessed by several people of very high status. If we look at it in detail, we can see that the basic story is stated no fewer than three times,49 in association with eleven separate documents. The charter is very worn in parts and hence sometimes illegible. 1

The first document is a commitment by both parties (Victino represented by Vilifred) to the saio Karoso to appear in the court of Teodoredo and Felix on the 8th of the Ides of November (6 November), using the standard formula of commitment; Pedro also commits to bring witnesses: (Chrismon) Petrus frater qui acsere, in sua uoce, et Uilifredus frater qui respondet, in uoce de Uictino, tibi saionni nostro Karoso per unc nostrum placitum tibi conpromitimus qualiter die quod est die . . . viiio idus nobembris sic presentemus nostras personas in presentia Teoderedo Iustizi et Felix iudigum presentem ego Petrus mea persona et mea testimonia per quos firmetur super Uilifredo qui est uigarios de Uictino.


The second document is Pedro’s statement of his evidence, in other words the story: how Victino invited him to dinner, gave him land to build a monastery

48 See Davies, Windows on Justice, pp. 1–5, 146–51, for discussion and photographs, where I used the text of OvC26. The most recent edition is Documentos orixinales de los sieglos IX-X de los archivos del monesteriu de San Pelayo y de la catedral d’Uviéu, ed. A. M. Miranda Duque and X. Ll. García Arias (Oviedo: Universidad de Oviedo, 2011), pp. 61–7, no. 18; this diplomatic edition is a real improvement but my readings occasionally differ. 49 Note that the case is stated three times in the separate documents of Li192 prior, as it is also in Li34 (915). This may have been standard practice. These documents were not intended to be kept and hence such repetition is not a feature of most texts that are preserved.



and confirmed it on pain of a sanction of 100 solidi if he should default; the monastery was built but then Victino took it back, on the advice of his wife and Teodevera and Sister Gogina, and Victino also took a monastery that Pedro had acquired from his ancestors and a vineyard which Sister Gogina and Abbot Severo had given him; and then Abbot Severo confirmed Pedro’s legitimate possession in a meeting in Gordón before Teodoredo and Felix: quomodo segdente50 se Petrus in suo monasterio de Seuero de Seuero abba sic se leuauit51 Uictinus et inuitauit Petrum ad sua mensa et in alio die accessit ad illo uolumtas animo karo et spontania sua uolumtate et donauit ad Petrum terra uagua sua ratione et de sua muliere Fredildi et de Teodeuera uno consilio aderato et terra donada et assinada super placito et iuramentum que fecissent ibidem monesterio ita et fecerunt et roborabit placito Uictinus in solidos C et monesterio facto et eglesia et kasas et corte clusa et ortum et pomare postum et aqua metuta que ipso monesterio regat syc se leuauit Uictinus per consilio de Uilifredo qui ipsa uoce autorigauit ad integritate et per consilio de serore Gugina et de Teouera et de Fredildi bracio forcioso animo orrido et tullerunt ad frater Petrum ipso monesterio et ipsa ereditate unde placito rouoraberunt et tulliuit mici alia eglesia quem abeba de auios meos et uinia que donauit Seuero appa et serore Gogina ad Petrum et autoriauit Seuero appa ad Petrum in concilio de Gordone ante Teoderedo Iustiz et Felix iudigum in conuentum eglesie Sancte Marine. Si se leuauit Uilifredus et Uictinus et tullerunt mici ipsa uinia et ipso uino. 3

The third document, which is a statement by Vilifred (the spokesman for Victino and his sisters) that he will receive [Pedro’s] evidence, with a sanction of 5 solidi for detractors, is incomplete. It is in fact the end of the first document; together parts 1 and 3 represent the commitment of both parties to present and receive evidence in court, with that evidence (i.e. part 2) inserted between the commitments of each party:52 et ego Uilifredus qui sum uigarius de Uictino et de suas sorores mea persona pro ipsas testimonias accipere et qui minime fecerit pariet solidos V que Petro et Uilifredo in oc placito manus nostras ++ fecimus.


The fourth is a witness statement for Pedro by thirty-three people, three of whom are named, the words of their statement being introduced by the standard formula for such cases. Their statement is in fact a repetition of Pedro’s statement (part 2), with some very slight variants, adding the location of the land and ending with the signa of the three named witnesses:

50 Italics indicate the start of the repeated passage. 51 For this expression see also n. 22 above and n. 57 below. 52 Cf., with a much larger penalty, Li192, part 1, above, where commitment to present and receive evidence is part of a single document.



Ueruum testium que testifigaberunt testimonias de parte Petrum contra Uilifredo qui est uigarius et mantator de Uictino et de suas sorores id est Felix, Agila, Onorigus et alios plures numeros XXXa. Quia occulos uidimus et presente fuimus quando adduxo Uictinus frater Petrum ad sua mensa et sedente se Petro in suo monesterio su mano de Seuero appa si se leuauit Uictinus et accesit ad illum uolumtas animo karo et espontania sua uolumtate consilio aderato quum Teodeuera et Fredildi et rouorauit placito ad fratrem Petrum in solidos C de ipsa ereditate de Teodeuera et de Fredildi que donaberunt ad frater Petrum et donaberunt et assinaberunt ipsa ereditate que abebant in Paragiao terra uagua que fecisset ibidem monesterio, ita et fecerunt, et monesterio facto eglesia et kasas et coorte clusa et ortum et pomare postum et aqua que ipso monesterio regat si se leuauit Uitinus per consilio de Uilifredo et de sorore Gugina et de Teodeuera et de Fredildi bracio fortioso et animo orrido et tulliuit ad Petrum ipsa ereditate et ipso monesterio et alia eglesia quam abebat de suos auios et uinia que donauit Seuero appa et sorore Gugina ad Petrum et autogauit illa Seuero appa in concilio de Gordone ante Teoderedo Iustizi et Felix iudigum et si se leuauit Uictinus et Uilifredus et tullerunt ipsa uinia et ipso uino ad Petrum et que testifigamus et iurare deuemus Felix + Agila + Onorigus +. 5

The fifth is the Oath, of one of the two standard types, introduced by the standard formula: (Chrismon) Conditiones sagramentorum adque ex ordinatione Teoderedo Iustizi Felix iudigum iuraturi sumus nos testes prolati qui sumus de parte de frater Petro testifigantes contra Uilifredo qui est uigarius de Uictino, id est Felix et Agila et Onorigo, iuramus inprimis . . . Deum Patrem omnipotentem qui fecit celum et terra, mare et omnia qui in ei sunt, iuramus pro XII profetas et pro XIIm apostolos et sancta IIIIor Euangelia, id est, Marcus, Mahteus, Lugas et Ioannes, iuramus pro reliquias Sancta Maria uirginis et genetrix Domini nostri Ihesu Christi, iuramus pro reliquias Sancte Marine uirginis in cuius loco as conditiones manus nostras tenemus sibe et pro diuina omnia que sunt sancta, quia occulis uidimus et presentes fuimus et bene nobis cognitus manet in ueritatem.


And the sixth is the sworn evidence within the Oath, which is another repetition of Pedro’s statement (part 2), with the location of the property (and is in fact the best – that is, the most coherent – version of this story). It ends with a slightly longer version of the ending of part 2, followed by a long sanction against anyone who should contravene the Oath or swear falsely, in which event he/she should submit to the judgment of God and the ordeal of hot water: quanto aduxo Uictinus Petrum ad sua mensa et sedente se Petrum in suo monesterio sub mano de Seuero appa et si se leuauit Uictinus cuius 69


uigarius est Uilifredus pro consilio de Teodeuera et de Fredildi et accessit ad illos uolumtas animo karo spontania sua uolumtate et donaberunt et assinnaberunt suas raciones quem abebunt in Paragiao sic Uictino quomodo etiam et Teodeuera et Fredildi ad frater Petrum spontania sua uolumtate et rouoraberunt placitum ad frater Petrum de omnia sua ereditate quem abebant in Paragiao ad integritate et roboraberunt placitum in solidos C et donaberunt et assinaberunt ipsa terra uagua quo fecissent monesterium ita et fecerunt et monesterio facto et eglesia et kasas et corte clusa et ortum et pomare postum et aqua mittuta qui ipsum monesterium inregat et ipso lauore facto sic se leuauit Uictinus pro consilio de Uilifredo qui ipsa uoce autorcauit ad integritate in concilio et pro consilio de sorore Gugina et de Teodeuera et de Fredildi et tulliuit ipsum monesterium ad Petrum et alia eglesia quam abebat Petrus de suos auios et uinia que donaberat Seuerus appa et serore Gugina ad Petrum et autorcauit illa Seuero abba ad Petrum in concilio de Gordone ante Teoderedo Iustiz et Felix iudigum in conuentum eglesie Sancte Marie et si se leuauit Uictinus et Uilifredus et tullerunt ipsa uinia et ipso uino ad frater Petrum et abent illo contra se et mentiuit ipsum placitum et abet et ereditate et placito contra se et qui iuramus recte et fideliter iuramus et in anc iuramento nunllo qui fraudis ingenio non interponimus et si mentimus et nomen Domini in falso tangimus conprendat nos iudicius Dei et pena kaldaria. 7

Part 7 is a statement that the oath was sworn in the church of Santa María at the gate of Gordón, and thereafter an innocent was taken to the ordeal. The statement continues that the innocent’s hand was clear three days later, the 12th of the Kalends of December (20 November) 923,53 with confirmation of this by three special witnesses (of whom the third appears to have guarded the innocent during the three days of waiting), the two judges, another person, the innocent and the saio: Iuraberunt istas testimonias in eglesia Sancte Marie qui exitata est ad porta Gordonens et pro eorum iuramentum ingresus est innocens ad pena kaldaria nomine Fredenandus qui in tercio die limpidus apparuit coram multorum testium qui II sunt conscripti uel signa facturi late conditiones, notum die XIImo kalendas dicembris era DCCCCLXIa. Monnisso ubi fidelis fuit de amborum partibus et rogitus de frater Petrum manum mea roborabit (signum). Assuri ubi fidelis fuit de amborum partibus et rogitus de

53 The date is clearly problematic and does not match the floruit of the subscribers. The text appears to read 923, although the diplomatic edition can see an additional V and hence reads 928; the subscriptions of the king and his court are consistent with 953, García Larragueta’s preferred date. See the discussion of Documentos orixinales de los sieglos IX–X, ed. Miranda Duque and García Arias, p. 61, who note an Abbot Severo recorded elsewhere in 932. A date for the court-case in the 920s or early 930s seems more likely than the projected 953.



frater Petrum manu mea +. (Long Chrismon) Furakasas ubi fidelis de amborum partibus rogitus de frater Uilifredo manum mea et nocentem sub custodia tenuit et in IIIo die limpidum presentabit mea +. Teoderedus Iustizi quos iudicabi (signum). Felix quos iudigaui +. Gaudemiro . . . (Chrismon) Fredenandus ubi innocens fuit et aserias conditiones iurauit et manum in kaldaria misit et . . . eiecit et in IIIo die in concilio limpidus apparuit manum mea + (Chrismon) Karosus ubi saio fuit. 8 The eighth document is a list of witnesses of the oath, perhaps including the saio, in the church of San Martín: Qui preses fuerunt ad iuramentum id est Roderigus presbiter (signum). Sunliam presbiteri (signum). (Long Chrismon) . . . relius Cenderigus Taurellus Gello . . . Karosus Pepinus Placinus . . . in conuentum eglesie Sancti Martini episcopi. 9 The ninth is a (different) list of witnesses of the innocent’s entry to the ordeal, including Vilifred: Qui preses fuerunt quando innocens ingressus est ad pena kaldaria id est Furakasas, abba Romanus confirmat. Ioannes confirmat. Uilifredus confirmat. Quintila confirmat. Tedisclus. 10 The tenth is a list of witnesses of the emergence of the innocent with a clear hand; it is for the most part different from the confirmers of part 7, although it includes the two judges: Qui ipsum nocente limpidum uiderunt id est Teoderedo Iustizi. Felix Rubio. Maximynus Nikeiti. Aurelius Ualerianus Azmon Nunillani presbiter. Donnellus. Sabgildus. Seuero. 11 The eleventh is the formal commitment of the two parties to the saio to appear in Gordón before judges with evidence to be sworn, with a sanction of 5 solidi for default but no witnesses; the formulas are standard commitment formulas:54 Petrus et Uilifredus tibi sayonni nostro Karoso per unc nostrum placitum tibi compromitimus qualiter die quod erat die Va f . . . sic presentemus personas nostras ic in Gordone in presentia iudigum presentem ego Petrum mea persona . . . ipsa testimonia pro ad iuramentum et ego Uilifredus mea persona per ipsas testimonias a iuramenture et qui minime fecerit parie solidos V que Petrus et Uilifredus in placito manus nostras (signa). 12 Part 12 is the formal commitment of the two parties to the saio to appear in Gordón before judges, on which occasion Vilifred will formally concede to 54 Cf. Li192 and T66 above.



Pedro, with a sanction of 5 solidi for default but no witnesses; the formulas are standard commitment formulas: Petrus et Uilifredus tibi saionni nostro Kaoroso per unc placitum tibi conpromitimus qualiter die quod sabbato sic presentemus personas nostras in presencia iudigum in Gordone presentem ego Uilifredus mea persona et meum pactum quantum lex mandat55 per sentencia et ego Petrus mea persona pro accipere ipsum pactum per manum saionis et qui unc placitum exerserit pariet soldos V que et insuper conpleat ipsu pactum. Petrus et Uilifredus in placitu manus nostras + + fecimus. 13 The thirteenth is the formal commitment of the two parties to the saio to go to Erias and Paragiao, where Vilifred was to consign the property, with a sanction of 5 solidi for default but no witnesses; the formulas are standard commitment formulas: Petrus et Uilifredus tibi sayonni nostro Karoso per unc nostrum placitum tibi compromitimus qualiter die quod quo die Va feria sic presentemus nostras personas ic in Ayrias pres (sic) in Paragiao presentem ego Uilifredus mea persona et assinnem ipsa uilla que in iudicio resonat per manum saionnis et ego Petrus mea persona pro ipsa uilla accepere et qui minime fecerit pariet solidos V que Petrus et Uilifredus in placito manus nostras + . . . fecimus. 14 The fourteenth is a list of witnesses and subscriptions of King Ordoño III and lay and clerical members of his court: Sub Christi nomine Gundesalbus aepiscopus sedis Legionensis confirmat (signum). Gundesindus comes palatii confirmat. Uermudus Nunnizi comes palatii. Fortes Iustizi confirmat. Dextera uirtus [sic – recte In Christi virtute?] Armentarius Dumiense sedis aepiscopus (signum). (Long Chrismon) Serenissimus princeps Hordonius (signum). Altissimi manu protectus Gudesteus Ouetense sedis aepiscopus (monogram). Sanctius serenissimus princeps confirmat (monogram). Geluira prolis Ranemiri confirmat (signum). Nebutianus presbiter (signum). 15 The fifteenth is an endorsement of 981, by a woman on behalf of Gogina, that the judgment was true, with confirmation and three witnesses: (Chrismon) In era XVIIII post millesima auidante domna Florentina in Sancta Eulalia in Lena anouit me ego Florentina comodo iste iudicio super Uictino est ueriuigo et de sua mano rouorado et ego in persona de domna Gogina et mea manum mea confirmo et autorigo manum mea. Qui preses fuerunt id est Neboziano presbiter (signum), Ranmondo presbiter (signum), Donnido presbiter cos noduit (signum). 55 Cf. ‘que les mantare’ in the Agreement consequent on a confession, OD39 (995).



Both part 14, the royal witness list, and part 15, the 981 endorsement, are added by different hands from the main hand. The contents of the three columns below the body of the text (which are not laid out in strict columns) are as follows: in the righthand column are parts 9 to 13, that is, the fourth and fifth witness lists and the last three commitments, still in the main hand; the chrismons indicate that this was part of the main design of the charter. The middle column and upper part of the left column contain the royal and aristocratic subscriptions, the king being focal. At the bottom left lies the 981 endorsement. This is clearly a retrospectively constructed document, intended to impress (by size and length, and royal and aristocratic signa). It equally clearly has corrupt elements; for example the split of parts 1 and 3 is very odd, the dating is inconsistent, while some of the witnesses of part 14 cannot be pre-981, the date of the endorsement below it (Gudesteo went to his bishopric c. 992 and Armentario in 985); and it is clearly partly garbled: the repetitions of the long statement of Pedro’s case include errors, omissions and nonsense. However, the text has equally clearly been constructed from something. It includes seven different witness or confirmer lists; it includes four different commitments, in standard form, using standard formulas; it includes a witness statement of evidence, introduced by the standard formula, an Oath of standard type, and an account of an ordeal of standard type, using standard wording. The arrangement of the fourth and fifth witness lists and the last three commitments in a separate appendix to the main charter, bottom right, in itself suggests that these were independently existing elements that were available to the redactor; and the main story, given the nature of its variants, must have been copied from a written text – the wording of the introduction to the story that Pedro was in his monastery under Abbot Severo is a good example of this.56 So, this extremely important charter shows us both that at tertiary stage a redactor could use different strategies to create a credible-looking charter – from invoking standard formulas to copying old documents to repeating them; but also that shorter, closer-to-court documents must have existed: someone must have been making notes of different sets of witnesses and someone must have been drawing up the commitments. There is in fact no reason to suppose that such a court-case did not happen.

Conclusion These examples demonstrate that composite texts have been preserved which are essentially composed of primary, close-to-court, records. These have been brought together more or less coherently. Hence Li192 is still coherent; T66 much less so, although since it survives in a late copy it is impossible to assess the original compiler’s intention; and ACO B.1.11 looks coherent at first sight but is less so on perusal. Together, however, they make it abundantly clear that there was some

56 See Davies, Windows on Justice, pp. 150–1, for a detailed comparison of the three versions.



element of court recording in the ninth and tenth centuries, although it looks as if these records were not intended for long-term preservation. Most of the judicial dispute records that are still preserved are post-court constructions, sometimes using primary records and at others not. This is particularly well demonstrated by Accounts. For example, the tertiary-stage Account that follows has some narrative background on the disputed properties, a brief indication of the court-case, a reference to the oath taken, a reference to the subsequent confession and a solid conveyance with a guarantor and sanction: Hec est agnicio ueritatis facta sub era MXXXVIa, anno imperii domni nostri Garseani Gomiz comite et Zahbascorta ven Abolhauz sedente in Toro ipsas kalendas marcias. Denique orta fuit intemptio inter fratres de Domnos Sanctos, domno Uincentio abbate uel suos fratres contra Uela Uelaz pro Uilla de Petro qui est inter flumen Ceia et riuo Aratoi quia sic tenete fratres de Domnos Sanctos ipsa Uilla de Petro de testamento de Ansur maiordomo qui habuit ipsa uilla comparata de domna Salomona in La solidos et tenentes fratres de Domnos Sanctos ipsa uilla in suo iure sic se leuauit57 Uela Uelaz et sacauit ipsa uilla de iure de fratres et presumpsit ipsa uilla et suas herencias;58 et deuenerunt inde ad iudicium ante Garsea Gomiz et Zahbascorta in Uilla quam uocitant Alpando et dederunt fratres de Domnos Sanctos suo mandatore Andreas;59 et petiuit ipse Uela Uelaz pro sua erencia et pro ipsa uilla iudicium et ordinarunt ad comite Garsea Gomiz et Zahbascorta uel omni concilio quia sic dederunt fratres de Domnos Sanctos suos iuratores;60 et cognouit se Uela Uelaz in ueritate61 qui duplare uel componere quantum per mendacium presumpserat sane constricti in iudicio; et non habente Uela Uelaz unde componeret ipsa uilla cum sua erentia et tradidit illa uilla per manu fidiatoris Gelmiro presbiter; et firmauerunt ipsa uilla cum sua erencia ad fratres de Domnos Sanctos secundum quod in concilio habuerunt.62 Ita tamen ut si deinceps ausus fuerit talia comittere ipse Uela Uelaz aut filiis uel neptis uel aliquis homo qui hoc destruere uoluerit quomodo pariat uobis uel uoci uestre ipsa hereditate duplata uel quantum uobis fuerit meliorata.63 Notum die quod erit ipsas kalendas marcias, era suprascripta. 57 58 59 60 61 62 63

For this expression, see also Li192 and ACO B.1.11, nn. 22 and 51 above. The background. The court-case. Reference to the oath. Reference to the confession. The conveyance. The sanction.



Garsea Gomiz confirmat; Zabascorta confirmat; Ouecco Telliz confirmat; Fredenando Monniz confirmat; Garsia Fortuniz confirmat; Scape Pepiz confirmat; Argeuado confirmat. Felix diaconus exarauit.64 The difference between close-to-court and post-court construction has a bearing on the veracity of the record. Deconstruction of T66 and ACO B.1.11 made clear that it was easy for corruption to enter copied text, whether intentional or not, just as the corpus of single sheets that might be regarded as ‘originals’ contains plenty of additions, corrections and mistakes. Corruption is common and does not have to invalidate the content of a text. That is not the same point as bias. Most of the surviving records of judicial disputes were very obviously constructed by or for the principal beneficiary; and even the closer-to-court records must for the most part have been constructed by representatives of the victors. Even where a record was drawn up by order of the saio (as we know some were, for example oaths65) or by a judge (as a few could have been66), there still may have been beneficiary influence. Would that we could unpack the complex of local relationships that influenced the redactors; but we cannot, and there are inevitable unknowns. Plenty of bias is nevertheless clear. Two thirds of retrospective Accounts have a monastic beneficiary and most appear to have been written to record a real or imagined ecclesiastical victory. They are inevitably biased in favour of the ecclesiastical body that produced them. On the other hand, deliberate falsification is a different matter and is much more difficult to spot in dispute records of pre-1000. While one might identify, on palaeographic grounds, a pseudo-original which claims to be a document fashioned much earlier than it actually was, it does not follow that the content of such documents was fabricated or invented. For the most part we have to depend on accumulation of glaring anachronisms to identify the deliberate falsification. Most of the content of early medieval Iberian dispute records is beyond proof of any kind. Given the opportunities for inadvertent corruption and the straightforward bias of many of these texts, we need to retain a healthy scepticism about their contents. That means that in making our assessments we have to deal in terms of greater or lesser credibility, not proof. However, understanding the construction of the records makes it perfectly clear that elements of those texts reflect record-making that was itself part of judicial process and that even the most retrospective of records has a relationship, though complicated, with actual events.

64 S356 (1 March 998). 65 For example, the explicit reference to this in Sob109. 66 For example, OD31 (991), OD43 (997).



This paper is concerned with all the major charter collections in the kingdom of Asturias-León up to the year 1000: all of them, because comparison between collections is very instructive (and, indeed, looking at a single collection can be misleading).1 Asturias-León is a modern term for the biggest kingdom in northern Iberia in the tenth century, a kingdom which covered a quarter of the peninsula – a large geographical area, which is physically very diverse. The year 1000 is the chosen end point because of the changes in the political and ecclesiastical geography of northern Iberia in the eleventh century: the León kingdom expanded to take in territory to the south and east from Muslim control; the kingdom of Navarre, which lay to the north-east east, expanded to east and west to take in much of Aragón and Castile; and new charter collections were begun for new or revived institutions. Comparing like with like across the year 1000 is therefore difficult. Nevertheless, to assist comparison with other parts of western Europe, a quick survey of some of the most significant eleventh-century material is appended. The pre-1000 corpus of charters is large, in number approximately 2700, although not all collections were used for this paper. Most charters survive in eleventh- and twelfth-century cartularies but there are several hundred originals on single sheets of parchment; for the purpose of this analysis there appears to be no significant difference between exchange charters from cartularies and those on single sheets. The León material cited in the following table (5.1) comes from a number of different collections, both episcopal and monastic, which were eventually gathered in the cathedral archive; the Portuguese material (PMH DC) comes from many different sources, though that from the monastery of Guimarães is important in the context of exchange; and the collection called ‘Floriano’ in the table is a twentieth-century collection of charters from mixed sources, from the period 718–910 – these figures refer to charters not included in more recent editions (texts 1 First published in I. Fees and Ph. Depreux (eds), Tauschgeschäft und Tauschurkunde vom 8. bis zum 12. Jahrhundert/L’acte d’échange du VIII au XII siècle, Archiv für Diplomatik, Beiheft 13 (2013), pp. 471–89.


E X C H A N G E C H A RT E R S I N A S T U R I A S - L E Ó N

coming largely from Lugo and Santiago de Compostela). The raw data in this table show that surviving numbers of charters are extremely uneven, for a third of the total comes from León whereas just a fortieth comes from Otero. Numbers of charters are not the same as numbers of transactions, since a note of previous transactions sometimes prefaces the main record of transfer, since some charters record several separate transactions and since some texts do not record any. Over 90% of charters come from the tenth century; hence, what follows necessarily focusses on that century. The material comes from the different regions of Galicia in the west, the high plateau of the meseta north of the river Duero, Castile and its political focus of Burgos in the east, and beyond the Cantabrian Mountains in the north. Table 5.1 Number of charters used from the kingdom of Asturias-León 701–1000 Source of collection

Region Total charters Charters Charters Charters Total transactions 701–1000 701–800 801–900 901–1000 901–1000

Celanova Samos Sobrado PMH DC León Sahagún Otero Dueñas Cardeña Valpuesta Santo Toribio Oviedo San V Oviedo Cath. Floriano TOTALS

Galicia Galicia Galicia Galicia meseta meseta meseta Castile Castile North North North mixed

223 64 117 184 580 357 50 207 46 77 25 29 69 2028

10 6 18 11 13 4 1 1 7 15 1 11 53 151

1 1 1 1

2 9 15

213 57 98 172 566 353 49 206 39 60 24 18 7 1862

225 58 99 190 587 436 46 202 36 62 24 17 6 1987

Gift, sale and exchange This paper focusses on charters which describe transactions by using ‘exchange’ words such as commutare and concambiare. Of course, virtually all transactions involve exchange of some kind and in an ideal world one might reasonably consider all kinds. However, charters that use words signifying ‘exchange’ are a distinctive subset of the corpus and have some consistent characteristics, even if we allow that the meaning of commutare can signify different things and can change over time.2

2 See E. Magnani, ‘L’échange dans la documentation diplomatique bourguignonne’, in Fees and Depreux (eds), Tauschgeschäft und Tauschurkunde, pp. 403–26; also I. Rosé, ‘Commutatio. Le vocabulaire chrétien de l’échange au haut Moyen Âge’, in J.-P. Devroey, L. Feller, R. Le Jan (eds), Les élites et la richesse au haut Moyen Âge (Turnhout: Brepols, 2010), pp. 113–38.



I will begin with an overview of the principal types of exchange recorded by charter. Gift and sale transactions predominate in this corpus, as well as in the tenth-century charters of Navarre and Aragón to the east; some types of transaction that are common in collections elsewhere – like loans or precaria(e) – do not feature in any identifiable way.3 Just over half the total number of transactions are recorded as gift and rather fewer than half as sale, although more sales than gifts are recorded on the meseta (see table 5.2). The language of gift and sale is very strongly differentiated in nearly all records, although a few at a first reading appear to be confused: hence scriptura donationis vel vendicionis in an apparently straightforward sale text or carta vendicionis of a charter which describes a pro anima gift to a monastery.4 Apparent confusions between the two concepts are nearly always capable of explanation: given that there is a spectrum of transactions from commercial exchange to gifts for no consideration, and given that there are basically only two formats for the record (gift or sale), transactions along the spectrum have to be assigned to one format or the other. In effect, transfers for something done in the past, like paying off debts or compensating for past assistance, are recorded as sale; and transfers for things to be done in the future, like paying fines or giving in order to secure some explicit benefit, are recorded as gift. There remains just a handful of apparent confusions that cannot be explained. Exchange charters are in fact a hybrid form. Modern editors have sometimes noted or counted exchange charters in their introductions to editions, and there is some comment by legal historians, but there is no large historiography on Iberian exchange.5 This is probably because the number of exchange charters in the corpus considered here is extremely low, constituting just over 4%, far fewer than those from for example Lucca or Freising; only four come from the ninth century, an even lower proportion. Some collections have none – like those from Otero de las Dueñas and Valpuesta; some have slightly more than 4% – like those from Celanova, Sobrado and León. Although on the whole more survive from Galicia and fewer from Castile, it is difficult to see any regional significance in the differences: in Galicia, Celanova has relatively many and Samos relatively few; on the meseta, the very large collection from León has relatively many but the comparably large collection from Sahagún has very few. With such low numbers 3 For analysis of gift and sale, see ch. 6 following, ‘Sale, price and valuation’; W. Davies, Acts of Giving: Individual, Community, and Church in Tenth-Century Christian Spain (Oxford: Oxford University Press, 2007), especially cc. 1, 5 and 6; ch. 7 following, ‘When gift is sale’. 4 Cel535 (953); SM90 (972). 5 For example, Emilio Sáez in his introduction to Li (p. xlv), commenting that the purpose of exchanging was to consolidate property, and José-Miguel Andrade in his introduction to Cel (p. xxii), noting numbers. Note R. Fernández Espinar, ‘La compraventa en el derecho medieval español’, Anuario de historia del derecho español, 25 (1955), pp. 293–528, at pp. 373–5; G. Martínez Díez, ‘Terminología jurídica en la documentación del reino de León. Siglos IX–XI’, in Orígenes de las lenguas romances en el reino de León, Siglos IX–XII, 2 vols (León: Centro de Estudios e Investigación ‘San Isidoro’, 2004), vol. 1, pp. 229–72, at pp. 237–8, who, surprisingly, sees ‘juridical precision’ in the records.


E X C H A N G E C H A RT E R S I N A S T U R I A S - L E Ó N

these differences may well not be significant at all and are probably the product of chance: the chance of the strategies adopted by major proprietors, the chance of survival and the chance of whose transactions were recorded in the first place. There are passing references to more exchanges, mentioned in comments on the previous history of a property conveyed by gift or sale;6 these have only been included in table 5.2 if there is detail of the actors involved and of both parts of the exchange.7 Table 5.2 Percentage of tenth-century gift, sale and exchange charters in collections from Asturias-León Source of collection

Total number of transactions 901–1000

% gift

% sale

% exchange

Total number of exchanges

Celanova Samos Sobrado PMH DC León Sahagún Otero Dueñas Cardeña Valpuesta Santo Toribio Oviedo San V Oviedo Cath. Floriano TOTALS

225 58 99 190 587 436 46 202 36 62 24 17 6 1987

48 88 60 55 42 37 37 58 70 58 33 88 66 51

41 0 30 33 50 58 57 38 28 34 63 6 33 43

7.1 1.7 6.1 4.2 5.5 1.8 0 2 0 4.8 4.2 5.9 0 4.03

16 1 6 8 32 8 0 4 0 3 1 1 0 80

Exchange charters: form and formulas The format of exchange charters is very similar across the whole region, from Galicia to Castile. The form is simple, with a tendency to include plenty of detail of boundaries and of the physical characteristics of land conveyed. The constituent parts of the charter are nearly always invocation, salutation, disposition – the latter often preceded by a ‘without pressure/of my own free will’ clause, as in sales, and often followed by a ‘freely into your power’ clause – then penalty clause, date, witness list and/or list of confirmers. There are no arengas. Occasionally there is no invocation or salutation, and occasionally the disposition is introduced

6 In PMH DC 77 (959), for example. 7 Exchanges did sometimes occur as a result of court-cases: for example, the outcome of the very long record of dispute of PMH DC 183 (999), where one party lost the disputed estate but was given another instead, although this is not framed as an exchange in the record; such cases are only included in the figures cited in table 5.2 if they are recorded in exchange format. See further ch. 15 following, ‘Judges and judging’, pp. 259–60.



by the gift formula magnus est titulus donationis in qua nemo potest actum largitatis irrumpere. Here, for example, is a standard record of an exchange between two lay parties, which survives on a single sheet in the León cathedral archive: In Dei nomine. Ego Leuuildi vobis Munio Nunizi et uxore vestre Paterne. Placuit mici volumtas, nemineque conientis imperio neque suadentis artigulo, set propria mici acesit volumtas,8 ut facerem vobis karta concambiacionis de omnem mea ereditatem que abeo in villa Talecias, locum predictum ubi abolus meus Stevanus et Leocadia avitavit; omnem mea racione et de iermano meo Ermegildo, que nos continet inter meos iermanos, concedimus vobis ad integritatem. Et acepimus de vos alia terra in Talecias, in loco predicto in Campizo, ubi Lalli abitavit: de termino de pumare maiore, et per termino de Asela, et per termino Dagamiri, et afige in termino de Sisverti; ipsa terra ex integra. Ida ut de odie tempore ipsa ereditatem de nostro iure exitum, et in vestro dominio sit confirmatum, adeatis, teneatis, iure quietum in perpetuo vindicetis, tam vos et omnis postegritas vestra. Et si vos aliquis isrumpendum veneri, tunc abeatis potestate adprendere de nos ipso que in karta resona duplatum, et vobis perpetum abiturum. Facta karta concambiacionis VIIII kalendas abrilis, era DCCCCLXXXVII. Ego Leuuildi et Ermegildus in ac karta manus nostras fecimus. Amu Paternus ts, Amu Kardellu ts, Andosindus ts, Ofrasius presbiter ts, Andulfus ts, Sanctus ts, Iuannes ts, Donellus ts, Aurelius ts, Bega ts, Senior tresbiter ts, Stevanus presbiter ts.9 In the light of practice in other parts of western Europe, it should be noted that there are no charters recording royal confirmation of the exchanges of other parties in this corpus, although kings certainly participated in exchange themselves.10 The main verb used in exchange charters is comutare (74% of cases), with concambiare and contramutare substituted in 10% and 16% of cases respectively. Corresponding nouns are also employed (kartula/scriptura/titulo comutationis or (con)cambiacionis or contramutationis), particularly in dating clause and witness list, in the following proportions: 65%, 14%, 21%. In this the contramutare/ contramutatio variant is regionally specific to southern Galicia: it is used in 81% 8 This is a version of the exceptionally common ‘freely without pressure’ formula (placuit nobis atque convenit nullius cogentis imperio sed propria nostra voluntate accessit nobis ut); I am currently aware of forty different versions in northern Iberian charters of this period; see further W. Davies, Windows on Justice in Northern Iberia, 800–1000 (Abingdon: Routledge, 2016), pp. 97–106. 9 Li204 (949) (= ACL 90). 10 As demonstrated by Ph. Depreux for ninth- and tenth-century Francia, ‘The development of charters confirming exchange by the royal administration (eighth-tenth centuries)’, in K. Heidecker (ed.), Charters and the Use of the Written Word in Medieval Society (Turnhout: Brepols, 2000), pp. 43–62; see also Depreux, ‘Le souverain, maître de l’echange?’, in Fees and Depreux (eds), Tauschgeschäft und Tauschurkunde, pp. 45–64.


E X C H A N G E C H A RT E R S I N A S T U R I A S - L E Ó N

of Celanova and 62% of Portuguese cases.11 This may well indicate regional scribal practice. By contrast, the formulas which are characteristic of Catalan charters, cited by Professor Vones, do not feature;12 nor do the very common parique tenore or econtra in recompensationem found in many parts of Frankia.13 One might have thought that Astur-Leonese formulas would reflect the much cited provision of Visigothic law, that exchange should have the same security as sale, as the Catalan charters do, but they do not, even though other parts of Visigothic law are quite frequently cited, and sometimes, quoted, in the corpus as a whole.14 One could argue that the tiny number of texts using words of exchange and sale together are some kind of reflection of Visigothic law, but if so it is but a vague reflection. There is no reason to suppose, however, that exchange transactions had any less security than sale at this time. It is also notable that Astur-Leonese formulas do not reflect the exchange formula of the Ripoll formulary (from Catalonia) nor the two Visigothic exchange formulas in Gil’s collection, although some of his first formula is partly reflected in some later charters.15 It would be a mistake, in any case, to suppose that writers in this northern Iberian culture followed legal prescription in a literal way: where they used written law, they would modify and massage it to suit the purpose of the moment.16

11 Compare 42% of Sobrado cases, in northern Galicia. 12 See L. Vones, ‘Zwischen Tausch und Teilung: Besitz- und Herrschaftsstrukturen in Katalonien vom 9. bis zum 11. Jahrhundert’, in Fees and Depreux (eds), Tauschgeschäft und Tauschurkunde, pp. 451–70. 13 See H.-W. Goetz, ‘Die St. Galler Tauschurkunden’, and G. Bührer-Thierry, ‘De la traditio à la commutatio: sens et pratiques de l’échange à Freising du VIII au XI siècle’, in Fees and Depreux (eds), Tauschgeschäft und Tauschurkunde, pp. 171–200, 217–37; Cel370 (974) has in hanc tenore. 14 Forum Iudicum, 5.iv.1: Leges Visigothorum, ed. K. Zeumer, MGH LL nat. Germ. I (Hanover: Hahn, 1902), p. 218; even in this early, Visigothic, formulation, although the title is headed De conmutationibus et venditionibus, it is only the first of its twenty-two chapters that focusses on exchange. See further Vones, ‘Zwischen Tausch und Teilung’; also S. Esders, ‘Die frühmittelalterliche „Blüte“ des Tauschgeschäfts’, and U. Vones-Liebenstein, ‘Qui utiliter commutat, nullatenus alienat’, in Fees and Depreux (eds), Tauschgeschäft und Tauschurkunde, pp. 19–44, 427–50. 15 M. Zimmermann, ‘Un formulaire du Xème siècle conservé à Ripoll’, Faventia, 4:2 (1982), pp. 25–86, at p. 77; ‘Formulae Wisigothicae’, ed. J. Gil, in J. Gil (ed.), Miscellanea Wisigothica (Seville: Universidad de Sevilla, 1972), pp. 69–112, at pp. 99–100. Gil cites late eleventh- and early twelfth-century charters from Sahagún in the context of the first formula. 16 For discussion of this habit, see especially A. Iglesia Ferreirós, ‘La creación del derecho en Cataluña’, Anuario de historia del derecho español, 47 (1977), pp. 99–423, especially at pp. 215–29; this paper deals with Catalonia but its extremely subtle treatment of the difference between quoting, abbreviating and modifying the law in Catalan charters has implications for the whole of northern Iberia. See also I. Alfonso Antón, ‘Resolución de disputas y prácticas judiciales en el Burgos medieval’, in J. J. García González, F. J. Peña Pérez, L. Martínez García (eds), Burgos en la plena edad media (Burgos: Asociación Provincial de Libreros de Burgos, 1994), pp. 211–43, at p. 218; P. Martínez Sopena, ‘La justicia en la época asturleonesa: entre el Liber y los mediadores sociales’, in A. Rodríguez (ed.), El lugar del campesino. En torno a la obra de Reyna Pastor (Madrid/València: CSIC, Universitat de València, 2007) pp. 239–60, at pp. 242–3.



In these charters transactions are usually recorded in a one-sided way, as in the example quoted above: hence, ‘I, A, exchange X with you, B; and I receive Y from you, B, in return’; this was normally witnessed by A and others but not by B. There are some occasional variations to this pattern: there are no witnesses recorded for an early Liébana exchange; in a few cases B witnesses too; in one tenth-century Celanova case the format is ‘I, A, exchange X with you, B; I, B, exchange Y with you, A’; and in a later tenth-century Celanova example the format is largely inverted and there is supplementary information on the land transferred by B.17 On two occasions, two different charters survive for each transaction. In an exchange between two monasteries of a plot of arable for a vineyard on the meseta, the documents are of ‘mirror’ type with some slight variations.18 There are only two common witnesses; they have different scribes and are preserved in two different cartularies. It looks as if each monastery made its record, and was keen to keep it. Ego Iulianus abba . . . vobis Bellitus abba . . . Dedimus vobis terras in Valle de Morella . . . Nos Bellitus abba . . . dedimus vobis terras in Matella . . . Facta scriptura comutacionis sexto kalendas marcii, era DCCCCLX Iulianus abba hanc comutacionem a nobis facta. Donnadeo, Abhativa, Ranosindus abba, Salvatus confessor, Vimara presbiter, Salamon, Asco de Billaseca, Martinus, Mudarrefe, Manna. Salamon notuit.19 Ego Vellitus abba . . . vobis Iulianus abba . . . dedimus vobis terras in Matella . . . Nos Iulianus abba . . . dedimus vobis terras in Valle de Morella . . . Facta scriptura commutacionis sexto kalendas marcii. Vellitus abba in hanc comutacionis a nobis facta, Bonus, Lazaro, Salomoni, Pinniolus, Falgarosus, Cresceturo, Donnellus, Iohannes, Vimara, Vivitiza, Elyas, Felix, Maruan. Iustus presbiter notuit.20 In the second case, an exchange of estates and a church in the Cea valley, between King Alfonso III and three men, the records are of ‘mirror’ type and are essentially the same, although Alfonso’s charter has four additional witnesses and the spelling of proper names varies between the two; there could well have been two scribes, but both charters are preserved in the Sahagún cartulary.21 These two pairs constitute a very small proportion of the exchange documents that survive and the keeping of two records looks abnormal. However, there is a reference in a

17 T16 (884); Li185 (944), S102 (945), Cel370 (974), Cel336 (995) and perhaps Li69 (926); Cel370; Cel375 (997): A to B, B gives Y, A gives X, how B acquired Y, witnessed by B but not A. 18 See E. Huertas, ‘Des actes en miroir à Lucques au VIII siècle’, in Fees and Depreux (eds), Tauschgeschäft und Tauschurkunde, pp. 159–69, for ‘mirror’ types. 19 Li244 (24/2/952). 20 Li245 (24/2/[952]). 21 S9, S10 (909).


E X C H A N G E C H A RT E R S I N A S T U R I A S - L E Ó N

Portuguese donation charter to the fact that it was the custom to exchange charters (referring to a previous exchange between aristocrats, sometime before 959);22 it may be chance that so few pairs survive: perhaps there were once many more.

What was exchanged? The exchange format was used for swapping lands, whether tiny or huge, whether a discrete property or several parcels. Normally, if land was transferred in return for goods or metal, then the transfer was classified as sale, sale formulas were used and a price was stated.23 There are no estimates of comparative value in exchange charters and very few valuations cited in records of any kind of conveyance of property, in marked contrast to practice in some other parts of Europe.24 Of those rare valuations, one of the vineyards exchanged in 994 was valued at 5 solidi, an apparently unequal exchange between two lay parties;25 and a few of the southern Galician parcels listed in Celanova inventories were valued in modii – land at 15 modii, a house at 3 modii, a vineyard at 3 modii, a property at 7 soldos and 2 quartarii.26 There are also occasional references to assessment of ‘just price’.27 Clearly some people had a sense of value for property, and there were people available to make assessments, although we do not often hear of them: one Portuguese charter of sale of a half property provides for the other half of the property to be sold to the same purchaser if the vendor subsequently wanted to sell it and for boni homines to assess (apreciaberunt) a just price in that event; and one northern Galician exchange agreement provides for pervisores to assess the size of the lands exchanged so that one party could even them up – ‘peacefully’ – if necessary.28 In most cases we have no way of knowing if exchanges were of equivalent value or not: although many properties have boundaries detailed, they are unlocatable now because they are largely defined by 22 PMH DC 77 (959). 23 Exceptions are a sale between lay parties near Oviedo in the far north, in which small plots were transferred, apparently as part of the price, together with objects: a cloak and grain, together with one and a quarter orchards, Ov21 (978); and a sale of land to the monastery of Cardeña, for which the price paid was a vineyard and a valuable carpet, C199 (988). Arguably also another sale, for some arable, two fine furs and a fine horse (although treated as an exchange in this paper – see further in the following, n. 43), C31 (939). 24 One could read price as a kind of valuation, but price was clearly not the same. 25 Liii562. The solidus was a common unit of account, usually conceptualized as silver, perhaps influenced by knowledge of the key unit of the reformed Carolingian currency; neither the solidus nor any other unit of account was minted as coin in northern Iberia at this time. 26 Cel197 (975–1011), Cel368 (975–1009). The modius was also a common unit of account, of lower value, although probably also a standard measure, usually of grain. For discussion of these units of account, see ch. 6 following, ‘Sale, price and valuation’. 27 Cel180 (1010), for example; and see below. 28 PMH DC 79 (960); Sob122 (960). For boni homines as estimators, cf. I. Fees, ‘Tauschurkunden in Venedig und im Veneto’, and F. Bougard, ‘Commutatio, cambium, viganeum, vicariatio’, in Fees and Depreux (eds), Tauschgeschäft und Tauschurkunde, pp. 99–128, 65–98. See ch. 16 below.



reference to other people’s properties.29 However, given the availability of assessors for the Galician exchanges, we can reasonably suggest that there was some intention of equivalence. Reading the texts gives the same impression: a Castilian exchange has uno orto per ipso nostro and una terra per una nostra terra and an early Galician exchange has terra nostra concedo . . . et accepit de vos alia terra similem tanta.30 There are certainly no obviously unequal exchanges, to the benefit of a church body.31 However, there is the unequal exchange of vineyards between lay parties cited above and there is one case of an additional gift to a church, for a count gave five oxen and five cattle in calf to the bishop of Oviedo, on top of an exchange of lands between them. There are also very occasional sales in which the vendor was offered more payment than he thought necessary, with the surplus to remain with the church purchaser.32 The main objects of an exchange were always land but sometimes some movables were added; that they were added is a point that could be made explicitly: a goat insuper, as quoted below; another goat adisti nobis super mensuram; and a cow and two calves insuper adidisti michi.33 These extra movables are found in 23% of tenth-century exchange charters. They could be animals (a mule, a goat, a horse), objects (a shroud, wine, a mattress) or a valuation in silver, which may well in practice have been handed over as objects.34 It is particularly interesting that there are none from Galicia; one comes from Oviedo and three from Sahagún; but all the rest are from León collections (78%). In other words, this is a characteristic of transactions which took place on the meseta, not elsewhere, and overwhelmingly of transactions in and around León. The reason for adding these extra movables is rarely explicit and often ambiguous. In a few texts there is a comment that the movables were given in order to confirm the transaction and make it solid. Hence the following example, in which the lay couple who exchanged land with the abbot of Santos Justo y Pastor was given a goat explicitly in order that they confirm the exchange: 29 An attempt to assess comparative value was made by A. García Leal, in the case of the exchange transactions of two counts in the early eleventh century, concluding that the exchanges were of equivalent value: ‘Los condes Fruela Muñoz y Pedro Flaínez: la formación de un patrimonio señorial’, Anuario de estudios medievales, 36:1 (2006), pp. 1–110, at pp. 31, 78–9. 30 C36 (941), Floriano 121 (878), which repeats recepimus eius similem tanta. 31 Cf. Bougard, ‘Commutatio’, on transactions which were ‘melioratio au profit de l’église’. I do not agree with the editor of Ov23 (982), who interprets pariatio in a very difficult text as a reference to compensation to even up the value of lands exchanged with a priest; rather, this seems to refer to fines, i.e. gifts made to cover outstanding debts. Liii591, however, refers to 2 solidi given in meliora to a lay party; this was clearly an extra, on top of an exchange; see further in what follows. 32 OvC33 (991); S73 (939): ‘donastis nobis in xii solidos et prendimus inde xx argenteos et quod restat dimisimus vobis pro remedio anime nostre vel parentibus nostris’, from a lay family to the monastery of Saelices on the sale of a water course in the Cea valley. What seems to be implied here (very unusually) is that the church negotiator arrived to pay with a bag of silver; the vendor took some pieces and sent the rest back. 33 Lii325 (960), Li161 (943), Li206 (949). 34 For example, Li153 (942): mule; Lii340 (961): mattress; S177 (960): 3 argenteos.


E X C H A N G E C H A RT E R S I N A S T U R I A S - L E Ó N

In Dei nomine. Ego Ermorigus et coniuge mea Lupa vobis Iulianus abba et collegium fratrum Sanctorum Iusti et Pastoris, salutem. Placuit nobis, bone pacis volumptas, ut comutaremus vobis terra quem abemus in villa vocitata Matella . . . [boundaries] Et accepit de vos alia terra iusta villa ipsa Matella . . . et affigit in termino de fratres; ipsa terra integra, secundum illa obtinuit frater Manilani. Et pro confirmandam hunc comutacio insuper accepi de vos, ego Ermorigus, narrunu cum quornua quatuor. Ita ut unusquisque quod accepit firmiter obtineat per secula cuncta.35 We find the same kind of movables given, as countergifts, to secure some straightforward gifts.36 In the case of the countergifts, the gift of animals was very common (horses, mules, sheep, cattle) and that of clothing or textiles was not uncommon (cloth, pelts, cloaks, a mattress, a carpet).37 Nearly half of the seventy or so recorded countergifts were explicitly given so that the donor of the primary gift would confirm his or her donation and make it secure, in order to prevent, for example, reversion to the family of the donor.38 Hence: Et pro confirmandam hanc nostram preceptionem accepimus de te offertionem cavallo nobis placivile, where King Vermudo II receives a horse from a trusted aristocrat to whom he had given an estate.39 So, confirmation is one explanation for the addition of movables. Of the other exchange cases, one extra was clearly an additional gift to the church – that cited above of cattle to the bishop of Oviedo. One can also suggest, given the clear intention that some swaps should be of equivalent value, that some extras were added in order to even up the transfers, particularly when what was transferred was a composite of small parcels – like the five plots of arable, two meadows, garden, enclosure and three houses in Villobera which were exchanged for arable in two places near León, and a colt, in 954.40 One can also make a case that some of them were ‘sweeteners’, given to encourage one party to make the exchange, particularly where one set of lands exchanged was adjacent to monastic property and the monastery gave the extra. Half of the León cases of added movables concern exchanges between a lay party and an abbot; two more concern exchanges between other clerics and an abbot; and two concern exchanges between a lay party and a priest (as also do two of the examples from Sahagún). Nearly all the movables

35 Lii325 (960). 36 For countergifts see ch. 8 following, ‘Countergift in tenth-century northern Iberia’, and ch. 6, pp. 100–01. 37 For example: S196 (962), S293 (978), Liii576 (997) (animals); Li168 (943), C54 (945), Lii265 (954), S178 (960) (objects). 38 Note similar practice in tenth- and eleventh-century Italy: C. Wickham, ‘Compulsory gift exchange in Lombard Italy, 650–1150’, in W. Davies and P. Fouracre (eds), The Languages of Gift in the Early Middle Ages (Cambridge: Cambridge University Press, 2010), pp. 193–216. 39 Liii530 (989). 40 Lii275.



come from abbots (or occasionally priests) to the other parties. In view of their direction of movement, from ecclesiastic to other party, I think it likely that in practice some of these were gifts to persuade the party to go ahead with the transaction. Abbots, especially, were in the business of amassing and consolidating properties and it is not improbable that they set about it by distributing presents. There are also four cases of the gift of such additional movables in transactions between lay parties, so this does not have to have been a mechanism peculiar to the church.41 In at least two of the latter the donor of the extra consideration was a high aristocrat; high aristocrats presumably had similar interests to abbots in amassing and consolidating property and the gifts may well have been made for similar reasons. Overall, however, since the framing of so many of these texts is unspecific, several could have been made for any of the above four reasons. There is a further complication: although it is extremely inconsistent, there is some tendency for the extras to be called ‘price’ and for transactions with extras to be described at some point as an exchange and sale. (The text might read comuto vel vendo or kartula vendictionis sive comutacionis once or twice but not throughout.) Some of these are in fact composite transactions, recorded in a single document – literally both an exchange and a sale of different sets of property, like the two exchanges of lands in Villobera for others in Villacete and Valderaduey, together with sale of a house and enclosure in Villobera for silver, recorded in one text of 936.42 But others are not, or apparently not, and it looks as if to some writers – though not all – the handing over of an additional movable on top of the swap of lands conveyed the idea of sale. This addition of vendere or venditio only ever occurs in exchange charters if there was an extra movable; and by no means all exchange charters with movables have this feature.43 It is therefore tempting to suppose that these transactions with extras were simply commercial sales with a price. It is clear, however, that most writers in tenth-century Iberia saw sale and exchange as different transactions: land was sold in return for a price; land was exchanged for other land. Given the large number of sale records, exchange was certainly not some kind of substitute for sale, as it arguably was in some other parts;44 nor was the occurrence of exchange transactions anything to do with weak monetization: there were vastly more sales than exchanges, despite 41 Li78 (928), Lii388 (965), Liii591 (986–99), S353 (997). 42 Li106. 43 C31, of which there are two versions, is a possible exception (see above, n. 24). Both versions occur in the Cardeña cartulary, although the monastery of Cardeña was not a party to the transaction; no scribes are named. The parties occur in the same order in each charter (i.e. A gives X to B and receives Y from B), but the first in the cartulary (fo. 34r) uses sale words and has four additional witnesses, while the second (fo. 90v), which is in effect an abbreviated version of an otherwise identical text, uses exchange words. I have treated this as an exchange charter but it is possible that it was in the first instance conceived as sale and subsequently rewritten as an exchange. One could argue this in either direction. 44 See Esders, ‘Die frühmittelalterliche „Blüte“’, and Bougard, ‘Commutatio’; contrast the tiny number of sales in, for example, Saint Gall records (Goetz, ‘Die St. Galler Tauschurkunden’).


E X C H A N G E C H A RT E R S I N A S T U R I A S - L E Ó N

the absence of coin. For the most part writers chose to formulate the record of the relatively small number of land swaps in a different way. That making such a distinction was a deliberate choice is also indicated by some inventories of accumulated property, such as those emanating from the monastery of Celanova late in the tenth century. In these inventories property is listed in the context of its mode of acquisition: in effect, the inventories are summaries of charters, stripped of initial and final protocol, stuck together in a long continuous narrative of acquisition. Some of these properties were acquired as gifts; many were acquired by purchase; but a few were acquired in exchange. The distinction between sale and exchange, and the relatively low occurrence of exchange, is maintained even in this kind of text. For example, one list notes twenty-two sales, nineteen gifts (of which six were to deal with debts and twelve were to meet the obligations of judicial fines) but only two exchanges.45 The same point is made by incidental comments on the past history of property bought or sold: the property given in 959 which had previously been exchanged between two aristocrats; another given in 968, which had been acquired in an exchange with a sister-in-law; the latter mentioned again in 983, as well as another acquired by purchase and exchange.46 There is also one particularly instructive case, which again reinforces the point, in which one party to an exchange between lay persons asked for and received it back – redescanbiavimus; he agreed, with two sureties appointed to make sure he kept his word, that if he wanted to get rid of it again he would only do so to the party to the initial exchange.47

Who was involved in exchanges? The people involved in exchange transactions were of many kinds. Almost a third of tenth-century exchanges were between lay parties, a considerably higher proportion of lay participation than we find in many western European collections;48 over a half were between a lay and an ecclesiastical party; and the remainder were between ecclesiastics. In fact, we can break this down further into four kinds of activity, which are status- and management-related. Firstly, in 20% of cases, peasants exchanged small plots with other peasants, and with local priests, in what looks like the normal to-and-fro of agricultural life; there was probably some pressure here, some consolidation there, some practical convenience, some social and economic climbing. Such parcels are small: Hego Elpericus et Braolio tibi Vistremiro. Placuit nobis . . . ut comutaremus tibi nostras porciones in Uxio; or Placuit mici 45 46 47 48

Cel368 (975–1009). PMH DC 77, 99, 138. Lii477 (pre-980). Contrast M. Stofella, ‘Gli atti di permuta nella Toscana occidentale’, and B. Kasten, ‘Tausch- und Prekarieurkunden in Lotharingien bis 1100’, in Fees and Depreux (eds), Tauschgeschäft und Tauschurkunde, pp. 129–57, 326–80.



volumtas . . . ut facerem vobis karta concambiacionis de omnem mea ereditatem que abeo in villa Talecias . . . Et acepimus de vos alia terra in Talecias, in loco predicto in Campizo . . . de termino de pumare maiore, et per termino de Asela, et per termino Dagamiri, et afige in termino de Sisverti; or Alia hereditas que nobis ibidem dedit Eita et uxor sua Sabegoto et ipsi abuerunt eam de contramutatione quam contramutaverunt cum Zallico pro alia quam ei dederunt in Billi.49 Secondly, monasteries exchanged with peasants, and with aristocrats too, clearly to consolidate properties and create larger workable space, in other words for reasons of land management on some scale (42% of cases). What was exchanged in these cases often lay ‘by the monks’ land’ or ‘by the church’, and where there were several exchanges in the same district the plots were often adjacent, as in Marialba (near León) in the 920s and 940s.50 Half of all the exchange documents in León collections relate to the monastery of Abellar, and they run from 926 to the 990s: Abellar seems to have been using exchange as a strategy to accumulate property. Thirdly, aristocrats also exchanged with peasants, for similar reasons of consolidation. There are fewer of these cases (about 6% of the total) but Bagaudano and his wife did so in the Liébana in the 930s; Count Hermenegildo did so in northern Galicia at the same time; and Count Munio Fernández did so in Valdevimbre much later in the century.51 The remaining kinds of activity are those of monasteries swapping with other monasteries; monks with monks and other clerics; aristocrats with other aristocrats. Each of these groups occurs in relatively small numbers and the swaps may well have been largely for practical convenience: for example Bishop Rosendo swapped estates with his sister and her husband in 949 (here and elsewhere it is clear that the rationalization of inherited portions within families is one factor stimulating some of the exchanges).52 Table 5.3 Lay and ecclesiastical parties to ninth- and tenth-century exchanges Parties to the exchange

Number of ninth-century exchanges

Number of tenth-century exchanges

Lay/lay Lay/bishop Lay/abbot or monk Lay/priest or deacon Abbot/bishop or priest Abbot/abbot or monk Monk/monk Priest/deacon


25 9 26 10 3 5 1 1


49 Li23 (909); Li204 (949); Cresconio will, Cel180 (pre-1010). 50 Li69 (926), Li70 (926), Li78 (928), Li161 (943), Li185 (944), Li195 (947), Li209 (950). Cf. Depreux, ‘The development of charters confirming exchange’, p. 52. 51 T41 (932); Sob111 (936); Liii562 (994), Liii591 (986–99). 52 Cel357; cf. aristocratic family swaps cited in PMH DC 88, 99, 138 (c. 964).


E X C H A N G E C H A RT E R S I N A S T U R I A S - L E Ó N

Table 5.4 Parties to ninth- and tenth-century exchanges, by status Parties to the exchange

Number of ninth-century exchanges

Number of tenth-century exchanges

Peasant/peasant Peasant/priest Peasant/monastery Peasant/aristocrat Aristocrat/aristocrat Aristocrat/monastery Aristocrat/bishop Monastery/monastery Other ecclesiastical


11 5 23 5 10 10 7 6 3

1 2

Looking at status in this way, we can note that over half of exchanges involved peasant transactors and over a third involved secular aristocrats.

The eleventh century While it would be impractical to attempt to survey all collections into the twelfth century, I have taken an overview of the León material, as the largest collection, and of that from Otero de las Dueñas, as an unusually important collection from a lay archive. In both, the form and words used for exchange documents are essentially the same as those used in the tenth century. For Otero de las Dueñas, from which no exchange charters survive from the tenth century, there are nine exchanges in the period 1002–1039, that is 4.6% of all charters of 1001–1050; there is none from the second half of the century. The main verb used for exchange is concambiare, but half of these charters are unusual in being in sale format, with the exchanged lands presented as price. There are no additional payments or ‘sweeteners’ of movables. Almost all are exchanges between lay parties and all involve the Flaínez comital family, some transacted with peasants and some with other aristocrats.53 Clearly this is a very small group, but it is useful in emphasizing the fact that aristocrats did make exchanges, just as monasteries did, when accumulating property. It is particularly interesting because a high proportion of the total archive could have survived (the material comes from the Flaínez family archive). The eleventh-century collections from León are large, comprising at least 711 charters. There are thirty eleventh-century exchange charters, twenty-one from 1001–1050 (4.4% of earlier eleventh-century charters) and nine from 1051–1100 (3.9% of later eleventh-century charters). Two thirds use the verb comutare and 53 OD62 (1002), OD69 (1006), OD74 (1007), OD82 (1009), OD97 (1014), OD172 (1025), OD173 (1025), OD188 (1029), OD224 (1039). OD69 is the exception to the lay exchanges: one party was a priest, probably a local priest.



the remaining third uses concambiare; several are rather wordy texts and one includes a very long narrative.54 In two cases, in 1073 and 1092, there survive pairs of documents for a single transaction, both involving the bishop of León; they vary considerably and the later pair, both of which have the same scribe, has largely different witnesses.55 Eight of these thirty charters have additional payments: they use the same kinds of object as the tenth-century gifts of movables but all date from before 1028, and all are transactions involving either the king or an abbot/abbess. The additional movables appear to be functionally similar to the tenth-century movables (and indeed look like a continuing tenth-century practice). The transactors are of several types, but proportions of each group are similar to those of the tenth century: a third are transactions between lay parties; nearly a third are between a lay party and an abbot; there are six exchanges between a lay party and a bishop, and one between a lay party and a priest; and the rest are transactions between ecclesiastics. The bishop of León is a much more prominent party from the 1030s (and all the verbose texts are records which involve the bishop). Much of this looks like a continuation of tenth-century practice, with exchange as both a mechanism that any landowner could use, for convenience, and also one strategy among many for those accumulating property, depending on local circumstances. However, after the first decade we see much less of monasteries exchanging lands with peasants; that phase and aspect of accumulation appears to have finished.

The place of exchange transactions I have pointed to small surviving numbers of charters but use of a standard form across a very large area. This probably in itself implies that there were once rather more exchange charters; add to that the Portuguese reference to the exchange of charters as a norm when property was exchanged; and bear in mind the number of exchanges between peasant lay parties in León collections, preserved by chance, and the number of exchanges between aristocratic lay parties in Celanova and Portuguese collections, preserved because of strong aristocratic connections with Celanova and Guimarães.56 All this suggests that we may have quite a small proportion of the exchange charters originally produced. The survival rate may perhaps be low because many lay exchange documents were not preserved in any archive and because ecclesiastical exchange (records of which have a greater chance of survival in an archive) depended on the accumulation strategy the ecclesiastical body preferred. Different monasteries used different strategies: whereas Abellar clearly used exchange in the tenth century, Sahagún overwhelmingly used

54 Wordy texts: Liv942 (1035), Liv1188 and Liv1189 (both 1073); Liii737 (1015). 55 Liv1188 and Liv1189; Liv1268 and Liv1269. 56 PMH DC 77; see above.


E X C H A N G E C H A RT E R S I N A S T U R I A S - L E Ó N

purchase; Celanova used purchase for properties near Celanova but exchange for properties farther away, in Quiroga and Portugal. So, the survival of documents depends on the nature of the archive, on the strategic approach to property accumulation and – as ever – on chance. The exchange charters show two main, and rather different, processes: firstly, lay people swapping with other lay people, as part of normal agricultural life, as had perhaps gone on for centuries; secondly, ecclesiastical (and to a lesser extent aristocratic) accumulation of property – an extra-normal process, because accumulation was not a steady constant and tended to happen in phases – with the initiative coming strongly from the accumulators, as sometimes emphasized by the extra gifts of movables. In the former case the exchanged properties tended to be near each other but others could be many kilometres apart. Comparison between Iberian patterns and those of other parts of western Europe is interesting. There are some obvious similarities, for example in the basic form of the records and in the occurrence of ‘mirror’ pairs, but plenty of contrasts. Numbers, and proportions, of exchange charters are tiny by comparison with those from some of the Italian and German collections, although the proportion of exchange between purely lay parties in northern Iberia is surprisingly high. This was clearly not a practice deliberately designed to meet an ecclesiastical need, particularly given the occurrence of exchanges between peasants in some of the earliest surviving charters. There is therefore no sense in this material that exchange was a mechanism designed to overcome prohibitions on alienating church property; in practice there seem to have been no such prohibitions – perhaps in itself a reason for the relatively low number of exchange charters. There was also much less concern to record estimates of value than found in some parts of Europe, although many people clearly had a sense of value and there were ways of making assessments; and we find very little in the way of deliberately unequal exchanges for the benefit of a church. It is possible that some of the movables added to exchange transactions were intended to benefit a church in this way, although some of those extras were clearly added for quite different reasons, from a desire to get confirmation of title by an alienator to the desire to persuade a perhaps reluctant party to go ahead with an exchange. The additions were made in less than a quarter of surviving records of exchange, and there are some ambiguities in about 10% of exchange records over whether the addition of a movable turned the exchange transaction into sale: they remain ambiguities because these records are not explicit enough to know what was in the mind of the contracting parties. What is clear, though, is that the addition of movables in exchange transactions only happened in and around the city of León, and not across great swathes of rural Spain and Portugal; in other words, it happened in the most commercial of northern Iberian regions. This may simply be because movables, and increasingly silver, were more readily available in this area. It is also clear that the practice of adding movables did not expand in the rapid commercialization of the eleventh century; indeed, after 1028 it seems to have petered out, particularly as the consolidating activity of certain abbots came to an end. Exchange, then, 91


was a useful mechanism for some restricted purposes but it was not the preferred method of transferring property in this area in this period; gift and sale were much more common. In the transfer of property northern Iberia was noticeably different in some respects from other parts of western Europe; these seem to have been real differences in practice and not simply differences in documentation.57

57 I am extremely grateful to Irmgard Fees and Philippe Depreux for initiating this study and to conference participants for their helpful comments; I am also indebted to François Bougard and Paul Fouracre for their comments on subsequent drafts of this paper.


Part II

Production and exchange


It is in the tenth century that charter material suddenly becomes plentiful in northern Spain.1 Asking the questions of these charters that one often asks of such texts, it immediately becomes apparent that several different systems of valuation were in daily use in the area: solidi, tremisses, modii, quarters, cattle, silver pieces. What could these terms have meant to people on the ground? Were they realities, solidly related to well tried systems of exchange? Or did they belong to the realm of ideas or to the restricted world of notarial jargon? Further, was this use of different valuation systems characteristic of most of northern Spain – Christian Spain – or was it a relic of some Atlantic backwater, steeped in archaic practice, cut off from the thriving and rapidly growing economies of the Mediterranean? Indeed, do these systems of valuation and the practices to which they relate have anything at all to do with the economic mainstream of western Europe? This paper sets out to explore these issues and to consider their implications for exchange, exchange mechanisms and the market in Christian Spain at a time when contemporary narratives were overwhelmingly preoccupied with the theme of conquest and reconquest. To begin with some essential context: following the rapid Muslim conquest from southern Spain in the eighth century, the tenth, eleventh and early twelfth centuries saw the to-and-fro of ‘Christian’ campaigning from the north and Muslim campaigning from the south.2 Discussions of issues of ethnicity and religious persuasion, of the reality of either conquest or reconquest and of the key stages in the formation of the polities that emerged have been many, but it remains the standard view that much of the Duero basin was incorporated within the Christian kingdom of León by 900, while the southern part of the basin and the upper waters of the river were incorporated in the late eleventh century, leaving the Ebro valley to the east within Muslim hands until the mid-twelfth. Meanwhile, the kingdom 1 First published in W. Davies, Early Medieval Europe, 11 (2002), pp. 149–74. 2 For example, T. F. Glick, Islamic and Christian Spain in the Early Middle Ages (Princeton: Princeton University Press, 1979); H. Kennedy, Muslim Spain and Portugal: A Political History of al-Andalus (London: Routledge, 1996); B. F. Reilly, The Contest of Christian and Muslim Spain 1031–1157 (Oxford: Blackwell, 1992), for broad general surveys in English.



of Asturias, in and beyond the mountains in the north and north west, and that of Pamplona (Navarre) in the north and north east, sustained the Christian rulerships that appear to have controlled those areas since the early eighth century in the former case and the early ninth in the latter. The Duero and Ebro river systems are crucial to the story, and the mapping of now Muslim, now Christian, control by tradition follows their lines. It is also part of the dominant tradition that the Duero basin was depopulated as a consequence of Muslim conquest, repopulation following in the wake of successful and sustained Christian reconquest in the later ninth, tenth and subsequent centuries.3 Whatever the dominant historiographical trends, it does seem clear that there was frequent military campaigning across the basin in the tenth century, especially in the first and last thirds of the century: Christians were campaigning in southern Spain as early 913 and 914, while Muslims were on the Duero at Toro in 916 and in Burgos in 931 and were campaigning as far north as Astorga and León in 995 and Santiago in 997. The reality of depopulation, however, is very questionable, and has been challenged in recent years.4 The real issues in fact centre on the density and distribution of population in the phase before ‘Christian reconquest’ and colonization, not on total desertion. It is in any case perfectly clear that people were living in and going about their business in the Duero basin throughout the tenth century.

3 See for example, C. Sánchez-Albornoz, Despoblación y repoblación del valle del Duero (Buenos Aires: Instituto de Historia de España, 1966); id., ‘Repoblación del Reino Asturleonés’, in his Viejos y nuevos estudios sobre las instituciones medievales españolas, 3 vols (Madrid: Espasa Calpe, 1976–80), vol. 2, pp. 581–790 (first published 1971); C. Estepa, ‘La vida urbana en el norte de la Península Ibérica en los siglos VIII y IX. El significado de los terminos “ciuitates” y “castra”’, Hispania, 38 (1978), pp. 257–73, at pp. 257–61; J. Mestre Campi and F. Sabaté, Atlas de la ‘Reconquista’. La frontera peninsular entre los siglos VIII y XV (Barcelona: Península, 1998), p. 14; Reilly, Contest, pp. 28–9; R. Collins, ‘The Spanish kingdoms’, in T. Reuter (ed.), The New Cambridge Medieval History, vol. III c.900–c.1024 (Cambridge: Cambridge University Press, 1999), pp. 670– 91, at p. 673. 4 See J. Escalona Monge, ‘Transformaciones sociales y organización del espacio en el alfoz de Lara en la Alta Edad Media’ (Doctoral Thesis, Complutense University Madrid, 1996); M. C. Rodríguez González and M. Durany Castrillo, ‘Ocupación y organización del espacio en el Bierzo entre los siglos V al X’ and J. M. Mínguez, ‘Continuidad y ruptura en los orígenes de la sociedad asturleonesa. De la villa a la comunidad campesina’, both in Studia Historica, 16 (1998), pp. 45–87, 89–127; J. L. Quiroga and M. R. Lovelle, ‘Ciudades atlánticas en transición: La “ciudad” tardo-antigua y alto-medieval en el noroeste de la Península Ibérica (s.v-xi)’, Archeologia Medievale, 26 (1999), pp. 257–68; S. Castellanos and I. Martín Viso, ‘The local articulation of central power in the north of the Iberian peninsula (500–1000)’, Early Medieval Europe, 13 (2005), pp. 1–42. The ideology of desertion occurs early in Spanish texts, at least from the ninth century and arguably from the eighth, and profoundly influences all subsequent historiography; some tenth- and eleventh-century charters also pick up this theme; and without question grants are recorded in charter texts ad populandum; see Sánchez-Albornoz, ‘Repoblación’, for references, but note the cautionary words of C. Estepa Díez, Estructura social de la ciudad de León siglos XI–XIII (León: Centro de Estudios e Investigación ‘San Isidoro’, Archivo Histórico Diocesano, Caja de Ahorros y Monte de Piedad de León, 1977), pp. 72–3; see also above, ch. 1.



Much of Spain is a land of high plateau; mountain ranges project above it, fragmenting this plateau: Pyrenees in the north east, Cantabrian mountains in the north, separating the narrow coastal strip of the north coast from the Duero basin, León mountains in the north west, separating Galicia from central Spain, the Iberian system bounding the Ebro valley on its south and west sides, the Central system fringing the south of the Duero basin. While the high rainfall of the northwest makes Galicia wet and green, much of the Duero basin (modern Castile-León) falls on the plateau, the meseta, which is hot and arid in summer.5 The detail of locality and transaction is the subject of what follows, but it is worth remembering the context sketched here – both the long hot summers, and often dry winters, and the fact that a campaigning army might sometimes come rushing past, as was the case in Sahagún in 988 and León itself in the same year and in 995. Both points are very easy to forget when concentrating on the bustle of the market.6

The material Although this material has often been discussed,7 there are still aspects to explore, particularly in respect of transaction patterns and regional difference.8 For comparative purposes there are three monastic collections which include substantial and very useful tenth-century material: that of Celanova, in Galicia in the west (not far from the present-day northern border of Portugal), that of Sahagún on the plateau of central León, and that of Cardeña, just south-east of Burgos, beneath the sierras that separate the Duero basin from the Ebro valley to the east.9 Although 5 M. Terán and L. Solé Sibarís (eds), Geografia general de España (Barcelona: Ariel, 1978). 6 On which, classically, see C. Sánchez-Albornoz, Una ciudad hispano-cristiana hace un millenio (Buenos Aires: Nova, 4th [rev.] edn, 1947); also Estepa, Estructura social de la ciudad de León. 7 For example, C. Sánchez-Albornoz, ‘El precio de la vida en el reino asturleonés hace mil años’, in his Viejos y nuevos estudios, vol. 2, pp. 811–52 (first published 1944); id., ‘Moneda de cambio y moneda de cuenta en el reino asturleonés’, in Viejos y nuevos estudios, vol. 2, pp. 855–83 (first published 1960); E. Sáez, ‘Nuevos datos sobre el coste de la vida en Galicia durante la Alta Edad Media’, Anuario de Historia del Derecho Español, 17 (1946), pp. 865–88; P. Beltrán Villagrasa, ‘Dinero de vellón de Fernando I el Magno, en la colección “Los Arcos”’, in his Obra Completa, II. Numismatica de la Edad Media y de Los Reyes Catolicos (Zaragoza: Universidad de Zaragoza, 1972), pp. 585–604 (first published 1952); J. Gautier Dalché, ‘L’histoire monétaire de l’Espagne septentrionale et centrale du IXe au XIIe siècles: quelques réflexions sur divers problèmes’, Anuario de Estudios Medievales, 6 (1969), pp. 43–95; cf. C. Wickham, ‘Pastoralism and underdevelopment in the early middle ages’, in L’uomo di fronte al mondo animale nell’alto medioevo, Settimane di Studio sull’alto medioevo, 31 (1985), pp. 401–51, at pp. 434–48. And see now E. Manzano and A. Canto, ‘The value of wealth: Coins and coinage in Iberian early medieval documents (9th–11th centuries)’, in S. Barton and R. Portass (eds), Beyond the Reconquista: Constructing and Representing Social Power in Medieval Iberia, c.700–1300 (Leiden: Brill, forthcoming 2020). 8 Gautier Dalché, ‘L’histoire monétaire’, is however a valuable pioneering paper on regional differences; recently, I. Martín Viso, Poblamiento y estructuras sociales en el norte de la Península Ibérica (Salamanca: Universidad de Salamanca, 2000), has a significant comparative section. 9 Cel, S, C.



they have fewer tenth-century records, there is relevant material in plenty of other collections too, particularly from Valpuesta, in the mountains to the north of Burgos, and from San Millán in the rich agricultural lands of the modern Rioja to the east.10 On a different scale is the enormous collection of material from León cathedral, although – as an episcopal collection including records from small rural monasteries which subsequently came into the hands of the bishopric – it is rather different in kind.11 I intend to focus on the comparisons between the three large monastic collections, because of their size and their reference to three different kinds of landscape, but will take note of others (especially the three aforementioned) where relevant. 222 transactions are recorded in tenth-century Celanova material, 424 in that of Sahagún, and 215 in that of Cardeña – i.e. more than 850 in all (comparative numbers for Valpuesta are 45, San Millán 123, León over 600).12 Since transactions are relatively rarely recorded in the first decade of the tenth century, this analysis will focus on the years 910–99 inclusive, by decade. In fact, however, the generation 930–60 AD is the first period which is well evidenced enough to allow significant comparison. The number of transactions peaked in the 950s (20% of all transactions from the six collections mentioned above), thereafter tailing off to the 990s (in which 8% occurred); more than half of all transactions (51%) took place between 940 and 969 (see figure 6.1). 350 300 250 200 150 100 50 0 910s




All transactions

950s All sales





All donations

Figure 6.1 The incidence of all transactions, sales and donations, 910–99, in the six collections

10 SM; V. 11 Li, Lii, Liii. 12 Some texts record more than one transaction; hence, transaction numbers do not mirror the numbers of records.



The transactions recorded in these texts include donations, confirmations of donations, settlement of disputes, exchanges, sales and manumissions. Not surprisingly, since these are mainly monastic collections, donations (often, but not invariably, to the monasteries) are common, constituting 45% of all transactions recorded in the six collections across this period. However, there are relatively few in the 960s (37% of transactions in that decade) and relatively many in the 990s (62% in the decade). Sales are even more frequently recorded (47% of all transactions), with relatively many in the 930s and 960s (61% and 59% respectively of transactions in those decades) and notably few in the 920s and 990s (31% and 29% respectively). However, there was considerable regional variation in sales per decade: of the three main collections, there were many sales in Sahagún but rather few in Celanova and Cardeña in the 940s; absolute numbers of Celanova and Cardeña sales peaked in the 960s, as did the proportion of Celanova sales to donations, but the proportion of Cardeña sales did not peak until the 980s (see table 6.1). Correspondingly, particularly few donations are recorded round about Sahagún in the 930s and 940s (17% and 21% of transactions in those decades) and round about Celanova in the 960s (18%); however, there were relatively many donations in and around Cardeña in the 940s right through to the 970s (60%–72% of transactions per decade). León patterns were very similar to those of nearby Sahagún but Valpuesta and San Millán have few sales recorded throughout this period (22% and 16% of all transactions) and clearly had different kinds of socio-economic relationships with their hinterlands. Sahagún, then Celanova, then Cardeña, by contrast, appear to have embarked on policies of acquisition at different points in the tenth century, building up their properties through deliberate purchasing.13 Table 6.1 Sales as a percentage of all transactions, by collection, by decade

910–19 920–29 930–39 940–49 950–59 960–69 970–79 980–89 990–99


Cardeña Sahagún León San Millán Valpuesta All six Total collec- number tions of sales

0 33 33 37 35 80 18 29 27

33 38 47 22 22 39 21 59 55

75 22 83 70 51 63 47 62 17

35 31 65 73 62 63 49 46 33

0 0 13 3 35 0 0 0 12

33 50 67 0 17 33 25 0 0

36 31 61 48 46 59 39 48 29

18 23 93 100 145 165 65 89 37

There are, then, some striking facts that emerge, even from the basic statistics: more transactions – of all kinds – are recorded for the thirty years from 940 than 13 See J. M. Mínguez Fernández, El dominio del monasterio de Sahagún en el siglo X (Salamanca: Universidad de Salamanca, 1980).



from before or after (or both put together); and far higher proportions of sales are recorded for the 930s and the 960s than at other periods. The year 930 is a clear point of change. The particular decades marked by sales partly reflect regional differences: exceptionally large numbers at central Sahagún for the forty years from 930 and exceptionally large numbers at western Celanova in the 960s; but neither the totality of transactions nor the proportion of sales steadily increased across the century. It is the mid-tenth century that is marked by exceptionally high levels of recorded activity.

