William the Conqueror 0300118759, 9780300118759

Citation preview

W I L L I A M T H E C O N Q U E RO R

i

ii

WILLIAM THE C O N Q U E RO R David Bates

YALE UNIVERSITY PRESS NEW HAVEN AND LONDON

iii

Copyright © 2016 David Bates All rights reserved. This book may not be reproduced in whole or in part, in any form (beyond that copying permitted by Sections 107 and 108 of the U.S. Copyright Law and except by reviewers for the public press) without written permission from the publishers. For information about this and other Yale University Press publications, please contact: U.S. Office: [email protected] yalebooks.com Europe Office: [email protected] yalebooks.co.uk Set in Baskerville by IDSUK (DataConnection) Ltd Printed in Great Britain by Gomer Press, Llandysul, Ceredigion, Wales. Library of Congress Cataloging-in-Publication Data Names: Bates, David, 1945– author. Title: William the Conqueror / David Bates. Description: New Haven : Yale University Press, 2016. Identifiers: LCCN 2016030229 | ISBN 9780300118759 (cl : alk. paper) Subjects: LCSH: William I, King of England, 1027 or 1028–1087. | Great Britain—History—William I, 1066–1087. | Great Britain—Kings and rulers—Biography. | Nobility—France—Normandy—Biography. | Normandy (France)—History—To 1515. | Normans—Great Britain—Biography. Classification: LCC DA197 .B3423 2016 | DDC 942.02⁄1092 [B] —dc23 LC record available at https://lccn.loc.gov/2016030229 A catalogue record for this book is available from the British Library. 10 9 8 7 6 5 4 3 2 1

iv

This book is dedicated to the memory of Professor Frank Barlow, with grateful thanks

v

vi

CONTENTS

viii x xiv xvii xviii

LIST OF PLATES PREFACE LIST OF ABBREVIATIONS MAP GENEALOGY

Prologue: Writing a Life of William the Conqueror

1

1

The First Years

16

2

From Childhood into Adolescence

49

3

The Shaping of Things to Come

91

4

The Making of a Reputation

127

5

On to the Attack

164

6

The Year of Victory

211

7

King of the English

258

8

Victory Transformed

295

9

From Crisis to Triumph

329

10 Cross-Channel Rule

373

11 Political Consolidation and Personal Loss

423

12 The Final Years

451

13 Death and Legacy

483

Epilogue

513 529 568

BIBLIOGRAPHY INDEX

vii

P L AT E S 1.

William of Jumièges presenting the Gesta Normannorum Ducum to William from Orderic’s autograph manuscript of the Gesta Normannorum Ducum (Rouen, Bibliothèque Municipale, ms. 1174 (Y.14), fo. 116r). © Collections Bibliothèque Municipale de Rouen. 2. Diploma of Count Baldwin V of Flanders (13 November 1056) recording Harold’s presence at Saint-Omer (Gent, Rijksarchief te Gent, Sint-Pietersabdij, charter, no. 133). © Gent, Rijksarchief te Gent. 3. The tomb slab of Queen Matilda in the choir of the abbey of La Trinité of Caen. © Région Normandie – Inventaire général – Manuel de Rugy. 4. The Bayeux Tapestry’s representation of Harold taking an oath at Bayeux. © Detail of the Bayeux Tapestry, eleventh century, with special permission from the city of Bayeux. 5. (a) and (b) The Bayeux Tapestry’s representations of Harold crossing the Channel to France and his apparent seizure on landing by Guy, count of Ponthieu. © Detail of the Bayeux Tapestry, eleventh century, with special permission from the city of Bayeux. 6. (a) and (b) The Bayeux Tapestry’s representations of King Edward the Confessor’s deathbed and King Harold’s coronation. © Detail of the Bayeux Tapestry, eleventh century, with special permission from the city of Bayeux. 7. William’s coronation by papal legates in 1070 as represented in the late eleventh-century manuscript known as the Ramsey Benedictional (BNFr, ms. latin, 987, fo.111r). © Bibliothèque Nationale de France. 8. (a) and (b) The west front and the nave and crossing and choir of the abbey of La Trinité of Caen. Photographs taken by François Neveux. 9. The manuscript of the version of the Pseudo-Isidorean Decretals associated with Archbishop Lanfranc showing the section describing how it was obtained for Christ Church, Canterburg (Trinity College, Cambridge, ms. B.16.44). © The Master and Fellows of Trinity College, Cambridge. 10. William’s writ confirming to the citizens of London all the laws of which they were worthy in King Edward the Confessor’s day. © London Metropolitan Archives. viii

plates

ix

11. (a), (b) and (c) Pennies of the first, second and (possibly) last issue (the PAXs type) minted in William’s name as king of the English. © The Fitzwilliam Museum, Cambridge. 12. A cast of the majesty side of William’s double-sided seal, made in the nineteenth century by Louis-Claude Douët d’Arcq. © Archives Nationales de France. 13. The remains of the west front of the abbey church of Notre-Dame of Jumièges consecrated in William’s presence on 1 July 1067. © Conway Library, The Courtauld Institute of Art, London. Photograph by James Austin. 14. (a) and (b) The abbey church of Saint-Etienne of Caen. © Conway Library, The Courtauld Institute of Art, London. Photograph (b) taken by F. R. P. Sumner. 15. The gatehouse of the castle enclosure at Exeter built in the years after the 1068 siege. © R. A. Higham. 16. (a) and (b) The White Tower of London and the remains of the keep at Ivry-la-Bataille in Normandy. © Dr Roland B. Harris 17. The surviving remains of the west front of Bishop Remigius’s cathedral at Lincoln. © Conway Library, The Courtauld Institute of Art, London. 18. A twelfth-century illuminated initial showing William sending a letter to Abbot St Hugh of Cluny to request monks assist the Church in England (BNFr, ms. latin 17, 716, fo.37r). © Bibliothèque Nationale de France. 19. Diploma recording the restitution of Gisors to Rouen cathedral by Simon, count of Amiens, Valois and the Vexin, in the presence of Matilda, Roger de Beaumont and other nobles (AD Seine-Maritime G8739). © Archives Départementales de la Seine-Maritime. 20. Two of the great original pancartes recording property acquired by the abbey of Saint-Etienne of Caen (AD Calvados H1830, nos. 1 and 1bis). © Archives Départementales du Calvados. 21. Aerial view of the site north of Sainte-Suzanne known as the Camp de Beugy, which was, in some form, used by William to conduct the siege. © Gilles Leroux, Inrap. 22. (a), (b) and (c) Reproduction from the Alecto facsimiles of two folios of Great Domesday Book for Bedfordshire, and the colophon at the conclusion of Little Domesday Book. © Reproduced by kind permission of the Syndics of Cambridge University Library.

P R E FAC E

I have lived with the responsibility of writing a full, scholarly biography of William the Conqueror for many years. The advice that I should do so was given by the late Marjorie Chibnall in a typically wise review of the ‘popular’ biography of William that I published in 1989. The suggestion that I should actually write this book came from Robert Baldock in 2000, with the signature of the contract following rapidly afterwards. Robert has been a consistent supporter of the project over the fifteen years of intermittent and interrupted progress that have followed. His suggestion in 2013 that I should write a prologue reflecting on David Douglas’s book and the changes of perspective that have occurred in the fifty years since it was published was a catalyst of decisive importance in this book’s completion. The 1989 William the Conqueror that I published, although informed by my ongoing work on William’s charters and a fully up-to-date knowledge of the secondary literature in terms of how I then understood it, now feels like a book written within a narrow historiographical tradition. This one is intended to bring together much broader perspectives on its subject. It attempts to locate William within the major changes in the writing of history that have taken place during my professional lifetime as well as the history of both eleventh-century Europe and the wider themes of human history over a much longer period. It is deliberately as multidisciplinary as I can make it. Even as I have delivered the manuscript to Yale University Press, I am left thinking that there is more that I must reflect on and learn about. It is my hope that others will be persuaded to think as much as I have. Since 2000, I have benefitted from the support of many institutions and individuals. My employer in 2000, the University of Glasgow, was a generous supporter, as has been the University of East Anglia, my employer from 2007 to 2010 and an exceedingly generous champion of my work since my retirement. Between 2009 and 2012, I held a Chaire d’Excellence at the Université de Caen Basse-Normandie, financed by the Conseil Régional de la Basse-Normandie. This gave me indispensable access to the university’s unique libraries and to seminars at a university with which I

x

p r e fac e

xi

have long been closely associated, and of which I am honoured to hold a doctorate honoris causa. I held a British Academy Marc Fitch Research Fellowship from 2001 to 2003, which released me from all duties at the University of Glasgow, and a Leverhulme Trust Emeritus Fellowship from 2013 to 2015, which funded the final stages of research and writing. I was also privileged to hold a Visiting Fellow Commonership at Trinity College, Cambridge, in 2002–03, and a Visiting Professorship in Paris at the École Pratique des Hautes Études in 2003. Since 2003, the Faculty of History of the University of Cambridge and Clare Hall (of which I was elected a life member with Marjorie Chibnall’s support), and Cambridge University Library have provided support, stimulus and working conditions that have been invaluable. The apparent disruptions of the onerous duties of the Directorship of the Institute of Historical Research from 2003 to 2008, and the invitation to give the Ford Lectures in the University of Oxford in 2010 (subsequently published in 2013 as The Normans and Empire), were hugely important to the broadening of intellectual horizons that have produced this book. The library of the Institute of Historical Research, where David Douglas’s private library was deposited during my time as Director, has remained my favourite place to work. I have incurred many debts of gratitude over the long period since I signed the contract to write this book. Drafts of all the chapters in the final manuscript have been read by Elisabeth van Houts and Helen Bates. Both have made many suggestions that have made huge improvements to both the presentation and quality of argument. At Yale University Press, Rachael Lonsdale has provided support, guidance and, sometimes, entirely justified pressure, all of which have been indispensable in the process of completion. Melissa Bond and Samantha Cross have seen the book through its production stage with exemplary efficiency and good humour. My family – Jonathan, Shiho and Amy Bates-Kawachi, and Rachel Bates – have been strong supporters, if occasionally impatient ones. Over such a long period, I have benefitted greatly from the advice, stimulus and criticism of many scholars. Inevitably, given my age, some of them are no longer with us. Marjorie Chibnall, Tim Reuter and Patrick Wormald all fall into this category. It is a measure of their influence that I still ponder what they might be saying to me now. Jinty Nelson has been a great and long-term supporter of the project, and John Gillingham a consistent stimulus to thought. Simon Keynes and Tessa Webber were hugely helpful to me during my time at Trinity, and Tessa has subsequently advised on many palaeographical matters. Pierre Bauduin and Véronique Gazeau were centrally important to making possible my large stretches of work in Caen, where many colleagues welcomed me warmly. Dominique Barthélemy

xii

p r e fac e

was immensely supportive during my time in Paris. I have also discussed several major aspects of the book’s narrative and analysis with my UEA colleague Tom Licence to my great profit, and Martin Allen’s advice on all numismatic matters has been invaluable. Anna Asbury has helped with many problems of translation and made a large contribution towards compiling the bibliography. Richard Mason was an excellent copy-editor who did a great deal to shape some of the less confident prose in some sections of the book. Sophie Rixon has provided research assistance and worked on the index. Emily Ward has also worked on the index and provided good counsel. During the times that I have been writing and thinking about this book, I have had so many helpful conversations with people and received so much advice that the list of names that follows cannot possibly be complete. I therefore apologise to all who feel I have accidentally excluded them, and I hope that a statement on these terms will be a sufficient compensation for my forgetfulness. Those whom I must specifically thank are: Bill Aird, Laura Ashe, Debby Banham, Julia Barrau, Stephen Baxter, Pierre Bouet, Jean-Michel Bouvris, Elma Brenner, Martin Brett, Chris Briggs, Joe Canning, Cathie Carmichael, Helena Carr, Jean-Luc Chassel, John Clark, Bob Cowie, David Crouch, Johanna Dale, Els de Parmaentier, Frances Eustace, Anne-Marie Flambard Héricher, Jean-Hervé Foulon, Richard Gameson, Anna Gannon, George Garnett, Lindy Grant, Judith Green, Francis Grew, Ralph Griffiths, Mark Hagger, Elizabeth Hallam-Smith, Roland Harris, Sandy Heslop, Leonie Hicks, Bob Higham, George Hope, Edward Impey, Michael Jones, Katharine Keats-Rohan, Derek Keene, Pierre-Yves Laffont, Sylvette Lemagnen, Jacques Le Maho, Catherine Letouzey-Réti, Robert Liddiard, Aleksandra McClain, Christophe Maneuvrier, Stephen Marritt, Florian Mazel, Rory Naismith, François Neveux, Mark Ormrod, Jörg Peltzer, Alheydis Plassmann, Benjamin Pöhl, Susan Raich, Carole Rawcliffe, Sarah Rees Jones, Susan Reynolds, David Roffe, Pauline Stafford, Matthew Strickland, Kathleen Thompson, Rod Thomson, Nicholas Vincent, Paul Warde and Emily Winkler. I have given lectures and seminars on William over recent years in the universities of Bergen, Bristol, Caen Normandie, the Department for Continuing Education at the University of Oxford, and Rennes II, and must thank many people in the audiences at those institutions for their contributions. The final and largest expression of thanks must be to my PhD supervisor at the University of Exeter, Frank Barlow, to whose memory this book is dedicated. Frank was writing Edward the Confessor when I was his research student and gave me the manuscript to read. I willingly admit that I cannot remember any advice that I might have given him that could have been of any use. But the vote of confidence in my abilities and the way in which I

p r e fac e

xiii

was treated as a serious member of the historical profession did wonders for my self-esteem. The preparation of the British Academy’s memoir of Frank’s life after his death in 2009 at the age of ninety-eight brought his way of doing things back to the foreground of my memory and has been enormously beneficial in the writing of this book. Its publication and its dedication to his memory fulfil a promise I made to him in the last months of his life.

