Anglo-Norman Studies XLI: Proceedings of the Battle Conference 2018 [41] 1783273992, 9781783273997

Table of contents :
ILLUSTRATIONS AND TABLES vii
EDITOR’S PREFACE ix
ABBREVIATIONS x
Horses, Knights and Tactics (The R. Allen Brown Memorial Lecture, 2018) / Sally Harvey 1
Baldwin of Forde, Bartholomew of Exeter and the Authorship of the 'Liber de sectis hereticorum et orthodoxe fidei dogmata' / Sabina Flanagan 23
Evidence of the Ordinary: Wives and Children of the Clergy in Normandy and England, 1050–1150 / Hazel Freestone 39
Anthropology, Feud and 'De obsessione Dunelmi' / Tom Lambert 59
New Archaeologies of the Norman Conquest / Aleksandra McClain and Naomi Sykes 83
An Angevin Imperial Context for the Amboise–Anjou Narrative Programme / Nicholas L. Paul 103
The Noble Leper: Responses to Leprosy in the Twelfth and Thirteenth Centuries / Charlotte Pickard 119
Royal Taxation and Written Record in Eleventh-Century England and Ninth-Century West Francia / David Pratt 135
Early Royal Rights in the Liberty of St. Edmund (The Marjorie Chibnall Memorial Essay, 2018) / Richard Purkiss 155
Castle Construction, Conquest and Compensation (The Christine Mahany Memorial Lecture) / David Roffe 175
Four Scenes from the 'Chanson de Roland' on the Façade of Barletta Cathedral (Southern Italy) / Lucia Sinisi 193
‘The Jews are our Donkeys’: Anti-Jewish Polemic in Twelfth-Century French Vernacular Exegesis / Linda M. A. Stone 209

Citation preview

ANGLO-NORMAN STUDIES XLI PROCEEDINGS OF THE BATTLE CONFERENCE 2018

This year’s volume continues to demonstrate the vitality of scholarship in this area, across a variety of disciplines. There is a particular focus on the material culture of the Norman Conquest of England and its aftermath, from study of horses and knights to archaeologies to castle construction and the representation of a chanson de geste on an Italian church façade. The volume also includes papers on royal and private authority in Anglo-Saxon England, the relationship between AngloNorman rulers and their neighbours, intellectual history, priests’ wives, and noble lepers. Elisabeth van Houts is Honorary Professor of European Medieval History, University of Cambridge, and Fellow of Emmanuel College.

ANGLO-NORMAN STUDIES ISSN 0954-9927

Editor Elisabeth van Houts Editorial Board S. D. Church (University of East Anglia) Lindy Grant (University of Reading) Mark Hagger (Bangor University) Leonie V. Hicks (Canterbury Christ Church University) C. P. Lewis (Institute of Historical Research, University of London)

ANGLO-NORMAN STUDIES XLI PROCEEDINGS OF THE BATTLE CONFERENCE 2018

Edited by Elisabeth van Houts

THE BOYDELL PRESS

© Editor and Contributors 2018, 2019 All Rights Reserved. Except as permitted under current legislation no part of this work may be photocopied, stored in a retrieval system, published, performed in public, adapted, broadcast, transmitted, recorded or reproduced in any form or by any means, without the prior permission of the copyright owner First published 2019 The Boydell Press, Woodbridge ISBN 978–1–78327–399–7

ISSN 0954–9927 Anglo-Norman Studies (Formerly ISSN 0261-9857: Proceedings of the Battle Conference on Anglo-Norman Studies)

The Boydell Press is an imprint of Boydell & Brewer Ltd PO Box 9, Woodbridge, Suffolk IP12 3DF, UK and of Boydell & Brewer Inc. 668 Mt Hope Avenue, Rochester, NY 14620–2731, USA website: www.boydellandbrewer.com A catalogue record for this book is available from the British Library The publisher has no responsibility for the continued existence or accuracy of URLs for external or third-party internet websites referred to in this book, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate Aleksandra McClain and Naomi Sykes, “New Archaeologies of the Norman Conquest” is available as Open Access under the CC-BY licence This publication is printed on acid-free paper Typeset by www.thewordservice.com

CONTENTS

ILLUSTRATIONS AND TABLES vii EDITOR’S PREFACE ix ABBREVIATIONS

x

Horses, Knights and Tactics (The R. Allen Brown Memorial Lecture, 2018) Sally Harvey

1

Baldwin of Forde, Bartholomew of Exeter and the Authorship of the Liber de sectis hereticorum et orthodoxe fidei dogmata Sabina Flanagan

23

Evidence of the Ordinary: Wives and Children of the Clergy in Normandy and England, 1050–1150 Hazel Freestone

39

Anthropology, Feud and De obsessione Dunelmi 59 Tom Lambert New Archaeologies of the Norman Conquest Aleksandra McClain and Naomi Sykes

83

An Angevin Imperial Context for the Amboise–Anjou Narrative Programme 103 Nicholas L. Paul The Noble Leper: Responses to Leprosy in the Twelfth and Thirteenth Centuries Charlotte Pickard

119

Royal Taxation and Written Record in Eleventh-Century England and Ninth- 135 Century West Francia David Pratt Early Royal Rights in the Liberty of St Edmund (The Marjorie Chibnall Memorial Essay, 2018) Richard Purkiss

155

Castle Construction, Conquest and Compensation (The Christine Mahany Memorial Lecture) David Roffe

175

vi Contents Four Scenes from the Chanson de Roland on the Façade of Barletta Cathedral (Southern Italy) Lucia Sinisi

193

‘The Jews are our Donkeys’: Anti-Jewish Polemic in Twelfth-Century French Vernacular Exegesis Linda M. A. Stone

209



ILLUSTRATIONS AND TABLES

Hazel Freestone, ‘Evidence of the Ordinary: Wives and Children of the Clergy in Normandy and England, 1050–1150’ Figure 1

The hereditary archdeacons of Stow (c. 1090s–1187)

45

Figure 2

Family of the unnamed daughter of Dunprust, hereditary priest of 51 St Andrew’s chapel, Sutton, near Plympton (Plymouth), 1035–1176

Table 1

Number of married clergy by diocese and grade in Normandy

58

Table 2

Number of Married clergy by diocese and grade in England

58

Tom Lambert, ‘Anthropology, Feud and De obsessione Dunelmi’ Figure 1

Relevant descendants of Uhtred and Ecgfritha

61

Aleksandra McClain and Naomi Sykes, ‘New Archaeologies of the Norman Conquest’ Figure 1

Changing zooarchaeological representation of A) pigs and B) chickens through the Middle Ages, and by site type; from Naomi J. Sykes, The Norman Conquest: A Zooarchaeological Perspective, BAR International Series 1656, Oxford 2007

94

Figure 2

Temporal changes in age profiles for A) cattle and B) sheep from medieval rural assemblages; from Naomi J. Sykes, ‘From Cu and Sceap to Beffe and Motton: The Management, Distribution, and Consumption of Cattle and Sheep in Medieval England’ in Food in Medieval England: Diet and Nutrition, ed. C. M. Woolgar, Oxford 2005, 56–71

95

Figure 3

Scene from the Bayeux Tapestry, eleventh century, showing the Norman contingent cooking and dining; reproduced with special permission of the City of Bayeux

96

Nicholas L. Paul, ‘An Angevin Imperial Context for the Amboise–Anjou Narrative Programme’ Map 1

Main places mentioned in and around Amboise and in Anjou

102

Richard Purkiss, ‘Early Royal Rights in the Liberty of St Edmund’ Map 1

The eight-and-a-half hundreds of Bury St Edmunds within East Anglia © Richard Purkiss; hundred boundary data produced by the UCL Landscapes of Governance project

156

viii

Illustrations and Tables

Map 2

The distribution of servitium and consuetudo obligations in the eight-and-a-half hundreds according to LDB © Richard Purkiss; hundred boundary data produced by the UCL Landscapes of Governance project

164

Map 3

The distribution of avera and inward obligations in Cambridgeshire 167 according to GDB © Richard Purkiss; hundred boundary data produced by the UCL Landscapes of Governance project

Table 1

Dues recorded in LDB (1086) and Kalendar (1186 × 1191)

160

Table 2

LDB holdings in Bradfield, Thedwestry hundred

163

David Roffe, ‘Castle Construction, Conquest and Compensation’ Figure 1

Christine Mahany (1939–2016)

174

Table 1

Baronial castles and sake and soke in Lincolnshire

178

Table 2

Castles and compensation, 1066 to 1086

188

Lucia Sinisi, ‘Four Scenes from the Chanson de Roland on the Façade of Barletta Cathedral (Southern Italy)’ Figure 1

Charlemagne and boar, Barletta cathedral; photograph by Luisa Derosa

199

Figure 2

Boar, dog and young man (Roland?), Barletta cathedral; photograph by Luisa Derosa

201

Figure 3

Roland and Ferragut, Barletta cathedral; photograph by Luisa Derosa

203

Figure 4

Roland on horseback blowing an oliphant, Barletta cathedral; photograph by Luisa Derosa

205

The editors, contributors and publishers are grateful to all the institutions and persons listed for permission to reproduce the materials in which they hold copyright. Every effort has been made to trace the copyright holders; apologies are offered for any omission, and the publishers will be pleased to add any necessary acknowledgement in subsequent editions.

EDITOR’S PREFACE The forty-first Battle Conference of Anglo-Norman Studies took place from the 20th to the 23rd of July 2018 at Lucy Cavendish College, Cambridge. The Fellow Steward, Mrs Christine Houghton, arranged a warm welcome, exemplary logistical support, delicious meals and very welcome shade in the lush college gardens. We are most grateful for the financial support from the George Macauley Trevelyan Fund of the Faculty of History (University of Cambridge), and to the Master and Fellows of Emmanuel College for the use of the Old Library and Gardens for the drinks reception. We are indebted to Edward Mayes, Faculty of History, who dealt with the financial aspects of the conference. The fifth winner of the Marjorie Chibnall Memorial Essay Prize was Richard Purkiss (University of Oxford), whose research on the early royal rights of the Liberty of St Edmund was considered the unanimous choice of the panel of judges. This year’s Muriel Brown scholars were Hannah Boston, Catherine Healy and Daniel Talbot. For the first time this year, at the initiative of Dr Leonie Hicks, we held a successful poster session for graduate students. The University of Christ Church at Canterbury made available a prize for the best presentation, which was won by Daniel Talbot. As for the conference proceedings, as ever I am most grateful for the expert assistance provide by Caroline Palmer and Rohais Landon and the home team at Boydell & Brewer. After five years as director of the conference and editor of Anglo-Norman Studies I would like to express my warmest gratitude to my co-trustees of the Allen Brown Memorial Trust, Professors Stephen Church and Lindy Grant, and Drs Mark Hagger, Leonie Hicks and Chris Lewis, without whom these responsibilities would have been infinitely harder to carry out. I am delighted to hand over to Professor Stephen Church (University of East Anglia), who will succeed me in both posts. Elisabeth van Houts January 2019

ABBREVIATIONS AD AmHR ANS ASC

Archives départementales American Historical Review Anglo-Norman Studies Anglo-Saxon Chronicle, cited by year (corrected in square brackets if necessary) and manuscript; unless otherwise stated the edition is Two of the Saxon Chronicles Parallel, ed. Charles Plummer, 2 vols, Oxford 1892–9 ASC, trans. Swanton The Anglo-Saxon Chronicle, trans. and ed. M. J. Swanton, London 1996 Anglo-Saxon England ASE BAR British Archaeological Reports Bulletin of the Institute of Historical Research BIHR London, British Library BL BM Bibliothèque Municipale BnF Paris, Bibliothèque Nationale de France The Carmen de Hastingae Proelio of Guy, Bishop of Amiens, Carmen ed. and trans. Frank Barlow, Oxford 1999 Council for British Archaeology CBA Dugdale, Monasticon William Dugdale, Monasticon Anglicanum, new edn by Henry Ellis and Bulkeley Bandinel, 6 vols, London 1817–30 Eadmer, HN Eadmer, Historia novorum in Anglia, in Eadmeri Historia novorum in Anglia, et Opuscula duo; De Vita Sancti Anselmi et quibusdam miraculis ejus, ed Martin Rule, RS 81, 1884 English Episcopal Acta EEA Early English Text Society EETS English Historical Review EHR EME Early Medieval Europe English Lawsuits English Lawsuits from William I to Richard I, ed. R. C. van Caenegem, 2 vols, Selden Society 106–7, 1990–1 Freeman, Norman Edward A. Freeman, The History of the Norman Conquest Conquest of England, its Causes and its Results, 6 vols, 1st edn Oxford 1867–79, revised edn New York 1873–6 GDB Great Domesday Book, followed by the folio number, a or b (for recto or verso), cited from Domesday Book, seu Liber Censualis Willelmi Primi, 2 vols, London 1783, I, or from Great Domesday Book: Library Edition ed. Ann Williams and R. W. H. Erskine, Alecto Historical Editions, London 1986– 92; followed in parentheses by the abbreviated county name and the entry number (substituting an oblique for a comma between the first and second parts) used in Domesday Book, ed. John Morris and others, 34 vols, Phillimore, London 1974–86 Gesetze, Die Gesetze der Angelsachsen, ed. Felix Liebermann, 3 ed. Liebermann vols, Halle 1903–16

Abbreviations xi GND

The Gesta Normannorum ducum of William of Jumièges, Orderic Vitalis, and Robert of Torigni, ed. and trans. Elisabeth M. C. van Houts, 2 vols, Oxford 1992–5 Harmer, AS Writs F. E. Harmer, Anglo-Saxon Writs, 2nd edn, Stamford 1989 Howden, Chronica Chronica Rogeri de Houedene, ed. William Stubbs, 4 vols, RS 51, 1868–71 Howden, Gesta Gesta Regis Henrici Secundi [now attributed to Roger of Howden], ed. William Stubbs, 2 vols, RS 49, 1867 HSJ Haskins Society Journal Huntingdon Henry, Archdeacon of Huntingdon, Historia Anglorum: The History of the English People, ed. and trans. Diana Greenway, Oxford 1996 Inquisitio Eliensis, in Inquisitio Comitatus Cantabrigiensis IE [and] Inquisitio Eliensis, ed. N. E. S. A. Hamilton, London 1876 Journal of Ecclesiastical History JEH JL Regesta pontificum romanorum ab condita ecclesia ad annum post Christum natum MCXCVIII, ed. Philipp Jaffé, Wilhelm Wattenbach, S. Loewenfeld, and others, 2 vols, Leipzig 1885–8 Journal of Medieval History JMH John of Worcester The Chronicle of John of Worcester, ed. R. R. Darlington and P. McGurk, II-III, Oxford 1995–8 Codex Diplomaticus Aevi Saxonici, ed. J. M. Kemble, 6 vols, KCD London 1839–48 – cited by charter number Little Domesday Book, followed by the folio number and LDB a or b (for recto or verso), cited from Domesday Book, seu Liber Censualis Willelmi Primi, 2 vols, London 1783, II, or from Great Domesday Book: Library Edition ed. Ann Williams and R. W. H. Erskine, Alecto Historical Editions, London 2000; followed in parentheses by the abbreviated county name and the entry number (substituting an oblique for a comma between the first and second parts) used in Domesday Book, ed. John Morris and others, 34 vols, Phillimore, London 1974–86 Letters of Lanfranc The Letters of Lanfranc, Archbishop of Canterbury, ed. and trans. Helen Clover and Margaret Gibson, Oxford 1979 William of Malmesbury, Gesta pontificum Anglorum: The Malmesbury,   Gesta pontificum History of the English Bishops, ed. and trans. M. Winterbottom and R. M. Thomson, 2 vols, Oxford 2007 Malmesbury, William of Malmesbury, Gesta regum Anglorum: The  Gesta regum History of the English Kings, ed. and trans. R. A. B. Mynors, M. Winterbottom and R. M. Thomson, 2 vols, Oxford 1998–9 Malmesbury, William of Malmesbury, Historia novella: The   Historia novella Contemporary History, ed. Edmund King, trans. K. R. Potter, Oxford 1998 MGH Monumenta Germaniae Historica ODNB Oxford Dictionary of National Biography, usually cited from online edn (http://www.oxforddnb.com/view/article/), with article number and date accessed

xii Abbreviations OED OMT Orderic

Oxford English Dictionary Oxford Medieval Texts The Ecclesiastical History of Orderic Vitalis, ed. and trans. Marjorie Chibnall, 6 vols, Oxford 1969–80 P & P Past and Present PL Patrologia cursus completus, series Latina, ed. J.-P. Migne, 221 vols, Paris 1844–65 Poitiers The Gesta Guillelmi of William of Poitiers, ed. and trans. R. H. C. Davis and Marjorie Chibnall, Oxford 1998 PR The Great Roll of the Pipe for [regnal year, king], Pipe Roll Society; except for 2-4 Henry II, ed. Joseph Hunter, London 1844; 1 Richard I, ed. Joseph Hunter, London 1844; 26 Henry III, ed. Henry Louis Cannon, London 1918 Proc. Brit. Acad. Proceedings of the British Academy Recueil des actes des ducs de Normandie de 911 à 1066, ed. RADN Marie Fauroux, Mémoires de la Société des Antiquaires de Normandie 36, Caen 1961 Regesta Regum Anglo-Normannorum, 1066–1154, 3 vols (I, Regesta ed. H. W. C. Davis, Oxford 1913; II, ed. Charles Johnson and H. A. Cronne, Oxford 1956; III, ed. H. A. Cronne and R. H. C. Davis, Oxford 1968) Regesta: William I Regesta Regum Anglo-Normannorum: The Acta of William I (1066–1087), ed. David Bates, Oxford 1998 Robertson, A. J. Robertson, Anglo-Saxon Charters, Cambridge 1939   AS Charters RS Rolls Series (Chronicles and Memorials of Great Britain and Ireland during the Middle Ages, Published under the Direction of the Master of the Rolls). P. H. Sawyer, Anglo-Saxon Charters: an Annotated List and S Bibliography, London 1968, and with a revised and updated version largely edited by S. E. Kelly, available at http:// www.esawyer.org.uk/ Scriptores (in Folio) [in MGH] SS s.a. sub anno, annis (‘under the year, years’) sub verbo, verbis (‘under the word, words’) s.v. Tabularia Tabularia: Sources écrites de la Normandie mediéval [online journal: www.unicaen.fr/mrsh/craham/revue/tabularia/] Telma Chartes originales antérieures à 1121 conservées en France, ed. C. Giraud, J.-B. Renault et B.-M. Tock, Nancy: Centre de Médiévistique Jean Schneider; éds électronique: Orléans: Institut de Recherche et d’Histoire des Textes, 2010, http:// www.cn-telma.fr/originaux/ Date de mise à jour: Première version, 10 juin 2010 tempore regis Eadwardi (‘in King Edward’s time’) TRE TRHS Transactions of the Royal Historical Society TRW tempore regis Willelmi (‘in King William’s time’) VCH The Victoria History of the Counties of England [with county name], in progress Vita Ædwardi The Life of King Edward who Rests at Westminster, Attributed to a Monk of Saint-Bertin, ed. and trans. Frank Barlow, 2nd edn, Oxford 1992

Abbreviations xiii Wace, trans. Burgess The History of the Norman People: Wace’s Roman de Rou, trans. Glyn S. Burgess, Woodbridge 2004 Whitelock, AS Wills D. Whitelock, Anglo-Saxon Wills, Cambridge 1930 × [The form 1066 × 1087 indicates an uncertain date within the range]

The R. Allen Brown Memorial Lecture, 2018

HORSES, KNIGHTS AND TACTICS* Sally Harvey Much has been written on the subject of warfare, knights and horses since I published on the small knightly holdings of late eleventh-century England in 1970.1 Since then, the interpretation of late Anglo-Saxon England as a functioning state with a competent fiscal and coinage network has won general acceptance. Those competences, of course, served only to heighten the attraction of the country to the militarily ambitious from overseas and within. The successful Normans utilized all possible avenues of Anglo-Saxon governance; however, the two regimes differed most visibly in the construction of castles, and so in one area at least, that of military organization. Wide-ranging and nuanced studies of the knightly classes have appeared.2 And whilst it is not always easy to link the evidence and conclusions of specialist studies, several themes do recur. One is the indispensable role of ships and naval expertise in the late Anglo-Saxon military,3 and the list of Norman magnates who contributed ships for the invasion, now accepted to have an authentic origin, confirms the crucial part naval power played in Duke William’s preparations within Normandy for conquest.4 Another is the importance of the Norman kings’ extended military household, providing ‘a well-trained and adequately paid professional military force’ for the command and garrisoning of royal castles, and enabling reaction from his forces in the field as a mounted corps d’élite.5 Although horses have always featured picturesquely in the military story of the Norman conquest, their role remains surprisingly neglected in mainstream histories; yet the distance horses might normally travel, the power and protection they supplied, and the stresses of contemporary conditions on them were essential constituents of political decisions and military action. Exceptional were R. C. Smail’s *

I wish to thank Liesbeth van Houts for her generous assistance with the preparation of this article. ‘The Knight and the Knight’s Fee in England’, P & P 49, 1970, 3–43; reprinted in Peasants, Knights and Heretics: Studies in Medieval English Society, ed. R. H. Hilton, Cambridge 1976, 133–73, hereafter Harvey. Amongst ‘an explosion’ of publications, see Stephen Morillo, Warfare under the Anglo-Norman Kings 1066–1135, Woodbridge 1994, 5–7. His book analyses many of the practicalities of fighting and campaigns, including the conclusion that the role of cavalry should be re-evaluated (191). 2 E.g. D. Crouch, The Image of Aristocracy in Britain, 1000–1300, London 1992; P. Coss, The Knight in Medieval England 1000–1400, Stroud 1993; M. Strickland, War and Chivalry: The Conduct and Perception of War in England and Normandy, 1066–1217, Cambridge 1996. 3 N. Hooper, ‘Some Observations on the Navy in Late Anglo-Saxon England’, in Studies in late AngloSaxon and Anglo-Norman Military Organisation and Warfare: Studies in Medieval history presented to R. Allen Brown ed. C. Harper-Bill, C. J. Holdsworth and J. L. Nelson, Woodbridge 1989, 203–13; reprinted in Anglo-Norman Warfare, ed. M. Strickland, Woodbridge 1992, 17–27. 4 E. M. C. van Houts, ‘The Ship List of William the Conqueror’, ANS 10, 1987, 159–83. 5 M. Chibnall, ‘Mercenaries and the Familia Regis under Henry I’, History 62, 1977, 15–23; J. O. Prestwich, ‘The Military Household of the Norman Kings’, EHR 96, 1981, 1–35 at 6, 9–11, 20–7; reprinted in Anglo-Norman Warfare ed. Strickland, 93–127. 1

2

Sally Harvey

work on Crusading Warfare, which paid due attention to horse power, and Ralph Davis’s wide-ranging and under-used works on the Norman horse.6 Somewhat more recently, realistic studies have been produced by Howard Clarke, John Gillingham and Andrew Ayton for Maurice Keen’s Medieval Warfare;7 and substantial monographs by the equine expert Ann Hyland, and by the archaeological historian John Clark, have furthered our knowledge of the history of the horse.8 But their works are not yet fully integrated into the general perception of the eleventh and twelfth centuries. By contrast, better documentation of the warhorse in the later medieval period has enabled Andrew Ayton’s studies of the horses required for later assaults on France and Scotland, adding much to H. J. Hewitt’s books on breeding and the logistics of mounting English armies in the thirteenth and fourteenth centuries.9 Some logistical questions remain open: one concerns the balance in army personnel. The long overlooked, yet often crucial, role of bowmen and crossbowmen in the campaigns of the eleventh and early twelfth centuries, and their interaction with cavalry, have now been brought into focus.10 Questions persist over castles and seigneurial castleguard in post-Conquest England. Recorded castle-guard obligations were inadequate to provide the complement for castle garrisons and simply unworkable in war-time conditions.11 Were remote rural castles manned with a skeleton staff?12 Did castles serve primarily to intimidate, dominate, or protect a region?13 Many functioned rather as fortified residences and centres of consumption and rent-collecting.14 Matthew Strickland has depicted the ease with which the Scots caused havoc for significant distances into England in the eleventh and twelfth centuries, with raiding forces simply by-passing rural strongholds.15 Nevertheless, the established tenure of castles or fortified towns remained key to the continued economic exploitation of a region.16

R. C. Smail, Crusading Warfare (1097–1193) first edn Cambridge 1956, 2nd edn. Cambridge 1995; R. H. C. Davis, ‘The Warhorses of the Normans’, ANS 10, 1987, 67–82; R. H. C. Davis, The Medieval Warhorse, London 1989. 7 Medieval Warfare, ed. M. Keen, Oxford 1999. 8 A. Hyland, The Endurance Horse, London 1988; A. Hyland, Equus: The Horse in the Roman World, London 1990; The Medieval Warhorse, Stroud 1994; The Warhorse, 1250–1600, Stroud 1998; A. Hyland, The Horse in the Middle Ages, Stroud 1999; J. Clark, ‘Medieval Horseshoes, Finds Research Group 700–1700’, Datasheet 4, Coventry Museum 1986; idem, The Medieval Horse and its Equipment, Woodbridge 2004; first published London 1995. 9 A. Ayton, Knights and Warhorses: Military Service and the English Aristocracy under Edward III, Woodbridge 1994, 194–257, Tables, 263–71; H. J. Hewitt, The Organisation of War under Edward III, Manchester 1966, esp. chapter 5; H. J. Hewitt, The Horse in Medieval England, London 1983. 10 Although Smail Crusading Warfare, propounded the tactical interaction employed in the Latin kingdoms and R. Hardy’s, Longbow, Cambridge 1976, has been reissued many times. Norman and other engagements are studied in M. Strickland, ‘The Bow and the Crossbow in Early Medieval Warfare’ and ‘Dismounted Knights and Archers in Anglo-Norman England’, in M. Strickland and R. Hardy, From Hastings to the Mary Rose: The Great War-bow, Stroud 2005, 51–96. 11 J. Moore, ‘Anglo-Norman Garrisons’, ANS 22, 1999, 205–59 at 205–9. 12 M. Strickland, ‘Securing the North: Invasion and the Strategy of Defence in Twelfth-Century Anglo-Scottish Warfare’, in Anglo-Norman Warfare ed. Strickland, 208–29. 13 For examples, see Morillo, Warfare under the Anglo-Norman Kings, 94–7. 14 E.g. T. E. McNeill, ‘Great Towers of Early Irish Castles’, ANS 12, 1989, 99–117. 15 Similarly William the Lion in 1173 and 1174, see Strickland, ‘Securing the North’, 214–19. 16 See further, A. Ayton and J. L. Price, ‘Introduction’ to The Medieval Military Revolution, London 1995, 7; P. Edbury, ‘Warfare in the Latin East’, in Medieval Warfare, ed. Keen, 89–108, 98. J. Bradbury, ‘Battles in England and Normandy, 1066–1154’, in Anglo-Norman Warfare, ed. Strickland, 182–93 at 182; J. Gillingham, ‘William the Bastard at War’, in Anglo-Norman Warfare, ed. Strickland, 143–60, 150–7.

6



Horses, Knights and Tactics

3

A recurrent theme in recent studies is commanders’ avoidance of set battles, with their dangers to the chief protagonists through death or capture, which could be decisive or fatal to a campaign; for example the capture of Duke Robert ‘Curthose’ at Tinchebrai in 1106 led to his loss of Normandy and his lifelong imprisonment.17 The principles were expounded in the late Roman handbook on war, De re militari by Vegetius, frequently copied or translated throughout the Middle Ages.18 Every military textbook advised against risking full-scale battle: God’s judgement on the issue could not be foreseen. Successful commanders found that threatening principalities when they were disadvantaged by an aged or an inexperienced ruler – pillaging his hinterland, disrupting his economy, and undermining his credibility as over-lord – was generally the safer and often more lucrative strategy.19 Then there are the knights’ horses – my concern here. All commentators agree that being mounted was the sine qua non of being a knight,20 yet how does the motif of knights as heavily armed cavalry, mounted on expensively maintained horses trained to perpetrate or withstand a charge, fit with campaigns of surprise movements and pillaging? (The later term for a raid, la chevauchée, indicates how inseparable was the horse from this mode of warfare.) Also, how did knights, along with the horses that defined their status, fit into military campaigns focused on the defence or besieging of castles and towns? The presence of horses in a castle under siege would rapidly intensify demands on supply: apart from needing fodder in bulk, a healthy horse drinks between 6 and 24 gallons of water a day; however, its blood and meat formed a reservoir of liquid and food in real emergencies.21 Mounted knights were of little use to the besieged, except for getting messages to relief forces. Most of the first generation of keeps or mottes would not accommodate horses, which were presumably kept within relatively vulnerable banked and palisaded baileys. A Challenge Revived In 1970, I argued that in Domesday Book many knights (milites) were provided with a base on fragments of large demesne manors and that over a third had between ¾ and 1½ hides of land with a mean annual value of 33 shillings.22 (Here, I simply use the word ‘knight’, miles, to indicate any mounted man well-equipped for fighting, but not an archer.) That study of knights and their holdings gained widespread, but not total, acceptance, and received a head-on challenge.23 Allen Brown would not accept the existence of knights holding as little as 1–1½ hides of land, arguing that such an interpretation lacked understanding of the demands of the knightly ethos, 17

Tactics discussed Gillingham, ibid., 144–7; J. Gillingham, ‘An Age of Expansion, c.1020–1204’, in Medieval Warfare, ed. Keen, 59–88, 78–81. 18 P. Contamine, War in the Middle Ages, Oxford 1986, 210–12. 19 E.g. Malmesbury, Gesta Regum 1, 438. F. Barlow, William I and the Norman Conquest, London 1965, 26–36, is succinctly realistic; J. Gillingham, ‘Richard I and the Science of War in the Middle Ages’, in Anglo-Norman Warfare, ed. Strickland, 194–207, 197–207. M. Strickland discusses ravaging and attrition in War and Chivalry, 273–81. J. Barker, Agincourt: The King, the Campaign, the Battle, London 2005, 265; Smail discusses the question of battles in warfare, Crusading Warfare, 135–7, 165–8. 20 E.g. Crouch, Image of Aristocracy, 124–5 discussed and qualified Strickland, War and Chivalry, 19–24. 21 E.g. Gesta Francorum, ed. R. Hill, London 1962, 62. 22 Harvey, 15–16, 20–1. 23 Reviewed in N. Brooks, ‘The Archbishop of Canterbury and the So-Called Introduction of Knight-Service into England’, ANS 34, 2010, 41–62 at 42–4.

4

Sally Harvey

and especially of the size, quality and cost of their horses; the interest in my argument showed the extent of historians’, and my, ignorance of horses.24 Since Allen Brown long ago threw down the gauntlet, maybe it is fitting for me, as an almost life-long owner of horses ranging in size from 10 to 16 hands, somewhat belatedly to pay tribute to him and his vibrant views here by jousting with him on these tenets. (A ‘hand’ is 4 inches: a traditional measurement using the hand’s width upwards to a horse’s withers – the bony lump in front of the saddle – usually the highest point of its back.) It is time to consider the evidence for the type of horse brought by the Norman knights, and to address whether the calibre of their cavalry contributed in any direct way to the Norman conquest of England. This leads me to reflect on the less-publicized advantages of being mounted in military action generally. It is not easy to deploy the many contemporary clues to depict the fast-evolving late eleventh-society in the round. The aim of my 1970 study was to gain some insight into arrangements made with knights who perhaps had not received their land ‘through friends and money’ – as Roger de Lacy’s own charter records of his under-tenancy.25 Within the context of the first generation of Norman settlement, itself inherently unstable, Domesday Book provides very specific data on the type of land that some knights were allocated, which it would be wasteful not to utilize.26 That data is surprisingly consistent. Kent apart, the holdings between 1 and 2 hides in size are so recurrent that they seem to be contemporaries’ idea of a norm, however transitory that may have been.27 That is not to deny the existence of well-connected knights, often described by chroniclers as ‘well-equipped’ or ‘belted knights’. These were distinguished by a formal girding with the ‘belt of knighthood’, a ceremony acknowledging the social distinction and public function of the recipient, as has been noted from the Carolingian period onwards.28 When the future Henry I was knighted as a young man, Archbishop Lanfranc himself girded him with the ‘belt of knighthood’.29 Those so accoutred had received recognition from power, and their presence among the higher landholding levels of society and their sons is not in dispute. 24

R. A. Brown, Origins of English Feudalism, London 1973, 25–6; ‘The Status of the Norman Knight’, in War and Government in the Middle Ages, ed. John Gillingham and J. C. Holt, Woodbridge 1984, 18–32, 22–30; reprinted in Anglo-Norman Warfare, ed. Strickland, 128–42. He had later support in D. F. Fleming, ‘Landholding by Milites in Domesday Book: A Revision’, ANS 13, 1991, 83–98. 25 per amicos et per pecuniam requisiuit, V. H. Galbraith, ‘An Episcopal Land-Grant of 1085’, EHR 175, 1929, 352–72 at 352, 372. 26 Although Donald Fleming questioned my figures on the mean of the annual values of the lands of Domesday milites, yet on closer reading, and ‘following my assumptions’, our differences on these are marginal; similarly, his calculations on size in hides or carucates in land, ‘bear out the accuracy’ of mine, Fleming, 91, n. 41, 88, cf. Harvey, 15–16, 19–24. The disparity between us properly lies in our respective assumptions on the presence and costs of heavy cavalry, the lord’s contribution to expenses, and the use of Domesday evidence, see also Coss, The Knight, 10–12, 24–6. For Evesham Abbey’s military tenures of the Norman and Angevin periods traced from Domesday onwards, see H. B. Clarke, ‘“Those Five Knights which you owe me in respect of your Abbacy”. Organizing Military Service after the Norman Conquest and Beyond: Evesham and Beyond’, HSJ 24, 2012, 1–39: military tenures holding villein land at 11. 27 Even if Kentish values in general stand out from the rest, Kent also being blatantly under-taxed, F. W. Maitland, Domesday Book and Beyond, London 1960, 464–7, 536–7; S. P. J. Harvey, ‘Taxation and the Economy’, in Domesday Studies, ed. J. Holt, Woodbridge 1987, 249–64. 28 R. le Jan, ‘Frankish Giving of Arms and Rituals of Power: Continuity and Change in the Carolingian Period’, in Rituals of Power from Late Antiquity to the Early Middle Ages, ed. F. Theuws and J. L. Nelson, Leiden 2000, 301. Later, the belt of knighthood gave distinction above those who were simply distrained to knighthood in order to perform administrative services, Barker, Tournament, 113–14. 29 militiæ cingulum, Orderic 4, 120–1.



Horses, Knights and Tactics

5

England in the eleventh and twelfth centuries comprehended milites from all levels of landholding, including ‘one-hiders’.30 For some, their lords might well have furnished their mounts and equipment and contributed to their expenses, but they, nevertheless, went into action alongside the elite.31 Of course, every gradation of landholding occurred in between, and extra parcels of land would form excellent incentives or rewards.32 There was also the opportunity of spoils in the form of horses and precious metals.33 The smaller knightly landholders of Domesday, and other texts, should not be denied their function because of historians’ own ideals of knightly conduct.34 On their ‘minimum wage’ in the form of land, the nameless Domesday milites could have supported their old age should they reach it, and, before that, their small and active horses. Now to detail. First, holdings of 1 or 2 hides of land were not to be despised (a hide being a nominal 120 acres). King William did not do so. He gave his time and name to a charter restoring a single hide of land to a London moneyer; the moneyer, founder of a great dynasty, was evidently content to receive it.35 Moreover, by 1086, not one of the three surviving sons of William the Conqueror had been given so much as a hide of English land, when Robert, the eldest, was about 32 years of age and William Rufus about 26.36 One leg of the argument that a knight’s income must necessarily be substantial was the high cost of horses. Yet many knights had their horses provided, or at least replaced, by their lord. In the campaigns of William I, William Rufus and Henry I the king’s military household had a core role. These professional knights, of both high and low status, were augmented in 1101, inter alia, via a money-fief agreement with the count of Flanders: for a £500 annual payment, the count undertook to serve with 500 knights in Maine or with 1,000 knights, each with three horses, in England or Normandy.37 Its terms make two points clear: that both the count of Flanders and Henry I thought an annual payment of 10s. or £1 sterling a head was adequate contribution towards the maintenance, equipment and availability of these knights. Second, importantly, the agreement included a clause that the king should compensate the count’s knights for their losses (and horses were their most vulnerable assets), ‘as is his custom concerning his household troops (familia)’. It was an expected responsibility. New arrivals in the royal and other households would doubtless be accompanied by their own mounts, whose quality would vary with the individual; but they did not have to furnish replacements. Moreover, such an arrangement might well deter a ruler from contemplating a hopeless strategy or campaign, thereby providing indirect protection for the knights concerned.

30

F. Stenton, First Century of English Feudalism, Oxford 1961, 142. For example, in the Gâtine the word miles described whoever fought on horseback, G. T. Beech, A Rural society in Medieval France: The Gậtine of Poitou, Baltimore 1964, 94. 31 E.g. Smail, Crusading Warfare, 106–11. 32 S. D. B. Brown discusses fiefs rentes and future hopes in ‘Military Service and Monetary Reward in the Eleventh and Twelfth centuries’, History 74, 1989, 20–38, at 24–9. 33 Chibnall, World of Orderic, 45. 34 Compare, R. A. Brown, English Feudalism, 25 with the underhand actions revealed in S. Anglo, ‘How to Win at Tournaments: The Technique of Chivalric Combat’, Antiquaries Journal 68, 1988, 248–64. 35 Regesta: William I, no. 107, pp. 386–7; for other examples, see nos. 352, 268. 36 For the ages of William’s children, D. C. Douglas, William the Conqueror, London 1964, 391–5 and F. Barlow, William Rufus, London1983, 441–5. 37 Diplomatic Documents preserved in the PRO, ed. P. Chaplais, 2 vols, London 1964, I, no.1, 1–4; translated, E. Van Houts, ‘The Anglo-Flemish Treaty of 1101’, ANS 21, 1998, 169–74, caps 8 and 14, at pp. 171 and 172.

6

Sally Harvey

The tradition held good throughout the Middle Ages. In chivalric literature, and earlier, the mark of a good lord was to give warhorses to his knights, or horses to his followers, and the best lords were renowned as a source. Between 1008 and 1012, Alfwold, bishop of Crediton, willed to each of his retainers, hired men, the steed which he had lent him.38 When the legendary King Arthur dubbed 400 knights, sons of counts and kings, he gave them three horses each to demonstrate his generous lordship.39 The supply, or replacement of (or compensation for), horses was customary for those in royal and other employ in the thirteenth and fourteenth centuries, when the practice of recording was becoming more formalized.40 It also became common practice, following tournaments and jousts, for lords to provide or pay for replacement mounts for those who fought under their colours.41 The cost of his mount was not necessarily the responsibility of the knight. Nor, with exceptions to be made for the leaders’ mounts, was the cost of most horses as formidable as formerly maintained.42 Norman charters certainly feature horses changing hands for sums between 14s. 6d. and £20.43 Yet, Norman values are less impressive once converted into English currency. Norman coin was both debased – less than half fine – and light; it represented something of the order of only one third of the value of English coin or silver (although large sums were likely to have been taken by weight).44 Individual costly animals certainly feature in chronicles, too: on one occasion we are told that King William Rufus’s horse had just cost him £10 in a context implying a goodly price paid.45 But the likelihood is that such exchanges were confined to the ruling elite and to monastic houses, as in succeeding centuries.46 Whether the majority of fighting cavalrymen who fought at Hastings were accompanied by the highest level of horse-power is questionable. Estimates of total combatants range widely between 6,000 and 30,000.47 If only 2,000 putative knights of the Norman invasion did indeed boast two stallions worth £10 Norman apiece, adjusted merely to £5 sterling, their animals would have a total value of £20,000 sterling. This, when the total annual revenues to landholders from the much sought-

38

The Crawford collection of Early Charters and Documents, ed. A. S. Napier and W. H. Stevenson, Oxford 1865, no. xx; Davis, ‘Warhorses of the Normans’, 82. 39 Chretien de Troyes, quoted in L. Ashe, ‘William Marshal, Lancelot, and Arthur: Chivalry and Kingship’, ANS 30, 2007, 19–40 at 24. 40 M. Prestwich, War, Politics and Finance under Edward I, London 1972, 50–1; Ayton, Knights and Warhorses, esp. 86–7. Although practice might vary, Barker, Agincourt, 115. 41 J. R. V. Barker, The Tournament in England 1100–1400, Woodbridge 1986, 174. 42 Brown, ‘Status of the Norman Knight’, 28–30. 43 Davis, ‘Warhorses of the Normans’, 77–8; see also Regesta: William I no. 173, £10; no. 164, £30 (Le Mans). 44 P. Spufford, ‘Coinage and Currency’, Cambridge Economic History of Europe, Cambridge 1963, III, 582; F. Dumas, ‘Les Monnaies Normandes (x-xii siècles) avec un répertoire des trouvailles’, Revue Numismatique, 6 ser., 21, 1979, 84–103, 89–96, 103; P. Grierson, ‘The Monetary System under William I’, in Domesday Book: Studies, Alecto Historical Editions, London 1987, 112–18. 45 Malmesbury, Gesta Regum, I, 550. 46 Examples from Jumièges and Fécamp in Davis, ‘Warhorses of the Normans’, 77–8, and from Mont Saint-Michel and Saint-Evroult in M. Chibnall, The World of Orderic Vitalis, Oxford 1984, 55. Survey B shows that Burton Abbey maintained 70 brood mares and foals at Burton (1114–18), English Historical Documents, II, ed. D. C. Douglas and G. W. Greenaway, 2nd edn. 1984, 884. 47 Strickland, The Great Warbow, 63 suggests between 7,000 and 14,000 combatants. For a recent survey, see D. Bates, William the Conqueror, New Haven and London 2016, 228–30. See also The Battle of Hastings: Sources and interpretations, ed. S. Morillo, Woodbridge 1996, xxii-xxx and C. M. Gillmor, ‘Naval Logistics of the Cross-channel Operation, 1066’, ANS 7, 1984, 105–31 at 114–28.



Horses, Knights and Tactics

7

after agriculture of the whole of recorded England amounted to some £73,000 in 1086, after pressure from the Norman landlords.48 It is hardly plausible that the 1066 invasion forces were already equipped with horse-power of this value before they acquired their English revenues (even though William’s forces included adventurers of status from outside Normandy). Pack-horses were required in addition. Although an experienced general, William’s power on the formal battle-field was not invariably overwhelming.49 A variety of animals would, necessarily, be drawn into the English conquest enterprise, many more suitable for the prevailing conditions of travel and tactics than heavy cavalry. Further, as commonly practised, William needed to leave some men – and horses – in Normandy, in case of attack in his absence. Certainly, William had carefully covered his back.50 He had spent his ducal reign in subduing neighbouring states – or allying with them: with his own marriage to Matilda of Flanders, and that of his sister, Adelaide, to Eustace of Boulogne’s younger brother, and last but not least, in over-awing Guy of Ponthieu and Conan of Brittany, and probably by assassinating his rival, Ralph of Mantes. Anjou, too, had been repulsed and was now in no position to strike again.51 Both Eustace and Guy commanded shorter crossings to English ports than William himself had, and William took the further precaution of taking his rival, Eustace, and also forces from Maine with him on the conquest venture. Horse size and type affect logistical problems of transport and the costs of maintenance, including the quantity of fodder required. An animal of 13 hands might commonly be little over half the weight of one of 15 hands, needing half the food. At 11 hands, its needs are halved again. If the animal had either native or feral pony blood or Arab/Barb blood, its nutritional needs would be further reduced. The Barb is a hot-blooded horse from the Berber coast of the Mediterranean – Libya, Tunisia, Algeria and Morocco – rather smaller and of slighter build than the Arabian horse, and arguably a better performer on sloping ground.52 ‘Native’ ponies require some bulk fodder, but little grain; they not only do not need – they should not receive – over-rich food. Such ponies are not weaklings. Today, while small ponies with powerful necks and shoulders are avoided as mounts for children because these animals are too strong, those are respected features in stallions of several small ‘native’ British breeds, making them good mounts for adults. Size and weight also determine the manageability of horses on the ground: once a horse’s shoulders are higher than the handler’s the task becomes less easy. Here I argue that eleventh-century horses, however well bred, were small. With a hand being 4 inches, the difference in height between an 11-hand and a 15-hand horse is merely a nominal 16 inches; but the difference in total bulk and, thus, fodder needs is dramatic. It is true that William of Jumièges writes of William filling 48

H. C. Darby, Domesday England, Cambridge 1977, 227. Orderic, II, 350–1; D. Bates, Normandy before 1066, London 1982, 245–6; Gillingham, ‘William the Bastard’, 144–7. 50 E.g. Poitiers, 74–7. 51 W. Scott Jessee, ‘The Angevin Civil War and the Norman Conquest of 1066’, HSJ 3, 1991, 101–9 at 102–6. 52 Also, Davis, Medieval Warhorse, 34, 50, 119–22. The short-hand terms ‘hot-blooded and ‘coldblooded’ have nothing to do with blood temperatures. They denote horses of Arab/Thoroughbred types evolved in hot climates, light in build and with a fine coat, and those from northern and mountainous regions, stocky in build, resistant to cold, and heavily coated, respectively. Their temperament tends to differ: hot-bloods being fiery, cold-bloods more phlegmatic. ‘Warm-bloods’ have evolved from crosses between the two. 49

8

Sally Harvey

his fleet with ‘mighty horses and most valiant men with hauberks and helmets’,53 however, even a rise of 4 inches in horse height makes for a considerable difference in ‘substance’, in bone, breadth and length. Development of the ‘great horse’ came later, provoked by the priorities of the charge, the formalized tournament, the increasing weight of human armour, and the adoption of horse-armour. The medieval ‘great horse’ itself was not over-tall, possibly about 14 hands, simply more substantially built than his forerunners. The well-known 1330s depiction of Sir Geoffrey Luttrell setting out for a tournament shows a horse significantly more substantial than was the earlier medieval equivalent: the knight’s lady hands his helm up to the mounted figure.54 Yet were the Luttrell lady 5 feet 2 inches in height – that is, the average for medieval women excavated from cemeteries – even this undeniably ‘great horse’ might well be under 15 hands.55 Its ‘Roman’ nose indicates heavy animals in its pedigree – the antithesis of Arab blood. Heavier build gave the combination of knight and mount irresistible tank power, yet could make it difficult for the horse to wheel and manoeuvre: these conflicting qualities of conformation might well be responsible for the specialized selection during the fourteenth century of horses for jousting purposes.56 In contrast, a ‘courser’, of similar height but with speed and stamina, was often the preferred mount of those who had the choice.57 The desirable warhorse was an evolving type, changing with the demands made on it – often not quickly enough. It is worth emphasizing that height does not correlate directly with either stamina or speed. Small mounts are hardier, more sure-footed and less susceptible to sprains and strains, especially on uneven ground. Large working farmers in Iceland still ride local animals of about 12 hands for all-day mobility on their farms. And the longer the distances involved, the lighter in build the better. Caught on the move in ‘trappy’ country – woods, defiles, sand, wet ploughed fields or boggy terrain – heavily equipped cavalry were vulnerable to ambush, and knights could not easily fight on the march.58 Downhill at faster paces, smaller, nimbler animals are safer. Large animals tend to ‘over-reach’ downhill, shod hind feet slicing into the delicate parts of the front legs above the hoof, or into the canon bone or tendons; even in the open fields, ridge and furrow, approached at an angle, is not easily negotiated by a large horse at speed. A horse is a top-heavy entity anyway – more so with a rider, particularly one in armour: large horses are more prone to lose their balance and fall. Moreover, in hand-tohand fighting against foot soldiers, as in eleventh-century England, knights on small mounts were in a good position to inflict telling blows at the necks of foot-soldiers, as clearly depicted in the Bayeux Tapestry.59 Given the high pommel and cantle of the wooden saddles, leaning far out or down from a large horse was hazardous. A stallion, we should note, was not necessarily the outstanding animal in the eleventh century that it is in the twenty-first century, when stallions are kept for specialized racing, dressage or showing, with breeding the ultimate aim and when

53

GND II, 164–5: tam valentibus equis quam hominibus robustissimis cum loricis et galeis; for diagrams, see Clark, Medieval Horse, 170–2. 54 Frontispiece, The Luttrell Psalter, ed. E. G. Millar, London 1932. 55 Clark, Medieval Horse, 24–32. 56 Barker, Tournament, 174–5. 57 Ayton, ‘Arms, Armour and Horses’, in Medieval Warfare ed. M. Keen, 186–208, at 191. 58 E.g. Smail, Crusading Warfare 156, 179, 183–4. 59 The Bayeux Tapestry, ed. F. Stenton, New York 1957, no.70; The Bayeux Tapestry, ed. D. M. Wilson, London 1985, plate 69.



Horses, Knights and Tactics

9

an entire (ungelded) male horse requires a government licence. In the Middle Ages, probably few male horses would have been gelded because of the dangers of infection or accident during the process.60 Stallions, with their confidence and competitive aggression, were the chosen animals for battle. They can be fairly tractable if well schooled, regularly exercised, and not mixed with mares to compete over. However, a fit, warm-blooded stallion of size is a being of considerable strength; and even well-schooled horses are liable to panic when hurt – although their social instincts lead some horses to carry out trained manoeuvres despite being fatally injured. Closer focus on equine power is instructive. Horses do not like violence and, as their natural instinct is flight, tend to panic if trapped in an aggressive and unfamiliar situation. It was undoubtedly necessary for knights to train their horses in groups, since horses gain confidence from more experienced horses and get accustomed to what they do routinely. Joint schooling was certainly essential for horses involved in fighting units or tournaments.61 Relevant here is that knights in Domesday frequently appear in small groups with lands pertaining to demesne manors at central places. Importantly, the smaller or lighter-weight equines available in the eleventh century were more suitable for the Norman invasion than was the postulated warhorse. Many eleventh-century horses were probably about 10–12 hands. Indeed, Ralph Davis’s studies of the warhorse of the Normans came to similar conclusions, but they fell on stony ground. Even of the medieval warhorse in general he judged that: ‘only a handful of horses (those costing 20 or 30 pounds) would have approached the weight of 1,300 lbs or the height of 14 hands’.62 He was right about the height, but these alternatives were misleading, being two disparate assertions (which came about by following another historian’s figures in the course of disputing with him).63 In fact, it requires a horse of ‘substance’ approaching 16 hands to reach the weight of 1,300 lbs.64 Horses for today’s endurance competitions have an ideal weight of around 800–950 lbs.65 A light-weight 14-hand mount might be only about 850 lb. What is the evidence for this revision of size, now advocated in varying degrees by several authorities but still unrecognized generally? Most tangible is the archaeological. Anglo-Norman horse shoes from London excavations show that horse shoes of only about 3½ to 4 inches wide, or less, and very light at c. 3 oz, prevailed before the mid-twelfth century. The same shoes would also fit a taller, more lightly built animal. Study of horse shoes and skeletal remains has led John Clark to conclude that before the thirteenth century most horses were ‘decidedly small’, well under 13 hands. However, during the thirteenth century, horse shoes increased in both 66 size and weight, ranging from 4 oz to 7. Interestingly, C. Gleditz judges from his collected data that the difference between the purchase cost of an ordinary work-

60

Military horses were not usually gelded in the Roman world, although it was Scythian practice, Hyland, Equus, 80–2. 61 Brown, ‘Status’, 28, 32; Barker, Tournament, 19–22, 139–45. 62 Davis, ‘Warhorses of the Normans’, 80. 63 An error unfortunately compounded by a mis-print giving the alternative as 13,000 lb, ibid. 64 S. Pilliner, Getting Horses Fit, Oxford 1986, 40, 46. 65 Hyland, Endurance Horse, 91. 66 J. Clark, Medieval Horseshoes, Department of Medieval Antiquities, Museum of London, Finds Research Group 700–1700, datasheet 4, Coventry Museum 1986, 4; Davis, Medieval Horse, 29–32.

10

Sally Harvey

beast and a destrier increased in the century after the 1180s.67 Both weight and width of horse shoes increased again in the fourteenth and fifteenth centuries, and it is on these that images of the knightly horse seem predicated. Picturesque incidents in historical literature presuppose a small horse. In the Carmen story of Hastings, probably written within three years of the battle, on the death of his first horse, Duke William signalled to an injured knight from Maine to relinquish his own mount. The knight, caught in a terrifying dilemma, refused. But William, now on foot, turned on the knight furiously and, seizing him by the nasal protection on the helmet, pulled him head over heels off his horse, which William then speedily mounted.68 Even for William, probably about 5 feet 10 inches in height,69 this would be feasible only if the knight’s horse were 13 hands or under, and the incident credible only if small horses were commonplace. After William Rufus had had his horse killed under him by an enemy knight who did not recognize him, irate and in full armour, he vaulted onto a replacement, without waiting for a mounting aid.70 There is also the story of the Richard who became count of Aversa, Italy, in the 1050s, and subsequently prince of Capua, who preferred to ride with his feet able to touch the ground.71 It attracted comment, but was not beyond the pale as it would be today. Then there is the artistic evidence: first and foremost, the Bayeux Tapestry. This pictorial embroidery regularly depicts the knights’ feet, whilst mounted, to be at or even below the knee of the horse. Not only the Bayeux Tapestry so represents knights’ legs. All the manuscript illustrations or sculptures I have found of knightly riders in England, France, Italy or Normandy up at least to the 1180s have this feature.72 No one in the Bayeux Tapestry, except Duke William himself, exudes high status better than Odo of Bayeux: at one point he is represented in a manner copied from a depiction of Christ at the Last Supper; at another, fully armed, he encourages young knights into battle. However, Odo’s seal shows him with armour, sword and shield on a small lightly-built animal, high-stepping, with a high narrow head, reminiscent of a Barb horse, the rider’s feet close to the ground.73 Both William I’s and William II’s seals exhibit a similar type of horse – finely built with a small head – again with the royal feet near the ground.74 Similarly, the seal of King Stephen depicts the king in full battle array on a lightly built, quality, but very small animal.75 Other well-known examples, like the prob-

67

C. Gleditz, Horse Breeding in the Medieval World, Dublin 1997, 155. Carmen, lines 491–4, pp. 30–1. 69 Bates, William the Conqueror, 514. 70 Malmesbury, Gesta regum, I, 550–1. 71 G. Loud, The Age of Robert Guiscard; southern Italy and the Norman Conquest, Harlow 2000, 1. 72 E.g. Canterbury Psalter ed. M. R. James, London 1935, mid twelfth century, fols 7b, 22, 44b, 76, 90, 147; W. Greenwell and C. Hunter Blair, ‘Durham Seals’, Archaeologia Aeliana, 3rd ser. 12, 1915, nos 2812, 2804, 2905, 2806; cf. seals of the late thirteenth century and fourteenth centuries, 2814, 2767, 2808, 2809, Plates IV and III. For accessible illustrations, see I. Peirce, ‘The Knight, his Arms and Armour, c. 1150–1250’, ANS 15, 1992, 251–74 at 271 and 273. 73 Reproduced in Davis, Medieval Warhorse, 22, from the drawing for Sir Christopher Hatton’s Book of Seals. 74 Facsimiles of English Royal Writs to A.D. 1100 presented to V. H. Galbraith, ed. T. A. M. Bishop and P. Chaplais, Oxford 1957, plate 30 (from Eton College Library); Barlow, William Rufus, facing 74; John Gillingham, William II, The Red King, London 2015, plate 4. 75 The Cartulary of Worcester Cathedral Priory, ed. R. R. Darlington, PRS, 76, NS 38, 1962–3 (1968) Plate 2. 68



Horses, Knights and Tactics

11

able illustration of Robert of Gloucester and the seal of Henry I, are consistent 76 and could be portrayals of the Arab-type horse. It may be countered that to represent riders as relatively large compared to their horses was simply an artistic convention common in contemporary Europe: pictorially important as the horses were, the knights and magnates they bore were more so. Yet the kings and dukes so represented on their seals found their image acceptable, indicating that the height of the horse was not a factor in their status, as in royal portraits centuries later. But other features depicted evidently do matter: to be mounted on horses with fine legs – denoting speed and agility; horses that are high-stepping – denoting active temperament; that have ‘a good front’ – that is a long, well-crested neck and powerful shoulders; and with high head and tail carriage – denoting Arab blood or quality stock. The high-set neck and head carriage common in illustrations of rulers’ horses were of the essence. They prevented the rider’s forward momentum from becoming destabilized at a check, stumble, change of direction, or whilst lunging forward with a sword. The Bayeux Tapestry shows a Norman knight using his horse’s neck as a balance while he lunges forward with his sword in the thick of combat.77 In contrast, heavy horses, known as ‘cold-blooded’, do not usually have these high-set necks and tails. In William fitzStephen’s London of the 1170s, amongst the fine horses exhibited for sale at Smithfield, the warhorses are ‘elegant’, with ‘necks raised’, and with ‘large haunches’ that enabled them to wheel and turn easily: important for mobility.78 Louis Nolan, with extensive military experience in India and Europe in the nineteenth century, wrote comprehensively in his efforts to improve the selection and training of cavalry.79 After also reading a wide range of Continental sources on many campaigns including Magyar operations, Nolan maintained that heavy cavalry – large men on heavy horses – lacked speed and endurance: they were ‘more formidable in appearance than in reality’. A cavalryman needed only a serious wound in one arm and he became useless, and so agility of both horse and rider were of the essence. The lance was unusable in a melée, having little penetrative power unless carried home by the charge. And every charge, perforce, ended in a melée where the key to success lay in precise control of the horse and a sharp sword.80 Horses in Normandy There was undoubted Norman interest in horse breeding before and after 1066. Normandy’s good pastures and wide contacts might have enabled an improvement in their equine resources, which were probably at a low ebb after depletions by Vikings – who themselves employed equine speed, as well as marine, to devastating effect.81 Very small weight-carrying ponies were familiar to the Scandinavian regions and northern Europe, and during the Viking era interchange took place with

76

The former reproduced from History of Kings of Britain by Geoffrey of Monmouth dedicated to Robert of Gloucester, in Davis, Medieval Warhorse, 56. 77 The Bayeux Tapestry, ed. Stenton, no. 70; The Bayeux Tapestry, ed. Wilson, plate 69. 78 William Fitz Stephen, ‘Description of the City of London (1170–1183)’, in English Historical Documents, ed. Douglas and Greenaway, II, 1026–7. 79 E.g. L. E. Nolan, Cavalry: Its History and Tactics, London 1853. 80 H. Moyse-Bartlett, Louis Edward Nolan and his influence on the British Cavalry, London 1971, 145. 81 Davis, ‘Warhorses of the Normans’, 74–5; Davis, Medieval Warhorse, 55–8; H. B. Clarke, ‘The Vikings’, in Medieval Warfare, ed. Keen, 36–58 at 48–55.

12

Sally Harvey

western Europe. Yet, in the mid-eleventh century, the stock of Norman horses had wide-flung demands to meet: for ventures in Italy in the 1050s and 1060s and for the traffic between the two territories, albeit Lombardy and southern Italy were themselves a source of desirable mounts. Heavier horses were known in Lombardy and other Italian regions, so Norman incursions into Italy may well have enabled the eventual evolution of a somewhat more substantial warhorse in Normandy, but that was unlikely by 1066. It takes several generations to produce another type in significant numbers. Of course, the nature of conflict itself meant a huge turnover in horses, even injuries such as broken legs being fatal. Demand was open-ended when a single charge might result in the death of eighty horses, and archers deliberately targeted the then unarmoured horses in the battle lines.82 Moreover, prolonged campaigns in hostile country also took their toll. Horses, with their high metabolic rate, cannot manage long periods with low food and water intake nearly as well as men can. The idea of a campaigning season, perhaps best after harvest but before autumn set in, evolved as much to preserve equine life as to preserve troops. Also, multitudinous are the impairments to active service caused by poor food or abuse. Between July 1300 and the following February more than half the royal household’s warhorses on the Scottish borders fell sick or died on the march or in winter.83 However much horses cost, over-taxed leg tendons or broken wind through sustained galloping are irreversible conditions, and one frightening experience suffices to make horses unpredictable in similar circumstances – when once more well fed enough to protest. Furthermore, the evidence is that Norman rulers sought the distinctive mounts depicted on their seals not on the vaunted pastures of Normandy, but further afield. One of William’s mounts at Hastings, according to the late chronicler Wace, was a Spanish horse.84 William of Poitiers relates how early in his ducal rule William received prestigious gifts of ‘noble’ horses from Spain, Gascony and the Auvergne. ‘Spanish kings sought his friendship with these gifts among others.’85 Geoffrey of Anjou was – seemingly – not yet knighted when his marriage was arranged with Henry I’s daughter, and when Henry knighted him in 1128 at Rouen, he gave him a Spanish not a Norman horse.86 Why did the Norman and other western rulers so consistently seek Spanish horses? From the Moorish conquest the peninsula had acquired Arab and Barb and Turkmene blood-lines; and the peninsula’s indigenous animals afforded hardy, compact breeding stock.87 In the ninth century, Iberian animals had impressed Carolingian rulers, and in 876 Pope John VIII had sought the finest Arab breeding stock there.88 Even after western European rulers had devoted centuries of resources to equine breeding and to importing Arab stallions, the hardiness, stamina and speed of the true Arabian horse proved perennially outstanding. Not only in heat but in cold too. After the disastrous retreat from Moscow, Napoleon’s aide-de-camp was in no doubt. He wrote:

82

Orderic VI, 238–9, 350–1. Calendar of Documents relating to Scotland 2, Edward I, ed. J. McBain, Edinburgh 1884, no. 1190, 304–5; and e.g. Hewitt, The Horse, 70. 84 Wace, Le Roman de Rou, III, lines 7535–42, ed. A. J. Holden, 3 vols, Paris 1971–3, II, 164. 85 Poitiers, 16–7: Item reges ispaniae his donis inter alia eius amicitiam captabant. 86 Historia Gaufredi Ducis, in Chroniques de Comtes d’Anjou et des seigneurs d’Amboise, ed. L. Halphen and P. Poupardin, Paris 1913, 179; Barker, Tournament, 113. 87 Hyland, Medieval Warhorse, 55–7. 88 Davis, ‘Warhorses of the Normans’, 73. 83



Horses, Knights and Tactics

13

The Arab horse withstood the exertions and privations better than the European horse. After the cruel campaign in Russia almost all the horses the Emperor had left were his Arabs. General Hubert … was only able to bring back to France one horse out of his five, and that was an Arab … So it was with me also.89

Today’s little-publicized 100-mile endurance races, completed in about 7 riding hours (three stops taken for water, one including a compulsory half-hour rest), with horses carrying an obligatory 11 stone, are mainly entered and won by light-weight Arabian horses, usually 14.2 to 15.2 hands high, even in an age when larger competition horses are ubiquitous. Occasionally Welsh cobs do well. The history of the horse on prolonged campaigns argues for the efficacy of the light horse. In 1071, the key battle of Manzikert in the heat of August saw the heavy cataphracts of Byzantium defeated by light Turkoman horse.90 Even in less torrid conditions, lightly armed horsemen often bested the more heavily armed directly, or defeated them by indirect strategies. It was not often that the efforts of mounted knights created the turning point.91 Norman rulers were aware of the resilience and stamina of the Arab/ Barb/Spanish blood-lines, and sought them out; but only the ruling elite could be mounted on animals of such distinctive standards. Horses in England The finest horses were always costly to acquire and difficult to breed, but it is not evident why the quality of English and Norman horses of the ruling groups before 1066 should differ significantly. The royal line of Anglo-Saxon England evidently spent much of their leisure time hunting – in the companionship of those with expertise in hunting and breeding horses. In his will dated 1015, the Ætheling Æthelstan allocated two horses, one white, to his father, King Æethelred; a black stallion to Bishop Ælfsige; and, after gifting other individual horses, all his remaining stud went to his huntsman.92 The elites of England, like those of Normandy and Flanders, could easily have sourced stallions from elsewhere. Horses moved regularly between England and the Continent: the standard tariff at Dover before 1066 for carrying the horses of royal messengers across the Channel was 2d. in summer and 3d. in winter.93 Hereward the Wake, who had a formidable reputation in military contests in Flanders and Poitiers, as well as Ireland, acquired a mare he called Swallow – implying a dark-coloured horse – of ‘very great speed’, and her colt, of ‘outstanding beauty’, called Lightfoot; they were sourced following Hereward’s military venture under Robert ‘the Frisian’, the count of Flanders’s son, from an island on the northwest fringes of the Scheldt estuary, which had ‘a remarkable breed of very swift horses’.94 (The present-day Frisian breed is entirely black – recessive in horses – with an unusually long neck and high head carriage.) At the highest levels of society, there was little reason for the best horses of Englishmen and Normans to differ. 89

Quoted in C. Silver, Guide to the Horses of the World, Lausanne 1976, 137. Hyland, Medieval Warhorse, 53. 91 Gillingham, ‘William the Bastard’, 155–6. 92 Whitelock, AS Wills, 20. 93 GDB1a (KEN/ D3). 94 Gesta Herwardi , in Gaimar: L’estoirie des Engles I, ed. T. Hardy and C.T. Martin, 2 vols, 1888–9, I, 339–404 at 363: genus esse velocissimorum equorum in quadam terrae insula, ubi cum paucis … et exinde equam nimiae velocitatis et pullum spectabilis formae conduxit, que ‘Levipes’ cognominavit, et matrem Anglice ‘Hirundo’ vocavit. See also E. van Houts, ‘Hereward and Flanders’, ASE 28, 1999, 201–23 at 204–9. 90

14

Sally Harvey

Even if the English did not customarily fight on horseback, many rode to the site of the battle carrying arms – and with luck rode away – and their horses needed training suited to the task. The putative site of the battle of Stamford Bridge was reported to have produced numerous small horse shoes and pieces of sword after nineteenth-century ploughing.95 And, of course, horses together with arms formed the heriots of earls, king’s thegns and thegns: eight horses at an earl’s death, four horses from a king’s thegn, and one horse from a thegn, were owed, thus associating horses with rank and with armed conflict.96 One manuscript depicts a highstatus group with spears, on quality, proudly stepping horses, the leader on the most stylish animal.97 The subject of breeding is complex. On the one hand, horses were obviously internationally mobile, travelling long distances as traded goods, as prestigious gifts for rulers, as war mounts and as war booty. On the other, each region perforce utilized the mares of the locality as breeding stock. The two types could have been cross-bred at any time in the eleventh and twelfth century, and also crossed with heavier working horses to produce a larger animal. Here it is argued that the predominant height of horses in battle was no more than a hand above the figure for the century: namely, in the eleventh century generally up to about 12 hands; in the twelfth century 13 hands; and during the thirteenth century rising to 14 hands. In the course of this process, weight-carrying increased, but stamina often declined. But bear in mind: the weight of armour and the importance of the charge were not the sole motive for breeding somewhat taller horses: the mount of 14 to 15 hands tends to have a longer, more comfortable, stride than the undoubtedly more agile smaller mount. Present-day Welsh ponies and cobs have been bred from elements comparable with those available to twelfth-century leaders: Arabian blood introduced deliberately onto native ponies and small sturdy working animals to give quality, and hardiness with speed. Of the four types of Welsh equines, two – the Welsh Mountain pony, Section A (10 to 12 hands in height), and the more substantial Section C, Welsh pony of Cob type (12 hands to 13½ hands) – with their small intelligent heads, large eyes, large crests and powerful hind-quarters, recall those of the Bayeux Tapestry. Indispensable would be the small foundation mares of native blood, or generalpurpose working mares, crossed with a stallion preferably of Arab or Barb blood. In fact we know that at the end of the eleventh century one prominent Norman, the belligerent son of Roger of Montgomery, Robert of Bellême (a district in Normandy still noted for its studs), did explicitly that: he imported ‘fine’ Spanish stallions to set up ‘most excellent studs put apart for breeding’ horses on his new Welsh border lands. The region produced the animal eventually known as the Powys horse, much in demand in the thirteenth century, ‘on which account the horses sent from thence are remarkable for their majestic proportion and astonishing fleetness’.98 The description accords well with the modern Welsh Cob ‘Section D’

95

G. Fleming, Horseshoes and Horseshoeing, London 1869, 287. II Cnut, 71 and 71a, in The Laws of the Kings of England from Edmund to Henry I, ed. and trans. A. J. Robertson, Cambridge 1935, 208–9 and discussed by N. P. Brooks, ‘Arms, Status and Warfare in Late Saxon England’, in Communities and Warfare, 700–1400, London 2000, 138–61. 97 British MS. Cotton Cleopatra cviii, reproduced in R. I. Page, Life in Anglo-Saxon England, London 1970, 105. 98 Giraldus Cambrensis, Itinerarium Kambriae, in Giraldi Cambrensis Opera, ed. J. S. Brewer, James F. Dimock, George F. Warner, 8 vols, RS 1861–91, VI, 14: In hac tertia Gualliae portione, quae Powisia 96



Horses, Knights and Tactics

15

stallion: traditionally 14 hands plus in height, ‘an animal of great stamina, surefooted, strong and brave’, extremely strong, intelligent and fiery in action, though not in temperament. With a powerful presence, its handling requires a strong and experienced horseman.99 Breeding is a long-term and hazardous business: horses suffer high rates of mortality or injury within the first two years of life. Local stock formed the essential constituent in any breeding programme. Moreover, it is advisable to breed from proven stock whose legs, tendons, heart and wind have shown themselves strong enough in action for the required task. A new generation needs a minimum of five years from conception to relative maturity, about 7 or 8 years to full strength. The bigger the horse, the longer that takes. It takes several generations, further, to produce a new standard type, as first crosses often resemble one parent or the other. Whilst crossing two types almost automatically increases size through hybrid vigour, yet some curious shapes can be produced, which give weakness. Heavier types feature in wall-paintings of knights in the church at Claverley, Shropshire (a church connected with Roger of Montgomery), probably datable to the first half of the thirteenth century.100 There the horses of the apparently evil knights are ill-proportioned, with long backs and over-large heads, suggesting that this depiction was deliberate, not defective artistry. The good knights ride better-proportioned horses, perhaps a Powys horse – after all, Powys is merely a day’s ride from Claverley. But no breeding programme could keep up with the casualties. There were said to have been 700 horses killed in William Rufus’s attack on Chaumont in 102 1098.101 The crusaders were perennially short of horses. Records of horses’ ages from the more sedate agrarian sector suggest that medieval horses in general were not long-lived, probably because they were over-exerted too young. Demesne plough-horses averaged 5½ years and demesne cart-horses some 7 years, and more than half their replacements were bought in.103 Even in the thirteenth and fourteenth centuries, when royal studs were managed with forethought – with grazing rotated, animals dispersed away from the coast in times of conflict, and mares selected before the breeding season – they required continual replenishment with costly horses from overseas.104 At the beginning of the thirteenth century,

dicitur, sunt equitia peroptima; et equi emissarii. laudatissimi, de Hispaniensium equorum generositate, quos olim comes Slopesburiae Robertus de Beleme in fines istos adduci curaverat, originaliter propagati. Unde et qui hinc exeunt equi, cum nobili formae pictura, ipsa protrahente natura, tam membrosa sui majestate, quam incomparabili velocitate, valde commendabiles reperiuntur. trans. Lewis Thorpe, Gerald of Wales: The Journey through Wales: The Description of Wales, Harmondsworth 1978, 201. 99 Specifications include: ‘eyes bold and wide-set, ears small and pricked. Neck long and proudly carried, setting into strong shoulders; forelegs set square and forwards, … body strong, deep-girthed, and muscular, hind-quarters lengthy and powerful, tail high-set and proud … The action free, straight and forceful. At trot the knee should be bent and extended as far forward as possible; hocks flexed under the body in straight, powerful leverage.’ Silver, Guide to the Horses of the World, 65. The early sixth-century Lakenheath horse of about 14 hands (sacrificed to be buried with warrior, sword, spear and shield) has been likened to a Welsh cob, Letter of 5.9. 2008 from Joanna Caruth, excavator, Suffolk County Council. 100 G. Home, The Parish Church of All Saints, Claverley, l947, 4–7. 101 Orderic V, 216–18. 102 Ralph of Coggeshall, Chronicon Anglicanum; De expugnatione terrae sanctae libellus, ed. J. Stevenson, RS 66, London 1875, 46: siquidem illo exercitu non amplius quam sex equi cum una mula fuerant; Smail, Crusading Warfare, 188. 103 J. L. Langdon, ‘The Economics of Horses and Oxen in Medieval England’, Agricultural History Review 30, 1982, 31–40, 36; B. M. S. Campbell, English Seigniorial Agriculture 1250–1450, Cambridge 2000, 138–9. For the ages discussed, see Clark, Medieval Horse, 172–4. 104 H. J. Hewitt, The Organisation of War under Edward III, chapter 5; Hewitt, The Horse, 11–26.

16

Sally Harvey

King John imported 100 heavy stallions from Flanders, Holland and the river Elbe region (the last a long-established source of heavier equines), but he also continued to seek Arabian blood, importing Spanish horses, as did Henry III and his French rival, Louis IX, and Edward I, Edward II, Edward III after him.105 Yet with heavier horses – and horses worked heavily – came the problem of equine stamina. The dearth of horses suitable for military action was a perennial matter for English concern.106 Those sceptical of the argument on horse size here might reflect that so low was England’s stock of suitable horses that Henry VIII enacted legislation in 1535 attempting to compel landowners with parks to keep at least two mares of a minimum height of 13 hands that were to be covered only by stallions of a minimum height of 14 hands – legislation that confirms the small size of many horses inherited from the medieval centuries.107 A Range of Qualities All the sporadic evidence before and after the Norman era suggests that only the king and the greatest landholders had the resources to run studs that produced high-quality animals, and that their mounts were distinct from those commonly available. It was ever thus. King Oswine of Northumbria rebuked Bishop Aidan for giving away to a beggar the fine horse, equum optimum, that he had presented to the bishop, telling the bishop that ‘we had many cheaper/commoner horses’ (habuimus equos viliores plurimos) that would have made a more appropriate vehicle for the bishop’s charity.108 The distinction between the equines ridden by royal and aristocratic leaders and the generality was always present. The pro-active Henry II was avid for top-quality horses when negotiating his bargains: in 1173 he sought 40 ‘very valuable horses’ as annual tribute from Raymond V of Toulouse.109 Later in the Middle Ages, the inventories for warhorses valued prior to active service in France list a wide range of animals; they reveal that whereas equines of between £80 and £100 featured for the leaders, some were valued between £5 and £8: Ayton concludes that ‘there has been a tendency to dwell upon horses of the highest quality, without giving a clear impression of what an ordinary man-at-arms would need to pay for a typical warhorse’ – an error of which I, too, was partly guilty in 1970 – but no longer!110 Mounts assembled for campaigns varied, affected by the rank of the owner, the terrain of the campaign, and the animals available. Whilst chroniclers of the crusading wars distinguished between the well-protected knights and the lighter

105 W.

Childs, Anglo-Castilian Trade in the Later Middle Ages, Manchester 1978, 120–2; Hewitt, The Horse, 25–7. 106 D. Eltis, The Military Revolution in Sixteenth-century Europe, London 1995, 114. 107 1535–6 ‘An Acte Concernyng the breade of Horsys’, Statutes of the Realm, 3, ed. A. Luders, London 1810–28, 535–6; followed up by the act of 1540, 758 seeking somewhat taller horses in the southern counties. The 1580 ‘Proclamation for Horsemen and Breed of Horses’ suggests that previous acts had proved ineffective, see J. Thirsk, Horses in Early Modern England: For Service, for Pleasure, for Power, Reading 1978, 15. 108 Bede’s Ecclesiastical History of the English People, ed. B. Colgrave and R. A. B. Mynors, Oxford 1972, 258–9. 109 Robert of Torigni, Chronicle s.a. 1173, in Chronicles of the Reigns of Stephen, Henry II, And Richard I, ed. R. Howlett, 4 vols, IV, 255: Promisit [sc Raymond] ei [Henry II] se daturum equos magni pretii, quotannis quadriginta. 110 J. E. Morris, Welsh Wars of Edward I, Oxford 1901, 49, 53, 82; Ayton, Knights and Warhorses, 46–8, cf. Harvey, 40.



Horses, Knights and Tactics

17

equipped equites levis armaturae, these knights were not separated in action: there was no ‘heavy cavalry arm’.111 The late medieval indenture indicated the status of the men-at arms, as it included every social rank from royal dukes down to the humblest esquire who could afford horses and basic equipment.112 The less well-protected simply took their smaller chance of survival, aided perhaps by their mounts’ greater mobility through being less encumbered. Horses and Tactics Given the extent of the problems posed by the supply and maintenance of horses, what were the principal advantages of being mounted? Most obviously, speed: the ability to attack, manoeuvre, and re-form quickly. Cavalry was at its most effective for shock impact on disordered troops – a traditional medieval deployment.113 Mounted, an armed force has the ability to mobilize a strike at an opportune moment, often with the decisive element of surprise, so crucial in Viking successes.114 (Pack-horses enabled surety of equipment and food supplies.) Following Hastings, William of Normandy secured Dover, Canterbury and Winchester, and encircled London before attempting to enter it. The evident success of William’s campaigns in the vicinity of Hastings and when subduing and imposing his rule across large areas of wild country was due to his safe and rapid movements, whilst keeping himself well informed by reconnaissance.115 William’s unusual winter campaigns of 1069–70 – when he subdued Danish incursions in Yorkshire, then traversed the Pennines, quelled rebellion in Cheshire and Shropshire, and returned to Yorkshire in the spring – demanded adaptable, speedy mounts with stamina and hardiness, rather than heavy cavalry. Foraging while ravaging was commonplace, but there was unlikely to be ample stored fodder available in these uplands, and indeed the be-bogged horse casualties offered welcome sustenance for the troops.116 In all, in subduing England, there is little evidence – archaeological, artistic or documentary – that Duke William made the mistake of leading the anachronistic heavy cavalry that historians have given him. It was indeed the speed and mobility of being mounted that enabled the Norman victors to subdue their new hinterland effectively, aided by their castles. As William of Malmesbury put it later, ‘There were many castles throughout England, each defending its territory, or to tell the truth, plundering it. The knights from the castles carried off flocks and herds, sparing neither churches nor churchyards.’117 But William’s ravaging might have proved a short-term term strategy. Profits from the local peasants were key to the successful exploitation of the Normans’ new country; however, northern England still bore the evidence decades later of the devastation William wreaked in 1069–70.

111

Smail, Crusading Warfare, 106–11. Ayton, Knights and Warhorses, 45–6. Agincourt, 115. 113 Smail, Crusading Warfare, 113–15; P. Contamine, War in the Middle Ages, trans. M. Jones, Oxford 1984, 229–30. 114 With horses as well as ships, Clarke in Medieval Warfare, ed. Keen, 48–55. 115 E.g. Poitiers, 36–8; William of Jumièges in GND II, 118, 126–30; Gillingham, ‘William the Bastard’, 151–6. 116 Orderic II, 234–6. 117 Malmesbury, Historia novella, II, c. 36, pp. 70–1: Castella erant crebra per totam Angliam, quaeque suas partes defendentia, immo ut uerius dicam, depopulantia. Milites castellorum abducebant ab agris et pecora et pecudes, nec aecclesiis nec cimiteriis parcentes. 112 Barker,

18

Sally Harvey

Quite as sinister was the long-term effect of the armed and mounted Norman forces prevailing at seemingly properly conducted and recorded courts, thereby legalising the Norman take-over: even with witnesses or jurors present, and even against the greatest old English abbeys. The Abingdon portrayal of a suit between the abbey and the bishopric of Lincoln over a mill-sluice shows the royal clerk, Peter, later bishop of Lichfield/Chester, pursuing the case for the vacant bishopric, accompanied by a large band of well-armed and mounted men. We only know of the context of the court case held here because, extraordinarily, the abbot invoked the aid of the relics of St Victor, with the result that the horses of the armed men began to sink in the wetlands bordering the river!118 Yet the Domesday and earlier courts, which had sanctioned and recorded the transfer of land from Anglo-Saxons to Normans, and given the transfers the veneer of legal process, were supervised by the county sheriffs, who would invariably have had such armed and mounted men in attendance. To continue with the other, less trumpeted, advantages of being mounted: a horse might not necessarily enable a man to fight more effectively, but it did assist him in self-protection by allowing him to withdraw unwounded, or to avoid death or capture. Or, to put it another way, to flee. Foremost in the knight’s essential equipment, according to Archbishop Anselm of Canterbury, was his horse: ‘which is so necessary for him that it could rightly be described as his faithful companion. For with his horse he both charges and puts to flight the enemy, or, if need arise, escapes from his own pursuers’ (my italics).119 It might well have been this option that explains the much-quoted incident in the Confessor’s reign which is supposed to show conclusively that the English could not fight on horseback. In a campaign in which the newly settled Normans of the Welsh border were keeping the Welsh at bay, the Anglo-Saxon Chronicle states that ‘the English army fled, because they were on horseback’.120 Yet the Worcester Chronicler, whilst commenting that it was not the English custom to fight mounted, recounts how the French and Norman forces under Earl Ralph the Timid had taken flight first.121 It seems quite likely that, mounted, the English found it expedient to do likewise. We all know many later examples. At the crucial battle of Lincoln of 1141, while King Stephen had dismounted to fight hard and earnestly, many of his powerful magnates took advantage of being mounted to desert early, despite their expertise.122 At the Battle of the Standard, and fighting on foot, the Scottish king’s only means of escape was on horseback. Moreover, his one line of mounted knights retreated because, once their horses were wounded, they could not fight against dismounted knights in armour.123 Chroniclers of pre-battle speeches, whilst adopting classical models, did not shrink from portraying leaders warning their forces against the option of flight.124 Even after the developed chivalric ethos of the high Middle Ages, it was not unknown for leaders of the French cavalry corps to escape crucial encounters. Battles from 118 Historia

Ecclesiae Abbendonensis, ed. J. Hudson, 2 vols, Oxford 2007, I, 372–5. of St Anselm, ed. R. W. Southern and F. S. Schmitt, Auctores Medii Aevi I, London 1969, 97: Imprimis siquidem caballus, qui ei adeo est necessarius, ut recte dici possit suus fidelissimus socius. Cum illo namque et occurrentem aggreditur hostem et persequitur fugientem, et cum tempus postulat, fugit persequentem. 120 ASC C s.a.1055, ed. D. Whitelock, London 1961, 150. 121 John of Worcester, s. a. 1055: Anglos contra morem in equis pugnare iussit, sed cum prelium essent commissuri comes cum suis Francis et Nortmannis fugam primitus capessit, II, 576–7. 122 Orderic VI, 236–7: fugerunt igitur omnes equites regis et willelmus Yprensis. 123 Huntingdon, 716. 124 E.g. Huntingdon, 728, 732, 737. 119 Memorials



Horses, Knights and Tactics

19

Philip Augustus in the 1190s to Agincourt featured the phenomenon. In the latter action, a French herald and chronicler recorded the handful of French cavalry that were killed, whereas ‘all the rest failed to do their duty, for they fled shamefully and never struck a blow against the English’.125 In the sixteenth century, with the constant risk of death from a pistol fired at close range, ‘horsemen lightly galloped off rather than stay the full course on the field’.126 For troops without horses beneath them or arms denoting their status, more general slaughter was often the outcome as they were unable to make their escape. Relevant here is, of course, the battle of Hastings, where the first flight – promoted in part by rumour of William’s death – helped to break up the English formation. It was then deployed a second time as a deliberate tactic.127 In pursuing the apparently fleeing Normans, the native forces left their advantageous stand on the hill-top, with their impregnable shield-wall formation – impregnable certainly to horses, who will refuse to mow down humans deliberately. The English pursuit is wholly comprehensible in the light of those Normans established on the Welsh borders under Edward the Confessor employing their mounts to flee from the Welsh. William of Poitiers, an ex-knight himself, does not hesitate, in his encomium, to depict the Norman duke reminding his forces before Hastings that flight was not an option – on one side were armed forces and a hostile country, and the sea was on the other.128 William of Poitiers regarded the Normans’ first flight as genuine, but serendipitous: certainly not a culpable offence. ‘The Normans believed that their duke and lord had fallen, so it was not too shameful to give way to flight; least of all was it to be deplored, since it helped them greatly.’129 True, feigned flights already feature in studies on warfare as ‘a widespread but effective tactic’,130 yet the strategy could not have worked repeatedly unless there had been many more flights that were genuine. At what point a sound strategy of postponing confrontation became flight rather than wisdom was probably only a matter of obedience to commands or common assent.131 Eustace of Boulogne, no longer young, and an ambiguous supporter of William, attempted to leave the field of Hastings with his fifty knights.132 Frequent flights, real rather than feigned, do not perhaps accord with later romantic conceptions of knighthood, but then, neither do policies of wasting the crops of civilians, taking their livestock, and appropriating the revered treasures of the church.133 It is this ability to flee with speed to the protection of a stronghold that links the mounted knight to the domination of the surrounding hinterland.134 Cavalry required discipline in action to be successful; but its youthful character often meant some knights were over-impetuous.135 The Templars found it necessary 125 Barker,

Agincourt, 291, 295–7; see also the behaviour of Warwick, J. Gillingham, The Wars of the Roses: Peace and Conflict in Fifteenth Century England, London 2005, 200. 126 Eltis, Military Revolution, 64. 127 Poitiers, 128, 132. 128 Poitiers, 126. 129 Poitiers, 126–8: Credidere Normanni ducem ac dominum suum cecidisse. Non ergo nimis pudenda fuga cessere; minime ero dolenda, cum plurimum iuuverit. 130 Strickland, War and Chivalry, 130. Smail treats tactical uses of feigned flight, Crusading Warfare, 78–9. 131 E.g Poitiers, 74–5 (Conan of Brittany fled from William in order to raise further military support). 132 Poitiers, 138–9. 133 E.g. Poitiers, 26–7, 36–9, 48, 52–5, 138–9. 134 E.g. Poitiers, 50–1. 135 Smail, Crusading Warfare, 128–30.

20

Sally Harvey

to lay down the law on this point; their discipline for engagement of 1164 states that whereas knights were allowed the necessary warm-up short gallop for their horses before a battle, no breach of formation was to be tolerated.136 Well-horsed, well-equipped young knights were likely to be eager to engage battle too soon, or to make their own decision as to when to leave it. A drawback was that their horses enabled them to do just that. In battle, cavalry might well need to fall back and regroup; the less-trumpeted corollary was that a horseman might well evade battle altogether. Horses offered their knights other physical protection. Whilst with skill they might be manoeuvred to maximize the warrior’s effectiveness, at close quarters, more crucially, they simply took the brunt of the axe- and sword-wielding instead of the rider – horses’ legs, wind-pipe and belly offering the more accessible and vulnerable targets to sword and spear (and more visible to the archer). That was presumably the case for the two, or three (according to chronicler), horses killed under William at Hastings.137 On other occasions, King William and William Rufus had horses killed beneath them.138 At the battle of Brémule in 1119, eighty knights were captured after their horses ‘were quickly killed’ under them at the first charge.139 Being well mounted and armed gave the knights further limited liability: that of marking out those whose lord or family might pay a profitable ransom for their return. At Brémule, ‘they were all clad in mail and spared each other on both sides, out of fear of God and fellowship in arms; they were more concerned to capture than to kill the fugitives’. Of 900 knights who fought at Brémule, only three were killed; most of the losers fled or were taken prisoner.140 Aside from being a possibly lucrative ploy, taking hostages had a genuine strategic design, in providing political bargaining chips. Mounted men were by no means the answer to every military situation. William employed archers and crossbowmen for his enterprise of 1066, and their efforts proved a turning-point.141 It was known that infantrymen could out-face horses, or retreat despite them, if they kept in close order; they became vulnerable only once gaps appeared in the formation.142 Archers might cause chaos at a distance amongst cavalry; they helped to win the day for Henry I at Bourgthéroulde in 1124, where they brought down charging horses.143 But archers were not donors to great ecclesiastical institutions, nor did they hold high office, and whilst they are noted as present or even pivotal, chroniclers do not highlight their part. The horse, however, might offer swift superiority. Smail has shown how the cavalry charge, in appropriate terrain, was often decisive in concluding an inevitable confrontation; yet he also emphasized how crucial was the cover archers provided, and how essential and varied was the protection afforded by foot soldiers during the horsemen’s preparation and assembly before they were committed to a charge.144

136 La

Règle du Temple, ed. H. de Curzon, Paris 1886, nos. 162, 163. horses Carmen, line 509, pp. 30–1; three, Poitiers, 134–5. 138 John of Worcester, s.a 1079, III, 32–3; Malmesbury, Gesta regum I, 551. 139 Orderic VI, 238–41: equis eorum protinus occisis. 140 Orderic VI, 240–1: Ferro enim undique vestiti errant, pro timore Dei notitiaque conturbernii vicissim sibi parcebant, nec tantum occidere fugientes quam comprehendere satagebant. 141 Poitiers, 126–7, 132–3; H. E. J. Cowdrey, ‘Towards an Interpretation of the Bayeux Tapestry’, ANS 10, 1987, 49–65, 62. 142 Gillingham, ‘An Age of Expansion’, 78; Eltis, Military Revolution, 45–7. 143 Orderic VI, 348–50. 144 Smail, Crusading Warfare, 118–20, 130, 161, 174–5. 137 Two



Horses, Knights and Tactics

21

Although most vital to the knights themselves was the ability to fight another day, seasoned commanders knew that a crucial conflict might be best faced without their mounts.145 In 1124, when in rebellion against the English king near Rouen, before the battle of Bourgthéroulde, Amaury IV of Montfort cautioned his younger comrades who were exulting on seeing their opponents dismount, thinking that they now offered easy prey: it meant rather that the opposing forces intended to fight until they had won the day, and that ‘a warlike horseman, who has dismounted 146 with his men, will not flee from the field; rather he will die or conquer’. Amaury’s prediction proved accurate, and more than eighty over-confident magnates and knights were captured, and, confined to the king’s prison, paid the penalty for their folly for a long time. Life-threatening incidents remained, of course, a constant accompaniment to medieval life in general and to warfare and equine activities in particular: bravery was demanded of all concerned. But en passant it does not seem that knights in action invariably lived up to some modern expectations. Even ‘the ideal’ in manuals of chivalry, and in fictional narratives with their many noble exemplars, prompts the conclusion that it was ‘surprisingly flexible’.147 When Henry II captured Stephen’s castle of Crowmarsh, Stephen’s knights were spared but sixty archers were beheaded.148 At least the well-horsed knight from the eleventh to the fifteenth centuries, if unable to flee, stood a good chance of being ransomed, and, as Orderic puts it, ‘an honourable fortune’ might be made out of ransoms.149 Infantrymen and archers could have less hope of being spared if caught. To conclude: the primary purpose of the technical superiority of being mounted, like that of the latest armour, was largely self-preservation in the first instance, and, ultimately, effective control of less mobile populations. In the enterprise of conquest and the endeavour to hold what had been won, a wide range of horses – almost invariably small and light-weight – participated in military action, and furnished the speed, mobility and flexibility demanded of them. Appendix: Horse-Transport and the Bayeux Tapestry The Bayeux Tapestry’s depiction of horses jumping out of boats after the Channel crossing has been much debated.150 For embarkation, this mode was indeed unfeasible. The impact of a shod horse might well damage or splinter the timber, and cause broken legs or strains and sprains for the horses. To embark or disembark from a quayside required a ramp. Embarkation from or onto a beach could be solved simply by portable wooden ramps, as were used on the Byzantine expedition to Crete in 960.151 Even so, a fit horse after a stressful sea voyage might well leap forward into shallow waters on a sandy beach. Transit of horse across the sea was a proven tactic in Western warfare; the Danish great army crossed the English Channel from Boulogne ‘in one journey, horses and 145 Huntingdon,

716–19. VI, 348–51: Bellicosus eques iam cum suis pedes factus non fugiet sed morietur aut vincet. 147 Ashe, ‘Chivalry and Kingship’, 20; J. Gillingham, ‘War and Chivalry in the History of William the Marshal’, Thirteenth Century England 2, 1988, 1–13, at 1–2. 148 Robert of Torigni, Chronicle, s. a. 1153, pp. 173–4; Gillingham, ‘An Age of Expansion’, 83. 149 Orderic IV, 48–9: honorifice; also Barker, Agincourt, 284, 302–3. 150 The Bayeux Tapestry, ed. Stenton, no. 45; The Bayeux Tapestry, ed. Wilson, plate 43. 151 J. H. Pryor, ‘Transportation by Sea during the Era of the Crusades: Eighth Century to 1285 AD’, Mariners Mirror 68, 1982, Part 1, 9–26, 10. 146 Orderic

22

Sally Harvey

all’ in 892.152 Robert Guiscard’s Norman expedition to Sicily of 1061 had taken horses across the Straits of Messina to attack Muslim Sicily. From the number of vessels employed and the size of his force, the vessels may have been able to carry about twenty-one horses each.153 In 1091, to launch his attack on Malta, Guiscard’s brother, Roger of Sicily, seems to have had fourteen knights and their horses aboard his flagship.154 But a short crossing was crucial to retain the element of surprise and for the horses to function well on arrival. Curtailing the length of the sea voyage may have been a factor in Duke William’s decision to move his invasion fleet up the coast from the mouth of the Dives to Saint-Valéry-sur-Somme. Hobbles and muzzles could have been adopted for travelling stallions in close proximity, as they were later, but their effects would soon debilitate horses. It took horses a week to recover from the journey between southern England and Bordeaux, and when Edward III purchased horses from Spain, France, Germany and Lombardy, they were imported on the short journey via Wissant, just north of Boulogne, whenever possible.155 Both Ponthieu and Boulogne offered viable points for equine transit to England; their strategic importance may well have figured early in William’s plans to acquire the English crown. Time and distance remained critical factors in a well-conducted transportation of horses.