Sale and price There are at least 735 exchanges from this period in these collections which are recorded as commercial transactions. Most of these transactions are explicitly conceptualized as ‘sale’, with the active verb usually vendere, to sell, the document sometimes called a cartula vendicionis, charter of sale, and a price, pretium, being handed over. Hence, ‘we sell you a villa in Fontasquesa and received a price of one bay cow’ or ‘a price which satisfied me’.14 Occasionally the words emtores (buyers), comparare (to buy) and comparatio (a purchase) are used too – in references to land bought in the past or bought by charter;15 and there are references to the mercimonium/a (commerce/deals) by the cemetery at Cea and elsewhere.16 The language of sale was different from that of gift; things were handed over on condition that a specified, measurable price was received; failure to produce the price, at the right time, would invalidate the transaction. Some things that were handed over during or after transactions were not recorded in terms of price: a carpet and a coat given in recognition of the confirmation of a grant in 957; linen cloth for an act of corroboration in 960; a best mule in respect of a grant made in 962.17 In such cases not only was the language different but the timing of the return and the level of obligation to make it were different from the timing of the price payment and the necessity of making it. They are best seen as countergifts – presents, gestures of appreciation – and were clearly not a necessary condition for completion of the main transaction. In fact Cardeña scribes, though sometimes using the language of sale, were more prone to use the words donare, tradere and concedere (give, hand over, concede) than vendere, and more frequently recorded countergifts than did other scribes. Countergift was not the same as price: the latter was defined and agreed in advance,

14 S34: ‘vendimus . . . et accepimus de vos in precio una vacca laura’; S53: ‘et alio precio que michi complacuit’. 15 C25, S159, Cel168; S146 (cf. S125, S130). 16 S130; cf. ‘qui hanc mercatum fecimus’, of a sale between two lay parties (V36) and the mercato of ‘Sile’ (S155); also S214, S315, SM115. Cf. expressions like mercedem recipio ante Deum (‘I receive reward/favour before God’), C27, C33. 17 V35, S178, S196; cf. S26, S285, S302.



and it was essential to meet it in full (hence – so often – the record reads ‘I have accepted the price and nothing remains’).18 The price that was paid was sometimes expressed in terms of silver but frequently included or was entirely composed of other objects. Miscellaneous objects are cited as elements of price paid nearly 500 times in sales from the three main monastic collections (some prices cite none, some cite several) and they constitute the majority of price expressions; in all six collections considered here nearly half (49%) of all prices are cited in terms of objects only; 41% in terms of silver or gold currency only; and 10% in terms of both. For example, some land in Lemos was sold in 942 for a goat and three quarters of grain; a couple sold a vineyard at Canaleja in January 950 for 15 solidi of silver; and a man sold land on the River Cea in 955 for 20 solidi, a pair of shoes and two quarters of grain.19 Proportions went up and down through the century until use of miscellaneous objects without any precious metals dropped to 32% of expressions of price in the 990s (see table 6.2). Table 6.2 Percentage of prices cited as miscellaneous objects, or silver ‘currency’, or both, in six collections, by decade

910–19 920–29 930–39 940–49 950–59 960–69 970–79 980–89 990–99 910–99

a) Misc. objects

b) Silver currency

c) Both

Total no. of prices citing a) b) or c)

47 56 51 51 57 52 40 33 32 49

35 24 39 41 31 40 54 59 53 41

18 20 10 8 12 8 6 8 15 10

17 25 90 112 124 157 68 63 34 690

Again, there are some very clear regional differences: before 930 price was usually expressed in Sahagún texts in terms of farm produce or clothing, especially in 18 ‘et de ipso pretio aput vos nichil remansit sed completum est’, Cel237 (934); ‘et de pretio aput vos devitus nicil remansit’, S120 (950). See the excellent paper by W. I. Miller, ‘Gift, sale, payment, raid: Case studies in the negotiation and classification of exchange in medieval Iceland’, Speculum, 61 (1986), pp. 18–50, on the difference between gift and sale; and for a subtle exploration of reciprocities initiated by gift – and sometimes sale too – in Galicia at a slightly later period, A. Rodríguez López and R. Pastor, ‘Reciprocidades, intercambio y jerarquía en las comunidades medievales’, Hispania, 60 (2000), pp. 63–101; cf. R. Pastor, E. Pascua Echegaray, A. Rodríguez López, P. Sánchez León, Transacciones sin mercado: instituciones, propiedad y redes sociales en la Galicia monástica 1200–1300 (Madrid: CSIC, 1999), pp. 17–22. See further ch. 8, ‘Countergift in tenth-century northern Iberia’. 19 Cel451, C69, S149.



animals, which were sometimes given a valuation in silver solidi but more often not; and it was once expressed as a silver vessel.20 Thereafter price was often expressed in terms of solidi until the 970s, when this mode of expression came to predominate. Although the number of cases is smaller, Cardeña transactions more often used solidi to express price before 930, as they did from the 960s, but frequently used produce and objects in the 930s and 940s. Half of all tenth-century cases also had a valuation in solidi attached; records of Cardeña sales were more likely to use the concept of the solidus than records from the other collections, at any point in the tenth century.21 Celanova transactions, on the other hand, used metal concepts only rarely and they never predominated in this period. The contrasts between the three are therefore stark: western Celanova essentially nonmetal-based; eastern Cardeña more metal-based than not; and central Sahagún changing its practice across the century (see table 6.3).22 Beyond these three the Valpuesta zone, in the mountains of the north north east, was similar to Celanova: until the 960s only objects were cited as payment, and items were valued in solidi exceptionally rarely;23 thereafter metal concepts occasionally feature. Round San Millán, to the east, use of metal concepts was even more common than it was round Cardeña. And in and around León, just as it was in nearby Sahagún, practice was mixed but tended to change in favour of metal concepts as the century approached its end. Table 6.3 Percentages of ‘currency’-only expressions in all prices in the Celanova, Cardeña and Sahagún collections, by decade

910–19 920–29 930–39 940–49 950–59 960–69 970–79 980–89 990–99




0 20 22 20 0 0 0 38 0

100 67 38 13 25 63 60 63 60

0 0 41 36 35 38 55 74 33

20 S1–5, S11, S14, S15, S17–19, S34, S35. 21 C1, C3–4, C7, C9, C13 etc. 22 The numbers of instances in 910s, 920s and 990s are too low to be statistically significant (two, three and four respectively, of a total 148). 23 V27 and V29. Cf. the increasing references to metal payments in Navarre in the eleventh century, M. A. Zamanillo Arizabalo, ‘Circulación monetaria y sistemas de pago en Navarra en los siglos X a XIII’, in Primer congreso general de historia de Navarro (Pamplona: Institución Principe de Viana, 1988), pp. 239–45, at p. 243.



Three quarters of the objects that were handed over as price in the three main monastic collections – almost invariably as the price of landed property, in fact – were farm produce (livestock 37% of citations, skins 4%, dairy produce 2%, beverages 11% and bread or grain 21%); cloth, coverings, bedding and clothing were also, however, a significant proportion (a further 16%). In addition, silver dishes feature a few times, a precious sword and a brooch once each, saddles and harness now and then and bags of spice (sumac) occasionally.24 Horses feature frequently (23% of livestock); sheep, cattle and oxen also commonly occur (24%, 20% and 19% of livestock respectively); goats, asses and mules from time to time; and dogs and pigs rarely. Sometimes a long list of objects makes up the price: for a sale in Spinareto in 955, six measures of grain, bedding, a tunic or shirt, two rams, four cheeses; for a sale of land on the River Esla in the early 980s, bedding, an ox harness, two oxen, ten sheep and a skin; and from mountain Valpuesta, for a sale in Pando in 929, a belt, books, a cloak, linen, (?) cups, a hat, napkins, an ox, two green shirts and one white, and a pelt.25 Many of these objects are found throughout the century – the bedding, cloth, horses, oxen and cattle, and grain – but sheep and very precious objects tend to come earlier and harness items later. Probably more significant than the temporal variation is the regional: all three locations used the presumably prestige goods of bedding, clothing, skins and horses; but lengths of cloth and beverages were much more characteristic of Celanova than elsewhere and cheese, cattle, oxen, goats and rams were more characteristic of the Sahagún area than of other regions. Celanova texts often cite unspecified things (res, perhaps cattle); indeed, the formula ‘cibaria et bibere, res et pannos’ (in effect, food and drink,? stuff and clothes), and variations on it, frequently occurs.26 Grain and wine sometimes dominated the objects cited as payment, regardless of the formula: of the sixty-one objects separately itemized in payments during 961 and 962, almost two thirds (62%) involved the transfer of grain and wine. The fact that people near Celanova were twice as likely to use grain as livestock, and people near Sahagún nearly four times more likely to use livestock than grain, must be significant and – directly or indirectly – must have a bearing on farming régime (see table 6. 4).

24 Dishes S24, S26, S47, S55, S78, Cel254, Cel380; sword S166, brooch C43; harness etc. C23, C42, C50, C93, C96, C111, C113, S139, S188, S244, S302, S318, S359, S364; folle de zumake S36, S66, S237. Also books, V13. Cf. Sánchez-Albornoz, ‘El precio’, pp. 818–19. 25 S148, S302, V13. 26 E.g. Cel399. See above, p. xiii.



Table 6.4 Percentage of all citations of objects as price, 910–99, per monastery, by object

Livestock Skins Grain, bread Beverages Cloth Coverings Clothes Harness




17 2 32 21 16 2 1 0

53 4 15 5 2 5 5 4

29 8 7 5 5 7 19 13

Precious metals Though silver was the preferred metal when currency notions were employed, gold is very occasionally mentioned; for example, in a corrupt reference to payment of three morabitinos, arab gold coins, in 940 in Pegariña, in the Celanova orbit.27 Gold factiles occur with a sword and there are a few very general references to ‘silver and gold’, as in 956.28 Gold is also frequently cited in the sanctions attached to San Millán, Cardeña and Sahagún charters, right through the ninth, tenth and eleventh centuries. These sanctions express penalties for contravention of donations as well as of sales: the penalty is sometimes worded in terms of the doubling (or tripling) of the value of the transaction; sometimes it is cast in purely religious terms; and sometimes it is expressed in terms of metal, that is gold and silver. In San Millán sanctions pounds or talents of gold are the norm, from the early ninth to the early eleventh century (solidi occasionally from the mid-tenth century): 7 golden pounds (863), 5 talents of gold (938), 10 pounds of gold (953) and so on, but 100 solidi (947).29 Cardeña sanctions take many forms but frequently use precious metal expressions; pounds of gold (occasionally talents) are common until the eleventh century but solidi of silver become more frequent from about 950; there are also occasional references to solidi of gold.30 Sahagún sanctions nearly always use the double value expression following sales but metal-based expressions are common for donations, with use of both silver and gold. Clearly echoing the way they expressed payments, Celanova and Valpuesta texts tend to use non-metal sanctions, and Valpuesta material is especially prone to

27 Cel40; cf. renders expressed in morapetinos longos, Cel158. This is characteristic of the twelfth century, from 1112 (post-Almoravid attacks), and one therefore doubts the anthenticity of the record; see the lists published by Sánchez-Albornoz, ‘El precio’, Sáez, ‘Nuevos datos’, Gautier Dalché, ‘L’histoire monétaire’. 28 S153, S166; V12. 29 SM6, SM26, SM66; SM47. Cf. F. Mateu y Llopis, ‘Las clausulas penales pecuniarias de los “documentos para la historia de las instituciones de León y de Castilla (siglos X–XIII)”’, Anuario de Historia del Derecho Español, 23 (1953), pp. 579–92, at pp. 580–5, on their derivation from Visigothic legislation. 30 C1–4 etc.; C62, C64, C67 etc.; C36; cf. V1 (falso).



cite those of religious connotation, though gold features occasionally.31 It is extremely unlikely that gold was used in practice at this time for even in the following century, when kings were receiving huge tributes in gold, it does not feature much in transactions.32 Its use here is more likely to be formulaic. Sanctions requiring payment in gold often specify that this was due to the king: gold was explicitly associated with high status people and was the appropriate way to express payment to one who was king.33 References to silver payments are frequently ambiguous. They take the form solidi de argento/argenteo or solidi argenti (seventy-eight cases in prices in the three monastic collections, and more in valuations) and solidi alone (sixty-eight cases). Despite its late Roman origins as a gold coin, the solidus was a major feature of the reformed Carolingian silver currency, as the key unit of account: silver denarii were coined at the rate of 12 to the solidus and 264 to the pound weight. In Muslim Spain there was silver coin too: silver dirhems, coined from the eighth to tenth centuries, circulated widely in Europe.34 But coin was not minted in northern Spain until the eleventh century. The meaning of the word solidus in these texts has therefore provoked much discussion. Sánchez-Albornoz writes as if many late Roman coins, Frankish coins and silver dirhems too were in circulation in León in the tenth century, and more recent writers, while modifying this position, have also suggested the availability of contemporary foreign coinage at this time.35 Very few coin hoards are known from northern Spain but it is intrinsically likely that some Arab coins and some Frankish denarii found their way to Zamora and León and even Burgos: a San Millán charter of 986 records a gift of 150 solidi argenti to Muslims in ransom and farther north pilgrims on the route to Santiago clearly introduced some coin, like the small hoard of Anglo-Saxon pennies of c. 990 found in a pilgrim hostel at Roncesvalles.36 However, it must also be the case that some prices expressed in solidi were expressions of the value of the price and not expressions of the nature of the things handed over: when Godesteo and his wife Ledegundia paid two solidi for half a vineyard on the River Esla, they may well have actually paid goods to the value of two 2 solidi, neither coin nor silver.37 That this was common practice, even in coin-producing northern Europe, is implied by Charlemagne’s Saxon capitulary

31 V46 (975); cf. Cel223 (935), Cel116 (944), Cel403 (964), Cel175 (974), Cel354 (989); Celanova also – exceptionally – used a cattle fine as sanction: Cel494 (935), Cel421 (997). 32 J. J. Todesca, ‘Means of exchange: Islamic coinage in Christian Spain, 1000–1200’, in L. J. Simon (ed.), Iberia and the Mediterranean World of the Middle Ages (Leiden: Brill, 1995), vol. 1, pp. 232–58, at pp. 234, 249–50. 33 For example, S75, S206, SM48. 34 Todesca, ‘Means of exchange’, pp. 235–6. 35 Cf. Sánchez-Albornoz, Ciudad, p. 50, and Todesca, ‘Means of exchange’, for references to the literature; at p. 235, for example, Todesca states that the solidus argenti was a ‘Frankish unit of account’, sometimes paid (in Spain) in Carolingian coin and silver dirhems. 36 SM99. Alberto Canto (personal comment) notes hoards from Pamplona and Soria but also that there are very few from north of the Duero. F. Mateu y Llopis and R. H. Dolley, ‘A small find of AngloSaxon pennies from Roncesvalles’, British Numismatic Journal, 27 (1955 [for 1952–54]), pp. 89–91. 37 Lii355.



of 797, where variant solidus equivalents are cited.38 The solidus was a unit of account; it is extremely likely that some or many prices quoted in this fashion were paid in miscellaneous goods, notwithstanding the arguments of Sánchez-Albornoz.39 We really have little idea how many might have been paid in silver and how many in goods; however, where there are clearly changing and consistent trends in expressions of price, as noted above, this may well reflect some change in practice. Bearing these warnings in mind, nevertheless sometimes it is perfectly clear that silver was handed over in payment (apart from the silver artefacts – dishes, harness trappings – that occasionally occur): for example, in the Valdeduey valley in 935 a payment of 40 solidi was received, 20 in silver (‘in argento’) and 20 in cloth; and in 980, near the River Porma, a price of 160 solidi in pure silver was paid (‘argentum purum solidos CLX’).40 Sometimes it is explicit that the solidus was a measure of weight: ‘a silver dish weighing 5 solidi’, ‘a table service weighing 35 solidi’.41 At other times, when the expression is solidi de argento or solidi argenti rather than plain solidus, it is not unreasonable to suggest that silver was probably used here, by weight. Hence, ‘IIIes solidos de argento’, 3 solidi of silver, in 936, 14.5 in 937, 3 in 951, 15 in 997, 23 in 999 and so on.42 This silver is likely to have been bullion, probably of mixed quality and in mixed forms (including foreign coins). The solidus therefore had wide use both as a measure of silver by weight, and also as a unit of account. Over three quarters of payments expressed in metal terms from the three monasteries involve solidi. Apart from the reference to morabitinos,43 most of the remainder (21%) comprise arenzos, argenteos, argenzatas (rare), argenzeos and argenzos (forty cases). The argenzos (to use a single term, for ease of reference) occur from 930 until the end of the century but there are again some very significant regional variations in their occurrence: nearly all the cases come from Sahagún;44 there are none from Celanova, where reference to metal was in any case exceptionally rare, although there are a couple of valuations in argenzos – the silver dish exchanged in 942 in Amoroce was valued at ‘5 solidos and 3 argenteos’ and three oxen exchanged in 991 in Orga were valued at ‘18 solidos and 7 arienzos’;45 and there are two abbreviated cases in the Cardeña collection, in 981 and 993 (and a valuation).46 León material is very

38 ‘Note should be taken of the proper equivalents of the Saxon shillings [solidi]. . . . Of oats, the Bortrini must give 40 bushels for a shilling, and of rye 20 bushels; those to the south, however, must give 30 bushels of oats for a shilling and 15 bushels of rye’, Capitulare Saxonicum, ch. 11 (Boretius no. 27), translated in H. R. Loyn and J. Percival, The Reign of Charlemagne (London: Hodder & Stoughton Ltd, 1975), p. 56. 39 For example in his ‘Moneda’, p. 878, where he argues that all such references are to silver. 40 Li103; S309. 41 ‘scala argentea appendente Ve solidis’, S55, ‘servitio de mensa pensante solidis 35’, S183. 42 S36, S65, Cel430, C210, SM116. 43 See above. 44 S37 (930), then S46, S36 etc. 45 Cel254, Cel248. 46 C181 ‘VIIII solidos et II arcs’, C205 ‘XII arcs de argento’; C123 ‘in IIII arcs’ (all expanded in the Martínez Díez edition).



similar to that from Sahagún: references to argenzos (more than fifty in payments in our period) are present already in 910, run through to 999 and are noticeably common between 928 and 966.47 There are also references in a composite San Millán text of the 940s,48 and in three Valpuesta texts, between 960 and 975.49 The argenteus was a late Roman silver coin, of the early fourth century, equal to 25 denarii, and of which there were 24 to the pound.50 There is no need to suppose that the term argenzos used in tenth-century Spain refers to currency of Late Antique origin, but since the Roman solidus influenced the terminology of the tenth century the same might be argued of argenzos. That the word signified a concept used to express value is explicit in the Celanova and Cardeña cases cited above and occasionally in Sahagún charters.51 This was more common in León material, particularly in the 930s and 940s.52 A case has also been made for argenzos as a type of currency. Gautier Dalché thought the word referred to Arab silver dirhems and Carolingian denarii in use in northern Spain, and Beltrán that it referred to Arab dirhems used by weight.53 Textual references are certainly cast in a form which suggests that argenzos could be physical objects, and not simply a mental construct: payments were expressed in argenzos alone (2, 4, 21.5 and so on), in argenzos in combination with solidi (15 solidi and 2 argenteos), as were valuations, and in combination with other objects (a cow and 4 arenzos).54 In one case a value is quoted as 12 solidi and the text continues ‘we have taken 20 argenteos [from the 12] and what remains we leave to you for the health of our and our family’s souls’;55 the implication is that argenzos were the things actually used to make the silver payment expressed in solidi (as also that the vendor received some but not all of the price, but was content to leave it at that – 20 argenteos was less than 12 solidi).56 It is extremely difficult to see argenzos here as anything other than objects: the simplest explanation is that they were pieces of silver. They may indeed have been Arab or Frankish coins, used by weight: irregularly cut fragments of Arab dirhems do after all occur in Spanish dirhem hoards.57 In fact, the mode of reference to argenzos – 21.5, 20 argenzos as less than 12 solidi and so on – could even suggest a standard (or

47 48 49 50 51 52 53 54 55 56

Also an isolated reference in 899, Li14. SM59. V36, V37 and V46. A. H. M. Jones, The Later Roman Empire, vol. 1 (Oxford: Basil Blackwell, 1964), pp. 438–9. S327 (984). Cf. OD24, of 980. Li104, Li129 (argencios – in a charter surviving as an original) etc. Gautier Dalché, ‘L’histoire monétaire’, pp. 46–7; Beltrán Villagrasa, ‘Dinero de vellón’, p. 597. S37, S46, S82; S60; S94. S73. Cf. 8 argenzos = 1 solidus argenti, in a San Millán text of 1037, Beltrán Villagrasa, ‘Dinero de vellón’, p. 600 – but statements of equivalence are many and various. 57 M. Blackburn, ‘Money and coinage’, in R. McKitterick (ed.), The New Cambridge Medieval History, vol. II, c.700–c.900 (Cambridge: Cambridge University Press, 1995), pp. 538–59, at pp. 550–1. I am extremely grateful to Mark Blackburn for his personal comments. See also above, n. 36.



near standard) weight, something immediately recognizable by use of the term. Is it possible that cut dirhems occurred in the North too, and were even cut to a standard weight, just as finely cut up dirhems occurred in southern Scandinavia in the tenth century, although it has to be said that coin finds of any kind are exceptionally rare in northern Spain?58 and in any case, dirhems apart, is it possible that pieces of silver of any origin were cut to standard weights, as was common in Scandinavian contexts at this period (sometimes stamped)?59 The development of ‘metal-weight economies’ in the tenth century is characteristic of northern Europe in the zones beyond those that used coined money, particularly in the Scandinavian and Irish worlds; in some cases silver ingots were used as a means of payment; finds of balances for weighing the silver occur in many archaeological contexts; and several weight systems appear to have been in use.60 The contexts are so similar – in the availability of silver and the absence of local minting – that some comparable ‘metal-weight’ economic system is worth considering for the León/Sahagún zone, where argenzos feature frequently. There were therefore distinct regional differences in references to metal currency as well as consistent changes over time. Broadly, there was a stronger awareness of gold at San Millán than elsewhere, a sense that was sustained across two centuries; Cardeña moved from a preference for expressions using gold in the ninth and early tenth centuries to those using silver from the mid-tenth; Sahagún was more silver- than gold-conscious throughout the period (and apparently low value silver, at that) and in essence seems to have reflected the practice of León; while mountain Valpuesta and western Celanova made much less use of precious metal notions throughout. The differences are encapsulated by the widespread use of religious penalty clauses at Valpuesta.61 The use in practice of pieces of silver called argenzos in the tenth century seems to have occurred but was far from standard: they are especially characteristic of the Sahagún and León regions, both covering the hinterland of the city of León. Their

58 Blackburn, ‘Money and coinage’, p. 544. I am very grateful to Anna Balaguer for her comments on finds. 59 Much of the fundamental work on Scandinavian weights and silver has been done by H. Steuer; see for example, his ‘Geldgeschäfte und Hoheitsrechte in Vergleich zwischen Ostseeländern und islamischer Welt’, Zeitschrift für Archäologie, 12 (1978), pp. 255–60, at p. 258; id., ‘Feinwaagen und Gewichte als Quellen zur Handelsgeschichte des Ostseeraumes’, in H. Jankuhn, K. Schietzel, H. Reichstein (eds), Handelsplätze des frühen und hohen Mittelalters (Weinheim: Acta Humaniora, 1984), pp. 273–92, at pp. 276–7. 60 See S. E. Kruse, ‘Ingots and weight units in Viking Age silver hoards’, World Archaeology, 20 (1988), pp. 285–301, especially p. 286; ead., ‘Late Saxon balances and weights from England’, Medieval Archaeology, 36 (1992), pp. 67–95, especially pp. 79 and 88. Susan Kruse makes useful points about the interaction between ‘metal-weight’ and ‘coined-money’ economies: what happened in northwest England, where ingot-using Irish, coin-using English and bullion-using Scandinavians intermingled, is a good case; ‘Viking age silver ingots from England and Wales and their economic implications’ (PhD thesis, University of London, 1988), especially p. 231. 61 The match is not perfect, however, for Celanova texts also frequently use the ‘doubling’ sanction.



appearance at Sahagún coincides precisely with the point at which sales became more frequent there – as they were to remain for the next sixty or so years. It is particularly interesting that in the Cardeña region – not far from the city of Burgos – where it was common to express price in terms of silver, references to solidi were far more common than argenzos, even in the late tenth and eleventh centuries. This suggests a different kind of urban hinterland, one where they were using the unit of account but without readily available pieces of silver. The much more rural Celanova and Valpuesta regions, where even general references to gold and silver are rare and where solidi only feature exceptionally, provide a different kind of contrast. Argenzos are best explained as a fairly small measure of silver, perhaps of consistent weight but not necessarily size or shape. It is appropriate to find such material in sub-urban contexts near León; it is not surprising that we cannot find them in rural Celanova and Valpuesta nor in rich San Millán (where we know more about high-level than peasant transactions in the tenth century). Their rarity round Cardeña, in the clearly silver-conscious hinterland of Burgos, is – on the face of it – unexpected.

Valuation The solidus was also used as a measure of the value of objects in expressions like the following: ‘duas pelles obtimas in XX solidos’ (two best skins worth 20s) (939); ‘equa in IIIIor solidis et IIIIor solidos de civaria et vino’ (a horse worth 4s, 4s-worth of grain and wine) (949); ‘sex solidos in vino’ (6s-worth of wine) (961); ‘uno tapete ualente V solidos argenti’ (a carpet worth 5s of silver) (988); or even ‘duos solidos de arenzos et ceuaria in alios duos solidos’ (2 solidi of silver pieces and grain worth another 2s) (966) or ‘uno bove et Ia vaka et Io carnero et Io sestario de cibaria; totum in XII solidos’ (an ox, a cow, a ram and a sester of grain, all worth 12s) (965). High-value expressions like ‘spata obtima cum factiles deauratos valente solidos Cm’ (a best sword with golden ?ornament worth 100s) (959) or ‘cavallo castaneo apreciato in LXa solidos’ (a chestnut horse valued at 60s) (941) are less common;62 compare an ox worth (ualente) 8s in 975.63 Interestingly, a higher proportion of Cardeña objects used in payment was valued in solidi than proportions from other areas – 48% of all objects cited as price paid had a solidus valuation as did 96% of all the valuations, as against 22% and 96% respectively from Sahagún, and 16% and 23% respectively from Celanova.64 Given Charlemagne’s Saxon capitulary, with its precise statement about the difference between the Bortrini’s solidus valuations and those of the ‘southerners’, we probably should not expect the valuations in northern Spanish material to be 62 C31, S111, S185, C199, V37, S225, S166, S76. 63 V46. For values, see the long treatment of Sánchez-Albornoz, ‘El precio’, who expects consistency; nowadays we would not make that assumption and would expect variation. 64 Not all objects used in payment carried a valuation; hence the differences between the smaller percentages of objects cited as payment and the greater percentages of all valuations.



consistent.65 The very concept of Galician solidi in Celanova material – as in ‘solidos galicenses, solldos gallicenses’ – and solidi ‘of the present’ in Cardeña in 984 makes the point.66 Some items vary widely, with horses valued from 4s to 300s; obviously there were differences in quality, as is explicit in references to ‘good’ and ‘best’ horses and implicit in the colours sometimes specified (a bay horse, a chestnut horse and so on), but this may well not account for the whole variation.67 Valuation in the region dominated by Celanova was expressed with much greater variety and complexity than it was in the other collections considered here – with ‘une sorte d’anarchie monétaire’, as Gautier Dalché put it.68 In fact, expressions of price in the Celanova region were explicitly referred to three kinds of standard – those of solidi, as elsewhere, of cattle and of measurements by weight or volume – as well as very occasionally to a fourth, argenzos, and implicitly to a fifth, the tremissis. Only 16% of items valued in the Celanova region between 910 and 1000 AD were valued by solidi (although the practice is found from the 920s on). Examples include a horse worth 4 solidi, wheat worth 1 soldare, a silver dish worth 5 solidi, some silver articles (reiteles) worth 100 solidi.69 Use of the solidus to express the whole price (for example, a vineyard sold in 983 for 16 solidi) was even rarer and rarer than occurrences in the other collections considered here: only 9% of prices in the Celanova collection in this period are expressed solely in solidi; less than 3% of all the prices reviewed in this paper that are expressed solely in solidi come from Celanova.70 There are variations on the term solidus, as shown in the examples cited above, the word sol(i)dares being nearly as common, accounting for almost half of the 16% cited. Although at first sight puzzling, the word appears to mean ‘something worth a solidus’: ‘something weighing a solidus’ would not make sense, given valuation of animals in this way. The comparability of solidus and solidare seems to be explicit in ‘VIII soldares kalicenses pretium iustum, id est bove uno, vacca

65 See above, n. 38. See also Sáez, ‘Nuevos datos’; Gautier Dalché, ‘L’histoire monétaire’, pp. 49–51, pace Sánchez-Albornoz, ‘El precio’, p. 826, and elsewhere, who looks constantly for consistent economic explanations (the rise and fall of prices, the bearing of the frontier on price and so on). 66 Cel182, Cel342, Cel348, Cel384 (cf. solidos calicenses and solidos calizenses in the inventory of Cresconio’s property of c. 1005, Cel204); C191 ‘centum solidos de presenti’. ‘Galician’ solidi not Gallic (i.e. Frankish), as some earlier commentators thought; see Sánchez-Albornoz, Ciudad, p. 51, n. 77; id., ‘La primitiva organización monetaria de León y Castilla 887–928’, in his Viejos y nuevos estudios, vol. 2, pp. 894–5, and ‘Moneda’, pp. 862–3. 67 S111 and Cel176 (4s); S318 (300s); S19, S26, S261, S313 (good/best). It may not be significant but a few items do show a rough consistency: skins, for example, tend to be valued at between 3 and 7 solidi, while ‘best skins’ are valued at 10 solidi (e.g. C31). 68 ‘L’histoire monétaire’, p. 46. 69 Cel176, Cel500, Cel380, Cel234. 70 By the year 1000, however, inventories of the wealthy Cresconio have a higher proportion of valuations in solidi, though valuing by modius and quarter was common and by cattle occasional; Cel204; cf. Cel197.



I, in cibaria et vino et pannos in solidos VI’, given that ox and cow are elsewhere ‘vacca soldare, boves solidares’; hence, this appears to mean ‘[sold for] a just price of 8 Galician soldares, that is for an ox, a cow and 6 solidi-worth of grain, wine and cloth’.71 Valuation by tremissis (the late and post-Roman third of a solidus) was rare and mostly used for cloth in these charters: lentio tremisale (937), linteo tremisale (961), lenzio tremisale (967).72 This presumably means ‘a tremissis-worth of cloth’.73 Although it is found infrequently, the locations of these transactions are very scattered, occurring both near Celanova itself and to the north by the River Sil, perhaps implying an earlier widespread Galician (?or Asturian) system of valuation. The second explicit mode of valuation, and by far the most common, was to use measures of weight or volume – pricing, as it were, ‘by the sackful’, by implication a standard sack. 42% of all items used in payment in this region at this period have a valuation of this kind (55% of all valuations), and it occurs from 934 to 999. The concept does occur occasionally in Sahagún and León texts, as also in charters from Otero de las Dueñas 30km to the north of León, other north-western texts and Portugal.74 It does not occur in the Cardeña collection at this period. The usual measure was the modius, though quarters sometimes feature. While some usages of the terms modii and quartarii are most obviously understood as measures, since they are specified item by item – ‘3 modii and 2 quarters of wheat, 7 modii and 2 quarters of rye’, ‘1 modius of wheat, 3 modii of rye, 6 quarters of millet, 4 sesters of wine’, both paid in 96175– others only make sense as a valuation. Hence: ‘for a total price of 16 modii’, ‘worth 9 modii and 2 quarters’, ‘12 modii-worth of grain, drink and cloth’, ‘an ox and a cow, with its calf, priced at 24 modii’, ‘a cloak at 4 modii’.76 Wine, for example, was measured in sesters but valued in modii: ‘12 modii-worth of bread and wine’ but ‘2 sesters of wine, 1 quarter of rye . . . in total worth 6 modii and 2 quarters’.77 And there are some prices simply expressed in modii (11% of all Celanova priced sales), just as some are 71 Cel182; Cel176, Cel504. See further in what follows. 72 Cel565, Cel397, Cel164; also Cel223, Cel406; also a pig, Cel409 (990). 73 Beltrán Villagrasa, ‘Dinero de vellón’, p. 598, cites eleventh-century texts on equivalences of the tremissis. 74 S148, Lii404; OD10, OD11, OD47 (a small but interesting group of tenth-century charters from this monastery in the foothills of the Cantabrian mountains). See the published lists of Sánchez-Albornoz, ‘El precio’, Sáez, ‘Nuevos datos’, Gautier Dalché, ‘L’histoire monétaire’, for other north-western locations; M. del Pilar Laguzzi, ‘El precio de la vida en Portugal durante los siglos X y XI’, Cuadernos de Historia de España, 5 (1946), pp. 140–7. 75 Cel395, Cel390. 76 ‘sub uno pretio XVI modios’, Cel166; ‘modios VIIII quartarios II valentem’, Cel393; ‘in XII modios inter cibaria bibere et pannos’, Cel388; ‘bove et vacca cum sua agnicula [sic] preciatos in XX et IIIIor modios’, Cel402; ‘saia de IIII modios’, Cel410. In the examples given here some valuations use modius and quarter together; a few texts value in quarters without modii: e.g., Cel412, Cel386, Li200, Lii297, OD5, OD17, OD45. 77 ‘in panem et vino modios XII’, Cel326; ‘vinum sestarios II, centeno quartario Io . . . sub uno modios VI quartario IIo’, Cel406.



simply expressed in solidi: an inheritance sold for 5 modii in 957, land sold for 3 modii in 964, vineyards sold for 12 modii in 962, half a house sold for 8 modii in 954.78 Here it is clearly again an abstract construct of valuation. There are a few occasions on which price is expressed as a combination of solidi and modii, almost as if the modius was perceived as a fraction of the solidus: for example, ‘wine and grain, cloth and goods, to the value of 2 Galician solidi and 4 modii’.79 One case, of 960, appears to relate the value of the modius directly to that of the solidus, at 9 to 1: ‘(sold) for 5 modii of grain, 4 modii of drink and 9 modii of cloth and goods, in total 2 solidi’.80 It is extremely unlikely that this was consistently the case across north-west Spain, and the calculations are inconsistent with some direct statements of equivalences.81 We should expect variation. The third construct of valuation explicitly used in these Celanova texts was a cattle standard, found in a few transactions dating between 950 and 997 – rarer even than the solidus – and applied to 8% of the items used in payment and 13% of valuations in our period (although there are a few more from the early eleventh century). The meaning, however, is perfectly clear: ‘a skin worth 2 oxen’ in 950, and in 991 ‘three horses, one at 15 oxen, another at 5, and the third at 7’.82 Again, by implication, Celanova people had in mind a ‘standard’ cow for the purpose of valuation. There also occurs the somewhat surprising concept of the ‘vacca soldare’ and ‘bos solidare’, the solidus cow and ox. The price paid for some land in 941 was a bay-coloured cow (‘vaca soldare colore laurea’);83 the total value of a price paid in 950, which included mules, a cow, grain and cider, as well as the skin worth two oxen cited above, was ‘18 boves solidares’. While there is sometimes a suggestion that the ox was worth a solidus – as noted above – that clearly was not always the

78 Cel65, Cel379, Cel400, Cel73. Cabra modiale and ovelia modiale in prices presumably indicate a goat and sheep worth 1 modius, as also understood by Sáez, ‘Nuevos datos’, pp. 886–7; Cel451; Cel390, Cel403, Cel397. 79 ‘vinum et cibaria pannos et res in soldos gallicenses IIos et modios IIIIor’, Cel384; cf. Cel544. This is in fact a frequent practice in the inventories of Cresconio’s property, of the early eleventh century (Cel197, Cel204). 80 ‘cibaria modios V bibere modios IVor res et pannos modios VIIII sub uno solidos IIos’, Cel378. See also above, n. 78. 81 Sánchez-Albornoz made the equation 1 solidus = 1 sheep = 1 modius, on the basis of León texts of 951 (1 sheep = 1 modius) and 1008 (100 sheep = 100 solidi), Ciudad, p. 41, n. 46; also ‘El precio’, pp. 812–13, and ‘Moneda’, pp. 865–6. Quite apart from the fact that no direct relationship is cited between solidus and modius in these two texts, separated by fifty-seven years, the proposition is not borne out by the totality of available evidence, which makes it perfectly clear that one item could have varying values. Cf. above n. 80 for the solidus and modius relationship. Others have made the same point (Sáez and Gautier Dalché notably), but their criticism was not accepted by Sánchez-Albornoz. 82 ‘pelle valente duos boves’, Cel504; ‘caballos IIIes, uno de boves XV, alio de boves V. Ille tertius de boves VII’, Cel248. Cf., in 1001: ‘kavallo de sex boves adpreciatum, et duas equas adpreciatas in VII boves’, Cel555. 83 Cel452; cf. Cel176. See also Gautier Dalché, ‘L’histoire monétaire’, pp. 49–50.



case if three oxen could elsewhere be valued at over 18 solidi.84 The underlying concept of vacca soldare seems to be that of ‘a standard ox’, an ox appropriate for the purposes of accounting. More puzzling at first sight is the notion of the ‘agreeable’ ox or cow: a price of grain and cloth paid in 956 was valued at ‘7 boves placabiles’ and another in 981, which included a measure of wheat, was priced as a ‘vacca placibile’.85 The meaning here is presumably something like ‘agreed to the value of 7 (standard) oxen’ or ‘agreed to the value of a (standard) cow’. Even more interesting is the fact that those who framed valuations near Celanova could clearly use different standards of value at one and the same time. The 991 transaction which values horses in terms of cattle goes on to value three oxen at ‘18 solidos and 7 arienzos’, just as a transaction in Rabal in 961 valued animals and a covering in cattle but a silver dish in solidi: ‘cavallo de VII boves equa de duos boves tappere de III boves scala artentea de V solidos’.86 This is partly explicable by reference to different kinds of object: near Celanova silver dishes were valued in solidi, skins in cattle and sheep in modii. However, such a neat system was not consistently sustained. Horses could be valued by a cattle standard but they were valued in solidi too; grain and cloth were valued in three and four systems respectively; oxen could be valued in solidi and modii.87 It is tempting to suggest that different practice could be the result of local regional variation, since use of the cattle standard occurred in a confined area within about 10km of Celanova itself. However, since the very same places that used the cattle standard also used solidi and tremisses occasionally and modii often, it is quite explicit that different modes of valuation were used simultaneously.88 The same is true of Sahagún and Otero de las Dueñas, using solidi, modii and occasionally argenzos standards; and to a much smaller extent of urban León, where valuation was normally expressed in terms of silver (solidi especially but argenzos from time to time). In the absence of further evidence, the best explanation for the simultaneous use of several systems seems to be to do with total value – solidi for high-value things, modii for medium-value, cattle for lower-value – but the evidence is inconclusive. To the east, whether at Cardeña near Burgos, or San Millán in the Ebro valley, or Valpuesta in the mountains, valuation was almost entirely expressed in solidi in the tenth century.89 Here is a major difference in the construction of valuation within northern Spain at that

84 Cf. S195, an ox at 4 solidi; S241, an ox at 6 solidi; S229, three oxen at 12 solidi; C184, two oxen at 20 solidi. 85 Cel391, Cel571. Cf., for example, ‘in precio boue placiuile . . . et porcu placiuile’, in 986, OD26. 86 Cel380. 87 The Portuguese material cited by Laguzzi values cattle solely in modii, while other items (e.g. horses, cloth) were valued in modii and solidi; Laguzzi, ‘El precio’, p. 146. 88 Sánchez-Albornoz, ‘El precio’, p. 851, cites a few other cases of the cattle standard scattered across Galicia – in Santiago, Lugo, Sobrado. 89 A fact which raises the possibility that the more complex western systems were in some respects relics of an earlier age, as is in any case suggested by the use of archaic tremissis. Sánchez-Albornoz thought that use of the old systems continued into this period.



time: crudely, the farther west you travelled, the greater the number of valuation systems you would find in simultaneous use.

Thoughts for the future Tenth-century patterns are interestingly and surprisingly different. Of the three collections that have been the focus of this paper, that of Sahagún has the clearest development across the period. Lying in the Duero basin, in the hot, dry meseta and in the economic hinterland of the city of León, it was actively making acquisitions, especially in the period 930–70, both in the north/south valleys that cut the plateau and in the mountains to the north, cannily establishing an interest in much of the available arable.90 People of this area primarily used metal-based notions of value and they sometimes used metals to make payments; in mid-century these were often pieces of silver – argenzos – operating, as in León, as a kind of pre-currency. Where people used other things to make payments, they were more likely to use livestock than anything else. Whether that signifies that livestock were more readily available, or more scarce (and valuable), is arguable. Either way, it is perfectly clear that the context was a vigorous, active exchange economy, with rapidly developing exchange mechanisms and considerable market activity. The rate of change across the century was such that by 970 expressing price in terms of precious metals, rather than in other objects, had come to predominate; the implications are clearly that more metal was in use than it was earlier in the century. 340km to the west, beyond the León Mountains, lay Celanova, in the green and hilly Atlantic zone. Although it was barely 20km from Orense, a town that was the focus of a Visigothic bishopric, there is no sense anywhere in this material of the nearby presence of a thriving town.91 Indeed, the urban property that the monastery received was in Zamora, 200km away. Like Sahagún, Celanova acquired property – vineyards and orchards in the hilly landscape of its own neighbourhood (cattle-standard country), large estates 60 or 70km to the north by the River Sil and larger vineyards 30–40km to the south.92 This monastery actively made acquisitions too, in the 960s especially, a generation after its foundation, particularly in the rolling country near Celanova itself. Of the several modes of valuation 90 Later, Sahagún is known for its exploitation of pasture and development of large-scale transhumant systems; see Mínguez Fernández, El dominio del monasterio de Sahagún, pp. 119–33 (although he anticipated the pastoral exploitation – the lands acquired in mid-tenth century have a strong arable base, even those in the mountains, as Wickham, ‘Pastoralism’, p. 440, pointed out). 91 For urban changes in the post-Roman period in north-west Spain, see Quiroga and Lovelle, ‘Ciudades atlánticas en transición’. 92 M. C. Pallares Méndez, Ilduara, una aristócrata del siglo X (A Coruña: Do Castro, 1998), has a detailed examination, especially in cc. 1 and 3, of properties that eventually passed to Celanova. For a detailed analysis of landscape and settlement in two parishes near Celanova (Rabal and Bobadela), see E. Portela and M. C. Pallares, ‘La villa, por dentro. Testimonios galaicos de los siglos X y XI’, Studia Historica, 16 (1998), pp. 13–43, at pp. 37–43.



simultaneously in use there, those based on precious metals were not at all common, and use of precious metals in payments seems to have been rare throughout the century. The things used to make payments were more likely to be grain than silver, or even livestock – despite the fact that this is good pasture country; lengths of cloth and beverages were also very common. Again the question of what determined the mode of payment is open: did purchasers pay with high-value objects (cloth is likely to have been precious), or easily available objects (with so many orchards and vineyards, payment in beverages would fall into such a category)? One would expect there to have been plenty of livestock in this area: why use grain rather than animals to make payments? Perhaps they paid with things that the vendors of land were really anxious to get hold of: maybe grain and drink in years of famine (which would raise the interesting possibility that some or many transactions were initiated by the vendors)? The volume of property transfer was clearly a new development here, but it is particularly interesting that clearly articulated systems of valuation were in existence prior to the new development, at least some of which – like that based on the notional tremissis – must have been of considerable antiquity. Here is an utterly rural economy but one that does not remotely suggest self-sufficiency: it was equipped with the apparatus for ready exchange, and commercial exchange at that. The rural mode of its market transactions stands as a striking contrast to those of urban/sub-urban Sahagún. Lastly Cardeña, on the eastern edge of the Duero basin, under the sierras. Like Sahagún, it lay in an urban hinterland but one that was curiously different. Here silver-based notions of value were much more prominent; indeed no other systems of valuation were used at all. Here, when non-metal objects were used to make payments they were characteristically very high-value objects, and payments were more likely to be expressed in terms of precious metals than other objects from the 960s – that is relatively early, a decade earlier than at Sahagún. People in the Cardeña neighbourhood seem to have been thinking of making payments in terms of silver, rather than anything else, at least from mid-century. However, the monastery was not so actively buying, as Sahagún and Celanova did, although it benefited from donations large and small; indeed, it was the 980s before the number of recorded sales overtook that of donations, and numbers were not high even then. This is a different kind of sub-urban context, one that was much less commercially active, although clearly silver-conscious; significantly, the proportion of recorded countergifts in Cardeña’s collection is higher than that in the others – those who made the records, and doubtless many of those involved as actors, seem to have thought more in terms of personal relationships than of the market. They knew about sale and the mercantile mode of exchange, as the comment that Judas sold (vendidit) the Lord makes clear,93 but they did not choose this mode for most of their transactions: they chose instead the mode of continuing relationships and obli-

93 C109, C140.



gations (unlike the Sahagún records of purchases, emphasizing the finality of the transactions and the end of continuing obligations to the vendors).94 The Cardeña approach seems to be reflected at Valpuesta to the north and more obviously at San Millán to the east, with their greater proportions of donations. So, three different approaches to valuation, three different approaches to price and two different approaches to sale. Overall, one might broadly say that a change occurred from a world in which goods constituted the due price, and in which gold was the precious metal of greatest reference, to one in which silver was increasingly both mentioned and also handed over as payment, although the absolute chronology of the change varied with location. It was clearly happening in the neighbourhood of the city of León by 930 but, not surprisingly, it came much later – after the tenth century – in and beyond the mountains to the north and west.95 This must mean that, despite developments in later centuries, the shift to silver was tangible in the Duero basin by the mid-tenth century.96 The material of the 930–60 generation is therefore critical, offering a window on to the changing transaction practice of the region. It is particularly interesting that this change happened at a time of heavy raiding and political campaigning. While it is reasonable to suppose that the campaigning gave some stimulus to production and exchange, it must have been agricultural production that generated sufficient surplus to sustain these changing patterns. It is therefore extraordinarily unlikely that this agriculture was on lands that were literally deserted for the 200 years before: on the contrary, it is much more likely that the lands of the Duero basin had been occupied and worked throughout; after all, early tenth-century grants were of working agricultural properties. Different valuation systems clearly meant different things to different people in northern Spain in the tenth century. All those considered here were notional units of account – mental constructs – but one, the argenzos, was closely related to the objects used to make payments. In some parts – near León and Sahagún especially – they related to active systems of commercial exchange; in others – near Cardeña – the systems belonged to the world of writing and had less to do with buying and selling. The Atlantic systems were distinctive, but not wholly cut off: they certainly indicate thriving exchange economies, with some commercial aspects – the very existence of valuations makes this a sale culture. The systems of the meseta were even more obviously thriving, more commercial, as also more urban and less characteristically rural, and already by the late tenth century were 94 Cf. Miller, ‘Gift, sale, payment, raid’, p. 49. 95 Cf. Zamanillo Arizabalo, ‘Circulación monetaria y sistemas de pago’. 96 Discussion of comparable issues in the eleventh century is dominated by the presence of gold – particularly the enormous quantities of gold taken as tribute from Muslim states. There are, however, some good discussions about the rarity of gold in daily use in Castile-León (see Gautier Dalché, ‘L’histoire monétaire’, Todesca, ‘Means of exchange’). It is a matter for a different paper but is worth considering that the silver shift evident in the tenth century reflects the underlying long-term trend.



beginning to be related to wider economic networks – of Spain and of the world beyond. Their relationship to the economic mainstream of western Europe obviously cannot be assessed in respect of the tenth century alone and needs a forward look into the eleventh, but the fact that transaction practice was already changing in mid-tenth century is an important signal of wider and deeper change.97

97 I am very grateful to the British Academy for a grant which enabled me to travel in Castile-León in October 2000 and to secure the assistance of Dr Paul Kershaw in July/August 2000 to expand the corpus of data, as also to UCL’s Graduate School, History Department and S&HS Dean’s Travel Fund for grants which enabled me to travel in Galicia in May 2001. Grateful thanks are also due for some very helpful discussions to Cardiff University Archaeology Department, the ‘Bucknell’ charter group, UCL History Alumni and the London Medieval Society, and of course to Paul Kershaw. I am especially grateful to Julio Escalona and Chris Wickham for their pertinent comments on early drafts of this paper and for sustained help across several years.


7 WHEN GIFT IS SALE Reciprocities and commodities in tenth-century Christian Iberia

In 908 a woman called Trudilli gave her husband Evenando several estates near Guidões, about 25 km south west of Braga, in northern Portugal, along with six slaves, three of whom were ‘Moors’.1 The gift was recorded in a charter, which survives on a single sheet and which gives some further information about the transaction: we are told that the couple had married at the wish of their parents and that they had no children; Trudilli made her gift in order that her husband should look after her during her lifetime (aveatis de me cura in mea vita) and provide for masses to be said, candles to be lit and prayers to be offered at her death; her husband was free to dispose of the estates as he wished.2 In fact we also know, from another charter on a single sheet, that Trudilli’s brother Odoario had given her one of these estates, and the three Moorish slaves, the year before; he had acquired the estate from their father, an abbot, with an instruction that he should have lifetime use of it and his sister should have it after his death.3 His sister was to provide masses, candles and prayers at the point of Odoario’s death too, and he received a silver vessel and lengths of cloth from her in 907, in return for confirming the charter. Both charters were recorded by the same local priest. While much about the records of these transactions is unremarkable, both have aspects which are unusual. Trudilli’s gift to her husband might be thought to reflect some kind of dowry settlement,4 but it appears to have been made some considerable time after the marriage took place, and at least part of the gift had been used by her brother for his lifetime; Trudilli apparently had to pay her brother, with silver and cloth, to get the transaction in her favour confirmed. It was relatively unusual to make a record of transactions between family members at this period;

1 First published in W. Davies and P. Fouracre (eds), The Languages of Gift in the Early Middle Ages (Cambridge: Cambridge University Press, 2010), pp. 217–37. 2 PMH DC 16. For single sheets, see following p. 120. 3 PMH DC 14 (907). 4 Dowry, since the property ultimately came from father to daughter. These texts sometimes record provision of a dowry of property for the new household, by the parents of aristocratic women, when the woman married; see W. Davies, Acts of Giving: Individual, Community, and Church in Tenth-Century Christian Spain (Oxford: Oxford University Press, 2007), p. 168.



it was unusual to record provision for liturgical commemoration at death; and it was unusual to include, as both texts do, the phrase pro remedio anime mee in records of gifts between lay persons – the phrase normally occurs in records of gifts to religious institutions. While it is possible that Odoario, Trudilli and Evenando were all of the church, none of them is associated with terms that would indicate this, nor with terms indicating distinctive lay commitment to the church (confessus, confessa, servus dei, ancilla Christi, famula dei, famulus Christi and deovota commonly occur in the corpus of texts from the region). Both of these transactions represent deals done between people who knew each other well, in order to get something – liturgical commemoration. As such, they are good examples of the transactional nature of much recorded giving in this culture at this time. It is interesting both that the standard format of a record of gift is the format chosen and also that there are non-standard elements in the formulation. The examples are unusual but they are not unique and they merit further exploration. It is the purpose of this paper to explore the relationship between the language used in gift records from northern Iberia and the transactions that they record, for the relationship was far from straightforward. There survives a large corpus of charters recording transactions – mostly, but not exclusively, land transactions – from northern Iberia. This material begins in the ninth century but its volume becomes significant from the 930s and continues throughout the tenth century and beyond; there are in fact several thousand tenth-century charters, the material to be considered here. We therefore have an important body of material that is all about exchange in an early medieval society, and a body which is sufficiently large to be subject to statistical analysis. Just over half of this corpus records gifts and just under half records sales.5 It is therefore particularly appropriate to ask questions about gift-giving, gift-exchange and gift economies on the one hand, and about commodities, commodification and commercial exchange on the other, during this century when documentation became plentiful. As we have seen, the historiography of gift-giving emphasizes the social relationships in which gift transactions were embedded and which gifts reinforced, while the historiography of sale, strongly influenced by Marx’s characterization of commercial exchange as the ‘fetishism of commodities’, tends to suppose that the commercial transaction denies or conceals (or downgrades) social relationships.6 5 See the following for figures. Leases are not common in this material. 6 See Davies and Fouracre (eds), The Languages of Gift, and especially papers by J. L. Nelson, ‘The setting of the gift in the reign of Charlemagne’, and C. Wickham, ‘Compulsory gift-exchange in Lombard Italy, 650–1150’, pp. 116–48, 193–216. Note, however, Gregory’s demonstration that gift and commodity are not alternatives but co-exist – they are ‘co-eval’ – C. A. Gregory, Savage Money: The Anthropology and Politics of Commodity Exchange (Amsterdam: Harwood Academic Publishers, 1997), pp. 10, 41–70; as also that the modern currency options market is ‘commodity fetishism in its highest form’, ibid., p. 294. Cf. P. J. Geary, ‘Gift exchange and social science modelling. The limitations of a construct’, at p. 130, and L. Kuchenbuch, ‘Porcus donativus. Language use and gifting in seigniorial records between the eighth and the twelfth centuries’, at pp. 230–2, both in G. Algazi, V. Groebner, B. Jussen (eds), Negotiating the Gift: Pre-Modern Figurations



Some have argued, further, that there were purely gift economies in some past cultures, in which all transactions were conducted in gift mode; and many have assumed that the commercial economies which succeeded them were characterized by transactions divorced from social relationships.7 To what extent either set of concepts is appropriate in the early middle ages in northern Iberia, and their relative significance and relationship, is the subtext of the discussion here.

Texts and transactions Texts mostly survive in cartulary copies from the great monasteries of northern Iberia, although there are also – as in the two examples cited – several hundred charters on single sheets.8 The transactions which they record – on single sheets and in cartularies – peaked in the middle of the tenth century, with far more in the 950s and 960s than in earlier or later decades; indeed, tenth-century numbers before 920 total less than a hundred; from 930 the increase was steep until the decade of the 950s (during which there were several hundred transactions), partly because of a significant increase in the proportion of small-scale business; and from 970 numbers dropped off quite steeply. Different collections have different profiles, however, and there is considerable variation in regional distribution. In the following couple of pages, I shall indicate the basic statistics of the collections, with their regional associations. In fact, most collections include material from different sources, for a cartulary was often the ultimate result of a complex process of accumulation. There is a considerable corpus from northern Portugal, where we began, some 184 charters of pre-1000, recording just under 200 transactions; there is no modern edition of the corpus as a whole, although there are individual studies of some subsets; it can be difficult to differentiate the authentic from the false without the benefit of modern critical assessment, although some of the records of large aristocratic grants to great monasteries such as Guimarães are clearly suspect.9 Nevertheless, this material is so interesting, and so obviously of similar character to that from northern Spain, it has to be taken into account.

of Exchange (Göttingen: Vandenhoeck & Ruprecht, 2003), pp. 129–40, 193–246; M. Osteen (ed.), The Question of the Gift: Essays across Disciplines (London: Routledge, 2002). 7 For a robust dismissal of gift economies in early medieval Francia, see F. Curta, ‘Merovingian and Carolingian gift giving’, Speculum, 81 (2006), pp. 671–99; although it needs some update in the light of recent work, Philip Grierson’s classic paper ‘Commerce in the Dark Ages: A critique of the evidence’, Transactions of the Royal Historical Society, 5th ser., 9 (1959), pp. 123–40, is an effective debunking of the notion that all exchange in the early middle ages must have been commercial. 8 Charters on single sheets are often ‘originals’, but not necessarily so, for some are forgeries and some are copies; cf. n. 13 following. 9 PMH DC; for recent editions of Braga and Coimbra charters, see Liber fidei sanctae bracarensis ecclesiae, ed. P. Avelino de Jesus da Costa, 3 vols (Braga: Junta Distrital de Braga, 1965, 1978, 1990); Livro Preto. Cartulário da Sé de Coimbra, co-ord. M. A. Rodrigues (Coimbra: Arquivo da Universidade de Coimbra, 1999). For Guimarães, see now PMH DC Livro de Mumadona. Cartulário do Mosteiro de Guimarães, co-ord. L. C. Amaral (Lisbon: Academia das Ciências de Lisboa, 2016).



Figure 7.1 Location of sources of the main charter collections used

The late twelfth-century cartulary of the monastery of Celanova, in Galicia in the west, not far from the present northern border of Portugal, includes about 225 recorded tenth-century transactions (figure 7.1).10 Other important Galician collections, though smaller for the tenth century, are those from the monastery of Sobrado de los Monjes, 45 km north east of Santiago de Compostela, and from that of San Julián at Samos, 65 km to the south east of Sobrado, in the western foothills of the Cantabrian mountains; they record just under 100 and just under 60 transactions respectively.11 The next major group involves the several collections in the archive of the bishopric of León, to the east of Galicia, one of the most important foci in the Duero basin; this episcopal collection includes records from urban and rural monasteries which subsequently came into the hands of the bishopric and has over 580 recorded tenth-century transactions, of which about 200 are on single sheets.12 There is also a very large collection from the monastery of Sahagún, on the plateau, 50 km to the south east of León itself, with over 430 tenth-century transactions.13 Also from the same region, a small but important group of 50 charters, mostly on single sheets, comes from Otero de las Dueñas,

10 Cel – a late twelfth-century cartulary comprising, in the opinion of its editor, three earlier collections. 11 Sam, Sob. 12 Li, Lii, Liii – the core of the edition is a large cartulary of the first third of the twelfth century. 13 S – from two collections, one a set of single sheets (including both originals and copies made up to the early twelfth century) and the other a cartulary of c. 1110.



25 km north west of León, in the southern foothills of the Cantabrian mountains, including a lay archive of the eleventh century.14 Farther to the east, in Castile, lies the monastery of Cardeña, in the western foothills of the Sierra de la Demanda, on the edge of the plateau, lying just to the south of the city of Burgos, beside one of the main routes from central Spain to France, then as now; this late eleventh-century cartulary includes over 200 recorded tenth-century transactions.15 Although they have fewer tenth-century records than the large sets, there is useful Castilian material in some of the smaller collections, particularly from Valpuesta, in the mountains to the north of Burgos, and from San Millán de la Cogolla in the rich agricultural lands of the modern Rioja, to the east (comparative numbers of transactions for Valpuesta are 36 and for San Millán 121).16 There are in addition a few from Arlanza, also in Castile, a few from Albelda, like San Millán within the orbit of both Castile and Navarre, and even fewer from San Juan de la Peña, farther east in Aragón (comparative numbers for Arlanza are 25, Albelda 28, San Juan 11).17 Two small but very important collections come from much farther to the north: one, including important early material, comes from Santo Toribio in the Liébana, in Cantabria on the northern side of the Cantabrian mountains, and the other comes from San Vicente in Oviedo, again on the northern side of the Cantabrian mountains, but farther west, in Asturias, to the north of León (62 and 24 transactions respectively).18 There are sometimes small numbers of tenth-century charters in other collections, such as that from Leire; I have occasionally noted their contents but have not included them in statistical analyses.19 There are also several thousand charters from Catalonia, in the region of the Pyrenees, in north-east Spain; Catalonia is in many ways distinctive in northern Iberia, and its tenth-century history was directly influenced by a Frankish past, making its culture different from that of parts farther south and west. The charters have been the subject of extensive analysis in recent years and I will refer directly to those analyses, rather than to individual texts.20 The material is in some ways separate, like the culture, but needs to be taken into account.

14 OD. 15 C. 16 V – the meticulous analysis of this edition shows that the cartulary was mainly in eleventh- and twelfth-century hands, but that some folios were prepared in the tenth century. SM – an early twelfth-century collection, from several sources. 17 A; Ar; SJP. 18 T; Ov. 19 Leire; there are 13 charters of pre-1000, some of doubtful authenticity. 20 P. Bonnassie, La Catalogne du milieu du Xe à la fin du XIe siècle: croissance et mutations d’une société, 2 vols (Toulouse: l’Université de Toulouse-Le Mirail, 1975–6), and also his shorter assessment, P. Bonnassie, La Catalogne au tournant de l’an mil. Croissance et mutations d’une société (Paris: Albin Michel, 1990); A. J. Kosto, Making Agreements in Medieval Catalonia: Power, Order, and the Written Word, 1000–1200 (Cambridge: Cambridge University Press, 2001); id., ‘Laymen, clerics, and documentary practices in the early middle ages: The example of Catalonia’,



Altogether records of approximately 2,140 tenth-century transactions survive from northern Iberia excluding Catalonia, of which 1,085 are recorded as gifts and 912 as sales – very large numbers by northern European standards; by comparison, just over 100 survive for the same region from the ninth century.21 The breakdown by region is as follows: Table 7.1 Gift and sale, by region of 10th-century northern Iberia, as percentage of all transactions per region22

Castile Galicia Portugal The North Western meseta

% gift

% sale

Total numbers

67 57 55 51 40

27 32 33 42 54

412 381 190 86 1,069

Clearly, the records from the far West record more gift than sale, as do those from Castile; the western meseta, by contrast, provides more records of sale than gift, as does Catalonia;23 numbers from the North suggest rather less of a contrast between the two categories of transaction.

Gift and sale Western European charters of the early middle ages usually clearly differentiated between gift and sale, using a different standard format for each kind of record; that standard formats for documents were available in Iberia is clear from a surviving tenth-century Catalan formulary, as also from the sixth- and seventh-century Speculum, 80 (2005), pp. 44–74; Ll. Tó Figueras, ‘L’historiographie du marché de la terre en Catalogne’, in L. Feller and C. Wickham (eds), Le marché de la terre au moyen âge (Rome: Ecole française de Rome, 2005), pp. 161–80. 21 Numbers given for transactions are necessarily approximate, and I wish to avoid giving any spurious sense of precision. Numbers of transactions do not exactly mirror the numbers of records: what was or was not a transaction is sometimes arguable; some charters record more than one transaction; some charters record texts such as inventories or sworn witness statements and do not record any transactions; and some charters duplicate transactions recorded elsewhere. For this count I have excluded obviously corrupt documents; included boundary statements, records of disputes and exchanges (as transactions other than sale or donation); and included confirmations with donation figures. 22 San Juan de la Peña (Aragón): 7 gift 0 sale, of 11 transactions. 23 Catalonia: sales constitute 70% of about 2,000 recorded tenth-century transactions in Bonnassie’s graph, which shows a peak in the 990s and then a steady decline from 1030s to only 10% of transactions in the 1140s; see Bonnassie, La Catalogne, vol. 1, p. 364, vol. 2, pp. 893, 887 (several thousand more transactions are now known). See also Tó Figueras, ‘L’historiographie du marché de la terre’, p. 162, and passim for a survey of the extensive past discussion on Catalan sale. Bonnassie’s contribution is very substantial, but see also Salrach’s paper on the Catalan land market in the tenth century: J. M. Salrach, ‘El mercado de la tierra en la economía campesina medieval. Datos de fuentes catalanas’, Hispania, 55 (1995), pp. 921–52.



texts from central Spain that are preserved on slate.24 Indeed, it is clear that most authors of charters had access to the same standard repertoire of formulas, whether they were attached to a royal court, a great monastery or a minor rural religious household and whether they worked in Galicia, on the meseta or in Catalonia (the fact that transactions between lay persons, whether of gift or sale, from as early as we have charter texts, use the same formulas means that the formulas must have older origins).25 The repertoire of formulas was elaborated in different ways in the major lay and religious centres but used in simpler unelaborated forms to record peasant transactions in rural localities. Whether the record was made in a major centre or in a village, a distinction between gift and sale was maintained by tenth-century writers: the language of gift and sale was in most cases strongly differentiated, and the concept of price was central to sale. By ‘giving’ they meant transferring something to another party, without a pre-arranged price or return gift being necessary to validate the transaction (although a countergift might be added, as Trudilli did when her brother Odoario made his gift to her); by ‘selling’ they meant transferring something to another party, as a result of contract between the two parties, for a pre-determined price, to be paid at a pre-determined time. Hence, charters of gift used verbs like donare, tradere and concedere (give, hand over, concede) and carta/cartula donationis or occasionally concessionis (charter of gift), while sale charters usually had the active verb as vendere (to sell), with the document itself called a cartula vendicionis (charter of sale), and a price, pretium, being specified.26 Hence, ‘I agreed to sell you my garden in Villagoya . . . and I received a price of one ram from you, which was most acceptable to me’; texts often include the words ‘I have accepted the price and nothing remains due’.27 This concern with price also occurs in composite records, like the 940s text from San Millán de la Cogolla in which a long list of separate small-scale transactions has price associated with all those termed sales, but not with those termed gifts.28 If gifts provoked a material countergift in response, as in the case

24 M. Zimmermann, ‘Un formulaire du Xème siècle conservé à Ripoll’, Faventia, 4:2 (1982), pp. 25–86; I. Velázquez Soriano, Las Pizarras Visigodas. (Entre el latín y su disgregación. La lengua hablada en Hispania, siglos VI–VIII) (Madrid/Burgos: Real Academia Española, Fundación Instituto Castellano y Leonés de la Lengua, 2004), e.g. nos 8, 19, 40; cf. ‘Formulae Wisigothicae’, ed. J. Gil, in id., Miscellanea Wisigothica (Seville: Universidad de Sevilla, 1972), pp. 70–112, for a probably Visigothic formulary (whose formulas recur in ninth- and tenth-century charters), preserved in an early modern manuscript. 25 There were plenty of small family monasteries and lay religious households; see Davies, Acts of Giving, pp. 36–60. 26 See Davies, Acts of Giving, pp. 88–9, for detailed examples. 27 Li174 (943): ‘Placuit mici . . . ut vinderem tibi meo ortale . . . Et acepi de te in pretio, que mici vene conplacuit, id est karnario I’; ‘et alio precio que michi complacuit’, S53 (933); ‘et de ipso pretio aput vos nichil remansit sed completum est’, Cel237 (934); ‘ut de ipso pretio nichil apud vos remansit in debito’, PMH DC 78 (960). 28 SM59 (943–51).



of the chestnut mule given to a countess in recognition of a large gift of land in 935,29 the language was different from that of the price payment. Countergift was clearly viewed as different from price: failure to produce the price, at the right time, would invalidate the transaction; countergift was less explicitly contractual. In the vast majority of cases these distinctions between gift and sale, countergift and price, were sustained. But although the essence of the distinction was usually clear and consistent, there are a few texts in which it was not, and indeed where the concepts appear to be confused – hence we find kartula donationis vel venditionis (charter of gift and sale);30 or we sometimes find an apparently straightforward record of a gift, in gift language, which bizarrely terminates with a reference to this ‘carta vendicionis’, as happened in 972 when two brothers gave an enclosure to the monastery of San Miguel de Pedroso ‘for the salvation of their souls’.31 Even more confusingly, contracts for a price were sometimes called ‘gifts’, and transactions which in all other respects appear to have been gifts were recorded as ‘sales’. As we saw in the Portuguese cases with which I began, deals also appear to have been done, for something like price, but were recorded as transactions of gift, in gift language and gift format. It would also appear that the same kind of transaction could be recorded in one place as a gift and in another as a sale. Take the case of grants made to cover fines. When Gigulfo and his wife had to pay a fine for Gigulfo’s own adultery, and when Eirigu and his wife Seniorina had to pay a fine for stealing and eating sheep, the transactions are recorded as gifts in the language of gift (concedo, scripture concessionis; facimus kartam, damus . . . atque concedimus);32 whereas some fines are recorded in the language of sale: Goton ‘sold’ (ut venderem . . . et vendidi) to Bishop Rosendo his share in an orchard near Celanova because he had stolen Rosendo’s bees (abelias) and had been ordered to pay a fine (habebamus vobis ad pariare) to the value of nine modii and two quarters.33 Likewise, the record of Cisilu’s payment of a fine is called scriptura vendiccionis (charter of sale) – paid to a lay couple for a daughter’s adultery.34 Or, to take another kind of example, that of ‘helping out’: when Sanzone gave the priest Munnio a plot of arable, a garden and some fruit trees in gratitude because the priest had helped out with food and clothing in a very bad year, it was recorded in the language of gift; but

29 C27; for countergift, see ch. 8 following. 30 Lii357, for example; cf. Cel535, and the Catalan example of 989 noted by Tó Figueras, ‘L’historiographie du marché de la terre’, p. 163. 31 SM90. 32 Cel338 (989), Cel169 (962); cf. PMH DC 43 (937). 33 Cel393; abelias here may well be a corruption of ovelias – there are many cases of sheep being stolen and eaten (cf. n. 32 above). The modius was a standard unit of account, which I have interpreted as a ‘sackful’, presumably of grain; see ch. 6 above, ‘Sale, price and valuation’, at pp. 111–12. 34 PMH DC 58 (949).



when Matecia and her cousin Vicente gave a lay couple some land with nut trees in 962, it was recorded in the language of gift, but it had a price associated – goats, cheeses, grain and other goods that had been supplied in a bad year were called ‘price’; when Gogina and her children gave the priest Braolio half an orchard in 964 because of the grain he had provided for them, it was recorded in the language of sale; and when Adosinda gave a lay couple her portion of a farm in Portugal in 971 because they had supported her with food and clothing, it was recorded in the language of sale.35

Confusion or other ways of seeing things? How are we to deal with these confusions between two apparently different kinds of transaction? Events were not always as straightforward as the words of the records imply and, while we cannot always perceive the logic of the tenth-century framing, much is capable of explanation. Indeed, there are several kinds and several layers of explanation. Firstly, apparent confusions were not necessarily confused: a rationale can sometimes be detected. When a family exchanged property with the monastery of Cardeña in 941, the verbs used were ‘exchange, sell and hand over (conmutamus, vendemus, tradimus)’, the text was called a charter of ‘sale and exchange (cartula venditionis vel comutationis)’, and what they received was called a ‘countergift and price (in honore vel precium)’.36 From the detail of the text it looks as if a deal was done with the family to stop them building houses round the church of San Torcuato, which was owned by Cardeña, and as if they received a bit more than they handed over; hence, there was a contract of exchange, but they got more than the land swapped, as the ‘price’ for agreeing to the transaction, and also a sweetener (a skin) as countergift.37 Further, there is a consistency in recording the clearing of debts: using land to pay off debts was nearly always conceptualized as sale, regardless of region and regardless of record maker. Vineyards were handed over in Galicia in place of overdue renders of wine and grain; a church was handed over in Portugal in 973 in place of a very large grain rent; on the meseta a share of an enclosure was handed over in place of overdue grain, wine and cheese.38 In cases like this, and they are not unusual in the second half of the tenth century, the sale format and sale language are used, although there may not be a price clause. Sometimes, however, the debt is explicitly described as the price – retrospectively

35 36 37 38

V28 (950), T65, OD14, PMH DC 103. C36. See further, Davies, Acts of Giving, p. 137. Cel411 (989), Cel409 (990), PMH DC 111 (973) (cf. PMH DC 91 (965)); Lii371 (964). Cf., in Catalonia, Tó Figueras, ‘L’historiographie du marché de la terre’, p. 167, on debt as a principal reason for sale.



reconstituted as price, in effect.39 So too the cases of helping out: although Sanzone’s case cited above is recorded in the language of gift, this is very unusual; usually past assistance was retrospectively framed as debt, and for debt the sale format was more appropriate. Hence, land given in gratitude for past help was land sold; the grain that Braolio gave Gogina was ‘advanced (prestatit)’, a (retrospective) down-payment, as it were, on the land that was to come to him later.40 Another level of understanding comes from the realization that there were clearly more than two kinds of transaction: the alternatives of gift or sale are too crude. There certainly was sale, as we might understand it now, that is commercial exchange for price: land was bought to consolidate property, for example, and to exploit it more effectively; the great monasteries were involved in such transactions, especially in the second half of the tenth century and later, often buying from high aristocrats, although they bought small plots from peasants too.41 Prices in aristocratic transactions could be extremely high, like the mule, cloaks, gold vessel and skins, valued at 1,000 solidi, which the Portuguese monastery of Guimarães paid a woman for estates, saltpans, fisheries and a church in 953.42 Indeed, valuation was a concern of many texts: prices paid include ‘a carpet worth (valente) 5 silver solidi’, ‘an ox and a cow, with its calf, priced at (preciatos) 24 sackfuls of grain’, ‘three horses, one worth 15 oxen, another worth 5, and the third worth 7’.43 Indeed, one Portuguese charter of March 960, recording the sale of half a farm from one lay couple to another for a price of a cloak worth 30 sackfuls of grain, has a codicil recording the parties’ agreement that if the other half of the property had to be sold, they would reverse the 960 transaction for a ‘just price’,

39 Transactions which used land as collateral for a loan (‘prêts sur gage’), familiar in Italy in the eleventh century, may look superficially similar but were quite different: they were essentially about lending, with the prospect of repayment of the loan and return of the land; selling land to pay off debts or to get food was much more final. See C. Violante, ‘Les prêts sur gage foncier dans la vie économique et sociale de Milan au XI siècle’, Cahiers de civilisation médiévale, 5 (1962), pp. 147–68, 437–59. Bonnassie, La Catalogne, vol. 1, p. 399, notes the existence of ‘prêts sur gage’ in Catalonia, but not until the 970s and even then infrequently. 40 Cf. the many sales of small plots which are simply recorded as sales for food; see Davies, Acts of Giving, pp. 158–60. 41 Cf. the studies of monastic accumulation of property in, for example, J. M. Mínguez Fernández, El dominio del monasterio de Sahagún en el siglo X (Salamanca: Universidad de Salamanca, 1980); J. A. García de Cortázar y Ruiz de Aguirre, El dominio del monasterio de San Millán de la Cogolla (siglos X a XIII). Introducción a la historia rural de Castilla altomedieval (Salamanca: Universidad de Salamanca, 1969) – in which case the main buying falls in the eleventh century; S. Moreta Velayos, El monasterio de San Pedro de Cardeña. Historia de un dominio monástico castellano (902–1338) (Salamanca: Universidad de Salamanca, 1971). 42 PMH DC 67. 43 C199 (988), Cel402 (961), Cel248 (991); for detailed discussion of valuation systems, see above, ch, 6, pp. 109–14; and for their commercial significance, ch. 9 following, ‘Notions of wealth in the charters of ninth- and tenth-century Christian Iberia’.



which was to be assessed by local elders (boni homines).44 At the opposite end of the spectrum it looks as if there were gifts for no consideration – as far as we can see, gifts that brought the donor no return or benefit either in life or at the point of death. Although some of these transactions may of course have had some unrecorded return, where they detail gifts to family members, especially lay members, it might be reasonable to think that there were none:45 the aristocrat Flamula gave estates in Galicia to her sister Clixovara in 963; a man called Stefano, perhaps a peasant, gave his two grandchildren a vineyard in Castile in 988; a man called Nuño gave his son scattered plots in Oncina, just south of the city of León in mid-century, together with a few stock and some furniture.46 It may be rare but pure altruism can sometimes occur, as can the unreciprocated gift.47 It is even possible that some gifts to the church were made without expectation of any return: while most gifts to the church, if not expressly made for the salvation of the soul, were at least recorded with accompanying expressions of piety, a small percentage of tenth-century records of gifts to the church have no such accompanying phrases; since giving for the sake of the soul grew in frequency in records of the tenth century, it is worth considering the possibility that some of the earlier donors simply gave because they could. Between these two poles of sale and gift for no consideration there were many different kinds of transaction.48 As we have seen, transactions for the purpose of repaying debts were recorded in the language of sale, although they were clearly not simple commercial transactions; they were certainly, however, about economic relationships. Like the similarly recorded transactions designed to repay past assistance, they could have social consequences too: the debts owed to aristocrats, and the assistance proffered by aristocrats, could put peasants in their continuing power. Sales of land to get food look like commercial transactions, and were recorded in terms of sale and price and sometimes value, but had similar social, economic and political overtones. An isolated record does not say too 44 PMH DC 79. Cf., in Catalonia, the comments of Bonnassie, La Catalogne, vol. 1, p. 310, on the regulatory role of boni homines in Catalan village contexts. 45 Although see J. Parry, ‘The gift, the Indian gift and the “Indian gift”’, Man, 21 (1986), pp. 453–73, at p. 458, who opposes Malinowski on this point. 46 Cel74, C202, Lii488; I exclude the many cases in which peasants gave land to aristocrats for no stated reason but with the likelihood that they sought some kind of protection; see Davies, Acts of Giving, pp. 161–2. 47 For the categories of ‘pure’ gift-giving and the unreciprocated gift, see Parry, ‘The gift’, who maintained that the ‘ancient literate civilisations of Europe and Asia’ stressed the notion of the pure gift, as also that it is more likely to arise in highly differentiated societies. For this notion in central medieval contexts, see I. F. Silber, ‘Gift-giving in the great traditions: The case of donations to monasteries in the medieval West’, European Journal of Sociology, 36 (1995), pp. 209–43, at pp. 226, 235; cf. G. Algazi, ‘Introduction: Doing things with gifts’, in Algazi, Groebner, Jussen (eds), Negotiating the Gift, pp. 9–27, at p. 16. 48 Compare Silber’s ‘whole spectrum’, ‘Gift-giving’, pp. 218–19; or early Malinowski’s ‘continuum’, Parry ‘The gift’, p. 454; or alternatively Gregory’s ‘finely graded set of ten categories’ which ‘do not form a continuum’, Savage Money, pp. 52–3.



much in itself, but when we find a set of texts recording many sales of small plots to an aristocratic couple or to a monastery for food, we see a social as well as a commercial process and we begin to glimpse the dynamics of very local politics. Examples include the many transactions with the powerful couple Ermegildo and Paterna in northern Galicia, and with another powerful couple Bagaudano and Faquilona in the Liébana, and with monasteries like Galician Celanova, Leonese Abellar and Santos Justo y Pastor.49 The same process also happened at a smaller scale: the priest Vicente (sometimes with his wife) bought a series of small plots in a village near Oviedo between 937 and 949 for – variously – grain, meat, cider, pigs and goats.50 Perhaps surprisingly, transfers of property following court judgments were recorded in gift format, as in the examples cited above; Caradonna was found guilty of laying siege to one Belido and dealt with the judgment by giving (dabo) a vineyard to the court holder, later count, Flaino Muñoz and his wife; Fredino and his wife Leovina gave (damus, dabo) the same couple several parcels of land because one of their sons had raped a woman and another had assaulted and wounded a man.51 One might have expected, by analogy with payment of debts, that they would have been recorded as sale; but although payment of fines is occasionally recorded as sale, this is unusual. These transactions fall into a special category, however: the form is certainly more gift than sale, although some texts have relatively few giving words, some may use scriptura or cartula confirmationis (confirmation charter) rather than cartula donationis or concessionis (charter of gift), and most though by no means all make use of the word pariare (to pay a fine).52 The basic format, nevertheless, is gift format. So also gifts in respect of past judicial assistance tend to be cast in gift rather than sale form: Valentino and his wife Donna gave (damus, perfiliaremus) Ablavel Godesteoz some of their land near the river Cea because he had saved them from the judge.53 Gifts were made for other reasons. Kings handed over property to aristocrats, both for their ‘service’ (usually past) and for assuming an element of territorial control (usually future), as other aristocrats did for service.54 These are normally recorded in the language of gift. The gifts of partnerless people, both male and female, are another kind of case: they made gifts so that they would be cared for 49 Sob67, Sob79, Sob93 (920–34), for example; T21, T23, T25, T26, T27, T32, T36, T39 (915–30); Cel392 (961), Cel394 (956), Li170 (943), Lii283 (954), for example. 50 Ov8, Ov9, Ov10, Ov12, Ov13. 51 OD44 (998), OD33 (992), OD31 (991). For Flaino Muñoz and his family, see C. Estepa Díez, ‘Poder y propiedad feudales en el período astur: las mandaciones de los Flaínez en la montaña leonesa’, in Miscel·lània en homenatge al P. Agustí Altisent (Tarragona: Diputació de Tarragona, 1991), pp. 285–327. 52 For example Liii556 (993), Sob54 (930). 53 S332 (986) – ‘pro que saccastes nos de fisco et de mandacione et de iudice et de mandacione [sic]’; there is no detail of the problem; perfiliare means to give as to a son or daughter, to adopt as heir. Cf. OD27 (987), Sam6 (997). 54 S19 (920), Liii541 (990), Cel505 (935), for example.



in old age, as the man Nuño Sarracíniz gave to the priest Iñigo in 964; and in some cases so that they would be taken into the household of the beneficiary, as the woman Gesmira gave to the lay couple Flaino and Brunildi in 950.55 Other people, couples as well as single or widowed people, made gifts in order to provide for their burial or, as in the cases with which I began, liturgical commemoration: Nuño’s grant was to provide for burial as well as care in old age; the woman Eusicia made gifts to the monastery of Cardeña specifically to provide for her burial, as did the man Diego to the monastery of San Millán and a married couple to that of Buezo.56 Vimarano gave an estate to the monastery of Celanova so that masses be said for him, as did the nun Leodegundia to her cousin Tannito for the same purpose; Bishop Rosendo gave lands to his freed slave Muzalha so that candles be lit and offerings made for him every Christmas Day, and a woman called Severa gave to a lay couple to provide for regular lights and alms for her soul.57 Others, both lay and religious, gave so that they or family members could be commemorated every year, sometimes at specified festivals – the woman Emilo Hamita to the monastery of Sahagún, the couple Abzuleman and Gotu to the monastery of Sao Mamede in Lorvão near Coimbra, the monk Menendo to the church of Santa María de Triacastela, King Sancho Garcés Abarca and his wife to the monastery of Leire, the nuns Eilo and Goisenda to Celanova and more.58 All of these were gifts that not only expected something in return, but specified what was to come in return – far from the gift for no consideration and perhaps rather closer to classic models of gift-exchange. These are inevitably strongly embedded in a surrounding social network; indeed, gifts to religious institutions often carry the implication of joining a patronage system. A step away, but still belonging to the world of reciprocities, are gifts made to religious institutions or persons (for the most part59) for the sake of the donor’s or a close associate’s soul – pro anima mea/patris mei, pro remedio animae suae/filiae meae. Again recorded in gift format, in gift language, these gifts were made in expectation of a return, although the return was to come in the afterlife through the pardoning of sin, the achievement of a reward in heaven, the avoidance of hell and the ultimate salvation of the soul.60 Such expecta-

55 C117, Li222. 56 C39 (942), C82 (952), SM64 (952), V23 (950); cf. Cel7 (950), Sob4 (959), Lii425 (973), Ov18 (974), OD23 (978), C211 (999). 57 Cel177 (952), SamS-3 (961), Cel172 (943), T64 (?962). 58 S190 (961), PMH DC 95 (968), Sam164 (989), Leire 9 (991), Cel216 (993). Note that these examples range from Portugal and Galicia to the borders of Aragón. 59 There are occasional instances of pro anima gifts to lay persons – cf. the Portuguese example at the beginning of this paper – but they usually involve some associated or further transmission to a religious body. 60 Cf. D. Ganz, ‘Giving to God in the mass: The experience of the offertory’, in Davies and Fouracre (eds), The Languages of Gift, pp. 18–32, on the overtly commercial phrases used by some Christian writers in the earlier middle ages.



tions are explicit, and the causal relationship between the act of giving and the ultimate reward is often explicit too, in all kinds of charter – I give so that (ut) I earn God’s pardon for my sins.61 It is characteristic of a range of different recording sources, although it would be fair to say that records made by the great monasteries tend to be more explicit, and more elaborate, in this respect and that brief records of donation, locally made, often contain no such indications. In this corpus of material nearly half of all donations to the church were explicitly made for the sake of the soul, and another quarter for reasons such as avoiding hell, attaining paradise or winning the final reward, without explicit reference to the soul. Yet another kind of transaction is that characterized by gift and countergift, still in the world of reciprocities but a different kind of gift-exchange. The countergift – that is, something handed over to the donor after receipt of a gift – stands as a public acknowledgement that a gift has been made; it is a formal statement, a significant gesture, sometimes a present. In some cases the value of the countergift does not remotely match that of the gift – an ordinary horse in return for a church and an estate in El Bierzo in 988 or a rabbit skin in return for a set of scattered properties in different estates, and further associated fields, in 954.62 At other times, however, countergifts include rare and precious objects, like decorated horse trappings or silver plate – giving these was clearly much more than a simple gesture. We find two good horses and a special silver vessel given to the king for confirming the gift of a church in 921; a bag of precious spice given in recognition of the gift of some land in Otero Morisco in 947; a carpet and a coat given in recognition of the confirmation of a grant in ?957.63 But countergifts do not seem to have been required in all gift transactions and are as inconsistent in their occurrence as they appear to be in value; they only occur in just over three per cent of all texts (and in six per cent of gift records). Let us therefore look at countergift in some more detail. Countergifts occur more frequently in Castilian collections (over six per cent in Cardeña) and less frequently in Galician collections (under one per cent in Celanova) than in those from the meseta, although the objects used are much the same in all parts; they are much more common everywhere from the 940s 61 See PMH DC 119 (976) for a particularly obvious example: a woman gave ‘pro remedio anime mee ut donet mici dominus remisio peccatorum meorum’ (for the salvation of my soul so that the Lord will give me remission of my sins). See Davies, Acts of Giving, ch. 5, for detailed discussion; S. D. White, Custom, Kinship and Gifts to Saints: The Laudatio Parentum in Western France, 1050–1150 (Chapel Hill: University of North Carolina, 1988), for a classic discussion of the principles; and Ph. Jobert, La notion de donation. Convergences 630–750 (Paris: Université de Dijon, 1977), for origins of the concepts. 62 Sam171, Lii265. 63 S26, OD6, V35. While folle zumag might conceivably denote a leather bag, a bag of the Middle Eastern spice sumac (still used in Persian restaurants [and purchasable in Waitrose]) seems to make more sense in these contexts.



onwards than earlier and are even more common in the 980s and 990s.64 They feature in transactions between laity, of high or lower status; in transactions between religious institutions; and – especially – after gifts from lay people to a church or monastery; indeed, two thirds of recorded examples are countergifts returned by churches to donors, whatever the status of the donor. This contrasts strongly with the bulk of the Italian material discussed by Chris Wickham, where, prior to some extension of use in the eleventh and twelfth centuries, countergift was essentially a matter for transactions between lay persons.65 The words used for countergift in this Iberian material again contrast with the Italian: they are usually honor or offertio, although other words occur too (beneficium, prestacio, signum, confirmatio). There is no apparent significance in the choice of word: it does not vary consistently in relation to the kind of gift, the function of the return gift or the status of the parties. Indeed, usage essentially seems to reflect the training and writing habits of the record-makers; hence, Cardeña texts, for example, always use honor whereas texts from all sources in the far West virtually always use offertio; collections from the meseta have much more varied usage but texts that originate from the larger monastic houses in that region tend to use offertio whereas those from local sources frequently use honor. There is no noticeable shift in usage over time across the tenth century. There are, however, some close analogies with the Italian material in the function of the countergift. More than two fifths of known countergifts in northern Spain were gifts made explicitly in order to secure confirmation of a previous gift: when King Ramiro II gave the monastery of Sahagún an estate in 971, Sahagún responded with a fine bay horse and a length of very expensive silk for the king for confirming the charter (ad istu testamentu confirmante); when Fernando Díaz and his wife gave the monastery of San Martín of Turieno the church of San Esteban de Mieses in 980, Turieno gave them in return a valuable mule worth 50 solidi, and two mares and their foals, in confirmatione, and the text goes on to say that he (Fernando) was therefore making the transaction secure.66 In these cases, like many of those in Italy, the function of the countergift was to ‘fix’ the gift and make it firm and enduring; these are more than simple gestures; the countergifts perform a formal, public function. A wide range of objects was used to secure this confirmation – animals, cloth, clothing, bedding, carpets, skins, harness, precious metalwork; any one object could be of high or lower value but usually more than one item was handed over in these cases. Such confirmation gifts occur in all kinds of countergift transaction – between lay persons, between churches, between laity and the church – and were given to both low- and high-status donors. 64 They occur in a higher proportion of extant charters at the end of the century. For details see ch. 8 following, ‘Countergift in tenth-century northern Iberia’. 65 Wickham, ‘Compulsory gift-exchange’. 66 S261, T75 (et facio inde securitatem).