ABBR E V I AT I O NS

AAC

AAR AD AN ANS ASC

ASE B&C BIHR BL BNFr Breuis Relatio

BSAN Carmen CCCM CS DB

Les actes de Guillaume le Conquérant et de la reine Mathilde pour les abbayes caennaises, ed. Lucien Musset, Mémoires de la Société des Antiquaires de Normandie, vol. xxxvii (Caen, 1967) ‘The Acta Archiepiscopum Rotomagensium, Study and Edition’, ed. Richard Allen, Tabularia www.unicaen.fr/mrsh/ craham/revue/tabularia, ix (2009), Documents, 1–66 Archives Départementales Annales de Normandie Anglo-Norman Studies: Proceedings of the Battle Conference 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). For a convenient modern translation, The Anglo-Saxon Chronicle, trans. and ed. M. J. Swanton (London, 1996) Anglo-Saxon England Facsimiles of English Royal Writs to A.D. 1100 presented to V. H. Galbraith, ed. T. A. M. Bishop and P. Chaplais (Oxford, 1957) Bulletin of the Institute of Historical Research British Library Bibliothèque Nationale de France ‘The Breuis Relatio de Guillelmo nobilissimo comite Normannorum, written by a Monk of Battle Abbey’, ed. and trans. Elisabeth van Houts, in idem, History and Family Traditions in England and the Continent, 1000–1200 (Aldershot and Burlington, VT, 1999), chapter VII. Reprinted, with a translation added, from Camden Miscellany, 5th ser., x (Cambridge, 1997) Bulletin de la Société des Antiquaires de Normandie The Carmen de Hastingae Proelio of Guy Bishop of Amiens, ed. and trans. Frank Barlow, OMT (Oxford, 1999) Corpus Christianorum Continuatio Medievalis Councils and Synods with Other Documents Relating to the English Church, 1066–1204, ed. Dorothy Whitelock, Martin Brett, and C. N. L. Brooke, 2 vols. (Oxford, 1981) Domesday Book, seu Liber Censualis Willelmi Primi regis Angliae, ed. Abraham Farley, 2 vols. (London, 1783)

xiv

abbreviations

Douglas, WC DP EHR EME GC GG GND GP GR

HCY

HH HN

HR HRA HSJ JEH JMH JW Lanfranc Letters LDE MGH MSHAB

xv

David. C. Douglas, William the Conqueror: The Norman Impact on England, new edition (New Haven, CT, and London, 1999) Katharine Keats-Rohan, Domesday People: A Prosopography of Persons Occurring in English Documents, 1066–1166. 1. Domesday Book (Woodbridge, 1999) English Historical Review Early Medieval Europe Gallia Christiana in Provincias Ecclesiasticas distributas, ed. D. Samarthanus et al., 16 vols. (Paris, 1715–1865) The Gesta Guillelmi of William of Poitiers, ed. and trans. R. H. C. Davis and Marjorie Chibnall, OMT (Oxford, 1998) The Gesta Normannorum Ducum of William of Jumièges, Orderic Vitalis, and Robert of Torigni, ed. and trans. Elisabeth M. C. van Houts, 2 vols., OMT (Oxford, 1992–5) William of Malmesbury, Gesta pontificum Anglorum: The History of the English Bishops, ed. and trans. M. Winterbottom and R. M. Thomson, 2 vols., OMT (Oxford, 2007) William of Malmesbury, Gesta regum Anglorum: The History of the English Kings, ed. and trans. R. A. B. Mynors, M. Winterbottom, and R. M. Thomson, 2 vols., OMT (Oxford, 1998–9) Hugh the Chanter, The History of the Church of York, 1066– 1127, ed. and trans. Charles Johnson, revised by Martin Brett, C. N. L. Brooke, and Michael Winterbottom, OMT (Oxford, 1990) Henry, Archdeacon of Huntingdon, Historia Anglorum: The History of the English People, ed. and trans. Diana Greenway, OMT (Oxford, 1996) 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, no. 81 (London, 1884) Historical Research Historia Regum Anglorum, in Symeonis Monachi Opera Omnia, ed. T. Arnold, 2 vols., RS, no. 75 (London, 1882–5) Haskins Society Journal Journal of Ecclesiastical History Journal of Medieval History The Chronicle of John of Worcester, ed. R. R. Darlington and P. McGurk, 2 vols. (vols. ii and iii), OMT (Oxford, 1995–8) The Letters of Lanfranc, Archbishop of Canterbury, ed. and trans. Helen Clover and Margaret Gibson, OMT (Oxford, 1979) Symeon of Durham, Libellus de Exordio atque Procursu istius hoc est Dunhelmensis Ecclesie, ed. and trans. David Rollason, OMT (Oxford, 2000) Monumenta Germaniae Historica Mémoires de la Société d’Histoire et d’Archéologie de Bretagne

xvi NMPrinces NMProsopographie ODNB OMT OV PBA PL RADN Regesta Register Registrum RHF Roman de Rou RS TRHS VCH VEdR

abbreviations

Véronique Gazeau, Normannia Monastica. Princes normands et abbés bénédictins (Xe et XIIe siècle), Publications du CRAHM (Caen, 2007) Véronique Gazeau, Normannia Monastica. Prosopographie des abbés bénédictins (Xe et XIIe siècle), Publications du CRAHM (Caen, 2007) Oxford Dictionary of National Biography (http://www. oxforddnb.com/view/article/) Oxford Medieval Texts The Ecclesiastical History of Orderic Vitalis, ed. and trans. Marjorie Chibnall, 6 vols., OMT (Oxford, 1969–80) Proceedings of the British Academy Patrologia cursus completus, series Latina, ed. J.-P. Migne, 221 vols. (Paris, 1844–65) Recueil des actes des ducs de Normandie de 911 à 1066, ed. Marie Fauroux, Mémoires de la Société des Antiquaires de Normandie, vol. xxxvi (Caen, 1961) Regesta Regum Anglo-Normannorum: The Acta of William I (1066–1087), ed. David Bates (Oxford, 1998) The Register of Pope Gregory VII, 1073–1085: An English Translation (Oxford, 2002) Das Register Gregors VII, ed. Erich Caspar, 2 vols. MGH, Epistolae Selectae (Berlin, 1920–3) Recueil des historiens des Gaules et de la France, ed. Martin Bouquet et al., 24 vols. (Paris, 1738–1904) The History of the Norman People: Wace’s Roman de Rou, trans. Glyn S. Burgess (Woodbridge, 2004) Rolls Series Transactions of the Royal Historical Society The Victoria History of the Counties of England (with county name), in progress The Life of King Edward who rests at Westminster, attributed to a monk of Saint-Bertin, ed. and trans. Frank Barlow, 2nd ed., OMT (Oxford, 1992)

miles

200

0

km

320

Abernethy

Falkirk

S

C

O

T

L

A

N

D

0

D

N O R T H N

Durham

E

N

I R I S H Dublin S E A

Stamford Bridge Rhuddlan

G

York

S

E L

Shrewsbury

A

Lichfield

Lincoln Nottingham

A

N

St David’s

Beverley

Selby

Chester

L

I R E L A

S E A

Newcastle upon Tyne

W

Worcester Peterborough Norwich Ramsey Hereford Ely Thetford Chepstow Gloucester Bury St Edmunds Cardiff Oxford GERMAN Malmesbury Dorchester Colchester Bath London Southwark Utrecht Wells Glastonbury EMPIRE Salisbury Rochester Wilton Winchester Exeter Canterbury Sandwich Bruges Bosham Battle Dover Wissant Selsey Pevensey Saint-Omer Ghent Boulogne F L A N D E R S Aachen Cologne E n g l i s h C h a n n e l Thérouanne Arras Cherbourg Liège Saint-Valéry Cambrai Fécamp L O T H A R S Y Bonneville Le Bec Amiens D Bayeux N Rouen Coutances Caen A Mainz M Se Lisieux Mont St. Michel R Pontoise in N O e Evreux Speyer Avranches Rheims Verdun Mantes Dinan Dol Domfront Paris B R I T T A N Y Alençon Rennes Chartres Rh in e

D

A N J O U Le Mans La Flèche

Angers Nantes

Orléans re Loi Blois Dijon

Besançon

Poitiers

B a y o f B i s c a y

Limoges

Lausanne

Cluny

Y

Angoulême

Vercelli

U B

Rhô ne

ne on

R

G

U

Vienne Ga r

N

D

A Q U I T A I N E

Map of Normandy, the British Isles and western Europe in the eleventh century

xvii

Selective genealogical chart to illustrate William’s family Svein ‘Forkbeard’ king of Denmark,

986–1014

Gytha = Godwine

Earl = Estrith Úlfr

earl of Wessex

Ælfgifu = Cnut the Great king of England, of 1016–35 Northampton king of Denmark,

1018–35 others

Tostig

Edith = Edward the Confessor

king of Norway,

1030–35

Svein Osbeorn Estrithson

Harold II (Godwinson)

Beorn

Harold I (Harefoot),

king of England,

king of Denmark,

king of England,

1066

1047–76

1035–40

Harold ‘Hein’ king of Denmark,

1076–80

Cnut IV ‘The Saint’

Olaf others king of Denmark, king of Denmark 1086–95 1080–86 murdered while about to invade England

Drogo

Godgifu

=

= Eustace I

count of the Vexin

count of Boulogne

Alfred

Edward The Confessor

murdered,

king of England,

1036

1042–66 = Edith dau. of Godwine earl of Wessex

Walter

Ralph

count of the Vexin d. c.1063

earl of Hereford d. 1057

Adelida

Matilda

Cecilia, d. 1127 (nun) abbess of La Trinité of Caen

xviii

Constance, d. 1090 = Alan IV count of Brittany

Adela, d. 1137 = StephenHenry count of Blois

? Agatha

=

and the disputed succession to the English Kingdom Edgar ‘The Peaceful’

duke of Normandy,

Richard I

king of England,

942–96

959–75

Emma d. 1052

=

=

Robert

Richard II

king of England,

archbishop of Rouen,

duke of Normandy,

978–1016

989–1037

996–1026

Richard III

Robert I

duke of Normandy,

duke of Normandy,

Æthelred II = Ælfgifu ‘The Unready’

Harthacnut

Edmund ‘Ironside’

king of Denmark,

king of England,

1035–42

1016

king of England,

1040–42

1026–27

Edward The Ætheling, d. 1057

1027–1035 = Herleva dau. of Fulbert

William I the Conqueror duke of Normandy, 1035–87 king of England, 1066–87 = Matilda dau. of Baldwin V count of Flanders

Edgar the Ætheling, d. c. 1125

Christina (became a nun)

Robert II duke of Normandy,

1087–1106, d. 1134

St = Malcolm III Margaret king of

count of Ponthieu,

d. 1053 (2) Lambert of Lens (3) Odo of Champagne

Judith = Earl Waltheof, d. 1076

Scotland

Richard o. s. p. l. c. 1070

Adelaide = (1) Enguerrand II

William II

Henry I

king of England,

king of England, 1100–35 duke of Normandy, 1106–35

1087–1100

= Edith/Matilda, d. 1118 dau. of Malcolm III and St Margaret

xix

1. William of Jumièges presenting the Gesta Normannorum Ducum to William, as represented on Orderic’s autograph manuscript of the Gesta Normannorum Ducum.

2. Diploma of Count Baldwin V of Flanders (13 November 1056) recording Harold’s presence at Saint-Omer. The signa include not only Harold, but also Guy, count of Ponthieu, and Guy, archdeacon (later bishop) of Amiens, and the text concerns responsibilities exercised by Eustace, count of Boulogne.

3. The tomb slab of Queen Matilda in the choir of the abbey of La Trinité of Caen. Placed at the centre of the choir, it was at the heart of a mausoleum that was intended to commemorate the lives of the female members of William’s family.

4. The Bayeux Tapestry’s representation of Harold arriving at Bayeux and then taking an oath to William. The ambiguity of the accompanying narrative (UBI HAROLD SACRAMENTUM FECIT WILLELMO DUCI) is notable.

5. (a) and (b) The Bayeux Tapestry’s representations of Harold crossing the Channel to France and his apparent seizure on landing by Guy, count of Ponthieu.

6. (a) and (b) The Bayeux Tapestry’s representations of King Edward the Confessor’s deathbed and King Harold’s coronation.

7. William’s coronation by papal legates in 1070, as represented in the late eleventh-century manuscript known as the Ramsey Benedictional. It includes a depiction of the papal banner given to William in 1066.

8. (a) and (b) The west front and the nave and crossing and choir of the abbey of La Trinité of Caen. The choir was substantially rebuilt towards the end of the eleventh century, and the west front is also of later date. Only the western part of the choir is likely to remain substantially unaltered from the church that was being built at the time of the dedication in 1066.