152 ASC,

s.a. 892. of Monte Cassino, quoted in Pryor, ‘Transportation by Sea’, 12–13. 154 Malaterra, quoted in Pryor, ibid., 13. 155 Hewitt, The Horse, 55, 25–6. 153 Amato

BALDWIN OF FORDE, BARTHOLOMEW OF EXETER AND THE AUTHORSHIP OF THE LIBER DE SECTIS HERETICORUM ET ORTHODOXE FIDEI DOGMATA Sabina Flanagan This paper is a contribution to an ongoing debate about the nature and extent of the writings of Baldwin of Forde, the often-mentioned, but rather less often-studied, abbot of Forde (c. 1175–80), bishop of Worcester (1180–4) and second archbishop of Canterbury (1185–90) after Thomas Becket.1 The discussion has been conducted almost exclusively in the introductions to scholarly editions of Baldwin’s works or occasional journal articles, which serve to amplify them. My somewhat belated intervention in the discussion was prompted by the appearance in 2008 of an edition by José Louis Narvaja of what is purported to be Baldwin of Forde’s lost work, the Liber de sectis hereticorum et orthodoxe fidei dogmata,2 but which I will argue is more likely to have been written by his friend Bartholomew of Exeter (d. 1184). In short, my paper represents a response to Narvaja’s comments in 2008 on Bell’s 1984 dismissal of Dom Jean Leclercq’s claim in 1963 that the works of Baldwin are ‘assez vaste’ (quite extensive) and many await publication. From this sequence of events it might be thought that scholarship on Baldwin cannot itself be described as ‘assez vaste’ and tends to be prosecuted at long intervals and in far flung places: from Luxemburg to Newfoundland to Argentina and Australia. Two of the three scholars in question, Dom Jean Leclercq and Professor David Bell, of Memorial University, Newfoundland, hardly need an introduction. Not so familiar, however, is the third, Professor José Louis Narvaja. He was the first to identify the text in a Latin manuscript from the Bibliothèque Nationale de France, which had already been discussed as an anonymous work in studies of the conciliar movement, as that of a lost work by Baldwin.3 He then published a critical edition of the text in the series ‘Rarissima Mediaeualia’, under the auspices of the Theological Faculty of the St George Institute at Frankfurt. Narvaja is an Argentinian Jesuit and, as it happens, nephew of the present pope. Let us begin with the statement of Leclercq, found in his introduction to the 1963 edition of Baldwin’s De sacramento altaris.4 When discussing the author and his works Leclercq made the following claim: ‘Baldwin’s work is quite extensive. Only a part has been published, that is, the treatises or sermons, the books on the

1

The place from which Baldwin’s toponymic is derived was previously written as ‘Ford’ and located in Devon until geographic renaming rendered it Forde and removed it to Dorset. I retain the original spelling when quoting works written before the change. 2 Balduinus Cantuariensis archiepiscopus, Liber de sectis hereticorum et orthodoxe fidei dogmata, ed. José Luis Narvaja, Rarissima mediaeualia opera latina 2, Aschendorff 2008. 3 BnF lat 12264, fols 158ra–262va. 4 Baudouin de Ford, Le Sacrement de l’Autel, ed. J. Morson, trans. E. de Solms, 2 vols, Sources Chrétiennes 93, 94, Paris 1963, 9.

24

Sabina Flanagan

Sacrament of the Altar and the Commendation of Faith. [His] Other works deserve to be edited and studied.’5 When, in 1991, David Bell published what he took to be the remaining works of Baldwin not already furnished with modern editions, namely The Sermons or Tractates and The Commendation of Faith, he referred the reader to an article of 1984 in which he discusses the corpus of Baldwin’s works.6 His opening paragraph noted that Leclercq’s statement translated above ‘is an unfortunate one, both for what it says and for what it implies’, in other words, that there is a lot more available in manuscript waiting to be edited and studied.7 After reducing by about a third the nineteen works attributed to Baldwin by Pits in his Relationum historicarum de rebus Anglicis (Paris 1619) – which list is reproduced by subsequent commentators Henriquez (1626), de Visch (1649), Tissier (1662) and Canivez – and finding the et multa alia (‘and many others’) which concludes their lists to be unsubstantiated, Bell toned down his criticism of Leclercq somewhat. He concluded the article as follows: If the ‘other writings’ to which Dom Leclercq refers are restricted solely to two, the [sermons] De Obedientia and the De Cruce – then we cannot take issue with his statement; but if he is implying more than this, then although it may well be true that these other works deserve to be studied and edited, we cannot be absolutely certain of this until they have first been discovered.8

Narvaja, however, seems to have missed the conciliatory tone of Bell’s conclusion (and possibly the point) when he returned to Bell’s claim in his 2008 paper in the journal Stromata. After identifying the text in BnF lat 12264 as Baldwin’s missing De sectis hereticorum et orthodoxe fidei dogmata (henceforth Liber de sectis) he writes: In spite of the list of works of Baldwin having been reduced by one due to the combining of two works which were considered distinct, the material available to be edited and studied remains just as abundant as it was in 1963 and 1991. The 104 folios of lat. 12264 – a bit less than 300 pages in our edition, offer investigators an abundant source for the study of conciliar ideas … the reception of the Fathers of the Church in the Middle Ages and the translation and reception of Greek sources in the Latin West.9

As far as I know Bell has not rejoined the debate, but in this paper I want to raise a possibility that would cast doubt on Narvaja’s confident assertion. The possibility is that the Liber de sectis is not, in fact, a work written by Baldwin of Forde (or Baldwin of Canterbury, as Narvaja prefers to call him). Indeed, it seems to me that 5

‘L’oeuvre littéraire de Baudouin est assez vaste. Une partie seulement a été publiée: ce sont les Traités ou Sermons, les livres sur Le sacrament de l’autel et sur L’éloge de la foi. D’autres écrits mériteront d’être édités et étudiés.’ 6 Balduini de Forda, Opera, ed. David N. Bell, Corpus Christianorum Continuatio Mediaeualis [hereafter CCCM] 99, Turnhout 1991; D. N. Bell, ‘The Corpus of the Works of Baldwin of Ford’, Cîteaux 35, 1984, 215–34. 7 Bell, ‘The Corpus’, 215. 8 Bell, ‘The Corpus’, 226. 9 ‘A pesar de que la lista de las obras de Balduino se reduzca una vez más debido a la unidade de dos obras que se consideraban distintas, el material disponible a la edición y al estudio no dejaba de ser abundante, tanto en 1963 como en 1991. Los 104 folios de lat. 12264 – un poco màs de 300 páginas en nuestra edición critica – ofrecen a los investigadores una fuente abundante para el estudio … de la idea de concilio, la recepción de los Padres de la Iglesia en la Edad Media y la recepción y traducción de fuentes griegas en el Occidente latino.’ José Luis Narvaja, ‘El Liber de sectis hereticorum et orthodoxe fidei dogmata: El hallazago de una obra perdida de Balduino de Canterbury’, Stromata 64, 2008, 263–88 at 285–6.



The Authorship of the Liber de sectis hereticorum

25

Bartholomew of Exeter, Baldwin’s diocesan colleague from the time of his appointment as archdeacon of Totnes in 1162, fellow judge delegate since the 1170s and brother bishop since 1180, is a far more likely candidate. Bartholomew and Baldwin were not only colleagues but also close friends. It has often been noted that they dedicated their writings to each other. They also apparently worked together on compiling decretal collections and possibly the Penitential which is usually attributed to Bartholomew but in at least one manuscript is ascribed to them both.10 So I come to my first question: Why identify the unique text found in BnF lat. 12264 with Baldwin’s lost work? At first there seems to be no very obvious connection. The manuscript in question dates not from the twelfth or thirteenth century but from the late fifteenth century. Indeed, it can be more precisely dated to the period between 1477 and 1491, since it bears an autograph colophon by Thomas Basin (1412–91), then living at Utrecht, donating the manuscript to the cathedral of Lisieux. Basin had been bishop of Lisieux from 1447 until he was forced to retire in 1474 for political reasons, and subsequently held the honorary title of Bishop of Cesarea. Born in Normandy, Thomas Basin is chiefly remembered as an early humanist. He gives an account of his eventful life in his Breviloquium peregrinationis.11 In it he enumerates forty-two of the places in which he was forced to take up residence during his long career. From an even greater number of possibilities he settled on that particular figure to reference the forty-two stations of the desert wanderings of the Jews described in Numbers 33, and thus suggest the sufferings of his own life’s journey. In the Paris manuscript, the Liber de sectis appears in the company of several works by Taius, a seventh-century Spanish bishop, and Leonardus Brunus Aretinus (c. 1370–1444), which Basin had copyists prepare for his own use. A colophon directs that it is to be given on his death to Lisieux cathedral.12 The manuscript is heavily annotated by Basin, particularly the text of the Liber de sectis.13 However it contains no indication of authorship for the Liber de sectis, nor any prefatory letter of dedication. The text was first studied in 1948 by the Byzantinist Francis Dvornik, who mistakenly thought it was a fifteenth-century work.14 Subsequently, Hermann Joseph Sieben, the German doyen of conciliar studies, printed extracts from the text and suggested on internal evidence that it was probably of late twelfthcentury German provenance.15 However he has since wholeheartedly endorsed the ascription to Baldwin.16

10

Lambeth Palace MS 235. For the Penitential see Dom Adrian Morey, Bartholomew of Exeter, Bishop and Canonist: A Study in the Twelfth Tentury, Cambridge 1937, 175–300; Jason Taliadoros, ‘Bartholomew of Exeter’s Penitential: Some Observations on his Personal Dicta’, in Proceedings of the Thirteenth International Congress of Medieval Canon Law: Esztergom, 2008, ed. Péter Erdö, Vatican City 2010, 457–73. 11 BnF Lat. 5970, fols 59ra-62vb. 12 BnF Lat. 12264, fol. 262va. 13 Charles Samaran and André Vernet, ‘Les livres de Thomas Basin’ in Hommages à André Boutemy, Latomus 145, Brussels 1976, 324–39. 14 Francis Dvornik, The Photian Schism: History and Legend, Cambridge 1948, 359–60. 15 Hermann-Josef Sieben, ‘Der “liber de sectis haereticorum” und sein Beitrag zur Konzilsidee des 12. Jahrhunderts’, Annuarium Historiae Conciliorum 15, 983, 262–306 and ‘Der “tractatus de septem conciliis generalibus” (ms. Paris, Bibl. Nat. 12264). Eine griechische Konzilssynopse in lateinischer Übersetzung des 12. Jahrhunderts’, AHC 16, 1984, 47–58. 16 Hermann-Josef Sieben, Konzils-und Papstidee. Untersuchungen zu ihrer Geschichte, Paderborn 2017, 157.

26

Sabina Flanagan

Bibliography, Medieval Catalogues and Early Sightings Baldwin died in 1190 at the siege of Acre. Within his lifetime there seems to be no direct mention of the Liber de sectis, let alone any suggestion that Baldwin was its author.17 Nor is there any evidence that he was particularly interested in heresy. That he was interested in faith, as an intellectual, or even psychological concept, can be seen from his treatise, De commendatione fidei.18 But it should be noted that he was primarily interested in examining the concept of faith in relation to belief and knowledge, rather than its content. Such a matter as the nature of transubstantiation he was happy to leave as an unexplained ‘mystery’.19 The earliest surviving mention of the work is the one that links it to Baldwin in Prior Eastry’s catalogue of books at Christ Church Canterbury c. 1331.20 The entry reads simply: ‘Baldewinus de sectis hereticorum’ without incipit or further identifying information, such as the number of books it contained.21 The current printed source for this list and other Canterbury catalogues is the venerable 1903 edition of M. R. James, The Ancient Libraries of Canterbury and Dover.22 Obviously scholarship has moved on since then and we await the new definitive treatment of these catalogues in the Corpus of British Medieval Library Catalogues. In the meantime, thanks to the generosity of their editor, Dr James Willoughby, I have been able to avail myself of the latest research on the Canterbury catalogues (and also the catalogues of Exeter cathedral, which will be of importance later in the paper).23 Almost two centuries after Eastry’s catalogue was produced, we find in 1508 a second mention of the work in a list of around 300 books repaired at Christ Church, compiled by William Ingram, described as ‘Custos of the Shrine of St Thomas’.24 This list is important for identifying the De sectis with the text in the Paris manuscript as it gives the first few words of the second folio, viz: domus dei edificaretur. The almost identical phrase Domus dei edificabatur can be found twelve lines into the second folio of the text in Basin’s manuscript.25 Since the writing of this manuscript is quite small and heavily abbreviated it appears that a lengthy letter of dedication could not have been present in the Canterbury manuscript in Ingram’s time. Had there been a letter of dedication then the beginning of the second folio would have contained words from earlier in the text, rather than later. With regard to identifying medieval texts Richard Sharpe warns that, ‘In every case what the ancient, medieval or modern bibliographer had seen matters more than what such writers say … One has always to try and reach behind what they say to what they saw and then interpret that.’26 Presumably whoever compiled the catalogue for Prior

17

The question of whether Peter of Blois was alluding to the Liber de sectis and attributing it to Baldwin in Letter 77 of The Later Letters of Peter of Blois, ed. Elizabeth Revell, Auctores Britannici Medii Aevi 13, Oxford 1993, 323–9, is now dealt with more fully in the Appendix to this paper. 18 De commendatio fidei in Balduini de Forda, Opera, ed. David N. Bell, CCCM 99, Turnhout 1991, 343–458. See also Sabina Flanagan, Doubt in an Age of Faith, Disputatio 17, Turnhout 2008, chapter 4. 19 Morson, Le sacrament, I, 210. 20 BL MS Cotton Galba E IV. 21 Galba E IV, fol. 130r. 22 Montague Rhodes James, The Ancient Libraries of Canterbury and Dover, Cambridge 1903. 23 I should add, however, that Dr Willoughby does not necessarily agree with the uses to which I have put such information or the inferences drawn from it. 24 James, Ancient Libraries, 152–64. 25 See the facsimile of fol. 158r at 28 of Narvaja’s edition. 26 Richard Sharpe, Titulus: Identifying Medieval Texts. An Evidence-Based Approach. Brepols Essays in European Culture 3, Turnhout 2003, 121.



The Authorship of the Liber de sectis hereticorum

27

Eastry and later Ingram had seen the actual manuscript containing the De sectis.27 However, the most important person in this chain of sightings and citations is John Leland, the Tudor antiquary whose De uiris illustribus and other bio/bibliographical works have a complex production and publication history, now in the process of being untangled by James Carley.28 We know that Leland saw the manuscript in Canterbury on one of the tours he made of monastic libraries in early 1534. Leland is the first person to provide an incipit for the Liber de sectis hereticorum. He also mentions what appears to be another work, the ‘De orthodoxis fidei dogmatibus’, and provides its incipit. These two incipits correspond to different sections of the work in the Paris manuscript. But what was Leland looking for at Canterbury? He was not intending to make an inventory of the entire library but seeking books of particular interest to himself for his projected literary undertakings and possibly to the king under whose auspices he was working.29 It is notable that the few entries from the list he seems to have made soon after his visit to Christ Church Canterbury are strongly biased to English manuscripts and include several works supposedly by Baldwin.30 When Leland’s Collectanea was finally published in 1770 it also contained an expanded list of Canterbury books.31 According to Carley the original list has been supplemented from Henry Eastry’s catalogue, which was owned by John Twyne after the Dissolution and to which Leland had access.32 Indeed, both James and Carley have identified certain annotations in the Cotton manuscript of Eastry’s catalogue as being by Leland. However, on examining the manuscript I found that the annotations are of at least three different kinds, including a series of small crosses quite distinct from the rather messy trefoils identified as Leland’s. Leland was at Canterbury before its suppression, when Eastry’s catalogue was still in place. So if he intended to examine the library it is reasonable to assume that he had the librarian or custodian of the books look through the catalogue for works that might be of interest to him. Alternatively, he may have marked up the books he wanted to see in the catalogue – perhaps more neatly than was his wont since he was under the eye of the prior or designated keeper of the books. This supposition is supported by the congruence between the marked items and the first brief list he produced, which includes several works by Baldwin.33 We might wonder why Leland was particularly interested in Baldwin, apart from his being an English author. It is tempting to think that this interest arose from his enthusiasm for Joseph of Exeter, whose epic poem on the Trojan War, said to be in the style of Lucan, first came to Leland’s notice even before he went to Paris in the 1520s. Joseph of Exeter was probably Baldwin’s nephew, and Leland would presumably have been aware of their relationship. His entry on Joseph in De uiris illustribus is extensive and laudatory.34 So while there was no work of Joseph of 27

Strictly speaking, the compiler of Eastry’s catalogue might not himself have seen the manuscript if he were merely copying an older catalogue (as James suggested), though we are on firmer ground with Ingram. 28 John Leland, De uiris illustribus/On Famous Men, ed. by James P. Carley, Oxford 2010. 29 Leland, De uiris, li–liv. 30 See Collectanea 4. 10–11, cf. Oxford Bodl. Ms Top. Gen. c. 3, 7–8. 31 Collectanea 4. 120–1, Oxford Bodl. MS Top. Gen. c. 3, 189–91. 32 See Leland, De uiris, lxxvi. 33 Examples of the small crosses can be seen on Galba E iv, fol. 130, which lists several works of Baldwin’s. 34 See Leland, De uiris, 402–9, See also James Carley, ‘John Leland in Paris: The Evidence of his Poetry’, Studies in Philology 83, 1986, 21.

28

Sabina Flanagan

Exeter at Christ Church, Leland may well have been interested in finding works by his patron. The point of this excursus is that Leland may well have been depending on Eastry’s catalogue for his identification of Liber de sectis as a work by Baldwin in the absence of identifying descriptions on the manuscript itself. And the same goes for the Liber regum, another work he attributes to Baldwin, found on the same page of the catalogue.35 All mentions in subsequent catalogues and bibliographies until the publication of Narvaja’s edition depend on Leland, who depends partly on Eastry and partly on his own observations. That Leland actually had the manuscript of Liber de sectis in his hands is indicated by the fact that he not only notes the incipit of the Liber de sectis for the first time but also records the title and incipit of the second part of the work, the ‘Orthodoxe fidei dogmata’, which appears at fol. 174r of the Paris manuscript. It should be noted that the ‘Orthodoxe fidei dogmata’ is not listed as a separate work by Eastry. To recapitulate my observations so far: the attribution of the work to Baldwin dates only from the fourteenth century, and Leland, though he saw the manuscript, does not necessarily provide independent witness of its authorship, since he could well have been relying on the attribution in Eastry’s catalogue. Moreover, the catalogue is not itself without error, containing several misattributions. In fact James suggests that it is a not very good copy of an earlier catalogue.36 Another potential source of error is the confusion between books written by, and books owned by, named persons. There does not seem to have been a consistent distinction between the use of the nominative (indicating authorship) and genitive (indicating previous ownership of donated books) in catalogues or library lists. Indeed, when the genitive of a name is inscribed in a book it might indicate authorship, but sometimes it is a sign of ownership. Thus Eastry’s entry in the third column of the same folio as that which lists the Liber de sectis has ‘Liber Regum Baldewini l. iiii’. This was previously taken to be another lost work of Baldwin’s but is now generally accepted to indicate a work donated by Baldwin to the library. Given such lack of clarity in the conventions of ownership and authorship it would be very easy for the copyist to have written ‘Baldewinus’ instead of ‘Baldewini’ in the case of the Liber de sectis. It may be wondered, then, whether Eastry’s attribution of the Liber de sectis to Baldwin is ultimately authoritative now that the text has been discovered. Without wanting to look a gift horse in the mouth, it seems to me that this work differs so markedly from the rest of Baldwin’s writings that some sort of explanation is required. Any such explanation is conspicuously lacking from Narvaja. However, before venturing further some account of the work itself is necessary. The Liber de sectis hereticorum et orthodoxe fidei dogmata As already indicated, the text consists of two related parts, sometimes catalogued as two distinct works: the Liber de sectis, which deals with the heresies of the early church, starting with Simon Magus; and a second part, the ‘Orthodoxe fidei dogmata’, which introduces the twenty-two early councils of the church, from Nicaea to Seville, and their doctrinal pronouncements. Separating the two parts there are several lists of councils and popes. Thus in Narvaja’s pagination, 29–74 treat the heresies; 74–84 comprise lists of general councils and a list of the Roman

35 36

fol. 130 col 3 Liber Regum Baldewini l. iiii. James, Ancient Libraries, xliii.



The Authorship of the Liber de sectis hereticorum

29

pontiffs; the next portion of the work, 84–262, enumerates and comments on the first twenty-two councils; the work concludes, 262–334, with the confessions, that is the creeds, of the first fourteen Roman popes. The author’s method is to quote a larger or smaller chunk from his sources (which are conveniently identified in the original chapter headings, e.g. ‘From the third book of Eusebius’), and to follow it with his explanacio. Sieben claims that this is the sole example of such a conciliar text with commentary to be found in the Middle Ages.37 On further examination it turns out that the greater part, around 60%, of the edition’s nearly 300 pages consists of direct, though sometimes abbreviated, quotation from an interesting range of authorities, rather than the author’s own commentary. These somewhat disparate parts are not entirely discrete, since there are internal references from one part to the other. This is a further confirmation that we are dealing with a single work rather than two separate ones.38 Narvaja’s introduction to the edition devotes three and a half pages to the manuscript, two pages to the author, less than a page to sources, a page to Thomas Basin’s annotations, less than a page to previous editions (of which the extracts previously edited by Sieben are the only exemplar), a page and a half to the orthography of the manuscript (which is not useful for our purposes), and two pages to explaining the methodology of the edition. Clearly, such an introduction is not going to settle many queries arising from the identification of Baldwin as the author. However, Narvaja also directs the reader to his article in the journal Stromata, published by the Universidad del Salvador in Buenos Aires.39 This contains his examination of the evidence for Baldwin’s authorship using both internal and external arguments. External Evidence from the Catalogues In the Stromata article Narvaja devotes a good deal of space to an examination of the catalogue entries and later lists of authors and their works that attribute the Liber de sectis to Baldwin from the fourteenth century up to the present.40 But, as we have seen, the evidence of the early modern and later catalogues which were compiled after Leland’s visit in 1534 adds nothing to our understanding, since he was apparently the last bibliographer actually to have seen the Canterbury manuscript. All subsequent incipits are derived from Leland, as is also the division of the whole into two separate works, whether they are found listed sequentially or not. The evidence from the post-1534 catalogues, then, is less than compelling. Digital Humanities and the Comparison of Texts Narvaja next goes on to present a series of parallel texts that he believes are similar to other works of Baldwin in terms of both style and content.41 For purposes of comparison he relies heavily on the two prefaces from the beginning of the work and the start of the ‘Orthodoxe fidei dogmata’ and some passages from the commentary. Any selective comparison such as Narvaja has carried out – probably by means of the Brepols database, the Library of Latin Texts (LLT), although he nowhere

37 38 39 40 41

Sieben, ‘Der “Liber de sectis”’, 263. See Narvaja, De sectis, 66 (fol. 170ra). Narvaja, ‘El Liber’. Narvaja, ‘El Liber’, 266–70. Narvaja, ‘El Liber’, 272–85.

30

Sabina Flanagan

describes his methodology – will be suggestive rather than definitive. Unfortunately Narvaja’s comparison suggests rather more to him than it does to me. First there are virtually no examples of actual duplication of phrases, let alone larger units, with the exception of biblical quotations and echoes. In these few cases it is interesting to note that the citations often support different arguments or appear in different contexts.42 That the Liber de sectis uses similar arguments, or has similar attitudes, to Baldwin concerning the role of heresy for strengthening the faith and on certain aspects of the Trinity is hardly surprising either.43 But if such comparisons of the Liber de sectis with the work of Baldwin are not persuasive, can better results be achieved if we assume that the Liber de sectis was written by Bartholomew? There is a problem here since LLT has the complete works of Baldwin but only one work by Bartholomew in the database. A further complication is that if the Liber de sectis were completed by Bartholomew any time after the mid-1170s, which I, with Sieben, think is the most likely date, it could easily incorporate aspects of Baldwin’s thought and even his style, since the two men are known to have worked closely together and to have exchanged their writings. Such influence would apply especially to the prefaces which Narvaja uses extensively for his comparisons, since we may assume that they were written last. It is worth noting that many of Narvaja’s supposed parallels come from Baldwin’s De sacramento altaris, which bears a dedicatory letter from Baldwin to Bartholomew, who no doubt possessed a copy.44 But if such targeted comparisons using the imperfect tool of LLT (imperfect because it can perform only Boolean searches and because of its restricted database) have proved inconclusive, perhaps the use of newer methods developed by digital humanities practitioners may provide the answer. Could the techniques of ‘computational stylistics’, recently applied to the sermons of Bernard of Clairvaux, be used to settle the question of authorship?45 We would need to compare the 40% of Liber de sectis, which is the original contribution of the author, with the works of Baldwin and ideally the remaining works of Bartholomew. If the complete corpus of both authors were the basis of comparison it might overcome the problem of comparing different genres, which would entail comparing different writing styles and subject matter with each other. At present such a strategy might be possible for Baldwin, but not Bartholomew, only one of whose works is available in a modern edition in a database. Before a proper comparison could be undertaken the manuscripts containing Bartholomew’s sermons and Dialogus contra Judeos would have to be transcribed and digitized.46 Moreover, supposing Bartholomew were the sole author of the Penitential, his original contribution to the collection of extracts from previous writers in this work would have to be established.47 42 See, for example, Narvaja, ‘El Liber’ at the foot of 274 where the phrases secundum mensuram donationis sue/causam et mensuram donationis sue refer in the first instance to revealing secrets and in the second to answering petitions of the Lord’s Prayer. 43 Narvaja on 277 pairs Quod fides contradiccione hereticorum magis clarescit with Quod fides ex comparatione impietatis laudem habet et gloriam which is not quite the same thing. 44 Morson, Le sacrament, 70. 45 Jeroen De Gussem, ‘Bernard of Clairvaux and Nicholas of Montiéramey: Tracing the Secretarial Trail with Computational Stylistics’ Speculum Digital Humanities (http://www.journals.uchicago.edu/ toc/spc/2017/92/S1), 190–225. See also Mike Kestemont, Sara Moens and Jeroen Deploige, ‘Collaborative Authorship in the Twelfth Century: A Stylometric Study of Hildegard of Bingen and Guibert of Gembloux’, Digital Scholarship in the Humanities 30, 2015, 199–224. 46 Dialogus, Oxford, MS Bodley 482; Sermones, MS Bodley 449. 47 Fortunately some of this work has already been done, see Taliadoros, ‘Penitential’.



The Authorship of the Liber de sectis hereticorum

31

If this were not enough, some of the assumptions of these programs, such as the reliance on function words, might be questioned.48 How representative of the author’s ‘individual stylome’ are the texts in the existing databases used by the investigators? Alarm bells should start ringing when we learn in De Gussem’s article that the Patrologia Latina is part of this database.49 But even the ‘state-of-the art scholarly edition[s]’ of the Brepols Library of Latin Texts represent the editor’s choice from a series of different readings in a number of manuscripts. The words set out in De Gussem’s Tables 1 and 2 appear to be just those most easily altered by a scribe who was not copying entirely mechanically – and there were such – as the variants noted in critical editions amply attest.50 So if it seems that this sort of global comparison is not going to be feasible any time soon, we must resort to other methods.51 Evidence from the Sources Used Earlier commentators on the text, such as Dvornik and Sieben, have noted the unusual range of sources chosen for the commentary. They include: the Adversus haereses of Irenaeus of Lyon; Eusebius of Caeserea’s Historia Ecclesiastica in the translation of Ruffinus; Cassiodorus’s Tripartite History; the Historia Longobardorum of Paul the Deacon; and the Pseudo-Isidorian Liber canonum. While some versions of these texts were generally available in twelfth-century England, that of Irenaeus was extremely rare. Even more intriguing is the incorporation into the Liber de sectis of what has been described as a treatise on the Seven General Councils, which the author claims to have ‘had recently translated from the Greek’.52 The question is where and when such a translation, or indeed, translator, could have been available to the compiler of Liber de sectis, whether we settle for Baldwin or Bartholomew. On this question Narvaja writes: ‘Turning to the reference of the author to the translation that he had made of the Tractatus de septem conciliis generalibus we can find nothing against the attribution of the work to Baldwin as since the seventh century there was a tradition of Greek studies at Canterbury.’53 But while Greek studies at Canterbury were possible in the time of Archbishop Theodore, in the late seventh century, there was no continuing tradition, and by the time of Archbishop Theobold (1139–61) and his successors in the twelfth century there would have been no obvious candidate in Canterbury who could have made such a translation. 48

De Gussem, ‘Bernard of Clairvaux’, 4–5. De Gussem, ‘Bernard of Clairvaux’, 5. 50 See De Gussem, ‘Bernard of Clairvaux’, 11, for table of such ‘function’ words. For example, a quick search of the apparatus of Bartholomew’s Contra fatilitatis errorem, Bell, 1996 turns up the following variants on 140: alique/aliqua. non/nec, and the omission of interdum while at 161 non is omitted in two different manuscripts and sicut is substituted for sic. It might be noted that the interests of the early proponents of the data base seem to have been more lexicographical than statistical and thus minor variations between texts were not so important. 51 Despite the difficulties outlined above, a project for using stylometry to decide between Baldwin and Bartholomew as author of the De sectis has recently been initiated by Dr Sébastien de Valeriola of the Université catholique de Louvain. I thank Dr Nicolas Ruffini-Ronzani for putting me in touch with his colleague (personal communication). 52 Extat tractatus grece editus de septem conciliis generalibus grecis quem ego nuper in latinum transferri feci. Narvaja, Liber de sectis, 100. 53 Narvaja, Liber de sectis, 19: ‘Volviendo a la referencia del autor a la traducción que mandó hacer del Tractatus de septem conciliis generalibus, podemos señalar queno contradice la atribución de la obra a Balduino, ya que desde el siglo VII se puede encontrar en Canterbury una tradición de estudios griegos’. 49

32

Sabina Flanagan

Narvaja’s implication is that Baldwin wrote the piece when he was archbishop, although he does not explicitly attempt to situate the work among Baldwin’s other writings either conceptually or temporally. However this period (1184–90) is inherently unlikely given the nature of Baldwin’s ecclesiastical and secular duties at the time. Moreover, after a brief honeymoon stage immediately following his appointment as archbishop, the animosity between the monks of Christ Church Canterbury and their archbishop and titular abbot reached such a pitch that Baldwin complained he was unable to officiate in his own cathedral and was allowed to visit it only as a pilgrim, so there is little likelihood that he was afforded free run of the convent library.54 If Canterbury was not itself a locus of Greek scholarship in the mid-twelfth century, there is, however, a possible connection to such a place in the person of John of Salisbury, who was part of the household of Archbishop Theobald. Although it is now believed that John himself was no Greek scholar, through his contacts with Poitiers, where his friend John of Canterbury was bishop, he had access to someone who did know Greek. This was the enigmatic John Saracen, to whom he addressed a letter inquiring about the meaning of Greek words used in theological disputation.55 He also asked his friends at Poitiers about the progress of the translation John Saracen was making of Pseudo-Dionysius (1166/7).56 John of Salisbury knew and corresponded with both Baldwin and Bartholomew, though he was probably closer to Bartholomew, since they had both been members of Theobald’s familia at Canterbury. Purpose of the Work But more clues about the authorship of the work can be gained by asking what might have been the purpose of the compilation. Rather than ascribing it to the years c. 1185–90, when Baldwin was archbishop, I believe it fits better with an earlier date. Given the nature of the contents and its rather loose articulation, it could well have been conceived as a kind of dossier, worked on at various times and possibly finalized in anticipation of the Third Lateran Council, held in 1179. On this assumption, Bartholomew is a better candidate than Baldwin for several reasons. The first is his position. By the 1170s, Bartholomew had long been a bishop, while Baldwin was still merely abbot of Forde. Bartholomew was close to Pope Alexander III and attended the Council of Tours in 1163. He may well have shared Alexander’s enthusiasm for councils and been meditating this dossier for some time. In the event, he was not present at Lateran III since Henry II allowed only four English bishops to travel to Rome. However, such a dossier would have proved useful to anyone who attended, or who might have been expecting to attend. Indeed, the preface to the second part of the Liber de sectis, that is, the ‘ Orthodoxe fidei dogmata’, begins by extolling the role of councils, where ‘truth is built upon, religion asserted, falsity refuted, irreligion condemned, the falsehoods of heretics demolished, the makers of error and cultivators of perverse teachings subjected

54

Epistolae Cantuarienses ed. William Stubbs in Chronicles and Memorials of the Reign of Richard 1. 2 vols, RS 38, II, London 1865, 316. 55 John of Salisbury, The Letters of John of Salisbury: The Later Letters (1163–1180), ed. by W. J. Millor and C. N. L. Brooke, OMT, Oxford 1979, 268–75. 56 John of Salisbury, Letters, 92–3, 388–9, 424–5.



The Authorship of the Liber de sectis hereticorum

33

to well-deserved malediction’.57 After a good deal more in the same vein, the author concludes by claiming that out of the obscure and prolix writings by the Fathers on the subject he has collected in unum (in one place) what will be of benefit both to himself and to other potential readers as being more handy, and easier to understand and recall. In order to perform this service Bartholomew, as a member of the secular clergy, did not require permission to write. By contrast Baldwin, as a Cistercian, was not allowed to write anything without the express direction of his superiors.58 Evidence from Temperament If we consider the two men and the nature of their surviving works, another difference is apparent. On the evidence of his writings, Bartholomew emerges as a controversialist. This may have been connected with his episcopal status, since a bishop was meant both to promote the faith and combat its enemies, a task he took seriously as the titles of his works – Contra fatalitatis errorem, the Dialogus contra Judeos – and his didactic Penitential attest. Interestingly, in the letter of dedication to his late work of the 1180s, Contra fatalitatis errorem, Bartholomew mentioned fatalism as the error that was most troublesome in his pastoral role.59 The fact that he did not write about fatalism until this late date is usually explained by his being too busy with episcopal duties, but it could also be that in the 1160s and 1170s he was writing about heresies on a wider scale, more relevant to the upcoming council, than of specifically English concern.60 Baldwin’s titles and works are of a different nature. In the De commendatione fidei Baldwin is arguing for something, rather than against it. Leclercq seems to acknowledge this aspect of Baldwin’s work when he describes Baldwin’s writing in De sacramento altaris as ‘théologie admirative’, and Bell also remarks on its nonpolemical nature.61 In the light of these considerations it seems to me that the De sectis hereticorum et orthodoxe fidei dogmata aligns more closely with the works of Bartholomew than Baldwin. Evidence from Subject Matter We might also ask whether the content of the Liber de sectis engages with subjects of interest to the Third Lateran Council. First, as pope in time of schism, Alexander III needed to be accepted as the legitimate pope and also needed his councils to be recognized as genuine. It is significant that he took pains to declare the Council of Tours a general council in line with previously recognized councils. So matters of papal authority and the relation of this authority to councils was a current problem. The question of heterodoxy, and indeed heresy, was also topical. On matters of doctrine we find in the Liber de sectis reference to the problem of nihilianism. This was the doctrine that Christ, in his human nature, was ‘nothing’, his essential 57

Ibi ueritas astruitur, pietas asseritur, falsitas arguitur, impietas condempnatur, fraudes hereticorum deteguntur, fabricators errorum et cultores peruersorum dogmatum debite malediction subiciuntur. Narvaja De sectis, 84. 58 See, for example the preface to De commendatione fidei. Bell, 344. 59 Bartholomew of Exeter, Contra fatalitatis errorem, ed. David N. Bell, CCCM 157, Turnhout 1996, 11. 60 See David Runciman, ‘Bishop Bartholomew of Exeter (d. 1184) and the Heresy of Astrology’ Journal of Ecclesiastical History online (https://doi.org/10.1017/S0022046918001306). 61 Le sacrament, ed. Morson, I, 37; Bell, ‘Baldwin of Forde and the Sacrament of the Altar’, in Erudition in God’s Service, ed. John Summerfeldt, Kalamazoo 1987, 217–42 at 217.

34

Sabina Flanagan

being attributable to his godhead alone, an ongoing debate in the 1170s which was addressed at the Third Lateran Council.62 Heresy on a larger scale was also tackled, and anti-Albigensian measures were promulgated. The role of councils in electing popes was defined, too.63 Although it seems likely that the Liber de sectis was finally put together in the 1170s with the Third Lateran Council in mind, it may well have been started many years earlier, and here we must return to the identification of the sources used. We have already noted that the most unusual was Irenaeus of Lyon’s Contra haereses (in a third-century translation). Indeed, Sieben claims that this text was almost unknown in the Middle Ages.64 There was, however, a copy of Contra haereses at Christ Church Canterbury.65 As we have noted, Bartholomew had a long association with Canterbury, having joined the household of Theobald in 1142. He only left it in 1155 to become archdeacon of Exeter and thereafter bishop in 1161. He continued his association through John of Salisbury, and in due course preached the sermon of reconciliation in the cathedral in 1171 after the murder of Thomas Becket. If, then, the work depends on the Canterbury Irenaeus, Bartholomew could have begun his project any time he was in Theobald’s service at Canterbury from 1142 to 1155. Furthermore, if we look at the order in which the sources of Liber de sectis are utilised and plot their localities, an interesting possibility emerges. It seems that by the time Bartholomew takes up the second distinction, at chapter 50, he has moved to Exeter. He uses Eusebius, which was at Exeter, also Liber canonum (Pseudo-Isidore) and Isidore of Seville. Although there is some overlap in the middle, the sources used for the beginning of the work were at Canterbury and not Exeter, and the sources used at the end of the work were at Exeter and not Canterbury. Excerpts from the translation of the Greek treatise on the seven councils occur in both the Canterbury and Exeter sections, which suggests that Bartholomew was using his own personal copy (or that he inserted the earlier references into the compilation at a later date, always a possibility given the nature of the work). Where, then, might Bartholomew have obtained his translation of the treatise on the Greek councils? It is tempting to think that it might have been at Poitiers. There is a gap in his biography around 1169/70, when Bartholomew is said to have retired briefly to a monastery. Could it have been at Poitiers, where we know John of Salisbury had connections? Although the difficulty of leaving and re-entering England at the height of the Becket dispute might have made such a move problematic, it would explain how he could come to write that he had had the text ‘recently translated’. On the other hand, while it is true that Baldwin was also at Exeter from around 1155, he had no close or extended connection with Canterbury until his unfortunate association with it from 1184 to 1190. By 1170 Baldwin was monk, then abbot, at Forde, writing his sermons and treatises on faith and the Eucharist, at the time when I propose that Bartholomew resumed his work on heresies and councils, now using the books available at Exeter cathedral. There is one last point I would like to make about the copy of the Liber de sectis in the possession of Thomas Basin. Might he himself have had the manuscript of the Liber de sectis copied while it was at Canterbury? We know from his Breviloquium

62

See Marcia L. Colish, Peter Lombard, Leiden 1994, vol. 1, 427–38. See Raymonde Foreville, Latran I, II, III, IV, Histoire des conciles oecuméniques 6, Paris 1965, 134–58. 64 Sieben, ‘Der “Liber de sectis haereticorum”’, 264. 65 BL MS Cotton Galba E IV, fol. 131v, col 2. James thought that this manuscript was still in existence but the Arundel manuscript he mentions (James, Ancient Libraries, xl) has now been identified as a later German importation. 63



The Authorship of the Liber de sectis hereticorum

35

that he spent two months in London in 1438. He had been on his way to the papal court at Ferrara and decided to travel via England because of fighting in France. He says that he was delayed by illness in London, but unfortunately tells us nothing about what he did during his enforced stay.66 Could he have visited Canterbury and seen the manuscript? If he had had it copied there, it could have remained among the books that accompanied him on his travels until he ended up in Utrecht and had it recopied into the Paris manuscript. Given the timing, as Basin was on his way to a council which dealt with many subjects covered in the text, it would have been a most opportune find, but once again the evidence for such a happy conjunction is lacking. Conclusion This paper has ranged widely both in time and space. My arguments for suggesting that the Liber de sectis was not by Baldwin of Forde depend partly on questioning Narvaja’s evidence for the assertion that it was, and partly on my attempt to show why Bartholomew of Exeter is a more likely candidate. Of course, much of what I have been arguing in this paper is not susceptible of definitive proof, given the gaps in our knowledge of many events and circumstances. But in the absence of any investigation at all of the intriguing puzzles presented by Narvaja’s identification of the Liber de sectis as a lost work of Baldwin of Forde, I have thought it worth trying to produce some arguments for a reassignment of authorship to Bartholomew of Exeter, which are at least plausible, if not proof positive. Appendix One of the pleasures of attending a gathering like the Battle Conference is the opportunity to meet scholars at the beginning of their careers who are working in one’s particular field of interest. Furthermore, at the Battle Conference I found out about the recent graduate research by Dr Susanne Coley at the University of Southampton. She generously allowed me to read and quote from her recent PhD thesis on the Cistercians and heresy, in which she treats the Liber de sectis at length as a genuine work by Baldwin de Forde.67 I should like to address a couple of points that her thesis raises which have particular relevance for my paper. The Eusebius from Forde Dr Coley points to the fact that the twelfth-century Eusebius from Forde – now Oxford, Bodl. MS Laud Misc 450 – bears certain annotations and highlighting in a twelfth-century cursive hand. After some discussion of these she concludes, ‘There can be no doubt that this manuscript was used heavily during the composition of the Liber de Sectis Hereticorum. … At multiple points throughout the texts indicators were drawn to highlight relevant sections of the text which would be transcribed into the main body of his tract on heresy.’68 She also notes that the correspondence between the annotations and what was included in the text is not exact, and also that ‘there are a few differences in the spelling and ph[r]asing at times between the two versions, but they are reasonably small scale discrepancies which could easily 66

BnF Lat 5970, fol. 60v. Suzanne Coley, ‘Archbishop Baldwin of Canterbury and the Fear of Heresy in Late Twelfth-Century England’, PhD thesis, University of Southampton 2018 (embargoed until 2021). 68 Coley, thesis, 79. 67

36

Sabina Flanagan

be explained by scribal alterations or errors.’69 However, until a more thorough examination of such instances can be made it seems to me that any assertion that this particular text was used by Baldwin for the Liber de sectis must be open to doubt. Indeed, examination of the extant versions of the chief sources of the Liber de sectis, such as Eusebius and Pseudo-Isidore, which are known to have been available at specific locations at the appropriate times may prove helpful in deciding between Baldwin and Bartholomew. My suggestion that the order of the texts in the Liber de sectis was a reflection of Bartholomew’s move from Canterbury to Exeter might be strengthened if a comparison of the sources could clearly identify which manuscripts from Canterbury he was using in the first instance, and which manuscripts from Exeter later on. Of course, this would depend on the survival of, and sufficient differences between, the relevant manuscripts, which is unfortunately not the case for the Eusebius which was formerly at Exeter. However, the extracts from Eusebius in the Liber de sectis might be compared with the Canterbury Eusebius, Cambridge Corpus Christi College MS 187, and for good measure with the manuscript from Forde, Oxford MS Laud Misc. 450. Similarly, the texts of the Canterbury Pseudo-Isidore (Collectio Lanfranci), Cambridge Trinity College MS B16.44, could be checked against the Pseudo-Isidore that was at Exeter.70 But such investigation, like any further stylometric study, is beyond the scope of this paper. Peter of Blois’ Prologue to the De fide When I delivered this paper I was aware of Cotts’s discussion of the passage from a letter of Peter of Blois that mentions Baldwin by name.71 I did not think it was a sufficiently unambiguous connection of Baldwin with the Liber de Sectis to influence my argument, since it could be sufficiently explained by Peter’s obvious borrowings from the De commendatione fidei. However, Dr Coley’s examination of Oxford Jesus College MS 38 has established beyond doubt not only that Peter of Blois depends on Baldwin’s text on faith in his own De fide, but also that the portion on the heresies (chapters 45–112) is largely a distillation from the Liber de sectis. So if one already believes that the Liber de sectis was written by Baldwin, Peter’s statement in the preface to his own work can be read as a confirmation of Baldwin’s authorship: In this work I follow, indeed honour, the footsteps of the venerable fathers who wrote about the faith, and raised their conquering standards against heresies. Among these most shining in renown were Eusebius of Caesarea, Athanasius, Ruffinus, Irenaeus, Hilary, Jerome, Augustine, Epiphanius, and Baldwin, the primate of England, who although later in time, is hardly lesser in terms of his life, wisdom, and holiness.

However, if Baldwin’s authorship is not already assumed (and the burden of my paper has been to question this assumption) then the passage can be read differently – not against the grain perhaps – but chiastically. The grammatical/rhetorical figure of chiasmus occurs where the order of elements (words/clauses) is reversed so that they could be schematically represented as a diagonal cross (X). It is frequent in clas-

69

Coley, thesis, 80. Eton College MS 97. 71 See Note 17 above. Peter of Blois, The Later Letters, 323–9; John D. Cotts, The Clerical Dilemma: Peter of Blois and Literate Culture in the Twelfth Century, Washington, DC 2009, 237–41. 70



The Authorship of the Liber de sectis hereticorum

37

sical literature, the Bible, and in medieval writing.72 Indeed, Peter includes several examples elsewhere in his preface, ‘ponentes LUCEM TENEBRAS et TENEBRAS LUCEM’ being a simple example. The figure can be extended to larger elements, however, including structural blocks of an entire work. The passage, which I believe should be read chiastically, is more complicated than a couple of words but the principle is the same. The passage in question reads: In hoc autem opera sequor, immo adoro vestigia venerabilium patrum qui tractaverunt DE FIDE et CONTRA HERESES victrices aquilas erexerunt. Inter quos celebrioris fame titulus effulsere EUSEBIUS CESARIENSIS, ATHENASIUS, RUFFINUS, IRENEUS, HILARIUS, IERONIMUS, AUGUSTINUS, EPIPHANIUS, et PRIMAS ANGLIE BALDEWINUS, qui licet sit posterior tempore non multum tamen ab eis degenerat, vita, scientia, sanctitate.73

Read in this way, Peter is acknowledging his debt to Baldwin for the section on faith and to the earlier Fathers for the section on heresy. It is important to note here that Peter’s text also falls into two quite distinct parts: chapters 1–33 on faith, and 33–101 on heresy. However, while Peter might be seen as covering himself by this acknowledgement against the charge of plagiarism, a matter about which he seems to have been somewhat sensitive, the following passage, describing his method, leaves him open at least to a charge of disingenuousness. After employing a well-worn trope about producing honey from other men’s flowers he goes on to say: Et quia delicatis auribus solet esse onerosa prolixitas de illa pluralitate voluminum, ea duntaxat que moderne utilitatis exigentie congruebant contra hereticorum blasphemias quasi in fasciculum coartavi. And because prolixity is burdensome to delicate ears, from that abundance of volumes I have put together the things which exactly suited the demands of modern utility against the blasphemies of heretics in a small compass.74

The implication here is that he consulted the volumina of the fathers mentioned, laboriously and individually, rather than in a comprehensive, though unwieldy, compilation such as the Liber de sectis. The discrepancy might well have been noted by any recipient of the De fide who was familiar with the De sectis. But what if the Liber de sectis were by Bartholomew? Peter does not seem to have been so scrupulous in attributing his borrowings to Baldwin’s former friend and colleague. After completing his work on faith and heresies, Peter also wrote a treatise against the Jews as a kind of companion piece.75 This unpublished work has not received much study and no one, to my knowledge, has remarked that it depends on Bartholomew’s similarly unpublished dialogue against the Jews.76 Bartholomew’s work, like the Liber de sectis, is a storehouse of texts, mostly biblical in this instance, rather than patristic. Peter has utilized it for his own treatise, selecting from among the relevant texts and following closely the sequence found in Bartholomew. More significantly he has included (as in the case of the borrowings from De

72

It is a favourite trope of Bernard of Clairvaux and is also found in John of Salisbury, see Letter 276, 582–3 ‘PERPETUITATE FELIX et FELICITATE PERPETUUS gaudeant’ (May Felix enjoy perpetuity and Perpetuus felicity). 73 Peter of Blois, Later Letters, 325; Oxford Jesus College MS 38, fol. 84v. 74 Peter of Blois, Later Letters, 325–6; Oxford Jesus College MS 38, fol. 84v. 75 Oxford Jesus College MS 38, fols 70–83. 76 Oxford Bodleian Library MS 482, fols 1–79.

38

Sabina Flanagan

sectis) parts of Bartholomew’s original commentary and linking passages.77 Given these considerations, Peter’s prologue to the De fide, rather than being an unambiguous pointer to Baldwin as author of Liber de sectis, provides further evidence that it is more likely to have been the work of Bartholomew.

77

Compare, for example, Oxford Jesus College MS 38, fol. 79 and Oxford Bodleian Library MS 482, fol. 47.

EVIDENCE OF THE ORDINARY: WIVES AND CHILDREN OF THE CLERGY IN NORMANDY AND ENGLAND, 1050–1150* Hazel Freestone Clerical marriage has been the subject of debate for the best part of 1,700 years, and is as contentious now as it was in the eleventh and twelfth centuries. Hitherto, modern scholars have focused on the campaign to impose celibacy on the secular clergy primarily from the perspective of medieval ecclesiastical authorities intent on eradicating clerical marriage. Recent studies have overwhelmingly relied on normative sources and related texts written by ‘reformers’, the term, for the purposes of this paper, I use to describe those ecclesiastics intent on imposing celibacy. Modern historians have tended to treat the issue as a predominantly theological and ideological one, centred on the papacy in western Europe.1 As a result current scholarship perpetuates the same unquestioning, or at least unwitting, assumptions – that the ‘reformers’ were correct in their characterization of the failings of the church, and that their solution (the imposition of cultic purity) had indeed made the church ritually and spiritually purer.2 Following this interpretation the practice of clerical marriage necessarily undermined the church, and the drive to enforce celibacy on all secular clergy, alongside the idea of ‘reform’ more generally, has been assumed to be a ‘Good Thing’.3 Accordingly, a married cleric has been considered ‘bad’. Such attitudes are in part responsible for the relative neglect of the sources defending clerical marriage until recently.4 Awareness that the current grand narrative is at

*

I am very grateful to Elisabeth van Houts, Emily Ward and James Kane for their comments on early drafts of this paper. 1 A. Theiner, Ein Beitrag zur Kirchengeschichte, 2 vols, Altenberg, 1828; H. C. Lea, The History of Sacerdotal Celibacy in the Christian Church, 3rd edn, 2 vols, New York 1907; Augustin Fliche, La réforme Grégorienne, 3 vols, Louvain-Paris 1924–37, esp. vol. 1; Charles Frazee, ‘The Origins of Clerical Celibacy in the Western Church’, Church History 41, 1972, 149–67; John E. Lynch, ‘Marriage and Celibacy of the Clergy: The Discipline of the Western Church: An Historico-Canonical Synopsis’, Part I and Part II, The Jurist, 1972, 14–38; 189–212; Uta-Renate Blumenthal, The Investiture Controversy: Church and Monarchy from the Ninth to the Twelfth Century, Philadelphia, PA 1988; Medieval Purity and Piety: Essays on Medieval Clerical Celibacy, and Religious Reform, ed. Michael Frassetto, New York 1998; Kathleen G. Cushing, Papacy and Law in the Gregorian Revolution: The Canonistic Work of Anselm of Lucca, Oxford 1998; Leidulf Melve, ‘The Public Debate on Clerical Marriage in the Late Eleventh Century’, JEH 61, 2010, 688–706; Jennifer Thibodeaux, The Manly Priest: Clerical Celibacy, Masculinity and Reform in England and Normandy, 1066–1300, Philadelphia 2015. 2 Bernard Verkamp, ‘Cultic Purity and the Law of Celibacy’, Review for Religious 30, 1971, 199–217 at 206–7; Paul Beaudette, ‘“In the world but not of it”: Clerical Celibacy as a Symbol of the Medieval Church’, in Medieval Purity and Piety, ed. Frassetto, 23–46; J. Barrow, ‘Ideas and Application of Reform’ in The Cambridge History of Christianity, vol. 3, ed. T. F. X. Noble and J. M. H. Smith, Cambridge 2008, 345–62. See also M. Douglas, Purity and Danger, London 1966, esp. 8–50 and Maroula Perisanidi, Clerical Continence in Twelfth-Century England and Byzantium, London-New York 2019, 116–58. 3 Barrow, ‘Ideas and Application of Reform’, 346; Gerd Tellenbach, The Church in Western Europe from the Tenth to the Early Twelfth Century, trans. Timothy Reuter, Cambridge 1993, 157–8 n 78. 4 Anne Barstow, Married Priests and the Reforming Papacy. The Eleventh Century Debates, New York-Toronto 1982; Brigitte Meijns, ‘Opposition to Clerical Continence and the Gregorian Celibacy

40

Hazel Freestone

best problematic and at worst inadequate is gradually gaining traction, especially in the work of Maureen Miller, Conrad Leyser and Kathleen Cushing.5 As yet no new consensus has been reached. In respect of England and Normandy, the focus of this paper, the most recent studies examining aspects of clerical celibacy and marriage – by Jennifer Thibodeaux, Hugh Thomas and Maroula Perisanidi – have not questioned older narratives and assumptions.6 My central argument is that the married clergy, their wives and their children are missing almost entirely from the current debate, except in a piecemeal and selective fashion. If we are to understand how celibacy became the norm amongst them, we need to understand not only who these men and women were, but also when and where they can be found throughout the period. Only then can we offer conclusions as to what, if anything, changed over time. Polemical and normative sources need to be treated much more cautiously, and placed alongside other available evidence, such as charters, cartularies, Liber vitae, chronicles, letters and saints’ lives. Integrating evidence that is not primarily concerned with clerical marriage either in purpose or intent can provide a glimpse into the lived experience of married clergy across the period, within their local and regional context. Priority should be given to sources which were written contemporaneously with, or within living memory of, the events that they describe. Uncovering details of the lives of the wives and women of secular clergymen in Normandy and in England is fraught with difficulty.7 To avoid extrapolating broader conclusions from a narrow study, this paper is based on a systematic survey of all the dioceses in Normandy and England, with the exception of Carlisle, which was founded only in 1133 and is poor in sources for the first half of the twelfth century. I have identified priests’ wives where possible, and their children and husbands where direct evidence for them is lacking, whilst keeping in mind the distinctive histories of the Norman and English churches as well as the impact of the Conquest of 1066 on both Normandy and England. As far as possible, it is essential to see the period between 1050 and 1150 on its own terms, rather than framing research with the observations and views of late twelfth-century or thirteenth-century writers.8 Such an approach can transform our understanding of the importance and influence of the ideal of clerical celibacy, as reflected in glimpses of the lives of the wives, children and clergymen themselves. I focus on the ranks below that of bishop, much of which has been discussed elsewhere, though not systematically gathered together. My focus on women and children below episcopal rank is important because it is necessary to prove that marriages existed and continued to exist across the period, Legislation in the Diocese of Thérouanne: Tractatus pro clericorum conubio (c. 1077–1078)’, Sacris Erudiri 47, 2008, 223–90; Melve, ‘The Public Debate’, 688–706; Elisabeth van Houts, ‘The Fate of Priests’ Sons in Normandy with Special Reference to Serlo of Bayeux’, HSJ 25, 2013, 57–105; Thibodeaux, The Manly Priest. 5 Maureen Miller, ‘The Crisis in the Investiture Crisis Narrative’, History Compass 7/6, 2009, 1570– 80; Conrad Leyser, ‘Review Article: Church Reform – Full of Sound and Fury, Signifying Nothing?’, EME 24, 2016, 478–99; Kathleen Cushing, ‘Canon Law on the Peripheries: Rethinking Reform in Eleventh-Century Church Councils’, unpublished paper given at Leeds, International Medieval Congress, 2 July 2018. 6 Thibodeaux, The Manly Priest; Hugh Thomas, The Secular Clergy in England, 1066–1216, Oxford 2014; Perisanidi, Clerical Continence. 7 For what follows, see my PhD thesis: H. A. Freestone, ‘The Priest’s Wife in the Anglo-Norman Realm, 1050–1150’, unpublished PhD thesis, University of Cambridge 2018. 8 Both Jennifer Thibodeaux and Hugh Thomas frame their research in reference to the thirteenth century. Maroula Perisanidi dates her analysis from 1130 to 1215, but the majority of her evidence was written post-1180.



Wives and Children of the Clergy in Normandy and England

41

especially as many normative sources, and scholars relying on them, assume that they did not. In the first part of this paper, I will briefly outline the prosopographical data for Normandy and England, before turning to the evidence for married archdeacons and secular canons in both regions. I will then discuss hereditary succession amongst parish clergy, where my findings raise fresh questions regarding ‘hereditary’ churches. Finally, I will examine two narrative sources which proved unexpectedly useful in revealing attitudes towards priests’ wives and married clergy, countering the impression of the polemical writings that modern scholarship has accepted almost without question. Prosopographical Data from Normandy and England Prosopographical evidence reveals clearly that wives and children of the clergy form a group of significant size. Their presence in clerical environments was neither unusual nor hidden. In total, between 1050 and 1150, I identified 392 married clergymen: 76 in Normandy, and 316 in England. A total of 67 women were named and/or identified by a descriptor, such as uxor, coniunx, socia, compar, amica or concubina.9 Of these, 14 women were found in Normandy, and 53 in England. The vast majority of references concern a single man with one child – normally a son. These numbers most likely represent only a fraction of family members who we can reasonably assume lived in or near ecclesiastical communities.10 In Normandy, my findings are biased in favour of Bayeux and Rouen, since the surviving evidence is greater for these two dioceses than for the others.11 In England, the data is strongly weighted in favour of the diocese of Lincoln, where it proved possible to identify many more married clergy than elsewhere. This is due to the volume and accessibility of surviving evidence for the diocese, and does not necessarily indicate clerical marriage was more prevalent there than elsewhere. The initial data set for England was drawn from approximately 1,900 episcopal acta dated between 1050 and 1170 (forty-five of which provided evidence for clergy who were married or had children), and the Fasti Ecclesiae Anglicanae, which comprehensively lists and dates men holding offices within each diocese, and very occasionally provides additional biographical details.12 Both Norman and English data sets were supplemented with evidence resulting from close examination of cartularies, charter collections, Liber vitae, chronicles, saints’ lives, letters and unpublished manuscripts. Throughout this data-collecting exercise my working assumption was that clerical marriage was probably a norm at the beginning of the period and remained so until the evidence demonstrated otherwise. The evidence for the wives and children is substantial enough to demonstrate that the date at which clerical marriage was no longer the norm, especially in England, should be 9 Uxor, coniunx mean wife; socia, and possibly compar, are synonyms for wife; amica is usually translated as mistress or lover; while concubina refers to a woman, probably of lower status than her partner, in a long-standing affective relationship, but without rights of dower or inheritance. There remains a lack of research or analysis on the language used to describe women in this period – and many modern translations reflect modern misogyny rather than reflecting usage of the time. 10 See Appendix. 11 My initial data set was derived from David S. Spear, The Personnel of the Norman Cathedrals during the Ducal Period, 911–1204, London 2006; idem, ‘Additions and Corrections to David S. Spear, Personnel of the Norman Cathedrals during the Ducal Period, 911–1204’, Tabularia, Études 14, 2014, 151–94. Hereafter Spear, NP. And Orderic. 12 English Episcopal Acta. Fasti Ecclesiae Anglicanae, 1066–1300, new edn, 10 vols, ed. Diana Greenway et al. London 1968–. Hereafter Fasti.

42

Hazel Freestone

pushed much later than has been previously argued, just as Julia Barrow suggested in her important study of secular clergy in northwestern Europe.13 Archdeacons In Normandy and England, the office of archdeacon was not yet clearly defined, beyond the notion that the archdeacon was usually second in command to a bishop, with responsibilities regarding justice and discipline amongst the diocesan clergy as well as management of a bishop’s secular affairs.14 The office only gradually evolved over the course of the eleventh and twelfth centuries.15 It was not primarily a spiritual role and the archdeacon did not celebrate the Mass. With these reservations in mind, it is apparent that if an archdeacon himself was married or sympathetic to married clergy, he was in a position to profoundly affect the implementation of celibacy within his jurisdiction. Furthermore, many of the men initially appointed as archdeacons were married, something which is more visible in the English evidence, as we will see in due course. David Spear has identified and collated the core prosopographical work on archdeacons in Normandy.16 The exact number of archdeacons is unknown, although our best evidence suggests each diocese was divided into three or more archdeaconries. We can therefore estimate that there were between twenty-five and thirty-three archdeacons at any one time in Normandy. In total Spear has identified 113 individuals who acted as archdeacons, and of these, thirteen appear to have been married or had sons. Only Robert, archdeacon of Évreux, active c. 1079 × 1087 (and died before 1113), can be firmly linked with a named wife, Adelis. In a charter recording a gift to the church of St Martin de Rouvray, dated between April 1099 and 31 March 1100, the donor Gilbert (Gislebert) identifies himself as the son of Robert, the archdeacon of Évreux. The pro anima clause contains a dedication to the souls of both his parents, and, unusually, names the donor’s mother. The witnesses of the charter are acting on behalf of the archbishop of Rouen, who in 1099 was William Bona Anima (1079–1110), as well as representatives of the abbot and chapter of the abbey of Jumièges.17 William Bona Anima was himself the son of a bishop, Radbod of Sées (d. after 1032), and had begun his career as an archdeacon, before becoming a monk and later abbot of St-Étienne, Caen. William also had a son, identified by Richard Allen as Hugh ‘nostri filii [sic]’, albeit with reservations, not least regarding the use of the plural and the fact that the document itself is a transcrip13

Julia Barrow, The Clergy in the Medieval World: Secular Clerics, their Families and Careers in North-Western Europe, c. 800–1200, Cambridge 2015, 139. 14 Barrow, The Clergy, 49. 15 Concise Oxford Dictionary of the Christian Church, ed. E. A. Livingstone, 2nd rev. edn, Oxford 2006; online version 2013 accessed 9 February 2015; C. N. L. Brooke, ‘The Archdeacon and the Norman Conquest’ in Tradition and Change: Essays in honour of Marjorie Chibnall, ed. D. Greenway, C. Holdsworth and J. Sayers, Cambridge, 1985, 1–19; J. Barrow, ‘Grades of Ordination and Clerical Careers, c. 900–1200’, ANS 30, 2007, 41–61 at 50. 16 Spear, NP. Also see: David Bates, Normandy Before 1066, London-New York 1988, 260–2; David S. Spear, ‘L’administration épiscopale normande: Archidiacres et dignitaires des chapitres’, in Les évêques normands du XIe siècle, ed. P. Bouet and F. Neveux, Colloque de Cérisy-la-Salle, 30 Septembre–3 Octobre 1993, Caen 1995, 81–102. 17 Chartes de l’Abbaye de Jumièges, ed. J.-J.Vernier, 2 vols, Paris-Rouen, 1916, I, no. 40: Ego Gislebertus, filius Rotberti archidiaconi Ebroicensis, filium meum nomine Hugonem sub regula sancti Benedicti militaturum Summo Regi apud cenobium sanctę Marię Gemmetici, devota mente, obtuli. Postmodum vero, in eadem die, pro susceptione pueri, pro anima Balduini, episcopi Ebroicę urbis, patris quoque mei matrisque Adelidis parentumque meorum cęterorum …



Wives and Children of the Clergy in Normandy and England

43

tion of a now lost original. If the identification is correct, however, then Hugh was most likely born whilst Bona Anima was an archdeacon in Rouen.18 In addition to William Bona Anima we can we can identify a group of archdeacons and their sons active at Rouen in the middle of the eleventh century, when the campaign to impose clerical celibacy was especially fierce in Normandy. Benedict, canon, treasurer and archdeacon of Rouen and active c. 1055 to 1106, probably had two sons, John and Nicholas.19 Another possible married archdeacon was Ralph of Chaudry, who may have been an archdeacon of Rouen, active before 1070s. If the identification is correct, he had a son called Fulk or Fulbert, who was an archdeacon of Rouen active c. 1075.20 At the same time, Gauthier or Gother was active as an archdeacon of Rouen and had a son, Walter.21 Thus, at the time of the council of Rouen in 1072, the third of four councils in Normandy that dealt with clerical marriage before 1100, the majority of archdeacons in Rouen were married and/or had children. In one canon from the 1072 council, we can hear the voice of John d’Ivry, archbishop of Rouen (1167–79), addressing archdeacons ‘who ought to govern’, and who were banned from having ‘a concubine or supposed wife or harlot, but are to live chastely and properly giving an example of chastity and holiness to the clergy under them.’22 Far from expressing an abstract ideal or a general indictment of archdeacons, it seems to me that this should be read as John’s direct rebuke to his own clergy, whom he was unable to persuade to abandon their wives or remove from their positions. Turning to England, a significant number of archdeacons were married from the inception of the office after the Conquest of 1066. Prior to the Norman Conquest references to archdeacons are rare.23 Out of the fifteen English dioceses, married archdeacons, or at least archdeacons with children, have been identified in nine so far: Canterbury, Coventry & Lichfield, Ely, Exeter, Lincoln, London, Norwich, Winchester and York. Such a wide geographical distribution shows that married archdeacons were not restricted to one particular area.24 By the first quarter of the twelfth century, these nine dioceses were divided into thirty-four archdeaconries, in which thirty-one married archdeacons can be identified between 1070 and 1180. Christopher Brooke identified four married archdeacons in the diocese of London alone before 1150.25 In the diocese of Lincoln, the Liber vitae of Thorney abbey refers to the only woman identified as ‘wife of the archdeacon’ (uxor archidiaconi). The reference appears in a block of names on fol. 3v which reads: Adam, Agustinus, Gunnilda, 18 Spear, NP, 197, 271; D. Spear, ‘William Bona Anima, Abbot of St Stephen’s, Caen, 1070–1079’, HSJ 1, 1989, 51–60; for his son see BN, coll. du Vexin, vol. 4, 156–7 as cited by Richard Allen, ‘The Norman Episcopate, 989–1110’, 2 vols, unpublished PhD thesis, University of Glasgow 2009, Appendix D, 471. 19 Spear, NP, 207, 219, 247; Allen, ‘Norman Episcopate’, PhD, Appendix G, no. 73. 20 BnF Baluze MS 77, no. 13 as cited in Spear, NP, 208 – who notes that ‘Fulk’ could be a mistranscription for ‘Fulbert’, at present it is impossible to determine. 21 Baluze MS 77, no. 48 as cited in D. Spear, NP, 208, Additions, 177. 22 Orderic II, 290–1: Archidiaconi qui eos regere debent, non permittantur aliquam habere nec concubinam nec subintroductam mulierem, nec pelicem, sed caste et iuste uiuant, ex exemplum castitatis et sanctimoniæ subditis præbeant. Oportet etiam ut tales decani eligantur, qui sciant subditos redarguere et emendare. quorum uita non sit infamis, sed merito præferatur subditis. 23 Frank Barlow, The English Church, 1000–1066, 2nd edn, London 1979, 247. 24 See Freestone, ‘The Priest’s Wife’, Chapter Two. 25 C. N. L. Brooke, ‘The Composition of the Chapter of St Paul’s, 1086–1163’, Cambridge Historical Journal 10, 1951, 111–32 at 124–5. To these we can add, Walter, archdeacon of Waltham, c. 1102–1115 and his son Alexander, see: Feudal Documents from the Abbey of Bury St Edmunds, ed. D. C. Douglas, London 1932, no. 172.

44

Hazel Freestone

Emma, Cecilia, et uxor archidiaconi.26 Cecily Clark proposed that the Adam heading this list could perhaps be identified with Adam of Stukeley, son of Henry of Huntingdon, the author of the Historia Anglorum. If so, she suggested the unnamed woman is either Henry’s wife or the wife of his father, Nicholas, the first archdeacon of Huntingdon, Cambridge and Ely.27 The Thorney Liber vitae was most likely written between 1099 and 1113, which fits the known dates for either man.28 Thorney abbey fell under Nicholas and Henry’s archidiaconal jurisdiction, and on a personal level Henry and his son Adam held land at Stukeley, and can be counted as local landholders given Thorney abbey’s relative isolation. This makes Clark’s identification most likely correct. Adam of Stukeley was a parish cleric rather than a member of the chapter, and his son, Master Aristotle (d. 1199), was vicar of King’s Walden, and a bailiff and itinerant justice for Henry II.29 Nevertheless, the archdeaconry had descended through two generations of the same family. Similarly, another archdeaconry of Lincoln, that of Stow, was also held by one family (Fig. 1). Hugh’s appointment, which can be dated to the 1090s, makes it highly likely that he was the first holder of the post, and thus contemporary with Nicholas.30 Nicholas’s son Henry gives a pen portrait of the family in a letter known as De contemptu mundi written for another Lincolnshire archdeacon, Walter, archdeacon of Leicester, who was active after 1135. Significantly, Henry fails to mention that Hugh, Osbert and William are in fact related:31 The priest Hugh should not be omitted, a memorable man, the origin and, as it were the foundation-stone of the church. He was succeeded by Osbert, an extremely affable and endearing man. In their place there is now William, a young man of great natural talent.32

Undoubtedly Hugh was succeeded by his son, Osbert, and grandson, William. It was only Philip, the great-grandson (d. 1187), who failed to succeed to the archdeaconry in the mid-twelfth century, but he did maintain his position as a canon of the cathedral.33 Whilst it has proven impossible to identify any wives, we can be 26

The Thorney Liber Vitae, London, British Library, Additional MS 40,0000, fols 1–12r. Edition, Facsimile and Study, ed. L. Rollason, Woodbridge 2015, fol. 3v 22. Hereafter LVT. 27 Cecily Clark, ‘The Liber Vitae of Thorney Abbey: British Library, Additional MS 40,000, ff. 1v–12r’, in Words, Names and History, Selected Writings of Cecily Clark, ed. P. Jackson, Woodbridge 1995, 301–47 at 315. D. Greenway concurs, see Huntingdon, xxvii. 28 LVT, 8; see also R. Gameson, ‘The Thorney Liber Vitae: Planning, Production and Palaeography’, in LVT, 115–24. 29 Huntingdon, xxvii–xxviii; Twelfth-Century English Archidiaconal and Vice-Archidiaconal Acta, ed. B. R. Kemp, Publications of the Canterbury and York Society, Woodbridge 2001, no. 108; C. T. Clay, ‘Master Aristotle’, EHR lxxvi, 1961, 303–8. 30 The relationships for this family are derived from: Fasti III, 44; Registrum Antiquissimum of the Cathedral Church of Lincoln, ed. C. W. Foster and K. Major, 10 vols, Lincoln Record Society, 1931–73; II, no. 553, III, no. 921, hereafter RAL. Documents Illustrative of the Social and Economic History of the Danelaw, vol. V, ed. F. M. Stenton, London 1920, no. 218. For son Philip see Archidiaconal Acta, ed. Kemp, no. 188. 31 Huntingdon, lxxi, 584–619. D. Greenway notes, 584 n 2, that the recipient of De Contemptu Mundi died before its completion – therefore the archdeacon is unlikely to be Walter, archdeacon of Oxford, recipient of the dedication of Geoffrey of Monmouth’s Historia Regum Britanniae, who died in 1153; but rather Walter, archdeacon of Leicester who disappears from charters after 1133. 32 Huntingdon, 590–1: Nec tacendus est Hugo sacerdos uir memoria dignus, principium et quasi fundamentum ecclesie. Cui successit Osbertus, uir omnino comis et desiderabilis. In quorum loco iam Willelmus extat, iuuenis magne indolis. 33 Fasti III, 103; appendix 23; RAL IX, no. 2608.



Wives and Children of the Clergy in Normandy and England

45

Fig. 1: The hereditary archdeacons of Stow (c. 1090s–1187) Hugh, archdeacon of Stow c.1090s–? d. bef. 1188

Osbert, archdeacon of Stow c. 1133– d. c. 1154.

William, archdeacon of Stow c. 1132–1150s

Gilbert

Phillip, canon of Lincoln c. 1155–1187

confident that they must have existed. Another married archdeacon was Richard, archdeacon of Lincoln in the 1100s, whose son Gilbert became a canon of Lincoln.34 Married archdeacons can also be identified in the Lincoln diocese in the late twelfth century, for example, Baldric de Sigillo, archdeacon of Leicester c. 1162, whose son can be identified as Master Gerard.35 Elsewhere, at Exeter, two archdeacons of Totnes, Hugh d’Auco (c. 1137–1162) and Robert fitzGille (c. 1171–1186), were both married.36 Whilst most of the examples provided thus far have been identified from multiple generations of the same family in one ecclesiastical community, a few archdeacons had children who cannot be associated with their father’s cathedral chapter. For example, Richard de Bellofago, archdeacon of Suffolk and later bishop of Avranches (c. 1116–1119 × 1134/5), had three children whilst archdeacon: Alan, Roger and Robert, none of whom became canons.37 Richard’s near contemporary, Roger, archdeacon of Norfolk (c. 1121/3–1166), had a daughter as can be inferred from an undated letter of Archbishop Theobald to William, bishop of Norwich (1147–74). The letter mentions in passing that Roger’s son-in-law, R. de Pavilli, was guilty of murdering a priest.38 A mandate addressed directly to the archdeacon requires him to come before Archbishop Theobald to answer for his disobedience and part in the crime. Of greater interest for this paper is Theobald’s further admonition: ‘See to it also that you be not found guilty (which is far from our desire) of the manifold acts of incontinence with which sinister rumours have aspersed your reputation.’39 This 34

RAL III, no. 921. Fasti III, 33; Archidiaconal Acta, ed. Kemp, nos. 131, 132, 134. 36 For Hugh d’Auco see: Fasti X, 25; EEA XI, lxv, nos. 23, 32. For Robert fitzGille: Fasti X, 25; EEA XI, no. 94, 101; John of Salisbury, The Letters of John of Salisbury (1153–1161), ed. by W. J. Millor, S. J. and H. E. Butler, rev. C. N. L. Brooke, 2 vols, Oxford 1979–86, II, nos. 146, 147, 148; Frank Barlow, ‘John of Salisbury and his Brothers’, JEH 46, 1995, 95–109; D. Luscombe, ‘Salisbury, John of (late 1110s–1180)’, ODNB, (http://www.oxforddnb.com/view/article/14849 [accessed 10 December 2018]). 37 Fasti II, 67; EEA VI, nos. 17, 19, 35, 45, 50; The Chronicle of Battle Abbey, ed. and trans. Eleanor Searle, OMT, Oxford 1980, 238–48; Recueil des actes des ducs de Normandie de 911 à 1066, ed. Lucien Musset, Caen 1961, no. 229. 38 John of Salisbury, Letters, I, nos. 78, 79. 39 John of Salisbury, Letters, I, no. 79: Hoc quoque prouideas ne multiplicis incontinentiæ, qua nomen tuum sinistræ famæ rumor asperserit, quod minime optamus, reus inueniaris. 35

46

Hazel Freestone

is a rare charge made against any archdeacon, but again dates to the 1150s or later, and not to the first century of reform. Just as in Normandy, married archdeacons in England lived and worked alongside their celibate counterparts, and continued to do so up to and beyond 1150. Throughout our period bishops’ sons were appointed archdeacons. We have already noted William Bona Anima, son of Radbod, bishop of Sées who began his career as archdeacon of Rouen.40 Other examples can be found in England, such as Richard de Belmeis (Beaumais), elected bishop of London in 1108, who was the last bishop consecrated by Archbishop Anselm (d. 1109).41 He had two sons: William de Belmeis, who became archdeacon of London, and his brother, Walter de Belmeis, a canon of St Paul’s, London.42 Many of Richard’s kin were connected to St Paul’s and in Christopher Brooke’s view the Belmeis family maintained their position within the chapter until the death of Richard de Belmeis II, bishop of London in 1162, who was Richard I’s nephew and namesake.43 At Coventry & Lichfield, Bishop Robert Peche (c. 1121–1126) was almost certainly responsible for his son Richard becoming archdeacon of Coventry c. 1126.44 Richard held this post until he was elected to the same bishopric in 1160, and he himself had a son Philip, who appears to have been a layman.45 At Salisbury, another bishop consecrated by Archbishop Anselm was Roger, bishop of Salisbury (d. 1139), married to Matilda of Ramsbury, who survived him. Their son, Roger le Poer, became chancellor of King Stephen (r. 1135–53), although he was probably not archdeacon of Berkshire as Edward Kealey suggested.46 Bishop Roger was succeeded by Jocelin de Bohun (1139–84), whose son, Reginald the Lombard (or fitz Jocelin), held the office of archdeacon of Wiltshire, before becoming bishop of Bath (1174–9).47 Not only were some bishops’ sons promoted to archdeaconries or to episcopacies themselves, but at the very moment when bishops were present at councils in both England and Normandy legislating against clerical marriage, some of their sons attested charters openly identifying themselves as the sons of bishops. I provide three examples here, one from Normandy and two from England. First, Richard of Douvres, bishop of Bayeux (1107–33), was part of a formidable clerical dynasty with roots in Bayeux and connections at Worcester, York and in royal circles. He was the grandson of Muriel and Osbert, a priest of Bayeux and the son

40

The date of William Bona Anima’s appointment as archdeacon is unknown, he resigned 1057, see Spear, NP, 207. 41 Fasti I, 1; Eadmer, HN, 198. 42 Fasti I, 9, 65–6; Early Charters of the Cathedral Church of St Paul, London, ed. Marion Gibbs, London 1939, no. 218. 43 Brooke, ‘The Composition of the Chapter of St Paul’s’, 126. 44 Following the Episcopal Acta. The diocese was variously known as Lichfield, Coventry or Chester in this period. 45 Earliest attestation Regesta II, no. 544 dated 3 September 1101, and then nos. 684, 800, 1015, 1204, 1226 (charters dated 1116–20). Malmesbury, Gesta Pontificum, I, 470–1; EEA XIV, xxxviii; M. J. Franklin, ‘Peche, Richard (d.1182)’, ODNB (http://www.oxforddnb.com/view/article/21733 [accessed 10 December 2018]); for son Philip, see EEA XVI, xxvii; Staffordshire Historical Collections, Publications of the William Salt Archaeological Society, VIII, 170. 46 E. J. Kealey, Roger of Salisbury, Viceroy of England, Berkeley-London 1972, 272–3; the editors of the Fasti were unable to confirm Kealey’s suggestion, see Fasti IV, 29. 47 Materials for the History of Thomas Becket, Archbishop of Canterbury, ed. J. C. Robertson and J. B. Sheppard, 7 vols, London 1878–85, III, 524–5; B. R. Kemp, ‘Bohun, Jocelin de (1105 × 1110?–1184)’, ODNB (http://www.oxforddnb.com/view/article/50340 [accessed 10 December 2018]); EEA XVIII, lxiv–lxx, C. Duggan, ‘Reginald fitz Jocelin (c. 1140–1191)’, ODNB (http://www.oxforddnb.com/view/ article/9613 [accessed 10 December 2018]).



Wives and Children of the Clergy in Normandy and England

47

of a bishop, Samson of Worcester (1096–1112).48 His brother, Thomas II, became archbishop of York (1109–14), while his sister, Isabella, was the mistress of Robert of Gloucester, eldest illegitimate son of Henry I.49 Richard of Douvres’ son, John, appears as a witness in a charter of Henry I dated 1118, where he is identified as Iohannem Baiocensis episcopi filium.50 The charter records the settlement of a dispute between William, the previous count of Mortain and the abbots of Caen and Savigny, which had been brokered on the advice of the archbishops of Canterbury and Rouen, and witnessed by the bishops of Avranches, Coutances, Exeter and Le Mans. Other witnesses included Stephen, count of Mortain (later King Stephen); Richard, earl of Chester; Robert, the king’s son; and Ranulf the chancellor. John attests immediately after Ranulf. There can be no doubt that John was known, and accepted, as the son of the bishop of Bayeux by the upper clergy and senior laity in Normandy. The first English example of bishops’ sons appointed as archdeacon concerns Radulf, archdeacon of Durham, who issued a charter in 1146 in which he described himself as Radulfus filius Rannulfi Dunelmensis episcopi, the son of Ranulf Flambard, bishop of Durham (1099–28).51 The second case is that of William, son of an archbishop (Willelmus filius archiepiscopi), who regularly attested the acts of Hugh du Puiset, bishop of Durham (1152–95) in the last quarter of the twelfth century.52 Unfortunately, neither the content nor the context of the charters sheds light on exactly who William’s father might have been. The most likely candidate is Roger of Pont l’Évêque, archbishop of York (1154–81).53 Roger began his career as a clerk of Archbishop Theobald, before becoming archdeacon of Canterbury; he made firm friendships with Hugh du Puiset and Gilbert Foliot.54 Hugh du Puiset was suffragan to Roger, and had sons of his own residing in Durham. It is very likely that he would have been prepared to offer a home to Roger’s son. The earliest charter witnessed by William filius archiepiscopi is dated to 1172 – that is, to the latter half of Roger’s pontificate. William could therefore have been born before his father was elected archbishop of York in 1154 or shortly afterwards. The identification of William as the son of the archbishop in the documentary sources does not reflect any social shame. Yet, post-1150, it would perhaps have been unnecessarily provocative for a son to reside with or attest charters for the archbishop of York himself. Yet, it was tolerable, or even acceptable, for such a man to attest the charters of a bishop with the prestige of Hugh du Puiset.55 48

Orderic, V, 210–1; Hugh the Chanter, The History of the Church of York, 1066–1127, ed. and trans. Charles Johnson, rev. M. Brett, C. N.L Brooke, and M. Winterbottom, OMT, Oxford 1990, 46; Antiquus Cartularius Ecclesiae Baioncensis (Livre Noir), ed. V. Bourrienne, 2 vols, Rouen 1902–3, I, xlii, no. XLI; Spear, NP, 32–3. 49 Orderic VI, 428; Spear, NP, 33. 50 C. H. Haskins, Norman Institutions, Cambridge, Mass, 1918, 294; Regesta II, no. 1183. 51 Durham Episcopal Charters, ed. H. S. Offler, Gateshead 1968, no. 35b. 52 For example, Willelmo filio archiepiscopi attests EEA XXIV, nos. 5A, 8, 37, 39, 40, 41, 47, 63, 64, 70, 89, 93, 102, 121 and 139. 53 For Roger as father: The Register, or Rolls of Walter Gray, Lord Archbishop of York, ed. J. Raine, Durham 1872, xxviii; Fasti VI, 4; EEA XX, xxix, n. 7. The alternative candidate was William fitzHerbert, archbishop of York (1141–dep. 1147/48 reinstated: 1153–d.1154) suggested by G. V. Scammell, Hugh du Puiset. A Biography of the Twelfth-Century Bishop of Durham, Cambridge 1956, 237, without supporting evidence, and tentatively followed by M. G. Snape (EEA XXIV, xli) but rejected by C. Norton, St William of York, York 2006, 17 n 33. 54 F. Barlow, ‘Pont l’Évêque, Roger de (c. 1115–1181)’, ODNB, (http://www.oxforddnb.com/view/ article/23961 [accessed 10 December 2018]). 55 For Hugh’s career see Scammel, Hugh du Puiset.

48

Hazel Freestone

In presenting this evidence, I am not arguing that the ideal of clerical celibacy lacked influence. There is sufficient evidence to suggest that clergy were under pressure in Normandy not to marry and to abandon their wives, but the decision to do so was personal, driven by familial or individual choices. A key indicator of this pressure is that the numbers of identifiable fathers and sons within cathedral chapters decline by 1100.56 For example, only one married canon each can be identified at Coutances and Lisieux after 1100.57 At Rouen, there is evidence that sons who wished to have ecclesiastical careers moved away from their father’s institution. Gilbert I (Gislebert), chanter and canon of Rouen, and treasurer of Henry I, who died c. 1137, is possibly the son of Robert, archdeacon of Évreux, and his wife Adelis (see above). Gilbert I had four children. Three of his children had ecclesiastical careers, with two becoming canons, but only one maintained a connection with Rouen cathedral.58 Towards the end of the twelfth century, Benedict Groignet, a canon of Rouen active c. 1161 who died before 1214, had a son, Ralph Groignet, who is described as civis Rothomagensis, a citizen of Rouen, and therefore a layman.59 At the same time this decline in the number of identifiable priests and their sons in Norman cathedral chapters is matched by an increase in the number of uncles and nephews. As Julia Barrow has observed, this paradigm has a long history, probably first emerging during the fifth century.60 In Normandy we can provide evidence of the shift from father and son succession to uncle/nephew patronage occurring in practice thanks to Spear’s Norman Personnel. In the period 1050–1100 only four uncles/nephews can be identified as canons, a figure that rises to six in the period 1100–50, and then to twenty-three in the period 1150– 1200. Six of them appear in the decade 1150–60 alone. The writings defending clerical marriage make it apparent that the ideal of clerical celibacy was influential within Normandy in the last quarter of the eleventh century. The evidence we have suggests that it was harder for hereditary canons and their sons to remain in their ‘home community’. Yet there is little to support the idea that clerical celibacy was forcefully implemented by the bishop or the church hierarchy. Instead, the evidence suggests that it was the changing attitudes of the laity and clerical families themselves that altered familial policies and marked an accommodation with, if not acceptance of, celibacy.61 By contrast, in England, it is harder to discern a shift in the behaviour of clerical families at such an early date. Married clergy and the children of secular canons can be identified in the dioceses and/or cathedral communities of Canterbury, Chichester, Coventry & Lichfield, Durham (before 1083), Exeter, Lincoln, London, 56

Barrow, The Clergy in the Medieval World, 149. At Coutances, Tesceline had two sons, Gilbert and William, who also became canons there. This Tesceline was either Tesceline I (d. bef. 1113) or Tesceline II active c. 1132 × 1151 – see Spear, NP, 126; for sons, idem, 114, 129. At Lisieux, Roger son of Amisi [Aini] was a canon c. 1148, and had a son, Laurence, also a canon of Lisieux, c. 1230 – see Haskins, Norman Institutions, Appendix H, no. 2, 322–3; Spear, NP, 190. 58 Children: William (?of Évreux), canon of Rouen who became prior of Ste-Barbe-en-Auge, Walter, canon of St-Lô; Hugh, monk of Jumièges and Robert of Évreux (Spear, NP, 221); M. Arnoux, Des clercs au service de la Réforme: Études et documents sur les chanoines réguliers de la province de Rouen, Turnhout 2000, 11–17; Haskins, Norman Institutions, 109–10 and 109, n 111. 59 Archives Départmentales de la Seine-Maritime, Rouen, G4279, as cited by Spear, NP, 237. 60 Barrow, The Clergy of the Medieval World, 117–35, esp. 124–5 and for figures covering 904–1204 see idem, 149–50. 61 For an important study of the nobilities’ involvement in the implementation of reform that underpins this point see J. Howe, ‘The nobility’s reform of the medieval church’, American Historical Review 93, 1988, 317–39. 57



Wives and Children of the Clergy in Normandy and England

49

Rochester (before 1066), Hereford and York.62 Julia Barrow’s work on the diocese and cathedral clergy of Hereford found that, while married canons continued to be present, after the middle of the twelfth century the sons of secular canons ceased to succeed (or inherit) their father’s position.63 One named wife at Hereford is known: Margaret, daughter of Stephen the moneyer, an important burgess in Hereford, who married William Foliot, clerk to the bishop and later precentor of Hereford cathedral, after 1186. They had two children, an unnamed daughter, and a son, John Foliot, who became rector of Bockleton, Herefordshire, before himself becoming a canon of Hereford cathedral. Margaret’s identity was recorded in the obit book of Hereford cathedral.64 At York, aside from five married archdeacons, eleven clergymen were married and can be identified as canons of the chapter or attached to the bishop’s household. For example, Paulinus was active as a canon before 1142 alongside his son Thomas, active c. 1142–54.65 Another canon, Thurstan, was possibly at York around the 1160s and 1170s, and had a son, Richard.66 A third canon, Letold, had a son, John, who became archdeacon of Cleveland before 1171 and was later appointed archdeacon of Nottingham.67 Again, as at Hereford, we can identify only one wife, Pagana, wife of Tosti, a canon of York, who made an agreement with Whitby abbey at some point between the 1140s and the 1170s.68 Dating Tosti’s activity as a canon is problematic. He possibly witnessed a charter of Thomas I, archbishop of York (d. 1100), but more certainly he appeared in a charter dated c. 1125 × 1133.69 This later date seems to be more consistent with evidence for his wife, Pagana, and their son, William, also a canon of York, who appears in the record c. 1138 × 1173/74.70 Bearing these findings in mind, the most influential studies of married clergy, those of Christopher Brooke concerning St Paul’s, London, should not be seen as typical of England in the twelfth century, but rather as reflective of marked but subtle regional differences.71 Brooke found that a quarter of cathedral clergy were married at any one time, and their sons had at least some claim to their fathers’ prebends until the midtwelfth century. He also found that in London the presence of both married canons and hereditary archdeaconries and canonries more or less ceased by the mid-twelfth century.72 Yet, as I have shown, this was clearly not the case in Lincoln, where it was only archdeaconries that ceased to be ‘hereditary’. The sons of canons continued to be present in the chapter up to and beyond 1150, and elsewhere they remained visible 62

Freestone, ‘A Priest’s Wife’, Chapter Three. J. Barrow, ‘Hereford Bishops and Married Clergy, c. 1130–1240’, Historical Research 60, 1987, 1–8 at 3. 64 Fasti VIII, 14, 75 (but not identified as married); EEA VII, liv, nos. 160, 163, 172; for a summary of the family’s history and Margaret’s obit entry, see Oxford, Bodleian Library, MS Rawlinson B328 fol. 53, as cited by Barrow, ‘Hereford bishops and married clergy’, 4, n 15. 65 Fasti VI, 97–8 is wrong. Thomas is not the son of Magister Paulinus (who was the son of Ralph Nowell, Bishop of Orkney); Early Yorkshire Charters, 13 vols, ed. W. Farrer and C. T. Clay, Edinburgh 1914–65, I, no. 164 (hereafter EYC); see esp. Norton, St William of York, 233. 66 EYC I, no. 125. 67 Fasti VI, 44–5, 124; EYC I, no. 562; EYC II, no. 959, EYC III, no. 1566; The Charters of the Honour of Mowbray, ed. D. Greenway, Oxford 1972, no. 236. 68 Cartularium de Abbathiæ de Whiteby, Ordinis s. Benedicti, fundatæ anno MLXXVIII, ed. John Christopher Atkinson, Durham 1879, no. 260. 69 Fasti VI, 130; EEA V, nos 3, 74. 70 Fasti VI, 132; EYC II, no. 978. 71 Brooke, ‘The Composition of the Chapter of St Paul’s’, 111–32; idem, ‘Gregorian Reform in Action: Clerical Marriage in England, 1050–1200’, Cambridge Historical Journal 12, 1956, 1–21. 72 Brooke, ‘The Composition of the Chapter of St Paul’s’, 125; ‘Gregorian Reform in Action’, 18. 63

50

Hazel Freestone

at Hereford and York.73 Of course, it may be possible that these prebends were not strictly inherited, but were within the gift of the bishop (or the king), but it seems to me that where sons appear beside their father, or succeed to a prebend within their father’s chapter (we cannot always be certain it was the same prebend), the expectation must have been that they were acceptable members of the community and could assume their father’s place on his retirement or death. As compelling as Brooke’s findings for St Paul’s, London, are, they are not representative of the incidence of clerical marriage amongst other cathedral chapters. The variety of findings suggests that local and regional views and practices played a major part in the acceptance of married clergy. In England, therefore, if the years around 1150 appear to have been a turning point, it was by no means a definitive one. From the evidence presented thus far, it should also be apparent that a recent observation that priests’ sons disappear from the record post-1150 is erroneous.74 Married Parish Clergy and their Relations with Monastic Houses Having established that a substantial minority of archdeacons, charged with enforcing church discipline, continued to marry and have children throughout the period, I now turn to the largest group of married secular clergy: the parish clergy who served local parishes, ostensibly under the archdeacons’ oversight. One of the contemporary complaints made against an hereditary priesthood was that it inevitably lead to the alienation of church property, that is that priestly fathers passed churches and lands on to their sons, which were then lost to the church. This complaint may have reflected events elsewhere in Europe, but in England, and to a lesser extent in Normandy, despite the assertions of the ‘reformers’ and evidence of canons, it is problematic. Monastic abbots and chapters often controlled or oversaw the appointment of clerics to parish churches within their landholdings.75 In the records for a number of abbeys I found that multi-generational families ‘inherited’ churches with the consent of abbots and their communities, and for the most part maintained cordial relationships. For example, Plympton Priory, Devon, maintained its right in its dependent chapel of St Andrews from the 1030s up to 1176, even though the chapel passed from father to son, to grandson and great-grandson before finally passing to an illegitimate nephew, whose parents were the last priest’s sister and the new Norman lord of the manor (see Fig. 2). It was only on the death of the illegitimate nephew that the priory’s possession of the chapel was challenged by the local lord, a dispute that was finally settled in the priory’s favour by the papacy.76 Further examples can be found in the cartulary of St Benet of Holme, Norfolk, where we can identify three named wives and six clerical families. The three women and their husbands were Clericia, wife of Master Roger le Simple, cleric of the church of Felmingham; Agnes, wife of Philip of Fordham (chaplain and priest); and 73

Freestone, ‘The Priest’s Wife’, Chapters Two and Three; Barrow, ‘Hereford Bishops and Married Clergy’, 3. 74 Thibodeaux, The Manly Priest, 82. 75 The other issue is the complex web of relationships that individual parish priests may have had to negotiate, notably who owned the land and who controlled appointment to the benefice, making it less likely that a parish priest could alienate land even had he wanted to – see Barrow, The Clergy in the Medieval World, 323–5 for an overview of potential influences. 76 The original cartulary has not survived, but for the seventeenth-century transcript see: John Blair, Anglo-Saxon Church in Society, Oxford 2005, 519–22; also see, Allison D. Fizzard, Plympton Priory: A House of Augustinian Canons in South-Western England in the Late Middle Ages, Leiden-Boston 2008, 35–6, 102–8.