In another fifth of recorded countergifts, there is no explicit reference to confirmation and a single object was handed over, nearly always an animal; if there was more than one donor, however, they could each receive a countergift – a horse, an ox and a cow respectively to the three donors of a house (casa) and appurtenant land to San Esteban de Salcedo in 950.67 Again, these single countergifts occurred in transactions between people of higher and lower status, both lay and religious. The texts give no explanation for their occurrence and since countergifts do not feature in all gift transactions we can only speculate about why they were made. They have something of the feel of an expected, traditional gesture – a horse in recognition of the generosity of the gift – but their value should not necessarily be seen as low (‘best’ horses are not uncommon); they may well be performing an act of fixing and stabilizing comparable to the countergifts which were explicitly given for confirmation. There is also a further group – 29% of cases of recorded countergift – in which something different seems to have been happening. Here, the status of the transactors was more limited, for the transaction was virtually always between an aristocrat and a church or monastery, and the value of gift and countergift looks more comparable to that of the initial gift. The aristocrat made a gift; the monastery returned a countergift of very substantial value – three horses, 400 sheep, a saddle and a tunic in one case; decorated spurs and nine lengths of imported cloth worth 500 solidi in another.68 Occasionally it is explicit that a deal has been done, for the church is said to have made a request for the land given. These cases therefore look rather more like sales, high-value transactions in which churches acquired property that they wanted by handing over some rather special goods.69 Notably, the Cardeña cases fall in a period largely devoid of sale texts in that cartulary, up to the mid-960s; but sale texts do feature in the next, post-mid-960s, phase, when these high-value countergift transactions disappear; it is as if the Cardeña author thought in terms of gift for many different types of transaction until he started to use the sale model in the 960s. Notably the word most commonly used in all these cases is honor, perhaps ‘charge’ here, close to price. So, countergift is often about systems of guarantee, although the ‘quasi-sales’ are different. It was only one of several forms of guarantee available, however: guarantees in writing, by a special charter of guarantee, separate from the transaction document, occurred, although not often; guarantee by sureties or pledges

67 SM57. 68 C42 (943), C23 (932). 69 Cf. James Campbell’s comments on transactions in late seventh-, eighth- and ninth-century England; although recorded in gift format, substantial treasure was handed over in return for land, ‘The sale of land and the economics of power in early England: Problems and possibilities’, Haskins Society Journal, 1 (1989), pp. 23–37 (reprinted in id., The Anglo-Saxon State (London/New York: Hambledon and London, 2000), pp. 227–45).



was more common.70 There are fewer instances of the use of personal sureties (  fideiussores, fidiatores) than there are instances of countergift, but they feature in 1.5 per cent of all charters and their occurrence mirrors the same time-frame, to the extent that there are also relatively more in the 990s; there are also more from Castile and fewer from Galicia. There are differences in their functions, though, for they were used to guarantee sales and future obligations rather than gifts: in other words, sureties were an appropriate guarantee mechanism for one kind of transaction and countergifts for another.71 There are therefore a number of points of obvious comparison with northern Italian material, but equally obviously there are differences. Guarantors were used concurrently in Spanish society, and for the most part for different purposes, and were not a mechanism that took over from countergift; there was not much use of countergift to guarantee promises and agreements (though they are attached to a few post mortem grants) and obligations were more likely to be secured by sureties. Further, quite apart from the difference in status of the contracting parties, overall the value of the countergifts in these Spanish cases looks higher: when a valuation is given in solidi, it is much more frequently above 10 solidi than below, and it is often above 100 solidi.72 Significantly, countergift is not particularly common although it preponderantly occurs in transactions of gifts to the church. It cannot have been a ‘normal’ practice and cannot be explained away by supposing that there were many more unrecorded transactions of this type, given this distinctive preponderance and the hundreds of available texts. It must therefore have been used in special circumstances, presumably when the politics of the particular situation demanded special guarantees. ************ The range of transactions found in northern Iberia runs from the purely commercial dealing in commodities to the gift for no consideration, passing through transfers to pay off debts, sales to get food, varieties of gift-exchange, gifts for countergifts and gifts for a small consideration. Almost half are recorded in sale format, using sale language; and just over half in gift format, using gift language. Only these two basic formats were available and the recorder had to choose one. There are a few confusions between formats, some of which can be easily explained and some of which cannot. And some of the transactions in the middle of the spectrum clearly gave some recorders a problem of classification. Hence,

70 Confirmation charters: C88 (955). For suretyship see further following, ch. 17, ‘On suretyship in tenth-century northern Iberia’. 71 The exception is Castile, where sureties were occasionally used to guarantee gifts – in precisely the same period (947–66) that countergifts occur. (It looks as if the Cardeña record-maker used gift form with countergift to record sales in this period [the (few) exceptions can all be explained away] and used gift form with sureties to record gifts.) 72 Of course, perceptions of the value of a solidus may well have been different in Spain, but it would require a ten-fold difference to explain this away completely.



transfers of property in gratitude for assistance in the past were nearly always classified as sale, although very occasionally they are recorded as gift, as Sanzone’s was;73 a handful of other cases may look like transfers in gratitude for past assistance, but in fact, although they record such transfers, each of these cases also makes provision for future assistance.74 Deals and transfers of property to secure some benefit or action in the future were nearly always classified as gifts, like the transactions of Odoario and Trudilli with which I began; very occasionally they are recorded as sale, like the sales made to churches or clerics, with the price expressed as liturgical masses to be said in the future.75 Given the two available formats, there is an overall consistency: transfers of property in respect of things that happened in the past normally go into the sale format; transfers in respect of things to happen in the future go into the gift format. Transactions which include both temporal dimensions might well have an element of both. There remain only a handful of texts that do not respect these principles – Sanzone’s from Valpuesta, the three cases of sale for burial or memorial functions and one payment of a fine which broke the normal gift mode of recording and was recorded, rather awkwardly, as sale;76 all of these exceptions occur in very local records of small-scale transactions, the product of local scribes working with a smaller repertoire of formulas, rather than the product of an experienced scriptorium. The exceptions are a tiny proportion of the whole – much less than one per cent. Clearly the records that we have are conditioned by the limited number of formats that were available, formats which in fact owed much to a distant Visigothic past. But although the formats were limited, it is equally clear that they were used to record some very different kinds of transaction and that different kinds of transactions co-existed. That being the case, we must guard against taking the words at face value. The range of transactions also shows how inappropriate it is to think of market economy and gift economies as alternatives. Both existed to some extent, but given that both did exist, the totality of economic relationships cannot be described as one or the other. Taking the words at face value also leads us to suggest confusion where in reality there was none: it is not so much the incompetence of early medieval authors that is revealed but rather our own failing to read their rationale. There is, in the end, a surprising consistency of approach. Different places used different 73 As also transfers in gratitude for judicial assistance. 74 T64 (?962), T69 (963), Li222 (950), Lii357 (962). 75 Li137 (940), C29 (937); Hatita and Totadomna ‘sold’ the priest Mavia a small plot of land near León for the price (pretio) of thirty votive masses for their parents and for their commemoration with the ‘other’ dead, while Munnio ‘sold’ Anderquina a vineyard in Castile for the price (precio) of the shroud to be provided when he died and masses to be said on that occasion. Property near Oviedo which was handed over for candles to be lit for a dead husband’s soul was also recorded as sale; Ov22 (980). Leocadia’s sale to the monastery of Abellar, to provide for future care, however, also had a very large price paid by the monastery and looks close to a commercial sale; Lii462 (978). 76 Liii578 (997) – the record of the judgment is inserted into the price clause in this case.



diplomatic – large monasteries tended to have their own favourite formulas and their preferred modes of elaboration, while local priests tended to stick to the basic repertoire of inherited stock phrases – but they tended to respect the same principles over what to record as gift and what to record as sale, whether they were in Portugal, or Galicia, or the Liébana, or the western meseta, or Castile. There were some genuine regional differences: there is more commercial sale in and around León than anywhere other than Catalonia; there are lower proportions of female alienators in Castile than elsewhere; but these tend to be differences in underlying socio-economic structure rather than in record-making practice. Although there are some pitfalls, these texts can reveal a lot; thinking about the language can also help us understand practice in its social and economic context.77

77 Cf. Kuchenbuch, ‘Porcus donativus’, especially pp. 199 and 203: his materials, and his approach, are different, but the message is comparable. I am most grateful to all the members of the ‘Bucknell’ group for their rigorous comments on earlier drafts of this piece.



When I began to think about the occurrence of countergift in early medieval Spanish charters, I initially supposed that it must reflect some established system of reciprocities, such as one finds in all the classic models of gift-exchange;1 perhaps it was even very archaic, reflecting a world in which transactions were concluded through gift rather than through other exchange mechanisms.2 The gift-exchange model continues to stimulate new theory and has, after all, been used to elucidate early medieval contexts for the last half century.3 It has been valuably applied to the study of the spiritual reward for gifts to God and to the development of seigneurial obligations, although the last decade has seen some thoughtful revisionist contributions to the medieval application.4 A moment’s thought reveals that gift-exchange obviously cannot have been the dominant mode of exchange in tenth-century Iberia because there are hundreds of sale documents from precisely the same areas as produce countergift documents. Looking at the occurrence of countergift makes it clear that a more complex set of rationales lay behind it. 1 First published in A. Deyermond and M. J. Ryan (eds), Early Medieval Spain: A Symposium, Papers of the Medieval Hispanic Research Seminar, 63 (London: Dept of Hispanic Studies, Queen Mary, University of London, 2010), pp. 79–96. 2 M. Mauss, ‘Essai sur le don: forme et raison de l’échange dans les sociétés archaïques’, L’Année sociologique, new ser., 1 (1925 for 1923–24), pp. 30–186. 3 C. A. Gregory, Gifts and Commodities (London: Academic Press, 1982); J. Parry, ‘The gift, the Indian gift and the “Indian gift”’, Man, 21 (1986), pp. 453–73; M. Godelier, L’énigme du don (Paris: Fayard, 1996); M. Osteen (ed.), The Question of the Gift: Essays across Disciplines, Routledge Studies in Anthropology, 2 (London: Routledge, 2002); P. Grierson, ‘Commerce in the dark ages: A critique of the evidence’, Transactions of the Royal Historical Society, 5th ser., 9 (1959), pp. 123–40; W. Miller, ‘Gift, sale, payment, raid: Case studies in the negotiation and classification of exchange in medieval Iceland’, Speculum, 61 (1986), pp. 18–50; I. F. Silber, ‘Gift-Giving in the great traditions: The case of donations to monasteries in the medieval West’, European Journal of Sociology, 36 (1995), pp. 209–43. 4 G. Algazi, V. Groebner, B. Jussen (eds), Negotiating the Gift: Pre-Modern Figurations of Exchange, Veröffentlichungen des Max-Planck-Instituts für Geschichte 188 (Göttingen: Vandenhoeck & Ruprecht, 2003); A.-J. A. Bijsterveld, Do ut des. Gift Giving, Memoria, and Conflict Management in the Medieval Low Countries (Hilversum: Verloren, 2007); F. Curta, ‘Merovingian and Carolingian gift giving’, Speculum, 81 (2006), pp. 671–99.



This paper is about the tenth century, that is the first early medieval century of extensive local documentation, offering the first opportunity to look at rural society as it emerges from a long period of darkness. It is about northern Iberia, including northern Portugal, because that is where the documents were produced; and it excludes Catalonia, because of its Frankish background and leanings. The documents are charters – records of the transfer of property rights. About 2000 charters survive from the tenth century from the area considered, largely in cartulary copies of the eleventh and twelfth centuries, although there are also several hundred surviving originals. They come from across the North, from different collections in Portugal, Galicia, the meseta, and Castile, with rather fewer from Navarre and Aragón. There does not appear to have been extensive discussion of early medieval countergift by Spanish scholars, although those writing in the early twentieth century referred to it; they thought that the incidence of countergift indicated the appearance of features of Germanic law in Spain.5 Nowadays our views on legal systems and ethnic identities have changed and we are much more interested in the way institutions functioned at the time of reference than in their origins.6 Recent editors of Spanish charter collections – interestingly – use a range of different terms to describe countergift in their headings for individual charters but rarely discuss it: confirmación, contradonación, oferción, robra, en homenaje, even launegildo.7 The most pertinent modern treatment known to me is grounded in recent social theory, and Reyna Pastor, in particular, uses the notion of countergift explicitly;8 the import of the reciprocities inherent in unequal exchange is further developed by Rodríguez López and Pastor in a subsequent paper.9 However, these works relate to a later period than that considered here and therefore do not engage with the tenth-century material; they are in any case concerned with a smaller area and a wider range of social and economic issues.

Countergift: the primary data By countergift I am referring to something given to a donor in return for receiving a gift; this was not the same as price, which was agreed in principle before a 5 E. de Hinojosa, ‘El derecho en el Poema del Cid’, in E. de Hinojosa, Estudios sobre la Historia del Derecho Español (Madrid: Imprenta del Asilo de Huérfanos del Sagrado Corazón de Jesús, 1903), pp. 70–112, at pp. 109–10; C. Sánchez-Albornoz, Estampas de la vida en León durante el siglo X (Madrid: Espasa Calpe, 3rd edn, 1934), p. 135. 6 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: School of History, University of Leeds, 1977), pp. 105–38; W. Pohl and H. Reimitz (eds), Strategies of Distinction: The Construction of Ethnic Communities, 300–800 (Leiden: Brill, 1998). 7 Emilio Sáez, in his introduction to Li, pp. xli–xlii, is a notable exception. 8 R. Pastor, E. Pascua Echegaray, A. Rodríguez López, P. Sánchez León, Transacciones sin mercado: instituciones, propiedad y redes sociales en la Galicia monástica 1200–1300 (Madrid: CSIC, 1999), pp. 21, 222. 9 A. Rodríguez López and R. Pastor, ‘Reciprocidades, intercambio y jerarquía en las comunidades medievales’, Hispania, 60, no. 204 (2000), pp. 63–101.



transaction and was handed over in order to get possession of goods purchased. In this set of tenth-century texts countergift was differentiated from price: different words were used and gift documents had a different format from sale documents. Sale format was one kind of diplomatic, and normally included price; gift format was a different kind of diplomatic and sometimes, though by no means always, included a countergift. Of course, in practice the difference between gift and sale was not nearly so clear-cut and there was a decidedly grey area between the two.10 Here, however, is an example of a standard sale of property, of 7 March 950, by a married couple to the monastery of Sahagún: Ego Frunimius una cum uxore mea Eilo. Vobis domno Vincentio abbati vel omni collegio fratrum Sanctorum Facundi et Primitivi. In Christo salutem. Placuit nobis propria nostra voluntate ut vinderemus vobis iam dicto abbati proprias nostras vineas et terras et casas in loco certo in Villa de Helias. [. . .] Et pro hoc accepimus de vos in precio xviim solidos de argento quantum nobis bene complacuit; et de precio nichil remansit in debito aput vos. Ita ut de isto die ipsa hereditate in vestro iure permaneat et de Sancto Facundo faciatisque ex ea quod vestra extiterit voluntas. [. . .] Facta carta vendicionis nonas marcii, era DCCCC LXXX VIIIa. Regnante rege Ranemiro in Legio. [. . .]11 And an example of a record of gift with countergift, from a family to the monastery of Cardeña, of 22 April 945: Ego Armentario Didaz, prona et spontanea mea volumtas, una pariter cum uxor mea Tegridia et filiis nostris Rodrico et Munnio, ut pro remedio animabus nostris sibe pro animam pater meus Didaco Didazi, qui iam obiit e seculo, ut in vestris orationibus sacris illum Deo conmendetis, tibi patri nostro domno Stefanus, abba, vel omne collegium fratrum cohabitantes in domum Sanctorum Apostolorum Petri et Pauli, qui est situm cenobium locum Caradigna, tradimus vobis in ipso pozo, qui fuit ex patre meo, pro ipsius anima quartam partem, et aliam quartam rationem in ypso pozo tradimus et concedimus vobis [. . .]. Et accepimus ex vobis in honore, id est, sella in XXXa solidos et mantella; ut de odie die abeatis, vindicetis et ex proprio iure nostro sit extraneo et in vestra potestate maneat et confirmatum per secula. [. . .] Facta carta traditionis seu donationis die notum X kalendas maias, era DCCCC.LXXX. IIIa, regnante rex Ranimiro in Obeto et comite Fredinando Gundissalbiz in Castella. [. . .]12

10 W. Davies, Acts of Giving: Individual, Community, and Church in Tenth-Century Christian Spain (Oxford: Oxford University Press, 2007), pp. 135–8, 218–21. 11 S122. 12 C50.



The word used for countergift in these texts was usually either honor or offertio, occurring in almost equal numbers, although others occasionally feature (confirmatio twice; beneficium, prestacio, and signum once each). Niermeyer’s Dictionary recognizes twenty-six different uses of honor, which had a very wide range of meaning in early medieval texts, including ‘charge’;13 its use is easily understandable in this context. The use of offertio is more surprising: one might naturally expect it to refer to a gift to a church, given the frequent use of offere as the main verb of giving in charters of donation to an ecclesiastical body and given the significance of the liturgical offertory;14 however, when used to describe a countergift, offertio can refer to objects exchanged by two lay parties and can be used to describe a gift from a church to a lay person, rather than to a church. I cannot see any intrinsic significance in the choice of one word or the other: there is no difference in date range, nor in the nature or status of the transacting parties, nor in the kinds of gift for which it is a return. In fact, the choice of one word or the other seems to depend on the habits, training, and preference of the record-maker. So, for example, Cardeña scribes use honor; Galician and Portuguese scribes virtually always use offertio; practice on the meseta is more mixed, although the larger ecclesiastical houses tend to use offertio whereas local records often use honor. (This is a pattern which occurs in other respects: Castilian writing habits tend to differ from Galician, while those of the meseta are more mixed – the use of regnante dating clauses is a case that makes the point well.)15 The number of references to countergift in this corpus of charters is not large: they occur in just over 3% of the records considered (I have noted 72 cases, including references in three suspect charters). Percentages per individual collection often reflect this low proportion, but countergifts occur significantly more often in the substantial body of Cardeña material and significantly less often in some of the large Galician collections, Sobrado and Celanova most notably.16 They are more common from the 940s onwards than before; that largely reflects the availability of records, since there are more records after the 930s;17 however, they are disproportionately common in the 980s and 990s, when the incidence of

13 J. F. Niermeyer, Mediae Latinitatis Lexicon Minus, 2 vols (Leiden: Brill, 2nd rev. edn, 2002). 14 D. Ganz, ‘Giving to God in the mass: The experience of the offertory’, in W. Davies and P. Fouracre (eds), The Languages of Gift in the Early Middle Ages (Cambridge: Cambridge University Press, 2010), pp. 18–32. 15 W. Davies, ‘The incidence of princeps in the ninth- and tenth-century charters of northern Spain’, in J.-M. Picard, H. Oudart, J. Quaghebeur (eds), Le prince, son peuple et le bien commun (Rennes: Presses Universitaires de Rennes, 2013), pp. 217–32. 16 The Oviedo and Eslonza percentages look high, but the total number of tenth-century charters in each collection is small and the percentages cannot be statistically significant; the San Millán percentage looks low, but many of those records are truncated and may not reflect the entire content of originals. Navarre and Aragón examples are notably absent, unless one counts the San Millán occurrence as Navarre; the text of Leire 8 could be interpreted as gift and countergift (Leire is in Navarre), although the words isto mercato push the transaction more towards sale. 17 See ch. 6, ‘Sale, price and valuation’, pp. 98–100.



countergift increases, although this was a time when the total number of surviving records declines. It looks as if the practice of recording a countergift was becoming more common at the end of the tenth century, with the curious exception of Castile, to which I shall return. Table 8.1 Percentage of charters with countergift, per collection, with presence of countergifts in 980–99 Region





Sobrado Samos Celanova Portugal Otero León Eslonza Sahagún Oviedo S. Toribio Cardeña Valpuesta San Millán All

1.0 3.4 0.9 4.0 2.2 3.6 5.0 3.9 8.3 3.2 6.4 2.8 0.8 3.2

× ×


Far North Castile Castile/Navarre Overall

× × × × ×


The function of countergifts Nearly half (43%) of these countergifts, whatever the Latin term used, were explicitly given in order to secure confirmation of a previous gift, a point made by Sánchez-Albornoz many years ago.18 In other words, A gave something to B; B then gave A a countergift in order that A should confirm the gift B had received and make it permanent. The following example, from King Vermudo II to a trusted aristocrat in 989, and vice-versa, is typical: Veremudus serenissimus serenissimus [sic] princeps tibi Munnio Fredenandiz, salutem in Domino semper, amen. Magnus est enim titulus donationis in quo nemo potest actum largitatis inrumpere sed quicquid grato animo offertur debetur libenter amplecti. Quapropter annuit namque huius serenitati regni nostri glorie ut daremus tibi, sicut et donamus, villam iuri nostro devitam [. . .] et est ipsa villa nuncupata Toral. Sic tibi eam concedimus ab intecro per terminis suis cum omnem populationem suam et omnibus prestationibus suis quantum ad ipsam villam pertinere videtur. Et pro confirmandam hanc nostram preceptionem accepimus de

18 Sánchez-Albornoz, Estampas de la vida, p. 135.



te offertionem cavallo nobis placivile. Ut quiete obtineas ipsa villa ex dato nostro omnibus diebus vite tue et post relinquas cui tua institerit volumtas vel cui successorio iure per legem debitum fuerit. [. . .]19 Or these two, between – respectively – two lay persons in 960, and King Ramiro III and the monastery of Sahagún in 971: [. . .] Et ad corroborandam cartam istam accepimus de vos in vestimentum linteum lineum que nobis bene complacuit ut scriptura istam plenam habeat firmitatem. [. . .]20 [. . .] Et Seriqus abba una qum coleium nostrum damus ad vobis Ranimiru rex et regina Gelvira kaballu baio obtimum et pannu de sirgu valiente C soll[i]dus de argentum ad istu testamentu confirmante quantu ad vos plaquit. [. . .]21 This practice of fixing gifts was not peculiar to northern Iberia; it happened in Italy in the early and central Middle Ages, where for a period it was an important mechanism for making gifts legally stable and for fixing the obligations of the initiating donor in the event of a change of mind or of claims by third parties; it was cited in charters from the eighth century onwards, was particularly common in the eleventh century, but petered out in the twelfth, as other forms of guarantee came to replace it.22 In Spain, confirmatory countergifts occurred in response to gifts from lay person to lay person, from church to church, and from lay person to church, and they involved parties of higher and lower status. The primary gifts which they confirmed were land or churches or both. The objects given in response could be of high value, like the fine bay horse and expensive silk given by Sahagún to King Ramiro III cited above; or rather lower, like the ten sheep and fine mattress given by the monastery of Santos Justo y Pastor to a lay couple in 959; or much lower, like the single colt given by the monastery of San Salvador de Matallana to two nuns in 990 in response to their gift of land and houses in Alija and other locations in the Cantabrian Mountains, or the ox and skin given by a lay couple to the laywoman Leodegundia, also in 990, for her gift to them of a parcel of land in northern Portugal.23 Countergifts which were explicitly given in confirmation thereby secured the initial gift and made it permanent. Another fifth (21%) of recorded countergifts involve the return of things that seem to be of more symbolic value, usually one thing per transaction, and usually an animal, but occasionally cloth or a skin (but not the very low value countergifts, 19 20 21 22

Liii530. S178. S261. C. Wickham, ‘Compulsory gift-exchange in Lombard Italy, 650–1150’, in Davies and Fouracre (eds), The Languages of Gift, pp. 183–216. 23 Lii313, Liii535, PMH DC 159.



e.g. handkerchiefs, found in some Italian transactions). Countess Mumadonna, her son, and her grandchildren gave the monastery of Cardeña a stretch of mountainside in 935 and received a mule in return; the woman Egilo gave the monastery of Abellar substantial property in 954 and received a rabbit skin in return; a lay couple gave San Rosendo of Celanova their share in the villa Cobreiros in 955 and got a ‘best animal’ in return.24 These records do not have explicit statements about confirmation but the acts look like standard gestures – a standard response to a gift by returning an animal. The function of such gestures could well be confirmatory; indeed, that is the best explanation of them, since some countergifts of single animals have a confirmation phrase in any case; we also have another record of the Alija transaction noted above, and this record does not mention confirmation, which makes it perfectly clear that such records can nevertheless record confirmatory transactions.25 We cannot always be certain, however. Again, these countergifts were made by people of every kind of status, throughout the period under consideration, in all parts. A further 29% have a different profile. Most of the countergifts in this group were given by churches in response to gifts from aristocrats to those same churches. The primary gifts include both landed property rights and churches and monasteries. It is impossible to assess value in absolute terms, because there are insufficient data to do so, but it is nevertheless striking that nearly all of these cases involve countergifts of very high value by comparison with other valuations – of the order of 100 to 500 silver solidi, where valuations are given; compare prices of 4, 5, or 6 solidi that are frequently cited in sales, or the ox worth 8 solidi in 975.26 It looks as if the value of the countergifts in this group may often have approached the value of the initial gift: in 943, for example, two horses, 400 sheep, a saddle, and cloth were given by the monastery of Cardeña to Count Assur Fernández in return for land and a watercourse.27 These transactions look much closer to sales – contracts agreed for a price – although they are recorded in gift format without a price clause. Sometimes it is made quite clear that a deal had been done, since the monastery is recorded as asking for the initial gift in the first place, as Abellar had asked for land in Marialba.28 Sometimes a deal, or pressure, seems to lurk below the surface of a text, as in the record of the deacon Sisnando Menéndez, who gave half a church to the monastery of Santa María de Valdevimbre in 951, for the salvation of his soul, and gave the other half for an offercio of 20 silver solidi; this record is entirely framed in terms of gift but the content seems more like sale.29 It is also notable, firstly, that many of these cases come from the Cardeña cartulary; and, secondly, that hardly anything is recorded as sale in that cartulary until the mid-960s, whereas from the mid-960s the sale 24 25 26 27 28 29

C27, Lii265, Cel570. Liii534. See ch. 6 above; V46. C42. Lii446. Li231.



format is used and this kind of transaction disappears. In other words, it looks as if Cardeña scribes were using gift format for virtually all transactions until that period of change in the 960s. In other collections, as we have seen, sales were usually differentiated from gifts by their different diplomatic form and by their use of the concept of price. The apparently surprising decline in Castilian examples of countergift late in the tenth century, at a time when evidence elsewhere was increasing, is therefore more the end of a quirk in recording habit than the disappearance of a standard practice. The remaining small group of records is miscellaneous in character and has a range of unusual features. Once, in 944, in the presence of the people of Burgos, King Ramiro II gave Cardeña some land that had once been an orchard; the syntax is unclear but the people may have received the honor from Cardeña, in order to reinforce the initial gift and add extra guarantees.30 On other occasions an additional (counter)gift went to the saio (a legal officer who, for example, took oaths from suspects and supervised the ordeal), who had formally handed over the primary gift;31 or a countergift went to the spokesman for a group.32 All three of these cases look like actions intended to establish additional guarantees. On another occasion, in 999, a third party provided the countergift for the beneficiary to give to the donor: a laywoman provided two fine bridles and expensive bedding for the monastery of Sahagún to give to the bishop of León, who had given the monastery three churches; this case looks like another ‘concealed sale’, the ‘price’ being provided by the real benefactor, the woman.33 In other cases the record is an unusual mixture of gift and sale format, or the word offertio seems to mean something other than countergift.34 Whatever we might think of these categories, countergift was clearly unusual in tenth-century Iberia; there are relatively few cases. It certainly cannot have been used as any kind of norm, as it seems to have been in eleventh-century Italy for special categories of gift, such as gifts between lay persons and charters of promise. We have no direct evidence on why there was sometimes a countergift in northern Spain and at most times there was not. We can only presume that sometimes the beneficiary was concerned about the solidity of the transaction and for reasons unknown wanted extra guarantees.

30 C46. 31 C42; cf. E13, of 938, in which ‘ad Uirmundus Nunioni Sala Caremiz’ is probably a misreading of a saia carmez (cf. C54) given to Vermudo as an additional (counter)gift, where the main countergift was to the donor, King Ramiro II. 32 C54. 33 S359; cf. Leire 6, in which King Sancho Garcés I gave Bishop Basilio goods in order to enable him to give tithes to Leire, in about 918; this text is doubtful as it stands, because of its precocious reference to tithes, but the dynamics of the reported transaction are very interesting. 34 Liii584; Ov23, PMH DC 160.



Guarantees Let us look in more detail at some of these cases. The confirmatory function of much countergift is emphasized by the facts that some records are primarily confirmations: the offertio/honor was given in order that some separate transaction be confirmed, for example by the king or by a dead donor’s family; and by the fact that on one occasion the countergift (honor) seems to have been given in respect of an agreement not to construct houses near a church.35 One Liébana charter is very explicit that countergift was intended to provide extra guarantees: when Fernando Díaz and his wife gave their monastery of San Esteban in Mieses to the nearby monastery of San Martín de Turieno, they received a substantial countergift of an expensive mule, two horses, and their colts, for confirming the gift; the text then continues ‘et facio inde securitatem sicut tenedes in vestrum pactus’.36 That the donor must stand by his or her gift and support it, whatever the challenge, is emphasized by the fact that a group of donors might receive a countergift each, as Sarracino Ovecoz was given a valuable horse, his brother an ox, and his mother a cow, for their gift of a house and church to the monastery of San Esteban de Salcedo in 950.37 It seems odd to find countergift used as a guarantee mechanism in northern Iberia at this time, for there were other guarantee procedures in use: it was not uncommon to use a personal surety if guarantees were required, literally a third party whose function was to guarantee the obligations of the principal parties. As in the case of countergifts, there are not many cases in the corpus of tenth-century texts considered here and they occur only in 2% of the corpus. 39% of these references to personal sureties come from Castile, 30% from the meseta and 27% from Portugal.38 As in the case of countergift, there are relatively more in the 990s than at earlier periods – in fact they occur at twice the average frequency. These sureties did a mixture of things: in part they acted as semi-private policemen, responsible to the community as a whole but appointed by the contracting parties; in part they took on financial responsibility themselves and paid up if something went wrong with the transaction; and in part they were a formalization of family responsibility, public representatives of the wider family: when a married couple defaulted in paying a debt of 72 measures of bread and wine, late in the tenth century, the wife’s brother, who was the surety for the debt, paid up from family property in Orga, near Celanova.39 By the end of the century, sureties

35 S293, V36; C36, cf. W. Davies, Acts of Giving: Individual, Community, and Church in TenthCentury Christian Spain (Oxford: Oxford University Press, 2007), p. 137. 36 T75. 37 SM57. 38 See ch. 17 following, ‘On suretyship in tenth-century northern Iberia’. 39 Cel368.



were also being used to guarantee that one party would also in effect answer to representatives of the state, as the surety Gelmiro did when he handed over an estate, the ownership of which had been disputed in a court-case heard before Count García Gómez.40 Suretyship was extending in the late tenth century, as it continued to do in the eleventh and twelfth centuries, with new uses in urban contexts, like procedures for handling strangers. One therefore has to ask if there was a difference between the guarantees provided by sureties and those provided by countergift. Looking at the recorded cases, it is clear that sureties were used to guarantee things that were to happen in the future – sales, debts of one kind or another, court procedures, and outcomes of court-cases (many of which can be seen as a kind of contract); their use in gift transactions is extremely rare. Countergift, by contrast, was used to guarantee transactions that had already happened – a confirmation of gifts that had already been made, making them secure. Sureties seem to have been the mechanism for guaranteeing sale and future obligation; countergift was the mechanism for guaranteeing gift and the continuing security of past actions. Very interestingly, the three exceptions to the use of sureties in gift transactions all come from Castile, and one specifically from Cardeña records, of 947.41 Despite other possible explanations, the text makes it seem likely that this gift involving a surety followed a court-case. The other two cases are both exceptional: sureties were used to provide guarantees that a group of friends and neighbours would not reclaim their contributions to a monk’s burial expenses in 939; and sureties were used to guarantee the gift of a man to San Millán de la Cogolla – in effect this was a guarantee that the man would pay his rent to the monastery; though the form of the record is unusual, this is really a case of guaranteeing future obligation, a standard surety function.42 While the exceptions underline the fact that Castilian practice was sometimes distinctive, essentially they do not detract from the proposal that sureties were used to guarantee future obligation while countergift was used to secure a past action in the tenth century.

Discussion In discussing function I drew a distinction between countergift as confirmation mechanism and countergift as de facto price. While the distinction is useful for the purposes of analysis, it would be a mistake to suppose that tenth-century people saw a stark difference between the two. Real situations were clearly much more complex. In fact the categories overlap: just as symbolic objects might explicitly be given to secure confirmation, so could extremely valuable goods; just as some people devoted significant resources to securing confirmation, others devoted

40 S356. 41 C59. 42 V17; SM93.



significant resources to buying property that they wanted. We do not know why some countergifts were symbolic and some had economic value – low and high value countergifts occur in all categories, however one defines them. It is clear from León material, in particular, that the countergift mode of transacting was one acceptable way of dealing with low status, low value transactions, just as it is particularly clear from much Cardeña material that it was equally acceptable as a way of dealing with some high-value transactions.43 Nor do we know why some transactions were recorded as sale and some as gifts with countergifts of comparable value. We do not have sufficient information about individual circumstances to answer these crucial questions. The answer to the latter lies partly in the recording tradition and in the preferences of the scribes who drew up the texts. But some scribes were making choices about whether to use sale format or gift/countergift, particularly on the meseta; did they perceive a difference in the quality of the personal relationships of the actors? There are a few clues about what might have conditioned their choice of format, always bearing in mind that far more transactions were recorded as sale than as gift/countergift. Three of the countergift transactions are explicitly between relations: brother to sister, nephew to aunt, and wife to husband; and one is between two bishops.44 It is not difficult to see how it might have been thought more appropriate to cast transactions between closely connected individuals within the giving rather than the commercial framework, although there are plenty of other cases between such parties which are recorded as sale. Secondly, and differently, there is a group of countergift transactions between kings and high aristocrats. In three of these the king explicitly gave the aristocrat land his father had had before him; the countergift in these cases seems to have been in practice a consideration in order to secure confirmation of the aristocrat’s possession of the land – rather like an entry fine for a heritable fief. It is inconceivable that, at that time, such a transaction would be presented as the purchase of a fief; gift and countergift was the more appropriate format for socio-political networks of this kind.45 These are three explicit cases. Several of the other cases of royal donations of property to aristocrats, with countergift, could be interpreted in the same way, although they lack explicit detail of background. There are also other kinds of explanation. Without losing sight of the fact that countergifts occurred in transactions between lay parties too, it remains the case that two thirds of gifts which attracted recorded countergifts were gifts to an ecclesiastical individual or community, and three quarters of those were to bishops or powerful monasteries. Many were pro anima gifts, in contrast to Italian practice, where it was rare to provide a countergift in response to a pro anima gift. In this Iberian world, it must be likely that these countergifts arose from beneficiary

43 E.g. Li168, Lii325. 44 PMH DC 14, S313, S370; Lii333. 45 Cel234, S285, and by implication S365.



initiatives, as a result of their desire to secure their acquisitions. The accumulation of property by leading monasteries is striking in the later tenth century and this seems to be another product of that drive. The fact remains that the use of both countergift and surety was unusual, although they both continued in use through the eleventh and into the twelfth centuries. There were also other modes of confirmation – for example by writing, as in the case of C88, which confirms the sale recorded in C87; or, as became more common in the twelfth and thirteenth centuries, by payment of a robra to the vendor in cases of sale, a practice prefigured by the vino in albaroc of C123 (965) and the flagon of wine in honore of Valpuesta 36 and OD47 (960–1 and 999).46 Why were these mechanisms of countergift and suretyship sometimes used in apparently straightforward transactions in the tenth century? As I indicated above, proportions were small – 3% and 2% respectively of all records (which means that about 6% of gift records have countergifts and 4% of sale records have sureties). The background may hold some further clues. Visigothic law had proclaimed that sureties needed to be given to the buyer in sale transactions if the vendor was not of good character.47 Visigothic law was well known in the tenth century; copies of texts were made, such as that in the Codex Albeldensis of 974–76; and it was sometimes cited in charters.48 This legal background provides good reason for the occasional use of sureties. Perhaps we can extend this, by analogy, to gift and countergift. Were countergifts used to add security when beneficiaries doubted the donor, or, more importantly, his or her family? There are, then, many things we do not know, but use of the notion of countergift throws some interesting light on tenth-century perceptions of relationships, and reveals that we have some clear evidence on what many might call proto-feudal relationships. There are clearly several different explanations for the occurrence of countergift in these texts: doubting the commitment of a donor’s family provides one kind of explanation for its occasional use; perceiving the relationship between the transacting parties as one which lay within the dynamic of friendship rather than that of commerce provides another; ecclesiastical initiative to get donations secured provides another; and scribal idiosyncrasy provides yet another. None of this was unwaveringly systematic. There were variants and oddities, and occasional confusions. A few countergifts confirm sales and exchanges rather than gifts, just as sureties occasionally guaranteed gifts rather than sales, and a confirmatory

46 R. Fernández Espinar, ‘La compraventa en el derecho medieval español’, Anuario de Historia del Derecho Español, 25 (1955), pp. 293–528, at pp. 486–95. 47 Forum Iudicum 5.iv.2, in Leges Visigothorum, ed. K. Zeumer, MGH LL nat. Germ. 1 (Hanover: Hahn, 1902), pp. 35–456, at p. 218. 48 A. Millares Carlo, Corpus de códices visigóticos, prep. M. C. Díaz y Díaz, A. M. Mundó, J. M. Ruiz Asencio, B. Casado Quintanilla, E. Lecuona Ribot, 2 vols (Las Palmas de Gran Canaria: Gobierno de Canarias, Universidad de Educación a Distancia, Centro Asociado de Las Palmas de Gran Canaria, 1999), no. 49. Liii587, which has a countergift, Sam132.



payment was occasionally made to a vendor in addition to the price.49 But the broad lines of the distinction between use of countergift and surety in the tenth century are clear enough, and mostly consistent. Providing a countergift was largely about keeping what had been received; its somewhat rare use must in the end be a comment on trust.50

49 The countergift in PMH DC 151 was made to a third party to confirm a sale; while this may technically be closer to a robra, it is worth considering in the context of countergifts because the word used is offertio and it was explicitly given for confirmation; cf. Lii325. 50 I am grateful to Chris Wickham for stimulating me to think about Iberian countergift; to participants in the Queen Mary Symposium for helpful reactions to the preliminary version of this piece and for a stimulating day; and in particular to Alan Deyermond for his initiative and for very helpful comments on a written draft.


9 N O T I O N S O F W E A LT H I N T H E C H A RT E R S O F N I N T H - A N D T E N T H - C E N T U RY C H R I S T I A N IBERIA This paper is about words and about values, using the extant charter material from northern Iberia.1 This material records about 100 transactions from the ninth century and over 2,000 from the tenth (excluding Catalonia), especially from the 930s onwards. These come from localities across northern Spain and Portugal, although there are relatively few from Navarre and Aragón. The biggest collections are from the monastery of Celanova in Galicia; the episcopal collections from León; the monastery of Sahagún on the meseta; and the monastery of Cardeña in Castile.2 There are some smaller collections which are important, however, especially those from the monasteries of Santo Toribio in the far North and Otero de las Dueñas on the northern edge of the meseta – important for its collection of originals from a lay archive – and the collection of material from miscellaneous sources in northern Portugal.3

Words Let me begin with words, as a way into concepts and perceptions. These Iberian charters use several words for wealth and riches – divitiae (as they also use dives), opes, thesaurum. For the most part the words occur in the preambles of donation charters (that is, words of pious introduction and of explanation for making the gifts), and they express variations on the theme of abandoning wealth in order to store a better treasure in heaven. There are plenty of explicit references to Christian notions that it was undesirable to have wealth, which would become an obstacle to entering heaven. Hence, when the famula [Dei] Leocadia made a gift of extensive properties to the monastery of Abellar in 951, the record is preceded by several biblical quotations, or versions of them, including ‘Sell everything you have, give alms, and find your treasure (thesaurum) in heaven, where you will

1 First published in J.-P. Devroey, L. Feller, R. Le Jan (eds), Les élites et la richesse au haut moyen âge, Collection Haut Moyen Âge, 10 (Turnhout: Brepols, 2010), pp. 265–84. 2 Cel; Li, Lii, Liii; S; C. 3 T; OD; PMH DC.



not lose it’, an easily recognizable version of Luke 12.33: ‘Sell that ye have and give alms; provide yourself [bags which wax not old] a treasure in the heavens that faileth not, where no thief approacheth, neither moth corrupts’.4 The grant by another woman, in northern Portugal, in 961, is prefaced by an extended admonition to take worldly riches (divicias corporales) and make ‘from Mammon’ a permanent treasure in heaven (thesaurum permanentem in celis).5 When Ermigia, another lay woman committed to religion, gave property in La Tierra de Campos on the meseta to the nunnery of Santiago in León, in June 970, the record is preceded by similar citations, including one – wrongly attributed to St Paul – which advocates renouncing the search for profit (lucrum).6 There are other examples, including those associated with male donors (this was not a distinctively female approach). However, these expressions are much more common in charters recording the transactions of religious actors, especially the committed laity – confessi/-ae, ancillae, famuli/-ae Dei – than in those of others; the authors of such charters had a habit of expanding on standard formulas in elaborate and creative ways.7 A distinctive subset of the élite was therefore conscious of, and invoked, the New Testament injunction, a subset whose members often seem to have drafted their own charters, which had individual characteristics and varied widely from the standard models.8 Concerns about riches, rich people and the moral danger of wealth were part of the common currency of religious writing across northern Iberia. By contrast, however, these expressions are not characteristically associated with the grants of the wider lay aristocracy, whoever drafted the records. Although preambles are by far the commonest context of explicit references to wealth, occasionally ‘riches’ were associated with living people: the priest Melic endowed the church of San Salvador on the river Porma with considerable wealth (largissimas divitias), sometime before 960;9 the powerful royal, Elvira, the León king’s aunt, referred to a rich man (dives), with great wealth (multa opes), whose property she had come to control by 970.10 Significantly, however, none of these expressions is particularly common. Much more common are reward words (merces, remuneratio, praemium), again in preambles, in the context ‘do this and get

4 Li236; cf. Cel2 (942), Li256 (952), and the blatantly commercial Lii276 (937–54): ‘Vendite terrena ut ematis celestia [. . .] Dum tempus abetis, et pretium in manus vestras tenetis, redemitte animas vestras’. 5 PMH DC 82. 6 Lii413 – ‘set cum dicit Paulus Apostolus: “Ab bono operum animum sollicitat et mundiali lux contempnendo abrenunciandoque nos ab se lucri adpeteret, nec miser homo de hac vita sine fructu bonorum operum animum discedat”’. 7 Cf. H.-W. Goetz, ‘Idéologie (et anti-idéologie) de la richesse au Haut Moyen Age’, in Devroey, Feller, Le Jan (eds), Les élites et la richesse, pp. 33–58, on ‘aversion’ to wealth. 8 See further W. Davies, Acts of Giving: Individual, Community, and Church in Tenth-Century Christian Spain (Oxford: Oxford University Press, 2007), pp. 102, 177–9. 9 S183. 10 S255.



your reward in heaven’.11 It was more usual to express the quest for virtue in terms of reciprocities, which were not tied to any relative or absolute notions of wealth. Such expressions of ultimate reward are common in records of aristocratic grants, especially – though by no means exclusively – those of kings, where the drafting of the text was done by members of the royal household or of the monasteries that were beneficiaries. These expressions are much rarer in records of peasant grants, which were often drafted by priests or other locals. There are then two points of note: firstly, the discourse of renunciation and heavenly reward is strongly associated with writing that comes from the high culture of courts and powerful monasteries. Secondly, the ethic of reward is much more commonly expressed than that of the dangers of wealth. Wealth and its dangers were nevertheless a characteristic concern of some kinds of religious writer.

Rich people What, then, made a person rich? Firstly, and obviously, there were plenty of people who, from our point of view, were rich in landed property. For example, the man Pelayo who handed over his share of twelve estates to Countess Ilduara in 940 (as penalty for homicide); or Velasco Muñoz who, with his wife and sons and two female relations, gave the monastery of Boñar six estates, a mountain, a valley and four vineyards in 996 (Velasco was the brother of a count, from a powerful family operating on the southern side of the Cantabrian Mountains).12 In both cases the properties were presumably the tip of the iceberg of a much greater landed estate, since the donors continued to live as aristocrats and support families thereafter. Other aristocrats made extensive post mortem donations, which may represent a larger proportion of their total landed assets: the Galician count Rodrigo and his wife Elvira gave over thirty estates and appurtenances to the monastery of Sobrado; the religious woman, ancilla Dei, Eiloni, gave eleven estates and at least twelve portions of others to the Galician monastery of Santa Cruz in 985; while Count Munio Fernández bequeathed four estates and other tracts on the meseta, near León, to the monastery of San Juan Bautista, reserving similar proportions for each of his four children.13 There were also many who made successive gifts, like those of the Navarre kings to the monastery of San Millán de la Cogolla; and many who kept buying new lands, like Bagaudano and his wife Faquilona in the Liébana in the 910s to 930s, or Ansur and his wife

11 Cf. B. Jussen, ‘Religious discourses of the gift in the middle ages: Semantic evidences (second to twelfth centuries)’, in G. Algazi, V. Groebner, B. Jussen (eds), Negotiating the Gift: PreModern Figurations of Exchange (Göttingen: Vandenhoeck & Ruprecht, 2003), pp. 173–92; he argues, from a wide range of European material, that the language of reward was elaborated in sermons and tracts of the eighth to eleventh centuries, especially in the sense of God rewarding man’s opus and labor, pp. 187–9. 12 Cel456, S352. 13 Sob4 (959); Cel462; Liii701 (1011), Liii743 (1016).



María a generation later on the meseta, or the monastery of Sahagún for much of the century. They look rich, from our point of view, and we may define them as wealthy, because they had relatively much land and dealt in quantities that vastly exceeded those of peasants. Their landed resources clearly enhanced their power and status, but they tend not to be described as rich in these texts; they simply occur as people with much property.14 If the landed were not called rich, it is worth taking a look at movable wealth and its written associations. I will do this in three ways: by considering church treasure, things handed over as price or countergift and things with an associated valuation. A) Church treasure ‘Treasure’ (thesaurum) is a word primarily used in these charters to describe the physical, movable property of a church – hence thesaurum ecclesie and occasionally tesaurum templi.15 What was church treasure? Not bags of money, since coin was not minted in northern Spain until the later eleventh century. Rather, it was the furnishings and plate of the church, primarily for liturgical use. A number of charters provide lists of these treasures, sometimes short, sometimes very long; sometimes rather generalized, sometimes extraordinarily specific. When Velasco and family gave landed property to Boñar, they also gave three altar cloths, a silk stole, two chasubles (one red, one linen), a silver chalice, a bronze candelabra, a bronze cross, a bronze thurible, two bells, three glass chandeliers and seven books. So far, this might be what one would expect as church treasure. However, there were also, in addition to ‘all kinds of utensil’, four ornate bed sets (one silk), a fine bedspread, ten feather mattresses, five tablecloths, two sets of linen, a cushion or pillow, three cups, a cauldron, four spoons, a large dish, a roasting dish, a ladle, a saltshaker, two flasks, two bowls, two jugs, un-numbered barrels, chairs and beds, and seven kinds of stock.16 Much more elaborately, when Bishop Rosendo endowed the monastery

14 As Juan José Larrea points out (personal comment), if we had narrative texts as well as charters, we might have more descriptions of rich people. 15 S22 (921), T34 (925), for example; Cel558 (952); a cartario de thesauro domini Salvatoris, drawn up for the monastery of Samos after a period of destruction, appears to deal with landed as well as other resources: Sam35 (944). Cf., for different uses, D. Iogna-Prat, ‘Préparer l’au-delà, gérer l’icibas: les élites ecclésiastiques, la richesse et l’économie du christianisme (perspectives de travail)’, in Devroey, Feller, Le Jan (eds), Les élites et la richesse, pp. 59–70. 16 S352 (996): ‘cupas, lectos, sedilias, vel omnia utensilia, lectos II palleos, alio tramisirgo, laneo I, almuzalla I, plumazos X, mutas litones de mensa IIII, literatos I, pares de lineas II, fazale I de argento, copos II, caldera I, culiares IIII, offertoria I, frixorio I, trulione I, salare I, tazola erea I, aredoma erea I, espanesca I, concos II, aquamaniles II, alfagara facenzal moreda I, frontales II, oral de sirgo I, casulla erac vermelia I, alia de lino, calice de argento I, candelabrum ereum I, cruce erea I, incensario ereo I, antifonario I, psalterios II, mistigo I, ordino I, prego I, comico I, campanas II, coronas vitreas III, equas V’ and other animals. One could quibble over the meaning of some



of Celanova in 942, his list was longer but similar and included two silver crosses (one with gold and gems), one bronze and two silver candelabras, three silver chandeliers (one with gold and gems), a lamp, gold thuribles, two silver reliquaries (with gilt decoration), five silver chalices (three with gilding and another with gems), bells, silver and gold belts, ten albs, twenty-three chasubles, nine stoles (four with silver and gold thread) and at least twenty-one books; together with bed covers,17 twenty-one blankets (seven ornate), twenty-three feather mattresses (ten ornate), six fine bedspreads, furs, two ornate rugs, two silver flasks, a gilt tureen, gilded and bronze goblets, six gilded and twelve bronze dishes, three gilt mugs, a gold cup, ivory and glass tumblers, two further tureens (bubalinas), a dish, a silver table service, four pitchers, seven engraved bowls, two candlesticks, two bowls from the East, nine flasks (of eastern origin) and twenty tablecloths – amongst much more.18 In other words, ‘church treasure’ included the books, church plate and vestments for liturgical use that one would expect, but also precious, luxury objects for the household; these household goods of extra fine quality were often deemed to be imported – metal vessels from al-Andalus, Moorish blankets and cloaks, metalwork and cloth from ‘Iraq’.19 One can occasionally find very similar lists associated with the laity. For example, there is a late tenth-century inventory of goods which a couple of high-ranking lay aristocrats gave to their daughter as dowry (these were people who moved in

of these terms but there is a rationale in the order of the items, which allows some deductions; helpful guides can be found in: C. Sánchez-Albornoz y Menduiña, Estampas de la vida en León durante el siglo X (Madrid: Espasa Calpe, 3rd edn, 1934), App. III and IV, pp. 173–99; E. Sáez, ‘Nuevos datos sobre el coste de la vida en Galicia durante la alta edad media’, Anuario de Historia del Derecho Español, 17 (1946), pp. 865–88, at pp. 886–8; M. C. Torre Sevilla-Quiñones de León, ‘Aproximación al estudio terminológico de la indumentaria nobiliaria leonesa a través de los documentos (ss. ix–xiii)’, in M. Pérez González (co-ord.), Actas II Congreso Hispánico de Latín Medieval (León: Universidad de León, 1998), pp. 865–74. ‘Bed sets’: lecto/leito seems to indicate a rather fancy bedroll; the precise meaning was in one case specified as blanket, mat and pillow, S183 (960). Mutas – Torre Sevilla-Quiñones de León gives ‘underwear’ as the normal meaning, but the mensa here seems to indicate a different kind of changeable cloth. Cf. the semantic range of ‘linen’ in English (i.e. both underwear and table covers), an observation I gratefully owe to James Campbell. 17 stragmina lectulorum; I am grateful to Ian Wood for the suggestion of ‘tomb covers’, but they may simply be bed covers since they are listed along with furniture rather than with liturgical vessels. 18 Cel2. 19 aredoma [. . .] espanesca (S352); almuzallas and alleapes (Li220), alliphaphes (Cel2); concas aeyraclis (Cel2), cf. casulla erac vermelia (S352). Cf. also Byzantine cloth: plumazos greciscos, in panno grecisco (OD50). Cf. Sánchez-Albornoz, Estampas de la vida, p. 18, n. 5; for imported textiles, see L. Serrano-Piedecasas Fernández, ‘Elementos para una história de la manufactura textil andalusí (siglos ix–xii)’, Studia Historica. Historia Medieval, 4 (1986), pp. 205–27. Some of the cloth could have been local versions of imported styles; there were certainly some workers making andalusí cloth in the neighbourhood of León in the early eleventh century; see C. Estepa Díez, Estructura social de la ciudad de León (siglos xi–xiii) (León: Centro de Estudios e Investigación ‘San Isidoro’, Archivo Histórico Diocesano, Caja de Ahorros y Monte de Piedad de León, 1977), p. 417.



royal circles and exercised judicial functions for the king): in addition to slaves and stock, she got the same kind of fine furs, blankets and mattresses as were given to churches; also the silver spoons (twelve), ladle, saltshaker, dishes, silver vessels, tumblers, bowls and jugs, as well as the more characteristic bride’s presents of gold and other jewellery, clothes, headgear, furs (some decorated with gold thread), towels and carpets and a mule which had a silver saddle and bridle.20 There is also a Portuguese list of goods given by husband to wife, as morning gift at marriage, in 946: as well as substantial landed properties, stock and furniture, there were a horse with a saddle and silver bridle, silk blouses, a precious cup and dish, lambs’ fleeces, silk stoles and jewellery.21 The very fact that churches often acquired their treasure as gifts from the lay aristocracy – like King Sancho Garcés I’s lavish gifts to the monastery of Leire and to Bishop Basilio round about 918 or Countess Mumadona’s even more lavish gifts to the monastery of Guimarães in 959 – makes it clear that these goods were owned by lay and clerical élites alike.22 Overall, then, ‘treasures’, ‘riches’, ‘wealth’, as owned by lay and clerical élites, seem to have had more to do with accumulations of movable goods, and especially precious things, luxuries, exotic special objects; insofar as ‘treasure’ was conceptualized, it was perceived to lie in movables rather than in the basic landed resource; to possess (and doubtless display) the visible, countable, graspable treasure was to have riches, as in many other societies, ancient and modern.23 B) Things handed over as price or countergift If we look at objects which were handed over in transactions of gift and sale of land – as countergift and price – there are some points of contact with these lists of movable wealth. There are just under 1,000 recorded tenth-century sales from the region; while some records which were in sale format clearly conceal negotiations which were far from commercial, many seem to have been straightforward commodity transactions.24 In many of these sales the price was recorded as paid in objects: A sold land to B for objects 1, 2, 3, as Dalla sold some parcels of land in Aleje to Munio Flaínez for two rams and a cheese and Godesteo sold a deacon a plot of arable land in Moreda for a goat and three quarters of grain.25 While 20 OD50. 21 PMH DC 56. It is tempting to suppose that these goods are gendered: however, although women were certainly given chattels, they were also given land, sometimes much land; and plenty of men gave and received chattels. 22 Leire 6 – there are suspect elements in this charter but it still makes the point; PMH DC 76. 23 Cf. J. Campbell, ‘The sale of land and the economics of power in early England: Problems and possibilities’, in J. Campbell, The Anglo-Saxon State (London and New York: Hambledon and London, 2000), pp. 227–45, at pp. 240–1: ‘the weight of interest and emotion is directed not towards land but towards treasures’. 24 See ch. 7 above, ‘When gift is sale’. 25 S163 (959), Cel451 (942). See ch. 6 above, ‘Sale, price and valuation’, for more detail of prices and valuations.



many of the sales were of modest parcels, and many of the objects handed over in payment were also modest – foodstuffs, basic clothing, single animals – the transactions of aristocrats were quite different. We find precious carpets, ornate bed sets, fine clothes, silver dishes, in other words, exactly the same things as constitute church treasure, although there were also, often, fine horses (as opposed to old nags or carthorses).26 So, for example, a woman received two silk robes from a lay couple in 933 when she sold them an estate in Lores; two couples who sold property to the monastery of Santos Justo y Pastor in the 950s were paid with a red carpet, a decorated mattress, a Moorish blanket and some packs (saccos – perhaps packs of cloth); a lay couple received a fine bay horse from the well known aristocrat Ansur in 974 for selling a substantial establishment by the walls of León; another lay couple received two carpets, a yellow tunic and a bed set in 976, for selling land to the monastery of Cardeña; and a woman received a precious sword, with gilt decoration, from a priest in 959 for selling her town house in León, by the bishop’s gate.27 The latter case may look surprising, with sword passing from priest to woman, but it makes a nice point; this was not so much a weapon as a luxury good, valuable and desirable because portable and exchangeable. Weapons are, in fact, only rarely mentioned in these transactions.28 Similar messages come from countergifts, that is gifts made in recognition of receipt of a gift, often to secure the primary donation. Offering a countergift was a standard mechanism for getting a guarantee that a gift would be stable and would not revert, for example, to the donor’s family. It was clearly differentiated from price, which was in principle agreed in advance of a sale as part of a contract, whereas the countergift followed a transaction, retrospectively securing it.29 Hence, the language of price and countergift in these texts was different, as was the entire format of sale and gift texts.30 Again, many countergifts consisted of single, relatively modest, objects – a mule, some cloth, a hide, a goat; but those associated with aristocratic gifts were often much more elaborate and included silks, fine horses, silver dishes, fine harness, cloaks, carpets, bed sets, spices. When two nuns gave the monastery of San Salvador land and houses in 990, they

26 Cf. the feather mattresses, linen, gold and silver vessels, ladle etc., used to purchase land and privileges in early Anglo-Saxon England; Campbell, ‘Sale of land’, pp. 232–3. 27 S52, Lii350, S278, C169, S166. 28 Two swords and two shields paid to a lay couple for an estate by a confessus (Sam23 (982)); two swords (following, n. 33); sword and spurs (following, n. 35); gold spurs (PMH DC 99 (968)); but spurs, swords, shields and lances twice in Portugal (PMH DC 114 (974) and PMH DC 147 (985)). 29 See ch. 8 above, ‘Countergift in tenth-century northern Iberia’. Cf. comparable processes in early medieval Italy, which provides many more examples: C. Wickham, ‘Compulsory gift-exchange in Lombard Italy, 650–1150’, in W. Davies and P. Fouracre (eds), The Languages of Gift in the Early Middle Ages (Cambridge: Cambridge University Press, 2010), pp. 183–216. For the temporal distinctions between gift and sale, cf. C. A. Gregory, Gifts and Commodities (London: Academic Press, 1982), pp. 46–8. 30 Of course, in practice the distinction was not so black and white; there were plenty of grey areas. See further ch. 7 above, ‘When gift is sale’.



were given a colt; but when King Sancho Ordoñez restored five or more estates to the aristocrat Odoario in 928, Odoario gave him a special horse and a special pale-coloured (literally ‘yellow’) mule, a fine bed set, three hides and some decorated silver objects; Ramiro, probably a royal, got a fine bed set and a silver dish from a Portuguese lay couple in 926, in acknowledgement of his gift of an estate; and the monastery of Sahagún gave a fine horse and silk to King Ramiro III and his aunt, Elvira, in return for their gift of an estate in 971.31 It is important to note the directions of movement in these examples: objects moved from aristocrat to other aristocrats, from churches to aristocrats, from clerics to clerics and from aristocrats to churches. These were not treasures locked up in a treasure chest; these precious goods circulated, and had practical socio-economic uses.32 It is impossible, of course, to quantify the relative volume of these transactions on the basis of the available data, but it is noticeable that high-value countergifts are more often recorded as moving from churches to aristocrats and that high-value gifts of movables more often move from aristocrats to churches than in other directions. It is also noticeable that more modest gifts of precious movables passed from what look like minor aristocrats to churches, and occasionally to lay persons. Hence, a group of lay people made a series of small grants of local property on entry to the monastery of Piasca in 957, including two silver dishes, three beds with blankets and covers, six sets of table (or altar) linen from one person; two feather mattresses, two swords, two sets of table linen from another; two fine bed sets, thirteen mattresses, two silver dishes from another.33 A woman and her two sons gave the monastery of Lorvão in northern Portugal gold, silver, iron and ‘all kinds of metal’; silken, woollen and linen cloth; buckets, barrels, beds, chairs, tables, food and wine; as well as landed property.34 And in a rare gift of military equipment, a man gave his son an estate in Oncina (near León), with appurtenant vineyards, a meadow and a share in a mill; a yoke of oxen, unnumbered sheep, nine cows, seven horses, together with a riding horse with saddle and bridle, a good sword, spurs and cloth; and barrels, beds, chairs and equipment for the house.35 Although our principal interest in this volume is in the aristocracy, it is also worth noting that other lay people also dealt in precious goods, especially those one might term middle-range people and rich peasants. They received, amongst other things, fleeces, mattresses, Moorish bed covers, cloaks and overcoats from monasteries as the price or part of the price for the sale of land, as well as the

31 Liii534 and 535, Cel234, PMH DC 31, S261. 32 The imaginative reconstruction of the León market in Sánchez-Albornoz, Estampas de la vida, pp. 16–43, is overstated for modern tastes, but remains well worth reading. 33 S153. See J. Montenegro Valentín, Santa María de Piasca. Estudio de un territorio a través de un centro monástico (857–1252) (Valladolid: Universidad de Valladolid, 1993), pp. 50, 52, 55, 66–7. 34 PMH DC 44 (937). 35 Lii488 (944–82).



occasional silver dish and carpet.36 These goods were paid for small-scale properties like plots of arable or patches of vines.37 This was not uncommon from the 920s onward on the meseta (although it was much more characteristic of the meseta than of other regions, and even on the meseta it was more usual to make payments for small properties in food and basics than to make them in precious objects). In these cases of sale for precious goods by middle-range people, the direction of movement of the goods was overwhelmingly from monasteries to the laity (as also tended to be the case for higher value sales). However, even so, there are some cases in which the luxuries passed from lay party to lay party in smallscale transactions – saddles, bridles, linen, once a shield of some kind, cloaks, tunics, blankets, fleeces.38 Hence, the circulation of precious goods reached social levels well below the aristocracy; and, while directions of movement were several, the tenor of surviving records is that the main direction of the flow was from high aristocrats to monasteries to non-aristocrats. C) Valuation It is not uncommon, although not invariable, for these texts to attach a valuation to objects that were handed over in transactions. Often the value was relatively low, expressed in terms of cattle, or sackfuls of grain, or, if in solidi, merely a handful of them (3, 5 or 6).39 (The solidus was a unit of account and was clearly not minted as coin in northern Spain at this time; the unit was probably influenced by knowledge of the key unit of the reformed Carolingian currency, for which silver denarii were coined at the rate of 12 to the solidus, of which there were initially 22 to the pound weight.)40 But the silk which Sahagún gave to King Ramiro was valued at 100 solidi; the bay horse which Ansur paid for the León town house was also valued at 100 solidi; King Vermudo II got a horse worth 300 solidi as countergift from Fernando Núniz in 994; Asur González, from a comital family in Castile, got a countergift of nine lengths of exotic imported cloth and decorated spurs, worth 500 solidi, from Cardeña in 932 for the gift of a church; and the

36 For example, Li58 (923), Li104 (935), Li119 (937), Li156 (942), Li157 (942), Li253 (952), Li254 (952), S80 (941), S88 (943); silver dishes: S55 (934), Cel380 (961) – this is rare in the Celanova collection. 37 Vines: S180 (960); V16 (939), for example. 38 Li60 (923), Li129 (939 – a scutum of relatively low value), Li193 (946), S174 (960), for example. 39 Contrast the vast hoards of silver dirhems found in southern Spain; see E. Manzano, ‘Circulation de biens et richesses entre al-Andalus et l’Occident européen aux VIIIe–Xe siècles’, in L. Feller and A. Rodríguez (eds), Objets sous contrainte. Circulation des objets et valeur des choses au Moyen Age (Paris: La Sorbonne, 2013), pp. 147–80, at p. 150. 40 Despite the well considered view of J. Gautier Dalché, ‘L’histoire monétaire de l’Espagne septentrionale et centrale du IXe au XIIe siècles: quelques réflexions sur divers problèmes’, Anuario de Estudios Medievales, 6 (1969), pp. 43–95, at p. 46, that the Carolingian system did not replace earlier systems in Galicia; he did however note finds of Carolingian denarii sometime after 899 at Santiago de Compostela, ibid., p. 48, n. 28; see ch. 6 above, pp. 109–11, for full discussion.



Portuguese monastery of Guimarães paid precious goods to the value of 1,000 solidi for land in 953.41 This is important. As I indicated earlier, some people in these cultures were certainly familiar with abstract concepts of wealth, but they were not very commonly expressed. Some people, if not most, also had very clear ideas of value, and of the difference between high and low value, and these by contrast were often expressed. This conceptual framework of valuation allowed for a wide range of wealth levels: there was a cattle standard, a modius standard (i.e. ‘sackfuls’) and units of account like argenzos (small pieces of silver) and solidi, as well as some less common standards; hence, a horse worth six oxen, a tunic worth four modii, a coat worth ten argenzos, a carpet worth five solidi of silver.42 Different standards were used contemporaneously for high- and low-value goods, as skins were valued in cattle, sheep in sackfuls and silver dishes in solidi in tenth-century southern Galicia. Using a range of different standards at the same time is by no means peculiar to tenth-century Spain; it happens in many parts of the world.43 They were not absolute standards, of course – any one standard might vary and we should expect them to have done so.44 Any single value could be partly socially conditioned: in other words, value was not simply determined by supply and demand but by other factors, like rank or kinship or relative affluence; sometimes, for example, the rich were expected to pay more: richesse oblige, as Sahlins put it.45 There was, nevertheless, some regional variation in the incidence of the valuation systems. Whereas valuations in solidi occurred in all parts, they were in fact rare in Galicia, common on the meseta and almost invariable in Castile; valuations in modii were very common in Galicia and Portugal, also occurring on the meseta and north of the Cantabrian mountains; valuations in cattle were largely

41 References above, nn. 31 and 27, for the first two cases (S261, S278); Liii565, C23, PMH DC 67. 42 Cel555 (1001), Cel410 (987), Li157 (942), C199 (988). Although the word signified several things at this time, by the thirteenth century the modius became a liquid measure; see J. Clemente Ramos, La economía campesina en la corona de Castilla (1000–1300) (Barcelona: Universidad de Extremadura/Crítica, 2004), pp. 21–3. 43 See C. A. Gregory, Savage Money: The Anthropology and Politics of Commodity Exchange (Amsterdam: Harwood Academic Publishers, 1997), pp. 7–8, 23–6. 44 A point made strongly by Sáez, ‘Nuevos datos’, pp. 865–9, and Gautier Dalché, ‘L’histoire monétaire’, pp. 49–51; in contrast to C. Sánchez-Albornoz, ‘El precio de la vida en el reino asturleonés hace mil años’, in his Viejos y nuevos estudios sobre las instituciones medievales españoles, 3 vols (Madrid: Espasa Calpe, 1976–80), vol. 2, pp. 811–52, and ‘Moneda de cambio y moneda de cuenta en el reino asturleonés’, ibid., pp. 855–83, who proceeded from an assumption that economically consistent principles informed Spanish valuation systems in the tenth century, which could be related to each other consistently; he looked for demand/supply explanations alone for price differentiation – ‘la ineludible ley de la oferta y la demanda’, ibid., p. 819. 45 M. Sahlins, Stone Age Economics (London: Taylor and Francis, 1974), pp. 211–15, 268, 277–9. Or the rich chose to pay more: for an early medieval example, see J. Hannig, ‘Ars donandi. Zur Ökonomie des Schenkens im früheren Mittelalter’, in R. van Dülmen (ed.), Armut, Liebe, Ehre. Studien zur historischen Kulturforschung (Frankfurt am Main: Fischer, 1988), pp. 11–37, at pp. 32–6, for prestige price against market price in Italy and France c. 900.



confined to Galicia.46 The very strong regional differences in valuation practice are emphasized by the facts that most sales between lay parties in the Celanova collection (Galicia) are priced in food and basic cloth, measured in modii and quarters; and that even high-value transactions in that collection tend to be paid for in similar goods, like the estate sold for fifty lengths of cloth and forty goats.47 ‘Valuation is essentially a comparative process by which two unlike entities are compared and judged.’48 Value systems necessarily relate to exchange, and in most cultures they relate to commodity exchange, that is exchange as a system of purchase and sale, exchange in which the primary relationship is between the objects transacted rather than between the transactors. This is exchange with more than a touch of the commercial.49 The very existence of value systems means that exchange must have been normal – an expected part of regular socio-economic relationships – even if some goods arrived as the consequence of raiding. While we cannot possibly quantify volume, the widespread occurrence of value systems means that exchange must have been socially and economically significant in the tenth century. There were hundreds of sales for value, both high and low value, which must in itself suggest active participation in the value systems at different social levels. The range of social levels involved in these transactions with attached valuations was indeed wide – high aristocrats certainly, but also middle-range people (some urban-dwellers, like the woman and the priest), some rich peasants and some not-so-rich peasants.50 A peasant couple sold small plots for goods valued in argenzos in 923; a peasant and his mother sold a water course for goods valued in solidi and modii; two brothers sold small plots for goods valued in solidi; some cousins sold more substantial plots for goods valued in modii in 955; a woman and her children sold plots for food valued in modii, as did two brothers and their wives; and a group of siblings sold their share of land for food and basics valued (unusually) in Galician solidi and in modii in 961.51 There are many more examples. These valuation systems were not just a book-keeper’s concept; they functioned across many localities, in richer and poorer circles. There are even cases in which the notion of ‘just price’ was invoked in peasant transactions – for example, in 960 a just price was to be assessed by local elders in northern Portugal if a previously 46 Gautier Dalché, ‘L’histoire monétaire’, pp. 69–71, 90, notes a few cattle valuations in eleventhand twelfth-century Asturias and one in Otero de las Dueñas; PMH DC 124 (978) also appears to have a cattle valuation. Some of the earliest Galician valuations may relate to a gold rather than a silver standard; see further ch. 6 above, p. 104. 47 Cel161 (965), Cel164 (967); Cel535 (953). For regional difference, cf. Gautier Dalché, ‘L’histoire monétaire’, pp. 47–8, 56–8. 48 Gregory, Savage Money, p. 13. 49 Cf. Gregory, Gifts and Commodities, pp. 19, 42. 50 Although some peasant sales were certainly not commercial – rather, a means of getting food and other necessities; see Davies, Acts of Giving, pp. 158–60. 51 Li60, Li124 (937), Li254 (952), S148, Cel373 (956), Cel377 (996), Cel384. Galician solidi feature in Portuguese texts too.



exchanged property had to be sold.52 Accordingly, valuations feature in records made by local priests, as well as those made by great monasteries or by royal or episcopal households; so, the priests Ababdella, Zenzitus, Iohannes and Arrian all made valuation records for small-scale sales between lay persons (and occasionally local clergy), as did others whose status is not indicated.53 When people were recording transactions they were accustomed to think in terms of value.