9. The version of the Pseudo-Isidorean Decretals associated with Archbishop Lanfranc. The section on the right-hand page contains the statement that the manuscript had been obtained for Christ Church, Canterbury, in all probability from Lanfranc’s former monastery of Le Bec. It is followed by a letter from the anti-pope Clement III.

10. William’s writ confirming to the citizens of London all the laws of which they were worthy in King Edward the Confessor’s day. Although formally undated, it is normally assigned to early in the year 1067, and represents an important ideological statement about William’s rule as king of the English.

11. Silver pennies of William I: (a) Penny of the Profile/Cross Fleury type, the first coinage issued on William’s behalf as king of the English (1066–c.1068), with a portrait similar to the coins of King Harold II, minted in 1066. This coin was produced at Hastings by a moneyer named Dunnic, who had also issued coins of Harold II. It shows that the English coinage was at first unaffected by William’s conquest; (b) Penny of William’s second (Bonnet) coinage, made by the York moneyer Aleif, c.1068–70. Eleven moneyers issued this coinage in York, but the number fell to four in the next issue. This is another indication of just how badly York was affected by the rebellions of 1068–9 and William’s Harrying of the North; (c) Penny of the Paxs type, issued by the Bristol moneyer Brihtword. The dating of this coinage is controversial. It might be William’s last coinage related to the attempted reconciliations associated with Domesday Book, or it might be a proclamation of the king’s peace corresponding with William II’s coronation in 1087.

12. A cast of the majesty side of William’s double-sided seal made in the nineteenth century by Louis-Claude Douët d’Arcq. It was copied from a seal attached to a diploma in favour of the abbey of Saint-Denis.

13. The remains of the west front of the abbey church of Notre-Dame of Jumièges, consecrated in William’s presence on 1 July 1067. The church was started in the early eleventh century, probably constructed according to a single design, and drew heavily on motifs from the architecture of the former Carolingian Empire.

14. The abbey church of Saint-Etienne of Caen: (a) The nave and crossing; (b) The west front. The nave and the crossing illustrate how far Saint-Etienne is an innovatory building, in all probability built according to a grander and more unique design after the conquest of 1066. The east end may well have only been in construction at the end of William’s life and was the subject of major changes in the thirteenth century.

15. The gatehouse of the castle enclosure at Exeter was built in the years soon after the 1068 siege. The mixture of stylistic influences, which include triangular-headed arches and cushion capitals typical of late Anglo-Saxon architecture, indicates newcomers and natives working on its design. The original entrance was blocked in later centuries and a new one created alongside it.

16. (a) The White Tower of London; (b) The remains of the keep at Ivry-la-Bataille in Normandy. The illustration of the White Tower shows one of the greatest secular buildings to survive that can be securely dated to William’s time. The apse is visible, which is closely linked to the one at Ivry, as are the blind arches at firstfloor level, which resemble those at Saint-Etienne of Caen. Although the remains at Ivry are fragmentary, the apse and the ground plan show that it was, in significant respects, likely to have influenced the construction of the White Tower.

17. The surviving remains of the west front of Bishop Remigius’s cathedral at Lincoln. There are several points of comparison in its construction and style that suggest parallels with the Arch of Constantine in Rome, notably the existence of the frieze.

18. An illuminated initial in a late twelfth-century manuscript showing William sending a letter to Abbot St Hugh of Cluny. The manuscript contains Abbot Hugh’s letter refusing William’s request for some of his monks to be sent to England.

19. A diploma recording the restitution of Gisors to Rouen cathedral by Simon, count of Amiens, Valois and the Vexin, in the presence of Matilda, Roger de Beaumont and other nobles. The crosses of William and Matilda’s son Robert and of Roger de Beaumont are placed below the text that records the ceremony.

20. Two of the great original pancartes recording property acquired by the abbey of Saint-Etienne of Caen. Although the earlier left-hand text is damaged, close reading shows that the interlineations and erasures were part of an editorial process that produced the second right-hand document.

21. Aerial view of the site north of Sainte-Suzanne known as the Camp de Beugy, which was, in some form, probably used by William to conduct the siege of Sainte-Suzanne. Ongoing archaeological excavations currently suggest that the site was already an aristocratic residence in the tenth century, but with major developments taking place in the twelfth century.

22. (a) The bottom of the right-hand column of one of the folios of the section of Great Domesday Book devoted to Bedfordshire that includes the return for part of Sharnbrook; (b) The bottom of the right-hand column of the reverse of the same folio that has, at its foot, an addition to the main text mentioning the English priest who performed a mass every Monday for William and Matilda’s souls; (c) The colophon of Little Domesday Book that dates the carrying out of the descriptio to 1086.

xx

PROLOGUE

WRITI N G A L I F E O F W ILLIAM T HE C O N QUE ROR I only met David Douglas once, when, in 1970, he was the external examiner for my PhD thesis. Incomprehensible as it may be to members of a modern generation accustomed to almost immediate feedback on a PhD, it was not until 2013, when working on the late Frank Barlow’s papers, that I learnt that Douglas thought highly of my work.1 Many who knew him better than I did regard Douglas as having been an inspiring personality.2 Published in 1964, his William the Conqueror was the first book to appear in the Yale English Monarchs series. He was also General Editor of the series. To revisit his hugely successful book while writing this one has been a thought-provoking experience. Although it is arguably unsurprising that Yale University Press should have commissioned a replacement over thirty years after the initial publication of Douglas’s William the Conqueror, for all the changes in the study and interpretation of the past that have taken place over the last half-century, much about the way in which he approached his subject remains relevant. Douglas was acutely aware of the difficulties and the responsibilities involved in writing about William the Conqueror. To re-read the Prologue to his book is not only to be reminded of this, but also to be struck by just how pertinent some of his observations remain in 2016. Thus, to cite two passages: The posthumous career of William the Conqueror in controversial literature is almost as remarkable as his actual career in eleventh-century history.3 1 Frank Barlow to Andrew MacLennan, 8 October 1974 (‘The thesis was examined by David Douglas, the doyen of Anglo-Norman scholars, who thought very well of it’). The exchange of letters is now among my own papers. 2 For David Douglas’s life and career, R. H. C. Davis, ‘David Charles Douglas, 1898–1982’, PBA, lxix (1983), 513–42. My comment is also based on several personal conversations. For Douglas’s work on the Normans, see also M. T. Clanchy, ‘Introduction’, in David C. Douglas, The Normans (London, 2002), xi–xix. This book contains reprints of David C. Douglas, The Norman Achievement 1050–1100 (London, 1969) and The Norman Fate, 1100–1154 (London, 1976). 3 Douglas, WC, 5.

1

2

william the conqueror

Briefly, my aims have been to eschew the controversies of the past; to bring French and English scholarship here into closer relation; and throughout to base my study upon the original testimony . . .4 These two statements might almost feature verbatim as the mission statement for this book. To bring English and French scholarship into closer relation remains as necessary and admirable an aim as it has always been, even if collaboration between scholars on the two sides of the Channel has become much more extensive than it was in Douglas’s day. There has also been much reflection since that time on how national historiographies have been, and arguably still remain, a barrier to understanding.5 It is nonetheless very important how the scholarly study of early medieval rule has internationalized, with, above all, the long-running debates initiated by Gerd Althoff ’s Spielregeln about the nature of communication between rulers and ruled having crossed national boundaries.6 There are opportunities here to be exploited. In relation to the second of Douglas’s aims, to base a book about William on ‘the original testimony’ is nowadays a complex aspiration. What exactly might ‘the original testimony’ mean? While it is certainly something that has the verifiable potential to inform us factually or interpretatively about William and those around him, exactly how this material is to be used is not at all straightforward. Those who were writing about William during his lifetime were aware that they were living in momentous times. Stories about him were being invented while he was alive. The literary productivity that followed immediately after the victory of 1066 presents problems analogous to those tackled by contemporary historians of our own times – or indeed of any times. William divided opinion in his moment of triumph, just as he does now. The notion that the influences of the times in which a historian lives are inescapably important for one writing about William the Conqueror, with the historian often not consciously aware of what these influences are, or have been, applies as much to those writing before William’s death on 9 September 1087 as to all those who have come afterwards. And in the same way that the ‘original testimony’ is difficult to

4

Ibid., xiii. This is epitomized by the publications that have followed from the CNRS ‘France-ÎlesBritanniques’ projects. For an interesting attempt at dialogue and synthesis, Christopher Fletcher, Jean-Philippe Genet, and John Watts (eds.), Government and Political Life in England and France, c.1300–c.1500 (Cambridge, 2015). 6 Gerd Althoff, Spielregeln der Politik im Mittelalter: Kommunikation in Frieden und Fehde (Darmstadt, 1997). The relevance of these debates to my subject will be explained later in this Prologue and throughout the book. For a notably perceptive early review, see Timothy Reuter in Bulletin of the German Historical Institute, xxiii (2001), 407. 5

prologue

3

analyze in relation to William, so it is also difficult in relation to those whom he defeated. For all who wrote about this subject and, very specifically, about King Harold and Edgar the Ætheling, God’s judgement on them required moral and ethical perspectives that distort every bit as much as they do in relation to William.7 As will be suggested, above all in Chapters 6 and 7, while factual accuracy most definitely is present in this material, we certainly do not find objectivity in it. So viewed, the controversies of the past, the present, and indeed of the future, are inescapable aspects of the topic, a situation that has the most profound implications for the way in which we interpret ‘the original testimony’. William’s life has always been, and will forever remain, a morally difficult subject. It brought about massive changes. But it did so at the expense of thousands of lives and it was the cause of much misery. The ethical implications of this were a problem in the 1060s and 1070s, let alone today. As clear an expression as possible of the issues that a biographer of William the Conqueror faces can be found in the book written by FrançoisThéodore Licquet, published posthumously in 1835: But then, alas! When a more philosophical line of thought leads us to the examination of means and of results, our admiration falls, and cold seizes us. William frightens us, and, we are, in sum, almost reduced to lamenting the good that he did.8 Licquet’s words lie within a long tradition. When, for example, Orderic Vitalis, one of the writers on whom we are heavily dependent, set out to write about William in the early twelfth century, the knowledge that William had successfully subdued the English created the paradox that, while his triumph must have been pleasing to God, it entailed events which Orderic viewed as being a human catastrophe, ‘a mournful theme of ruin for the pen of true historians’.9 In the face of a situation in which a multiplicity of – often very extreme – opinions have been expressed, no biographer’s personal beliefs can be entirely put to one side. All we can

7 For a recent essay on this theme, Mary Frances Giandrea, Episcopal Culture in Late AngloSaxon England (Woodbridge, 2007), 7–34. 8 Georges-Bernard Depping, Histoire de la Normandie, depuis les temps les plus reculés jusqu’à la conquête de l’Angleterre en 1066, 2 vols. (Rouen, 1835), ii, 173–4 (‘Eh bien! Que la pensée philosophique nous ramène à l’examen des moyens et des résultats, l’admiration tombe, le froid nous saisit; Guillaume nous fait peur, et nous sommes presque réduits à déplorer le bien qu’il a fait’); cited in Véronique Gazeau, ‘Imagining the Conqueror: The Changing Image of William the Conqueror, 1830–1945’, HSJ, xxv (2014, for 2013), 245–64, at 249. I have used Professor Gazeau’s translation. 9 OV, ii, 190–1 (unde flebile tema de sua ruina piis historiographis ad dictandum tribuit).

4

william the conqueror

do – as David Douglas did – is try to be fair.10 And find a way of doing so. We must remember that the ethical dimensions inherent in the study of the past mattered to Douglas, just as they did to William’s contemporaries and to those who wrote about him in the twelfth century, and as they should to us. In the selection of his essays that he chose for publication as he approached his eightieth birthday, Douglas began by thanking his school history master (who later became bishop of Coventry) for introducing him to these ethical dimensions.11 The schoolmaster/bishop whom Douglas did not name was Neville Gorton, bishop of Coventry from 1943 until his death in 1955. The date of his appointment says most that needs to be said about him to the author of this book, who was born in the city towards the end of the Second World War, and introduced to the charred cross in the ruined cathedral as a young child.12 In the end, for all who wrote about him during his lifetime and afterwards, including the unknown mastermind of Domesday Book, William’s remarkable career had to be seen as the result of a providentially destined purpose and, in England after 1066, the basis in which current legitimacy and law were rooted. His achievements were therefore deeply embedded in the sometimes incompatible belief systems of the medieval Church and of secular society. While the writers of histories could interpret very differently the personality and the actions of the man at the centre of events, one question was unavoidable. Why had he been allowed to win a decisive battle against a crowned king and subject a prosperous Christian population to his will and rule? And how had this been done? If the end had to be accepted, the means were open for debate. Moral and interpretative issues of this kind were present even in the way William’s birth was treated. For two of the major twelfth-century historians, Orderic Vitalis and Wace, the fact that he was born of parents who were not technically married was the cause of great personal and political insecurity. For a third, William of Malmesbury, who, like them, had read all the available eleventh-century narrative sources, it seems not to have mattered. Others injected motifs from saints’ lives to help them understand the birth of a child whose life was going to be so remarkable. And when we come to what was written after 1066 about the conquest of England, the contrast between the contemporary apologia written by William’s chaplain William of Poitiers and the sustained critique in the equally contemporary ‘D’ version of the Anglo-Saxon Chronicle is very 10 Douglas’s fairness was commented on by Frank Barlow in his Foreword to the 1999 Yale Edition of William the Conqueror, xi (‘Douglas was very fair-minded’). 11 David C. Douglas, The Time and the Hour (London, 1977), 9. 12 See further Gorton’s obituary in The Times (1 December 1955), 14.