Wives and Children of the Clergy in Normandy and England

51

Fig. 2 Family of the unnamed daughter of Dunprust, hereditary priest of St Andrew’s chapel, Sutton, near Plympton (Plymouth), 1035–1176 Ælfheah

Sladda the priest

Ælfnoth d. bef. 1066

Dunprust

William Bacini d. c. 1107–1127

Robert the cleric

Unnamed daughter

Reginald Vautort

Thomas the cleric

Agnes, wife of William the clerk. All three couples and the other clerical families provide evidence that married parish clergy and their wives had cordial relations with the abbey of St Benet of Holme from early twelfth century through to the 1180s.77 Space does not permit a discussion of every case, but it is important to note in respect of the first couple that Master Roger le Simple (magister Rogerus simplex) held the church of Felmingham at the presentation of the abbot, and he was confirmed in his position as clericus of the church by William Turbe, bishop of Norwich (1147–74), in the period 1164 × 1168.78 Moreover, one of the two grants that survive in favour of Roger’s daughter, Anne, states that she holds the land in feodum et hereditatem.79 On this basis, we can infer that Clericia was Roger’s legitimate wife, not a concubine, and that Anne was their heir. Other cases from the cartulary confirm this pattern. In the case of the second couple, Philip of Fordham was not only married but also the nephew of Abbot William II (c. 1154–68) of St Benet of Holme. The confirmation of the grant of the church of St Lawrence in South Walsham to Philip by Bishop William Turbe confirmed that Philip was to hold the land, men and possessions which had been held by his father, Geoffrey, clerk of Fordham, and his unnamed grandfather.80 An earlier example of father/son succes77

EEA VI, nos. 105, 106, The Register of St Benet of Holme, 2 vols, ed. J. R. West, Norwich-Fakenham 1932, II, nos. 88, 170–1, 220, 227, 240, 264, 307. 78 EEA VI, no. 106. 79 St Benet of Holme, II, no. 220. 80 EEA VI, no. 105; St Benet of Holme, II, no. 88: … confirmavi Philippo presbitero filio Galfridi clerici de Fordham ecclesiam sancti Laurencii in Suthwalsham cum omnibus ad eam pertinentibus. Terram quoque et homines et vniuersas possessiones quas pater suus et auus suus possederunt in eadem uilla.

52

Hazel Freestone

sion occurs in a charter dated 1126/7, confirmed this time by William, archbishop of Canterbury, which recorded that the church of St Michael in Norwich was a gift to the abbey from Stigand and his son, Thurvert (also identified as Turbern, priest and dean).81 The church of St Michael and the churches of Potter Heigham and Ludham were later granted to Thurvert’s son, Thomas, c. 1153–68.82 Between 1175 and 1186, Ivo, priest of Potter Heigham, inherited lands from his father, Hugh, and the abbot of St Benet of Holme agreed that Ivo’s son Adam, or his younger brother Benedict (if Adam had died), could inherit them on their father’s death.83 In the same period, Henry the clerk, son of Henry, the dean of Fordham, was granted some tithes. On his death, the next of his brothers, who was also a clerk, was to succeed him – that is retain the right to the tithes.84 These cases taken together illustrate that at least one clerical family had close ties to the abbots of St Benet of Holme, and that even without discernible familial ties to the abbey successive generations of other clerical families maintained good relations with it. No evidence survives to suggest that the abbey lost its right to consent to each generational transfer. Moreover, episcopal approval was secured in at least two cases. None of these documents refers to the status of married priests, and all accept the children of priests in successive generations as legitimate heirs. Acceptance or toleration of married clergy changed at St Benet of Holme on receipt of a letter from Pope Lucius III (1181–5), who condemned the practice of allowing the sons of priests to inherit.85 In a subsequent papal letter Lucius III confirmed the appropriation of the income from the churches of Potter Heigham, Ludham and Ashby to the sacrist of the abbey.86 Therefore only after papal intervention do we find evidence of action against hereditary priests in the form of the appropriation of income that had formerly belonged to married clergymen. Yet, as Maroula Perisanidi has recently illustrated, in the last quarter of the twelfth century the issue of financial support for clerical families and the (potential) alienation of church property by married clergymen for the benefit of their families was an important subject of discussion amongst English commentators on Gratian’s Decretum.87 It is possible that this broad debate over the exploitation of church property may have been due to efforts to displace older customs and relationships. Such attempts tried to reach a new consensus by replacing older practices with new norms more in keeping with papal and ‘reform’ concerns. However, in the light of my findings, alienation of church property by married clergy in England may not have been as significant an issue as the modern study of canon law or commentaries implies. In Normandy it has not been possible to identify more than a few parish clergy, but these few examples reveal a range of responses that reflect similar relationships between monastery and parish priest to those found at St Benet of Holme. Two cases appear in the cartulary of the abbey of Saint-Pierre-de Préaux, both of which accept the inheritance of priests’ sons. The first charter, dated before 1120, concerns a gift from Radulf, the priest of Bourneville, of two tenants and their lands to the abbey. A second charter, dated 1125 × 1146, refers to the restitution of goods by his son, Osbern the cleric, on the orders of Waleran II, count of Meulan (d. 1166), as it 81 82 83 84 85 86 87

St Benet of Holme, II, no. 76. St Benet of Holme, II, no. 183. St Benet of Holme, II, no. 212. St Benet of Holme, II, no. 253. St Benet of Holme, II, no. 70. St Benet of Holme, II, no. 71. For a similar letter by Pope Clement III, see, no. 72 (1187–91). Perisanidi, Clerical Continence, 76–83.



Wives and Children of the Clergy in Normandy and England

53

seems he had not abided by the earlier agreement.88 The charter between the abbot and Osbern the cleric, son of Radulf the priest, demonstrates that he was considered the rightful heir of his father. The second case appears in two charters dated 1145/46 and concerns a pair of siblings, Richard Harenc and his brother William, a married priest with lands close to the abbey. The first charter concerns the sale of land to the abbey and the second is the confirmation charter of Walter, son of William the priest (Walter vilanus filius Willelmi presbiteri), confirming the sale of the vineyard. Walter was thus accepted as his father’s heir.89 Yet, not all abbots in Normandy were so accepting. For instance, a charter dated 1112 × 1147 gives details of the agreement between an unnamed abbot of Troarn who gave the care of the church of Réville, in the Cotentin, to a priest called Roger and demanded an oath that he would not exercise hereditary succession.90 The abbot clearly disapproved of Roger’s (possible) marriage, but that was insufficient to bar his appointment as priest. The evidence pertaining to parish priests in both Normandy and England suggests that the transference of churches and associated lands between generations was understood by and largely acceptable to both monasteries and parish priests, as well as, we must presume, their local bishop. In light of this evidence, the idea that the marriage of a cleric or priest automatically resulted in the alienation of church lands needs to be revisited.91 Normative and Non-Normative Sources As noted in my introduction, a dominant trend within current scholarship is to rely on normative and polemical sources. In this final section I want to illustrate how broadening the range of source material studied can enrich our understanding of clerical marriage and the influence of clerical celibacy, alongside our understanding of the church as an institution in practice. Normative sources, for example polemical tracts or canons that emerged from papal or regional church councils, are poor guides to events in individual dioceses. They are inherently biased and need to be treated with caution and considered in the context of when, where, and, if possible, why they were issued. The manner by which ideas and attitudes shifted towards clerical marriage and celibacy over time is, of course, important to trace. But without attention to the circumstances of each individual text and understanding the dissemination or corroborating evidence for its reception, normative sources are misleading guides to what happened in practice. Without such evidence it becomes too easy to assume that clerical families were affected by the moves against clerical marriage because they ought to have been. Clerical celibacy was, and remains, a defining characteristic of the Roman Catholic church as opposed to Byzantine church and the Protestant churches that emerged in the sixteenth century, and the study of the 88

Cartulaire de l’abbaye bénédictine de Saint-Pierre-de-Préaux, 1034–1227, ed. D. Rouet, Paris 2005, nos A145; A146. 89 Cartulaire … Saint-Pierre-de-Préaux, nos. A134, A135. 90 Chartrier rouge de Troarn, fol. 207v, as cited by Grégory Combalbert, ‘Gouverneur de l’église: Evêques et paroisses dans la province écclesiastique de Rouen (v. 1050–v. 1280)’, 2 vols, unpublished thesis, Université de Caen/Basse-Normandie (2006), I, 158 n 150: Juravit … quod hereditatem non clamabit in ipsa ecclesia nec aliquis per eum et quod nobiscum se tenebit. A further example occurs at Bény-Bocage, c. 1137 when the monks agree that Arnold the priest and his son after him should have the service of the church there. Chartrier rouge de Troarn, fol. 78v, as cited by G. Combalbert, ‘Gouverneur de l’église’, I, 158. 91 See Susan Wood, The Proprietary Church in the West, Oxford 2006, 873–4 for the difficulties of applying the concept of ‘ownership’ or ‘possession’ in Norman churches.

54

Hazel Freestone

subject until very recently has reflected this divide. Yet in the eleventh century celibacy was a signifier only of monastic and eremetic clergy – the secular clergy were and had been married for centuries. The starting point of any study must therefore be that clerical marriage was the norm, rather than an inherent signifier of a ‘failing’ church. Modern scholars have tended to rely on normative sources and this has led to the widespread assumption that priests’ marriages were invalidated and therefore wives were reduced in status to that of concubines by the canons of either the First Lateran Council (1123) or the Second Lateran Council (1139).92 Martin Boelens’s study, published in 1968, has been especially influential in this regard. He observed that the language of the papal canons used to describe women became pejorative over time: the wives (uxores) of earlier canons become concubines (concubine).93 He argued that the thrust of the decrees of the late eleventh and early twelfth centuries was to forbid priests’ marriages, but noted that it was unclear whether the decrees should be seen primarily as disciplinary (i.e. to enforce continence) or that clerical marriages were made invalid in principle.94 In this respect Boelens is circumspect, but other scholars have been less so.95 The evidence presented here is that papal decretals had no bearing on the legitimacy of priests’ marriages or their children in England and Normandy in the twelfth century. This does not mean that the canons had no effect elsewhere in Europe. It is crucial to understand that my analysis of non-normative sources written in Normandy and England has shown that there was no comparable shift either in language or in attitude to priests’ wives outside the reformers’ literature, as a few case studies here will show. Priests’ wives can be identified in the three surviving English Liber vitae for: the abbey of New Hyde (Winchester), written immediately before or after the Conquest of 1066; for Thorney abbey, written throughout the eleventh and early twelfth centuries; and for Durham, composed throughout our period, 1050–1150, and beyond.96 Each book contains lists of the names of men and women, lay and clerical, whom the monks wished or were asked to remember for past gifts and patronage, and possibly those who could be looked to for future support. The books do not record details of the gifts or the links with the benefactors, only names. Entries for married clergy are recorded in each text, in the same manner as married lay benefactors, as husband and wife, or in family groups, sometimes written neatly, in other cases 92

Decrees of the Ecumenical Councils, Volume One: Nicaea I to Lateran V, ed. Norman P. Tanner, Washington 1990, for First Lateran Council (1123), c. 21, p. 194; for Second Lateran Council (1139), c. 6 and 7, p. 198. For example, Brooke, ‘Gregorian Reform in Action’, 18; Barstow, Married Priests and the Reforming Papacy, 102–4; J. Brundage, Law, Sex & Christian Society, Chicago 1987, 216–17, 220; Ruth Mazo Karras, Unmarriages: Women, Men, and Sexual Unions in the Middle Ages, Philadelphia 2012, 118–9; Thomas, The Secular Clergy in England, 180–2; Jennifer Thibodeaux, ‘The Defence of Clerical Marriage: Religious Identity and Masculinity in the Writings of Anglo-Norman Clerics’, in Religious Men and Masculine Identity, ed. P. H. Cullum and K. J. Lewis, Woodbridge 2013, 46–63 at 48. 93 M. Boelens, Die Klerikerehe in der Gesetzgebung der Kirche, Paderborn 1968, 117. 94 Boelens, Die Klerikerehe, 167–8. 95 See n. 92. 96 The Liber vitae of the New Minster and Hyde Abbey, Winchester: British Library Stowe 944: Together with leaves from British Library Cotton Vespasian A. VIII and British Library Cotton Titus D. XXVII, ed. S. Keynes, Copenhagen 1996; Cecily Clark, ‘The Liber vitae of Thorney Abbey: British Library Additional MS 40,000, ff. 1v–12r’, in Words, Names and History, Selected Papers of Cecily Clark, ed. P. Jackson, Cambridge 1995, 301–47; The Thorney Liber Vitae, London, British Library, Additional MS 40,000, fols 1–12r. Edition, Facsimile and Study, ed. L. Rollason, Woodbridge 2015; The Durham Liber Vitae: London, British Library, MS Cotton Domitian A.VII. Edition and Digital Facsimile with introduction, codicological, prosopographical and linguistic commentary, and indexes, including the Biographical Register of Durham Cathedral Priory (1083–1539), 3 vols, ed. D. and L. Rollason et al., London 2007, see esp. vol. III.



Wives and Children of the Clergy in Normandy and England

55

crammed amongst many entries seemingly without order. Thus, in the Liber vitae of New Hyde abbey, c. 1090, two priests’ wives can be identified: Emma, coniunx of Rodbertus the priest, along with their daughter Adelina and son Odinus; and Ælfgifu, wife of Ealdred the priest.97 Above I mentioned the uxor archidiaconi in the Liber vitae of Thorney abbey. In addition to her, in entries dated before 1100, we find four other priests’ wives: Wulfled, uxor of Sumerlede the priest; Isþara, uxor of Gunni the priest, with one child, Merɜet; Speterun, uxor of Inɜolf the priest; and Godgifu, uxor of Wul[f]wi[g] the priest. At the slightly later date of c. 1100–45 we find Leofsige and his son Wluricus as successive priests of Tydd. Tom Licence has tentatively identified them as the husband and son of the so-called notorious concubine (nefandissima concubina) in the life of Godric of Throckenholt. This vita is one of the few saints’ lives that is explicitly hostile to clerical wives, but, since it was written most likely in the 1170s, it is a relatively late text in respect to other sources discussed here.98 In the Durham Liber vitae, a mother and daughter-in-law are listed amongst the confraternity. The mother, Leofrun, was married to Algod, a canon at St Paul’s in London before 1086. She had two sons who also became canons of St Paul’s, Edmund and Ralph fitzAlgod.99 Leofrun’s daughter-in-law, Mahald, is described as the socia of Ralph fitzAlgod, a canon from 1086/7 until c. 1132, as well as an alderman of London.100 Mahald and Leofrun are named as part of a family commemoration in the Durham Liber vitae, which is itself unusual since most commemorations in this Liber vitae refer to men alone. Mahald’s inclusion alongside her husband, her sons, brother and her mother-in-law is a clear indication of her status within the family, and supports my view that she was a legitimate partner, almost certainly a wife. A later entry lists Aliza, wife (uxor) of Walter the chaplain.101 Walter attested a number of charters for Bishop Hugh du Puiset between the 1150s and 1190s, providing a clear record of a married chaplain active over a number of decades in the service of a bishop well into the late twelfth century.102 From the eleventh century through to the late twelfth century, monks of these abbeys evidently had no difficulty in accepting gifts from married clergy or conferring on them the spiritual benefit of being recorded in Liber vitae. The wives of parish clergy recorded in saints’ lives are admittedly few in number, but the examples that we find illustrate that non-normative sources reflect attitudes that were much more supportive of married clergy than modern scholarship has led us to expect. For example, Popelina, wife (socia) of Anskar or Anger, canon of St Paul’s, before 1104, was mother of two famous sons. The first, Audoen, succeeded his father c. 1104 before eventually becoming bishop of Évreux in 1114. The second, Thurstan, was appointed archbishop of York in the same year, a post he held until his death in 1135. The Vita Thurstini refers to Popelina as ‘a worthy woman’ (digna femina), clearly indicating that BL, Stowe MS 944 fol. 28v: ‘rodbertu[s] presbiter & eius co[ni]u[n]x /nomine Amma & filia ei[us] Adelina/ & Odinus filius eius / Ealdredd p[res]b[ite]r* et c[on]iunx/Ælfgifu. NB: *the abbreviation is an addition squeezed into the line above. 98 Tom Licence, ed., ‘The Life and Miracles of Godric of Throckenholt’, Analecta Bollandiana 124, 2006, 15–43; Rachel Koopmans, Wonderful To Relate: Miracle Stories and Miracle Collecting in High Medieval England, Philadelphia 2011, 134 suggests a date post–1172. 99 LVD III, 560; EEA XV, no. 7; Brooke, ‘The Composition of the Chapter of St Paul’s’, 123. 100 LVD, III, 560; Fasti I, 74. 101 LVD III, 119–20; LVD III, 107: Hugh, clericus of North Allerton and his sons, Richard and Thomas, c. 1154; Hugh, clericus of Pickering and his sons, Simon, Robert, Rannulf/Ralph, c. 1170. 102 LVD III, 120; Walter as witness: EEA XXIV, nos. 126, 128, 161. 97

56

Hazel Freestone

her marriage to a cleric had not diminished her reputation.103 Other saints’ lives portray priests’ wives as integral to their communities. A woman called Leviva played a small but pivotal role in The Life and Passion of William of Norwich written by Thomas of Monmouth, a monk at Norwich cathedral priory, between 1150 and 1174.104 The Life recounts how in 1144 the body of William was discovered in Thorpe Wood on the outskirts of the city of Norwich, and the boy’s family believed that he had been murdered by Jews. Leviva was William’s maternal aunt, married to Godwin Sturt, a priest. In the Life she has a prophetic dream that she recounts to her husband in private on hearing the news of William’s murder, before swooning – only to recover and mourn her nephew.105 The Life’s author, Thomas of Monmouth, was an educated monk and could not have been ignorant of the debate over clerical marriage, yet he has no difficulty in placing a prophetic vision in the mouth of a clergyman’s wife. It is hard to avoid the conclusion that priests’ wives must have been common-place enough for Thomas of Monmouth to identify Leviva as such and to assume that his audience would accept her as a credible witness. Her husband’s actions within the story are equally dramatic and public. It was Godwin who denounced the murder of his nephew in the presence of Everard of Calne, bishop of Norwich (1121–46), and offered a trial by ordeal to prove the Jews guilty of the murder.106 In Godwin’s testimony, he refers to his wife as witness to his accusation. The bishop and the chapter of Norwich must have been aware of a married priest living with his wife in the diocese. If the repeated canons against clerical marriage that emanated from the papacy in the first half of the twelfth century – especially those of 1123 and 1139, which are assumed to have invalidated clerical marriage, as well as canons announced at the councils in England of 1102, 1108 and 1129 – had been effective, the appearance of a married priest and his wife active in the 1140s should have undermined Thomas of Monmouth’s case, not supported it. As it is, this evidence shows that in practice married clergy and their wives in Norwich in the mid-twelfth century had not been affected by canons designed to impose clerical celibacy. And Leviva is not alone in her prominence. A second priest’s wife acts as a silent witness and subject of one of the stories designed to illustrate Wulfric’s holiness in John of Forde’s Life of Wulfric of Haselbury.107 Godida was the wife of Brihtric, parish priest of Haselbury, and they had a son, Osbern, who inherited his father’s position. John of Forde, monk and later abbot of the Cistercian abbey of Forde, described how Wulfric had been given a piece of linen cloth to have an alb made.108 While sewing the alb for him, Godida had made a ‘careless mistake’. Wulfric had seen her making this mistake in his mind’s eye and sent his servant, a young lad, with a message asking her to pay 103 Historians

of the Church of York, 3 vols, ed. J. Raine, London 1879–94, II, 261; John of Worcester, III, 134; Fasti, I, 36, 43; Fasti VI, 1; David S. Spear, ‘Une famille écclesiastique anglo-normande: L’évêque Ouen d’Évreux et l’archévêque Thurstan d’York’, Études Normandes 35, 1986, 21–7. 104 Thomas of Monmouth, The Life and Passion of William of Norwich, ed. and trans. M. Rubin, London 2014; for most recent discussion see: E. M. Rose, The Murder of William of Norwich. The Origins of the Blood Libel in Medieval Europe, Oxford 2015. 105 Thomas of Monmouth, Life and Passion, 29. 106 Thomas of Monmouth, Life and Passion, 31–2. 107 John of Ford, Wulfric of Haselbury, ed. M. Bell, London 1933; John of Forde, The Life of Wulfric of Haselbury, Anchorite, ed. and trans P. Matarasso, Collegeville, MN 2011. 108 An alb (alba) or talaris/tunica talaris is a long white linen/cotton long-sleeved tunic that was the foundation garment worn during the Mass or other liturgical ceremonies, see Maureen C. Miller, Clothing the Clergy. Virtue and Power in Medieval Europe, c. 800–1200, Ithaca-London 2014, 247.



Wives and Children of the Clergy in Normandy and England

57

more attention to what she was doing.109 This story may be brief, but it is significant, for John of Forde, writing in the 1160s, is describing a priest’s wife carrying out a ‘usual’ task. The implications are two-fold. First, the author accepts Godida’s status as a priest’s wife as normal and virtuous. Second, cloth-making and sewing were female tasks, and it was part of a woman’s customary role to sew for the church. In the eyes of most ecclesiastical authors, only a virtuous woman could do so.110 Maureen Miller has argued that the piety of the activity of sewing could be overstated, and that it was largely a construct of a male clerical author – a portrayal which did not necessarily accurately reflect a woman’s own experience.111 Nevertheless, in this instance John of Forde can only have included the story to illustrate Wulfric’s spiritual gifts. It would undermine his purpose if Godida, the priest’s wife, could not be considered virtuous. As it is, John of Forde’s portraits of Godida’s husband and her son Osbern reflect an honourable, conscientious and pious family serving their community and their church. Conclusion The evidence for priests’ wives and their children throughout the eleventh and twelfth centuries has proven to be both fruitful and unsettling. Fruitful in that non-normative sources reveal far more women and children than I, or indeed others, could have anticipated. Unsettling because much of this seems to contradict the assumptions of current scholarship on clerical marriage with regard to the presence of priests’ wives and children in society. Even from this limited survey of the evidence for priests, their wives and their sons in non-normative sources, it is obvious that normative texts are only of partial use as markers for social change, however much they reflect the intentions and ambitions of papal reformers. When we explore sources beyond the ideological arguments embedded in normative texts, a far more family-oriented church emerges. Where change was driven by local and regional concerns, shaped by its own history and internal structures, and only intermittently affected by the political ambitions of rulers or decisions emanating from the papacy. In my view, in light of the continued presence of married clergy at all levels of the church in England and Normandy in the period 1050–1150, it is necessary to reconsider the issues of clerical celibacy and the dissolution of clerical marriage. Such an effort should now turn from ideas to people.

109 Life

of Wulfric of Haselbury, trans. Matarasso, 193; Wulfric of Haselbury, ed. Bell, 109: Data est linea tela viro Dei ad albam faciendam. Cumque ex more incisam Godida mater Osberti presbyteri consueret, accidit ut consuturam unam invertens per incuriam erraret. Accersito ergo vir Dei ministro suo : ‘Vade’, ait, ‘et dic Godidae quia non satis attendit ad opus manuum suarum.’ 110 For the importance of women in textile production see David Herlihy, Opera Muliebra. Women and Work in Medieval Europe, Philadelphia 1990, esp. 1–24; Miller, Clothing the Clergy, 149–53. 111 Miller, Clothing the Clergy, 151–3, 166–7.

58

Hazel Freestone Appendix: Summary Tables of Married Clergy by Diocese and Grade

Table 1: Number of married clergy by diocese and grade in Normandy Diocese Avranches Bayeux Coutances Évreux Lisieux Rouen Sées Unidentified Total

Total 2 26 10 9 9 14 5 1 76

Bishops 1 4 3 2 1 3 1 – 15

Archdeacons – 2 4 1 2 4 – – 13

Secular Canons – 9 2 2 2 5 3 – 23

Parish Clergy 1 11 1 4 4 2 1 1 25

Table 2: Number of married clergy by diocese and grade in England Diocese Bath & Wells Canterbury Chichester Coventry & Lichfield Durham Ely Exeter Hereford Lincoln London Norwich Rochester Salisbury Winchester Worcester York Unidentified Total

Total 1 8 4 16

Bishops – – – 4

Archdeacons – 1 – 1

Secular Canons – 6 2 3

Parish Clergy 1 1 2 8

10 3 12 19 85 49 29 4 10 17 5 43 1 316

3 1 – – 1 2 2 – 2 1 1 1 – 18

– 1 2 – 8 6 3 – 1 3 – 5 – 31

4 – 2 4 10 24 2 1 3 2 – 11 – 74

3 1 8 15 66 17 22 3 4 11 4 26 1 193

ANTHROPOLOGY, FEUD AND DE OBSESSIONE DUNELMI* Tom Lambert The corpus of evidence for Anglo-Saxon history contains only one narrative of an extended real-life feud. In its most detailed form it appears in a short tract known to historians as De obsessione Dunelmi.1 The story begins with events that may have taken place in 1006. Uhtred of the House of Bamburgh, the ruling family of the English region north of the River Tees, won a great victory against a Scottish army which had besieged Durham. After mounting the heads of the slain Scots on stakes around Durham’s walls, he was summoned south to meet King Æthelred II and rewarded with an earldom that encompassed not just his family’s northern heartlands but also what had, until the expulsion and killing of Erik Bloodaxe in 954, been the Scandinavian kingdom of York. Driven, perhaps, by the need to secure local allies in order to rule this region, Uhtred dismissed his first wife – Ecgfritha, daughter of the bishop of Durham – and instead married the daughter of a prominent Yorkshire nobleman named Styr. Styr, however, had a condition for the match: Uhtred was to kill his enemy, Thurbrand the Hold. But Thurbrand struck first. As England succumbed to Scandinavian conquest in the mid-1010s, Uhtred (as De obsessione tells it) remained staunchly loyal to King Æthelred, whose daughter had by this point become his third wife. However, in 1016, once Æthelred had died and Cnut had seized the throne, he had little option but to submit. Having secured safe-conduct from Cnut he travelled to the king’s hall and entered his presence, whereupon, apparently as a result of Thurbrand’s unspecified treachery, the king’s soldiers sprang from their hiding place in a curtained-off section of the hall and slaughtered Uhtred and the forty men who accompanied him. The next act in the drama is disappointingly vague: after a brief interlude Uhtred’s earldom, now shorn of Yorkshire, passed to his son Ealdred, who killed Thurbrand in circumstances which are left obscure. Following this De obsessione provides more detail. There was a prolonged period of hostility in which Earl Ealdred and Carl, Thurband’s son, made efforts to ambush and kill one another. Eventually, in 1038, they were persuaded by friends to make peace and each made amends to the

*

This essay was written as a tribute to my friend and mentor, the anthropologist Paul Dresch, in the wake (and hopefully also in something of the spirit) of the workshop in his honour that led to The Scandal of Continuity in Middle East Anthropology: Form, Duration, Difference, ed. Judith Scheele and Andrew Shryock, Bloomington 2019. 1 De obsessione Dunelmi, in Symeonis Monachi Opera Omnia, ed. Thomas Arnold, 2 vols, RS 75, 1882–5, I, 215–20 esp. at §2, 5, 7 (the text is cited here by paragraph number). Christopher J. Morris, Marriage and Murder in Eleventh-Century Northumbria, York 1992, provides a translation (1–5) and a detailed historical commentary (19–25) from which much of the contextual information in the following summary derives. The text’s misleading title (‘On the siege of Durham’) derives from the rubric in its late twelfth-century manuscript. The core of the feud story appears in broadly similar form in other Durham texts loosely associated with Symeon of Durham (fl. c. 1090–c. 1128). See Historia Regum, in Opera Omnia, ed. Arnold, II, 3–283 at 197–8, 200; De Primo Saxonum Adventu, in Opera Omnia, ed. Arnold, II, 365–84 at 383.

60

Tom Lambert

other; they went so far as to become sworn brothers and to begin a joint pilgrimage to Rome. However, forced to abandon their journey by storms at sea, they returned to Yorkshire, where Carl received Ealdred in his home with lavish hospitality and undertook to escort him on his onward journey, presumably northwards back to his earldom. It was on this journey, in some woodland, that Carl struck, treacherously killing Ealdred. The final act in the narrative comes thirty-five years later, in the winter of 1073–4. Earl Waltheof, the son of one of Ealdred’s five daughters, sent a large band of young men to Settrington in Yorkshire, where the unsuspecting sons and grandsons of Carl were feasting together. They carried out a bloody massacre, killing all save Carl’s son Cnut, whom they spared because of his ‘innate goodness’.2 The story ends with their return home, loaded with plunder. De obsessione is not a straightforward text. The feud story is not in fact its main focus; its anonymous author, a member of Durham’s cathedral community, appears to have been tasked with composing a memorandum setting out the history of six estates to which his church had a claim. They had, it is explained, originally been used as a dowry for Ecgfritha, Uhtred’s first wife and the daughter of Bishop Aldhun of Durham (d. 1018), on the condition that they be held only for as long as her husband remained married to her, and would to revert to the church (it is implied but not explicitly stated) either upon her death or in the case of divorce. In the event, Ecgfritha was renounced not just by Uhtred but also by her second husband, and she took religious vows. Durham’s claim was based on the assertion that the six estates ought to have been returned at this point. In fact, Ecgfritha had children with both men, and across the eleventh century the estates passed acrimoniously between her female descendants, possession alternating between the progeny of her two marriages. The details of this form what the text treats as its main story. It begins by detailing Ecgfritha’s marriages and ends with an account of how the various estates (which were separated into two groups early on) came to be held by her daughter, one of her granddaughters, and eventually a great-granddaughter.3 Mastering the necessary genealogical detail (Fig. 1) and writing this up coherently was a task complex enough, but from the text’s outset it is apparent that its author had no intention of sticking tightly to his remit. The story of Ecgfritha’s marriage to Earl Uhtred, with which it opens, is placed in the context of Uhtred’s heroic military and political exploits: he defeats the Scots at Durham, is rewarded with the extension of his earldom, and divorces Ecgfritha so as to make the politically advantageous marriage to Styr that led to the feud.4 Next Uhtred disappears from view temporarily so that Ecgfritha’s subsequent marriage, its breakup, and the consequences for the six estates can be discussed,5 but he then returns to the fore in an even more extensive digression. This relates Uhtred’s heroic defiance of King Cnut, the tragedy of his betrayal and death, and the bloody series of revenge killings it prompted.6 None of this has anything to do with the six estates, as the author acknowledges, almost apologetically, when he returns to the subject at the end of his text. Bernard Meehan speculated that this rich seam of extraneous material ‘may represent the recension of some form of oral saga’ focused on Uhtred, and to my

2 3 4 5 6

De obsessione, §7 (cui pro insita illi bonitate vitam permiserunt). De obsessione, §1–3, 8. De obsessione, §1–2. De obsessione, §3. De obsessione, §4–7.

Arkil

=

=

Ecgfritha

=

son

Eda

Ælfsige of Tees

unknown son

=

Ealdgyth

Granted Earldom 1072 Killed sons and grandsons of Carl 1073-4 Executed 1076

Earl Osulf

=

Judith

Niece of King William I

=

• •

Ælfgifu

Eadulf Rus

Uhtred

Gospatric

Key

Maldred

Gospatric

Feud story protagonists in bold Possessors of disputed estates in italics

Waltheof

Dolfin

Earl Gospatric

=

Bought earldom 1067 Deposed 1072

Ealdgyth

Daughter of King Æthelred II Third wife of Uhtred

Killed Bishop (also Earl) Walcher 1080

Killed Earl Copsi March 1067 Killed by bandit later same year

Killed Earl Eadulf 1041 Seized three estates from Sigrid and three from Durham

=

Earl Waltheof

Ælfflæd

Sige

Daughter of Styr Second wife of Uhtred

Earl Eadulf

=

Second husband of Sigrid (left) Killed by Siward in 1041

Earl Siward

Earl Uhtred

Marries Sige on condition he kills Thurbrand Killed by Thurbrand 1016

Earl Ealdred

=

Killed Thurbrand c. 1020s Killed by Carl 1038

Ælfflæd

Ælfflæd

Son of Uhtred (right) Second husband of Sigrid Killed by Siward in 1041

Held two estates 1109x14 ‘Recently’ had to fight Gospatric unknown

Waltheof

Ecgfritha

Daughter of Bishop Ealdhun First wife of Uhtred Held all six estates

Earl Eadulf

Æthelthryth

=

Seized two estates sometime after 1070

Orm

son of Gamel

Sigrid

=

Held three estates until Siward seized them, then all six after Siward’s death

‘Recently’ had to fight Waltheof son of Ælfsige

Gospatric

Gospatric

Son of Ecgfrith Third husband of Sigrid Seized all six estates in 1055 Returned three to Durham on Sigrid’s death Fled England 1070

Arkil

Son of Fridegist First husband of Sigrid

Second husband of Ecgfritha

Kilvert

Relevant descendants of Uhtred and Ecgfritha

62

Tom Lambert

mind this is indeed the most convincing explanation for its presence.7 That is, when investigating the history of the six estates, the author found their origins bound up with a story told about Uhtred, and about the feud in which his descendants became embroiled. And once he had related the start of that story, the part which intersected with the descent of the six estates, he felt compelled to bring it to its conclusion. If this is accepted, we can infer that a version of this feud narrative was circulating in Northumbria, most plausibly in an oral form, when our author undertook his legalhistorical research. The issue of when exactly that was need not detain us here. An appendix to this essay provides reasons for placing its composition in the last two decades of the eleventh century; others have suggested 1074–5 or the first two decades of the twelfth century. This essay is not fundamentally about De obsessione itself, but rather considers a much broader set of historiographical problems that the text serves to highlight. De obsessione preserves what is by some margin the most detailed feud narrative to survive from the period, and though the text’s oddities present interpretative challenges we nevertheless have a fairly good idea of this narrative’s provenance. One might reasonably expect it to be the core source around which the history of feuding in eleventh-century England has been constructed. But it is not, and this puzzling state of affairs makes De obsessione a good historiographical case study: it calls for an explanation. The most productive approaches to De obsessione have tended to mine it for information on northern politics and Anglo-Scottish relations; these topics are obscure enough in the surviving evidence for De obsessione to be a valuable source, despite the manifold interpretative problems it poses. The feud story is often irrelevant to such discussions, especially where the primary focus is Scotland, as here the crucial issues are details incidental to the killings, such as the dating of the eponymous siege of Durham and the reliability of the text’s brief account of the ceding of Lothian to the Scots.8 However, the most ambitious attempt at political contextualization to date, that of William Kapelle, tackles the story head-on, recasting it in terms of a power struggle between the ethnically English House of Bamburgh – sandwiched uncomfortably between the Scottish kingdom and the Scandinavians of Yorkshire, and thus natural allies for England’s ruling Cerdicing family – and a Scandinavian faction in Yorkshire, hostile to the expansionist Cerdicings and favourable to Cnut and his successors.9 Kapelle’s reconstruction is debatable in its specifics, but his broader insight that these events must have been entangled with contemporary power politics can scarcely be contested, and the probability that our feud story glosses over much of the political context surrounding these killings provides part of the explanation for why historians have struggled to make much of positive use of De obsessione for other purposes. Kapelle’s assessment of the broader significance of his reading of the De obsessione narrative was uncompromising: with the events it described revealed as matters of inter-ethnic Realpolitik, it was no longer possible to interpret them otherwise. ‘There was no northern blood feud.’10 Such incautious phrasing has made him an easy target for later writers, who have chided him for drawing an 7

Bernard Meehan, ‘The Siege of Durham, the Battle of Carham and the Cession of Lothian’, Scottish Historical Review 55, 1976, 1–19 at 10. 8 De obsessione, §6. 9 William E. Kapelle, The Norman Conquest of the North: The Region and Its Transformation, 1000– 1135, London 1979, at 17–26. 10 Kapelle, Norman Conquest, 24.



Anthropology, Feud and De obsessione Dunelmi

63

anachronistically sharp distinction between political and social spheres.11 The criticism is certainly justified on one level. It is surely the case that if we are to make use of the De obsessione feud story for purposes other than the reconstruction of political narrative, we need to accept that there is nothing intrinsically implausible about the notion that to contemporary eyes a series of killings could simultaneously be both highly politicized and part of an inter-familial feud. Yet it is debatable how far this insight has got us in practice. In one respect the results have been impressive. Christopher Morris and Richard Fletcher – the latter expansively, for a mass-market audience – have produced excellent elucidations of the text itself, using their broader understandings of tenth- and eleventh-century society, culture and economics (as well as politics) to inform sensitive readings of De obsessione. But attempts to do the reverse, to make De obsessione bear meaningfully on wider interpretative issues, are notably thin on the ground. Indeed, the text – our only narrative of an extended real-life feud – is surprisingly peripheral in literature on feuding. Paul Hyams’s Rancor and Reconciliation, a book inspired in part by irritation at ‘the way Anglo-Saxon historians seemed to pay lip-service to feud without taking it seriously as an integral part of the political culture’, discusses it for only two pages.12 And though my own book – inspired by Hyams’s work from a very early stage and driven by a similar sense of frustration – places feuding at the centre of its account of the Anglo-Saxon legal order, it makes no reference at all to the De obsessione story.13 This reticence reflects broader anxieties about De obsessione’s representativeness. Can we reasonably extrapolate from the behaviour of such powerful families? Even if we could, might it not be that northern society was rather different from that of England south of the Humber? But Kapelle’s basic objection can scarcely be ignored: De obsessione may well provide a misleadingly partial account of the motivations of those involved, dressing opportunistic political assassinations up as the vindication of family honour. For all that his trenchant language has rendered him vulnerable to criticism, his underlying logic has largely held firm. The likelihood that these killings were politically charged has indeed made it impossible to understand them as a ‘straightforward’ feud, so the history of Anglo-Saxon feuding has largely been written from other sources.14 Indeed, somewhat ironically, the De obsessione feud narrative is rather more conspicuous in arguments that seek to minimize the prevalence and relevance of feuding in Anglo-Saxon culture. The manifold grounds on which the story’s representativeness can be doubted make it a good fit in such contexts. John Niles makes much both of political motivations and of the potential Scandinavian ancestry of those involved on both sides.15 Peter Sawyer follows Kapelle in taking the likelihood of political motivation as they key disqualifying factor – ‘this Northumbrian “feud” appears to have been, in fact, a political dispute’16 – while Patrick Wormald, 11

Morris, ‘Marriage and Murder’, 22; Richard Fletcher, Bloodfeud: Murder and Revenge in AngloSaxon England, London 2002, 51–2; Paul R. Hyams, Rancor and Reconciliation in Medieval England, Ithaca 2003, 279; Hudson, ‘Feud’, 31. 12 Hyams, Rancor, 75–7. 13 Tom Lambert, Law and Order in Anglo-Saxon England, Oxford 2017. 14 Kapelle, Norman Conquest, 19. 15 John D. Niles, ‘The Myth of Feud in Anglo-Saxon England’, Journal of English and Germanic Philology 114, 2015, 163–200 at 180–2. It is unclear on what basis Niles attributes Scandinavian origin to the name Uhtred. 16 Peter Sawyer, ‘The Bloodfeud in Fact and Fiction’, in Tradition og Historieskrivining: Kilderne til Nordens Ældste Historie, ed. Kirsten Hastrrup and Preben Meulengracht Sørensen, Acta Jutlandica 63, Aarhus 1987, 27–38 at 31.

64

Tom Lambert

even more briskly, writes it off as ‘one of many indications that conditions north of the Humber were not at all like those in the realm’s southern heartland’.17 The tale receives top billing in John Hudson’s reassessment of the prevalence of feuding in England, which opens with a sensitive discussion of the feud story as a narrative construct. Ultimately, however, this gives way to its interpretation as an artefact of an unusually brutal political culture specific to the north of England: The violent dispute narrated by De obsessione may be the product of particular circumstances rather than a rare survival of a more general English custom. At the highest level of Northumbrian society killing was certainly more frequent than elsewhere in England … the death of even post-Conquest rulers of Northumbria took place in a context of insult, killing, negotiation, and vengeance.18

In short, while De obsessione may at first sight appear to be an unusually rich and promising source for the study of late Anglo-Saxon feuding culture, in fact very little positive use has been made of it for that purpose. Indeed, it has mainly functioned as a weak point upon which those sceptical about the existence of such a feuding culture can seize, because it is readily dismissible as unrepresentatively politicized, Northumbrian, or Scandinavian. This essay’s main contention is that our difficulties in making De obsessione bear usefully on feuding in eleventh-century England are best understood not as the unavoidable consequences of a flawed source but as symptoms of deeper problems with the interpretative assumptions and analytical priorities we traditionally bring to the subject. Its hope is that a critical review of the historiography of feuding will both help us understand those problems and highlight some potentially more productive lines of analysis. The discussion will return to De obsessione at its conclusion, to illustrate how the essay’s methodological suggestions might play out in concrete terms, but what follows is primarily concerned with the interpretative lens through which historians have viewed feuding in this period. It is an attempt to understand why the literature tends to frame feud in the ways it does, and to draw attention to the more problematic aspects of those framings. Medieval feud historiography’s most famous feature is its long-standing tradition of drawing on anthropology – dating back, according to the sub-discipline’s semi-legendary origin story, to a friendship between J. M. Wallace-Hadrill and Max Gluckman19 – so such an endeavour inevitably raises questions about the nature and significance of interdisciplinary borrowings. These questions form the focal points around which the discussion here is organized; what follows thus amounts to a case study in the historiographical reception of anthropology and might perhaps have some wider value as such. In the 1950s the unique intellectual contribution made by the anthropology of feud seemed clear-cut, at least to Max Gluckman: All over the world there are societies which have no governmental institutions. … Yet these societies have such well-established and well-known codes of morals and law, of convention and ritual, that even though they have no written histories, we may reasonPatrick Wormald, ‘Giving God and King Their Due: Conflict and Its Regulation in the Early English State’, in idem, Legal Culture in the Early Medieval West: Law as Text, Image and Experience, London 1999, 333–57 at 340. 18 John G. H. Hudson, ‘Feud, Vengeance and Violence in England from the Tenth to the Twelfth Centuries’ in Feud, Violence and Practice: Essays in Medieval Studies in Honor of Stephen D. White, ed. Belle S. Tuten and Tracey L. Billado, Farnham 2010, 29–53 at 49. 19 Ian Wood, ‘“The Bloodfeud of the Franks”: A Historiographical Legend’, Early Medieval Europe 14, 2006, 489–504. 17



Anthropology, Feud and De obsessione Dunelmi

65

ably assume that they have persisted for many generations. They clearly do not live in unceasing fear of breaking up in lawlessness. … Therefore when anthropologists came to study these societies, they were immediately confronted with the problem of where social order and cohesion lay. Studies of societies in which private vengeance and self-help are the main overt sanctions against injury by others, and where this exercise of self-help is likely to lead to the waging of feuds, led to one of the most significant contributions which social anthropological research has made to our understanding of human relations. Anthropologists have been able to see the situations which give rise to internecine fights, and, more importantly, to examine the mechanisms which lead to settlements. The critical result of their analysis is that these societies are organized into a series of groups and relationships, so that people who are friends on one basis are enemies on another. Herein lies social cohesion, rooted in the conflicts between men’s different allegiances.20

Gluckman’s idea was that anthropological studies of feuding had managed to solve the conundrum posed by the existence of stateless societies: ‘the problem of where social order and cohesion lay’ if it was not imposed by centralized authority. The pressure to make and maintain peace exerted by the ‘cross-cutting ties’ that linked feuding groups was the missing piece of the puzzle – the hidden force which allowed these societies to cohere. No modern anthropologist would accept the underlying reasoning here. The most fundamental problem is that social cohesion in the absence of the state only constitutes a puzzle if one presupposes that humans are inherently asocial beings, whose natural propensity to violent disorder needs to be counteracted by some powerful force.21 This, the logic of Hobbes’s Leviathan, is deeply and dangerously embedded in contemporary common sense, and it remains a feature of much that has been written about feuding and early medieval law. It is worth dwelling on briefly. Gluckman was guilty of what Marilyn Strathern scathingly critiqued as ‘discovering social control’: Social control is frequently discovered in other cultures as the specific counterpart to regulatory mechanisms in western society. … [E]ven where we cannot find counterparts in other societies to our formal judicial structures we nevertheless like to discover regulatory mechanisms, in however a diffuse or generalised form. They may be uncovered as ‘customary law’. Yet it is not always appreciated, perhaps, the extent to which such discoveries simultaneously underwrite western law as a specialised and sophisticated mechanism whose primitive counterparts are uncoverable everywhere, and also underwrite the legitimacy of such law as dealing with fundamental human processes. Societies everywhere are seen to have similar regulatory problems. [… Yet] the very idea that law is a societal mechanism which meets basic human needs for regulation is part of the ideology of law. This ideology in turn draws on a model of social life and human behaviour which belongs very much to the industrial west, as well as to state systems of government. … As expanded by anthropologists, this model classifies societies as everywhere attending to the same ends, including the management of disruptive behaviour manifested at the level of individual conflict.22

Though medievalists are mostly now conscious that Gluckman bears the taint of functionalism, and therefore tend carefully to distance the notion of ‘peace in the feud’ from their own analyses, historians remain prone to discovering social control. 20

Max Gluckman, ‘The Peace in the Feud’, P & P 8, 1955, 1–14 at 1–2. Paul Dresch, ‘Legalism, Anthropology, and History: A View from the Part of Anthropology’, in Legalism: Anthropology and History, ed. Paul Dresch and Hannah Skoda, Oxford 2012, 1–37 at 9–10. 22 Marilyn Strathern, ‘Discovering “Social Control”’, Journal of Law and Society 12, 1985, 111–34 at 112–13 (emphasis in original). 21

66

Tom Lambert

Early medievalists often conceive of ‘horizontal’ feuding practices as direct substitutes for ‘vertical’ state punishment, framing the two as alternative mechanisms by which regulatory problems common to all human societies might be solved.23 As Daniela Fruscione explains, in a direct exposition of a theoretical position more usually discernible as an unexamined assumption, ‘Anglo-Saxon laws allow us to observe the progression between two methods of dealing with offences: restitutive, which is typical of family-centred societies; and punitive, which is typical of civilisations where a central power is emerging.’24 The anthropologist Simon Roberts’s ruminations when confronted with Jenny Wormald’s account of the inter-relationship of feuding and central authority in early modern Scotland illustrate how this logic can pan out. Anthropologists should find very interesting what … [social historians] have to say here about the treatment of feud where central government has been established. On the whole anthropologists have assumed that fighting is seldom an approved mode of handling disputes under central government. Rulers tend to object strongly to sustained fighting among their subjects, to present themselves as authoritative agents of dispute resolution and to do their best to make sure they are treated as such. Where significant resort to retaliatory violence has been found in association with central government, this has been taken as an indication of the uncertain extent to which government is established. I am not sure how far this view is challenged by Wormald’s account of the incorporation and regulation of feud in early modern Scotland. For any ruler struggling to establish or extend his authority an alternative to attempting direct suppression must be to associate himself closely with indigenous institutions in the first instance and gradually subject them to regulation.25

Roberts’s reaction to this puzzling historical case study was to find a way of resolving the apparent problem without questioning the assumption that the aims of ‘central government’ remain constant across cultures.26 Scottish kings might have associated themselves with ‘indigenous institutions’, but the supposedly universal tendency of rulers to object strongly to violent feuding made it seem most natural to posit that they did so with the hidden agenda of gradually regulating those institutions so that in the long term feuding could be suppressed entirely. And, logically, if this was really what was going on, the fact that kings had to resort to such a roundabout approach to achieve their goal can only indicate their relative weakness. Thinking of feuding and compensation on the one hand, and state punishment of crime on the other, as two possible mechanisms by which a society might accomplish the same, universal, task of maintaining order thus pushes us inexorably to impose a developmental framework. Primitive ‘horizontal justice’ precedes its more advanced ‘vertical’ counterpart. Problems arise, however, when we try to fit our evidence to this framework. The Anglo-Saxon legal tradition, broadly in line with

23

Lambert, Law and Order, 5–6, provides a brief discussion and some references. Daniela Fruscione, ‘Beginnings and Legitimation of Punishment in Early Anglo-Saxon Legislation from the Seventh to the Ninth Century’, in Capital and Corporal Punishment in Anglo-Saxon England, ed. Jay Paul Gates and Nicole Marafioti, Woodbridge 2014, 34–47 at 34–5. 25 Simon Roberts, ‘The Study of Disputes: Anthropological Perspectives’, in Disputes and Settlements: Law and Human Relations in the West, ed. John Bossy, Cambridge 1983, 1–24 at 9. He is referring to Jenny Wormald, ‘The Blood Feud in Early Modern Scotland’, P & P 87, 1980, 54–97, an abridged version of which appears in Disputes and Settlements, ed. Bossy, 101–44. 26 Cf. J. M. Wallace-Hadrill, ‘The Bloodfeud of the Franks’, Bulletin of the John Rylands Library 41, 1959, 459–87 at 486: ‘There is no strong and continuous royal pressure against the principle of feud, as I see it. There is no “Kampf gegen die Fehde”. Even the pressure of the Church should be subject to most careful interpretation.’ 24



Anthropology, Feud and De obsessione Dunelmi

67

most of its European contemporaries, was fiercely punitive in its treatment of thieves while simultaneously accepting the legitimacy of vengeance killing (albeit within certain limits and alongside all manner of initiatives evidently intended to encourage peace settlements). This combination of ‘horizontal’ and ‘vertical’ elements is all too readily understood as some sort of uneasy hybrid or compromise, the (diagonal?) by-product of a prolonged struggle between royal power and engrained social practices that gradually pushed the system as a whole from one end of the horizontal–vertical spectrum to the other. This can leave us, with Roberts, puzzled by the apparent incoherence of strong rulers tolerating feuding, and instinctively resistant to the notion that they might have conceived of their role – or indeed of order – differently, in ‘indigenous’ terms that might themselves be worth investigating. Instead, this framework encourages us to interpret the early medieval period as one in which law was in flux, the relative balance of horizontality and verticality in the system reflecting the balance of power between state-building rulers and reactionary social forces. In English historiography this is apparent in the way that discussions of feuding have been bound up with arguments about the precocious centralization of the medieval English kingdom from the late ninth century onwards. Patrick Wormald was keen to banish feuding from an impressively state-like pre-Conquest period, presenting it as something tenth-century kings had all but eliminated only for it to be reintroduced by England’s Norman conquerors. So the same sources or types of source which tell us about Norman feuding offer nothing for the late Anglo-Saxon period except punished homicide. This could of course be a result of the far richer wealth of post-conquest evidence. But it might also indicate that Edmund’s strictures on feud took effect; even that the Old English ‘state’ had begun to claim Weber’s ‘monopoly of legitimate violence’. If so, it was a monopoly that kings had no hesitation in exploiting vigorously. There is, to say no more, reason to think that the English scene was quite unlike Professor White’s Touraine: until, that is, conquest superimposed the French scene on it.27

In what is effectively a toned down version of the same argument, Hudson abandons Wormald’s somewhat strained chronological framework and places greater weight on the uncertainties involved in assessing the prevalence of feuding in the period. He concludes his review of the available sources with the unimpeachably cautious observation that ‘in terms both of definition and evidence, it is thus very problematic to say how common or how significant feud was in England from the tenth to the twelfth centuries’.28 Nonetheless, he then proceeds to point out that England was regarded as an ‘area of particular peace’ by contemporaries, to argue that this represented a genuine contrast with ‘not just Celtic Britain but much of Continental Europe’, to contend that ‘the strongest reason for the restriction of violent disputes in England seems to have been the power of the monarchy throughout at least most of its small realm’, and to suggest that this twelfth-century situation might be understood as the result of periodic extensions of royal authority under strong kings from the late ninth century onwards.29 Feuding in this influential strand of scholarship is less a subject of interest in its own right than a proxy for larger arguments about the state-building achievements of England’s kings. The underlying interpretative principles align with those of

27 28 29

Wormald, ‘Giving God and King Their Due’, 341. Hudson, ‘Feud’, 49. Ibid., 49–52, quotations at 49, 50, 51.

68

Tom Lambert

Roberts and Gluckman. If we assume that feuding is a relatively primitive ‘horizontal’ method of maintaining order that rulers instinctively wish to replace with more advanced ‘vertical’ legal regimes like our own, then questions about the prevalence of feuding do indeed become questions about the capabilities of contemporary governmental structures; when the accretion of incremental institutional reforms finally makes it possible for rulers to impose such major legal changes, they will soon seize the long-awaited opportunity to suppress horizontal justice in favour of vertical. Therefore, once ‘strong’ government is achieved, feuding will inevitably decline in significance until it is little more than a residual folk custom, lingering at the unregulated margins of state power. Within this framework of interpretation the many strands of evidence for the English kingdom’s generic governmental ‘strength’ – for the political power of monarchs, for the sophisticated administrative structures at their disposal, and especially for the harsh punishments prescribed and meted out to thieves – all seem to weigh heavily against the notion that feuding might have been prevalent, despite having no direct bearing on the issue. The corollary of this is that sources which suggest that feuding was a social and legal reality, such as De obsessione, constitute awkward signifiers of governmental weakness, seemingly at odds with the wider picture of centralized strength. It makes sense to reconcile them with that wider picture by explaining them away as reflections of an unusually brutal northern political culture, as consequences of Scandinavian influence, or as ideologically motivated archaizing on the part of late Anglo-Saxon legislators. A second strain of anthropological thought with significance for medieval historians’ thinking on feud grew out of the study of segmented tribal structures. The crucial point of departure here is Emrys Peters’s analysis of violent conflict among the Bedouin of Cyrenaica. One of the key ways Bedouin society in the late 1940s differed from medieval Western Europe was its segmentary structure. Tribes were divided into primary sections, which were subdivided into secondary sections, and these were in turn subdivided into tertiary sections. These groups of 200–700 individuals were the practical social units within which people lived: ‘tertiary tribal sections correspond to the camps which cluster around the larger watering points during the drought season’.30 In line with his Manchester colleague, Gluckman, Peters recognized that when incidents of violence occurred between closely linked groups, such as two tertiary sections belonging to the same secondary section, the need to maintain established social and economic relationships created powerful pressures for rapid resolution, leading to peace settlements of one sort or another. However, he also observed that when violence occurred between tertiary sections that belonged to different secondary sections, which characteristically had no established links or practical need to cooperate, there were no such pressures. Conflicts between such groups might well settle into long-term relationships of hostility, punctuated by reciprocal killings. There might be attempts at peace-making through compensation, but because these hostile relationships typically mapped onto competition for resources those peace settlements inevitably proved temporary. The conflict would reignite when the need for scarce water or grazing brought the tribesmen into contact and tempers flared. Peters proposed a typology of violent conflict in Bedouin society in which ‘feud’ referred to these indefinitely long-lasting enmities between groups from

30

E. L. Peters, ‘Some Structural Aspects of the Feud among the Camel-Herding Bedouin of Cyrenaica’, Africa 37, 1967, 261–82 at 262.



Anthropology, Feud and De obsessione Dunelmi

69

different secondary sections, while the more rapidly suppressed hostilities between collateral tertiary sections (i.e. those belonging to the same secondary section) he framed as matters for ‘the penalty of vengeance or the payment of blood money’.31 In this typology hostilities between primary sections could be characterized as raids, whereas those between full tribes Peters classified as warfare.32 At the other end of the scale, killings within tertiary sections were so problematic as to be deeply sinful and shameful, matters to be hushed up and forgotten about, killers sent into a semivoluntary and temporary exile for a number of years to return only once feelings had cooled.33 Probably because of the difficulties involved in applying a schema designed for this context to medieval societies that lack segmentary tribal structures, Peters’s direct influence on medievalists has been fairly small. However, in 1969 Jacob Black-Michaud was awarded an MPhil for a thesis in which attempted to adapt Peters’s insights into a general theory of feuding applicable to the entirety of the Mediterranean and Middle East. In 1975 this thesis was published, without significant revision, as Cohesive Force: Feud in the Mediterranean and the Middle East, to which Peters, having examined the original thesis, contributed a notably reserved foreword.34 Cohesive Force is a difficult work. Dense, convoluted and habitually hyperbolic, it operates on a rather abstract plane throughout, perhaps in part reflecting the fact that, as one of its more hostile reviewers acerbically observed, it was not ‘informed by any field experience on the part of the author in a community with feuding processes going on’.35 It insists, for instance, that feud be understood as ‘a type of generalized struggle extending into all spheres of social life at once, … [which] can only materialise in conditions of what I shall call “total scarcity”’, a concept that ‘may be summarised as the moral, institutional and material premise of a certain type of society in which everything felt by the people themselves to be relevant to human life is regarded by those people as existing in absolutely inadequate quantities’.36 For medievalists’ purposes, however, the book’s most salient contention has been that feud generally, not just in segmented tribal societies, should be comprehended as a ‘hostile relationship’ with a ‘structurally or geographically defined partner’ in which ‘each episode of violence … perpetuates the relationship through time’.37 That is, according to Black-Michaud, feud was by definition interminable. A potential stumbling block for this argument was the consensus in the ethnography on which it was based that (in contexts other than those of conflicts between secondary tribal sections among Cyrenaican Bedouin) feuds could be, and frequently were, ended through peace settlements. Black-Michaud dealt with this by arguing that these other anthropologists were mistaken. He set out to show ‘briefly from the ethnographic facts’ that, contrary to the suppositions of the ethnographers with firsthand knowledge of the societies he was discussing, ‘neither ceremonial reconciliation

31

Ibid., 264–9, quotation at 267. Ibid., 269. The schematic diagram on this page is particularly instructive. 33 Ibid., 263–4. 34 Jacob Black-Michaud, Cohesive Force: Feud in the Mediterranean and the Middle East, Oxford 1975, ix–xxvii. For instance: ‘no attempt is made … to debate the author’s … views of my writings on feud. The divergences between us are evident in what we both have to say’ (ix). 35 Louise E. Sweet, ‘Review of: Jacob Black-Michaud, Cohesive Force: Feud in the Mediterranean and the Middle East’, Middle East Journal 31, 1977, 494–5 at 494. 36 Black-Michaud, Cohesive Force, 121–2, emphasis in original. 37 Black-Michaud, Cohesive Force, 30–1. 32

70

Tom Lambert

after the exaction of vengeance nor peaceful settlement sanctioned by compensation can be truly said to terminate a violent episode’.38 He accomplished this in four pages, comprising a series of ethnographic and historical accounts of situations in which compensation settlements had been reached but subsequently collapsed into further bloodshed.39 This undeniably serves to demonstrate that peace settlements are sometimes ineffective. Exactly how Black-Michaud imagined it supported his contention that they were always ineffective, such that ‘neither vengeance nor compensation can truly conclude a feud’, is not clear.40 Whether it makes sense to extend Peters’s typology of violence in Bedouin society into a generalized theory in this way is debatable. Feuding has not been a significant focus of ethnographic study since the 1960s but there is little in the subsequent literature to suggest that Black-Michaud’s theoretical framework has been influential among anthropologists. Christopher Boehm explicitly set out to refute it in his 1984 ‘ethnohistorical’ book on feuding in highland Montenegro.41 Nonetheless, Cohesive Force has been significant in medieval historiography. Peter Sawyer took it as his sole guide to modern anthropological understandings of feud in his 1987 article ‘The Bloodfeud in Fact and Fiction’ and, understandably enough, noted ‘many fundamental differences between medieval societies and those in which feud has flourished recently’, not least of which was the manifest inappropriateness of Black-Michaud’s concept of total scarcity.42 He concluded that in the medieval period ‘the bloodfeud flourished best, not in the real world, but in the fictions of poets, storytellers and lawyers’.43 In 1997, Guy Halsall drew on both Black-Michaud and Sawyer in arguing for a sharp conceptual distinction between ‘customary vengeance’ – which involves an original offence being satisfactorily resolved through either vengeance or a compensation settlement motivated by the threat of vengeance – and feuds, which necessarily involve at least three violent encounters and reflect long-term hostile relationships.44 Most recently, John Niles’s article ‘The Myth of Feud in Anglo-Saxon England’ adopts Halsall’s definition.45 A central consequence and attraction of these stringent definitional approaches was that they permit historians to cast feuding as an absent or extremely marginal phenomenon in most or all medieval societies. Because real-life feud accounts like that in De obsessione are both scarce and easily doubted as retrospective narrative constructs, requiring a minimum of three connected violent incidents to constitute a feud makes it difficult to confirm even the existence of feuding with medieval evidence. This is convenient from one perspective: as both Halsall and Niles point out, the term feud is heavily loaded in general usage, carrying strong connotations of primitivism which serve to reinforce unhelpful stereotypes about our period’s characteristic barbarism.46 These concerns are not lightly dismissed. As Stephen White has recently shown, for all that the key unifying theme running through

38

Ibid., 13. Ibid., 13–16. 40 Ibid., 15. 41 Here cited from Christopher Boehm, Blood Revenge: The Enactment and Management of Conflict in Montenegro and Other Tribal Societies, 2nd edn., Philadelphia 1987. 42 Sawyer, ‘Bloodfeud’, 32. 43 Ibid., 36. 44 Guy Halsall, ‘Violence and Society in the Early Medieval West: An Introductory Survey’, in Violence and Society in the Early Medieval West, ed. Guy Halsall, Woodbridge 1998, 1–45 at 19–29. 45 Niles, ‘Myth of Feud’, 167. 46 Ibid., 171–3; Halsall, ‘Violence’, 28. 39



Anthropology, Feud and De obsessione Dunelmi

71

serious studies of feuding over the last sixty years has been the stripping away of these problematic associations within academic discourse, the success of this project is belied by the stubborn historiographical resilience of the view that feuding reflects an irrational and dysfunctional aspect of human nature that finds unchecked expression in relatively primitive, unregulated societies.47 Indeed, as White also notes, these ideas have recently received a fillip from heavy-handed attempts to draw on neuroscience and evolutionary psychology, reframing historical violence as a manifestation of human males’ innate vengeful instincts.48 Gregory Hanlon’s recent call for a turn towards ‘post-cultural’ history provides a good example: Vengeance should be seen as part of a universally understood principle of reciprocity. Culture and the modern state have not erased the social processes at work here, and so they resurface each time the state becomes weak. … These patterns of retaliatory killing were (and still are) easy for males to acquire and therefore must be rooted in our genetic makeup, for they rebound wherever the modern state gives way.49

The problem here is not the underlying ideas. Has it ever been doubted that the strong cross-cultural, indeed (as Hanlon notes) trans-species, association of violent behaviour with males must to some extent be rooted in biology? Does anyone seriously contend that modern states play no role in the suppression of violence? But the privileging of these banalities and the side-lining of both non-violent aspects of human nature (presumably the behavioural sciences could equally demonstrate that peaceful co-operation has a biological basis?) and all forces for pacification other than the modern state result in something at once worryingly crude and dangerously alluring. Seemingly common-sense assumptions rooted in our own cultural environment are once again set up as universal truths, this time buttressed by a reassuring scientism. This is just the latest manifestation of what Strathern aptly characterized as ‘a model of social life and human behaviour [in which] society itself is imposed on individuals who are by natural propensity asocial beings’: the basic need for social control discovered afresh in neuroscience and big data.50 These patterns of thought are probably so deeply rooted as to be ineradicable. Social control is perhaps destined to be continually rediscovered as a necessary function of rulers, and whenever it is the evidence of feuding in the Middle Ages will serve to reinforce stereotypes of violent barbarity and primitive government. The urge to banish feud’s pejorative connotations from our period by means of restrictive definitions is understandable. But whether we can reasonably expect this expedient to have much impact on wider perceptions of medieval society is debatable, and such definitions’ significance as an obstacle to historical understanding should not be underestimated. Though Halsall’s distinction between ‘customary vengeance’ and ‘feud’ evidently fits nicely with Peters’s typology of violence among

47

Stephen D. White, ‘The Peace in the Feud” Revisited: Feuds in the Peace in Medieval European Feuds’, in Making Early Medieval Societies: Conflict and Belonging in the Latin West, 300–1200, ed. Kate Cooper and Conrad Leyser, Cambridge 2016, 220–43. For this period specifically, see David Crouch, The Birth of Nobility: Constructing Aristocracy in England and France, 900–1300, Harlow 2005, 95, where Anglo-Norman aristocratic mores are defined in opposition to the ‘Iron age societ[y] of pre-Conquest England’ that they had ‘left behind’, whose ‘aristocratic habitus was harsh, unreflective, unforgiving and violent’ by comparison. 48 White, ‘Feuds in the Peace’, 242. 49 Gregory Hanlon, ‘The Decline of Violence in the West: From Cultural to Post-Cultural History’, EHR 128, 2013, 367–400 at 396. 50 Strathern, ‘Social Control’, 113.

72

Tom Lambert

the Bedouin of Cyrenaica and with the general theorizing of Black-Michaud, it is much less apparent, and cannot safely be assumed, that any medieval culture understood conflicts that encompassed one or two killings to be conceptually distinct from those which persisted for three or more. Boehm objected to imposing this rigid schema on violent conflicts in highland Montenegro on the grounds that that doing so involved ‘fragmenting what is an organic system of symbols and behaviour’.51 This concern is no less pressing for our period. The anthropology that I, at least, have found most helpful and intellectually satisfying in coming to terms with feuding has been that which prioritizes the elucidation of the concepts and categories that framed local understandings of violent acts, rather than that which purports to construct universalizing models. An excellent example is Peters’s later analysis of debt relationships among the Bedouin, which shows how debts created by hospitality, gift-giving and some kinship relationships were understood within the same conceptual field as debts ‘of blood’ and ‘of a bone’ in contexts of feud, vengeance and compensation.52 Likewise, Paul Dresch’s exposition of ‘the language of honour’ among Yemeni tribesmen does much to emphasize the important point that cases of violent antagonism (he eschews the vocabulary of feud) can unfold in all manner of unpredictable ways, all of which are understood locally within a common framework of categorization and moral evaluation.53 This is what I take Boehm to be getting at with the phrase ‘an organic system of symbols and behaviour’; feud for him is a much broader and more flexible concept, encompassing not just long-term cycles of reciprocal vengeance killings but also much less dramatic incidents in which a potentially violent conflict provoked by an original offence is swiftly and successfully settled through the intervention of prominent local figures. What matters is the existence of social and cultural context that makes an escalating cycle of vengeance a realistic prospect, one which ‘indigenous problem-solvers’ need to work to avert.54 And rapidly settled conflicts can be just as reflective of such a context as intractable long-term enmities. It has long been recognized that the medieval period can furnish us with plenty of examples of feuding understood in the more flexible way favoured by Boehm.55 Indeed, both William Ian Miller, writing on feuding in Icelandic sagas, and Paul Hyams in an English context, have drawn directly on Boehm in framing the concept.56 But how can we go about reconstructing something akin to ‘an organic system of symbols and behaviour’ using medieval evidence? Miller has the best claim to have achieved this. With the astonishing richness of the sagas to draw upon finding sufficient relevant material was not an issue. Perhaps his biggest challenge was persuading readers that ‘the sagas describe a real world external to themselves, a world we can recover fairly well in those areas where the saga description is thick enough’.57 Anticipating scepticism, Miller also articulated a fall-back position: ‘we are at least recovering the world of the sagas and the laws, 51

Boehm, Blood Revenge, 221. Emrys L. Peters, ‘Debt Relationships’, in idem, The Bedouin of Cyrenaica, ed. Jack Goody and Emanuel Marx, Cambridge 1990, 138–87 at 169–72. 53 Paul Dresch, Tribes, Government, and History in Yemen, Oxford 1989, 38–70. 54 Boehm, Blood Revenge, 121–42 and quotation at 206. 55 For example, Wallace-Hadrill, ‘Bloodfeud’. 56 Hyams, Rancor, 6–11; William Ian Miller, Bloodtaking and Peacemaking in Saga Iceland, Chicago 1990, 179–81; see also Jeppe Büchert Netterstrøm, ‘Introduction: The Study of Feud in Medieval and Early Modern History’, in Feud in Medieval and Early Modern Europe, ed. Jeppe Büchert Netterstrøm and Bjørn Poulsen, Aarhus 2007, 9–67 at 49–59. 57 Miller, Bloodtaking, 76. 52



Anthropology, Feud and De obsessione Dunelmi

73

a world that, furthermore, has the virtue of looking very much like some worlds that we can prove to have existed and in fact have been carefully studied’.58 Early medieval England, sadly, is not blessed a similarly rich vein of fictional material. Much can be done to excavate values and assumptions from what we have (as Peter Baker has recently demonstrated)59 but no-one would think to describe Beowulf as ‘consciously realistic’, let alone argue that it depicts ‘a real world external to itself’.60 The deficit cannot be made up with narratives of real-life conflicts such as that of De obsessione, for these are rare, particularly before the late twelfth century.61 What England does have is a substantial corpus of laws. From the late ninth to the early eleventh century, the Cerdicing rulers of Wessex and England were moderately enthusiastic legislators, such that King Eadred (946–55) is unique in having ruled for a significant period without leaving an identifiable legislative record. With laws, however, we enter problematic territory. Looking back on the last thirty years of ‘medieval conflict studies’ in the USA in 2001, Warren Brown and Piotr Górecki offered a summary of what medievalists had learnt from the works of Simon Roberts and John Comaroff in the late 1970s and early 1980s. Roberts, they explained: introduced, for the first time explicitly, two approaches to analysing, comprehending and describing those dimensions of ‘order’ that come into play in the context of dispute, but which he declined to identify as ‘the law’ – the normative and the processual. The normative approach takes as given, that is, assumes an autonomous existence and importance of norms, and then enquires into their convergence with the behaviour they purport to regulate. The processual begins by reconstructing the behaviour of participants in disputes, and then situates the role of norms within that behaviour as a dependent variable.62

As they note, Roberts was ‘meticulously evenhanded’ when he set out these two approaches for the benefit of historians in 1983, despite the fact that his own and Comaroff’s work ‘falls squarely on the processual side of [the] schema’, bridging ‘the distinction between norm and process by essentially annexing the one into the other’.63 Perhaps this characterization might be debated, but it strikes me as fairly representative of what early medieval historians took from this anthropological encounter: a pronounced distrust of laws as a source for social history (with a concomitant shift to examining them as a product of elite intellectual culture) and an enthusiasm for the project of emulating anthropologists by analysing the behaviour of participants in real-life disputes, usually having reconstructed that behaviour from legal documents.64 The implications of this distrust of laws are discernible in approaches to feuding in pre-Conquest England. Wormald devoted most of his life’s work to the analysis 58

Ibid. Emphasis in original. Peter S. Baker, Honour, Exchange and Violence in Beowulf, Cambridge 2013. 60 Miller, Bloodtaking, 46. 61 This is amply illustrated by the effort to assemble a corpus based on inclusive criteria in Hyams, Rancour, 267–308. 62 Warren C. Brown and Piotr Górecki, ‘What Conflict Means: The Making of Medieval Conflict Studies in the United States, 1970–2000’, in Conflict in Medieval Europe: Changing Perspectives on Society and Culture, ed. Warren C. Brown and Piotr Górecki, Aldershot 2003, 1–35 at 7. 63 Brown and Górecki, ‘What Conflict Means’, 8; likewise, Dresch, ‘Legalism, Anthropology, and History’, 6 n; see also Roberts, ‘Study of Disputes’. 64 An excellent critical overview can be found in Alice Rio, ‘Introduction’, in Law, Custom and Justice in Late Antiquity and the Early Middle Ages: Proceedings of the 2008 Byzantine Colloquium, ed. Alice Rio, London 2011, 1–22. 59

74

Tom Lambert

of Anglo-Saxon law texts but, somewhat paradoxically, he was sceptical about their practical relevance. On the subject of feuding this was particularly pertinent. His argument was that the laws (which fairly straightforwardly accept the legitimacy of vengeance killing for homicide, at least in situations where the killer’s family obdurately refuse to pay compensation) are at their most misleading as a guide to the realities of legal practice here. They were produced by and for intellectuals and, for ideological reasons Wormald never articulated precisely, they make no reference to the establishment of royal punishment for homicide, generalized and severe enough to have effectively outlawed feuding.65 Both Hyams and Hudson have since sought to modify this position, declining to follow (without explicitly disavowing) Wormald in attributing decisive significance to two terse allusions to ‘unjust slaying’ in late tenth-century documents.66 Like Wormald, they emphasize the laws’ ideological content and caution readers against assuming them relevant to legal practice, and their interpretative strategies seem to reflect these anxieties.67 The central analytical goal for both writers is to find the right balance between what Hudson characterizes as, on the one hand, ‘the aims of the king and his advisers, often ecclesiastical’, and on the other, ‘non-royal attitudes and activities’.68 Both explicitly frame their conclusions in terms of a conflict between the aspirations of kings and the resistance their initiatives prompted – Hyams emphasizing the latter much more strongly than Hudson – and grapple with the problem of how much weight to assign to each, while recognizing the impossibility of doing so accurately with the evidence available.69 The outcome, in both cases, is essentially a restatement of Wormald’s interpretation, but with the addition of different degrees of doubt about kings’ ability to overcome social counter-pressures and impose their will. The assumed conflict between ‘horizontal’ and ‘vertical’ justice is lurking here, giving this interpretative framework a common-sense appeal, but closer to the surface and perhaps just as significant is the dominant historiographical view of laws as potentially highly unrealistic prescriptive texts. This naturally permits their use to illustrate royal aspirations, and it seemingly also licenses attempts to uncover something of laws’ unrealism by reading them against the grain, in the hope of 65

Wormald, ‘Giving God and King Their Due’, 336–42; Patrick Wormald, The Making of English Law: King Alfred to the Twelfth Century. Volume I: Legislation and Its Limits, Oxford 1999, 483: ‘The development of English law from Alfred to Cnut and beyond did not run along the tracks that legislation marked out, but it did run in parallel with them’ (emphasis in original). Patrick Wormald, ‘Papers Preparatory to The Making of English Law: King Alfred to the Twelfth Century. Volume II: From God’s Law to Common Law’, ed. Stephen Baxter and John Hudson, London 2014 (http://www.earlyenglishlaws.ac.uk/reference/ wormald/ [accessed 1 November 2018]), 10–12. 66 Wormald, ‘Giving God and King Their Due’, 339–40. Cf. Lambert, Law and Order, 195–7. 67 Hudson, ‘Feud’, 34–5: ‘To escape the silence, one is compelled to turn to the laws, reading them not merely for signs of royal intention or aspiration but also of social practice. The laws, however, are highly problematic texts. As Patrick Wormald has shown, they survive very much as ecclesiastical and ideological documents … Their overall significance in terms of revealing widespread practice may therefore be exaggerated’. Hyams, Rancor, 78: ‘Scholars have now learned to understand these [laws] as kingship treatises not actually used in litigation and thus of most significance as indicators of an ideology of royal governance.’ 68 Hudson, ‘Feud’, 35. 69 Ibid., 40: ‘[kings’] very efforts and aspirations reveal counter-pressures, the continuing use of violent self-help motivated by vengeance, the continuing involvement of kin and others. There must have been resistance, unconscious and conscious, to the extension of royal authority.’ Hyams, Rancor, xvi: ‘I do not deny … the ideology of a kingship with vast aspirations toward the extension of central authority and the creation of an administrative apparatus that looked capable of its implementation. … My scepticism doubts the king’s capacity to face down the natural resistance of the noble class on whom all royal power ultimately rests.’



Anthropology, Feud and De obsessione Dunelmi

75

catching sight of the modes of thought and behaviour against which kings were pushing. It encourages, that is, a view of laws’ contents as inherently contradictory: the assumption that they reflect a tense encounter between two coherent worldviews, the ideology of centralized authority in full view, that of society glimpsed only occasionally. By contrast, it obstructs, perhaps even precludes, analyses starting from the premise that the laws constitute coherent wholes, and lends an air of theoretical naivety to the not obviously unreasonable supposition that elite lawmakers shared many of their assumptions with the societies to which they belonged. The idea that we might use these texts to reconstruct what Miller might have termed ‘the world of the laws’ – something closer to Boehm’s ‘organic system of symbols and behaviour’ than to even a broad conception of ‘royal ideology’ – is difficult to reconcile with this received historiographical wisdom. Indeed, as Dresch has observed, the focus on reconstructing behaviour in real-life disputes tends to push categories and values – ‘the terms in which people act’ – out of the limelight.70 Feuding becomes significant as a ‘process’, instances of which can be unearthed from narrative sources with reference to a set of defining characteristics, rather than as a manifestation of a shared framework of ideas. And once conceptualized in this way, it is all too easy to descend, as Hudson and Wormald do for England, into trying to assess whether feuding was more or less prevalent in some places and periods than others by counting up the number of occurrences found in them, as though the exercise might somehow manage to prove informative despite the obvious problems posed by small sample sizes and divergent patterns of source production and survival. As Timothy Reuter once observed of attempts to perform essentially the same exercise in late tenth- and eleventh-century France, ‘trying to assess a quantitative problem using qualitative evidence means that impressionism and anecdotalism masquerade as hard data instead of fulfilling their proper functions within the rhetoric of historical explanation’.71 So, what should be taken from all of this? It is easy to assume that what anthropologists have to offer historians are universalizing ‘theories’, and that the anthropologically literate medievalist’s role is to comprehend these theories and sensitively apply them to medieval evidence. But while some anthropologists, like Jacob BlackMichaud, do indeed engage self-consciously in the elaboration of generalizing theories, there are others who sympathize with Paul Dresch’s lament: There are few things more irritating for the reader than larding one’s text with what Stephen Potter would have called ‘OK anthropology names’, or bracketing what one has to say between portentous discussions of ‘theory’ as if the latter had a life of its own apart from ethnography. Perhaps, unhappily, it does. But the ethnography is what matters. We seem to be divided into those who argue theory with ethnography as illustration and those who identify theoretical sophistication with avoiding the asking of faulty questions. My sympathies lie all with the latter group.72

Dresch means to be read literally here. This is not a denunciation of the pursuit of theoretical sophistication but of the belief that it inheres in a discrete, abstract discourse distinct from the discussion of concrete ethnographic detail.73 His point is that true theoretical sophistication lies in careful scrutiny of the questions one poses 70

Dresch, ‘Legalism, Anthropology, and History’, 3–9, quotation at 6. Timothy Reuter, ‘Debate: The “Feudal Revolution”: III’, P & P 155, 1997, 177–95 at 178. 72 Dresch, Tribes, vii. 73 For more detailed critical reflection on the ‘distorting effects’ or the ‘exteriorisation of theory’ see Judith Scheele and Andrew Shryock, ‘Introduction: On the Left Hand of Knowledge’, in Scandal of Continuity, ed. Scheele and Shryock [forthcoming]. 71

76

Tom Lambert

and of the assumptions lying behind them. It is developed reflexively, in dialogue with both the society under study and whatever intellectual resources the anthropologist or historian can productively bring to bear on the task. Understood thus, theoretical sophistication is indistinguishable from the analytical acuity, careful conceptual reflection and eclectic inspiration that has long characterized the best historical scholarship, and quite different from the intellectual passivity implied by any attempt simply to ‘apply’ abstract theory derived from another discipline. Such abstract theories might usefully prompt us to interrogate established interpretative frameworks, identifying problematic questions and suggesting possible refinements, but it would be naive to imagine that they can be exploited in any straightforward way, somehow ‘used’ to produce improved answers to existing questions. This essay’s underlying hope is that its review of various strands of the anthropology and historiography of feuding has cleared some conceptual ground; recognizing that there are some conventional questions that we should avoid asking of medieval feuding ought to prompt us to focus our collective efforts on more productive avenues. It might help to return briefly to the example with which we began, the De obsessione Dunelmi feud narrative, to get a sense of what this might look like in practice. First, then, faulty questions. Perhaps the key point that has emerged here is that we should not expect the De obsessione narrative specifically, or evidence of feuding more generally, to tell us about the strength of the state, in eleventh-century Northumbria or anywhere else. The varying extent of English kings’ political control over different northern regions in this period is one topic; the question of how cases of violent antagonism among the non-elite populations of these regions were usually conducted is quite another. If killing was demonstrably a more frequent feature of political competition in Northumbria than in the south, this could simply reflect the political turbulence arising from southern appointees’ often limited ability to command the loyalty of entrenched local elites. It need not imply that Northumbria’s free population was any more prone to violent feuding that those of the Mercia or Wessex. And even if that were the case, it would be problematic to frame this as indicative of local resistance to feud-suppressing initiatives pursued elsewhere by the English state. There is no evidence that such initiatives existed, nor is there good reason to suppose kings wanted to introduce them. And while the idea that royal authority was episodically strengthened by a series of powerful rulers across the tenth, eleventh and twelfth centuries clearly has merit, we need to be alert to the risk of conceptual imprecision when diagnosing increases in such qualities as ‘royal authority’ or ‘governmental strength’. We should not lightly assume that our perception of a generalized growth in central power in this period implies any concurrent shift in non-elite practices of vengeance and compensation. We also need to be wary of expecting feuding to function as a mechanism of social control, as a form of ‘horizontal justice’ which necessarily constituted a positive contribution to the maintenance of order or somehow promoted social cohesion. These ideas begin from assumptions that are difficult to defend, and De obsessione’s bloody narrative would be exceptionally hard to reconcile with such characterizations in any case. If we approach it expecting feuding to perform social functions we will inevitably find it a disappointing case study, and perhaps even be tempted to read it as symptomatic of some sort of wider dysfunction. Indeed, it is probably best to avoid trying to make this text bear on general questions about the broader social consequences of feuding practices. It is not simply that De obsessione is an unrepresentative case study, but that our wider source-base for real-life feuding is so thin as to render the whole question of whether feuding served more to undermine or



Anthropology, Feud and De obsessione Dunelmi

77

uphold community cohesion moot. If we are interested in contemporary perceptions of feuding’s impact on order, on the other hand, we need to acknowledge that this is not a salient theme in De obsessione, whereas laws often have to make assumptions on the issue and occasionally address it directly. If we want to find out whether elite Anglo-Saxon socio-legal theory involved an equivalent to ‘The Peace in the Feud’ we should make use of this much richer seam of evidence.74 The question of whether the events described in De obsessione can really be considered a feud is also unlikely to lead anywhere productive. The most honest answer is uninteresting: it depends on how you define a feud. And even if everyone could agree a single definition, what would be the point? It would allow us to count up recorded feuds, certainly, but it would do nothing to make the project of quantitative comparison across regions with diverse source-bases more viable. We need to accept that we are ill equipped to gauge the prevalence of feuding, however defined, in places and periods that lack systematic judicial records. And we should confront the fact that historians have been drawn down this unproductive avenue by dubious assumptions – by the idea that an authoritative demonstration of feud’s prevalence or absence in a society can tell us something significant about the strength of its central government, or its level of civilisation. The most intellectually fertile explorations of feuding have not been those which focus on putting historical examples in modern phenomenological categories but those, like Miller’s work on Icelandic sagas, which seek to place vengeance and compensation in their cultural context, showing how they are enmeshed with pervasive notions like honour, masculinity, hospitality, debt and balance. Perhaps, then, it would make sense to analyse De obsessione primarily as a cultural artefact, a repository for some of the categories and values that framed contemporary understandings of violent conflict. A good starting point is Hudson’s observation that our narrative is a construct, made up of events from which very different feud narratives might have been wrought, and indeed which need not have been interpreted as vengeance killings by all contemporaries. As he notes, there is a sense in which ‘feud can be said to be narrative, generally legitimizing narrative’.75 Once one starts down this surprisingly underexplored avenue progress is swift. There can be no doubt whose violence the De obsessione feud narrative serves to legitimize. The final act in the drama, Earl Waltheof’s sending a band of men to slaughter the sons and grandsons of Carl while they feasted in their hall in the winter of 1073–4, would look like a reprehensible piece of opportunistic political assassination if taken in isolation. But as it is framed in the narrative it is the culmination of an inter-generational feud stretching back to Earl Uhtred. Moreover, it is a feud in which the family of Carl and Thurbrand has twice been responsible not just for killing members of the House of Bamburgh, but for doing so treacherously. Uhtred had received safe conduct and was a guest in Cnut’s hall when he was slain because of Thurbrand’s treachery. Ealdred’s vengeance killing of Thurbrand is portrayed neutrally – we are told nothing about it except that it happened – but when Carl in turn kills Ealdred he is killing his own guest, his sworn brother, and a fellow pilgrim. This backstory provides ample justification of Waltheof’s slaughter. The fact that his hall massacre showed no respect for the protection that, in late Anglo-Saxon legal theory, was supposed to cover people within houses could even be read as a calculated display of contempt. Carl’s family had repeatedly shown how 74

For an attempt to do this: Lambert, Law and Order, 227–30, 236 n. Hudson, ‘Feud’, 33. The next sentence – ‘However, it is also a phenomenon, or better a process.’ – shifts his discussion to another analytical plane. 75

78

Tom Lambert

little value they placed on the protection implied by their hospitality, so why would Waltheof recognize any claim to safety at their own feast? The narrative serves to make a staggeringly brutal, somewhat underhand, and perhaps rather grubbily politicized mass killing into condign, even glorious, vengeance. This observation focuses attention on the context for which this narrative was framed, and on the vexed question of how far it distorts Waltheof’s true motivations. The idea that he was driven solely by a burning urge to avenge Earl Ealdred’s murder, thirty-five years previously, stretches credulity. If Waltheof’s personal honour really was bound up so closely with that of his mother’s family, the House of Bamburgh, he must have struggled with the fact that his father – Earl Siward, a Scandinavian appointed to rule Yorkshire under Cnut – was reputed to have treacherously killed its leading member, Earl Eadulf, before he seized the family’s earldom for himself.76 When Siward died in 1055 Waltheof was perhaps five years old and had already lost his mother, and thus his most important link to the House of Bamburgh. His upbringing is obscure, but there is little to suggest her family were closely involved; he first enters the historical record in the Midlands in 1065, when he seems to have been made earl over Northamptonshire and Huntingdonshire, as his father had been.77 The first reference to his presence in a northern context comes in the autumn of 1069, when both Waltheof and, interestingly, ‘the four sons of Carl’ appear in a list of the leaders of the rebel army that stormed York in conjunction with a Danish fleet.78 When this rebellion failed Waltheof managed to turn the situation greatly to his advantage: he submitted and was not only pardoned by the king but even brought into the ruling family. He married William’s niece, was granted further lands in the south and, following the expulsion of his cousin Gospatric in 1072, was made earl of Northumbria.79 The Historia regum portrays him as subservient in that role, humbly and obediently doing as he was bidden by Walcher, a much-hated Lotharingian appointed bishop of Durham by William.80 It would be plausible, then, to portray Waltheof as a marginal figure to the House of Bamburgh – tainted by association with his father and perhaps a southern upbringing – and as a Norman quisling. As C. P. Lewis has noted, one possible reading of the massacre of Carl’s progeny is that Waltheof was ‘acting for King William in destroying enemies who had refused to come to terms’.81 But if he was regarded with suspicion by his maternal relatives, then reigniting a feud stretching back to Earl Uhtred would have served as a statement that even though Waltheof was Siward’s son and William’s appointee, his loyalties lay firmly with the House of Bamburgh. It seems unlikely to be accidental that our narrative serves not only to glorify these dubious killings but also to constitute Waltheof as unproblematically a member of the family which claimed descent from Earl Uhtred, perhaps even its natural head. 76

Historia Regum, in Opera Omnia, ed. Arnold, II, 198; Symeon of Durham: Libellus de Exordio atque Procursu istius hoc est Dunhelmensis Ecclesie, ed. David Rollason, Oxford 2000, 170. 77 Forrest S. Scott, ‘Earl Waltheof of Northumbria’, Archaeologia Aeliana 30, 1952, 149–215 at 155–64; C. P. Lewis, ‘Waltheof, Earl of Northumbria (c. 1050–1076)’, ODNB, no. 28646, accessed 1 November 2018. 78 Orderic II, 226; Scott, ‘Earl Waltheof’, 170–84. 79 Scott, ‘Earl Waltheof’, 185–92; Lewis, ‘Waltheof’. 80 Historia Regum, in Opera Omnia, ed. Arnold, II, 260; Henrietta Leyser, ‘Walcher, Earl of Northumbria (d. 1080)’, ODNB, no. 28428, accessed 1 November 2018. 81 Lewis, ‘Waltheof’; conversely, Kapelle, Norman Conquest, 135, imagines that Waltheof was in rebellion against William, characterising his raid into Yorkshire to kill Carl’s descendants as a ‘traditional Northumbrian gesture of defiance of southern authority’.



Anthropology, Feud and De obsessione Dunelmi

79

De obsessione may thus provide some insight not just into how Waltheof’s actions were framed by his apologists but also into how he sought to construct a favourable reputation and identity for himself. We could read the killing of Carl’s descendants as an exercise in self-fashioning through symbolic violence: a public statement of his commitment to a culturally significant set of ideals about masculinity and violence, which simultaneously asserted his position as the primary custodian of Bamburgh family honour. Whatever other motivations may have underlain the massacre, Waltheof must have appreciated that it would assume such significance when framed as the climax of this feud narrative. Of course, it is uncertain how far contemporary audiences were convinced by this. Though the likelihood that the story was circulating in the Bamburgh family’s heartlands in the decades after Waltheof’s execution for treason suggests that they happily claimed him as one of their own in retrospect, whether he achieved such acceptance in his lifetime is impossible to know. But we can safely assume that the story’s intended audience understood and at least partially subscribed to the framework of values to which the narrative appeals, regardless of what they made of Waltheof’s manoeuvrings. There would have been no point trying to legitimize and valorize his violence in these terms were this not so. The story therefore also allows us to infer something about the cultural context which made Waltheof’s actions rational and his message legible to contemporaries. If a feuding culture is one in which vengeful violence is recognized as a legitimate and properly masculine response to an affront, then De obsessione implies that late eleventh-century Northumbria was a feuding culture. It is doubtful that this tells us anything about the barbarity and lawlessness of the north relative to the south, or about the strength of the English state; but it may have implications for our understanding of what it meant to be male and noble in this society, for contemporary conceptions of honour, hospitality and debt, and for the social meaning of the many acts of political violence that litter our narrative sources for the period. Appendix: Dating De obsessione Dunelmi The dating of De obsessione is a vexed issue. The text survives in a single copy in Cambridge Corpus Christi College MS 139, the relevant section of which dates to 1161–7.82 At the outside, it must have been composed after the final massacre of the feud, which took place in the winter of 1073–4, and before a series of letters dated 1109–14 which documents two of the alienated estates De obsessione discusses returning to Durham’s possession.83 However, modern commentators have generally accepted Christopher Morris’s conclusion that ‘there are elements [in the text] that point to composition either in the early 1070s or the first two decades of the twelfth century, without either date being entirely convincing’.84 The early date, originally advocated by Bernhard Meehan, is based primarily on the idea that the text would have mentioned Earl Waltheof’s execution in 1076, had it taken place.85 This logic is not wholly compelling. Earl Waltheof appears in the text only because of his role in the feud story, which the author explicitly treats as a digression from his main topic, the history of the estates. Given that Waltheof’s 82

Morris, ‘Marriage and Murder’, 5–6. The Charters of the Honour of Mowbray, 1107–1191, ed. D. E. Greenway, London 1972, 6–12. 84 Morris, ‘Marriage and Murder’, 7–10, quotation at 10. See Hudson, ‘Feud’, 31; Fletcher, Bloodfeud 4; David Rollason, ‘Symeon’s Contribution to Historical Writing in Northern England’, in Symeon of Durham: Historian of Durham and the North, ed. David Rollason, Stamford 1998, 1–13 at 11. 85 Meehan, ‘Siege of Durham’, 18–19; Morris, ‘Marriage and Murder’, 7–8.

83

80

Tom Lambert

execution appears unrelated to both the feud and the estates, discussing it may simply have been a digression too far. Moreover, to accept a composition date this early it is necessary to explain away the text’s two references to the present day, seemingly written much later, as copyists’ interpolations. The later date range of 1100–20, suggested as an alternative by Morris, derives from one of these references to the present. A fight involving a different Waltheof is said to have taken place ‘recently’, and it is implied elsewhere in the text that this Waltheof’s mother, Ecgfritha, could not herself have been born before 1070.86 By what date we can plausibly imagine Ecgfritha’s son to have reached an age when he, rather than his still-living father, was fighting a different branch of the Bamburgh family is debatable. Morris reasoned that 1100 was the earliest possible point, but in extending his range as far as 1120 he implicitly recognized that this requires a suspiciously compressed chronological sequence – both Waltheof’s parents and grandparents reproducing at the earliest possible opportunity, then Waltheof assuming the leading role in an intra-familial conflict in his mid-teens – and that sometime after 1110 would be more plausible.87 As Morris himself notes, the key problem with placing the composition of De obsessione in this later period is that there are several indications that by then ‘the information it contains was largely out of date’.88 He points out, for instance, that Bishop Ranulf Flambard was forced to return the estates of Carlton and School Aycliffe to the church in 1109. De obsessione tells us simply that these estates were returned to Durham before 1066 and remained in its possession. Likewise, in De obsessione’s present day, the estates of Barmpton and Skirningham were possessed by Ecgfritha and her husband, Ælfsige of Tees, who had seized them on their own initiative and authority; the text provides information about possible heirs, presumably for future reference, naming their eldest son Waltheof and their daughter Eda. Sometime between 1109 and 1114, however, Nigel d’Albini was returning those estates to Durham, Waltheof having held them as his tenant.89 Quite a number of significant events – Nigel d’Albini’s asserting superiority over Barmpton and Skirningham, Waltheof’s inheriting from his parents, Ranulf Flambard’s abstraction of Carlton and School Aycliffe – appear to have taken place in these estates’ history between the composition of De obsessione and the early 1110s. While it is logically possible to posit another compressed chronology here, such that all these things happened in rapid succession, Morris was rightly concerned about the plausibility of this; a more natural reading is that cumulatively they suggest the landholding information in De obsessione belongs to an earlier period. This led him to wonder whether De obsessione really was intended to provide up-to-date information on the estates to which Durham had a claim, in spite of the lack of satisfactory alternative explanations for its purpose.90 Given that both early and later suggestions for dating are problematic, we perhaps ought to look again at the intervening period. Morris rejected the 1080s and 1090s because of this account of the ancestry of Waltheof, described elsewhere as having ‘recently’ been involved in fighting: Afterwards, King William came to England and Arkil fled into exile. And so that land lay waste once more. Later, a thegn from Yorkshire called Orm, son of Gamel, married Æthel-

86 87 88 89 90

De obsessione, §3 (nuper), 8. Morris, ‘Marriage and Murder’, 7, 10. Ibid., 10. Ibid., 8. Ibid., 8–10.