Conclusion If we go back to our organizers’ initial questions about the contradictions inherent in the Christian approach to wealth, about contemporary criteria for richness in the early middle ages, about the acquisition and maintenance of wealth and about the economic awareness of élites, there are some clear answers from this corpus. There is certainly explicit moralizing in this material: there are plenty of exhortations to give away personal riches in order to achieve the superior treasure of heaven. Despite that, many individuals did in practice accumulate both landed and movable goods, as did religious bodies, and there is little to suggest any awareness of the contradictions in this. As for criteria, the texts are inclined to associate precious movables with being rich; and I suspect that most people who thought about ‘wealth’ in this area at this time thought of it in terms of movables rather than in terms of land or total resources.54 In effect, the most consistent, and the most common, expressions of wealth lie in the statements of relatively high value attached to movables. As for acquiring wealth and becoming rich, there is plenty to suggest that high aristocrats sought precious objects, as kings of earlier centuries had done – doubtless as a convenient vehicle of social display, reinforcing their social and political status. Indeed, it looks as if some aristocrats were so keen to acquire precious objects that they used land to do so. This is the best explanation for the transactions in which aristocrats exchanged land for very high value countergifts (apparently close to the value of the property initially conveyed) or for a price composed of high-value, often exotic, objects.55 These look as if they were deals between aristocrats (lay and ecclesiastical) to move goods about and to acquire rare and precious things. This does not in itself have to imply economic awareness, nor indeed competence. Perhaps the contrary. The texts certainly do not talk in terms of profit (apart from the occasional wicked lucrum in a sub-biblical quote56), nor in terms of any desire to increase surplus, much less do they make

52 PMH DC 79; cf. Cel182 (1000). Although this doubtless reflected a concern about making undue profit, rather than about establishing a market price. 53 Li162 (943 – not so small), Li200 (948), Li237 (951), Li238 (951); others: Li158 (942), Li173 (943), Li234 (951); all but two of these examples are recorded on single sheets. 54 Cf. Sahlins, Stone Age Economics, p. 218; Campbell, ‘Sale of land’, pp. 240–1. 55 See further ch. 8 above. 56 See above, n. 6.



explicit reference to credit.57 But we can nevertheless see from analysis of the texts both the presence of families rich in landed estates, and, especially in the second half of the tenth century, the beginnings of the establishment of the great monastic estates:58 their purchasing strategies, and their deliberate consolidation of properties, must mean that some of their stewards were beginning to think in terms of surplus accumulation and control. Indeed, it is clear by the 990s that a handful of lay families were also accumulating properties – through purchase, gift and the proceeds of fines in court cases – in such numbers that it is difficult to interpret except as intentional.59 With the benefit of hindsight it looks as if they too were thinking about accumulating and developing landed resources, even if conceptualization of wealth itself was still grounded in movables. How does this fit into the wider issues of economic development? It is customary to focus on the development of great estates in the tenth century, alongside isolated references to merchants and to (weekly) markets, the import of manufactured goods from the South by magnates and low levels of urbanization. Spanish writers have argued that commercial circulation was essentially to feed the needs

57 J. A. García de Cortázar, La sociedad rural en la España medieval (Madrid: Siglo XXI, 1988), pp. 34, 43, interpreted some records of sale as being essentially about the provision of credit. I am not persuaded that these were credit arrangements, in the absence of evidence that the properties in question were subsequently recovered and that any of the ‘prices’ paid (which are largely food) were returned to the purchasers; most charters also give the purchasers rights of free disposition of their acquisitions, which suggests that the transactions were considered final. The paucity of references to objects or to property put down as security, i.e. as collateral for a loan (usually called pledges or gages), is very striking in northern Spain by contrast with references in other parts of western Europe at this time; see F. Bougard, ‘Le crédit dans l’Occident du haut Moyen Age: documentation et pratique’, in Devroey, Feller, Le Jan (eds), Les élites et la richesse, pp. 439–78. 58 The classic example is J. M. Mínguez Fernández, El dominio del monasterio de Sahagún en el siglo X (Salamanca: Universidad de Salamanca, 1980); but the significant build-up of estates by the bishopric of León, by Celanova and by Cardeña is also notable in the tenth century; for León, see Estepa, Estructura social de León, pp. 197–203, and G. del Ser Quijano, ‘La renta feudal en la alta edad media. El ejemplo del Cabildo catedralicio de León en el período asturleonés’, Studia Historica. Historia Medieval, 4 (1986), pp. 59–75; for Celanova, although it covers a longer period, there is much in J.-M. Andrade Cernadas, El monacato benedictino y la sociedad de la Galicia medieval (siglos x al xiii) (A Coruña: Do Castro, 1997); for Cardeña, see S. Moreta Velayos, El monasterio de San Pedro de Cardeña. Historia de un dominio monástico castellano (902–1338) (Salamanca: Universidad de Salamanca, 1971). See García de Cortázar, Sociedad rural, pp. 34–6, 48, for an overview of the progressive submission of peasants to aristocracy and consequent relative increase in aristocratic wealth; C. de Ayala Martínez, ‘Relaciones de propiedad y estructura económica del reino de León: los marcos de producción agraria y el trabajo campesino (850–1230)’, in El reino de León en la alta edad media, 6, Estudios (León: Centro de Estudios e Investigación San Isidoro, 1994), pp. 133–408, gives an overview of proprietorship and tenant relationships in the provinces of León and Zamora, but identifies little economic development in the period 850–1050. 59 Munio Fernández and Flaino Muñoz, for example: see Davies, Acts of Giving, pp. 146–8. There are earlier examples too, although those we know most about tend to have alienated their property by endowing monasteries; Ermenegildo and Paterna in northern Galicia, for example: M. del C. Pallares Méndez, El monasterio de Sobrado: un ejemplo de protagonismo monástico en la Galicia medieval (A Coruña: Diputación Provincial de La Corun̂ a, 1979), pp. 29–43.



of great proprietors; that, while there are traces of development in the later tenth century, the volume of goods circulating at that time was still small; and that the main development came in the eleventh century.60 Clearly there was plenty of change in the eleventh century, and one could not possibly disagree that development intensified then, particularly given the advent of minting coin, the clear development of pottery manufacture for widespread distribution and the stimulus of the tributes received from al-Andalus.61 But it is worth suggesting some adjustment of the balance for the tenth century. There is, after all, plenty of evidence of extending cultivation throughout the century, which must imply higher levels of production – witness many disputes about peasants ‘encroaching’ on monastic land and many charters recording grants of land ad populandum;62 there was competition over mills and water rights, notably in the later tenth century;63 there was exploitation of salt pans, by Cardeña especially, which can only have been for distribution;64 there were some towns, even if few, and some shops, as well as merchants, markets and artisans.65

60 García de Cortázar, Sociedad rural, pp. 37–46; Estepa, Estructura social de León, pp. 373–424; cf. Mínguez Fernández, El dominio del monasterio de Sahagún, p. 244, for ‘espectaculares progresos’ in the eleventh century. See A. Isla Frez, La alta edad media. Siglos viii–xi (Madrid: Síntesis, 2002), pp. 219–37, for a useful overview. 61 For pottery, see J. A. Quirós Castillo and B. Bengoetxea Rementeria, Arqueología (III) (Arqueología Postclásica) (Madrid: Universidad Nacional de Educación a Distancia, 2006), p. 360. See further following, nn. 66–70. 62 Disputes: Sam247 (909) implicitly, Li184 (944), Cel446 (959), for example; pro populare/ad populandum: Li79 (928), Cel7 (950), S256 (970), for example – this was not so much about populating the land for the first time but about dominating it, i.e. controlling surplus production; cf. J. A. García de Cortázar, La Historia rural medieval. Un esquema de análisis estructural de sus contenidos a través del ejemplo hispanocristiano (Santander: Universidad de Santander, 1982), pp. 141–59, and, for ‘peasant dynamism’ in the ninth and tenth centuries, J. J. Larrea, La Navarre du IVe au XIIe siècle (Paris/Brussels: De Boeck, 1998), pp. 165–9, 196–7, 590. Cf. G. G. Astill, ‘Exchange, coinage and the economy of early medieval England’, in J. Escalona Monge and A. Reynolds (eds), Scale and Scale Change: Western Europe in the First Millennium (Turnhout: Brepols, 2011), pp. 253–72, for intensification of production prior to major economic shift. 63 J. A. García de Cortázar, ‘El equipamiento molinar en la Rioja Alta en los siglos x al xiii’, in Homenaje a Fray Justo Pérez de Urbel, OSB (Studia Silense, 3) (Silos: Abadía de Silos, 1976), pp. 387–405, at pp. 392, 401, notes many more references to mills in the eleventh century but warns that many must go unrecorded; T. F. Glick, Islamic and Christian Spain in the Early Middle Ages (Princeton: Princeton University, 1979), pp. 92–3, 96–8. 64 Note the many purchases of salt rights by the monastery in the later 970s/980s, a distinctively new interest: C170, C173, C175, C176, C179, C186, C194, C195, C197 (a gift); cf. Mínguez Fernández, El dominio del monasterio de Sahagún, pp. 193–4, on the early development (early tenth century) of Sahagún’s salt interests; Larrea, La Navarre, p. 295, for Albelda’s; there is a helpful survey of increasing exploitation of salt rights in E. Peña Bocos, La atribución social del espacio en la Castilla altomedieval. Una nueva aproximación al feudalismo peninsular (Santander: Universidad de Cantabria, 1995), pp. 72–85. 65 Estepa, Estructura social de León, pp. 373–5, and above n. 19; Quirós Castillo and Bengoetxea Rementeria, Arqueología (III), p. 361; Isla Frez, Alta edad media, pp. 223–30 – Isla notes how quickly the towns recovered from attacks.



Thanks to a valuable recent study of pottery in the southern Basque country (Álava), we can see that pottery production and distribution in the ninth and tenth centuries was more complex than previously thought and that new wares and new forms became prevalent in some zones.66 There was domestic production in the household, as one might expect, but the proportion of such wares gradually petered out across the ninth, tenth and eleventh centuries;67 pottery also seems to have been sold by itinerant potters, across a wide area of 100 km or more;68 and, from the ninth century, some pottery was being made in local workshops for limited distribution, as for example in a zone of 20–30 km around Vitoria and a slightly smaller zone in the Rioja Alta.69 There was also a tiny proportion of imports.70 While it remains demonstrable that the bases of production changed significantly in the eleventh century, it is also clear that there was some change and development in the preceding three centuries; that there were regional zones of market activity before the eleventh-century changes; and that there were supra-regional networks supplying non-local wares at the same time.71 None of this material could be described as precious and it indicates some element of market activity at social levels below that of aristocrats. Moreover, as indicated above, precious goods circulated more widely across the North than from magnate to magnate; and there was increasing use of silver, not just increasing reference to it, for transactions in the second half of the tenth century, with use of the small pieces of silver known as argenzos on the meseta in the hinterland of León and Sahagún.72 Despite the customary comments, there is plenty of evidence of local, non-élite exchange in the North-west (though rather less in Castile), not least in the existence of multiple valuation systems; and on the meseta there is plenty of evidence – from the 920s onwards – of non-élite participation in the exchange of imported goods. If we had the regional pottery studies for the meseta, this pattern would surely be reinforced. There are many indicators both of an increase in production and of an increase in the volume of exchange

66 J. L. Solaun Bustinza, La cerámica medieval en el País Vasco (siglos viii–xiii) (Vitoria: Gobierno Vasco, 2005), pp. 380, 310, 316. 67 Ibid., pp. 354–6, 393–9. 68 Ibid., pp. 371–4. 69 Ibid., pp. 357–9, 367–71. 70 Ibid., pp. 318, 320–2, 379. Estimates of the relative production methods for the central nucleus round Vitoria in the ninth and tenth centuries are: 8% domestic, 8% itinerant, 78% local workshops, 3% foreign; and in the Rioja Alta in eighth to tenth centuries: 10% domestic, 23% itinerant, 57% local workshops, 7% foreign, with much higher proportions of domestic production in western Álava and in Vizcaya to the north; ibid., pp. 380–3. 71 Excavations in the last couple of years, e.g. at Aistra (east of Vitoria), have produced new material, whose analysis may lead to some revision of Solaun’s dating framework (personal comment, Andrew Reynolds); while any refinement of dating will be welcome, Solaun’s suggested complexity of pottery production before the eleventh-century changes will still provide an important insight. 72 See ch. 6 above, pp. 106–9, 116.



activity in northern Spain, as in many other parts of western Europe, in the tenth century. This was not evenly spread, of course: indications of increasing intensity of exchange are much more pronounced on the meseta. Focussing on precious goods demonstrates that they circulated, in several directions, and that, although they were unquestionably dominated by the élite, their possession was not confined to the élite. Participation in commodity exchange seems to have been wider than is customarily allowed, especially on the meseta.73 What makes northern Spain look so different from some other parts of western Europe is the absence of locally minted coin, as also of fiscal pressures to convert surplus. But if we were able to quantify production and exchange, maybe the meseta, at least, would not look quite so different. Further, if one were looking for aristocratic demand, as the stimulus for economic step-change, there is plenty of evidence of it across northern Spain.74 Economic behaviour was unquestionably changing by the year 1000; the change did not yet inform the conceptual apparatus and was not reflected in recorded expressions of wealth. A few people were nevertheless ‘economically competent’. And the widespread awareness of high versus low value suggests that many were on the verge of ‘competence’.75

73 Cf. Isla Frez, Alta edad media, p. 226: there must have been trade in local produce in the second half of the tenth century. 74 Cf. C. Wickham, ‘Rethinking the structure of the early medieval economy’, in J. R. Davis and M. McCormick (eds), The Long Morning of Medieval Europe: New Directions in Early Medieval Studies (Aldershot: Ashgate, 2008), pp. 19–31, at pp. 24, 30–1; G. Astill, ‘Exchange, coinage and the economy’, pp. 269–72. 75 I am extremely grateful to Grenville Astill, Julio Escalona, Juan José Larrea and Chris Wickham for their helpful comments on a draft of this paper, and for some references; and to the History Faculty of the University of Oxford for a grant to support travel to the conference.


10 WAT E R M I L L S A N D C AT T L E S TA N D A R D S Probing the economic comparison between Ireland and Spain in the early middle ages Hector Munro Chadwick was an early hero of mine and I am particularly honoured by the invitation to give this lecture.1 I am going to spend some time on northern Iberia but I shall range across cultures and across disciplines (and focus on some words), which I hope will meet the spirit of his legacy. Let me begin with my title and some Irish and Iberian comparanda. Firstly, water mills: one day, on the meseta, the central plateau of Spain, not far from the city of León, a dispute arose between the people of San Juan en Vega (led by a man called Gondemaro) and the abbot of the nearby monastery of Valdevimbre. The dispute was about the water course that fed the abbot’s mill. Gondemaro said that he and his neighbours had been drawing water from that stream for generations (avis and trysavis) but the abbot protested against this. The dispute over the water went to the king’s court, which was close by in León, and the king sent four men he trusted (  fideles) to investigate. They went to the stream and measured the water levels, which the abbot had claimed were not high enough to power his mill. Using some stakes to check the levels, they found the water running in equal measures for both parties, and so the king proclaimed that Gondemaro and his neighbours could go on drawing water from that stream. The same thing happened the next year; this time three prominent royal judges went to investigate; again the water levels were high enough for all; the king again confirmed in favour of Gondemaro and his neighbours and on 25th June 938 ordered the monastery not to stop them.2 This is but one of many such disputes. An extraordinary number of early medieval water mills has been discovered in Ireland; they include the early seventh-century mills at Little Island, Co. Cork, in the south west; at Strangford Lough, by Nendrum, Co. Down, in the north; and

1 First published as a separate lecture, H. M. Chadwick Memorial Lectures 21 (Cambridge: Department of Anglo-Saxon, Norse and Celtic, 2012). I am very grateful to Máire Ní Mhaonaigh and her ASNC colleagues for the opportunity to deliver the lecture and to the enthusiastic and responsive audience; as also to Máire, Jinty Nelson and Thomas Charles-Edwards for comments on the paper. 2 Li128.



the substantial wheel pit at Killoteran, near the early Viking site of Woodstown, not far from Waterford.3 Physical evidence apart, there is written evidence too: the Spanish story of Gondemaro and his neighbours seems to be the kind of situation for which the Irish eighth-century law tract Coibnes Uisci Thairidne (Kinship of Conducted Water) was providing; quite apart from fixing fees for the cutting of new water channels to power mills, it provided that water should be allowed to flow where two generations had been used to drawing it: Ar ata[a]t .iii. cáini nad chumgat comarbai do chumscuchud dia n-árdamat a n-athair [┐a senathair] fria n-aimsir: cáin cach uisci t[h]airidne . . . Neoch ma ‘d-rodmatar (neoch) dle(a)ga[i]r a mb[u]ith samlaid co bráth, im deólad fa lóg dlegar . . . ‘For there are three rules which heirs are not capable of altering if their father and grandfather have acknowledged [them] throughout their lifetime: the rule of every water course . . . If they have been acknowledged, they are to remain for ever, whether they be gratis or whether a fee is due . . .’4 It provides a nice mirror, in its emphasis on past generations, of the reported Gondemaro case. As for cattle standards: on 18 October 991 a man called Munio Ovequiz sold part of a farm in Orga, in southern Galicia, to the monastery of Celanova; he sold it for three horses, three oxen, some silver and a skin. One of the horses he received was worth fifteen oxen, the second five oxen and the third seven oxen.5 Many calculations in cattle feature in the early Irish law tracts, of the eighth century, and in subsequent commentaries, although the Irish cattle standards have more refinements than the Iberian – they distinguish between the milch-cow, the two-year-old heifer, the three-year-old dry heifer and so on. The Old Irish tract 3 C. Rynne, ‘The craft of the millwright in early medieval Munster’, in M. A. Monk and J. Sheehan (eds), Early Medieval Munster: Archaeology, History and Society (Cork: Cork University Press,1998), pp. 87–101; T. McErlean and N. Crothers, Harnessing the Tides: The Early Medieval Tide Mills at Nendrum Monastery, Strangford Lough, Northern Ireland Archaeological Monographs, 7 (Norwich: Stationery Office, 2007); D. Murphy and S. Rathbone, ‘Excavation of an early medieval vertical watermill at Killoteran, County Waterford’, in J. O’Sullivan and M. Stanley (eds), Settlement, Industry and Ritual: Proceedings of a Public Seminar on Archaeological Discoveries on National Road Schemes, September 2005 (Dublin: National Roads Authority, 2006), pp. 19–28. 4 D. Binchy, ‘Irish law-tracts re-edited: I. Coibnes Uisci Thairidne’, Ériu, 17 (1955), pp. 52–85, at pp. 68, 69. See R. Sharpe, ‘Dispute settlement in medieval Ireland: A preliminary inquiry’, in W. Davies and P. Fouracre (eds), The Settlement of Disputes in Early Medieval Europe (Cambridge: Cambridge University Press, 1986), pp. 169–89, at p. 180. 5 ‘placuit mihi, ego Munio Ovekiz cum uxore mea Guldregoto facerem tibi cartulam venditionis, sicuti et facimus hanc tibi fratri Cresconio de ipsa villa [Orga] . . . Abeas firmiter ipsas villas de meo dato pro que accepimus de te pretium, id est, caballos tres, uno de boves XV, alio de boves V. Ille tertius de boves VII, boves . . . ariento . . . pelle’, Cel248.



on land values, Cis lir fodla tire? (How many kinds of land are there?), proposes, for example: Etham rem[i]-bi ethamnaib, cumal .iiii. bo fichit mblicht a log; etham taulcach, cumal fichit mbo mlicht . . . Antren(n)d, cumal da bo ndec seisc n-airi. Andomain, cumal ocht mbo seisci n-airi . . . Mad sligi ad-cuma o thir co flaith no mainistir, do-formaig .iii. bu. ‘First-class cultivable land, its value is a cumal of twenty-four milch cows (i.e. a cumal [measure] of it is worth twenty-four); upland cultivable, a cumal of twenty milch cows . . . Very rough [land], a cumal of twelve dry cows for it. Shallow [land], a cumal of eight dry cows . . . If it be a way that extends from [the] land to [the dwelling of] a lord or to a monastery, it adds three cows’ (as do a sea, a stream or a mountain).6 In both regions such references are to standards of value, using cattle to set a notional standard against which payments could be calibrated.7 This lecture is about asking questions. Are these similarities in the use of cattle standards, and in collisions over access to water to power mills, real or illusory? Were these economies actually functioning in similar ways or are these superficial similarities? I could cite other kinds of similarity: the use of graded penalties for different qualities of wound; the use of paying and enforcing sureties to guarantee the terms of a settlement and so on; although, of course, there are plenty of differences. But if things that look similar really were similar, what does this signify? What, if anything, does it tell us about economic relationships and the operation of production, distribution and exchange? It is striking that, unlike much of western Europe, neither region minted coin until the eleventh century and neither seems to have used much coin from external sources before the tenth century, although both were certainly using metal by weight to make payments at that time. In each case the difficulty of making a rational assessment is complicated by source problems. There is no Domesday Book and there are no polyptyques; there is nothing in the way of a ‘comprehensive’ survey, however imperfect. That means that we cannot quantify or measure in any robust fashion. The available evidence is also extremely uneven and often not strictly comparable: there are several thousand charters from late ninth- and tenth-century northern Iberia but only a handful of charter fragments (of earlier date) from Ireland.8 Charters have 6 G. Mac Niocaill, ‘Tír Cumaile’, Ériu, 22 (1971), pp. 81–6, at pp. 82, 85. 7 See further following for the cumal and other standards of value. 8 For northern Iberia, the most important collections of texts are noted in the abbreviations list above pp. xvii–xix. For Ireland, the most important collection of fragments are those of the ‘Additamenta’ to Tírechán’s collection in The Patrician Texts in the Book of Armagh, ed. and trans. L. Bieler, Scriptores Latini Hiberniae, 10 (Dublin: Dublin Institute for Advanced Studies, 1979), pp. 166–79; but note the further fragments mentioned in W. Davies, ‘The Latin charter-tradition in western Britain, Brittany and Ireland in the early mediaeval period’, in D. Whitelock, R. McKitterick, D. Dumville (eds),



especial significance because they are often localizable, taking us, literally, to ground level. Hardly any texts from post-Visigothic Iberia survive before the late ninth century; but Ireland has a very substantial corpus of eighth-century written material – annals, computation, exegesis, genealogy, hagiography, laws, penitentials, poems.9 As for archaeology: there has been a remarkable quantity of (recent) excavation of early medieval sites in Ireland, of high quality: 2,214 excavation summaries (all periods) for the single year 2006, for example, and 2,208 early medieval sites now (March 2011) on the EMAP (Early Medieval Archaeology Project) database.10 There is a growing body of early medieval archaeology from Iberia, which is important, but the volume is small by comparison – tens of cases rather than thousands. So, there are considerable problems of evidence and of comparing evidence. Notwithstanding these problems, in what follows I shall consider five themes: the persistence of tradition, transaction language, cereals, valuation and silver. Before that, a word on basic geography: Ireland is smaller than northern Iberia, an island with many waterways, with a wetter, very moderate climate. Northern Iberia has a much more varied terrain and more extreme climates; it has high mountain ranges, a central plateau which is hot and very dry in summer (the meseta), although the west – Galicia – is wetter and greener. But both are in some sense ‘Atlantic’ and both lie west of Francia and beyond the direct sphere of Frankish influence. For northern Iberia I include Galicia (including northern Portugal), the far north, the Duero basin, Castile, Navarre and Aragón, but not Catalonia; the land to the south was Muslim al-Andalus.

The persistence of tradition I borrow the phrase from H. M. Chadwick, who, in the preface to Studies on AngloSaxon Institutions, wrote: ‘the persistence of tradition, popular as well as ecclesiastical, is a force which seems to me to have been underrated by writers on ancient English institutions’.11 The insight can well apply elsewhere. The force of tradition is familiar in Ireland, where the copying of and commenting on earlier law texts,

Ireland in Early Mediaeval Europe (Cambridge: Cambridge University Press, 1982), pp. 258–80, at pp. 261, 269, 273; reprinted as Essay XI in W. Davies, Welsh History in the Early Middle Ages: Texts and Societies, Variorum Collected Studies (Farnham: Ashgate, 2009). 9 For a good overview, see D. Ó Cróinín (ed.), A New History of Ireland, I, Prehistoric and Early Ireland (Oxford: Clarendon Press, 2005), especially the essays at pp. 331–548, by T. Charles-Edwards, D. Ó Cróinín, P. Russell, J. Carney, W. O’Sullivan. 10 I. Bennett (ed.), Excavations 2006: Summary Accounts of Excavations in Ireland (Dublin: Wordwell, 2009); The rate of excavation declined after 2006, and very sharply from 2008. 11 H. M. Chadwick, Studies on Anglo-Saxon Institutions (Cambridge: University Press, 1905; rep. New York: Russell & Russell, 1963), p. viii.



for centuries, is well known.12 Perhaps not so well known is the fact that Visigothic law, that is law deriving from the sixth and seventh centuries, continued to be cited, summarized and evoked in court cases of the tenth and eleventh centuries in northern Spain. Charters which record the settlement of disputes can make vague comments such as ‘as lex gotica provides’, but can also specify book, title and chapter of the law (often known as Liber Iudicum) and quote them. In a record of a property dispute of 952 the law on freedom of alienation of property was quoted from book 4, title 2, chapter 19, declaring that every free man and woman who left no children should have the freedom to do whatever he or she wished with his or her property; this was followed by a quotation from book 5, title 2, chapter 6 on post-mortem donation.13 These writers of the tenth and eleventh centuries could abbreviate or massage the texts, as well as quote in full; although it was associated with a distant Gothic past, it was still deemed relevant – its antiquity gave it both authority and relevance. One important manuscript of Visigothic law was written by a practising Catalan judge, Judge Bonushomo, in 1011. It is heavily glossed, the glosses occupying 164 pages of the modern edition, a fact which clearly emphasizes the continuing use and relevance of this ancient tradition but also indicates some comparability with approaches to written Irish law across a similar period.14 There are other kinds of ancient notion with continuing relevance. Take the quinta: this is an extremely common word in Iberian charters and refers, in the tenth century, to ‘our portion’, ‘our share’ of property rather than a strict fifth: placuit mihi atque convenit ut facerem vobis scriptura venditionis sicut et facio de quinta quam abeo de marido meum Illallo in ipsa sua villa ubi dicent Carioga. ‘I have agreed to give you a record of sale and I do so with regard to the quinta I have from my husband Illallo in his farm called Quiroga’.15 Quinta is a word which also derives from Visigothic law and it originally referred to that portion of family property that could be alienated (in other words, over which there were free powers of disposition). It clearly did not retain its strict legal meaning in subsequent centuries but it was often used in the second half of the tenth century to refer to portions that were given away or sold – an appropriate

12 See F. Kelly, A Guide to Early Irish Law, Early Irish Law Series, 3 (Dublin: Dublin Institute for Advanced Studies, 1988), pp. 225–31. 13 ‘proclamaverunt se uterque ambo ad Librum, et qualem ex his Liber mandasset stabilire, abuisset roborem et stabilitatem firmissimam, sicut et de presente mox impletum fuit in libro iiii, titulo secundo, kapitula xviiii: “Omnis ingenuus vir atque femina, qui filios . . . non relinquerit, faciendi de rebus suis quod voluerit indubitanter licenciam abebit”’, Li256 (1/8/952). 14 Liber iudicum popularis. Ordenat pel jutge Bonsom de Barcelona, ed. J. Alturo, J. Bellès, J. M. Font i Rius, Y. García, A. M. Mundó (Barcelona: Generalitat de Catalunya, Departament de Justícia i Interior, 2003); glosses at pp. 256–70 and 635–785. 15 Cel163 (932).



word because of the connotation of free disposability (rather like ‘bookland’ in England).16 Interestingly, one of the texts attached to Tírechán’s Irish collection of material about Saint Patrick, the so-called ‘Additamenta’, of a date no later than the mid-eighth century, uses the same term: ‘et postquam babtitzavit obtulerunt filius Cairthin et Caichan quintam partem Caichain Deo et Patricio’.17 This is one of the fragmentary Irish charter texts, written in recognizable charter language, and records the gift of a quinta of Caíchán’s property to Patrick, and God, free from obligations to anyone else. It reflects the occurrence of the word quinta in the Hibernensis, the eighth-century collection of Irish canonical material, where it occurs in a context of making gifts to God, quoting Old Testament Leviticus.18 Caíchán’s property was clearly land, since its boundaries were quoted. In the longer term the ‘fifth’ (cóiced) survived in Irish place-names, just as quinta, quintana and quintanilla survive in the toponymic landscape of Spain.19 Boundaries are another case of the survival of ancient tradition. In Galicia, especially, boundaries could be expressed as full perambulations, like those attached to English and Welsh charters (e.g. ‘it runs from the stream to the tree and up the hill to the crossroads’); they are sometimes called terminos antiquos, the ancient boundaries, and they are sometimes marked by stones, petras, but also petras scriptas and in one text of 988 at least four sets of petras sculptas et scriptas (sculptured stones with writing).20 This must mean inscribed stones. They do not appear to survive, but they cannot help but remind us of the very large corpus of early medieval inscribed stones from many parts of Ireland that do survive, as also of the citation of stones as legitimizing boundary markers in a number of different Irish texts.21

16 For further discussion see W. Davies, Acts of Giving: Individual, Community, and Church in Tenth-Century Christian Spain (Oxford: Oxford University Press, 2007), pp. 76–8, and references there to the large Spanish bibliography; of the latter, the following are classic treatments: L. G. de Valdeavellano, ‘La cuota de libre disposición en el derecho hereditario de León y Castilla en la alta edad media’, Anuario de Historia del Derecho Español, 9 (1932), pp. 129–76; J. Orlandis, ‘“Traditio corporis et animae”. Laicos y monasterios en la alta edad media española’, in J. Orlandis, Estudios sobre instituciones monásticas medievales (Pamplona: Universidad de Navarra, 1971), pp. 217–378, at pp. 277–83. 17 ‘Additamenta’ 8, in Patrician Texts, ed. Bieler, p. 172. ‘And after he had baptized them, Macc Cairthin and Caíchán offered to God and Patrick the fifth part of Caíchán(’s estate)’. 18 The precise context relates to the man who has given his house to God and wishes to redeem it; he may do so but must leave a fifth of its value with the church. R. Flechner, The Hibernensis: Book 1: A Study and Edition (Washington, DC: The Catholic University of America Press, 2019), 17.2. I am extremely grateful to Roy Flechner for giving me a copy of his edition of Hibernensis prior to publication. 19 See E. Hogan, Onomasticon Goedelicum locorum et tribuum Hiberniae et Scotiae (Dublin: Hodges, Figgis & Co., 1910), pp. 279–80; R. Ruiz Agullo, R. Ruiz Serramalera, A. Ruiz Serramalera, Pueblos de España: nomenclator comercial (Madrid: R. Ruiz, 15th edn, 1997), pp. 576–8. 20 LaC102; cf. the interpolated no. 76 (968). 21 R. A. S. Macalister, Corpus Inscriptionum Insularum Celticarum, 2 vols (Dublin: Stationery Office, 1945–9); cf. Kelly, Guide to Early Irish Law, p. 186; F. Kelly, Early Irish Farming, Early Irish Law Series, 4 (Dublin: Dublin Institute for Advanced Studies, 1997), p. 409.



Transaction language Iberian texts use the concepts of gift and sale (donare and tradere, especially; vendere; and cartulae donationis/venditionis) and to a lesser extent exchange (comutare, concambiare). Here ‘gift’ refers to a transfer without the necessity to receive something in return; and ‘sale’ to a transfer for price, agreed in advance, its receipt necessary to validate the transfer. Price, pretium, is virtually always associated with sale, and cited. There are a few ‘exchanges’, which in practice meant the particular exchange of swapping land for another piece of land. Nearly half of all charters from before 1000 record sales and just over half record gifts. Now, it is true that some gifts provoked a countergift in return but most of these had a very specific function: to secure the primary gift. When a laywoman, Leodegundia, gave a lay couple a parcel of land in northern Portugal, in 990, the couple gave her an ox and a pelt to confirm the charter of gift – ad cartam confirmando.22 These countergifts are rare – they occur in 6% of gift records; they appear to have been made when beneficiaries sought extra security because of particular, unstated, concerns, presumably concerns that the gift might revert to the donor or his/her heirs. They are certainly not evidence of some kind of gift-exchange system. For Ireland we have very few texts recording actual transactions and we are largely reliant on law texts, for the most part recording principles. These are extremely important, but different. However, there is a handful of recorded transactions in the ‘Additamenta’ to Tírechán’s collection. These transactions were mostly gifts but there were a few sales. The language of record is both Latin and Irish. The Latin language of giving uses words like offerre (‘offer’) and immolare (literally, ‘make a sacrifice of’).23 The language is again reflected in the Latin of Hibernensis: oblationes and immolare occur in respect of gifts, as does testamentum (as well as vendere and pretium of land sold, citing an Irish synod).24 These Latin giving words have a different flavour from the Iberian Latin words: offerre can sometimes be found in Iberian texts but immolare not at all; the word is very particular to giving to the church in the north-west of Europe (it also occurs in texts from Wales and Scotland).25 The vernacular language of transactions in Ireland has both gift and sale words.26 Giving words include immransat (they entrusted/assigned/bequeathed), adopart (he offered) and tubart (he gave/brought); edoct, a idacht, usually translated ‘testament’, although its literal meaning is closer to ‘declaration’, a verbal commitment. Selling words include dirróggel (purchased), ríthae (was sold) and forlóg (additional price, in the context of a land purchase).27 As for the law tracts,

22 PMH DC 159. 23 ‘Additamenta’ 3.2, 5.4, 8.1, 10.1, 10.2; 1.6, 5.1, 5.2, 7, in Patrician Texts, ed. Bieler, pp. 168, 170, 172. 24 Flechner, The Hibernensis, Book 17 especially; also 42.6; 2.5; 16, 31.22, 40.4; 16.11, 35.4, 41.7. 25 For this charter language, see Davies, ‘Latin charter-tradition’, pp. 269–70, 273. 26 I am particularly grateful to Thomas Charles-Edwards for discussing this terminology with me. 27 ‘Additamenta’, 11.1, 16.1, 16.2; 16.1; 11.2, 11.4, in Patrician Texts, ed. Bieler, pp. 174, 178.



although there is a heavy emphasis on transmission of land through inheritance, with a possible implication that alienation of land (whether by gift, sale or exchange) was in practice extremely rare, in fact the law tracts and commentaries employ a vocabulary of gift and sale, especially in the context of contract (which does include gift). The principal words are compounds of berid (bears, gives, brings), crenaid (buys) and renaid (sells); and the jingle reicce na creicce (selling and buying) occurs quite frequently. Hence, in one of the several cases which specify those who cannot make a contract: ‘Ní túalain[g] reicce na creice na cuir na cuind(u)ruda sech óen a c(h)enn . . .’ ‘She is not capable of sale nor of purchase nor of contract nor of bargain without one of her guardians’, that is, a woman cannot sell or buy anything without the involvement of one of her guardians.28 Heptad 50 specifies the purchases a son can make without his father, such as an article of equipment for his farm or joints of meat.29 As in the latter case, many of the sale contexts relate to the sale of movables but transactions in land were allowed for by, for example, Córus Béscnai, in the context of gifts to the church.30 Beyond that, there is also a wide related vocabulary: cor (contract), lóg (value, price), tindscrae (payment), saíthiud and díupart and derb-díupart (overpayment), imnae (gift), ón and airliciud (loan), ráth (paying surety).31 I emphasize the number and range of vernacular words because this is a native vocabulary, not a borrowed vocabulary; words for contract run through many kinds of law text (for example, Coibnes Uisci Thairidne) and are not confined to the text termed Di Astud Chor (On the Securing of Contracts).32 These are Irish words for Irish actions. We cannot possibly quantify transactional activity, nor relative proportions of transactions in movables and land, but there is plenty of evidence of the availability of the conceptual apparatus of transacting and some evidence of transacting in practice in early medieval Ireland.

Cereals In both areas there is some reasonable evidence of the intensification of cereal production in the ninth and tenth centuries, although this intensification appears to have begun earlier in Ireland than in Iberia. In Ireland there is a good deal of surviving physical evidence of water mills, from the early seventh century 28 Bandíre-tract 444.5–6 (Corpus Iuris Hibernici, ed. D. A. Binchy, 6 vols (Dublin: Dublin Institute for Advanced Studies, 1978), vol. 2, 440.32–444.11, numbering consecutive through the six volumes, by page and line number); see N. McLeod, Early Irish Contract Law, Sydney Series in Celtic Studies, 1 (Sydney: Centre for Celtic Studies, University of Sydney, 1992), p. 71. 29 McLeod, Early Irish Contract Law, p. 70. 30 Ibid., pp. 84–5; cf. Flechner, The Hibernensis, 40.4. 31 See McLeod, Early Irish Contract Law, pp. 124–96, 34, 182, 43–6, 84–5, 84, 16–21, for examples; cf. Kelly, Guide to Early Irish Law, pp. 158–72, 120–2, 117–19. 32 McLeod, Early Irish Contract Law; see L. Breatnach, A Companion to the Corpus Iuris Hibernici, Early Irish Law Series, 5 (Dublin: Dublin Institiute for Advanced Studies, 2005), pp. 244–6, for important comment on McLeod’s edition.



onwards, as noted above: there were 38 pre-eleventh-century mills at the last published count, with the important horizontal mill of Kilbegly (near Ballinasloe) added since then.33 There are also plenty of references to mills and milling in Irish law tracts; these are not confined to the issues arising when running a water course across a neighbour’s land, for they also include references to appropriate shares in mills according to status and to millwrights.34 For northern Iberia the physical evidence is lacking – not one mill – but there are many textual references to individual cases. There are the disputes over water courses to power mills but there are also shares in mills (time shares, as it were); these occur as a normal part of property transactions: a gift of one day in eight in a Castilian charter of 962; or a sale of three portions of two mills on the river Arnoia in southern Galicia in 964; or a gift to the monastery of Piasca of a day and a night’s milling every Wednesday (or Thursday) c. 980.35 These transactions might be recorded on a scrap of parchment or on the back of an unrelated text; they are treated as normal and unremarkable, and certainly not sufficiently notable to justify a large and elaborate record.36 Now, water mills are about grinding corn, and they imply grinding at some scale. We see it in both regions, through different kinds of evidence. Corn-drying kilns and grain storage pits echo the message about scale, and scale suggests production for distribution. There are some single corn-driers but, for example, the early medieval complex at Raystown, Co. Meath, in the eastern midlands of Ireland, included five ‘figure-of-eight’ corn-driers, as well as its six early medieval mills, with an increasing volume of wheat processed in phases 3, 4 and 5 (that is, from the seventh century onwards).37 Recently four corn-driers have been found

33 N. Brady, ‘Mills in medieval Ireland: Looking beyond design’, in S. A. Walton (ed.), Wind and Water in the Middle Ages: Fluid Technologies from Antiquity to the Renaissance, Penn State Medieval Studies, 2 (Tempe, AZ: Arizona Center for Medieval and Renaissance Studies, 2006), pp. 39–68, at pp. 47–9; N. Jackman, ‘Early medieval food-processing technology at Kilbegly, Co. Roscommon: The miller’s tale’, in M. Stanley, E. Danaher, J. Eogan (eds), Dining and Dwelling: Proceedings of a Public Seminar on Archaeological Discoveries on National Road Schemes, August 2008, NRA Monograph Series, 6 (Dublin: National Roads Authority, 2009), pp. 9–18. 34 For example, the share of the ócaire in Críth Gablach: Críth Gablach, ed. D. A. Binchy, Mediaeval and Modern Irish Series, 11 (Dublin: Dublin Institute for Advanced Studies, 1941), 4.96. See Kelly, Early Irish Farming, pp. 482–5. 35 C102, Cel403, S305. 36 The Piasca record is upside down in the top margin of a long record of an unrelated gift; AHN, Madrid, clero Sahagún, carpeta 875, no.16 (published as S305, after 19 April 980); the principal gift on this parchment is S242, of 966. 37 M. A. Monk and E. Kelleher, ‘An assessment of the archaeological evidence for Irish corn-drying kilns in the light of the results of archaeological experiments and archaeobotanical studies’, Journal of Irish Archaeology, 14 (2005), pp. 77–114; M. Seaver, ‘Through the mill: Excavation of an early medieval settlement at Raystown, County Meath’, in O’Sullivan and Stanley (eds), Settlement, Industry and Ritual, pp. 73–87; M. Seaver, ‘Early medieval settlement and burial on the Blackhill: Excavations at Raystown, Co. Meath’, in C. Corlett and M. Potterton (eds), Death and Burial in Early Medieval Ireland (Dublin: Wordwell, 2010), pp. 261–79, at pp. 268–70, 275.



at Roestown, Co. Meath; and another four at seventh- to tenth-century Kilgobbin, Co. Dublin, processing rye, barley, oats and plenty of bread wheat; and there are three at Gortybrigane, Co. Tipperary, and two at Sallymount, Co. Limerick.38 Of course, most of these groups of kilns were in successive rather than simultaneous use but the numbers clearly indicate sustained production over the later part of the early middle ages. So far there are no corn-drying kilns of this period from Spain, although there are some earlier kilns (sixth and seventh century) from the Duero basin. Of considerable significance, however, are the grain storage pits from early medieval Spain. Nearly a hundred were found at Villaoreja (Palencia), starting in the tenth century; at Zornoztegi (Álava), in the southern Basque country, there were groups of pits from the eighth to the eleventh centuries (thirty were excavated), and a series of freshly cut terraces for arable cultivation.39 Sopelana and Zapata’s graph shows a large volume of winter wheat from the tenth-century pit, but less of this grain from both eighth- and eleventh-/twelfth-century features. New terraces for cereal growing were established at Monte Gaias, just south of Santiago de Compostela in Galicia.40 There is, then, some limited but nevertheless very useful archaeology. On the other hand, there is a mass of textual evidence from Iberia of new planting and of extending areas of cultivation: ‘the vineyard Diego Muñoz planted [sometime before 963]’; cleared land (terra calva); and a host of stories about peasant intrusion into monastic land.41 Typically, in the latter case, peasants entered a stretch of monastic mountainside, started to cultivate it and take a crop; this provoked complaints and eventually an agreement whereby the peasants were allowed to continue working the land but had to return some of the produce to the monastery. For example, just before 955, on the northern edge of the meseta, twenty-one named peasants, from three different settlements, agreed to pay the monastery of Pardomino a quarter of their produce every year, because they had entered Pardomino’s land, had ploughed it, pastured their animals and cut down trees.42 This was peasant expansion – literally the cultivation of uncultivated land; stretches of woodland and

38 R. O’Hara, ‘Early medieval settlement at Roestown 2’, in M. Deevy and D. Murphy (eds), Places along the Way: First Findings on the M3, NRA Monograph Series, 5 (Dublin: National Roads Authority, 2009), pp. 57–82; T. Bolger, ‘Excavations at Kilgobbin church, Co. Dublin’, Journal of Irish Archaeology, 17 (2008), pp. 85–112, at pp. 95–7, 105–7; P. Long, ‘Food for thought: Newly discovered cereal-drying kilns from the south-west midlands’, in Stanley, Danaher, Eogan (eds), Dining and Dwelling, pp. 19–28. 39 J. A. Quirós Castillo, ‘Early medieval landscapes in north-west Spain: Local powers and communities, fifth-tenth centuries’, Early Medieval Europe, 19 (2011), pp. 285–311, at p. 301; I. Sopelana and L. Zapata, ‘Primeros resultados de los estudios carpológicos del despoblado de Zornoztegi (Salvatierra-Agurain, Álava)’, in J. A. Quirós Castillo (ed.), The Archaeology of Early Medieval Villages in Europe (Bilbao: Universidad del País Vasco, 2009), pp. 437–45. 40 P. Ballesteros-Arias and R. Blanco-Rotea, ‘Aldeas y espacios agrarios altomedievales en Galicia’, in Quirós Castillo (ed.), Archaeology of Early Medieval Villages, pp. 115–35, at pp. 130–4. 41 T69, cf. Cel338 (989); Cel398 (963), Cel174 (964). 42 Li184 (944); Lii290.



scrub and grassland form what the texts call terra indomita ‘untamed land’; these references to untamed land decline across the tenth century. There was action by major landowners too – much accumulation of property, notable from the second quarter of the tenth century by big monasteries and from the last quarter by high aristocrats. The property was acquired in various ways but much was by purchase, with land deliberately bought for deliberate exploitation; some was by exchange, especially of peasant plots, for consolidation of the arable into a larger cultivable space. This was the beginning of farming for profit.

Valuation Although primarily interested in social stratification, H. M. Chadwick devoted the two opening chapters of Studies on Anglo-Saxon Institutions, with a further ‘Excursus’, to monetary systems. There has been a vast amount of work on Anglo-Saxon coinage since he wrote, but it is notable that he had a very good grasp of the difference between coin, unit of account and unit of weight and of the importance of differentiating between them – a point one cannot always make of recent comment. I drew attention, at the outset, to the use of a cattle standard in both early medieval Ireland and early medieval Iberia. A cattle standard was used to measure the value of objects exchanged in payment (or to measure the amount of compensation due, for example, for wounding) in northern Spain; and to measure the properties which gave value to land in Ireland (or, again, compensation for wounding or to measure honour price). These were valuation systems, mechanisms for assigning value when it was necessary to make comparisons between unlike things; they did not require cattle to be exchanged, although they could have been, for they were in essence notional systems of value. Now, in neither case were cattle the only standard of value. In Iberia there was a silver solidus unit – a carpet worth five solidi, six-solidi-worth of wine or a price of nine solidi, paid in cattle, grain and cider.43 There was also a modius unit, literally a measure, in practice something like a sackful (of grain); hence, a cloak of four modii; an ox, a cow and a calf priced at twenty-four modii; a goat worth a modius, cabra modiale.44 There were also quarters, which tend be associated with modii, for example, a total value of six modii and two quarters (paid in wine, rye, wheat,

43 ‘uno tapete valente V solidos argenti’ (C199 [988]); ‘sex solidos in vino’ (S185 [961]); ‘in precium in viiii solidos, in bove et cevaria et sicera’ (Li143 [941]). Solidi were not minted and do not appear to have been available as coin; they were simply a unit of account, presumably derived from knowledge of the late Roman gold solidus or perhaps from the reformed Carolingian currency, in which silver denarii were coined at the rate of 12 to the solidus, of which there were initially 22 to the pound weight. 44 ‘saia de IIII modios’ (Cel410 [987]); ‘bove et vacca cum sua agnicula preciatos in XX et IIII modios’ (Cel402 [961]); Cel451 (942). The word modius occurs as a measure of weight as well as a unit of account, hence modii of wheat, rye and millet. For much longer discussion of this, see ch. 6 above.



sheep and cloth).45 The cloth in the latter example is valued by tremissis, literally a third of a late Roman solidus, so lentio tremisale, a tremissis-worth of cloth. Some of these, like the tremissis, are extremely rare; some, like the modius, are very common. Some were regionally specific: solidi valuations were more common in Castile and cattle valuations in Galicia, although they were not alternatives. Different units of account were used concurrently; for example, horses were valued in cattle and silver was valued in solidi, in one transaction.46 This is partly because different units were considered appropriate for different kinds of object, but also because there were different cultural habits in different micro-regions. There were also several units of account in use in early medieval Ireland. Firstly there was the bó (cow) and the sét (often, though not always, half a milchcow): land worth twenty-four milch cows (cumal .iiii. bo fichit mblicht a log) and ‘three sét are his honour price’ (tri séoit lóg a enech).47 There was the míach, like the modius both a measure of weight and a unit of value: for Fergus Kelly a bushel of several possible cereals; for Thomas Charles-Edwards a standard measure of malted barley.48 Hence, a bullock worth two míach: ag lóige dá míach.49 There is also the unit of the cumal, literally a female slave; honour price could be expressed in cumala: ‘seven cumals are his honour price’ (vii cumala a enechclann).50 And there was the screpul (from Latin scripulum), a twenty-fourth of an ounce of silver; the value of cows could be expressed in scruples, the more calves a cow had the more scruples it was worth, thirty for a 4-calf cow, forty for a 5-calf cow, fifty for a 6-calf cow, and so on: ‘thirty scruples for a 4-calf cow’ (bo ceathramad loig air .xxx.s).51 Another unit is the ungae (from Latin uncia), an ounce: ‘half-an-ounce-worth of pigs’ (lóg leith ungae di muccib).52 Although there is a certain comparability between the two systems, in that silver, sacks and cattle are prominent elements in both, I would not want to push this too far: since there was considerable variation, in both systems, they were not directly parallel. Although lawyers had a tendency to systematize and to look for strict equations, one can find a cumal equated to four, five, six and ten cows, as well as the more common three.53 Where there is plenty of evidence of practice,

45 ‘vinum sestarios II centeno quartario I triigo quartario I ovelia quartario I lentio tremisale sub uno modios VI quartario II’ (Cel406 [961]). 46 Cel380 (961). 47 Mac Niocaill, ‘Tír Cumaile’, 82; Críth Gablach, ed. Binchy, 5.120. 48 Kelly, Early Irish Farming, pp. 560–99, at p. 588; T. M. Charles-Edwards, Early Irish and Welsh Kinship (Oxford: Oxford University Press, 1993), pp. 478–85, at p. 478. 49 Kelly, Early Irish Farming, p. 588. 50 Críth Gablach, ed. Binchy, 18.450–1. 51 Kelly, Early Irish Farming, p. 506. 52 ‘Additamenta’, 11.3, in Patrician Texts, ed. Bieler, p. 174. 53 See C. A. Gregory, Savage Money: The Anthropology and Politics of Commodity Exchange (Amsterdam: Harwood Academic Publishers, 1997), p. 8, on the co-existence of multiple value systems; and M. Sahlins, Stone Age Economics (London: Taylor and Francis, 1974), p. 279, on the way equivalences swing up and down within a community.



as there is in northern Iberia, there was great variation; the use of these notional units will have varied with local circumstances and we should not expect strict equivalences. What is important is that both cultures used units of account; they both had systems of valuation. The very existence of such systems means that things were exchanged (there is no need for a unit of account unless things which are dissimilar are regularly exchanged), for value systems necessarily relate to exchange; that most commonly means commodity exchange, exchange as a system of buying and selling, exchange with a touch of the commercial.54 Exchange must have been normal in these societies, as is in any case suggested by the terminology of transactions. We cannot hazard a guess at volume in either case, because of the nature of the evidence, although the large body of Irish archaeological evidence of focused production is suggestive and in both regions more intensive production appears to have been regionally limited. Niall Brady’s map of water mills, producing cereals for distribution, shows a regional concentration, in the south, middle and east of Ireland; and there are far more kilns in the east midlands than elsewhere.55 In northern Iberia, as well as evidence of the existence and use of valuation systems, we have hundreds of records of actual transactions. There were hundreds of sales for value, high and low value, especially on the central plateau of the meseta. The range of social levels involved in such transactions was wide, from high aristocrats through middle-range people to peasants: a woman and her children sold plots in Rabal for food, valued in modii; a peasant and his mother sold a water course on the river Porma (meseta) for fleeces valued in solidi and wheat valued in modii; and so on.56 These ideas about value were used.

Silver As mentioned above, in neither culture was coin minted in the early middle ages. The first coining in Ireland came round about 997, by Dublin Vikings; that in northern Iberia in the mid-eleventh century, by King Fernando I. By the tenth century, however, there was plenty of silver available in both areas: silver, not gold. There was some gold about – gold objects have been found in Ireland and there are references to gold objects in Spanish lists – but silver appears to have been much more common; there are no references to the (gold) mancus in either case.57

54 C. A. Gregory, Gifts and Commodities (London: Academic Press, 1982), pp. 19, 47–9: ‘commodities are those things that arise as things pass from house to market’; Gregory, Savage Money, p. 13. 55 Brady, ‘Mills in medieval Ireland’, p. 46; Monk and Kelleher, ‘Assessment of the archaeological evidence for Irish corn-drying kilns’, p. 82. 56 Cel373 (956); Li124 (937). 57 For references to the mancus (which occur overwhelmingly in Italian and Anglo-Saxon texts), see M. McCormick, Origins of the European Economy: Communications and Commerce, A.D. 300–900 (Cambridge: Cambridge University Press, 2001), pp. 323–42, 812–14.



In Ireland, as we have seen, there were silver units of account in the screpul and ungae, cited in eighth-century law tracts and later commentaries (and sometimes elsewhere); and there have been finds of finely wrought silver objects, especially from the eighth-century onwards (finds such as the Derrynaflan hoard of liturgical vessels or the Tara brooch).58 Silver was available and there were some skilled silver smiths from an early date. But, from the mid-ninth century there is some increase in the material evidence of silver availability, largely from silver hoards; this is part of a much wider, northern, phenomenon – a cultural marker of the Viking world. The silver was of foreign origin (recycled foreign coin and other objects) and the hoards began to be deposited round about 850, coincident with early Viking settlement in Ireland. There is every reason to suppose introduction of this silver was a direct consequence of Viking activity. There has been much important work, on the entire northern corpus by James Graham-Campbell, and on the Irish corpus by John Sheehan.59 The hoards are of bullion or of coin or both; there were initially bullion hoards, deposited from c. 850 to c. 950 – ingots, armrings, hack silver (fragments); there was a smaller group, according to Sheehan’s analysis, of bullion and coin hoards, from the early tenth century; and then some coin-only hoards (of foreign coin) from the mid-tenth century onwards, although the latter tend to be very small. Interestingly, most of the bullion hoards were found in areas not controlled or settled by Vikings; in other words, even if the Vikings introduced the silver, it was being collected and by implication used as bullion by Irish people. The hoard evidence from northern Iberia is not remotely comparable. There is a small hoard from Roncesvalles (of six Anglo-Saxon pennies of c. 990, on a pilgrim route) and a hoard of 205 eighth- and ninth-century silver dirhems from San Andrés de Ordoiz (Navarre) but that is all; Carolingian coins are said to have been found in Santiago.60 I do not know of anything comparable to the Irish bullion 58 Derrynaflan: Journal of the Royal Society of Antiquaries of Ireland, 110 (1980), pp. 91–148; The Derrynaflan Hoard I: A Preliminary Account, ed. M. Ryan (Dublin: National Museum of Ireland, 1983); R. Ó Floinn, ‘Secular metalwork in the eighth and ninth centuries’, in S. Youngs (ed.), ‘The Work of Angels’: Masterpieces of Celtic Metalwork, 6th-9th centuries AD (London: British Museum, 1989), pp. 72–124, especially pp. 77–108; cf. H. Richardson, ‘Visual arts and society’, in Ó Cróinín (ed.), A New History of Ireland I, pp. 680–713. 59 J. Sheehan, ‘Early Viking Age silver hoards from Ireland and their Scandinavian elements’, in H. B. Clarke, M. Ní Mhaonaigh, R. Ó Floinn (eds), Ireland and Scandinavia in the Early Viking Age (Dublin: Four Courts Press, 1998), pp. 166–202; J. Sheehan, ‘Ireland’s Viking-Age hoards: Sources and contacts’, in A.-C. Larsen (ed.), The Vikings in Ireland (Roskilde: Viking Ship Museum, 2001), pp. 51–9; J. Sheehan, ‘The form and structure of Viking-age silver hoards: The evidence from Ireland’, in J. Graham-Campbell and G. Williams (eds), Silver Economy in the Viking Age (Walnut Creek CA: Left Coast Press, 2007), pp. 149–62, at pp. 149–54, updates his earlier figures and introduces a new classification. J. Graham-Campbell, Viking Artefacts: A Select Catalogue (London: British Museum, 1980); J. Graham-Campbell, The Viking-Age Gold and Silver of Scotland (AD 850–1100) (Edinburgh: National Museums of Scotland, 1995). 60 F. Mateu y Llopis and R. H. Dolley, ‘A small find of Anglo-Saxon pennies from Roncesvalles’, British Numismatic Journal, 27 (1955), pp. 89–91; J. M. de Francisco Olmos, ‘El nacimiento de



hoards. There are very few chance finds, since, as in Ireland, metal detecting is illegal. Contrast the south – al-Andalus – where there have been finds of thousands of tenth-century Arabic dirhems: 6,800 from Haza del Carmen (Córdoba), 3,632 from Fontana and so on.61 That might suggest that the north was coin- and silver-free. However, we have seen silver solidi (solidi argenti) as units of account; these were commonly cited in some areas in the tenth century, especially in and around the city of León on the meseta. Now, while textual references to solidi are usually to the valuation system, there is reason to think some texts refer to the use of metal. As already mentioned, silver artefacts were sometimes used to make payments, silver scalae especially, but also silver bowls and silver-decorated saddles and bridles. Some descriptions of prices paid are explicit that silver was used by weight: a payment in 935 of forty solidi, twenty in silver and twenty in cloth (viginti in argento et viginti in panno); a payment in 980 of 160 solidi ‘in pure silver (argentum purum)’.62 There is another expression which also implies the use of something with silver content: some payments (and occasional valuations) were made in units called argenzos, arienzos, arenzos, argenteos, argencios and other variants. So, a price of twenty-one and a half argenzeos: ‘in precio argenzeos XXI et medio’; a price of four solidi, paid in cloaks and arenzos: ‘in pretio IIII solidos in saiale et in arenzos’; a price for an arable plot on the river Cea of four bags of spice and two and a half argenzos: ‘in precio pro ipsa terra IIII folles zumakes, argenzos II et medio’.63 The manner of reference to these arienzos suggests that they were objects, presumably pieces of silver, presumably used by weight; they could have been bullion – hack silver, foreign coin, cut coin. The precision of the references (especially the halves) suggests the possibility of a standard weight (or at least a notional standard weight): dirhems were regularly cut to a quarter and a fifth weight in the Muslim south.64 Now, references to arienzos are very restricted regionally. They are common in León material from 910, common in Sahagún material (to the south-east of León, also on the meseta) from 930 – precisely the point when sales became more common – and extremely rare in Galicia; they are also rare in Castile and Navarre and in the far north. León was a city. It must be likely that silver was available as bullion, and used in transactions in the city and


62 63 64

la moneda en Castilla: de la moneda prestada a la moneda propria’, in I Jornadas Científica sobre Documentación jurídico-administrativa, económico-financiera y judicial del reino castellanoleonés, siglos X–XIII (Madrid: Universidad Complutense de Madrid, 2002), pp. 303–46, at pp. 311, 308. See A. Canto, ‘Al-Andalus: sus monedas’, in D. R. López Guzmán (co-ord.), El Zoco. Vida económica y artes tradicionales en al-Andalus y Marruecos (Granada: El legado andalusí, 1995), pp. 35–41; E. Manzano Moreno, Conquistadores, emires y califas. Los Omeyas y la formación de al-Andalus (Barcelona: Crítica, 2006), pp. 62–3, 311–16; on chance finds, see the following, n. 67. Li103; S309. S82 (942), Li193 (946), Li197 (947). For much fuller discussion of these units, see ch. 6 above, pp. 107–9; for cut dirhems, Canto, ‘Al-Andalus: sus monedas’, p. 37 (p. 169 for illustration of a fragment), and Manzano Moreno, Conquistadores, p. 316.



its hinterland, for example in places like the market at Cea. It is exceptionally frustrating that we have not so far found the physical evidence in the ground but it must surely be there. Indeed, a couple of years ago, four tenth-century andalusí dirhems were turned up by an illegal metal detectorist in Álava, the southern Basque country. These were cut dirhems, precisely the sort of things one would expect to be described as arienzos.65 Meanwhile, the Scandinavian analogy is very helpful. Cut dirhems occurred in southern Scandinavian hoards of the tenth century; some silver, of whatever origin, was cut to standard weights in Scandinavian contexts, although the weight standard (in the order of 25 grams) is arguable; silver was clearly used in transactions as bullion. I can envisage something comparable to the Scandinavian practice in and around León, using, especially, cut coin from the south, but also hack silver. Indeed, it is worth considering if some of the references to silver scalae, in payments, could be to ingots (i.e. something to put on the scales): normal practice is to translate this word ‘dishes’, as is clearly the meaning in some lists of church treasure, but ‘ingots’ would make much sense in transactional contexts.

Conclusions There are many differences between early medieval Ireland and early medieval Iberia – differences in scale, climate, cultural and political background, neighbours and so on. But there are some real similarities. Both societies were conceptually prepared for exchange, and particularly commodity exchange, that is sale and purchase; and the richness of the vocabulary in the one case and the richness of the written evidence of transactions in the other suggests active exchange economies in both. Secondly, in both regions there was intensification of production – production of cereals on some scale in the south and east of Ireland and on the meseta in Iberia, which must have been for distribution (the disputes over water courses near León presumably arose in the context of milling grain to feed the city). Niall Brady made a perceptive comment, when considering the distribution of Irish water mills, that one could begin to see the rational mind at work – just as one can see the beginnings of farming for profit in Spanish texts. All this was well underway by the tenth century, the Irish take-off beginning at least two centuries earlier.66 Thirdly, the availability of silver bullion in some parts (again south and

65 J. A. Quirós Castillo, personal comment 26/3/2009, for which I am very grateful. 66 The Irish take-off point looks to have been the earlier (because the greatest numbers of mills are from eighth and ninth centuries), but this may be an issue of uneven evidence, rather than a substantive difference, given the very few written texts from eighth- and ninth-century Iberia and the limited archaeological record. However, ploughs with sock and coulter do not appear to have been used in Ireland before the tenth century, and Brady argues for increasing landlord control of resources in eighth and ninth centuries and increasing production for a growing population in the tenth; N. Brady, ‘Food production in medieval Ireland, aspects of arable husbandry’, Ruralia, 8 (2011), pp. 137–43.



east Ireland and the Iberian meseta), from the mid-ninth in the one case and early tenth century in the other, suggests exchange at a volume worth facilitating with silver. Silver could easily be a medium of exchange, without being coined. In the northern world ‘bullion’ and in some cases ‘metal-weight’ economies were practical. Finds of balances for weighing silver are widespread in northern archaeological contexts, and there are lead weights too.67 There is no need to suppose these practices were confined to the north – one tenth-century Beato (that is, an illustrated copy of the work of Beatus of the Liébana), the Beato of Valcavado, from the meseta (Palencia), clearly shows a very standard balance in its representation of one of the four horsemen of the Apocalypse. While this must ultimately depend on the biblical comment that the third horseman carried a balance (Revelations 6:5), as is indeed cited on the folio, the artist could hardly have drawn so convincing a balance if he was completely unaware of its characteristic features.68 Further, in northern Spain and Portugal rich people, lay and ecclesiastical, amassed goods in the tenth century – goods of liturgical use for ecclesiastical corporations (chalices, vestments, bells, books) but household goods for both those and the laity (bedrolls, bedspreads, feather mattresses, furs, rugs, table cloths, cups, cauldrons, dishes, flasks, bowls, jugs, table service). The latter could be silk, highly decorated, silver, gilded or ivory; and they could come from the andalusí south, from ‘Iraq’, or from the Greek (i.e. Byzantine) east.69 These goods clearly circulated, for they were used to make payments; so, for example, a monastery would buy land with a carpet or a mattress or a Moorish bedcover. Goods moved from churches to lay people, lay people to other lay people, lay people back to churches and so on. These transactions were not confined to the aristocracy – saddles, bridles, linen, cloaks, tunics, blankets and fleeces passed from lay party to lay party in small-scale transactions, from quite early in the tenth century; there are cases from 923. There is also a little evidence of the circulation of really lowvalue goods: in an interesting study of excavated pottery from the Basque country, J. L. Solaun has demonstrated that in ninth- and tenth-century Álava, although there was some domestic production in and for the household, pottery was also supplied by itinerant vendors and, importantly, was made in local workshops for limited distribution. For example, in a zone near Vitoria, in the two hundred or so years to c. 950, 19% of the ceramics came from domestic production, 30% from itinerant traders, 25% from local workshops and 8% from supra-regional sources; in the following hundred years those proportions changed so that proportions

67 For illustrations of a discovered balance, lead weights and cut Arab coins from Viking contexts, see Graham-Campbell, Viking Artefacts, nos. 306, 345, 346, pp. 83, 264, 99–100, 274. 68 Beato de Valcavado, Library of the University of Valladolid; page reproduced in M. Sartori Mana (co-ord.), In Principio Erat Verbum. El reino de León y sus beatos (León: Sociedad Estatal de Conmemoriaciones Culturales, 2010), p. 79. I am extremely grateful to Professor Jill Mann for pointing me to the appropriate Revelations verse: ‘And I beheld, and lo a black horse; and he that sat on him had a pair of balances in his hand’. 69 See further ch. 9 above.



from domestic and itinerant sources dropped while those from local workshops increased nearly three-fold. There seem to have been regional zones of market activity for low-value goods.70 There are then some real similarities between Ireland and northern Iberia in the early middle ages. Both areas had active, growing economies (or complexes of economies) and shared in wider western European trends. It clearly was not necessary to mint coin to participate in those trends. Indeed, the absence of coin is probably more to do with the absence of sustained fiscal demands than with the absence of production or of surplus. The circulation of material goods suggests that there was some demand for them.71 In Iberia the desire for things is demonstrable across social strata. It is more difficult to demonstrate this for Ireland but when the mass of recently excavated material is analysed, it will be surprising if one cannot detect local distribution patterns of low-value goods such as iron and bone objects (knives, hooks, awls, needles, keys, pins, combs, beads) – the Irish analogue for Basque pottery from local workshops. There is, after all, plenty of evidence of stone-working, bone- and antler-working and ironworking.72 The butchering of relatively old (three-year-old) cattle to provision settlements, on a suggested semi-commercial basis, as at tenth- and eleventh-century Clonmacnoise, Dublin and probably Knowth, clearly shows a different kind of demand.73 For higher value goods the closest one can get to indicators of demand are the many lists in the Book of Rights. This is a later text (probably twelfth century) but it includes lists of goods either distributed or required – horses, silver-decorated bridles, saddles, cloaks, tunics, rings, bracelets – many of which are similar to those listed in Spanish texts (although with a stronger military flavour overall).74 Let me end by returning to the persistence of tradition. As I hope I have shown, both societies were societies in which ancient things were valued and had a continuing relevance. But these were not societies that were bound by tradition, much less were they ‘traditional societies’. They were changing and rapidly evolving. They had complex economies, sharing European trends. For all their differences, they were similar in having active exchange economies without using money. And recognizing that is a useful lesson.

70 J. L. Solaun Bustinza, La cerámica medieval en el País Vasco (siglos viii–xiii) (Vitoria: Gobierno Vasco, 2005), pp. 380, 310, 316. 71 For the significance of demand, see C. Wickham, ‘Rethinking the structure of the early medieval economy’, in J. R. Davis and M. McCormick (eds), The Long Morning of Medieval Europe: New Directions in Early Medieval Studies (Aldershot: Ashgate, 2008), pp. 19–31. 72 E. g. M. Comber, The Economy of the Ringfort and Contemporary Settlement in Early Medieval Ireland, British Archaeological Reports, International Series 1773 (Oxford, 2008), pp. 173–80, 187–92. 73 F. McCormick and E. Murray, Excavations at Knowth, 3, Knowth and the Zooarchaeology of Early Christian Ireland, RIA Monographs in Archaeology (Dublin: Royal Irish Academy, 2007), pp. 51–8, 116. 74 Lebor na Cert: The Book of Rights, ed. and trans. M. Dillon, ITS 46 (Dublin: Irish Texts Society, 1962), e.g. pp. 84–5, 88–9, 104–5.



In this tribute to Professor Morimoto I will compare two regions which lie on the continental mainland and for the most part outside the Carolingian Empire: eastern Brittany and northern Iberia (west of Catalonia).1 I want to ask how far they share characteristics of land management and how far these contrast with Carolingian practice.2 The areas are of very different size (in the order of 17,000 km2 against more than 190,000 km2) but since both lie just to the west of the Empire, and since I have done detailed work on the land-use of both, it is interesting and potentially instructive to consider them together. A few essentials need stating at the outset. These are areas of very different terrain, as well as size. Brittany is predominantly a land of broad plateaux, mildly undulating and sloping gently to the sea; there are no great heights and the heavily indented coastline is long; in the east the brown earth soils are susceptible to erosion but the soft schistes on which they lie are a constant source of new material, so there is plenty of good arable farming land. The moderate climate is moist, with relatively warm winters and relatively cool summers. Iberia, on the other hand, is a land of high plateaux, punctuated by high mountain ranges which are broken by the two great river systems of the north, the Duero and the Ebro. Much of the Duero basin lies on the plateau, the meseta, which is hot and arid in summer, although it is crossed by long river valleys running south to the Duero from the Cantabrian Mountains. Beyond the Cantabrian Mountains to the north is a narrow coastal strip; to the west, the high rainfall of Galicia makes it wet and green for much of the year; to the east lie the high sierras bordering the Rioja and beyond them the mountains and hills of Navarre. It is a very diverse land, with mountainous regions that are unsuitable for cultivation, though farmed in parts, and a hot plateau that requires careful management. However, these regions had predominantly mixed farming in

1 First published in J.-P. Devroey and A. Wilkin (eds), Autour de Yoshiki Morimoto. Les Structures agricoles en dehors du monde carolingien, formes et genèse, Revue belge de philologie et d’histoire, 90 (2012), pp. 361–80. 2 Brittany may reasonably be considered part of the Carolingian Empire in the period c. 819–842, but only then.



the early middle ages – arable and pastoral, orchards and vineyards, probably with a greater element of arable in eastern Brittany and a greater element of viticulture in northern Iberia, in both cases quite unquantifiable. For both regions the available source material is not strictly comparable with that of the Frankish heartland and has limitations: there are no polyptyques, although there are a few brief inventories; there are charters, but the availability of charters is confined to limited periods. Most Breton charters come from the ninth century, especially from two generations in the middle of the century, and from the cartulary of the monastery of Redon, which was founded in 832.3 There are no surviving originals and the number of charters is small – only 345 – but they relate to a well defined area and provide unusually dense coverage at an unusually local level. There are a few other Breton charters, including Landévennec material from the tenth century, but there are not enough of these to support sustained analysis and they have several corrupt elements.4 Iberian charters are much more numerous and come from many different sources (monasteries, bishoprics and lay archives). Over 2,330 ninth- and tenth-century charters survive from west of Catalonia (including Portugal), surviving as cartulary copies and on single sheets. However, there are few from the ninth century (less than 200) and the series do not become numerous until the 930s. Although all the charters are of recognizable European types, and record transactions in similar ways, there are subtle differences between them. The Breton texts are usually easily localizable, specifying meeting places and topographically recognizable details of property. Iberian charters very rarely specify meeting places, and although they include plenty of statements of boundaries, the boundaries are often defined by reference to other people’s properties, which is of limited use to the modern investigator. This makes precise localization extremely difficult. Breton charters therefore give narrow but deep coverage – there are many transactions for a few adjacent places; whereas Iberian charters provide very wide, but shallower, coverage. In what follows I will try to address the fundamental questions posed by our organizers: if the ‘classical’ landed estate model does not apply in these zones beyond the Frankish world, how did aristocratic elites derive their income and by what, if any, mechanism did they control production? Given that there were aristocratic elites, to what extent did an independent peasantry have a role in the exploitation of the land and did they benefit from it? To do this I shall comment on peasant properties, large estates and the organization of labour, other sources of aristocratic income and changes in the systems.

3 In addition to CR, see also the facsimile edition Cartulaire de l’abbaye Saint-Sauveur de Redon (Rennes: Association des amis des archives historiques du diocèse de Rennes, Dol et Saint-Malo, 1998), and the fundamental comments therein of H. Guillotel, ‘Le manuscrit’ and ‘Répertoire chronologique’, pp. 9–25, 71–8. 4 Le cartulaire de l’abbaye de Landévennec, ed. A. de la Borderie (Rennes: Ch. Catel (for Société archéologique du Finistère), 1888). See now the facsimile in Cartulaire de Saint-Guénolé de Landévennec, ed. S. Lebecq (Rennes: Presses universitaires de Rennes, 2015).



Brittany The river Vilaine, in eastern Brittany, constituted a major cultural divide in the ninth century: east of the Vilaine the nomenclature (personal and place) was overwhelmingly Romance or Frankish; west of the Vilaine it was Breton; and several early medieval texts treat the river as a boundary between Franks and Bretons. Despite that, the modern administrative region of Brittany includes land to the east of the river and the political history of more than a millennium has linked east and west. Most of the charters deal with lands to the west of the river, but the few from east of the river suggest some significant differences in land management there. Peasant properties West of the Vilaine there was a primary unit of civil association in the ninth century, always termed the plebs. Peasants identified themselves, and located their properties, in relation to the plebs. The word itself suggests that such groupings may in origin have been ecclesiastical but by the ninth century the word plebs in these texts meant the village community and its territory.5 Within the plebs the basic agricultural working unit was known as the ran (plural rannou), which literally means ‘portion’ or ‘share’ (occasionally tigran, that is ‘ran plus house’, and sometimes bot, that is ‘house’). The ran was clearly the working unit of a household, and it makes most sense to think of it as a small farm (‘smallholding’ in English).6 We can make reasonable estimates of size and say something about its composition. Size would of course have varied with the nature of the landed resources available but the mean size lay between 10 and 25 hectares, with an extreme range of 5 to 60 hectares. The ran had two different forms: it was often a discrete unit, and might include meadow and pasture as well as arable; it had a mixed agricultural character. Sometimes, however, the ran was not discrete but the term referred to a plot of arable, lying beside other arable plots in a large arable zone, the owner having rights in meadow, pasture and woods beyond this arable zone. It is perfectly clear that many ran-owners were free peasants: they engaged in transactions as independent actors, buying, selling and mortgaging property; they played a vociferous part in community business, arguing and taking each other to court. Some of them had more than one ran: depending on the location, an estimated 3% to 7% of the total known free population (that is, as a proportion of all free tenants and proprietors) and 30% to 56% of known proprietors had accumulations of rannou, most usually two, occasionally three or four and exceptionally five.7 This means that there were some rich peasants, who 5 See W. Davies, Small Worlds: The Village Community in Early Medieval Brittany (London: Duckworth, 1988), pp. 63–7. 6 The word ‘smallholding’ in English does not in itself imply that the unit was a tenancy. 7 Some local priests, who in many ways were part of the rural peasant community, had up to eight properties; Davies, Small Worlds, p. 95.



accumulated property and, since one household could not itself work three or four rannou, we often find serfs tied to such properties. So, rich peasants could be serf owners too. The indications are that the serf managed the farming of the ran, and returned most of the produce to the peasant owner. There was usually one serf (and family) per ran (as happened in more than half the cases) but sometimes there were two or three. A quarter of all properties that were the subject of transactions had associated serfs; that is probably a reasonable estimate of the overall proportion of serf-managed rannou. There was great variation from plebs to plebs, however: some plebes had relatively many free peasants and some had relatively many serfs. It looks as if there were roughly equal numbers of free and servile peasants in Augan (i.e. 50% and 50%), whereas in Carentoir there were about 90% free and 10% serfs and in Bains 70% free and 30% serfs.8 Large estates and labour Firstly, a warning: although there were plenty of arguments about weirs, water and water rights, there were none about mills. Given the ubiquity of cereal production, this suggests that production was neither for wide distribution nor was it controlled by large-scale producers in this area.9 Nevertheless, it is perfectly clear that there were some properties that were larger than the peasant working unit, as also that some individuals owned many, evidently scattered, properties. For example, Count Pascwethen (who became co-ruler of Brittany for a short period in the 870s) owned properties from Plélan in central east Brittany to Guérande on the south coast to Grand Fougeray to the east of the Vilaine – lands that were 70 km apart; these included ten salt pans on the coast, which are likely to have been a commercial enterprise, producing salt for distribution.10 We can also see some lesser aristocrats with smaller property interests, in narrower zones, for example of about 30 km across, on both sides of the river. One woman had interests in Médréac and surround, to the northwest of Rennes, and another was dealing in property in and around Lusanger, to the south of Rennes.11 Another of these aristocrats, with the Germanic name Raginbald, alienated an estate with eighteen named serfs and children as well as another estate with five named serfs.12 The latter are exceptional and they lie well to the east of the river Vilaine. One could argue that these must have been managed rather like the classical bi-partite estate, with serfs

8 Davies, Small Worlds, pp. 91–104. 9 Three fragments of granite millstones were recovered from the excavation of an eighth- to ninth-century farm at La Talvassais, near Montours, although this site, which is well to the north and east of the area considered in this paper, is on the north-eastern edge of Brittany, not far from Avranches; I. Catteddu, Les habitats carolingiens de Montours et La Chapelle-Saint-Aubert (Illeet-Vilaine) (Paris: Maison des Sciences de l’Homme, 2001), pp. 206–7. 10 Davies, Small Worlds, pp. 164–6. 11 Ibid., p. 167. 12 CR41 (845). Cf. the eight named serfs at Coiron of CR59 (849).



providing some of the labour for a réserve/demesne, or one could suppose that they were entirely directly managed. However, we have no detail of the internal management here nor on any other estate. Nor indeed do we have any explicit comments on the size or shape of an estate and it is extremely rare to have a statement of estate boundaries.13 Aristocratic properties west of the Vilaine sometimes had named serfs associated too, as was the case in Augan, although sometimes there was just a generic reference to manentes. West of the Vilaine the number of serfs per property was smaller than in the east: often one, occasionally two or three and once four. In fact, the properties of aristocrats here are very often called rannou, as could be the case east of the Vilaine, though terms like villa, manse and alod are as common in the latter region; aristocratic rannou seem to have been the same kind of smallholding as those owned by peasants, in this case worked by a servile household for the aristocrat rather than for (or by) a free peasant. For example, Pascwethen’s transactions involved saltpans at the Loire mouth, villas in the east and several separate rannou, with associated serfs, in the west;14 and a gift of four rannou in Caro had named serfs associated in three instances – ‘Ran Anaumonoc cum colonis Anaumonoc, cum filiis suis Drecon et Rietoc, et Ranmorenoc cum colono suo Haeluidoe, et Ranuuoranau cum colonis suis Uuoranau et Uuethanau et Driuualoe et Iohan, et Ranroch’ – note that in two cases the ran takes its name from the serf; in another case a ram, two pigs, fifteen loaves, fifteen pence, three measures of oats, eight sesters of rye and one sester of wheat were due in 895 to the owner of a ran in Marzan (on the north side of the estuary of the Vilaine) from its one serf.15 It looks as if, although there might have been a version of the classical estate just to the east of the river Vilaine, west of the Vilaine and at the Loire mouth aristocrats had accumulations of small, separate units, each unit worked by servile farmers, with obligations to return either a specified annual rent in perpetuity or most of the produce. The aristocratic ‘estate’, then, was in practice a collection of small peasant farms – the ran in Marzan, with the serf who was super eam, was just one element of an aristocratic woman’s property. There is no hint that these accumulations were managed as integrated economic enterprises. An element of the labour to work these ‘estates’ therefore came from serfs. But some also came through tenants – although less than 10% of properties transacted throughout the ninth century seem to have been tenanted. Before the 960s tenanted properties were particularly scarce, and tended to lie in aristocratically dominated or commercially exploited parts – Guérande and Batz salt pans, Lusanger to the east of the Vilaine, Plélan and Augan to the west. Some of the tenants look quite

13 There is a detailed perambulation of the lands of the small monastery of St Ducocca, which was given to Redon sometime before 871; these apparently comprised a discrete block (west of the Vilaine) and – from references to Silfiac church, Cléguérec and the river Blavet – must have constituted an area of at least 35 km2 (CR247). 14 CR22 (854), CR23 (859), CR26 (857), CR35 (852), CR72 (858–62), CR85 (862), CR260 (876), CR262 (875), CRA30 (851–57), CRA39 (854). 15 CR269 (878–89); CR266.



wealthy: Pricient, who was a tenant and serf owner from the lower Vilaine, took income from at least three different plebes near Redon; he came into conflict with the monastery over these properties, a dispute which was only settled through the mediation of Pascwethen and two other high aristocrats.16 On a smaller scale, but still notable, Catlowen ran several salt pans on behalf of their owner Duil.17 What was due from tenants, whether a wealthy tenant or the tenant of a single holding, was a return to the landlord, either in kind or in cash. Rents due from free tenants are often accounted in cash and, judging from the anxiety to raise money by mortgaging properties, must sometimes have been paid in cash: 18 pence from a holding in Augan, 2 solidi from a ran in Carentoir, 2 solidi from a tigran in Pipriac.18 2 solidi per year per ran is in fact a common rate. Rents were expressed in kind too – in wheat, oats, rye, loaves, sheep and honey especially – and reductions could be negotiated. Tenants paid rent in an agreed amount of produce or cash, not in labour. I see no evidence of labour service anywhere in this material.19 Landlords managed their properties as an accumulation of peasant properties, using the labour of serfs or tenants or both. The tenants either behaved like free peasant proprietors, whether labouring themselves or using serf labour, or, in the case of those like Pricient, they behaved like minor aristocrats. Aristocratic farming looks like peasant farming, but with more numerous, larger and more scattered accumulations. Other sources of aristocratic income For some aristocrats there was another source of income, although ill defined: these were the dues owed to those who had control (potestas or dominium) of some plebes. We do not always know what this meant, and it only appears to have applied to some selected plebes, but it could involve controlling vacant properties, taking a specific due or – as in the most famous case – taking tolls: he who controlled (haberet in potestatem) Bains had a right to take tolls at Balrit (on the river Vilaine) on ships and commerce.20 We have no idea how such obligations were allocated within the plebs, although if it was something very specific, like a toll, it would have been obvious where the obligation fell.21 It operated concurrently with

16 17 18 19

CR242 (869). CR86 (865), CR169 (863), CR170 (866). CR68 (867), CR63 (863), CR128/219 (834) [doublets]. The word opera occasionally occurs but, whatever its origins, by the ninth century it referred to a set of dues, like the sine opere alicui homini sub caelo nisi Sancto Salvatori of 875, which was – explicitly – a payment of pigs, sheep, oats, rye and wheat (CR262); see further, Davies, Small Worlds, pp. 48–9. 20 CR106 (843–51). 21 This is a difficult subject for there is insufficient evidence available to resolve the many ambiguities, although it is much discussed. For my much longer treatment see Davies, Small Worlds, pp. 174–5, 207–8; but note the views of A. Le Moyne de la Borderie, Histoire de Bretagne, 6 vols (Rennes/Paris: J. Plihon & L. Hervé, Alphonse Picard, 1896–1914), vol. 2, pp. 155–61; M. Planiol, Histoire des institutions de la Bretagne, 5 vols (Mayenne: Association pour la publication



straightforward proprietary rights and was not an alternative: plenty of individuals owned property in the plebs of Bains, although it also had a ‘controller’. One could think of these powers as a kind of ‘extensive lordship’ or as a kind of seigneurial power, although they do not quite fit either model:22 since their origins are lost we cannot determine the source of their authority; they are evidence, however, of the uneven operation of an inherited superiority which could enhance the power of particular aristocrats in particular localities. In other words, some aristocrats could call on other kinds of resource in addition to the profits of direct proprietorship. There is also the issue of ‘tax’. There are occasional references to dues which fell on large estates associated with rulers – sine pastu caballum et canum, lochcaballis, loth and angarium (feeding and transport services).23 These seem to refer to charges on specific lands rather than to any general and widespread obligation to pay dues to the ruler, although they could well have derived from ancient fiscal land. Further evidence for any fiscal obligation is fragmentary but there is one explicit reference to a censum regis and one explicit reference to the ruler’s public resources, that is the Breton ruler: Salomon endowed the monastery of Saint-Sauveur and Saint-Maxent (in Plélan) from nostro publico.24 The censum regis is unambiguous: it occurs in a record of a sale from a layman to a (local) priest; the text, which employs many non-standard formulas, goes out of its way to make clear that the property was unencumbered with obligations to anyone, except for this censum regis. It is quite clear that this was not some kind of proprietary due. Again, there is no reason to think the obligation fell on all properties, and there is not enough to let us see how the obligation was distributed, but it offers a tantalizing hint that rulers, as rulers, were involved in surplus extraction from the land. Changes in the system As indicated above, the number of tenants and tenanted properties visible before the 860s is very limited. Whatever the precise numbers, proportions of tenancies

du manuscrit de M. Planiol, 2nd edn, 1981–4), vol. 2, pp. 80–96; J. M. H. Smith, Province and Empire: Brittany and the Carolingians (Cambridge: Cambridge University Press, 1992), pp. 134– 5; N.-Y. Tonnerre, Naissance de la Bretagne. Géographie historique et structures sociales de la Bretagne méridionale (Nantais et Vannetais) de la fin du VIIIe à la fin du XIIe siècle (Angers: L’université d’Angers, 1994), pp. 238–40. 22 Cf. R. Faith, ‘Social theory and agrarian practice in early medieval England: The land without polyptyques’, in Devroey and Wilkin (eds), Autour de Yoshiki Morimoto, pp. 299–314; and C. Wickham, Framing the Early Middle Ages: Europe and the Mediterranean, 400–800 (Oxford: Oxford University Press, 2005), pp. 320–6, who (following Maitland) writes in terms of ‘superiorities’. 23 CR49 (866), CR50 (866), CR52 (866), CR126 (858), CRA29 (pre-857); CR241 (869), CR242 (869), CR272 (892); most of these cases are associated with rulers. 24 CR136 (942) (cf. the wine due to the ruler Nominoe in CR192 [834–37]); CR241 (869). See the discussion of A. Chédeville and H. Guillotel, La Bretagne des saints et des rois, Ve–Xe siècle (Rennes: Ouest France, 1984), pp. 326–9; also Davies, Small Worlds, pp. 144–6, 205–6, on the possibility that people called maiores, to whom there are occasional references, were ‘tax’ collectors.



certainly increased from the 860s, especially in the period 860–872. That was directly because of the activities of the monastery of Redon, and of its abbot Conwoion in his later years and his successor Ritcant, both of whom appear to have set out deliberately to consolidate and increase Redon properties and manage them for profit.25 There was a rush of peasant donations to the monastery of properties which were received back by the donors on terms of tenancy (with rents now usually expressed in pence); and there were also new tenancies created as a consequence of disputes over ownership: when the peasant Fomus argued with Redon over ownership of one property in Augan, the matter was settled because the monks gave him tenancy of another property in return for an annual rent of wheat and pennies.26 Thereby the property interests of what was now a major monastery increased dramatically: by the late 860s Redon had properties in every plebs in an area 30 km x 55 km around it, and in many beyond. Many free peasant proprietors became free peasant tenants. It is unlikely to have made much difference to the organization of work at the time – families continued to work their small farms as they had always done; but it had long-term consequences: until the eighteenth century Redon continued to have very extensive property in the region and collected rents from a very dependent population.27 And it meant, in accordance with a familiar process, that control of surplus was shifting away from the community of the plebs to a few major landlords.