prologue

5

striking indeed. The authors’ individual biographies were also a factor in the way they wrote. Malmesbury, a clever man massively learned in the Latin classics and with an interest in biography, was often very insightful, yet also capable of adjusting situations to fit classical and biblical models in a way that misleads. And Orderic never forgot that he himself was angligena, born of an English mother (and a French father). Domesday Book, it must be added, was certainly not an objective census.13 For the anonymous monk who wrote the famous homily in the ‘E’ version of the Anglo-Saxon Chronicle, William’s life was one on which all ought to reflect and from which they could learn. It contained a bewildering mix of good and bad. In reading how such writers tried to deal with these matters, we may sometimes think that they are laying bare William’s own anxieties. In the sources written during William’s lifetime and in those slightly later ones that had access to material transmitted directly by those who had witnessed events or through social memory, we can sometimes think that we are listening to the instruction that William was receiving. Thus, ‘the original testimony’, when used carefully, can become a means to try to explore William’s private thoughts. But in thinking in this way we must be aware that, in terms of the history of power and the responsibilities of rulership, several differing value systems were operating, and that nothing should be simplified. Although the first phase of Douglas’s professional career was located within the English social history tradition associated with his doctoral supervisor Sir Paul Vinogradoff and with Sir Frank Stenton, he consistently believed that all the more important developments in England’s history derived from the Latin tradition going back to Rome, rather than from the Germanic and Anglo-Saxon ones. This belief, very evident in Douglas’s teaching in his first academic post, at the University of Glasgow, probably went back to his undergraduate days.14 It placed him at odds with some of his contemporaries and it gave his work a quality that was very different from theirs. The depth of this commitment to seeing William’s life and the general history of the Normans through continental, as well as English, sources is epitomized by Douglas’s acquisition of French-language publications for his personal library and the libraries of the universities he served. The story of how a large consignment of printed cartularies, which he had persuaded the University of Leeds to purchase, left Rotterdam in 1940 just before Hitler’s army captured the city, to be incorporated into the Brotherton Library, is justly famous. The collections of the library of his 13 Most recently and very effectively, Sally Harvey, Domesday: Book of Judgement (Oxford, 2014). 14 Davis, ‘David Charles Douglas’, especially at 513–14, 519.

6

william the conqueror

fifth and final university, Bristol, were also enriched by similar endeavours. Douglas’s personal library, of which the French-language publications are now located at the Institute of Historical Research, also demonstrates the remarkable scale of his efforts.15 Understanding the history of the history of the Norman Conquest was his main preoccupation in the 1930s and 1940s. Douglas published so extensively on its anglophone historiography as to cause anxiety among his contemporaries that this was preventing him from doing the serious work on the history of Normandy and the Normans that they wanted him to do.16 Because awareness of the controversies of the past is an indispensable aspect of the subject of his book and mine, his publications should be consulted as a preliminary by all who take the subject seriously. But in the midst of this work Douglas did evidently become exasperated with the pernicious influence of polemic and unhistorical assessments. In a lecture given in 1946, a cutting reference to ‘a haze of heated polemical vapour’ was followed by an unacknowledged quotation from Ezekiel (‘The fathers had eaten sour grapes and the children’s teeth were set on edge’).17 Douglas’s abandonment of historiography was followed in the 1940s by a remarkable series of articles on the duchy of Normandy and by the preparation of a book entitled William the Conqueror and the Rise of Normandy. Draft chapters were dispatched to Eyre & Spottiswoode in 1958 and a contract was sent out to him on 13 September of that year. Through a process that cannot now be traced in any detail, this project evolved into William the Conqueror, with the invitation to give the 1963 Ford Lectures clearly being the catalyst for the decision to set his subject in a broader context. One significant early change of direction within this process was caused by an awareness that Marie Fauroux was at work on her École des Chartes thesis devoted to an edition of the pre-1066 charters of the Norman dukes, completed in 1950 and published in 1961, itself a major landmark for the writer of a biography of William the Conqueror. This led Douglas to abandon his own efforts in that direction. He turned instead to major interpretative works on the history of the Normans, of which William the Conqueror was the first. The Norman Achievement and The Norman Fate followed.

15 For the bequest of Douglas’s personal library to Keble College, Oxford, and the transfer of the French books to the IHR, www.history.ac.uk/library/bequests/douglas. 16 Davis, ‘David Charles Douglas’, 523. 17 David C. Douglas, The Norman Conquest and British Historians (Glasgow, 1946), with the quotation at 8–9. See also idem, English Scholars (London, 1951). The former was reprinted among Douglas’s selection of his essays, The Time and the Hour, 57–76, with the quotation at p. 58.

prologue

7

Douglas was once moved to speculate why Shakespeare did not write a play about the Norman Conquest.18 Taking into account the parallels between the lives of Edward the Confessor and Hamlet, prince of Denmark, and also the manifest relevance of King Lear, it has often occurred to me while writing this book that Shakespeare might have chosen Hamlet and Lear because their stories were rather less complicated to stage than Edward’s and William’s. With both Edward and William, so much was taking place within their minds and the minds of others that the roots of the drama are ultimately unknowable. The thread that runs through everything is Edward and William’s personal relationship, a story in which nothing discernible happened for years on end, yet which was constantly ongoing; that they apparently met only once between 1041 and 1066 should give pause for thought. And, in relation to other actors, the thoughts of Harold Godwineson, earl of Wessex and, in 1066, king of the English, and Ealdred, bishop of Worcester and then archbishop of York, are also of central importance. It can feel as if the biographer’s task is to make visible the invisible, and to create a context that makes the narrative that emerges plausible. Of course Shakespeare’s reasons for not writing about Edward and William may have lain elsewhere. A play that involved a disputed succession and a successful invasion from France sanctioned by the pope would certainly not have gone down well with either of the two English monarchs who reigned during his lifetime. The difficulties of reconstructing the hidden Edward/William narrative are also central to the relationship between Douglas and his younger nearcontemporary Frank Barlow, whose Edward the Confessor was added to the English Monarchs series at Douglas’s invitation. Barlow signed the contract on 20 July 1965 and delivered the manuscript on 15 April 1969. In Edward, he thanked Douglas for his generosity in asking him to write the book and acknowledged that ‘its coolness towards the Norman party might even cause him pain’. That it did so appears manifest in a passage in Douglas’s otherwise wholly positive report on the book, which is preserved both in the English Monarchs series archive and among Frank Barlow’s papers: I happen to disagree personally with some of his interpretations of the evidence. That does not matter a bit. But I think Barlow might get more out of the Norman sources without relinquishing his dislike and distrust of everything Norman.

18 David C. Douglas, ‘The Norman Conquest and British Historians’, x (The Time and the Hour, 59).

8

william the conqueror

To someone who knew Frank Barlow well, and whose writing of Edward the Confessor coincided almost exactly with the period when he was teaching me the Norman Conquest Special Subject and was my PhD supervisor, it feels as if Douglas was being mischievous as well as critical; there is a strong element of parody in the comment. Acquaintance with Frank Barlow’s publications rapidly shows that he did not dislike everything Norman. He did, however, dislike William the Conqueror and he profoundly distrusted William of Poitiers. The inter-relationship between William and Edward’s lives, whatever form it actually took, means that Frank Barlow’s Edward is every bit as important an element in the background to this book as Douglas’s William. In addition to their differences of opinion about William and the available sources, the intellectual core of the difficulties between the two of them also resided in another statement in Douglas’s Prologue to the effect that the ‘transference of power and influence’ associated with the so-called Norman achievement ‘was a prime factor in the making of Europe’.19 With Frank Barlow’s The English Church, 1000–1066, first published in 1963, constituting in its day a major contribution to the rehabilitation of the late Old English state, we are back to an extent with the extravagant claims that were still being made in those days by so-called Normanists and Romanists on the one side and by so-called Germanists and AngloSaxonists on the other. Within those debates a sophisticated argument that has elements in common with Douglas’s approach appears in the work of another younger contemporary, Sir Richard Southern, except that the Normans become a conduit for French civilization. The title of the essay based on his 1966 Creighton Lecture, ‘England’s First Entry into Europe’, says most that needs to be said for now, except to notice that the resilience and voices of the defeated play a more prominent part in his work than they did in Douglas’s, and as, I hope, they do in this one.20 Given that my book draws on francophone and continental European scholarship in a way of which I hope Douglas would have approved, it is more than somewhat ironic that at the time of the publication of Douglas’s William perceptions of the early history of Normandy were being radically changed in Normandy itself, above all by the publications of Lucien Musset, as well as by those of Michel de Boüard and Jean Yver. Their great achievement was to create a different and more meaningful evolutionary framework of continuity and change from Normandy’s Carolingian past. In consequence, Musset thought Douglas’s approach to early Norman Douglas, WC, 7. R. W. Southern, ‘England’s First Entry into Europe’, in Medieval Humanism and Other Studies (Oxford, 1970), 135–57. 19 20

prologue

9

history illogical, mostly because Douglas saw it as a preparation for eleventh-century greatness, rather than as a complex process of assimilation and adjustment.21 Musset had also read many more unpublished Norman charters than Douglas. Yet Douglas was greatly honoured in France, receiving a doctorate honoris causa from the University of Caen in 1957. It was to do justice to this Norman historiography, to seek out charters, and ultimately, guided by then novel approaches to identity and power, to set ‘The Rise of Normandy’ into a very different European perspective from Douglas’s, that I went to France in 1976 and 1978 for two archival visits which laid many of the foundations for my own career. A major result of the work done by Musset, his contemporaries, and those who followed, including myself, is that although points of detail and most especially the degree of integration of Lower and Upper Normandy will have to be discussed, we can start from the assumption that the social, cultural, and political resilience of the later tenth- and early eleventh-century Norman principality can be taken for granted. No one nowadays could surely subscribe to Douglas’s statement that the conquest of England ‘came from a province which forty years before the Norman Conquest of England showed few signs of its future achievement’.22 Reflections on this state of affairs appear throughout this book, mostly in Chapters 5 and 13. It would appear that Douglas found it difficult to engage with William. Towards the end of the Epilogue of William the Conqueror, he remarked that ‘He remains then something of an enigma: admirable; unlovable; dominant; distinct.’ Slightly earlier he noted that ‘There was in fact an element of paradox in his character.’23 That the paragraphs around these sentences convey very eloquently the predominant characteristics of the man makes the statements all the more surprising. They also suggest that Douglas could not in the end bring himself to like William (‘the personal portrait which emerges is undoubtedly repellent’). There is indeed here an expressed aversion to the exercise of violent and arbitrary power that runs through Douglas’s writings. In a notable piece of self-plagiarism, for example, he writes of William, Robert Guiscard, Count Roger of Sicily, and Bohemond that ‘they were all men of great ability and vastly 21 Lucien Musset, ‘Gouvernés et gouvernants dans le monde scandinave et dans le monde normand (XIe–XIIe siècle)’, in Nordica et Normannica. Recueil d’études sur la Scandinavie ancienne et médiévale, les expéditions des Vikings et la fondation de la Normandie (Paris, 1976), 426 (‘Posé sous cette forme chronologique, le problème nous paraît insoluble et même illusoire’). 22 Douglas, WC, 7. 23 Ibid., 375, 376.