Anthropology, Feud and De obsessione Dunelmi

81

thryth, one of the five daughters of Earl Ealdred, and they had a daughter called Ecgfritha, who by Ælfsige of Tees had a son, Waltheof, two other sons, and a daughter, Eda. Because Ecgfritha was descended from Ealdred and the daughter of Bishop Aldhun, she claimed hereditary right and, with her husband, Ælfisge, seized Barmpton and Skirningham.91

If we take this literally, Arkil fled and his estates were wasted, probably in the winter of 1069–70, and it was only after this that Æthelthryth married Orm.92 We therefore have to allow time for Ecgfritha to be born, come of age, and marry Ælfsige before the final events in the account. However, it is striking that the nongenealogical events in the narrative seem to form a causal sequence: because Arkil fled into exile, leaving his lands lying vacant and wasted, they were available to be claimed by the opposing branch of the family. It is possible that this is the story which the author meant to tell. That is, we might suppose that he did not intend to imply that Ecgfritha was born only after Arkil’s flight in 1069–70, with Barmpton and Skirningham lying abandoned and wasted until she came of age and married in perhaps the late 1080s. Rather, he meant to convey that Ecgfritha claimed these estates soon after Arkil’s departure, probably in the early 1070s. If so, the genealogical information on who Ecgfritha was should be read as an explanatory aside, intended to help readers understand the nature of Ecgfritha’s claim and to identify her possible heirs, which has accidentally slipped into the main narrative sequence, sowing confusion. If this reading is correct, Waltheof could easily have been of an age to fight much earlier than 1100: there is no obstacle to dating the text to the 1080s or 1090s, where it is a rather more comfortable fit. A date in the 1070s, perhaps even the early 1080s, would make the text’s other reference to the present day difficult to explain. This is a statement that Sumerled, the only son of Carl not present at the massacre of 1073–4, ‘survives to this day’.93 The implication must be that some considerable time had passed since the massacre, and probably also that Sumerled was now an old man. Domesday Book lists a Sumerled as the 1066 holder of one estate in Yorkshire and several in Lincolnshire, but he is not named as a 1086 landholder. If we could infer from this that the Sumerled mentioned in De obsessione died before the Domesday survey the most likely date for the text would be the early 1080s, but how secure such a deduction would be is debatable.94 Dating De obsessione to the 1080s or 1090s necessitates the supposition that its author became muddled while trying to integrate complex genealogical information into his narrative, or perhaps that a later copyist introduced an error while trying to clarify his prose. This is not something one should suppose lightly, but in the circumstances it is arguably the least problematic interpretative course open to us. Meehan’s earlier dating requires two passages to be dismissed as later interpolations, and if we opt for Morris’s later period we must either reconcile ourselves to some implausibly compressed chronologies or imagine that our author neglected to include current information on the estates he was discussing.

91

De obsessione, §8: Postea, Willelmo rege veniente in Angliam, ipse Arkillus fugiuens exul factus est. Sicque iterum terra ipsa vastata remansit. His ita peractis, quidam tein in Eoverwicschire, nomine Orm, filius Gamellonis, accepit uxorem unam ex quinque filiabus Aldredi comitis Etheldritham, ex qua genuit filiam nomine Ecgfridam, ex qua Eilsi de Teise genuit Waltheof, et duos ejus fratres, et Edam sororem eorum. Et quia illa Ecfrida erat de genere Aldredi comitis et filiæ Alduni episcopi, hæreditario jure ipsa Ecgfrida, cum marito suo Eilsi, arripuit Dermetum et Skirningeim. 92 Orderic II, 218, 222, 226–8; Ann Williams, The English and the Norman Conquest, Woodbridge 1995, 40. 93 De obsessione, §7 (usque hodie superest). 94 Sumerled’s Domesday holdings are discussed, with references, in Williams, The English, 31 n.

This chapter is available as Open Access under the CC-BY licence

NEW ARCHAEOLOGIES OF THE NORMAN CONQUEST Aleksandra McClain and Naomi Sykes It is becoming clear that the Norman Conquest both initiated and intensified farreaching changes at all levels of society, including culture and identity, social structure, economy, diet, art and architecture, portable material culture, rural and urban settlement, manorial and community landscapes, religion and mortuary practice, and the management of the environment. Many of these elements are either inaccessible from documentary evidence alone or have distinct material implications, and recent archaeological research is beginning to show that they have the potential to complicate traditional historical narratives of the Conquest, or take our understanding of the period in new directions.1 Nevertheless, the vast majority of academic scholarship on the Conquest has been carried out without reference to the abundant archaeological evidence from the eleventh and twelfth centuries that has been recovered in excavation and still survives above ground. As a result, both academic and public understanding of the Conquest has been predicated primarily on its impact on the elite social classes, and on narratives that have been derived almost entirely from the wealth of documentary history available for the period. The comprehensive archaeology of the Norman Conquest and Anglo-Norman transition is therefore a story that is yet to be told. This paper articulates key themes, research directions, and a new case study emerging from the project ‘Archaeologies of the Norman Conquest’, an AHRC research network led jointly by the authors.2 The chief objective of the project has been to create and sustain a research community linked by an interest in revitalizing archaeological research in the eleventh and twelfth centuries, in order to examine the cultural, social, and political implications of the Norman Conquest and its aftermath through an explicit focus on material culture. The network has brought together the humanistic, scientific, academic, professional and public-engagement arms of archaeology, as well as key participants from cognate disciplines, in a range of workshops focusing on themes of interpretative agendas, methodologies, international perspectives and public outreach. The central aim has been to create a materially focused research framework for the period which can be engaged with and taken forward by both archaeological and interdisciplinary audiences. This article is a first step in that direction. The network and its activities address a prominent gap in the current research environment focusing on the eleventh and twelfth centuries. The Battle and Haskins Naomi J. Sykes, The Norman Conquest: A Zooarchaeological Perspective, BAR International Series 1656, Oxford 2007; Ben Jervis, ‘Conquest, Ceramics, Continuity and Change. Beyond Representational Approaches to Continuity and Change in Early Medieval England: A Case Study from Anglo-Norman Southampton’, Early Medieval Europe 21, 2013, 455–87; Michael Fradley, ‘Scars on the Townscape: Urban Castles in Saxo-Norman England’, in The Archaeology of the Eleventh Century: Continuities and Transformations, ed. Dawn M. Hadley and Christopher Dyer, Abingdon 2017, 120–38; Aleksandra McClain, ‘Rewriting the Narrative: Regional Dimensions of the Norman Conquest’, in The Archaeology of the Eleventh Century, 203–27. 2 Project reference: AH/P006841/1; http://www.normanarchaeology.org 1

84

Aleksandra McClain and Naomi Sykes

conferences have made strides in recent years by including more work from the fields of archaeology and material culture, including the extremely welcome addition of an annual archaeology-focused memorial lecture at Battle (see Roffe, this volume). Nevertheless, archaeology and archaeologists have customarily been a featured extra instead of core members of these two most prominent events of Anglo-Norman scholarship. The root causes of archaeology’s lack of engagement with the eleventh and twelfth centuries and the Conquest itself will be discussed in more detail below, but on the whole, the problem stems less from other disciplines refusing to listen to what archaeologists have to say than from archaeology doing too little to make itself heard. For a variety of reasons, archaeologists have not sufficiently appreciated that the period is not only brimming with unexplored potential in and of itself, but can also inform much broader questions about the interactions between society, material culture, ethnicity, culture contact and political transition that are enormously relevant to many other periods of human history. While innovative archaeological scholarship on the Conquest is taking place, it has been carried out by a relatively small number of scholars often isolated by methodology or sector, without a like-minded community of material-culture specialists interested in the period to listen, read, comment and spur on further research. The network has thus offered a space in which dialogue and collaboration between the range of specialisms within archaeology can be fostered, focusing specifically on what we can do together to interrogate more fully how and why the Norman Conquest and Anglo-Norman transition happened, and the significance of the material dimensions of that process. The Current State of Norman Conquest Archaeology The supposed invisibility of the Norman Conquest in the material record, at least when it comes to the ‘stuff of everyday life’ which is regarded as the primary purview of archaeology, has posed a longstanding quandary for the discipline.3 Trevor Rowley has written that archaeologists have always been ‘more conservative’ in our attitudes to the impact of the Conquest than cognate historical disciplines, despite acknowledging that the spread of castles across the English landscape and the mass rebuildings of churches in Romanesque styles after 1066 were undoubtedly significant, tangible signatures of the arrival of the Normans.4 He argued that this conservative perspective has stemmed primarily from the lack of immediate, obvious stylistic or technological changes in more prosaic material elements such as pottery, tools, and craft materials like wood, metal, bone and leather, suggesting that life for the masses went on much as before, despite the political upheaval.5 Instead of facing this apparent contradiction head on, or seeing it as an opportunity to investigate a period of particular complexity, the response of the discipline was instead to avoid the subject, and leave most nuanced interrogation of the process and effects of the Conquest to documentary historians. Despite the fact that the Conquest is easily the most widely recognized event in the English Middle Ages, Rowley’s 1997 work Norman England: An Archaeological Perspective on the David Griffiths, ‘The Ending of Anglo-Saxon England: Identity, Allegiance, and Nationality’, in The Oxford Handbook of Anglo-Saxon Archaeology, ed. David A. Hinton, Sally Crawford, and Helena Hamerow, Oxford 2011, 62–78 at 62–3. 4 Trevor Rowley, Norman England: An Archaeological Perspective, London 1997, 12. 5 Rowley, Norman England, 134; Trevor Rowley, Norman England (Shire Living Histories), London 2010, 1. 3



New Archaeologies of the Norman Conquest

85

Norman Conquest was the first general academic book on the Norman Conquest with a specific focus on archaeology.6 Remarkably, there were no further attempts in the intervening twenty years, until the arrival of Dawn Hadley and Christopher Dyer’s 2017 edited volume The Archaeology of the Eleventh Century: Continuities and Transformations.7 This is in marked contrast to the many Norman Conquest overviews from a historical perspective which have been published relatively recently.8 The lack of scholarship is emphatically not the result of a lack of archaeological data. On the contrary, a wealth of buildings, material culture and landscape evidence exists from the period, as well as human, animal and plant remains. However, this data is most often subsumed within wide-ranging considerations of a particular type of material culture or evidence, emphasizing the developmental trajectories of a class of settlement, landscape, building or artefact over the longue durée of the whole of the medieval period, or the early or late Middle Ages – frustratingly, often ending at 1066 or beginning after it. Vital data pertinent to the period is also contained within large, multi-period excavations of rural sites, of which Wharram Percy, Goltho and Faccombe Netherton are three of the best known, as well as major urban excavations at such sites as Winchester, Lincoln, York and Wallingford.9 There is of course considerable value in couching eleventh- and twelfth-century material within the context of longer and broader developmental trends, rather than looking at it in relative isolation. Nevertheless, consistently having taken the (very) long view means that we have lacked a body of research which specifically marshals material evidence to ask probing questions about how the Conquest happened, the ways in which the material dimensions of this socio-cultural transition were important, or even about distinctive elements of the eleventh and twelfth centuries and the role the period played in longer trajectories. The lack of a coherent intellectual framework has resulted in work which often undertakes the essential first step of characterizing the material culture from the period, but stops short of analysing its significance in the context of wider research questions, preventing archaeologists from playing a substantive role in scholarly debates on the Norman Conquest. Neither Rowley’s, nor Hadley and Dyer’s, book explicitly seeks to lay out research agendas for the archaeology of the Conquest, although some key themes can be discerned, as can the impact of broader theoretical and disciplinary developments within academic archaeology over the past two decades. Rowley focuses consistently on the coexistence of continuity and change throughout the period, emphasizing continuity at ‘everyday’ levels and watershed-worthy change at elite levels, and he cites the arrival of a more systematic and consistently defined 6

Rowley, Norman England. The Archaeology of the Eleventh Century: Continuities and Transformations, ed. Dawn M. Hadley and Christopher Dyer, Abingdon 2017. 8 Amongst others, Hugh M. Thomas, The Norman Conquest: England after William the Conqueror, Lanham 2008; George Garnett, The Norman Conquest: A Very Short Introduction, Oxford 2009; Richard Huscroft, The Norman Conquest: A New Introduction, Harlow 2009; Brian Golding, Conquest and Colonisation: The Normans in Britain, 1066–1100, Basingstoke 2013. 9 A History of Wharram Percy and its Neighbours, ed. Stuart Wrathmell, Wharram: A Study of Settlement on the Yorkshire Wolds XIII, York 2012; Guy Beresford, Goltho: The Development of an Early Medieval Manor, c. 850–1150, London 1987; J. R. Fairbrother, Faccombe Netherton: Excavations of a Saxon and Medieval Manor Complex, vols I and II, London 1990; Martin Biddle, Object and Economy in Medieval Winchester, vols I and II, Oxford 1990; Michael Jones, David Stocker, and Alan Vince, The City by the Pool: Assessing the Archaeology of the City of Lincoln, Oxford 2003; Patrick Ottaway and Nicola Rogers, Craft, Industry and Everyday Life: Finds from Medieval York, The Archaeology of York 17/15, York 2002; Neil Christie and Oliver Creighton, Transforming Townscapes: From Burgh to Borough, the Archaeology of Wallingford, Abingdon 2013. 7

86

Aleksandra McClain and Naomi Sykes

feudal social system as a major Norman contribution.10 His stress on the widespread evidence for continuity was likely more groundbreaking in archaeological circles at the time than it seems to us twenty years on, although it should be noted that it still lagged somewhat behind the critical assessments of the variable impact of the Normans conducted by some documentary scholars.11 Throughout the volume, Rowley’s discussions focus primarily on the material evidence of castles, churches, and urban and rural planning. He cites the supposed ‘invisibility’ of the Conquest in other areas of archaeology, particularly small finds, and does not probe it much further, conceding that material culture may be a less-than-ideal medium for tracking short-term or subtle change.12 Rowley also flirts with the idea that cultural exchange around the Conquest was more complicated than is often supposed, and although he does not address the intricacies of the issue in full, just acknowledging the reflexive influence of conquered upon conquerors was an innovative theme, and one that still remains to be substantively explored in subsequent archaeological scholarship.13 Hadley and Dyer’s volume demonstrates that while archaeology has undoubtedly moved on in the intervening decades, some concerns remain the same. The Archaeology of the Eleventh Century, like Rowley’s book, focuses heavily on the duality of continuity and change in the period, as the subtitle communicates. The lack of precision in dating around 1066 is an issue that has still not been satisfactorily resolved, nor is it likely to be when style is the primary diagnostic element, although advances in the availability and precision of absolute dating such as radiocarbon have been noted.14 However, on the whole it is clear that those twenty years have brought archaeologists greater confidence in their ability to contribute to scholarly dialogue on the Conquest. In their introduction, Hadley and Dyer state outright that archaeology can provide insights to elements of the period which have no connection to political events, documentation or ‘great men’, and they offer no concession to archaeological ‘invisibility’ in the period.15 The complete lack of any mention of feudalism in the volume, even during considerations of castles, manors and the countryside after 1066, stands in stark contrast to Rowley’s discussions, and demonstrates how much more comfortable modern medieval archaeology has become in moving away from the parameters of traditional historical debates. When reviewing Norman England, Robert Higham wrote that of the ample archaeological evidence for change that Rowley cited, ‘it would be interesting to distinguish … those changes specifically arising from the “Normanness” of the Conquest from those which flowed more generally from the economic, artistic, military and intellectual growth which the twelfth century witnessed on a wider

10

Rowley, Norman England, 21, 123, 134. Robin Fleming, Kings and Lords in Conquest England, Cambridge 1991; Ian Short, ‘“Tam Angli quam Franci”: Self-Definition in Anglo-Norman England’, ANS 18, 1995, 153–75; Hugh M. Thomas, ‘The significance and Fate of the Native English Landholders of 1086’, English Historical Review 118:476, 2003, 303–33. 12 Rowley, Norman England, 13. 13 Rowley, Norman England, 13, 32, 110. 14 Dawn M. Hadley and Christopher Dyer, ‘Introduction’, in The Archaeology of the Eleventh Century: Continuities and Transformations, ed. Dawn M. Hadley and Christopher Dyer, Abingdon 2017, 1–13 at 2. For commentary on radiocarbon dating, see Elizabeth Craig-Atkins, ‘Seeking “Norman Burials”: Evidence for Continuity and Change in Funerary Practice Following the Norman Conquest’, in The Archaeology of the Eleventh Century, 139–58 at 155. 15 Hadley and Dyer, ‘Introduction’, 2. 11



New Archaeologies of the Norman Conquest

87

front’.16 It is heartening to see that Higham’s directive eventually became a central theme in The Archaeology of the Eleventh Century, which throughout stresses the strength of archaeology to contribute to our understanding of the context of broader eleventh- and twelfth-century trajectories.17 However, a full understanding of these developments requires looking beyond England to a wider European context, both to Normandy and the other countries of the British Isles, obviously, but also to regions where 1066 was a footnote rather than a watershed. Both Rowley’s and Hadley and Dyer’s volumes acknowledge the importance of an international perspective, and offer some parallel history of what was happening on the Continent at the time, yet there are no chapters in the edited volume which consider evidence from outside England.18 Archaeological scholarship published in English which encompasses this period on the Continent is thin on the ground, and few productive connections or discussions have as yet been made with archaeologists writing on the period in other countries – another situation which the network project has sought to improve.19 The Value of Material Approaches to the Norman Conquest A significant part of archaeology’s lack of engagement with the Norman Conquest has stemmed from our historically complicated relationship with documented periods of the past, which has been covered extensively in other work.20 This tension is particularly pronounced in the eleventh and twelfth centuries, due to the explosion of documentation that resulted from the establishment of Norman bureaucracies, legal frameworks and religious institutions. This remarkable wealth of written evidence has of course fuelled the long tradition of historical work on the period, but it has also perhaps intimidated archaeologists into thinking there is little we can add to the discussion, and that our efforts might be better spent where the documentary record is less comprehensive. Alongside the large body of textual scholarship that has been generated on the period, the apparent confluence between the most obvious forms of ‘Norman’ material culture (e.g. castles, major churches) and a story of successful Norman imposition can also give the impression that the

16

Robert A. Higham, ‘Review of Trevor Rowley, Norman England: An Archaeological Perspective on the Norman Conquest’, Archaeological Journal 155, 1998, 410–11 at 411. 17 Hadley and Dyer, ‘Introduction’, 2. See also the case study below for commentary on longer-term processes. 18 Rowley, Norman England, 13–19; Hadley and Dyer, ‘Introduction’, 6–7. Continental contexts and the ‘European aristocratic diaspora’ are also touched on in Keith D. Lilley, ‘Urban Landscapes and the Cultural Politics of Territorial Control in Anglo-Norman England’, Landscape Research 24:1, 1999, 5–23 at 18. 19 Two examples published in English are David Petts, ‘Churches and Lordship in Western Normandy AD 800–1200’, in Churches and Social Power in Early Medieval Europe: Integrating Archaeological and Historical Approaches, ed. Jose C. Sanchez-Pardo and Michael Shapland, Turnhout 2015, 297–338 and Elisabeth Zadora-Rio, ‘The Making of Churchyards and Parish Territories in the Early-Medieval Landscape of France and England in the 7th-12th Centuries: A Reconsideration’, Medieval Archaeology 47, 2003, 1–19. 20 The best-known discussion is still found in David Austin, ‘The “Proper Study” of Medieval Archaeology’, in From the Baltic to the Black Sea: Studies in Medieval Archaeology, ed. David Austin and Leslie Alcock, London 1990, 9–42. For two more recent perspectives, see also Martin O. H. Carver, ‘Marriages of True Minds: Archaeology with Texts’, in Archaeology: The Widening Debate, ed. Barry Cunliffe, Wendy Davies and A. Colin Renfrew, Oxford 2002, 465–96 and the chapter ‘Archaeology and its Discontents’ in Guy Halsall, Cemeteries and Society in Merovingian Gaul: Selected Studies in History and Archaeology, 1992–2009, Leiden 2010.

88

Aleksandra McClain and Naomi Sykes

period is both straightforward and already well understood. Until very recently, the Conquest seemed to incite in us what Richard Bradley has called archaeology’s ‘loss of nerve’ – the silence that results when we fear we cannot say anything valuable about the past, or in this case, anything that history has not already said.21 In general, historians have been considerably better at parsing the impact of the Conquest and critiquing the ‘story’ told by the documentary evidence than have those who deal with material culture. Particular patterns or lacunae in late eleventh- or early twelfth-century material evidence have been explained by attributing them to the simple existence of the Norman Conquest or some feature of it, such as the Harrying of the North, without further interrogation.22 A lack of thorough engagement with the relevant historical scholarship on the period has resulted in archaeologists at times framing our interpretations around grand narratives which have already been critiqued, questioned, or deconstructed by historians – a problem which Guy Halsall has noted in other historical periods as well.23 While archaeology should not be subordinated to history’s research questions any more than history should be subordinated to ours, the solution is not for archaeology to avoid engaging with historical agendas and evidence, as David Austin once argued.24 In the eleventh and twelfth centuries, it is actually becoming clear that archaeology’s agendas increasingly align with or complement those of historians, especially as economic, social and cultural histories have come to the fore alongside the traditional subjects of legal, military and political history, and, most significantly, with the advent of the material turn in history and other humanities.25 Best practice will allow these areas of confluence to be identified and developed in collaboration with each other, while still leaving each discipline free to explore research questions and avenues which are more particular or disciplinarily specific. As an antidote to archaeology’s sometime intimidation in the face of highly documented periods, Halsall has argued instead that the more extensive the written record in a period, the greater the potential of archaeology to complement that record, but also to question and challenge it, in order to provide better insight into social and ideological structures. A much richer understanding is possible in periods where the material and documentary records are both substantial, rather than when one or the other dominates.26 In raw material terms, therefore, the untapped collaborative and interdisciplinary potential of the eleventh and twelfth centuries is clear. However, the volume of documentary evidence, or scholarship, is in some sense irrelevant to the value of material culture to this period. Material culture has much to add to the study of the Norman Conquest, or indeed any period of transition, not only because

21

Richard Bradley, ‘Archaeology: The Loss of Nerve’, in Archaeological Theory: Who Sets the Agenda?, ed. Norman Yoffee and Andrew Sherratt, Cambridge 1993, 131–3 at 133. 22 Pamela Allerston, ‘English Village Development: Findings from the Pickering District of North Yorkshire’ Transactions of the Institute of British Geographers 51, 1970, at 106; Christopher Harper-Bill, ‘The Anglo-Norman Church’, in A Companion to the Anglo-Norman World, ed. Christopher Harper-Bill and Elisabeth van Houts, Woodbridge 2002, 165–90 at 171; Nigel Saul, English Church Monuments in the Middle Ages, Oxford 2009, 24. 23 Halsall, ‘Archaeology and its Discontents’, 76. 24 Halsall, ‘Archaeology and its Discontents’, 87; Austin, ‘The “Proper Study”’, 12–13. 25 Hugh M. Thomas, ‘History, Archaeology and the Norman Conquest’, in The Archaeology of the Eleventh Century, 283–300 at 283; History and Material Culture: A Student’s Guide to Approaching Alternative Sources, ed. Karen Harvey, Abingdon 2009; Dan Hicks and Mary C. Beaudry, ‘Introduction. Material Culture Studies: A Reactionary View’ in The Oxford Handbook of Material Culture Studies, ed. Dan Hicks and Mary C. Beaudry, Oxford 2010, 1–21 at 1. 26 Halsall, ‘Archaeology and its Discontents’, 84.



New Archaeologies of the Norman Conquest

89

it is valuable to us now, as some of the best surviving evidence of past social behaviour, but also because it was enormously important to people at the time. Humans are uniquely material animals, set apart by our creation of artefacts, art, technology and an extensive built environment and anthropogenic landscape. Indeed, ‘material culture’ itself may be a redundant term, as it is not really possible to define a form of human culture which is not materially enacted in some way.27 In periods of social pressure or intense change – for which the Norman Conquest undoubtedly qualifies – material expression would have been especially communicative and meaningful.28 It provided a tangible outlet through which to display identities and allegiances, and was a means of social competition and orientation, both of which have already been noted as increasingly important features of eleventh- and twelfth-century lordship and elite culture.29 Material culture was also both a symbol and a tool of power and control, and an important means of maintaining, subverting or negotiating social norms.30 In periods of transition, material culture could be central to processes of stabilizing and reifying political and cultural change, through reorganization of the natural and built landscapes, monetary systems, and cultural practices, such as personal adornment, fashion and style, and cuisine. As strategic remembering and forgetting of meaningful places, monuments and cultural practices were a part of every political transition, the necessity of engagement with material evidence in order to fully interrogate processes of social change in the past is clear.31 Archaeological evidence is also valuable in that it allows us to access a greater proportion of society than we can through documents alone, providing insights into modes of communication with diverse audiences that did not rely on literacy of either party. This potentially opens up much wider avenues of enquiry into how change happened and who it affected. In classic terms arguing for the value of archaeology, this has typically been the lower and marginalized classes, who are rarely documented (‘the voiceless’),32 but there is also considerable opportunity to reveal the huge swathes of the middling sort and lower elite, who are also less often represented in documents. Of course, even the high elite who are frequently present in textual evidence also had rich material lives. All strata of medieval society had material dimensions which we need to understand, not just those absent from documents. It is important to emphasize here that while material culture does offer access to a wider spectrum of society in the eleventh and twelfth centuries than contemporary texts do, we do not take the view that it is necessarily a superior form of evidence, or that it offers a more ‘real’ view of the past than documentary history. Material culture is no more objective, apolitical, unbiased, or able to ‘speak for itself’ than documents, although it has at times been treated that way.33 They are simply

27

Dan Hicks, ‘The Material-Cultural Turn: Event and Effect’, in The Oxford Handbook of Material Culture Studies, ed. Dan Hicks and Mary C. Beaudry, Oxford 2010, 25–98 at 27. 28 McClain, ‘The Archaeology of Transition’, 25. 29 Gardiner, ‘Manorial Farmsteads’, 100. 30 Following the recursive social model of structuration in Anthony Giddens, The Constitution of Society, Oxford 1984, 9; John C. Barrett, ‘Fields of Discourse’, Critique of Anthropology 7, 1988, 5–16 at 10. 31 Paul Gready, ‘Introduction’, in Political Transitions: Politics and Cultures, ed. Paul Gready, London 2003, 1–26 at 2. 32 Kenneth L. Feder, Linking to the Past: A Brief Introduction to Archaeology, Oxford 2007, 77. 33 James Deetz, In Small Things Forgotten: The Archaeology of Early American Life, 2nd ed., New York 1996, 259; Catherine M. Hills, ‘History and Archaeology: Do Words Matter More than Deeds?’, Archaeological Review from Cambridge 14, 1997, 29–36 at 33. Both cite material culture as more ‘objective’ than other forms of evidence.

90

Aleksandra McClain and Naomi Sykes

different forms of evidence, with distinct strengths and weaknesses. As such, documents and archaeology may well tell different stories about the Conquest, and that is not inherently problematic.34 While there inevitably will be disconnects between the interpretations drawn from texts and those drawn from material culture, the interpretations of different types of material culture do not always easily align either. The pasts spoken to by eleventh-century pottery and a Romanesque cathedral are potentially as distinct from each other as those spoken to by pottery and the AngloSaxon Chronicle. Even the same class of material culture is fully capable of offering distinctive insights and apparently conflicting narratives about the period. For example, Gareth Perry’s examination of Torksey ware pottery from Lincolnshire has shown that it flourished in Lincoln and York from the ninth to the mid-eleventh centuries, becoming those towns’ dominant Anglo-Scandinavian ceramic type by c. 1000. It then steadily declined beyond that date and had ceased production by the lateeleventh century, its decline apparently aligning with changing styles and modes of production after the Conquest.35 Stamford ware, another Lincolnshire type, also arose c. 900 and was distributed widely across eastern and midland England in the period immediately before the Conquest.36 In contrast, however, production of Stamford ware continued uninterrupted not only through the Conquest and transition, but as late as c. 1250, and seems to have responded little to the arrival of the Normans or the production of new ceramic types.37 These apparent disconnects that we see arising between texts and material culture, and between different types of material culture, emerge precisely because the texts and artefacts were created by a wide array of people, for variable audiences, and for specific purposes. They are the products of complex people doing complex things in complex situations, so we should not expect the stories they tell us to be simple ones. Challenges and Opportunities Improving material approaches to the eleventh and twelfth centuries involves a range of methodological challenges and opportunities. A primary hindrance in maximizing our knowledge and use of relevant datasets for the eleventh and twelfth centuries has been the lack of integration between the academic, professional and heritage sectors of archaeology. Contract archaeologists digging for units often generate relevant material, but their ability to fully publish new excavations is limited, so results are held in grey literature repositories, and old excavation archives often sit in storage. Similarly, there are a number of important sites curated by heritage bodies, as well as artefacts held in national and regional museums, but the pressures of time and public funding limit the scope of these institutions to initiate interaction with academic archaeology. Excavation archives, grey literature reports, 34

See McClain, ‘Rewriting the narrative’ for an extensive discussion of the contrasting pictures of the aftermath of the Harrying of the North as portrayed in chronicles versus the evidence from churches and commemoration. 35 Gareth J. Perry, ‘Pottery Production in Anglo-Scandinavian Torksey (Lincolnshire): Reconstructing and Contextualising the Chaîne Opératoire’, Medieval Archaeology 60, 2016, 72–114 at 77; Ailsa Mainman, Anglo-Scandinavian Pottery from 16–22 Coppergate, The Archaeology of York 16/5, York 1990, 427. 36 Chris Cumberpatch et al., ‘A Stamford Ware Pottery Kiln in Pontefract: A Geographical Enigma and a Dating Dilemma’, Medieval Archaeology 57:1, 2013, 111–50 at 112. 37 Kathy Kilmurry, The Pottery Industry of Stamford, Lincolnshire, c. AD 850–1250, BAR British Series 84, Oxford 1980, 201–3.



New Archaeologies of the Norman Conquest

91

and museum and site stores therefore represent a considerable opportunity to make better use of extensive and so-far unexploited existing datasets. Due to the dominance of 1066 as an epoch-dividing date, accurate dating has always been a primary concern in the period. Because much archaeological dating has been reliant on typologies and chronologies built on progressions of form and style, hybrid products and technological or stylistic continuity through the late eleventh and early twelfth centuries have made it difficult to assign a wide range of material culture to either late Anglo-Saxon or Norman contexts.38 Uncertainties in these chronologies have then had knock-on effects on relative stratigraphic sequences, as it has often been difficult to correlate archaeological horizons, based primarily on ceramic and numismatic chronologies, with pre- and post-Conquest activity.39 To more effectively engage with the Norman Conquest, we therefore need both improved chronologies and dating wherever possible, and also to find more productive ways of dealing with inevitable chronological uncertainties. The inability to place a particular object, building, or artefact on one side of 1066 or the other should not be taken as precluding us from saying anything useful about the Conquest, or as a shortcoming of archaeological data or methods. Rather, the fact that a particular class of material culture is difficult to place on one side or the other of a major cultural and political breakwater should be seen as interesting in and of itself, and an opportunity to ask why. The primary opportunities concerning absolute dating for the period involve recent methodological innovations in radiocarbon dating. The considerable numbers of surviving stone buildings from the eleventh and twelfth centuries provide opportunities for the dating of mortars used in these structures, allowing us to confirm or challenge current stylistically derived architectural chronologies, particularly given that some buildings (e.g. major churches, castles) are known to be post-Conquest, while others (e.g. parish churches) are uncertain.40 A previous study dating Carolingian mortars has also been able to elucidate patterns of the reuse of broken-down Roman limestone in early medieval mortars, which has considerable potential for illuminating debates in England over new quarrying and the reuse of Roman materials in both the Anglo-Saxon and Anglo-Norman periods.41 Another advance has been in the dating of both charred food and lipid residues on pottery sherds, which 38

David Stocker and Paul Everson, Summoning St Michael: Early Romanesque Towers in Lincolnshire, Oxford 2006 and Richard Gem, ‘The English Parish Church in the 11th and Early 12th Centuries: A Great Rebuilding’, in Minsters and Parish Churches: The Local Church in Transition, 950–1200, ed. John Blair, Oxford 1988, 21–30 have addressed this dilemma for church architecture. McClain, ‘Rewriting the narrative’ has done so for funerary sculpture, and Craig-Atkins, ‘Seeking “Norman Burials”’ for burial practice. Rosie Weetch, ‘Tradition and Innovation: Lead-Alloy Brooches and Urban Identities in the 11th Century’, in The Archaeology of the Eleventh Century, 263–82 and David Hinton, Gold, Gilt, Pots and Pins: Possessions and People in Medieval Britain, new ed., Oxford 2006, 1have considered dating issues in metalwork. 39 Rowley, Norman England, 12. 40 Methodological advances are detailed in Asa Ringbom et al., ‘19 Years of Mortar Dating: Learning from Experience’, Radiocarbon 56, 2014, 619–35 and Roald Hayen et al., ‘Mortar Dating Methodology: Assessing Recurrent Issues and Needs for Further Research’, Radiocarbon 59, 2017, 1859–71; medieval case studies in Jan Heinemeier et al. ‘Successful AMS 14C Dating of Non-Hydraulic Lime Mortars from the Medieval Churches of the Åland Islands, Finland’, Radiocarbon 52, 2010, 171–204 and Juan Antonio Quiroz-Castillo et al., ‘Dating Mortars: Three Medieval Spanish Architectures’, Arqueologia de la Arquitectura 8, 2011, 13–24. 41 Sophie Hueglin, ‘Time Framing Earl Medieval Stone Building North of the Alps--A Discussion of Recent Challenging Results’, Radiocarbon 59, 2017, 1657–75; Richard Morris, ‘Churches in York and Its Hinterland: Building Patterns and Stone Sources in the 11th-12th Centuries’ in Minsters and Parish Churches: The Local Church in Transition, 950–1200, ed. John Blair, Oxford 1988, 191–9.

92

Aleksandra McClain and Naomi Sykes

has been carried out primarily in prehistoric periods.42 The direct dating of lipids particularly is opening up new avenues for archaeological chronologies, and has the potential to clarify Saxo-Norman pottery sequences, as well as illuminating whether usage changes even if the ceramic style or form does not. Even just gaining more dates from traditional radiocarbon dating sources, as seen in Jim Leary’s recent project coring Norman mottes, will be exceptionally useful for clarifying eleventhand twelfth-century sequences, given that the period has comparatively rarely been a target for investigation.43 In addition to dating methods, there are a number of scientific analytical techniques which have been used to great effect in prehistoric archaeology, but which have not often been applied to the medieval period generally, or the eleventh and twelfth centuries particularly. These techniques can be employed to inform us about past health and diet, animal breeding and husbandry, the animal products used to make artefacts, leather and parchment, the sources of pigments used in wall paintings and illuminated manuscripts, and the movements of human and animal populations and individual people over their lifetimes.44 The application of scientific approaches to newly generated evidence, as well as to the extant datasets and archives discussed above, greatly expands the scope of how we can interrogate past people and things in the context of the Norman Conquest. Some of the implications of these techniques for the eleventh and twelfth centuries will be discussed in the case study below, hopefully illustrating that new archaeologies of the Norman Conquest will not be Peter Sawyer’s ‘expensive way of telling us what we know already’, but rather untapped sources of evidence opening up novel lines of enquiry about people and society, which all historical disciplines should have an interest in.45 Case Study: Animals, Economy and Cultures of Cuisine In 2005, Anglo-Norman Studies published a paper that summarized Naomi Sykes’s doctoral thesis, which was later published in its entirety as a monograph in 2007.46 This research synthesised all available published zooarchaeological data from midfifth- to mid-fourteenth-century ad sites, as well as the results from original analyses of archived material, the aim being to see if any Conquest-related shifts in assemblage composition were observable. The main conclusions of the study were that, indeed, there was a dramatic increase in wild-animal exploitation on post-Conquest sites of high status, namely castles and manors. This was accompanied by evidence for new hunting rituals that were apparently associated with the arrival of the exotic fallow deer (Dama dama). In turn, the fallow deer’s introduction brought the necessity of emparkment, the fashion for which increased in the post-Conquest period, directly in correlation with rising fallow deer populations. Sykes argued that all of 42

Melanie Roffett-Salque, ‘From the Inside Out: Upscaling Organic Residue Analyses of Archaeological Ceramics’, Journal of Archaeological Science: Reports 16, 2017, 627–40; Dai Kunikita, ‘Dating Charred Remains on Pottery and Analyzing Food Habits in the Early Neolithic Period in Northeast Asia’, Radiocarbon 55:3, 2013, 1334–40. 43 https://roundmoundsproject.wordpress.com/ 44 Relevant examples include G. Marucci et al., ‘Raman Spectroscopic Library of Medieval Pigments Collected with Five Different Wavelengths for Investigation of Illuminated Manuscripts’, Analytical Methods 10, 2018, 1219–36, and Matthew D. Teasdale et al., ‘The York Gospels: A 1000-Year Biological Palimpsest’, Royal Society Open Science 4, 2017. 45 Quoted in Philip Rahtz, ‘New Approaches to medieval Archaeology, Part 1’, in Twenty-Five Years of Medieval Archaeology, ed. David A. Hinton, Sheffield 1983, 12–23 at 15. 46 Sykes, ‘Zooarchaeology of the Norman Conquest’ and The Norman Conquest.



New Archaeologies of the Norman Conquest

93

these elements – hunting rituals, fallow deer and parks – formed an elite cultural package introduced from the Norman kingdom of Sicily following the Conquest.47 In many ways, Sykes’s Battle Conference paper was an important marker. Not only was it one of the earliest archaeological articles to be published in AngloNorman Studies, but it also set out the intellectual agenda upon which a number of subsequent research programmes were built. For instance, it underpinned the interdisciplinary project ‘Dama International’, which sought to explore the extent to which the biocultural history of the fallow deer was a legacy of the Norman empire.48 The project’s results are now forthcoming and, intriguingly, they are challenging the interpretations presented in the 2005 article, providing unexpected new data and requiring new narratives to be constructed for our understanding of trade and human–animal–environment relationships through the eleventh and twelfth centuries. In this section we will present the data from this and other recent research projects, with the aim of highlighting how archaeological investigations – in particular those that combine multiple forms of evidence, and facilitate collaborations between the humanistic and scientific arms of archaeology – can generate powerful new data to reinvigorate old questions and interpretations. Before presenting these new results, it is first necessary to examine some of the broader economic and dietary trends that were presented in Sykes’s book of 2007, but not considered in the 2005 article. Figure 1 shows basic zooarchaeological representation data for pigs and chickens on medieval sites. They indicate that both species are better represented on elite sites dating to the Norman period, with pigs also demonstrating a post-Conquest increase on low-status rural sites. These subtle shifts have been suggested to reflect Norman dietary preferences, since French medieval sites show high frequencies of pigs and chickens on sites of all types.49 The evidence for cattle and sheep representation shows no Conquest-related change, but instead suggests a long-term, gradual trend (seventh to fourteenth centuries ad) towards increased sheep, which is undoubtedly linked to the growth of the wool industry. An emphasis on wool is also reflected by the sheep age profiles, which demonstrate that animals were kept to increasingly older ages in order to obtain the maximum number of wool-clips from each individual.50 Against these overarching trends, however, there are some noticeable anomalies in the eleventh and twelfth centuries which are particularly observable on low-status rural sites. Figure 2 illustrates the cattle and sheep ageing data from these rural ‘producer’ sites, and it can be seen that the mid-eleventh to mid-twelfth centuries represent a point of inflection, whereby the number of animals slaughtered under 12 months of age drops substantially, with more animals being slaughtered between 1 and 3 years. These variations suggest a post-Conquest shift away from dairying, indicated by the high frequency of animals slaughtered prior to 6 months of age in pre-Conquest assemblages, and towards meat production. An emphasis on meat production would also be consistent with the already noted increase in pigs and chickens, both of which are predominantly food-producing animals, in the form of meat and eggs.

47

Sykes, ‘Zooarchaeology of the Norman Conquest’, 197; Naomi J. Sykes et al., ‘Wild to domestic and Back Again: The Dynamics of Fallow Deer Management in Medieval England (c. 11th–16th century AD)’, STAR: Science & Technology of Archaeological Research 2:1, 2016, 113–26. 48 https://gtr.ukri.org/projects?ref=AH%2FI026456%2F1 49 Sykes, The Norman Conquest. 50 Naomi J. Sykes, ‘From Cu and Sceap to Beffe and Motton: The Management, Distribution, and Consumption of Cattle and Sheep in Medieval England’ in Food in Medieval England: Diet and Nutrition, ed. C. M. Woolgar, Oxford 2005, 56–71 at 62.

94

Aleksandra McClain and Naomi Sykes

Figure 1: Changing zooarchaeological representation of A) pigs and B) chickens through the Middle Ages, and by site type A) Pigs 60 % Total Assemblage

Rural

Urban

Elite

Religious

45 30 15 0

5-7th

7th-mid-9th

B) Chickens

mid-9th-mid-11th mid-11th - mid-12th mid-12th-mid-14th

France 10-12th

Century AD

% Total Assemblage

30

22.5

15

7.5

0

5-7th

7th-mid-9th

mid-9th-mid-11th

mid-11th-mid-12th mid-12th-mid-14th

France 10-12th

Century AD

These zooarchaeological findings are given greater weight when combined with evidence from pottery lipid analysis, a technique that allows fats preserved on the interior of pottery sherds to be extracted and analysed to reconstruct the ingredients that were cooked within the vessel.51 Traditionally, this method has been applied almost exclusively to prehistoric material culture, but the potential for exploring medieval ceramics is now being recognized.52 For example, the study of vessels from the rural Anglo-Norman site of West Cotton (Northamptonshire) was able to highlight a shift from dairy fats in the Anglo-Saxon pottery towards adipose (body) fats in the post-Conquest period, indicative of the consumption of beef, lamb and pork. In line with the zooarchaeological evidence, they also found no evidence for dairying in the post-Conquest ceramics.53 A similar observation has recently been made in a forthcoming analysis of ceramics from Saxo-Norman Oxford. Dairying markers were found only in the pre-Conquest pottery, and high quantities of pig fat were found only in post-Conquest sherds.54 51

H. R. Mottram et al., ‘New Chromatographic, Mass Spectrometric and Stable Isotope Approaches to the Classification of Degraded Animal Fats Preserved in Archaeological Pottery’, Journal of Chromatography A, 833, 1999, 209–21. 52 M. S. Copley et al., ‘Direct Chemical Evidence for Widespread Dairying in Prehistoric Britain’, Proceedings of the National Academy of Sciences, 100:4, 2003, 1524–9; Alan K. Outram et al., ‘The Earliest Horse Harnessing and Milking’, Science, 323:5919, 2009, 1332–5. 53 Richard P. Evershed et al., ‘Identification of Animal Fats via Compound Specific δ13C Values of Individual Fatty Acids: Assessments of Results for Reference Fats and Lipid Extracts of Archaeological Pottery Vessels’, Documenta Praehistorica 29, 2002, 73–96. 54 Elizabeth Craig-Atkins, Ben Jervis, Allie Taylor, Helen Whelton, Sandra Nederbragt, Lucy Cramp and Richard Madgwick, ‘The Dietary Impact of the Norman Conquest: A Multi-Proxy Archaeological Investigation of Medieval Oxford’, in preparation.



New Archaeologies of the Norman Conquest

95

Figure 2: Temporal changes in age profiles for A) cattle and B) sheep from medieval rural assemblages

At the late Saxon manorial – and potentially ecclesiastical – site of Flixborough (Lincolnshire), a slightly different pattern was observed. Dairy-dominated ceramics were more common in the eighth- and ninth-century wares, whereas the ninth- and tenth-century ceramics overwhelmingly revealed mixed cattle/sheep dairy and body fats, suggesting that methods of food preparation and cooking at the site were already changing by this time. One particularly important finding was that chicken fats were identified in Flixborough pottery, which also tallies with the zooarchaeological record, in that very large quantities of chickens were recovered from this site.55 Indeed, it would seem that a high frequency of chickens is a feature of late Saxon ecclesiastical assemblages, as a similar emphasis has also been noted at other major sites, such as Lyminge (Kent).56 Figure 1 highlights that increased frequencies of chickens are first observable in assemblages from religious houses dating to the mid-ninth to mid-eleventh centuries. Fothergill et al. have argued that this period witnessed the first attempts to actively manage high levels of chicken production, a Andre C. Colonese et al., ‘The Identification of Poultry Processing in Archaeological Ceramic Vessels Using In-Situ Isotope References for Organic Residue Analysis’, Journal of Archaeological Science 78, 2017, 179–92. 56 B. Tyr Fothergill, Julia Best, Alison Foster and Beatrice Demarchi, ‘Hens, Health and Husbandry: Integrated Approaches to Past Poultry-Keeping in England’, Open Quaternary 3, 2017, 1–25. 55

96

Aleksandra McClain and Naomi Sykes

situation which may even be responsible for the shifts in chicken genetics which others have been able to date to approximately ad 1000. It may be that religious beliefs, in particular the idea that chickens and eggs could be eaten during periods of fasting, may have prompted the move to intensive poultry farming.57 Similar arguments have been proposed to explain the ‘fish event horizon’, which saw large-scale marine fishing taking place across northern Europe, also around ad 1000.58 The widespread rise in commercial fishing may also have been instigated by monasteries, where the first evidence for large-scale marine fishing is found.59 In this way, neither an increase in fish consumption in the eleventh century, nor the increase in chicken frequencies seen on elite sites, can be seen as purely ‘Norman’ innovations; both trends were already present and developing independently in pre-Conquest England. That does not mean, however, that these trends were not influenced or accelerated by the Conquest. Given the Norman proclivity for ostentatious expressions of religiosity, most often seen through church building and monastic patronage, the post-Conquest secular elite may well have adopted the frequent consumption of both fish and chicken as a way to overtly demonstrate their piety. With this in mind, it is noteworthy that in the Norman dining scene on the Bayeux Tapestry, the assembled company are sitting down to a meal of fish and spit-roasted chickens (Fig. 3).

Figure 3: Scene from the Bayeux Tapestry, eleventh century, showing the Norman contingent cooking and dining; chickens, along with other meats, are being prepared on spits over the fire, and fish have been served at Bishop Odo of Bayeux’s table The deployment of pre-existing symbols of power and authority may be equally pertinent with regards to the fallow deer. Dama International’s research, combining zooarchaeology and biomolecular analysis (isotopes, genetics and radiocarbon dating), has demonstrated that fallow deer were introduced not from Sicily, which possessed fallow deer genetically different from those which are found in England,

57

Fothergill et al. ‘Hens, Health and Husbandry’, 21; For dating evidence of genetic change, see L. Loog et al., ‘Inferring Allele Frequency Trajectories from Ancient DNA Indicates that Selection on a Chicken Gene Coincided with Changes in Medieval Husbandry Practices’, Molecular Biology and Evolution, 34, 2017, 1981–90. 58 James H. Barrett, A. M. Locker and C. M. Roberts, ‘Dark Age Economics Revisited: The English Fish Bone Evidence AD 600–1600’, Antiquity 78, 2004, 618–36; James H. Barrett, ‘Medieval Sea Fishing, AD 500–1500: Chronology, Causes and Consequences’ in Cod and Herring: The Archaeology and History of Medieval Sea Fishing, ed. James H. Barrett and David C. Orton Oxford: Oxbow, 2016, 250–72; David Orton, James Morris and Alan Pipe., ‘Catch Per Unit Research Effort: Sampling Intensity, Chronological Uncertainty, and the Onset of Marine Fish Consumption in Historic London’, Open Quaternary 3, 2017, 1–20. 59 R. Reynolds, ‘Food for the Soul: The Dynamics of Fishing and Fish for Consumption in Anglo-Saxon England: c. A.D. 410–1066’, unpublished PhD thesis, University of Nottingham 2015.



New Archaeologies of the Norman Conquest

97

but rather the eastern Mediterranean. Furthermore, their programme of radiocarbon dating indicated that fallow deer were likely present in Britain prior to the Norman Conquest. The earliest dated specimen suggests an introduction date of around ad 1000, which may also help us to date the first deer parks, which we know existed in some numbers prior to the Conquest.60 Together, the findings from these independent studies all point to c. 1000 as being a more significant date than 1066 for the origins of particular trends in animal and food cultures that have often been associated with the coming of the Normans to England, a chronology that chimes with interpretations from continental Europe. Despite the lack of alignment with 1066, the evidence nevertheless offers significant insight into the values and priorities of the Normans, the cultures they chose to advance in Anglo-Norman England, and the social motivations behind them. Findings such as these demonstrate the advantages of unshackling archaeological interpretations from the confines of the grand ‘Norman Conquest’ narrative, enabling us to deconstruct assumptions about what the Conquest involved, improve our understanding of Norman innovation versus acceleration in terms of materials and behaviours, and to tie Normandy and England into their wider European contexts. New Agendas for the Archaeology of the Eleventh and Twelfth Centuries A number of themes highlighting the potentially unique contributions of archaeology to our understanding of the Norman Conquest specifically, and the eleventh and twelfth centuries generally, recur in the research which has been carried out in the early twenty-first century, and especially the individual papers in The Archaeology of the Eleventh Century, which has gathered together the most recent work by archaeological scholars publishing on the subject. To conclude, we shall draw out some of these themes to demonstrate how they may feed into future research agendas. A focus throughout most recent archaeological research is the ability of material evidence to access the experience of the total population during the Anglo-Norman transition, or at least a wider segment of it than is available through documents alone. The effects of the Conquest on the peasant communities of rural settlements have been discussed alongside the choices surrounding settlement and manorial organization that were made by the Anglo-Saxon and Norman elite, as have the impacts of Norman urban reorganization of Anglo-Saxon towns on the considerable numbers of people who continued to live and work in those towns through the transition period.61 Other contributions have highlighted practices such as food choices and dress accessories on multiple levels of society, particularly in urban areas, illustrating how market forces and cultural changes began to create post-Conquest cultures where the difference between urban and rural may have mattered far more than the difference between Saxon and Norman.62 The potential for archaeology to

60

Robert Liddiard, ‘The Deer Parks of Domesday Book’, Landscapes 4, 2003, 4–23. Oliver Creighton and Stephen Rippon, ‘Conquest, Colonisation and the Countryside: Archaeology and the mid-11th to mid-12th-Century Rural Landscape’, in The Archaeology of the Eleventh Century, 57–87; Letty Ten Harkel, ‘The Norman Conquest and its Impact on Late Anglo-Saxon Towns’, in The Archaeology of the Eleventh Century, 14–29. 62 Ben Jervis, Fiona Whelan and Alexandra Livarda, ‘Cuisine and Conquest: Interdisciplinary Perspectives on Food, Continuity and Change in 11th-Century England and Beyond’, in The Archaeology of the Eleventh Century, 244–62; Jervis, ‘Conquest, Ceramics, Continuity and Change’; Weetch, ‘Tradition and Innovation’. 61

98

Aleksandra McClain and Naomi Sykes

draw out the rise of the elusive ‘middling sort’ of medieval society may be particularly pronounced in research on urban small finds and the built environment. In all of these cases, however, it is clear that we have just begun to scratch the surface of archaeology’s potential with non-elite communities. Some of the most interesting new data in the future may come from osteological data and the scientific techniques which we can apply to it. Skeletal evidence from burial excavations has given us access to the ‘everyday people’ of urban and rural parochial communities, but also unparalleled insight into unusual and marginalized groups – for instance, infants, criminals and even lepers – and how their lives and treatment in death could be affected by such a sociocultural change as the Norman Conquest.63 Even within the elite strata of society, however, there is room for significant contributions from archaeology. A number of articles have addressed the complex array of manorial lords and lesser elite who flourished in late Saxon and AngloNorman England, and this promises to be a rich vein of enquiry for understanding the Conquest. Unlike the royal and baronial ranks, these minor elite do not as often appear in the documentation of the eleventh and twelfth centuries, yet they were equally important players in the process of Conquest, as they were in daily contact with the materials and realities of manorial and village life, interacted frequently with those both above and below them in the social hierarchy, and made day-to-day choices ‘on the ground’ which drove post-Conquest continuities and transformations.64 In both rural and urban areas, and in elite and non-elite segments of society, future archaeology in the period will benefit from a broadening of the types of material culture that we consider to be relevant to questions about the Norman Conquest. It is not only such obviously ‘Norman’ things as castles and Romanesque architecture that can be harnessed to tell stories about the Anglo-Norman transition, but also more mundane things, such as coins, pottery, undecorated fonts and uninscribed funerary monuments. The key to exploiting this evidence is the better use of more systematic and comparative methods of analysis, deploying large-scale databases, mapping and spatial analysis, and other digital humanities technologies which allow us to take advantage of the large datasets we are capable of collecting, and to characterize them within temporal and spatial contexts. A number of recent articles focusing on the Conquest exhibit the influence of advances in the use of social theory in archaeology, particularly surrounding the concept of identity, but also considering themes of agency and materiality. They highlight the role of material culture in constructing identities in a period of flux, a theme which echoes earlier work by Naomi Sykes with Norman animal, food and hunting cultures, and by Robert Liddiard on the elements of the ‘Norman package’ found in castle building and elite landscape design.65 Unsurprisingly, the ethnic

63

Craig-Atkins, ‘Seeking “Norman Burials”’; Simon Mays et al., The Churchyard, Wharram: A Study of Settlement on the Yorkshire Wolds XI, York 2009; Simon Roffey, ‘Charity and Conquest: Leprosaria in Early Norman England’, in The Archaeology of the Eleventh Century, 159–76. 64 Mark Gardiner, ‘Manorial Farmsteads and the Expression of Lordship Before and After the Norman Conquest’, in The Archaeology of the Eleventh Century, 88–103; Michael G. Shapland, ‘Anglo-Saxon Towers of Lordship and the Origins of the Castle in England’, in The Archaeology of the Eleventh Century, 104–19; McClain, ‘Rewriting the Narrative’, 224 and see also Aleksandra McClain, ‘Patronage in Transition: Lordship, Churches, and Funerary Monuments in Anglo-Norman England’, in Churches and Social Power in Early Medieval Europe, 185–225 at 215. 65 Robert Liddiard, Landscapes of Lordship: Norman Castles and the Countryside in Medieval Norfolk, 1066–1200, BAR British Series 309, Oxford 2000; Sykes, ‘Zooarchaeology of the Norman Conquest’, and Sykes, The Norman Conquest.



New Archaeologies of the Norman Conquest

99

and cultural identities of Normans and natives have received attention, but so too do cross-cultural ‘spatial’ identities which derived from the particular settlement or regional context in which an individual lived, as well as lordly identities of status and power, all of which could supersede the ethnic identities which it is often assumed would take precedence in an environment of foreign conquest.66 The evidence suggests that cultural allegiances did not always align in a straightforward manner with post-Conquest priorities, as there are instances where Normans as well as natives saw particular value in maintaining continuity from before 1066, and strategies of assimilation, adaptation and accommodation became as key to Norman success as forcefully imposed cultural change.67 For the native population, it is usually assumed that the imposed changes wrought by the Conquest were felt as detrimental, for example in the confiscation of land and destruction of housing stock for urban castle building, as seen in York, Norwich and Lincoln, or the frequent sweeping away of old buildings and their familiar architectural styles in favour of alien ones.68 However, castles, churches, markets and other elements of Norman infrastructure may not always have been perceived negatively by the native population. They did fulfil military, economic and religious functions required by the Norman ruling classes, but they also may have been seen by a settlement’s inhabitants as improving services and facilities and building a sense of community around shared, updated amenities. They could thus be a means of practically and symbolically enforcing the Conquest while also eroding opposition to it.69 In the rural milieu, some native landowners who survived the Conquest took advantage of opportunities to acquire additional tenancies, construct new manor houses, and found or rebuild churches. In these instances, the Conquest was not a negative, but rather offered tools of advancement for those seeking to establish themselves and thrive within the new hierarchy.70 This body of research highlights the complexity of identity and its relationship with material culture in the period around the Norman Conquest, as well as its further potential. For example, although archaeologists have explored material culture in relation to masculine and feminine gender identities extensively in later medieval periods, the varying experiences of both Norman and native men and women during the Conquest have so far not been considered in our assessments of the Anglo-Norman transition. Any one person could hold various identities, which were multivalent and malleable dependent on

66

Jervis, ‘Conquest, Ceramics, Continuity and Change’ and Jervis et al., ‘Cuisine and Conquest’ consider ethnic identities. Weetch, ‘Tradition and Innovation’ focuses on urban identities. McClain ‘Rewriting the Narrative’ and ‘Local Churches and the Conquest of the North’ highlight regional identities. McClain, ‘Rewriting the Narrative’, 219; Lilley, ‘Urban Landscapes’, 18–19; Shapland, ‘Anglo-Saxon Towers of Lordship’, 104; Gardiner, ‘Manorial Farmsteads’, 97–100, and Liddiard, Landscapes of Lordship, all comment on lordly identities and social competition. 67 Michael Lewis, ‘The Bayeux Tapestry: Window to a World of Continuity and Change’, in The Archaeology of the Eleventh Century, 228–43 at 236 and 240 for the preservation of English artistic styles in the tapestry and Norman assimilation strategies; McClain, ‘Rewriting the Narrative’, 223 for hybrid styles in transitional grave monuments; Ten Harkel, ‘The Norman Conquest and its Impact’, 23 for the maintenance of pre-Conquest moneyers and minting practices; Fradley, ‘Scars on the Townscape’ 135 for the reuse of English systems of civic governance and their physical remnants by urban castle-builders. 68 Keith D. Lilley, ‘The Norman Conquest and its Influences on Urban Landscapes’, in The Archaeology of the Eleventh Century, 30–56 at 35; Ten Harkel, ‘The Norman Conquest and its Impact’, 27. David Stocker and Paul Everson, ‘Archaeology and Archiepiscopal Reform: Greater Churches in York Diocese in the 11th Century’, in The Archaeology of the Eleventh Century, 177–203 at 196. 69 Ten Harkel, ‘The Norman Conquest and its Impact’, 27; See also Creighton and Rippon’s ‘What have the Normans ever done for us?’ commentary in ‘Conquest, Colonisation and the Countryside’, 57–8. 70 McClain, ‘Rewriting the Narrative’, 217.

100

Aleksandra McClain and Naomi Sykes

circumstance, and different individuals and groups had distinct needs, motivations and roles to play in the post-Conquest social world, all of which could shape their particular responses to it.71 Future research directions must therefore recognize that both Normans and natives had agency during the Conquest and subsequent transition, and should seek to understand how various individuals and groups experienced and knowledgeably utilized objects, buildings and the landscapes they inhabited to deal with a changing society. Two further emerging themes have the potential to allow us to reconfigure our understanding of the Norman Conquest in time and space. In 2005, Robert Liddiard, following models put forward in documentary history, insisted on the necessity of understanding the Conquest not as an event, but as a much longer, more complex process. He advocated for archaeology to adopt the concept of a ‘long’ Norman Conquest, disaggregating it into the relatively swift military and political takeover, and the much longer period of settlement and sociocultural and material adaptation which lasted well into the twelfth century.72 The 2017 articles by Oliver Creighton and Stephen Rippon on the Anglo-Norman countryside and Mark Gardiner on manorial development both take this perspective, examining evidence through to the mid-twelfth century in order to highlight traditions of settlement reorganization and manorial building which began around or even before the Norman Conquest and continued for over a century after it, influenced not only by the Conquest itself, but also by other changing social and economic forces.73 Linked to this reconceptualization of the Conquest as a contingent process instead of a monolithic event is the necessity of understanding regional differences and variable responses to the Conquest based on pre-existing regional and local circumstances, material traditions, and populations.74 Rather than speaking of ‘the Norman Conquest’ and attempting to characterize its effects on England holistically, we would do much better to understand it as a multiplicity of disparate and complex ‘Norman Conquests’, which could vary substantially in trajectories and outcomes dependent on context.75 The above research clearly demonstrates that much of our longstanding perception of the Conquest’s archaeological ‘invisibility’ has resulted not from inherent shortcomings in the evidence, but rather from us not asking the right sort of questions, nor thinking sufficiently critically about what the material manifestations of political and cultural change might look like, and the timeframes over which they might take shape. The Norman Conquest is only invisible if we expect to see rapid, responsive shifts across all cultural and material practice, regardless of the medium, its role in society, and the motivation for or necessity of change in that medium within the specific context of the Conquest and the subsequent transition. It is a given that in any period of transition, some things changed while others apparently did not, and archaeology has in the past two decades become adept at cataloguing 71

Aleksandra McClain, ‘The Archaeology of Transition: Rethinking Material Culture and Social Change’, in The Art, Literature and Material Culture of the Medieval World: Transition, Transformation and Taxonomy, ed. Meg Boulton, Jane Hawkes and Melissa Herman, Dublin 2015, 22–41 at 40–1; Gardiner, ‘Manorial Farmsteads’, 99; Jervis et al., ‘Cuisine and Conquest’, 259. 72 Robert Liddiard, Castles in Context: A Social History of Fortification in England and Wales, 1066– 1500, Macclesfield 2005, 14. 73 Creighton and Rippon, ‘Conquest, Colonisation and the Countryside’, 78; Gardiner, ‘Manorial Farmsteads’, 97. 74 The theme is touched on in urban environments in Lilley, ‘The Norman Conquest and its Influences’ and Ten Harkel, ‘The Norman Conquest and its Impact’, and in rural environments in Creighton and Rippon, ‘Conquest, Colonisation and the Countryside’ and McClain, ‘Rewriting the Narrative’. 75 McClain, ‘Rewriting the narrative’, 204.



New Archaeologies of the Norman Conquest

101

these instances of continuity and change on either side of 1066. However, acknowledging the coexistence of continuity and change is no longer innovative, and it is now incumbent on us to move beyond this overly simplistic model. Instead, we should ask questions about which particular things changed, which stayed the same, and why. We must also enquire about the forces and agents which drove continuity or change, and recognize that a wide range of knowledgeable actors, both Normans and native, could play these roles.76 What motivations lay behind their choices? Who benefitted from these choices? Which audiences were targeted with particular actions or materials? Did specific changes and continuities affect different groups of people in varying ways? We must even acknowledge that the concept of ‘continuity’ in a sociocultural transition may be somewhat misleading. For example, recent research has demonstrated that substantial changes in foodstuffs and cooking techniques following the Conquest did not necessitate changes in the style or form of the pottery vessels in which food preparation was carried out, and quite distinctive changes in the styles and forms of commemorative markers did not necessarily mark fundamental alterations in burial practices, the class of the patrons in question, or the religious and social purposes they were intended to serve.77 Even something which ‘stayed the same’ may have done so within a radically different social context, so its role, audience, significance and connotations might have shifted profoundly despite superficial impressions of consistency. On the whole, for archaeology to move forward in the eleventh and twelfth centuries, the ways in which we conceptualize the relationship between material culture and sociopolitical change must become more nuanced. It is a certainty that the Norman Conquest happened – archaeologists do not need to ‘find’ it in the material record in order to prove it, or to validate their worth to scholarship on the period. Instead, archaeology’s task is to reveal the wide range of ways in which people at all levels of society engaged with a hugely varied physical world during what was a long and multifaceted process of transition, and most importantly, why they might have done so in the ways that they did, considering what was going on around them.78

76

McClain, ‘The Archaeology of Transition’, 23, 40. Jervis, ‘Conquest, Ceramics, Continuity and Change’, 478; McClain, ‘Local Churches and the Conquest of the North: Elite Patronage and Identity in Saxo-Norman Northumbria’, in Early Medieval Northumbria: Kingdoms and Communities, AD 450–1100, ed. Sam Turner and David Petts, Turnhout 2011, 151–78 at 178. 78 This article was supported by the AHRC-funded research network Archaeologies of the Norman Conquest (AH/P006841/1). The authors also wish to thank the Dama International project (AH/I026456/1), the Cultural and Scientific Perceptions of Human-Chicken Interactions project (AH/L006979/1), and the Oxford Diet Project for their input, and access to unpublished data. 77

Map: Main places mentioned in and around Amboise and in Anjou

AN ANGEVIN IMPERIAL CONTEXT FOR THE AMBOISE–ANJOU NARRATIVE PROGRAMME Nicholas L. Paul Among the thousands of books of uncertain provenance housed in the collection of Latin manuscripts in the Bibliothèque Nationale de France in Paris are two twelfthcentury codices, MSS Latin 6218 and 6006, which contain the same sequence of three historical narratives. These works are entitled in their rubrics Liber de compositione castri Ambaziae et ipsius dominorum gesta (hereafter Liber), Chronica de gestis consulum Andegavorum (hereafter Chronica), and Gesta Ambaziensium dominorum (hereafter Gesta). All three works concern the history of events and people in the Loire valley, particularly the chief ruling dynasties of the county of Anjou and the fortified settlement of Amboise (Indre-et-Loire, arr. Loches), strategically located on the left bank of the Loire.1 The first of the three, the Liber, begins with the history and geography of the Touraine region, specifically the land around Amboise, from the time of Julius Caesar through the rise of the Franks and ultimately until the reign of the Capetian king Louis VII of France (r. 1137–80). The second narrates the dynasty of the Angevin counts from their legendary progenitor, the Breton forester Torquatius, offering a series of biographies of each count until the time of Count Geoffrey le Bel, count of Anjou and duke of Normandy (r. 1129–51). The third text begins the dynastic history of the Amboise dynasty in Hugh, a godson of Hugh Capet whose son was the first to gain control of Amboise in the time of Count Fulk III Nerra of Anjou (d. 1040), and tracks the family through a further three generations until the period after the death of the lord Sulpice II in 1153. The three works have been edited twice since the middle of the nineteenth century, once in 1856 by Paul Marchegay and André Salmon (republished with an introduction by Émile Mabille in 1871) and a second time in 1913 by Louis Halphen and Renée Poupardin, based on a re-evaluation of the texts and manuscripts published by Halphen in 1906.2 The stories that the Anjou and Amboise texts tell have been widely cited within historical scholarship on western Francia in the eleventh and twelfth centuries, and popularized by the writings of Georges Duby, who relied heavily upon them for details of twelfth-century life, the experiences of women, and the cultures of historical writing in high medieval ‘France’.3 1

For an historical overview of the region with separate thematic discussions see Jean-Michel Matz and Noël-Yves Tonnerre, L’Anjou des princes: Fin ixe-fin xve siècle, Paris 2017, 21–218. The basic narrative of Angevin history has long been stood as an exemplar of feudal France, see R. W. Southern, The Making of the Middle Ages, New Haven 1953, 81–20; Thomas Bisson, The Crisis of the Twelfth Century: Power, Lordship, and the Origins of the European Government, Princeton 2009, 129–42. 2 Chroniques des comtes d’Anjou, ed. Paul Marchegay and André Salmon with an introduction by Émile Mabille, Paris 1871; Louis Halphen, Etude sur les Chroniques des comtes d’Anjou et des seigneurs d’Amboise, Paris 1906; Chroniques des comtes d’Anjou et des seigneures d’Amboise, ed. Louis Halphen and Renée Poupardin, Paris 1913. 3 Duby, Le chevalier, la femme, et le prêtre, Paris 1981, tr. Barbara Bray as The Knight, the Lady, and the Priest: The Making of Modern Marriage in Medieval France, Chicago 1983, 227–52; idem, Dames

104

Nicholas L. Paul

Despite their modern popularity, however, the precise relationship between the three narratives is still not fully understood. Perhaps as a result, they have received almost no attention as works of historical literature, and have occupied a peripheral position in studies of historical writing from the Plantagenet domains. In part, the problem would seem to be the desire to treat them as separate works, the products of distinct, individual authors, when in fact much evidence suggests that they should be read as a coherent whole or as a deliberately plotted sequence. It is the purpose of this article to reconsider these works as part of one historical programme, just as a medieval reader encountered them in the two surviving twelfth-century manuscripts. When read together, it becomes clear that this programme emerged from and responded to both local events and an affair of global significance: the birth of the Angevin empire. By establishing the likely context for the assembly of these works we can, in turn, begin to understand their place within the larger tradition of historical writing in the Angevin world. In order to see how and why these texts should be read together, however, it is necessary first to see how it was that they came apart. The Deconstruction of the Amboise and Anjou Histories Apart from markings and bindings corresponding to the collection of Colbert, nothing is recorded of the earlier provenance of the twelfth-century Latin manuscripts 6218 (hereafter MS A) and 6006 (hereafter MS B). The pricking and ruling of the manuscripts, the general approach to the writing space, the quality of the parchment and the consistency of the scripts would suggest that the two were produced at different scriptoria, but the programme of writing in both manuscripts is nonetheless similar: the second manuscript clearly imitated the first: it has nearly identical rubrics, and coloured initials highlight the openings of the same passages. Even though the quantity of text in the two manuscripts is quite different, in both manuscripts the graphic space has been managed and the rubrics and initials added so that the transitions between works and sections has an almost identical visual effect.4 That the two manuscripts are indeed directly related to one another was demonstrated by Louis Halphen in 1906. In a masterful study, Halphen showed that markings, corrections and marginal notations in MS A correspond precisely to what we find in the main text of MS B.5 Substantial additions were made to the texts in MS B which suggest that the programme was moving (if it had not already moved completely) into the orbit of the great Benedictine abbey of Marmoutier near Tours.6 A second set of markings and annotations, appearing this time in MS B, correspond precisely to the greatly enlarged version of the Anjou dynastic history written by a monk of Marmoutier named John in 1172–3.7 du XIIe siècle tome 2: Le souvenir des aieules, tr. Jean Birrell as Women of the Twelfth Century 2: Remembering the Dead, tr. Jean Birrell, Chicago 1997, passim, esp. 125, 144–5; idem, ‘Remarques sur la littérature généalogique en France aux XIe et XIIe siècles’, Comptes rendus des séances de l’Académie des Inscriptions et Belles Lettres, 1967, 335–45 at 337–40. 4 Compare, for instance, MS A ff. 14v–15r with MS B f. 8v–9r (for the transition between Liber and Chronica) and MS A 34r, with MS B 33r (for the transition between Chronica and Gesta). 5 Halphen, Etude, 6–11 and ‘Introduction’, in Chroniques, ix–xiv. 6 The influence of Marmoutier is clear in the additions entitled (by Halphen and Poupardin) ‘The Piety of Fulk the Good’, in Chroniques, 140–3 and ‘Geoffrey the Bearded and the Monks of Marmoutier’ in Chroniques, 152–5. 7 The date of John of Marmoutier’s composition is established by his comments about the state of political affairs in Henry II’s lands (see Chroniques, 171). For the argument that John intended his text



The Amboise–Anjou Narrative Programme

105

In their respective introductions to the three texts, Mabille and Halphen offered explanations about the texts’ authorship and constructed their editions based on their conclusions about the relationship between these two manuscripts and others bearing witness to the textual tradition. Both men were guided by what appears to be key evidence in the prologue of the version of the monk John of Marmoutier. Addressing King Henry II in the preface to his own version of the Anjou dynastic history, John explained that: In case either you or other hearers or readers of this history have occasion to doubt that I might leave something out in these things that I wrote, I will take a moment to make known from which writers I learned these things … The first writer was Thomas of Loches, who discovered brief chronicles written under the name of Abbot Eudes and added much from what he had heard from his mouth and what he knew because it was widely reported. Second was Robin and Brito of Amboise who improved those chronicles and added that which they had heard said in their presence. I am the third, and I added much from many histories … [he goes on to name his other sources].8

With the words of John of Marmoutier ringing in our ears, the temptation has always been for modern historians to try to detect the various phases of composition and to associate them with the names provided by John. The temptation is made all the stronger by two additional, tantalizing factors. First, we do hear several distinct voices in the texts: the prologue and the main outline of the Anjou dynastic history would seem to have been written in around 1106–1109, late in the life of the troubled Count Fulk IV le Réchin (r. 1069–1109).9 A hastily assembled continuation, taking the narrative up to the reign of Geoffrey IV le Bel of Anjou, was evidently written at the time of the latter’s death in 1151. The authorial voice which surfaces toward the end of the Liber speaks sorrowfully of the failure of the Second Crusade, alluding to its defeat before the walls of Damascus in 1148 almost as if the wound were still fresh.10 Eleanor of Aquitaine and Louis VII of France are described as married, and there is no reference to their separation in 1152. The authorial voice in the Gesta, by contrast, opens the work by lamenting news of the capture of Sulpice, lord of Amboise, in 1153, and in the closing chapters announces the news his death later in the same year, as if the author had been writing through the family’s ordeal.11 But that same text mentions the succession of Henry II in England and the death of the lady of Amboise, Elizabeth of Jaligny, in 1154.12 The second factor inviting us to follow John of Marmoutier’s lead is that most if not all of the writers he listed can be identified with men resident in Anjou and the Touraine: ‘Abbot Eudes’ with Eudes I of Marmoutier (r. 1124–37); ‘Thomas of Loches’ with one of the men named Thomas who acted as chaplain and notary of for presentation at Henry’s Christmas court at Chinon in 1172, see Nicholas L. Paul, To Follow in Their Footsteps: The Crusades and Family Memory in the High Middle Ages, Ithaca 2012, 240. 8 Chroniques, 163–4: Ut autem in his que scripsimus vel tibi vel ceteris auditoribus sive lectoribus hujus historie occasionem dubitandi substraham, quibus hec auctoribus didicerim breviter intimare curabo … Primus scriptor extitit Thomas Luchensis, qui breves cronicas nomine Odonis abbatis intitulatas, ut ab eius ore audivi, repperit et multa, que fama vulgante cognovit, addidit. Secundus extitit Robinus et Brito Ambaziacensis, qui ipsas cronicas emendaverunt et quedam, ut viva voce ab ipsis audivi, addiderunt. Tertius ego ex multis historiis multa addidi. 9 For the context of the earliest version of the Anjou Chronica see Nicholas L. Paul, ‘Origo Consulum: Rumours of Murder, a Crisis of Lordship, and the Legendary Origins of the Counts of Anjou’, French History 29:2, 2015, 139–160. 10 Chroniques, 24. 11 Chroniques, 74–5, 129–30. 12 Chroniques, 131.

106

Nicholas L. Paul

two successive counts of Anjou; and ‘Brito of Amboise’ with a canon of St Florentin of Amboise by that name active in 1141.13 The temptation to identify the two Paris manuscripts with these named and historically verifiable individuals has proven irresistible. Émile Mabille identified the version in MS A as the work of Abbot Eudes, although this required the existence of an anonymous continuator, since the text of the Anjou Chronica in that manuscript continues through the death of Geoffrey the Fair (1151). Mabille identified a lost manuscript copied by André Duchesne with the work of Thomas of Loches, and MS B with Robin and Brito of Amboise.14 Halphen, in the 1914 edition, rejected this sequence, proposing his own in which MS A was the work of Thomas of Loches, revised by Robin, and MS B was the work of Brito. The markings and annotations in MS B must, to Halphen, logically be the hand of John of Marmoutier working to prepare his own recension.15 While that last identification – of the markings in MS B with John of Marmoutier – seems beyond question, the precise roles of Eudes, Thomas, Robin and Brito cannot be confirmed by anything within the manuscripts themselves. It is, moreover, unclear if John, who was working across all three of the texts in MS B, was naming the contributors only to the Anjou Chronica, or to the whole textual programme that he consulted in MS B. While there can be no doubt about the value of the kind of information provided by John with regard to his historical process, there is also no doubt about the limits that it has placed on the inquiries posed regarding these texts. As if in a centrifuge, the body of work in our manuscripts is pulled apart – first to separate the Anjou Chronica from the Amboise Liber and Gesta, and then to reconstruct the hypothetical preceding versions of the Chronica (assigned to the named authors). Always medial, manuscripts A and B are valued only as witnesses of a tradition in motion from textes primitifs to the version of the Anjou Chronica addressed to King Henry II himself. But the sequence we find in MSS A and B comprises a distinct project, one that was unrelated to John and his objectives in 1172–3. The three components of this sequence almost certainly did have separate origins, but at the moment we encounter them, they are already deeply interwoven. The prologue to the Anjou Chronica, for instance, refers to a preface, now missing, described as ‘the foregoing exposition regarding the kings of the Franks’.16 That work would seem to have been removed and replaced with the Amboise Liber. But the Amboise Liber itself seems to have been changed, as it somewhat awkwardly incorporates a history of the kings of the Franks (sections rubricated ‘De Clodoveo’ and ‘De Karolo’), perhaps the same text or part of the same text that had originally prefaced the Anjou Chronica.17 The Amboise Gesta draws heavily upon the Angevin Chronica, using it as both a stylistic source and a narrative scaffolding through which to tell its story,18 but in the Angevin Chronica itself, Amboise and its lords occupy a suspiciously prominent 13

Chroniques, xxvii–xxxix. On the identity of Thomas of Loches, see Henri Martin, ‘Autour de Thomas Pacitius, prieur de la collégiale de Loches’, Bulletin Trimestriel de la Société Archéologique de Touraine 14, 1986, 389–95; Kathryn Dutton, ‘Geoffrey, Count of Anjou and Duke of Normandy, 1129–1151’, PhD dissertation, Glasgow 2011, 173–6. 14 Mabille, ‘Introduction’, pp. iv–xiv (for Eudes) pp. xiv–xxv (for Thomas) p. xxv–xxviii (for Robin and ‘Le Breton’). 15 Chroniques xx–xxvi. 16 Chronica, 25: Quoniam in ante expositis de regibus Francorum qui huic operi precedent maximeque sequenti necessaria esse puto explanavi. 17 Chronica, 13–24. 18 Instances of the Gesta borrowing from the Chronica identified by Halphen: Chroniques, 75, 77, 80, 82, 84, 85, 88, 90, 104, 105, 108, 131.



The Amboise–Anjou Narrative Programme

107

position in certain passages, so suspicious in fact that Halphen concluded that these passages could not be original to the Chronica itself and must be interpolations.19 It is, so to speak, a project of narrative reciprocity, in which the Anjou Chronica comes to owe a debt to the Amboise texts, which themselves relied upon the Anjou Chronica. The process of composition mimics the reciprocal relationship between lord and vassal which, as we will see, is also spelled out in these works. All three narratives rely heavily upon classical, patristic and classicizing sources for stylistic embellishment.20 The Liber and the Gesta both begin with quotations from the letters of Sidonius Appolinaris.21 Both the Chronica and the Gesta show a preference for Cicero, Sallust and Lucan. Neil Wright has also demonstrated a major reliance in both texts on Baudri of Bourgueil’s history of the First Crusade, Historia Ierosolimitana, which is used in the same allusive manner as the classical sources.22 So strong is the similarity between the use of stylistic sources in these three works that we can either conclude with Halphen that the allusions were inserted in both texts by the same reviser, or that the author of the Gesta deliberately imitated and sought out the same types of sources that had been used in the Chronica, which was clearly a major source for narrative content. In either case, the result is a programme of writing that is highly integrated in terms not only of its content but also its style. An Amboise–Anjou Narrative Programme Interpolators and scribes worked hard to present the Amboise–Anjou programme as a coherent whole. Even as the three component texts – the Liber, the Chronica and the Gesta – retained distinct narrative strategies, these distinct strategies actually complement an overall project with clear and consistent objectives. The first objective was to emphasize the value of Amboise itself, as a stronghold, a community and a lordship, and to illustrate the virtues of friendship between the lords of Amboise and the counts of Anjou. A second objective was to glorify the Anjou dynasty while at the same time denigrating and delegitimizing their enemies, the counts of Blois. Finally, the textual programme helped to contextualize and rationalize the dramatic crescendo of Angevin power that came with the accession of Henry II as king of England and duke of Aquitaine against deeper narratives of royal and imperial history. The first work in the sequence, the Liber de compositione castri Ambaziae, spans a period of over a millennium. Along the way, the author tells us that he is working primarily from three sources: a Gesta Romanorum, unidentified, but believed to be the kind used by other writers, such as Orderic Vitalis;23 a Historia Britonum which is either the work of Geoffrey of Monmouth or a work derived from it (possibly the First Variant or Wace’s Roman de Brut);24 and a Historia Francorum which is unidentified.25 Throughout, Amboise is invoked as a central, strong, reliable fortress, Instances of apparent interpolations into the Chronica from the Gesta, identified by Halphen: Chroniques, xxi n. 1. 20 Chroniques, lxvi–viii. 21 The Liber cites Sidonius, Epist. I, 2 (see Chroniques, 1); the Gesta invokes Sidonius, Epist. I, 6 and 7 (see Chroniques, 74). 22 Neil Wright, ‘Epic and Romance in the Chronicles of Anjou’, ANS 26, 2004, 177–89. 23 The Roman source is repeatedly called ‘Gesta’: Chroniques, 1, 6, 8, 19. For its relationship with other lost Gesta Romanorum see Halphen, ‘Introduction’, in Chroniques, lii and n. 2. 24 Chroniques, 10 and see Halphen, ‘Introduction’, in Chroniques, liii and n. 1. 25 Chroniques, 15; Halphen [‘Introduction’, in Chroniques, lv] offers several comparisons with surviving Frankish histories and genealogies that could have been sources or at least part of the same tradition with a source text. 19

108

Nicholas L. Paul

it is described as being fundamentally attached to Anjou, and it is associated not just with counts but with kings and conquerors too. Julius Caesar himself chooses the future site of Amboise (formerly Ambaquis) as the location for his winter camp as he subdues the surrounding tribes.26 Embracing an anachronistic view of political geography and ancient demography, the Liber notes that it was specifically from Amboise that Caesar was able to conquer ‘Anjou and Brittany [Andegaviam et Armoricamque], the cities of Tours and Le Mans, some through fear and others by alliance’.27 His conquests extend to Germania, to the Rhine and finally – ‘with his most beautiful fleet [pucherrima classe]’ – to Britain.28 The text stresses Amboise’s loyalty to imperial rule. Amboise remained in Roman hands from Caesar until Diocletian, at which time it was completely destroyed in the uprising of the Bagaudae.29 Even in ruins, it is loyal, providing the stone with which Tours is refortified. When Avicianus, ‘count of Tours’, decided to refortify the location, calling it by its modern name Ambazia, the memory of Caesar’s great fortress still hung over the site: Avicianus is said to have realized that ‘he could not repopulate that great town of Caesar with his own people’.30 After the defeat of the Roman army by Alaric – divine punishment for Rome’s sinful embrace of Arianism– the Liber shifts into an Arthurian storytelling framework in a section rubricated ‘Cronica de Arturo’. The Liber relies here on the work of Geoffrey of Monmouth, but it departs from Geoffrey and all other texts dependent upon him in important ways.31 In Geoffrey of Monmouth, Arthur’s conquests on the Continent and his victory over the Gallic leaders are followed by a triumphal gathering at Paris, where he distributed lordship over the provinces to members of his court. In the Liber, the story is roughly the same, except that the Liber pauses to relate that at that time ‘Clodius’ (possibly Clovis), king of the Franks, ruled the greater parts of Germany to Cambrai and Tournai, and that he was much beloved of Arthur.32 Where Geoffrey of Monmouth had written of Arthur’s gift of Anjou to his steward Kay, the Liber makes this a gift of Anjou and the Touraine (Andegaviam et Turoniam), associating them for the first time as if the two principalities naturally belonged together.33 It was for Kay, the Liber makes clear, that the castle of Chinon (Kainon) was named, adding also the detail from Geoffrey that Kay was buried at Chinon. Alluding to some sort of blood relationship between Kay and an earlier ruling dynasty of Tours, the Liber describes how Kay bestowed Amboise upon his cousin Billeius and imagines this greater lordship of Amboise as constituting the lands bounded by Chinon, the Indre river, and a wood at Clion. Amboise, from its ancient beginnings a strong, reliable seat of conquerors and emperors, a lordship created by the count of Anjou and Tours through the authority 26

Chroniques, 1–5. ‘Chroniques, 5: ab Ambaquis recedens, Andegaviam Armoricamque adgressus est regionem, urbe Turonica et Cenomannica, paitim federe, partim metu, adquisita. 28 Chroniques, 6. 29 Chroniques, 7. 30 Chroniques, 8: sciebat enim magnum oppidum Cesaris sua plebe impleri non posse. 31 For commentary on the Liber as a witness to the transmission of the Arthurian tradition, see Mildred Leake Day, Latin Arthurian Literature, Cambridge 2005, 17 and 34; Martin Aurell, ‘Henry II and Arthurian Legend’, in Henry II: New Interpretations, ed. Christopher Harper-Bill, Cambridge 2007, 362–94 at 383; idem, The Plantagenet Empire, 1154–1224, trans. David Crouch, Harlow 2007, 154. 32 Chroniques, 9–10. Cf. Geoffrey of Monmouth, The History of the Kings of Britain, ed. Michael D. Reeve and tr. Neil Wright, Woodbridge, 2007, 208–9. 33 Chroniques, 10. 27



The Amboise–Anjou Narrative Programme

109

King Arthur, is then juxtaposed in the Liber with the origins of the neighbouring county of Blois in a section rubricated ‘De Blesi Cronica’: Returning from battle together with Count Boso of Chartres, a certain young man from Brittany named Ivomadus who had with him a thousand men asked for a place in his county where he might stay. Deceiving [the count] with his lying and lisping talk, [Ivomadus] obtained a place to his pleasing on the banks of the Loire where he built not a villa but a most heavily fortified town, lest it be taken away by Boso or anyone. When, some time later, Boso saw it, he was angered: ‘I would not have conceded this to you, had I only remembered that wise saying told from father to son: ‘Beware fawning and lisping words,/ simplicity is the mark of truth and talkativeness fraudulent fictions’ [Cato, Distichs, III: 4]. Ivomadus, soothing his anger, offered him the castle with supplications and entreaties; but Boso, as he was a kind man, conceded the castle to him, accepting homage from him with oaths, and on account of the deception, he called it Blois (Blesim).34

The contrast with the foundation narrative of the castle-town of Amboise and (as we will see) its noble line could not be more striking. The inclusion of the story itself makes no narrative sense, as it does not follow from the section before it, and neither the Blois family nor Blois itself is mentioned again in the Liber. Its purpose seems only to denigrate and delegitimize Blois and its rulers. Where the story of Blois is stunted and disconnected, that of Amboise continues to flow, alongside the Loire, continuously through history, proceeding in its final sections through the Frankish royal history. This lightly annotated genealogy of the royal Frankish dynasties from the fifth up through the mid-twelfth centuries is still peppered with references to local history, and Amboise is once again emphatically described as an important royal stronghold, held directly by Frankish kings from Clovis until Charles the Bald. The text even seizes on the story, which first appears in Gregory of Tours, that it was at a location near Amboise that the Frankish commander Clovis and the Visigoth Alaric met and made peace.35 In the Liber, the Touraine is transformed into a rich realm of memory: one that explains the etymologies of place names, the conversion of pagan towns and sacred sites to Christianity, and the foundation and growth of settlements and fortifications. The linchpin of it all is Amboise itself. The place brings coherence to these narratives of triumph, calamity and supersession. The walls and the people are strong and reliable, but the stories of Amboise (and presumably those who tell the story) are also useful for the way they brought concordance to so many discordant narratives, rationalizing within a single framework Roman, barbarian, Arthurian, Merovingian, Carolingian and Capetian history. In the Liber, a sense of recursion and persistence prevails, enhanced by the periodic return to the geographical centre (Amboise) and by the reappearance of particular figures and lineages. Appropriately, when the Liber reaches the time of the author’s own life in the mid-twelfth century, it recounts a story that literally transcends time. Charlemagne’s squire ‘Johannes de Temporibus’ 34

Chroniques, 11: Ivomadus quidam juvenis de Britannia, secum habens mille viros, a prelio cum Bosone Carnotensi consule rediens, locum in comitatu suo ubi remaneret petiit. Qui blandis blesisque sermonibus eum decipiens, locum super ripas Ligeris ad libitum impetravit, ubi non villam sed oppidum firmissimum, ne a Bosone vel alio eriperetur, erexit. Quod cum diu post Boso aspiceret, iratus ait: “Hoc tibi non concessissem si verbum sapientis patris filio dictum memoriter retinuissem: sermones blandos blesosque vitare memento, simplicitas veri fama est, fraus ficta loquendi.” Ivomadus, iram eius mitigans, supplicando obtestandoque castrum optulit; sed Boso, ut erat benignus, hominium cum iureiurando ab eo suscipiens, castellum illud concessit et a deceptione Blesim vocavit. 35 Chroniques, 14. For the event in context see Guy Halsall, Barbarian Migrations and the Roman West, 376–568, Cambridge 2007, 298.