Northern Iberia Despite the great differences in scale, landscape and terrain, there are some obvious and some not so obvious similarities between northern Iberia and Brittany in the exploitation of landed resources in the early middle ages. Peasant properties The free peasant proprietor is a major actor in Spanish historiography of the twentieth century, although that historiography has been much questioned in the past generation because of its insistence on the colonization of depopulated land.28 We 25 There are no early inventories from Redon, but Ritcant’s organization of the archive of charters is activity of inventory-making type; W. Davies, ‘The composition of the Redon cartulary’, Francia, 17 (1990), pp. 69–90, at p. 81. 26 CR127 (852). The implication is that Fomus gave up his claim to the disputed property. 27 See the extensive collections in Rennes, Archives Départementales Ille-et-Vilaine, series 3H; for example, the Ruffiac Priory rent-roll for 1649–50 lists forty tenancies, many with multiple tenants (AD Ille-et-Vilaine 3H187 L1). 28 The classic statement is by C. Sánchez-Albornoz, ‘Pequeños propietarios libres en el reino Asturleonés. Su realidad histórica’, in Agricoltura e mondo rurale in occidente nell’alto medioevo, Settimane di studio del centro italiano di studi sull’alto medioevo, 13 (1966), pp. 183–222; id., ‘El régimen de la tierra en el reino astur-leonés hace mil años’, in id., Viejos y nuevos estudios sobre las instituciones medievales españolas, 3 vols (Madrid: Espasa-Calpe, 2nd edn, 1976–80),



may no longer be so sure of depopulation and repopulation but if we look at the voluminous charter evidence of the tenth century, it is still perfectly clear that there were many free peasant proprietors, whether we look at the east, centre or west. The charters record their small-scale property deals in all parts – twenty cherry trees and two apple trees, a quarter of a vineyard, a seventh of some water rights, a patch of water meadow, a plot of arable and so on – all within the home locality: properties lying beside others in an intensively worked landscape; properties of low value, bought for slices of bacon and small quantities of grain; proprietors who were involved in small-scale disputes – stealing a sheep, arguing over the boundary of a field. Although there are hundreds of examples, it is difficult to localize these peasant properties now. However, there are some pointers to size and shape that are consistent: firstly, there are many indications that a household’s property consisted of a number of scattered plots, rather than a discrete block; there is no equivalent of the Breton ran as a term for property unit.29 So, for example, one peasant couple’s property included: a plot of arable, lying beside someone else’s house, with an orchard at its head; more arable by the road; a vineyard on the road to the parents’ house; uncultivated land with apple trees around it; and an orchard beside the parents’ house.30 That of another couple included: a small house; several plots and half plots of arable; half a mill; water rights; half a meadow; half a bramble patch.31 Many had fractions of plots in the assemblage – a half, a quarter, a ninth – especially of gardens, orchards and meadows; and in the course of two or three generations families can be seen both splitting and sharing the inheritance, and family members can be seen alienating or acquiring portions. There seems to have been a constant process of small-scale shift in the shape of the unit of ownership.32 Sharing scarce resources was also common, especially in mills and water rights, the shares frequently expressed as a number of days per week (or some other period), often for a limited number of months: so, four sevenths of a saltpan every seven days, a third of a saltpan every 26 days, a mill for one in every eight days, the stretch of water that ran from the mill to the church.33 It is difficult to estimate absolute size or any kind of norm but it is not unreasonable to think of a house


30 31 32


vol. 3, pp. 1313–521. For a more recent assessment, J. Escalona Monge, ‘De “señores y campesinos” a “poderes feudales y comunidades”. Elementos para definir la articulación entre territorio y clases sociales en la alta edad media castellana’, in I. Álvarez Borge (ed.), Comunidades locales y poderes feudales en la edad media (Logroño: Universidad de la Rioja, 2001), pp. 115–55. For an effective analysis of peasant land-use in Galicia, see M. del C. Pallares Méndez, Ilduara, una aristócrata del siglo X (A Coruña: Do Castro, 1998), pp. 38–47, at pp. 31–41; E. Portela and M. C. Pallares, ‘La villa, por dentro. Testimonios Galaicos de los siglos X y XI’, Studia Histórica. Historia Medieval, 16 (1998), pp. 13–43. Cel380 (961). Lii388 (965). For much more detailed discussion, see W. Davies, Acts of Giving: Individual, Community, and Church in Tenth-Century Christian Spain (Oxford: Oxford University Press, 2007), pp. 17–19, 189–93. C175 (978), C195 (984), C102 (962), S338 (987).



plus six plots, of one kind or another, as a common standard. For Rabal, near the monastery of Celanova in Galicia, rather like Ruffiac in Brittany, so many records of transactions survive that it is possible to estimate size as a mean of 15 hectares per household property and 2 hectares per plot.34 In practice, of course, there would have been wide variation, poorer peasants with less, richer with more; smaller plots on very good land, larger on worse and so on. The strong implication of all the material is that peasants provided their own labour to work these plots. There are hardly any references to serfs working for peasants, although some peasant proprietors certainly took on tenancies in addition to what they owned; they must have used others to work them; we have no detail of how. Estates Despite the ubiquity of the peasant proprietor, there were plenty of estates too, and some very wealthy proprietors, lay and ecclesiastical. It is exceptionally difficult to estimate size and shape of these, not least because the term villa, which often means ‘estate’ or ‘farm’, can also mean a single settlement or a territory.35 Estimates of size are very limited: in southern Galicia there were some estates that can have been no more than 2 or 3 km across (because they are adjacent to each other), which would make them 500 to 900 hectares in area; some to the west and south west of Sahagún, in the very different landscape of the meseta, may have been up to 5 km across (2,000+ hectares), although here too there were many within 2–3 km of each other, like the villas in Valdefresno.36 Listings of stock may also give some sense of size: a group of four estates just south of León was given by the bishop to the monastery of San Juan de Vega in 950, and was provided with four mules, ten draught horses, fifteen yokes of oxen, forty horses and a stallion, ten donkeys, six foals, fifty-six cattle, two bulls and 720 sheep.37 The Villa Cova given

34 Davies, Acts of Giving, pp. 190–2. 35 M. del C. Carlé, ‘Gran propiedad y grandes propietarios’, Cuadernos de Historia de España, 57–58 (1973), pp. 1–222, attempts an estimate of sizes, but most of her examples are of the eleventh century. 36 Cel95 (950–1); cf. M. C. Pallares and E. Portela, ‘El lugar de los campesinos. De repobladores a repoblados’, in A. Rodríguez (ed.), El lugar del campesino: en torno a la obra de Reyna Pastor (València/Madrid: Universitat de València, CSIC, 2007), pp. 61–87, at p. 66. For the proximity of villas on the meseta, see the maps of Sahagún acquisitions in the mid-tenth century in J. M. Mínguez Fernández, El dominio del monasterio de Sahagún en el siglo X (Salamanca: Universidad de Salamanca, 1980), pp. 232, 235, 242; Valdefresno: S123 (c. 950), which has a detailed boundary statement, S225 (965), S262 (971). Cf. J. J. Sánchez Badiola, who estimates sizes of eleventh-century administrative territories, which contain many estates, as well as other forms of exploitation: in most cases less than 100 km2 but exceptionally more than 300 km2, El territorio de León en la edad media. Poblamiento, organización del espacio y estructura social (siglos ix-xiii), 2 vols (León: Universidad de León, 2004), vol. 1, pp. 549–51. 37 Li220.



to the church of the Virgin Mary in northern Galicia in 935, with three attached villas, only had ten oxen, fifteen cows, thirty sheep and twenty goats listed; while a single villa on the north coast, in Asturias, given to Sahagún in 980, had four oxen, fifty sheep and unnumbered pigs, hens and geese.38 There were obviously large, middling and small estates. Small could be very small: the endowment of a small monastery in Castile in 968 was three arable plots, three enclosures, a pasture, five vineyards, a terraced plot, two houses and two gardens – hardly more than the property of a rich peasant.39 The very wealthy often seem to have had large numbers of relatively small estates rather than very large single estates: a woman gave the church of Santa Cruz of Baroncelli (Verín) eleven estates in 985; Jimena, daughter of a high aristocrat, gave sixteen estates to Sahagún, also in 985 – some of these had been inherited but she had also bought several.40 In practice the wealthy often seem to have managed ‘peasant’ working units rather than develop integrated estates. This is strongly suggested by the fact that they dealt in fractions of groups of estates – a quarter of seventeen villas, given by one layman to another in 961; a half, a sixth and a tenth of villas given to Bishop Rosendo by a lay couple in 936.41 It is also suggested by the fact that lay and ecclesiastical landlords accumulated property through gifts and purchases from peasants – in all parts. These gifts were of small plots, rather than of whole farms – a field here, an orchard there, a small vineyard, rather different from the Breton pattern. The accumulations of the monastic officer Cresconio, for whose acquisitions we have some inventories, offer a good example: half a villa, a share in a villa, five separate vineyards, an inheritance in a villa, a share in a vineyard, an enclosure, a third of an enclosure, a quarter of a vineyard, a plot of arable and some apple trees were successively acquired in the late tenth century for his monastery of Celanova.42 Some, perhaps most, estates were a haphazard composite of peasant plots, parts of small farms. There is no hint of specialized production, beyond salt, and that came from small units of production. One can, however, make two kinds of case for different kinds of land management: where there was a significant element of dependent labour, as happened in a few cases, it must be likely that there was a higher degree of management in the landowner’s interest. Secondly, there was some tendency towards consolidation across the tenth century. This is well demonstrated by exchanges of property, which appear to have been made in order to rationalize exploitation; the transactions of the León monastery of Abellar are a good example.43 However, it should be noted that numbers were small – records of exchange constitute only 4% of total transactions. Consolidation is also suggested by the 38 39 40 41 42 43

Sob118; S308. C134. Cel462, S328 (four of these are half villas). Cel477, Cel33. Cel368 (975–1009). See further ch. 5 above, ‘Exchange charters in the kingdom of Asturias-León, 700–1000’.



few inventories that survive: with the exception of the several Cresconio inventories for the monastery of Celanova, these are all very brief, but they certainly indicate that a few landlords were listing, and presumably assessing, their resources.44 Labour Where the wealthy acquired property from peasants it is sometimes explicitly stated that the peasant donors and vendors continued to work the plots as free tenants, making a regular return to the new owner (similar to the Breton peasant donations of the 860s and later): the record of a gift of their inheritance in Bubal from a couple to Bishop Pelayo in 992 states that thereafter half of the wine and a quarter of the bread that the land produced was due (to the bishop) every year, and half the bread if the bishop should provide stock.45 At least some labour, then, came from free peasant tenants, who made an annual return of rent; their landlords lived off rent, without involving themselves in management. However, there are also references to slaves, serfs and tied labour of one kind or another. Some people were slaves: they could be sold and given, without attachment to a property, as were two children in northern Portugal in 995, in a gift from one lay couple to another; slaves were given to aristocratic women at marriage, often recorded as ten named boys and ten named girls, reflecting the requirements of Visigothic law; and at times slaves were freed, as the deacon Ermegildo provided for those who had been born free and those to whom he had restored freedom to retain their freedom (‘de servis meis et libertis, secundum suas habent ingenuitates hac restaurationes a me factas, ingenui permaneant’) when he bequeathed large estates on the meseta to the monastery of Abellar in 936; and as a hundred slaves on estates on the river Douro were to be freed when the very aristocratic Froilo, daughter of Gondesindo and Enderquina, died.46 Some of these (like Ermegildo’s house boys) were clearly household slaves but others equally clearly worked on the land, like those listed with the stock of an estate: the villa at Piasca given by a lay couple to the community of Piasca (on the edge of the Liébana) was given together with the mancipia Anastasius with his children and Hildosindus, seven horses and a stallion, ten cows and a bull, twenty sheep, twenty goats and thirty pigs: ‘offerimus et donamus . . . villam quam dicunt Piasca . . . sive et mancipia nominibus designata: Anastasium cum

44 OD22 (976), which is a list of charters relating to different acquisitions (orchards, patches of arable and so on) in a villa compiled for an aristocratic married couple; SM64 (952), which is a record of a gift to San Millán de la Cogolla by a layman of his share of an inheritance, listed by plots, houses and men, in different locations. Of course, the grants of groups of estates, like those cited above, imply some listing and record-keeping. Cresconio: Cel181 (1005), Cel192 (1001), Cel197 (975–1011), Cel203 (1007), Cel204 (1005), Cel205 (975–1011), Cel368 (975–1009) and embedded in Cel348 (1075). 45 Cel490. Cf. Cel446 (959), Lii371 (964). 46 PMH DC 174, Sob119 (887), Li109, PMH DC 12 (897).



filiis suis et Hildosindum; necnon etiam ex rebus nostris equas numero septem cum suo amissario’;47 three villas given to the church of St Christofor in León included the gift of three pueros who looked after a hundred cows and a hundred sheep; and when Zenserigo gave a quarter of his estates to Menendo González he also gave a quarter of his mancipias and mancipios.48 There is good evidence of slavery in northern and southern Galicia and Portugal, as is always emphasized in the literature, but also in the Liébana (north of the Cantabrian Mountains) and on the meseta, in ninth as well as tenth centuries. Where particular individuals are associated with specific estates, like the Piasca case above, they may well have dealt with the agricultural routine themselves. There were other kinds of tied labour, however. There are very considerable ambiguities over status and obligations in the texts, and there are many extremely vague references to workers associated with property, whose status has to remain uncertain – plenty of properties were alienated together with a workforce of residents. However, some references to tied labour are unambiguous. Since some of this tied labour is explicitly described as free, it seems appropriate to differentiate between free and unfree tied labour. For example, King Vermudo II in 996 ‘gave’ the monastery of Pardomino (on the slopes of the Cantabrian Mountains) ten named men and their families, on condition that they remain free (‘homines . . . vel qui de eorum progenie nati fuerint concedimus ad ipsum locum sanctum . . . ut ipsi homines . . . ingenui existant’).49 Further references suggest tighter and looser ties. In 917 some tenants on the meseta were quite free to leave the estate on which they worked, but if they did they had to leave half of their goods behind; but in Portugal in 991, in an agreement between a working couple and a female landowner, it was stated that the couple could not leave ‘without her permission, knowledge or order’, on pain of a fine of 5 solidi.50 For the most part we do not know what those labourers had to do beyond ‘serve’. ‘Serve’ can often mean to pay an annual render: in the 917 case above, the tenants paid 12 modii of barley, 12 lengths of linen and 6 ploughshares a year. So, tenants could be tightly tied or rather loosely tied, but even if tightly tied their obligation as tenants could simply be to return an annual rent, managing the agricultural routine themselves. Many tenants, however strongly tied, paid rent, but there are a few charters in which some labour element is indicated; it should be stressed that this occurs in a very small proportion of the available texts. One couple in mid-tenth-century Portugal gave themselves to a priest, to labour at will; thereafter they were not free to leave him.51 There is one Castilian and one Galician reference to men who look as if they were obliged to work saltpans – such were the fifteen homines kasatos 47 48 49 50

S39 (930). Li43 (917), Cel477 (961). Liii574. Li43, PMH DC 164. Cf. SM30 (943), whereby the tenants were free to leave but had to leave all their goods behind if they did. 51 PMH DC 70 (956).



of Poza in Castile and the omuntiolos who lived by the saltpans in Requeixo; there is a Portuguese reference to a servus whose hereditary duties were to live on an estate and look after boats at its port; and a Galician reference to men (mellarios) in four places who were committed to bee-keeping to provide wax candles – strictly this is a specialized render but it must have a labour implication.52 There are also two explicit references to what can only be labour service, although neither is precisely dated. A Celanova list of named descendants of the pistores who were established by Bishop Rosendo, several of which run to five generations, implies a hereditary commitment by those descendants to grind or otherwise prepare grain; in addition, explicitly, two of them are associated with looking after pigs, washing barrels and making baths for the brothers to bathe in.53 A Leire list of people’s obligations, from Aragón, is for the most part a list of what was due to the king of Navarre from each person on a royal estate, but there are two labour obligations included, one to harvest grapes and put wine in barrels and the other to plough, sow, cut and thresh, using their own oxen – all the stages of cereal planting and harvesting; here tenants paid dues together with a labour obligation to work at heavy times in the agricultural year.54 While the first five cases need imply no more than that the labour element of rent was organized by peasant initiative on peasant holdings, in the latter two cases some direct management by the landlord or his agent is implied; and the last could easily be read as an obligation to labour at heavy times on the landlord’s directly managed part of the estate. There are two general points to make about this labour service element: nearly all of these cases relate to royal estates, or estates with a former royal connection; and overall there are very few of them – less than ten of over 2,000 charters. The overwhelming message is that, for the most part, landlords were not involved in direct management of labour and they derived their income from the rent generated by peasant plots, plots that were held by free tenants, tied tenants and sometimes slaves. There was a very wide range of arrangements to profit from peasant labour, with perhaps some implication that those with the very tightest ties paid some of their dues in labour. While the wide range of arrangements is

52 C188 (981), Cel92 (968), PMH DC 139 (984), Cel5 (986) – the mellarios were given by the king to Celanova. Cf. an undated Galician reference to eight named karpentarios and their families in the control of one Froián, Cel146; and a list of those due to provide wine and of the quantities required, Cel240 (1004). 53 Cel158, dated to 942–77 by Andrade (reflecting Rosendo’s dates) but surely later, given the five generations; pistores could mean bakers (cf. references to De cocina), but grinding corn seems to make as much sense in this context, particularly given the number of them; dated to [977] in Colección diplomática del monasterio de Celanova (842–1230), ed. E. Sáez and C. Sáez, 3 vols (Alcalá: Universidad de Alcalá, 1995–2006), vol. 2, no. 183. Cf. SM88 (971), which cites a requirement to serve for two days a week, although this cannot entirely be a text of 971 since it includes a very anachronistic immunity. 54 Leire 12, dated to 991 by the editor, following Ubieto’s dating of the preceding charter, but this text could well be later; however, J. J. Larrea, La Navarre du IVe au XIIe siècle. Peuplement et société (Paris/Brussels: De Boeck, 1998), pp. 244–50, follows this dating.



quite clear, we have no idea of respective proportions; however, it looks as if the tied labour element was relatively small and more likely to be associated with royal resources. Other sources of aristocratic income There is a kind of donation which looks more than purely proprietary and which has implications for the income of some high aristocrats. Kings allocated control of some specified territories to lay aristocrats and to bishops ad imperandum, with a requirement that the inhabitants be ‘obedient’ to the person receiving the delegated authority: so, King Ordoño III gave a district to the bishop of León in these terms ‘ordinamus atque concedimus vobis, ad imperandum, comissum quod vocitant Valle de Ratero . . . ut omnis ipse populus, qui ibidem habitant vel ad habitandum venerint, ad vestram concurrant hordinationem pro nostris utilitatibus peragendis; et quicquid iniunctum vel ordinatum acceperint, omnia inescusabiliter adimpleant atque peragant’.55 These powers were usually assigned by territory (that is by comissos and mandationes), territories which included constituent properties of various kinds, which are sometimes listed, like the villas in the comisso of Santa María;56 however, in two tenth-century cases the subject is (the same) castellum with mandationes and in another two it is the men living within a specified distance of a monastery and in a villa.57 The number of these texts is not large (there are more in the eleventh century) and in the tenth century the beneficiaries are predominantly ecclesiastical; however, though relatively infrequent, such grants happened from early in the century and are recorded in different charter collections, in both cartulary copies and originals. The charters employ a distinctive diplomatic (some clearly drawn up by royal scribes), which is different from that employed in straightforward proprietary grants (including those of the latter which convey a workforce).58 They are clearly intended to convey a different kind of power. Since the donor of such territories was always the king in tenth-century texts, it looks as if a distinctively royal power was being alienated; and since the territories included estates and farms of varying ownership, these powers seem to have been more

55 Li257 (952); cf. ‘ordinamus adque concedimus vobis ad inperandum . . . castellum . . . qum mandationibus suis . . . et villar Petrunio cum omnes auitantes in ea vel qui ad avitandum venerint, ad vestram concurrant iussione’, Lii300 (951–56) (a charter on a single sheet), Cel207 (929), Cel206 (977). See C. Estepa Díez, ‘Formación y consolidación del feudalismo en Castilla y León’, in En torno al feudalismo hispánico. I Congreso de estudios medievales (Ávila: Fundación Sánchez-Albornoz, 1989), pp. 157–256, for seminal treatment of ad imperandum grants, especially pp. 161–2, 164–180; cf. Pallares and Portela, ‘El lugar de los campesinos’, p. 82. 56 Lii301 (952–56), referring to the mandationes of Li257. 57 Lii300 (951–56), Liii588 (999), Sam38 (937), S6 (904). 58 The diplomatic of obedience influences the texts of some grants of estates with workforce (e.g. Lii461 (978), Liii549 (991), S93 (944), S255 (970), all charters on single sheets); since these are all royal grants it is possible that the intention is to convey the same special power, but the point is arguable.



than the power of the king as landlord. What ‘ruling’ and ‘obedience’ mean is not specified in these charters, but occasionally a charter indicates that the holder of a mandatio or comisso should get the tribute due to the king, so this was a power that could generate income.59 We do not know if all land was subject to royal encumbrances of this distinctive type but it seems unlikely, given that such indications are rare and given that we lack references to general or consistent taxation.60 By the early eleventh century, holders of mandationes were clearly holding courts, and taking fines, as were their (lay) ancestors in the late tenth century.61 Passing references to lay as well as ecclesiastical holders of mandationes in the late tenth century reinforce this late tenth-century judicial dimension.62 Clearly the act of delegation by these tenth-century kings put new sources of income, of several kinds, into the hands of some high aristocrats, ecclesiastical and lay. There was, then, a specific royal power over some territories; it could be delegated to others (not in perpetuity, at least initially – there are several cases of the power being regranted in the second and later generations);63 in practice it could involve payment obligations by the population. It is different from the Breton power over plebes, in that the source of authority was very clearly royal, and recent, but it too could be seen as a kind of ‘extensive lordship’ or a kind of seigneurial power, sitting above proprietorship, although again it does not quite fit either of these models. Changes in the system There are several indications of change during the tenth century, of different types. Firstly, there was change in agricultural practice: new planting of vines and orchards, for example in the Liébana and on the meseta in the generation before the 970s and in Galicia in 970s and 980s. Transactions in vineyards increased 59 See Estepa Díez, ‘Formación y consolidación del feudalismo’, p. 176, for thoughts about what it could mean. For tributes (tributum, obsequium, regali debita): Sam38 (937), Lii300 (951–56), Cel54 (955); cf. Cel92 (968), Cel264 (949), Sob106 (958), Lii453 (978), Lii461 (978), although the latter were not ad imperandum grants. 60 Cf. Larrea, La Navarre, pp. 251–5, who, while stressing that the notion of fiscal obligation can be found across Christian Spain, and differentiating it from Francia c. 1000, also points out that the economic importance of any return was weak. 61 OD56 (1001), OD99 (1014) and OD116 (1019), for Pedro Flaínez; cf., in precisely the same area, the courts held by his father Flaino Muñoz, OD31 (991), OD34 (993), OD43 (997), OD44 (998); and those held (nearer to León) by a count from a different family, Munio Fernández, Liii556 (993), Liii561 (994), Liii578 (997). For Flaino Muñoz, see C. Estepa Díez, ‘Poder y propiedad feudales en el período astur: las mandaciones de los Flaínez en la montaña leonesa’, in Miscel·lània en homenatge al P. Agustí Altisent (Tarragona: Diputació de Tarragona, 1991), pp. 285–327, at pp. 288–300. 62 E.g. S332 (982–86), a gift from one layman to another because the latter had saved the former ‘de fisco et de mandacione et de iudice’; Liii560 (994). 63 E.g. Cel499 (942) and Cel54 (955), re Cel207 (929); Liii588 (999); and claimed by Li257 and Lii300.



two- or three-fold between the second and fourth quarters of the tenth century. Land was used more intensively: references to ‘untamed land’ (terrae indomitae) and to large tracts almost disappear in the second half of the tenth century, and references to cleared land (terra calva) come in, especially in Galicia.64 This trend is reflected in the growing body of archaeological environmental evidence of changes happening from the ninth century: the creation of terraces for cereal growing at Monte Gaias, just south of Santiago de Compostela in Galicia, for example, or the many storage pits for wheat associated with small settlements of this date in the Basque country.65 Secondly, there are many cases of peasant ‘intrusion’ into monastic land (usually into large tracts rather than plots of arable): they began to cultivate such land, planting and taking a crop; there were disputes, then court-cases, and the outcome was usually that the peasants remained and henceforth paid rent; thereby there seems to have been an increase in the cultivated area, essentially because of peasant activity and initiative.66 In one case it is explicitly stated that peasant lands were worked out (veteres erant et lavatas), representatives negotiating with an Aragonese monastery to have access to monastic land.67 Thirdly, both lay and ecclesiastical landlords acquired more properties in the tenth century, by receiving gifts and making purchases. Neither giving nor buying was new but both high lay aristocrats and the greater monasteries appear to have amassed property to a much greater degree than before: lay aristocrats acquired peasant properties in lieu of judicial fines (they were given land when fines were calculated in cattle or in solidi and those convicted could not pay) as well as receiving gifts in lieu of debts or for past ‘assistance’, like helping out with food in hard times.68 Receiving land instead of a fine is particularly noticeable from the last quarter of the tenth century. In a case of wounding for which the penalty was assessed at ten oxen, the penalty was first reduced to four oxen, and the court-holder finally accepted orchards in lieu.69 Even more evident, and evident from an earlier time (from the 930s), is the accumulation of property by monasteries. This occurred for similar and for different reasons: acquisition of smaller proprietary churches and monasteries, with their appurtenant lands; pious donation, of estates as well as peasant plots; acquisition of plots for the same reasons as the laity acquired them – debts, assistance, the occasional fine and so on; and straightforward purchase.70 The purchases make the intention of land 64 See Davies, Acts of Giving, pp. 6–11. 65 R. Blanco and P. Ballesteros, ‘Aldeas y espacios agrarios altomedievales en Galicia’ and J. A. Quirós, ‘Arqueología del campesinado altomedieval: las aldeas y las granjas del País Vasco’, both in J. A. Quirós Castillo (ed.), The Archaeology of Early Medieval Villages in Europe (Bilbao: Universidad del País Vasco, 2009), pp. 115–35, at pp. 130–4, and pp. 385–403, at pp. 394–5, 400. 66 Cf. Larrea, La Navarre, pp. 196–7, 210, 590. 67 SJP32 (s.d. but c. 1000). 68 See Davies, Acts of Giving, pp. 127–30, 146–54, 216–18. 69 Sob103 (952). 70 The classic examples are J. A. García de Cortázar, El dominio del monasterio de San Millán de la Cogolla (siglos X a XIII). Introducción a la historia rural de Castilla altomedieval (Salamanca:



management clear, as do the (rarer) exchanges – this was the beginning of management for profit. In the long term these accumulations of diverse properties had a bearing on control of labour but within the tenth century the tendencies were for labour on aristocratic properties to continue as before, with new owners, and for that on peasant plots also to continue as before, although with some surplus now being diverted through payment of rent.

Conclusions The act of comparison between the small zone of eastern Brittany and the much larger expanse of northern Iberia is interesting, although the areas are really very different from each other. Some things are clear: a case can be made for the existence of something like the classical bi-partite estate in both, but if it is persuasive it only applies to the eastern parts of eastern Brittany and it is but a tiny element of the proprietary landscape in each. Secondly, although some of the apparent differences are differences in the way the texts are written rather than in the practices described, there are nevertheless some real differences between them: quite apart from differences in size and terrain, the absence of mills is notable in Brittany, as well as of labour service, as is the fact that rich peasants owned serfs; all this might suggest both a steeper peasant social hierarchy and a less wealthy aristocracy. The ubiquity of mills in northern Iberia (despite the pistores) suggests production at a greater scale, and the small element of labour service on some royal estates could suggest areas of more integrated management. Picking up Professor Morimoto’s point about the importance of inventories, some movement towards deliberate management in northern Iberia is also suggested by the few inventories that we have,71 although it remains significant that neither in Brittany nor in Iberia do we have anything like the polyptyques of the Frankish heartland. There are also some striking similarities. In both areas the estate was overwhelmingly an accumulation of separate units worked and managed by peasant households paying rent: landlords took the income but had little involvement with the management of the farming regime. In both cases there were significant phases of acquisition from peasant proprietors by landlords, thereby reducing numbers Universidad de Salamanca, 1969), and Mínguez Fernández, El dominio del monasterio de Sahagún; but the significant build-up of estates by the bishopric of León, by Celanova and by Cardeña is also notable in the tenth century; for León, see C. Estepa Díez, Estructura social de la ciudad de León (siglos xi-xiii) (León: Centro de Estudios e Investigación ‘San Isidoro’, Archivo Histórico Diocesano, Caja de Ahorros y Monte de Piedad de León, 1977), pp. 197–203, and G. del Ser Quijano, ‘La renta feudal en la alta edad media. El ejemplo del Cabildo catedralicio de León en el período asturleonés’, Studia Historica. Historia Medieval, 4 (1986), pp. 59–75; for Celanova, although it covers a longer period, there is much in J.-M. Andrade Cernadas, El monacato benedictino y la sociedad de la Galicia medieval (siglos x al xiii) (A Coruña: Do Castro, 1997); for Cardeña, see S. Moreta Velayos, El monasterio de San Pedro de Cardeña. Historia de un dominio monástico castellano (902–1338) (Salamanca: Universidad de Salamanca, 1971). 71 See above, n. 44.



of free peasant proprietors and increasing numbers of tenants (though in Brittany whole farms were acquired whereas in northern Iberia it was usually single fields or orchards). There were therefore very significant accumulations of property by major monasteries, in east Brittany from the mid-ninth century and in northern Iberia from the mid-tenth, and by some lay aristocrats in northern Iberia by the late tenth; in the longer term this made for change in socio-economic structures. In both cases there were powers that could bring additional resources to some aristocrats, powers which could be described in terms of ‘extensive lordship’ or in terms of proto-seigneurial authority, although neither pattern fits precisely. However, the fact that they were additional to proprietary powers, not a substitute for them, is important and is an element to add to the complexity of the models of land management beyond the regions of the classic bi-partite estate. The systems of exploitation in both cases were in the end mixed; we do not have to suppose a single mode of exploitation in any region, as Jean-Pierre Devroey, Alexis Wilkin and Professor Morimoto have all stressed.72 There were free peasant proprietors, larger and smaller estates, free tenants and tied labour and additional superiorities.73 The character of the economic organization, and the strength of the ties of dependence, will have varied from district to district. The principle system of exploitation of wealthy resources was through free and tied peasant labour, based on the peasant working unit, but the mix of free and tied, proprietor and tenant, varied. And sometimes, particularly on royal estates in Iberia, there was some direction, and specialization, of rents due; and occasionally a hint of direct management.

72 J.-P. Devroey and A. Wilkin, ‘Diversité des formes domaniales en Europe Occidentale’ and Y. Morimoto, ‘Du “modèle évolutif du régime domanial” aux “vertus du comparatisme”. Quelques considérations introductives’, in Devroey and Wilkin (eds), Autour de Yoshiki Morimoto, pp. 249– 60, 261–72. 73 The mix is not unlike that suggested by Wickham, Framing the Early Middle Ages, pp. 540–1, for Francia and Italy at a slightly earlier period – some large estates and some areas of substantial groups of peasant proprietors – with rather less of landlord management of the estates in these cases.


12 THE MANAGEMENT OF LAND-USE IN OLD CASTILE The early strands of the Becerro Galicano of San Millán de la Cogolla co-authored with David Peterson1

Northern Iberia has left no early medieval polyptyques, which means that we do not have the kinds of written source about land-use and land management that Jean-Pierre Devroey has done so much to illuminate.2 Indeed, thinking about northern Iberia in the early middle ages is in any case problematic because there is an acute hiatus in available source material for the eighth and ninth centuries. That hiatus occurs in the first instance because of the Muslim conquest of the early eighth century and the creation of the Emirate of al-Andalus in mid-century, encompassing more than half of the peninsula.3 However, there is some written material that can shed light on land-use and even some that can throw light on land management before the development of the great monastic estates of the central middle ages.4

1 First published in A. Dierkens, N. Schroeder, A. Wilkin (eds), Penser la paysannerie médiévale, un défi impossible? (Paris: La Sorbonne, 2017), pp. 47–68. 2 For a magisterial synthesis, J.-P. Devroey, Puissants et misérables. Système social et monde paysan dans l’Europe des Francs (VIe–IXe siècles) (Brussels: Académie Royale de Belgique, 2006). 3 See H. Kennedy, Muslim Spain and Portugal: A Political History of al-Andalus (London: Routledge, 1996); E. Manzano Moreno, Conquistadores, emires y califas, Los Omeyas y la formación de al-Andalus (Barcelona: Crítica, 2006). 4 For the great estates, see for example: J. Gautier Dalché, ‘Le domaine du monastère de Santo Toribio de Liébana: formation, structure et modes d’exploitation’, Anuario de Estudios Medievales, 2 (1965), pp. 63–117; S. Moreta Velayos, El monasterio de San Pedro de Cardeña. Historia de un dominio monástico castellano (902–1338) (Salamanca: Universidad de Salamanca, 1971); M. del C. Pallares Méndez, El monasterio de Sobrado: un ejemplo del protagonismo monástico en la Galicia medieval (La Coruña: Diputación provincial de la Coruña, 1979); J. M. Mínguez Fernández, El dominio del monasterio de Sahagún en el siglo X (Salamanca: Universidad de Salamanca, 1980). Also n. 7 following.



The monastery of San Millán de la Cogolla lies in the Rioja in the hills on the south-western side of the Ebro valley. Its origins may well lie in the late sixth century but its growth into a powerful institution dates from the tenth century, following extension of political control of the Ebro valley by kings of Pamplona (Navarre) from round about 923, kings whose patronage of San Millán becomes clear from the 970s.5 Documentation of the monastery’s interests survives from the mid-tenth century and by the early eleventh century the foundation was receiving major gifts from aristocratic patrons.6 During the eleventh century it established a substantial portfolio of property, by gift and purchase, becoming one of the major landowners of the region, a process thoroughly analysed by Professor García de Cortázar.7 The Becerro Galicano is a late twelfth-century cartulary including records of these properties, arranged to meet the demands of eleventh- and twelfth-century political interests;8 it includes twelfth-century forgeries, designed to emphasize the patronage of the kings of Navarre and counts of Castile, especially in the early to mid-tenth century; but it also includes material from a lost early twelfth-century cartulary, the Becerro Gótico, into which were copied records acquired by San Millán in the eleventh century as it enveloped many smaller monasteries of the region, together with their properties and their archives.9 It is these earlier archives, and the information they reveal about the property interests of the monasteries of Old Castile, that form the subject of this paper.

The texts Early medieval records of property conveyance and acquisition are for the most part charters, of standard format. Throughout western Europe charter-writing was

5 See R. Collins, ‘The Spanish kingdoms’, in T. Reuter (ed.), The New Cambridge Medieval History, Volume III c.900–c.1024 (Cambridge: Cambridge University Press, 1999), pp. 670–91, at pp. 687– 91; A. Isla Frez, La alta edad media. Siglos VIII–XI (Madrid: Síntesis, 2002), pp. 61–74; J. J. Larrea, La Navarre du IVe au XIIe siècle (Paris/Brussels: De Boeck, 1998), pp. 213–26. Royal patronage: for example BG122 (972), BG323 (984), BG54 (992). BG is based on the edition of F. García Andreva, El Becerro Galicano de San Millán de la Cogolla. Edición y estudio (San Millán de la Cogolla: Cilengua, Fundación San Millán de la Cogolla, 2010); the older published edition of charters to 1076, SM, is organized by suggested date of each text rather than in cartulary order. 6 BG606 (933) has suspicious elements but is suggestive; BG532 (942), BG531 (948), BG358 (952), BG689 (959). 7 J. A. García de Cortázar y Ruiz de Aguirre, El dominio del monasterio de San Millán de la Cogolla (siglos X a XIII). Introducción a la historia rural de Castilla altomedieval (Salamanca: Universidad de Salamanca, 1969). 8 D. Peterson, ‘Reescribiendo el pasado. El Becerro Galicano como reconstrucción de la historia institucional de San Millán de la Cogolla’, Hispania, 69 (2009), pp. 653–82; D. Peterson, ‘Rebranding San Millán: The Becerro Galicano as a rejection of the monastery’s Navarrese heritage (1192–95)’, Journal of Medieval Iberian Studies, 5 (2013), pp. 184–203, DOI:10.1080/17546559.2012.762538. 9 D. Peterson, ‘El Becerro Gótico de San Millán. Reconstrucción de un cartulario perdido’, Studia Historica. Historia Medieval, 29 (2011), pp. 147–73.



heavily formulaic: whether a longer or a shorter text, a charter was largely composed of a great number of standard formulas, the texts only varying in the specifics of persons and of property conveyed. Across northern Iberia the same formulas were used over and over again in the ninth and tenth centuries, from central Portugal to Asturias and from Galicia to Cataluña, formulas which have analogues in the practice of many places beyond the Pyrenees. Although there are plenty of variants, the level of standardization overall is very striking.10 Remarkably, most of these very common formulas do not occur in the pre-eleventh-century charters of Becerro Galicano at all and those that do feature occur very rarely. So, for example, the exceptionally common disposition formula, which announces that a donor or vendor is making the transaction of his or her own free will, does not occur at all: placuit nobis atque conuenit nullius cogentis imperio nec suadentis articulo sed propria nobis accessit uoluntas ut . . .; nor does the formal handover formula, de meo iure sit abrasa et in tuo iure sit tradita. The almost invariable liberam habeatis potestatem only occurs four times and all of those occurrences are in fabricated charters.11 The common sanction clause que in iudicio/concilio uindicare non ualuero ([for which] I cannot provide a valid confirmation in court) does not occur at all; the very common disclaimer at the start of a sanction, quod fieri minime credimus (which we scarcely believe could happen), only occurs twice; and the sanction formula listing those prohibited from interfering in a transaction, an ego an quemlibet subrogata persona (whether I myself or any substitute), only occurs twice. The latter exceptions are in royal charters of the 990s and one 959 record of a gift to San Millán.12 By contrast, the extremely common meseta sanction Si quis contra hanc factum meum ad disrumpendum uenerit uel uenero is exceptionally common throughout the Becerro Galicano, occurring 260 times in the infinitive version, and looks as if it was added at the time of compilation of this cartulary;13 for example, the fabricated charter of 929 attributed to King García Sánchez has Si quis homo, ex meis propinquis aut extraneis, hoc meum factum disrumpere voluerit, fiat a Domino Deo maledictus et confusus.14 The absence of many standard formulas is both striking and remarkable, given their wide-spread occurrence elsewhere. If standard formulas are absent, what can we say about the pre-eleventh-century charters of this cartulary? While the forgeries certainly do have charter form, as do the credible royal grants of the late tenth century, with dispositions, dates, sanctions,

10 See W. Davies, Windows on Justice in Northern Iberia, 800–1000 (Abingdon: Routledge, 2016), ch. 4; and the standard manuals, A. Giry, Manuel de Diplomatique (Paris: Félix Alcan, 2nd edn, 1925); H. Bresslau, Handbuch der Urkundenlehre für Deutschland und Italie, 2 vols (Leipzig: Veit, 2nd edn, 1912–31). See also M. Milagros Cárcel Ortí (ed.), Vocabulaire international de la diplomatique (València: Universitat de València, 1994). 11 BG167, BG302/337, BG328, BG522 (‘You shall have free power’, i.e. freedom and full power). 12 BG54 (both), BG689, BG96. 13 Fernando García Andreva commented at the San Millán conference, October 2013, on changes made to sanctions in the course of writing the Becerro Galicano. 14 BG11.



witness lists and usually some initial protocol, many of the remaining pre-eleventhcentury texts of the Becerro Galicano cannot be called charters. They do not even approximate to charter form and lack its essential elements. Some, however, do and there is a small group of texts that we could reasonably call charters. For the most part they record gifts in favour of monasteries and churches absorbed by San Millán in the eleventh century, and they date from 867 – to Orbañanos, Acosta, Obarenes, Quijera, Hiniestra, Barticare, Herrán, Arce.15 The content of these unusual charters is intrinsically credible, partly because they have a very individual diplomatic but also because, with one exception, they come from a limited area lying to the north west of San Millán, from the Sierra de la Demanda to the valley of Tobalina, largely within the modern province of Burgos and the primitive county of Castile.16 This distribution appears to represent a coherent zone of scribal practice distinct from the standard tradition. Given these two attributes, it seems reasonable to propose that these charters derive from credible products of the ninth and tenth centuries. However, much of the pre-eleventh-century material does not have charter form. This material consists of notes of acquisitions made by the beneficiaries of a property transfer. These notes are inconsistent in form: occasionally they have an associated sanction or a date and sometimes a few witness names, but most do not have witness lists; they lack formal protocol; they have very few formulas and they are – in particular – short on expressions conveying the freedom to control the property acquired. So, for example, there are two sets of bare tenth- and early eleventh-century notes from Salcedo, the first set with dates and the second set without, neither having witnesses; or there is a list of individual sales to Hiniestra, and three gifts, with an attached sanction; or a set of three notes in favour of Santiago de Villapún-Maurdones, each note without witnesses, date or sanction.17 So, for example, the Hiniestra list of sales takes this form: Ego Nunnu Durgania dono pro me anime in molina de Aslanzone, ad Sancti Emiliani, una vice. Ego Gonzalvo Nunniz vendo in III solidos alia vice et III argenzos. Ego Argelo et meo filio vendimus in II solidos duas vices. Ego Alla vendo in XI argenteos una vice. Ego Feles Nunniz dono vice. Ego Didaco Albura vendo tres vices in IIII solidos. Ego Alvaro Zema dono tres vices. 15 BG424 (867), BG220 (871), BG396 (?873), BG505 (c. 913), BG378 (947), BG689 (959), BG467 (978), BG411 (988). Cf. García de Cortázar, El dominio del monasterio de San Millán de la Cogolla, p. 112, who counted eight monasteries with a pre–San Millán history – probably an underestimate. 16 The major exception is Acosta, BG220, in Álava. G. Martínez Díez, ‘El Monasterio de San Millán y sus monasterios filiales. Documentación emilianense y diplomas apócrifos’, Brocar, 21 (1998), pp. 7–53, at pp. 23–4, considered this charter spurious but his reasons are not persuasive. 17 BG523 (937–1001, 899/912–1004/1035); BG384 (943–51); BG544 (949). Where a text has no date cited, dates have been supplied through references to known individuals.



Ego Alvaro Ovecoz et Beila Didaz vendimus duas vices in solido et VI argenteos. Ego Blasco Gomiz et mea soror vendimus media vice in una novellam. Ego Munnio Vacoda vendo una vice in VI argenteos. Abite vendidit una vice in VI argenteos. Ego Monnio Ferrero et meo germano vendimus una vice in VI argenteos. Ego Beila Monnioz, I vice in VI argenteos. Ego Godemiro de Aslanzone, una vice in VI argenteos et duos solidos. Ego Endura, una vice in I solido. Ego Lili, I vice in VI argentos. Ego Monnio Nunniz, una vice in V aranzatas de cera, et tres vices sunt de Salitus abba. Ego Nunnu Durgania, cum filiis meis, qui sumus venditores atque, pro animabus, donatores firmamus perhenniter ad tibi, Salitus abba. Et sunt XXV vices. Nos, omnes supra scriptos, unicuique singillatim ad honorem Sancti Emiliani de Fenestra roboramus supra scripta hereditatem. Si quis homo ex nostris propinquis vel extraneis hanc nostram offertionem in aliquo disrumpere voluerit, sit a Deo omnipotentem maledictus, et a parte comitis exsolvat LXa solidos argenti; et ad regula parte, duplet hereditatem.18 While one of the Villapún-Maurdones notes has this form: In Dei nomine. Ego Nunu placuit mihi, pro salute anime mee, trado ad regula Sancti Iacobi me ipsum cum facultate mea, id est: uno orto et una ferragine in loco qui dicitur Rateziella, iuxta aqua qui currit ad Sancti Martini: de alia pars, penniella de eras super villa; et in illa tova de parras, mea tercia parte. Et in molino de Robuela, II feria, die et nocte, de VIIIo in VIII dies. Et una vinea in foio de Robuela, iuxta rivo maiore Flumenziello. Et uno agro latus Rivo Maiore. Et in molino de Biggas, III feria, die et nocte, de octo in octo dies. Et una faza deorsum Calzata de Mirone. Et agrum de Dolquiti, latus villa: de alia pars, Beila. Isti ab omni integritate serviant ad Sancti Iacobi.19 Whereas charters record transactions, these notes are different for they record acquisitions: the principal interest of the recorder was in the goods acquired rather than in the moment of acquisition. All of this suggests that it was the standard practice of monasteries in this area to keep records of acquisitions. There is nothing intrinsically incredible about the practice nor about the content of the notes, particularly since their subjects are often of very small scale. Take the example of six notes made in favour of the 18 BG384 (dated by reference to Abbot Salitus). 19 BG544.



monastery of Taranco, running from the early ninth to the early tenth century: they include a gift of an apple orchard and of a small time-share in a mill.20 The circumstantial detail is so particular, and the properties so small, it is extremely unlikely that such content could have been made up from nothing. There is of course no corroborative evidence, and there could well be errors here and there, but the notes offer credible texts. If monasteries were making notes of their acquisitions, in effect they were keeping registers – a practice similar in principle though not in scale to compiling a polyptyque. There seem to have been several such monasteries already in the ninth century – Taranco and San Miguel de Pedroso from early in the century, Salcedo, Oca, Acosta and perhaps Herrán and Hiniestra, as well as those which are not evidenced in text before the tenth century but which, like San Millán, may have pre-tenth-century origins; most of these places were keeping records until the early eleventh century.21 They lie in a coherent block, north west of San Millán, in the modern provinces of Burgos and Álava, and their number is significant. From a purely diplomatic point of view, it is evident that there was a further stage to record keeping before much of this material was copied into either Becerro: different monasteries gathered separate notes together into composite documents, recording the acquisition of a range of different properties, at different times. From their appearance in the Becerro Gótico this often seems to have been an early eleventh-century stage, as is in any case implied by the dates of the final items of some lists – the long list from Obarenes, ending with a date of 7 May 1009, for example, or the two Salcedo composite documents, ending with a date between 1004 and 1035, and the Hiniestra composite, which has entries to 1035.22 One of the Salcedo lists is rounded off with a confirmation clause, naming virtually all the donors mentioned, and a sanction; and the Obarenes list is rounded off with sanction, date and confirmation list. The three Villapún-Maurdones notes were put together and completed as a composite with date, sanction and list of witnesses.23 What the redactors seem to have been trying to do in constructing these composites was to create something that looked like a charter rather than a set of notes, using confirmation clauses which looked like witness lists. These monastic houses were therefore developing their own recording practice, using lists and notes to compile an overview of their holdings. As an aspect of Iberian recording and archiving in the early middle ages, this tells us that there was a lot more to record-making than creating charters. But it also has an economic significance because the purpose of making and keeping such lists and notes was

20 BG553 (807–912). We see no strong reason to classify these as fiction, as suggested by Martínez Díez, ‘El Monasterio de San Millán’, p. 21. 21 BG553, BG301 (759), BG523 (899–1035, 937–1001), BG355 (864) and BG361 (869), BG220 (871), BG356/468/548 (872); BG355/361 and BG356/468/548 have elements of corruption; for Hiniestra, see the following, although the 899 date of BG382 appears to be an error for 949. 22 BG421, BG523, BG382 (943–1035). 23 BG544 (949).



to know about each monastery’s property. The scale was of course much smaller than that of the spread of properties of the great monasteries of northern France and the Low Countries, but the kind of interest was comparable, as was the intention to manage and profit from the acquisitions. These records were far from being polyptyques but their construction into composite documents functioned as a kind of landlord survey.

Early monasteries and their estates It is quite clear that the early ninth-century monastery of Taranco, situated 100 km north-west of San Millán in the Valle de Mena, had landed possessions, just as other ninth-century monasteries did. It is also clear that the land was cultivated; references to planting, barns, mills and planted pasture imply that cereals and green fodder were grown.24 The properties specified in the ninth-century notes relating to Taranco also include apple orchards and a stand of enclosed woodland for pannage, presumably for pigs;25 and there is one unusually explicit reference to a time-share in a mill of one day and night per week, every Tuesday.26 The so-called ‘foundation charter’ of 800, which may well be a retrospective construction, adds horses, cattle, sheep, goats, pigs; some vineyards and wine presses; and pasture and meadows.27 All of this emphasizes a mixed but planned agricultural regime. Since the arable and orchards were in scattered locations, and since Taranco absorbed several churches together with their appurtenant properties during the ninth century, it is clear that the totality of this endowment was considerably more than could be worked by the eight persons of the foundation charter and must have produced more than they needed to sustain themselves.28 The Becerro Galicano includes a number of would-be ‘foundation charters’, like that of Taranco. Whether or not they actually reflect a single moment of foundation we cannot know, but it is perhaps unlikely; what they certainly offer is a constructed view of property holdings, presumably from a member of the community, at a significant stage in the monastery’s early history. An initial endowment has been plausibly reconstructed by Juan José Larrea for another monastery in this region, San Román de Tobillas, from a later charter with questionable elements: Abbot Avito’s supposed bequest in 822 shows the monastery’s interests stretching some 10 km to the north, with arable plots in a series of villages, woodland, sernas and also

24 BG552: sernas . . . culturas nostras . . . cellarios, orreos . . . plantavimus . . . fabricavimus molinos; BG553: terra . . . agro . . . defesas (enclosures) . . . ferragine . . . molino. 25 BG553: mazanares . . . (807, 912 and s.d.) defesa de glandiferos (s.d.). 26 BG553: cum ipsa vice in molino . . . die et nocte, die IIIa feria, de VIIIo in octo dies, per omnia ebdomada. 27 BG552. 28 A point made by García de Cortázar, El dominio del monasterio de San Millán de la Cogolla, p. 106.



some salt pans farther to the south east.29 Another suggestion comes from the less questionable ‘foundation charter’ of Orbañanos, which offers an unusually precise description of a ninth-century endowment; it is worth consideration in some detail.30 The village of Orbañanos sits on the south bank of the river Ebro, in the Merindades region of Burgos province, where the Ebro cuts canyons through the limestone of a landscape reminiscent of the Dordogne. Protected from both the harsher climate and the Arab raiding parties of more exposed areas to the south, these fertile, orographically complex valleys consistently offer the earliest testimonies of post711 communities in northern Iberia. In Orbañanos the valley floor has been submerged as a result of the modern damming of the Ebro in the nearby Sobrón gorge, flooding some of the most valuable land and with it – in all probability – some of the place-names preserved in the 867 text which forms the basis of this section. Nonetheless, modern cartography allows a remarkably high proportion of the early endowment of the monastery of San Juan, by the ‘founding’ abbot Guisandus, to be identified, revealing a coherent and compact estate in Orbañanos itself (at 600 m above sea-level), complemented by a second endowment on a similar scale in the more arid Obarenes (at 750 m), a now abandoned village 8 km to the south east. The topography of the Orbañanos estate is marked by the proximity of the Ebro and is reflected in the record: cannare, irrigation(?) channels; rivo, the river bank itself; saucto, a riverside wood, ‘soto’ in modern Spanish; Aqua Fierco; nave, a boat – presumably, in this context, metonymically a landing place. The properties gifted by Guisandus consist of one ager, five terrae, the channel mentioned above, a vegetable garden (hortus), a ferragine (field sown with green forage) and four vineyards. All lie within a radius of one kilometre from the church, except, rather incongruously, for a single field some 6 km due east in the orographically very aptly named Recuenco (bowl) valley. In what would appear to be a separate gift, Guisandus then added further properties (4 terrae, 1 ager and 1 vineyard) in Orbañanos and again a few kilometres east, this time also encroaching across the Ebro to Sobrón (Villa Semprun). To the south east in Obarenes there are four more terrae, and only one vineyard, but several references to apple-trees amidst a variety of other possessions: another ferragine (containing apple-trees), another hortus (also with apple-trees), an apple-orchard in its own right and a flax-field. Finally, a mill (with its own hortus) is also included, though the phrase ‘four turns, every eight days’ indicates it was shared, perhaps with the rest of the village community, many of whose lands were neighbouring properties of those gifted.31 29 J. J. Larrea Conde, ‘Construir iglesias, construir territorio: las dos fases altomedievales de San Román de Tobillas (Álava)’, in J. López Quiroga, A. M. Martínez Tejera, J. Morín de Pablos (eds), Monasteria et territoria. Elites, edilicia y territorio en el Mediterráneo medieval (siglos V–XI), BAR International Series, S1720 (Oxford, 2007), pp. 321–36. For sernas, see below. 30 BG424, classified by Martínez Díez, ‘El Monasterio de San Millán’, pp. 24–5, as suspect, for reasons which we do not find convincing. 31 ‘Et molino in Rivo de Ovarenes, IIIIor vices, de octo in octo dies, cum suo orto’, BG424. For time-sharing in mills, see further below. For gardens, see ch. 13 following, ‘Gardens and gardening in early medieval Spain and Portugal’.



The topography and geography of Hiniestra A particularly good example of a monastery with a range of economic interests is provided by Hiniestra. Hiniestra lies at 1,000 metres above sea level and is today a near deserted hamlet situated on the northern flank of the Montes de Oca, a relatively lightly inhabited but heavily wooded area of rolling hills straddling the watershed between the Duero and Ebro river valleys; it lies between the fertile Bureba basin to the north east and the alluvial slopes of the Arlanzón valley to the south west and is some 20 km east of the city of Burgos. The Montes de Oca is an agriculturally marginalized area, valued for grazing and timber, and was notoriously dangerous for pilgrims passing through on the route to Santiago.32 Another indicator of the marginality of these hills is the fact that Basque immigrants settled there around this time, as reflected in place-names such as Galarde and Zalduendo. What little we know of the monastery of Hiniestra, dedicated jointly to the saints John and (confusingly) Millán, results from its incorporation into the domain of San Millán de la Cogolla in 1052 – in what follows ‘San Millán’ will refer exclusively to San Millán de la Cogolla, while we will refer to San Juan and San Millán de Hiniestra simply as ‘Hiniestra’.33 According to San Millán’s Becerro Galicano, Hiniestra was founded shortly before 947 under the protection of the prestigious count Fernán González, although some of the Hiniestra texts imply an earlier history, such that we might think of the 947 event as more of a prestigious restart.34 By the early eleventh century Hiniestra had acquired numerous possessions in a number of rural communities through dozens of donations and purchases.35 Taken together the Hiniestra material, as yet unstudied as a block, must constitute one of the largest and earliest dossiers bearing on rural Castilian society in the tenth century. In what follows we will focus on three contrasting and complementary aspects of Hiniestra’s possessions: the monastery’s core estates in the Montes de Oca hills; the acquisition of milling rights on the Arlanzón river, 8 km to the south; and the acquisition of arable land and vineyards around the town of Briviesca, some 20 km to the north east (see figure 12.1). The monastery of Hiniestra itself should be associated with the microtoponym San Millán, situated a kilometre south of the village of Hiniestra, from which it is clearly distinguished in the charter BG377 (via qui discurrit ad Fenestra villa). In this initial record of endowment the monastery received a mixture of grazing rights, vineyards, orchards, meadows and mills, but the only element described in any detail was a rhomboid area of approximately 250 hectares around the monastery, termed its defensa.36 32 This was the justification for the founding of the hospice at San Juan de Ortega, just 2 km south, in the early twelfth century. 33 BG375. 34 BG377. For Gonzalo Martínez Díez this is one of only two authentic instruments of the count to be found in the cartulary, ‘El Monasterio de San Millán’, p. 33. See BG382 for pre-947 transactions. 35 BG378–BG384. 36 ‘cum introitus et exitus, terras, vineas, ortos, pomiferos, pratos, defesas et molinos’.



Figure 12.1 San Millán and principal sites discussed Source: courtesy of Google Maps

This is delimited with a fair degree of precision, the terms of reference being paths (represented by arrows on figure 12.2) heading towards a series of identifiable cardinal points: Milanes, circled; Hiniestra, circled; and Oca, a much more important centre 10 km due east and thus beyond the map.37 At the north-western extreme of the demarcated area the microtoponym La Dehesa occurs, seemingly marking the limit between the monastic defensa and the village of Barrios de Colina to the north. Similarly, the Alto de la Ermita may well have marked the limit between the village of Hiniestra and monastery’s core estate. 37 ‘ipsa defesa iuxta casa, latus via qui discurrit ad Milanes; de alia pars, latus via qui discurrit ad Auca; et de alia pars, latus via qui discurrit ad Fenestra villa et pergit ad Auca’.



Figure 12.2 Hiniestra’s core estate Source: based on 1:25,000 maps, Instituto Geográfico Nacional

The conditions set out for the use of this defensa refer exclusively to logging, with severe penalties of 5 solidi per tree felled illegally. Rather more problematically, a second dehesa with the same conditions applying is then delimited, seemingly some 3 km farther east but rather harder to locate.38 Regardless of the precise geography of this second area, what is clear is that the main resources being protected are the trees themselves; grazing rights, for example, go unmentioned. Elsewhere too the defensa was primarily an instrument for the management of woodland. The reference to glandiferos in the early Taranco text mentioned above suggests pannage, but aside from references in texts which are almost certainly later forgeries, the first mention of grazing rights in dehesas in

38 ‘Et illa alia defesa, id est: latus via qui discurrit ad Villam de Orovio; et de alia pars, latus via publica qui discurrit ad montem de Auca; et de alia pars, via qui exit de Coscorrita et pergit ad Montem Maiorem’. The road ‘towards’ Villorobe, some 8 km south, could refer to any of a series of possible north-south routes. Presumably the defensa’s southern limit is marked by the ‘public road’ but, since the whole area is that of the Montes de Oca, again it is unclear which. The third delimiter referring to a road that ‘leaves’ Cuzcurrita suggests this is not a reference to Cuzcurrita de Juarros, some 15 km south west, but another unidentified, much closer Cuzcurrita (that there is another Cuzcurrita in the Rioja Alta sugggests that such homonimity is possible). The only readily identifiable cardinal point is Monte Mayor, probably to be associated with an area still identified by that name 3 km east of Hiniestra, although it too is worryingly generic.



this material dates from 1006.39 Timber and firewood were undoubtedly important resources, and indeed, within the same Montes de Oca, we later see San Millán coming into conflict with neighbouring village communities over such matters.40 Beyond the dehesa, in the surrounding constellation of villages, though only in those to the north, sporadic acquisitions by the monastery are recorded during the century of documented independent existence, at the rate of one a decade, with a concentration in the early eleventh century (Hiniestra 6 times, Milanes 3 times, [Barrios de] Colina and Arraya) (see table 12.1). From such notices it is difficult to make out any systematic process of management or expansion, except perhaps a tendency to consolidate the monastery’s holdings with purchases on the southern edges of these villages and/or just outside (auditus) them, and on one occasion explicitly next to one of Hiniestra’s existing possessions (see table 12.1).41 The purchases (as also most of the donations) are all small scale, and generally of terrae or agri, with two mentions of apple-trees, and one turn in a mill. Table 12.1 Hiniestra’s acquisitions in the villages immediately to the north (BG382) Year

Previous owners




943 951 991 995 1003 1005 1013 1013 1015

Muñina & Ablazar Olio & Muñina Muño Oveco Nuño Álvaro & Oveco Pedro & Eilo Oveco Álvaro & Momadoña Guntroda, Oveco, Justo & Soña Alvaro & wife

purchase donation donation donation purchase purchase donation purchase donation

Milanes Hiniestra Hiniestra Milanes Colina (auditus) Hiniestra (auditus) Hiniestra (subtus) Hiniestra (subtus) Arraya


terra ager terra terra terra, 2 mazanos terra terra 2 terrae 2 vineae, 1 ager, 17 mazanos terra


vice de molino

1017 1035

Hiniestra (latus agro de S. Emiliani) Milanes

Hiniestra also received donations in another dozen locations which are either more scattered (Rábanos, Alarcia (Falariza)?, Cerezo) or impossible to identify now. More significantly, there are also occasions on which it is recorded that certain individuals made donations to Hiniestra of possessions in different villages of the Montes de Oca (Colina, Atapuerca and Hiniestra itself) and simultaneously of other holdings in or near Briviesca, some 20 km to the north east; they thereby mirrored Hiniestra’s own split holdings between these two non-contiguous and ecosystemically complementary areas. It is striking that some lay people had

39 See above, n. 25, for glandiferos. Likely forgeries: BG1 (929), BG11 (929), BG337 (945), BG340 (979). Grazing rights: cum hereditatibus . . . unum agrum ad defesam de herba de Iunkera, BG321 (1006). 40 BG330 (1142). 41 All examples are taken from BG382; the names of the protagonists are included for ease of reference within this lengthy text.



similarly distributed properties, and these appear to have been wealthier individuals making donations on a grander scale than the modest ones so far discussed. This pattern is repeated four times, and there is not a single case of a hybrid donation involving other locations (see table 12.2). Table 12.2 Donations to Hiniestra in both the Montes de Oca and Briviesca (BG382) Year 991 991

Previous owners

In Montes de Oca

In Briviesca

Fernando Núñez de Colina Gómez

terra in Colina

vinea ad illa Spina, in Birviesca in Birviesca, 3 romas vinea in Birviesca


Amuña & Gómez


Álvaro Díaz & Allo

meas casas . . . in Colina, terras, vineas terras et vinea in Atapuerca . . . Et alia terra in Audita de Atapuerca 3 sernas . . . in Valle Sancio, exiente de Fenestra

vinea in Comas

Hiniestra itself had very extensive interests around Briviesca, traditional capital of the agriculturally rich Bureba basin, situated 725 m above sea level.42 Hiniestra would acquire here four different privately owned churches, on each occasion by donation from individuals of no stated social rank: St Clement’s, St Tirso’s, St Sebastian’s and St Leocadia’s.43 There are references to other neighbouring monastic holdings in these brief donation texts, giving the impression of a densely populated and perhaps highly coveted area: one of Hiniestra’s terrae was next to another owned by Froncea, a monastery some 30 km to the south west on the Arlanzón river; another borders possessions of the brothers of Bovatella, an otherwise unidentified monastery.44 Both the count of Castile and the king (of Pamplona?) also had property here.45 There are at least a dozen further acquisitions, as many as in the immediate surroundings of the monastery (see table 12.3). Around Briviesca two things are striking. Firstly, the concentration of Hiniestra’s possessions suggests both very extensive and highly concentrated holdings: on three different occasions the newly acquired land bordered on a piece already owned by Hiniestra. Secondly, the majority of these acquisitions (and it is true of all the early ones) are from donations rather than purchases. In other words, it is not a case so much of Hiniestra buying into the area, as that the monastery seems to have had

42 It is difficult to gauge the relative importance of proto-urban centres from the exclusively monastic and generally rural early documentation, but Briviesca, a strategically well placed old Roman town which subsequently would re-emerge as the urban centre of the Bureba region, is a likely candidate for the dominant centre in the early medieval period too. 43 BG378 (947), BG379 (959), BG380 (997), BG383 (1013). 44 BG378, BG380. 45 BG382 (see n. 55 and table 12.3).



significant influence around Briviesca, despite the distance between the two and the fierce competition from other churches. We can also note the much higher proportion of references here to vineyards when compared to the Montes de Oca, which is in accordance with the ecosystemic contrast between the two areas already noted. Table 12.3 Hiniestra’s acquisitions in and around Briviesca (BG382) Year

Previous owners



943 947

Muño & Gutier Beila & Muño

donation donation


Muñina Paterne


10th C

Oveco de Briviesca








Sarrazina & Muñina Aznar

terra inter Cameno et Birviesca hereditate, terras, vineas, ortos, pomiferos, casas, cum introitus et exitus; et in molino de Tovas, VI vices agro de perale, et media vinea que est in Ecclesia Comase, et casa que est iuxta ferragine vinea et ortu, cum sua pomifera, et vice in molino de Tovas, in Birviesca terra in Cameno . . . duas vineas . . . vinea in Valle de Rota terra in Birviesca, in Valdezonio

1015 1015

Gonzalo & Gota María

donation donation


Muño abbate





terra apud Ecclesia Comase . . . vinea in Valle de Rota vinea in Valle de Rota Birviesca ad Sancti Tirsi una terra, latus terra de Munnio: de alia pars, de fratres de Fenestra terra in Besga, in Sancti Tirsi, latus terra de Monnio: de alia pars, fratres de Fenestra una vinea in territorio de Birviesca, in Fontanellas, iuxta limite de vinea de rex: de alia pars, vinea de Sancti Emiliani de Fenestra

According to the 947 ‘foundation text’, the abbot Salitus endowed Hiniestra with all his possessions, including very specifically – in an otherwise rather general list – the mills at Arlanzón.46 This is complemented in a separate undated text by a list of 18 transactions detailing Salitus’s acquisition for Hiniestra of turns in the Arlanzón mill (see table 12.4).47 In contrast to Hiniestra’s acquisitions in both Briviesca and the Montes de Oca, the mill list mainly comprises purchases,

46 BG377. 47 BG384; for the full text see above, pp. 206–7. The principle here is that different individuals owned a ‘turn’ in the shared resource of a mill, most commonly a day and a night, often once a week, but the frequency varies considerably.



though three donations are included as well as Salitus’s personal holding (and donation) of three turns. Two different units of account are mentioned (solidi and argenteos), and in at least two cases Salitus paid in kind (a quantity of beeswax and a calf). A turn in a mill tended to cost between 1 solidus and 6 argenteos, although those of Gonzalo Núñez and Gudmir of Arlanzón seem to have been significantly more valuable; the latter’s exceptional usage of the locative surname suggests a superior social status. Similarly, two of the four donations (those of Alvaro Zema and Salitus) were of multiple shares, again suggesting more prosperous members of the community. Another multiple share-holder, Diego Albura, is perhaps remembered in the name of the nearby hamlet of Villalbura. From their names just two of the sellers, Argelo and Alla, appear to have been women (though in the latter case it is unclear), in addition to the anonymous sister of Blasco Gómez. Table 12.4 Turns in the Arlanzón mill acquired by Hiniestra (BG384, c. 951) Original owner



Nuño Durgania Gonzalo Núñez Argelo & son Alla Feles Núñez Diego Albura Álvaro Zema Álvaro Ovecoz & Beila Díaz Blasco Gómez & sister Muño Vacoda Abite Muño Ferrero & brother Beila Muñoz Gudmir de Arlanzón Endura Lili Muño Núñez Salitus TOTAL

1 1 2 1 1 3 3 2 0.5 1 1 1 1 1 1 1 1 3 25.5

donation 3 solidi, 3 argenteos 2 solidi 11 argenteos donation 4 solidi donation 1 solidus, 6 argenteos 1 novella (calf) 6 argenteos 6 argenteos 6 argenteos 6 argenteos 6 argenteos, 2 solidi 1 solidus 6 argenteos 5 aranzatas de cera (measure of beeswax) donation 12 solidi, 53 argenteos

The list is laconic and contains no topographical information other than the reference to the mill and to Arlanzón itself, the name common to both village and river. An attempt has been made, presumably retrospectively, to turn the list into something resembling a charter, repeating the name of the first donor, Nuño Durgania (whose surname is unusual), and presenting the others as if they were his children. An even more laconic list from the same period insists on the interests of Salitus and Hiniestra in acquiring mill turns in Arlanzón, with 30 turns being bought from Nuño, Alvaro, Oveco and Gudmir of Arlanzón for 100 solidi in 951, 217


a significantly higher price, three times the rate paid in the undated tenth-century record cited earlier.48 Gudmir was clearly the same man as in the earlier text, in which his selling price was exceptionally high. If we similarly regard the Nuño of 951 as the same as Nuño Durgania, here Alvaro could well be Alvaro Zema, the donor of three turns in the earlier record. In other words, these may have been the more prosperous and prominent members of the community, selling larger shares (unless the price differential simply relates to a different and more valuable mill). At the same time, Hiniestra was also acquiring milling rights in Briviesca, at the unidentified Tovas mill. There are records of two separate donations, the second of which, from 947, outlines a steep penalty for appropriating water from the dam during the winter months.49 A couple of generations later, it is evident that Hiniestra’s interest in milling rights on the Arlanzón had not diminished for it purchased two thirds of the mill of Villabáscones (today Castañares), just outside Burgos, for 100 solidi, in 1017. The price was no greater than the earlier prices, but now the resource was concentrated in the hands of just three parties, where previously it had been shared much more evenly within a group: the sellers themselves, brother and sister Enneco and Totadonna, the monks of Froncea and a Jew called Citiello.50 Indeed, just two years previously Enneco and Totadonna had acquired a significant two-day time-share for 20 solidi.51

Concluding thoughts At least from the early ninth century some little-known monasteries were managing substantial properties, such that one has to ask whether they were producing a surplus, even in the ninth century. The detailed study of Hiniestra suggests that by the mid-tenth century monasteries of this kind could have been producing for distribution. In that case timber from the core estate was clearly highly valued; the 48 ‘Ego Nunnu, et Alvaro, et Oveco et Gudemiro de Aslanzone, cum alios nostros homines, vendimus XXX vices in molino sito in Aslanzone, in C solidos argenti, ad tibi, Salitus abbate. Era DCCCCLXXXVIIII, Ranimirus rex’, BG382; BG384 is the earlier record, for which see above. 49 ‘Ego Beila et Munnio presbiter tradimus nos medipsos ad ipsa regula, cum hereditate, terras, vineas, ortos, pomiferos, casas, cum introitus et exitus; et in molino de Tovas, VI vices, cum tale usu ut de quando aquas crescunt usque medio aprile, si quis aqua furaverit de illa presa, pactet pro die carnero, et pro nocte V solidos. Et cauto ad rex, quinque libras auri. Era DCCCCLXXXVa, Ranimirus rex’; BG382 (SM48). 50 ‘Ego Enneco fratre et germana mea Totadonna, sp[ont]ania voluntate, vendimus nostro molino de illo medio ad tibi, abbate Munnio de Fenestra, in Burgus, in flumen de Aslanzone, in illo que dicunt Molino de Bascones, cum suis aquis et molas, et illo directo aque molente, in C solidos argenti; et tercia tenent parte fratres de Faranducia et Citiello iudeo. Et molino introitu et exitu serviat in Sancti Emiliani. Si quis retemtaverit, ad parte comitis exsolvat CC solidos; et ad tibi abbate, illas duas partes de molino duplatas. Era MLVa’; BG382 (SM167). 51 ‘Ego Iohannis Didaz et uxor mea domna Tia vendimus ad tibi, domno Enneco fratre, et tua soror Totaduenna, in illo Molino de Bascones duos dies et duas noctes, in XX solidos; in Burgus, in rivo de Aslanzone. Era MLIIIa’; BG382 (SM162).



monastery also accumulated arable plots in different locations within a 2–3 km radius, as well as some vines and orchards. Also essential to its monastic economy were more distant properties 20 km away near Briviesca, where the interest in vineyards is notable, although not exclusive. Many other monasteries of this region had interests in vineyards too, interests which are evidenced from as early as the earliest records, with the rate of acquisition increasing across the tenth century. Hiniestra, like some larger monasteries, was establishing an interest in salt pans by the mid-tenth century;52 and acquiring mills also appears to have been deliberate policy from the mid-tenth century, with the monastery taking the significant initiative of buying multiple turns of mill time.53 It is very difficult to see this exceptional interest in mill rights as anything other than a deliberate attempt to develop capacity to process cereals on some scale.54 Despite the fragmented nature of the source material, what emerges is intensive use of resources and a high degree of spatial specialization, as illustrated by some repetitive but illuminating passages.55 This was also the case with monasteries that do not directly feature in the Becerro Galicano, like that of Froncea, which had possessions ranging from Burgos to Briviesca.56 With one significant exception, that of the Ebro itself, water was everywhere being harnessed for mills (Obarenes, Briviesca (Tovas), Milanes, Arlanzón, Burgos), and access to it regulated, protected, traded and shared. All of this raises critical questions about labour. The number of properties in a monastery’s portfolio, and their distance from the monastic centre, must mean that some of the essential agricultural labour came from beyond the monastery itself, although it does not exclude the possibility that monks themselves participated. It is also clear that some lay donors like Gómez, who gave Hiniestra arable and vineyards in two separate areas in 991, had multiple properties without being great magnates.57 The texts provide no explicit detail of working arrangements and offer very few clues. However, we might note that the gifts to San Millán in 952 by Diego Beilaz – a man who was a very wealthy owner with many scattered 52 BG382 (SM55 [949]); cf. BG531 (948) for competition between San Millán, Salcedo and Cardeña over salt rights. 53 García de Cortázar, El dominio del monasterio de San Millán de la Cogolla, pp. 73–4, noted that Hiniestra was the only monastery making purchases at this period. 54 Cf. the enormous investment of Irish monasteries in water mills in the eighth to tenth centuries, W. Davies, ‘Economic change in early medieval Ireland: the case for growth’, in L’Irlanda e gli irlandesi nell’alto medioevo, Settimane di Studio di Centro Italiano di Studi sull’alto medioevo, 57 (2010), pp. 111–33, and ch. 10 above, ‘Water mills and cattle standards’, and references there cited. 55 ‘una vinea in vineas de Amiugo, iuxta vinea de comite Fredinando Gondissalvez: de alia pars, vinea de Scemeno; de alia, vinea de Lopatone’ (BG382 (951)); ‘et vinea in vinearum de Birviesca, in Valle de Rota, latus vinea de fratres de Fenestra’ (BG380 (997)). 56 ‘et una terra in Valle de Sancti Genesi (in Birviesca), latus terra de Faranlucea’ (BG378 (947)); ‘nostro molino . . . in Burgus, in flumen de Aslanzone . . . et tercia tenent parte fratres de Faranducia’ (BG382 (1017)). For this monastery of Froncea there exists an as yet unpublished and virtually unknown fourteenth-century cartulary containing tenth-century material, Oviedo University Library, Signatura #456. 57 BG382; see above, table 12.2.



properties – included people called meos homines and at least three casatos.58 Casatos here must mean dependent workers; homines, since they are differentiated, were perhaps free tenants.59 Reference to casati is exceptionally rare in this material but BG505 (913) lists four casas and BG342 (971) names seven collazos. The donors of these three properties were evidently aristocrats and one might suppose dependent labour more likely on large aristocratic estates. (There are also references to sernas in this material, especially in association with the owners of multiple properties, but although some sernas had explicit arable use, there is no suggestion in this material that they were necessarily worked by dependent labour.60) By contrast, the many ninth- and tenth-century gifts of single parcels of arable (often bounded by those of other named individuals), as also of fruit trees and patches of vines, look as if they came from peasant proprietors; where these properties were distant from the monastery it must be likely that the donor continued to work the land given, making a regular return to the new owner, in effect becoming a (free) tenant in respect of the donation. Overall it is likely that in the ninth and tenth centuries there was a mixture of arrangements for working the land: there could have been agricultural workers who were members of the monastic household (although there is no evidence of this); there appear to have been some dependent (i.e. unfree) workers, living on or near the lands they worked, especially in association with large aristocratic estates; and there appear to have been free tenants as well as free proprietors. On balance, since there are so few references to dependents and so many small donations and some small sales, it looks as if free tenants dominated the monastic landscape. This area, like so much of northern Iberia pre-1000, seems to have shared the system of exploitation of wealthy resources through free and tied peasant labour, based on the peasant working unit, managed by peasant households paying rent, landlords having little involvement with the management of the farming regime.61

58 BG358; it is possible that the references in this text to a named individual cum suas casas, or casare de, imply the same. 59 Cf. BG382 (951): ‘Ego Nunnu, et Alvaro . . . cum alios nostros homines’. 60 For example, BG544 (949). There is a large literature on sernas, which were subject to seigneurial charges in later centuries and which have been interpreted as a forum for collective labour in a kind of ‘common field’ system; there is debate on whether or not these were characteristically large areas of arable worked by dependent labour for a lord. See I. Alfonso, ‘Las sernas en León y Castilla. Contribución al estudio de la relaciones socio-económicas en el marco del señorio medieval’, Moneda y Crédito, 129 (1974), pp. 153–210; E. Botella Pombo, La serna: ocupación, organización y explotación del espacio en la edad media (800–1250) (Santander: Tantín, 1988), who comments, p. 24, that until the mid-tenth century serna meant essentially the same as terra. 61 See ch. 11 above, ‘Free peasants and large landowners in the west’. As a result of work on this present paper, I would date the ‘significant accumulation of property by . . . monasteries . . . in northern Iberia’ earlier than I suggested in 2012.