10

william the conqueror

ambitious’ and that ‘all were in varying degrees personally repellent, cruel and coldly unscrupulous’.24 As someone who also finds violence and unchecked power unattractive, if not reprehensible, I find myself in sympathy with Douglas. Yet at the same time I do not think that there is either enigma or paradox in William’s character. It is certainly difficult to penetrate his hard outer shell, especially since, when it comes to the exercise of power in a world where the ruler has a central autocratic place, the distinction between the public and the private will always be blurred. But it can be done, and a coherent interpretation can be created. William ultimately belongs within a world into which violence – sometimes extreme violence – was culturally structured and in which the succession to kingdoms, principalities, and lordships was shaped by norms rather than rules. Frank Barlow and David Douglas both regarded the writing of history as a branch of literature. But their approaches to biography differed radically: Barlow was interested in biography in ways in which Douglas was not. Barlow certainly believed that it was possible to understand William’s personality and his psychological development. In a passage in the midst of several pages of remarkable prose, he wrote that William ‘learned to survive in a society riddled with intrigue and deceit. His experiences made him a man who believed in guile and gave his confidence to few whom he had not known long and tested extensively.’25 In the same book Barlow wrote: ‘It is a common fallacy . . . that notable achievements must have involved notable men.’26 All this is as may be, but he was certainly able to engage with the subject of his biographies. His remarkable contribution to the writing of biography – and not just medieval biography – has to be crucial to the way I try to write about William, even if the portrait of William that emerges is different from the one he would have painted.27 Barlow’s apparent belief that William was less accessible to a biographer than either of his two sons, Robert Curthose and William Rufus, is not one I share; charters and other narratives provide anecdotal material that has not been fully exploited.28 For reasons identical to Douglas’s, historiography will feature scarcely at all in this book. It pays homage to the masters of the past and is written in full awareness of just how important their contribution was. It does from Douglas, The Norman Achievement, 6. Frank Barlow, William I and the Norman Conquest (London, 1965), 12. 26 Ibid., xvi. 27 David Bates, Julia Crick, and Sarah Hamilton (eds.), Writing Medieval Biography, 750–1250: Essays in Honour of Professor Frank Barlow (Woodbridge, 2006), vii–x, 1–13; David Bates, ‘Frank Barlow, 1911–2009’, PBA, clxxii (2011), 3–24, at 16–17, 21–3. 28 Thus, in the Preface to the 1999 Yale edition of Douglas’s William the Conqueror, x. 24 25

prologue

11

time to time refer back to Douglas to foreground changes in the general approach that are needed to explain my argument. It is of course also written in full awareness of just how much generations before Douglas’s contributed to the subject. Ultimately, however, it is deliberately only occasionally located within a vertical historiographical dimension, but instead is set within a horizontal one, that is, within a modern framework that draws on anthropology and other social sciences, as well as on the latest work on the primary sources. Its intellectual framework is as much of the vast literature devoted to culture, social norms, gender, power, trust (sociologically defined), and identity as I have been able to master.29 Taken together with a rigorous approach to chronology, the careful use of many excellent new editions of the main primary sources, and engagement as far as is possible for me with art and architectural history, archaeology, liturgy, medical history, and numismatics, there is, I think, a way to get somewhere near knowing the man. It has also come as something of a surprise while writing this book to discover how many apparently accepted dates in William’s life require re-examination. With the exception of Douglas’s book and Michel de Boüard’s important 1984 biography, I have not referred extensively to the many treatments of William’s life that have appeared since 1964, except where they have made a significant contribution to the understanding of a particular point. These include one that I wrote myself.30 In the end, the central problems that we must try to unravel are to understand what it was that made William so uncompromisingly convinced that he had a right to succeed Edward the Confessor; what it was that convinced

29 For an essay that admirably sets out the principles of such an approach, Pauline Stafford, ‘Writing the Biography of Eleventh-Century Queens’, in Bates, Crick, and Hamilton (eds.), Writing Medieval Biography, 99–109, with a succinct statement at 109. 30 Michel de Boüard, Guillaume le Conquérant (Paris, 1984). Among the many books about William that have appeared since Douglas’s book was published are Paul Zumthor, Guillaume le Conquérant (Paris, 1964; repr., 2003); David Bates, William the Conqueror (London, 1989; repr. Stroud, 2001); Gilles Henry, Guillaume le Conquérant (Paris, 1996); Pierre Bouet, Guillaume le Conquérant et les Normands au XIe siècle (Condé-sur-Noireau, 2000); Stéphane W. Gonduin, Guillaume Roi (Condé-sur-Noireau, 2002); Philippe Maurice, Guillaume le Conquérant (Paris, 2002); Alain Hervé, Guillaume le Conquérant (Paris, 2003); Henri Suhamy, Guillaume le Conquérant (Paris, 2008); Yann Coz, Guillaume le Conquérant (Paris, 2011); Peter Rex, William the Conqueror: The Bastard of Normandy (Stroud, 2011); Stéphane W. Gonduin, Guillaume le Conquérant et les femmes (Cahors, 2012); Mark Hagger, William: King and Conqueror (London and New York, 2012); François Neveux and Claire Ruelle, Guillaume le Conquérant. Le bâtard qui s’empara de l’Angleterre (Rennes, 2013); Gilduin Davy, Guillaume le Conquérant. Le bâtard de Normandie (Paris, 2014). A French publisher also thought it worth reprinting in 1993, unaltered, the 1861 edition of Louis-Phocion Todière, Guillaume le Conquérant (Tours, 1856; 2nd edn., Tours, 1861; Paris, 1993). William’s wife Matilda has been the subject of one recent book: Tracy Borman, Matilda: Queen of the Conqueror (London, 2011).

12

william the conqueror

so many people to share this conviction and join him in 1066 and afterwards in a very risky enterprise; why many people feared, respected, and even liked him, and were prepared to follow him loyally; and why many others disliked him so intensely that they were prepared to fight him when the odds against such resistance succeeding were extremely long. Moreover, it is important to understand what it was about him that so profoundly shaped the history of nations and people after the conquest of 1066. What was it about William the Conqueror that caused events to turn out as they did? Why was the outcome one that produces remarkable statements from historians of a comparative bent, like John Gillingham’s characterization of the process that was to decimate the landholding elite of pre-1066 England as ‘an event unparalleled in European history’?31 Equally remarkable is Peter Heather’s description of the Norman Conquest of England as ‘small-scale elite transfer’, a valuable comparative assessment because of the contrast he draws with the migrations of the previous millennium and the absence of disruption to existing structures in the conquered lands because of the overall control that was maintained.32 Both pose the questions of what was different about William and what made his achievement distinctive. And, in doing so, they underline again the importance of the genre of biography. To attempt to make sense of William’s life and give it the requisite broad context, we need to find allies in places where earlier historians of William’s life have not much ventured, namely in, for example, the histories of power, ritual, feud, socially constructed violence, and trust. It is here that the socio-cultural history of the last four decades must come to the fore, with trust – sociologically defined – being in many ways the cement that binds.33 Social memory, with its emphasis on the capacity to make factual claims about past events that have no objective reality, can also be a good friend.34 The model of early medieval warrior kingship, with its emphasis not just on military leadership, but on the capacity to distribute rewards and to create an ambience of legitimacy and confidence, supplies a convincing 31 John Gillingham, ‘Problems of Integration within the Lands Ruled by the Norman and Angevin Kings of England’, in Werner Maleczek (ed.), Fragen der Politischen Integration in Mittelalterlichen Europa, Vorträge und Forschungen, vol. lxiii (Ostfildern, 2005), 85–135, at p. 85. 32 Peter Heather, Empires and Barbarians: Migration, Development and the Birth of Europe (London, 2009), 23, 298–9, 327, 350–1, 614. 33 For trust, see Geoffrey Hosking, Trust: A History (Oxford, 2014), with the general issues set out at pp. 1–8. A pioneering essay on the Middle Ages does not really venture into the areas that I think are crucial: Susan Reynolds, ‘Trust in Medieval Society and Politics’, in idem, The Middle Ages without Feudalism: Essays in Criticism and Comparison on the Medieval West (Farnham, 2012), chapter 13. 34 A convenient survey for historians is supplied by James Fentress and Chris Wickham, Social Memory (Oxford, 1992), 20–31.

prologue

13

framework; it is here, of course, without underestimating the complexities of the sociology, that trust and performance are central to understanding the conscious and unconscious thought processes that made William a convincing ruler and war leader for so many of his contemporaries.35 His capacity to instil fear, to gain respect, and to maintain power and authority in the novel circumstances created by conquest, has parallels in the preceding centuries, Charlemagne being the obvious comparator.36 The complementary, if at first acquaintance seemingly contradictory, attributes of rule – ‘anger’ (ira) and ‘mercy’ (misericordia) – that were at the heart of an individual ruler’s practice of power in the eleventh and earlier centuries have both an appearance of arbitrariness and at the same time can derive an inherent, deeply moral, sense of responsibility; their relationship within any ruler’s actions is crucial to understanding both his (or occasionally her) personality and the impact they had in the short, medium, and long term. Although we lack the literature devoted to the uncertainties of life at court that exists for the Carolingian and the Angevin periods, we can be certain that William’s court was a place where favour was dispensed apparently arbitrarily, but according to reasons that, while normally hidden from us, would have been well understood by William and his contemporaries.37 That much of what he did was communicated through the theatrical performances that rule required must be borne in mind. A physical and metaphorical barrier always existed within the palace and around the man when he was on the road from one place to another. Yet at the same time the choices that shaped performance and policy can often be surmised. All rulers operated within the practice of consultation and abstract notions of right and wrong that placed even the most formidable of them within processes that acknowledged norms and rules.38 Kings were both 35 For valuable general statements, Timothy Reuter, ‘Plunder and Tribute in the Carolingian Empire’, TRHS, 5th ser., xxxv (1985), 75–94 (reprinted in Medieval Polities and Modern Mentalities, ed. Janet L. Nelson [Cambridge, 2006], chapter 13); cf. Janet L. Nelson, Charlemagne and the Paradoxes of Power, The Reuter Lecture 2005 (Southampton, 2006), 4–11; Guy Halsall, Warfare and Society in the Barbarian West, 450–900 (London, 2003), 1–39. 36 See recently, Jennifer R. Davis, Charlemagne’s Practice of Empire (Cambridge, 2015), 128–66. 37 For an essay on this theme, with multiple references, David Bates, ‘Anger, Emotion and a Biography of William the Conqueror’, in Janet L. Nelson, Susan Reynolds, and Susan M. Johns (eds.), Gender and Historiography: Studies in the Earlier Middle Ages in Honour of Pauline Stafford (London, 2012), 21–33. 38 Thus, with references to many other publications, Timothy Reuter, ‘Assembly Politics in Western Europe from the 8th Century to the 12th’, in Peter Linehan and Janet L. Nelson (eds.), The Medieval World (London, 2001), 423–50; Susan Reynolds, ‘Assembly Government and Assembly Law’, in Nelson, Reynolds, and Johns (eds.), Gender and Historiography, 191–9. Levi Roach, Kingship and Consent in Anglo-Saxon England, 871–978: Assemblies and the State in the Early Middle Ages (Cambridge, 2013), deals with all these issues in relation to England before 1066 and breaks down barriers between national historiographies in an exceedingly helpful way.

14

william the conqueror

military leaders and the bringers of peace, an apparent paradox that is ultimately no paradox at all given the multifaceted nature of power and the confrontations and stresses it generated. In this context, there was nothing new in the way in which William and those around him set out to legitimate the conquests of Maine and England; violence was for them and others the only route by which peace and right could be assured, a point that of course applies to Harold as well.39 All who wrote about William and rulers like him shared the assumptions that while peace and justice were impossible in what they knew to be an imperfect world, they were best upheld by strong kingly rule buttressed by the support of a unified aristocracy, the moral leadership of bishops, and the prayers of monks. But at that time, just as now, what constituted good governance was a controversial subject on which irreconcilable opinions would inevitably be held.40 In terms of English history, the relationship of kingship and law that appears to be so much more articulately analyzed in the later twelfth century was actually an issue fundamentally present in and after 1066.41 For this reason William’s life was assessed in very different ways by contemporaries. It is these contrasts and contradictions that are at the heart of this book. For all that the norms, scripts, and rules of social power are crucial to understanding the life of William the Conqueror, it is only through the genre of biography that we can comprehend what made his career so remarkable. As a result, it is important to spell out that this is a book about William the Conqueror and not a book about the Norman Conquest. Hence, unlike Douglas’s book, it has no subtitle. This is partly because I want to put William in his place within a process and in his time; and partly because, while I think that the role of the individual in history manifestly does matter, I want to react strongly against the notion that William alone was the single cause of everything that happened. He was an agent within processes, not a deus ex machina. The ways in which he interacted with the prevailing cultures of his time and the people around him are crucial. In the end, this book has to consider the subjects of the Norman Conquest and William’s place within the history of Europe. Hence, some of the book’s final chapter and the Epilogue are devoted to them. But it does so on the 39 The general point is very well made in Karl Leyser, ‘Warfare in the Western European Middle Ages: The Moral Debate’, in Timothy Reuter (ed.), Communications and Power in Medieval Europe (London and Rio Grande, 1994), 189–203, with 1066 discussed at pp. 192–3. See also Paul J. E. Kershaw, Peaceful Kings: Peace, Power and the Early Medieval Political Imagination (Oxford, 2011), 1–28. 40 For a succinct statement, Sigbjørn Olsen Sønnesyn, ‘In vinea Sorech laborare: The Cultivation of Unity in Twelfth-Century Monastic Historiography’, ANS, xxxvi (2014), 167–87, at 177. 41 Again for a succinct statement, John Hudson, The Oxford History of the Laws of England, volume II: 871–1216 (Oxford, 2012), 844–7.

prologue

15

terms set out above. It is also written within a framework that sees England and Normandy inside a wider European framework of social and cultural change in which both had distinct identities, but within which both were deeply embedded. To place Normandy within such processes as they were then understood was the ambition of my Normandy before 1066, even if its approach to, among other things, the supposedly globalizing ‘feudal revolution’ is one to which, almost forty years later, I would no longer subscribe. Much writing on later Anglo-Saxon England over the past half-century has been constructed in precisely these terms, a situation that enables an analysis to be developed that takes full account of political cultures and of social and institutional distinctiveness.42 As a result – to look back at old historiographies and opinions that occasionally resurface – no one nowadays can possibly view the conquest of 1066 as one that rescued the English from barbaric isolation, as was once often done.43 So-called ‘English exceptionalism’ nonetheless remains a subject of fruitful ongoing debate and discussion.44 As also is the argument about whether England should be placed in ‘Outer Europe’, even if the coiner of the phrase also refers, with tantalizing use of inverted commas, to ‘ “Carolingian” England’.45 Both are unavoidable subjects for a biographer of William the Conqueror. Indeed, it is a central proposition of this book that neither subject can be left to historians of the so-called Early Middle Ages. It is hoped that this approach will enable this book to move beyond old controversies.46

42 For a recent commentary on the massive historiography, Conrad Leyser, ‘Introduction: England and the Continent’, in David Rollason, Conrad Leyser, and Hannah Williams (eds.), England and the Continent in the Tenth Century: Studies in Honour of Wilhelm Levison (1876– 1947) (Turnhout, 2010), 1–13. James Campbell, ‘Observations on English Government from the Tenth to the Twelfth Century’, TRHS, 5th ser., xxv (1975), 39–54 (repr., idem, Essays in Anglo-Saxon History [London, 1986], 155–70), remains a seminal landmark. 43 Gazeau, ‘Imagining the Conqueror’, is an important essay illustrating this theme. 44 Most recently and very stimulatingly, George Molyneaux, The Formation of the English Kingdom in the Tenth Century (Oxford, 2015). 45 Chris Wickham, The Inheritance of Rome: A History of Europe from 400 to 1000 (London, 2009), chapters 19 and 20. 46 Thus, Tom McNeill, ‘Davison versus Brown, quarante ans après’, in Anne-Marie Flambard Héricher and Jacques Le Maho (eds.), Château, ville et pouvoir au Moyen Age, Publications du CRAHM (Caen, 2012), 41–50.