110

Nicholas L. Paul

(‘Timeless John’?), is said in the Liber to be still alive in the early twelfth century, having lived for 361 years.36 With the shift to the second text, Chronica de gestis consulum Andegavorum, our programme adopts the new narrative mode of dynastic history. Like other dynastic histories that were being written at roughly the same time in Barcelona, Normandy and Flanders, the Anjou Chronica charts the story of the dynasty’s rise and its struggle to acquire and defend its honores. A closing envoi makes explicit that the purpose of the work is didactic.37 In several ways, the transition from the Liber to the Chronica would seem jarring and inelegant. The Liber promised in its opening lines that it would relate the history of the lords of Amboise, but failed to do so, and ended with vague concerns about the failure of the French king on the Second Crusade. The Chronica begins by stepping back in time again, but now concentrates on a considerably shorter period: 300 years as compared to 1,100. The structure of the work is also different, with rubrics dividing not by periods or dynasties (as in the Liber) but instead by short biographical chapters devoted to individual princes. In important other ways, however, there is also continuity. The prologue refers explicitly to ‘the foregoing exposition of the kings of the Franks’, which does accurately describe the final two sections of the Liber that immediately preceded it.38 Moreover, the foreshortening of time means that the Anjou narrative is in some ways nested within the Amboise account: the Chronica begins its first chapter with a reference to the departure of the forester Torquatius from Armorica Gallica after the expulsion of the Armorican Bretons by the Roman governor Maximianus, an event chronicled in the Arthurian section of the Liber.39 Charles the Bald, the Carolingian king in whose service the dynasty begins its rise, had been the focus of considerable attention in the Carolingian portion of the Liber: a reader of the two texts in sequence would have known him from the Liber as the great king who travelled to Constantinople to defeat the Persians and drive back the Saracens, returning with many relics.40 He had accomplished, in other words, precisely what the Liber notes that the king of France had failed to do on the Second Crusade. His conquests and relic acquisitions would, however, resonate with the Chronica’s stories about the pilgrimages of Count Fulk III Nerra and the crusading exploits Count Fulk V, who became king of Jerusalem in 1131.41 Importantly, although the focus has shifted from the strategic location of Amboise to the lineage of the counts of Anjou, a reader continuing on from the Liber to the Chronica would be reassured to find that Amboise retains a place of importance in the narrative. In making an early claim on the dynasty’s right to the Touraine, for instance, the Chronica strikingly names Amboise as one of the original landholdings bestowed upon Ingelgarius, the ninth-century ancestor and earliest official predecessor of the counts of Anjou.42 The role of Amboise as a key to the Touraine and a bulwark against Blois is invoked early in the narrative, during the biography of ‘Count Maurice’ (a shadowy half-brother of Count Fulk III Nerra who is said to have

36

Chroniques, 24. Chroniques, 25. On the didactic nature of the Chronica and other dynastic histories, see Paul, To Follow, 65–9. 38 Chroniques, 25. 39 Chroniques, 9 (Liber), 26 (Chronica). The story to which both allude is given in Geoffrey of Monmouth, History of the Kings of Britain, 104–5. 40 Chroniques, 20. 41 Chroniques, 50–1 (Fulk III), 68–71 (Fulk V). 42 Chroniques, 30. 37



The Amboise–Anjou Narrative Programme

111

ruled briefly in the early eleventh century). This chapter and the next one, dedicated to Fulk Nerra, told the story of the rebellion of the lord Landric of Chateaudûn, who seized Amboise in the early eleventh century and treacherously allied himself with Blois.43 The Chronica tells of a pair of brothers who loyally defended Amboise against Landric, attacking him both from their own fortified house in Amboise and from the residence of the count (domicilio comitis).44 As the war between Anjou and Blois escalated, Amboise is the strategic location from which Fulk III Nerra rode to his great victory over Odo II of Blois at the battle of Pontlevoy in 1016, whence he returned with plunder and captives.45 Following this victory, Hersendis, the niece of one of the loyal defenders, is given in marriage by the count to his faithful knight, a Breton adventurer named Lisois of Bazouges. Lisois is the first of the line of the lords who still held Amboise at the time of the text’s composition in the mid-twelfth century, and it is striking that the Angevin Chronica goes out of its way to stipulate, with a strongly legalistic inflection, that he succeeded to the stone fortress together ‘with all its appurtenances’ (omnibus appendiciis eius).46 One of the most pivotal events in Angevin history was the victory of Count Geoffrey Martel of Anjou over Count Thibaud III of Blois at the battle of Nouy in 1044, leading to the latter’s capture and the surrender of Tours to the Angevins. Given the momentous nature of this battle, it is striking to find that the single most prominent individual in the Anjou Chronica’s account is not the count of Anjou himself, but Lisois, the lord of Amboise. It is Lisois, according to our text, who is said to warn the count of the movement of Thibaud’s troops, painstakingly collecting information so that he would know exactly the disposition of Thibaud’s forces before riding to Count Geoffrey and making a lengthy speech advising him to meet Thibaud in open battle rather than simply besieging Tours.47 The speech is one of the longest in the Chronica, and it depicts the lord of Amboise as the man of action and wisdom at a pivotal moment in Angevin history. Like other passages in the Chronica identified by Neil Wright, this speech was lifted directly from the crusade chronicle of Baudri of Bourgueil, and it may have been intended to invoke a specific episode in the story of the First Crusade.48 If the echo was meant to be palpable, then Lisois of Amboise is here envisioned as the crusading hero Raymond of St Gilles, arguing with the other leaders in January 1099 that they should leave off their siege of Jabala in favour of pushing onwards toward Jerusalem. The writers and editors who assembled the narratives of Amboise and Anjou could not avoid the fact that in recent memory relations between the two families had sunk into open hostility. Both the Chronica and the Gesta go to some lengths to establish that the first seeds of trouble had been sewn in the reign of Count Fulk IV le Réchin. As I have shown elsewhere, the story of Fulk IV’s reign was deeply troubling to writers in the region, but nowhere did he receive less sympathy than in the Amboise–Anjou narrative programme.49 These narratives recalled how Fulk had imprisoned his elder brother and was rumoured to have murdered his own eldest son, Geoffrey Martel II, in 1106. Prominent among his crimes in the Chronica (and 43

Chroniques, 52–3 (for the attacks on Blois from Amboise). Chroniques, 46. 45 Chroniques, 52–3. 46 Chroniques, 54. 47 Chroniques, 55–6. 48 For the speech, cf. The Historia Iherosolimitana of Baldric of Bourgueil, ed. Steven Biddlecombe, Woodbridge 2014, 98; Wright, ‘Epic and Romance’, 186. 49 Paul, ‘Origo consulum’, 139–44. 44

112

Nicholas L. Paul

the Gesta) is his dispossession and expulsion of the man tasked with guarding his domicile in Amboise.50 In the Chronica, the wounds inflicted in Fulk IV’s reign are mended by the triumphant careers of his son younger son, Fulk V, and grandson, Geoffrey le Bel. The chapters dedicated to them are essentially the stories of Angevin accession to Maine, Jerusalem and the Anglo-Norman realm, including the necessary genealogical information and details of succession. As the counts of Anjou triumph, however, and the damage of the past is repaired, the Chronica does not neglect to mention the implications for the lords of Amboise. As in the description of the siege of Nouy in 1044, the account of Fulk V’s accession to the county of Maine allows the lord of Amboise remarkable prominence: The distinguished man Fulk, powerful in arms, orthodox in his faith, benevolent to worshippers of God, having succeeded to the two counties, he lifted up his friends and pressed upon evil ones and his adversaries, in short he achieved glory and the highest fame equal to none. Receiving homage, he conceded all Amboise to Hugh of Chaumont, which had been given to him by [Fulk’s] brother [Geoffrey] Martel and returned Montrichard, which had been unfairly taken away from his ancestors.51

The wrongs of the reign of Fulk IV, no matter how disturbing, can be set right, and honour restored. This was, as we shall see, an important message to keep in mind as a reader moved between the Chronica and the Gesta. As part of the sequence of narratives in the two Paris manuscripts, the Gesta complements the material that precedes it in several ways. First, it continues the narrative of Amboise begun in the Liber, picking up where it left off, thematically, in the domain of Capetian history. The continuity between the two texts is so smooth that the rubricator of MS B even changed the first rubric of the text from ‘De Hugone’ to ‘De Hugone Capet’, making more explicit the continuation of the sequence of Frankish kings from the Liber.52 As in the transition from the Liber to the Chronica, the transition from the Chronica to the Gesta marks a shift in timescale, with the period under consideration now covering only 150 years, with the majority of the focus on the last fifty – the living memory of the author and much of the audience. Moving rapidly to the eleventh century and the life of Lisois of Bazouges, first ancestor of the lords of Amboise, the Gesta also uses the political narrative already provided in the Anjou Chronica as a kind of narrative scaffolding, paraphrasing and copying out its account of Angevin history and then expanding on the narrative at much greater length, providing many more details related either to Amboise or to the other Touraine lordships. Though some of the scaffolding is obviously repetitive, the total effect is complementarity. The account of the battle of Nouy in 1044, for example, draws on the Chronica. But the Gesta’s version of the battle eschews the Chronica’s classicizing epic tone and avoids the passages from Baudri of Bourgueil, providing new details and finding new ways to emphasize the Amboise contribution. Where the Chronica noted that Count Thibaud III of Blois was ultimately apprehended by men from Amboise, for instance, in the Gesta, this scene is expanded to explain that it was Lisois of Amboise himself who captured the count of Blois.53 50

Chroniques, 64. Chroniques, 67–8: Vir honestus Fulco, armis strenuus, fide catholicus, erga Dei cultores benivolus, adepto utroque consulatu, amicis exaltans, malignos et sibi adversarios opprimens, gloria et optima fama infra nulli in brevi effectus est. Qui Hugoni de Calvomonte Ambazium totum, a fratre eius Martello ei datum, accepto hominio concessit et ipsi Montricardum, antecessoribus suis olim iniuste ablatum, reddidit. 52 BnF MS Lat. 6006, fol. 33r. 53 Chroniques, 57 (for the Chronica account of the capture) and 84–5 (for the Gesta account). 51



The Amboise–Anjou Narrative Programme

113

The Gesta seems to be just as concerned with the implications of the Angevin victory for Amboise, for it is in the period of peace and restitution of rights following the battle that the Gesta places the foundation of the collegiate church of St Florentin in Amboise (a religious community thus founded on the cooperation between Anjou and Amboise).54 Another result was the eventual succession of Lisois’ son Sulpice not only to Amboise but also to the lordship of Chaumont-sur-Loire, a strategically important fief of the counts of Blois. The unification of Amboise with Chaumont-sur-Loire was a tremendous coup for the Angevins and their allies in the Touraine. Kimberly LoPrete has argued that the union of these two lordships, and the need for the Amboise family to demonstrate faithful service to both lords, led the author of the Gesta ‘to ton[e] down the triumphalistic rhetoric’ of the Angevin Chronica.55 While this may be true (and may explain the more muted treatment of the battle of Nouy in the Gesta), it is also the case that faithful service is one of the central themes of the entire Amboise– Anjou programme, which emphasizes steadfastness, loyalty and the rewards of fidelity. These virtues are underlined in the Gesta’s account of the career of the new Amboise in-law Geoffrey of Chaumont, who supported William of Normandy’s invasion of England in 1066 and served as a loyal counsellor to both William and the counts of Blois, helping to arrange the marriage between William’s daughter Adela and Stephen-Henry of Blois.56 In this, once again, the Gesta is complementary, describing the Norman conquest which the Chronica had omitted. The Gesta is also complementary in that it brings to the fore a new genre of historical narrative that is totally absent from either of the previous texts. In its prologue, which is otherwise reminiscent of both the Liber (in quoting Sidonius Appolinaris) and the Chronica for what it says about its sources and methods, the Gesta breaks from either by invoking contemporary circumstances.57 There is a great sense of anxiety and urgency: Sulpice II of Amboise (r. 1129–53) is, as the writer begins his work, currently a captive of Count Thibaud V of Blois (d. 1191). Unlike the Liber and the Chronica, which each devote only a few lines to the most recent events known to them, more than a quarter of the Gesta narrative is devoted to the lordship of Sulpice II of Amboise, a period which must have corresponded with the lived experience of many of the text’s readers. As in the Chronica, much of the evil in the Gesta’s account is laid squarely at the feet of Count Fulk IV le Réchin. Where the Chronica had relayed the story of the count’s unjust removal of the guardian of his domicile in Amboise, the Gesta expands greatly on this account, first introducing the good men from Amboise in turn, praising them (in words from Cicero’s De amicitia) for their uprightness, and then launching into a ferocious invective against the evil Count Fulk.58 In the end Fulk had entered Amboise under the pretence of asking for hospitality, only to seize the domicile’s guardian and replace him with his preferred vassal.59 Worse still, this event is said to have occurred as Fulk was returning from Paris, where he had just

54

Chroniques, 85. Kimberly Loprete, Adela of Blois: Countess and Lord (c. 1067–1137), Dublin 2007, 36. 56 Chroniques, pp. 97–8. See Jean Dunbabin, ‘Geoffrey of Chaumont, Thibaud of Blois, and William the Conqueror’, ANS 16, 101–16. 57 For the parts of the prologue of the Liber reminiscent of the Chronica cf. Chroniques, 75 and 25. 58 Chroniques, 89–90. 59 Chroniques, 90–1. The story of the count’s domicile in Amboise was analysed in depth by in Louis Halphen, Comté d’Anjou au Xie siècle, Paris 1906, 148–51 and 172–3 and Olivier Guillot, Le comte d’Anjou et son entourage au XIe siècle, Paris 1971, 297. 55

114

Nicholas L. Paul

surrendered his ancestral holdings at Chateau-Landon, another moment of great shame. Fulk is a model of bad lordship, ultimately envious of his own men and wishing their destruction. The conflict between Amboise and Anjou in the time of Fulk le Réchin inaugurates a mode of storytelling in the Gesta which highlights not epic or heroic conflict, but instead the suffering of local people. Describing the fighting in the suburbs of Amboise and the Touraine more generally, the Gesta employs the language of devastation (devastans … vastabat) for the suburbs of Amboise and the lands around Tours; those who lived around Loches, Montrichard and Montresor are entirely annihilated (omnia … deleverunt); and two men, noted for their cruelty, were said to have mutilated (deturpabant) many knights.60 Amboise was itself besieged by Fulk, the town’s churches burned, and ‘the wounding and slaughter was exceedingly great’ (vulneratioque et occisio nimia erat).61 At a particularly dramatic moment in 1106, the lord of Amboise, Hugh of Chaumont-sur-Loire, decided to seize the domicile and raze it to the ground.62 Whether the story is true or not, it is easy to understand that within a larger narrative sequence that begins with the building of Amboise in the Liber, and which places so much emphasis on cooperation between the two families, the domicile’s destruction is freighted with significance. The Gesta makes clear that there was nothing inevitable about the erosion of relations between Amboise and Anjou. In the first instance, as we have seen, the blame is assigned to Fulk IV le Réchin, but the longer saga of suffering and division, recounted at length and taking the narrative up to the time of writing, is associated with the troubled lord of Amboise Sulpice II.63 The Gesta describes him as envious (twice quoting Horace, ‘the envious man gets thinner as another man grows fat’) and violent. Playing the counts of Vendôme, Blois and Anjou against one another, Sulpice was the spark that ignited the already combustible region. As the Gesta narrative draws to a close, news has arrived that (in 1153) Sulpice has been tortured to death in prison by allies of Count Thibaud V of Champagne.64 The authorial voice in the Gesta breaks at this point into the first person: ‘I have confided to writing the sequence and the truth of the aforesaid treachery lest it be lost to future generations.’65 This was not the only expression of shock. A charter of Saint-Julien of Tours is dated to ‘the year in which Sulpice of Amboise was captured’, and Manasses, archbishop of Orléans, was remembering him in prayer five years later.66 The flawed lord died imprisoned in Chateaudûn, the home, some readers might realize, of the treacherous lord Landric whom the founders of the Amboise lineage had vigorously opposed.67 The story of the dynasty of the lords of Amboise was one that had begun with, and could potentially end on, accounts of the villainy of Chateaudûn.

60

Chroniques, 107. Chroniques, 91. 62 Chroniques, 104. 63 For details of the continuing conflict during the reign of Fulk V, see LoPrete, Adela of Blois, 317–22. 64 Chroniques, 129–30. 65 Chroniques, 130: Supradicte vero proditionis seriem veritatemque, ne latere posteros queat, scripto mandavi. 66 Chartes de Saint-Julien de Tours (1002–1227), ed. L. J. Denis, Le Mans 1912, 119 no. 94. The charter also notes Henry II’s crossing to England; R. Porcher, ‘Histoire de l’Abbaye de Pontlevoy: pièces justificatives’, Revue de Loir-et-Cher 15, 1902, cols 145–52, 177–81, 201–7 and 16, 1903, cols 30–2, 33–42, 51–63, 65–75, 92–6 at co1. 92 no. 27. 67 Chroniques, 129–30. 61



The Amboise–Anjou Narrative Programme

115

Context and Time of Composition: The Birth of an Empire The Liber–Chronica–Gesta narrative sequence that we find in MSS A and B celebrated Amboise as a strategic site, Anjou’s gateway to Tours and its bulwark against Blois. Amboise had hosted great emperors and conquerors, like Caesar and Arthur, both of whom had wielded power in Britain and in western France. The ancestors of the lords of Amboise were loyal supporters of the counts of Anjou and (through Chaumont-sur-Loire) of the Anglo-Norman royal dynasty. That loyalty could be strained, but only through the fault of malevolent or misguided individuals, whose decisions and actions the text condemns. In a series of telescoping chronographies, each period more focused than the one before, the sequence brought the narrative to the dramatic events of the present (1154). King Stephen (of Blois) was dead, and Henry, count of Anjou and duke of Normandy, had acceded as the anointed ruler of England. Who were the makers of this programme, where were they working, and why did they undertake such a difficult and complex project? It is clear that the entire sequence greatly privileges Amboise, with the Angevin dynastic history girded by the two Amboise narratives. This would seem strongly to suggest composition in a context favourable to Amboise. Mabille, thinking of the composition of the separate texts, proposed the abbey of Pontlevoy, foundation and burial place of the lords of Amboise, as the site of the composition of the Liber and the Gesta.68 Under the influence of John of Marmoutier’s reference to ‘Brito’, Halphen pointed to the ‘master Brito’ active at the collegiate church of St Florentin at Amboise in 1141, and suggested him as a possible author of the Gesta.69 It seems to have gone unnoticed by either scholar that St Florentin had actually been given to Pontlevoy in 1096 by the co-lords of Amboise Hugh of Chaumont-sur-Loire and Aimery of Courron at the moment of their departure on the First Crusade.70 St Florentin was not the only religious community in Amboise with direct links to Pontlevoy. The community most closely associated with Pontlevoy was the priory of St Thomas, a church located within the walls of Amboise. St Thomas is mentioned in both the Gesta and the 1141 charter of ‘Master Brito’ of St Florentin.71 In 1140 or 1141 a monk named Arnulf, who identified himself as a monk of Pontlevoy and who was also the prior of St Thomas of Amboise, was responsible for the creation of a manuscript that included his own original treatise on computus, the rule of St Benedict and the necrology of Pontlevoy.72 While the manuscript was produced at least a decade earlier than our historical project, it is nonetheless very suggestive of the capabilities of the personnel of the priory (and possibly also the college) of Amboise in conjunction with their mother house of Pontlevoy to produce both 68

Mabille, ‘Introduction’, xliii–xlv. Halphen, ‘Introduction’, xxxvi, lviii. The charter to which Halphen referred is Porcher, ‘Histoire de l’Abbaye de Pontlevoy’ 16, cols 55–7, no. 41. Sincere thanks to Alexis Durand at AD Loir-et-Cher for kindly sending a scan of the Revue. 70 Acte n°2308, in: Chartes originales antérieures à 1121 conservées en France, ed. Cédric Giraud Jean-Baptiste Renault and Benoît-Michel Tock, Nancy 2010 (Telma). 71 Chroniques, 114. Like the other churches in Amboise, virtually no records survive for St. Thomas, but the buildings do appear in Dubuisson-Aubenay, Plan de la ville d’Amboise, 1635 repr. in Lucie Gaugain, Amboise: Un Chateau dans la Ville, Tours 2014, fig. 1. 72 Blois, Bibliothèque Municipale, MS 44 (unseen); see Charles Samaran and Robert Marichal, Catologue des manuscrits en écriture latine portant des indications de date, de lieu ou de copiste tome VII: Ouest de la France et Pays de Loire, Paris 1984, 103 ; Auguste Molinier, Les obituaires français au Moyen Âge, Paris 1890, 176–7, no. 120 ; RHGF : Obituaires de la province de Sens II :Diocèse de Chartres, Paris 1906, 208–19. 69

116

Nicholas L. Paul

manuscripts and works of literature. That Pontlevoy possessed both a library and the facility for book production is equally attested by a twelfth-century addition to the obituary for John Italus, a monk of both Marmoutier and Pontlevoy, who ‘bound together the great number of our books that is called the library, certain of them especially well and others suitably and he made beautiful books through the art of binding’.73 If the Amboise–Anjou collection was a project of a learned individual like Arnulf, drawing on the book-production resources of Pontlevoy and its daughter communities and clerics in Amboise (and possibly also Chaumont-sur-Loire, where the abbey had other possessions), when and why was it undertaken?74 As we have seen, the latest datable moment in the narrative sequence is the death of Elizabeth of Jaligny, mentioned at the close of the Gesta after the accession of Henry II (19 December 1154). The narrative voice of the Gesta expresses fear about what more could befall Amboise after the killing of Sulpice II, and within three years those fears were realized. The absence of Henry in England, first at the time of his accession (December 1154–January 1156), the disruption caused by his brother’s revolt (February–July 1156), and his return to England (April 1157–August 1158), as well as the growing alliance between the Capetian king Louis VII of France and the three Blois brothers (Thibaud V of Blois, Henry I of Champagne, and Stephen of Sancerre), all strengthened the Blois position in the Loire valley.75 By 1158, rising tensions between Henry and Louis over control of the Vexin, and the seeds of chaos already sown by Sulpice II, had apparently led to an assault from Blois, for according to Robert of Torigni Thibaud was negotiating with Henry to return Amboise, together with Fréteval.76 A year passed, during which Henry II led a force assembled from across his domains – an army that must have seemed truly imperial in its composition – to confront Count Raymond V of Toulouse over Queen Eleanor’s claims to western Occitania.77 In 1160, he struck. For what follows, we are entirely in the hands of Robert of Torigni, but his account of events in the Touraine for 1160 is extremely detailed: The three counts [Thibaud of Blois and his brothers Henry and Stephen], together with their men, began to strengthen the defences at Chaumont-sur-Loire, which was a fief of the castle of Blois, that they might occupy the Touraine … that castle was a possession of Hugh, son of Sulpice [II], of Amboise who held it of Count Thibaud (the chief part of his lordship, Amboise, he held of King Henry II).78

73

RHGF: Obituaires, 216 citing Blois BM MS 44, fol. 146v: Ipse enim maximum librorum nostrorum, qui Bibliotheca dicitur, et quosdam alios tam specialiter quam decenter conjungens ligavit et arte ligatoria speciosos codices effecit. 74 For the existence of ‘the churches of St. Nicholas and St. Martin’, churches belonging to Pontlevoy at Chaumont, see R. Porcher, ‘Histoire de l’Abbaye de Pontlevoy: Pièces Justificatives’, in Revue de Loir-et-Cher 16:185, 1903, col. 66. 75 John Hosler, Henry II: A Medieval Soldier at War, 1147–1189, Leiden 2007, 56–7; Warren, Henry II, Berkeley 1973, 88–90. 76 Robert of Torigni, Chronique, ed. Leopold Delisle, Rouen 1873, I, 314. For the connections between the earlier behaviour of Sulpice and the violence that followed his death see Dominique Barthèlemy, La Société dans le Comté de Vendôme de l’An Mil au XIVe Siècle, Paris 1993, 728–9. 77 Perhaps not coincidentally, Thibaud V of Blois was also in the south, on pilgrimage to Santiago de Compostela. Returning from their southern sojourns, Henry II attempted to provide hospitality for Thibaud at Limoges, but was rebuffed according to Geoffrey of Vigeois. Geoffrey of Vigeois, ‘Chronique (première partie)’, ed. Pierre Botineau, Thesis, École des Chartes, 1968, 125. 78 Robert of Torigni, Chronique, I, 329–30: Exinde hi tres comites, coadunatis viribus suis, coeperunt firmare munitionem Calvi Montis, quae erat de feudo castri Blesensism ut exinde pagum Turonicum



The Amboise–Anjou Narrative Programme

117

Henry took Chaumont-sur-Loire, capturing at the same time thirty-five knights and eighty sergeants. ‘And’, concludes Robert, ‘he returned the fortifications to Hugh of Amboise, who had been shut out by Count Thibaud because he imprisoned the aforesaid Sulpice, father of the same Hugh, who died there wretchedly’.79 Before retiring to Rouen with Eleanor to celebrate Christmas 1160 Henry had accomplished a stunning victory: the reconquest of the Touraine, the crushing defeat of the count of Blois, the restoration of Hugh II of Amboise as rightful lord of Amboise and Chaumont-sur-Loire, and the refortification of Amboise itself.80 This moment of rebuilding and repairing both buildings and friendships is as likely a context as could be imagined for the creation of the Amboise–Anjou narrative project. With its emphasis on the history of construction and reconstruction both of the castle of Amboise itself and of the alliances between Amboise and Anjou, and its careful enfolding and interleaving of the history of the two dynasties, the collection would represent a powerful and creative response to Henry’s triumph over his ancestral enemies in Blois and his recuperation of Amboise and Chaumontsur-Loire. If it is correct to imagine 1160 or sometime soon afterwards as the appropriate context in which to understand the Amboise–Anjou programme, however, it is also important to acknowledge that the sequence was more than just the celebration of a single victory; it was also a meditation on the ways that the world had changed from the point of view of a community that now found itself at the heart of a radically altered political context. The restoration of Amboise to Angevin control also marked its incorporation in a much larger political world. The count of Anjou was now not only king of England, but also the ruler of a territorial empire. In 1160, Henry was fresh from his campaign against Toulouse, leading an army so diverse it included the king of the Scots. The king of France, who was himself fitfully attempting to recoup old imperial forms through issuing the collective peace, was overawed by this Angevin king who was moving from strength-to-strength throughout his domains.81 The Liber–Chronica– Gesta programme was history composed for precisely this moment, situating the Angevin dynasty within much larger paradigms of imperial history. Whether it is appropriate to call the vast agglomeration of lordships ruled by Henry II and his sons an ‘empire’ is a question that still vexes the historiography of twelfth-century Europe.82 It is not a question, however, that is usually answered with reference to the many novel forms of cultural production that appeared in the Angevin domains, chief among them new modes of historical writing.83 The Amboise–Anjou programme is one that explicitly enfolds Angevin dynastic history within the history of empires, and its efforts to do so were not ignored. Shortly infestarent … quia ellud castrum erat de casamento Hugonis filii Supplicii de Ambazia, quod tenebat de comite Teobaldo (caput autem sui honoris, scilicet Ambaziam, tenebat de rege Henrico). 79 Robert of Torigni, Chronique, I:330: et reddidit eandem munitionem Hugoni de Ambazia, qui aversabatur pro posse suo comiti Tedbaldo, quia in carcere praedicti comitis Sulpicius, ipsius Hugonis pater, nequiter extinctus fuerat. 80 For commentary see Jacques Boussard, Le Gouvernement d’Henri II Plantagenêt, Paris 1956, 423–4; Hosler, Henry II, 188–9. 81 Jean Dunbabin, ‘Henry II and Louis VII’, in Henry II: New Interpretations, ed. Vincent, 47–62 at 53. 82 Fanny Madeline, ‘The Idea of “Empire” as Hegemonic Power Under the Norman and Plantagenet Kings (1066–1204)’, ANS 40, 179–96; Aurell, The Plantagenet Empire, 1–22. 83 See most recently (but focusing on England) Michael Staunton, Historians of Angevin England, Oxford 2017. For the continental writing in the vernacular see Charity Urbanski, Writing History for the King: Henry II and the Politics of Vernacular Historiography Ithaca 2013. For a survey of historical writing in the Angevin domains see Aurell, The Plantagenet Empire, 13–14, 134–62.

118

Nicholas L. Paul

after its composition, the programme was so successful that MS A was soon duplicated and enlarged in MS B. Richard Barton has demonstrated that an account of the battle of Alençon (1118) interpolated into the Chronica in MS B recast earlier Angevin history to support the contemporary claims of Henry II in the 1160s and early 1170s, and this shows that the programme continued to be used as a mirror on Angevin policies.84 The compilation also came to the attention of Ralph of Diceto, who incorporated its Angevin history into his own chronicles.85 Even when John of Marmoutier set about making this second expanded version, removing the Liber and the Gesta and addressing his version of the Chronica to King Henry himself, he tacitly admitted the success of the compilation, for into his new history of the counts of Anjou he copied long extracts from the texts concerned with Amboise.86 Taking the vantage point of a stronghold that was sited both on a frontier and near the archaic heart of the Angevin world, the Amboise–Anjou histories encompassed empires and royal dynasties from the Roman republic to the royal house of Jerusalem. This was a history that worked to make sense of a multitude of stories while at the same time providing didactic lessons, setting a landscape on fire with memories and legends, and warning about the consequences of disloyalty. If we are to imagine an explicitly Angevin historical culture, perhaps its beginnings are not to be found at the royal court, in the venerable abbeys of Normandy, or among the more famous Cistercian chroniclers of England, but in the old stone fortress that guarded the Loire and had already watched several empires come and go.

84

Richard Barton, ‘Writing Warfare, Lordship, and History: The Gesta consulum Andegavorum’s Account of the Battle of Alençon’, ANS 27, 2005, 32–51. 85 For examples see Ralph of Diceto, Opera, I, 135–6, 139, 148–9, 159, 161–5, 183–4, 200, 239, 293, 365–6. 86 Chroniques, 168, 171.

THE NOBLE LEPER: RESPONSES TO LEPROSY IN THE TWELFTH AND THIRTEENTH CENTURIES* Charlotte Pickard As has been written about at length, medieval attitudes towards and responses to leprosy were complex and, at times, deeply ambiguous, even contradictory.1 This paper uses case studies, predominantly drawn from the northern French nobility, to consider how responses to leprosy among the nobility can be detected in their interactions with lepers and in their religious patronage. It will consider whether the patronage of leper houses can be attributed to the commemoration of family members, or whether donations reflect familial patronage traditions. In particular, it will discuss the families of two known noble lepers from northern France and the ways in which their families appear to have reacted to their illness: the first, Theobald VI, count of Blois; and the second, Ralph II, count of Vermandois. How far did reactions to their illness reflect a sympathetic approach to lepers? Medieval perceptions of leprosy have long been the subject of speculation among historians, with the majority acknowledging that attitudes were both complex and contradictory. Historiography traditionally focused on the issues surrounding segregation and biblical attitudes towards leprosy. Until the 1990s, scholarship predominantly argued that lepers were socially excluded, and that their illness was seen as a form of punishment for their sins.2 Traditionally, it was believed that the emphasis on exclusion and the location of leprosaria outside the city walls was an indication that lepers should be set apart from society.3 Saul Brody and others have argued that while the church might have viewed leprosy as a sign that God had chosen to grant the sufferer salvation, it was also a diagnosis of sin and spiritual corruption.4 Brody emphasized that this was a disease of the soul, as well as the body.5 However, since the work of Touati, Rawcliffe and, as we shall see, Demaitre, the notion that those suffering with leprosy were excluded and stigmatized has been dispelled; instead greater emphasis has been placed on more compassionate atti-

*

I am indebted to Lindy Grant for her thoughtful comments and suggestions on this paper. I would also like to extend warm thanks to Liesbeth van Houts for inviting me to speak at the 41st Battle Conference in Cambridge and for obtaining funding from the George Macaulay Trevelyan Fund to support my attendance. 1 Catherine Peyroux, ‘The Leper’s Kiss’, in Monks and Nuns, Saints and Outcasts: Religious in Medieval Society, Essays in Honour of Lester K. Little, ed. Sharon Farmer and Barbara H. Rosenwein, Ithaca and London 2000, 172–88 at 174. 2 Jeffrey Richards, Sex, Dissidence and Damnation: Minority Groups in the Middle Ages, London and New York 1991, 150. 3 Elma Brenner, ‘Outside the City Walls: Leprosy, Exclusion, and Social Identity in Twelfth- and Thirteenth-Century Rouen’, in Difference and Identity in Francia and Medieval France, ed. Meredith Cohen and Justine Firnhaber-Baker, Farnham 2010, 139–55 at 139. 4 Saul Nathaniel Brody, The Disease of the Soul: Leprosy in Medieval Literature, Ithaca and London 1974, 61. 5 Ibid., 107.

120

Charlotte Pickard

tudes towards the illness and the ways in which those suffering were still integrated in society.6 The focus has shifted away from the belief that lepers were shunned and excluded, to one that recognizes the ways in which they remained a part of society. Touati, in particular, propounds the more positive idea that, rather than exclusion, leprosaria were instead intended for the conversion of lepers to religious life. Touati has also argued that lepers were increasingly viewed as a special group, in the thirteenth century, chosen by God to suffer on Earth and be saved.7 As such, they attracted patronage. Carole Rawcliffe has supported this view, arguing for variability in the way that we view medieval attitudes towards lepers and lepers’ experiences. As Rawcliffe has emphasized, lepers were the recipients of compassion and charity from both the lay and ecclesiastical communities, who offered them alms as well as spiritual support.8 While patronage of leper houses may say more about the patron than the recipients, it would seem to support the more positive approach to lepers and leprosy proposed by Touati. The varying approaches found in the historiography are to be expected when the evidence is so ambiguous, demonstrating diverse responses to the disease.9 Attitudes towards Leprosy When assessing biblical attitudes towards leprosy, it is evident why leprosy and the status of lepers in medieval society were so diverse. God is presented as striking people down with the illness in response to their actions. Miriam, sister of Moses, was made leprous as punishment for speaking against her brother. Miriam was only healed once she had spent seven days outside the camp.10 King Uzziah also contracted leprosy as punishment from God. Uzziah entered the temple to burn incense on the altar, when only the priests should have done so. As a result, he too became leprous. Uzziah lived the rest of his life apart from his people, while his son ruled over the kingdom.11 The emphasis on segregation is also set out in Leviticus, in which we are told that lepers should live ‘without the camp’.12 From these examples alone, it quickly becomes apparent whence the concept of segregation emerges. However, stories of Christ touching and healing lepers present a more compassionate approach towards those suffering from the disease. In the Gospel of Matthew, after coming down from the mountain, Jesus encountered a man with leprosy who asked Jesus to make him clean; Jesus put out his hand, touched the man and healed him.13 There are other such stories, which highlight Jesus’ compassion towards

6

François-Olivier Touati, ‘Les Léproseries aux XIIème et XIIIème siècles?’, in Nicole Bériou and François-Olivier Touati, Voluntate Dei leprosus: Les lépreux entre conversion et exclusion aux XIIème et XIIIème siècles, Spoleto 1991, 1–32 at 3–19; François-Olivier Touati, Maladie et société au Moyen Age: La lèpre, les lépreux et les léproseries dans la province ecclésiastique de Sens jusqu’au milieu du XIV siècle, Paris 1998; Carole Rawcliffe, Leprosy in Medieval England, Woodbridge 2006. 7 Touati, Maladie et Société au Moyen Age, 631–748. 8 Carole Rawcliffe, ‘Learning to Love the Leper: Aspects of Institutional Charity in Anglo-Norman England’, ANS 23, 2000, 231–50 at 231. 9 Elma Brenner, ‘Recent Perspectives on Leprosy in Medieval Western Europe’, History Compass 8, 2010, 388–406 at 388. 10 Numbers 12:1–16. 11 2 Chronicles 26:16–23. 12 Leviticus 13:46. 13 Matthew 8:1–4.



The Noble Leper

121

those suffering with the illness.14 Clearly, the Bible presents contradictory ideas regarding lepers: one model marks lepers for exclusion, another considers them deserving of charity.15 Taking the Old Testament view, leprosy was believed to be a sign of God’s judgement, a personal punishment in which the decaying body became a representation of the soul.16 In his sermons, Jacques de Vitry used leprosy as a metaphor for sin, attempting to transfer the revulsion associated with leprosy to sin.17 Similar ideas can be found in the sermons of Maurice de Sully, who directly linked leprosy to sins including fornication, adultery, usury, robbery, theft, gluttony and drunkenness.18 The concept that leprosy, and illness more broadly, might be linked to sin is reflected in the fact that the first line of treatment was spiritual healing.19 Churchmen evidently attempted to use leprosy as both a byword for sin and a warning to the faithful. The Fourth Lateran Council supported this view, stating that, ‘sickness of the body may sometimes be the result of sin’.20 The Council thus decreed that it was first essential to take care of spiritual health, ‘for when the cause ceases so does the effect’.21 The direct connection made between illness and sin would appear to be damning for the leper. However, a more prevalent belief would seem to be that lepers were sinners who had been marked out by God for redemption, and that leper houses were places for conversion to the religious life. Churchmen, including Bernard of Clairvaux, supported this idea, putting forward the notion that illness was a gift from God and therefore an opportunity; the sick were in a kind of living purgatory,22 ensuring their eternal salvation. Arguably, one of the main reasons for a renewed focus on the plight of those suffering from leprosy in the twelfth and thirteenth centuries was the Third Lateran Council of 1179. The council made explicit provision for leper houses, likely in reaction to the advance of leprosy in the West. Canon 23 decreed that leper communities should have their own churches, priests and cemeteries and that they should not pay tithes for their gardens or the pasture of animals.23 The decree focused attention on the plight of leper houses and the special position of those suffering from leprosy – a sign of the papacy showing and encouraging compassion.24 It also suggests that this was not necessarily commonplace before this date, and underscores that it was a Christian duty to provide for lepers. Clearly, the church envisioned that lepers would reside in leprosaria, in a community. Leper houses were predominantly religious foundations, with many following Augustinian rule – those residing within leprosaria participated in the

14

Mark 1:40–5, Luke 5:12–16. Courtney Krolikoski, ‘Saints and Sinner: The Role of the Saints in the Life of the Leper in the Thirteenth Century’, Annual of Medieval Studies at the CEU 18, 2012, 66–78 at 70–3. 16 Rawcliffe, Leprosy in Medieval England, 6. 17 Jessalynn Bird, ‘Medicine for Body and Soul: Jacques de Vitry’s Sermons to Hospitallers and their Charges’, in Religion and Medicine in the Middle Ages, ed. Peter Biller and Joseph Ziegler, York 2001, 91–108 at 92. 18 Maurice de Sully, Maurice of Sully and the Medieval Vernacular Homily. With the text of Maurice’s French Homilies from a Sens Cathedral Chapter MS, ed. Charles Alan Robson, Oxford 1952, 91. 19 Bird, ‘Medicine for Body and Soul’, 103. 20 Cum infirmitas corporalis nonnumquam ex peccato proveniat (Decrees of the Ecumenical Councils, ed. and trans. Norman P. Tanner, 2 vols, London 1990, I, 245–6). 21 … cum causa cessante cesset effectus (Ibid., I, 245–6). 22 Bernard of Clairvaux, Opera Omnia, PL 184, cols. 1264–5. Rawcliffe, Leprosy in Medieval England, 58. 23 Decrees of the Ecumenical Councils, I, 222–3. 24 Rawcliffe, ‘Learning to Love the Leper’, 237. 15

122

Charlotte Pickard

house’s religious life.25 The prayers of the leprous were thought to be especially effective.26 This meant that they were expected to play a part in the Christian community – clearly the marginality of lepers was not clear-cut.27 Within leper houses strict segregation of the sexes was typically enforced – this was, no doubt, partly motivated by beliefs surrounding the sexual causes of the spread of leprosy. Segregation also ensured chastity, in line with expectations of religious communities.28 However, not all lepers entered into leper houses – it has been assumed that the wealthy could arrange seclusion in their own houses, or buildings in the grounds of leprosaria.29 Traditionally, historians viewed lepers as living in a kind of prison, forced to be segregated from the outside world, often banished outside of the city walls. However, leprosaria were not closed communities – a wide range of people, including family members, might come and go.30 Building on the work of Touati and Rawcliffe, Luke Demaitre has argued that until the late thirteenth century leprosaria served as charitable shelters, rather than a means of isolation.31 Moreover, as discussed by Elma Brenner, the founding of leprosaria outside of the city walls, while undoubtedly linked to ideas set out in Leviticus, also had a more practical reason – land there was cheaper and more readily available.32 The fact that many houses were situated on roads, close to city gates, would suggest that, rather than being segregated, leprosaria were instead well placed to receive charitable gifts from those passing by, as well as being a visible symbol of the founder’s prestige and piety.33 Following broader trends towards an increase in religious patronage, there was a rise in the popularity of gift-giving to leprosaria. As well as the usual gifts of money and rents, it was also common to make donations that would provide sustenance. This might include donations of cloth and wheat, as well as the right to take wood, used to provide heat.34 Leper houses could also be afforded the right to hold fairs – the community was not as separate from society as has traditionally been believed.35 Secular Reactions to Leprosy The notion that leprosy was a sign of God’s judgement and an indicator of sin certainly had the potential to be problematic for the nobility, especially when one of their own was suffering from the illness. However, we have surviving accounts of secular interactions with lepers, of men and women choosing to seek out and embrace those suffering from the illness. 25

Bird, ‘Medicine for Body and Soul’, 103. Elma Brenner, Leprosy and Charity in Medieval Rouen, Woodbridge 2015, 5. 27 Ibid., 133. 28 Elma Brenner, ‘Leprosy and Identity in the Middle Ages’, in The Routledge History of Disease, ed. Mark Jackson, London 2017, 461–73 at 468. 29 Richards, Sex, Dissidence and Damnation, 153. 30 François-Olivier Touati, ‘Les groupes de laïs dans les hôpitaux et les léproseries au Moyen Âge’, Les mouvances laïques des ordres religieux: Actes du troisième colloque international du CERCOR, Sainte Etienne 1996, 137–62 at 147–8; Simon Roffey, ‘Medieval Leper Hospitals in England: An Archaeological Perspective’, Medieval Archaeology 56, 2012, 203–33 at 217. 31 Luke Demaitre, Leprosy in Premodern Medicine: A Malady of the Whole Body, Baltimore 2007, 140–1. 32 Brenner, Leprosy and Charity, 21. 33 Roffey, ‘Medieval Leper Hospitals’, 222. 34 Brenner, Leprosy and Charity, 97. 35 Brenner, ‘Outside the City Walls’, 153. 26



The Noble Leper

123

One of the earliest royal examples is that of St Radegund, queen of the Franks (d. 587). St Radegund not only offered hospitality to lepers but also kissed their faces and washed them, as well as feeding them and treating their sores.36 In doing so, Radegund was demonstrating humility and piety.37 The touching and kissing of lepers became an established way for kings and queens to display extreme religious devotion. Examples of similar behaviour can be detected in later centuries. One of the most notable is that of Queen Matilda (d. 1118), Henry I of England’s first wife, who was said to have washed and kissed the feet of lepers.38 Matilda is also known to have founded leper houses, including that of St Giles at Holborn.39 Similarly, Louis IX of France (1226–70) made efforts to come into contact with lepers and to serve them, in the model of St Radegund. It was said that the more serious the illness, the more likely he was to touch them. Louis recognized lepers for what they represented.40 Such an overt display of compassion towards lepers, and the desire to be associated with their cause, would have undoubtedly set an example for others. There are also models for such behaviour from the nobility below royalty. One of the best-known stories relates to Theobald IV, count of Blois and (as Theobald II) Champagne (d. 1152; hereafter referred to as Theobald IV). Theobald was the son of Adela of Blois (d. 1137), William the Conqueror’s daughter, and Count Stephen II of Blois. He was also the brother of King Stephen of England (1135–54) and Henry, bishop of Winchester (d. 1171). Theobald, then, belonged to one of the foremost noble families, possessing power on both sides of the Channel. He had vast land and wealth at his disposal and was a prominent religious patron and a friend and supporter of Bernard of Clairvaux. Theobald was renowned for his commitment to caring for lepers. Walter Map recounts the story of how Theobald regularly visited the home of a leper. During one of these visits Theobald found the leper close to death and asked the provost of the village to care for him. Some days later, Theobald returned to the hut to find that the leper had recovered and that the stench of his sores was replaced by a sweet odour. When Theobald thanked the provost for the leper’s recovery, the provost informed him that the leper had died and that he had arranged his funeral, offering to take Theobald to the grave. When Theobald returned to the hut, after visiting the grave, he found nothing but an empty building, whereat he ‘rejoiced that he had beheld Christ’.41 Versions of this story were also included in the exempla collections of Jacques de Vitry, Thomas of Cantimpré and Caesarius of Heisterbach.42 36

Venatius Fortunatus, ‘The Life of Holy Radegund’, in Sainted Women of the Dark Ages, ed. Jo Ann McNamara and John E. Halborg, Durham and London 1992, 78. 37 Peyroux, ‘The Leper’s Kiss’, 181–5. 38 Aelred of Rievaulx, ‘Genealogia regum anglorum’, PL 195, ed. J.-P. Migne, Paris 1844–64, 736, trans. Lois L. Huneycutt, Matilda of Scotland: A Study in Medieval Queenship, Woodbridge 2003, 104–5. 39 Henry II also supported Matilda’s foundation of Saint Giles. Huneycutt, Matilda of Scotland, 106. 40 Katie Phillips, ‘Saint Louis, Saint Francis and the Leprous Monk at Royaumont’, in Selected Proceedings from ‘The Maladies, Miracles and Medicine of the Middle Ages’, March 2014. The Reading Medievalist: A Postgraduate Journal 2, 2015, 71–84 at 72–4. 41 gauisus est se uidisse Christum (Walter Map, De Nugis Curialium, ed. and trans. Montague Rhodes James, Oxford 1983, 462–5). 42 Jacques de Vitry, The Exempla or Illustrative Stories from the Sermones Vulgares of Jacques de Vitry, ed. Thomas Frederick Crane, London 1890; reprint, New York 1971, 43–4, no. 94. Thomas of Cantimpré, Bonum Universale de Apibus, Douai 1627, 254–5. Caesarius of Heisterbach, Dialogus Miraculorum, VIII, c. 31, ed. Joseph Strange, Cologne 1851, 105. For a discussion of the story’s transmission, see Paul Acker, ‘Harvard MS Eng 1032: Exemplary Verse Containing Theobald and Charlemagne’, Neuphilologische Mitteilungen 87, 1986, 354–64.

124

Charlotte Pickard

Theobald’s generosity towards the leper marked him out as being worthy of a close relationship with Christ. This story is reflective of other exempla, in which physical contact with lepers – often to the revulsion of the person involved – resulted in Christ manifesting himself as a leper.43 Similarly, the Blessed John of Montmirail (d. 1217) was praised for his interactions with lepers. A nobleman turned Cistercian monk, much of what we know about John survives in a vita produced to support his canonization.44 John was lord of Montmirail and lord of Oisy and was married to Helvide of Dampierre, with whom he was a generous religious patron. Among his many works, John founded an Hôtel-Dieu in Montmirail for the benefit of the poor.45 He further demonstrated his commitment to the religious cause by participating in the Third Crusade, where he was renowned as a knight who served King Philip Augustus (1180–1223).46 Later, in 1210 John chose to take the Habit, leaving his wife Helvide and their children, who were minors, to enter the Cistercian monastery of Longpont – a decision that was greeted by Helvide with such disapproval that she refused to speak to him.47 John’s piety was especially evident in his care for lepers, as reported on in the Vita. As well as patronizing leper houses, the author of the Vita records one occasion, when John was staying at his castle of Oisy, when he met lepers on the road and gave them alms.48 The author tells another story of how John heard of a leper who was deformed and decided to spend the day with him, generously giving him alms and kissing him.49 The stories relating to John’s interactions with lepers were intended to support the application for his canonization and so naturally cast him in a favourable light. Nevertheless, it is likely that John took the time to care for lepers personally. Clearly, for the author of the Vita, John’s care of lepers marked him out at being especially devout and deserving of canonization. While some churchmen may have perceived lepers as representations of Christ, we should not assume that this attitude was shared by all laymen. In the Vie de Saint Louis, Jean de Joinville recalls how, when asked by Louis IX if he would prefer to be a leper or to have committed mortal sin, Joinville chose mortal sin. Louis responded by saying that there is no leprosy as foul as being in a state of mortal sin, as leprosy is healed upon death, whereas if a man dies after committing a mortal sin, he can never be sure if he has sufficiently repented during his lifetime.50 Although Louis’ beliefs align with those of the church, Joinville’s revulsion at the mere thought of leprosy serves to highlight the fears of laymen. It is important to note that all of these examples of royal and noble laity interacting with lepers are marked out as displaying special piety – they were the exceptions, rather than the rule. From the accounts that we have, it is clear that these figures were viewed as overcoming a natural and expected fear of lepers. The fact that they were willing to do so made them more pious. This suggests that, for the majority, leprosy was still an illness to be feared and that lepers were widely seen as repulsive.

43

Anne E. Lester, Creating Cistercian Nuns: The Women’s Religious Movement and its Reform in Thirteenth-Century Champagne, Ithaca and London 2011, 197. 44 ‘Vita de B. Joanne de Monte Mirabili’, in Acta Sanctorum: Septembris, 71 vols, Paris 1861–1940, VIII, 186–255. 45 Alexandre Clément Boitel, Histoire du Bienheureux Jean, Paris 1859, no. 34. 46 Vita, c. 1. 47 Ibid., c. 3. 48 Ibid., c. 2. 49 Ibid., c. 2. 50 Jean de Joinville, Vie de Saint Louis, ed. Jacques Monfrin, Paris 1998, 13–15.



The Noble Leper

125

Religious Patronage of leprosaria The patronage of leper houses was prioritized by those with wealth, including the nobility and royalty. We know that the kings of France made great efforts to engage with leper houses and offer them support.51 The French kings supported leprosaria, including those situated in Normandy, after 1204. Perhaps most notably, they patronized Mont-aux-Malades – a house under Augustinian rule at Rouen.52 Louis VIII (1223–6) showed a special commitment to supporting lepers when drawing up his will in 1225; he left substantial amounts of money to the church, in particular to Premonstratensian and Victorine canons and the Cistercians, as well as to hospitals and leper houses. Louis famously provided 100 solidi to each of the 2,000 leper houses in his kingdom.53 The contents of Louis’ will are indicative of his religious devotion – his prioritization of leper houses reflects wider trends emerging in both England and France. Henry I of England’s second queen, Adeliza of Louvain (d. 1151), followed in her predecessor’s footsteps in founding the leper house of St Giles at Wilton. Such instances of queenly piety helped to emphasize a queen’s generosity, projecting a powerful image for their royal successors, as well as the wider lay nobility.54 The Empress Matilda followed the example set by Henry I’s queens in patronizing leper houses.55 Similarly, Countess Adela of Blois was a known religious patron who established a tradition of patronage for successive generations of countesses.56 There was clearly a female tradition of supporting leprosaria.57 This model of piety was also adopted by their male relatives. Henry I of England patronized the leper house of Grand-Beaulieu in Chartres, founded by the counts of Blois in 1054.58 Henry granted 10 livres of Rouen ‘for the soul of my father and the souls of my kindred, and for remission of my sins and the welfare and security of my kingdom of England and my duchy of Normandy’.59 While this particular grant to a leper house, situated outside Henry’s lands, may at first appear unexpected, Grand-Beaulieu had been placed under comital protection by Countess Adela of Blois, Henry’s sister.60 Henry I’s grandson Henry II followed in their footsteps and was a great supporter of leprosaria – including founding the leper house of Salleaux-Puelles, at Rouen, which was founded for leprous noblewomen.61 Evidently,

51

Brenner, Leprosy and Charity, 24. Ibid., 17. 53 Layettes du Trésor des Chartes, ed. Alexandre Teulet, 5 vols, Paris 1863–1909, II, no. 1710. 54 Rawcliffe, Leprosy in Medieval England, 147. 55 Brenner, Leprosy and Charity, 25. 56 Kimberly LoPrete, ‘Adela of Blois: Familial Alliances and Female Lordships’, in Aristocratic Women in Medieval France, ed. Theodore Evergates, Philadelphia 1999, 7–43. 57 Ibid., 26. For discussion of the religious patronage of the Anglo-Norman and Angevin royal families, including leprosaria, see Lindy Grant, ‘Le patronage architectural d’Henri II et de son entourage’, Cahiers de civilisation médiévale 37, 1994, 73–84 and Lindy Grant, ‘Aspects of the architectural patronage of the family of the counts of Anjou in the twelfth century’ in Anjou: Medieval Art, Architecture and Archaeology, ed. John McNeill and Daniel Prigent, Leeds 2003, 96–110 at 98. 58 Grand-Beaulieu was a mixed leper house, although it predominantly housed men. Touati, Maladie et Société, 316. 59 Pro anima patrum et parentum meorum et pro remissione peccatorum meorum et statu et incolumitate regni mei Anglie et ducatus mei Normannie (Cartulaire de la Léproserie du Grand-Beaulieu et du prieuré de Notre Dame de la Bourdinière, ed. René Merlet and Maurice Jusselin, Chartres 1909, 1). 60 Kimberley A. LoPrete, ‘Adela of Blois and Ivo of Chartres: Piety, Politics and the Peace in the Diocese of Chartres’, ANS 14, 1991, 131–52 at 145. 61 Brenner, Leprosy and Charity, 59; Leonie V. Hicks, Religious Life in Normandy, 1050–1300: Space, Gender and Social Pressure, Woodbridge 2007, 100. 52

126

Charlotte Pickard

there was a familial tradition of supporting leper houses, both particular foundations and leprosaria more widely. In addition to the miracle story, Walter Map praised Theobald IV of Blois for his generosity towards leprosaria. Walter remarks that Theobald ‘used to support lepers more willingly and with more pleasure than other poor persons … the more abjectly contemptible and the more unbearably importunate they were, the more pleasing he hoped, was the service he offered to God, and the more lovingly was it accepted.’62 Theobald was a patron of Grand-Beaulieu, where he established a fair lasting eight days. By 1151, the fair had been transferred from a suburb into the city itself. Grand-Beaulieu also received all commercial taxes on products entering Chartres while it took place.63 Theobald’s role as a religious patron fits into familial patterns set by his mother, Adela. However, as Walter Map suggests, Theobald seems to have had a particular sympathy for those suffering from leprosy. The tradition was followed by successive generations: his son, Henry the Liberal (d. 1181), his son’s wife, Marie (d. 1198), countess of Champagne, and their son Henry II (d. 1197), count of Champagne, were all patrons of leper houses. Henry the Liberal gave to several leper houses in Troyes and a leper house in Bar-sur-Aube. In 1186, Marie, with the consent of the chapter of Saint-Etienne, gave two prebends of the chapter to the leper house of Deux-Eaux, south of Troyes.64 Interestingly, the patronage carried out by Henry the Liberal and Henry II of Champagne coincided with their crusading. Henry the Liberal donated to leper houses in Troyes and a hospital in Bar-sur-Aube around the time that he was due to go to the Holy Land in 1179.65 His son Henry II is also recorded as having patronized a leper house in Troyes whilst in Palestine in 1191.66 The foundation and patronage of leper houses has often been associated with the rise in crusading. While it is true that the expansion in the foundation of leper houses in the twelfth and thirteenth centuries coincided with the height of the crusading movement, there is evidence of earlier foundations.67 The connection between crusading and lepers can also been seen in Odo of Deuil’s account of Louis VII visiting a leper colony outside Paris before going on crusade. Odo tells us that Louis left his retinue outside and that he was accompanied only by two companions. Although Odo does not provide us with details of how Louis spent his time in the leper colony, he emphasizes that Louis remained there for ‘a long time’ (per longam moram).68 Although contact with lepers and donations to leprosaria were undoubtedly acts of personal piety, it is also possible that such actions were motivated by a fear among crusaders that they may contract leprosy in the Holy Land.69

62 Walter Map, De Nugis Curialium, 462–3: Leprosus lecius et libencius exhibebat quam alios pauperes … illos autem ideo precipue quia, quanto sunt despicabiles abiectius et intolerabilius improbi, tanto se sperat obsequium prestare Deo placencius, et affectuosius acceptari. 63 Brenner, ‘Outside the City Walls’, 153. 64 Henri d’Arbois de Jubainville, Histoire des Ducs et des Comtes de Champagne, 7 vols, Paris 1859– 69, III, 386, nos. 344 and 345; Theodore Evergates, Marie de France, Countess of Champagne, 1145– 1198 Philadelphia 2019, 57. 65 d’Arbois de Jubainville, Histoire des Ducs et des Comtes de Champagne, III, 244, nos. 167 and 309. 66 Idem, III, 398, no. 411. 67 Simon Roffey has speculated at the possibility of a pre-Conquest leper hospital in England. However, the earliest recorded leper house in England was Lanfranc’s foundation in the 1080s. Roffey, ‘Medieval Leper hospitals’, 211. 68 Odo of Deuil, De Perfectione Ludovici VII in Orientem, ed. Virginia Gingerick Berry, New York 1965, 116–17. 69 David Marcombe, Leper Knights: The Order of St. Lazarus of Jerusalem in England, c. 1150–1544, Woodbridge 2003, 10.



The Noble Leper

127

Noble Lepers It is difficult to know how prevalent leprosy was among the nobility, primarily because it is likely that many cases were hidden or simply not reported. However, there are a few notable examples about which we have information. Carole Rawcliffe has argued that the increase in instances of the nobility contracting leprosy may have been a powerful factor in promoting an alternative image of the disease.70 However, churchmen were prepared to use the negative connotations of leprosy as a weapon against the lay nobility. According to the Life of Saint Hugh of Lincoln, Robert Fitzpernel, earl of Leicester, contracted leprosy as punishment for unjustly confiscating an estate lately belonging to Hugh of Avalon, bishop of Lincoln, which he seized immediately after Hugh’s death in 1200. The author of the Vita informs us that Robert was struck down with leprosy as a punishment for committing this sin, dying shortly after in 1204.71 From an ecclesiastical perspective, Robert’s pride and avarice offered an explanation for his illness.72 As with so many stories of this kind, we should not take it at face-value. If Robert had indeed unjustly obtained one of Lincoln’s estates, it is perhaps unsurprising that an account of his leprosy written by a monk would present his illness as punishment for his sins, thus offering a cautionary tale to anyone considering taking similar action. Yet the circumstances surrounding Robert’s supposed leprosy are exceptional. While we have ecclesiastical texts that place judgement on the causes of noble leprosy, lay attitudes are more difficult to identify. Instead, we have to use alternative evidence, such as religious patronage, to reconstruct familial responses to a relative’s illness. Theobald VI, count of Blois, who died in 1218, offers a northern French example of noble leprosy. Theobald was the son of Louis I, count of Blois, and Katherine, countess of Clermont-en-Beauvaisis, and great-grandson of Theobald IV. As we have seen, the counts and countesses of Blois were renowned religious patrons, not least of leprosaria, following the example set by their ancestor Adela of Blois. Theobald is said to have contracted leprosy while fighting against the Moors in Castile and is believed to have chosen seclusion in his own castle, La Ferté-Villeneuil, rather than entering into a leper house. La Ferté-Villeneuil was an important centre for the counts of Blois and is believed to have been a favoured residence of Theobald’s, with charters regularly being issued there.73 Theobald supported the town’s religious houses, including its leprosarium.74 Despite seemingly being a favoured residence, it is surprising that Theobald eschewed his major urban centres, especially when aristocratic life was typically itinerant; the town’s position, approximately halfway between the comital centres of Blois and Chartres, may account for Theobald’s decision to reside there during his illness. Theobald’s decision to remain in his favoured residence of La Ferté-Villeneuil supports wider assumptions that the nobility chose to reside in their own homes, rather than enter leper houses. Theobald was married twice, first to Maud of Alençon and then to Clémence of Roches. However, he died without an heir and the county was divided between his aunts: the elder, Margaret, inherited the county of Blois, while the younger, Isabelle, 70

Rawcliffe, Leprosy in Medieval England, 54. Magna Vita Sancti Hugonis. The Life of St Hugh of Lincoln, ed. Decima L. Douie and Dom Hugh Farmer, 2 vols, London 1961–2, II, 84. 72 Rawcliffe, Leprosy in Medieval England, 52. 73 Cartulaire de la Léproserie de Grand-Beaulieu, 80–1, no. 208; charter in Noel Mars, Histoire du Royal Monastère de Sainct-Lomer de Blois, Blois 1869, 168–9; L’Abbé J. Augis, Essai Historique sur la Ville et Chatellenie de La Ferté-Villeneuil, Châteaudun 1902, 376–7, Pièces Justificatives XIII and XIV. 74 Augis, Essai Historique, 127–30. 71

128

Charlotte Pickard

became countess of Chartres. The Blois–Chartres heiresses were known religious patrons, Isabelle especially so. The sisters were keen to support the leper house of Grand-Beaulieu, as their ancestors had before them. Grand-Beaulieu benefitted from Margaret’s generosity when she exempted one of the servants from all taxes.75 Isabelle also made a gift in 1235 of 40 solidi to be taken from her toll at Chartres for the celebration of her anniversary.76 Isabelle was thus commemorated in their obituary lists.77 In addition to Grand-Beaulieu, it is believed that Isabelle was one of the earliest patrons of the leper house of Saint-George-de-la-Banlieue in Chartres.78 One could speculate that Margaret and Isabelle’s patronage of leprosaria was motivated by a sense of responsibility to provide for those who were suffering from an illness that had so greatly affected their own family. However, we should not overlook the fact that Margaret and Isabelle’s provision for leprosaria aligns with family tradition, fitting with the sisters’ wider religious patronage.79 It is evident that the countesses were consciously following the examples set by their predecessors, and it is perhaps in this context that their support of leprosaria should be viewed. Possibly the best-known northern French example of noble leprosy is Ralph II, count of Vermandois (c. 1147–c. 1167). As count of Vermandois Ralph ruled a vast territory, which included Amiens, Saint Quentin, Péronne, Montdidier, Ribemont, Chauny and Valois. The counts of Vermandois were descended from a cadet branch of the Capetians and could claim descent from Charlemagne – they were at the very pinnacle of the French aristocracy.80 Ralph was born c. 1147 and was the product of the controversial marriage between Ralph I, count of Vermandois, and Petronilla of Aquitaine, Eleanor of Aquitaine’s younger sister. It is said that upon meeting Petronilla at the royal court, Ralph was besotted with her. Consequently, Ralph left his first wife, Eleanor, the niece of the count of Blois, in order to marry Petronilla in 1142.81 However, at a council held in Lagny later that year, Ralph’s first marriage was declared valid and Ralph and Petronilla were ordered to separate. When they refused, they were both excommunicated along with the three bishops who had allowed the marriage to take place.82 Eventually, at the Council of Reims in 1148, Ralph obtained the dissolution of his marriage to Eleanor and had the excommunication lifted.83 Bernard of Clairvaux was a particularly vocal opponent of the marriage and is reported, by John of Salisbury, to have prophesied that, ‘no issue springing from that bed would bear worthy fruit among the people of God, and that they themselves would not enjoy each other for long’.84

75

Table Chronologique des Diplômes, Chartes, Titres Et Actes Imprimés Concernant l’Histoire de France, ed. Louis Georges Oudart Feudrix de Bréquigny, 8 vols, Paris 1846, V, 165; Robert Perrichon, La Ferté-Villeneuil au long des siècles: Histoire d’un village de Beauce, Versailles 1970, 39, 66, 81. 76 BnF Nouv. Acq. Lat. 608. Cartulaire de Grand-Beaulieu, no. 306. 77 Obituaires de la Province de Sens, ed. Auguste Molinier, 4 vols, Paris 1902–23, II, 406. 78 Eugène de Bouchère de Lépinois, Histoire de Chartres, 2 vols, Chartres 1854, I, 365. 79 For discussion of Isabelle and Margaret’s patronage, see Michelle Armstrong-Partida ‘Mothers and Daughters as Lords: The Countesses of Blois and Chartres’, Medieval Prosopography 26, 2005, 77–107. 80 Constance Brittain Bouchard, Those of My Blood: Creating Noble Families in Medieval Francia, Philadelphia 2001, 52. 81 Ralph V. Turner, Eleanor of Aquitaine: Queen of France, Queen of England, New Haven 2009, 63–7. 82 ‘Historiae Tornacensis’, in MGH SS, XIV, 343. Stephen Robson, “With the Spirit and Power of Elijah” (Lk 1,17): The Prophetic-Reforming Spirituality of Bernard of Clairvaux as Evidenced particularly in his Letters, Rome 2004, 221. 83 John of Salisbury, Historia Pontificalis, ed. and trans. Marjorie Chibnall, London 1956, 12–13. 84 John of Salisbury, Historia Pontificalis, 14: quod nunquam erat de lecto illo soboles egressura que laudabilem fractum faceret in populo Dei.



The Noble Leper

129

Ralph and Petronilla’s children made a series of high-ranking marriages, with all three marrying into the family of the counts of Flanders. The eldest, Elizabeth, married Philip, count of Flanders; Ralph II, count of Vermandois, married Margaret of Flanders (Philip of Flanders’s sister); and the younger sister, Eleanor, married, as her third husband, Matthew, count of Boulogne (Philip of Flanders’s brother). While these marriages had the potential to permanently unite the two great counties, all three marriages failed to produce an heir. Ralph II’s leprosy is reported on in the Continuatio Bruxellensis of the Flandria generosa – a history of the county of Flanders, likely started after 1164 by a monk of the abbey of St Bertin in Saint-Omer.85 Ralph is said to have contracted elephantiasis, an ancient Greek term used to describe Hansen’s Disease (leprosy).86 In his telling of Bernard of Clairvaux’s prophecy, John of Salisbury states that the prophecy was partly fulfilled, as ‘the son became a leper in his boyhood and now drags out a wretched life’.87 In her edition, Marjorie Chibnall argues that John of Salisbury was writing in 1164, making his observations on Ralph concurrent with his illness.88 What is notable here is that John implies that Ralph’s leprosy was directly associated with his parents’ sin. John also highlights Ralph’s state of wretchedness – evidently leprosy was not, by any means, a desirable state. However, there is, surprisingly, little or no mention of Ralph’s illness and death in other chronicles, with Gislebert of Mons merely stating that ‘Raoul the young count of Vermandois began to grow ill and died young’.89 It is unclear exactly when Ralph contracted leprosy, and the date of his death has also been the subject of debate.90 It seems highly unlikely that Ralph’s illness had been diagnosed before 1163, the year in which he married Margaret of Flanders.91 We can presume that the marriage would not have taken place had Ralph’s illness been diagnosed. The chapter of Saint Quentin lists Ralph’s date of death in their necrology as 26 January 1164.92 It has also been argued that acts issued by the young count cease to exist after 1163, attesting to a date of death of 1163/4.93 If Ralph did die in 1164, it would appear that his health had rapidly declined after his marriage the previous year. However, there is evidence that disputes this date, suggesting that although Ralph had ceased to act as count of Vermandois in 1163, he was still very much alive. We 85

Barbara N. Sargent-Bur, ‘Alexander and the Conte du Graal’, Arthurian Literature 14, ed. James P. Carley, Cambridge 1996, 1–18 at 3. 86 Radulfum fratrem ejusdem uxoris sue, qui elephantie infirmitatem incurrerat. (‘Continuatio Bruxellensis de la Flandria generosa’, in MGH SS, IX, 325). For a discussion on the terminology used for leprosy, see E. V. Hulse, ‘The Nature of Biblical ‘Leprosy’ and the use of Alternative Medical Terms in Modern Translations of the Bible’, Palestine Exploration Quarterly 107, 1975, 87–105. 87 John of Salisbury, Historia Pontificalis, 14: sed filius in puericia leprosus factus/ miserrimus uiuit adhuc. 88 John of Salisbury, Historia Pontificalis, xxviii. 89 Radulphus autem juvenis comes Viromandie egrotare cepit, et juvenis mortuus est. (La Chronique de Gislebert de Mons, ed. Léon Vanderkindere, Brussels 1955, 86; Gilbert of Mons, Chronicle of Hainaut, ed. and trans. Laura Napran, Woodbridge 2005, 51). 90 Louis Duval-Arnould, ‘Les dernières années du comte lépreux Raoul de Vermandois (v. 1147–1167) et la dévolution de ses provinces à Philippe d’Alsace’, Bibliothèque de l’École des Chartes 142, 1984, 81–91. 91 Marriage was clearly something that could be greatly affected if one of the partners contracted leprosy. Although leprosy was not deemed to be a reason to seek annulment, it is evident that many did. Brody, The Disease of the Soul, 85. 92 Léon Louis Borrelli de Serres, La réunion des provinces septentrionales a la couronne par Philippe Auguste: Amiénois, Artois, Vermandois, Valois, Paris 1899, ix–xiii. 93 For full discussion see Duval-Arnould, ‘Les Dernières Années’.

130

Charlotte Pickard

know that Philip of Flanders assumed the title ‘count of Vermandois’ in 1163/4, a title that he held through his wife, Elizabeth. However, there are charters issued by Ralph after this date. One dates from 1167, in which Ralph is styled ‘Ego Radulfus, Viromandorum comes’. The charter was approved by Philip, ‘count of Flanders’, suggesting that Philip’s authority was required to validate the charter.94 In another charter, dated sometime after 1163, Ralph made donations for the souls of his parents.95 This charter was ratified by ‘Philippus, Flandrensis et Viromandensis comes, et uxor mea’.96 While this charter indicates that Ralph was still active, it is interesting to note that when granting fishing rights to the monks of Saint-Arnoul, the rights were dependent on Philip of Flanders’s agreement.97 These charters support the argument that Ralph did not die in 1163. However, it is apparent that Philip of Flanders held authority and that his participation was required to confirm Ralph’s donations. A much later date of death is suggested on a seventeenth-century inscription in the cloister at the abbey of Longpont, Ralph’s burial place. The inscription records Ralph’s death in 1176, more than a decade after the county had passed to his sister Elizabeth and brother-in-law Philip. There is some dispute as to the accuracy of this inscription, not only because it is a much later addition, but also because his sister Eleanor’s death is listed as 1214,98 when in fact she is believed to have died in 1213, suggesting inaccuracies in the dating.99 It is unlikely that we will ever have a definite date of death for Ralph. However, what is important is that, from 1164 onwards, Ralph was no longer deemed capable of ruling. Ellen Shortell has speculated that Ralph’s illness forced him to enter a leprosarium in 1163, but to date I have not found evidence that this was case.100 Indeed, the fact that Ralph was still issuing charters, no matter how limited, would seem to indicate that he had not. It was not impossible for a leper to rule, as shown by King Baldwin IV of Jerusalem (1174–85), who sat on the throne of Jerusalem for more than ten years with leprosy. No attempt was made to segregate Baldwin – out of necessity, a pragmatic approach was taken towards his illness.101 And yet, Baldwin is an exceptional example. Ralph, unlike Baldwin, had an obvious successor. Philip of Flanders would have been eager to take control of the county, especially as it was now certain that Ralph would not be fathering any heirs. Ralph’s illness and death were a crucial turning point in the history of the county of Vermandois, as it caused his sister Elizabeth to inherit. Elizabeth also died childless, in 1183, which resulted in a protracted inheritance struggle between her younger sister Eleanor and Philip of Flanders. The dispute was eventually settled with the king’s intervention, whereby it was agreed that Philip of Flanders could continue to hold the title until his death, while the territories were divided between The charter relates to land sold to the Cistercian abbey of Longpont. Ibid., 90, Pièce Justificatives I. The donations in this charter cover a wide range of religious houses, including both nunneries and monasteries, with gifts ranging from wheat and fishing rights, through to 30 solidi. 96 Philip’s use of the Vermandois title places the charter after 1163. 97 si voluntas comitis Flandrensis Philippi et Viromandensis affuerit (Duval-Arnould, ‘Les Dernières Années’, 91, Pièces Justificatives II). 98 Hic jacent Radulphus Junior eiusque soror Elienor, Viromandie ac Valesiae comites … qui obierunt, Radulphus quidem 15 calendas julii 1176, Elienor autem 11 calendas julii 1214. 99 Duval-Arnould, ‘Les dernières années’, 88–9. 100 Ellen Shortell, ‘Saint-Quentin, Chartres, and the Narrative of French Gothic’, in New Approaches to Medieval Architecture, ed. Robert Bork, William W. Clark and Abby McGehee, Farnham 2011, 35–44 at 40. 101 See Bernard Hamilton, The Leper King and his Heirs: Baldwin IV and the Crusader Kingdom of Jerusalem, Cambridge 2000 and Stephen Lay, ‘A Leper in Purple: The Coronation of Baldwin IV of Jerusalem’, Journal of Medieval History 23, 1997, 317–34. 94 95



The Noble Leper

131

Eleanor, Philip of Flanders and the king, Philip Augustus. In a later agreement with Philip Augustus, Eleanor consented to the king being her heir should she die childless. Considering that Eleanor was childless, and in her thirties, there can have been little doubt about who would inherit the county on her death. Elizabeth of Vermandois, Ralph’s older sister, acted as a religious patron alongside her husband, Philip of Flanders. Although Elizabeth rarely acted as an independent religious patron, there is evidence that she commemorated Ralph by cultivating connections with leper houses, something that her sister Eleanor would continue.102 One of the leper houses to benefit from her generosity was that of Péronne, situated on her ancestral lands. The concern with maintaining links to her brother extended to the abbey of Longpont, Ralph’s burial place. Philip and Elizabeth issued a charter in 1166 exempting the abbey from tolls.103 While these donations would suggest a personal response to commemorating Ralph, in both cases we have to accept that they also reflect traditional patterns of patronage set by previous counts of Vermandois.104 The evidence for Elizabeth’s patronage is relatively limited, and its implications uncertain. However, evidence for an enhanced interest in leper houses, and a determination to commemorate Ralph, is much clearer in Eleanor’s case. Eleanor of Vermandois was born c. 1148/9, and is renowned as being one of the great religious patrons of her age.105 She gave to virtually all religious orders and is especially known as the founder of the Cistercian nunnery of Parc-aux-Dames and the Fontevraudine priory of Longpré. Eleanor’s generosity as a religious patron is evident in a treaty with the king made in 1191/2 and three successive memoranda that set out her donations up to 1194. The intention of these documents was to record Eleanor’s donations up to that date and restrict her future gifts of income-producing property. In doing so, King Philip Augustus was attempting to prevent the loss of patrimony through extensive donations, securing his own inheritance.106 One thing that is noticeable from the memoranda is Eleanor’s generosity towards leper houses. It is for this reason that we know far more about Eleanor’s patronage of leper houses and the reaction to her brother’s death. Eleanor’s donations were largely focused on leprosaria in her Valois territories (the main bulk of what she inherited), although she also made a donation to Chauny, a foundation in the county of Vermandois. In line with wider trends of gifts designed to provide sustenance,107 Eleanor’s donations to leper houses typically consisted of grants of wheat. The leper houses of La Ferté-Milon, Houillon, Crépy, Chauny and Morienval, as well as the leper house of Chambly, situated on her husband’s

102 André

Moreau-Néret, A. and Marcel Levoy, ‘Les maladreries de la région de Villers-Cotterêts et le comte lépreux Raoul V de Crépy’, Fédération des Sociétés l’Histoire et Archéologie de l’Aisne 16, 1970, 138. 103 Elie Brun-Lavainne, Franchises, lois et coutumes de la ville de Lille, Lille 1842, F° 296. 104 In 1144, Ralph I, count of Vermandois gave funds to build the church at Longpont. Louis Paul Colliette, Mémoires pour servir à l’histoire ecclésiastique, civile et militaire de la province du Vermandois, 3 vols, Cambrai 1771–3, II, 264–5. Antoin Muldrac, Compendiosum Abbatiae Longipontis Suessionensis Chronicon, Paris 1652, 12–15. 105 For discussion of Eleanor of Vermandois’ religious patronage, see Constance Hoffman Berman, ‘Two Medieval Women’s Property and Religious Benefactions in France: Eleanor of Vermandois and Blanche of Castile’, Viator 41, 2010, 151–82 at 153. 106 Hoffman Berman, ‘Eleanor of Vermandois’, 152. 107 Rawcliffe, Leprosy in Medieval England, 324.

132

Charlotte Pickard

Beaumont-sur-Oise lands, received gifts of wheat from the countess.108 Eleanor also provided for their chaplains. There is evidence that Eleanor made a gift of 4 muids of wheat to the chaplain of Houillon, while the chaplain of La Ferté-Milon was granted 21 muids of wheat from the mill at Pont de Val, reflecting Eleanor’s dedication to supporting that particular leper house. Although Eleanor’s patronage of leprosaria largely took the form of wheat, her benefactions were not limited to one form of donation. We know, for instance, that she also made donations of dead wood – used as firewood to provide heat, something that was almost as important as food.109 One such example can be seen in donations to both the leper house of La Ferté-Milon, and the chaplain serving there, of a cartload of dead wood, which they were permitted to collect on a weekly basis. The Chauny leper house also received an alternative gift when they were given an income to be taken from the toll at Chauny. From existing records, though, it would appear that La Ferté-Milon was a particular favourite of Eleanor’s, as she is recorded as making the gift of a meadow at the same castle – one of her residences.110 Eleanor’s commitment to leper houses is especially noticeable in comparison to her patronage of other religious houses. In fact, there are very few establishments that received the same level of generosity. Other than leprosaria, the religious houses that received the most support from Eleanor were the priory of Longpré, a Fontevraudine priory, which she founded in 1180; the collegiate church of Saint Thomas, in Crépy, founded by her sister Elizabeth and Philip of Flanders; and the Cluniac priory of Saint-Arnoul in Crépy, her parents’ burial place. These religious houses had deeply personal connections. The fact that Eleanor’s patronage of leper houses was on a similar scale suggests that she considered them to be of great importance, and that they held a similar claim on her affections. Further evidence of Eleanor’s affection for and dedication to her brother can be viewed in her choice of burial place. Like Ralph, Eleanor was buried at the abbey of Longpont.111 One might have expected her to be buried in the priory of Saint-Arnoul in Crépy with her parents, or perhaps at one of the religious houses that she founded. Eleanor’s decision to be buried at the abbey of Longpont would seem to reflect a desire to be near her brother in death, affirming the commitment to his memory that she had demonstrated in her patronage of leprosaria. Ralph of Vermandois’ leprosy does not appear to have been a source of shame or embarrassment for his family. They certainly did not attempt to hide his condition. However, Saint Bernard’s condemnation and the connection drawn by John of Salisbury between Ralph’s leprosy and his parents’ transgression casts Eleanor’s patronage in new light. While Eleanor’s patronage of leper houses served as a commemoration, if not a celebration, of her brother, it may also have served a deeper penitential purpose. Conclusion In the twelfth and thirteenth centuries, the patronage of leper houses had become an established convention, one that marked a nobleman or woman as being particularly pious. The fact that lepers’ situation was so desperate made the nobility’s 108

For a list of Eleanor’s donations to leprosaria see Louis Duval-Arnould, ‘Les aumônes d’Aliénor, dernière comtesse de Vermandois et dame de Valois’, Revue Mabillon 60, 1984, 395–463, memoranda 2 and 3. 109 Peter Richards, The Medieval Leper and his Northern Heirs, Cambridge 1977, 30. 110 Duval-Arnould, ‘Les aumônes’, memorandum 3. 111 Alexandre Eusèbe Poquet, Monographie de l’abbaye de Longpont, Paris 1869, 132–4.



The Noble Leper

133

donations a greater demonstration of religious devotion. By caring for lepers, or granting generous religious gifts to provide for the sick, the nobility were able to secure their own salvation. Interesting parallels can be drawn between the Vermandois and the Blois–Chartres heiresses, as they found themselves in similar situations. Both families had experienced a difficult inheritance due to leprosy, and both were known patrons of leprosaria, though the patronage of the countesses of Blois and Chartres may be a case of entrenched family tradition, rather than a special commitment to supporting leper houses. Through their patronage, the Vermandois heiresses demonstrated a true commitment to providing for those suffering from leprosy. Eleanor of Vermandois’ donations to leprosaria are the most striking. While Eleanor’s patronage appears to comply with wider trends of supporting leper houses, as a means of demonstrating charity, its extent goes beyond that of the average noblewoman. Eleanor took great pains to ensure the prosperity of leper houses situated not only on her ancestral lands but also in her husband’s territories. What then was her motivation for making such vast donations to so many leper houses? While there may have been broader, political considerations, the numerous donations made to leprosaria imply a commitment to something deeper than merely fulfilling lordly duty. The evidence suggests that Eleanor and Elizabeth’s donations had a more emotional dimension, demonstrating a true affection for a brother who had died at a young age from a disfiguring illness. Clearly attitudes towards leprosy were problematic – the evidence is often vague at best, and shows that reactions to the disease were wide-ranging and ambiguous. It is unsurprising that noble responses to leprosy strayed away from church teaching, especially as instances of noble leprosy increased. It is understandable that the nobility chose to concentrate on the compassionate model offered by the New Testament, rather than the Old Testament focus on leprosy as a punishment for sin. The fact that prominent members of the lay nobility contracted leprosy is likely to have helped the cause of lepers and leprosaria, encouraging greater acts of generosity, forcing a reassessment of how human suffering was viewed.112 What emerges from the evidence is that, while leprosy posed a problem for the nobility, with churchmen sometimes exploiting its negative connotations against the lay nobility, it was not necessarily a source of shame for the noble family. In some instances, it was something to be highlighted, if not celebrated.

112 Guenter

B. Risse, Mending Bodies, Saving Souls: A History of Hospitals, Oxford 1999, 180; Rawcliffe, ‘Learning to Love the Leper’, 233.

ROYAL TAXATION AND WRITTEN RECORD IN ELEVENTH-CENTURY ENGLAND AND NINTH-CENTURY WEST FRANCIA David Pratt This paper explores Continental contexts for the understanding of taxation in eleventh-century England. The subject is widely recognized as important but fraught with difficulty. The central idea is the relevance of Carolingian practices of surveying and the keeping of written records for later Anglo-Saxon and Anglo-Norman government. Such a European perspective has been important for Anglo-Saxon historians: one thinks especially of the model emphasizing Carolingian influence on later Anglo-Saxon government advanced by James Campbell and Patrick Wormald.1 It has also been important for historians of Domesday, when one considers the comparative richness of the Carolingian administrative record, comprising the corpus of polyptychs, plus capitulary material indicating government interest in surveys and list-making. The Carolingian sources have been seen as the best comparanda for the two Domesday volumes and ‘satellite’ texts, as explored in classic studies by Campbell, John Percival, Henry Loyn and R. H. C. Davis.2 Are these indications of a shared early medieval culture of surveying and information gathering? There has been awareness that Anglo-Norman administrators came from a world familiar with polyptychs and certain legacies of Carolingian government, though ultimately commentators have placed more emphasis on the distinctiveness of the Domesday record. One should stress the necessary limits of knowledge on this difficult subject. Firstly, one encounters the dominance of Domesday in the surviving English record. The use of shire and hundredal structures in 1086 is highly suggestive of some administrative continuity from Anglo-Saxon government, but how much should one envisage in respect of written documentation? It was generally harder for AngloSaxon material to survive. This is especially shown by particular forms of AngloSaxon document which we know to have been widespread, now attested in rare or unique instances, such as certain types of writ.3 Yet it remains difficult to assess

1

J. Campbell, ‘Observations on English Government from the Tenth to the Twelfth Century’, in idem Essays in Anglo-Saxon History, London 1986, 155–70, at 159–70; P. Wormald, ‘Frederic William Maitland and the Earliest English Law’, and ‘Engla Lond: The Making of an Allegiance’, both in idem Legal Culture in the Early Medieval West: Law as Text, Image and Experience, London 1999, 45–69 at 54–6 and 51–3, and 359–82, at 366–71; idem, Papers Preparatory to The Making of English Law: King Alfred to the Twelfth Century. II: From God’s Law to Common Law, ed. S. Baxter and J. Hudson, London 2014, 98–191, available online. 2 Campbell, ‘Observations on English Government’, 163–7; J. Percival, ‘The Precursors of Domesday: Roman and Carolingian Land Registers’, in Domesday Book: A Reassessment, ed. P. H. Sawyer, London 1985, 5–27; H. R. Loyn, ‘The Beyond of Domesday Book’, and R. H. C. Davis, ‘Domesday Book: Continental Parallels’, both in Domesday Studies, ed. J. C. Holt, Woodbridge 1987, 1–13 and 15–39. 3 S. Keynes, ‘Royal Government and the Written Word in Late Anglo-Saxon England’, in The Uses of Literacy in Early Mediaeval Europe, ed. R. McKitterick, Cambridge 1990, 226–57, at 247–8; J. Campbell,

136

David Pratt

silences. Secondly, the Carolingian material is relatively extensive, but heavily skewed towards documents produced and preserved by ecclesiastical houses. The existence of centrally kept records is strongly implied by capitularies but harder to demonstrate.4 One risks the danger of idealizing the practices of certain bishoprics and monasteries, but perhaps also of under-estimating forms of central control. The problems were aptly summarized by Campbell: ‘No one knows when (no one indeed really knows if) surveys ceased to be used by the Continental rulers; no one knows when English rulers began to use them.’5 The silences may help to explain why the subject has become a slight dead-end. Both Loyn and Davis were forced to focus on the polyptychs and Domesday Book, in awareness that these were types of document produced for rather different purposes; they therefore faced the difficulty of finding a positive point of connection.6 In this paper, I seek to return to these issues of European context, but to approach them from a new angle: namely, the operation of systems of land-tax in England and West Francia. My aims are to take advantage of what is relatively well understood and to develop implications for areas of uncertainty. The English land-tax or geld is reasonably well documented and has proved ever useful for the interpretation of Domesday. One may compare the tributes raised by Charles the Bald in ninth-century West Francia: the nature of these payments, especially those for 866 and 877, is quite well documented but the possible implications for the polyptychs have been insufficiently considered. Through a clearer understanding of the West Frankish land-tax, certain points of contrast can be identified with AngloSaxon administrative practices. At the same time, certain points of connection can be identified with the geld as it operated after the Conquest, under William the Conqueror. Overall, my analysis highlights some distinctive features of the AngloSaxon geld, and makes the case for specific Continental influence on the English tax system, and on associated forms of record, in the post-Conquest period. One may begin by considering Domesday material and taxation, an important area for a number of reasons. As Galbraith famously showed, Domesday Book could not have functioned as a ‘geld-book’, but the last thirty years have also shown the benefits of understanding Domesday material in the light of fiscal concerns.7 Here, Round’s work has been of continuing relevance.8 Both Sally Harvey and David Roffe have, despite certain differences in interpretation, viewed Domesday material as reflecting royal interest in the productive capacity of the rural economy, under the pressures of taxation. From differing perspectives, each has floated the hypothesis that the Conqueror may have been considering a major reform of taxation at the time of his death.9 Moreover, if one looks beyond 1086 we have docu-

‘Some Agents and Agencies of the Late Anglo-Saxon State’, in idem The Anglo-Saxon State, London 2000, 201–25, at 219–20. For issues of survival, see Campbell, ‘Observations on English Government’, 157–8. 4 J. L. Nelson, ‘Literacy in Carolingian Government’, in idem, The Frankish World 750–900, London, 1996, 1–36, at 14–17 and 19–22. 5 Campbell, ‘Observations on English Government’, 164. 6 Loyn, ‘The Beyond of Domesday Book’, 13; Davis, ‘Domesday Book: Continental Parallels’, 23–9. 7 V. H. Galbraith, The Making of Domesday Book, Oxford 1961, 17–27. For Domesday historiography, see now S. Baxter, ‘The Domesday Controversy: A Review and a New Interpretation’, HSJ, 29, 2018, 225–93, an important contribution which appeared as this paper reached the final stages of preparation. Stephen Baxter’s new reconstruction of the making of Domesday, informed by the ‘Exon Domesday’ project, offers a perspective compatible with the present enquiry. 8 J. H. Round, ‘Danegeld and the Finance of Domesday’, in Domesday Studies, ed. P. E. Dove, 2 vols, London, 1888 and 1891, I, 77–142; idem, Feudal England, London 1895, 3–157. 9 S. Harvey, ‘Taxation and the Ploughland in Domesday Book’, in Domesday Book: A Reassessment,



Royal Taxation

137

ments which clearly did act as forms of tax record. Firstly, the Northamptonshire Geld Roll is an Old English document antedating Domesday, recording provisions for the payment of geld by each hundred in Northamptonshire.10 It may represent a more widespread form of document associated with the shire-reeve. Secondly, the Geld Account or Inquisitio Gheldi, a Latin document from 1083–4 or 1086, records liability to geld at hundredal level for several of the southwestern shires.11 All commentators are agreed that understanding how the geld was levied casts important light on Domesday material. At the same time, Anglo-Saxon historians have been impressed by this use of written records, since heregeld (‘army-money’) had its origins late in the reign of King Æthelred II. First instituted in 1012 for the purpose of paying Thorkell’s mercenary fleet based in London, the tax was levied on the hide, the standard unit of land assessment, and came to be expected regularly.12 From an Anglo-Saxon perspective, the existence of post-Conquest geld documents has seemed suggestive, in raising the possibility that the geld had from its inception made use of written records. Campbell characteristically regarded this as likely but unprovable.13 One can but acknowledge the likely significance of the document known as the County Hidage, a list of assessments for some thirteen shires in western and midland England. The combination of assessments would place the document in the early eleventh century, since it seems to postdate the reorganization of Gloucestershire under Eadric Streona, late in Æthelred’s reign.14 We simply cannot know the context which generated the Hidage, but we can imagine its utility for various purposes. My own work on exemption from geld has further explored links between taxation and Domesday material. My 2013 article focused on the exemption of manorial demesne, and grew out of my engagement with Rosamond Faith’s work on the Anglo-Saxon rural economy.15 The problem which I was wrestling with might best be expressed as a question: how far do the Northamptonshire Geld Roll and the Geld Account reflect the manner in which geld was levied in the pre-Conquest period? For it is clear from both post-Conquest documents that manorial demesne or inland was then routinely treated as exempt from geld. Yet many issues arise when one considers Domesday material directly. Again, in the last thirty years there has been increasing awareness of the importance of differ-

ed. Sawyer, 86–103; idem, ‘Domesday Book and Anglo-Norman Governance’, TRHS 5th series 25, 175–93, at 186–92; idem, Domesday: Book of Judgement, Oxford 2014, 223–6 and 235. D. Roffe, Domesday: The Inquest and the Book, Oxford 2000, 234–42. 10 Anglo-Saxon Charters, ed. A. J. Robertson, 2nd edn, Cambridge 1956, 230–6. C. Hart, The Hidation of Northamptonshire, Leicester 1970, 16–21. 11 See R. W. Finn, Domesday Studies: The Liber Exoniensis, London 1964, 97–123; idem, The Domesday Inquest and the Making of Domesday Book, London 1961, 137–51; and the ‘Exon Domesday’ project website. 12 S. Keynes, ‘The Historical Context of the Battle of Maldon’, in The Battle of Maldon AD 991, ed. D. Scragg, Oxford 1991, 81–113, at 101–2 cf. 90–100; A. Wareham, ‘Fiscal Policies and the Institution of a Tax State in Anglo-Saxon England’, Economic History Review 65, 2012, 910–32, at 917–22. 13 Campbell, ‘Observations on English Government’, 157–8 and 166–7; J. Campbell, ‘The Significance of the Anglo-Norman State in the Administrative History of Western Europe’, in idem Essays in AngloSaxon History, 171–89, at 178–9. See also A. Williams, Kingship and Government in Pre-Conquest England, c. 500–1066, Basingstoke 1999, 144–5. 14 P. H. Sawyer, From Roman Britain to Norman England, London 1978, 228–9; C. S. Taylor, ‘The Origin of the Mercian Shires’, in Gloucestershire Studies, ed. H. P. R. Finberg, Leicester 1957, 17–51, at 25–9 and 43–5. 15 D. Pratt, ‘Demesne Exemption from Royal Taxation in Anglo-Saxon and Anglo-Norman England’, EHR 128, 2013, 1–34; R. Faith, The English Peasantry and the Growth of Lordship, London 1997.

138

David Pratt

ences in recording procedure across the Domesday record, reflected in the work of Harvey, Roffe and Stephen Baxter.16 The kingdom was idiosyncratically divided into seven circuits or groupings of shires, typically formed of five shires. While broadly similar data were gathered for each circuit, the manner in which the information was represented differed significantly. Typical formulae for the assessment of estates in each circuit were listed in my earlier article.17 Ultimately, it is very difficult to reconcile the language of all circuits if read literally. Hence, this is necessarily an area of debate: I should acknowledge here some very interesting and constructive email correspondence with David Roffe on this subject a few years ago, in which we agreed to the existence of differing interpretations. My interpretation takes as its starting-point Sally Harvey’s emphasis on the formulae for Circuit I, which record two hidage assessments for each estate, one tunc or in the time of King Edward, and another, invariably lower, modo, in the time of King William. As Harvey has shown, these reductions in assessment for Circuit I seem to have applied to the manorial demesne of tenants-in-chief, as a minimum reduction.18 Equally, for Circuit II, covering the southwestern shires, the hidage assessments are represented as being from the time of King Edward; however, the Geld Account shows that in the post-Conquest period royal and baronial demesne had been treated as exempt.19 Conversely, for Circuit III, covering Middlesex, Hertfordshire, Buckinghamshire, Bedfordshire and Cambridgeshire, estates are assigned a single hidage assessment but also commonly include a separate record of hides in demesne (in dominio).20 I therefore stand with Harvey in postulating a major reassessment of geld liability, probably located close to the time of Domesday. Indeed, I would go slightly further in suggesting that it probably applied to the kingdom in its entirety. Northamptonshire lay elsewhere, in Circuit IV; the representation of inland in Yorkshire and Lincolnshire, in Circuit VI, would also make sense in this light.21 This interpretation has several significant implications. Certain elements of the fiscal structures represented in Domesday were probably of recent origin. One should regard the Northamptonshire Geld Roll and Geld Account as both reflecting the recent reassessment, and see the Geld Roll as perhaps dating from the early 1080s.22 The late Anglo-Saxon geld had a wider scope, with no general exemption of demesne land, but instances of specific exemption or ‘beneficial hidation’.23 The general exemption of demesnes formed part of the Norman contribution to post-Conquest government. It had strong tenurial dimensions in applying to the manorial demesne of manors held in dominio by the king and tenants-in-chief, and not applying to subtenants.24 The reassessment thus related to, and was shaped by, post-Conquest Norman structures of royal lordship exercised over tenants-in-chief and their landholding.

16

Harvey, ‘Taxation and the Economy’, in Domesday Studies, ed. Holt, 249–64; Roffe, Domesday: The Inquest and the Book, 169–85; S. Baxter, ‘The Representation of Lordship and Land Tenure in Domesday Book’, in Domesday Book, ed. E. Hallam and D. Bates, Stroud 2001, 73–102 and 203–8. 17 Pratt, ‘Demesne Exemption’, 31–2. 18 Pratt, ‘Demesne Exemption’, 10; Harvey, ‘Taxation and the Economy’, 257–60. 19 Pratt, ‘Demesne Exemption’, 10. 20 Harvey, ‘Taxation and the Economy’, 260–1. 21 Pratt, ‘Demesne Exemption’, 10–14. 22 Ibid., 12–13. 23 Ibid., 16–20. 24 Ibid., 10–11, cf. 15–16 and 27–8.



Royal Taxation

139

One may now turn to ninth-century West Francia, where a lengthy series of tribute payments was made to Viking armies between 845 and the end of the century. Of these, the tributes of 866 and 877 under Charles the Bald have special significance, since they are recorded in detail in the Annals of St-Bertin by Hincmar of Rheims.25 The 877 payment is additionally known from a tax edict issued by Charles, attested in two forms.26 This remarkable document records procedures similar to those described by Hincmar. Both tributes had the purpose of buying peace with Viking armies threatening the Seine; 4,000 pounds of silver were raised in 866, and 5,000 pounds in 877. In view of parallels with some details of an earlier payment, in 860–1, the two tributes probably reflect methods used more generally in the ninth century.27 Overall, the West Frankish payments show some striking contrasts with the geld, which appears to have been levied at a standard rate per hide of land. The West Frankish tax, conversely, had varying rates for different types of land unit. In particular, it specified rates for certain forms of dependent tenure, in terms which implied a strong manorial context for these forms of tenure. The West Frankish land-tax was also supplemented by other forms of levy, due to be paid by priests, by traders, by those subject to military service, and from the treasuries of churches. It is helpful to focus on the land-based taxation. According to the Annals of St-Bertin, in 866 ‘six denarii were required from each free manse’, namely, a peasant holding which owed substantial ploughing and labour services; ‘three from each servile one’, that is, a peasant holding which owed predominantly manual services; and ‘one from each accola and one denarius from every two hospitia’, in reference to two types of minor peasant holding.28 ‘Then one denarius was taken from every manse, whether free or servile’: this may describe an additional charge, perhaps necessary to reach the tribute sum.29 ‘And at length by two turns’ – that is, including the additional charge – ‘in accordance with what each of the magnates of the kingdom had by way of honores, collected in silver and wine, he collected what was agreed to be paid to the Northmen.’30 In 877, one finds a modified approach, with a reference to demesne as well as dependent manses, and an overall reduction in the burden of payments falling on the peasantry. The shortfall was now drawn from the landholder, in the form of payments assigned from the lordly rent (census dominicus). According to the Annals of St-Bertin, ‘from manses in demesne one solidus, from each free manse four denarii from lordly rent and four from the resources of the manse-holder, from every servile manse two denarii from lordly rent and two from the resources

25

See J. L. Nelson, Charles the Bald, Harlow 1992, 28–9, 213 and 250–1. The early analysis of the tributes by E. Joranson, The Danegeld in France, Rock Island IL 1923, 62–110 is subject to extended revision by S. Coupland, ‘The Frankish Tribute Payments to the Vikings and their Consequences’, Francia 26, 1999, 57–75. 26 Capitularia Regum Francorum, ed. A. Boretius and V. Krause, MGH Capitularia, Legum Sectio 2, 2 vols, Hanover 1883–97, II, 354. 27 Annales de Saint-Bertin, ed. F. Grat, J. Vielliard and S. Clémencet, Paris 1964, hereafter AB, 82–3, cf. 105; J. L. Nelson, The Annals of St-Bertin, Manchester 1991, 92, cf. 112. 28 De unoquoque manso ingenuili exiguntur sex denarii et de seruili tres et de accola unus, et de duobus hospitiis unus denarius (AB, 125–6; Nelson, Annals of St-Bertin, 130). 29 Inde de unoquoque manso, tam ingenuili quam seruili, unus denarius sumitur (AB, 126; Nelson, Annals of St-Bertin, 130). Nelson, Charles the Bald, 213; Coupland, ‘Frankish Tribute Payments’, 63. 30 Et demum per duas uices, juxta quod unusquisque regni primorum de honoribus habuit, coniectum tam in argento quam in vino ad pensum quod ipsis Nortmannis pactum fuerat persoluendum contulit (AB, 126; Nelson, Annals of St-Bertin, 130, with n. 7).