In the end this is debatable and there can be no certainty about working arrangements at this date. What we can be sure of, however, is that long before the creation of the great monastic estates that are so well known, there were many monasteries managing the mixed agricultural regimes of their property portfolios, with an eye on specialized land-use for grape production and on controlling cereal production. By 950 many of these monasteries must have been producing for the market – which presumably explains why San Millán de la Cogolla was so interested in acquiring them across the following century.



The pools, fountains, pavilions and exotic plants of southern Spain may well be the image that comes to mind when the question of gardens and gardening in early medieval Iberia is raised.1 But gardens come in many forms and serve many purposes, from basic food provision to ornamental areas for relaxation. I offer here a first attempt at an overview of the range of gardens, garden plants and garden work to be found in early medieval Spain and Portugal, as also of the kinds of relevant source material available and of the principal problems that such a study poses. This paper deals with the whole of Iberia, with written and material sources and with the early Middle Ages up to about AD 1000, although in practice the focus will be on the ninth and tenth centuries. This is because there is a wide range of written sources from that time, from both north and south, and because the bulk of the archaeological evidence comes from the eighth century and later; however, there will sometimes be a look back to earlier centuries. In late antiquity Iberia was part of the west Roman empire; most of it fell within a single Visigothic state from the late fifth century until 711. Thereafter, following invasion by Arabs and Berbers, southern and central Iberia became the single Muslim emirate and then caliphate of al-Andalus, until replaced by the several taifa states of the eleventh century. Meanwhile, there were two principal kingdoms in the north and several autonomous counties in the north-east; Portugal was not a separate political entity at this time.2 Iberia is a very large peninsula and is geographically extremely diverse, with high flat plateaux, high mountain ranges, rolling hills, forest cover, coastal lowlands.

1 First published in W. Davies, Early Medieval Europe, 27:3 (2019), pp. 327–48. I am extremely grateful to Ann Christys, Hugh Kennedy and Eduardo Manzano for assistance with Arabic texts; to Juan Antonio Quirós for the results of recent archaeology; to Marios Costambeys and the EME editorial board for giving me this opportunity; and to the contributors to that volume for much advice. 2 For surveys of al-Andalus: H. Kennedy, Muslim Spain and Portugal: A Political History of al-Andalus (London: Routledge, 1996); E. Manzano Moreno, Conquistadores, emires y califas. Los Omeyas y la formación de al-Andalus (Barcelona: Crítica, 2006). For broad general surveys: R. Collins, Early Medieval Spain: Unity in Diversity, 400–1000 (Basingstoke: Macmillan, 2nd edn, 1995); A. Isla Frez, La alta edad media. Siglos VIII–XI (Madrid: Síntesis, 2002); E. Manzano Moreno, Historia de España vol. 2: Épocas medievales (Barcelona: Crítica, 2011).



It also has an extremely wide climatic range, from very hot and dry in the southeast to wet and green in the north-west; there are major variations in the level of annual rainfall and the south has a much higher mean average temperature than the centre or north. We should expect plenty of differences across the peninsula, then, in produce and in productive capacity. Dealing with early medieval gardens is not straightforward and there are some real problems to keep in mind. Firstly, the distribution of evidence is very uneven: the archaeological evidence is patchy and tends to focus on zones surrounding universities with active medieval archaeologists; written material from the south has a disproportionate focus on Córdoba and its surrounds; that from the north is voluminous for some parts, for example Catalonia and the region around León, but is virtually non-existent for others. Of the surviving material evidence, the pollen record, which one might have thought useful for these purposes, is unreliable, given the disparities in volume and range of pollen that different plants disperse, and given that it can be difficult to distinguish between the species of a genus, for example between types of cereal. And while the archaeobotanical record is immensely valuable, many species do not preserve well unless waterlogged. It is therefore difficult to make valid comparisons. Secondly, the character of Arabic texts can present problems because, in that highly literate world, any one known author may have drawn upon the works of other authors, both local and eastern, and may have done so silently as well as explicitly. Further, the transmission of texts is such that we may only know of one author’s work because of its subsequent circulation, with emendations, by generations of students or because of its citation by a later author. For example, the work of the Anonymous Andalusī Agronomist used fourth- to fifth-century Byzantine treatises in ninth-century Iraqi translations, and the received text of Ibn al-Qūṭīya’s History of the Conquest of Spain post-dates his death in 977. Likewise, the work of the most relevant historians of tenth-century al-Andalus is only preserved within the later eleventh-century works of Ibn Ḥayyān, for example Aḥmad al-Rāzī (d. 955) and his son ʿĪsā al-Rāzī (d. 989). The latter provides five years of detailed observation of activity in Caliph al-Ḥakam II’s court (AD 970–5), known by the term ‘Anales Palatinos’ coined by their modern translator.3 It can therefore be difficult to differentiate local contemporary observation from earlier borrowing or later accretion: that means a reference to sugar cane in a tenth-century Córdoban text might indicate that sugar cane was grown in or near

3 Kitāb fī Tartīb awqāt al-Girāsa wa-l-magrūsāt. Un tratado agrícola andalusí anónimo, ed. and trans. A. C. López y López (Granada: CSIC, Escuela de Estudios Árabes,1990), for which see further, M. Forcada, ‘The garden in Umayyad society in al-Andalus’, Early Medieval Europe, 27:3 (2019), pp. 349–73; D. James, Early Islamic Spain: The History of Ibn al-Qūṭīya (Abingdon: Routledge, 2009); Anales Palatinos del califa de Córdoba al-Hakam II, por ʽῙsā ibn Aḥmad al-Rāzī, trans. E. García Gómez (Madrid: Sociedad de estudios y publicaciones, 1967).



tenth-century Córdoba, or might arise because the author had read an Egyptian manual or other earlier source.4 Lastly, and most obviously, what is a ‘garden’? Contemporary experience in the English-speaking world might initially suggest the cultivated space around the domestic home, planted with grass, flowers and shrubs, and perhaps some vegetables – sometimes described as a ‘home garden’ or in North America a ‘yard’.5 But there are also allotments, primarily for growing vegetables, at a household scale; there are market gardens, on a larger scale, providing fruit and vegetables for distribution and sale; there are botanic gardens, devoted to rare and exotic plants; there are grand gardens, of much larger scale, with a wide range of plants and trees and maybe a ‘kitchen garden’ for vegetables, often associated with a great house or palace. In English these are all gardens; the one term covers all kinds, including both productive and ornamental space. They can be differentiated from cultivated fields by their smaller size; by their range of crops, as opposed to a single crop; by their higher labour requirement, because more processes are involved (for example, repeat sowing of quick-maturing vegetables); and by their need of a regular water supply. The garden represents a more intensive form of cultivation than that normally required by agriculture. In Arabic, by contrast, there are several words for ‘garden’, such as janna, bustān, rawḍa and ḥadīqa. There are also several words for ‘garden’ in Modern Spanish, primarily jardín, huerto and huerta (and, with different connotations, jardim, horto and horta, also pomar (orchard), in Portuguese), and they do not easily translate into English. Jardín is first and foremost a beautiful garden of special plants, flowers and running water; huerto is primarily a productive garden of fruit and/or vegetables, to provide food; huerta is primarily a productive garden on a larger scale, more like a market garden.6 However, in contemporary speech the semantic range of all three can and does overlap. Further, in Modern Spanish huerto (from Latin hortus) signifies both ‘garden’ and the separate English concept of ‘orchard’, a physical space devoted to fruit trees; huerto might therefore refer to either of these concepts or to a physical space with both cultivated plants and fruit trees. Given the derivation, this raises the question of what hortus, or more usually orto, might signify in medieval Iberian texts. In some regions of the north references in early medieval Latin texts (which are overwhelmingly charters) use ortos, but in others there are considerably more pomares (orchards) than

4 But for guidance, see Forcada, ‘The garden in Umayyad society’; and, for the extremes of nuance in usage, E. García Sánchez, ‘Cultivos y espacios agrícolas irrigados en Al-Andalus’, in L. Cara Barrionuevo and A. Malpica Cuello (eds), Agricultura y regadío en Al-Andalus. Síntesis y Problemas (Almería: Instituto de estudios almerienses, 1995), pp. 17–38. 5 See further, L. Peña-Chocarro and G. Pérez-Jordà, ‘Garden plants in medieval Iberia: the archaeobotanical evidence’, Early Medieval Europe, 27:3 (2019), pp. 373–93. 6 See S. Gutiérrez Lloret, ‘The case of Tudmīr: Archaeological evidence for the introduction of irrigation systems in al-Andalus’, Early Medieval Europe, 27:3 (2019), pp. 394–415, for some discussion of their distinctive character.



ortos. Do such distinctions signify differences in land use or simply differences in scribal habit?7 In what follows all kinds of garden are considered, with firstly an overview of the types to be found in al-Andalus, and then those found in the north. Plants and produce, in all parts for which we have evidence, come next; and then the physical and technical requirements of gardening, together with a brief look at labour issues.

Gardens in al-Andalus There can be no doubt about the existence of the ornamental garden in ninth- and tenth-century al-Andalus. There is a large bibliography on Islamic gardens: it is often said that these gardens belong to a long tradition, with origins in the paradise gardens of Ancient Persia, especially adapted to reflect the shade and water of Quranic texts.8 Tenth-century texts, such as the History of Ibn al-Qūṭīya, and many others too, refer in passing to the palace gardens in Córdoba and to the Garden Gate:9 the palace gardens were a standard point of reference, a known feature of the local environment. And tenth-century texts occasionally refer to the gardens of rural estates beyond the city too, the almunias, aristocratic estates that may have included space for ornamental gardens, productive gardens and also agricultural exploitations (although the extent to which they were productive areas at this early period is debatable).10 Al-Rāzī, for example, mentions the commentaries of earlier historians on the well-irrigated gardens of a beautiful munya (estate) given to the Caliph on 19 May 973.11 Indeed, there survives some of the fabric of Madīnat al-Zahrā, a city founded by Caliph Ἁbd al-Raḥman III in 936–41, which was destroyed in 1010–13. This is a whole city, on descending terraces of the hills outside Córdoba, with a large palace complex including gardens at the top level. Excavation has confirmed the layout and the tenth-century context.12 We can still

7 See further in what follows. 8 For example, J. Dickie, ‘The Hispano-Arab garden: Notes towards a typology’, in S. K. Jayyusi and M. Marín (eds), The Legacy of Muslim Spain (Leiden: Brill, 1992), pp. 1016–35; G. D. Anderson, The Islamic Villa in Early Medieval Iberia: Architecture and Court Culture in Umayyad Córdoba (Farnham: Ashgate, 2013), pp. 107–9. But see Forcada, ‘The garden in Umayyad society’, for the applicability of the ‘paradise’ term before the eleventh century. 9 James, Early Islamic Spain, pp. 113, 114, 124. 10 See F. López Cuevas, ‘La almunia cordobesa: entre las fuentes historiográficas y arqueológicas’, Revista Onoba, 1 (2013), pp. 243–60; J. F. Murillo Redondo, ‘Grandes residencias suburbanas en la Córdoba omeya: estado de la cuestión’, Revista al-Mulk, 12 (2014), pp. 85–108. These estates have been a focus for archaeological attention in recent years and substantial publications are expected. 11 Anales Palatinos, trans. García Gómez, pp. 136–8. 12 For example, J. E. Hernández Bermejo, ‘Aproximación al estudio de las especies botánicas originariamente existentes en los jardines de Madīnat al-Zahrā’, Cuadernos de Madīnat al-Zahrā, 1 (1987), pp. 60–79; A. Vallejo Triano, La ciudad califal de Madīnat al-Zahrā’. Arqueología de su



see the terraces, walkways, service quarters, reception rooms and foundations of the Oriental Salon, where al-Ḥakam II received delegations in splendid ceremonies above the gardens. The so-called Anales Palatinos provide pages of vivid detail of these ceremonies: the hierarchical order of processions, the status of the participants, the clothes they wore. They tell of the lawns and platforms, of pavilions, of buildings big enough for horses, of olive and fig trees. But they do not say much about other plants – just an occasional reference to the Rose Gate.13 There were certainly gardens, then – formal gardens, architectural in design – but they are the gardens of rulers and their officers; hardly common and hardly the norm for ordinary people. The exceptional nature of such gardens is reinforced by the observations of the tenth-century geographers who described the land of al-Andalus. Travellers would journey for days through pasture and arable land, past orchards and vineyards, but it is extremely rare to find reference to a garden. These rare mentions include a few by al-Rāzī (the Córdoba historian wrote a description of Iberia): in detailing the individual characteristics of many different places he notes gardens at Zaragoza and Ocsonoba; and flowers in the Beja and Cabra districts.14 Al-Muqaddasī, writing c. 985 (a foreign geographer who collected reports from Iberian informants), notes gardens at Cazalilla.15 Ibn Ḥawqal, a tenth-century Iraqi who travelled in the west, refers to cultivated gardens in Seville, as well as vineyards and figs.16 And the Chronicle of Ἁrīb ibn Sa‘īd, a physician and civil servant who died in 980, has gardens near Toledo.17 These are all isolated cases. By contrast, all of these writers constantly note trees, especially fruit trees, wherever they or their informants travelled. Ornamental gardens unquestionably existed in the south in the tenth century; but they were very much an expression of ruler power, and of aristocratic life style; and there were not that many of them. Moreover, it looks as if we should visualize them as places with trees, shade, water and buildings, rather than as an explosion of flowering plants. It is notable that, while the Anales Palatinos provide copious detail of the colours of clothes worn at ceremonies, they do not dwell on the form or colour of flowers. The Arabic words that designate ‘garden’ do not usually indicate whether a text refers to an ornamental or a productive garden. We can deduce that other

13 14 15 16 17

excavación (Córdoba: Almuzara, 2010), especially pp. 234–6, 267, 288; Anderson, The Islamic Villa, pp. 27–30, 111–13, 152–5. Anales Palatinos, trans. García Gómez, p. 252. E. Lévi-Provençal, ‘La “description de l’Espagne” d’Aḥmad al-Rāzī’, Al-Andalus, 18 (1953), pp. 51–108, at pp. 78, 91, 87, 65. Al-Muqaddasī, Description de l’Occident musulman au IVe–Xe siècle, ed. and trans. C. Pellat (Algiers: Carbonel, 1950), p. 35. Ibn Ḥawqal, Configuración del mundo, trans. M. J. Romani Suay (València: Anubar, 1971), p. 68. La Crónica de Ἁrīb sobre al-Andalus, trans. J. Castilla Brazales (Granada: Impredisur, 1992), p. 217.



kinds of garden – productive gardens, and/or productive areas within ornamental gardens – must have existed too, because there are references in the Calendar of Córdoba and in the work of the Anonymous Andalusī Agronomist to times for planting and times for harvesting different kinds of fruit and vegetable.18 Cordoban society was clearly concerned to grow food, whether in spaces that could have been modest productive gardens or in aristocratic almunias. Archaeology has also shown that some irrigation systems date from this early period and they must have been intended to support intensive production.19

Gardens in the north There is very little to suggest the paradise or pleasure garden in the north at this time: although there are occasional references to palatia in some texts, these appear to be storage sites rather than places of elegant living.20 That does not have to mean that there were none, but they are neither described nor implied. And of the many references to gardens in charters, relatively few are associated with kings or aristocrats – there are only six kings among the ninety-three actors transacting in gardens in León charters, and the gardens that are associated with them tend to occur in standard formulas. The closest, apparently, occurs in the record of a gift by King Ramiro III to the monastery of Sahagún in 977, which has gardens near the castrum regis, but this is a suspect charter and in any case its gardens are associated with mills and fields of forage on the river Torío – it looks like a working garden.21 The construction of ornamental gardens does not seem to have been a feature of northern land use at this time. There are, however, hundreds of references to gardens in the several thousand charters of ninth- and tenth-century northern Iberia. These references do not describe the gardens, for they simply refer to them, almost always in a formulaic, rather than a narrative, context. But despite the formulas there are, nevertheless, some clues about the nature of these gardens. Firstly, the references are of two principal kinds: those where the garden (orto) is itself the subject of a transaction; and those in which gardens are part of an appurtenance formula, along with fields, meadows, woods, water courses and so on. Hence, (1) ‘Ego domna Bona et filio meo Belasco tibi Munnio presbiter . . . nobis accessit uolumtas ut

18 On these texts, see further in what follows. 19 For full discussion, see S. Gutiérrez Lloret, ‘De Teodomiro a Tudmīr. Los primeros tiempos desde la arqueología (s. vii–ix)’, in De Mahoma a Carlomagno: Los primeros tiempos (siglos vii–ix). XXXIX Semana de Estudios Medievales, Estella, 17–20 de julio de 2012 (Pamplona: Gobierno de Navarra, 2013), pp. 229–83, at pp. 256–70. 20 J. Escalona and I. Martín Viso, ‘Los palatia, puntos de centralización de rentas en la meseta del Duero (siglos IX–XI)’, in A. Vigil-Escalera Guirado, G. Bianchi, J. A. Quirós (eds), Horrea, barns and silos: Storage and Incomes in Early Medieval Europe (Bilbao: Universidad del País Vasco, 2013), pp. 103–26. 21 S290.



uinderemus tibi nostra ratjone in illo nostro horto qui est iusta orto de Sanzone’; but (2) ‘Offerimus omnia nostra facultate quicquid augmentauimus uel augmentare potuerimus terris, uineis, ortis, pomeribus, aquis, molinis, aqueductis, fontibus, pratis, padulibus [sic], iumentis, uestimentis, auro et argento, domibus’.22 Appurtenance clauses should not, by the way, be dismissed as simply formulaic: they vary by region and that variation is a significant pointer to both contemporary observation and contemporary expectations. There are in addition to these two main types of reference, much more rarely, some references to a garden as a boundary point.23 The second clue as to the nature of the gardens is the marked regional difference in the occurrence of both main types of reference. Gardens as the subject of a transaction and as included in appurtenance formulas are both rare in Galicia and Portugal; for example, they only occur in 4.5% of pre-eleventh-century charters in the Celanova cartulary, from a major monastery in southern Galicia, and in 2.6% of the Sobrado cartularies, from a north-eastern monastery in Galicia.24 They are extremely rare in northern Iberia, that is in Cantabria and Asturias – they only occur in 1% of pre-eleventh-century charters from Santo Toribio in the Liébana and there are none in the San Vicente collection from Oviedo, although there are some in the Oviedo cathedral collections.25 But references are very common in charters from the central meseta in and around León (15%) and Sahagún (24%); as also in Catalonia (16% in the pre-eleventh-century charters of Pallars and Ribagorça or 30% in the ninth-century charters of Osona and Manresa).26 They are even more frequently cited in Castilian charters, for example in the collections of Cardeña (36%) and Valpuesta (48%), with similar proportions in other collections from Castile. It is here that the terminological problem demands attention. Galician charters do not often mention gardens but they do mention fruit trees constantly: ‘ut uenderem uobis hereditate mea propria . . . terras cultas uel incultas, arbores fructuosas et infructuosas’; not

22 V29 (950) ‘I, the lady Bona, and my son Belasco want to sell to you, the priest Munnio, our share in our garden, which is next to Sanzone’s garden’; S162 (959) ‘We offer all of our property, however extended in the past or to be extended in the future – arable lands, vineyards, gardens, orchards, streams, mills, water courses, springs, meadows, marshes, horses [perhaps draught animals], clothes, gold and silver, houses’. 23 For example, ‘per sepe de orto de Pepi et inde in carrale’, Cel176 (951) ‘along the boundary of Pepi’s garden and from there along the road’. 24 Cel, Sob. 25 T, Ov, OvC – 9.4% but that includes a number of falsa. 26 Catalunya Carolíngia volum III. Els comtats de Pallars i Ribagorça, ed. R. d’Abadal i de Vinyals, 2 vols (Barcelona: Institut d’estudis catalans, 1955); Catalunya Carolíngia volum IV. Els comtats d’Osona i Manresa, ed. R. Ordeig i Mata, 3 vols (Barcelona: Institut d’estudis catalans, 1999). The Catalan gardens, alongside other land use types, are mapped in the excellent series of early medieval Catalan atlases: for example, J. Bolòs and V. Hurtado, Atles del comtat d’Osona (798–993) (Barcelona: Rafael Dalmau, 2001), pp. 42–3; J. Bolòs and V. Hurtado, Atles dels comtats de Pallars i Ribagorça (v806–v998) (Barcelona: Rafael Dalmau, 2012), pp. 66–9.



infrequently they specify fruit types, as in ‘hereditate . . . [cum] linares . . . uineis pumares figares cesares’.27 Could it be that scribes used ortos (gardens) in some areas but pomares (orchards) in others for essentially the same kind of land use? It is also clear that ortos could have plenty of trees – for example, ‘hortale iusta uestro horto cum quinque mazanos et tres nocetos et uno perare’, or ‘orto meo obtimo cum suo pomifero bono’, or ‘orto clauso cum suos pomiferas pumares perales ceresales morales’.28 There is no sure way to determine the character of most of the gardens mentioned, but on balance orto and pomare appear to signify two different things in these charter texts. This is because, although orto and pomare/pumare are occasionally synonymous, much more often they occur together, as if the two words had different connotations, as in: ‘tam in uillis . . . quam in terris, pomeriis ac uineis, montibus seu siluis, pratis, pascuis, ortis et aqueductis’, or ‘et pratum cum suos ortos et pumares qui ibidem sunt plantati’.29 There is also a very strong association between ortos and water, implying irrigation and therefore some intensive production, which is not so much the case with pomares: ortos are on river banks, or associated with mills, or are at the end of a water channel, and they often lie beside the specialized production of flax fields (linares). A Catalan gift of houses, curtilage and a garden in 904 includes water a rigandum, for watering; one charter of 925 from the northern edge of León records the concession of a right to take water in the future from a channel previously given to a monastery, so that the original donors could continue to water their garden, flax and hemp. Another charter, of 978 from the neighbourhood of León, records a gift of access to draw water for two and a half days and nights a week, with the obligation to give some of the water to the poor on Sundays for the purpose of irrigation.30Orto therefore normally seems to have

27 Cel399 (961) ‘to sell you my own hereditary property [including] cultivated and uncultivated land, fruiting and non-fruiting trees’; PMH DC 166 (992), from Guimarães, ‘hereditary land with flax fields, vineyards, (apple) orchards, fig trees, cherry trees’. 28 V28 (950) ‘a garden beside your garden with five apple trees, three walnut trees and one pear tree’; S89 (943) ‘my excellent garden with its good apple-tree’; Lii486 (982) ‘an enclosed garden with its fruit trees – apples, pears, cherries, mulberries’. 29 Li109 (936) ‘as much in settlements as in arable, orchards and vineyards, mountains and woods, meadows, pasture, gardens and water channels’; Li236 (951) ‘and a meadow with its gardens and orchards planted there’. But contrast the synonymous C44 (944): ‘Trado uobis ipsa terra qui est iusta flumen Aslanzon . . . et ipsa ratione in pumare . . . Si quis . . . inquietare ausus fuerit, sit extraneus fiat a sancta ecclesia. Et ypsa terra et ipso orto dupplato’, ‘I give you some land by the river Arlanzón and a share in the orchard . . . If anyone should dare to disturb this arrangement, let him be excluded from the holy church. And [repay] that land and that garden twofold’. 30 Catalunya Carolíngia volum IV. Els comtats d’Osona i Manresa, ed. Ordeig i Mata, no. 58: ‘cum ipso caput aquis a rigandum’; Li67: ‘ut demus uobis de ipsa aqua pro uestro adtimplum regare, ortum, linum etiam et uestrum cannamum’; Lii454: ‘duos dies et medio cum suas noctes, in illa domada, ad inrigandum de illas aquas, et illo die dominico dent illa aqua pro rigare ad mesquinos’; cf. Cañ23 (912): water ad rigandos ortos uestros.



implied, and been associated with, intensive production, whereas a pomare did not usually have such associations. Lastly, there are a few further clues about what was distinctive about gardens. They were often explicitly enclosed, as one might expect from the term. Also, although towns were few and far between at this date, such as there were had a higher proportion of gardens than can be found in the countryside: in León, Burgos, Zamora, Coimbra, urban charters are more likely to mention gardens.31 Further, very occasionally gardens were explicitly associated with vegetables, although this is extremely rare: ‘iuxta olerum ortum’ (‘beside the vegetable garden’) in 959, ‘ortis et oleris’ (‘gardens and vegetables’) in 960 and again in 990, ‘uineis, ortis, oleris’ (‘vines, gardens, vegetables’) in 984, all in the neighbourhood of León although outside the town. There are also references to legumina (perhaps pulses) in another León charter, in Sobrado charters of 963 and 989, in a Celanova charter of 959 and in a Coimbra charter from Portugal.32 The cortinae, cortinales and cortiniares mentioned in some Galician charters may possibly refer to small vegetable plots.33 Recent excavation has added the existence of what might reasonably be called gardens, in some rural contexts, in association with small-scale residences. Firstly there is Gózquez, in central Spain south of Madrid, an earlier settlement of the sixth to eighth centuries but worth noting in this context.34 Here, in a settlement of ten to twelve households, there were stable ‘planting plots’ beside each set of structures. These plots are quite large, at 78 x 34 m, but are worth considering as gardens because of their very direct associations with residential structures. The excavator, Alfonso Vigil-Escalera, argued that cultivation must have been privately rather than centrally managed – at household level – because of the spacing of residential units and because cultivation areas were within easy reach of houses. Likewise, the storage of produce was in small (rather than large) silos (storage pits), which were associated with each structure. There are comparable indications at other sites in the same region, like the evidence of water capture – for irrigation – at El Pelícano.35 Secondly, at the later period of the eighth to tenth

31 For example, Burgos: C5 (913), beside another garden; Zamora: Cel427 (1010); Coimbra: Lor17 (957); cf. Caroline Goodson on Italian urban gardens, ‘Garden cities in early medieval Italy’, in R. Balzaretti, J. Barrow, P. Skinner (eds), Italy and Medieval Europe: Papers for Chris Wickham (Oxford: Oxford University Press, 2018), pp. 339–55 . 32 Lii317, Lii333, Liii539, Lii498; Lii406 (968), Sob45, Sob114, Cel446, PMH DC 185 (1001). 33 Cañ23 (912); Sob16 (950); L. Vazquez de Parga, ‘Los documentos sobre las presuras del obispo Odoario, de Lugo’, Hispania, 10 (1950), no. 5 (993). 34 A. Vigil-Escalera Guirado, M. Moreno-García, L. Peña-Chocarro, A. Morales Muñiz, L. Llorente Rodríguez, D. Sabato, M. Ucchesu, ‘Productive strategies and consumption patterns in the early medieval village of Gózquez (Madrid, Spain)’, Quaternary International, 346 (2014), pp. 7–19. 35 A. Vigil-Escalera, ‘Formas de parcelario en las aldeas altomedievales del sur de Madrid: una aproximación arqueológica preliminar’, in H. Kirchner (ed.), Por una arqueología agraria. Perspectivas de investigación sobre espacios de cultivo en las sociedades medievales hispánicas,



centuries, at Zaballa in the Basque Country, while the plots are not so clearly defined, the archaeobotanical evidence from the occupation layers of residential structures suggests intensive cultivation in the neighbourhood.36 While it is impossible to prove local cultivation, there is nothing to suggest participation in extensive exchange networks at this site at this period, and local cultivation remains the best explanation for the residue of food consumed. Archaeological evidence of this type is rare but it is important because it associates intensive production with what can only be peasant households both earlier and later in the early Middle Ages.

Plants and produce What was grown in these gardens, north and south? Whether grown in garden or field, the overwhelming weight of all types of evidence, everywhere, is of cereal cultivation: there are remains of wheat especially, both naked and hulled (Triticum aestivum/durum and, for example, Triticum dicoccum); hulled barley (Hordeum vulgare spp. vulgare); rye sometimes (Secale cereale); millet sometimes (both broomcorn millet (Panicum miliaceum) and foxtail millet (Setaria italica), although in referring to millet the texts only use one word, milio).37 Northern texts do not often specify cultivated produce but when they do they usually identify cereals. However, there are also references to vines in many places across the north, and to new plantations of vines in the tenth century; there are many references to flax fields; and there are very occasional references to pulses.38 Fruit trees feature in charters from all parts of the north, repeatedly – apples especially, pears, cherries, plums, figs, occasionally mulberry – with a wider range in Galicia and a much narrower range in the Liébana and Asturias (which lack the figs and cherries); and there are occasional references to nut crops like walnuts and chestnuts.39 Andalusī texts present some problems but the widespread occurrence of olives, vines, figs and fruit trees is clear and consistent, because mentioned frequently by geographers and travellers. Andrew Watson’s argument for a ‘green revolution’ introduced by Islam, published more than a generation ago, will be familiar to many: while the argument applies to all Islamic regions, it includes the introduction of many new exotic species to Spain from the east, as well as new cultivation

36 37 38 39

BAR International ser. 2062 (Oxford, 2010), pp. 1–9, at p. 5; cf. A. Vigil-Escalera Guirado and J. A. Quirós Castillo, ‘Un ensayo de interpretación del registro arqueológico’, in J. A. Quirós Castillo (ed.), El poblamiento rural de época visigoda en Hispania (Bilbao: Universidad del País Vasco, 2013), pp. 357–99, at pp. 382–6, for further examples from the same early period. J. A. Quirós Castillo (ed.), Arqueología del campesinado medieval: la aldea de Zaballa (Bilbao: Universidad del País Vasco, 2012); see further below on this evidence. Cf. tritico, centeno [rye], milio et ordio, Cel180 (pre-1010); triigo, centeno, milio, Cel390 (961). Eruelias: Cel571 (981); le[c]gumina: see above n. 32. See above, nn. 27, 28.



techniques.40 I do not intend to discuss this at any length because, despite Paolo Squatriti’s recent paper of support, the current consensus among Iberianists is that Watson overstated the range and volume of introductions and suggested too early a date for those that are credible.41 While some new species may have been introduced in the tenth century, the eleventh century and later often provide a better context. One of the texts on which Watson relied was the so-called Calendar of Córdoba, a version of a work written by Ἁrīb ibn Sa‘īd, who was born in Córdoba c. 900 and died round about 980, a provincial governor for Ἁbd al-Raḥman III and a court official for al-Ḥakam II. Ἁrīb’s work, Kitāb al-Anwā’, derives from the Arabic agronomist tradition and was probably written in 973.42 The Calendar, an attractive text, has long sections of astronomical calculations as well as a calendar of cultivation tasks, spread across the year, advising when to plant vegetables and when and how to store produce; it mentions a very wide range of plants – from rice, sugar cane and bananas, to carrots, lettuce and onions, to roses, violets and jasmine.43 The Calendar belongs to an established ethnoastronomical genre, of eastern origin, and the writer clearly drew on existing literature of eastern origin in constructing his work. That does not have to mean that it contains no local knowledge: aspects like the routine for harvesting olives or the time for buying horses could well derive from personal experience as governor; and products that recur often, in several varieties (like apple, pear, quince), are likely to reflect what was available, as Ángel López y López has argued.44 But it is unreasonable to suppose that all of the text was a description of what could be seen growing in and around Córdoba in the mid-tenth century. There are similar problems with the work of the Anonymous Andalusī Agronomist: Kitāb fī Tartīb (Book of Planting Times). This is neither precisely dated nor provenanced but appears to be of Andalusī origin of about the year 1000 and to use Ἁrīb’s work.45 This is more practical and less literary than the Calendar: for example, to plant pine, put three pine nuts on end in a hole in the ground, having soaked them for three days. But the very wide range of plants mentioned is very

40 A. M. Watson, ‘The Arab agricultural revolution and its diffusion, 700–1100’, Journal of Economic History, 34 (1974), pp. 8–35; A. M. Watson, Agricultural Innovation in the Early Islamic World: The Diffusion of Crops and Farming Techniques, 700–1100 (Cambridge: Cambridge University Press, 1983). 41 P. Squatriti, ‘Of seeds, seasons, and seas: Andrew Watson’s medieval agrarian revolution forty years later’, Journal of Economic History, 74 (2014), pp. 1205–20. 42 See further, Forcada, ‘The garden in Umayyad society’. 43 Le calendrier de Cordoue, ed. R. P. A. Dozy, trans. Ch. Pellat (Leiden: Brill, 1961), pp. 60, 74, 76, 132, 144, 158. 44 A. C. López y López, ‘Estudio particular de las especies botánicas que se citan en el Calendario de Córdoba de Ἁrīb ibn Saʽīd’, in E. García Sánchez (ed.), Ciencias de la naturaleza en al-Andalus. III. Textos y estudios (Granada: CSIC, 1994), pp. 43–78; A. C. López y López, ‘Vida y obra del famoso polígrafo cordobés del s. X Ἁrīb ibn Saʽīd’, in E. García Sánchez (ed.), Ciencias de la naturaleza en al-Andalus. I. Textos y estudios (Granada: CSIC, 1990), pp. 317–47. 45 Kitāb fī Tartīb awqāt al-Girāsa wa-l-magrūsāt, ed. and trans. López y López, pp. 18–20, 31–4.



similar to those of the Calendar, and the author also demonstrably used existing written sources, especially from the Greek geoponic tradition.46 There is another text which may allow us to view local practice more directly, because the horticultural references are incidental to the main purpose of the collection. This is Ibn al-Ἁṭṭār’s formulary, a legal text compiled by a learned Cordoban lawyer across the 980s and 990s as a set of models for recording many kinds of legal action, very similar in purpose to the models of the seventh- and eighth-century Frankish formularies familiar to western early medievalists.47 This large collection is very wide-ranging: it has material on slavery, inheritance, court procedure and legitimate evidence, for example, but it also includes practical things like contracts for hiring a person to irrigate a garden, or for crop-sharing, or for estimating damage caused by birds eating figs.48 In so doing it mentions many plants in passing, sometimes explicitly localized to Córdoba. As well as the inevitable cereals, there are fruits such as apples, olives, plums, pears, figs, grapes and mulberry; and there are also ‘cucumbers’, aubergines, pumpkins, gourds, beans, peas, chickpeas, radishes and carrots. The range is credible and much of it is reflected in the macro-botanical evidence – that is, the evidence of seeds, bark and fragments of wood.49 A securely dated archaeological sequence can offer localized evidence of food that was consumed nearby. At very early medieval Gózquez, from several contexts – pits, silos, occupation layers – apart from the large quantities of wheat, barley, rye and oats, there were tiny quantities of vetch, grass or red pea, figs, grapes and walnuts.50 And at eighth- and ninth-century Zaballa, in the occupation layers of structures, apart from large quantities of wheat, barley and millet (both types), there was some broad bean, vetch, lentil, grass or red pea and grapes.51 Note that the pulses are not the highly bred peas and beans we know now, but a substantial proportion came from Lathyrus (grass or red pea – compare the pod of the modern sweet pea flower) and Vicia (vetch) species, whose bitter seeds need soaking to remove toxins. Although vetch was widely consumed across Asia, the Near East and Europe from at least Neolithic times, it is rarely used for human consumption now, but Lathyrus sativus is still widely consumed in East Africa and Asia, as also in central Spain.52 At other

46 A. Dalby, Geoponika: Farm Work: A Modern Translation of the Roman and Byzantine Farming Handbook (Totnes: Prospect Books, 2011); R. Rodgers, ‘Κηποπoιία – Garden making and garden culture in the Geoponika’, in A. R. Littlewood, H. Maguire, J. Wolschke-Bulmahn (eds), Byzantine Garden Culture (Washington, DC: Dumbarton Oaks, 2002), pp. 159–75. 47 Ibn al-Ἁṭṭār, Formulario notarial y judicial andalusí del alfaquí y notario cordobés m. 399/1009, trans. P. Chalmeta and M. Marugán (Madrid: Fundación Matritense del Notariado, 2000). 48 Ibn al-Ἁṭṭār, Formulario notarial y judicial, nos. 28, 23, 159 (pp. 190–1, 166–73, 618). 49 However, no remnants of aubergine have so far been found, and the ‘cucumber’ is problematic; see Peña-Chocarro and Pérez-Jordà, ‘Garden plants in medieval Iberia’. 50 Vigil-Escalera Guirado et al., ‘Productive strategies’, pp. 15–17. 51 Quirós Castillo (ed.), Arqueología del campesinado medieval, pp. 459–69. 52 The seeds of grass and red pea are difficult to differentiate; the red pea was probably more often used for animal fodder than for human consumption; see Peña-Chocarro and Pérez-Jordà, ‘Garden plants in medieval Iberia’.



sites there was more fruit, like cherry seeds from tenth- or eleventh-century Santiago, with cherry stones on display in the new museum.53 Leonor Peña-Chocarro’s 2017 paper provides detail of the range of food plants discovered and their relative distribution in time and space for the whole of Iberia from the Roman period until the late Middle Ages.54 As she points out, an assemblage of macro-botanical remains may well include fruits that were gathered from the wild as well as cultivated; there is therefore an issue over how much of this recovered food material was garden produce. In the case of walnuts and chestnuts, however, the textual evidence is quite explicit that they were cultivated and not simply gathered.55 Although it is difficult to generalize from the relatively small number of sites so far investigated, and differential preservation may distort the picture of what was actually available, for the medieval period overall (eighth to thirteenth century) there is an unsurprising predominance of cereals and a very strong presence of fruit, and there are very small quantities of other plants, even up to the thirteenth century. For the Islamic period in the south and east of the peninsula, at the moment the macro-botanical evidence indicates the same range of fruit and vegetables as found on Christian sites but grape, fig and olive are more numerous than other fruits.56 The main difference indicated between ‘Islamic’ and ‘Christian’ sites, and between southern and northern Iberia, are the higher proportions of fruit than cereals in the totality of assemblages in the south and the higher proportions of cereals than all other remains in the totality of assemblages from ‘Christian’ sites.57 Natàlia Alonso’s paper comparing the macro-botanical material from eighth- to twelfth-century Christian and tenth- to eleventh-century Islamic sites in the Rioja provides some useful detail and a striking comparison: there are rather few vegetables from either Islamic or Christian sites, but there is a much wider range of crops, fruit especially, from the Islamic sites than from the Christian.58

53 Á. Rodríguez Resino, ‘Sistemas subterráneos de almacenamiento en la Galicia medieval. Una primera tipología y consideraciones para su estudio’, in Vigil-Escalera Guirado, Bianchi, Quirós (eds), Horrea, barns and silos, pp. 193–208, at p. 196. 54 Cf. L. Peña-Chocarro, N. Alonso, G. Pérez-Jordà, F. Antolín, A. Teira Brión, J. Pedro Tereso, E. M. Montes Moya, D. López Reyes, ‘Roman and medieval crops of the Iberian Peninsula: A first overview of seeds and fruits from archaeological sites’, Quaternary International, 499 (2019), pp. 49–66 (available online from 2017). 55 For example, walnuts: V28 (950), Lii378 (964); chestnuts: Cel446 (959), PMH DC 60 (950), and cf. Lévi-Provençal, ‘La “description de l’Espagne”’, p. 83. 56 ‘Christian’ in this paper denotes sites under Christian political control at the date from which samples were recovered and ‘Islamic’ those under Islamic political control. Hence Melque, in the centre, is ‘Medieval Islamic’ for ninth- to tenth-century layers but ‘Medieval Christian’ for tenthto thirteenth-century layers; see Peña-Chocarro et al., ‘Roman and medieval crops’, Table 1. 57 The exception is that there is a higher proportion of pulses from the north-west. 58 N. Alonso, F. Antolín, H. Kirchner, ‘Novelties and legacies in crops of the Islamic period in the Northeast Iberian Peninsula: The archaeobotanical evidence in Madîna Balagî, Madîna Lârida, and Madîna Turṭûša’, Quaternary International, 346 (2014), pp. 149–61, at p. 155.



Gardening Arabic texts like the Calendar of Córdoba and the Anonymous Andalusī Agronomist have the most direct information on cultivation techniques – the large and well-known corpus of important works by Andalusī agronomists comes a little later, from the eleventh century and beyond.59 There is much detail on when to plant and how to plant different species; on transplanting, taking cuttings and grafting; on harvesting and on turning produce into something storable. Crops, for instance, can be stored in salt or straw, and syrups can be made from fruit; walnuts are picked in September, while olives are harvested in October, and chestnuts in November.60 That kind of detail is suggestive: whatever the sources those texts used, the procedures are credible for practice in southern Iberia. The contracts of the Formulary reinforce this message with references to digging, irrigating and protecting from birds and other pests.61 The Formulary also has a range of different model contracts on the relative responsibilities of owner and worker: who was to provide the seed and in what proportions? how was the fruit to be shared out? And it includes a range of model assessments for compensation for loss of olives, grapes, figs, apples, pears, plums, leaves of pumpkins and of aubergines, whether occasioned by rain, flood, drought, heat, pests, grubs or locusts. Formula no. 29, which is explicitly about gardens, is especially interesting because not only does it specify the work the contractor has to do, but it clearly indicates that a garden may have fruit trees and arable land as well.62 Northern texts do not offer this kind of detail (except for indicating the planting of vines and fruit trees) but they do sometimes make reference to procedures which must have been comparable. There is a Barcelona sale charter that refers to tenants with complantatio agreements, with echoes of the co-planting contracts of the Andalusī Formulary.63 And one can make comparable points about water: some transactions for providing access to water have already been cited. There are also records of disputes over water channels that drained too much from the river, depriving some people of the resource, and agreements to keep channels clear of silt.64 Since, in more than one case, quite small quantities of water are

59 L. Bolens, Agrónomos andaluces de la edad media (Granada: Universidad de Granada, 1994) for the well-known eleventh-century agronomists Ibn Wāfid, Ibn Bassāl and Abū l-Jayr; cf. E. García and J. M. Carabaza, ‘Studies on the agronomy of Al-Andalus’, Revue des mondes musulmans et de la Méditerranée, 126 (November 2009) [online 31 July 2014] [accessed 28 September 2017]. 60 Le calendrier de Cordoue, ed. Dozy, pp. 144, 158, 172. 61 For the introduction of irrigation systems to al-Andalus, see Gutiérrez, ‘The case of Tudmīr’. 62 Ibn al-Ἁṭṭār, Formulario notarial y judicial, pp. 192–4. 63 Els pergamins de l’arxiu comtal de Barcelona de Ramon Borrell a Ramon Berenguer I: estudio i edicio, ed. G. Feliu i Montfort and J. M. Salrach, 3 vols (Barcelona: Fundació Noguera, 1999), vol. 1, no. 46 (1000). 64 For example, Li94 (932), Li128 (938), Li144(941); C89 (956).



specified – as much as will pass through a hole the size of a clenched fist – this is presumably small-scale domestic supply for a few households.65 Access to water was clearly critical, in both north and south, and the Arab geographers often detail whether water supply came from rain or was channelled: at Cazalilla it was pumped up from wells; for Jaén there were many streams; at Arjona there was rain water only; whereas for Coimbra the river regularly flooded and so there was very good land for cereals in the vega.66 Weather was also critical and Anales Palatinos, like other texts, is full of very precise comments on the weather in Córdoba: torrential rain on 28 March 972; three days of frost in April 973, causing much damage to vines and figs; so much rain on 8 March 975 that the river rose and people drowned; whereas on 3 September 974, after a long period of waiting, it finally rained and work in the fields could start.67 Weather could do plenty of damage but heat and rain were nevertheless essential. Less colourfully, weather damage could be captured by northern texts too, like the note of a Hiniestra transaction of 947 that includes penalties for allowing the stream to flood.68 The year 950 in Castile, anno pessimo, was so bad that peasant households faced starvation and handed over their lands to wealthier landowners in order to get food for their children, thereby reducing themselves to dependence.69 The cultivation procedure on which one might expect plenty of detail is that of keeping the land in good heart. In fact there is very little on manuring and fertilizing. Lucie Bolens, distinguished historian of the agronomist tradition, argued that there was little use of any fallow system or of animal manure in the south until the fourteenth century, although vegetal manuring was widespread before that; archaeologists currently insist that they do not find surface scatters suggesting manure spreading until the eleventh century.70 A charter from San Juan de la Peña, near the Pyrenees, records that local peasants asked for land from the monastery because their own land was old and worked out – they do not seem to have known about fertilizing.71 There are, however, a few relevant references. The Anonymous Andalusī Agronomist advises to plant fruit trees in big holes in which a little well-rotted manure is mixed with earth, as also that, citing the sixth-/seventh-century Byzantine agronomist Cassianus Bassus, bird manure, especially from pigeons, was the best (a comment which ultimately appears to have derived from classical Roman

65 C22 (932), C89 (956). 66 Al-Muqaddasī, Description de l’Occident musulman, pp. 35, 37; Lévi-Provençal, ‘La “description de l’Espagne”’, p. 89. 67 Anales Palatinos, trans. García Gómez, pp. 95, 129, 249–50, 226. 68 BG382/SM48. 69 V28. 70 Bolens, Agrónomos andaluces, p. 212; personal comment by J. A. Quirós Castillo. Gózquez is the exception, where it is possible that the planting plots were manured, Vigil-Escalera Guirado et al., ‘Productive strategies’, pp. 12–13; Vigil-Escalera Guirado and Quirós Castillo, ‘Un ensayo de interpretación del registro arqueológico’, p. 382. 71 SJP32; undated but apparently of c. 1000.



treatises on agriculture).72 The Formulary also has occasional references to cultivated and manured land, for example in the context of sharing cultivation tasks in a share-cropping contract, which indicates that the principle of manuring was clearly known.73 In the late eleventh century Ibn Bassāl reports a tenth-century story about the particular fertility of Évora, resulting from manuring and a distinctive ploughing technique.74 What may have been more common in practice than animal or bird manure was green manure: recent work on the deposits from the early medieval agricultural terraces outside Santiago de Compostela suggests green manuring (gorse in the sixth and seventh centuries, animal bedding in the eighth and ninth), although of course this relates to fields rather than gardens.75 Other vegetal matter may well have been used: there is a large ‘Islamic-period’ mixed fruit dump in Lisbon, which the excavators suggest could have derived from fruit collected for fertilizing.76 Northern texts include many references to fields of forage (ferragines), in other words to growing a green crop, often explicitly in association with references to gardens.77 They were perhaps as likely to be growing vetches for digging into the ground as for feeding to animals. Without engaging with Watson’s idea of the intensification of labour as a result of the green revolution, I will make a couple of simple points about work and workers. Notices of the disasters which drove peasant households to dependence provide one kind of clue about labour, be it in fields or gardens: many peasant households across northern Iberia were dependent for survival on the capacity of their own land to feed them, as their transactions with powerful people to provide support in sickness or old age demonstrate;78 they must have provided their own labour. The planting plots of Gózquez are the clearest demonstration of an association between cultivated area and residence; the household must for the most part have depended on its own labour to provide food.79 But there were clearly other

72 Kitāb fī Tartīb awqāt al-Girāsa wa-l-magrūsāt, trans. López y López, pp. 147, 174; R. Jones, ‘Why manure matters’ and D. Varisco, ‘Zibl and Zirā᾽a: Coming to terms with manure in Arab agriculture’, in R. Jones (ed.), Manure Matters: Historical, Archaeological, and Ethnographic Perspectives (Farnham: Ashgate, 2012), pp. 1–11, at p. 7, and pp. 129–43, at p. 134. 73 For example, Ibn al-Ἁṭṭār, Formulario notarial y judicial, no. 27, p. 183. 74 M. Forcada, ‘Ṣāʽid al-Bagdādī y los antecedentes de la agronomía andalusí’, Al-Qanṭara, 16 (1995), pp. 163–71. 75 C. Ferro Vázquez, S. González Prieto, A. Martínez Cortizas, F. Criado Boado, ‘Deciphering the evolution of agrarian technologies during the last ~1600 years using the isotopic finger-print (δ13c, δ15n) of a polycyclic terraced soil’, Estudos do Quaternário, 12 (2015), pp. 39–53, at p. 48. 76 J. Bugalhão and P. Queiroz, ‘Testemunhos do consumo de frutos no período islâmico em Lisboa’, in Al-Ândalus.Espaço de mudança (Mertola: Campo arqueológico de Mértola, 2006), pp. 195–212. 77 For example, BG424 (867), BG544 (949), C79 (950), V21 (950), Li252 (952), C192 (984), S355 (997) – in this case city properties with gardens and associated forage fields; cf. Peña-Chocarro and Pérez-Jordà, ‘Garden plants in medieval Iberia’, on growing lucerne in the south. 78 For further discussion, see W. Davies, Acts of Giving: Individual, Community, and Church in Tenth-Century Christian Spain (Oxford: Oxford University Press, 2007), ch. 6. 79 Richer peasants in eastern Brittany in the ninth century had dependent serfs to assist and it is not impossible that this sometimes occurred in Iberia too; W. Davies, Small Worlds: The Village Community in Early Medieval Brittany (London: Duckworth, 1988), pp. 87–9, 95–9, 103.



kinds of labour too. There are two references to gardeners in northern charters, one (ortulanus) working for a monastery on the northern edge of León, and the other (ortolano) witnessing a transaction just outside the city.80 They suggest that urban garden plots (corte cum orto) and suburban plots could be worked by hired labour. Elsewhere, especially in rural contexts, there is some limited evidence of dependent labour providing for religious institutions and this may have had a consequence for garden work.81 In al-Andalus, while presumably there were independent and dependent households providing their own labour too, a variety of urban arrangements for cultivation is indicated by the Formulary: contracts for cultivation could involve the contractor him- or herself doing the work for a term of years. However in no. 29, there are also references to the contractor sub-contracting seasonal workers, and, as in no. 20, sub-contracting a team of jornaleros, day labourers. There was clearly a wide range of arrangements for gardening, with more complex organization in urban contexts.

Conclusion There were different kinds of garden in early medieval Iberia, from the architectural gardens of Andalusī rulers and officials, with their water features and pavilions, to the gardens growing plants from which textiles were made, to urban gardens in a town house complex (in both north and south), to peasant plots in the countryside. Fine gardens were closely associated with an elite, were southern rather than northern and at this date seem to have been characterized more by water and buildings than by exotic plants. Productive gardens could be found all over the peninsula, however, north as well as south. Vegetables were clearly grown in them, with the emphasis on pulses. At the moment, the weight of evidence, everywhere, would suggest that vegetables constituted a small proportion of produce, with cereals overwhelmingly the major element. Peasants may well have been growing cereals in the plots beside their houses, as is implied by the many silos sited beside houses in the north, while the model formula no. 29 from al-Andalus is explicit that in the south a garden might include cereals as well as fruit and other cultivated plants. Indeed, where northern charters specify the constituents of a peasant household’s property, house and arable often lie side by side and other kinds of land use, like vineyard, garden or orchard, lie in scattered plots at a distance from the residence.82 Beyond cereals, fruits seem to have been more important than vegetables as a source of vitamins, sugar and mineral salts. Every kind of evidence, north and south,

80 Lii450 (977), Li156 (942). 81 See ch. 11 above, ‘Free peasants and large landowners in the West’, pp. 195–8. 82 For example, Lii435 (974): house plus arable, share of garden, cherry tree, three more arable plots, shares of water meadow and of vineyard. See further, Davies, Acts of Giving, pp. 189–93.



points to the economic, dietary and cultural significance of growing fruit: there is a repeated refrain in the early tenth-century campaign accounts of Ἁrīb’s Chronicle that the army went forward and destroyed the enemy’s fruit trees, in order to target an essential in the food supply;83 the traveller Ibn Ḥawqal commented that medium-quality fruit was accessible to everyone, and not too expensive, in tenth-century al-Andalus.84 References to gardens in northern charters clearly indicate intensive cultivation. How common it was for a peasant household to have a garden plot is unknowable, but it looks as if there was regional variation across the peninsula – some did, some did not. If they did have a garden, whether it was to supplement their diet beyond cereals is arguable, although present evidence, in indicating that garden plots might be situated at some distance from the residence, near a water source, suggests cultivation of something more specialized than cereals. It is too soon to be certain what that specialization might have been – we await, in particular, more archaeological and archaeobotanical evidence – but the question is important, and the physical relationship between residence and garden suggests that our twenty-first-century model of house surrounded by garden may not be appropriate for early medieval Iberia. For the future, we have to hope for some solution to the problems of underrepresentation and varied contexts of paleobotanical remains. The fact that we can recover any seeds, fruits and bark of food plants is exceptionally important and it offers great potential for future understanding. But preservation conditions vary considerably and they affect the survival of archaeobotanical material: it is possible that the low percentage of vegetables in the archaeobotanical record is a matter of under-representation of vegetable material, conditioned by size of seeds or differential survival rates or preservation conditions. The absence of aubergine traces in the south, despite credible references in the texts, is an extreme case; future work will probably change this picture. Further, the contexts in which archaeobotanical remains have been found – whether they be occupation layers, planting plots, rubbish dumps, cess pits, ditch fills, harbours or stream sediments – need even closer interrogation in order that the relationship between find spot and cultivation spot be better understood. It is easy to argue that these are the remains of food consumed nearby, but was that food grown locally or far away, arriving by some form of exchange? If local, how much was cultivated and how much gathered from the wild? Again, future work will define more patterns. While the ornamental gardens of al-Andalus have benefitted from decades of distinguished research, there is more work to be done with the northern charters. Looking at the social status of those involved in garden transactions suggests peasant participation of the order of a third; an argument for peasant enterprise

83 La Crónica de ‘Arīb sobre al-Andalus, trans. Castilla Brazales, for example pp. 126, 131, 163, 172, 175, 196, 210, 217. 84 Configuración del mundo, trans. Romani Suay, p. 66.



presents itself but it needs probing and testing. Moreover, many of the beneficiaries of garden transactions were monasteries. Looking at the patterns of garden acquisition by monasteries should tell us more about monastic economic strategies, for it is quite clear that some monastic houses were deliberately acquiring gardens in the course of the tenth century. Was this in order to feed the growing monastic community or to develop commercial enterprise (or both)? I hope these questions will be actively pursued and that this Iberian evidence will be compared with that from other parts of Europe. It could yet be the case that an increase in gardening contributed to increased productivity and to economic upturn.


Part III

Judicial practice

14 SETTLING DISPUTES IN E A R LY M E D I E VA L S PA I N A N D P O RT U G A L A contrast with Wales and Brittany? Let us take two late tenth-century Spanish cases.1 In the summer of 991 a man called Ermegildo attacked and wounded another called Padre.2 The wounds were assessed by a local law officer, the saio Frogia, at a value of ten (silver) solidi.3 Ermegildo appears to have denied this (or refused to pay), because he was taken before the judge Braolio by Padre’s father. Having sworn on oath in court that he had not made those wounds, Ermegildo was required to provide a trustworthy supporter, a fidelis, to undergo the hot water ordeal (to test the veracity of the denial).4 Both arrived at the court on the agreed day; once the fidelis had submitted his hand to the hot water, Ermegildo admitted the truth, that he had indeed caused wounds of the assessed value. As a result his parents, Fredino and Leobina, made a land grant of two arable plots and half of nine orchards, in Valdoré, on the southern slopes of the Cantabrian Mountains, on his behalf. Although some aspects of the matters recorded are unclear, the principle of graded compensation for wounding is explicit and unambiguous, as is the family responsibility. 1 First published in R. Griffiths and P. Schofield (eds), Wales and the Welsh in the Middle Ages (Cardiff: University of Wales Press, 2011), pp. 89–107. The paper began as the Jones Pierce Lecture for 2008, delivered in Aberystwyth. T. Jones Pierce, ‘The laws of Wales: The kindred and the bloodfeud’, in J. Beverley Smith (ed.), Medieval Welsh Society: Selected Essays by T. Jones Pierce (Cardiff: University of Wales Press, 1972), pp. 289–308, at p. 298 (first published in the University of Birmingham Historical Journal, 1952), is especially relevant to what follows. I am grateful to the Aberystwyth community, especially Natalie Roberts and Phillipp Schofield, for much help; to the audience for some extremely pertinent questions; and to Phillipp, Isabel Alfonso, Thomas Charles-Edwards, John Hudson and Susan Reynolds for their valuable comments on subsequent written drafts. 2 OD31. 3 The solidus was the key unit of account of the reformed Carolingian silver currency, and silver denarii were coined at the rate of 12 to the solidus in the Carolingian Empire; neither was coined in northern Spain. For saiones, see what follows. 4 One might argue that the supporter took a supporting oath rather than that he went to the ordeal, because the term fidelis is unusual in this context and the language of the text is ambiguous: ‘Et abut Ermegildo plazum a dare fiidelem, pro manum mitere ad kalida, et gonplibit a die plazum, gum suo fiidelem. Et de post fiidelem datum pro manum mitere ad innocentiam et gonobiset Ermegildo in veritatem, de post gonditionet iuratam, que fecit Ermegildo ipsa plagam in x solidos adprecitum’. However, given what happened in comparable texts, my summary makes sense.



The following year, in November 992, another charter records that Fredino and Leobina again made payments of fields and orchards in Valdoré, this time because another son, Argemiro, had gone to the house of a woman called Licinia at night and violently raped her – all the worse because Licinia was a relation in the third degree (a second cousin).5 This time they made the payments to the local count, Flaino Muñoz, who presided at courts in this mountain region (though there is no reference to a court in this text) and who was probably also the recipient of the first grant, since he is greeted at the start of that record.6 So, in the kingdom of León in the 990s – by then encompassing all of north-west Spain – there were judicial procedures, law officers, compensation systems, fines, offences of wounding and rape, family responsibility and at least one very troublesome family. In what follows I want to look at some of the recorded mechanisms for dealing with disputes in northern Iberia in the tenth century, dealing with them in practice rather than in theory, and think about how they compare with Welsh and with Breton procedures of that time.7 My focus is on the tenth century, since that is the first well documented century in Spain and Portugal and since it is a period before some changes in practice.8 By ‘dispute’ I mean, following Brown and Górecki, ‘the phase of conflict which is articulated as a claim, between two or more parties, concerning some specific subject matter’.9 Considering the procedures of these far western countries together has a particular value since, as we shall see, there are more similarities than one might expect; that being the case, they should have a bearing on our approach to western European practice as a whole. Welsh contemporary evidence for settlement procedure in the early middle ages is very limited, although there is one extremely important mid-ninth-century text in the Lichfield Gospels, recording settlement of a property dispute in west Wales by local elders; and there are relevant texts in the Llandaff charters.10 Although 5 OD33. ‘in tercio grado’: second cousin according to the family tree diagram of the tenth-century Madrid, Real Academia de la Historia, Codex 25, fo. 145v. 6 See the following for compensation to victims. I use the word ‘court’ in this paper for meetings that dealt with explicit judicial business. 7 Beverley Smith’s paper on Welsh judgment is especially pertinent to the matters discussed here, ‘Judgement under the Law of Wales’, Studia Celtica, 39 (2005), pp. 63–103. 8 Cf. the comments of A. García Leal that Count Fruela Muñoz’s acquisition of property as a result of judicial decisions decreased (in the early eleventh century) because fines were paid in metal rather than in land: ‘Los condes Fruela Muñoz y Pedro Flaínez: la formación de un patrimonio señorial’, Anuario de Estudios Medievales, 36:1 (2006), pp. 1–110, at p. 33; and P. Martínez Sopena, ‘La justicia en la época asturleonesa: entre el Liber y los mediadores sociales’, in A. Rodríguez (ed.), El Lugar del Campesino. En Torno a la Obra de Reyna Pastor (Madrid/València: CSIC, Universitat de València, 2007), pp. 239–60, at p. 258. 9 W. C. Brown and P. Górecki, ‘What conflict means: The making of medieval conflict studies in the United States, 1970–2000’, in W. C. Brown and P. Górecki (eds), Conflict in Medieval Europe: Changing Perspectives on Society and Culture (Aldershot: Ashgate, 2003), pp. 1–35, at p. 1. 10 For the most accessible edition and the most sustained recent comment, see D. Jenkins and M. E. Owen, ‘The Welsh marginalia in the Lichfield Gospels’, part 1, Cambridge Medieval Celtic Studies, 5 (1983), pp. 37–66, at pp. 50–2, and part 2, ibid., 7 (1984), pp. 91–120; the date of this text is arguable, but I follow Jenkins and Owen. Llandaff charters: LL.



some of the narrations of the Llandaff charters are certainly later creations, others are not and they provide important incidental cameos of procedure, especially in later ninth-, tenth- and eleventh-century contexts.11 The Breton contemporary evidence, again in charters, focusses on the ninth century, since there are few tenth- and early eleventh-century charters and not many post-ninth-century court cases; although limited in time-span, the ninth-century material is very rich and detailed, allowing an insight into procedure both at aristocratic and at peasant, village, level – unusual for European material of that period.12 The Iberian material comes from northern Spain and northern Portugal (at that time part of Galicia), since it is the north that produces the texts (southern Iberia was Muslim-controlled). The Iberian texts begin in the ninth century but become very numerous in the tenth century and thereafter. The character of the evidence is similar to the Breton, in that it largely comes from credible charters, but it is much more varied in range and there is vastly more of it.13 There are more than 2,000 tenth-century charters (excluding Catalan material) and in the order of 250 recorded dispute cases in that century;14 comparative figures for ninth-century Brittany are 345 charters and 49 dispute cases.

Occasions of criminal dispute The cases cited above hung on the offences of rape and wounding, offences we would call criminal in modern contexts. I am using the term ‘criminal’ for such cases, since parties other than those directly involved clearly had an interest in their resolution; it might be argued that the term is anachronistic, but they were certainly not purely private matters.15 There are other kinds of criminal offence

11 See W. Davies, The Llandaff Charters (Aberystwyth: National Library of Wales, 1979), for a guide to suspect narrations. I don’t share the scepticism of K. L. Maund, Ireland, Wales, and England in the Eleventh Century (Woodbridge: Boydell, 1991), pp. 183–206, because, although she makes many valuable points, she fails to take the diplomatic character of the charters into proper account. And now P. Sims-Williams, The Book of Llandaf as a Historical Source (Martlesham: Boydell & Brewer, 2019). 12 CR; see also the facsimile edition: Cartulaire de l’abbaye Saint-Sauveur de Redon (Rennes, 1998), and the fundamental comments therein of H. Guillotel, ‘Le manuscrit’ and ‘Répertoire chronologique’, pp. 9–25, 71–8. There are some eleventh-century court records (e.g. CR290 (1089), CR309 (1081–2), CR345 (1084)), but they are procedurally very different from the ninth-century cases. 13 Although the narrative of the disputes detailed in the charters might sometimes be fictitious, created to ‘produce a written judgment of lawful possession’ (P. J. Geary, ‘Land, language and memory in Europe 700–1100’, Transactions of the Royal Historical Society, 6th ser., 9 (1999), pp. 169–84, at pp. 170–1), these texts remain a valuable guide to procedural principles and are full of useful circumstantial detail. 14 For updated numbers, see W. Davies, Windows on Justice in Northern Iberia, 800–1000 (Abingdon: Routledge, 2016). 15 See John Hudson’s useful discussion of the term ‘crime’, in an English context, in ‘Kings and crime: Ideology and practice in the tenth and twelfth centuries’, in Ph. Chassaigne and J.-P. Genet (eds), Droit et société en France et en Grande-Bretagne (XIIe–XXe siècles). Fonctions, usages et



in the Iberian material: breaking and entering, for example, as in the case of the couple Reparado and Trasvinda who, being childless, designated two cousins as their heirs; once the document of conveyance had been drawn up, the cousins attacked them, ejected them from their house, tied them up and held them by force for two days and a night, in 936.16 There is also arson: a man called Lupi and his friends surrounded a local priest’s house in 963, on the meseta south of León, laid siege and set fire to it, doing damage to the value of 60 solidi.17 There are plenty of examples of straightforward homicide – another Padre terrorized the neighbourhood of Valdávida, in the hills bordering the valley of the river Cea, in the 940s, along with his two sons, four nephews and a nephew’s son, killing and ‘doing much evil’.18 This disruptive family was so bad it was handled by the king of León himself, who confiscated its property and exiled the evildoers, but there are plenty of simpler cases which were handled by more local courts: when a man called Pelayo battered a young retainer to death in 940, the damage was assessed locally at a value of five oxen.19 There is theft too. In Spanish texts there are recurrent cases of peasants thieving stock, especially sheep, which were stolen and eaten – a kind of poaching; the unfortunate couple Gonzalo and his wife Elo stole and ate two sheep ‘furtively’ and were taken before a judge at the court of Munio Fernández, on the meseta, and made a penalty payment of two patches of vines in Toral, in 993.20 Some such disputes were settled out of court, by negotiation and compensatory land grants, but plenty went to court and were settled by the court. Iberian practice does not seem to be significantly different from what happened in early medieval Wales or Brittany, but the evidence of court activity in Iberia is much stronger. For Wales there are indications in the Llandaff narratives and elsewhere of offences that might be called criminal – theft and homicide especially, as well as assault and damage to property, insult, bloodshed, rape and abduction – but there is much less contemporary indication of any consequent secular court proceedings, although this must be implied both by proceedings in church courts

16 17 18 19 20

représentations (Paris: La Sorbonne, 2003), pp. 15–38, at pp. 18–23 especially; and, in a Welsh context, D. Jenkins, ‘Crime and tort and the three columns of law’, in T. M. Charles-Edwards and P. Russell (eds), Tair Colofn Cyfraith. The Three Columns of Law in Medieval Wales: Homicide, Theft and Fire (Bangor: Welsh Legal History Society, 2005), pp. 1–25, at pp. 1–2. ‘Crime’ is also used of Iberian material up to 1038 by G. Martínez Díez, who makes useful points on procedure, ‘Terminología jurídica en la documentación del reino de León. Siglos IX–XI’, in Orígenes de las lenguas romances en el reino de León, Siglos IX–XII, 2 vols (León: Centro de estudios e investigación ‘San Isidoro’, 2004), vol. 1, pp. 229–72, e.g. at p. 254. Cel228. Lii360; also slaughter to the value of 130 solidi in this case. S84. Cel456. Use of a cattle standard was not unusual in tenth-century Galicia; see ch. 6 above, ‘Sale, price and valuation’. Liii556.



and by the twelfth-century text known as ‘Braint Teilo’.21 The latter is a claim to rights of jurisdiction over such cases by the church of Llandaff. For Brittany the evidence is also more flimsy, simply because the records of dispute settlement deal overwhelmingly with property disputes. However, assault and homicide, and again theft of stock, do occur.22

Occasions of property dispute Spanish and Portuguese texts provide plenty of evidence of property disputes too and they constitute just over half of all recorded dispute settlements. Sometime before 985 a woman, Fernanda, was taken to court in Galicia by her sister-in-law, Astotia, on the grounds that the husband should not have given his wife family property but should have passed it on to herself, as his sister.23 Fernanda won, before judges in Samos, because the local abbot supported her and because her husband said that the property had been properly conveyed, at her marriage, as her morning gift (‘in nuptias per cartam de beisadigo’). Farther to the east the bishop of León took action against a man called Pedro and three associates for appropriating a stretch of land near León.24 Denying it, they were required to take an oath before judges, but confessed instead. They signed one document, declaring this; and another, on 17 April 946, formally returning the land and agreeing not to repeat the claim, on pain of the enormous penalty of 500 solidi to the saio. There are disputes over water rights and mill rights too, like the case of Gondemaro and his neighbours against the abbot of Valdevimbre over water channels on the meseta to the south of León.25 Gondemaro and his neighbours seem to have won against the monastery: the abbot’s case is three times described as ‘malicious’. The dispute was about whether the water channel feeding Gondemaro’s mills diverted water from the abbot’s mills; trusted men were sent from the court to inspect the channel and measure the water levels as water was drawn off; in the end, in 938, they found that there was enough for everyone. There are plenty of comparable Breton property cases, both between lay parties and between church officers and laity: for example between brothers formally disputing a share-out of family property; between peasants disputing how much

21 LL237b (c. 925), LL223 (c. 940), LL222 (c. 942), LL233 (c. 905), LL263 (c. 1040), LL271 (c. 1075), LL259 (c. 1040). The text of ‘Braint Teilo’ can be found in LL pp. 118–21; see W. Davies, ‘Braint Teilo’, Bulletin of the Board of Celtic Studies, 26 (1974–6), pp. 123–137, reprinted in ead., Welsh History in the Early Middle Ages: Texts and Societies (Farnham: Ashgate, 2009), Essay III. An earlier section of ‘Braint Teilo’ (c. 1000) claims less specific rights of jurisdiction for the bishop of Llandaff. See now P. Russell, ‘Priuilegium Sancti Teliaui and Breint Teilo’, Studia Celtica, 50 (2016), pp. 41–68. 22 CR202 (858), CR107 (844), CR163 (860), CR32 (862); CR32 was settled before it reached court. 23 Sam239. 24 Li191. 25 Li128.



land one had sold the other; between the abbot of Redon and a priest whose uncle had endowed the monastery.26 There are Welsh parallels too, but they are more limited in range: although there are stories of disputes between laymen, they are stories recorded in a form which may well not derive from a contemporary record, unlike those of Iberian and Breton charters.27 There are certainly recorded cases of judgment and penalty – for damage to church property, violation of sanctuary or control of properties – but these are cases of church officers against laymen, in which the church parties win; these records look like records deliberately made by ecclesiastical beneficiaries.28

Procedure The cases cited above indicate that there were procedures for court settlement in Spain and Portugal and officers of the law to administer them. The law officer we encounter most is the saio – a man who supervised the ordeal and the taking of oaths, who could initiate a court case on receipt of a complaint and who sometimes took a fine. It was the saio Teodesindo who took Gonzalo and his wife to the count’s court in the sheep-stealing case cited above; here he was part of the procedure for beginning a case. For ending cases, as for example in the dispute between the bishop of León and Pedro, there could be formal, written agreements.29 There were also different procedural stages along the way, which could last across several occasions; the trusted men sent out by the court to measure water levels were collecting data in the course of a series of hearings. Some records detail many stages of court procedure. In an assault and wounding case in 946, on the meseta, the priest Berulf gave notice to the saio on 26 March that he had a case against a man called Matheo, who had invaded the land of the church of San Esteban and wounded its men, to the value of 72 solidi and Matheo agreed to turn up to hear the charge, on pain of 100 solidi if he defaulted; there follows the verbatim evidence of two witnesses on behalf of Berulf; then come the formal conditiones sacramentorum, a record of the oaths of Berulf’s witnesses, swearing by God, twelve prophets, the twelve apostles, the four Gospels, the relics of the Virgin Mary and the relics of San Claudio. This is followed by Matheo’s confession and 26 CR192 (834–7), CR139 (860), CR56 (865–6). 27 E.g. LL212, LL214, both relating to transactions of c. 862. 28 This does not deprive them of value, although such records clearly have a different character. Some of the Llandaff records of church victories are clearly later fabrications, but others lack the characteristic later formulas and appear to derive from tenth-century records: e.g. LL222 (c. 942), LL217 (c. 960), LL231 (c. 910). 29 The amount of documentation generated is well discussed by R. Collins, ‘“Sicut lex Gothorum continet”: Law and charters in ninth- and tenth-century León and Catalonia’, English Historical Review, 100 (1985), pp. 489–512; R. Collins, ‘Visigothic law and regional custom in disputes in early medieval Spain’, in W. Davies and P. Fouracre (eds), The Settlement of Disputes in Early Medieval Europe (Cambridge: Cambridge University Press, 1986), pp. 85–104. But see my further, detailed, comment, Davies, Windows on Justice, Part I.



his witnessed agreement to restore the land to Berulf within six days, on pain of 100 solidi; and finally a further commitment, on 7 May, that Matheo and family would pay a fine of 200 solidi for any further incursions and a list of witnesses of the consignment of the property.30 Note that there are at least three, and perhaps more, separate meetings here within a six-week period; note that there are oaths and witnessed agreements, as well as witness evidence: two different sets of witnesses are recorded for this one case, in addition to the witnesses who gave evidence. A Portuguese charter of 999 records that a man called Lovesindo Abenazar raised a case with the local count against the monastery of Guimarães over ownership of his father-in-law’s farm, Villa Sautello (Soutelo, 5 km north of Braga).31 The saio called the monastery to court and, having heard the opening statements, required both sides to appear with five witnesses each, on 1 November, at an appointed place, Pedralva. There each set of five gave evidence. Then the saio sent everyone to the count’s court, where the judges read all the documents and ‘understood everything’ and found that the monastery’s representative had the truth. They ruled that the monastery’s five witnesses should then take the oath. Subsequently everyone arrived at the appointed church for the oath, where Lovesindo refused to give in but gave assurances to the saio that he would give back the farm and make no further claim, on pain of the large penalty of 500 solidi, given that he was offered another farm instead, and 28 solidi as well. In effect a deal was done. In these Iberian cases there is much more detail of the formal stages of a court case and of the written depositions, confessions and agreements that they might produce. There are no early medieval Welsh parallels for the production of written documents on this scale in a court case, and there are barely parallels to the procedure, although the changes made to the original judgments recorded in some of the Llandaff narrations must imply some staged processes: for example, in one early tenth-century case, since the gold required as penalty for fighting could not be found, the payment was made in land instead.32 Nor are there many Breton parallels for the generation of written documents. There is one Breton sworn witness statement, but there is nothing like the formal confessions or agreements of the Iberian corpus.33 Given the quality and the detail of these surviving Breton texts, the implication must be that the process did not generate the same range or number of written texts for individual cases. In Brittany we are occasionally told that the accused was given time to appear with evidence – the ruler Salomon gave Ratfred ten days to gather documents and witnesses – so that must imply staged processes similar to those that were common in Iberia, at least for some cases.34 30 31 32 33 34

Li192. See further ch. 4 above, ‘Creating records of judicial disputes’. PMH DC 183. I am very grateful to André Marques for assistance with identifications. LL233. CR185 (858); cf. the note, CR205 (848–67). CR105 (857–8).



Progress towards a conclusion There are plenty of judges in the Iberian corpus of charters, often termed iudices, and they characteristically although not invariably worked in panels. While the texts often assign them a judging function (‘qui iudicavit’), their actions might reasonably be described as progressing a case towards a conclusion rather than making a final decision.35 Assessment of written evidence, as in the Guimarães case, was one means of taking a case forward but the written word did not necessarily override other evidence, for documents could be pronounced false;36 indeed it was the oath about documents that clinched the Guimarães case, not the documents themselves. In striking contrast, Welsh records of disputes never refer to written texts as justification for the case of one or other party.37 In Brittany the role of written evidence was strongly emphasized by the texts at all social levels – ‘vanquished as much by the charter as by the witnesses and guarantors’, as the record of one disputed property case proclaims in 860.38 If anything, the value of written evidence is more prominent in the Breton record than the Iberian. Taking evidence from witnesses was very common in Iberia, with witness statements frequently recorded and preserved; they were those who saw and heard, vidimus et audimus. It is particularly notable in disputed boundary cases, though by no means confined to those: local people, with local knowledge, would join members of the court and walk a boundary, searching for the markers as well as proclaiming from their own knowledge of past practice. There is a series of cases relating to the valley of Verín, in southern Galicia: in 950–1, for example, elders of neighbouring Pazos appeared to determine the bounds between Verín and Abedes.39 Breton material has good examples of similar procedure: the ruler Salomon along with others walked the disputed bounds of a large estate at Cléguerec in 871.40 There are no direct parallels in Welsh material, but the detailed perambulations of bounds attached to many Welsh charters, especially of the tenth and eleventh centuries, perhaps imply similar procedures, particularly when such perambulations follow records of disputes.41 It is also interesting that later Welsh law texts called witnesses gwybyddiaid, the ‘knowers’, those who saw and heard.42

35 For judges, see ch. 15 following, ‘Judges and judging’. 36 Cited by Collins, ‘“Sicut lex Gothorum continet”’, p. 508 (Oña 3). Cf. C55 (945) – in which a false charter was burned; Lii410 (968), in which the authenticity of a charter was questioned. See also I. Alfonso, ‘Judicial rhetoric and political legitimation in medieval León-Castile’, in I. Alfonso, H. Kennedy, J. Escalona (eds), Building Legitimacy: Political Discourses and Forms of Legitimacy in Medieval Societies (Leiden: Brill, 2004), pp. 51–87, at pp. 80–1. 37 This is as true of the Lichfield marginalia as of the Llandaff narrations. 38 CR139. 39 Cel95, cf. Cel93, Cel502 (940); see Davies, Windows on Justice, pp. 221–6; also R. Pastor, Resistencias y luchas campesinas en la época del crecimiento y consolidación de la formación feudal. Castilla y León, siglos X–XIII (Madrid: Siglo XXI, 2nd edn, 1990), p. 79. 40 CR247. 41 E.g. LL170 (c. 850), LL218 (955), LL233 (c. 905), LL239 (c. 925), LL244 (c. 980). Cf. Smith, ‘Judgement’, p. 67: hynefyddion, ‘old men, elders’. 42 Smith, ‘Judgement’, p. 3.



Taking a solemn oath was common in Spain and Portugal, as in several of the examples above, although the function of the oath can look ambiguous to us now.43 In Iberia the oath was very frequently a confirmation, or test, of witness evidence and rarely seems to have been ‘oath-helping’ (or compurgation). (An oath-helper, supporter of an accused person, characteristically swore to the reliability and good character of the accused rather than to the particular facts of a case; sworn witnesses did the latter.) The record of a judicial decision in favour of two brothers over water rights for a mill in the river Tirón comes close, however: the brothers brought twelve witnesses to the court and three fiadores; the witnesses testified and the fiadores swore oaths; the best explanation is that the latter swore to the truthfulness of the brothers.44 Oaths are not especially common in Breton records, although there are a few clear cases of oath-helping and there are some sworn witnesses too.45 There are no early medieval Welsh examples, although a twelfth-century addition to the late eleventh-century Life of Cadog refers to supporting denial of a charge with the oath of sixty men, and later medieval Welsh law has denial by compurgation;46 use of the oath is very strong in – the sometimes analogous – Irish law of the early middle ages, which also employs finer distinctions in the terminology of witnessing than are common in any of the systems considered here.47 And then there is the ordeal. In the opening example, from Valdoré, the accused had a fidelis who appears to have undergone the ordeal on his behalf.48 More usually people other than the contesting parties went to the ordeal, so that the judgment of God be revealed through their innocence: a case of the same period (986–99) between the bishops of Lugo and of Iria (forerunner of the bishopric of Santiago), about control of the income of Santa María in Présaras, focussed on whether it pertained to the monastery of Sobrado (within the jurisdiction of Iria).49 Two abbots (one for Lugo and one for Sobrado) were appointed to conduct a local

43 Cf. the comments of I. Alfonso Antón, ‘Resolución de disputas y prácticas judiciales en el Burgos medieval’, in J. J. García González, F. J. Peña Pérez, L. Martínez García (eds), Burgos en la plena edad media (Burgos: Asociación provincial de libreros de Burgos, 1994), pp. 211–43, at p. 217. 44 SM27 (940); fiadores could have signified sureties, given that the word fiador is sometimes used for this function (e.g. Cel368 or cf. T66 – fiator); however, fidiator is more common for surety and there is no reference in this text to an earlier transaction which had been guaranteed; in the context, oath-helping seems more likely. Cel144 (s.d.) seems to refer to an oath-helping case: a late tenth-century list of persons attached to Celanova refers to a woman whose father had denied paternity – the man he accused took an oath with twelve others in order to clear himself. 45 Oath-helpers: CR191 (801–12), CR46 (860–71); sworn witness: CR185. 46 Vita Sancti Cadoci, ch. 69: Vitae Sanctorum Britanniae et Genealogiae, ed. and trans. A. W. Wade-Evans (Cardiff: University of Wales, 1944), p. 138. R. C. Stacey, The Road to Judgment: From Custom to Court in Medieval Ireland and Wales (Philadelphia: University of Pennsylvania, 1994), pp. 156–7, 208–9; Smith, ‘Judgement’, p. 67. 47 See F. Kelly, A Guide to Early Irish Law (Dublin: Institute for Advanced Studies, 1988), pp. 198–202. For fíadnaise (eye-witness evidence), teist (character-witness evidence) and imthach (compurgatory oath), see Críth Gablach, ed. D. A. Binchy (Dublin: Institute for Advanced Studies, 1941), e.g. lines 19, 535 and Binchy’s comments, pp. 90–1, 99–100. 48 See the words of caution above at n. 4. 49 Sob109.



inquest about the matter; they spent fifteen days taking evidence from respected men of the locality but could not find anyone who knew what had happened in the past. There was then a further meeting of the bishops; the abbots reported their findings but the bishop of Lugo thought the local men must know about the affiliations of the property but were concealing it; since the abbots had found no reason to suppose deception, two men from each settlement were therefore required to take an oath that nothing was known; five pairs from five settlements did so and the formal oaths were recorded under the supervision of the saio, including the statement that they did not know of anyone with any rights in the church other than the dues which had been given to Sobrado by King Ordoño. An innocent then had to go to the ordeal, to test the veracity of their oaths, and trusted men were appointed to witness it; come the appointed day, more than fifty people gathered at the church of Sant Olalla: the innocent removed the hot stones and three or four days later his hand was clear, as presented by the saio in a council at San Julián and certified by the appointed witnesses, again recorded verbatim. The bishops then acknowledged that the truth had been revealed and Sobrado got the rent. In these cases the ordeal was intended as a test of the quality of the sworn evidence – proof that the oath was good, just as in the Ermegildo case, the ordeal was a test of the credibility of Ermegildo’s oath. As a procedure this is extremely unusual, in two senses: it is markedly different from anything we find in Wales or Brittany before 1100; and its use to test witnesses is at first sight surprising, for in Catalonia the ordeal was reserved for the accused.50 The ordeal was by no means used in every case; it appears to have been invoked where there was an element of doubt, particularly where there was reason to doubt the honesty of some actors. Many cases end with a confession, though not all: compare the case of Gondemaro and the abbot of Valdevimbre, who lost the case he brought once evidence had been gathered. Confession is in part a convention of the record-making tradition – revelation of ‘truth’ tends to be associated with confessions – but there is no reason to suppose that all confessions were imagined. We can only guess at the local politics, factions and tensions which led to this.