Chapter 1

T HE FI R S T YEARS

THE SON OF A CONCUBINE The fact that William the Conqueror’s parents, Robert, duke of Normandy from 1027 to 1035, and Herleva, were not married according to the rules of the Christian Church has over the years massively influenced interpretations of his life, his prospects, and his personality.1 It has been transmitted across the centuries into modern parlance with the nicknames ‘William the Bastard’ and ‘Guillaume le Bâtard’. Yet, with a host of recent publications having shown that Robert and Herleva’s relationship was of a type that was common and, for many, socially acceptable at the time of his birth, the whole subject is wide open for reconsideration. Young male members of the aristocracy often had a long-term partner with whom they did not go through a marriage ceremony, and whose status is for convenience frequently described as that of a concubine in the modern literature. Such a partner might in due course become a wife, be supplanted by a wife, or might form a ménage à trois with one. Many breaches of the Church’s expectation that heirs would be produced within a marriage sanctified by an ecclesiastical ceremony had occurred for centuries before William’s birth. They continued to occur up until the middle of the twelfth century.2 1 Thus, Douglas, WC, 31. (‘William, although to be in due course styled “the Conqueror” or “the Great”, was for his contemporaries emphatically “William the Bastard”. Nor is there any reason to suppose that during his infancy he was ever considered as a possible successor to the Norman duchy.’) In describing Herleva as Robert’s frilla (friedelehe) and in his general approach, Michel de Boüard, Guillaume le Conquérant, 84–92, did see William’s prospects in more positive terms. But cf. Ruth Mazo Karras, ‘The History of Marriage and the Myth of Friedelehe’, EME, xiv (2006), 119–51, with Robert and Herleva discussed at pp. 137–8. 2 See, for example, Stuart Airlie, ‘Private Bodies and the Body Politic in the Divorce Case of Lothar II’, Past and Present, 161 (1998), 3–38, at 14–20 (repr. in idem, Power and its Problems in Carolingian Europe (Farnham and Burlington, VT, 2012), chapter 10; Pauline Stafford, Queen Emma and Queen Edith (Oxford, 1997), 66–75; Régine Le Jan, Famille et pouvoir dans le monde franc (VIIe–Xe siècle): étude d’anthropologie sociale (Paris, 1995), 271–85; Megan McLaughlin, Sex, Gender, and Episcopal Authority in an Age of Reform, 1000–1122 (Cambridge, 2010), 16–31, 36–8; Ruth Mazo Karras, Unmarriages: Women, Men and Sexual Unions in the Middle Ages (Philadelphia, PA, 2012), with the specific case of Robert and Herleva discussed at pp. 42–3; Hudson, The Oxford History of the Laws of England, 226–37.

16

the first years

17

That William’s beginnings mattered to him and to his contemporaries is shown by there being no mention of Herleva by name as William’s mother before Orderic Vitalis in his early twelfth-century interpolations into the Gesta Normannorum Ducum, initially composed by William of Jumièges in the 1050s and 1060s, and subsequently continued by Orderic and Robert of Torigni.3 The absence of any direct reference to William’s mother in the two principal eleventh-century Norman histories of his life, Jumièges’ Gesta Normannorum Ducum and William of Poitiers’s Gesta Guillelmi, the latter written in the 1070s to justify William’s actions and the conquest of England, is certainly an indicator of sensitivities.4 Furthermore, the numerous contemporary and near-contemporary references to William as a bastard (nothus or bastardus) indicate that his birth remained a subject of comment across much of western Europe throughout his life, and for decades afterwards.5 However, only one historian made William’s birth into a significant political factor and converted it into a source of weakness, namely Orderic Vitalis, writing in Normandy in the first decades of the twelfth century. Although he mostly used the word nothus as a simple nickname or descriptor, Orderic was so keen to give it prominence that he included it in the rubric at the head of what was for him the sixth book of the Gesta Normannorum Ducum. He also on several occasions made William’s birth a justification for rebellion.6 It is manifestly wrong to give authoritative status to a single author writing almost a century after William’s birth, especially when, as we shall see, others saw the situation very differently. The starting point for any discussion of how Robert and Herleva’s contemporaries might have viewed William’s birth and prospects is the eulogy of Christian marriage in Dudo of Saint-Quentin’s Historia Normannorum, written in the late tenth and early eleventh centuries, and the earliest of the histories on which this book relies so much. Although her name was only included in later manuscripts, the eulogy was directed to Gunnor, the concubine who famously became 3 GND, ii, 96–7. A reference in a late eleventh-century charter, while making no mention of her being William’s mother, confirms that Herleva was indeed her name (Regesta, no. 158: Et pro sua anima et sue uxoris Herleve). 4 The beginning of the Gesta Guillelmi is lost, but since writers who later used it make no mention of William’s birth, this suggests strongly that no reference was made to it there. 5 For the assembled references, E. A. Freeman, The History of the Norman Conquest of England: Its Causes and Its Results, 2nd revised edn., 6 vols. (Oxford, 1867–79), ii, 172–4, 608; Elisabeth van Houts, ‘The Norman Conquest through European Eyes’, EHR, cx (1995), 832–53 (idem, History and Family Traditions in England and the Continent, 1000–1200 [Aldershot, 1999], chapter 8), passim. 6 GND, ii, 90 (Incipit de Willelmo notho ex Normannorum duce Anglorum rege effecto). For usage as a descriptor or nickname, OV, i, 158; ii, 2–3, 148–9; iii, 336–7; iv, 110–11; v, 26–7, 38–9, 158–9; vi, 166–7 (x2), 210–11. As a justification for rebellion, GND, ii, 96–7 (Roger de Tosny); OV, ii, 312–13 (1075); iii, 254–5 (William of Arques); iv, 82–3 (Guy of Brionne), 84–5 (William of Arques and Archbishop Malger).

18

william the conqueror

the wife of William’s great-grandfather Count/Duke Richard I (942–96), thereby exemplifying the Norman rulers’ determination to portray themselves as models of Christian legitimacy.7 Its very existence is to an extent indicative of how far Norman society had adjusted to the cultural norms of the medieval West since the time of the Viking settlements in the first half of the tenth century that had formed the foundations of the duchy of Normandy; concubinage remains present, but it is marriage that is idealized. The same motivation, this time directly expressed by the rulers themselves, lies behind the survival of the texts of the dowries given to their wives by Dukes Richard II (996–1026) and Richard III (1026–7).8 Presentation is, however, often far from reality. With Gunnor living on until 1031, and in her old age presumably meeting the child William, the continuing vibrancy of the norms of both marriage and concubinage was personalized at the ducal court. Furthermore, one of her children, Robert, archbishop of Rouen from 989 to 1037, who was both married and the holder of extensive lands around Evreux and in the Seine valley effectively as a lay member of the duchy’s secular elite, was Normandy’s leading churchman and a dominant figure in the duchy’s politics throughout the first decade of William’s life. The archbishop’s eminence must have kept this past even more alive for the boy’s contemporaries. As a result, at the time of his birth, no one could possibly have seen William’s prospects in the simplified terms that have often been suggested, even if, in the most recent generations, the normal pattern had been for a Norman duke both to have a concubine and to marry a wife drawn from the aristocratic elite of northern France. Thus, Duke Richard II had married Robert’s mother Judith, a sister of Geoffrey, count of Brittany (992–1008), and Duke Richard III had married Adela, who was a daughter of the French king Robert the Pious (996–1031); she later married Count Baldwin V of Flanders (1035–67).9 It was therefore very likely that Robert would contract a formal Christian marriage that might then produce a son, with the near-certainty that any such son would automatically be considered Robert’s successor as duke. However, it is also possible that Herleva would have had prospects of becoming Robert’s wife, 7 Dudo of Saint-Quentin, De Gestis Normannie Ducum seu de moribus et actis primorum Normanniae Ducum, ed. Jules Lair, Mémoires de la Société des Antiquaires de Normandie, vol. xxiii (Caen, 1865), 289–91 (Dudo of Saint-Quentin, History of the Normans, trans. Eric Christiansen [Woodbridge, 1998], 163–4). 8 RADN, nos. 11, 58. On all this, Pierre Bauduin, ‘En bon usage de la dos dans la Normandie ducale (Xe–début du XIIe siècle)’, in François Bougard, Laurent Feller, and Régine Le Jan (eds.), Dots et douaires dans le haut Moyen Age (Rome, 2002), 429–65, at 432–7. Also, Elisabeth van Houts, ‘Countess Gunnor of Normandy (c.950–1031)’, Collegium Medievale, xii (1999), 7–24, at 13–14. 9 For conclusive arguments that the Adela who married Richard III was King Robert’s daughter, Bauduin, ‘En bon usage de la dos’, 445.

the first years

19

if the precedents of Gunnor and, possibly, of Papia, who became Richard II’s second wife after Judith’s death in 1017, were followed.10 As a commentary on William’s beginnings, a contemporary of his early years who was well informed about Normandy, the Burgundian monk Ralph Glaber observed that, while Duke Robert had no son born of a Christian marriage to succeed him, it was a long-standing custom of the Normans to take their rulers from their princes’ unions with concubines. He described Robert’s nomination of William as his successor as a source of distress to the Normans, but then listed biblical and classical examples to provide historical parallels to show why he should succeed his father, as well as pointing out that the correct procedures to secure political legitimacy had been followed. Ralph’s close connection to William of Volpiano (died 1 January 1031), abbot from 1001 to 1028 of Richard I and Richard II’s great monastic foundation at Fécamp, and to his disciples, and therefore to the duchy’s religious elite, makes his perspective extremely important. Abbot William and those associated with him had been responsible for the renewal of monasticism in Normandy in Richard II’s time and had continued to have moral responsibilities that paralleled those of their religious equivalents elsewhere, in, for example, England and the Empire. William’s successor at Fécamp, John, abbot from 1028 to 1078, and others associated with them were influential not just in Normandy, but with the papacy and across much of western Europe, thereby giving Ralph’s opinions an almost official status. In short, while Ralph may have been excusing unusual practice on the grounds that the Normans were historically still relative newcomers to Christianity, he was also saying, his personal disapproval notwithstanding, that according to standards long prevalent, and in his day still prevalent, the young William should be regarded as Duke Robert’s legitimate son who had been correctly installed as his successor.11 While believing that William’s beginnings were not ideal, Ralph did not think him disqualified from becoming duke of the Normans.12 10 There is no contemporary evidence that Papia had been a concubine or that her sons, Malger and William, later to be archbishop of Rouen and count of Arques respectively, were the children of one. William of Malmesbury does, however, describe Count William as nothus, GR, i, 432–3. Her background as the member of a minor Upper Norman aristocratic family is at least suggestive. See further David C. Douglas, ‘Some Problems of Early Norman Chronology’, EHR, lxv (1950), 289–303, at 291–2. 11 Rodulfus Glaber Opera, ed. John France, Neithard Bulst, and Paul Reynolds, OMT (Oxford, 1989), 204–5 (Raoul Glaber: Histoires, ed. Mathieu Arnoux [Turnhout, 1996], 258–9). On this passage, Mathieu Arnoux, ‘Un observateur indulgent: Raoul le Glabre, les ducs de Normandie, leurs concubines et leurs enfants’, in François de Beaurepaire and Jean-Pierre Chaline (eds.), La Normandie vers l’An Mil, Société de l’Histoire de Normandie (Rouen, 2000), 88–92, at 88–9; Raoul Glaber, ed. Arnoux, 17–18. 12 Karras, Unmarriages, 42–3. For this argument developed somewhat differently, George Garnett, ‘ “Ducal” Succession in Early Normandy’, in George Garnett and John Hudson