140

David Pratt

of the manse-holder’.31 The use of the lordly rent here has the appearance of a compromise which may have enabled Charles, quite impressively, to shift the burden towards landholders.32 One should observe the intriguing issue of manses in demesne. The mansus indominicatus had not been mentioned in 866, where the tax is associated only with dependent peasant tenures. The 877 arrangements were perhaps unusual in requiring payment from manses in demesne, thus taxing landholders directly in some form. In a perceptive comment prompted by the West Frankish tributes, Campbell briefly raised the possibility of comparison with the English situation, on the basis that the mansus indominicatus may have been generally exempt from the Frankish land-tax.33 I wish to develop this point further by making two observations. Firstly, for English historians, familiar with the geld, one might assume that the Frankish taxation applied to the kingdom in its entirety, but this seems unlikely.34 The fuller and more general recension of the 877 tax edict, known as version B, refers to ‘bishops, abbots, counts and vassi dominici’ giving from their honores.35 This grouping of elites is a recognizable formulation in Carolingian administration. In 807, Charlemagne had instructed his missi to make surveys of all benefices and bring these lists to the king; none has survived, but the episode was probably the key watershed for Carolingian list-making.36 As Janet Nelson has argued, it is clear from a related capitulary that Charlemagne’s survey covered the benefices of ‘bishops, abbots, abbessess, counts and royal vassals’, implying the exclusion of benefices held on the allods of lay landholders.37 Charles the Bald himself had drawn on a similar range of faithful men in compiling surveys (breues) in 869 to assist in the building of fortifications on the Seine at Pîtres.38 It therefore looks as if the 877 tax may have excluded certain forms of benefice held of lay landholders. Such an interpretation would be consistent with version A of the tax edict, which refers to vassals, since the document is limited to describing arrangements for ecclesiastical landholding. The possible variations hint at the practical limits of royal authority. Secondly, the hypothesis that demesnes were generally excluded from taxation receives strong support from the corpus of polyptychs. Much hinges on the use of the mansus as a unit of assessment. From an English perspective, one might be tempted to see the mansus as equivalent to the hide, but that would be problematic. The fiscal use of the mansus was the subject of an important article by Walter Goffart in 2008.39 The overall argument should be treated with caution, because it forms part of Goffart’s thesis of major continuity from late Roman to Carolingian taxation, aligning Goffart with claims which have generally been found problematic 31

De mansis indominicatis solidus unus, de unoquoque manso ingenuili quattuor denarii de censu dominico et quattuor de facultate mansuarii, de manso vero seruili duo denarii de censu dominico et duo de facultate mansuarii (AB, 213; Nelson, Annals of St-Bertin, 200). 32 Nelson, Charles the Bald, 249–50; Coupland, ‘Frankish Tribute Payments’, 67. 33 Campbell, ‘Observations on English Government’, 160. 34 Cf. Coupland, ‘Frankish Tribute Payments’, 65–6. 35 Capitularia Regum Francorum, ed. Boretius and Krause, II, 354. 36 Capitularia Regum Francorum, ed. Boretius and Krause, I, 136, cf. 134; Nelson, ‘Literacy in Carolingian Government’, 18–19; Davis, ‘Domesday Book: Continental Parallels’, 19. 37 Beneficia episcoporum, abbatum, abbatissarum atque comitum sive vassallorum nostrorum (Capitularia Regum Francorum, ed. Boretius and Krause, I, 177); Nelson, ‘Literacy in Carolingian Government’, 19. 38 AB, 152–3; Nelson, Annals of St-Bertin, 153–4. 39 W. Goffart, ‘Frankish Military Duty and the Fate of Roman Taxation’, EME 16, 2008, 166–90.



Royal Taxation

141

in Anglophone historiography.40 Yet Goffart makes the important point that in the Frankish world what counted for the purposes of valuing an estate were dependent mansi, that is, mansi held by individual peasant tenants.41 The mansus indominicatus, conversely, did not function as a unit of value, but rather denoted a single ‘demesne farm’ of variable size, upon which multiple peasant mansi were dependent.42 The point is borne out by the polyptychs. My investigation relies on a representative sample, comprising those for St-Germain-des-Prés, St-Bertin and Rheims, together with the information for St-Wandrille and St-Riquier, known from later summaries of lost documents.43 Entries typically start with the mansus (in)dominicatus or the variant term casa (in)dominicata, the specific size and nature of the demesne is then described, before the listing of individual dependent mansi. Crucially, as Goffart’s observations would imply, demesnes are never measured in numbers of mansi.44 One should additionally note that the terms mansus (in)dominicatus or casa (in) dominicata are also employed to describe an estate held in benefice, a usage which includes examples of small benefices with only a handful of dependent mansi.45 All this helps to contextualize the levy on mansi indominicati in 877, which seems unlikely to have represented a significant attempt to tax demesnes. That major landholders fell within the measure is implied by version A of the tax edit, but at best this would have exposed them to a simple headcount of numbers of demesne farms, each charged at one solidus irrespective of territorial scale.46 Goffart is probably right to see the measure as also referring, given the usage of polyptychs, to small demesnes held by certain types of benefice-holder. As Goffart observes, the taxation of small demesnes may have effectively replaced the role played by the heribannum, which had been levied on free men in 866.47 So it looks 40

J. Durliat, ‘Le polyptyque d’Irminon et l’impôt pour l’armée’, Bibliothèque de l’Êcole des Chartes 141, 1984, 183–208; idem, Les finances publiques de Diocletian aux Carolingiens (294–888), Beihefte der Francia 21, Sigmaringen 1990; E. Magnou-Nortier, ‘Le grand domaine: Des maîtres, des doctrines, des questions’, Francia 15, 1987, 659–700. For responses, see C. Wickham, ‘The Fall of Rome Will Not Take Place’, in Debating the Middle Ages: Issues and Readings, ed. L. K. Little and B. H. Rosenwein, Oxford 1998, 45–57; J.-P. Devroey, ‘Polyptyques et fiscalité à l’époque carolingienne: Une nouvelle approche?’, Revue Belge de Philologie et d’Histoire 63, 1985, 783–94. 41 Goffart, ‘Frankish Military Duty’, 167–73. 42 Ibid., 171–3. See also D. Herlihy, ‘The Carolingian Mansus’, Economic History Review 33, 1960–1, 79–89; G. Duby, Rural Economy and Country Life in the Medieval West, trans. C. Postan, Columbia SC, 1962, 28–42; A. Verhulst, ‘Economic Organization’, The New Cambridge Medieval History III: c. 900–c. 1024, ed. T. Reuter, Cambridge 1999, 481–509, at 493–7. 43 Das Polyptychon von St-Germain-des-Prés, ed. D. Hägerman, K. Elmshäuser and A. Hedwig, Cologne 1993; ‘Le Polyptyque de l’Abbaye de Saint-Bertin’, ed. F. L. Ganshof, F. Golding-Ganshof and A. de Sinet, Memoires de l’Institut National de France. Academie des Inscriptions et Belles-Lettres 45, 1975, 57–209; Le polyptique et les listes de cens de l’abbaye de Saint-Remi de Reims (IXe-XIe siècles), ed. J.-P. Devroey, Travaux de l’Académie Nationale de Reims 163, 1984. Chronique des abbés de Fontenelle (Saint-Wandrille), ed. P. Pradié, Belles Lettres 40, Paris 1999, 133; Hariulf. Chronique de l’abbaye de Saint-Riquier (Ve siècle-1104), ed. F. Lot, Paris 1894, 87–97 cf. 307–8. See in general J.-P. Devroey, Êtudes sur le grand domaine carolingien, Aldershot 1993; Y. Morimoto, ‘État et perspectives des recherches sur les polyptyques carolingiens (ca. 1980–1986)’, in idem, Études sur l’économie rurale du haut Moyen Âge: Historiographie, Régime domanial, Polyptyques carolingiens, Brussels 2008, 31–80. 44 Goffart, ‘Frankish Military Duty’, 170. 45 Le polyptique et les listes de cens de l’abbaye de Saint-Remi, ed. Devroey, 58–60; ‘Le Polyptyque de l’Abbaye de Saint-Bertin’, ed. Ganshof et al., 76–85. Goffart, ‘Frankish Military Duty’, 176; Y. Morimoto, ‘Problèmes autor du polyptyque de Saint-Bertin (844–859), in idem Êtudes sur l’économie rurale, 399–424, at 405–14. 46 Capitularia Regum Francorum, ed. Boretius and Krause, II, 354. 47 Goffart, ‘Frankish Military Duty’, 172–3, cf. 180–1. For the heribannum, see also Coupland, ‘Frankish Tribute Payments’, 64.

142

David Pratt

as if large-scale demesne resources contributed only lightly in 877, just as they had escaped taxation entirely in 866. It is helpful to see the Frankish use of the written word for fiscal purposes in this light. One must remember that Charles’s land-tax took its place in a world already familiar with surveys and lists.48 The compiling of polyptychs was clearly a widespread practice. Although sometimes seen to have been encouraged ‘from above’ by royal example, the recording of dues and services also had inherent utility for landholders seeking to regulate, and perhaps to uphold, the customary obligations of dependent peasantry.49 Royal estates were a further focus for the generating of written records, seemingly encouraged, as Nelson has pointed out, by practices of dynastic partition and by the conflicts of the 840s.50 There is an interesting question whether one should envisage the keeping of central records for administering the tributes.51 Perhaps one should for the taxation of traders: the Annals of St-Bertin seems to imply a formal valuation of the assets of traders in 860.52 Such a document could have been reused or updated in 866. One should also consider the arresting passage for 869 on preparations for the building of fortifications: [Charles] had despatched letters throughout the realm, requiring the bishops, abbots and abbesses to see to the drawing up of breues of their honores, showing how many manses each held. These were to be brought to the king at the beginning of May following. The royal vassals were to draw up similar surveys of the benefices held by counts, and the counts were to do likewise for the benefices held by their own vassals.53

Might some of this material have assisted in levying the tax? I think one has to answer here ‘probably yes’, while noting the limited nature of Charles’s tax demands. As argued above, demesne resources remained largely unaffected, while the main burden in respect of mansi fell directly or indirectly on the dependent peasantry. The existing character and format of polyptychs, focusing on dependent mansi, would have been well suited to the fiscal task. Nor, given the treatment of demesnes, would landholders have had substantial incentives to misrepresent their holdings. Even the taxing of the lordly rent in 877 may have broadly aligned with seigneurial interests. Furthermore, missi could sometimes take an interest in the compiling of a polyptych: Hincmar mentions a lost survey pertaining to the abbey of Hautvilliers (Marne), where ‘missi dominici could find nothing false there’.54 In practice it is very difficult to associate any surviving polyptych with Charles’s landtax. The closest case may be a lost survey for St-Vaast, reportedly made by three missi of Charles the Bald and dated 866, of which no details survive.55 Moreover, 48

Nelson, ‘Literacy in Carolingian Government’, 14–21; Davis, ‘Domesday Book: Continental Parallels’, 15–23. 49 Percival, ‘The Precursors of Domesday: Roman and Carolingian Land Registers’, 13–17; M. Costambeys, M. Innes and S. MacLean, The Carolingian World, Cambridge 2011, 252–6, with references. Cf. also Duby, Rural Economy, 50–1. 50 Nelson, ‘Literacy in Carolingian Government’, 15–16; J. L. Nelson, ‘Henry Loyn and the Context of Anglo-Saxon England’, HSJ, 19, 2007, 154–70, at 166–9. 51 Nelson, ‘Literacy in Carolingian Government’, 18. 52 AB, 152–3; Nelson, Annals of St-Bertin, 153. Nelson, Charles the Bald, 31 and 37. 53 Per omne regnum summ litteras misit, ut episcopi, abbates et abbatissae breues de honoribus suis, quanta mansa quisque haberet, futuras kalendas mai deferre curarent, uassalli autem dominici comitum beneficia et comites uassallorum beneficia inbreuiarent (AB, 152–3; Nelson, Annals of St-Bertin, 153); Nelson, Charles the Bald, 218–19. 54 Missi dominici nichil ibi falsum possint invenere (Flodoardus Remensis: Historia Remensis Ecclesiae, ed. M. Stratman, MGH SS 36, Hanover 1998, 356). 55 Cartulaire de l’abbaye de Saint-Vaast d’Arras, ed. A. van Drival, Arras 1875, 5.



Royal Taxation

143

the most detailed surviving polyptychs provide most if not all of the information that would have been needed for tax purposes. Rather than the use of central records per se, therefore, perhaps the best model is one of taxation administered by polyptych, subject to a degree of central inspection or scrutiny. This is probably what was being described for 869. One may now turn to the implications of these West Frankish experiences for thinking about England, covering four areas. The first of these concerns English tribute payments to Viking armies, both in the later ninth century and around the millennium under King Æthelred II. We are badly under-informed on the means by which English tributes were raised.56 The West Frankish material forces one to think further about the patchy English evidence, where there are some suggestive parallels. The liberation of bullion from the treasuries of churches, described for 877, for example, chimes with reports relating to the time of Æthelred. Hemming gives an account of the melting down of goblets and crucifixes by the church of Worcester to meet Swein’s tribute; Archbishop Wulfstan, in his Sermo Lupi ad Anglos, includes a complaint about the despoiling of God’s houses.57 Later Anglo-Saxon tribute payments also drew upon burdens falling in some way upon land, though the mechanisms involved cannot wholly be reconstructed. A central document is the will of King Eadred from the early 950s, in which the king made a series of substantial payments for the benefit of his people, ‘that they may redeem themselves from famine and from a heathen army if they need’.58 The payments, amounting to 1,600 pounds, were to be held by a number of ecclesiastics on behalf of particular southern shires, and for the Mercians. These arrangements strongly imply that the shire, with its associated hides, might be liable for tribute payment. One may compare the institution known as ‘ship-scot’ (scypgesceot), a monetary payment levied on the hide as a contribution towards the costs of building and provisioning ships. Ship-scot was organized in units of 300 hides; the system of payment may be connected with the ship-building measures reported in the Chronicle for 1008, though could also be earlier.59 Ship-scot is principally known from a letter by Æthelric, bishop of Sherborne, complaining of difficulties in assembling ship-scot due to encroachments on the bishopric’s estates.60 One finds a further glimpse in the will of Ælfric, archbishop of Canterbury, datable

56

For useful discussion, see M. K. Lawson, ‘The Collection of Danegeld and Heregeld in the Reigns of Æthelred II and Cnut’, EHR 99, 1984, 721–38; Keynes, ‘The Historical Context of the Battle of Maldon’, 99–102; Wareham, ‘Fiscal Policies’, 916–22. 57 Hemmingi Chartularium Ecclesiae Wigorniensis, ed. T. Hearne, 2 vols, Oxford 1723, I, 148–9; The Homilies of Wulfstan, ed. D. Bethurum, Oxford 1957, 268; Lawson, ‘Collection of Danegeld’, 726–7; Keynes, ‘The Historical Context of the Battle of Maldon’, 101, 110, n. 54 and 111–12, n. 67; L. Roach, Æthelred the Unready, New Haven CT 2016, 223–4. 58 To þan ðæt hi mege magan hungor and hæþenne here him fram aceapian gif hie beþurfen (S 1515: Charters of the New Minster, Winchester, ed. S. Miller, Anglo-Saxon Charters, hereafter AS Charters 9, Oxford 2001, 76–7, with English Historical Documents c. 500–1042, ed. D. Whitelock, English Historical Documents 1, 2nd edn, London 1979, hereafter EHD, I, 554–6); Keynes, ‘The Historical Context of the Battle of Maldon’, 112, n. 69; Wareham, ‘Fiscal Policies’, 916. 59 C. W. Hollister, Anglo-Saxon Military Institutions on the Eve of the Norman Conquest, Oxford 1962, 111–15; A. Williams, Æthelred the Unready: The Ill-Counselled King, London 2003, 80–1; P. Taylor, ‘The Episcopal Returns in Domesday’, in Domesday Now: New Approaches to the Inquest and the Book, ed. D. Roffe and K. S. B. Keats-Rohan, Woodbridge 2016, 197–217, at 203–8. 60 S 1383: Charters of Sherborne, ed. M. A. O’Donovan, AS Charters 3, Oxford 1988, 46–7; S. Keynes, ‘Wulfsige, Monk of Glastonbury, Abbot of Westminster (c. 990–3), and Bishop of Sherborne (c. 993–1002)’, in St Wulfsige and Sherborne: Essays to Celebrate the Millennium of the Benedictine Abbey 998–1998, ed. K. Barker, D. A. Hinton and A. Hunt, Oxford 2005, 53–94, at 73.

144

David Pratt

between 1002 and 1005, which shows much interest in defensive preparations. By the will’s terms, Ælfric ‘in accordance with God’s will forgave the people of Kent the debt which they owed him, and the people of Middlesex and Surrey the money which he advanced them’.61 This again indicates a liability at shire level, whether for tribute or ship-scot. Ælfric also bequeathed a ship to the people of Kent and another to Wiltshire. It remains difficult, nevertheless, to assess the relative importance of land-based contributions as opposed to other possible means of raising cash. These would include the king’s own income of all kinds, from the profits of justice, from royal and personal estates, and from dues of various kinds owed at local level. The profits of minting and recoinage could have been another source. The ‘emergency’ conditions of Viking pressure have been plausibly seen as a contributory factor in the debasement of Mercian and West Saxon coinage in the later 860s and early 870s, while Æthelred II’s rule benefited from the practice of periodic recoinages.62 The melting down of bullion, a known response, is inherently difficult to quantify. There are uncertainties over another feature of the evidence, namely, charters which show the conveying of land by landholders in order to raise cash for tribute.63 The corpus is listed in Appendix 1, below. Two examples survive from the late ninth century, and rather more from the reign of Æthelred, including some from the period before 1012, when heregeld was instituted. It is hard to assess the precise mechanisms involved: whether, for example, the action should be seen as cash-raising to meet a specific liability, or amounted to some other form of contribution. Some instances suggest circumstances of distressed sale by churches: these include the two ninth-century examples, and also the well-known case of an estate at Risborough, Buckinghamshire, sold in 994 by Sigeric, archbishop of Canterbury, to Æscwig, bishop of Dorchester, for 90 pounds of silver and 200 mancuses of gold.64 Yet one should also note the remarkable series of charters, first identified by Simon Keynes and listed in Appendix 2, below, in which Æthelred himself is represented as conveying land in return for payment.65 The sequence begins in 1002, with no fewer than six known transactions, and suggests concerted efforts to raise cash on the king’s behalf.66 It is clear that churches and ecclesiastics could themselves 61

He forgeaf on Godes est Centingan þæne borh þe hy hym sceoldan, and Middelsexon and Suðrion þæt feoh þæt heom foresceat (S 1488: Charters of Abingdon Abbey, ed. S. E. Kelly, 2 parts, AS Charters 7–8, Oxford 2001–2, 518, with EHD, I, 589–90); N. Brooks, The Early History of the Church of Canterbury, Leicester 1984, 286–7; Williams, Æthelred the Unready, 80–3. 62 D. M. Metcalf and J. P. Northover, ‘Debasement of the Coinage in Southern England in the Age of King Alfred’, Numismatic Chronicle 145, 1985, 150–76; R. Naismith, ‘The Coinage of Æthelred II: A New Evaluation’, English Studies 97, 2016, 117–39; Wareham, ‘Fiscal Policies’, 922–5. 63 For this phenomenon, see Keynes, ‘The Historical Context of the Battle of Maldon’, 100 and 102, with references; J. Campbell, ‘The Sale of Land and the Economics of Power in Early England: Problems and Possibilities’, in idem The Anglo-Saxon State, 227–45, at 231–2 and 237; D. Pratt, The Political Thought of King Alfred the Great, Cambridge 2007, 97–8 and 101; R. Naismith, ‘Payments for Land and Privilege in Anglo-Saxon England’, ASE 41, 2013, 277–342, at 293–4 and 311–12. 64 S 882: Charters of Christ Church, Canterbury, ed. N. Brooks and S. E. Kelly, 2 parts, AS Charters 17–18, Oxford 2013, 993–5, with EHD, I, 69–70; Brooks, Church of Canterbury, p. 283; Williams, Æthelred the Unready, 46; For valuable discussion of the pressures bearing on ecclesiastical landholders from tribute and geld payments, see Lawson, ‘Collection of Danegeld’, 723–34; cf. also A. Wareham, ‘The Redaction of Cartularies and Economic Upheaval in Western England c. 996–1096’, ANS 36, 2014, 189–219, at 212–13 and 216–18; Roach, Æthelred, 221–5. 65 S. Keynes, The Diplomas of King Æthelred ‘the Unready’ 978–1016: A Study in their Use as Historical Evidence, Cambridge 1980, 107–9, cf. 202, n. 182. See also Naismith, ‘Payments for Land’, 293 and 336. 66 See Roach, Æthelred, pp. 220–1, for the suggestion that some transactions may have involved the pledging of royal land for what was effectively a loan.



Royal Taxation

145

benefit from such transactions. The will of Ælfric shows the archbishop bequeathing an estate at Dumbleton, Gloucestershire, which he had previously purchased from Æthelred for 50 talents, and brokering the return to the king by St Albans of the estate of Eadulfingtun, which the abbey had earlier received as security against a payment of 200 pounds.67 One should also take seriously the notion that some contributions towards tribute may have been voluntarily offered. This is strongly implied by the wills of King Eadred and of Archbishop Ælfric. The practice might seem counter-intuitive, but it should be remembered that until 1016 tribute payments were generally being made under the pressure of ravaging. There were potentially strong incentives for magnates and churches to contribute when estates faced immediate threats to livestock and capital. Ships could also be given voluntarily as a contribution to the king’s military needs, and there were probably strong political incentives to do so, as explored in the case of Earl Godwine by Simon Keynes and Rosalind Love.68 One would not wish to push the point, but there is a possible contrast between forms of voluntary payment and the collecting of heregeld in 1012, in that the latter was needed to fund Thorkell’s mercenary fleet. It is possible that methods for collecting heregeld, as a general levy on the hide, may have related most closely to the system of ship-scot.69 Did these circumstances generate a need to enforce the liability of landholders with new legal teeth? Several commentators have observed the notable association between the geld and Cnut’s law making a landholder’s title to land contingent on the discharging of an estate’s incumbent obligations.70 The second implication of West Frankish experiences for England concerns the striking contrast between the fiscal treatment of demesne land in West Francia and Anglo-Saxon England. This observation depends on my existing case for the liability of demesnes to the geld in the pre-Conquest period, which I should briefly recapitulate. My position is that the Conqueror’s reassessment, in the early 1080s, newly exempted from geld the manorial demesne of land held in dominio by the king and tenants-in-chief. From this one may infer that inland had ordinarily been subject to geld in King Edward’s day.71 One may compare the evidence for preConquest kings granting specific concessions or reductions in hidage applying to individual estates: eighteen instances are known from Domesday, together with a small group of writs relating to hidage and geld assessment.72 Especially significant is the Bury writ giving notice that King Edward had freed the abbey’s inland ‘from heregeld and from every other render’.73 Such specific acts of exemption imply 67

S 1488: Charters of Abingdon Abbey, ed. Kelly, 518, with EHD, I, 589–90; Cf. S 901: Charters of Abingdon, ed. Kelly, 513–15; S 912: Charters of St Albans, ed. J. Crick, AS Charters 12, Oxford 2007, 179–82; Brooks, Church of Canterbury, 286–7; P. Taylor, ‘Eadulfingtun, Edmonton, and their Contexts’, in The English and their Legacy, 900–1200: Essays in Honour of Ann Williams, ed. D. Roffe, 95–113, at 95–101 and 113. Cf. the will of the Atheling Æthelstan (1014), which records three estates previously bought from King Æthelred for substantial bullion payments, S 1503: Charters of Christ Church, Canterbury, ed. Brooks and Kelly, 1037–40, with EHD I, 594–6. A facsimile is available on the ‘Kemble’ website. 68 S. Keynes and R. Love, ‘Earl Godwine’s Ship’, ASE 38, 2010, 185–223, at 187–203. 69 Wareham, ‘Fiscal Policies’, 925. 70 Harvey, ‘Domesday Book and Anglo-Norman Governance’, 183–4; Lawson, ‘Collection of Dangeld’, 723–32 and 734; Williams, Kingship and Government, 130. 71 Pratt, ‘Demesne Exemption’, 3–17. 72 Ibid., 19–20 and 33–4; D. Pratt, ‘Charters and Exemption from Geld in Anglo-Saxon England’, in Writing, Kingship and Power in Anglo-Saxon England, ed. R. Naismith and D. A. Woodman, Cambridge 2018, 181–201, at 183–8. 73 Fram heregelde. and fram eghwilc oðer gaful (S 1075: AS Writs, 158).

146

David Pratt

that estates enjoyed no general form of exemption, and that inland had ordinarily been liable. Certain instances within Domesday of territory regarded as geld-free or as having never paid geld may be satisfactorily explained as the result of specific concessions or special arrangements.74 All this reinforces the impressive nature of the pre-Conquest geld, in applying more comprehensively and in a more uniform manner than the West Frankish land-tax, which left demesne resources largely unaffected by taxation. The geld was also typically levied at higher rates. Yet the comparison indicates a still deeper contrast, in that, differing from the Continental approach, demesne land had been routinely assessed in hides in England. This system of assessment appears to have been long-established, being particularly well attested in the West Saxon law-code of King Ine (688–726). The laws in question are those concerning noblemen leaving or abandoning their estates (cc. 63–8), in various respects revealing of early land tenure and agrarian organization.75 These regulations include the requirement that a certain proportion of an estate should be left as gesett land, a term probably meaning ‘tenanted land’.76 A nobleman with 20 hides should show 12 hides of tenanted land; one with 10 hides should show 6 hides of tenanted land; one with 3 hides should show 1½ hides of tenanted land.77 These laws have generally been taken to imply some sort of bipartite manorial structure, with the remaining, non-tenanted land forming inland or demesne.78 Hence one may observe here demesne land being measured in hides, representing the distinctive, English system of assessment. It is tempting to see the context as providing possible clues to the source of divergence from Continental practice, in the deep tenurial and fiscal interests exercised by early Anglo-Saxon kings in secular landholding specifically. In other words, one might tentatively connect the English system with the pre-Viking world of competing Anglo-Saxon kingdoms, an age characterized by the use of folkland, forms of multiple estate, and networks of royal vills.79 Further important evidence is provided by the Old English survey of Tidenham, Gloucestershire, preserved in the twelfth-century cartulary for Bath abbey.80 The document constitutes a remarkable survival, as it is the only Anglo-Saxon estate survey to provide anything close to the detail of the polyptychs. It is indeed quite possible, as Campbell suggested, that the tiny surviving corpus of later Anglo-Saxon estate surveys may owe something to knowledge of the polyptych tradition; if so, the English system of assessment remained resilient.81 Tidenham had been held by Bath 74

Pratt, ‘Demesne Exemption’, 17–18; Pratt, ‘Charters and Exemption from Geld’, 188–94. See Pratt, ‘Demesne Exemption’, 22–4. 76 T. H. Aston, ‘The Origins of the Manor in England’, in Social Relations and Ideas: Essays in honour of R. H. Hilton, ed. T. H. Aston, P. R. Coss, C. Dyer and J. Thirsk, Cambridge 1983, 1–43, at 7–11; Pratt, ‘Demesne Exemption’, 23. Cf. Faith, The English Peasantry and the Growth of Lordship, 103–4. 77 Die Gesetze der Angelsachsen, ed. F. Liebermann, 3 vols, Halle 1903–16, I, 118–19. 78 Aston, ‘Origins of the Manor’, 7–11 and 38; H. P. R. Finberg, ‘Anglo-Saxon England to 1042’, in The Agrarian History of England and Wales, I.ii: A.D. 43–1042, ed. H. P. R. Finberg, Cambridge 1972, 385–525, at 339–42 and 446–8; H. R. Loyn, Anglo-Saxon England and the Norman Conquest, 2nd edn, London 1991, 170–2. 79 See Pratt, ‘Demesne Exemption’, 23–7 and 29–30. 80 S 1555: Charters of Bath and Wells, ed. S. E. Kelly, AS Charters 13, Oxford 2007, 147, with English Historical Documents 1042–1189, ed. D. C. Douglas, English Historical Documents 2, 2nd edn, London 1981, 879–80; R. Faith, ‘Tidenham, Gloucestershire, and the History of the Manor in England’, Landscape Hisory 16, 1994, 39–51. 81 Campbell, ‘Observations on English Government’, 165–6; J. Campbell, ‘What is Not Known about the Reign of Edward the Elder’, in Edward the Elder 899–924, ed. N. J. Higham and D. H. Hill, London 2001, 12–24, at 12–14. See also R. Naismith, ‘The Ely Farming Memoranda and the Economy of Late Anglo-Saxon Fenland’, ASE 45, 2017, 333–77, at 372–5. 75



Royal Taxation

147

abbey since the mid-tenth century, having been granted by King Eadwig in 956.82 Since the survey is transmitted with a lease from the early 1060s, conveying the estate to Archbishop Stigand for one life, it is most naturally read in that chronological context.83 The survey opens with a strikingly clear statement that the estate of 30 hides comprised 9 hides of inland and 21 hides of tenanted land. As is then described, the land at Tidenham was in fact divided between five different settlements between the Severn and the Wye. It is probable, as Faith has explored, that each settlement had its share of inland.84 The approach taken differs significantly from the mansus indominicatus of the polyptychs. If only more documents of this type had survived: various features of the estate’s local economy may have been unusual. The Domesday record can also assist in reconstructing pre-Conquest inland and showing that it was assessed in hides. Two circuits, Circuit II for the southwest, and Circuit III covering Middlesex, Hertfordshire, Buckinghamshire, Bedfordshire and Cambridgeshire, have special value in that their procedures included the recording of demesne hides for some landholders.85 There is thus the possibility of matching up an estate with the pre-Conquest charter record, a viable methodology but one which presents various difficulties. These relate to the impact of conquest, to uncertainties over an estate’s tenurial history after the latest pre-Conquest charter, to the scope for the reorganization of an estate, and to the difficulty of identifying ‘beneficial hidation’, which is typically not explicitly recorded. If one takes account of these potential pitfalls, some striking instances can be identified where the Domesday assessment can be precisely matched with the terms of a pre-Conquest charter. These include Tidenham itself, which in 1086 formed part of the fitzOsbern lands then in the king’s hands due to their confiscation from Roger de Breteuil in 1075. The abbot as pre-Conquest landholder is recorded as having held 30 hides, of which ten were in demesne, very close to the 9 hides of inland recorded in the Old English survey.86 The multiple components of the estate are not outlined in Domesday, but strongly implied by the recording of fisheries. A second example is Wheathampstead, Hertfordshire, an estate well known from an original diploma, dated 1060, by which King Edward conveyed 10 hides at Wheathampstead to Westminster abbey in terms conveying exemption from the common burdens.87 This accords with the Domesday entry, which records the estate as held by the abbot and assessed at 10 hides, of which 5 hides were in demesne.88 A further case is the 30 hides at Risborough, Buckinghamshire, sold by Archbishop Sigeric in 994 in exchange for money to pay tribute; the land was subsequently acquired by Archbishop Ælfric and bequeathed to Christ Church, Canterbury, in his will.89 The estate is recorded in Domesday as held by Archbishop Lanfranc, again assessed at 30 hides, of which 16 hides were in demesne.90 Finally, some especially telling examples may be drawn 82

S 610: Charters of Bath, ed. Kelly, 98–9. S 1426: Charters of Bath, ed. Kelly, 145. 84 Faith, ‘Tidenham’, 45 and 47–8. 85 Harvey, ‘Taxation and the Economy’, 260–1; Pratt, ‘Demesne Exemption’, 10, 14 and 31. 86 GDB 164a (Gloucs. 1/56). 87 S 1031 reproduced in Facsimiles of Anglo-Saxon Charters, ed. S. Keynes, AS Charters, Supplementary Series 1, Oxford 1991, no. 22; F. Barlow, Edward the Confessor, London 1970, 334–5; Pratt, ‘Charters and Exemption from Geld’, 196. 88 GDB 135a (Herts. 9/1). 89 S 882: Charters of Christ Church, Canterbury, ed. Brooks and Kelly, 993–5, with EHD, I, 69–70; S 1488: Charters of Abingdon, ed. Kelly, 518, with EHD, I, 589–90; Brooks, Church of Canterbury, 283 and 286–7. 90 GDB 143b (Bucks. 2/3). 83

148

David Pratt

from a vernacular document recording estates conveyed by Æthelwine the Black to St Albans in the mid 1040s, comprising 5 hides at Granborough, Buckinghamshire, 7 hides and 1 virgate at Redbourn, Hertfordshire, 5½ hides at Langley, Hertfordshire, and 3½ hides at Þwangtun (Fawn Wood in St Albans).91 The first three estates can be identified in Domesday with the same historic assessments. Granborough is recorded as held by the abbot and assessed at 5 hides, of which 2 hides were in demesne; Redbourn was similarly assessed at 7 hides and 1 virgate, with 3 hides and 1 virgate in demesne; Langley had also been assessed at 5½ hides in the time of King Edward, but now at 3 hides, with 2½ hides in demesne.92 These correlations appear elegant: they demonstrate what I would see as ‘fiscal normality’ in the pre-Conquest period, with demesnes assessed in hides and forming part of the overall assessment of the estate. They may be contrasted, by way of illustration, with two examples of estates which are known to have been privileged in the pre-Conquest period, and thus depart from this norm. The first is the well-known case of Chilcomb, Hampshire, held by the Old Minster, Winchester; this large estate of 68 ploughlands had been assessed at 1 hide in the time of King Edward.93 The Domesday entry bears comparison with a considerable body of forged pre-Conquest charter material, produced around the time of the millennium, purporting to support the assessment.94 It seems likely that Chilcomb enjoyed some sort of special arrangements which the forged material sought to defend. A second example is the large estate centred on Sherborne, Dorset, held in the late Anglo-Saxon period by the bishop of Sherborne and his monastic community.95 The Domesday entry records two components of this estate which are represented as unhidated and never having paid geld: namely, 16 carucates of land held by the bishop in demesne, and 9½ carucates held by the community.96 The privileged treatment probably reflected a form of ‘beneficial hidation’ within the larger Sherborne estate, and may possibly be connected with King Æthelred’s foundation charter for Sherborne of 998.97 The third implication of West Frankish experiences reaches beyond the Conquest, in that in all likelihood one should view the treatment of demesne land in Charles the Bald’s tributes and West Frankish polyptychs as significant precedents for the Conqueror’s reform of the geld. On my reading of the evidence, William inherited from Anglo-Saxon government a system of assessment more comprehensive that that of the Continental mansus, and a land-tax which generally applied equally to demesne and tenanted land. This should be contrasted with William’s reassessment, in the early 1080s, which exempted the manorial demesne of tenants-in-chief across the kingdom.98 One should note the broader debate about the Conqueror’s intentions in the early 1080s. David Roffe, interpreting the circuit formulae differently, has argued that William may in 1086–7 have had the intention of cancelling the demesne exemptions, with the aim of taxing more comprehensively.99 While now 91

S 1228: Charters of St Albans, ed. Crick, 205. GDB 145b (Bucks. 8/1); GDB 135b (Herts. 10/10); GDB 135b (Herts. 10/9). 93 GDB 41a (Hants. 3/1). 94 See Pratt, ‘Charters and Exemption from Geld’, 192–4, with references. 95 Ibid., 189–91, with references. 96 GDB 77a (Dorset 2/6); GDB 77a (Dorset 2/7). 97 Pratt, ‘Charters and Exemption from Geld’, 190–1. S 895: Charters of Sherborne, ed. O’Donovan, 39–40, with S. Keynes, ‘King Æthelred’s Charter for Sherborne Abbey, 998’, in St Wulfsige and Sherborne: Essays to Celebrate the Millennium of the Benedictine Abbey 998–1998, ed. K. Barker, D. Hinton and A. Hunt, Oxford 2005, 1–14. 98 Pratt, ‘Demesne Exemption’, 9–14. 99 Roffe, Domesday: The Inquest and the Book, 234–42. 92



Royal Taxation

149

accepting that the demesne exemptions may have been a recent concession, Roffe has continued to posit a further plan to cancel the exemptions in 1086, and has suggested a tenurial context, in the firming up of military obligations owed by tenants-in-chief.100 Meanwhile, Sally Harvey, placing the exempting of demesnes in the early 1080s, made a connection with the declining profitability of large-scale demesne farming, suggesting that the Conqueror’s reassessment may have eased the pressure on demesnes.101 The significance of the years immediately preceding 1085 for administrative innovation is now implied by Baxter’s reconstruction of the making of Domesday, which on his interpretation would have needed planning ‘probably over a long period’.102 My own view, in accepting Harvey’s case, has been to suspect nevertheless a significant tenurial context for the reassessment, in the introduction of tenure by knight service as part of the Conquest process.103 This phrase is employed here to refer to the emergence of arrangements between the Conqueror and his tenants-inchief for the supplying of warriors that were formally tenurial and personal.104 These nascent Anglo-Norman structures differed significantly from late Anglo-Saxon military recruitment, which depended on a territorial system based on hides of land.105 In response to this line of thinking, Roffe has questioned why, if the exemptions related to tenure by knight service, they came late in the Conqueror’s reign.106 One must, however, make allowances for the fact that the geld had been inherited from late Anglo-Saxon government, for the complexity of the Conquest process, and for the time needed for new structures to become embedded. As Roffe implies, it is possible that developments of the early 1080s, including the events which culminated in the invasion scare of 1085, may have prompted the Conqueror and his administrators to reform the geld in the light of the prevailing new political and tenurial relationships.107 On the present interpretation, William’s reassessment of the geld in the early 1080s involved generous reductions for tenants-in-chief, through the exempting of demesnes, in recognition of the military contribution of knights’ fees. There is the natural assumption that these military structures had Norman origins; another fundamental unit, the manerium or ‘manor’, may, as Stephen Baxter and Chris Lewis have argued, also have been a post-Conquest concept.108 It therefore seems significant that a central component of William’s reassessment, the non-liability of demesne land, had Continental precedents. It is important to remember that polyptychs remained in circulation in eleventh-century France. The continuing utility and influence of certain Carolingian polyptychs has been

100 D.

Roffe, ‘Domesday Now: A View from the Stage’, and ‘Talking to Others and Talking to Itself: Government and the Changing Role of the Records of the Domesday Inquest’, both in Domesday Now, ed. Roffe and Keats-Rohan, 7–60, at 27–30 and 45–6, and 289–303, at 297–300. 101 Harvey, ‘Taxation and the Economy’, 259–60 and 262–4. See also idem, Domesday: Book of Judgement, 213–15, 218–19 and 221–2. 102 Baxter, ‘The Domesday Controversy’, 292–3, cf. 278–92. 103 Pratt, ‘Demesne Exemption’, 15–16 and 27. 104 J. C. Holt, ‘The Introduction of Knight Service in England’, ANS 6, 1984, 89–106. 105 Pratt, ‘Demesne Exemption’, 8–9; Hollister, Anglo-Saxon Military Institutions, 64–5 and 101–2; F. M. Stenton, Anglo-Saxon England, 3rd edn, Oxford 1971, 557–60, 634–5 and 680–3. 106 Roffe, ‘Domesday Now’, 29. 107 Roffe, ‘Domesday Now’, 28–30; idem, ‘Talking to Others and Talking to Itself’, 297–300. 108 S. Baxter, ‘The Making of Domesday Book and the Languages of Lordship in Conquered England’, in Conceptualizing Multilingualism in Medieval England, c. 800–c. 1250, ed. E. M. Tyler, Turnhout 2011, 271–308, at 299–303; C. P. Lewis, ‘The Invention of the Manor in Norman England’, ANS 34, 2012, 123–50; Pratt, ‘Demesne Exemption’, 14–15.

150

David Pratt

shown by Robert F. Berkhofer in his reassessment of French monastic administration in the central Middle Ages.109 From another angle Jean-Pierre Devroey has explored the tendency for earlier surveys to receive additions or to be redacted, such interventions typically occurring in the tenth and eleventh centuries.110 Overall, the production of polyptychs seems to have been in decline, but earlier texts were still being copied, and the polyptych of Rheims includes some late tenth- and early eleventh-century components.111 From a Norman perspective, Davis rightly emphasized the significance of the lost polyptych from St-Wandrille, known from a later summary.112 Might one even envisage some memory of Charles the Bald’s taxation connected with this material? A lost survey for St-Vaast, made in 866, was still known in the twelfth century.113 The Anglo-Saxon system of assessment, measuring the demesne in hides, would not have made sense to any administrator familiar with polyptychs, nor would a land-tax levied on that basis. It is probably no coincidence that within twenty years of the Conquest one finds a major adjustment being made, which treated demesnes as exempt and shifted the burden of taxation onto the peasantry and subtenants. Not only did this bring the Anglo-Saxon system closer to Continental practices: it also supported, in the privileging of tenants-in-chief, other aspects of Norman tenurial change. Finally, the fourth implication of West Frankish experiences has to be developed most tentatively, but one should take seriously West Frankish precedents for the tenurial component within the organization of Domesday material. One is considering here the inherently hybrid nature of the way landholding is represented, organized territorially by shire but then tenurially, grouping together the holdings of each tenant-in-chief. Inevitably, many uncertainties surround the possible use of central records for geld purposes in the pre-Conquest period.114 There are, nevertheless, strong indications that the shire remained the key administrative unit. One thinks of the County Hidage, and the evidence that individual shires could be liable for payment; this bears comparison with the Northamptonshire Geld Roll and Geld Account, which were themselves shire-based taxation records.115 It is harder to envisage a role in the late Anglo-Saxon period for records organized tenurially, as in Domesday, since the notion of a ‘tenant-in-chief’ reflected the new post-Conquest structures of royal lordship.116 One must remember the different configuration of late Anglo-Saxon land tenure, involving bookland, folkland and forms of loan.117 109 R. F. Berkhofer III, Day of Reckoning: Power and Accountability in Medieval France, Philadelphia PA 2004, 10–55, commented on by Nelson, ‘Henry Loyn’, 165–6. 110 J.-P. Devroey, ‘Au-delà des polyptyques: Sédimentation, copie et renouvellement des documents de gestion seigneuriaux entre Seine et Rhin (IXe-XIIIe siècle)’, in Décrire, inventorier, enregister entre Seine et Rhin au moyen âge, ed. X. Hermand, J.-F. Nieus and E. Renard, Mémoires et documents de l’Êcole de Chartes 92, Paris 2012, 53–86. 111 Davis, ‘Domesday Book: Continental Parallels’, 22–3 and 36–9; Le polyptique et les listes dce cens de l’abbaye de Saint-Remi, ed. Devroey, 72–3 and 77–119; Devroey, ‘Au-delà des polyptyques’, 76–80. See also Campbell, ‘Observations on English Government’, 164–7; idem, ‘The Significance of the Anglo-Norman State’, 183–5. 112 Davis, ‘Domesday Book: Continental Parallels’, 18 and 31; Chronique des abbés de Fontenelle, ed. Pradié, 133. 113 Cartulaire de l’abbaye de Saint-Vaast, ed. van Drival, 5. 114 See above, p. 137, n. 13. 115 See above, pp. 136–8. 116 Baxter, ‘The Representation of Lordship’, 80; idem, ‘The Making of Domesday Book’, 292–9. 117 S. Baxter and J. Blair, ‘Land Tenure and Royal Patronage in the Early English Kingdom: A Model and Case Study’, ANS 28, 2006, 19–46; Pratt, ‘Demesne Exemption’, 23–5, 27 and 29–30.



Royal Taxation

151

It is instructive to compare the arguments put forward by Harvey, Roffe and Baxter for the direct participation of landholders in the recording of Domesday information.118 Baxter’s case concerned the fee of the bishop of Worcester, and also Bury St Edmunds; as he suggests, the practice was probably more widespread.119 Baxter has now postulated a two-stage process whereby information was initially gathered from landholders on a hundred by hundred basis (as in the Inquisitio comitatus Cantabrigiensis), then reconstituted tenurially, grouped by tenant-in-chief.120 One may observe an emerging habit of thinking tenurially within the written record. This was doubtless encouraged by a range of practicalities and incentives, but also had West Frankish precedents. One should recall Charlemagne’s survey of benefices in 807, which could have been known to Norman administrators by various means, including the references to it in Ansegisus’s capitulary collection.121 Another precedent would be Charles the Bald’s assembling together of breues from his faithful men in 869, known from the Annals of St-Bertin.122 If, as argued above, this involved the inspection of polyptychs offered up by individual landholders, then the process would be broadly analogous to the procedure which Baxter and others have posited for certain Domesday tenants-in-chief. This is not to equate Domesday fees with the typical textual structures of polyptychs, and one must allow for the uncertainties over what, precisely, had been envisaged in the 860s and 870s. Nevertheless, one can but wonder whether some memory of these earlier arrangements may have contributed to the habit of thinking tenurially, as well as territorially, within the central written record.123 This encourages a thought only previously hinted at by Loyn and Davis: namely, that the Conqueror and his administrators may have taken more general inspiration from some memory of Carolingian activity, thus providing a context for the ambitious nature of the Domesday enterprise. 118 Harvey,

‘Domesday Book and Anglo-Norman Governance’, 191–2; Roffe, Domesday: The Inquest and the Book, 140–6; Baxter, ‘The Representation of Lordship’, 79 and 82–93; Harvey, Domesday: Book of Judgement, 58–63. 119 For Bury St Edmunds, see S. Baxter, ‘Lordship and Justice in Late Anglo-Saxon England: The Judicial Functions of Soke and Commendation Revisited’, in Early Medieval Studies in Memory of Patrick Wormald, ed. S. Baxter, C. Karkov, J. L. Nelson and D. Pelteret, Farnham 2009, 383–419, at 415–16. 120 Baxter, ‘The Domesday Controversy’, 280–1 and 283–5. 121 See above, p. 140. Collectio capiutularium Ansegisi, ed. G. Schmitz, MGH Capitularia, nova series 1, Hanover 1996, 610–11. It should be remembered that Ansegisus had been abbot of St-Wandrille (823–833/4). 122 AB, 152–3; Nelson, Annals of St-Bertin, 153–4. Campbell, ‘The Significance of the Anglo-Norman State’, 183. For the transmission history of the Annals, see Nelson, Annals of St-Bertin, 15–16. Although Hincmar’s text enjoyed a limited dissemination, it was known and used at Rheims and St-Vaast in the tenth century, and formed a notable element within St-Omer, Bibliothèque Municipale, 697 + 706, a broader historical compilation written at St-Vaast or St-Bertin in the late tenth or early eleventh century: see McKitterick, History and Memory in the Carolingian World, Cambridge 2004, 50–1; S. Vanderputten, ‘Universal History as Process? Shaping Monastic Memories in the Eleventh-Century Chronicle of St-Vaast’, in Universal Chronicles in the High Middle Ages, ed. M. Campopiano and H. Bainton, York 2017, 43–64, at 57–9; H. Reimitz, ‘The Providential Past: Visions of Frankish History in the Early Medieval History of Gregory of Tours’ Historiae (sixth-ninth century)’, in Visions of Community in the Post-Roman World: The West, Byzantium and the Islamic World, 300–1100, ed. W. Pohl, C. Gantner and R. Payne, Farnham 2012, 109–35, at 130. A now-lost copy of Hincmar’s manuscript of the Annals is known to have been made for Aimoin of Fleury in the late tenth or early eleventh century. Writing in the 990s, Richer of Rheims represented his Historiae as a continuation of Hincmar’s history: Richer of Saint-Rémi: Histories, ed. J. Lake, 2 vols, Cambridge MA 2011, I, 2, cf. xvi–xvii; J. Lake, Richer of Saint-Rémi: The Methods and Mentality of a Tenth-Century Historian, Washington DC 2013, 14, 16 and 55–8. 123 I am most grateful to Chris Lewis for mentioning that a similar view of the probable influence of the polyptych tradition on Domesday will be advanced by Stephen Baxter as part of their ‘Profile of a Doomed Elite’ project. It is notable that our respective arguments have been developed independently.

152

David Pratt

Ultimately, this paper presents an argued case on a subject which tantalizes, presenting many uncertainties: the losses of Anglo-Saxon material, the limited survival of polyptychs, the inevitable debate over recording variations within the Domesday volumes, and the difficulties of pinning down particular channels of influence. This paper has explored the advantages of taking a fiscal approach to survey material: it has thus avoided any simple comparison between Frankish and English documents, by taking account of the operation of the respective land-taxes. It has been illuminating to consider the limited scope of Charles the Bald’s land-tax: this fell predominantly on peasants and minor landholders in dependent tenure, and left the direct landholding of magnates and churches largely unscathed. In particular, demesnes were not assessed in mansi, and those of major landholders either escaped taxation entirely or were levied lightly. Rather than the extensive use of centrally kept records, it seems likely that polyptychs would have been sufficient to administer the tribute payments. Indeed, the treatment of demesnes ensured a degree of confluence between royal and aristocratic interests. Comparison with West Francia highlights certain distinctive features of the Anglo-Saxon geld: the territorial nature of its organization, with liabilities at shire level; its uniformity and comprehensiveness, being ordinarily levied on inland and tenanted land; and the use of specific rather than general forms of exemption. There are also stronger signs that the geld had been administered using centrally kept records. The County Hidage, when considered alongside the Geld Roll and Geld Account, indicates the probable use in the late Anglo-Saxon period of documents organized territorially. Yet Frankish comparison also reveals a more fundamental contrast with the long-standing English system of assessment, in that inland or demesne land had been routinely assessed in hides throughout the Anglo-Saxon period. Such arrangements had origins deep in the pre-Viking past, and probably reflected the strong interests of early Anglo-Saxon kings in the fulfilment of public obligations and burdens by all types of landholder. Furthermore, for the period immediately preceding the Conquest, the recording of demesne hides in certain Domesday circuits can be used to confirm that the hidages recorded in late AngloSaxon charters ordinarily included hides of inland. Conversely, one may observe strong parallels between the Continental treatment of demesne land and the Conqueror’s reform of the geld in the early 1080s. The relevance of polyptychs and the memory of Charles the Bald’s measures may be here suspected. The Anglo-Saxon system of assessment and liability would not have made sense from a Continental perspective. Interestingly, at around this same time one finds the earliest Anglo-Norman estate survey reflecting a revised view of the demesne. The survey in question comprises an early account of the core estate of Battle abbey, studied by Eleanor Searle.124 Whereas Domesday recorded 6½ hides for the estate, of which 2½ hides were in demesne, the Battle survey reassigned the full 6½ hides of the estate to the abbey’s tenants, leaving the demesne unhidated.125 One should regard this practice, also attested in some other estate surveys, as reflecting Continental practices.126 The Conqueror’s reassessment dovetailed with another Norman innovation, the emerging practice of tenure by knight service. There were also West Frankish precedents for the tenurial component in the organization

124 E.

Searle, ‘Hides, Virgates and Tenant Settlement at Battle Abbey’, Economic History Review 16, 1963–4, 290–300. 125 Ibid., 291, 297 and 300. GDB 17b (Sussex 8/3–16). 126 For a different view of this phenomenon, cf. Roffe, Domesday: The Inquest and the Book, 157.



Royal Taxation

153

of Domesday material, notably Charlemagne’s survey of benefices and Charles the Bald’s scrutiny of the landholding of his faithful men. Finally, the overall argument of this paper supports a broader ‘cultural’ point, to do with the probable awareness within William’s circle of Carolingian precedents. At the very least, the administrative ambitions of Carolingian government remained enshrined within the capitulary material. In a tradition going back to Stubbs, then Campbell, it has been conventional to emphasize the impact of Carolingian models on later Anglo-Saxon government.127 Yet one cannot help thinking further about William’s ‘very deep speech’ at Gloucester in the winter of 1085. Thankfully, due to the impressive body of Domesday material and ‘satellites’, it is abundantly clear that William’s aims were answered. We should, nevertheless, recognize the debts that Anglo-Norman government owed to the Carolingians’ administrative legacy. Appendices: Charters and Payments in Anglo-Saxon England during Periods of Viking Invasion

1. Instances of land being conveyed in order to meet payment of tribute S 1278, dated 872. Lease by Wærferth, bishop of Worcester, of 2 hides at Nuthurst, Warwickshire, to Eanwulf, king’s thegn, for 20 mancuses of tested gold, due to the ‘immense tribute of barbarians in the same year when the pagans stayed in London’. S 354, probably late 880s or early 890s. Alfred, king of the Anglo-Saxons, to Denewulf, bishop of Winchester, and the church of SS Peter and Paul, Winchester; regrant of the reversion of 50 hides at Chisledon, Wiltshire, and 60 at Hurstbourne Priors, Hampshire, plus a payment to Denewulf of 50 pounds of pure silver, in exchange for 100 hides at Cholsey, Hagbourne and Bæstlæsford (Basildon), Berkshire. Reversion had originally been ceded on account of the scale of tribute, by implication in the 870s. S 882, dated 994. Confirmation by King Æthelred for Æscwig, bishop of Dorchester, of an estate of 30 hides at Risborough, Buckinghamshire, granted to Æscwig by Archbishop Sigeric in exchange for 90 pounds of silver and 200 mancuses of gold, in order to pay tribute to the Danes. S 912, dated 1005. King Æthelred to St Albans abbey; grant of 1 hide at Flamstead and 5 at St Albans, Hertfordshire. Abbot Leofric had given 200 pounds of gold and silver to the king to pay off the Danes, receiving in return these two estates together with 55 hides at Eadulfinctun; but the latter has now been returned for 200 pounds. S 943, dated 1006 × 1011. King Æthelred to Toti, a Dane; grant of 1 hide at Beckley and 5 at Horton, Oxfordshire, in return for 1 pound of silver for paying the tribute. S 933, dated 1014. King Æthelred to Sherborne abbey; grant of 13 or 16 hides at Corscombe, Dorset. The estate had belonged to Sherborne in Edgar’s reign, but had been leased for two lives, then sold to Ealdorman Eadric on account of the preda-

127 J.

Campbell, ‘Stubbs and the English State’, in idem The Anglo-Saxon State, 247–68, at 256–8. In respect of Anglo-Saxon legacies reflected in Domesday, cf. Campbell, ‘Observations on English Government’, 166–7.

154

David Pratt

tions of the Danes (by implication 1009 × 1012), and subsequently bought back by Wulfgar, famulus of Sherborne. 2. Diplomas of 1002 conveying land in return for payment to King Æthelred II S 900, dated 1002. King Æthelred to Ælfhelm, his faithful minister; grant of 5 hides at Codicote, Hertfordshire, in return for 152 mancuses of pure gold. Ælfhelm then grants the land to St Albans. S 901, dated 1002. King Æthelred to Ælfric, archbishop; grant of 24 hides at Dumbleton, Gloucestershire, in return for 50 talents. The land had been forfeited by a woman for adultery. S 902, dated 1002. King Æthelred to Godwine, his faithful minister; grant of 10 hides at Little Haseley, Oxfordshire, in return for 30 mancuses of gold. S 903, dated 1002. King Æthelred to Westminster abbey; grant of 2 hides at Berewican, near Tyburn, Middlesex, in return for 100 mancuses of gold, with provision for masses and psalms on the king’s behalf. S 905, dated ‘1003’ for 1002 (11 July). King Æthelred to Æthelred, his faithful man; lease, for his life and that of his wife, of land in Canterbury and six fields outside the city, in return for 7 pounds, with reversion to Christ Church, Canterbury. A facsimile is available on the ‘Kemble’ website. S 916, dated 1007. King Æthelred to St Albans abbey; grant of land at Norton, 1 hide at Rodanhangron and land at Oxhey, Hertfordshire, previously forfeited to King Æthelred by Ealdorman Leofsige, and purchased for St Albans by Archbishop Ælfric and Abbot Leofric. Purchase probably occurred in 1002. A facsimile is available on the ‘Kemble’ website. Also later examples of payments to King Æthelred: S 910, 912, 915, 919, 923, 930 and 943 (cf. S 960). See Keynes, Diplomas, pp. 107–9.

The Marjorie Chibnall Memorial Essay, 2018

EARLY ROYAL RIGHTS IN THE LIBERTY OF ST EDMUND* Richard Purkiss Pre-Conquest kings laid claim to a variety of general rights. Apart from longstanding prerogatives such as the minting of coin, the booking of land and the ‘common burdens’, these had by the eleventh century come to include several which were regularly granted to subjects as franchises.1 Few received franchisal rights to compare with those of the abbots of Bury St Edmunds, whose later claims have been taken to preserve those of Anglo-Saxon kingship itself. This paper reviews evidence from the abbey’s liberty and further afield to limit the content of soke, the most common franchise, before examining the nature of apparently public obligations. Bury St Edmunds acquired what would become its liberty, meaning the territory in which public functions were delegated to the abbey, from Edward the Confessor. The king announced his grant of ‘the sokes of the eight and a half hundreds pertaining to Thingoe’ to St Edmund in a writ of 1043 × 1044, formally submitting a set of royal administrative units to the abbot.2 Thingoe was a barrow – the ‘assembly mound’ – just north of the town, where the hundreds met as a group.3 Between them, the hundreds of Thingoe, Thedwestry, Blackbourn, Bradmere, Lackford, Risbridge, Babergh (double) and Cosford (half) covered the western third of Suffolk and would make up the Liberty of St Edmund until the Dissolution (Map. 1). The eight-and-a-half hundreds lend themselves to close study thanks to an unusual concentration of documentary material. Evidence for the workings of a particular franchise at a given time or place can rarely be held up against independent records, but at Bury St Edmunds a well-preserved archive sheds light on the abbey’s privileges from several directions.4 Royal writs survive in a coherent run from the beginning of Edward the Confessor’s reign, proclaiming grants of rights and land. An extensive series of vernacular wills begins in the mid-tenth century and, along with benefactor lists, reveals the provenance of many estates. Both landholdings and rights are recorded in three separate post-Conquest surveys. Earliest among these is Little Domesday Book (LDB), incorporating information reaching back to 1066. Although a royal record, LDB seems to have been influenced by the abbot where his holdings were concerned.5 The Feudal Book of Abbot Baldwin *

I am grateful to John Blair, Ros Faith, Sarah Foot, Paul Hyams, Tom Lambert and Richard Sharpe for their comments on earlier versions of this paper. Sarah Foot and Richard Sharpe, as well as Susan Oosthuizen, have shown me unpublished material. The views expressed are my own. 1 Rights alienated in surviving documents are described by Harmer, AS Writs, 73–82. 2 S 1069. 3 O. S. Anderson, English Hundred-Names, Lund 1934, 83, 95–6. 4 A new edition of the pre-Conquest documents of Bury St Edmunds is in preparation: Charters of Bury St Edmunds, ed. K. Lowe and S. Foot, Oxford forthcoming. 5 S. Baxter, ‘The Representation of Lordship and Land Tenure in Domesday Book’, in Domesday Book, ed. E. Hallam and D. Bates, Stroud 2001, 73–102 at 93; idem, ‘Lordship and Justice in Late Anglo-Saxon England: The Judicial Functions of Soke and Commendation Revisited’, in Early Medieval Studies in

156

Richard Purkiss

(1087 × 1098) is related to, but largely independent of, LDB and records the abbey’s holdings for domestic purposes.6 Finally, the Kalendar of Abbot Samson (1186 × 1191) describes revenues which the abbey received from the free population of its hundreds.7 Nowhere else can privileges conferred upon a late Anglo-Saxon religious house be traced in such detail. Map 1: The eight-and-a-half hundreds of Bury St Edmunds within East Anglia

Memory of Patrick Wormald, ed. S. Baxter, C. E. Karkov, J. L. Nelson, and D. Pelteret, Farnham 2009, 383–419 at 412–16. 6 Feudal Documents from the Abbey of Bury St Edmunds, ed. D. C. Douglas, Records of the Social and Economic History of England and Wales 8, 1932, 3–44. The Feudal Book falls into three distinct sections, now sometimes labelled Bury A–C. On its internal structure, see l–lxi. 7 The Kalendar of Abbot Samson of Bury St Edmunds and Related Documents, ed. R. H. C. Davis, Camden Society 3rd series 84, 1954.



Early Royal Rights in the Liberty of St Edmund

157

This paper’s initial concern is with soke, a franchise that has proved resistant to satisfactory definition. More exclusive rights acquired by Bury St Edmunds (the so-called ‘reserved pleas’) will not be addressed here.8 Kings began to grant soke, invariably by writ, early in the eleventh century.9 It was unusual among AngloSaxon franchises in that its name was not obviously descriptive, and there has been much debate as to its precise content. The Old English socn (‘a seeking’) was often combined with sacu (‘a cause, dispute’) into the alliterative formula sacu ⁊ socn (‘sake and soke’).10 ‘Soke’ and ‘sake and soke’ are usually held to have been synonymous.11 A grant entitled the beneficiary to take royal fines (wites) from those over whom his soke extended.12 Discussion of soke long concentrated on justice, the main point at issue being whether it conferred the right to hold a private court.13 It is now widely accepted that evidence for truly private courts, or judicial immunities, is lacking and that soke carried a right only to judicial profits.14 Apart from fines, however, soke has also become associated with customary dues owed to the king. Paul Vinogradoff was the first to suggest that it could encompass feorm, the traditional royal food-rent. On his reading, soke might denote ‘all the rights accruing to the king from his subjects’.15 This wider definition gained currency through R. H. C. Davis’s edition of the Kalendar of Abbot Samson of Bury St Edmunds. The Kalendar lists in minute detail the payments which the abbey claimed by virtue of holding the eight-and-a-half hundreds. For Thedwestry, Thingoe and Blackbourn (by 1186 × 1191 a double-hundred incorporating Bradmere), information is recorded in a regular manner. The letes into which vills were grouped for geld purposes are described first, followed by the suits owed from each vill to the hundred court. The third and longest section lists landholdings liable for public obligations and the dues with which they were burdened. The account of Cosford and Babergh hundreds is less ordered and apparently in draft form, while Lackford and Risbridge are not covered. A wide variety of dues were collected: hidage (hidagium, a money rent or tribute paid by freemen), guard-service (commuted to warpeni), carrying-service (commuted to averpeni), horse-feed (foddercorn), mowing-service (falcare), sheriff’s aid (auxilium), reliefs (releva), merchet (gersuma pro filiabus), and payments for suit of court and view of frankpledge.16 8 Bury St Edmunds had the pleas of hamsocn, griðbryce, forsteal, fyrdwite, æbære þeof, and fihtwite, known locally as the ‘six forfeitures’. On the significance of the ‘reserved pleas’, see F. W. Maitland, Domesday Book and Beyond: Three Essays in the Early History of England, Cambridge 1897, 87–9, 282–3; N. Hurnard, ‘The Anglo-Norman Franchises’, EHR 64, 1949, 289–327, 433–60; T. Lambert, ‘Royal Protection and Private Justice: A Reassessment of Cnut’s “Reserved Pleas”’, in English Law Before Magna Carta: Felix Liebermann and Die Gesetze der Angelsachsen, ed. S. Jurasinski, L. Oliver, and A. Rabin, Leiden 2010, 157–75. 9 Earlier documentary references (S 659, 681, 1453) do not concern soke as a free-standing privilege, but rather land the soke of which lay elsewhere. References in S 779 may represent later usage. 10 J. Bosworth and T. N. Toller, An Anglo-Saxon Dictionary, Oxford 1898, s.v. ‘sacu’, ‘sócn’. 11 Maitland, Domesday Book and Beyond, 84. Cf. D. Roffe, Domesday: The Inquest and the Book, Oxford 2000, 28–46. 12 J. Hudson, The Oxford History of the Laws of England, Volume II: 871–1216, Oxford 2012, 59–61. 13 Maitland, Domesday Book and Beyond, 86–103 thought that soke-holders could convene their own courts, while J. Goebel, Felony and Misdemeanor: A Study in the History of Criminal Law, New York 1937, 339–61 argued that their right was to profits only. The debate was reviewed by P. Wormald, ‘Lordship and Justice in the Early English Kingdom: Oswaldslow Revisited’, in Property and Power in the Early Middle Ages, ed. W. Davies and P. Fouracre, Cambridge 1995, 114–36 at 114–20. 14 Wormald, ‘Lordship and Justice’, 120–30; S. Baxter, The Earls of Mercia, Oxford 2007, 211. 15 P. Vinogradoff, English Society in the Eleventh Century, Oxford 1908, 122 n. 1. 16 Kalendar, xxxii–xl.

158

Richard Purkiss

Davis was certain that all these dues were hundredal, meaning that they would have gone to the king had the hundreds not been granted to the abbot.17 His conclusions can be summarized as follows: (1) The nature of the dues and their association with socage land suggest that they derived from services once owed by the free population to itinerant kings.18 They therefore probably originated in the pre-Viking kingdom of East Anglia.19 (2) As the dues were assessed on carucates and ware acres belonging to a hundred, and usually collected by the hundred-reeves, they were attached to the hundreds.20 They were not manorial in origin but were sometimes collected at manors for convenience.21 (3) The dues are concealed within LDB records of soca held by the abbot, which demonstrably reflected possession of the hundreds. Soca thus encompassed both ‘jurisdiction’ and other customs which the land owed to the king.22 This analysis was widely accepted and now underpins a consensus that defines soke as judicial profits plus customary dues.23 It has also shaped perceptions of what counted as a public obligation. Customs such as guard-service or cartage, owed by freemen, have long been associated with extensive territories based on royal estates.24 The Kalendar depicts apparently archaic dues demanded from the free population as general rights tied to the hundred. In this it has helped to shape the influential inland–warland model of rural organization: inland representing heavily exploited demesnes worked largely by unfree labour, and warland meaning all the lands beyond, cultivated by freemen owing light, defined and respectable services to the king.25 Current thinking on royal rights before the Conquest plainly owes a great deal to conclusions drawn from the Kalendar’s account of the Bury hundreds, but these have not been paralleled since Davis’s edition appeared in 1954. All references to customary dues falling within soke ultimately rely on the Kalendar, sometimes quoting circumstantial evidence from other sources in support. Similarly, Abbot Samson’s survey remains the clearest illustration of such dues being treated as an 17

Kalendar, xiii–xv. Ibid., xxxv–vi, xli–ii. 19 Ibid., xlvi–vii. 20 Ibid., xxxii–iii. 21 Ibid., xxxiv–v. 22 Ibid., xl–xli. 23 C. A. Joy, ‘Sokeright’, PhD thesis, University of Leeds 1975, 22–3, 247–8; S. P. J. Harvey, ‘Domesday England’, in The Agrarian History of England and Wales, Volume II: 1042–1350, ed. H. E. Hallam, Cambridge 1988, 45–136 at 72; M. Chibnall, Anglo-Norman England 1066–1166, Oxford 1993, 137; R. Faith, The English Peasantry and the Growth of Lordship, London 1997, 90, 118; D. Roffe, ‘Introduction’, in Little Domesday Book: Norfolk: Introduction, Translation & Indices, ed. A. Williams and G. Martin, London 2000, 9–45 at 27–8; Baxter, The Earls of Mercia, 210–11; T. Williamson, Environment, Society and Landscape in Early Medieval England: Time and Topography, Woodbridge 2013, 111–12. 24 F. W. Maitland, ‘Northumbrian Tenures’, EHR 5, 1890, 625–32; J. E. A. Jolliffe, ‘Northumbrian Institutions’, EHR 41, 1926, 1–42; G. W. S. Barrow, ‘Pre-Feudal Scotland: Shires and Thanes’, in idem, The Kingdom of the Scots, 2nd edn, Edinburgh 2003 (first published 1973), 7–56 at 8–22, 38–9; G. R. J. Jones, ‘Multiple Estates and Early Settlement’, in English Medieval Settlement, ed. P. H. Sawyer, London 1979, 9–34 at 20–9; T. Williamson, The Origins of Norfolk, Manchester 1993, 94–102; Williamson, Environment, Society and Landscape, 29–30. 25 Faith, English Peasantry, ch. 2 and 4 respectively; most recently, see T. Lambert, Law and Order in Anglo-Saxon England, Oxford 2017, 119–23. D. Pratt, ‘Demesne Exemption from Royal Taxation in Anglo-Saxon and Anglo-Norman England’, EHR 128, 2013, 1–34, argues that the inland-warland scheme distorts the incidence of public obligations. However, it captures a division of real consequence in rural society, between people who worked towards someone else’s obligations and individuals who might encounter public authority more directly. 18



Early Royal Rights in the Liberty of St Edmund

159

overarching royal right. With much resting on the Kalendar, arguments rooted in this late account of Bury’s rights are worth re-examining. Three questions suggest themselves for this purpose: were the recorded dues part of soke as understood in the eleventh century; were they an overarching royal prerogative organized through public institutions; and were they ‘archaic’? For each in turn, we shall examine the evidence from the eight-and-a-half hundreds before considering whether it finds support elsewhere. Customary Dues and Soke As a result of Edward the Confessor’s grant of the eight-and-a-half hundreds to the abbey, LDB attributes soca to Bury St Edmunds by default within their bounds.26 If the Kalendar dues formed a subset of soca, this would have been how the abbey acquired them. Such an interpretation sees the Kalendar as a uniquely detailed, if late, description of soke rights, but assumes that no detailed evidence survives from before the late twelfth century. The possibility of identifying the Kalendar dues more specifically in LDB has not previously been recognized. This earlier point of comparison will clarify their relationship to soke and provide a means of tracing their origins. The key lies in a distinctive feature of LDB entries relating to Bury’s interests in the eight-and-a-half hundreds. Unusual characteristics have been attributed to Abbot Baldwin’s fee for some time. Scribal peculiarities, and frequent references to freemen owing both soke and commendation to the abbot, have been taken to suggest that Baldwin dictated a favourable account.27 The fee has also been thought to avoid recording freemen’s names.28 A further, hitherto overlooked, feature is the care taken to record dues claimed from the free population. Entries recording freemen and sokemen under the Bury fee and within the eight-and-a-half hundreds commonly describe them as owing servitium or consuetudo. Freemen usually owed servitium, and sokemen consuetudo, although the rule is not absolute. Domesday Book as a whole employs the word servitium only sporadically, in various contexts.29 Only the portion of the Bury fee relating to the eight-and-a-half hundreds uses it in a deliberate and regular manner, to denote dues levied on the free population. Consuetudo is more common, particularly in LDB, and can refer to a wide range of obligations. It has in the past been treated as consuetudo regis, supposedly an early royal tax pre-dating the geld.30 But ‘custom’ describes the basis for a due’s collection, rather than its nature. LDB has examples of consuetudo representing soke, commendation, higher forfeitures, suit to the hundred, or extorted payments. In the Bury St Edmunds fee, within the Bury hundreds, it is consistently used in a way that indicates dues collected from a particular social group.31 It was evidently

Blanket soca reflects the alienation of the hundreds: M. Bailey, ‘Introduction’, in Little Domesday Book: Suffolk: Introduction, Translation & Indices, ed. A. Williams and G. Martin, London 2000, 9–30 at 10. The same can be observed elsewhere in East Anglia. 27 Baxter, ‘Lordship and Justice’, 415–17 and n. 161. People seem to have avoided commending themselves to the lord who had their soke, as this could work against their interests. 28 Ibid. 29 See J. D. Foy, Domesday Book 38: Index of Subjects, Chichester 1992, s.v. ‘service’. 30 E. B. Demarest, ‘Consuetudo Regis in Essex, Norfolk and Suffolk’, EHR 42, 1927, 161–79. 31 A series of entries in Hartismere hundred form a partial exception, but their consuetudo was rendered at the Bury manor of Rickinghall [Inferior], which lay in the abbey’s hundred of Blackbourn (see Map 2): LDB 370b (Suffolk 14/129–36). 26

160

Richard Purkiss

important to the abbey that its right to servitium and consuetudo from the free population should be recorded. What was meant by these terms can be inferred from the GDB accounts of nearby Cambridgeshire and Hertfordshire. In many Cambridgeshire entries, sokemen are recorded as rendering avera (cartage), inward (guard), or a sum in commutation. These services plainly resembled those commuted for the Kalendar’s averpeni and warpeni.32 Most of the relevant entries describe obligations as being discharged to ‘the sheriff’ or ‘the King’s sheriff’, but some use the phrase in servitio regis.33 Hertfordshire often describes commuted avera or inward as a consuetudo collected by the sheriff. These parallels suggest that the servitium and consuetudo of the eightand-a-half hundreds mean cartage or guard-service. If so, two of the Kalendar dues can be located in LDB. To secure this identification, we must show that specific holdings which owed servitium or consuetudo in 1066 still owed averpeni, warpeni or both in 1186 × 1191. Concrete examples, as opposed to likely ones, are few. More than a century of enfeoffments and partible inheritance had reshaped the free communities, but the extracts reproduced in Table 1 put the proposition on a firm footing.34 Table 1: Dues recorded in LDB (1086) and Kalendar (1186 × 1191) West Stow, Bradmere [Blackbourn] hundred LDB Kalendar In [West] Stow 21 free men with 2 carucates of land. Always 6 ploughs between them all. Meadow, 2 acres. These [free men] belong to St [Edmund’s] with sake and soke and every customary due, and serviunt in Lackford. Value always 20s.

32

In West Stow, which is half of the third lete [of Blackbourn double-hundred] as against Norton [which is the other half], there is 1½ carucate of socage land which is divided thus: [List of individual holdings] All the aforesaid sokemen except six pay reliefs and merchet, and attend pleas at the hall of Lackford, and give warpeni there, to wit 14d., and collect among themselves 3s. which they give to the hall of Lackford for ploughing and reaping. And the hall of Risby shares all incidents of West Stow with the hall of Lackford, except those 3s. and except half a mod [?of oats] which the hall of Lackford has in West Stow from a certain rented holding which is called Kingsland. The advowson is the lord abbot’s. Sum of foddercorn: 13 loads. Sum of hidagium: 15s. 7d.

In one case, the commutation is explicitly called warpeni: GDB 190a (Cambs. 1/18). Unlike the Kalendar, the Cambridgeshire Domesday also mentions heueward, albeit only in four instances: GDB 200b (Cambs. 32/4; 7). Inquisitio Comitatus Cantabrigiensis, ed. N. E. S. A. Hamilton, London 1876, 34 adds one more. See Bosworth and Toller, Anglo-Saxon Dictionary, s.v. ‘heáfod-weard’. 33 E.g. GDB 189d (Cambs. 1/12); 197b (Cambs. 22/6); 202b (Cambs. 41/3). Further afield, a Kent entry refers to avera id est servitium: GDB 9d (Kent 5/138). 34 LDB 364a (Suffolk 14/71) and Kalendar, 40–1; LDB 362b (Suffolk 14/55) and Kalendar, 15–16. LDB translations adapted from the Phillimore edition; Kalendar translations my own.



Early Royal Rights in the Liberty of St Edmund Woolpit, Thedwestry hundred LDB [In Woolpit] 40 free men with 1 carucate of land. Always 6 ploughs. Meadow, 3 acres. These [free men] could all grant and sell their lands TRE, but the soke over the land, and every servicium, always belonged to St [Edmund’s], and all belong to the fold. Value 10s. 8d.

161

Kalendar In Woolpit there is 1 carucate of socage land which these [individuals] hold. John de Herst and his two parceners, to wit [blank], hold 15 acres but of this John alone holds 10 acres. He renders 16d., 1[d.] warpeni, and 2d. to the sheriff. These 15 acres are no freer than the other [holdings of] 15, and owe 1 load of oats, but it is kept back by the reeves. Robert son of William, and Edric, and Richard son of Arnold hold 15 acres. They render 16d. and 1 load of oats. [Six further 15-acre holdings paying the same amount] The men of this vill collect among themselves 17d. in warpeni, and as much in sheriff’s aid. [List of other holdings in the same vill] The advowson of this vill is the lord abbot’s. Woolpit mows 9 rods of meadow. [It renders] 4[d.] averpeni.

The case of West Stow has particular weight because it shows a consistent anomaly. In 1066, the abbey collected 20s. from 2 carucates by virtue of having soke. This payment is recorded at 1d. per acre or 16d. per 15 acres in both LDB and the Kalendar (where it is called hidagium), and probably originated as a fixed royal ‘tribute’.35 By the time of the Kalendar, the socage land at West Stow had shrunk to 1½ carucates and its hidagium to 15s. 7d. This basic rent went directly to the abbot, reflecting the soke ascribed generally to St Edmund. Servitium, on the other hand, was conspicuously rendered at a certain place, and so were 14d. in warpeni a century later. Guard-service due at Lackford hall evidently remained distinctive and noteworthy. At Woolpit, holdings and dues saw even less change in the intervening period. A carucate held by freemen in 1066 yielded the abbot 10s. 8d., and the Kalendar lists eight 15-acre holdings each paying 16d. in hidagium, for an identical total. In addition, LDB notes servitium and the Kalendar shows that the men of the vill collectively gave 17d. in warpeni and 4d. in averpeni. Where LDB mentions servitium in the eight-and-a-half hundreds, it is undoubtedly referring to the guardobligations (and possibly cartage) later itemized in the Kalendar.36 Similar examples can be traced for consuetudo, although this term was probably capable of encompassing further obligations, such as agricultural boon-works.37 The implications are significant. Most immediately, the argument that warpeni and the like formed a subset of soke cannot be sustained. Wherever servitium and

35

Feudal Documents, cxxii–cxxvii; Kalendar, xxxvii–xl; D. Roffe, Decoding Domesday, Oxford 2007, 246. 36 R. H. C. Davis himself traced several Kalendar statements back to LDB notices of servitium. He was also aware of GDB’s records of avera and inward in Cambridgeshire, but did not notice descriptions of these dues as servitium or consuetudo. The possibility that LDB’s employment of these terms in the eight-and-a-half hundreds might be deliberate therefore escaped him. 37 Thus consuetudo seems to record warpeni in Ingham: LDB 364a (Suffolk 14/69) and Kalendar, 43. Its meaning in Pakenham may be less specific: LDB 361b (Suffolk 14/49) and Kalendar, 8–10.

162

Richard Purkiss

consuetudo appear, it is alongside soke.38 A typical entry shows the abbey having sake and soke, commendation and servitium over several freemen, while a group of sokemen may be held with sake and soke and every consuetudo. Soke and the dues marked by servitium and consuetudo must therefore have been separate. This removes the chief foundation for the view that soke comprised customary services as well as judicial rights. Miscellaneous documents which have been adduced in support are unconvincing without the Kalendar behind them. Vinogradoff took the phrase socn to anre niht feorme, in a charter purportedly issued by Edgar, to mean that soke could encompass feorm.39 This isolated expression is perhaps better understood as ‘entitlement to feorm’, rather than as a statement bearing on the content of soke. More decisively, soke was a universal royal right, but Cnut explicitly gave up any claim to collect feorm from the general population just as soke was beginning to be treated as a grantable franchise.40 Terms such as fold-soke (soca faldae) have also been quoted in favour of a wide definition of formal soke.41 LDB’s treatment of fold-soke suggests a literal ‘seeking the lord’s fold’ rather than anything connected with a royal privilege.42 The main approach has been to point to the ample evidence for sokemen owing agricultural and other dues.43 While these obligations are not in doubt, it does not follow that soke equalled whatever people described as sokemen owed. Only a particular interpretation of the Kalendar supports such a link, and once it is discounted, customary dues can safely be removed from soke. Practical Services as Royal Rights If customary dues were separate from soke, the question arises whether they were royal prerogatives in the same sense. No authentic document pre-dating the reign of William II (1087–1100) mentions averpeni or warpeni, or treats any similar due as a franchise.44 Cartage and guard-service have nonetheless been interpreted as general rights claimed by the king over warland holdings.45 As such 38

R. Welldon Finn, Domesday Studies: The Eastern Counties, London 1967, 151–2, noted this but saw no way of identifying what might lie behind servitium/consuetudo. 39 Vinogradoff, English Society, 122 n. 1; S 779. 40 II Cn 69:1 (Gesetze, ed. Liebermann, I, 356–7). Cnut’s secular code was issued in 1020 or 1021. The first authentic writ granting soke, S 986, survives from 1020. The Northamptonshire Geld Roll’s separation of gewered land and kynges agen ferme land implies a continued distinction between public obligations and feorm: Robertson, AS Charters, Appendix I, no. 3. 41 Baxter, Earls of Mercia, 211. 42 Fold-soke meant a tenant’s obligation to place his sheep in the lord’s fold, so that the lord’s acres would receive their manure: Maitland, Domesday Book and Beyond, 76–7; D. C. Douglas, The Social Structure of Medieval East Anglia, Oxford Studies in Social and Legal History 9, Oxford 1927, 78–9. 43 Joy, ‘Sokeright’, 245–8; Baxter, Earls of Mercia, 259–61 and n. 220; Williamson, Environment, Society and Landscape, 111–12. 44 For grants by William II, see Early Charters of the Cathedral Church of St. Paul, London, ed. M. Gibbs, Camden Society 3rd series 38, 1939, nos. 10, 14, 17. Some of these documents assume that their franchises had existed in earlier reigns. Pre-Conquest grants of weardwite are not authentic: Harmer, AS Writs, 84. 45 Kalendar, xxxv; Joy, ‘Sokeright’, 240–8; R. P. Abels, Lordship and Military Obligation in AngloSaxon England, Berkeley 1988, 144; Roffe, Domesday, 41; D. M. Hadley, The Northern Danelaw: Its Social Structure, c. 800–1100, London 2000, 183–4; J. Campbell, ‘Hundreds and Leets: A Survey with Suggestions’, in Medieval East Anglia, ed. C. Harper-Bill, Woodbridge 2005, 153–67 at 162; S. Oosthuizen, ‘Sokemen and Freemen: Tenure, Status and Landscape Conservatism in Eleventh-Century Cambridgeshire’, in Anglo-Saxons: Studies Presented to Cyril Roy Hart, ed. S. Keynes and A. P. Smyth, Dublin 2006, 186–207 at 196–7.



Early Royal Rights in the Liberty of St Edmund

163

they would have been amenable to alienation by writ and might therefore, albeit silently, have passed to the abbey with the hundreds. The extent of servitium and consuetudo obligations, and their relationship to the hundreds, offer indications as to their status. It is immediately apparent that servitium and consuetudo cannot have been universal rights akin to soke. Holding the hundreds meant that St Edmund had soke over all freemen not explicitly subject to someone else, but LDB shows that only some of them owed customary dues. Over much of the free population, the abbey had soke and commendation, or soke alone, with no mention of further obligations. Given that records of servitium and consuetudo are distinctive to the Bury fee and the eight-and-a-half hundreds, and benefited the abbey, they most likely derive from material provided by the monks.46 There is, therefore, every reason to think that LDB records all instances of these obligations current at the time of the survey. Conversely, where soca alone is recorded, the freemen in question probably owed no such services. This is best illustrated by tabulating LDB entries over a set area, such as the sprawling vill of Bradfield, which shows the pattern in miniature (Table 2). Holdings in Bradfield make it clear that servitium and consuetudo were not universal to the local free population.47 Table 2: LDB holdings in Bradfield, Thedwestry hundred Holding Bradfield I

Acreage 20 acres

Holder Population 1 free man commended to Bishop Æthelmær St Edmund 15 villani; 18 bordarii; 1 slave 3 free men commended to St Edmund

Bradfield IIa Bradfield IIb

3 carucates (manor) 24 acres (valued in manor)

Bradfield III

1 carucate

9 free men

Bradfield IV

2½ carucates

10 free men

?holders; 2 bordarii ?holders; 12 bordarii

Obligations Soke [not described] Sake and soke; every consuetudo (could not sell lands without permission) Soke; servitium (could sell lands) Soke (could sell lands)

This fact could in itself suggest that the dues had not passed to the abbey through the hundredal grant, and their uneven distribution points to the same conclusion. If customary dues had been public and hundredal, one might expect individual hundreds to have owed them at somewhat comparable levels, but this was not the case. Holdings in the Bury fee, within the eight-and-a-half hundreds, which LDB associates with servitium or consuetudo run to 9,928 acres: 738 in Babergh (double-hundred), 1,890 in Blackbourn, 420 in Bradmere, 1,140 in Cosford (half46

See n. 5 above. The holdings tabulated here are described in the following entries: LDB 291a (Suffolk 2/3) for ‘Bradfield I’; LDB 362a (Suffolk 14/52) for ‘Bradfield IIa’, ‘Bradfield IIb’, and ‘Bradfield III’; LDB 362b–363a (Suffolk 14/59) for ‘Bradfield IV’. The vill covered what would become the ancient parishes of Bradfield St George, Bradfield St Clare, and Bradfield Combust. 47

Map 2: The distribution of servitium and consuetudo obligations in the eight-and-a-half hundreds according to LDB



Early Royal Rights in the Liberty of St Edmund

165

hundred), 534 in Lackford, none in Risbridge, 4,026 in Thedwestry, and 1,180 in Thingoe.48 In some places, the hundredal framework was manifestly ignored. Arrangements such as those at West Stow, where obligations were due at a manor in a different vill, disregarded hundred boundaries (Map. 2). If customary dues did not come to the abbey with the hundreds, they probably did not originate with the king at all. While the hundredal grant was not Bury’s only royal privilege, other grants can be examined with customary dues in mind. Several series of writs by Edward the Confessor in favour of the abbey survive.49 It seems that from early in the reign each abbot had writs granting him: (1) the monastery and everything belonging to it in land and in ‘sake and soke’;50 (2) the soke of the eight-and-a-half hundreds;51 and (3) the exemption of Bury’s demesne lands from scot and geld.52 Individual pre-Conquest writs also proclaim the grant, at one time or another, of: (4) ‘sake and soke’ over the abbey’s men;53 and (5) ‘sake and soke’ over any lands bequeathed to the abbey.54 The first series announced the transfer of existing assets to a new abbot, the second concerned the hundredal grant already discussed, and the third was a strictly fiscal privilege. The fourth and fifth types gave the abbey soke over its interests irrespective of location, even outside the hundreds. In view of the distinction established between soke and the Kalendar dues, therefore, none of Bury’s royal writs is likely to conceal a grant of customary services. There is little indication that the abbey itself considered its customary dues royal in the eleventh century. Bury’s thirteenth-century Customary states that the rights held ‘because they are royal’ by Abbot Robert II (1102–7), comprised hidagium, foddercorn and suits to the hundred.55 If accurate, this would imply that warpeni and averpeni were not considered royal at this time. Although the Customary is a late source, it receives corroboration from Abbot Baldwin’s Feudal Book. The third section of this survey (Bury C), which gives a detailed account of socage holdings, lists payments representing hidagium but nothing to match the servitium or consuetudo recorded for the same lands in LDB. Measures of oats appear alongside pence per acre in the more extensive accounts for the first six vills, but then cease to be recorded.56 Why subsequent vills were treated more briefly, and why foddercorn should have been thought royal in the first place, is unclear.57 Estate centres provide a more plausible focus for customary dues than the king, for LDB reveals a pronounced tendency for servitium and consuetudo to be owed at particular manors. Nine Bury manors collected dues from at least one other vill.58 Of the 9,928 acres associated with servitium or consuetudo in the eight-and-a-half These figures are maximum values. Not all the acres would have been burdened but LDB does not break down its entries any further. 49 On the Bury writs, and the circumstances of their issue and survival, see R. Sharpe, ‘The Use of Writs in the Eleventh Century’, ASE 32, 2003, 247–91. 50 Ibid., 270–9. 51 Ibid., 262–70. 52 Ibid., 279–83. 53 S 1071. 54 S 1072. 55 Kalendar, xiv; The Customary of the Benedictine Abbey of Bury St. Edmunds in Suffolk, ed. A. Gransden, Henry Bradshaw Society 99, 1973, 1–2. 56 Feudal Documents, 25–33, cf. 33–44. 57 Foddercorn usually appears as an ordinary manorial food-rent: N. Neilson, Customary Rents, Oxford Studies in Social and Legal History 2, Oxford 1910, 23–7. 58 Risby, Lackford, Hengrave, Culford, Ingham, Rougham, Stanton, Coney Weston, and Rickinghall. 48

166

Richard Purkiss

hundreds, 2,820 (over 28%) owed their dues at a manor in a different vill.59 Such lands were mostly held by freemen owing servitium, who served at the manor but stood outside it, free to grant or sell their land. Further services were due from lands which were valued within a manor. These accounted for 2,618 acres (over 26%), and their holders were usually described as sokemen owing consuetudo.60 More than half the acreage recorded in connection with customary dues thus stood in a formal relationship to a Bury manor. Another 1,770 acres (almost 18%) were located in vills where Bury had a manor. Those who held them were usually freemen, valued separately.61 Such lands were recorded in the same entry as the manor, and their holders, although able to sell, might still be bound to the abbot’s fold. Only 2,720 acres (just over 27%) owed services without being adjacent to a Bury manor or associated with one by LDB. It is likely, although impossible to prove, that people recorded within or near a manor by LDB owed their services there. Private deeds and indeed the Kalendar record examples of dues including warpeni being rendered ‘at the hall’.62 Arrangements of this kind have been interpreted as departing from hundredal organization for reasons of convenience, but their distribution followed no obvious pattern (see Map. 2).63 Nor did it bear any relation to the grouping of manors for the abbey’s food-farm.64 The available evidence therefore associates services with manors rather than with some larger framework. It may be significant that Risbridge, the only one of the Bury hundreds in which the abbey possessed no demesne, was also the only one to record no servitium or consuetudo at all.65 Bury probably gained such rights as a matter of course whenever, and from whomever, a large manor was acquired. These conclusions present a contrast to the avera and inward services of eleventh-century Cambridgeshire, which were unambiguously public although still distinct from soke.66 Sokemen owing them might occasionally be reckoned within manors, but the services themselves were associated with the king and sheriff. No instance of avera or inward is recorded as being owed at a specific royal manor. It has been suggested that services from the Bourn Valley were originally performed at the royal vill of Haslingfield, but this was not the case in 1066.67 Manors with appendant hundreds, which often reflected this kind of early territorial organization, were notably absent from Cambridgeshire.68 Another reason for dissociating the local avera and inward from royal manors is that holdings thus burdened were spread widely across the shire (Map. 3) outside the Isle of Ely, where GDB records none. In mainland Cambridgeshire, service-owing holdings are recorded in six of the ten vills containing royal manors, and sixty-four out of 122 other vills. The figures are, again, maximum values: see n. 48. E.g. ‘Bradfield IIb’ (this example refers to freemen rather than sokemen). 61 E.g. ‘Bradfield III’. 62 Feudal Documents, nos. 142, 144. Kalendar, 9, 22, 31–2, 47; and e.g. nos. 42–3, 98–9. Many further examples in Samson’s charters are indexed under ‘Rents, payable to the abbey’s manors’. 63 Cf. Kalendar, xxxiv–v. 64 Robertson, AS Charters, no. 104. 65 The Feudal Book lists several holdings in Risbridge as ‘manors which St Edmund has in his demesne’, but these were groups of freemen over whom the abbot had soke: Feudal Documents, 8. 66 In Cambridgeshire too, not all free holdings owed services, whereas all naturally owed soke to someone. The services were still claimed for the king even when soke had been alienated to a third party: GDB 189d (Cambs. 1/12). 67 Oosthuizen, ‘Sokemen and Freemen’, 204–7. 68 H. M. Cam, ‘Manerium cum Hundredo: The Hundred and the Hundredal Manor’, EHR 47, 1932, 353–76 at 369–70. 59 60

Map 3: The distribution of avera and inward obligations in Cambridgeshire according to GDB

168

Richard Purkiss

Taken in isolation, the Cambridgeshire evidence might support the idea that cartage and guard-service were common warland obligations to the king. Such a view assumes that they were widely levied on independent freemen but become historically visible only occasionally, as in the GDB accounts of Cambridgeshire and Hertfordshire. But it is not clear that any general scheme existed. Had services levied directly upon unattached freemen been commonplace, one might expect GDB to record them consistently at least within a circuit. Cambridgeshire and Hertfordshire lay in Circuit III, along with Bedfordshire, Buckinghamshire and Middlesex. None of the other three shires records avera and inward, and in Hertfordshire these services were confined to the six northeastern hundreds nearest the Cambridgeshire border.69 Their monetary value was considerable. The standard rate of commutation for avera was 8d. in Cambridgeshire and 4d. in Hertfordshire, while Cambridgeshire inward was worth 4d.70 The 300 or more instances of avera or inward owed in Cambridgeshire alone therefore indicate an annual revenue of more than £9, equivalent to a substantial manor.71 It seems unlikely that GDB should have omitted royal revenues of this order in all shires but two. Mutually inconsistent rates of commutation, and the attachment of services to the shrieval office, suggest arrangements between kings and two particular shire communities. This mode of organization appears to have been exceptional. As a rule, customary services levied in the king’s name were attached to specific royal manors. In Norfolk, the sokemen of Little Snarehill owed summagium (akin to avera) in Kenninghall.72 In Suffolk, sokemen from Finborough owed summagium in Thorney (Stowmarket).73 Avera and inward obligations recorded in Kent were always associated with lathe capitals or other large royal manors, while customary payments from the king’s Bedfordshire estates included sums for packhorses.74 Disputes arose where unaccustomed services were exacted. Thus two Hertfordshire sokemen were made to provide avera and inward ‘by force and wrongfully’ in Hitchin when Harold held it.75 A similar imposition provoked skirmishes between the abbot of Abingdon and a royal reeve based at Sutton Courtenay.76 While the king’s men were no doubt often tempted to commandeer transport, this seems to have been considered an abuse where not enshrined in local custom. The last word on this point belongs to the barons of King John, for article 30 of Magna Carta lays down that, ‘No sheriff or bailiff of ours or anyone else is to take horses or carts of any free man for carting without the agreement of that free man.’77 We should look for the origins of practical obligations in the lordly economy rather than in any general duty imposed on freemen. Such a derivation finds support in the formulaic scheme of the Rectitudines singularum personarum, where it is the services performed by the geneat at the lordly estate that most closely resemble those in the Kalendar: carrying-service, reaping and mowing, guarding the lord and

69

Braughing, Broadwater, Edwinstree, Hertford, Hitchin, and Odsey. J. H. Round, ‘Introduction to the Hertfordshire Domesday’, VCH Hertfordshire, I, 263–99 at 270–1. See e.g. GDB 196c (Cambs. 18/8); GDB 134c (Herts. 5/13); GDB 195c (Cambs. 14/69). 71 An exact count is impossible because it is not always clear how many services were levied. 72 LDB 178a (Norfolk 9/75); Dictionary of Medieval Latin from British Sources, ed. R. E. Latham, D. R. Howlett, and R. K. Ashdowne, Oxford 1975–2013, s.v. ‘avera’, ‘summagium’. 73 LDB 404a (Suffolk 29/1). 74 GDB 1b (Kent D/24), 2d (Kent 1/3), 9d (Kent 5/138), 14c (Kent 13/1); GDB 209b–c (Beds. 1/1–3). 75 GDB 132d (Herts. 1/9). 76 Historia Ecclesie Abbendonensis, ed. and trans. J. Hudson, 2 vols, OMT, Oxford 2002–7, II, 14–16. 77 J. C. Holt, Magna Carta, 3rd edn, Cambridge 2015, 387. 70



Early Royal Rights in the Liberty of St Edmund

169

guarding horses.78 Services to the king consisted chiefly of the ‘common burdens’ – fortress-work, bridge-work and military service – but these were the responsibility of the geneat’s lord, the thegn. Manorial associations should come as no surprise once cartage, guard-service and the like are considered on their own terms. It was in the nature of such tasks to be performed at estate centres, and any lord of substance needed carts and watchmen. Similar customs were undoubtedly levied on peasants whose subjection to the hall was more complete, for a later villein’s week-work could include watching the lord’s granary, while carting was a particularly common manorial obligation.79 If cartage and guard-service were lordly rather than royal privileges, implications follow for our understanding of warland. LDB’s account of the eight-and-a-half hundreds usefully illustrates these. Many entries describe manors held by Bury St Edmunds, beginning with the number of carucates and enumerating their villani, bordarii, servi and demesne livestock. Within this section is concealed the inland, presumably cultivated for the abbey’s direct sustenance by servile labour and weekwork. In the summary representation of Table 2, this core unit is ‘Bradfield IIa’. Formulaically, everything from here on outward represents warland. Still valued within the manor, requiring permission to grant or sell their lands, and usually bound to the fold, were a class of sokemen or freemen burdened with soke and consuetudines. In their somewhat restricted mobility and customary obligations (probably rendered at the hall), these free peasants shared some characteristics with the more heavily burdened individuals clustered around the inland. ‘Bradfield IIb’ represents constrained freemen of this kind. Beyond them were freemen valued separately from the manor, free to sell, yet still under soke and owing servitium. These were free landholders whose only link to the manor was cartage or guard-service. They are found in ‘Bradfield III’. But, as we have seen, many freemen were wholly unattached and owed the abbot nothing except soke and sometimes commendation. Their obligations were apparently limited to the hidagium that made up their value, and the ‘common burdens’. In other words, those who were warland peasants in the fullest sense were largely free of ‘warland services’. This outermost stratum is represented by ‘Bradfield I’ and ‘Bradfield IV’. We can infer that cartage and guard-services were not characteristic of warland in general, but of warland which was reckoned within or connected to a manor. When they were recorded against certain freemen, this was less a badge of status than a way of marking out those who owed more than others. Differentiation between shades of warland is still evident in the Kalendar. Large numbers of free tenants owed averpeni or warpeni, other services like ploughing, reaping and mowing, and were further subject to reliefs and merchet. Their obligations were often explicitly due ‘at the hall’ of a manor. Some, however, were recognizably distinct. Variously described as holding libere or de alto socagio, they characteristically owed no merchet, paid reliefs directly to the abbot, and stood guard at Bury St Edmunds. Freedom from merchet was a clear sign of detachment from manorial lordship, and the payment of reliefs to the abbot probably derived from heriot payments to the king.80 Most significantly, these tenants’ guard-duty at Bury was not the guard-duty commuted for warpeni: the fact that some performed it while still paying warpeni indicates that it was not merely the same obligation 78

Gesetze, ed. Liebermann, I, 444–5. See also the Tidenham survey: S 1555 (Robertson, AS Charters, no. 109). 79 Neilson, Customary Rents, 50–1, 60–7. E. Day, ‘Sokemen and Freemen in Late Anglo-Saxon East Anglia in Comparative Context’, PhD thesis, University of Cambridge 2011, 291–4 reaches similar conclusions. 80 In Kent, the king collected relevatio terrae when allodiarii died: GDB 1b (Kent D/17).