Penalties The ordeal was a test not a penalty. If a person lost a property dispute in northern Iberia, the usual outcome was restitution of the property to the declared owner,

50 Cf. Collins, ‘Visigothic law and regional custom’, p. 87. Appearance of this practice in some texts of Visigothic law is a ninth- or tenth-century addition; see J. M. Salrach, ‘Prácticas judiciales, transformación social y acción política en Cataluña (siglos IX–XIII)’, Hispania, 57:3 (1997), pp. 1009–48, at p. 1020. Whether or not use of the ordeal was always a straightforward proof is arguable; see S. D. White, ‘Proposing the ordeal and avoiding it: strategy and power in western French litigation, 1050–1110’, in T. N. Bisson (ed.), Cultures of Power: Lordship, Status, and Process in Twelfth-Century Europe (Philadelphia: University of Pennsylvania, 1995), pp. 89–123, who argues that its use was strategic, to ‘gain a tactical advantage’, pp. 91–2, 99–101, 122.



with no necessary penalty but with heavy sanctions if there should be a further attempt to gain the property. Criminal cases were different. For wounding, penalties were assessed in solidi or in cattle and assessed by a competent officer. The manner of reference implies that these were compensatory payments, presumably assessed in accordance with the scale of the injury (and possibly in accordance with the status of the person injured), and due to the injured person. On the other hand, for rape, arson, siege and theft, penalties were levied, payable to the court holder, who was usually a count or other aristocrat. In extreme cases property could be confiscated and bad men exiled but there seems to be no evidence of a death penalty. In practice, whatever the terms of the penalty, the fines were often paid in land.51 There are fewer records of compensation for the victims although in one case the due penalty was payment of a horse worth 40 solidi to the injured party (compensation) and of another horse worth 40 solidi to the court holder (fine) – the latter, as so often, commuted to land.52 Payment to victims may well have been much more common than is recorded: we know about payments to court holders because they were wealthy aristocrats and kept records; record-keeping habits mean that there is information about one beneficiary but not the other. In any case, some part of the payment to the court holder could have been passed on to the victim. A few examples support the likelihood that payment to victims was more common than recorded: a Galician case of assault, resolved in 952, hung on wounds assessed at a value of ten cattle.53 After assessment the victim went to court and those who had committed the assault came to court to confirm that they would pay, on an appointed day. They duly came, and made a payment to the value of four cattle, in the form of orchards for the court holder and his wife. The very fact that the victim went to court after his injury had been assessed indicates that he expected to receive something. The principle of compensation is clear, as is that of a scale of values. This case is also useful in showing that penalties could be reduced. Indeed, there are a number of cases in which boni homines (literally ‘good men’, i.e. ‘worthies, elders’) petitioned for leniency in order to achieve this. One such is a Portuguese record which both notes a compensatory payment made to a family by a priest who had killed one of its members and indicates that boni homines petitioned the priest’s lord for help.54 Both Welsh and Breton material provide clear parallels, although not many examples. In both cases there were scales of assessment of personal worth: a Llandaff charter of the early tenth century, for example, provides that reparation be made in accordance with a bishop’s status; and a unique Breton charter of 858

51 Cf. Martínez Díez, ‘Terminología jurídica’, pp. 251, 263–5; García Leal, ‘Los condes Fruela Muñoz y Pedro Flaínez’, p. 89; Martínez Sopena, ‘La justicia en la época asturleonesa’, p. 241, on the iudicato and other payments. 52 Liii578 (997); cf. Liii624 (1002), in which one vineyard went to Munio Fernández, the court holder, and another to the victim. 53 Sob103. 54 PMH DC 53 (943).



records that the worth (precium) of the offender Anau was to be paid if he committed a further offence.55 In both regions there was clear family responsibility for offences committed by its members (although there is less evidence of this in Brittany, and the responsible family circle seems to have been narrower).56 As for boni homines: there are a few panels of boni viri in Breton court cases, although senior (literally ‘elder’) is a more common term, while the degion (literally ‘good men’) of the Lichfield marginalia who mediated in west Wales are also a direct semantic (and perhaps functional) parallel for the Iberian boni homines.57 One Spanish case of 994 mentions penal slavery: a woman guilty of adultery, because she had slept with her father-in-law, should rightly have been reduced to the status of hereditary slave (ancilla origenale), but was permitted to make a gift of property instead; all her inheritance in Valdevimbre went to the court holder, Munio Fernández.58 There are Welsh references which may well imply penal slavery: an important Llandaff charter of 955 indicates an initial judgment that men be handed over as the penalty for breach of sanctuary, although this was later commuted to a land grant.59 I am not aware of any Breton evidence of penal servitude at this time.

Enforcement Good evidence of enforcement mechanisms in northern Iberia comes from the use of sureties, personal guarantors, although both the activity of the saio and the role of the court holder indicate that they had a part in enforcement too. Suretyship is also extremely well evidenced in practice in Breton village contexts where, as well as guaranteeing sale transactions, sureties were used, for example, to guarantee future payment of rent after default, to guarantee that a loser would not bring another case and to guarantee that disputing parties would in future stick to an agreed boundary line.60 Suretyship is also a strong theme in Welsh law codes of the later middle ages, some part of which clearly had early medieval origins, although I know of no recorded cases in practice in the early middle ages.61 In Spain, when the boundaries of properties on the slopes

55 LL233 (cf. LL237b, of c. 925); CR202. 56 E. g. LL223 (c. 940), LL263 (c. 1040); CR107 (844), CR261 (874–6). 57 Jenkins and Owen, ‘The Welsh marginalia in the Lichfield Gospels’, part 2, pp. 91, 101; CR129 (c. 834), CR180 (843–49). The term gwyrda (‘good men’) is still common in later medieval Welsh law texts, where it can be differentiated from hynefyddion ‘elders’; Smith, ‘Judgement’, pp. 65, 67. The Breton panels were not quite the same: they were involved in judging rather than mediating. 58 Liii561. 59 LL218; cf. ‘in servitio perpetuo’ for a thief in Vita Sancti Cadoci, ch. 33. 60 CR134 (868–71), CR29 (854–67), CR47 (893). 61 See W. Davies, ‘Suretyship in the Cartulaire de Redon’, in T. M. Charles-Edwards, M. E. Owen, D. B. Walters (eds), Lawyers and Laymen: Studies in the History of Law Presented to Professor Dafydd Jenkins on His Seventy-Fifth Birthday: Gŵyl Ddewi 1986 (Cardiff: University of Wales, 1986), pp. 72–91 (reprinted in W. Davies, Brittany in the Early Middle Ages [Farnham:



of the Cantabrian Mountains were agreed between seven village communities and the monks of Pardomino in 944, each village provided a surety that it would stick to the agreement.62 Sureties were the people who saw that the terms of an agreement or of the resolution of a case were carried out. They might do this by paying themselves if the loser defaulted or by forcing a defaulting party either to pay or to appear in court – just like the classic paying and enforcing sureties of Irish and Welsh law.63 They also performed practical functions, transmitting payments and formally handing over due land. So, when a man called Salomon was found guilty of theft in 931, in northern Galicia, and was required to pay 30 cattle, six sureties were named that he would not disappear before doing so, and if he did the sureties were to pay the court 20 cattle.64 When Revel committed homicide on the meseta in 979, he was required to pay half a vineyard; two sureties were assigned to make sure that he did, and they were to pay 100 solidi if he failed.65 When García Refugano was ordered to appear in court by the saio in a property dispute over the church of Tubilla del Lago in Castile, one surety had to give the court holder – the count – 300 solidi if García failed to turn up in court with his charters, clearly an extremely strong incentive to the surety to get him there.66 Legal historians have tended to situate suretyship within the domain of private law.67 In fact, although sureties represented and had obligations to private individuals, and although they often seem to have carried out their obligations without the supervision of an officer of state, public endorsement of their obligations was extremely important to the functioning of the system. It is a mistake to see suretyship as purely private.68 One can see it working in very similar ways in Breton villages and across the Iberian countryside, although by the late tenth century sureties were being used for a wider range of actions in Spain and Portugal.69

62 63 64 65 66 67

68 69

Ashgate, 2009]). Robin Stacey discussed Irish suretyship in the same volume, ‘Berrad Airechta: an Old Irish tract on suretyship’, pp. 210–33, and developed her views on the Welsh material in her book Road to Judgment. Li184. Stacey, Road to Judgment, pp. 34–45, 146–57. Sob21. This is what the text says – it may not be an error, since penalties did get reduced; see above and Martínez Sopena, ‘La justicia en la época asturleonesa’, pp. 256–8. Lii463, Lii464 (a separate document recording the obligations of the sureties). C90 (957). J. Gilissen, ‘Esquisse d’une histoire comparée des sûretés personnelles’, in Les sûretés personnelles. Première partie. Synthèse générale; civilisations archaïques, antiques, islamiques et orientales, Recueils de la Société Jean Bodin pour l’histoire comparative des institutions, 28 (1974), pp. 5–127, at p. 29. For a good discussion of the interplay between public and private roles, see Robin Stacey on Irish hostage-sureties, Road to Judgment, pp. 82–111. See ch. 17 following, ‘On suretyship in tenth-century northern Iberia’.



Conclusions Given the differences in historical background, in geography and in economy, there are more similarities between the Iberian and the Brittonic experience of dispute settlement procedure in the early middle ages than one might expect. Of course, we cannot know how many disputes went unrecorded and unresolved – the homicides that begin recorded settlement processes were clearly also the ends of largely unseen disputes – but there were well developed settlement procedures in all of these areas, many of which took place in a formal court setting. There are many similarities, but there are differences too. Some of the differences are differences in the nature and quantity of evidence available, so that has to be borne in mind. Hence, whereas we can see a court in almost every village in eastern Brittany covered by the documentation – in other words, a village meeting that could function as a formal court – this is not evident in northern Iberia. There we tend to see higher-level courts, although they could deal with peasant cases. The question is therefore: is the village level missing in Spain and Portugal, because dealt with elsewhere, at a higher level? Or is it simply that evidence of village-level courts does not survive? There are tiny hints of lower-level courts in the Iberian material, but not many of them: the case of the cousins who tied up their benefactors seems to have been settled in a local concilium, a common word for a local meeting in parts of northern Spain;70 a Portuguese case of 991 over control of the church of Guilhabreu (just north of Porto) involved action in another concilium before it got to court; while many of the examples cited above indicate action taken in advance of referral to a count’s court, particularly actions by saiones.71 Secondly, there are many more records of criminal cases from Iberia, where the corpus of criminal cases is extremely important for early medieval Europe as a whole. In Spain and Portugal records of criminal cases tend to survive because fines went to the court holders and in the tenth century they were often paid in land.72 There is also much less evidence of secular court activity in Wales than in Iberia or Brittany – certainly a consequence of the documentation available but perhaps also a comment on practice. Thirdly, despite the many similarities (the value of evidence, especially oral evidence, importance of local knowledge, scales of personal worth, family responsibility for offences, use of sureties), there do seem to be some real differences, 70 See W. Davies, Acts of Giving: Individual, Community, and Church in Tenth-Century Christian Spain (Oxford: Oxford University Press, 2007), pp. 202–7. Also J. J. Sánchez Badiola on the territorial concilium as judicial assembly, El Territorio de León en la edad media: poblamiento, organización del espacio y estructura social (siglos IX–XIII), 2 vols (León: Universidad de León, 2004), vol. 1, pp. 458–9; Martínez Sopena, ‘La justicia en la época asturleonesa’, p. 249. 71 PMH DC 163. 72 Later commuted to metal payments – see above n. 8; cf. García Leal, ‘Los condes Fruela Muñoz y Pedro Flaínez’, p. 31: participation in judicial acts was a means of acquiring property in the tenth century.



in procedure as also in values. These include: the use of advocates to speak for one or other or both parties in Spain, but not in Brittany (though this may well be a function of the comparative social levels of the courts that are evidenced); the use of clerics in panels of judges, which is quite common in Iberia, but virtually unknown in Brittany; the range of documents generated by Spanish and Portuguese cases – the careful recording and preservation of witness statements, oaths, renunciations, agreements, perhaps a recognition of the value of compromise, a recognition which may be culturally distinctive; the significance of written evidence in Iberia and especially Brittany but not apparently in Wales; the frequent citation of written law in Spain, but no hint of such a thing in Brittany, although one lawyer’s principle is cited in Wales;73 the prominence of the confirmatory oath in Spain and Portugal in contrast with Wales and Brittany; and the unusual use of the ordeal to test witnesses in the kingdom of León. Fourthly, there were dedicated and pro-active legal officers in Spain and Portugal, who administered much court procedure and who could operate as a channel of communication to the ruler or his agent, although they were not necessarily appointed by either; there is no precise parallel to this in Welsh or Breton material of the early middle ages. Further, although the holders of courts, who took fines, could be direct agents of the king, many were not and simply exercised the superiority in their own right.74 For Wales, although we can see both church courts and the very local procedures of the Lichfield Gospels, there is little contemporary evidence to locate these, nor secular equivalents, within a comparable process.75 In Brittany the holders of courts could be agents of the ruler, but more often than not they were machtierns, local petty aristocrats with responsibility primarily to their families and to some extent to the local community. Although the place of the local is common to all these systems, much of the Iberian system looks to have happened in the context of a more apparent royal power, especially in cases of the later tenth century. It is instructive to consider these western parts of Europe alongside each other, given that dispute settlement procedure of the early middle ages is so often situated within a Carolingian (or English) framework. Some practices are common to all three and some to the Carolingian world as well (for example, the presence of 73 LL218. 74 See C. Estepa Díez, ‘Poder y propiedad feudales en el período astur: las mandaciones de los Flaínez en la montaña leonesa’, in Miscel·lània en homenatge al P. Agustí Altisent (Tarragona: Diputació de Tarragona, 1991), pp. 285–327; García Leal, ‘Los condes Fruela Muñoz y Pedro Flaínez’; Martínez Sopena, ‘La justicia en la época asturleonesa’, p. 241; Davies, Windows on Justice, pp. 202–3. For some ‘interdependence . . . of royal justice, arbitral justice and negotiated resolutions’, see I. Alfonso Antón, ‘The language and practice of negotiation in medieval conflict resolution (Castille-León, eleventh-thirteenth centuries)’, in B. S. Tuten and T. L. Billado (eds), Feud, Violence and Practice (Farnham: Ashgate, 2010), pp. 157–74, at p. 169. 75 Thomas Charles-Edwards argues for the Welsh king’s interest in homicide by the late eleventh century: ‘The three columns of law: A comparative perspective’, in Charles-Edwards and Russell (eds), Tair Colofn Cyfraith, pp. 26–59, at pp. 30–4.



boni homines) – a nice warning not to treat the Brittonic as distinctively ‘Celtic’, much less as quaint and exotic. Where there are differences they tend to lie in record-keeping habits and in the value placed on different ways of arriving at a conclusion. Writing and keeping a wide range of documents, and the weight accorded to the oath and the judgment of God, are especially characteristic of Iberian practice, differentiating it from the Brittonic as well as from much of the rest of Europe (though long and detailed descriptions of court procedure are very characteristic of Italian records). What is also interesting, in the wider European context, is that some of the so-called characteristics of the Carolingian world – the public court, collective judgment – are as characteristic of the Iberian world.76 The suggestion that they disappeared in the Frankish world, to be replaced by ‘private’ mechanisms and ‘private’ power, has played an enormous role in the debates about the changing nature of power in the central middle ages that have characterized the last two generations – feudal revolution and feudal mutation, in shorthand.77 To suppose the public court distinctively Carolingian will not work – it clearly existed beyond the Carolingian world; to suppose collective judgment a vanishing early medieval norm will not work either – it clearly did not disappear – a point well made by Beverley Smith in the later medieval Welsh context.78 Hence, to argue that the disappearance of the public court is intrinsically tied to the fragmentation of Carolingian power makes little sense, quite apart from the arguments against this on purely Frankish terms.79 There was more to the ‘public’ court than the presence of the state: looking at the dispute settlement procedures of Europe west of Frankia should make us think again about the nature of the public.

76 Some recent Spanish historiography does locate Iberian practice within the Frankish paradigm: it postulates a late Roman judicial tradition collapsing with the disappearance of central power, from the Alps to Galicia; Salrach, ‘Prácticas judiciales’; J. M. Mínguez, ‘Justicia y poder en el marco de la feudalización de la sociedad leonesa’, La giustizia nell’alto Medioevo, secoli IX–XI, 2 vols, Settimane di Studio del Centro Italiano di Studi sull’alto medioevo, 45 (1997), vol. 1, pp. 491–546. Against all this, Martínez Sopena, ‘La justicia en la época asturleonesa’, p. 255. For collective judgment, the best overview is S. Reynolds, Kingdoms and Communities in Western Europe, 900–1300 (Oxford: Oxford University Press, 1984), pp. 23–34. 77 See T. N. Bisson, ‘The “feudal revolution”’, Past and Present, 142 (1994), pp. 6–42; Debate: ‘The “feudal revolution”’, ibid., 152 (1996), pp. 196–223 – D. Barthélemy, S. D. White; and ibid., 155 (1997), pp. 177–225 – T. Reuter, C. Wickham, T. N. Bisson; cf. S. D. White, ‘ From peace to power: The study of disputes in medieval France’, in E. Cohen and M. B. de Jong (eds), Medieval Transformations: Texts, Power, and Gifts in Context (Leiden: Brill, 2001), pp. 203–18. 78 ‘Judgement’, pp. 78–86: on the introduction of collective judgment. 79 For which, see S. D. White, ‘Tenth-century courts at Mâcon and the perils of structuralist history’, in Brown and Górecki (eds), Conflict in Medieval Europe, pp. 37–68; cf. Alfonso Antón, ‘Resolución de disputas y prácticas judiciales’, p. 220.


15 JUDGES AND JUDGING Truth and justice in northern Iberia on the eve of the millennium

In northern Portugal, in 999, a man called Lovesindo Abenazar raised a case with the local count against the monastery of Guimarães (at that time one of the most powerful of Portuguese monasteries, and backed by the principal comital family); the case was over a farm, the Villa Sautello, present-day Soutelo, 5 km north of Braga.1 The monastery said it had given the farm to Lovesindo’s fatherin-law, out of charity, on strict conditions of not passing it on nor selling it nor harming the monastery’s men. However, so they said, the father-in-law had gone into other monastic properties under the protection of this agreement, had taken over their port and prepared his boat there, had pressganged the monastery’s men into boat service, had had rows with their boat-builder and caused their men much trouble – so much so their original agreement had been voided. The saio (legal officer) required both sides to appear before him, at Pedralva 10 km to the south-east, with five witnesses each, on 1 November. Each set of five duly gave evidence (a woman and an abbot on each side). The saio then sent everyone to the count’s court at Villa Gondesindi. There the judges read the documents, listened to the witnesses and ‘understood everything and found the truth’ for Guimarães.2 Then the judges pronounced that the monastery’s representative (Vimara) should provide evidence that the monastery’s charters were earlier and required its five witnesses to take the oath, at a nearby church. Everyone assembled at the church; Lovesindo ejected Vimara from the church and refused to swear that what the monastery’s witnesses said was true, but he gave assurances that he would make no further claim (on pain of a penalty of 500 solidi to the

1 First published in W. Davies, Journal of Medieval History, 36:3 (2010), pp. 193–203: the Reuter Lecture, University of Southampton, June 2009. I was more than usually conscious of the responsibility of living up to the honour of delivering this Lecture: Tim Reuter was a very good friend, for more than thirty years, and I hope that what follows would both have interested him and caused him to reflect. 2 ‘Intellexerunt omnia invenerunt veridicum.’



monastery);3 he assigned the farm to the saio so that the latter could pass it back to Guimarães.4 Nowadays we are used to thinking of dispute settlement and conflict resolution as an important part of normal social process. Indeed, given the generation to which I belong, I am used to thinking of medieval courts as fora for negotiation, and as places for performance – even, in some contexts, mechanisms for securing emotional satisfaction;5 indeed, I myself contributed to a Past and Present volume on the ‘moral world’ of the law.6 This is not to say that the tensions surrounding conflict have had no place in interpretation of the texts, within the court as well as outside: the extent to which the strategies of power relations determined deals and lay behind the records of agreement remains a matter of debate.7 Our expectations of process are of course conditioned both by our contemporary experience and by our knowledge of the medieval past: our contemporary experience leads us to expect courts to be organs of state, established by the state, administering the law of the state, with judges who are learned in the law doing the presiding and sentencing; a jury of non-professionals often serves in criminal but not in civil cases, and trained non-professionals may have a role in judgment in minor cases; however, the mix of professionals and non-professionals varies considerably in different modern systems. Our knowledge of the early middle ages leads us to expect a

3 The solidus was a common unit of account, perhaps influenced by knowledge of the key unit of the reformed Carolingian currency, for which silver denarii were coined at the rate of 12 to the solidus, of which there were initially 22 to the pound weight; neither was minted as coin in northern Iberia at this time. 4 PMH DC 183. 5 On the importance of the paradigm of negotiation, see W. C. Brown and P. Górecki, ‘Where conflict leads: On the present and future of medieval conflict studies in the United States’, in W. C. Brown and P. Górecki (eds), Conflict in Medieval Europe: Changing Perspectives on Society and Culture (Aldershot: Ashgate, 2003), pp. 265–85, at pp. 283–4; on performance, P. J. Geary, ‘Land, language and memory in Europe 700–1100’, Transactions of the Royal Historical Society, 6th ser., 9 (1999), pp. 169–84, at pp. 175, 183; on emotions, D. L. Smail, The Consumption of Justice: Emotions, Publicity, and Legal Culture in Marseille, 1264–1423 (Ithaca: Cornell University Press, 2003). 6 ‘Local participation and legal ritual in early medieval law courts’, in P. Coss (ed.), The Moral World of the Law (Cambridge: Cambridge University Press, 2000), pp. 48–61. 7 ‘Conclusion’, in W. Davies and P. Fouracre (eds), The Settlement of Disputes in Early Medieval Europe (Cambridge: Cambridge University Press, 1986), pp. 207–40, at pp. 228–40; but especially Stephen White’s many papers, for example: ‘Proposing the ordeal and avoiding it: Strategy and power in western French litigation, c.1050–1110’, in T. N. Bisson (ed.), Cultures of Power: Lordship, Status, and Process in Twelfth-Century Europe (Philadelphia: University of Pennsylvania, 1995), pp. 89–123, at pp. 90–2; id., ‘From peace to power: The study of disputes in medieval France’, in E. Cohen and M. B. de Jong (eds), Medieval Transformations: Texts, Power, and Gifts in Context (Leiden: Brill, 2001), pp. 203–18; id., ‘Tenth-century courts at Mâcon and the perils of structuralist history: Re-reading Burgundian judicial institutions’, in Brown and Górecki (eds), Conflict in Medieval Europe, pp. 37–68. Cf., citing Foucault in an Iberian context, judicial enquiry as ‘a means of exercising power’, I. Alfonso, ‘Judicial rhetoric and political legitimation in medieval León-Castile’, in I. Alfonso, H. Kennedy, J. Escalona (eds), Building Legitimacy: Political Discourses and Forms of Legitimacy in Medieval Societies (Leiden: Brill, 2004), pp. 51–87, at p. 84.



strong emphasis on the local and on collective involvement in judgment – to some extent prefiguring the modern jury of citizens – where collective involvement is often taken to be an expression of a collective identity, and thereby an expression of public as opposed to private interest.8 The gradual erosion of this public mechanism by private interests is one of the grand themes of recent western European, as also of Spanish, historiography of the central middle ages.9 We might, then, expect performance, negotiation, the local, collective involvement and a sense of the public in our early medieval courts.10 However, the Portuguese case I have just quoted expects truth as an outcome; and other texts use words like ‘just’ and sometimes invoke an abstract concept of justice, like a León record of 952 in which the whole court recognized truth and justice – ‘tam pontifex quam omne concilio, agnoscentes veritatem et iustitiam’.11 So, I want to pose a question: rather than seeing the court as theatre, as opportunity for the expression both of individuality and of collective identity (through local involvement in decision-taking), did tenth-century commentators in Iberia think that going to court was essentially about finding truth and justice, grounded in abstract moral principle? To do that I am going to focus on judges and judging, since it is the process of judgment which was seen to give authority to the outcomes recorded. I shall begin with some basics on time and place, the nature of the records and the main aspects of procedure. Then proceed to address the questions: what did judges do? what was involved in judging? who were the individuals that did it? and what did contemporaries think they were doing? before giving some thought to the language of justice.

Time, place and records This paper is about northern Iberia, that is northern Spain and Portugal, excluding Catalonia (because of its Carolingian overlay);12 the north of Iberia only, because 8 On collective involvement, see S. Reynolds, Kingdoms and Communities in Western Europe, 900–1300 (Oxford: Oxford University Press, 1984), pp. 23–35. 9 Classically G. Duby, La société aux XIe et XIIe siècles dans la région mâconnaise (Paris: A. Colin, 1953), with much subsequent comment; for a survey, see White, ‘Tenth-century courts at Mâcon’. For some Spanish examples: P. Bonnassie, La Catalogne du milieu du Xe à la fin du XIe siècle: croissance et mutations d’une société, 2 vols (Toulouse: Université de Toulouse-Le Mirail, 1975–6); J. M. Salrach, ‘Prácticas judiciales, transformación social y acción política en Cataluña (siglos IX-XIII)’, Hispania, 57:3 (1997), pp. 1009–48; J. A. García de Cortázar and E. Peña Bocos, ‘Poder condal ¿y “mutación feudal”? en la Castilla del año mil’, in M. I. Loring García (ed.), Historia social, pensamiento historiográfico y edad media (Madrid: Orto, 1997), pp. 273–98; J. M. Salrach, ‘Les féodalités méridionales: des Alpes à la Galice’, in E. Bournazel and J.-P. Poly (eds), Les féodalités (Paris: Presses Universitaires de France, 1998), pp. 313–88. 10 I use the word ‘court’ for public assemblies whose main purpose was the performance of judicial business; of course, some may have dealt with other business too. 11 Li256; see further below for more examples. 12 Though it does share some common features of procedure; valuable recent work on Catalan courts and disputes includes: A. Iglesia Ferreirós, ‘La creación del derecho en Cataluña’, Anuario de



of Muslim control of the rest of Spain and Portugal and their more limited range of contemporary records; it is therefore about Christian and Latin Iberia, a large area notwithstanding. The time under consideration is up to the year 1000; in effect this means the late ninth and tenth centuries, since that is the period of rich documentation, very rich from the 930s onwards; documents continue thereafter, but the eleventh century was a time of considerable change in political, economic and social context.13 My interest is in the system before the changes. The records are charters, records of the transfer of property rights, in Latin, and there are over 2,300 from the area considered, from that period. There is very little relevant material in other kinds of source: for example, by contrast with some other parts of Europe, there are very few narrative sources. For the most part the charters record gifts and sales, but records of disputes sometimes precede detail of gifts, as explanatory background. The charter collections also sometimes include full or partial accounts of particular dispute cases, detached from any record of gift, in addition to the gift charters with extended narratives. It is appropriate to remember Tim Reuter’s comments here, published a decade ago: he pointed out that the narratives in charters all over western Europe became more detailed, precisely at this time, his ‘long’ tenth century; ‘Every charter tells a story’ and understanding the story (what and why it was told – the narrative strategies) was just as important as considering the authenticity of the text.14 These records are extremely mixed in format. There is no standard form, although two elements feature often: the conditiones sacramentorum, that is supposed verbatim records of oaths sworn; and formal statements of confession.15 An unusual example of the latter is written on a corner of an (apparently unassociated) record of a royal gift to the monastery of Sahagún, of 971, instructive because it suggests that making such records may well have been common, many now lost because they were not associated with any transfer of property: ‘In the presence of [three named] judges and the saio, Meme, I declare that I Haio was historia del derecho español, 47 (1977), pp. 99–123; Salrach, ‘Prácticas judiciales’; A. J. Kosto, Making Agreements in Medieval Catalonia: Power, Order, and the Written Word, 1000–1200 (Cambridge: Cambridge University Press, 2001); J. A. Bowman, Shifting Landmarks: Property, Proof, and Dispute in Catalonia around the Year 1000 (Ithaca: Cornell University Press, 2004); J. Jarrett, Rulers and Ruled in frontier Catalonia, 880–1010: Pathways of Power (London: Royal Historical Society, 2010); J. M. Salrach, Justícia i poder a Catalunya abans de l’any mil (Vic: Eumo, 2013). 13 Cf. A. García Leal, ‘Los condes Fruela Muñoz y Pedro Flaínez: la formación de un patrimonio señorial’, Anuario de estudios medievales, 36:1 (2006), pp. 1–110, at pp. 33, 79. 14 T. Reuter, ‘Introduction: Reading the tenth century’, in T. Reuter (ed.), The New Cambridge Medieval History: Volume III c. 900-c. 1024 (Cambridge: Cambridge University Press, 1999), pp. 1–24, at p. 5. 15 For detailed consideration, see W. Davies, Windows on Justice in Northern Iberia, 800–1000 (Abingdon: Routledge, 2016), Part I. Note the very useful comments of J. A. Fernández Flórez on conditiones sacramentorum, ‘Los documentos y sus scriptores’, in Monarquía y sociedad en el reino de León de Alfonso III a Alfonso VII, 2 vols (León: Centro de estudios e investigación San Isidoro, 2007), vol. 2, pp. 97–139, at pp. 136–8.



accused by Godleo . . . of my diabolical sin: [Godleo said that Haio] came upon a man from Montesone drinking water from the spring at Fuentes; he killed him with his sword and took the man [and] his ass to Kasigatu de Pice; and the saio found him at home with what he had thieved. I Haio have confirmed this sign with my own hand in this manifestum. Noted by the priest Placenti’.16 There are about 250 records of disputes from this period, some brief, some very detailed.17 There are large numbers from Galicia (which included northern Portugal at this time) and the province of León, and rather few from Navarre and Aragón; the collections from Sobrado, Celanova, Otero de las Dueñas, León, Sahagún and Cardeña are especially important. Cases could start with an appeal from one party to king or count, or because the saio, having heard an accusation from one party, commanded the other (or both) to appear before him. Court hearings were often large meetings, with many named people in attendance and many participating; they might take place at a royal or episcopal centre, or in a monastery or a church, or on disputed property. One case often spread over several meetings, in different locations. Advocates sometimes spoke for one or both parties; at other times the parties spoke for themselves. Hearings proceeded with a statement of the case and the production of evidence, whether written or oral or both. If there was no evidence, an enquiry might be held, with questions asked of local people. Failing all else, the judgment of God might be invoked in the ordeal of hot stones, usually applied to test witness evidence. Oaths were normally sworn by selected witnesses at some point in the proceedings, and these had an especially solemn significance. In the end confessions are common in these records, as are formal, witnessed, concluding agreements; if there was no confession, then liability was assigned; the losing party then agreed to hand over something that had been disputed, or made compensatory payments to victims or paid fines to the courtholder, and sometimes a combination of all three. Sureties could be appointed to make sure the resolution was carried out and texts often include further sanctions 16 S261. Julio Escalona has suggested the reading inkarigatu de pice, ‘loaded with fish’. 17 There are two useful papers on procedure by Roger Collins, although he was not using the full corpus considered here: ‘“Sicut lex Gothorum continet”: law and charters in ninth- and tenth-century León and Catalonia’, English Historical Review, 100 (1985), pp. 489–512; ‘Visigothic law and regional custom in disputes in early medieval Spain’, in W. Davies and P. Fouracre (eds), The Settlement of Disputes in Early Medieval Europe (Cambridge: Cambridge University Press, 1986), pp. 85–104; I discussed procedure at much greater length in Davies, Windows on Justice. There is a large Spanish historiography, of which see, for example, the following recent papers: J. M. Mínguez, ‘Justicia y poder en el marco de la feudalización de la sociedad leonesa’, in La giustizia nell’alto medioevo (secoli IX-XI), 2 vols, Settimane di Studio del Centro Italiano di Studi sull’alto medioevo, 44 (1997), vol. 1, pp. 491–546; I. Alfonso, ‘Litigios por la tierra y “malfetrías” entre la nobleza medieval castellano-leonesa’, Hispania, 57:3 (1997), pp. 917–55; G. Martínez Díez, ‘Terminología jurídica en la documentación del reino de León. Siglos IX-XI’, in Orígenes de las lenguas romances en el reino de León, Siglos IX-XII, 2 vols (León: Centro de estudios e investigación ‘San Isidoro’, 2004), vol. 1, pp. 229–72; P. Martínez Sopena, ‘La justicia en la época asturleonesa: entre el Liber y los mediadores sociales’, in A. Rodríguez (ed.), El Lugar del Campesino. En Torno a la Obra de Reyna Pastor (València/Madrid: Universitat de València, CSIC, 2007), pp. 239–60.



in the event of default.18 The matters under dispute were often about property (land, churches, rents, water rights, mills) but there were also many cases that we would term criminal in a more modern world: theft, fornication and rape, assault, breaking and entering, homicide, occasionally arson. Much of this is what might be found elsewhere in Europe in the early middle ages, but the presence and roles of a dedicated legal officer (the saio) is unusual, as is citation of written law; the volume of criminal cases contrasts strongly with numbers from other parts, witness the ‘zone d’ombre’ of the criminal of contemporary Italy.19

What did judges do? By ‘judges’ I am referring here to people who were termed iudex (literally ‘judge’) in the texts and iudices in the plural; as well as to people who were described as judging in witness lists – qui iudicabat, qui iudicaverit (he who was judging, he who has judged); and also to those who presided at court and are explicitly said to have joined in the judging. I will come back to these designations later. The charters often record that this or that happened ‘in the presence of (in presentia) Judges A, B, C’ or ‘before (ante) Judges D and E’, where both expressions occasionally, though by no means always, imply that one or more judges presided over the hearing; there does not seem to be any substantive difference in the role of judges between cases heard ‘in their presence’ and those heard ‘before them’.20 This applies to all regions, throughout the period (from the earliest cases we have), and to all types of case, whether property or criminal. One extremely common action performed by the judges was to proclaim what should happen next in the proceedings. So, for example, in a dispute in the Rioja over water to power a mill on the river Tirón, in 940, the three judges judged that, since twelve witnesses had testified and three others had sworn an oath in the church of Santa María of Settefenestre, there was no need to proceed to the ordeal.21 They might order a given number of witnesses to be produced or a

18 For examples, see ch. 17 below, ‘On suretyship in tenth-century northern Iberia’; I discussed sanctions at greater length in W. Davies, ‘Summary justice and seigneurial justice in northern Iberia on the eve of the millennium’, Haskins Society Journal, 22 (2010 [2012]), pp. 43–58. 19 F. Bougard, La justice dans le royaume d’Italie. De la fin du VIIIe siècle au début du XIe siècle (Rome: Ecole française de Rome, 1995), pp. 233–52. 20 Ante: Li191 (946), V33 (956), S159 (958), S295 (978), Liii556 (993); in/ad presentia: Floriano 74 (861), Li34 (915), Li128 (938), Sob129 (942), Sob130 (992), Cel252 (1000), for example. Both expressions can also include the person who presided over the court (for discussion of presidents, see W. Davies, ‘Holding court: Judicial presidency in Brittany, Wales and northern Iberia in the early middle ages’, in F. Edmonds and P. Russell (eds), Tome: Studies in Medieval Celtic History and Law in Honour of Thomas Charles-Edwards (Woodbridge: Boydell & Brewer, 2011), pp. 145–54). 21 SM27. Perhaps fiadores and fidiatores are ‘oath-helpers’ here rather than the more usual ‘sureties’; oath-helping is extremely rare in this corpus but this may be one of the few recorded cases in which it is cited.



given number of witness oaths to be sworn: in a northern Galician case of 985 between a monastery and an aristocratic woman over control of some freedmen, five judges ordered twelve witnesses to be provided by each party, and then a further three judges reviewed the evidence and required one set of witnesses to swear the oath;22 and in a southern Galician case of 1000 about ownership of a church, between the powerful monastery of Celanova and a man called Alfonso, the monastery came to court with 356 witnesses – against Alfonso’s ten (  fallacissimis et negligentiosis) – and so the judges ordered Celanova to provide just ten, but they were to take the oath.23 In the opening example in this paper, having read the documents presented, the judges ordered that the five witnesses produced by one party should take the oath. In effect, like some modern magistrates, they were deciding who was to be given the chance to demonstrate his or her case, and they were determining what parties had to do to make a case, as well as assessing the primary evidence. Sometimes – as in other parts of medieval Europe – judges made primary investigations themselves, as in 938 when three judges inspected and measured the rise and fall of water levels at San Juan en Vega near León, in another dispute about the control of water to power mills. As others had done before, they inspected and then pronounced that there was enough water for everyone (against the claims of the monastery, which had been arguing that local peasants were monopolizing and diverting the water supply).24 At other times judges decided very specific outcomes, although we do not often have explicit evidence of this. Defining the boundaries of disputed property is not uncommon, however, as happened in a Portuguese case of 936 in which two groups of local people argued about a boundary: they went to the local count for resolution, and the count together with a panel of three went to the land, walked it, found markers and did the division.25 Setting fines involved comparable decision making: in cases of stealing sheep or cattle or cheeses, the judges determined the penalties – simple repayment, or repayment two-fold, or even more. In a case of intrusion into and ploughing some monastic land, the intruders (Velito and Calendo) were liable to pay a very large fine, but in 997 they negotiated with the judges and achieved a substantial reduction.26 And at other times they just ‘decided’; in other words, the narrative explicitly credits the judges with reaching the final decision. For example, in an assault case on the tenants of the monastery of Santos Justo y Pastor in the Cea valley in 946, where wounds had been suffered to the value of 72 solidi, after citation of the oaths which had been sworn, the text then says ‘they testified and the judges judged’; in effect the judges accepted the evidence presented; after that the 22 23 24 25 26

SamS-8. Cel252. Li128. PMH DC 42; compare Sob129 (942); and Sam46 (933), Cel502 (940), Li89 (931). E.g. Sob75 (858), Sob23 (949), Sob31 (951); OD43.



accused confessed.27 In a property case in 958 concerning the southern foothills of the Cantabrian Mountains the judges, having required evidence to be produced, waited three days for one party to appear; since her representative did not do so, the judges permitted (permisimus) the advocate for the other party to go and claim the land – and laid a large fine on the woman for non-appearance.28 The judges were therefore people who performed several functions, regardless of who may have been presiding over the court. They determined what should happen to take the case forward; they decided in effect who was right, if there was no confession; and they could pass sentence. Judges did sometimes preside in court, although more usually it was a count who did so, or sometimes the king, and occasionally a bishop or an abbot or a secular lord.29 The court holder seems to have chaired proceedings and he (or occasionally she) often took a fine.30 Nevertheless, fines could go to the saio as well and occasionally to the iudex: the large fine cited above which should have been paid by Velito and Calendo for ploughing monastic land was due to be split, half to the monastery and half to the potestas (local political authority) and the iudex.31

Who were the judges? Firstly numbers. I am aware of 160 named individuals other than presidents who judged, of whom all but 23 occurred between 912 and 1000 (the 23 are spread across the ninth century); the ninth-century proportion reflects the distribution of surviving charters.32 At least fifteen individuals appeared more than once, often occurring in different charter collections. Many more are mentioned, but not named, so these numbers are minima. The numbers of judges per case varied between one and twelve; there does not appear to be any significant change over time in this, but there is a suggestion of a regional difference: the larger numbers are more frequent in Galicia, the smaller on the meseta. Some of these judges were royal judges, explicitly termed iudex regis, the king’s judge. Three such – Abaiub, Abozekar and Bello – heard a property case in León in 958; four such – Gudesteo Didaz, Mito Arias, Pelayo Arvaldiz and Tructesindo Nantildiz – heard a property case in Galicia in 1000; and unnamed royal judges (iudices de reie) heard a case by the river Órbigo in 994.33 Others were royal companions: not termed iudex regis explicitly but often found in the king’s presence, accompanying him on business; such were Leander and Maurello in and around

27 28 29 30

Li192: ‘testes testificaverunt et iudices iudicaverunt’. S159. For more detail on presiding, see Davies, ‘Holding court’. Presidents were nearly always male, although one or two powerful royal women did perform this function. 31 OD21 (976). 32 For updated figures see Davies, Windows on Justice, p. 163, n. 43. 33 S159, Cel252, OD37.



León in the 930s and 940s; Gutier Osoriz, judging the freedmen case of 985 in Galicia and witnessing gifts near León in 980–6; Lope Jiménez, judging a church case in Pando (Castile) in 956 and witnessing royal gifts on the meseta in 945; Suario Gundemariz, judging another church case in Galicia in 992 and witnessing on the meseta in 986.34 There were also five counts, a saio and a notary (who, as iudex, was scribe of a royal donation made in León) who were involved in judging and were regular or intermittent royal companions.35 It is instructive to look at some of these in detail. Take the king’s judge Abaiub ibn Tebite, the most frequently recurring. He appears at least eighteen times in the period 921–59, and was judge at least five times between 938 and 959, in and around León. He witnessed sales to Sahagún, 50 km to the south east, and two royal gifts in the 930s and 940s; and he witnessed gifts and sales to the monasteries of Sahagún, Abellar and Santos Justo y Pastor, and to the bishop of León, in the 950s, often in and around León. It looks as if the Valderaduey valley, running south of Sahagún for about 30 km, was his home base: this is where he appeared at the outset and later, in 970, we find a Froila Abaiubiz, also iudex, witnessing gifts in that area – perhaps his son, apparently continuing in the same role.36 Or there is the judge Godino Zelimiz, who appears at least ten times between 968 and 977; he was called judge in half of these cases, although we only see him judging once, in 968, in the royal court near León. He witnessed gifts and sales, often in the presence of the king, either in and around León or in the same area as Abaiub in the Valderaduey valley, south of Sahagún.37 We also have in him a practical glimpse of the way the king might use his judges: in 977 a man called Rapinato killed a Sahagún monk when he was drunk (in fact breaking down the doors of a dependent church before doing so); Godino was sent by King Ramiro III and his mother, Teresa, to the murderer’s property, which lay in the Villa Pedro; then, the text says, following customary practice, Godino seized all his land on behalf of the king.38 Some other judges were ecclesiastical office holders: at least sixteen abbots, seven monks, four bishops, twelve priests and three deacons. Abbot Cisila was one of ten judges who judged a property case in northern Galicia in 942; Brother

34 Li128 (938), Li99 (934); SamS-8 (985), S307 (980), S313 (982); V33 (956), S98 and S99 (945); Sob130 (992), S333 (986). 35 Fofus saio, Sob129 (942); Frimunnio notarius, OvC34 (992). 36 Iudex and sometimes witness: Li128 (938), S88 (943), Li230 (950), Li253 (952), S145 (955), S146 (955), S147 (955), S159 (958), Lii312 (959); witness only: OvC22, S49, S80, S93, S95, S98, Li144, Lii263, Lii270 (921–54); Froila: S255. Cf. the comments of C. Estepa Díez, Estructura social de la ciudad de León (siglos XI–XIII) (León: Centro de estudios e investigación ‘San Isidoro’, Archivo Histórico Diocesano, Caja de Ahorros y Monte de Piedad de León, 1977), pp. 156–7, J. J. Sánchez Badiola, El Territorio de León en la edad media: poblamiento, organización del espacio y estructura social (siglos IX-XIII) 2 vols (León: Universidad de León, 2004), vol. 1, p. 460. 37 Iudex and sometimes witness: Lii410 (968), Lii412 (970), Lii413 (970), S255 (970), Lii434 (974); witness only: S268 (973), S269 (973), S272 (973), S284 (976). 38 S287.



Munnio, iudex, was the scribe who recorded a gift in Oncina in 990; Bishop Froarengo was the judge who set the fine for theft and adultery in 858; in 980 the priest Christofor witnessed a sale on the meseta as iudex; and the deacon Adaulfo was one of the four who determined boundaries in the Portuguese case of 936.39 Many of the remaining 52% appeared in this or that locality, without any stated royal or ecclesiastical associations. Flacino, for example, was a judge in and around Cofiñal (at the head of the arable on the upper reaches of the river Porma on the southern slopes of the Cantabrian Mountains), between 949 and 955. He also appears as witness to sales in that area from 930 and it seems to be his property in Noanca, just south of Cofiñal, that was mentioned in 962.40 Others include Alrameliz, in Grajal, in the Tierra de Campos, in the 980s, with property there; and Eita Godesteiz in the late 970s outside León, in Quintanella to the east, his heirs selling from there in the next generation.41 Always men, and appearing in limited regions, these seem to have been local people of some standing and skill. Beyond the clear indications that some were landed aristocrats, and that many were propertied, we cannot generalize about their socio-economic status. The above analysis includes all the named people who were called iudex in charter texts or who were described as having judged (qui iudicavit, etc.) in witness lists. A further twenty-one individuals, who presided over courts, are recorded as joining the judges in making judgment; these were all figures with political power – kings, counts and bishops. While the label iudex was attached to some of the judges – like Flacino iudex, Judge Flacino – it was not applied to all. The group doing the judging, the group of iudices in the plural, might include, or indeed be entirely composed of, individuals who carried the label iudex, but it might also include others. So, a case in León in 959 (a property dispute between two monasteries) was heard by a iudex and two priests; that of 958 between a woman and Sahagún was judged by three people termed iudex, a bishop and three priests, all together referred to as iudices (‘nos iam dictos iudices’); and in a property case in Galicia, heard before the iudex Diego, unnamed judges (iudices) determined the case.42 The apparent inconsistencies in this usage are quite easily explained: being called a iudex was a marker of status – the label was applied to such people when, for example, they witnessed uncontested sale transactions; to do the judging you did not need to be a iudex, although you might be; in other words, the label iudex and the act of judging are separable.43 A iudex (in the singular) was a person of special status and special skill – a kind of professional; he must usually have

39 Sob129, Liii541, Sob75, Lii478, PMH DC 42. 40 Iudex and sometimes witness: S115 (949), S116 (949), S148 (955); witness only: S41 (930), S57 (936), S92 (943), S106 (947), S117 and S118 (949); termino de Flacino: S198 (962). 41 S329 (986), S316 (983), cf. terminum Hali Rameliz; Lii458 (978), S341 (989). 42 Lii312, S159, Sam239 (985); cf. V11 (919). 43 Sánchez Badiola makes a comparable point, El Territorio de León, vol. 1, p. 460; compare the Lombard texts which refer to the duke or gastald as iudex and the functionaries carrying out judicial activities as iudices, Bougard, La justice dans le royaume d’Italie, p. 141.



been literate (given the number of cases in which a scribe is termed iudex) and he is likely to have known some law. Doing the judging was something in which other leading men of a locality could participate; hence the common references to iudices, in the plural, as the people doing the judging; in the case about control of freedmen in Galicia, the additional three judges were selected from the assembled court to probe the witness evidence.44 Very occasionally there are references to choosing the judges from assembled boni homines, that is ‘worthies’, although that is rare (and the texts do not specify who made the choice).45 Having identified the typical actions of the judge (ordering what happens next, making primary investigations, reviewing evidence and making decisions), we can see that sometimes precisely the same kinds of action were performed by people who were neither called iudex nor even noted as a group of iudices: in the boundary case of Rio Covo, in northern Portugal, in which the count and panel of three went to inspect the land and find the boundary markers, judgment words are not used of their actions and none of them is called a judge, even though the case began in the count’s court, the outcome was witnessed and they resolved the case in the characteristic manner of the courts. This is a useful warning that the language of the records is neither consistent nor stable: it allows us to see something of what happened but by no means all; indeed, we may be seeing only the tip of an iceberg.

What did they think they were doing? We can only investigate contemporaries’ approach to judgment, however, through the language of the records. Despite all the things we cannot know, the language is interesting. The preambles to charters (words of pious justification for the deeds recorded) and the sanction (i.e. penalty) clauses of charters – all charters, not just dispute records – frequently invoke images of judgment. The discourse is strongly biblical: the day of judgment – diem magni iuditii – occurs often. For example, in 993 a priest’s gift to a nunnery on the meseta has quite a long preamble including the words ‘paveo presentia Domini Ihesu Christi et diem magnum iudici’, ‘I quake in the presence of the Lord at the great day of judgment’, as also a sanction invoking the wrath of God and eternal damnation in the event of default.46 A lay couple’s gift to the monastery of Celanova in 969 includes, among the giving words of the disposition: ‘ut ad diem magni iuditii subveniat nobis’, ‘[we give this] so that at the day of judgment he can help us’.47 More curiously, we find in a record of a gift to a nunnery near Otero de las Dueñas in 988: ‘we give, as it is written, at the Day of the Holy Tremendous Judgment, as the Apostle 44 Sam S-8. 45 E.g. Cel502 (940). See the following, ch. 16, ‘Boni homines in northern Iberia’, for more detailed discussion. 46 Liii555. 47 Cel246.



said: Give Lord, because we give; we have done what you ordered; give us what you promised’.48 Rather less demanding is the comment, in the record of Bishop Diego’s gift to the church of Valpuesta, in northern Castile, in 929: ‘in the love of and with the inspiration of God, I decided to give for the salvation of my sins and on account of the day of judgment’.49 The Day of Judgment was a present thought, often inspiring fear. There are also variations on the theme of standing before the tribunal of the Lord – ante tribunal Dei/Domini: we are all going to stand before the Lord (ante Deum), and prayers should be said so that we do not stand condemned.50 In a sanction following the record of a gift we find: ‘if anyone breaks this, he shall return what is owed before the tribunal of our Lord (at the time of his terrible judgment)’.51 More gently, in a gift text of 943 to the monastery of Abellar, there occurs: ‘I have agreed to put myself into the hands of Abbot Severo so that he will lead me – with his own hands – before the tribunal of our Lord’.52 Similar ideas were used in different ways: the notion of the final judgment is insistently present, often terrifying, sometimes mitigated through an intercessor. Judgment was about being called to account. The tone is strongly penal, and condemnatory; it is the occasion for just deserts. So, we also find: ‘Domine, tu es iustus iudex’, a just judge, as in the Celanova foundation document of 936.53 Justice was about being fair: the person who did wrong deserved to be punished; the person who did well deserved to be rewarded. Hence ‘recognizing justice’ (cognoverunt se per iustitiam) is cited as the reason for being lenient; or discernentes iusticia, ‘perceiving justice’, when judges decided not to allow a monastery’s claim against peasants.54 The idea of judgment evokes both terror and fairness – equitable dealing – as when the ‘whole council’ recognized truth and justice in 952 and let the provisions of some original documents stand without challenge.55 In this last case, veritas, truth, is associated with justice, as it also was in my opening Portuguese example, in which the judges ended up by ‘finding the truth’. There are many similar statements: agnovit in veritate, connoscimus nos in veritate, cognovimus in veritatem.56 One charter even develops the concept in its witness list, noting alongside his confirmation that the king (Alfonso IV)

48 OD29: ‘sicut exeratum est, ad diem sancti tremendi iudicii, sicut Apostolus ait: “Da, Domine, quia dedimus; nos fecimus quod iusisti, rede nobis quod promisisti”’. 49 V12. 50 Cel36 (889). 51 S39 (930): ‘ut quisquis . . . hanc nostram voluerit in aliquo frangere devotionem nobiscum (ante) tribunal Domini nostri Ihesu Christi reddat in debita tempore terribili iudicii sui’. 52 Li177: ‘trado me in manus abbati . . . ut per suas manus ducat me ante tribunal Domini nostri Ihesu Christi’. 53 Cel256. 54 C22 (932), Li144 (941). 55 Li256. 56 PMH DC 163 (991), T30 (922), OD43 (997), Cel197 (975–1011); cf. Floriano 74 (861) and Ast5 (878).



recognized truth (‘Adefonsus princeps quod veritatem cognovimus manu propria confirmamus’); and in another record the judges commented to the winner of a case against locals: ‘you have the truth’ (vestra est veritas).57 There are many variations: agnicio veritatis facta (‘recognition of truth was made’), iudicassent veritatem (‘[so that] they should assess the truth’), deatis mici veridatem (‘you gave me truth’), quod veritas docet (‘as truth teaches’), secundum veritatem (‘according to the truth’, in effect, what was right), veridice profitentes (‘declaring truth’);58 and it was lex et veritas that set a fine on a couple guilty of theft.59 There is also: veritatem vidimus, ‘we saw the truth’, in a record whose witness list adds ‘who saw’ (quos vidi [i.e. qui vidit]) to those of the panel involved in the judgment, as a variant of the more usual qui iudicavit.60 Or there is previdissent veritatem (‘they foresaw truth’) in a property case; and many variations on seeing truth: videntes plena veritate, potuimus veritatem prospicere.61 Finding the truth was in part a process of examining evidence but in looking backwards it was also a kind of foresight: vidimus, previdimus, previsores. We might even suggest that there was a touch of divination in the process of arriving at truth.

Truth and justice What should we make of all of this? What is interesting, given that the state was undeveloped, is that there was a public system, from east to west, north to south, which had recognized procedures, experts, written law, officers, scales of penalty, counts and others with potestas (in these contexts, legitimate capacity to hold a court). There was a strong sense of the public, although differently conceptualized from either ancient or modern notions. There was certainly local participation in this process – but the role of the locals had more to do with providing evidence (especially through enquiries, though also through the evidence of tenants brought out by this or that monastery, sometimes in large numbers) than with participating in collective judgment; judgment was a matter for panels, for the most part, but panels of leading (and not necessarily local) men. There was plenty of compromise, and thereby negotiation; and, as we have seen, many cases ended up with a ‘confession’ by one party, and deals were done: the Portuguese case with which I began ends by Lovesindo being given another farm, in lieu of the claimed property, and twenty-eight solidi as well. His stand at the church seems to have had the desired effect. So what is truth in all this? It is the language of the regular chanting of the Psalms which informs much of the language of the records: ‘laetentur et exultant 57 Li89 (931), Cel94 (987). 58 S356 (998), PMH DC 183 (999), OD27 (987), S287 (977), Sob23 (949), Sob31 (951), S159 (958 – cf. OvC26 (953)). 59 T41 (952). 60 PMH DC 42 (936). 61 Sob129 (942), Sob130 (992), Cel502 (940).



gentes quoniam iudicas populos in aequitate’; ‘confige timore tuo carnes meas a iudiciis enim; tuis timui [ain] iudicium et iustitiam . . . Miserere mei secundum iudicium diligentium nomen tuum’.62 The ‘truth’ comments of these texts are less to do with abstract concepts of justice than with rhetoric, a rhetoric of arriving at a conclusion: they are introduced as the resolution of the case is recorded.63 Note the role of ‘truth and justice’ in the victory speech by Nick Griffin at Manchester in the United Kingdom on 7/8 July 2009, celebrating the success of the British National Party in elections to the European Parliament: ‘the waters of truth and justice and freedom are once again flowing over this country. It is a great victory’. The same rhetoric continues, even in the modern world, as do the semantics: just as texts like the 999 Portuguese case use veridicum and other texts use variations on verum dicere, in the English language we still use the word ‘verdict’ of jury decisions. I am reminded here of Tim Reuter’s comments on assemblies, assemblies in which disputes were settled, along with much other business. He stressed the importance of ritual in reaching closure, as well as the staging and the stage managers.64 The idea of staging is particularly helpful here. In the end, much of this Iberian court procedure was performance: the confessions are as staged as it is possible to be. Note how much was decided in advance: the evidence was reviewed and decisions were in effect taken before oaths were sworn, long before the formal conclusion. And yet the participants went through procedure after procedure; they uttered complex oaths over and over again; they went through many meetings; and – in the language of the records – they voiced a vision (or prevision) of truth as the closing rhetoric for ultimate conclusion. Of course, not everything was performance: deals were done, oaths were serious public business and occasionally even judges could be declared to be in error. In the town of Burgos, in Castile, in 972, before the count and the entire local council (omni concilio), a man and a priest confessed that in a previous case they had sworn a false oath, thereby ‘rendering the judges and the men who asserted the truth in error’.65 It was

62 Psalms 66: 5; 118: 120–1, 132 (Vulgate). Cf. Authorised Version 67: 4 ‘O let the nations be glad and sing for joy: for thou shalt judge the people righteously’; 119: 120–1, 132 ‘My flesh trembleth for fear of thee; and I am afraid of thy judgments. I have done judgment and justice . . . be merciful unto me as thou usest to do unto those that love thy name’. I am extremely grateful to Julia Barrow for pointing out the importance of celebration through psalms in churches in western Europe before celebration of the Mass became widespread; as also the large numbers of lower clergy in those churches; see her The Clergy in the Medieval World: Secular Clerics, their Families and Careers in North-Western Europe, c.800–c.1200 (Cambridge: Cambridge University Press, 2015), ch. 2. 63 Cf. Alfonso, ‘Judicial rhetoric and political legitimation’, p. 78, on the ‘presentation of judicial process as a process of truth’. 64 T. Reuter, ‘Assembly politics in western Europe from the eighth century to the twelfth’, in P. Linehan and J. L. Nelson (eds), The Medieval World (London: Routledge, 2001), pp. 432–50; rep. T. Reuter, Medieval Polities and Modern Mentalities, ed. J. L. Nelson (Cambridge: Cambridge University Press, 2006), pp. 193–216. 65 C151: ‘et fecimus iudices et homines qui verum auctorificabant fallaciosos’.



acknowledged that judges, and the process of judgment, could sometimes fail to close a case.66 But the significance of the meetings was not diminished by the fact of performance; public performance was essential, not only to inform the community but to sanction the outcome; performance had important social functions. And the particularity of the judges’ decisions was essential to the construction of the performance.67

66 Cf. the false and mendacious judges of Cel189, Cel190. 67 I am very grateful to all those who made this lecture a memorable occasion, including the lively audience. I am indebted to Isabel Alfonso, John Hudson and Susan Reynolds for their extremely helpful comments on subsequent written drafts.


16 B O N I H O M I N E S I N N O RT H E R N IBERIA A particularity that raises some general questions The relationship we suggested between boni homines and adstantes in the ‘Glossary’ of our book The Settlement of Disputes in Early Medieval Europe has made me increasingly uncomfortable.1 An adstans was a ‘free man present at a public tribunal, to give it public status’, while boni homines was ‘a term applied to adstantes or “law-worthy” males of legal competence’.2 The implication that all adstantes were boni homines is what makes me uncomfortable. Recent work on the appearance of boni homines, literally ‘good men’, in northern Iberian texts of the early middle ages suggests some conclusions that appear to contrast with those of the long historiography of boni homines in western Europe.3 Where Duby saw boni homines disappearing from the mid-tenth century, Iberian charters provide an increase of references in the late tenth century.4 Where Chris Wickham saw that the ‘concept of the public hung directly on the participation of “good men”, boni homines and their innumerable variants’, the concept of the public in Spain and Portugal had more to do with the power of inherited tradition and, indeed, is evident long before use of the term boni homines becomes common;5 and in some Iberian usages boni homines did not refer to the members of a judicial court but to those performing specific, not necessarily court-based, functions. Practice that was apparently different in Iberia raises the questions of whether or not Iberia was in fact significantly different from the rest of western Europe or whether we are simply observing some differences in the recording of procedures which were essentially the same. It is a particularly good time to consider these

1 First published in R. Balzaretti, J. Barrow, P. Skinner (eds), Italy and Early Medieval Europe: Papers for Chris Wickham (Oxford: Oxford University Press, 2018), pp. 60–72. 2 W. Davies and P. Fouracre (eds), The Settlement of Disputes in Early Medieval Europe (Cambridge: Cambridge University Press, 1986), p. 269. 3 For full discussion of Iberian examples, see W. Davies, Windows on Justice in Northern Iberia, 800–1000 (Abingdon: Routledge, 2016), ch. 6, which is summarized in the following. 4 G. Duby, ‘Recherches sur l’évolution des institutions judiciaires pendant le Xe et le XIe siècle dans le sud de la Bourgogne’, Le moyen âge, 52 (1946), pp. 149–94, and ibid., 53 (1947), pp. 15–38, at p. 152, but see further in what follows. 5 ‘Conclusion’, in Davies and Fouracre (eds), Settlement of Disputes, p. 231.



questions because a recent useful paper by Thomas Szabó on the history of boni homines does not take the Iberian evidence into account.6 In what follows I will summarize the incidence of the term boni homines in pre-eleventh-century Iberian charters, excluding Catalonia; explore comparabilities elsewhere in Europe (including Catalonia); then consider the issues raised and end with some questions to pursue.

Boni homines in Iberian texts Textual references to the actions of boni homines occur, as they do in most of western Europe, in charters. The corpus of surviving Iberian charters is rather late by European standards, there being about twenty from the eighth century, 200 from the ninth and 2,500 from the tenth.7 Boni homines have stimulated a significant Spanish historiography, scholars tending to emphasize the role of these people as mediators and their status as peasant élite.8 The term does not feature unambiguously in Iberian charters before the 940s. However, there are references to their role as evaluators of the worth of slaves in Visigothic law texts from the seventh century and it could be chance that there are no earlier charter references.9 Even when it does begin to feature, use of the term is rare, occurring in about 2 per cent of the corpus of charters. References clearly increased in the later tenth century, however: two fifths of pre-eleventh-century references come from the 990s and four fifths are noted from 960 onwards; there are many more references in the eleventh to thirteenth centuries in charters and in the several Spanish fueros.10 It is clear that some scribes, in some regions, either did not know the term or did not choose to use it: it is disproportionately common in Galician texts, in the west (nearly half of all cases), very rare in Asturian and Cantabrian texts, in the north, and rare in Castilian and Navarre texts, in the east; and although it is common enough on the central meseta, it is rare in Sahagún material.

6 T. Szabó, ‘Zur Geschichte der boni homines’, in D. Balestracci, A. Barlucchi, F. Franceschi, P. Nanni, G. Piccinni, A. Zorzi (eds), Uomini paesaggi storie: studi di storia medievale per Giovanni Cherubini, 2 vols (Siena: Salvietti & Barabuffi, 2012), vol. 1, pp. 301–22. 7 See Davies, Windows on Justice, ch. 1, for a full overview. 8 M. del C. Carlé, ‘Boni homines y hombres buenos’, Cuadernos de Historia de España, 39/40 (1964), pp. 133–68; P. Martínez Sopena, ‘La justicia en la época asturleonesa: entre el Liber y los mediadores sociales’, in A. Rodríguez (ed.), El lugar del campesino. En torno a la obra de Reyna Pastor (València/Madrid: Universitat de València, CSIC, 2007), pp. 239–60; J. M. Mínguez, ‘Justicia y poder en el marco de la feudalización de la sociedad leonesa’, in La giustizia nell’alto medioevo (secoli IX–XI), 2 vols, Settimane di studio del centro italiano di studi sull’alto medioevo, 44 (1997), vol. 1, pp. 491–546, at pp. 541–3; F. Luis Corral and M. Pérez Rodríguez, ‘Negotiating fines in the early middle ages: local communities, mediators and the instrumentalization of justice in the kingdom of León’, Al-Masaq, 2017. doi:10.1080/09503110.2017.1349980 9 Leges Visigothorum, ed. K. Zeumer, MGH LL nat. Germ. 1 (Hanover: Hahn, 1902), at VI.1.4, IX.1.21 and X.1.17 (pp. 253, 365, 389). 10 See Carlé, ‘Boni homines y hombres buenos’, pp. 144–7, 153.



In these texts the term boni homines is an expression used of people gathered together, nearly always – though not exclusively – in the plural and nearly always – though not entirely – men. However, the words are employed in three related but distinctly different ways. Firstly, they are used as a general term to refer to respectable people from whom oath-takers might be selected or in the presence of whom judicial proceedings were heard, like a confession made in 972 in the presence of Count García Fernández, the council of Burgos and multorum bonorum ominum, or a Galician property dispute of 995 in which King Vermudo II ordered one party to provide suum iuratorem hominem bonum to take the oath.11 In this usage the label seems to signify the presence of ‘acceptable people’, those of suitable standing to take part in judicial action, the label adding an aura of respectability and authority to the proceedings; in some contexts it refers to great men and in others to local worthies. On the meseta and in Galicia this general usage is often expressed as filii bonorum hominum, the sons of law-worthy men, where the primary meaning seems to be ‘other respectable people’ rather than specific descendants.12 It is also common, for example, for judges to be named and the clause to conclude with ‘and other boni homines’. All but one of these cases with individual names are Galician and all but one are recorded in retrospective narrative accounts of court cases. Disputed boundaries were also fixed by named boni homines in the Verín area of southern Galicia in the mid-tenth century, men who were not local but sent for the purpose from the count’s and the king’s courts.13 In a second usage, boni homines exercised the specific function of intercession in court (on which occasions they are never named). In such cases they would speak on behalf of an accused person in order to get a penalty reduced, like those who spoke for people who had debts to the aristocrat Fruela Vimáraz in the mid990s;14 or those who got the penalty for incursions on to monastic land in Valdoré moderated in 997;15 or, in a Portuguese case fifty years earlier, those who spoke for an acknowledged killer and persuaded the dominus Ansur to make up the deficit in the compensation due to the dead man’s family.16 This use of boni homines clearly did not refer to judges, since it was the judges or the court holders who were petitioned; they appear to have been other people who went to court. Except for the Portuguese case of 943, all the examples of intercession come from the meseta and from the 990s, and there are many more in the eleventh century.

11 C151; Cel215. 12 For example, a gift on the meseta done in the presence of filii multorum bonorum hominum (Lii490 (983)); cf. multi filii bene natorum, sons of the well-born, which is used in a similar way (LaC71 (965)). Cf. nobilibus atque bene natis viris in ‘Formulae Visigothicae’, 14, in Formulae Merowingici et Karolini Aevi, ed. K. Zeumer, MGH Leges (Hanover: Hahn, 1886), p. 581. 13 Cel502 (940), Cel93 (950). 14 OD37 (994), OD40 (995), OD41 (995). 15 OD43; cf. S358 (998). 16 PMH DC 53 (943).



This kind of reference in part simply reflects a developing diplomatic habit, for the phrase rogare cum bonis hominibus is a standard formula for describing a petition.17 But occasionally there is a sense of performance about the process: Salvador prostrated himself, with boni homines, in the king’s court before coming to an agreement.18 Given this performance, and since there were several ways of referring to the act of intercession, it looks as if the diplomatic habit in reality reflects a developing practice of intercession in the later tenth century. A third, quite different, usage is as follows: in certain circumstances it was boni homines who fixed a price for a transaction or who saw to the fair distribution of property. When half a property was sold in Portugal, the record of the sale provides that if the other half of the property were to be sold the boni homines would fix a just price; and in another Portuguese case, a sick woman was taken to the boni homines, who led her to the monastery of Guimarães where she gave instructions about what should happen to her property after her death.19 Several charters name boni homines who witnessed or corroborated apparently straightforward gifts and sales.20 All of these witness lists date from the mid-tenth century onwards and most from the late 970s onwards; two thirds of them relate to properties near León.21 There are even a few such ‘witness’ lists with no names at all, but just a reference to the good men present: ‘In collegium Sanctorum Iusti et Pastoris ubi fuerunt multorum filii bonorum hominum’, concluding a charter of sale to the monastery of Santos Justo y Pastor, whose scribes seem more prone to this habit than others;22 one might think this the abbreviation of a cartulary copyist but the format can be found on single sheets.23 This was clearly for the most part meseta practice (but nevertheless from a number of different recording sources). Why these scribes chose to emphasize the suitability of the witnesses is unclear, for it was much more common simply to say ‘and others’ in such circumstances, but it is of course possible that there was some hidden background of division and distribution concerning the properties transacted. One late tenth-century Galician charter records that, following proceedings in a case of theft, an estate valued at 100 modii was paid by the convicted thief, et asignamus ipsa ereditate pro apreciatura de omines bonos; in this court case the boni homines are explicitly associated with the valuation (apreciatura) of the 17 See G. Martínez Díez, ‘Terminología jurídica en la documentación del reino de León. Siglos IX–XI’, in Orígenes de las lenguas romances en el reino de León, Siglos IX–XII, 2 vols (León: Centro de estudios e investigación ‘San Isidoro’, 2004), vol. 1, pp. 229–72, at pp. 260, 261–3; Martínez Sopena, ‘La justicia en la época asturleonesa’, pp. 241, 245, 248. 18 Liii559 (993) – ‘prostrauit se Saluatorem ad rocum cum ommines bonos in ipso concelio de regem’. 19 PMH DC 79 (960) – ‘pro pretium iustum quod omines bonos apreciaberunt’; PMH DC 81 (960), cf. Lor17 (957). 20 For example, Liii551 (992), OD23 (978), A8 (?941), Leire 13 (997), AHN 1325A fo. 4v (987). 21 A8 is the earliest; not León: A8, C88 (955), Leire 13. 22 Liii547 (991); cf. Liii590 (999). 23 As Liii547; also OD23.



penalty.24 Just as most scribes referred to a collectivity of witnesses without calling them boni homines, and hundreds, if not thousands, of people witnessed transactions without being termed boni homines, it could well be that some scribes referred to the function of price-fixing without associating it with the distinctive term. When charters refer to a sale of property ‘priced at’ (adpreciatum) such and such a value, someone must have done the pricing. This kind of language is notable in charter collections which do not refer to boni homines, like those of Oviedo, Valpuesta and Santo Toribio, although it is not confined to them.25 One Celanova record of a gift includes a note on a debt owed; the witness list begins ‘Qui presentes fuerunt et ipsa uilla apreciauerunt’ (Those who were present and valued this property), and goes on to name fourteen individuals. Were these fourteen people men and women who might elsewhere have been described as boni homines?26 The implication of this set of material is that suitable people characteristically assisted in pricing property transactions, whether called boni homines or not; and, given the evaluation references of Leges Visigothorum, that the practice was of long standing. Since the number of references to boni homines is small, one cannot be sure that differences in distribution reflect any realities, but they are consistent. For what it is worth, peopling the judicial court was essentially Galician; interceding was essentially a feature of the meseta; being present at transactions and fixing prices occurred in many parts, although prominent on the meseta. One might think that references to boni homines were simply a way of saying ‘other people’, not only because numbers are small but because some references are vague. But reference to them cannot simply be dismissed as the odd practice of a few scribes or monasteries. They occur in texts referring to different regions and different locations within regions (Valdoré, Astorga, León, Santiago de Compostela, Santa María de Mezonzo, Verín and so on). They occur on single sheets as well as in cartulary copies; and the records were made by different kinds of named scribe, from scribes working for lay aristocratic households to local priests to smaller and larger monasteries.27 It was clearly a widespread habit. And although most references are framed in the context of a standard diplomatic, there are occasions when a scribe used the term in a non-standard note, like that attached to an exchange of 980 or the Portuguese cases of price-fixing and distribution;28 it is as if these were the words that increasingly came to mind in the late tenth century when describing useful and responsible collective action. But it was nevertheless a diplomatic habit, a way of writing – people

24 Cañ65 (996). Modii: a widely used standard measure of account; see ch. 6 above, ‘Sale, price and valuation’, at pp. 111–12. 25 Ov19 (978), V36 (960–1) – pretium quantu ipsa ualuit, for example; cf. T68 (963), per mesura. 26 Cel353 (999); the list includes two women. 27 For example, OD40 and OD41 (995); Liii516 (987); Lii445 (976); Liii559; Lii490 and Lii492 (983). 28 Lii477; Lor17; PMH DC 79 and 81.



performing exactly the same functions were sometimes called boni homines and sometimes not. Overall, then, the Iberian texts reflect some differences in function, which are on the whole differentiated by region: firstly, an increasing habit of referring to participants in court proceedings, especially high-level aristocratic proceedings in Galicia, as boni homines; this was essentially a diplomatic habit, a way of writing, invoking a sense of legitimacy surrounding proceedings. Secondly, an increasing habit of attributing the function of intercession in court to a distinctive group of boni homines; this was a diplomatic habit too, but one that seems to reflect changing practice of the late tenth century. And, thirdly, a substratum of references to the many people who assisted in the proper conduct of transactions, usually at local meetings. The import of the majority of references, and especially of participants in court proceedings, is that boni homines were far more likely to be aristocrats than peasant farmers;29 but for the ordinary business of day-to-day (or month-to-month) transactions boni homines must have come from a lower social level: they were presumably property holders of good local reputation – something closer to the ‘local worthy’ or ‘elder’ that one might expect – the people who came to local meetings.30 Many of the people who performed these functions, and who had been performing the functions for some considerable time, were not called boni homines: what the tenth-century texts seem to show, then, was an increasing tendency to classify those who took on such roles as ‘suitable’ and ‘qualified’. This suggests that there was a current of feeling by the year 1000 that sought to designate group presence and group action by reputable people as legitimate; by so doing in its turn it enhanced the legitimating function of the actions.

Boni homines elsewhere Boni homines was a term widely used in Europe in the early and central middle ages to refer to trusted people who were called to witness contracts, commitments and transactions or from whom judges were chosen, in other words, respected people (nearly always men) of good reputation. Boni homines, and variants like boni viri and other expressions invoking the notion of good faith, feature in early medieval texts from Brittany, Italy and the Frankish Empire, as well as Spain

29 See Davies, Windows on Justice, ch. 7, for investigation of the social status of those who went to meetings. 30 Cf. the note recording the presence of filii bonorum patres de Saresacensi uallem testibus, law-worthy people from the valley of Salazar (where the property was located), added to Leire 13; J. J. Larrea, ‘La infanzonía en una perspectiva comparada: infanzones y arimanni del ordenamiento público al feudal’, in P. Bonnassie (ed.), Fiefs et féodalité dans l’Europe méridionale (Italie, France du Midi, Péninsule ibérique) du Xe au XIIIe siècle (Toulouse: CNRS, Université de Toulouse-Le Mirail, 2002), pp. 363–96, at pp. 381–2, points out that there is a direct Basque equivalent of filii bonorum patrum.



and Portugal;31 while the terms occur frequently in charters, they also occur in Frankish formularies, Lombard laws, Carolingian capitularies and Lex Romana Curiensis, and they can be found from the sixth century until as late as the fourteenth century in some parts.32 The legal texts situate boni homines in public courts, sitting together with a count or judges, their presence noted in cases of homicide, of putting down securities, of taking evidence and of hearing oaths; and the formularies show them acting as guarantors for several kinds of private transaction, especially where they are drawing on a restricted range of individual formulas. They also supply the public context for acknowledging agreements between individual parties: the Angers formulary, the earliest of the Frankish formularies, possibly originally of late sixth-century date, provides for anyone who brought an unsuccessful case against someone else to meet before bonis hominibus and provide a written assurance that he would not take such action again;33 the laws of King Liutprand, of the early eighth century, provide that an agreement made before three or four homines boni should not be denied; Charlemagne’s capitulary on robbers provides for the evidence of (free) bonorum hominum to be heard or the judgment of God to be invoked.34 Invoking their presence and assistance at court cases or at less formal meetings was therefore part of a widespread practice in western Europe. Karin Nehlsen-von Stryk showed that the term boni viri, good citizens, had occurred in pre-fifth-century Roman texts as well as in the Theodosian Code, Justinian’s Novels and the Digest, in contexts, for example, of candidature for office, of assessing damage, of assessing values in the process of testamentary disposition, of setting allowances for minors, of proper conduct of defence in court, of providing reliable evidence;35 many of the jurists cited by the Digest used the notion of boni viri as the exemplar of responsible behaviour. Their appearance in the early sixth-century Edict of Theoderic, assessing the costs and expenses of those falsely accused, in an elaboration of 31 For comprehensive consideration of Frankish sources of the early middle ages, see K. Nehlsen-von Stryk, Die Boni Homines des frühen Mittelalters unter besonderer Berücksichtigung der fränkischen Quellen (Berlin: Duncker & Humblot, 1981); Szabó, ‘Zur Geschichte der boni homines‘, pp. 308–9, stresses the range of terms used in exchange charters from early medieval Lucca. 32 For occurrence in thirteenth-century southern Italy, see P. Skinner, Medieval Amalfi and Its Diaspora, 800–1250 (Oxford: Oxford University Press, 2013), p. 53; and in fourteenth-century Italy, C. Giardina, ‘I “Boni homines” in Italia: contributo alla storia delle persone e della procedura civile e al problema dell’origine del consolato’, Rivista di storia del diritto italiano, 5 (1932), pp. 28–95, 313–94, at pp. 386–9, who put the earliest medieval reference as Theoderic’s Edict (p. 392) – on which see n. 36 following. For formularies, A. Rio, Legal Practice and the Written Word in the Early Middle Ages: Frankish Formulae, c. 500–1000 (Cambridge: Cambridge University Press, 2009). 33 Formulae Merowingici et Karolini Aevi, ed. Zeumer, p. 6; for the date, see Rio, Legal Practice and the Written Word, pp. 67–80. 34 Leges Langobardorum, ed. F. Bluhme, MGH LL 4 (Hanover: Hahn, 1868), p. 110; Capitularia Regum Francorum, vol. 1, ed. A. Boretius, MGH Capit. (Hanover: Hahn, 1883), pp. 180–1. 35 Nehlsen-von Stryk, Die Boni Homines, pp. 258–85, for antique and late antique sources; cf. Szabó, ‘Zur Geschichte der boni homines’, pp. 315–17.



a provision of Ulpian, shows that this term was still current in legal circles in Italy c. 500.36 Nehlsen-von Stryk also showed the persistence of the notion that witnesses should be suitable, qualified – idonei – by tracking the excerpting and glossing of Roman legal texts in the early middle ages.37 As many people have noted, the concepts of suitable witnesses and boni homines come together in the law of the Lombard king Liutprand which deals with agreements witnessed by three or four boni homines, for the qualities that made a witness suitable are also specified in this clause.38 Boni homines in charters can be seen to act in several capacities. They attended judicial courts, acting as judges, witnesses, confirmers and advisors; they could be mediators and they sometimes functioned as intermediaries between ruling powers and local people.39 In ninth-century Brittany panels of judgment-finders, who demonstrably came from the immediate locality of a village-level court and from the peasant population, were sometimes referred to as boni viri, in charters that preserve local diplomatic practice.40 In ninth- and tenth-century Catalonia boni homines would be present at judicial hearings; they might accompany the president of a court and could act as witnesses; in the eleventh and twelfth centuries they could play a major part in extra-judicial resolution by suggesting compromises.41 Their presence is noted in the charters of southern Lombardy, where in due course they might facilitate the negotiation of a conbenientia between rival

36 ‘sub aestimatione scilicet iudicis, aut bonorum virorum ex delegatione noscentium’; the point is made by S. D. W. Lafferty, Law and Society in the Age of Theoderic the Great: A Study of the Edictum Theoderici (Cambridge: Cambridge University Press, 2013), pp. 76–8, 145, 271. Its occurrence in the Latin version of Justinian’s Novels suggests the same for Byzantine legal circles a couple of generations later. 37 Nehlsen-von Stryk, Die Boni Homines, pp. 274–85, 321, 346–8. 38 See Szabó, ‘Zur Geschichte der boni homines’, p. 311, for example, noting the relationship between this clause and the Epitome Juliani (a mid-sixth-century Latin summary of the Novels). Cf. the suitable witnesses (idonei testes) noted at Leges Visigothorum II.4.3 and II.5.16: in the former case, their essential qualities are specified as free status, good reputation and wealthy; and a Breton charter of 892 on what constituted the suitability of a witness who was to provide evidence: good behaviour, good character, an unbribable nature, with a knowledge of right and truth and with no tendency to swear false oaths, CR271. 39 Nehlsen-von Stryk, Die Boni Homines, pp. 108–67; Giardina, ‘I “Boni homines” in Italia’, pp. 57–95; F. Bougard, La justice dans le royaume d’Italie de la fin du VIIIe siècle au début du XIe siècle, Bibliothèque des écoles françaises d’Athènes et de Rome, 291 (Rome: École française de Rome, 1995), pp. 152–3, 206, 226, 242, 342, 344; J. Bowman, Shifting Landmarks: Property, Proof, and Dispute in Catalonia around the Year 1000 (Ithaca: Cornell University Press, 2004), pp. 112–13. 40 CR129 (pre-834), CR180 (843–9). 41 J. M. Salrach, Justícia i poder a Catalunya abans de l’any mil (Vic: Eumo, 2013), pp. 21, 29, 77, 174; Bowman, Shifting Landmarks, pp. 111–13; P. Benito i Monclús, ‘Marché foncier et besoin d’expertise dans la Catalogne des Xe–XIIe siècles: le rôle des boni homines comme estimateurs de biens’, in C. Denjean and L. Feller (eds), Expertise et valeur des choses au moyen âge: I le besoin d’expertise (Madrid: Casa de Velázquez, 2013), pp. 153–65, at p. 154.



parties.42 Their presence and actions in court cases in the kingdom of Italy is noted from the ninth century but increases from the late tenth century.43 Across southern France and Catalonia too references to boni homines as witnesses, mediators, intercessors increased from round about 1000: they were the people at court who might guarantee a judgment or even in some circumstances intervene after a judgment, rather than judge themselves.44 In other words, the presence of boni homines at judicial proceedings, as anticipated by the legal texts, extended into a range of practical functions in court and beyond. Those practical functions encompassed a range of public activities, not all explicitly judicial: in the mid-ninth-century Tours case discussed by Janet Nelson, boni homines were local men, the neighbours who could be asked about the background of a disputed property.45 Breton boni viri did more than act as members of judging panels: they were also the people who confirmed donations, like the nine who confirmed a gift in the county of Nantes in the mid-ninth century;46 here, as elsewhere, they could guarantee regular transactions. Boni homines carried out functions like this in western Europe from at least the seventh century: charters show them acting as estimatores – surveyors and valuers – in Catalonia and Italy as well as in Spain and Portugal.47 Parcels of land to be exchanged were measured by bonis idoneis homines (sic) in ninth- to eleventh-century Verona charters;48 boni homines walked the boundaries of property and assessed its value; they fixed a just price for portions of land that remained when properties were split; they estimated the value of damage. The message is consistently that these were

42 P. Delogu, ‘La giustizia nell’Italia meridionale longobarda’, in La giustizia nell’alto medioevo, vol. 1, pp. 257–308, at p. 294; cf. H. Taviani-Carozzi, La principauté lombarde de Salerne. IX –XIe siècle. Pouvoir et société en Italie lombarde méridionale, 2 vols (Rome: École française de Rome, 1991), vol. 1, pp. 605–6; and C. Wickham, ‘Comunidades rurales y señorío débil: el caso del norte de Italia, 1050–1250’, in I. Álvarez Borge (co-ord.), Comunidades locales y poderes feudales en la edad media (Logroño: Universidad de la Rioja, 2001), pp. 395–414, at p. 399, on intermediaries 1000–1250. 43 F. Bougard, ‘“Falsum falsorum judicum consilium”: l’écrit et la justice en Italie centro-septentrionale au XIe siècle’, Bibliothèque de l’école des chartes, 155 (1997), pp. 299–314, at pp. 301–4. For more examples, Giardina, ‘I “Boni homines” in Italia’, pp. 72–91. 44 M. Bourin, ‘Les boni homines de l’an mil’, in C. Gauvard (ed.), La justice en l’an mil (Paris: La documentation française, 2003), pp. 53–65, at pp. 58–61. 45 J. L. Nelson, ‘Dispute settlement in Carolingian West Francia’, in Davies and Fouracre (eds), Settlement of Disputes, pp. 45–64, at p. 56. 46 CR75 (854–67). 47 Salrach, Justícia i poder a Catalunya, p. 147; Benito i Monclús, ‘Marché foncier et besoin d’expertise’, pp. 156–64, demonstrates that in eleventh- and twelfth-century Catalonia boni homines (and others who are described by comparable terms) intervened to fix a just price in particular circumstances; Giardina, ‘I “Boni homines” in Italia’, pp. 62–71; Skinner, Medieval Amalfi, p. 54. Cf. CR250 (?834), a sale, with price valued at 14 solidi, confirmed by twenty-two boni viri. 48 I. Fees, ‘Tauschurkunden in Venedig und im Veneto’, in I. Fees and Ph. Depreux (eds), Tauschgeschäft und Tauschurkunde vom 8. bis zum 12. Jahrhundert: l’acte d’échange, du VIIIe au XIIe siècle, Archiv für Diplomatik Beiheft 13 (Köln: Böhlau, 2013), pp. 99–128, at p. 111.



respectable people, idonei, who performed a range of socially useful functions, whether judicial or transactional, for much of the middle ages.