20

william the conqueror

William’s very name is a further indicator of how he was viewed. As both Glaber and, almost a century later, William of Malmesbury, noticed, it was selected from the ducal family’s stock of names, and in particular in imitation of that of the child’s great-great grandfather, William Longsword, count of Rouen from c.923 to 942.13 The principal names of earlier Norman rulers were Richard and Robert, but with William being most definitely a name within the stock on which the ruling family drew, it shows that, from the beginning, the newly born child had a future as a member of Normandy’s aristocratic elite. That the name of Adelaide, possibly his full sister, was also drawn from this stock reinforces the point further; it was anticipated that she too would have an aristocratic future, through a marriage at an elevated social level within the network-forming alliances typical of contemporary politics.14 Astonishing as the coincidence might seem, a near-contemporary parallel to William’s parents’ relationship is that of Harold, in 1066 king of the English, and Edith Swan-Neck (probably also known as Edith the Fair). Harold had at least five children by her, only marrying Ealdgyth, sister of Earls Edwin and Morcar, sometime between 1063 and 1066.15 A second relevant parallel is the relationship of Cnut, king of the English (1016–35), with his ‘wife’ Emma, who was also the sister of Duke Richard II of Normandy, and his ‘concubine’ Ælfgifu of Northampton, the sons of both unions being so much considered as potential heirs that Harold Harefoot, son of Ælfgifu, and Harthacnut, son of Emma, succeeded one another as kings of the English after Cnut, albeit following a great deal of turbulence after Cnut’s death. Like William and Adelaide, Cnut’s two known sons by Ælfgifu drew on the pool of names used by the Danish kings, being named after Cnut’s father (Svein) and grandfather (Harold).16 Furthermore, relationships involving concubines were certainly not confined to the Scandinavian-influenced lands of the North Sea world in the eleventh century. Within the kingdom of France, for example,

(eds.), Law and Government in Medieval England and Normandy: Essays in Honour of Sir James Holt (Cambridge, 1994), 80–110, at 108. See also Trevor Rowley, The Man behind the Bayeux Tapestry: Odo, William the Conqueror’s Half-Brother (Stroud, 2013), 24–7. 13 GR, i, 426–7. 14 Kathleen Thompson, ‘Being the Ducal Sister: The Role of Adelaide of Aumale’, in David Crouch and Kathleen Thompson (eds.), Normandy and its Neighbours, 900–1250: Essays for David Bates (Turnhout, 2011), 61–76, at 63–4. Adelaide is first identified as a sister with the same mother as William in Robert of Torigni’s interpolations into the Gesta Normannorum Ducum, written before 1139, GND, ii, 270–1 (soror uterina Willelmi regis Anglorum). 15 For the parallel, Robin Fleming, ‘Harold II [Harold Godwineson] (1022/3–1066), king of England’, ODNB. 16 For Ælfgifu, Timothy Bolton, ‘Ælfgifu of Northampton: Cnut the Great’s Other Woman’, Nottingham Medieval Studies, li (2007), 247–68, and especially at pp. 253–8.

the first years

21

children who had been born of unions that were not technically marriages remained within the family’s inner circle and had prospects of inheriting family land.17 Whether they could inherit kingdoms and duchies was, however, the subject of much soul-searching.18 Yet, although a postConquest text possibly written during William’s reign as king, containing a translation of pre-Conquest English law, could state that an heir to the kingdom or an ætheling, that is, a princely member of the royal kindred, should be the son of a marriage, it was still possible several decades later for King Henry I’s son born outside marriage, Robert, earl of Gloucester, to be considered as a possible successor when it became clear that Henry’s only available heir born of one of his marriages was going to be female.19 In England, as in Normandy, there was a significant gap between the ideal and the actual. There is surely enough in all this to indicate that any notion of William having a socially disadvantaged childhood because of the circumstances of his birth is a total myth. He was not necessarily always destined to be Robert’s successor, and his actual succession was certainly the consequence of circumstances and of events working in his favour, the main one being his father’s death on pilgrimage in 1035 at a time when there was no possible serious alternative to him available. But, once robust good health was apparent, at the very least his future was certainly going to be that of a high-status aristocrat. To normalize William’s parentage and beginnings in this way is not to say that they did not affect his personality and his behaviour or how he was viewed by others. It is actually very clear that they did have a considerable effect, but not necessarily in ways that disadvantaged him politically and socially. To think otherwise is to be excessively influenced by Orderic’s moral perspective on William’s life. Moreover, somewhat provocatively, it is also surely the product of attitudes to birth outside wedlock predominant in the nineteenth century and for much of the twentieth century – attitudes that for the most part have only been swept aside over the past forty years, and which for many have remained 17 See, for example, Ruth Mazo Karras, Sexuality in Medieval Europe: Doing Unto Others, 2nd edn. (London and New York, 2012), 161–2; Amy Livingstone, Out of Love for My Kin: Aristocratic Family Life in the Lands of the Loire, 1000–1200 (Ithaca, NY, and London, 2010), 51–2, 214–15. 18 For England, Catherine Cubbitt, ‘Bishops and Succession Crises in Tenth- and Eleventh-Century England’, in Ludger Körntgen and Dominik Waßenhoven (eds.), Patterns of Episcopal Power: Bishops in Tenth- and Eleventh-Century Western Europe (Bayreuth, 2011), 111–26, at pp. 111–18. 19 Instituta Cnuti, in Die Gesetze der Angelsachsen, ed. Felix Liebermann, 3 vols. (Halle, 1903– 16), i, 521, at p. 521; Gesta Stephani, ed. and trans. K. R. Potter and R. H. C. Davis, OMT (Oxford, 1976), 12–13. For a proposed date in William’s reign for the Instituta Cnuti, Patrick Wormald, The Making of English Law: King Alfred to the Twelfth Century, volume 1: Legislation and its Limits (Oxford, 1999), 404–5.

22

william the conqueror

influential in the way William’s life is interpreted for longer than they should have done. That William was described as a bastard in sources contemporary with his life does not mean that the term had the same significance for everyone. It is also probable that the more successful William became, the more likely it was that this would influence what was written about him. The reference to his bastardy in the 1066 annal in the ‘D’ version of the Anglo-Saxon Chronicle probably does have a morally pejorative meaning, since, as we shall see, the writer’s treatment of the years immediately after 1066 must be read as a commentary on the ethical failings of William’s rule.20 Elsewhere, however, the usage is described specifically as being a nickname by Adam of Bremen, who was associated with one of William’s enemies, Svein Estrithsson, king of Denmark from 1047 to 1074.21 Furthermore, in the early twelfth century, the genealogist Lambert of Saint-Omer, writing in Flanders, while describing William as nothus, added that he obtained the kingdom of the English ‘by hereditary right’ (iure hereditario) because he was Edward’s kinsman. This again resoundingly underlined the point that children born of concubines had a place in the future planning of eleventhcentury aristocratic families and that, without any doubt whatsoever, William had in 1034–5 been legitimated for political purposes.22 Even Orderic’s attitudes were not entirely consistent, since, for example, in the twelfth-century case of Eustace de Breteuil he accepted that a son born of a concubine could inherit.23 And, for William of Malmesbury, William was entirely worthy of becoming king of the English, partly because of his good character and partly because he was the nearest of Edward’s kin by blood.24 For him, William’s parentage did not matter at all. The most complex issues involve the apparent silences in the accounts of the two main historians writing about William in Normandy, William of Jumièges and William of Poitiers. In Jumièges’s case, the more straightforward of the two, when describing the domestic arrangements of the tenthcentury Norman rulers as marriages ‘in the Danish custom’ (more Danico), despite their being identical to those of William’s parents, he recognized the existence of a tradition that had produced rulers whom he regarded as ASC, ‘D’, 1066. Adam of Bremen, Gesta Hammaburgensis ecclesiae Pontificum, ed. Bernhard Schmeidler, MGH Scriptores (Hanover and Leipzig, 1917), 394 (Adam of Bremen: History of the Archbishops of Hamburg-Bremen, trans. Francis J. Tschan, with a new Introduction and Notes by Timothy Reuter [New York, 2002], 159) (cui pro oblico sanguine cognomen est bastardus); cited in van Houts, ‘The Norman Conquest through European Eyes’, 836, note 2. 22 Lambert of Saint-Omer, Liber Floridus, ed. Albert Derolez (Ghent, 1968), 153–5; cited in van Houts, ‘The Norman Conquest through European Eyes’, 844–5, note 5. 23 OV, vi, 40–1. 24 GR, i, 416–17 (preterea proxime consanguineus). 20 21

the first years

23

legitimate, but, of which, as a monk he had to disapprove.25 At the same time, however, his presentation of William’s life is always that of a ruler whose succession was politically legitimate. On the basis of what he wrote it is clear that, as we shall see, Jumièges did have problems with other aspects of William’s conduct. But he did not present his birth as being one of them. Poitiers’s treatment is more complex because he had been William’s chaplain and, although there is no certainty that William knew what he was writing, it is hard to believe that Poitiers was not to an extent either reflecting his subject’s thoughts or, at the least, what he surmised they were. The readiness of Poitiers and Abbot John of Fécamp, both of them exceptionally close to William, to condemn concubinage, as the latter did in a letter to Pope Leo IX (1049–54) in late 1050, suggests that they could take this position because both thought William sufficiently fireproofed, and that his audience would recognize why.26 Or perhaps that, in relation to the Gesta’s rationale, the subject simply did not matter; William was politically legitimate, so there was no need to dwell on the issue. It is also possible that the conventions shaping the construction of what was as much a work of rhetoric as one of history may have rendered it unnecessary for Poitiers to dwell on William’s birth. His use of classical referencing was aimed both to present William as an exemplary ruler and, by implication, to answer criticisms: ‘His assertions of truth were indicators of conflict’, as has been acutely observed.27 We shall see later that Poitiers was extremely good at raising topics on which he believed criticism had to be answered and, as a result, we often know what they were. Hence, what may therefore lie beneath the surface in relation to William’s birth was a belief that William and his parents had atoned or, in William’s case, was continuing to atone, and that, in relation to Poitiers’s specific purpose, there was no reason to comment. It is finally important to underline that William’s succession was not only deemed to be as politically legitimate as anyone’s could be, but that it must also be located in the context of the contemporary norm whereby kinship 25 GND, i, 58–9, 78–9. In general, Pierre Bauduin, ‘L’insertion des normands dans le monde franc, fin IXe–Xe siècles: l’exemple des pratiques matrimoniales’, in Anne-Marie Flambard Héricher (ed.), La progression des Vikings, des raids à la colonisation, Cahiers du GRHIS, 14 (Rouen, 2003), 105–17. 26 GG, 74–5, 114–15. For Abbot John’s letter, PL, cxliii, cols. 799–800. For his importance within the Norman Church, NMPrinces, 281. 27 For the Gesta Guillelmi as a work of rhetoric, Jeanette M. A. Beer, Narrative Conventions of Truth in the Middle Ages (Geneva, 1981), 13–22, with the quotation at p. 22. See also Nicholas Vincent, ‘The Strange Case of the Missing Biographies: The Lives of the Plantagenet Kings of England 1154–1272’, in Bates, Crick, and Hamilton (eds.), Writing Medieval Biography, 237–57, at 237–8. Emily Winkler, ‘The Norman Conquest of the Classical Past: William of Poitiers, Language and History’, JMH, xlii (2016), 456–78.

24

william the conqueror

was a socially constructed phenomenon which could be manipulated, a situation that was also reflected in the flexibility of the language of inheritance.28 While there were others for whom a claim to legitimate succession to the duchy of Normandy could have been made in 1035 – and perhaps such claims were made, although, if they were, we do not know about it – the situation epitomizes a social and cultural fluidity that was also present within the many succession disputes everywhere, with those in Maine, Anjou, and Flanders in the 1060s and early 1070s, and in England in 1016, 1035, 1042, and 1066, having a major impact on William’s life. Thus, as far as the English kingdom was concerned, both Harold Harefoot and Harthacnut succeeded against the claims of the older children of King Æthelred the Unready (978–1016), one of whom was Emma’s son, the future King Edward the Confessor; and Edward himself later succeeded in 1042 with no apparent account being taken of the children of Æthelred’s first marriage and its progeny, above all the descendants of his son King Edmund Ironside, who was briefly king in 1016.29 The result of this was that, with Cnut’s death in 1035 initiating a phase of great political uncertainty, a series of legitimating techniques was deployed in the English kingdom to justify pragmatic political acts that enabled its political elite to achieve a consensual agreement about who should be the next king in 1035–7, 1040, and 1041–2.30 However, the necessity for these consultations made it as good as certain that, at each succession, claimants would come forward who had to be bought off or defeated. At certain points in England’s eleventh-century history, these mechanisms produced remarkably peaceful transitions. However, in 1016, when Æthelred’s children were supplanted by Cnut, they had been only partially successful. And in 1066, at least from a structural perspective, they were going to be a total failure. This situation was to be of the greatest importance in determining the course of William’s life.