170

Richard Purkiss

uncommuted.81 These same individuals were often explicitly exempt from averpeni.82 If one traces the relevant holdings in LDB, some can duly be identified with entries that record no servitium or consuetudo, but soke only.83 The obligation to guard the town evidently rested on something that LDB does not record, and a charter of 1121 × 1138 hints at what this was.84 Its statement that ‘free sokemen’ worked to repair the ditch around the town unmistakably evokes the fortress-work of the ‘common burdens’, which kings reserved from all land. Such work was perhaps exactable only at a centre like Bury St Edmunds and may have remained uncommuted for a long time.85 The libere tenentes of the Kalendar suggest that the only practical services levied on the freest warland were the king’s ‘common burdens’. Practical Services as Archaic Survivals Practical services are often treated as stable features, and their supposedly royal nature has only added to the impression, particularly where they appear in late, non-royal contexts. Much work on early territorial organization proceeds backwards from twelfthor thirteenth-century sources, using apparently archaic dues as a guide.86 There were certainly forces discouraging change: fixed and honourable services could underscore the status of those who owed them and adjustments might be resisted. But to assign most warland services a manorial focus is to place them in a setting where they may have been viewed as active resources rather than residual customs. There are signs that dues in the Bury hundreds were more susceptible to modification than once thought. Abbot Samson’s Kalendar has been taken as a window on East Anglian society before the Scandinavian invasions.87 While there can be no doubt that similar services existed in the ninth-century kingdom, stability over the intervening three centuries is doubtful. With the Kalendar dues partly located in LDB, it is possible to gain some idea of their development across one century. Comparing LDB records of servitium or consuetudo with the Kalendar’s warpeni and averpeni suggests a net increase in the number of vills where such dues were collected. In a number of places, the Kalendar shows obligations where LDB has none.88 On the other hand, LDB records the abbot claiming servitium or consuetudo in several places where no one owed warpeni or averpeni a century later.89 Some dues in the Kalendar appear to have been confined 81

Davis thought that warpeni commuted guard-obligations like those at Bury (Kalendar, xxxv). But warpeni and guard-duty were owed from the same holdings at Thurston (Kalendar, 10–12), Hessett and Beyton (Kalendar, 13–14), and Drinkstone (Kalendar, 14–15). Only people who held libere or de alto socagio performed actual guard-duty. 82 At Thurston (Kalendar, 10–12), Stanton (Kalendar, 37–9), and Hepworth (Kalendar, 45). 83 At Pakenham, see Kalendar, 8–10 and LDB 361b (Suffolk 14/49); at Drinkstone, Kalendar, 14–15 and LDB 362b (Suffolk 14/56); at Sapiston, Kalendar, 43 and LDB 366a (Suffolk 14/83). 84 Kalendar, xxxv–xxxvi; Feudal Documents, no. 113. 85 Similarly pontage, originally bridge-work, could still be exacted in works in the thirteenth century. It sometimes related specifically to the bridge in a county town and might be collected by an exempt religious house. Neilson, Customary Rents, 137–41. 86 E.g. Cam, ‘Manerium cum Hundredo’; W. Rees, ‘Survivals of Ancient Celtic Custom in Medieval England’, in Angles and Britons: O’Donnell Lectures, ed. H. Lewis, Cardiff 1963, 148–68; Jones, ‘Multiple Estates and Early Settlement’; Barrow, ‘Pre-Feudal Scotland’. 87 Kalendar, xlvi–xlvii. 88 Lackford (Thingoe hundred); Drinkstone, Great Livermere, and Great and Little Whelnetham (Thedwestry); Great Ashfield, Barnham, Coney Weston, Honington, Hunston, Langham, Little Livermere, Troston, and Walsham le Willows (Blackbourn/Bradmere). 89 Chevington and Manston (Thingoe hundred); Barningham, Culford, Stowlangtoft, and Market Weston (Blackbourn/Bradmere).



Early Royal Rights in the Liberty of St Edmund

171

to particular hundreds, suggesting that the recorded arrangements cannot pre-date the introduction of hundreds in the mid-tenth century.90 Thus mowing-service was only demanded in Thedwestry hundred, where it is mentioned in connection with twelve vills, for eight of which allocations were based on a 4½-rod unit.91 Averpeni was collected in Thedwestry and Blackbourn, but not in Thingoe. It seems to have been possible to reorganize ostensibly stable obligations or impose new ones. This impression is only strengthened when one looks beyond the Bury hundreds. The avera and inward obligations of Cambridgeshire and Hertfordshire probably arose in the tenth century. Their association with the shire and sheriff, rather than estate centres, suggests that they can hardly pre-date these institutions in their recorded form. Tenth-century bishops of Worcester, when issuing new leases from their own bookland holdings, were accustomed to stipulate riding and hunting work as well as the king’s military service.92 Some leases added carrying-service for the church to the standard reservation of the ‘common burdens’.93 It was plainly worth establishing works of this kind, even for the mere three-life term of a standard lease. Another revealing case has been described, with a different emphasis, by Robert Liddiard.94 It concerns the Suffolk manor of Eye, where provisions for maintaining the park pale were made no later than 1101. Maintenance of seigneurial enclosures is often considered a typically archaic due.95 This particular obligation was incumbent upon Eye, ‘the men of the soke’, and six other vills. All the lands concerned were in the hands of Robert Malet in 1086, whose father had taken them over wholesale from their pre-Conquest holder, Eadric of Laxfield. None of the outlying vills had any further connection with Eye and some in fact lay across a watershed from it, indicating that obligations at the park reflected an arbitrary eleventh-century tenurial pattern.96 The Malets, Eadric, or whoever first held the same group of estates and needed a deer-fence, established new services of an ancient type. It seems clear that the nature of a service is no guide to its date. Nor do the Old English names of certain obligations constitute sure evidence of antiquity: a point notably illustrated by warpeni. In the Isle of Ely, thirteenth-century surveys record comparatively free tenants called ‘hundredors’ owing payments of this name. These surveys have been interpreted, similarly to the Kalendar, as showing freemen owing an archaic public guard-duty.97 Earlier evidence raises difficulties. In 1123 × 1129, Henry I had declared Bishop Hervey and his successors quit of the 40s. of warpeni which had previously been collected from Ely’s lands and men.98 Two significant points arise from this: that the bishop’s lordship over hundreds had not entitled him to collect warpeni, and that the due had been laid upon his fee in a round

90

On the chronology see G. Molyneaux, The Formation of the English Kingdom in the Tenth Century, Oxford 2015, 141–6. 91 The vills mowing multiples of 4.5 rods were Beyton, Bradfield St George, Hessett, Rushbrooke, Stanningfield, Thurston, Tostock, and Woolpit. Less regular, but possibly assessed for units of 7 rods, were Ampton, Fornham St Genevieve, Great Livermere, and Pakenham. These groups were geographically distinct. 92 S 1368, the ‘Lex equitandi’. 93 S 1309, 1332. Robertson, AS Charters, 334–5 suggests this as the meaning of the term circanlade. 94 R. Liddiard, ‘The Deer Parks of Domesday Book’, Landscapes 4, 2003, 4–23 at 16–18. 95 Faith, English Peasantry, 107. 96 On the importance of watershed boundaries in East Anglia, see Williamson, Origins of Norfolk, 14–19 and Environment, Society and Landscape, 82–94. 97 Kalendar, xli–xlii; E. Miller, The Abbey and Bishopric of Ely, Cambridge 1951, 116–9; S. Oosthuizen, The Anglo-Saxon Fenland, Bollington 2017, 19–23. For the relationship of the hundredors to the hundred: P. Vinogradoff, Villainage in England, Oxford 1892, 188–96, 441–52. 98 Liber Eliensis, ed. E. O. Blake, Camden Society 3rd series 92, 1962, 252, 256–7.

172

Richard Purkiss

sum. Corroboration comes from the 1130 pipe roll, which records a total of 41s. 8d. in warpeni as pardoned from Ely lands.99 The use of the fee as an organizing frame suggests that the imposition was a recent one. Nor does the thirteenth-century Ely Coucher Book show the due in an archaic light. Although the hundredors paid warpeni, it was not confined to their holdings. Unfree censuarii and cotarii, who owed week-work, gave it as well. The payment was assessed rigidly at 1d. per plot, doubled where two held jointly, and confined to Stretham, Wilburton and Linden [End].100 Perhaps these vills had been singled out to contribute to the 40s. originally demanded from the bishop, pennies being laid indiscriminately upon local tenants. Elsewhere, post-Conquest kings began to treat warpeni as a hundredal franchise, notably Henry I in a writ in favour of Bury St Edmunds.101 We have therefore encountered warpeni as a manorial due, a payment for shire-based service, a payment levied on a fee, and a franchise. Further references connect it with prerogatives which arose only after 1066, such as castle-guard and franchisal markets.102 Archaic terminology can deceive, and some instances of this particular word mark guard-obligations introduced only after the Conquest. Conclusions It is clear that the rights of Bury St Edmunds in the twelfth century are no reliable guide to those of Anglo-Saxon kings in the tenth or eleventh. Our three opening questions have all returned negative answers. Dues described in the Kalendar of Abbot Samson did not fall within the franchise of soke, as they appear separately in LDB. There is therefore no need to complicate this privilege by making it encompass customary renders. The precise content of soke grants must for the present remain an open question, although it may not be helpful to approach soke as a single set of rights.103 As for the Kalendar dues, there is little evidence that they were considered intrinsically royal or organized through public institutions before the Conquest. Later, it may be that the unique depth of the abbey’s control as hundredal franchisee, tenant-in-chief, and often demesne lord made it convenient to collect payments through the hundred-reeves.104 But in keeping with the nature of the obligations involved, their usual association, both within the Bury hundreds and elsewhere, was with estate centres. Freemen’s services were of value to the lords of such places and consequently endured in the living economy: not all were fossilized relics of earlier structures. The implications of these findings for the inland–warland model have already been touched upon. They do not so much dilute the model as adjust the place of ‘warland services’ within it. On the evidence of the Bury hundreds, freemen’s services were owed not to the king but to particular manors; in other words, they were essentially private rather than public. Nor were all freemen thus burdened, but only some, and to varying degrees. The general category of warland can be envis-

99

PR 31 Henry I, 34, 41, 72. Miller, F. Willmoth, and S. Oosthuizen, The Ely Coucher Book, 1249–50: The Bishop of Ely’s Manors in the Cambridgeshire Fenland, Cambridgeshire Records Society 22, Cambridge 2015, 67–76 (Stretham), 79–85 (Wilburton), 87–100 (Linden). 101 Feudal Documents, no. 46. 102 Neilson, Customary Rents, 131–7; R. H. Britnell, ‘English Markets and Royal Administration before 1200’, Economic History Review 31:2, 1978, 183–96. 103 The Leges Henrici Primi suggest that there were different qualities of ‘sake and soke’: Hn 61:9a (Gesetze, ed. Liebermann, I, 581). 104 See also Day, ‘Sokemen and Freemen’, 225–6. 100 E.



Early Royal Rights in the Liberty of St Edmund

173

aged as a series of concentric bands, whose ties to the manorial centre diminished outwards until services formed the only bond, before giving out entirely. All bands were, of course, liable for truly public exactions, and the outermost for nothing else. The independent freemen who occupied it were expected to find the ‘common burdens’, geld, an acreage-based tribute and heriots, but not the services hitherto considered diagnostic of warland. Although their holdings might be small, their obligations were thegnly and to the king alone. Where works and payments are recorded they may, but need not, preserve very old relationships. Freemen’s services reflected a mode of social and economic organization rather than a particular historical period, and continued to be negotiated for as long as the appropriate conditions prevailed (which in remote areas, such as northwest Scotland, could be as late as the eighteenth century).105 The type of due alone cannot be taken as a sure sign of antiquity. In contexts where obligations were attached to offices or territories and lacked a physical focus, they are likely to have resulted from deliberate acts of government, and would consequently have remained exposed to further changes. Customs rendered at estate centres may have been more inclined to stability, given that such centres tended to survive political upheaval.106 The royal status, or otherwise, of practical services was not necessarily immutable over time. Late Anglo-Saxon kings plainly received them, but as holders of great estates rather than as kings. It is conceivable that certain works once had more distinctively royal connotations but if so, only in an earlier period, when the lords of individual territorial centres could aspire to kingly status.107 Kingship steadily increased in scale, and a parallel shift can be discerned in the rights attached to it. The claims of early kings, from practical works to feorm, were bound up with estate centres. But as kingship became rarer, more exalted, and geographically more extensive, it began to make blanket claims by right. The ‘common burdens’, land taxation, and eventually soke, were privileges of this kind. Charters from the early eighth century onwards may bear witness to this shift in their reservation clauses, claiming one or more ‘common burdens’ while freeing the land in question from other such exactions as the provisioning of royal officials or work at the royal vill.108 We seem to see kings giving up miscellaneous rights collected through estate centres in return for universal and institutional ones – in other words, for royal prerogatives. This last term is unaccustomed in pre-Conquest settings but accurately describes the privileges of Anglo-Saxon kingship. Thinking of prerogative rights rather than public obligations may prove salutary in limiting their extent.

105 O.

Mackenzie, A Hundred Years in the Highlands, London 1949, 22, describes how a carrying boon supported the seasonal movements of a gentry family into the 1750s. 106 A striking example is Bensington (Benson) in Oxfordshire, possibly held by British, West Saxon, Mercian, English, and Norman kings in turn: ASC, trans. Swanton, s.a. 571, 777 [779]; S 217, 887; GDB 154c (Oxon. 1/1). 107 The scale of early kingship is discussed by S. Bassett, ‘In Search of the Origins of Anglo-Saxon Kingdoms’, in The Origins of Anglo-Saxon Kingdoms, ed. S. Bassett, London 1989, 3–27, and T. M. Charles-Edwards, ‘Early Medieval Kingships in the British Isles’, 28–39 in the same volume. 108 F. M. Stenton, The Latin Charters of the Anglo-Saxon Period, Oxford 1955, 56–9; N. P. Brooks, ‘The Development of Military Obligations in Eighth- and Ninth-Century England’, in England Before the Conquest: Studies in Primary Sources Presented to Dorothy Whitelock, ed. P. Clemoes and K. Hughes, Cambridge 1971, 69–84. Charters issued by Mercian kings in the late eighth and ninth centuries often provide an explicit indication of the services being given up: see for example S 134, a privilege of Offa for the churches of Kent.

Figure 1: Christine Mahany (1939–2016)

The Christine Mahany Memorial Lecture

CASTLE CONSTRUCTION, CONQUEST AND COMPENSATION* David Roffe For older members of the Battle Conference the most abiding memory of Christine Mahany (15 January 1939–13 July 2016) (Fig. 1) will almost certainly be of her sitting at the bar in The Chequers, pint in hand, deep in conversation with Allen Brown, while George, her beloved labrador, patiently languished at her feet. Dogs were a frequent subject of passionate discussion. It was, after all, a time when canine delegates Matilda and Offa were cast in the crucial role of celebrity judges who voiced their opinion of tendentious arguments and windy papers with a yawn or a howl. As often as not, though, the chat was about archaeology or history, or, more precisely, archaeology and history. As odd as it may seem today, the relationship between the two subjects was something of an issue in the 1970s and 1980s. Archaeology as the handmaiden to history was still a cheap put-down that was bandied about in historical circles. Chris came to Battle as an advocate of equality and common interest. After something of a false start, her background was copper-bottomed archaeology. She studied zoology as a student but spent much of her time in Leicester’s New Walk Museum sorting pottery sherds and the like, and so after graduation she was drawn to archaeology as a career. She cut her teeth under the tutelage of Philip Rahtz, the great exponent of open-area excavation, digging on seminal early medieval sites such as Cheddar and Cannington. Thereafter she became an itinerant archaeologist for the Ministry of Public Buildings and Works in her own right, supervising a wide range of sites. Her excavation of the small Roman town of Alcester in Warwickshire was a particular achievement. Temperamentally, however, her interests had always gravitated towards medieval archaeology, and her appointment as director of the Stamford Archaeological Research Committee in 1966 gave her the opportunity to indulge them. She was to stay in Stamford for the rest of her career. Her appointment came at an opportune time. After a decade of inappropriate post-war development – vandalism is probably a more appropriate term – the historic core of Stamford was designated as the first conservation area in Britain.1 Chris’s brief was to recover its archaeology and produce a coherent account of its origins and growth. Twenty or so excavations and numerous watching briefs followed, covering the full range of medieval life. Her findings were arresting but *

I am indebted to Steven Brindle for bringing to my attention the reference to the payment from Windsor castle to the manor of Clewer; the fact was instrumental in crystallizing a number of thoughts and speculations that had been going round in my head for a number of years. Katharine Keats-Rohan, Rob Liddiard, Rachel Swallow and Ann Williams kindly commented on the argument that ensued at various stages in its development. Nevertheless, all errors of fact and interpretation remain my own responsibility. 1 For the background, see Institute of Historic Building Conservation, ‘The History of Conservation Areas’ (https://www.designingbuildings.co.uk/wiki/The_history_of_conservation_areas [assessed 31 August 2018]). Stamford was designated just two months after the act was passed in 1967.

176

David Roffe

did not always sit easily with the received historical accounts of Stamford, and Chris quickly became aware that a more syncretic approach to the history and archaeology of the town was required. As a result, she fought for the appointment of one of the first historians in an archaeological unit to work in partnership with her. That historian was myself: Chris and I worked on a new synthesis of the history of the town in which the archaeological and historical data were allowed to play off against each other on equal terms. The results were presented at the Battle Conference in 1982.2 At the core of the analysis was Chris’s most famous site, that of Stamford castle. Her discoveries there have had a profound impact not only on our understanding of the history of Stamford, but also more widely on the nature of the castle as a form. The earliest reference to Stamford castle (NGR: TF028070) is found in Domesday Book, where it is recorded that five mansiones were destroyed in its construction.3 The excavation was unable to finesse the terminus ante quem of 1086, but it seems very likely that the castle was built in 1068 at the same time as those at Cambridge, Huntingdon and Lincoln: certainly by 1070 Stamford seems to have been considered a safe haven for Thorold, the incoming Norman abbot of Peterborough, in the face of a hostile English monastic community.4 Thereafter the castle developed along familiar lines. There was what we have conventionally called a shell keep and a stone curtain by the mid-twelfth century, and the excavation revealed extensive timber buildings in the bailey.5 A fine masonry hall and solar were added in the mid- to late twelfth century and remodelled in the thirteenth. By 1327, however, the castle was said to be in ruins and increasingly it was used only as a manorial centre and leet court. We owe the survival of a fragment of the great hall complex into the present day to its continuing local utility in these terms.6 The development of Stamford castle is in broad outline shared by many another site. What was of particular interest to historians and archaeologists alike was its antecedents. Vestiges of two concentric defensive ditches and an internal palisade of the late ninth century were discovered beneath the conquest layers, which, if extended, would seem to have originally surrounded the site of the castle. Whether the complex was of Mercian or Danish origin was, and remains, undetermined. Nor was there any evidence that it survived into the eleventh century. By 1066, though, the site was adjacent to the nucleus of a royal manor in the hands of Queen Edith. It would seem that the construction of the castle perpetuated a centre of authority of some 200 years standing.7

2 Christine Mahany and David Roffe, ‘Stamford: The Development of an Anglo-Scandinavian Borough’, ANS 5, 1983, 199–219. Reprinted with updated references in Anglo-Saxon History: Basic Readings, ed. D. A. E. Pelteret, Garland Reference Library of the Humanities 2108, New York and London 2000, 387–417. 3 NGR = National Grid Reference; GDB 336v: LIN S2. 4 Orderic, II, 218; ASC, 1067 [1068] D, 1070E. 5 The term ‘shell keep’ has recently been subject to critical review by Robert Higham (Shell-Keeps Revisited: The Bailey on the Motte? (http://www.castlestudiesgroup.org.uk/page131.html [accessed 21 February 2017]). The structure in Stamford was demolished without record in the 1930s to build the present bus station, and so its form is unknown. 6 C. M. Mahany, Stamford Castle and Town, South Lincolnshire Archaeology 2, Stamford 1978; eadem, ‘Excavations at Stamford Castle, 1971–1976’, Chateau Gaillard 8, 1976, 223–45. 7 Mahany and Roffe, ‘Stamford: The Development of an Anglo-Scandinavian Borough’, 199–219; D. R. Roffe and C. M. Mahany, ‘Stamford and the Norman Conquest’, Lincolnshire History and Archaeology 21, 1986, 5–9.



Castle Construction, Conquest and Compensation

177

This was a conclusion that fed into what was already a burgeoning revisionary assessment of the nature of the castle. Pre-Conquest antecedents had been found at such rural sites as Sulgrave and Goltho/Bullington,8 and the higher-status site in Stamford added grist to the mill. More sites came to light in the 1980s and 1990s in the wake of renewed archaeological investigation, and many more were suggested. An understanding of the castle as a symbol of authority, as well as a military resource, began to emerge, and its relationship to what went before was reconsidered. It became clear that what was new about Norman castles was not the fact of private fortification in itself – as much is witnessed in Anglo-Saxon England in the early eleventh-century ‘promotion law’ – but the marking of it with a motte and then a keep.9 How common was actual continuity of site remains unclear. In a rural context it may have been considerable. In addition to Sulgrave and Goltho, there is sound archaeological evidence for Castle Acre (Norfolk), Castle Bromwich (Warwickshire), Castle Neroche (Somerset), Eynesford (Kent), Laughton-en-le-Morthen (Yorkshire), Middleton Stoney (Oxfordshire), Portchester (Hampshire) and Trowbridge (Wiltshire),10 and suppositions of varying degrees of confidence, based upon topographical analysis and the like, for numerous other sites. Documentary evidence, implicit as it is, points in the same direction. In my own study of the early Lincolnshire castles, for example, I found that at least eleven out of fifteen baronial castles built before 1150 were sited in manors which were held with sake and soke in 1066 (Table 1).11 Sake and soke was the principal marker of superiority before the Conquest, and the cornerstone of title in the Domesday inquest, and so the fact suggests a high degree of continuity of elite authority and the sites through which it was exercised.12 By way of contrast, there was no such correlation for the often somewhat later castles of the honorial baronage. In towns continuity was probably more limited. At Southampton the castle bailey was shown in excavation to overlie a pre-Conquest timber hall and associated ditches.13 Likewise at Stafford a similar intramural burh was perpetuated by the castle.14 St George’s Tower in Oxford castle may be a surviving burhgeat of the 8

B. K. Davison, ‘Excavations at Sulgrave, Northamptonshire, 1968’, The Archaeological Journal 125, 1968, 305–7; idem, ‘Five Castle Excavations: Reports on the Institute’s Research Project into the Origins of the Castle in England. Excavations at Sulgrave, Northamptonshire, 1960–76: An Interim Report’, The Archaeological Journal 134, 1977, 105–14; Guy Beresford et al., Goltho: The Development of an Early Medieval Manor c. 850–1150, London 1987; Paul Everson, ‘What’s in a Name? ‘Goltho’, Goltho, and Bullington’, Lincolnshire History and Archaeology 23, 1988, 93–9. 9 For a discussion of the emergence of the new paradigm, see Robert Liddiard, Castles in Context: Power, Symbolism and Landscape, 1066–1500, Oxford 2005. Michael Shapland, ‘Anglo-Saxon Towers of Lordship and the Origins of the Castle in England’, in The Archaeology of the 11th Century: Continuities and Transformations, ed. Dawn M. Hadley and Christopher Dyer, Society of Medieval Archaeology Monographs 38, London 2017, 104–19, suggests that keeps owe at least something to pre-Conquest ‘tower-nave’ churches. 10 Liddiard, Castles in Context, 28–30. 11 David Roffe, ‘An Englishman’s Home: The Early Castles of Lincolnshire’ (http://www.domesdaynow.co.uk/horncastle.htm [accessed 2 February 2017]). 12 Cf. Andrew Lowerre, ‘Why Here and not There? The Location of Early Norman Castles in the SouthEastern Midlands’, ANS 29, 2008, 121–44 at 141, who found less than a third of castles in his study area of Bedfordshire, Cambridgeshire, Huntingdonshire, and Northamptonshire exhibited any degree of continuity of site from the pre-Conquest period. However, he did not distinguish between baronial and honorial castles. 13 J. Oxley and J. Boudillon, Excavations at Southampton Castle, Southampton Archaeology Monograph 3, Southampton 1986, 47–8. 14 R. Cuttler, J. Hunt, and S. Rátkai, ‘Saxon Burh and Royal Castle: Re-thinking Early Urban Space in Stafford’, Transactions of the Staffordshire Archaeological and Historical Society 43, 2008, 39–85.

178

David Roffe

Table 1: Baronial castles and sake and soke in Lincolnshire Castle Barrow Bolingbroke Bourne Bytham, Castle Carlton, Castle Folkingham Frampton Gainsborough Hough on the Hill Owston Redbourne Sleaford Stainby Tattershall Welbourn

Type* mb rw mb mb mb mb rw rw mb mb rw rw mb rw rw

Lord 1066 Earl Morcar Stori Earl Morcar Earl Morcar Godric Ulf Fenisc Adestan ? Ralf the staller Guerde Agemund Bardi Siward Godwin Godwin

Lord 1086 Drew de Beuvriere Ivo Taillebois Oger the Breton Drew de Beuvriere Ansgot of Burwell Gilbert de Ghent Walter de Aincurt ? Count Alan Geoffrey de la Guerche Jocelin son of Lambert Bishop of Lincoln Alfred of Lincoln Eudo son of Spirewic Robert Malet

Sake and soke† yes yes yes yes no yes yes no yes yes yes yes ? yes no

* mb = motte and bailey, rw = ringwork † The holders of sake and soke in Lincolnshire are listed at GDB 337: LIN T5.

type that characterized such sites before the Conquest, or a comparable seigneurial tower-nave church.15 Again, if the small number of properties destroyed at Stamford reflects a site under the direct control of the crown, then it may do so at Warwick and Wallingford too, where 4 masurae and 8 hagae respectively were waste.16 The first two were extramural and all three were adjacent to high-status churches.17 But these are probably the exception. Urban castles typically incorporated an angle within the existing defences of the town and so it must often have been utility rather than status that dictated the choice of site.18 In consequence the destruction of the properties was the more considerable. Domesday records that castle construction accounted for the wasting of a whole shire, that is a quarter, at York, 166 properties at Lincoln, 98 at Norwich, 51 at Shrewsbury, 27 at Cambridge and 16 at Gloucester.19 With the different terms employed – domus, mansio, ma(n) sura, haga – these figures may not be directly comparable, but archaeology has revealed that the destruction could be extensive. Excavation at Lincoln, Norwich, 15

Daniel Poore, Andy Norton, and Anne Dodd, ‘Excavations at Oxford Castle: Oxford’s Western Quarter from the Middle Saxon Period to the Eighteenth Century’, Oxoniensia 74, 2009, 1–18; John Blair, Building Anglo-Saxon England, Princeton 2018, 399. 16 GDB 238: WAR B2; GDB 56: BRK B1. The tenements in Warwick appear to have belonged to the church of Coventry. 17 Mahany and Roffe, ‘Stamford: The Development of an Anglo-Scandinavian Borough’, 203–4; Steven Bassett, ‘Anglo-Saxon Warwick’, Midland History 34, 2009, 123–55 at 140–2; David Roffe, ‘Wallingford in Domesday Book and Beyond’, in The Origins of the Borough of Wallingford. Archaeological and Historical Perspectives, ed. K. S. B. Keats-Rohan and D. R. Roffe, British Archaeological Reports British Series 494, 2009, 27–51 at 41. 18 Christopher Drage, ‘Urban Castles’, in Urban Archaeology in Britain, ed. J. Schofield and R. Leech, Council for British Archaeology Research Report 61, London 1987, 117–32. 19 GDB 298: YKS C1; GDB 336v: LIN C26; LDB 116v: NFK 1, 61; GDB 252: SHR C14; GDB 189: CAM B1; GDB 162: GLS G4. The castle of Gloucester was on the first, intramural, site.



Castle Construction, Conquest and Compensation

179

Pontefract and Winchester has shown that whole quarters, including churches and graveyards, might be swept away.20 If the castle qua institution represents continuity of sorts with the Anglo-Saxon past, it undoubtedly often saw a traumatic disruption of settlement after 1066. It has been estimated that there were as many as 600 castles in England and Wales by 1100.21 About a quarter of those surviving were ringworks and so could, conceivably, have had preConquest antecedents.22 The rest were either completely new or major modifications of exiting sites. The fact appears to fit into a narrative of the castle as an instrument of conquest. Orderic Vitalis, writing within fifty years of the Conquest, asserts that the English were not able to resist the Normans because they did not have castles. If he is discounted as a late and biased source, then the Anglo-Saxon Chronicle’s account of their construction in the late 1060s demonstrates their importance in securing the Midlands and the North.23 Their exact function is disputed. Revisionary analyses now see them as primarily centres of authority rather than fortifications.24 Either way, they were a major instrument of conquest. Narrative sources, few and far between as they are, seemingly attest the concomitant destruction they occasioned. William of Poitiers and Guy of Amiens report that Duke William unceremoniously ordered the inhabitants of Dover to evacuate their houses before the castle was built in their place.25 Even more graphic was the experience at Worcester. There the construction of the castle, entrusted to the notorious sheriff Urse d’Abitot, took in the cemetery of the church of Worcester. The protests of the monks against this encroachment were to no avail, and famously Archbishop Ealdred of York cursed Urse for his desecration.26 The possibility of compensation in a small number of instances has sometimes been recognized,27 but it is widely understood that the experience of Dover and Worcester was the norm.28 The conclusion is, indeed, apparently consonant with

20

‘Archaeology’ (https://www.lincolncastle.com/content/archaeology [accessed 1 September 2018]); Elizabeth Popescu, ‘Norwich Castle’, in Castles in the Anglo-Norman World, ed. John A. Davies, Angela Riley, Jean-Marie Levesque, and Charlotte Lapiche, Oxford 2016, 3–30 at 6–8; Ian Roberts, Pontefract Castle: Archaeological Excavations 1982–86, West Yorkshire Archaeology Service, Leeds 2002, 401–4; Martin Biddle, ‘Winchester 1961–68’, Château Gaillard 4, 1969, 19–30. 21 Richard Eales, ‘Royal Power and Castles in Norman England’, in Medieval Knighthood 3, ed. C. Harper-Bill and B. Harvey, Woodbridge 1990, 49–78 at 54–7. 22 B. K. Davison, ‘The Origins of the Castle in England: The Institute’s Research Project’, Archaeological Journal 124, 1967, 202–11; idem, ‘Early Earthwork Castles: A New Model’, Château Gaillard 3, 1969, 37–47; Barbara English, ‘Towns, Mottes and Ring-works of the Conquest’, in The Medieval Military Revolution, ed. A. Ayton and J. L. Price, London 1995, 45–61. 23 Orderic, II, 219; ASC, 1067 [1068] D. Orderic’s understanding has been questioned by proponents of the ‘peaceable power’ construct in castle studies. Although the remark may have been derived from the lost portion of William of Poitiers’ Gesta Guillelmi, it is argued that it still represents an essentially late and regional perspective (Liddiard, Castles in Context, 37–8). 24 Liddiard, Castles in Context, chapter 2. 25 Poitiers, 38; Carmen, 36. 26 Malmesbury, Gesta pontificum, 253. 27 Ella Armitage, The Early Norman Castles of the British Isles, London 1912, 116, 164, 169, 195; English, ‘Towns, Mottes and Ring-works’, 46; Robert Liddiard, Landscapes of Lordship: Norman Castles and the Countryside in Medieval Norfolk, 1066–1200, BAR British Series 309, Oxford 2000, 25; Michael Fradley, ‘The Old in the New: Urban Castle Imposition in Anglo-Norman England, AD 1050–1150’, unpublished PhD thesis, University of Exeter 2011, 346. N. J. C. Pounds, The Medieval Castle in England and Wales: A Social and Political History, Cambridge 1990, 209, recognises the possibility of compensation, but asserts that the evidence is wanting. 28 For recent statements of the apparent fact, see David Bates, William the Conqueror, New Haven and London 2016, 297–300, 325–5; Richard Eales, ‘Castles and Borders in England after 1066’, Château Gaillard 26, 2014, 149–57.

180

David Roffe

contemporary Norman propaganda. William claimed the throne of England by inheritance, but, by their resistance to his right, the English had forfeited their lands. In consequence, the only title they had after 1066 was by grant of the Conqueror. By definition the principle also applied to the Normans preferred by the new king. Domesday Book shows time after time that the king’s will superseded any other claim on land. William was apparently free to commandeer whatever lands he wished in order to construct castles and, it is assumed if only tacitly, he did so. Given carte blanche by their lord, his barons did likewise. Castle building, it would seem, was overwhelmingly an exercise in force majeure and was of a piece with a conquest and settlement in which the main motor was aggressive free enterprise.29 A re-examination of the limited evidence suggests a more nuanced picture. The primary analysis that follows is based on the documentary record and is largely confined to the sixty-nine castles that are known to have been built between 1066 and 1086.30 The archaeological evidence informs a secondary analysis of the wider context of castle construction and its impact on the settlements in which they were sited and the former inhabitants who were displaced. Castles were undoubtedly disruptive elements in late eleventh-century English society, but what emerges from this study is that their builders were ultimately far more constrained than has previously been appreciated. We can best begin with Windsor in Berkshire. Domesday Book records that the castle occupied half a hide of the manor of Clewer, which was held in chief by a certain Ralph fitzSeifrid.31 The site was up-river from the Anglo-Saxon royal palace at Old Windsor, above a narrow crossing of the Thames. Whether a bridge was built at the same time is unrecorded. Likewise, the exact date of construction is unknown – Domesday provides the earliest reference – but it is likely that it was early in the reign of the Conqueror. Recent research has shown that the castles of Oxford, Wallingford and Windsor, held respectively by the sheriff, the heir to the throne and the monarch, had complementary functions from the twelfth century onwards.32 It is therefore likely that, like the first two, Windsor was constructed in the late 1060s or early 1070s to control a strategic crossing of the river.33 Indeed, the Abingdon chronicler brackets all three together.34 The lord of Clewer, Ralph, is intriguingly mysterious. His only other known properties were the manors of Freefolk and Polhampton in Hampshire and two non-customary tenements in Wallingford, presumably an appurtenance of his manor of Clewer.35 His successors, the Sifrewast family, held Clewer in chief of the king for the service of a knight fee, subsequently reduced to a half.36 Ralph’s brother Roger adds something more to this

29

George Garnett, Conquered England: Kingship, Succession and Tenure, 1066–1166, Oxford 1997, 24–33. C. G. Harfield, ‘A Hand-list of Castles Recorded in the Domesday Book’, EHR 106, 1991, 371–92. Winchester, omitted from the list of castles not in Domesday but documented elsewhere, has been added. 31 GDB 62v: BRK 49, 1. 32 David Roffe, ‘Introduction’, in Wallingford: The Castle and the Town in Context, ed. K. S. B. KeatsRohan, Neil Christie, and David Roffe, BAR British Series 621, 2015, 1–8 at 4; K. S. B. Keats-Rohan, ‘“Most Securely Fortified”: Wallingford Castle 1071–1540’, in Wallingford: The Castle and the Town in Context, 54–115 at 36, 55–6, and passim. 33 Steven Brindle, ‘The Foundation of the Castle, c. 1066–1100’, in Windsor Castle: A Thousand Years of a Royal Palace, ed. Steven Brindle, Royal Collection Trust, London 2018, 16–23. 34 Historia Ecclesie Abbendonensis: The History of the Church of Abingdon, ed. John Hudson, OMT, Oxford 2002, II, 5–7. 35 GDB 41, 41v: HAM 3, 5; 10. GDB 56: BRK B3. 36 The Red Book of the Exchequer, ed. Hubert Hall, 3 vols, London 1896, I, 308; The Book of Fees, 2 vols in 3, Public Record Office, 1920–31, II 849, 851. 30



Castle Construction, Conquest and Compensation

181

mix. A minister of the crown, he held Fulscot and Purley in Berkshire, and Brightwell Baldwin in Oxfordshire, and these manors were subsequently held by a family with the surname of Huscarle, that is housecarl.37 It is possible, then, that Ralph and Roger had pre-Conquest antecedents. All of this would be of little moment for the present purposes were it not for the fact that the crown paid 12s. every year to the lord of Clewer for the site of the castle of Windsor until 1546.38 The sum represents just over the value of half a hide: the remaining 4½ hides were valued at £4 10s. in 1086.39 The castle did not become a palace until Old Windsor was abandoned in the early twelfth century.40 Nevertheless, it is striking that one of the main residences of the kings and queens of England was for the first 400 years of its history held in fee farm of a minor tenant-in-chief next door. The monarch of Sue Townsend’s novel The Queen and I, exiled on the Beaumont Leys council estate in Leicester in the early 1990s, was not the first to pay rent for her abode.41 The constraints on the king’s freedom of action at Windsor were clearly considerable. At the same time, or conceivably somewhat later, William also had to compensate Walter fitzOther or his son for land taken into the borough that grew up at the castle gates.42 The settlement was apparently no tabla rasa on which the king’s will had free rein. It did not stand alone in this respect. Domesday carefully records that the king’s castle of Stafford was built on the land of Henry de Ferrer’s manor of Chebsey.43 The burh excavated under the castle is most likely not a king’s hall, but the town residence of Henry’s predecessor (he is not named in Domesday).44 I have so far identified no reference to a rent in recognition of this relationship,45 but it might be supposed that the mere fact of the Domesday record supposes some expectation thereof. That was certainly the case of the royal castle of Nesse in Gloucestershire.46 The site is usually identified with Berkeley.47 However, the name, signifying a promontory, is more consistent with a location on the Severn, perhaps 37

GDB 62v, 63: BRK 49, 2–3; GDB 160v: OXF 58, 7; VCH Berks, III, 417–22; David Roffe, ‘Wallingford in Domesday Book and Beyond’, in The Origins of the Borough of Wallingford: Archaeological and Historical Perspectives, ed. K. S. B. Keats-Rohan and David Roffe, BAR British Series 494, 2009, 27–51 at 40–1. 38 The payment is not witnessed in the Pipe Roll of 1130–1, but appears thereafter from 1159–60 in the form ‘to Richard de Sifrewast, 12s in exchange for his lands (pro escambio terre sue)’. See, for example, Pipe Roll 5 Henry II, 37. The manor was bought by the crown in 1546 and in consequence the rent was discontinued (VCH Berks, I, 72–7). 39 GDB 62v: BRK 49, 1. 40 VCH Berkshire, III, 5–6. 41 Sue Townsend, The Queen and I, London 1992. 42 PR 31 Henry I, 100. 43 There has been some confusion as to the meaning of terra de Stadford (The Electronic Edition of Domesday Book: Translation, Databases and Scholarly Commentary, 1086, 2nd edn, at https://hydra. hull.ac.uk/resources/hull:domesdayTranslation or http://discover.ukdataservice.ac.uk/catalogue?sn=5694, STA 10, 9n [accessed 5 September 2018]. There is no indication that this was something like a berewick of Chebsey: the natural reading is that it was an urban appurtenance of the manor. For the location of the castle in Stafford rather than Castle Church, see Deborah Youngs and Philip Morgan, ‘The Documentary Evidence’, in Stafford Castle: Survey, Excavation, and Research 1978–1998, Volume I: The Surveys, ed. John Darlington, Stafford Borough Council, Stafford 2001, 32–82 at 32. 44 Fradley, ‘Old in the New’, 301–2. 45 Philip Morgan (personal communication) tells me he has found no evidence of a rent either but stresses that the extent of his researches were limited by the exigencies of time. 46 GDB 163: GLS 1, 19. 47 David J. Cathcart King, Castellarium Anglicanum: An Index and Bibliography of the Castles in England, Wales, and the Islands, 2 vols, Millwood, NY, 1983, I, 180.

182

David Roffe

the present Sharpness, which is close to Oldminster where the pre-Conquest church of Berkeley probably stood,48 and so it may represent an early incarnation of the later castle at the estate centre. If so, Nesse can be added to the short list of castles – Canterbury and Gloucester – which are known to have moved sites in the late eleventh or early twelfth century. Here Domesday records that Roger of Berkeley claimed the 5 hides on which it was built. Whether Roger made good his claim is unrecorded: his successors were to hold Sharpness by the late twelfth century, but in the meantime the whole of the manor had been granted to them.49 It is clear, however, that litigation was possible in such cases and could be successful. The abbey of Fécamp successfully pursued a claim against William de Braose in 1085. The judgement relates that William had built his castle of Bramber on 1 hide of his manor of Washington in Sussex.50 The watercourse that supplied his moat, a park and warren, by contrast, were on the abbot’s land in Steyning. The matter was brought before the king at Lacock in Wiltshire and it was decided that the watercourse should be filled in and the land restored to the abbot. The burial rights of the abbey’s church of St Cuthman were also confirmed, the burials at St Nicholas’s church in Bramber being disinterred and reburied at Steyning, and the abbot was freed from tolls at the bridge of Bramber.51 There is at least one other castle of a tenant-in-chief that was built on the land of another and through which a feudal relationship was probably forged. The first caput of the honour of Bayeux, that is the land of the Domesday tenant-in-chief Alfred of Lincoln, was most likely situated in Stainby in Lincolnshire where a motte and bailey can still be seen.52 By the early twelfth century, however, the honorial castle was a ringwork at Welbourn, and there it was constructed on sokeland that belonged to the honour of Eye. Here a family tie underpinned the relationship, for Robert Malet, the Domesday lord of Eye, was a kinsman of Alfred of Lincoln.53 But tenurial subordination of this kind was probably a rarity. An accommodation in the form of an exchange of land seems to have been preferred. We can cite five unequivocal examples before 1086. Domesday records that the king’s castle of Wareham in Dorset was built on 1 hide of the abbess of Shaftesbury’s manor of Kingston in the same county.54 The land in question was not an urban appurtenance of the rural estate like that of Chebsey in Stafford. Rather ‘Wareham’ references the castle of Corfe within the bounds of the manor, for in subsequent records the fact is explicit.55 In return for the hide the king gave the abbey the church of Gillingham in Dorset with its appurtenances, which was worth 40s. The deal seems to have been

So identified in the Alecto edition of Great Domesday Book. For the site of the minster, see Ann Williams, ‘The Piety of Earl Godwine’, ANS 34, 2011, 237–56 at 243. 49 I. J. Sanders, English Baronies: A Study of their Origin and Descent, 1086–1327, Oxford 1960, 13. 50 GDB 28: SSX 13, 9. 51 Calendar of Documents Preserved in France, 918–1206, ed. J. H. Round, Public Record Office 1899, 37–8; Regesta: William I, no 146. 52 David Roffe and Hilary Healey, ‘South Lincolnshire Earthworks Survey: Tower Hill, Stainby’, (http://www.roffe.co.uk/earthworks/castles/stainby.htm [accessed 15 February 2017]). 53 K. S. B. Keats-Rohan, ‘Domesday Book and the Malets: Patrimony and the Private Histories of Public Lives’, Nottingham Medieval Studies 41, 1997, 13–53; David Roffe, ‘Welbourn Castle, Lincolnshire’, Nottingham Medieval Studies 41, 1997, 54–6. 54 GDB 78v: DOR 19, 10. Kingston was assessed at 15 hides c. 1170 (Charters and Customals of Shaftesbury Abbey, 1089–1216, ed. N. E. Stacy, Records of Social and Economic History, new series 39, British Academy 2006, 136. However, the hide at Arne may be a sixteenth which might suggest a continuing tenurial link (ibid., 134 and n). 55 The Book of Fees I, 91. 48



Castle Construction, Conquest and Compensation

183

an advantageous one for Shaftesbury, for the modal value of a hide was half that, its 16 hides yielding only £16 in 1066. Haxton and Benton in Devon, assessed at 2 virgates and 1 virgate respectively, were given to the bishop of Exeter by the count of Mortain in exchange for ‘the castle of Cornwall’ (castelli de Cornualia).56 The castle in question is usually identified with Dunheved, that is Launceston, noticed in the Cornish folios of GDB, since it is sometimes so identified in later sources.57 The bishop of Exeter did indeed hold land to the south at Lawhitton, but the 30s. at which Haxton and Benton were valued in 1066 and 1086 hardly seem adequate compensation for the £20 that Dunheved rendered TRE, or even the £4 in 1086. Moreover, Dunheved may not have belonged to the bishop in 1066. Exceptionally, no TRE holder is recorded in the Domesday entry. According to the Exon Domesday, though, the twin settlement of Launceston had been held by Earl Harold in 1086.58 Dunheved had probably long been an administrative centre – there was a mint there both before and after the Conquest59 – and so it is likely that it too had been held by the earl. It is possible, then, that Haxton and Benton were given in exchange for land lost in the construction of the castle of Trematon, also noted in Domesday Book.60 The bishop held the manor of St Germans to the west and, significantly, it is recorded that 1 hide had been taken away from it. Further, Domesday adds that the bishop’s market there had been reduced to nothing because of the count’s market nearby, which, Exon adds, was in his castle.61 If this identification is correct, then a fair deal was struck. The land lost in St Germans rendered a barrel of ale and 30d., compared with the 30s. that Haxton and Benton rendered in 1086. No doubt the loss of the ale was still lamented. Athelney may not have done quite so well. Purse Caundle in Dorset was given to the abbey in exchange for Bishopstone where the count of Mortain had built his castle of Montacute.62 No TRE value is given for either holding, but in 1086 Bishopstone was worth in total £9 3s. where Purse Caudle was rated at only £3 7s. 6d. It is significant that compensation, however inadequate, was nevertheless given. According to Waltham abbey tradition, first recorded in the twelfth century, the holy rood that the church housed was found on St Michael’s Hill in Montacute, whence it was transported to Waltham by Tovi the Proud sometime in the early eleventh century.63 The construction of a castle there has been widely understood as a calculated insult to English religious sensibilities.64 However that may be, it would seem that the property rights of an English house were still observed.65 By c. 1102 the castle and the associated borough were given to Montacute priory on its foundation, perhaps in recognition of the cult status of the site.66 Tintinhull 56

GDB 101v: DEV 2, 9–10 and note. GDB 121v: CON 5, 1, 22. 58 Exon 264 b1. 59 Martin Allen, Mint and Money in Medieval England, Cambridge 2012, 387. 60 GDB 122: CON 5, 2, 11. 61 GDB 120v: CON 2, 6; Exon 199b. 62 GDB 78v: DOR 15, 1. GDB 93: SOM 10, 86. 63 The Waltham Chronicle, ed. and trans. Leslie Watkiss and Marjorie Chibnall, OMT, Oxford 1994. 64 S. J. Prior, ‘Strategy, Symbolism and the Downright Unusual: The Archaeology of Three Somerset Castles’, in People and Places: Essays in Honour of Mick Aston, ed. M. Costen, Oxford 2007, 76–89 at 82–4. 65 The count of Mortain held two hides in Ilton and the two in Ashill which were claimed by Athelney, but the context was probably loanland (GDB 91: SOM 10, 1; 6). 66 Two Cartularies of the Augustinian Priory of Bruton and the Cluniac Priory of Montacute in the County of Somerset, ed. H. C. Maxwell Lyte et al., Somerset Record Society 8, 1894, 119; VCH Somerset, III, 210–24. 57

184

David Roffe

immediately to the north, exchanged with Glastonbury abbey for Camerton and probably intended as a park or a demesne, was given to the priory at much the same time as Purse Caundle.67 In the two remaining cases there is insufficient evidence for comparative value. The first castle in Canterbury, marked now by the extramural motte known as Dane John, apparently encroached on the land of St Augustine’s abbey. In exchange, according to Domesday, fourteen burgesses who that had paid landgabel to the king in 1066 were transferred to the abbot.68 This seems not to have been the full extent of the settlement. A memorandum in the twelfth-century British Library Royal MS 1B xi records the names of eleven of the burgesses and the rents they paid, but adds that Abbot Scolland also accepted the churches of St Andrew the Apostle and St Mary de Castro.69 Domesday Book provides no data for a balance sheet of loss and gain. Nor does it for Rochester. In 1086 the bishop held land in Aylesford in Kent to the value of 17s. 4d. ‘in exchange for the land on which the castle stands’.70 The property taken by the king cannot be identified in the various references to the city. Boley Hill, on which stands a motte-like structure, has been canvassed as a likely site. However, it is now thought more likely that it was a later outwork or siege work which, encircled with a stone wall in the thirteenth century, was pressed into service as an outer bailey.71 The present castle, built by Bishop Gundulf after 1088, most likely marks the land lost. The site was described as ‘the fairest part of Rochester’ in the Textus Roffensis.72 This is the extent of the explicit evidence for compensation following the loss of land in the construction of castles before 1086. It is perhaps of some relevance to add the circumstances surrounding the building of the king’s palace in Winchester in 1070. Whether this was an extension of the Anglo-Saxon palace or a new-build is unclear. Nevertheless, it encroached on an existing street and took in a considerable area of land to the south of the High Street. Part was already within the king’s fee but the new structure also encroached on the cemetery of the monks of New Minster. In marked contrast to Worcester’s experience, Alton and Kingsclere were given to New Minster in exchange for the land lost.73 Likewise, Westminster abbey was given Battersea in exchange for the palace at Windsor that had been granted to it by Edward the Confessor on its foundation.74

67

GDB 91v: SOM 19, 9. GDB 2: KEN C1. 69 BL Royal MS 1B xi, f.146v, printed in William Urry, Canterbury under the Angevin Kings, London 1967, 445. There is a thirteenth-century version of the same memorandum in the White Book of St Augustine’s, TNA E164/27, f.15v.0+. Urry, Canterbury under the Angevin Kings, 66, asserts that the list of burgesses relates to the properties destroyed and so this memorandum relates to a separate transaction. Despite the disparity in numbers, the text more likely refers to the tenements received. They rendered 9s 6d in landgabel. 70 GDB 2v: KEN 1, 2. 71 Richard Peats and Paul Drury, Rochester Castle: Conservation Plan, I Understanding and Significance, Paul Drury Partnership, Twickenham 2009, 15; Jeremy Ashbee, Peter Purton, and Hugh Doherty, ‘Rochester Castle and the Great Siege of 1215’, Castle Studies Group Autumn Conference October 2015, Castle Studies Group Journal 29, 2016, 16–17. 72 Textus Roffensis: Rochester Cathedral Library Manuscript A.3.5, ed. Peter Sawyer, Early English Manuscripts in Facsimile 6 and 11, Copenhagen 1958–62, II, 201v. 73 GDB 43: HAM 6, 1; 9; Winchester in the Early Middle Ages: An Edition and Discussion of the Winton Domesday, ed. Martin Biddle, Winchester Studies 1 Oxford 1976, Winton Domesday, nos 57, 80/1, pp. 293–302. 74 GDB 32: SRY 6, 1. Twelfth-century Westminster tradition maintained that Battersea was given in exchange for Edward the Confessor’s crown, while lands in Essex were the compensation for the loss of Windsor (Regesta: William I nos 290, 317). 68



Castle Construction, Conquest and Compensation

185

Further references to exchange and grants of land may hint at other instances. The largest exchange of all was what was known as ‘the exchange of the castle of Lewes’ or, more concisely, ‘the Lewes exchange’. William de Warenne was given extensive lands in East Anglia, notably Norfolk, in compensation for estates that he had lost to the Rapes of Bramber and Pevensey.75 This transaction, however, represented the restructuring of his castlery rather than the mitigation of loss occasioned specifically by castle construction. There was a similar restructuring of Ilbert de Lacy’s castlery of Tanshelf/Pontefract at some time before 1086.76 Probably more specific was the compensation Fécamp received in respect of Hastings. The abbey had been granted Rameslege, apparently representing Brede, and its port in the early eleventh century by King Cnut.77 In 1085 King William gave the abbey Bury in West Sussex in consideration of the monks’ claim against him for ‘the possessions in Hastings which they had held in King Edward’s day’. If it was worth more than the rents (redditus) they had lost, they were nevertheless to keep the manor with all laws and customs and with sake and soke, but, if less, William undertook to discharge his obligation to them through an exchange of equivalent value.78 Here, it has been assumed, Fécamp had held the whole of the town of Hastings and that William had deprived them of it on account of its strategic importance. This charter, though, suggests that by the late eleventh century the monks had held only renders. The port of the OE diplomas, if Hastings, had either been lost or, more likely, Fécamp’s interest in it had always been confined to its dues. Some of the properties from which those rents were paid were situated in the Old Town to the east of the castle; the manor of Brede continued to hold land there, along with the advowson of two churches, into the nineteenth century. The castle fee, the land of the king, was confined to a narrow strip of land encompassing the castle on West Hill and a park to the northwest. It was probably for loss of this, or a part of it, that Fécamp received compensation.79 With somewhat less confidence we can cite five further exchanges that may have been related to castle construction. In Kent the abbot of St Augustine’s, Canterbury, was given half a sulung of the manor of Leeds, worth 10s., in exchange for the park of the bishop of Bayeux.80 Park and castle often went together in what is now termed a landscape of lordship.81 The castle of Leeds is first documented c. 1114, when it was held by the Crevecoeur family, but this transaction may indicate that it was already in existence by 1086. The earliest surviving structures are stone cellars of the early twelfth century, but the footprint of the castle is suggestive of an earlier ringwork or motte and bailey castle.82 Again, Robert de Verly received Field Dalling in Norfolk for Roding in Essex.83 As is common in LDB the other party is not named and the land cannot be precisely identified: Roding is an area 75 L. F. Salzman, ‘The Rapes of Sussex’, Sussex Archaeological Collections 72, 1931, 20–9; VCH Sussex, I, 354ff; J. F. A. Mason, ‘The Rapes of Sussex and the Norman Conquest’, Sussex Archaeological Collections 102, 1964, 79–86; idem, William I and the Sussex Rapes, Historical Association, Bexhill 1966. 76 GDB 373v: YKS CW1–3. 77 S 949, 982. 78 Regesta: William I 144. 79 Mark Gardiner, ‘Shipping and Trade between England and the Continent during the Eleventh Century’, ANS 22, 1999, 71–93 at 91. 80 GDB 7v: KEN 5, 57. 81 Robert Liddiard, ‘The Deer Parks of Domesday Book’. Landscapes 4, 2003, 4–23. 82 Robert Liddiard, personal communication, email 23 April 2017. 83 LDB 262: NFK 38, 3.

186

David Roffe

name which is applied to no fewer than sixteen adjacent holdings in Domesday Book which were subsequently represented by ten separate vills.84 Whether they embodied the full extent of eleventh-century Roding is unclear. However, if the complex extended beyond the area of the later vills, then the exchange could have related to the construction of the castle of Pleshey immediately to the east or Great Canfield to the northwest. Both were motte and bailey castles first documented in the twelfth century but conceivably of the late eleventh.85 Third, sometime before 1086 Ralph Taillebois exchanged Ware in Hertfordshire for land in Goldington, Cople and Salph in Bedfordshire which had been held by Hugh de Grandmesnil. No reason is given for the transaction, but it may relate to the construction of the motte and bailey castle of Risinghoe in Goldington. It is first noticed in the late twelfth century, but was then said to be ‘old’.86 The exchange of Ware may have been in preparation for its construction.87 Fourth, Ernulf de Hesdin’s exchange of Holcot and Biddenham in Bedfordshire for William Speke’s manor of Toddington is most likely related to the construction of the motte and bailey castle now known as Conger Hill to the east of the church.88 There are no unambiguous references to this structure, but, lacking any trace of masonry, it is thought most likely to date from the late eleventh century.89 Fifth and finally, the exchange of half of Princes Risborough for Ellesborough in Buckinghamshire by Ansculf de Picquigney, ‘by order of the king’ according to Domesday Book, may have been related to the construction of the motte and bailey now known as Cymbeline’s Castle in the latter.90 There are no known medieval references to the site. Other exchanges are tantalizingly vague. The bishop of Salisbury held Charnage in Wiltshire and Chernel, Bardulfeston, Athelhampton, Bowood, Buckham and Wellwood in Dorset in exchange for a Scepeleia/Scipeleia.91 The transaction was apparently an important one; in total the land was worth some £16 in 1086. Unfortunately, Scepeleia/Scipeleia cannot be identified and it is therefore impossible to determine whether it was related to castle construction. Compensation might indeed be given for losses incurred in other circumstances. Ælfric held Milford on Sea in Hampshire of the king in exchange for land that he had held TRE, probably in Barton, which had subsequently been taken into the forest.92 But straight swaps may more usually attest consolidation of assets. Robert de Tosny, for example, probably proposed to exchange his peripheral manor of Marston for two manors in Great and Little Ponton because they were closer to his core estates in the Kesteven division 84

Steven Bassett, ‘Continuity and Fission in the Anglo-Saxon Landscape: The Origins of the Rodings (Essex)’, Landscape History 19, 1997, 25–42. 85 Cathcart King, Castellarium Anglicanum, I, 143, 146. 86 Cartulary of the Abbey of Old Wardon, ed. H. G. Fowler, Bedfordshire Historical Record Society 13, 1930, 125, 131. 87 Lowerre, ‘Why Here and not There?’, 127–30. 88 GDB 212, 214v: BDF 20, 1.25, 1; 4. 89 Albion Archaeology, Extensive Urban Survey – Bedfordshire and Luton: Toddington Archaeological Assessment, Bedfordshire County Council and English Heritage 2003, 14. Available online through https://archaeologydataservice.ac.uk/archive/ [accessed 20 December 2017]. Matthew Paris, Matthaei Parisiensi, Monachi Sancto Albani, Chronica Majora, ed. Henry Richard Luard, 7 vols, Rolls Series 1872–83, V, 242) describes a dwelling belonging to Paulinus Pever in Toddington as ‘like a palace, with state rooms, chapel, bed chambers and other apartments of stone covered with lead, and environed with orchards and parks in a manner which astonished all beholders’, but this is more likely to refer to Toddington Old Park to the northeast since there is no evidence of stonework in the castle site. 90 GDB 148v: BUK 17, 2. 91 GDB 66: WIL 3, 5; GDB 77: DOR 3, 14–18. 92 GDB 51, 51v: HAM NF3, 12. NF9, 40.



Castle Construction, Conquest and Compensation

187

of Lincolnshire.93 In this case, though, Countess Judith apparently failed to honour the bargain. In LDB most references to exchange do not even notice the second parcel of land concerned. Unless the reason is explicit, simple grants are likewise impossible to contextualize and explain. One in Lincolnshire, however, is particularly resonant. It is recorded in Domesday that: Kolsveinn has in the city of Lincoln 4 tofts of his kinsman Cola’s land; and outside the city he has 36 houses and 2 churches to which nothing belongs, which he built on the wasteland that the king gave him, and that was never before built upon. Now the king has all the customs from them.94

Kolsveinn was one of the very few Englishmen who were major tenants-in-chief in 1086. He is not known to have held land before the Conquest, but in addition to the four tofts in the city his kinsman Cola was in possession of a manor in Barlings.95 Some of his predecessors in the rest of his fee may have also been family members. In the twelfth century his descendants were hereditary constables of Lincoln castle and of Lincolnshire and it seems likely that Kolsveinn had some similar ministerial role in 1086.96 The land that the king gave him was adjacent to the suburb of Butwerk, to the east of the Lower City.97 Cola’s tofts were probably situated in the Bail, for subsequently the Constable’s honour met in Bardolfhalle in the northwest corner.98 The castle, however, took in much of the Upper City, as the record of the destruction of 166 mansiones attests. Indeed, the whole of the area appears to have been its outer bailey.99 It is not unlikely that Kolsveinn lost land and the wasteland given to him by the king was compensation. I have come across no further grants that are particularly suggestive. Our total evidence base for compensation for the construction of castles is summarized in Table 2. We have two explicit cases, and a further two probable, in which a tenurial relationship was forged or expected; four in which an exchange of land was agreed, along with two possible examples; one plea that led to restitution; and one likely grant of land in lieu. At best there is evidence for twelve instances of compensation before 1086. The eight definite cases represent 12% of the sixty-nine castles documented within the first thirty-six years of the Conquest, and the remainder a further 6%. These figures, even at the lower level I would suggest, are more significant than they might at first appear. There is, of course, no comprehensive record of castles in the late eleventh century, much less of the circumstances of their construction. 93

GDB 377: LIN CK19. GDB 336v: LIN C22: Colsuen habet in Lincolia ciuitate.iiii. toftes de terra Cole nepotis sui. et extra ciuitatem xxxvi. domos et ii. æcclesias in quibus nihil adiacet. quas hospitauit in wasta terra quam rex sibi dedit. et quæ numquam ante hospitata fuit. Modo habet rex omnes consuetudines ex eis. Both the Phillimore and Alecto editions of Domesday Book translate nepos as ‘nephew’. The logic of the text, however, suggests that Cola was a predecessor of some kind and so the alternative meaning of ‘cousin’ or ‘kinsman’ is to be preferred in this context (Dictionary of Medieval Latin from British sources, sv ‘nepos’ 3 and 4). 95 GDB 356v: LIN 26, 7. 96 J. W. F. Hill, Medieval Lincoln, Cambridge 1948, 48, suggests that he was merely a quisling, but, it is possible that he inherited something like the role of staller. See Paul Everson and David Stocker, Custodians of Continuity: The Premonstratensian Abbey of Barlings and the Landscape of Ritual, Lincolnshire Archaeology and Heritage Reports Series 11, Heckington 2011, 375–6. 97 Hill, Medieval Lincoln, 133–4. 98 Hill, Medieval Lincoln, 105. 99 David Stocker and Alan Vince, ‘The Early Norman Castle at Lincoln and a Re-Evaluation of the Original West Tower of Lincoln Cathedral’, Medieval Archaeology 41, 1997, 233. 94

188

David Roffe

Most that are known (forty-seven out of sixty-nine) are largely incidental references in Domesday Book; the Domesday commissioners had no brief to record them systematically. The rest are found in diverse sources that again are rarely discursive: the Worcester record of depredation is the exception. It is remarkable that so many instances of compensation are known.

mb mb mb mb mb mb mb mb mb mb rw mb

yes 1066? 1066 1068 1069 1070 1066?

Date

Grant

Litigation

Exchange

Tenurial link

Foundation

Castle Parties Bramber* William de Braose/Fécamp Canterbury* King/St Augustine’s Corfe* King/Shaftesbury Hastings King/Fécamp Lincoln King/Kolsveinn Montacute* Mortain/Athelney King/Roger of Berkeley Nesse Rochester* King/Rochester Stafford King/Henry de Ferrers Trematon* Mortain/King Welbourn Alfred of Lincoln/Malet Windsor* King/Ralf fitzSeifrid * = compensation explicit † mb = motte and bailey, rw = ringwork

Type†

Table 2: Castles and compensation, 1066 to 1086

1085

yes yes yes

1085 yes

yes yes yes yes

1068? 1070?

yes yes yes

Our sample, small as it is, thus implies that the need for compensation was more widely recognized, and more often satisfied, than heretofore appreciated. Its characteristics tell us something of the circumstances and timing. First, despite limited evidence for dating, it seems that compensation was as likely to be offered for castles built early in the reign of the Conqueror as later. The construction of Hastings, one of our probable sites, can be assigned to 1066 by witness of the Bayeux Tapestry.100 Domesday records that Nesse was built by William fitzOsbern and so it can be dated between 1066 and 1070, when William left England for the last time.101 According to the Anglo-Saxon Chronicle Lincoln was built in 1068 and Montacute was in existence by the following year, when it was besieged by the English.102 Otherwise, the date of construction is a matter of supposition. Canterbury and Rochester, like Dover, may be early in William’s reign, marking the Conqueror’s first approach to London.103 Dunheved and Trematon, if built by the count of Mortain, must postdate the grant of Cornwall sometime in or after 1068.104 Stafford may be slightly later.105 In contrast, Worcester, where there was no immediate compensation, was built sometime before 1069, when Archbishop Ealdred died. The monks had to wait until 100 David

M. Wilson, The Bayeux Tapestry, London 1985, plates 49–50. P. Lewis, ‘William fitzOsbern’, ODNB. 102 ASC, 1066D; Orderic, II, 228–9. 103 Fradley, ‘The Old and the New’, 340–1. 104 Brian Golding, ‘Robert of Mortain’, ANS 13, 1990, 119–44 at 126–8; Brian Golding, ‘Robert, count of Mortain’, ODNB. 105 Youngs and Morgan, ‘Documentary Evidence’, 32. 101 C.



Castle Construction, Conquest and Compensation

189

1217 before restitution was made. In that year Henry III granted them half of the castle including the keep.106 This transaction attests a very long memory on both sides. In the present context it also serves to show that compensation was not necessarily contemporary with the construction of a castle. For most of the exchanges recorded in Domesday Book we have only a terminus ante quem of 1086. It is salutary to note, however, that Fécamp, as evidenced by its charters, only made good its claims on Hastings and Bramber in 1085, probably some twenty or so years after its loss. Roger de Berkeley and probably Henry de Ferrers were still awaiting satisfaction at the same time. Such evidence as we have, then, suggests that compensation was characteristic of the Norman settlement rather than conquest. It was, nevertheless, retrospective: it stretched back to the earliest period of castle building. By and large, it also seems to have been equitable. The Hastings exchange explicitly states the principle that land of equal value ought to be given in return for assets lost, and suggests that there were mechanisms to correct aberrant settlements. Where comparable data is available, our sample bears out its accuracy. The rent for the half hide in Clewer on which Windsor was built in particular was just over the going rate for such a piece of land. It is probably only Athelney that enjoyed a less than advantageous deal. Compensation was no anodyne fiction. Not surprisingly, the recorded beneficiaries were all adherents of the new regime. Fécamp was most closely identified with it. The abbey had been granted its lands well before the Conquest, but Earl Harold had usurped many of them. It had been one of Duke William’s staunchest supporters in the proposed invasion of England in early 1066 and no doubt looked forward to recovering its lost lands as a quid pro quo. Henry of Ferrers and Roger of Berkeley were Norman tenants. St Augustine’s, Canterbury, and Rochester were under the authority of Englishmen until 1070 and 1075 respectively, but it is likely that it was their Norman successors who negotiated compensation. Abbot Egelsin of St Augustine’s in particular was in conflict with the Conqueror, and it was his successor, Scolland, who benefited from a settlement. The abbot of Athelney is unknown at this time. Kolsveinn was English but clearly a trusted minister of the king and a substantial tenant-in-chief. The lowest in the social scale was Ralf fitzSeifrid. He too, like his brother, was probably a minister but almost certainly little more than a tainus regis in status. His successors were minor tenants-in-chief, probably originally holding by sergeancy. Unless we take the experience of Dover at face value and as the norm, the fate of the ordinary burgess, free man, or villein displaced by castle building is largely unrecorded. It may have been less cavalier than at first appears. According to Orderic Vitalis, William posted guards at the gates of Exeter after the siege in 1068 to prevent his troops from looting the city.107 There, however, the castle was built on royal land, so the question of compensation may not have arisen. At Barnstaple, some burials, probably recent, appear to have been carefully exhumed before the construction of the castle there, while in Norwich St John Berstrete was probably consciously built to replace a cemetery subsumed beneath the bailey of the castle.108

106 Patent

Rolls of the Reign of Henry III Preserved in the Public Record Office, 1216–1225, London 1901, 46, 52. There may have been a tacit acceptance of a right before this date: in 1141 the Empress Matilda granted to William de Beauchamp ‘the castle with the motte (castellum de Wigornia cum mota)’, a formulation that probably attests a separate title (Regesta, III, no 68). 107 Orderic, II, 210–11. 108 T. J. Miles, ‘The Excavation of a Saxon Cemetery and Part of the Norman Castle at North Walk, Barnstaple’, Journal of the Devon Archaeological Association 44, 1986, 59–84 at 68; Popescu, ‘Norwich Castle’, 8.

190

David Roffe

Elsewhere there is little evidence of gratuitous disturbance of graves. It is far from clear that the churches destroyed at Lincoln, Pontefract and Trowbridge were in use after the Conquest. The destruction of housing, by contrast, is unequivocal. Nevertheless, castle construction might also be the occasion for new initiatives that provided replacements, if only eventually. At Norwich and Nottingham French boroughs were founded nearby to meet the needs of the garrisons and new settlers alike.109 New markets were set up at Bramber, Launceston and Trematon. Elsewhere topographical analysis suggests that castle and borough often developed symbiotically.110 The borough of Bristol, for example, saw the replanning of the Old Market area in the post-Conquest period, presumably in tandem with the construction of the castle to the west.111 Examples could be multiplied, but detail is rarely supplied. Durham sources provide more than for most settlements. There the castle was built by King William in 1072, probably on the site of the earl’s residence.112 Some thirty years later Bishop Ranulf Flambard extended its defences to join up with the new cathedral and cleared the space in between, the present Palace Green, of housing. There is no explicit evidence for the fate of those who were displaced. But probably at much the same time Ranulf built the Old Bridge, the present Framwellgate Bridge, across the River Wear and presumably was responsible for laying out the burgage plots of the Old Borough to the west with which it communicated.113 It is possible that we see here a concerted act of town planning that met the needs of both the lord and those he had displaced. The supposition that castle building was an arbitrary droit de seigneur is not supported by the evidence that we have. Where castles were sited on the demesne we have little information. Nevertheless, wholesale impoverishment of the inhabitants of the estate was not in the long-term interests of any lord, and it seems intrinsically likely that provision was made for them in one way or another. Where it encroached on the rights of others compensation is more evident. It was rarely immediate. As one might expect, there was probably no negotiation in the initial phases of the Conquest. But rights were subsequently asserted and, perhaps equally often, acknowledged. The process was probably predominantly a legal one. Enfeoffment and grant speak for themselves. Exchange was equally formal. Where the matter has been considered at all, it has been viewed as either a cover for illicit seizure or an essentially private arrangement.114 Neither view is supported by the 160 or so recorded examples in Domesday Book.115 On the one hand, there was 109 GDB

280: NTT B2; LDB 118: NFK 1, 66. a recent review, see Keith D. Lilley, ‘The Norman Conquest and its Influence on Urban Landscapes’, in The Archaeology of the 11th Century: Continuities and Transformations, ed. Hadley and Dyer, 30–56. 111 Nigel Baker, Jonathan Brett, and Robert Jones, Bristol: An Archaeological Assessment, Oxford 2018, 101–3; Lilley, ‘Norman Conquest and Urban Landscapes’, 37–8. 112 Symeonis Monachi Opera Omnia, ed. Thomas Arnold, Rolls Series 75, 2 vols, 1882–5, II, 199–200; M. Leyland, ‘The Origins and Development of Durham Castle’, in Anglo-Norman Durham, 1093–1193, ed. D. Rollason, M. Harvey, and M. Prestwich, Woodbridge 1994, 407–24 at 408. Symeonis Monachi, I, 215, ascribes the construction of the castle to Earl Waltheof. 113 Symeonis Monachi, I, 140; Margaret Bonney, Lordship and the Urban Community: Durham and its Overlords 1250–1540, Cambridge 1990, 29–31, 34; M. O. H. Carver, ‘Early Medieval Durham: The Archaeological Evidence’, in Medieval Art and Architecture at Durham Cathedral, British Archaeological Association Conference Transactions 3, 1980 for 1977, 11–19. 114 R. Welldon Finn, The Eastern Counties, London 1967, 30–2; Marten, ‘Impact of Rebellion’, 39–40. 115 For an analytical list, see Robin Fleming, Domesday Book and the Law: Society and Legal Custom in Early Medieval England, Cambridge 1998, 510. 110 For



Castle Construction, Conquest and Compensation

191

no claim to exchanged land on any basis other than the fact of the exchange itself. On the other, deliverers or the like are noticed in a handful of cases. Exchanges were apparently transactions that were officially sanctioned; indeed, a number are recorded in the extant charters.116 Where they ensued from castle construction, they often most likely followed a legal procedure of one kind or another. Neither the king nor his barons, it seems, were free to build castles willy nilly. Sooner or later they had had to square their actions with prior claims to the land on which they built. What has been perceived as the most iconic instrument of conquest was subject to the rule of existing law and custom. This is a conclusion that goes a long way to bridging the gulf between advocates of the traditional understanding of the castle and those of so-called ‘peaceable power’. It recognizes the reality of the military phase of the Conquest as well as acknowledging that it had consequences which demanded attention and action thereafter. The castle may often have been born of necessity, but ultimately it had to demonstrate its legitimacy. This is surely a resounding token of a settlement that adhered to firm principles of tenure and right. Developed by Archbishop Lanfranc in 1070, the Norman narrative of conquest was propaganda which was belied by the mechanisms of settlement. The English who fought with King Harold and died at the battle of Hastings had unceremoniously forfeited their lands. Otherwise submission to William confirmed tenure. Since at the outset William determined to govern England through English agents and English law, this introduced no new principles of tenure: the initial settlement represented continuity with the Anglo-Saxon past, albeit with the king assuming the right to grant and confirm. So did the conferment of the lands of the English who subsequently forfeited: the new lord was to hold by inheritance from his antecessor, the lord who had held in 1066.117 Compensation for castle construction reveals that these principles were widely understood and duly applied. The claims that arose were probably among the least contentious of the issues that had to be resolved. Much of the litigation between 1066 and 1086, along with the cases that came to light in the Domesday inquest itself, hinged on the intricacies of loanland and commendation. In comparison the straightforward claims to land to which castle building gave rise were comparatively simple to resolve. Nevertheless, they graphically illustrate the means by which the Norman settlement was expeditiously effected. Despite the messy edges to which the intricacies of English tenure were conducive, free enterprise depredation was limited by the interests and rights of equals. One tenant-in-chief’s illicit invasion was another’s loss and so the settlement was essentially self-policing.118 Even royal grants that cut across such rights could be subject to mitigation. Sometime between 1081 and 1083 William undertook to recompense Ely abbey for loss of demesnes which he had gifted to other lords.119 The various pleas of the Conqueror’s reign have usually been seen as evidence of the chaotic character of the Norman settlement. We can now appreciate that they more convincingly attest the working out of an ordered process. We have come a long way from Stamford castle. It is a sad fact that Chris Mahany never contributed, in print at least, to the reassessment of the castle to which her findings contributed. More’s the pity. She had an encyclopaedic knowledge of the subject, as much else, and an acute intelligence that was capable of cutting through 116 Regesta:

William I, nos 74, 199, 144, 158, 204, 205, 215, 251, 290, 317. Roffe, Decoding Domesday, Woodbridge 2007, 162–76. 118 David Roffe, ‘Domesday Now: A View from the Stage’, in Domesday Now: New Approaches to the Inquest and the Book, ed. David Roffe and K. S. B. Keats-Rohan, Woodbridge 2016, 7–60 at 25–6. 119 Regesta: William I, no 119. 117 David

192

David Roffe

to the essence of a problem. What she would have made of the present findings is debateable. Chris was, I know, unsympathetic to the Jacobin wing of modern castle studies which has all but denied the military functions of the castle. However, I am sure that she would have been intrigued by the methods, if not the conclusions, of the present investigation. Historical and archaeological data interweave to show that the castle was indeed an instrument of conquest, but it fitted into an English tradition of lordly display and the legal apparatus that underpinned it. Its advent must at times have been irksome, but, with compensation recognized as a right, it was the annoyance of compulsory purchase rather than confiscation.

FOUR SCENES FROM THE CHANSON DE ROLAND ON THE FAÇADE OF BARLETTA CATHEDRAL (SOUTHERN ITALY)* Lucia Sinisi The chansons de geste of French tradition have had a significant impact on the culture and imagination of the populations with whom the Normans came into contact as a result of their conquests in both the north and south of Europe. Among the cycles in which the chansons de geste are traditionally grouped, the epic of Roncevaux is undoubtedly the one that has achieved greatest popularity.1 This paper is inspired by Rita Lejeune’s studies of representations of the Roland legend in the sculptures on the lintel of the cathedral of Angoulême, and on many other church walls throughout Europe.2 By proposing an interpretation of the enigmatic sculptures on an archivolt of the cathedral of Santa Maria Maggiore in Barletta in light of the propagation of Roland’s Song in a large area of Europe, this discussion aims at inserting the city of Barletta into the broad catalogue of churches on whose façades the deeds and ‘martyrdom’ of the paladins of France are celebrated. The Vikings became notorious in the chronicles because of their raids and sacking of churches and monasteries, so much so that in the liturgy of the ninth to the eleventh centuries the invocation that God might protect the faithful from the fury of the Vikings was introduced.3 In c. 911, Charles the Simple signed an agreement with the jarl Hrólfr (Rollo) of Norway, allowing Vikings to settle in part of the region to which they gave their name, in exchange for a vassalage relationship and the willingness to receive baptism. In the 150 or so years that followed, Rollo’s Vikings not only abandoned the language of their ancestors (that variety of Northern Germanic spoken by the Scandinavian populations) to embrace the idiom of the Franks, but also assimilated the Franks’ social structure and values. The evidence is – if we give credit to the historians of those times – that they soon adopted and tried to actualize the ideals professed by the Church of Rome at the turn of the first millennium: to liberate the territories in the hands of the Arab Muslims, in particular Spain and the Holy Land.

*

I am very grateful to Luisa Derosa for allowing me to use her photos in this article. Alfonso D’Agostino, L’epica di Roncisvalle, Linguista e filologia, Testi, I, Milan 2006 (= Linguistica e filologia). 2 Rita Lejeune, ‘Trois épisodes de la Chanson de Roland sur un linteau de la cathédrale d’Angoulême’, Comptes rendus des séances de l’Académie des Inscriptions et Belles-Lettres, 2, 1961, 381–99 ; idem, ‘Le linteau d’Angoulême et la Chanson de Roland’, Romania 82, 1961, 1–26 ; Rita Lejeune and Jacques Stiennon, The Legend of Roland in the Middle Ages, 2 vols, London 1971 (orig. tit. : La Légende de Roland dans l’art du Moyen Âge, Bruxelles 1966). 3 C. Vivanti, ‘Pio II e la cultura geografica del suo tempo’, in Europa e Mediterraneo tra Medioevo e prima età moderna: L’osservatorio italiano, ed. S. Gensini, Pisa 1992, 135–6, as cited in Salvatore Tramontana, ‘Popoli, etnie e mentalità alla vigilia della conquista di Sicilia’, in I caratteri originari della conquista normanna. Diversità e identità nel Mezzogiorno (1030–1130), Atti delle sedicesime giornate normanno-sveve, Bari, 5–8 ottobre 2004, eds Raffaele Licinio and Francesco Violante, Bari 2006, notes 34 and 93. 1

194

Lucia Sinisi

The attitude of the church did not vary between the time of the bishop of Cordoba Eulogius (d. 859), who aspired to liberate Spain from the Muslims, and the preaching of the First Crusade by Pope Urban II, which was preceded in 1063 by Pope Alexander II’s granting remission of sins to all those who went to Spain to fight the Saracens: the warrior killed while fighting in a war ordained by God was to be considered a martyr.4 ‘Nowhere is this pious conflict more apparent than in the interminable Old French songs of geste, which present, as in a mirror, the reflection of militant and missionary Christianity as the Middle Ages conceived it.’5 The Chanson de Roland It is in this cultural climate that the legend of the battle of Roncevaux takes shape and spreads, as it is handed down to us by the Chanson de Roland, which narrates the defeat of Charlemagne’s rearguard at the hands of the Saracen troops headed by King Marsile of Zaragoza, following the betrayal perpetrated by Ganelon, stepfather of the brave paladin Roland.6 In essence, it narrates the fight of Christianity against the païen, the Muslims. The Chanson is extant in ten manuscripts, of which six are codices, and three are just fragments.7 To these some versions in different languages should be added, all of them linked to the protograph of the Chanson de Roland.8 The oldest version is preserved in the famous manuscript, Digby 23 (Part 2), Oxford, Bodleian Library,9 which is generally thought to date from the second half of the twelfth century, though this has been widely debated.10 The manuscript is composed of two parts. The first contains Plato’s Timaeus in the Latin version of Calcidius, dating back to the first half of the twelfth century, of French origin. It was probably a gift to the abbey of Osney from Master Henry of Langley (probable year of death 1263), at which time it was already joined to the second part, which contains La Chanson de Roland. The version offered by the Oxford codex, which is made up of 4,002 lines in assonated epic decasyllables, divided into 291 laisses, is nonetheless – as we shall see – somewhat different from those offered by the other witnesses, not least because it is the only version in Anglo-Norman. 4

Lejeune, ‘Le linteau d’Angoulême’, 3–5. Meredith C. Jones, ‘The Conventional Saracen of the Songs of Geste’, Speculum, 17, 1942, 201–25 at 201. 6 Paul Aebischer, Préhistoire et protohistoire du Roland d’Oxford, Bern 1972, 75–87. 7 O (Oxford, Bodleian Library, Digby 213, part.2 (Anglo-Norman)), V4 (Venice, Biblioteca Marciana, ms 225 (Franco-Venetian)), C (Châteauroux (Indre), Bibliothèque Municipale, ms 1 (Franco-Venetian)), V7 (Venice, Biblioteca Marciana, ms 251 (Franco-Venetian)), P (Paris, BnF, ms fr. 860 (Old French)), T (Cambridge, Trinity College Library, ms R 3–32 (Old French)) L (Lyon, Bibliothèque de la Ville, ms 984 (Old French)), l (Fragments Lavergne (Old French) (two fragments)), F (BnF, nouv. acq. fr. 5327 (Old French lorenese)), B (BL, Add. 41295 G (Old French) (fragment)). 8 n (Old Norse translation contained in branch VIII of the Karlamagnis saga, the so-called Saga af Runzivals bardaga, preserved in three MSS), K (Konrad, Ruolantes liet, original, not a translation), w (Campeu Charlymaen, a prose romance in Welsh), h (Roelantslied, fragments preserved in various libraries, in Flemish), Carmen (Carmen de prodicione Guenonis (BL, Cottonian, Titus A, XIX) edited in La Chanson de Roland: Edizione critica, ed. Cesare Segre, Milan – Naples 1971, xxxvii–xlvii. 9 The digital version of MS Oxford, Digby 23 (Pt 2) is available at Early Manuscripts at Oxford University: Digital facsimiles of complete manuscripts, scanned directly from the originals Oxford Digital Library (http://image.ox.ac.uk/show?collection=bodleian&manuscript=msdigby23b). 10 La Chanson de Roland, xxxvii. 5

The Chanson de Roland on the Façade of Barletta Cathedral

195

William of Malmesbury, in his work Gesta regum Anglorum, written roughly between 1120 and 1130, reports that on the Hastings battlefield a cantilena Rollandi was chanted for the troops of William the Conqueror: Tunc, cantilena Rollandi inchoata, ut martium viri exemplum pugnaturos accenderet, inclamatoque Dei auxilio, praelium conserto, bellatumque acriter, neutris in multam diei horam cedentibus.11 Then, having chanted the song of Roland in order to inflame those ready to fight with his warlike example, and having invoked God’s help, they started the battle, and having harshly fought for many hours, nobody from either side surrendered.12

Thus also in Wace, who, writing about the battle of Hastings a few decades later than William of Malmesbury (between 1160 and 1174), refers to chansons de geste of the Carolingian cycle sung by the Norman minstrel Taillefer on the eve of the battle of Hastings: Taillefer, qui mult bien chantout, sor un cheval qui tost alout, devant le duc alout chantant de Karlemaigne et de Rollant, e d’Oliver e des vassals qui morurent en Rencevals.13 (lines 8035–40) Then Taillefer who sang right well, rode mounted on a swift horse before the duke, singing of Karle maine, and of Rollant, of Oliver and the vassals who died in Renchevals.14

These two accounts constitute evidence of the widespread circulation of a ‘song’ in which, already in 1066, the deeds of the heroes who died in the ambush of Roncevaux were narrated, long before its writing in the oldest manuscript that is handed down to us. Above all they testify to how the Normans at that time had already assumed the ideology at the heart of this chanson de geste, by identifying themselves with the heroes of the Carolingian cycle, and raising themselves to the role of defenders of Christianity.15 The Normans in Apulia The circumstances that brought the Normans to southern Italy were quite different from those that led them to attack and conquer England, as was the political and administrative situation they found on their arrival: Langobards and Byzantines were ever contending the territories of Apulia, Calabria and Campania, while Sicily was in the 11

Malmesbury, Gesta regum, I, 454. My translation. 13 Le Roman de Rou de Wace, ed. A. J. Holden, 3 vols, Paris 1970–3, II, 183. 14 Master Wace his Chronicle of the Norman Conquest from Roman de Rou, translated by Edgar Taylor, London 1837, 189. 15 On the inclusion of the Chanson de Roland in the canon of French literature in the nineteenth century, see Andrew Taylor, ‘Was there a Song of Roland?’, Speculum 76, 2001, 28–65, at 35: ‘The Chanson de Roland, then, was more than just a literary monument. Its editorial construction was part of the quest for national origins that dominated French romantic philology’. 12

196

Lucia Sinisi

hands of the Arabs; moreover, influential particularisms and autonomies were strong enough to oppose central powers.16 After the constitution of the county of Aversa in 1024, the first Norman county in continental southern Italy, and in 1042 that of the county of Melfi, by concession of the prince of Salerno, Robert Guiscard managed to conquer large territories in continental southern Italy.17 The Normans faced significant resistance, especially from the urban classes of the coastal Apulian towns, such as Bari, Trani, Otranto and Brindisi.18 It was a lengthy conquest, which, especially at the beginning, carried high costs in terms of destruction and lives. In Apulia, for example, the inhabitants of Cannae were forced to flee to escape the rage of Robert Guiscard, who destroyed their town in 1083; the inhabitants of Bari were forced to do likewise after their city was devastated by William I in 1156. In contrast, other towns, such as Barletta, took advantage of the newly established political condition, and flourished.19 Barletta and the Cathedral of Santa Maria Maggiore As early as 1126 the Normans promoted the construction of the cathedral of Santa Maria Maggiore in Barletta, assigning the task to the protomagister Simiacca and to his son Luca, as is attested in a document written in 1162.20 The Normans’ energetic development of Barletta is closely linked to the first crusades in the Holy Land. As Pope Urban II’s appeal in Clermont in 1095 to reconquer the holy sites of the Christians aroused great fervour in every echelon of Apulian society, the town of Barletta, along with Bari, Brindisi and Otranto, became among the most active ports from which the crusaders sailed to the Holy Land.21 The building of the church of the Holy Sepulchre (Santo Sepolcro), with its xenodochium which very likely hosted pilgrims ready to sail for the Holy Land, mentioned in a document of 1130, might be related to this spiritual climate.22 The cathedral itself bears the mark of the

16

For the conquest of southern Italy and its traits, see I caratteri originari della conquista normanna. Diversità e identità nel Mezzogiorno (1030–1130), Atti delle sedicesime giornate normanno-sveve, Bari, 5–8 ottobre 2004, ed. Raffaele Licinio and Francesco Violante, Bari 2006; in particular, in this volume, Mathieu Arnoux, ‘I Normanni prima della conquista. Costruzione politica e identità nazionale’, 51–66, and Cherubini, ‘Popoli, etnie e territorio alla vigilia della conquista. Il Mezzogiorno continentale’, 67–86; John Julius Norwich, The Normans in the South 1016–1130, London 1967; Gordon S. Brown, The Norman Conquest of Southern Italy and Sicily, Jefferson, North Carolina – London 2003; Graham Alexander Loud, The Age of Robert Guiscard: Southern Italy and the Norman Conquest, Harlow 2000; R. Allen Brown, The Normans, Woodbridge 1984; Hubert Houben, I normanni, Bologna, 2013 (updated version of Die Normannen, Munich 2012), just to mention a few. 17 Errico Cuozzo and Jean Marie Martin, ‘Intorno alla prima contea normanna nell’Italia meridionale’, in Cavalieri alla conquista del Sud. Studi sull’Italia normanna in memoria di Lèon-Robert Ménager, Rome – Bari 1998, 171–93. 18 Storia della Puglia: Antichità e Medioevo, ed. Giosuè Musca, Bari 1987, 119–24; Mario Gallina, ‘Gli stanziamenti della conquista. Resistenze ed opposizioni’, in I caratteri originari della conquista normanna, ed. Licinio and Violante, 151–79. 19 Sabino Loffredo, Storia della città di Barletta, Trani 1983, 129. 20 Codice Diplomatico Barese VIII. Le pergamene di Barletta. Archivio capitolare (897–1285), ed. Francesco Nitti di Vito, Bari 1914, no. 93: in the document they both sign as witnesses, the first as Simiacca protomagister frabice eclesie sancte Marie testatur, the second as Lucas magister filius eius testatur. 21 Angelo Ambrosi, Architettura dei crociati in Puglia. Il Santo Sepolcro di Barletta, Bari 1976, 10; Vittorio Franchetti Pardo, ‘Le città portuali meridionali e le Crociate’, in Il Mezzogiorno normanno-svevo e le Crociate, Atti delle quattordicesime giornate normanno-sveve, Bari, 17–20 ottobre 2000, ed. Giosuè Musca, Bari 2002, 313–14. 22 Luisa Derosa, ‘Barletta e la Terrasanta. Bilancio storiografico e prospettive di ricerca’, in Archeologia Storia Arte. Materiali per la storia di Barletta (sec. IV a.C. – XIX d.C.), ed. Victor Rivera Magos, Saverio Russo, and Giuliano Volpe, Bari 2015, 143–62.