The issues The issues raised at the start of this paper are essentially to do, firstly, with the chronology of explicit references – why do references to boni homines appear to decline in some parts of western Europe when they increase in others; secondly, with participation in judicial process in court; and thirdly with the additional perspective provided by the Iberian peninsula. Firstly, chronological issues. It is striking that although the weight of legal citations falls in the pre-tenth-century period, the frequency of charter references, in southern Europe especially, increased in the late tenth century and beyond, to such an extent that Patricia Skinner suggested that boni homines ‘emerged’ in southern Italy in the eleventh century.49 The idea that ‘good men’ were honourable trusted people acting in the public interest has its origins in antiquity, as demonstrated above.50 The earliest uses of boni homines in post-Roman texts already indicate the several contexts in which such action might be necessary – evaluating, witnessing, authorising, both within a judicial court and beyond; good faith, honourable action and good-quality evidence all feature – bone fidei homines, testimonium bonum.51 Use of the notion appears to have declined in northern Europe from the late tenth century, just as it became more common in southern Europe, that is in Italy, the Midi and the Iberian peninsula. Perhaps appropriately, Monique Bourin observed that its use was strongest in areas of Visigothic influence.52 Georges Duby’s indication of the fact that by the year 1000 references to boni homines in judicial courts had tended to peter out in the Frankish world has had a powerful influence on the historiography of public power in medieval Europe.53 Duby was in fact citing a paper of Ganshof’s, of 1928.54 Duby pointed out that this change was at one level just a matter of change in terminology, although his main point was that the great men who used to assist the count in court, specifically

49 Skinner, Medieval Amalfi, p. 53. Cf. Bourin, ‘Les boni homines de l’an mil’, p. 53; Benito i Monclús, ‘Marché foncier et besoin d’expertise’, pp. 156–61. 50 See above, nn. 35–7. Cf. Giardina, ‘I “Boni homines” in Italia’, pp. 392–4, citing boni viri in Cicero, and Nehlsen-von Stryk’s comment on this ‘republikanisches Bürgerideal’, Die Boni Homines, p. 258. 51 Nehlsen-von Stryk, Die Boni Homines, pp. 239–40. 52 Bourin, ‘Les boni homines de l’an mil’, p. 57. 53 See above, n. 4, for Duby. For the historiography, see T. N. Bisson, ‘The “feudal revolution”’, Past and Present, 142 (1994), pp. 6–42; S. D. White, ‘Tenth-century courts at Mâcon and the perils of structuralist history: Re-reading Burgundian judicial institutions’, in W. C. Brown and P. Górecki (eds), Conflict in Medieval Europe: Changing Perspectives on Society and Culture (Aldershot: Ashgate, 2003), pp. 37–68. 54 F. L. Ganshof, ‘Contribution à l’étude des origines des cours féodales en France’, Revue historique de droit français et étranger (1928), pp. 644–65.



in Burgundy, disappeared by the early eleventh century, to be replaced by the count’s fideles, signs of that change being evident from the mid-tenth century. His interest was not in boni homines as such, nor in their various functions, but in comital courts and scabini. Ganshof’s paper, which has a similar main message, is concerned with scabini and the mallus and he was essentially discussing the disappearance of scabini from the mid-tenth century onwards; he extended his comments from Burgundy to other parts of France. He barely mentioned boni homines and when he did he added in a footnote that it was unclear if the term referred to scabini or rather generally to notables.55 Both Ganshof and Duby were more interested in scabini than in boni homines but the combined citations of both of them nevertheless do suggest that the term boni homines dropped out of use in much of France. What clearly changed in France was the mode of describing those responsible for making decisions in the public court. Choosing to describe the actors in judicial or any other process as ‘good men’ was to endow their activity with a sense of legitimacy and to do so deliberately. Why this signal should have been dropped in northern Europe is a major question; that the signal was sustained and indeed extended in southern Europe has been effectively demonstrated by many scholars.56 That there is a contrast between north and south does therefore suggest some difference in attitudes to the legitimacy of public action. Indeed, legitimation is precisely the point that Chris Wickham made in discussing changes to public assembly in the central middle ages.57 Duby had emphasized that the Carolingian character of the count’s court changed completely in the early eleventh century.58 What changed, in France, according to Chris, was the size of the judicial assembly as well as the relationship of its members to the count. The contrast between north and south, then, lies in the changing character of judicial courts in the north in the tenth century, not in any change in the functions of boni homines there. What makes for further difference between north and south, however, is not simply that words disappear in the north and continue in the south but that the valuation function of boni homines continued widely in the south. This function had never been prominent in the north in any case, whereas it was characteristic of the south over many centuries. The combination, therefore, of two quite different contrasts provides the explanation for the temporal differences in the use of the term.

55 Ibid., p. 648. 56 See above, nn. 41–4. Not totally dropped in the north: the term gwyrda (‘good men’) is common in later medieval Welsh law texts, where it can be differentiated from hynefyddion ‘elders’, B. Smith, ‘Judgement under the Law of Wales’, Studia Celtica, 39 (2005), pp. 63–103, at pp. 65, 67. 57 C. Wickham, ‘Public court practice: The eighth and twelfth centuries compared’, in S. Esders (ed.), Rechtsverständnis und Konfliktbewältigung: gerichtliche und aussergerichtliche Strategien im Mittelalter (Köln: Böhlau, 2007), pp. 17–30; cf. T. Reuter, C. Wickham, T. N. Bisson, ‘Debate: The “feudal revolution”’, Past and Present, 155 (1997), pp. 177–225, at pp. 203–5 (Wickham). 58 Duby, ‘Recherches sur l’évolution des institutions judiciaires’, pp. 153–63.



Secondly, participation in judicial process in court. Many texts make it clear that people other than the major actors could be present in early medieval courts. These could be called adstantes, those present, standing by, or their presence could be described in other ways – qui adstabant, qui ibidem aderant, in circuitu sistentes, auditores, vel aliis pluribus circumstantibus.59 Chris recently commented that ‘The adstantes were there to give a public authority to legal decisions which were not taken by them. They had importance as a local political aggregation; they made things public’;60 these adstantes are precisely the members of the large public assembly that conferred legitimacy on the assembly, according to the Wickham model. Now, the notion behind such expressions is to do with presence, being there. This is quite different from the notion behind boni homines, which is to do with the personal qualities of the individuals – the good citizen, the reliable person who will undertake a public duty. While an adstans might well be a bonus homo, and vice-versa (as in the Angers formulary example cited above), it was not necessarily so. Indeed, it is perfectly clear that those present at public meetings were not necessarily all drawn from the reliable – some meetings were attended by people identified as decidedly unreliable.61 Moreover, it is clear from the evaluation material that boni homines could act outside the context of a public meeting: use of the term is neither confined to judicial assemblies nor public meetings. Adstantes and boni homines, then, are different kinds of term. Our comment in Settlement of Disputes that boni homines was a term applied to adstantes was therefore misleading. However, this does not mean that there were no adstantes, no free men present at public tribunals. The Wickham model of the ‘local political aggregation’ remains reasonable (and it does not need citation of the word adstantes to make the point), for it was the presence of the free population, not the qualities of boni homines, that conferred authority. In the Wickham model the change in the character and size of the assembly that happened in France from the mid-tenth century extended to Italy c. 1100 and to most other parts of western Europe (England excluded) in the twelfth century.62 In other words, in the long term Italy saw the same kind of change in court practice as did France. But despite this, use of the term boni homines in Italy continued

59 Nehlsen-von Stryk, Die Boni Homines, pp. 91, 135–9. 60 C. Wickham, ‘Consensus and assemblies in the Romano-Germanic kingdoms: A comparative approach’, in V. Epp and C. H. F. Meyer (eds), Recht und Konsens im frühen Mittelalter, Vorträge und Forschungen, 82 (Ostfildern: Thorbecke, 2017), pp. 389–425. 61 For example, those who lied in court, e.g. CR195 (841), C151 (972), Cel189 (c. 1000), Cel190 (1000). 62 Wickham, ‘Public court practice’, and Wickham, ‘Justice in the kingdom of Italy in the eleventh century’, in La giustizia nell’alto medioevo, vol. 1, pp. 179–250, at pp. 237–9, 249–50. S. Reynolds, Kingdoms and Communities in Western Europe, 900–1300 (Oxford: Oxford University Press, 1984), argued strongly for the continuity of collective participation in judicial process in the kingdoms of England, France, Germany and Italy right across the central middle ages. I have argued that this was not the case in northern Iberia, since I do not see the evidence for large judicial assemblies – as a norm – in the ninth or tenth centuries, Windows on Justice, ch. 9.



and extended, where it disappeared in France. This must suggest that the forum for legitimate public action by good citizens was wider than the judicial court and public assembly. And indeed, since the large public assembly declined, it looks as if it was the good citizenship of individuals that came to be seen to confer legitimation rather than the attendance of the free male population; as Chris commented, for different reasons, ‘the history of the end of the placitum [was] one of changes in public legitimation’.63 What happened to the use of the term boni homines belongs with the history of changing public legitimation. Before turning to the Iberian perspective, it is worth devoting a few words to the Breton. In the examples cited above, it may be noted that the preferred Breton term was not the common medieval boni homines but the more characteristically antique boni viri.64 This makes sense in the light of the close relationship between formulas used in Redon charters from eastern Brittany and those of the formulary of Tours.65 As Szabó pointed out, the Tours formulary, unusually, uses both boni viri and boni homines.66 As recent work by Alice Rio and others has made clear, the Tours formulary exists in many manuscripts and many forms, so it cannot offer a snapshot of the ‘bridge’ whereby usage shifted from one term to the other, as Szabó suggested.67 What is important, however, is that since the formulary reflects diplomatic practice of the sixth and seventh centuries, the ninth-century Breton echoes of this formulation suggest some continuation of that diplomatic tradition into the ninth century; it is particularly important that the terms occur in charters deriving from east Breton practice that was local – it was not an importation associated with any new monastic foundation. It also suggests that the well evidenced judging and confirming functions of good men in this region in the ninth century derived from practices current in late antiquity and not from any new development. As for the Iberian perspective: the references to boni homines in many different northern Iberian contexts of the early middle ages, performing precisely the same functions as performed in early medieval Italy and the Midi, do not make Iberia look very different; they must ultimately derive from the same late antique origins. It is unthinkable that people in Spain and Portugal independently invented the same words and the same range of functions. The valuation function, especially, was a widespread local practice and did not derive from the requirements of 63 Wickham, ‘Public court practice’, p. 28. 64 See text and references cited above, nn. 35, 40, 46; Szabó, ‘Zur Geschichte der boni homines’, pp. 314–17, on the antiquity of the term. 65 See W. Davies, Small Worlds: The Village Community in Early Medieval Brittany (London: Duckworth, 1988), pp. 136–7; W. Davies, ‘The composition of the Redon cartulary’, Francia, 17 (1990), pp. 69–90, at pp. 71–4, 77, 82. 66 Szabó, ‘Zur Geschichte der boni homines’, p. 315. 67 Rio, Legal Practice and the Written Word, pp. 112–17; cf. Ph. Depreux, ‘La tradition manuscrite des “Formules de Tours” et la diffusion des modèles d’actes aux VIIIe et IXe siècles’, in Ph. Depreux and B. Judic (eds), Alcuin de York à Tours: écriture, pouvoir et réseaux dans l’Europe du haut moyen âge = Annales de Bretagne et des pays de l’ouest, 111 (2004), pp. 55–71.



any governing authority. That must suggest that adoption of the term boni homines did not reflect new practice in the ninth and tenth centuries but some kind of continuation of earlier practice – though no doubt with many elements of change since late antiquity.68 Secondly, if the best explanation for increasing references to the activity of boni homines in judicial courts in northern Iberia in the late tenth century is an increasing desire to endow those activities with a sense of legitimacy, then this must suggest a similar desire for legitimation elsewhere in southern Europe. As it happens, a shift from the expectation that it was the large public assembly that conferred legitimacy to an expectation that it was the good citizenship of individuals that conferred it fits rather neatly into this scenario. This therefore leaves us with some difficult but interesting questions. Why should the legitimating actions of good citizens have been so much less evident in northern Europe from c. 1000? Are there other words that convey the notion? Was this change simply about the disappearance of Carolingian terminology? Or could the pervasive legacy of Roman law – in all its many manifestations, formal and informal, through exerpting, glossing and reformulating – have been a factor behind the similarities of the south? Why is there so little evidence of the valuation function of boni homines in northern Europe? How far were there some real differences in practice as well as in words?

68 Pace Szabó, ‘Zur Geschichte der boni homines’, p. 315.



This paper is about personal sureties – individuals who played a part in agreements between two other parties and who thereby guaranteed contracts, payment of debts, good behaviour, and so on, for other people.1 In formal terms, personal suretyship was ‘une institution par laquelle un tiers garantit à une personne qu’une autre personne fera ou ne fera pas quelque chose à son bénéfice ou à son détriment’.2 In practice sureties operated both with reference to wrongs which had to be put right and to duties owed as the result of undertakings or agreements.3 Systems of suretyship are common in many parts of the world – Europe, America, Africa, Asia – and are common across a very long period, featuring in ancient societies and classical antiquity as also in our contemporary world: still today parents will guarantee a mortgage for their children, as states can decide to guarantee the solvency of banks.4 Sureties were also common in many parts of early medieval Europe – Merovingian and Carolingian Francia, Lombard and

1 First published in J. Escalona and A. Reynolds (eds), Scale and Scale Change: Western Europe in the First Millennium (Turnhout: Brepols, 2011), pp. 133–52. 2 J. Gilissen, ‘Esquisse d’une histoire comparée des sûretés personnelles’, in Les sûretés personnelles. Première partie. Synthèse générale; civilisations archaïques, antiques, islamiques et orientales, Recueils de la Société Jean Bodin pour l’histoire comparative des institutions, 28 (1974), pp. 5–127, at p. 29; adapted by D. B. Walters as ‘the fulfilment or performance of expectations arising out of someone’s undertaking to transfer something, to do some act, or to refrain from some action for the benefit of some other or others’, in his ‘The general features of archaic European suretyship’, in T. M. Charles-Edwards, M. E. Owen, D. B. Walters (eds), Lawyers and Laymen: Studies in the History of Law Presented to Professor Dafydd Jenkins on his Seventy-Fifth Birthday. Gŵyl Ddewi 1986 (Cardiff: University of Wales Press, 1986), pp. 92–116, at pp. 93, 109. 3 Gilissen, ‘Esquisse d’une histoire’, pp. 97–8. 4 See the overview by Gilissen, ‘Esquisse d’une histoire’, and the three Jean Bodin volumes devoted to personal suretyship, Les sûretés personnelles: Première partie; Deuxième partie. Moyen âge et temps modernes, Recueils de la Société Jean Bodin, 29 (1971); Troisième partie. Période contemporaine, ibid., 30 (1969). For sureties in ancient and modern Bedouin law, see F. H. Stewart, ‘Schuld und Haftung in Bedouin law’, Zeitschrift der Savigny-Stiftung für Rechtsgeschichte. Germanistische Abteilung, 107 (1990), pp. 393–407.



Byzantine Italy, Anglo-Saxon England, and, it would seem, Ireland.5 In AngloSaxon England, for example, tenth- and early eleventh-century laws are precise about the need for guarantors for transactions: sales and exchanges had to have sureties as well as witnesses; suspects needed to have sureties for their appearance to answer charges; his surety was to pay if an accused man fled; and family members had to guarantee the future good behaviour of known thieves.6 The fact that we find sureties appearing in early medieval Iberian contexts is therefore neither surprising nor unusual. Suretyship is at one level a technical, legal, subject, but it has an important wider social and political significance. Using sureties was a way of getting things done, and making sure that they were done. This was especially important in societies where there was little or no state presence, because in those cases these non-state mechanisms of suretyship could be the only way to ensure an agreed outcome, especially if lordship was weak (although of course sureties can also play an important role in developed states like the first Indian empire, the Roman Empire, Tang-dynasty China, or our own). Suretyship could operate in the private sphere of relationships between individuals: two independent parties could agree to use a surety, whose identity was acceptable to both, without any further enforcement agency being involved; but personal suretyship could also be used by lords, rulers, and governments, adopted into systems of regulation and control, as in the Anglo-Saxon case. That being so, it becomes relevant to issues of scale and scale change:7 as lordly power, and state power and organization, increased, did those changing power structures affect the character of suretyship mechanisms? In what follows, I will sketch the operation of suretyship in early 5 For example, Marculf’s formulary in Formulae Merowingici et Karolini Aevi, ed. K. Zeumer, MGH Leges (Hanover: Hahn, 1886), pp. 32–106, at pp. 60, 67; cf. pp. 88, 92; Die Urkunden Pippins, Karlmanns und Karls des Grossen, ed. E. Mühlbacher, MGH DD Kar. I (Hanover: Hahn, 1906), p. 199; cf. A. Vandenbossche, ‘Le cautionnement en Gaule dans les droits franc et burgonde’, in Les sûretés personnelles II, pp. 27–33. J.-O. Tjäder, Die nichtliterarischen lateinischen papyri Italiens aus der Zeit 445–700, 3 vols (Lund: Gleerup, 1954–82), vol. 1, p. 232; Leges Langobardorum, ed. F. Bluhme and A. Boretius, MGH LL 4 (Hanover: Hahn, 1868), pp. 46–7, 125–6, 161. F. Kelly, A Guide to Early Irish Law (Dublin: Dublin Institute for Advanced Studies, 1988), pp. 167–73. Sureties are frequently referenced in Irish law texts, although it is extremely difficult to find examples of their use in practice; sureties (rata) do, however, feature in the witness lists of the eleventh-century charters written into the Book of Kells: Notitiae as Leabhar Cheanannais 1033–1161, ed. G. Mac Niocaill (Dublin: Cló Morainn, 1961), pp. 12–16; and their existence is implied by references to ratas in some of the eighth-century Irish material in the canonical collection, R. Flechner, The Hibernensis: Book 1: A Study and Edition (Washington, DC: The Catholic University of America Press, 2019), e.g. ch. 33. 6 Die Gesetze der Angelsachsen, ed. F. Liebermann, 3 vols (Halle: Max Niemeyer, 1903–16), e.g. vol. 1, pp. 154, 190, 202, 218, 230, 334, 336, but throughout tenth- and early eleventh-century legislation; II Athelstan 1.3 and 6.1, III Edmund 7.1, III Edgar 6, I Aethelred 1.7, III Aethelred 6.2, II Cnut 30, 33, especially. 7 For influential theories of scale change, see J. Escalona, ‘The early middle ages: A scale-based approach’, in Escalona and Reynolds (eds), Scale and Scale Change, pp. 9–30.



medieval Brittany; then I will consider the incidence and character of suretyship in tenth-century northern Iberia; and finally I will make some observations about its relationship to issues of scale.

Suretyship in eastern Brittany In order to give Iberian suretyship some early medieval context, let me begin by indicating how suretyship functioned in another early medieval society, that of eastern Brittany. This region is particularly useful as a model not only because it is well documented but because suretyship was used to guarantee personal obligations without any oversight by agencies of state. In lawyers’ terms, the mechanism was private.8 There is a good corpus of material from two generations during the ninth century (830s–870s), located in the Oust-Vilaine watershed (there are very few tenth-century charters to compare). Charters from that area include 292 citations of sureties; indeed, sureties feature in 25% of all ninth- and tenth-century charters and the corpus is unusually rich in local and circumstantial detail.9 Two words for surety are common in these texts: Latin fideiussor and, less frequently, a Latinized vernacular word dilisidus; the term securator was also used three times.10 All three terms seem to refer to the same set of functions. Sureties were used in transactions of sale and ‘pledge’ or gage (in other words, using property as collateral for a loan – peasants, for example, handed over parcels of land, typically to the local priest, in order to raise cash; if they repaid the cash, the property was returned; if they did not, the property was not returned). Sureties were not used in transactions of gift. They were therefore used when there was a price (and in fact the price was normally expressed, and as far as we can see paid, in money: coin was available). So, to take one example: a peasant sold a property in the village of Ruffiac to the local blacksmith and his wife on 1 March 847 for 17 solidi and 4 denarii, and assigned three named sureties from that village for the security of the transfer; this was done on the land in question, before many witnesses.11 Sureties were also used in the settlement of disputes, especially where they were formally settled in a local court (usually a village court, not an arm of the state); hence, they guaranteed payment of agreed rents, or guaranteed that an unsuccessful party and his family would not bring another case, or guaranteed that both parties would stick to an agreed boundary line.12

8 Cf. Gilissen, ‘Esquisse d’une histoire’, p. 6: personal suretyship belongs to the domain of private law. 9 For detail of the charters, and of sureties, see W. Davies, ‘Suretyship in the Cartulaire de Redon’, in Charles-Edwards, Owen, Walters (eds), Lawyers and Laymen, pp. 72–91. 10 Ibid., pp. 74–5. 11 CR64. The solidus was the key unit of account of the reformed Carolingian silver currency, and silver denarii were coined at the rate of 12 to the solidus. 12 For example, CR134, CR29, CR47.



Explicit rulings on how, precisely, the sureties were to execute their duties are rare but there are plenty of practical examples of what sureties did in practice: sureties might turn up in court and give especially valued evidence if a case arose subsequent to the agreement; if there were problems with completion of an agreement, sureties could pay up from their own resources to make good any deficiencies in payments; in other cases, they enforced the terms of a judgment – like modern bailiffs, they forcefully took movables from those who did not meet their agreed obligations. Sometimes, in effect, their suretyship was a public acknowledgement that they had resigned any interest of their own (especially if they were related to the party with obligations). There is also one unique case, that of a cleric called Anau who had assaulted a priest: the final agreement included compensation but also provided that the sureties should pursue Anau to death if he should offend again and should give his worth to the abbot of Redon – a very private mechanism for dealing with an offence of that kind.13 What kinds of people served as sureties in ninth-century Brittany? They were for the most part free, male, respected peasants – solid citizens of the rural community; usually they had some property – say 10 or 20 hectares – but they were not the most wealthy in the village and usually they were not the most prominent peasants. They were ordinary free people, acting for each other and for family; they were not, of course, serfs. They were also very local people: Augan people acted for Augan transactions, Carentoir people for Carentoir transactions, and so on; they virtually always came from the village where the party who had named them lived and where any property in question lay. Indeed, 90% of sureties are only ever recorded, in any capacity, in a single village; and the remaining 10% are found in a neighbouring village, but no more distant place (apart from the exceptional case of two troublesome aristocrats, the brothers Ratfred and Ratvili).14 Normally sureties were people who did not travel far.15 So, the scale of trust here, and the scale of reputation, was very limited. The

13 CR202. 14 The brothers came from a family of machtierns; machtierns were minor aristocrats heavily involved in peasant culture; see W. Davies, Small Worlds: The Village Community in Early Medieval Brittany (London: Duckworth, 1988), pp. 175–83. Ratfred and Ratvili were frequently in conflict with the abbot of Redon, who had probably displaced them from much of their expected inheritance (their father had given the original land to set up the monastery); unlike most machtierns, they ranged far and wide and did sometimes attend the ruler’s court. Following one dispute with the abbot, Ratvili gave two sureties that he would in effect resign his interest in the abbot’s portion of a property on the river Chère (CR215). In the completely exceptional case of CR202 just noted, both brothers were among the sureties whom Anau named for his future good behaviour, in addition to the local men who were sureties for the vineyard he handed over; very unusually, the terms were settled at Redon itself; presumably these powerful men of the Redon locality were drafted in to give some weight to the restraint of Anau. All of this is exceptional, but it does show that aristocrats could be involved in the mechanisms of suretyship in this culture. 15 For further detail see Davies, Small Worlds, pp. 126–9.



scale of the arena in which public business was conducted was that of the village and all its lands – an area of 40–50 km2 on average, although it could be as small as 10–15 km2 or exceptionally more than 100 km2; hence it was usually 6 or 7 km from one side to the other – one and a half hours walk – and exceptionally 3–4 km or 12–13 km.16 This analysis of suretyship in ninth-century eastern Brittany could be undertaken because the surviving charters happen to provide dense coverage of a very limited area (see figure 17.1). Many transactions were recorded from some villages; many different individuals were named, and many of these featured again and again; and there was much detail of locations, including detail of the places where transactions were performed or courts were held: this could be on land in question, in or outside churches, in village centres, occasionally in a private house.




B R I T TA N Y R. Vilaine


Plélan Augan

Maure Guer Carentoir ES Ruffiac DE Sixt LA Renac Langon Peillac NV Plaz Bains AU Avessac X Vannes REDON



Formal courts Other settlements

Plessé N


R. Loir

Nantes 0 0

10 10


20 miles 30km

Figure 17.1 Local courts in eastern Brittany

16 In most cases from periphery to the village (plebs) church, which was often focal, was within an hour’s walk; see further, Davies, Small Worlds, pp. 105–26.



Personal suretyship in northern Iberia Now, the Breton material is a very compact corpus, from a confined area, and from essentially two generations in the mid-ninth century. It relates overwhelmingly to the transactions of close-knit peasant communities; aristocrats appear in the corpus but we do not often see them using suretyship mechanisms. There is considerably more charter material from northern Iberia at the closest point of comparison (the late ninth and – overwhelmingly – the tenth century); that material is much more varied in character, covering a wider range of social contexts; and it covers a much wider area: there are charter collections from northern Portugal and Galicia,17 from the meseta north of the Duero,18 from Castile,19 a little from Navarre and Aragón,20 and some material from north of the Cantabrian mountains (see figure 7.1).21 The words used for ‘personal surety’ in these charters are fideiussor, as in Brittany, and, increasingly, fidiator; the verb fidiare (to guarantee) also features, as do fidiatura and fiatura for the actions of guaranteeing; classical pignus and caut(i)o also occur for gage or pledge, that is something put down to provide a guarantee – in technical terms, a ‘real’ as opposed to a ‘personal’ security.22 Castile provides the greatest body of evidence of sureties and no fewer than two fifths of tenth-century references derive from that region; there are also smaller but significant proportions from the western meseta (31%) and from Galicia and Portugal (26%); there is a little from the Liébana; and nothing from the heartland of Navarre nor from Aragón.23 The number of cases is, however, small (35) and they only occur in 1.5% of ninth- and tenth-century charter texts. It is therefore difficult to make meaningful comparisons between individual collections; however, references in Valpuesta material are extremely common at 20%, in Sobrado material are quite common at 4%, in Portuguese material are rare at 0.5%, while there seem to be none in Samos material. The references occur from the mid-ninth century, but numbers increase from the 930s (all the pre-930s cases come from Valpuesta or Sobrado). More than half come from the period 940s to 970s inclusive. That is not necessarily significant, because it reflects the overall distribution 17 18 19 20

PMH DC, Cel, Sam, Sob. Li, Lii, Liii, S, OD. C, SM, A; V. Leire, SJP. Both San Millán and Albelda collections (SM, A) cover material in the orbit of Navarre as well as Castile. 21 Principally Ov, T. Catalonia is not included in this study; if included it would more than double the corpus. 22 In V43 (?968) cauto could indicate the promise guaranteed by sureties, rather than an object put down; in C210 (997) the cauto is a fine payable to the king in case of default. For real security see Gilissen, ‘Esquisse d’une histoire’, pp. 87–94, and note his point that real and personal can in practice be difficult to distinguish, ibid., pp. 24–5. There is a substantial treatment of medieval real security in Spain by J. Orlandis Rovira, ‘La prenda como procedimiento coactivo en nuestro derecho medieval’, Anuario de Historia del Derecho Español, 14 (1943), pp. 81–183, although the weight of his evidence falls in the eleventh to thirteenth centuries. 23 It could be argued that San Millán and Albelda examples came from Navarre.



of charters across the tenth century, numbers rising to a peak in mid-century and declining thereafter. What is significant is that there are relatively many from the 990s (17% of references): given that total numbers of charters have declined significantly by then (the decade has about 9% of tenth-century charters), references to sureties are relatively more frequent in that decade. There is nothing to suggest that their functions were declining. Why and how were personal sureties used in Spain and Portugal? Their applications include guaranteeing sales, gifts, and exchanges; guaranteeing the payment of rent and debts; securing agreements; and securing both the procedure and the outcome of court cases. Of these, the commonest were guaranteeing sales (29% of cases) and guaranteeing the procedure or outcome of court cases (37%). There is a considerable amount of anecdotal evidence and this often allows us to see precisely what was expected of the surety. When sureties were used in sales, and occasionally exchanges, several kinds of action might ensue. Sometimes they paid up, from their own resources, if the transaction failed, or they might appear in court and give a specially valued kind of evidence. So for example, when Fernando, his wife, and his son sold a water channel from the river Cea to the monastery of Saelices, the surety had to pay 100 solidi (a very considerable sum) if something went wrong with the sale. In an exchange between lay people in 980, sureties were appointed to make sure that one party did not pass on the exchanged property to anyone else; if they did, the sureties were to provide goods to the value of 30 solidi (and the defaulting party was liable to pay compensation to the value of 60 solidi).24 In these cases the surety was sometimes explicitly the person who effected the sanction, implementing the standard penalty clause. Hence, in transactions of 955 and 977, the surety was to make a payment to the value of 100 solidi if something went wrong with the transaction, as in the Cea case; and in the records of Albelda and León sales, both of 944, the surety clause is embedded in the sanction formulas.25 On the other hand, the sureties might have a practical function and physically transmit the purchase price in a sale transaction: two sureties for parcels of property near Braga, in northern Portugal, sold by one lay couple to another in 965, carried the payment (cider, millet, rye, and some wine) from the purchasers to the vendors.26 In other cases, they might simply commit themselves or others to making no future claim on property which had been transferred – a simple kind of statement but an important guarantee that, for example, an alienator’s family would not claim the property back. One surety from Huércanos, in the Rioja, guaranteed that surviving 24 S73 (939); Lii477. For solidi, see above n. 11; the solidus was also a major unit of account in northern Spain, but silver denarii do not appear to have been coined there, although it is neither impossible that Frankish denarii were sometimes available nor that small measures of silver of consistent weight were in use; see further, ch. 6 above, ‘Sale, price and valuation’, pp. 106–9. 25 S366 (a case of guaranteeing rent), S289; A10, Li184; but cf. Liii590 (999), a promise to pay the sanction directly without using a surety. 26 PMH DC 91.



daughters would not have any future claims.27 There are both peasant and aristocratic cases of the use of sureties in sales and exchanges; they occur early in the century, mid-century, and late; they occur in all regions; and they occur in transactions between laity, as also those between laity and a church body. This was clearly a standard practice, and was widely used, although it was obviously not used for all sales: the vast majority of records of sale do not mention sureties. They must have been used in cases where there was some reason for doubt, and a need for extra guarantees, and hence in cases where there was dispute potential.28 Sureties were occasionally used to guarantee gift transactions; this only occurred in the mid-tenth century and in Castile, and only occurred in the case of gifts to monasteries, like the gift of an estate from Vermudo Gudéstioz to Cardeña in 947.29 It looks as if this gift may have followed a court case, and hence the surety function is standard. The other two cases are both exceptional: sureties were used to guarantee the gift of a man to San Millán de la Cogolla – in effect a guarantee that the man would pay his rent to the monastery, and really a case of guaranteeing future obligation, another standard surety function; and sureties were used to provide guarantees that colleagues and neighbours would not reclaim their contributions to a monk’s burial expenses in 939.30 When sureties were used to guarantee payment of rent or clearing of debts, characteristic peasant to lord transactions, the surety paid up from personal resources if the tenant defaulted or, more commonly, made up the deficit by assigning some of his or her own family property to the landlord. Among the many transactions noted in composite documents from late tenth-century Celanova, a peasant couple from Orga, about 3 km east of the monastery, named a surety for payment of a debt of 72 modii due in bread and wine.31 Guarantors might also secure other types of agreement, like assuring the good behaviour of a dubious character. In a case from northern Castile, two guarantors were appointed that a man called Vermudo would in future honour his obligations; if he failed to do this, then the sureties were to give cattle to the abbot of Valpuesta.32 Sureties were used in court cases in a number of different ways, and featured both in property disputes and in cases of offence against individuals. The surety might be the person who handed over disputed property when a decision was found, as the surety Gelmiro handed over the Villa Pedro to the monastery of Sahagún in 998; or he or she might be the person who forced an accused person to court to answer a charge, like the classic Irish ‘enforcing surety’, as the Spaniard

27 28 29 30

A10 (944). Compare the provisions of Visigothic law, following, n. 57. C59. SM93 (975), V17. It was more usual to use a countergift if guarantees were needed to secure gifts; see ch. 8 above, ‘Countergift in tenth-century northern Iberia’. 31 Cel368 (975–1009). The modius was a standard unit of account (which can often be rendered as ‘sackful’, e.g. of grain); for valuation systems, see ch. 6 above, pp. 109–14. 32 V43 (?968).



Allorito forced Siseguto c. 962.33 Sureties of an earlier agreement could appear in court, if the agreement unravelled, and could provide important evidence: an agreement that the priest Christoforo would hold the church of Santa María de Bonimento, and would make annual returns for it to the local count, lapsed when the count died; Christoforo refused to continue paying the rent to the count’s son, Munio Gutiérrez; Munio took the priest to court, where one of the two sureties for the original agreement, the woman Argilo, appeared and ‘spoke to the people’ about the agreement; her evidence decided the case.34 Because several sureties took an oath in favour of two plaintiffs claiming a mill on the river Tirón, against the claims of local lords, a court gave judgment in favour of the claimants, Sancho Gómez and Nuño Gómez, thereby demonstrating the special value of the sureties’ oaths.35 In the Allorito case in the Liébana, a complicated set of court appearances following a dispute, the surety had to provide a substitute vineyard if the vineyard Siseguto had to transfer was not handed over; he seems to have defaulted, but Allorito got him to court, where he undertook to pay a double penalty in the case of further default; in the end, Siseguto had to hand over two vineyards and two orchards into the hands of the surety and the court holder. The surety seems to have ended up with a vineyard and an orchard as a bonus.36 In cases of assault, bloodshed, theft, and other offences against persons, the surety might again perform several different functions. He might pay compensation to the injured person if the guilty party failed to do so: when Revel committed homicide, he was required to pay compensation of half a vineyard to Bishop Sisnando of León in 979; two sureties were appointed to make sure that he did this; if he failed, then the sureties were to hand over goods to the very substantial value

33 S356; T66. For Irish enforcing sureties, see Kelly, Guide to Early Irish Law, pp. 171–2. Early Irish law recognized three different kinds of surety, the paying surety, the enforcing surety and the hostage surety, in effect the three principal modes of exercising personal suretyship. These modes are common, and were not peculiar to Ireland or to the early middle ages, but the sustained treatment of the early Irish legal material is illuminating. One particularly important tract on suretyship, ‘Berrad Airechta’, has given rise to much discussion, although one might nowadays question its categorization as inherently ‘archaic’: R. Thurneysen, Die Bürgschaft im irischen Recht, Abhandlungen der Preussischen Akademie der Wissenschaften, Phil.-hist. Klasse, 2 (1928); D. A. Binchy, ‘Celtic suretyship, a fossilized Indo-European institution?’, in G. Cardona, H. M. Hoenigswald, A. Senn (eds), Indo-European and Indo-Europeans, Papers Presented at the Third Indo-European Conference at the University of Pennsylvania (Philadelphia: University of Pennsylvania, 1970), pp. 355–67; ‘Berrad Airechta: an Old Irish tract on suretyship’, trans. R. Stacey, in Charles-Edwards, Owen, Walters (eds), Lawyers and Laymen, pp. 210–33. 34 Sob130 (992). 35 SM27 (940). 36 T66; see further ch. 4 above, ‘Creating records of judicial disputes’, pp. 63–7, for this record. Early Irish law clearly provides for compensation to sureties in such cases (Kelly, Guide to Early Irish Law, p. 168), although capacity to compensate sureties developed in different ways in the wider world (e.g. A. Kiralfy, ‘History of the law of personal guarantee (suretyship) in England since 1500’, in Les sûretés personnelles II, pp. 399–424, at pp. 413–19).



of 100 solidi.37 Sometimes the surety’s function was to guarantee that an accused person would remain in the neighbourhood to answer the charge: six sureties were appointed in northern Galicia in 931 to guarantee that a thief would not run away; if he did, they were to pay 20 oxen to the court.38 At other times they might have to put down a gage or pledge of property as security, in advance.39 In one extremely unusual case of assault, heard in Valdoré on the southern side of the Cantabrian mountains, it looks as if a surety may have had to undergo the hot water ordeal in place of the accused, Ermegildo, who had attacked and wounded another man; the outcome of the ordeal was a confession by the accused and his parents then made a payment in accordance with the severity of the wounds their son had inflicted.40 Again, there are both peasant and aristocratic instances of the use of sureties in court and they occur in all regions; but these court cases do not occur at all times, for they mostly fall in the second half of the tenth century. If we ask what the sureties actually did, the answer is that they did a range of things. In part they were acting as ‘private policemen’, appointed by and agreeable to the parties to a contract. Their widespread use in dispute cases seems to indicate a familiar mechanism for securing outcomes in the later tenth century, making sure that decisions were followed through. In part they were carrying financial responsibility – making good any deficit either from their own funds or from their family property. And in large part they were a formalization of family responsibility for its members – public representatives of the wider family, as it were. As was also the case in ninth-century Brittany, it was often family members who acted as sureties.41 Galindo stood as surety for thirteen of his named heredes and sobrinos who were involved in the sale of Villanueva on the river Cea, on the road to Grajal, in 955.42 In southern Galicia, a married couple named the wife’s brother as surety for their debt, and another man named nephews and/or nieces; when they defaulted, the brother and the nephews/nieces paid, using their family property.43 A further twist to this view of the surety as part of the interest group 37 Lii463 and Lii464; Lii463 is the record of the judgment; Lii464, very unusually, is a separate document spelling out the sureties’ obligations, which they formally agreed, with witnesses. 38 Sob21. 39 E.g. the villa pledged in Cel215; cf. Gilissen’s caution about differentiating real and personal suretyship, above n. 22. Freedom from having gages/pledges (pignora) taken is a standard in immunity formulas; see V2 (804), SM67 (955), both suspect charters in respect of their purported dates, although probably reflecting formulation current in the eleventh century. 40 OD31 (991), an ‘original’. This is certainly what the text appears to say and what its recent editors understand. However, the text is ambiguous, and I think it is equally possible that Ermegildo may have undergone the ordeal himself, with the surety making sure he appeared in court on the due day. Surety here, unusually, is fidelis, and the surety may also have been asked to swear in support of Ermegildo’s denial; oath-helping was certainly one function undertaken by several unambiguous sureties. See further above, ch. 14, ‘Settling disputes’, pp. 243–4, for this case. 41 These were, however, clearly identified as separate individuals; this was not collective family responsibility, as much earlier discussion presumes. 42 S366. 43 Cel368 (975–1009), Cel204 (1005).



comes from an agreement by seven peasants of Fuentes de Payuelo that they would in future only pay rent to Sahagún; this was guaranteed by seven people: no. 4 of the guarantors stood surety for his son; nos 1 and 2 stood surety for no. 6 of the guarantors, who was also party to the transaction; no. 5 stood surety for no. 7 (a monk); no. 7 stood surety for his own brother, no. 5. In other words, at least three of the sureties had family connections with the principal parties, at least three were themselves party to the transaction, and at least five of the group were closely interconnected.44 As in Brittany, sureties were often interested parties.

Sureties and scale In the Breton material, with which I began, the scale of operation of sureties was exceptionally local. Nearly all of them acted within their own, home, village, community, not travelling much more than 6 or 7 km to exercise their duties; exceptions might venture to the next community but not beyond. In the Iberian cases, we very rarely know where a transaction took place and individuals rarely appear more than once; it is therefore difficult to localize individuals, particularly secondary actors; we often have no idea where the sureties came from.45 It is therefore impossible to make direct scale comparisons, since we see the former, as it were, in high resolution and the latter in rather lower resolution. There are, however, three kinds of observation that suggest that peasant use of sureties in northern Spain and Portugal may well have been comparable to the Breton pattern. Firstly, there are cases in which different peasant settlements made commitments to (monastic) landlords, each settlement producing one surety for its community: seven named settlements, with one fideiussor apparently acting for each, agreed to respect boundaries on the Pardomino mountain in 944; and the four sureties from three specified locations south of the river Ebro, north of San Millán de la Cogolla, who guaranteed a gift of property in that locality, must be local representatives from the lands transferred, given that two of the specified locations are only 2 km apart.46 Secondly, there are cases in which the surety was a member of a group which shared a small-scale inheritance, such as a set of parcels of land within a villa; his function was to hand over a parcel from the shared inheritance if it should be necessary – like Galindo and his thirteen named heredes and sobrinos noted above. Thirdly, suretyship quite often involved guaranteeing the actions of family members; of course, family members did not have to be local, but in peasant cases they usually were. For all those reasons, it appears that the scale of some

44 S289 (977); cf. the Breton CR96 (867). 45 This is partly a straightforward consequence of diplomatic style: whereas Breton charters commonly conclude with a formula indicating that this was done in such and such a place on such and such a date, the comparable Iberian formula simply indicates that this was done on such and such a date. Further, since the Iberian material is so much more diverse, covering a wider area more thinly and haphazardly, it is much more difficult to deduce the origins of individuals. 46 Li184; SM93 (975).



of the Iberian surety activity – not necessarily all of it – is entirely comparable to the operation of suretyship in ninth-century eastern Brittany. Such sureties might travel 6 or 7 km to carry out their obligations, a pattern which can be found all over the area, from Galicia across the meseta to Castile. Unambiguously aristocratic cases, however, were clearly quite different in their scale of operation. The dispute of Velasco Hánniz with the abbot of Abellar over properties on the river Araduey and in Marialba took them before the king in Simancas, where they were ordered to appear before him in León with their evidence, a surety called Fortes and the palace saio (a legal officer who, for example, administered oaths) guaranteeing their appearance; in fact they ultimately appeared before the bishop of León, rather than the king, in the church of San Félix de Torío, north of León, all this involving journeys in the order of 100 km.47 A Galician dispute between the two laymen, Ero Fofiz and Osorio Iohanniz, involved a commitment to appear before the king at Laias, on the river Miño west of Ourense; a surety guaranteed his appearance and the villa of San Paio de Albán – over 20 km away to the north-east – was named as security that Ero Fofiz would duly appear; he failed to do so and ultimately the surety’s oath ensured that the villa was handed over to Osorio.48 Lastly, the dispute of Vela Vélaz with the monastery of Sahagún, over ownership of the Villa Pedro in 998, was heard before the count sitting in Villalpando, 58 km to the south west; Vela lost and his surety, the priest Gelmiro, was the person who handed over the villa to Sahagún.49 There are clearly some major differences between ninth-century Breton and tenth-century Iberian practice. Sureties in Spain had a wider range of functions; these were not dissimilar but were more varied. For example, although they are called sureties, their functions in court could overlap those of oath-helpers (people who gave support through evidence of character), of saiones (that is, legal officers), and of expert witnesses (people who provided a special, expert, quality of evidence).50 This is not inconsistent with the wider European picture of medieval suretyship, but it is a dimension that is not much evidenced in Brittany in the ninth century.51 Secondly, the social range of those who used sureties was consistently wider, from peasants through middle-range people to high aristocrats; this appears to be a real social difference and not simply a reflection of the availability of different kinds of source material. Whereas in Brittany suretyship appears to have been a norm of intra-peasant relationships, it seems to have had wider social application in Iberia in the early middle ages. Thirdly, sureties could move across much wider areas in Iberia. What seems to have been happening in aristocratic cases was not simply that sureties covered longer distances in the context of much 47 Li256 (952). 48 Cel215 (995). Cf. R. Collins, ‘“Sicut lex Gothorum continet”: Law and charters in ninth- and tenth-century León and Catalonia’, English Historical Review, 100 (1985), pp. 489–512, at p. 503. 49 S356. 50 For example, SM27 (940), Li256 (952), Sob130 (992). 51 Gilissen, ‘Esquisse d’une histoire’, pp. 77–80.



wider property interests, which was clearly the case, but also some extension of the use of sureties. Sureties in this context did not simply guarantee obligations between two parties, for those parties, as in the Breton cases; in Spain and Portugal they certainly guaranteed obligations between parties in a similar way, but sometimes they also guaranteed that one party would answer to representatives of the ruler. We have seen in the examples above that disputants had to travel because they had to appear before the king or the count; sometimes the surety travelled for them; and sometimes the surety appeared before king or count to hand over disputed property. Hence, in the Sahagún case, the surety travelled and handed over the Villa Pedro; in the Galician case, it was the surety who went before the king to transmit the pledged Villa Albán, which had been identified as security; in the Marialba case, although the parties were ordered to appear before the king in León, the order was transmitted through the surety, who must have had to travel to Simancas. And in a well-known Castilian case about the church of San Pedro in Tubilla del Lago, in the south of the region, near Aranda, the surety had to transmit 300 solidi to the count, almost certainly 30–60 km away in Burgos, if the defendant, García Refugano, failed to produce his charters in court.52 We seem to have an extension of the surety mechanism here: the ruler and powerful aristocrats were using it to relate to, and indeed attempt to control, lesser aristocrats.

Conclusion The Iberian charter evidence of suretyship, like the Breton, is extremely important precisely because it provides evidence of personal suretyship in practice in all its varied forms and because it provides evidence of practice at an early stage, before the legal formulations of the eleventh century and later.53 One might say that in the Breton case we see something close to the private use of sureties, within local communities, at a very small scale. Of course, this was not entirely private since it was in the interest of the local community that suretyship worked, and the appointment of sureties was in effect endorsed by the community. Thereby the practice contributed to social stability. In Spain and Portugal we see that private use too, but we also see an extending use of the mechanism, and greater overlap with other mechanisms, and we see its adoption by powerful aristocrats as a tool in systems of control – partly by kings, partly by powerful religious bodies, partly by lay aristocrats. That happened at a time, in the late tenth century especially, when royal control systems were evidently extending, for example through the 52 C90 (957). For Tubilla del Lago, see J. Escalona and F. Reyes, ‘Scale change on the border: The county of Castile in the tenth century’, in Escalona and Reynolds (eds), Scale and Scale Change, pp. 153–83, at pp. 180–1; cf. R. Collins, ‘Visigothic law and regional custom in disputes in early medieval Spain’, in W. Davies and P. Fouracre (eds), The Settlement of Disputes in Early Medieval Europe (Cambridge: Cambridge University Press, 1986), pp. 85–104, at pp. 92–3. 53 See especially the survey of personal suretyship by F. T. Valiente, ‘Las fianzas en los derechos aragones y castellano’, in Les sûretés personnelles II, pp. 425–81, where the focus is on development from the eleventh century onwards.



use of mandationes, that is by delegating control of certain limited regions to trusted aristocrats.54 Of course, one might reasonably argue that aristocrats always operated on a wider scale anyway. What is interesting here is the use in high level courts of a mechanism which was particularly appropriate for local societies, in which reputation was paramount. Suretyship was harnessed for wider purposes, although it necessarily retained the element of personal guarantee. The comparison with tenth-century Anglo-Saxon England, where similar things were happening, is particularly instructive: here the record of royal legislation shows how kings deliberately used suretyship for state purposes, to underpin public order. Indeed, the laws not only indicate the duties of personal sureties when transactions went wrong, they also took a step towards the public institutionalization of suretyship: all men of ill repute were to have a surety (III Edmund 7.1, c. 942– 946, and frequently repeated thereafter); in Patrick Wormald’s words, in effect ‘neighbourhood sureties’ were created in the drive against ‘society’s bad risks’.55 In this case the legislation survives to demonstrate the rulers’ intentions, but there are fewer examples of practical application than survive in Iberia. The AngloSaxon comparison is particularly appropriate because of its clarity and its date; however, similar developments occurred in other European, and non-European, states at some point.56 I would make two final points, on the background and the future. None of this was new in the late ninth and tenth centuries. Visigothic law had made clear provision for the use of sureties in sales if the vendor was not of trustworthy character.57 There was in any case plenty of late Roman background.58 The sixth- and seventh-century texts preserved on slates from central Spain do not feature personal sureties, although they include records of sales; nor do the extant formularies, 54 See, for an especially clear treatment, C. Estepa Díez, ‘Poder y propiedad feudales en el período astur: las mandaciones de los Flaínez en la montaña leonesa’, in Miscel·lània en homenatge al P. Agustí Altisent (Tarragona: Diputació de Tarragona, 1991), pp. 285–327. 55 P. Wormald, The Making of English Law: King Alfred to the Twelfth Century, I, Legislation and its Limits (Oxford: Blackwell, 1999), pp. 290, 316, 328; id., Papers Preparatory to the Making of English Law: King Alfred to the Twelfth Century, Volume II: From God’s Law to Common Law, ed. S. Baxter and J. Hudson (London: University of London, 2014), ch. 9, ‘The Pursuit of Crime’, available at See above, n. 6, for law texts. D. Pratt, ‘Written law and the communication of authority in tenth-century England’, in D. Rollason, C. Leyser, H. Williams (eds), England and the Continent in the Tenth Century (Turnhout: Brepols, 2010), pp. 331–50, demonstrated that a new system of justice was developed under King Athelstan (d. 939), which meant that suretyship was to be administered by lords. 56 E.g. Ph. Godding, ‘Les sûretés personnelles dans les pays-bas méridionaux du XIe au XVIIIe siècle’, in Les sûretés personnelles II, pp. 263–364. 57 Forum Iudicum, 5.iv.2: Leges Visigothorum, ed. K. Zeumer, MGH LL nat. Germ., 1 (Hanover and Leipzig: Hahn, 1902), p. 218. 58 For example: The Theodosian Code and Novels and the Sirmondian Constitutions, trans. C. Pharr, T. S. Davidson, M. B. Pharr (Princeton: Princeton University, 1952), pp. 23, 60–1; Digesta Iustiniani Augusti, ed. T. Mommsen and P. Krueger, 2 vols (Berlin: Weidmann, 1868–70), vol. 1, 18.1.81 (p. 525). See also E. Levy, Weströmisches Vulgarrecht. Das Obligationenrecht (Weimar: Hermann Böhlaus Nachfolger, 1956), pp. 197–206, 221–3.



which to some extent reflect a sixth- and seventh-century past, although the formularies include plenty of impersonal (i.e. real) securities.59 That does not mean that personal sureties were never used; only that they were not a prominent feature of transaction models. After all, even in the tenth century, most sales were not secured by sureties – they occur in under 4% of sale records (although, notably, all the earliest documented cases relate to sales). Sureties were needed where trust was insecure, not as a general rule. Secondly, the use of sureties, although already well established in Late Antiquity, seems to have been extending at the end of the first millennium. That is in itself a comment on trust, as well as on extending systems of control, and is thereby a comment on a changing and diversifying society. It is also notable that the use of sureties to guarantee payment of rents and debts is a feature of the later period, from the 960s, and must reflect the strategy of increasingly powerful monasteries to secure their income.60 Nor did this process stop in the late tenth century; suretyship continued, and, without implying the evolution of a supposed ‘archaic’ suretyship into modern forms, the functions of sureties clearly continued to be adapted: by the twelfth century sureties were a common mechanism for avoiding seizure both of the goods of strangers and of those of people accused of homicide.61 This was still personal suretyship, but in an urban-regulated setting. The tenth century offers a window on a changing process. We can see sureties operating at smaller and wider scales, with scale by implication extending; we can note kings, counts, and powerful monasteries harnessing the mechanism for their own political purposes. The changes were related to the expanding arena of socio-political interaction and to a world in which buying and selling involved more and more strangers – hence the issues of trust. The social, economic, and political significance of suretyship was indeed wide.62

59 For the slates: I. Velázquez Soriano, Las pizarras visigodas. (Entre el latín y su disgregación. La lengua hablada en Hispania, siglos VI–VIII) (Madrid: Real Academia Española, Instituto de la Lengua Castellano y Leonés, 2004); note no. 92, pp. 338–40, ‘tuam securitatem’. For formularies: M. Zimmermann, ‘Un formulaire du Xe siècle conservé à Ripoll’, Faventia, 4:2 (1982), pp. 25–86, at p. 79; cf. ‘Formulae Wisigothicae’, ed. J. Gil, in J. Gil, Miscellanea Wisigothica (Seville: Universidad de Sevilla, 1972), pp. 69–112, at pp. 103, 105–6. 60 Perhaps paralleled in the Breton material: 31% of sureties occur in the 860s, a decade when the monastery of Redon was establishing many new tenancies; Davies, ‘Suretyship’, p. 75. 61 A. García-Gallo, ‘El fuero de León. Su historia, textos y redacciones’, Anuario de Historia del Derecho Español, 39 (1969), pp. 5–171, at pp. 161, 165; note the increase in references to sureties in twelfth-century recensions by comparison with that of 1017, e.g. pp. 154, 164 (Villavicencio version). Cf., in general, changing guarantee mechanisms in eleventh- and twelfth-century Italy: C. Wickham, ‘Compulsory gift-exchange in Lombard Italy, 650–1150’, in W. Davies and P. Fouracre (eds), The Languages of Gift in the Early Middle Ages (Cambridge: Cambridge University Press, 2010), pp. 183–216. Tracking the growing legal control of the use of (real) securities is one of the main themes of the Spanish historiography, with an emphasis on the eleventh century onwards: Orlandis, ‘La prenda’; Valiente, ‘Las fianzas’, p. 450. 62 I am extremely grateful to members of the Foundations of European Space project, and especially to Julio Escalona, for their helpful comments on earlier drafts of this paper.



ʽAbd al-Raḥman III 225, 232 Abellar, monastery of 23, 30, 88, 90, 143, 194, 270, 299 Accounts 53, 57, 74, 75 ACO B.1.11 67–73 adstantes 274, 285 adultery 125, 254, 268 Aḥmad al-Rāzī 223, 225, 226 al-Andalus 154, 180, 225–7, 234, 238–9 Álava 164, 175, 181, 208 Albelda, monastery of 43, 122, 163n64 Alfonso IV, king 270 al-Ḥakam II 223, 226, 232 almunias 225, 227 Amorino, priest 23, 28, 32, 50 Anales Palatinos 223, 226, 236 Andreas, magister 29, 30n52 Anonymous Andalusī Agronomist 223, 227, 232–3, 235, 236 apples 192, 194, 208, 210, 231 Arabic 223, 224 archaeobotany 231, 233–4, 239 archaeology 15, 48, 175, 187n9, 200, 223; excavations 230–1; Irish 166–7, 169 argenzos 106–9, 114, 159 ʽArīb ibn Saʽīd 226, 232, 239 aristocrats 31–2, 152, 201, 299; control by 128–9, 300, 302; households of 23–4, 32; resources of 189–90, 198–9; transactions of 88, 89, 127, 133, 147, 156, 157, 160, 161 arson 246, 253, 264 assault 62, 246, 247, 253, 264, 291, 296 assessors 84, 275, 277–8, 282, 283 Astorga, bishop of 45, 49 Asturias 43, 122, 194, 228 Bagaudano 88, 129, 152 baptism 43, 48

barley 175, 196, 231, 233 Becerro Galicano 204–9 Berulf 58–62, 248 bible, quotations from 150–1, 171, 182 bishops 34, 39, 40–1, 147, 198, 267; control by 44–5; see also Astorga; Braga; León; Lugo; Oviedo; Rosendo; Santiago de Compostela bloodshed 246, 296 boats 197, 210, 259 boni homines 47, 83, 128, 253–4, 269, 274–87 books 42, 45, 48–50, 103n24, 154 boundaries 37, 171, 228, 269, 276 Braga, bishop of 45; Martin of 35 Braolio, priest 24, 25, 28, 29 Brittany 22, 184, 186–91, 254, 257, 279; boni viri in 281, 286; sureties in 290–2; see also evidence; plebes Briviesca 214–16, 218 bullion 106, 107–8, 179, 182 Burgos 109, 144, 206, 208, 300 burial 43, 47, 48, 51, 130 Calendar of Córdoba 227, 232, 235 campaigning, military 3, 96, 116 candles 130, 197, 217 Cantabria 122, 228 capitularies, Carolingian 105–6, 109, 280 Cardeña, monastery of 100, 163; charters from 98–116, 122, 134n71, 140, 143, 228 cartularies 58, 63, 82, 121 Castile 9, 136, 141, 146, 194, 293; charters from 78–9, 132, 134; count of 211, 215 Castro, A. 2–3, 9, 15 Catalonia 7–8, 49n77, 123n23, 127n39, 136, 170, 223, 228; courts in 170, 261, 281–2



cattle-standard xiii, 112, 159, 160n46, 167–8, 176, 177, 246n19 Celanova, monastery of 90, 91; area surrounding 26; charters from 78–9, 98–115, 121, 140, 228; see also inventories cereals 173–6, 178, 219, 234; see also barley; millet; oats; rye; wheat Chadwick, H. M. 166, 169, 176 charters xi, 19–33, 34, 53–73, 77, 78, 83, 121, 206; Breton 185, 253–4; fabricated 43, 75, 120, 204, 205, 214n39, 248n28, 250; format of 54–5, 79–82, 90, 119, 129–30, 131, 134, 135, 217; foundation 209, 210, 211; hands of 25–30, 58, 63, 73, 75; Irish 168, 171; mirror-type 82, 91; suspect 47n62, 75, 227, 248; see also chrismons; Llandaff; signa cherries 192, 231, 234 chestnuts 231, 234, 235 chrismons 27, 30, 67, 71, 72, 73 Christians 1, 2, 9, 10, 13, 14, 96, 234 chronicles 12, 14, 35, 226, 239 churches 39, 40, 43, 44, 51, 259, 263; consecration of 45 climate 169, 184, 210, 223, 236 cloth 42, 103, 104, 111, 113, 154, 157, 158 clothing 85, 101, 103, 132, 155, 156 Cofiñal 22, 32, 268 Coibnes Uisci Thairidne 167, 173 coins 153, 165, 168, 176, 178, 179, 189, 290 Collins, R. 57, 248n29 colonization 3, 6, 12–13, 191 commemoration, liturgical 118–19, 130 commitments 54, 58, 59, 61, 62, 64, 67, 71, 72 commodification 119, 160, 165, 178 communities, local 3, 15, 24, 38, 46–7, 50–1, 255, 293 compensation 235, 243, 253, 294, 296 concilia 15, 31, 38, 46, 256 conditiones sacramentorum see oaths confessions 44, 54, 56, 252, 262, 263, 276 Confessions 53, 54, 57, 58, 64 contracts 173, 233, 235, 238, 288; see also sales Córdoba 180, 223, 225, 233, 236 countergifts 85, 100, 125, 126, 131–4, 137–49, 156–7; confirmation by 142, 145 courts, judicial 199, 261n10, 280, 283, 284, 292; actors in 31, 42; presidents of 266; procedure in 248–9 court-cases 50–1, 53–73, 79n7, 243–58, 263, 276, 282, 294, 295

court holders 257, 266, 271 crime 245–7, 256, 264 cumala 168, 177 debts 78, 126–7, 145, 288, 295 depopulation 3–4, 6, 11, 12, 96 Devroey, J.-P. 202, 203 dirhems 105, 107, 108, 158n39, 179, 180, 181, 182n67 display 155, 161 dowries 118, 154 Duby, G. 274, 283 Duero basin 3, 6, 96, 97, 116, 175 Durabiles, priest 23, 25, 27, 28, 29–31 Elvira 151, 157 endorsement 72, 73 England 133n69, 156n26, 289, 301 Eslonza, monastery of 37, 40, 140n16, 141 estates see property evidence 60, 61, 67, 68, 69, 250; Breton 245, 247, 249, 250, 256; Welsh 244, 246, 248, 250, 256 exchange 77–92, 95, 114, 119, 181, 194; commercial 115, 164, 165, 178, 221; unequal 84; see also gifts; markets; sales family 39, 85, 128, 147, 298; responsibility 145, 243, 254, 297–8; see also inheritance; property; religious households farm produce 42, 101, 103, 111, 158 ferragines 207, 209, 210, 227, 237 feudalism xiii, 2–3, 4–6, 11, 198n55, 258 figs 226, 231, 233 fines 64, 67, 78, 125–6, 129, 200, 253, 265 Flaino Muñoz 24, 27, 129, 162n59, 244 flax 210, 229, 231 fodder, green see ferragines folle zumag 103, 131n63, 180 formularies 55, 280, 286, 289n5, 302; Ripoll 26, 81; see also Ibn al-ʽAṭṭār formulas 54–5, 80–2, 124, 140, 206; appurtenance 227; examples of standard 59n21, 60n24, 65n43, 66, 67, 68, 69, 71, 72, 277; treatment of standard 25–7, 205, 297n39 Francia 202n73, 279, 282, 284, 285, 288 freedom see status Froncea, monastery of 215, 219 Fruela Vimáraz 24, 32, 276 fruit 226, 228–9, 231, 233, 234, 237, 238–9; see also apples; cherries



furnishings 14, 42, 84–5, 103, 153–5, 157, 180 furniture 49, 128, 155, 157 Galicia 52, 78–9, 80, 84, 134, 193, 196, 228, 278; charters from 131 Ganz, D. 28n48, 31n53, 33n60, 130n60 gardening 235–8 gardens 50, 85, 125, 192, 194, 210, 222–40; ornamental 225–6, 227 Gautier Dalché, J. 107, 110, 158n40 gift-giving 119, 128n47, 137 gifts 78, 87, 98–100, 115, 124–34, 139, 216; pro anima 130–1, 147 gold 104–5, 108, 116n96, 178 Gondemaro 166–7, 247 Gordón 67–72 Gózquez 230, 233, 237 grazing rights 211, 213, 214n39 guarantees 144, 145–6, 293, 295, 297, 302n61 guarantors see sureties Guilhabreu 43, 47, 256 Guimarães, monastery of 90, 120, 127, 155, 249, 259 hagiography 35, 41, 169, 251 hands see charters Hermenegildo 88, 129, 162n59 Hibernensis 171, 172, 289n5 Hiniestra 208, 211–18, 236 historiography 119, 137, 260–1, 274, 283; Spanish 2–6, 9–10, 138, 162–3, 191–2, 258n76, 302n61 hoards 105, 158n39, 179–80 homicide 21, 51, 152, 246, 247, 255, 280 Iberia 8–9, 121–3, 169, 184, 191–201, 226, 234 Ibn al-ʽAṭṭār, formulary of 233, 235, 237, 238 Ibn al-Qūṭīya 223, 225 Ibn Ḥawqal 226, 239 identity: collective 261; cultural 13–14; group 15, 46; individual 24–5; regional 8–9 idonei 281, 283 inheritance, family 39–40, 42, 88, 112, 173, 192, 233 inquest 31, 50, 252 intercession 270, 276–7, 279 inventories 83, 87, 110n70, 154–5, 191n25, 194, 195, 201, 206–7 Ireland 166–83, 289; suretyship in 296n33; see also law

irrigation 227, 229, 230, 233, 235–6; see also water channels Italy 22, 159n45, 258, 268n43, 289; boni homines in 279, 281–2, 285; countergift in 142, 144, 147 iudex 268–9 iudices 268–9 Jews 2, 9, 14–15, 218 judges 166, 170, 243, 250, 259–73; kings’ 266–7 judgment 268; collective 258, 261, 271; Day of 269–70 justice 270–1, 271–3 kilns, corn-drying 174–5, 178 kings 152, 215, 227, 300; claims of 12–13; control by 129, 147, 198, 204, 301; income of 105, 190, 197; of León 246, 299; see also Alfonso IV; judges; Ordoño III; Ramiro III; Sancho Garcés I; Vermudo II labour 188–9, 193, 195–8, 219, 237–8; service 197; tied 195–6 laity 21, 91, 158; archives of 121; see also religious households land: control of 163, 181n66; heritability of 171; management of 88, 186–9, 192n29, 209; values 168 land-use 203, 209–18 language 78, 119, 124–6, 126–7, 130, 172–3, 269–71, 272 Latin 24–5, 172 law: Anglo-Saxon 289, 301; expertise in 269; Irish 167–8, 169, 173, 174, 251, 255, 289n5, 296n33; Italian 280; Roman 280–1, 287, 301; Visigothic 81, 148, 170, 252n50, 257, 275, 301; Welsh 250, 255, 284 learned class, hereditary 41, 52, 268 Lebor na Cert 183 legitimation 12, 57, 279, 284, 285, 286, 287 Leire, monastery of 43, 140n16, 155 León, city of 91, 108, 113, 156, 181; area surrounding 26, 136, 154n19; bishops of 34, 299; charters from 77, 78–9, 84, 89, 98–116, 121, 223, 228; see also kings Li192 58–63, 65n43, 67n49, 73 Liber Ordinum 35, 48, 49 Lichfield Gospels 244, 250n37, 254, 257 Liébana 63, 122, 145, 196, 199, 228, 231, 293



liturgical vessels 42, 49, 153–4, 179 liturgy 40, 48–9 livestock 42, 84–5, 103, 104, 157, 193–4, 195–6 Llandaff charters 244–5, 246–7, 248n28, 249, 253, 254 lordship 5–6, 190, 199, 289; see also seigneurialization Lorvão, monastery of 37, 157 Lovesindo Abenazar 249, 259–60 Lugo, bishop of 39 Madīnat al-Zahrā 225–6 magistri 29, 41 mandationes 198–9, 301 manual see Liber Ordinum manuring 236–7 manuscripts 48, 58, 170 Marialba 22, 37, 63, 88, 143, 299 markets 97, 100, 115, 135, 163, 164, 181 marriage 118, 155, 195, 247 masses 118, 130, 135n75 meetings 15, 249, 263, 272, 279, 285 Melic, priest 32, 42, 151 meseta 97, 131, 132, 158, 164, 193, 196, 277, 278 millet 111, 231, 233, 294 mill rights 42, 163, 174, 192, 216–17, 218, 219 modii 83, 111–12, 113, 159, 177 monasteries 44, 88, 221, 240, 302; foundation of 209; small 36–7, 209–10; transactions of 89, 148; see also Abellar; Albelda; Celanova; Froncea; Guimarães; Leire; Lorvão; Piasca; property; Sahagún; Samos; San Juan de la Peña; San Millán; Santos Justo y Pastor; scriptoria; Sobrado; Taranco; Turieno; Valdevimbre; Valpuesta Montes de Oca 211, 214, 215 Moors 14, 118, 154 Morimoto, Y. 184, 201, 202 morning gift 155, 247 movables 84–6, 90, 91, 101, 153–8, 161; circulation of 163, 164, 178; imported 154, 164 Munio Fernández 24, 88, 152, 162n59, 246, 254 Munio Flaínez 23, 26, 31, 155 Muslims 1, 3, 10, 14, 40, 105 narrative 55, 57, 65 notary see scribe

oath-helping 251, 264n21, 297n40 oaths 51, 70, 74, 243, 251, 259, 262, 265 Oaths 53, 60, 69, 73 oats 106n38, 175, 188, 189, 233 olives 226, 231, 232, 235 Orbañanos 206, 210 orchards 21, 114, 192, 199, 219, 224, 226 ordeals 51, 71–2, 73, 243, 251, 252 ordination 40, 41n34, 48 ordines see liturgy Ordoño III, king 72, 73 orthography xi, 24, 25 ortos 224, 227, 229–30, 238 Otero de las Dueñas, charters from 77, 78, 79, 89 Oviedo: bishop of 85; cathedral 67; see also San Vicente parishes 36, 37, 38, 39 Parochiale Suevum 37–8 Pastor, R. 5, 138 pastoral care 43, 47–50, 51 pasture 114n90, 115, 194, 209, 226 peasants xi, 11, 20–2, 175–6, 201, 239, 291, 299; proprietors 3, 191; transactions of 50, 87, 88, 157–8, 160, 182, 192, 298; see also property penalties 252–4; enforcement of 254–5; see also fines; sanctions penitentials 49, 50n79, 169 Piasca, monastery of 157, 174, 195 plebes 46; Breton 39, 186, 189, 292n16 ploughs 181n66, 196 pomares 224, 229–30 populare 12, 96n4, 163 Portugal 81, 83, 118, 196, 228, 259; charters from 113n87, 120 pottery 163, 164, 182 preambles 150, 151, 269 precious metals see gold; silver precious objects 103, 115, 155–6, 157, 161, 164 presura 13, 38 prêts sur gage 127n39 prices 86, 100–9, 124–5, 155–6, 217, 218; fixing of 50, 159n45, 277; Irish 172, 173, 176, 177; ‘just price’ 83, 160, 277, 282; special uses of 83, 126–7, 139, 143, 290 priests, local 19–33, 34–40, 51–2, 87, 118, 161; appointment of 44; resources of 41–3; training of 41; see also pastoral care profit 151, 161, 191



property, landed 152; accumulation of 91, 148, 162, 191, 194, 214, 240; disputes over 247–8, 264, 295, 299; estate 187–9, 193–5, 209–10; family 42, 44, 52, 118–19; peasant 42, 186–7, 191–3, 238; see also villas proprietary rights 39–40, 44 Psalms 48, 49, 153n16, 271–2 public: justice 261; performance 272–3, 277; power 11; sense of the 271, 274 pulses 230, 231, 233 purchases 23, 32, 42, 116, 173, 200, 214, 216–17; see also sales quasi-presbiteri 40 quinta 170–1 Rabal 38, 114n92, 193 Ramiro III, king 157, 267 rannou 186–7 rape 129, 244, 246 reconquest 1, 9, 10 recording 22–4, 31, 50, 248, 258; see also scribal practice records: composite 58–73; dispute 59–72, 262; primary 56, 59; secondary 56–7 redactor see scribe registers 208, 217; see also inventories religious households, lay 21, 36, 151 rents 126, 188, 189, 191, 195, 196, 295 res xiii, 103 Reuter, T. 259n1, 262, 272 rewards 151–2 rich people 152–5; see also wealth ritual 55, 272 robra 138, 148, 149n49 Rosendo, bishop 88, 130, 153, 197 rye 175, 188, 231, 233 Sahagún, monastery of 90, 163n64; area around 26; charters from 78, 79, 98–116, 228; purchases by 22 saiones 50–1, 144, 243, 249, 252, 299; in composite texts 59–72 sales 50, 83, 84, 86–7, 98–109, 124–9, 133–4, 294–5 salt pans 163, 188, 189, 192, 194, 196, 219 Samos, monastery of 21, 247; area around 26; charters from 77, 78–9, 141, 293 Sánchez-Albornoz, C. 2–4, 9, 12n54, 105, 109n63, 110n65, 141, 159n44, 191n28 Sancho Garcés I, king 155

sanctions 63, 67, 104–5, 205, 208, 269, 294; examples of 61, 62, 65n43, 66, 68, 69, 71, 72, 74 San Juan de la Peña, monastery of 43, 122, 236 San Millán de la Cogolla, monastery of 203–21; charters from 98–116, 122, 140n16 San Román de Tobillas 38, 209–10 Santiago de Compostela 96, 179, 234; bishop of 38, 45 Santos Justo y Pastor, monastery of 58–63, 84, 129, 156, 267, 277 San Vicente, monastery of 122, 140n16, 228 scales, weighing 181, 182 Scandinavia 108, 181; see also Vikings schools see priests; scribes scribal practice 24–31, 53–73, 90, 147, 151–2, 198, 277, 278–9; of Cardeña 115, 134n71, 144; Castilian 206, 275; Galician 81, 275; local 135 scribes 19–20, 23–4, 50, 55, 73, 75, 267; training of 26, 30 scriptoria, monastic 35, 48, 55 seigneurialization 5, 11, 190 serfs 187, 188, 195, 201, 220 signa 28–9, 30, 58, 59, 61, 64, 65, 68,70, 71 silos 175, 200, 230, 238 silver 85, 101–2, 105–9, 157, 177, 178–81 slaves 155, 195, 254 Sobrado, monastery of 45, 251; area around 26; charters from 78–9, 140, 228, 293 solidi 104–6, 189; in fines 64, 68, 72, 300; in valuations 83, 102, 109–11, 112–13, 143, 158 status 239, 253; free 3, 12, 186, 191, 196, 291 storage pits see silos superiorities 189–90, 198–9, 202 sureties 51, 65–7, 74, 134, 146, 254–5, 288–302 sweeteners 85–6, 89–90, 126 Szabó, T. 275, 280n31, 286 T66 63–7, 73, 75, 296n36 Taranco, monastery of 208, 209 tax 11, 190, 199 tenants 188–9, 190–1, 195, 220 terra calva 175, 200 terraces, agricultural 175, 200, 237 terra indomita 176, 200 tertias 44, 45 theft 125, 246, 247



timber 213, 214, 218 Tírechán 168n8, 171, 172 tithes 36, 43, 144n33 towns 163, 230, 238, 302; see also Burgos; León; Santiago de Compostela; Zamora tradition, transmission of 41, 169–71, 183, 274, 286; learned 55 transactions 98–100, 122–3, 147; see also aristocrats; exchange; gifts; monasteries; peasants; sales; women treasure 133n69, 150, 151, 153–5 tremisses 110, 111, 113, 115, 177 trust 149, 276, 279, 291, 302 Turieno, monastery of San Martín at 63–7, 122, 132 units of account 109, 116, 159, 176, 177–8, 179, 217 Valdevimbre, monastery of 47, 143, 166–7 Valdoré 24, 29, 51, 243, 244, 276 Valpuesta, monastery of: charters from 78–9, 98–116, 122, 228, 293 valuations 83, 109–13, 143, 158–61, 176–8, 277–8, 284, 286 value, scales of 253, 256, 271 vegetables 230, 232, 233, 239; see also pulses verbatim reporting 55, 74 Verín 39, 250, 276 Vermudo II, king 85, 158 vestments 153–4, 182 vetch 233, 237

Vikings 167, 178, 179, 182n67 Vilaine, river 186, 187 villas 188, 193–4, 195, 211, 298 vineyards 64–6, 175, 192, 199, 210, 219, 231 Wales see Lichfield Gospels; Llandaff walnuts 231, 233, 234, 235 water channels 167, 210, 229, 235–6, 247, 294 water mills 166–7, 173–4, 178, 187, 192, 229; Arlanzón 217, 218; see also mill rights wealth 150–65 weapons 156, 157 wheat 175, 188, 231, 233 Wickham, C. J. 13, 202n73, 274, 284, 285 witnesses: in court 58, 59, 70, 71, 251, 259, 265; qualities of 281; statements of 249; of transactions 23, 46, 72, 82, 206 witness lists 37, 61, 62, 66, 68, 71; absence of 206; character of 80, 208, 277; construction of 59n17, 73 women 154–5, 187, 195, 265, 266, 295; in court 247, 254, 296; transactions of 118, 136, 151, 156, 173, 194, 216–17, 218; as witness 259, 278n26 woodland 175, 209, 213 wounding 176, 243, 248, 253 Zaballa 231, 233 Zamora, town of 105, 114, 230