The Chanson de Roland on the Façade of Barletta Cathedral

197

development of Barletta into a harbour and mercantile town, as on the capital of a column inside the church there is an inscription which regards a certain Moscato, who, after the successful siege of Ascalon by the Crusader Kingdom of Jerusalem, gives 200 ducats to the church for the construction of two columns. We can read on the abacus: Muscatus dedit in his duabus columnis ducentos ducales. Qui legit oret pro eo; and then on the palmettes of the annulet: Anno MCLIII, mense augusti, indictione prima, quando capta est Ascalona.23 As a result of this development, Barletta became a port of primary importance for those wishing to travel Outremer, as attested in the Chanson d’Antioche: Aujourd’hui vous pourrez entendre parler de Jérusalem, Et de ceux qui allèrent adorer le sépulcre du Christ. Comment ils firent assembler les armées de tous côtés ; De France, de Berri et d’Auvergne, De Pouille, de Calabre, jusqu’á Barlet-sur-Mer, Jusqu’au pays de Galle, ils demandèrent des renforts En maintes terres que je ne sais nommer.24 (laisse I, lines 12–18)

So now you can hear about Jerusalem, and about those who went to honour the Holy Sepulchre and how they brought together armies from all quarters: they summoned people from France and Berry and its equal the Auvergne, from Apulia and Calabria as far as Barlet on the coast, and beyond even as far as Wales and from innumerable lands whose names I do not know.25

The Archivolts Turning our attention to the façade of the cathedral of Santa Maria Maggiore, we notice its Romanesque architecture, with three portals. Among them only the two lateral ones go back to the first period of the building of the church, around 1150, as the central one was redone and now displays a Renaissance aspect. Both the original portals are surmounted by sculpted archivolts. The archivolt on the left portal is decorated with plant volutes entwined with fantastic creatures, which are held in the middle by a two-bodied centaur placed on the keystone. On both sides the archivolt rests on consoles that respectively portray a time-worn but still recognizable sphinx and a lion. The archivolt itself shows a series of scenes with men and animals engaged in various activities. Examining the sculpted scenes starting from the left console, in which a man is fighting with a huge beast, we can see in succession: a man with a long spear and a dog leaning on him; a man who pierces a bear with a sword, while a figure with a long hood holds the bear; a man with a sack or a mantel flowing on his shoulder, holding a club that is held at the other end 23 Raffaele Iorio, ‘«Ecclesia» e «civitas» barlettane nei documenti medievali’, Archivio storico pugliese 58, 2006, 157–278, at 162. 24 La chanson d’Antioche composée au XIIe siècle par Richard le Pèlerin, renouvelée par Graindor de Douai au XIIe siècle, trans Marquise de Sainte-Aulaire, Paris 1862, 2. 25 The Chanson d’Antioch: An Old French Account of the French Crusade, trans. Susan Eddington and Carol Sweetenham, New York 2011, 101. But note that erroneously, according to the translators, Barletsur-mer refers to the city of Bari.

198

Lucia Sinisi

by another man; a man on a horse, set apart from the following four scenes on the right-hand side of the archivolt by a dragon. The meaning of these scenes is obscure and has yet to be interpreted. The first sculpture on the right of the archivolt represents Samson rending the lion’s jaws, immediately followed by an eagle, after which comes the first of the four scenes we are going to analyse: it is a sleeping figure in front of whom a hirsute animal stands out. If we assume that the sculptor meant to represent somebody dreaming, more than sleeping, we cannot help thinking of Charlemagne’s four dreams which are narrated in the Anglo-Norman version of the Chanson de Roland. These can be divided into two groups: the first two are dreamt by the emperor before Ganelon’s betrayal, the other two after the battle of Roncevaux. In the first one, which occurs in laisse LVI, lines 717–24, Charlemagne, who is on his way back to France after having conquered towns and taken strongholds, during the night dreams he is among the major passes of Sizer, where he brandishes his ashwood spear, but Ganelon seizes it from the upper end and shakes it with such strength that it breaks into fragments. The second dream immediately follows the first (laisse LVII, lines 725–36). Charlemagne dreams he is in Aix-la-Chapelle, where he is bitten on the arm by a boar and attacked by a leopard that springs out from the Ardennes woods; but then a dog, a greyhound, jumping out of the palace, rushes to his aid, biting the boar’s ear, and furiously assailing the leopard. The third dream (laisse CLXXXIV, lines 2525–54) occurs the day after the two brave paladins Roland and Oliver have been killed, along with many others, at the battle of Roncevaux. In this vision Charlemagne’s ranks are hit by a rain of fire and at the same time attacked by hordes of fierce animals (bears, leopards, snakes, demons, gryphons and dragons). Charlemagne himself is attacked by a lion coming from the woods. The result of the struggle is not revealed to the emperor, though he is still sleeping. In the fourth dream (laisse CLXXXV, lines 2555–69), which occurs immediately after the third (just as the second dream ran on from the first), Charlemagne is in Aachen. There is a bear chained to a large rock, which thirty bears from the Ardennes come to reclaim. In this dream too the emperor is saved by a greyhound, which springs out of the palace and rushes at the oldest of the bears. Here too, Charlemagne does not discover whether the outcome of the struggle is favourable to him or not.26 The representation of a dreaming Charlemagne on the façade of the cathedral of Barletta is not the only extant one from the Middle Ages, as there is another fine example in stained-glass on Panel 9 of the so-called Charlemagne window of the cathedral of Chartres, probably made in 1225, where St James appears in the king of the Franks’ dream. The saint summons Charlemagne to follow the Milky Way, which will lead him to Galicia, where he should start the Spanish crusade.27 As we shall see, more depictions of a dreaming Charlemagne also occur in later manuscripts.28 26 Alexander Haggerty Krappe, ‘The Dreams of Charlemagne in the Chanson de Roland’, Publications of the Modern Language Association of America, 1921, 134–41; Frederick Whitehead ‘Charlemagne’s Second Dream’, Olifant, 3, 1976, 189–95; Benjamin M. Semple, ‘Recognizing Roland: The response of the medieval audience to the dreams of Charlemagne in The Song Of Roland’, in Dreams in French Literature, ed. Tom Conner, Amsterdam – Atalanta, GA 1995, 27–46; R. Bowen, ‘The dreams of Charlemagne’, Medium Aevum 23, 1954, 39–40; Wolfgang G. van Emden, ‘Another Look at Charlemagne’s Dreams in the Chanson de Roland’, French Studies 28, 1974, 257–71. 27 For the pictures on the glass windows of the cathedral of Chartres see http://www.medievalart.org. uk/chartres/07_pages/Chartres_Bay07_key.htm (accessed September 2018) and, for a comment on them, Clark Maines’ essay ‘The Charlemagne Window at Chartres Cathedral: New Considerations on Text and Image’, Speculum, 52, 1977, 801–23 at 808. 28 See below.

The Chanson de Roland on the Façade of Barletta Cathedral

199

Scene I

Fig. 1: Charlemagne and boar In the first scene (Fig. 1) a dreaming Charlemagne is portrayed with a hirsute animal in front of him; the representation is related to the second dream in the Chanson de Roland, which in laisse LVII of the Oxford codex is thus narrated: Aprés iceste altre avisiun sunjat: Qu’il ert en France, a sa capele, ad Ais, El destre braz li morst uns vers si mals. Devers Ardene vit venir uns leuparz, Sun cors demenie mult fierement asalt. D’enz de ‹la› sale uns veltres avalat, Que vint a Carles le‹s› galops e les salz; La destre oreille al premer ver trenchat, Ireement se cumbat al lepart.29 (lines 725–33)

So, according to the Oxford MS Digby 23, the animal which assails Charlemagne in his second dream is a vers, that is, a boar, whereas in the Châteauroux witness (C) the animal that endangers Charlemagne is a bear (ors, line 1053),30 as it is in the FrancoVenetian witnesses V7, which does not differ much from C,31 and V4 (line 661).32 This

29

Segre, La Chanson de Roland, 136–8. The Châteauroux Version of the «Chanson de Roland ». A Fully Annotated Critical Text, ed. Marjorie Moffat, Berlin-Boston 2014. Available through Google Books. See also Das altfranzösische Rolandslied. Text von Châteauroux und Venedig VII herausgegeben von Wendelin Foerster, Altfranzösische Bibliothek, 6) Heilbronn 1883, 53. 31 Ibid. 32 La chanson de Roland. Genauer Abdruck der Venetianer Handschrift IV, besorgt von Eugen Kölbing, Heilbronn – Paris 1877, 20 (https://archive.org/stream/lachansonderolan00kluoft#page/20/mode/2up). 30

200

Lucia Sinisi

discrepancy led Bédier33 to interpret the form vers as an Anglo-Norman spelling of urs ‘bear’; as a consequence many other editors and the various translators who have worked on those editions have been reluctant to acknowledge the lectio of Oxford MS Digby 23, amending vers with urs, according to the emendation proposed by Bédier, or with ors, according to the lectio of C, V7 and V4. Already in 1938, Lerch had pointed out that the noun vers occurs twice in the same laisse, as a nominative singular in line 727 (vers) and as an accusative singular (ver) in line 732. The accusative singular ver is strong evidence for accepting the Oxford lectio, because it does not show an -s, which is, on the contrary, present in the word urs.34 As Lerch pointed out, and as is elucidated in the Anglo-Norman Dictionary, the word vers (‘boar’) derives from Latin verres (acc. verrem) (‘wild boar’), already occurring with a denotative suffix -attum, in the form ferrat (glossed paerfach), in a Latin– German glossary compiled in Bavaria around the ninth century, the so-called Kassel Glossary, as in other works in Anglo-Norman.35 More recently van Emden has resumed the discussion,36 demonstrating that there is no reason for replacing vers with urs, as in the emendation advanced by Bédier, and adding that in the Middle English version the animal in question is a bore, Middle English for ‘boar’ (lines 94 and 101).37 To these arguments we can now add the evidence provided by the sculpture on the façade of Barletta cathedral, as the animal in this scene is clearly a boar, the animal denoted by the Anglo-Norman word vers as cited in Oxford, MS Digby 23. The identification is strengthened if we compare the Barletta boar with the representation of a bear on the left-hand side of the same archivolt, in a scene that has not yet been identified. Scene II The second scene highlights in an even more decisive way the close connection between the text of the Chanson and the Barletta cathedral carvings. Here the sculptor has portrayed the same boar with a smaller but very vicious animal on top of it, which is biting the boar’s ear with a wide-open mouth (Fig. 2). As we can read in the Oxford text, a veltres, Anglo-Norman for ‘hound’,38 is assailing the boar, biting him on his ear. Behind the dog, the figure of a young man is leaning towards the fighting animals. Is he Roland, who, according to the scholars who have interpreted Charlemagne’s dreams, is symbolically represented by the greyhound? I cannot myself answer this question; what I can say is that there is an astonishing similarity between the representation of Charlemagne’s dream on the façade of Barletta cathedral, and a drawing included in the set of thirty-nine illuminations which decorate a parchment manuscript preserved in the University Library

Joseph Bédier, ‘Remarques sur vingt passages difficiles de la Chanson de Roland’, in Mélanges Ferdinand Lot, Paris 1925, 25–40. 34 Eugen Lerch, ‘Le «verrat» de la Chanson de Roland: Une correction inutile’, Romania 64, 1938, 398–405. 35 Lerch, ‘Le «verrat» de la Chanson de Roland’, 399; Anglo-Norman Dictionary [hereafter AND] s.v. vers3 and ver2 (AND Online edition: http://www.anglo-norman.net/D/ver[2] [accessed 18 September 2018]). 36 Wolfgang van Emden, ‘Another look at Charlemagne’s dreams in the Chanson de Roland’, French Studies 28, 1974, 257–71 at 258. 37 Sidney J. H. Herrtage, The English Charlemagne Romances II: “The Sege off Melayne” and “The romance of Duke Rowland and Sir Otuell of Spayne”, now for the first time printed from the unique ms. of R. Thornton, in the British Museum, ms. Addit. 31,042, together with a fragment of “The song of Roland”, from the unique ms. Lansd. 388, ed. Sidney John Hervon Herrtage, EETS Extra Series 35, London, 1880 (repr. 2008), 107–36. 38 AND, s.v. ‘vealtre, vautre, veautre, veltre; vialtre’ with the meaning of “hound” (AND Online edition: http://www.anglo-norman.net/D/vealtre [accessed 27 September 2018]). 33

The Chanson de Roland on the Façade of Barletta Cathedral

201

Fig. 2: Boar, dog and young man (Roland?) of Heidelberg, MS Cod. pal. germ. 112.39 This manuscript, which consists of 123 folia, contains the rewriting of the Chanson de Roland in Middle High German, the Ruolantes Liet by Pfaffe Konrad. The date of its composition, according to one of the most eminent scholars who have devoted their attention to it, Dieter Kartschoke, should be around the year 1170. The drawing, which has been made with a quill, is on fol. 41v, immediately after line 3065, that is, immediately after the description of Charlemagne’s prophetic dream about Ganelon breaking the emperor’s spear into pieces, and it is related to lines 3066–81 of the Middle High German text.40 Both the Barletta cathedral carving and the image in the Heidelberg manuscript show a sleeping man at whose feet a hairy animal is standing. Whereas the lectio of Oxford MS Digby 23 identifies this animal as a boar, according to the representation in the manuscript it is a bear, as is confirmed by the word bere in the Middle High German text at line 3069, and also in the form pere (which was the pronunciation in Upper High German dialects, where the Second Consonant Shift had taken place in its entirety), at lines 3072 and 3074. The sleeping man is more easily identified in the Heidelberg manuscript, thanks to the crown which he is wearing; moreover, whereas the sculptor of Barletta cathedral depicts only Charlemagne and the animal at his feet, in the Heidelberg drawing the emperor’s sleep is guarded by his paladins. In the background another animal is depicted, which is on the verge of biting the hand or the arm of one of the guards (is it perhaps the hound that is ready to defend Charlemagne against the beast that symbolizes the traitor Ganelon?).41 39

The drawing described here can be seen online (https://digi.ub.uni-heidelberg.de/diglit/cpg112/0084). Das Rolandslied des Pfaffen Konrad, ed. Dieter Kartschoke, Frankfurt am Main – Hamburg 1970 (repr. 1993), 157. 41 On the quilt drawings in MS Heidelberg, pal-germ 112 see Peter Kern, who assumes that the illustrative cycle which characterises the manuscript was commissioned specifically for Konrad’s text (Peter Kern, ‘Bildprogramm und Text. Zur Illustration des Rolandsliedes in der Heidelberger Handschrift’, Zeitschrift für deutsches Altertum und deutsche Literatur 101, 1972, 244–70, at 251. 40

202

Lucia Sinisi

That Charlemagne’s dreams described in the Chanson de Roland were popular episodes from the epic poem is also underpinned by an illumination which decorates a manuscript containing the work Karl der Grosse, a rhymed version of the Ruolantes Liet, probably composed between 1230 and 1235, and attributed to a poet known by his epithet der Stricker (‘the knitter’). The work must have been very popular in its day, as is confirmed by the number of extant manuscripts which preserve it (there are more than forty, mainly in Upper High German). The parchment manuscript preserved in the library of the Abbey of Saint Gall (MS St. Gallen, Kantonsbibl., VadSlg Ms. 302) is adorned with eleven illuminations, including one, on fol. 25r, which describes the dreams of Charlemagne.42 The illumination is divided into four sections, three of which depict Charlemagne’s dreams. In the first section Charlemagne is sleeping on a lavishly draped bed; in the second, the dream relating to Ganelon’s breaking the emperor’s ash-wood spear is represented; in the third one a bear mauls Charlemagne’s right arm while some of his paladins try to defend him; in the last of the four sections a dog in the foreground is biting a wild animal. As a consequence of what has been said so far, we can assume that the sculptor of the archivolts of the cathedral of Barletta, or the person who commissioned the work, knew a version of the Chanson de Roland which was close to the Oxford witness. We can also assume that the manuscript tradition of the Chanson de Roland in the south of Italy was closer to an older version of it, far removed both from later versions in Old French and Franco-Venetian, and from reworkings in other languages. Scene III The third scene shows a warrior holding a spear in his right hand and a spade or club in his left hand, faced by a heavily worn figure which, although regrettably damaged, especially in the upper part of the body, shows traits of monstrosity (Fig. 3). It may represent a Saracen – identified by his draped breeches, worn according to the Saracen style – who has just been given a mortal downward blow.43 Although the figure on the right does not seem to be threatening his opponent, that is, he is not depicted on the verge of hitting him, it is quite evident that the figure on the left is trying to defend himself, warding off a blow. The warrior armed with a spear and club by his features appears to be young and beardless, which leads us to exclude a portrait of Charlemagne who – as is visible in the first scene of the cycle – has a thick beard and more markedly adult features. I am more inclined to think that the character portrayed in this scene might be Roland, not only because of the youthful features of the represented Christian hero, but also because of the monstrous features of his opponent, which recall Ferragut, the giant ‘of the race of Goliath’,44 champion of the pagans, who dared to challenge the Franks in a single combat outside the walls of the town of Nájera. According to Lejeune, this is ‘the most popular and

42

St. Gallen, Kantonsbibliothek, Vadianische Sammlung, MS VadSlg. 302, f. II_25r – Rudolf von Ems, Weltchronik. Der Stricker, Karl der Grosse (https://www.e-codices.ch/de/list/one/vad/0302 [accessed 27 September 2018]). 43 Medieval Islamic Civilization: An Encyclopedia, ed. by Josef W. Meri, 2 vols, Routledge encyclopedias of the Middle Ages, New York – London 2006, I, 159; Jones, ‘The conventional Saracen’, 203; Laura Cocciolo and Davide Sala, Storia illustrata della moda e del costume, Verona 2001, 63. 44 Historia Karoli Magni et Rotholandi, ou Chronique du Pseudo-Turpin, ed. Cyril Meredith-Jones, Geneva 1972, 146.

The Chanson de Roland on the Façade of Barletta Cathedral

203

Fig. 3: Roland and Ferragut most often recurring of all the illustrations of the Roland legend. It is one of the most spectacular representations of the fight between Vice and Virtue. It is almost as popular as its biblical inspiration – the fight between David and Goliath.’45 The duel with the Saracen Ferragut is not taken from the Chanson de Roland, but rather from the work known as Pseudo-Turpin, that is, the Historia Karoli Magni et Rotholandi. The episode in the Pseudo-Turpin describes the single combat between Roland and Ferragut. After a long dispute about the theological fundamentals which constitute the basis of the faiths followed by the two opponents, the giant Ferragut proposes that the outcome of the exhausting duel should be decided by God. Although Roland up to this point has fought with the strength of a lion, he is weaker than the giant 45

Lejeune and Stiennon, The Legend of Roland, I, 73.

204

Lucia Sinisi

Ferragut and he will certainly meet his death in the fight, but after having invoked the grace of Jesus Christ, he manages to strike Ferragut mortally, who, like the Greek hero Achilles, has one vulnerable part of his body, in his case the navel. The giant slumps to the ground.46 The hypothesis that the two figures on the Barletta archivolt may represent Roland and Ferragut is supported by the representation of the same duel in of one of the stained-glass windows of Chartres cathedral.47 As Duncan Robertson points out: ‘The joust scene does not literally represent the Pseudo-Turpin account of their combat, which takes place mainly on foot, with swords and stones. The image conveys rather the whole of their encounter, which is a theological disputation as well as a passage of arms.’48 The scene of the duel between Roland and Ferragut is also portrayed on one of the twenty-four bronze panels which compose the left portal of the basilica of San Zeno in Verona, which, being dated to 1138, is reckoned by Lejeune and Stiennon to be the first representation known to us of the combat.49 It is the last panel on the right, coming after some scenes from the New Testament, and represents a fight between two knights, very likely Roland and the giant Ferragut. Here, in the last panel, as is narrated in the Pseudo-Turpin, Roland manages to get the better of the arrogant Saracen, driving the sharp end of his sword into Ferragut’s navel. If our conjecture about the identity of the two figures on the archivolt of the cathedral of Barletta is right, it would be the first portrayal of the warriors Roland and Ferragut. The visual representation that best describes this episode, according to the FrancoItalian version, is perhaps the one on a historiated capital located on the south façade of the palace of the dukes of Granada de Ega, built during the reign of Sancho the Great (1150–94) in the town of Estella-Lizarra, in the region of Navarre.50 It does not leave any doubt as to the identity of the portrayed characters, as on one of the lower columns that ornament the façade there is an inscription FERA GUT and ROLLAN, along with the name of the sculptor, MARTINVS DE LOGRO NIO ME FECIT. As in the sculpture on the façade of Barletta cathedral, the duel between Roland and Ferragut here takes place not on horseback, as in many other church sculptures of the time, but rather on foot. The giant Ferragut is identifiable because he has a particularly massive head and frightening features. The Historia Karoli Magni et Rotholandi was written around 1140, therefore later than the bronze doors of the basilica of San Zeno, which are dated around 1138, and much earlier than the composition of the stained windows of Chartres cathedral. This may mean that around this time this episode narrated in the Pseudo-Turpin 46

Historia Karoli Magni, 160: Ita bellum ab utroque corrobatur, et illico Rotholandus paganum aggreditur. Tunc Ferracutus eiecit ictum spata sua super Rolholaudum, sed ipse Rotholandus saltavit ad laevam, et excepit ictum spatae eius in baculo suo. Interea absciso baculo Rotholandi, irruit in eum ipse gigas, et illum arripiens leviter inclinavit eum subter se ad terram. Statim agnovit Rothotandus quod tunc nullo modo evadere poterat, et coepit invocare in auxilium beatae Mariae virginis filium, et erexit se, Deo donante, aliquantulum, et revolvit eum subter se, et misit manum suam ad mucronem eius, et punxit eius parumper umbilicum, et evasit ab eo. 47 Duncan Robertson, ‘Visual Poetics: The Charlemagne Window at Chartres’, Olifant 6, 1978, 107–17; Clark Maines, ‘The Charlemagne Window at Chartres Cathedral: New Considerations on Text and Image’, Speculum, 52, 1977, 801–23; Mary Jane Schenck, ‘The Charlemagne Window at Chartres: Visual Chronicle of a Royal Life’, Word & Image 28, 2012, 135–60; Marcel Bulteau, Description de la cathédrale de Chartres, Chartres 1850, 238; Alphonse Vétault, Charlemagne, Tours 1877, no. 1, 547; Maines, ‘The Charlemagne Window’, 814–15. 48 Robertson, ‘Visual Poetics’, 117, n 10. 49 Lejeune and Stiennon, The Legend of Roland, I, 74. 50 Lejeune and Stiennon, The Legend of Roland, I, 92.

The Chanson de Roland on the Façade of Barletta Cathedral

205

coagulated around a narrative nucleus which was to become a central episode of the L’Entrée d’Espagne, a now-lost chanson de geste in Vulgar French, which would be later re-elaborated in a Franco-Italian version of the end of the thirteenth century, or the beginning of the fourteenth, of which only one copy survives. This marvellously illuminated copy is now preserved in the Biblioteca Marciana in Venice, catalogued with signatura Francese Z 21 (257).51 Scene IV

Fig. 4 Roland on horseback blowing an oliphant In the fourth scene a knight seated in the saddle of his horse is blowing an oliphant (Fig. 4). According to the legend, the portrayed knight should be Roland, who in extremis blows his horn in order to call Charlemagne’s troops, who are on their 51

Lejeune and Stiennon, The Legend of Roland, I, 73. For the edition of Venice, Biblioteca Marciana, MS Francese Z 21 (257), see L’Entrée d’Espagne: Chanson de geste franco-italienne, publiée d’après le manuscript unique de Venice, ed. Antoine Thomas, 2 vols, Société des Anciens Textes Français, Paris 1913; see also the digital edition by Serena Modena (http://www.rialfri.eu/rialfriPHP/public/testo/testo/ codice/entree.html).

206

Lucia Sinisi

way back to France and a long way from the battlefield, for help. This is a crucial episode in the Chanson de Roland, which is fraught with the tension deriving from Roland’s hesitation to blow the oliphant, conduct which will cost the lives of his brothers-in-arms. But Roland at last does blow his oliphant, as is narrated in laisse CXXXII, lines 1753–6 of the Chanson: Rollant al mis l’olifan a sa buche, Empeint le ben, par grant vertu le sunet. Halt sunt li pui e la voiz est mult lunge, Granz xxx liwes l’oïrent il respundre.52

On the now destroyed mosaics that decorated the floor of the cathedral of Brindisi, according to the description given by Francesco Ribezzo, among various other scenes depicted there was a knight in the act of blowing an oliphant.53 The Norman cathedral of Brindisi was built between 1132 and 1140, and boasted a mosaic floor which was destroyed by two earthquakes, one in 1743, the other in 1858, and can be reconstructed only through the awkward reproductions and descriptions made by Aubin-Louis Millin and Heinrich Wilhelm Schulz.54 The cycle of the mosaic showed a solid and imposing representation of the Roland epic on a band 2.5 metres high, which bordered a series of biblical scenes in the middle.55 As Lejeune has rightly pointed out, it was in some ways akin to the Bayeux tapestry, transposing the Chanson de Roland into pictures using a material distinctive of the Italian artistic tradition: stone.56 It showed a prelate/warrior on his horse between two knights; the one on his left is standing, the other is sitting in the saddle of a horse and blowing a horn. The inscription L’ARCEVESQUE TORPIN overhangs at the centre of the scene, above the head of the prelate.57 Undoubtedly, the artist who composed the Brindisi mosaic wanted to represent the well-known dispute between Roland and Oliver narrated in the Chanson, which was settled by the decisive intervention of the bishop Turpin (in lines 1737–52). The following scene in the mosaic showed Roland, identified by the inscription ROLLANT, carrying the corpse of a fighter on his shoulder, as is narrated in the Chanson (lines 2184–94). In the third scene the body of a knight lays supine on the blade of a sword, clutching the hilt with both hands against his chest. The identity of the character is indicated by the inscription ALVIER. He is on the verge of passing away, as can be inferred from the depiction of his soul, which, abandoning his body, flies into the sky (lines 2024–30).58 In the fourth scene Roland on horseback escorts another horse on the back of which Olivier lies mortally wounded. The episode has no equivalent in the 52

La Chanson de Roland, 338. Francesco Ribezzo, ‘Lecce, Brindisi, Otranto nel ciclo creativo dell’epopea normanna e della Chanson de Roland’, Archivio storico pugliese 5, 1952, 192–215 at 197. 54 The reproductions by Aubin-Louis Millin are preserved in the National Library of France (BnF, Cab. Estampes, Coll. Millin, Gb63), and are also available online at https://gallica.bnf.fr/ark:/12148/ btv1b69358304 (thumbnail ‘View 15’; accessed September 2018). Those by Schulz were published in 1860: Heinrich Wilhelm Schulz, Denkmäler der Kunst des Mittelalters in Unteritalien, Dresden 1860, I, 302–6 and IV, Tafel XLV, Fig. II; also online at http://digi.ub.uni-heidelberg.de/diglit/schulz1860bd1/0320 (accessed September 2018). 55 Melinda Mihályi, s.v. ‘Orlando’, in Enciclopedia dell’arte medievale – Treccani, 12 vols, Rome 1997, VIII; also online: http://www.treccani.it/enciclopedia/orlando_(Enciclopedia_dell’_Arte_Medievale)/ (accessed September 2018). 56 Lejeune and Stiennon, The Legend of Roland, I, 97. 57 Ribezzo, ‘Lecce, Brindisi, Otranto’, 196–7. 58 Lejeune observes that in the Chanson de Roland Oliviero dies face downwards, whereas in the representation we are talking about here the corpse faces upwards. It is hard to establish whether this was an 53

The Chanson de Roland on the Façade of Barletta Cathedral

207

Oxford version of the Chanson de Roland, but it is narrated in the Ruolantes Liet of Pfaffe Konrad, and in two Italian versions of the Roland epic, Rotta di Roncisvalle and Viaggio in Ispagna.59 In the last mosaic he depicts the ghastly episode regarding Roland, who, although very close to death, with a well-delivered blow of his oliphant kills a Saracen who is trying to steal his sword, smashing his skull and making his eyes pop out of their orbits (lines 2284–90). The scene is matched with the inscription ROLANT. The presence of a mosaic cycle of the Roland epic on the floor of the cathedral of Brindisi shows not only that Roland’s legend was very well known in southern Italy (but this is a matter which is by now unquestionably established),60 but also how the artist created his masterpiece following a version of the Chanson de Roland which was very close to the Oxford text, with certain variations, particularly in the order of the scenes. Could this be, as Pio Rajna supposes, an Italian, or rather a Franco-Italian, version made for pilgrims to Italy? This is the question that Lejeune poses, adding – consequently – ‘one must not forget that Brindisi was in the Norman kingdom of Sicily, where the court language in the second half of the twelfth century was Norman French’.61 Starting from the scenes of the fights between Christians and Saracens on the lintel of the first portal on the right-hand side of the western façade of SaintPierre cathedral in Angoulême, which are dated about 1120–40, the churches with sculpted representations of relevant episodes of the Chanson in Italy are manifold – and for this analysis Rita Lejeune and Jacques Stiennon’s work is still fundamental – but from now on we can add to the list the archivolts of the lesser-known cathedral of Barletta. Like the sculptor of the Angoulême lintel, the Barletta artist worked not according to his imagination, but rather followed and depicted a precise text, a renowned text, which narrates the geste of the paladins of Roncevaux. This text, which must have been almost faithful to the Anglo-Norman version in Oxford MS Digby 23, with some episodes that had gradually built up and coagulated around the nucleus of the story, was well known to the sculptor, or at least to the person who commissioned the work, whose intention was to transpose into stone, literally, the message of the Chanson, with the aim of arousing amazement and admiration among the faithful, to whom – we can assume – the story was also well known. Unfortunately – as far as I know – there is no evidence for the transmission and circulation of manuscripts containing such a chanson de geste in Apulia, except for a deed of donation in which two brothers, Johannes and Petrus, sons of Risando from Corato, a town only a few miles from Barletta, donate duos libros gestarum to the church of St Mary the Virgin in the same town of Corato.62 The editor of the document, Giovanni Beltrani, has interpreted the two books as two books of chronicles, whereas I would attribute to the word gestarum the meaning of chanson de geste.

original trait created by the artist or whether it was based on a version of the Chanson de Roland which no longer exists. (Lejeune and Stiennon, The Legend of Roland, I, 99). 59 Lejeune and Stiennon, The Legend of Roland, I, 99 and 102, n. 4. 60 Ettore Li Gotti, Sopravvivenza della leggenda carolingia in Sicilia, Biblioteca del Centro di Studi filologici e linguistici siciliani, 9, Palermo 1956, 8; Giovanni Palumbo, La Chanson de Roland in Italia nel Medioevo, Rome 2013. 61 Lejeune and Stiennon, The Legend of Roland, I, 100. 62 Codice Diplomatico Barese IX. I documenti storici di Corato (1046–1327), ed. Giovanni Beltrani, Bari 1923, no. 14.

208

Lucia Sinisi

Why did the Normans want to portray scenes of the Chanson de Roland on the façade of Barletta cathedral? We must not forget that at the beginning of their expansion in southern Italy the Church of Rome was somewhat distrustful towards the invaders. We can assume that for this reason the Normans were interested in legitimizing their presence in the eyes of the pope, promoting themselves as the champions of the struggle against the infidels. What better propaganda for them than being considered the heirs of the Roncevaux paladins?

‘THE JEWS ARE OUR DONKEYS’: ANTI-JEWISH POLEMIC IN TWELFTH-CENTURY FRENCH VERNACULAR EXEGESIS* Linda M. A. Stone Taken from a twelfth-century Psalms commentary, compiled for Laurette d’Alsace (d. c. 1175) and written in vernacular French, the reference to the Jews as donkeys (nos asnes) appears in the commentary’s exegetical interpretation of Psalm 40:14’s ‘Blessed be the Lord God of Israel’.1 In this explanation, the Jews are defined as the pack animals bearing the weight of Scripture, i.e. the Hebrew Bible, the prophetic contents of which justify Christianity’s veracity and supersession of Judaism. This interpretation of the Jews, albeit with some key differences, has its roots in the use by St Augustine (d. 430) of this particular verse and that of Psalm 58:12, with its injunction to ‘slay them not’,2 to develop his concept of Jewish service to Christianity, his testimonium veritatis, Jewish witness to the truth of the Christian faith.3 This paper discusses St Augustine’s construction of that concept, together with its impact on subsequent Christian–Jewish relations, and the various ways in which the concept was disseminated into twelfth-century Christian society, in particular via the vernacular tongue, and thus into literate lay society. The discussion focuses on three particular vernacular sources: first a selection of Psalms taken from Laurette’s Psalms commentary, including Psalm 40; second, the twelfth-century French vernacular religious poem Eructavit, based on Psalm 44 and dedicated to Marie de Champagne (d. 1198); and lastly the twelfth-century vernacular versions of the Latin sermons on the Song of Songs originally composed by Bernard of Clairvaux (d. 1153), a translation that is believed to have been for the purposes of a lay audience.4 St Augustine’s theory of Jewish service was no longer the purview only of learned Latin clerics; by the twelfth century it was also being disseminated to a literate lay society, both men and women. *

I would like to thank Professor Elisabeth van Houts for the opportunity to present this paper at the Battle Conference. My thanks, too, to Dr James Kane, University of Sydney, for his help with the vernacular translations. 1 Biblical references are taken from the Douai-Rheims edition of the Bible and, where necessary, from the Vulgate. Psalm 40:14 – ‘Blessed be the Lord the God of Israel from eternity to eternity. So be it. So be it’ (Vulgate, Benedictus Dominus, Deus Israel, a saeculo, et usque in saeculum. Fiat, fiat); The Twelfth-Century Psalter Commentary in French for Laurette d’Alsace (an Edition of Psalms I-L), ed. Stewart Gregory, 2 vols, London 1990, II, 443, line 371. 2 Psalm 58:12 – Douai-Rheims ‘God shall let me see over my enemies: slay them not, lest at any time my people forget. Scatter them by thy power; and bring them down, O Lord, my protector’ (Vulgate, Deus ostendet mihi super inimicos meos; ne occidas eos, nequando obliviscantur populi mei. Disperge illos in virtute tua, et depone eos, protector meus, Domine). 3 Anna Sapir Abulafia, Christian-Jewish Relations 1000–1300: Jews in the Service of Medieval Christendom, London 2011, 4–8. 4 Twelfth-Century Psalter Commentary; Eructavit: An Old French Metrical Paraphrase of Psalm XLIV Published from all the Known Manuscripts and Attributed to Adam de Perseigne, ed. T. Atkinson Jenkins, Dresden 1909; La traduction en prose française du 12e siècle des Sermones in Cantica de Saint Bernard, ed. Stewart Gregory, Amsterdam 1994, xiii.

210

Linda M. A. Stone

Much has been written on the growing use, in the twelfth century, of vernacular French and the importance of women in its development. Ian Short has extensively discussed the growth and spread of Anglo-Norman as a language and its related literature, while Susan Groag-Bell and Jocelyn Wogan-Browne have both written ground-breaking studies regarding medieval women’s book ownership and their roles as patrons of vernacular writings.5 The appearance and implications of anti-Jewish terminology in Anglo-Norman literary and religious texts have been discussed by Maureen Bolton, while Thelma Fenster has commented on the use of such vernacular texts by women, either as patrons or as readers, and the dissemination of anti-Jewish sentiments through their child-rearing and educative roles.6 The purpose of this discussion is therefore to extend this field of study by focusing on the way in which St Augustine’s concept of Jewish service appears in such vernacular texts, and further to consider how its presence in these texts can serve to reinforce his idea of Jews as servants of Christianity into twelfth-century society.7 St Augustine’s purpose in the construction of his testimonium veritatis was an attempt to reconcile what was, for Christianity, the ongoing paradox of the Jews. He needed first to resolve the conundrum of why the Jews continued to remain unpunished for their alleged involvement in the crucifixion of Christ. He then needed to explain why the Jews continued to exist; with the coming of the Messiah and thus the fulfilment of the Hebrew Bible’s prophecies, they should long ago have converted willingly to Christianity, but they continued to show no such inclination. St Augustine also needed to consider how these puzzles reconciled with Apostle Paul’s various prophecies, contained in his letter to the Romans, regarding the Jews’ eventual conversion to Christianity, albeit at the end of time, an action that would open the door for the final, world, that is Christian, salvation. There was, however, a further riddle, in that Paul’s prophecy in Romans 9:27 initially referred to a remnant of Israel being saved, while later, in Romans 11:26, he prophesied that all Israel would be saved.8 St Augustine, in his exposition on Psalm 40:14, used Genesis 25:23, relating to the birth of Esau and Jacob, and its reference to the elder serving the younger.9 He interpreted the psalm verse to show that the Jews, the elder, were like book-slaves 5

E.g. Ian Short, ‘Patrons and Polyglots: French Literature in Twelfth-Century England’, ANS 14, 1991, 229–49; Susan Groag-Bell, ‘Medieval Women Book Owners: Arbiters of Lay Piety and Ambassadors of Culture’, Signs: Journal of Women in Culture and Society 7, 1982, 742–67; Jocelyn Wogan-Browne, ‘“Our Steward, St Jerome”: Theology and the Anglo-Norman Household’, in Household, Women and Christianities in Late Antiquity and the Middle Ages, ed. Anneke B. Mulder-Bakker and Jocelyn WoganBrowne, Turnhout 2005, 133–59. 6 Maureen Bolton, ‘Anti-Jewish Attitudes in Twelfth-Century French Literature’, in Jews and Christians in Twelfth-Century Europe, ed. Michael A. Signer and John Van Engen, Notre Dame 2001, 234–54; Maureen Bolton, ‘Anti-Jewish Attitudes in Anglo-Norman Religious Texts’, in Christian Attitudes towards the Jews in the Middle Ages, a Casebook, ed. Michael Frassetto, London 2007, 151–65; Thelma Fenster, ‘English Women and their French Books: Teaching about the Jews in Medieval England’, in The French of Medieval England: Essays in Honour of Jocelyn Wogan-Browne, ed. Thelma Fenster and Carolyn P. Collette, Cambridge 2017, 175–89. 7 Anna Sapir Abulafia, ‘The Service of Jews in Christian-Jewish Disputations’, in Les dialogues adversos iudaeos: Permanences et mutations d’une tradition polémique, ed. Sébastien Morlet, Olivier Munnich and Bernard Pouderon, Paris, 2013, 339–49 at 339. 8 Romans 9:27 – Douai-Rheims, ‘If the number of the children of Israel be as the sand of the sea, a remnant shall be saved’ (Vulgate, … Si fuerit numerus filiorum Israel tamquam arena maris, reliquiae salvae fient); Romans 11:26 – ‘And so all Israel should be saved, as it is written …’ (Vulgate, et sic omnis Israel salvus fieret, sicut scriptum est …). 9 Genesis 25:23 – Douai-Rheims, ‘one people shall overcome the other, and the elder shall serve the younger’ (Vulgate, populusque populum superabit, et maior serviet minori).



Anti-Jewish Polemic

211

(capsarii) serving Christianity, the younger, using as his premise the idea that the books of the Hebrew Bible would provide evidence to pagans of the prophesied truth of Christianity. Moreover, such evidence would be all the more convincing since the books were those of the Jews, the enemy of Christianity.10 In his interpretation of Psalm 58:12’s exhortations to ‘slay them not’, and ‘scatter them by your power’, St Augustine alluded again to the Jews’ role as book-carriers, stating that the Jews were not to be killed, since, as they had the books which prophesied Christ, they were essential for others to believe in Christianity. Instead, as a result of their wickedness the Jews were to be dispersed among the nations, serving as witnesses to Christianity’s truth.11 St Augustine brought these exegetical interpretations of Psalms 40:14 and 58:12 together in his work City of God. First emphasizing God’s original love for the Israelites, St Augustine went on to declare that this love had been forfeited, initially because the Israelites had turned to the worship of strange gods, and eventually because of their sin in putting Christ to death.12 He acknowledged that some Jews had already converted – the Pauline ‘remnant’. He then explained that the punishment for the rest would be to be dispersed among the nations of the world, where they would serve not only as a warning to pagans, but also, via their role as bookcarriers, as proof against any challenges to the veracity and supersession of Christianity over Judaism. Such a role would be theirs until their Pauline-prophesied conversion to Christianity, thus bringing about eventual final salvation. Accordingly, St Augustine argued that the Jews, in providing such a service, and despite being Christianity’s enemies, should be neither persecuted nor compelled to convert, but instead should be tolerated.13 The casting of this hope for the Jews’ conversion as anti-Jewish polemic is based on the view that its purpose was in fact for the benefit of Christian salvation, rather than for the Jews per se; it was thus a hope that only served to emphasize the Jews’ subservient position to Christianity. St Augustine’s concept of Jewish service and its required toleration of the Jews was enshrined in the sixth century by Gregory the Great (d. 604) in his papal bull Sicut iudeis, in which he emphasized that Jews were to be allowed to continue their own manner of worship, without fear of punishment, a ruling which would be re-confirmed by subsequent popes. Gregory’s premise for Christianity’s toleration of the Jews might appear to have been defined by Roman law, since the bull had been issued in relation to an earlier letter (ad 598) chastising the bishop of Palermo for seizing synagogues and turning them into churches – Roman law protected existing synagogues. However, underlying Gregory’s actions, as Anna Abulafia has suggested, is the more likely Augustinian premise of toleration of the Jews, defined not by the tenets of the law but by the tenets of the Bible.14 The problem, nevertheless, was that such toleration necessarily inhibited or limited any pressure on the Jews to convert.15 With the exception of the attacks against the Jews that took place in Visigothic Spain, this biblical and institutional concept of toleration appeared to hold, with

10

Augustine, Enarrationes in Psalmos, ed. E. Dekkers and J. Fraipont, Turnhout 1956, 459. Augustine, Enarrationes, 745–6. 12 Augustine, De Civitate Dei, ed. B. Dombard and A. Kalb, Turnhout 1955, 127. 13 Augustine, De Civitate Dei, 644–5. 14 Abulafia, Christian-Jewish Relations, 20–1. 15 David Berger, ‘Mission to the Jews and Jewish-Christian Contacts in the Polemical Literature of the High Middle Ages’, The American Historical Review 91, 1986, 576–91 at 576. 11

212

Linda M. A. Stone

the relationship between Jews and Christians remaining relatively peaceful.16 By the twelfth century, however, a change in sentiment had begun. Incidents of Christian violence against the Jews started to increase, for example the virulence of the attacks that were witnessed at the outset of the First Crusade (1096), attacks which the church, in line with St Augustine’s premise, roundly condemned.17 Various factors were bringing about this societal shift in sentiment towards the Jews. The church and society in general were undergoing a period of significant and unsettling change. The economy was beginning to flourish as it moved from a mainly agricultural to an increasingly urban focus, a shift with a mix of political, social and financial implications. With its attendant growth in credit and usury, this development presented the church with a persistent source of anxiety.18 Worrying too for the church was the rise in heretical movements, a reflection that the earlier attempts to purify the church had lost impetus.19 Furthermore, the spiritual focus for Christians was changing, with more emphasis being placed on Christ’s humanity, resulting in his agony on the cross becoming more personal.20 Such societal changes did not augur well for the Jews, especially as twelfthcentury Christian society was faced with the increasing presence of ‘real’ Jews, often living close to their Christian neighbours.21 They were present and active within Christian society, active in that flourishing economy, and often involved in the church-despised business of money-lending, driven deeper into it by the church’s condemnation of Christian involvement in such activity.22 They were also challenging Christianity’s core beliefs, since there was also a Jewish intellectual ‘twelfth-century renaissance’ with Jewish scholars producing commentaries refuting Christianity’s interpretation of the Hebrew Bible.23 Moreover, while there had been some converts from Judaism to Christianity, the most prominent being Peter Alfonsi (d. c. 1140), the vast majority of Jews continued to remain true to their own faith.24 There were therefore no signs of the fulfilment of Apostle Paul’s prophecy that the Jews would convert en masse; in fact, many of those Jews in the Rhineland who had converted, admittedly by force, at the outset of the First Crusade, had reverted to Judaism.25 The Jews’ resistance to conversion was thus delaying the hopes of Christian final salvation. The result was that, for the church, the Jews were increasingly perceived to be outside of Christian society, having more in common with troublesome heretics.26 And yet, the core conundrum remained: in accordance with St Augustine and Gregory the Great, the Jews could neither be destroyed, nor compelled to convert. They had to be tolerated even though they remained intractable to Christianity. Abulafia, Christian-Jewish relations, 29. Robert Chazan, Daggers of Faith: Thirteenth-Century Christian Missionising and Jewish Response, Berkeley 1989, 13. 18 Lester K. Little, Religious Poverty and the Profit Economy in Medieval Europe, Ithaca 1978, 19–41. 19 Anna Sapir Abulafia, Christians and Jews in the Twelfth-Century Renaissance, London 1995, 55. 20 Rachel Fulton, From Judgement to Passion: Devotion to Christ and the Virgin Mary, 800–1200, New York 2002, 142–6. 21 Abulafia, Christian-Jewish relations, 48. 22 Robert Chazan, Medieval Stereotypes and Modern Antisemitism, Berkeley 1997, 23. 23 Abulafia, Twelfth-Century Renaissance, 69. 24 Abulafia, Twelfth-century Renaissance, 13. 25 Jeremy Cohen, ‘A 1096 Complex? Constructing the First Crusade in Jewish historical memory, medieval and modern’, in Jews and Christians in Twelfth-Century Europe, ed. Michael A. Signer and John van Engen, Notre Dame 2001, 9–26 at 10. 26 Abulafia, Twelfth-century Renaissance, 123. 16 17



Anti-Jewish Polemic

213

St Augustine’s concept of Jewish service had continued to be disseminated through later biblical exegesis. Cassiodorus (d. 585), writing in the sixth century, alluded to St Augustine in his exposition of Psalm 58:12, with his references to the Jews’ eventual conversion and role as witness to the truth of Christianity.27 Later, Remigius of Auxerrre (d. 908) expressed similar sentiments in his exposition on the same Psalm.28 In the twelfth century, however, a key channel for the dissemination of St Augustine’s concept regarding Jewish service was to be found in the contemporary glosses on the Psalms, glosses which would form part of what ultimately became known as the Glossa Ordinaria, a generic phrase covering all the glossed books of the Bible.29 These various glossed books, estimated to be extant in over 2,000 manuscripts, have been described by Christopher de Hamel as twelfthcentury bestsellers.30 They formed essential texts and reference books for scholars throughout the twelfth century and are considered to have been the standard tool for medieval exegetical teaching.31 The instigation of the pedagogical use of these various marginal and interlinear glosses, gathered from a plethora of exegetical writings from earlier centuries and incorporated into the teaching of exegesis, initially took place under the aegis of Master Anselm (d. 1117) at his renowned school at Laon.32 Many of his students would eventually become influential within the church and the royal courts of Europe. For example, prominent churchmen and courtiers educated at Laon and active within the Anglo-Norman realm included Bishop Alexander of Lincoln (d. 1148), his brother, Nigel (d. 1169), treasurer to Henry I and subsequently bishop of Ely, as well as Hugh, abbot of Reading and archbishop of Rouen (d. 1164), and Robert of Bethune, bishop of Hereford (1131–48).33 The parva gloss, the first set of twelfth-century glosses on the Psalms, is generally (although not universally) thought to have been compiled by Anselm of Laon himself.34 It was therefore his choice to use St Augustine, together with Cassiodorus, as the prime patristic sources for the glosses, and it was St Augustine’s concept of Jewish service that predominated in the glosses on the relevant verses of Psalms 40 and 58.35 It was thus St Augustine’s view of the Jews as servants to Christianity that percolated into the parva gloss and thence into twelfth-century Christian scholarly society. Anselm, using a mix of interlinear and marginal comments, incorporated St Augustine’s wording, albeit somewhat abbreviated, into his glosses. In Psalm 40:14, Anselm first stated that the reference to Israel was a confirmation of Christianity’s supersession of Judaism. He then confirmed the role of Jews as capsarii, whose 27

Cassiodorus, Expositio Psalmorum, ed. M. Adriaen, Turnhout 1958, 525–6. Remigius of Auxerre, Enarrationes in Psalmos, Bibliothèque municipale de Reims MS 132, fols 93r–93v. 29 Lesley Smith, The Glossa Ordinaria: The Making of a Medieval Bible Commentary, Leiden 2009, 1. 30 Mark Zier, ‘The Development of the Glossa Ordinaria to the Bible in the Thirteenth Century: The Evidence from the Bibliothèque Nationale, Paris’, in La Bibbia del XIII secolo: Storia del testo, storio dell’esegesi, ed. Giuseppe Cremascoli and Francesco Santi, Florence 2004, 155–84 at 156, n. 5; C. F. R. de Hamel, Glossed Books of the Bible and the Paris Book Trade, Cambridge 1984, 9. 31 Jenny Swanson, ‘The Glossa Ordinaria’ in The Medieval Theologians, ed. G. R. Evans, Oxford 2001, 156–67 at 166–7; Biblia latina cum glossa ordinaria: Facsimile Reprint of the editio princeps Adolph Rusch of Strassburg 1480/81, ed. Karlfried Froehlich and Margaret T. Gibson, 4 vols, Turnhout 1992, I, xi. 32 Smith, Glossa Ordinaria, 23. 33 C. Warren Hollister, ‘Anglo-Norman Political Culture and the Twelfth-Century Renaissance’, in Anglo-Norman Political Culture and the 12th-Century Renaissance, ed. C. Warren-Hollister, Woodbridge 1997, 1–16 at 10–11. 34 Smith, Glossa Ordinaria, 21. 35 Smith, Glossa Ordinaria, 48. 28

214

Linda M. A. Stone

testimony, as enemies, would serve to obviate any doubts as to Christianity’s truth.36 His glosses on Psalm 58:12 emphasized that the Jews must not be killed, in light of their role as witnesses to Christianity’s truth, including a further reference to the hope of eventual Jewish conversion, an expanded reference plucked from Cassiodorus’s Expositio.37 Two subsequent psalm glosses were written in the first half of the twelfth century, both of which made extensive use of the contents of the parva gloss. There are the media gloss of Gilbert of Poitiers (d. 1148), and the magna gloss of Peter Lombard (d. 1160).38 In their commentaries on Psalms 40 and 58, both scholars incorporated St Augustine’s concept of Jewish service, albeit refracted through the lens of Anselm’s parva gloss, emphasizing that the Jews, even though they were Christianity’s enemies, should not be killed since their role, as capsarii, was to serve as witness to Christianity.39 The extant parva glosses have been described as ‘numerous’ by Patricia Stirnemann.40 When these are combined with the number of extant media and magna glosses (in the case of Gilbert, some seventy-five manuscripts, while the Lombard’s magna gloss is extant in over 350), it is clear that the opportunity to reinforce the view of the Jews as Christianity’s servants to a large number of learned Latin clerics influential within Christian society was significant.41 But what of the concept’s dissemination into lay society, and not just to men, but also to women? The Psalms were not only vital within the liturgy of the Christian church, they were also important pedagogically and were often known by heart by clerics and laity alike.42 It should also be noted that they were important within the arena of twelfth-century Christian–Jewish relations, too. The renowned rabbi Rashi (d. 1105) frequently refuted the Christian interpretation of the Psalms in his own commentary.43 Moreover, the Psalter was not only limited to male clerics and male laity; women knew them too. Dhuoda (d. 843), the ninth-century Frankish writer of the Liber manualis, for example, frequently quoted the Psalms. She knew them by heart 36

Parva gloss, Troyes, MAT, MS 511, fol. 53r, marginal, col (b): Iudei sunt capsarii nostri qui nobis codices portant. Nos Israel. Aliter putassent pagani ficta quod dicuntur de Christo et ecclesia. Sed vincuntur testimonio inimicorum. 37 Parva gloss, Troyes, MS 511, fol. 75r, marginal, col (b): Ne occidas. Hoc est quod ostendit. Precatur ne iudei funditus permeant dispersi sunt; ut ad conversionem provocentur. Vel ut ecclesia ab inimicis testimonium veteris legis habeat; Cassiodorus, Expositio, 526. 38 Smith, Glossa Ordinaria, 76–7. 39 Psalm 40:14, Media gloss, Troyes, MAT, MS 488, fol. 48r, col (a); Magna gloss, Troyes, MAT, MS 543, vol. 1, fol. 96r. Psalm 58:12, Media gloss, MS 488, fol. 68r, col (b); Magna gloss, MS 543, vol. 1, fol. 133r, col (b). 40 Patricia Stirnemann, ‘Où ont été fabriqués les livres de la glose ordinaire dans la première moitié du XIIe siècle?’, in Le XIIe siècle: Mutations et renouveau en France dans la première moitié du XIIe siècle, ed. Françoise Gasparri, Paris 1994, 257–85 at 266–7. 41 Theresa Gross-Diaz, The Psalms Commentary of Gilbert of Poitiers: From lectio divina to the Lecture Room, Leiden 1996, 160–80; Patricia Stoppacci, ‘Le Glossae continuae in psalmos di Pietro Lombardo. Status quaestionis: Studi pregressi e prospettive di ricerca’, in Pietro Lombardo: Atti del XLIII convegno storico internazionale, Todi, 8–10 ottobre 2006, Spoleto 2007, 289–331 at 299. 42 John Harper, The Forms and Orders of Western Liturgy from the Tenth to the Eighteenth Century: A Historical Introduction and Guide for Students and Musicians, Oxford 1991, 67, 70; Pierre Riché, ‘Le rôle de la mémoire dans l’enseignement médiéval’, in Jeux de mémoire: Aspects de la mnémotechnie médiévale, ed. Bruno Roy and Paul Zumthor, Montreal 1985, 133–48 at 136–7. 43 Robert A. Harris, ‘Rashi and the “Messianic” Psalms’, in Birkat Shalom: Studies in the Bible, Ancient Near Eastern Literature and Postbiblical Judaism Presented to Shalom M. Paul on the Occasion of his Seventieth Birthday, 2 vols, ed. Chaim Cohen, Victor Avigdor Hurowitz, Avi Hurvitz, Yochanan Muffs, Baruch J. Schwartz and Jeffrey H. Tigay, Winona Lake 2008, II, 845–62 at 852–3.



Anti-Jewish Polemic

215

and recommended them as a spiritual teaching aid to her son, while, in the twelfth century, Christina of Markyate (d. 1161) is portrayed as irritating the devil as she read and sang the Psalms by day and night, her Psalter open for use at all hours.44 Given the ubiquity of the Psalter, it is unsurprising, as Ian Short commented, that the twelfth-century rise in vernacular French translations of the Psalms would play a pivotal role in breaking the monopoly of Latin.45 Geoff Rector further commented that such Psalter translations were the ‘single most comprehensive body’ of vernacular literature prior to the later explosion of narrative romance.46 Tony Hunt also commented on the fact that more than half of the extant twelfth-century French vernacular manuscripts can be linked to English Benedictine houses, and that a half of those texts are Psalters, some of which can be linked to women’s religious houses.47 Prime examples of such Psalter translations include the Oxford Psalter, dated to first half of the twelfth century, and the Eadwine Psalter, datable to the 1150s, which also contains the Latin parva gloss, and thus the relevant glosses on Psalms 40 and 58, referring to the Jews’ role as book-carriers.48 Among this glut of Psalter translations was a vernacular French Psalms commentary, linguistically linked to the Walloon dialect, which was prepared for Laurette d’Alsace. The commentary is extant in a number of manuscripts, dated from the mid-twelfth century to the late thirteenth century, and for the most part written in insular script. The majority of the manuscripts, however, are not complete copies of the commentary; the only complete copy is that held at Durham, believed to have been written for Bishop Hugh de Puiset (d. 1195).49 Stewart Gregory’s in-depth analysis of these various extant manuscripts and their content led him to divide the commentary into four parts, the commentary on the first fifty psalms having been written by one individual, who was possibly also the author of the commentary on Psalms 68–100. For Psalms 51–67, Gregory posited that the original commentary text may have been lost and then rewritten by another, while that on Psalms 100–150 was written later by a third individual.50 Gregory set the writing of the first block of commentary at some time around 1163–4, and his analysis of the text led him to assert that the underlying source for the commentary was based on, although not necessarily a literal translation of, the media gloss of Gilbert of Poitiers.51 A similar date of 1165–6 has been applied to the commentary on Psalms 68–100.52 The rewritten commentary on Psalms 50–67 was dated to between 1175 and 1185, while the last block, Psalms 100–150, appear to have Riché, ‘Rôle de la mémoire’, 137; Dhuoda, Manuel pour mon fils, trans. Bernard de Vregille and Claude Mondésert, ed. Pierre Riché, Paris 1975, 369; The Life of Christina of Markyate: A Twelfth-Century Recluse, trans. and ed. C. H. Talbot, Oxford 1959, 98. 45 The Oxford Psalter (Bodleian MS Douce 320), ed. Ian Short, Oxford 2015, 1. 46 Geoff Rector, ‘An Illustrious Vernacular: The Psalter en romanz in Twelfth-Century England’ in Language and Culture in Medieval Britain: The French of England c. 1100–c. 1500, ed. Jocelyn WoganBrowne, York 2009, 198–206 at 199. 47 Tony Hunt, ‘The Anglo-Norman Book’, in The Cambridge History of the Book in Britain, ed. Nigel Morgan and Rodney M. Thomson, 6 vols, Cambridge 2008, II, 367–80 at 369. 48 Oxford Psalter, 4–5; Margaret Gibson, ‘Conclusions: The Eadwine Psalter in context’, in The Eadwine Psalter, ed. Margaret Gibson, T. A. Heslop and Richard W. Pfaff, London 1992, 208–13 at 209; Margaret Gibson, ‘The Latin apparatus’, in The Eadwine Psalter, 108–22 at 108; Cambridge, Trinity College, MS R.17.1, fol. 73v, marginal col (b), fol. 102v, marginal col (b). 49 Twelfth-Century Psalter Commentary, I, 22-, 1–5, 1. I am grateful to Dr Janet Gunning at Durham Cathedral Library for arranging my access to the Durham manuscripts. 50 Twelfth-Century Psalter Commentary, I, 10–18. 51 Twelfth-Century Psalter Commentary, I, 11–12, 18–19. 52 Twelfth-Century Psalter Commentary, I, 19–20. 44

216

Linda M. A. Stone

been composed between 1191 and 1198. These latter two blocks have a closer, but not a literal, link to the magna gloss of Peter Lombard.53 Of note are Gregory’s comments, based on various timely references within the text, that the compiler of the first section of commentary, and thus relevant for the reference in Psalm 40:14 to the Jews as ‘donkeys’, may well have been one of Anselm’s students at Laon in the early part of the twelfth century.54 Gregory argued that this particular compiler may also have been Laurette’s confessor, whose purpose in writing the commentary was to provide a reassurance and comfort to her in her retirement to the convent of Forest-lez-Bruxelles, an event which took place in 1163, subsequent to the annulment of her marriage to Henry of Namur (d. 1196).55 The commentary’s tone, according to Gregory, implies that its recipient had not yet fully accepted the finality of a religious vocation, a view that might fit well with Laura Napran’s discussion of Laurette’s attempts to overcome the sentence of excommunication placed upon her following her flight from her marriage, her fourth, to Henry.56 It is not clear whether the commentary was written at Laurette’s behest or for her benefit. What is clear is that the compiler chose to write in the vernacular for his female recipient, and to include in his commentary on Psalm 40:14 St Augustine’s concept of Jewish service to Christianity, but with one key change. He had followed the media gloss and its emphasis on the psalm verse’s reference to Israel being interpreted as Christianity (Jacob, the younger) being served by Judaism (Esau, the elder). He then incorporated the media gloss’s further emphasis on the Jews’ importance, since, although enemies of Christianity, they were to be the younger faith’s servant, because their books stood as witness to any challenges to the truth of Christianity.57 The compiler’s significant alteration to the underlying media gloss, however, was that the Jews were no longer designated as capsarii; instead they were interpreted as donkeys, beasts of burden, who carried those truth-testifying books. Furthermore, veering away from the text of the media gloss, the compiler stated that while these donkeys, through their law, were also carrying the bread and wine of the Christian sacraments, they were unable to taste them, a comment that would seem to be the compiler’s own, since it does not appear to be present in the earlier relevant glosses, or in the usual patristic sources.58 What might have brought about the compiler’s choice to deviate from his underlying source and to portray the Jews as animals rather than humans, albeit still in a role as servants of Christianity? The decision may have had its roots in the late eleventh century, when a sea-change was occurring in Christianity’s perception of Jews. Christian theologians were beginning to strive to justify the concept of God becoming man, and they began, as discussed by Anna Abulafia, to employ the classical use of reason to demonstrate how such reason could justify the truth of Christianity and thus the falseness of Judaism. The result was that, since it was reason that separated man from animal, the Jews’ unwillingness to accept Christianity’s rational

53

Twelfth-Century Psalter Commentary, I, 22, 17. Twelfth-Century Psalter Commentary, I, 13. 55 Twelfth-Century Psalter Commentary, I, 18–19, 23–4. 56 Laura Napran, ‘Marriage and Excommunication: The Comital House of Flanders’, in Exile in the Middle Ages: Selected Proceedings from the International Medieval Congress, University of Leeds, 8–11 July 2002, ed. Laura Napran and Elisabeth van Houts, Turnhout 2004, 69–79 at 71. 57 Media gloss, Troyes, MS 488, fol. 48r, col (a); Twelfth-Century Psalter Commentary, II, 443, lines 367–71. 58 Twelfth-Century Psalter Commentary, II, 443, lines 371–6. 54



Anti-Jewish Polemic

217

arguments for its truth led to their being perceived as lacking in reason and thus less than human.59 One of the earliest twelfth-century proponents of the idea that Jews lacked reason was Odo of Cambrai (also known as Odo of Tournai; d. 1113). In his disputational discourse with a Jew regarding the veracity of the Virgin Birth, Odo emphasized the Jews’ unwillingness to accept such truth as a sign of their lack of spirituality and thus their lack of reason, since it was reason that opened up the Christian spiritual world.60 Similarly, in his work Against the inveterate obduracy of the Jews, Peter the Venerable (d. 1156) railed against the Jews for their refutation of the Virgin Birth and their refusal to accept Christian rational arguments, a refusal which in his view rendered them animals.61 Such Christian exegetical association of Jews with animals had not necessarily been uncommon in earlier exegesis, but it was gaining significant momentum in the twelfth century.62 By casting the Jews as donkeys, the compiler of Laurette’s commentary was therefore reflecting the contemporary Christian mindset. The Jews, with their perceived inability to accept Christianity, were no longer its human book-carriers, its capsarii, but, lacking reason, were instead consigned to be beasts of burden, donkeys, who carried the Christian truth. Furthermore, such references were not an unusual feature in Laurette’s commentary, not limited only to Psalm 40:14 and its connection to St Augustine. In the latter part of the commentary on Psalm 101:19, for example, the compiler described the Jews once again as donkeys and beseeched la bele seor not to be like those donkeys, the Jews, who were accused of madness (la folie), arrogance (hautesce) and hard-heartedness (durte).63 However, Laurette’s Psalms commentary also links the Jews with less prosaic animals. Psalm 21 is generally interpreted by Christianity to represent Christ suffering on the cross, and in verse 22 the interlocutor (i.e. Christ, from the Christian perspective) begs to be saved from the horn of the unicorn, a verse that would have been particularly significant in emphasizing Christ’s agony, since it would have resonated with the increasing tendency within twelfth-century Christian society to identify vicariously with his suffering.64 But, who was inflicting this suffering? In Laurette’s commentary, in the exegesis on this verse, it is the Jews, who at verse 22 are clearly defined as unicorns. Here, once again, the compiler used the media gloss, since Gilbert of Poitiers had commented that the unicorn inflicting the suffering represented the Jews. Gilbert embellished his comment with further references, drawn from Cassiodorus’s exposition on Psalm 28:6, to expand this idea. The animal’s one horn symbolized Judaism, a faith that followed only one Testament, and the representation rendered the unicorn the most proud of animals, an animal

Anna Sapir Abulafia, ‘The Intellectual and Spiritual Quest for Christ and Central Medieval Persecution of Jews’, in Religious Violence between Christians and Jews: Medieval Roots, Modern Perspectives, ed. Anna Sapir Abulafia, Basingstoke 2002, 61–85 at 62–3. 60 Abulafia, Twelfth-Century Renaissance, 108–10. Odo of Tournai, On Original Sin and A Disputation with the Jew, Leo Concerning the Advent of Christ, the Son of God, Two Theological Treatises, trans. and ed. Irven M. Resnick, Philadelphia 1994, 95–7. 61 Abulafia, ‘Intellectual and Spiritual Quest’, 63; Peter the Venerable, Adversus Iudeorum inveteratam duritiem, ed. Y. Friedman, Turnhout 1985, 57–8, 125. 62 E.g. Anthony Bale, ‘Fictions of Judaism in England before 1290’, in The Jews in Medieval Britain: Historical, Literary and Archaeological Perspectives, ed. Patricia Skinner, Woodbridge 2003, 129–44 at 142. 63 Durham, Cathedral Library, MS A.II.13, fol. 5r, col (b). 64 Psalm 21:22 – Douai-Rheims, ‘Save me from the lion’s mouth; and my lowness from the horns of the unicorns’ (Vulgate, Salva me ex ore leonis, et a cornibus unicornium humilitatem meam). 59

218

Linda M. A. Stone

who would rather die than be captured.65 The compiler of Laurette’s commentary chose to retain Gilbert’s reference to the overweening pride of the unicorn (and thus the Jews), but he also added the comment that the Jews were damned for all time (toz jors dampné) because of their refusal to believe in Jesus Christ, a reference that helped to render the Jews no longer solely biblical but also contemporary.66 These anti-Jewish polemical references in Laurette’s commentary, portraying the Jews as animals both prosaic and mythical, serve to reflect a changing and increasingly negative tone regarding the Jews’ service to Christianity. Moreover, these views were no longer restricted to learned Latin clerics. Rather, through choices made by the compilers, they were being disseminated, via the vernacular, into lay society, with the compiler ensuring that the reader would be aware of the risks of being stubborn, mad or arrogant, of being perceived to be like a proud and cruel Jew, forever instrumental in Christ’s suffering. Not only was the Jews’ service to Christianity being emphasized, but, at the same time, the use of anti-Jewish terminology averring the Jews’ lack of reason and casting them as animals was demonstrating how overweening pride could cause suffering to Christ, a suffering that was being increasingly personalized within Christian society. Laurette’s commentary, extant in a number of manuscripts, and written initially perhaps as a guide to a recalcitrant or stubborn entrant into a convent, reveals how the possibility of being perceived as a recalcitrant or stubborn Christian could lead to an individual’s being perceived as anathema, a Jew. But such ideas were not necessarily restricted to those entering the religious life. St Augustine’s concept of Jewish service was communicated beyond the Latin schoolroom, not only in vernacular religious texts related to the Psalms, but also in associated vernacular religious poetry, the audience for which, as Morgan Powell has commented, was often the secular courts.67 Such texts were often vernacular paraphrases of scriptural exegesis, of which the poem Eructavit, based on Psalm 44, which begins Eructavit cor meu (‘My heart hath uttered’), is an example. Psalm 44, with its opening dedication of a ‘canticle for the Beloved’ has often been seen as a form of epithalamium, and a psalm thus considered particularly suitable for women. The poem Eructavit, based on, but not a literal translation of, this psalm was dedicated to Marie de Champagne.68 Marie has been described as avantgarde in her taste for vernacular literature, and as an outstanding literary patron of her day, but it is not clear whether Marie commissioned the poem, or whether, like Laurette’s commentary, it was written as a guide or instruction for her.69 Its composition has been dated to the mid-1180s, with John Benton opting for pre-1187, since the poem makes no mention of the loss of Jerusalem.70 There has been some debate

65

Twelfth-Century Psalter Commentary, I, 259, lines 978–80, Unicornis est un beste ki si est orgoillose qu’ele mort bein por desdeing quant el est prise, et prise, et segnefiet les Judeus; Media gloss, Troyes, MS 488, fol. 26v, col (a): Unicornes est animal superbissimus quod dicitur mori ex indignatione si preter voluntatem teneatur. Dicitur autem unicornis ab uno cornu quod habent inare per hunc significantur iudei qui sunt quasi unicorns terrena tantum petentes vel de cultu dei singulariter se iactantes vel singulariter superbientes occiso filio dei et servis; Cassiodorus, Expositio, 252. 66 Twelfth-Century Psalter Commentary, I, 259, lines 981–3. 67 Morgan Powell, ‘Translating Scripture for Ma Dame de Champagne: The Old French ‘Paraphrase’ of Psalm 44 (Eructavit)’, in The Vernacular Spirit: Essays on Medieval Religious Literature, ed. Renate Blumenfeld-Kosinski, Duncan Robertson and Nancy Bradley Warren, Basingstoke 2002, 83–103 at 83. 68 Powell, ‘Translating Scripture’, 84. 69 John F. Benton, ‘The Court of Champagne as a Literary Center’, Speculum 36, 1961, 551–91, at 553, 586–7. 70 Benton, ‘The Court of Champagne’, 566.



Anti-Jewish Polemic

219

regarding the poem’s possible author. George McKibben suggested a Benedictine monk, and T. Atkinson Jenkins suggested Adam de Perseigne (d. 1221), a Cistercian monk and possibly Marie de Champagne’s chaplain, though John Benton has dismissed this particular option as unconvincing.71 Other twelfth-century French vernacular, scriptural, exegetical poetry has been assessed for its anti-Jewish sentiments, such as the poem, relating to the proverbs of Salomon, written in the second quarter of the twelfth century by Sansun de Nantuil, at the request of Alice de Condet, possibly for her son, Roger.72 Eructavit, extant in part or in full in fourteen manuscripts, however, has not, until now, been considered for its content regarding the Jews.73 And yet, anti-Jewish terminology is present, as seen in the poem’s prologue, where the poet takes issue with the Jews’ unwillingness to accept the principle of the Virgin Birth, reflecting the idea of the Jews lacking reason, i.e. denying the veracity of Christianity.74 Yet in the poet’s interpretation of Psalm 44:6, St Augustine’s concept of Jewish service to Christianity is clearly present.75 The verse’s references to ‘sharp arrows’ and the ‘king’s enemies’ allow the poet to discuss how, like those arrows, various Christian saints, such as St Simon and St Mathias, strove to bring conversion to countries as far flung as Persia and Mauritania.76 This discussion then leads the poet into echoing St Augustine and Psalm 58:12, stating: God commanded (do not doubt it at all) that we not kill the Jews, but rather, that we should let them live among us, because in their law are the books of our faith and the witness which we need.77

The inclusion of St Augustine’s concept of Jewish service in a section related to the conversion of pagans thus makes sense, since it demonstrates how the books of the Jews would stand as convincing proof to pagans of the truth of Christianity. What is pertinent is that earlier exegesis written on this psalm verse does not refer to such service. It does not, for example, appear in the relevant twelfth-century glosses on this particular verse, nor in the glosses’ key patristic sources. Jerome (d. 420) did not mention it in the commentary on Psalm 44 contained in his letter to Principia, nor did the Carolingian exegete Paschasius Radbertus (d. 865) in his specific exposition on Psalm 44.78 Atkinson Jenkins, writing in 1909, argued that this reference to the Jews was a reflection of the political situation in France at the time of the poem’s composition. For Jenkins, what he described as the poem’s ‘plea for humane treatment of the Jews’ was a reflection of, and response to, the actions in 1182 of Philip Augustus (d. 1223), banishing the Jews from his lands. Jenkins supposed that 71

George Fitch McKibben, The Eructavit, an Old French Poem; the Author’s Environment, his Argument and Materials, Baltimore 1907, 4; Eructavit, an Old French Metrical Paraphrase, xiii; Benton, ‘The Court of Champagne’, 582. 72 Fenster, ‘English Women and their French Books’, 178. 73 Eructavit: An Old French Metrical Paraphrase, xxix. 74 Eructavit: An Old French Metrical Paraphrase, 6, lines 126–7 de gesine et d’accouchement qu li Jue contralieus. 75 Psalm 44:6 – Douai-Rheims, ‘Thy arrows are sharp: under thee shall people fall, into the hearts of the king’s enemies’ (Vulgate, Sagittae tuae acutae, populi sub te cadent, in corda inimicorum regis). 76 Eructavit: An Old French Metrical Paraphrase, 35, lines 811–13 and 819. 77 Eructavit: An Old French Metrical Paraphrase, 36, lines 839–44, Deus commanda, n’en dotez mie, des Jüeus que nus nes ocie, ainz les laissons antre nos vivre por ce qu’en lor loi sont li livre de nostre foi et li tesmoing a quoi nos alons au besoing. 78 Augustine, Enarrationes, 504–5; Cassiodorus, Expositio, 406–7; Jerome, Epistulae, 3 vols, ed. Isidorus Hilberg, Vienna 1910, I, 616–47 at 629–30; Paschasius Radbertus, Expositio in psalmum XLIV, ed. Bedae Paulus, Turnhout 1991, 60–1.

220

Linda M. A. Stone

some may have found refuge in Champagne, where there were flourishing Jewish communities, particularly in Troyes, the main seat of the counts of Champagne.79 While the poet’s admonition that God’s command not to kill the Jews was not to be doubted in any way might reflect the contemporary political situation, the inclusion of St Augustine’s concept of Jewish service was also an exegetical choice related to the text of the verse – a reflection of those arrows, those various saints, who had converted pagans to the right path, with Christianity’s truth verified by Christianity’s enemies. The decision to include a reference to the Jews’ role vis-avis Christianity must therefore reflect the poet’s own choice, and perhaps a clerical or monastic education, in which St Augustine’s teaching regarding the Jews would have featured. The dissemination of the idea of Jewish service to Christianity within a learned environment was therefore also permeating Christian lay society and not only in moral, biblical commentary instruction but also into religious literature requested by, or written for, women. Furthermore, the idea did not remain restricted to those able to read vernacular French. By the early thirteenth century, Eructavit had been translated into a mix of Old French and Occitan, retaining those sentiments of Jewish service and thus percolating into a still wider sector of Christian lay society.80 While Psalm 44 is considered to be in some sense a wedding song, the most famous biblical wedding song is the Song of Songs. It is perhaps not unexpected, then, that Psalm 44 was one of the most frequently used psalms in Bernard of Clairvaux’s sermons thereon.81 The Song of Songs, with its erotic content, had always been troublesome for both Jewish and Christian faiths, since, as E. Ann Matter stated, it is the only biblical text that tells no sacred history, makes no theological or moral points, and does not mention God.82 It was not confirmed in the Hebrew biblical canon until the synod of Jamnia held in ad 90, when Rabbi Akiba (d. 135) used the justification of allegorical interpretation to aver that, since the Song expressed God’s love for Israel, it was in fact the ‘Holy of Holies’ in Scripture.83 Origen (d. 254), considered to be the founder of Christian exegesis on the Song, voiced the difficulties that the Song presented if taken at face value, since its sacred text had the potential to incite fleshly lust, which presented ‘no small difficulty and danger’.84 The only option was therefore to interpret the Song allegorically, eliminating any human, erotic elements; by the twelfth century, it had become the most frequently interpreted book of medieval Christianity.85 The Cistercian sermons on the Song of Songs – initiated by the eighty-six or so sermons by Bernard of Clairvaux, and continued by his successors Gilbert of Hoyland (d. 1170) and John of Ford (d. 1244) – all focused on the expression of love between God and the Christian community, but using allegory for the exegesis of the Song also provided an opportunity for the discussion of other matters, often reflecting the social, political, economic and religious issues of the time. As a result, such discussions often Eructavit: An Old French Metrical Paraphrase, x; Abulafia, Christian-Jewish Relations, 72–3. L’ “Eructavit” Antico-Francese secondo il ms. Paris B.N. fr 1747, ed. Walter Meliga, Alessandria 1992, lines 837–41; Keith Busby, Review of L’“Eructavit” Antico-Francese secondo il ms. Paris B.N. fr 1747, ed. Walter Meliga, in Romance Philology 49, 1996, 319–21. 81 Bernard of Clairvaux, On the Song of Songs, trans. Irene Edmonds, 4 vols, Kalamazoo 1980, IV, 234. 82 E. Ann Matter, The Voice of My Beloved: The Song of Songs in Western Medieval Christianity, Philadelphia 1992, 49. 83 The Cambridge History of the Bible, 3 vols, Cambridge 1970, I, From the Beginnings to Jerome, ed. P. R. Ackroyd and C. F. Evans, 134. 84 Ann W. Astell, The Song of Songs in the Middle Ages, Ithaca 1995, 1. 85 Matter, Voice of my Beloved, 4–6. 79 80



Anti-Jewish Polemic

221

focused on the Jews. The potential for the dissemination of such discussions can be demonstrated by the number of extant Latin manuscripts of Bernard’s sermons on the Song. Jean Leclercq estimated that there are at least 111 extant copies which date from the twelfth and early thirteenth centuries,86 while the late thirteenth-century register Registrum librorum Anglie demonstrates the breadth of such dissemination, with twenty-eight English cathedrals and monasteries holding copies.87 Bernard’s views regarding the Jews have been studied extensively by such historians as David Berger, Jeremy Cohen and Robert Chazan, often centring on Bernard’s use, in his letters, of Psalm 58:12 and its Augustinian interpretation that the Jews must not be killed, because of their service to Christianity.88 The debate has often focused on whether Bernard was simply anti-Jewish or whether, instead, he was following the theological path established by St Augustine.89 While this debate has concentrated mainly on the distribution of such ideas via Bernard’s writings in Latin, there is another route for their circulation, via the translation of Bernard’s sermons into the French vernacular, and their consequent dissemination into lay society. Gilbert Dahan states that many of Bernard’s sermons on the Song were translated from Latin into the vernacular in the twelfth century, and Stewart Gregory has transcribed one such manuscript.90 It is Gregory’s view that this translation was carried out by a cleric, possibly a Cistercian monk, in order to spread Bernard’s words among those who did not possess Latin. Furthermore, the text, extant now in only one manuscript, was composed, based on paleographical evidence, towards the end of the twelfth century, and comprises, in the Walloon dialect, a translation of forty-four of Bernard’s eighty-six sermons.91 Particular to the references to the Jews are contained in Bernard’s sermon XIV, relating to Song of Songs 1:2, ‘thy name is as oil poured out’.92 A comparison of the Latin to the vernacular shows that, where Bernard references the Jews, directly or indirectly, his wording is closely followed by the Cistercian cleric. Sermon XIV encapsulates various elements of generic anti-Jewish polemic, serving to remind Christians of the dangers of having characteristics or attitudes that could be associated with Jews, but it also contains references to that Pauline prophecy of the Jews’ final and crucial service to Christianity, their conversion. Throughout the Latin sermon, the Jews are consistently portrayed negatively, using generic anti-Jewish terminology, and these terms were duly translated into the vernacular. Synagoga, for example, in contrast to ecclesia, is described as blind and 86

Bernard of Clairvaux, Sermones super Cantica Canticorum, in Sancti Bernardi Opera, ed. J. Leclercq, C. H. Talbot and Henri Rochais, 8 vols, Rome 1957, I, xxiii. 87 Christopher Holdsworth, ‘The Reception of Saint Bernard in England’ in Bernard von Clairvaux: Rezeption und Wirkung im Mittelalter und in der Neuzeit, ed. Kaspar Elm, Wiesbaden 1994, 161–77 at 172. 88 David Berger, ‘The Attitude of St. Bernard of Clairvaux Towards the Jews’, American Academy for Jewish Research Proceedings, 40, 1972, 89–108; Jeremy Cohen, ‘“Witnesses of our Redemption”. The Jews in the Crusading Theology of Bernard of Clairvaux’, in Medieval Studies in Honour of Avrom Saltman, ed. Bat-Sheva Albert, Yvonne Friedmann and Simon Schwarzfuchs, Jerusalem 1995, 67–82; Robert Chazen, ‘Twelfth-Century Perceptions of the Jews: A Case Study of Bernard of Clairvaux and Peter the Venerable’, in From Witness to Witchcraft: Jews and Judaism in Medieval Christian Thought, ed. Jeremy Cohen, Wiesbaden 1996, 187–201. 89 Berger, ‘The Attitude of St. Bernard’, 89. 90 Gilbert Dahan, L’exégèse chrétienne de la Bible en Occident medieval: XIIe-XIVe siècle, Paris 1999, 16; La traduction en prose française, 2–283. 91 La traduction en prose française, ix–xiii. 92 Song of Songs 1:2 – Douai-Rheims, ‘Smelling sweet of the best ointments. Thy name is as oil poured out: therefore young maidens have loved thee’.

222

Linda M. A. Stone

quarrelsome (avougle et tenzouse).93 Such a description would also have resonated with contemporary art and architecture, wherein synagoga was often portrayed as blind and veiled, in contrast to the direct gaze of ecclesia.94 The Jews are portrayed variously as glorying in the law (glorïevet en la loi), arrogant (l’orguil Israël), and as friends of demons (les deables ses amis), references that all cohere with longstanding generic anti-Jewish terminology.95 There are, however, other elements of Jewish service contained within the Sermon. Both the Latin and thus the vernacular versions contain references to the ultimate service of the Jews to Christianity, their conversion. First, in its chastisement of synagoga, referencing in part Romans 11:25, and echoing St Augustine’s concept of Jewish service, the sermon makes reference to the view that the Jews, because of their pride and jealousy, will remain in error until the conversion of the Gentiles.96 The sermon then refers to Romans 9:27, commenting on how God will ultimately save the remnant, but the sermon emphasizes that such salvation would come about only through God’s mercy, which would redeem the Jews from what should have happened to them in light of their lack of mercy (i.e. in relation to Christ).97 The sermon, in both Latin and the French vernacular, resonates with elements of Jewish service to Christianity, not only in its direct anti-Jewish terminology but also implicitly in its references to the Jews’ final conversion. This sermon, and the other forty-three sermons, and their views on Jews were thus no longer the preserve of Latinate monks in the Cistercian monastery. Accessible in the vernacular, the sermons’ polemical views on Jews were available to a literate lay society. The vernacular version of Bernard’s sermons on the immensely popular Song of Songs, together with the evidence from Laurette’s Psalms commentary and the poem for Marie de Champagne on Psalm 44, all reinforce the idea of the castigated Jew, subservient, as St Augustine had striven to emphasize, to Christianity. But these texts and their chastisement of the Jews were not for a Jewish audience but for Christians, in these instances lay Christians, both men and women. The vast majority of Jews were secure in their own faith, so what was the purpose of disseminating such images and ideas into Christian society? Essentially, the Jew was performing another service, standing as a cautionary metaphor, highlighting for Christians the risks of not following the church’s teaching, of not accepting the correct order of things and the truth of Christianity, or of being too proud, too worldly, stubborn, even irrational – the danger of being perceived as a Jew. 93

Bernard of Clairvaux, Cantica Canticorum, Sermo XIV, 76, line 23: caeca et contentiosa; La traduction en prose française, 73, lines 47–8. 94 Bernhard Blumenkranz, Le juif médiévale au miroir de l’art chrétien, Paris 1966, 106–10. 95 Bernard of Clairvaux, Cantica Canticorum, Sermo XIV, 76, line 5: gloriante in lege; 77, line 23: supercilio Israel; 78, lines 14–15: cum amicis suis daemonibus; La traduction en prose française, 72, line 23; 74, line 88; 75, 118–19; Samuel Krauss, The Jewish-Christian Controversy: From the Earliest Times to 1789, 2 vols, ed. and rev. William Horbury, Tübingen, 1996, I, 19–26. 96 Romans 11:25 – Douai-Rheims, ‘For I would not have you ignorant, brethren, of this mystery (lest you should be wise in your own conceits), that blindness in part has happened in Israel, until the fulness of the Gentiles should come in’ (Vulgate, Nolo enim vos ignorare, fratres, mysterium hoc (ut non sitis vobis ipsis sapientes), quia caecitas ex parte contigit in Israel, donec plenitudo gentium intraret); Bernard of Clairvaux, Cantica Canticorum, Sermo XIV, 76, line 23: donec plenitudo gentium; La traduction en prose française, 73, line 47: de ci ke la plantez des paiens. 97 Bernard of Clairvaux, Cantica Canticorum, Sermo XIV, 76, line 29: reliquias salvaturus; La traduction en prose française, 73, line 56: anz salverat les remasilhes; Bernard of Clairvaux, Cantica Canticorum, Sermo XIV, 77, lines 32–3: iudicium profecto sine misericordia ei qui non facit misericordiam; La traduction en prose française, 73, lines 60–1: jugemenz senz mercit fuist a celui vraiement qui ne fist mercit.



Anti-Jewish Polemic

223

However, while the terminology used by clerics in their Latin exegesis regarding the Jews, and distilled into these vernacular texts, did serve as a warning to Christianity, it also contributed to the rendering of the Jews as ‘other’, outside of society, no better than animals, servants to Christianity. Increasingly negative attitudes towards the Jews were gaining traction, intellectually, theologically, politically and financially, within twelfth-century Christian society, and, furthermore, there were no signs of the Jews imminently converting.98 Of course, the presence of subservient and anti-Jewish terminology in a wide variety of both Latin and vernacular texts, and their dissemination into learned and lay society, were just two of many factors at work in creating such negativity. Nonetheless, for the Jews, their exegetically imposed role as St Augustine’s servants to Christianity would become increasingly ambiguous and, eventually, would cease to serve as a guarantee of their safety.99

98 99

Abulafia, Twelfth-Century Renaissance, 138–40. Abulafia, Christian-Jewish Relations, 220–2.

CONTENTS OF PREVIOUS VOLUMES For details of all previous volumes please go to www.boydellandbrewer.com An asterisk after a title indicates the R. Allen Brown Memorial Lecture. Volume 36 (2014) Elisabeth van Houts, The Planctus on the Death of William Longsword (943) as a Source for Tenth-Century Culture in Normandy and Aquitaine* Ilya Afanasyev, Biblical Vocabulary and National Discourse in Twelfth-Century England Mathieu Arnoux, Border, Trade Route, or Market? The Channel and the Medieval European Economy from the Twelfth to the Fifteenth Century Robert F. Berkhofer III, Guerno the Forger and his Confession Laura Cleaver, From Codex to Roll: Illustrating History in the Anglo-Norman World in the Twelfth and Thirteenth Centuries Matthew Hammond, The Adoption and Routinization of Scottish Royal Charter Production for Lay Beneficiaries, 1124–1195 Susan M. Johns, Women and Power in the Roman de Rou of Wace Catherine Letouzey-Réty, Literacy and Estate Administration in a Great AngloNorman Nunnery: Holy Trinity, Caen, in the Twelfth and Thirteenth Centuries Alheydis Plassmann, The King and his Sons: Henry II’s and Frederick Barbarossa’s Succession Strategies Compared Sigbjørn Olsen Sønnesyn, In vinea Sorech laborare: The Cultivation of Unity in Twelfth-Century Monastic Historiography Andrew Wareham, The Redaction of Cartularies and Economic Upheaval in Western England c. 996–1096 Teresa Webber, Monastic Space and the Use of Books in the Anglo-Norman Period Emily A. Winkler, 1074 in the Twelfth Century Volume 37 (2015) Edmund King, Henry of Winchester: The Bishop, the City and the Wider World* Richard Allen, Episcopal acta in Normandy, 911–1204: The Charters of the Bishops of Avranches, Coutances and Sées Pierre Bauduin, Richard II de Normandie: Figure princière et transferts culturels (fin dixième–début onzième siècle) Johanna Dale, Royal Inauguration and the Liturgical Calendar in England, France, and the Empire c. 1050–c. 1250 Jennifer Farrell, History, Prophecy and the Arthur of the Normans: The Question of Audience and Motivation behind Geoffrey of Monmouth’s Historia regum Britanniae Peter Fergusson, Canterbury Cathedral Priory’s Bath House and Fish Pond Sara Harris, Tam Anglis quam Danis: ‘Old Norse’ Terminology in the Constitutiones de foresta

226

Previous Volumes in the Series

Nicholas Karn, Quadripartitus, Leges Henrici Primi and the Scholarship of English Law in the Early Twelfth Century Lauren Mancia, John of Fécamp and Affective Reform in Eleventh-Century Normandy Eljas Oksanen, Trade and Travel in England during the Long Twelfth Century Gesine Oppitz-Trotman, The Emperor’s Robe: Thomas Becket and Angevin Political Culture Benjamin Pohl, The Illustrated Archetype of the Historia Normannorum: Did Dudo of Saint-Quentin Write a ‘chronicon pictum’? Volume 38 (2016) John Hudson, From the Articles of the Barons to Magna Carta* Anna Sapir Abulafia, Jews in the Glosses of a Late Twelfth-Century AngloNorman Gratian Manuscript (Cambridge, Gonville and Caius College, MS 283/676) Casey Beaumont, Monastic Autonomy, Episcopal Authority and the Norman Conquest: The Records of Barking Abbey Giles E. M. Gasper, Economy Distorted, Economy Restored: Order, Economy and Salvation in Anglo-Norman Monastic Writing Kate Hammond, Monastic Patronage and Family Disputes in Eleventh- and Early Twelfth-Century Normandy Alan V. Murray, Constance, Princess of Antioch (1130–1164): Ancestry, Marriages and Family Jean-François Nieus, Early Aristocratic Seals: An Anglo-Norman Success Story Jonathan Paletta, English Towns and Urban Society after the Norman Conquest Susan Raich, Wreck of the Sea in Law and Practice in Eleventh- and TwelfthCentury England Miri Rubin, Social Life and Religious Culture in Twelfth-Century Norwich and Norfolk Luigi Russo, Bad Crusaders? The Normans of Southern Italy and the Crusading Movement in the Twelfth Century Hugh M. Thomas, Turold, Wadard and Vitalis: Why Are They on the Bayeux Tapestry? Volume 39 (2017) Christopher Clark, Why Do Battles Matter? Julie Barrau, From Conquest to Commonwealth: Cross-Channel Circulation of Biblical Culture in the Anglo-Norman World Laura Cleaver, Documentation, Forgery and the Making of the Chronicle of Battle Abbey (British Library, MS Cotton Domitian A II) F. P. C. de Jong, Rival Schoolmasters in Early Eleventh-Century Rouen with Special Reference to the Poetry of Warner of Rouen (fl. 996-1027) Brian Golding, Remembering the Battle of Hastings: Memorialization, le Souvenir Normand, and the Entente Cordiale Simon Keynes, Earl Harold and the Foundation of Waltham Holy Cross (1062) Tom Licence, Edward the Confessor and the Succession Question: A Fresh Look at the Sources



Previous Volumes in the Series

227

Brigitte Meijns, England and Flanders Around 1066: The Cult of the English Saints Oswald and Lewinna in the Comital Abbey of Bergues Thomas O’Donnell, The Carmen de Hastingae Proelio: Politics and the Poetics of 1067 Alheydis Plassmann, England and Germany: Two Perspectives Élisabeth Ridel, Les préparatifs nautiques de la Conquête: Un héritage Viking? Les mots ont la parole … Christopher Whittick, Battle Abbey and the Vellomaniacs – Locating the Monastic Archive Ann Williams, Of Danes and Thegns and Domesday Book: Scandinavian Settlement in Eleventh-Century Berkshire Volume 40 (2018) Nicholas Vincent, English Kingship: The View from Paris, 1066–1204* (2017) C. P. Lewis, Audacity and Ambition in Early Norman England and the Big Stuff of the Conquest* (2016) Mathieu Arnoux, Ressources et croissance dans le monde Anglo-Normand: Sources et hypothèses James Barnaby, Becket vult: The Appropriation of St Thomas Becket’s Image during the Canterbury Dispute, 1184–1200 Dominique Barthélemy, La Bataille de Bouvines reconsiderée Scott G. Bruce, Abbot Peter the Venerable’s Two Missions to England (1130 and 1155/1156) Francis Gingras, La production manuscrite Anglo-Normande et la Bible d’Herman de Valenciennes: Usage et réception d’un livre vernaculaire (XIIe–XIVe siècles) Frédérique Lachaud, Ralph Niger and the Books of Kings Anne E. Lester, From Captivity to Liberation: The Ideology and Practice of Franchise in Crusading France Amy Livingstone, ‘Daughter of Fulk, Glory of Brittany’: Countess Ermengarde of Brittany (c. 1070–1147) Fanny Madeline, The Idea of ‘Empire’ as Hegemonic Power under the Norman and Plantagenet Kings (1066–1204) Emily Joan Ward, Child Kingship and Notions of (Im)maturity in North-Western Europe, 1050–1262 Thomas N. Bisson Note: A Micro-Economy of Salvation: Further Thoughts on the ‘Annuary’ of Robert of Torigni