28 Thus, with references to other literature, Charles West, Reframing the Feudal Revolution: Political and Social Transformation between Marne and Moselle, c.800–c.1100 (Cambridge, 2013), 52. The point is succinctly made in Le Jan, Famille et pouvoir, 43–5. For the language of inheritance, Pierre Bauduin, ‘Désigner les parents: le champ de la parenté dans l’œuvre des premiers chroniquers normands’, ANS, xxiv (2002), 71–84; Garnett, ‘ “Ducal” Succession’, 93–108. 29 Continental and anglophone literature is admirably surveyed in Judith A. Green, The Aristocracy of Norman England (Cambridge, 1997), 329–60. For an update, David Crouch, The Birth of Nobility: Constructing Aristocracy in England and France, 900–1300 (Harlow, 2005), 99–155. There are of course here echoes of Eleanor Searle’s ‘Predatory Kinship’, but as a western European not a Viking/Norman phenomenon (Predatory Kinship and the Growth of Norman Power, 840–1066 [Berkeley and Los Angeles, CA, and London, 1988]). 30 See, for example, Pauline Stafford, ‘Royal Women and Transitions. Emma and Ælfgifu in 1035–1042/3’, in Körntgen and Waßenhoven (eds.), Patterns of Episcopal Power, 127–44.

the first years

25

ROBERT AND HERLEVA The earliest source to provide usable information on William’s date of birth is the tract known as De Obitu Willelmi (‘On the Death of William’), whose composition is now dated to soon after William’s death in 1087. Although its complete text appears only in one manuscript, where it was added on at the end of a copy of Jumièges’s Gesta Normannorum Ducum produced under the supervision of Symeon of Durham, there are compelling reasons to think that the tract was not written at Durham. With production in Normandy being a possibility, it looks to have been created by someone close to events and to have been seen by its compiler as a logical conclusion to the Gesta. In it, we are told that William died in his fifty-ninth year, indicating a date of birth between 10 September 1027 and 9 September 1028.31 That the version of the Gesta Normannorum Ducum containing De Obitu Willelmi did not circulate widely may explain why the twelfth-century historians who wrote about William give different and contradictory dates for his birth. In the absence of any authoritative early text, they had nothing to guide them. Orderic seems to have thought that William was born in 1026 or very early in 1027.32 William of Malmesbury, on the other hand, says that William was in his eighth year when he became duke, thereby placing his birth between July 1027 and July 1028, a dating that was also independently inserted into two variant manuscripts of Jumièges’s earliest version of the Gesta Normannorum Ducum, presumably as an authorial corrective.33 And later, Robert of Torigni, in his interpolations into the Gesta Normannorum Ducum, completed by c.1139, like the author of De Obitu, stated that William died in his fifty-ninth year (fere sexagenarius).

31 GND, ii, 188–9. See now, Katherine Lack, ‘The De Obitu Willelmi: Propaganda for the Anglo-Norman Succession, 1087–88?’, EHR, cxxiii (2008), 1,417–56, at 1,423–5, 1,439–43. On the manuscript (BL, ms. Harley 491), Michael Gullick, ‘The Hand of Symeon of Durham: Further Observations on the Durham Martyrology Scribe’, in David Rollason (ed.), Symeon of Durham: Historian of Durham and the North (Stamford, 1998), 14–31, at 18–19, 27. 32 Orderic at first sight gives the impression of great confusion, stating, for example, that William was sixty-four years old when he died and eight years old when his father departed on his pilgrimage to Jerusalem in early 1035; the first would give a date between 10 September 1022 and 9 September 1023 for his birth, and the second, any time in 1026 or just possibly early in 1027. The figure of sixty-four is, however, the result of a mistaken belief on Orderic’s part that William was duke of Normandy for fifty-six years. For William as eight years old at the time of his father’s departure, OV, i, 158; ii, 10–11; iv, 82–3. For Orderic’s confusion about his age and the length of his reign, ibid., iv, 80–3. He later amended his figure on the length of the reign – as it happens again erroneously – to say that William was duke for fifty years, ibid., i, 158. 33 GR, i, 426–7; GND, ii, 81, note 5. For mss. C1 and C2 of the GND, ibid., i, xcix–c.

26

william the conqueror

But he elsewhere contradicted himself by saying that William was five when he became duke.34 The alternative approach of looking at the dates of birth of William’s male half-siblings Odo and Robert, later bishop of Bayeux and count of Mortain respectively, the sons of his mother’s marriage to Herluin de Conteville, and of William’s possible full-sister Adelaide, does not help because the evidence for their dates of birth is much vaguer than it is for William’s. With the earliest testimony and the majority opinion pointing to a date between mid-1027 and mid-1028, it is safest to say that William was born during that time. It is therefore very likely that William had not been born – indeed, may not even have been conceived – when his father became duke of Normandy on 6 August 1027 after the death of his brother Duke Richard III. This immediately pushes to one side the notion that Herleva would have been dismissed by Robert when he became duke. The dates of birth of Robert and Herleva are not known, but it is as good as certain that both were in their early twenties when their son was born.35 Both Orderic and the slightly earlier narrative known as the Quedam Exceptiones, written between 1101 and 1103, identify William’s place of birth as Falaise.36 Robert and Herleva’s relationship must therefore have started after Robert’s father, Duke Richard II, had given his son responsibility for the district of southern Normandy known as the Hiémois, a grant that must have been made before Richard’s death on 23 August 1026, and could have followed the assembly attended by Richard, his sons, and a large number of nobles and clergy at Fécamp in August 1025.37 Falaise in due course became so central to Robert’s power that he based himself there in 1026 during his rebellion against his brother, Duke Richard III.38 By the early twelfth century, Robert’s meeting with Herleva had become an element in the legends that developed around William; because Robert and Herleva produced a son who had a remarkable life, their relationship must have been a remarkable one. William of Malmesbury, who wrote of the relationship between William’s parents as a love-match, portrayed its beginning in terms of an instant attraction.39 Orderic is the main source of information about Herleva and her family. At one point, in his narrative of William’s siege of Alençon during a campaign GND, ii, 42–3, 194–5. The date of Robert’s birth is unknown because the date of his parents’ marriage is not known. Douglas suggested c.1003, Douglas, ‘Some Problems’, 289–91, and 291, note 3. Herleva, who bore at least five other children after William, was presumably of a similar age. 36 GND, ii, 80–1, 96–7, 299. 37 Ibid., ii, 40–1; RADN, nos. 34–6. 38 GND, ii, 44–5. 39 GR, i, 426–7. 34 35

the first years

27

that probably took place in the early 1050s, he has the defenders taunt William with the insult that Herleva’s relatives (parentes) had prepared corpses for burial.40 Elsewhere, however, he says that Fulbert (her father) was a ducal chamberlain.41 That this second passage was one of his interpolations into the Gesta Normannorum Ducum that was written over an erasure in a different ink and which extended into the margin shows that it was an expansion of what he had originally written, a statement that is now lost to us.42 While it is possible that Orderic had received new information about Herleva’s father, it is more likely that he had decided that Fulbert’s status at the ducal court was a necessary element in what he wanted to say about William’s parentage and that he must make an adjustment to achieve this; his only other insertion in this different ink was additional information that the command of a section of the ducal fleet was entrusted to Rabel, a warrior of exceptional strength, something that must be a clarification rather than an addition of any great significance.43 On the basis of William of Malmesbury’s story of instant attraction, it would appear that Herleva brought her father into the orbit of the ducal court. But it is certainly also possible that it was Fulbert’s presence in the ducal entourage as a chamberlain that led to his daughter becoming known to Robert; he might have been a recruit into Robert’s household when the latter established himself at Falaise. What the evidence probably does indicate is that some of Fulbert’s ancestors (i.e. parentes) had indeed been undertakers or tanners, but that he himself had risen socially some time before his daughter gave birth to the future William the Conqueror. In all probability Herleva and her family came from the ministerial class from which the Norman dukes traditionally recruited some of their lesser court servants, namely, the urban bourgeoisie, in this case, that of Falaise.44 As such, her standing would not have been unusual in the context of the social composition of the ducal court.45 Falaise was also so much 40 GND, ii, 124–5. On this passage, Elisabeth M. C. van Houts, ‘The Origins of Herleva, Mother of William the Conqueror’, EHR, ci (1986), 389–404; Klaus van Eickels, Vom inszenierten Konsens zum systematisierten Konflikt (Stuttgart, 2002), 73–4, note 63. 41 Herleua, Fulberti cubicularii ducis filia, GND, ii, 96–7. See also Searle, Predatory Kinship, 155. For a similar suggestion, Raoul Glaber, ed. Arnoux, 260, note 59. 42 Rouen, Bibliothèque Municipale, ms. 1174 (Y.14), fo. 135r. I am grateful to Claire Basquin and the staff of the Bibliothèque Municipale for their help in inspecting the manuscript. 43 Ibid., ms. 1174 (Y.14), fo. 134r (. . . Rabello fortissimo militi tradidit ac . . ., GND, ii, 78). 44 The classic study of this topic is Lucien Musset, ‘A-t-il existé en Normandie au XIe siècle une aristocratie d’argent?’, AN, ix (1959), 285–99. 45 Arguments similar to those here were entertainingly developed by Eleanor Searle, ‘Possible History’, Speculum, lxi (1986), 779–86, at 784 (‘I suspect that many historians have rather liked the notion of romantic, helpless, lowborn mistresses. It is only a suspicion, but it makes me wonder whether there are not more ladies who need rescuing from historians.’).

28

william the conqueror

integrated into the heartlands of ducal authority that the town’s layout has recently been seen in terms of a symbolic projection of that of Fécamp into the less securely held territories of southern central Normandy.46 Fulbert’s status as a chamberlain with constant personal access to Robert further meant that the relationship with his daughter must have been a publicly acknowledged one. Additionally, Orderic’s determination to treat the subject of William’s birth thoroughly led him to insert a second passage in which he mentioned that thirty-two of the individuals who mocked William at Alençon were mutilated on William’s orders. This insertion is again written in the margin, but this time in the same ink and style as the main text.47 He was presumably drawing attention either to William’s savagery or to his devotion to his mother; one does not of course rule out the other. It highlights a point that David Douglas made very well, namely, the importance of William’s mother and her family to his life. And it is – famously – an incident that shows how much William’s parentage matters for a biographer. But not necessarily because it was an embarrassment to him; it is very possible that William held his parents in high regard and that he did so as the product of a loving relationship. Orderic’s statement that Herleva’s marriage to Herluin de Conteville took place only after Robert’s death in 1035 indicates that he believed that Herleva and Robert’s relationship had been an enduring one, lasting perhaps for as long as ten years.48 Although William of Malmesbury said that the marriage took place before Robert’s death, he also wrote that Robert kept Herleva as if she was his lawful wife and loved her above all others.49 The good possibility that William’s sister Adelaide was their child, although only first alluded to by Robert of Torigni, argues in the same direction.50 While, as I argue below, Robert’s brief marriage to Estrith, sister of King Cnut, should surely be linked to an arranged marriage for Herleva in about 1030, because one of her sons by Herluin, Odo, was old enough to be made bishop of Bayeux in 1049/50, this does not mean that Herleva was banished from the court; while Odo must on any interpretation of the evidence for his birth have been appointed to his bishopric at well below the required canonical age of thirty, he must presumably have shown some aptitude for a position in the Church before his advancement, 46 Bernard Gauthiez, ‘Verneuil-sur-Avre, Falaise, Pont-Audemer et Lisieux en Normandie’, in Bernard Gauthiez, Elisabeth Zadora-Rio, and Henri Galinié (eds.), Village et ville au Moyen Age: les dynamiques morphologiques, 2 vols. (Tours, 2003), 35–95, at 48–50. 47 Rouen, Bibliothèque Municipale, ms. 1174 (Y.14), fo. 139r. 48 GND, ii, 96–7. 49 Deinceps unice dilexit et aliquamdiu iuste uxoris loco habuit, GR, i, 426–7. 50 GND, ii, 270–1 (soror uterina Willelmi regis Anglorum); Thompson, ‘Being the Ducal Sister’, 63–4.

the first years

29

and therefore have been some way beyond his mid- to late teens by 1049/50. Herleva’s husband’s responsibilities for the protection and exploitation of the coast on the south bank of the Seine estuary – he may well have overseen the construction of a canal named after him (the Fossa Herluini) – actually suggest that he was someone of sufficient importance to attend the court quite often.51 That Herleva and her new family turn up at significant points in William’s life throughout the 1030s and 1040s indicates that the marriage to Herluin extended the number of young William’s supportive kin rather than severing the links between mother and son.52 All this may also suggest a strong maternal bond between Herleva and her son. When we turn to other twelfth-century writers, the tradition that Robert and Herleva had been a close and loving couple is regularly present. In particular, in his unfinished verse chronicle written in the 1160s and early 1170s, the Roman de Rou, Wace elaborated the earlier stories with an account of the couple’s intimate lovemaking and of Robert’s personal tenderness towards Herleva, whom he said was a virgin when the relationship began. And, so he tells us, when Herleva had slept after intercourse, she awoke to announce that she had dreamt that a tree had grown out from her body and had covered the whole of Normandy in its shadow.53 Like others, this story is indicative of the fascination that Robert and Herleva’s relationship and William’s birth came subsequently to exercise. The tale of the impossibly precocious feat of the child William grasping the straw around him in his cradle soon after birth, which first appears in the so-called Brevis Relatio, an early twelfth-century narrative written at Battle Abbey, William’s English monastic foundation on the battlefield of Hastings, and which was also included by Malmesbury and Wace, is another. Malmesbury even has the midwife proclaim the infant William’s supposed precocity as a portent that the child was going to be a king.54 What mattered to them was to show how special William’s birth had to be, underscoring this with commonplace portents that drew on various biblical stories and on accounts of the foundation of imperial Rome.55 It is 51 For Herluin’s social status, Jimmy Mouchard, ‘De la voie navigable aux sites portuaires en basse vallée de Seine: maîtrise et gestion des accès (Antiquité-époque moderne)’, in Elisabeth Lalou, Bruno Lepeuple, and Jean-Louis Roch (eds.), Des châteaux et des sources: Archéologie et histoire dans la Normandie médiévale. Mélanges en l’honneur d’Anne-Marie Flambard Héricher (Rouen, 2008), 103–27, at 113–21. For the Fossa Herluini, RADN, no. 32. For documentary evidence, David Bates and Véronique Gazeau, ‘L’abbaye de Grestain et la famille d’Herluin de Conteville’, AN, xl (1990), 5–30, at 22–3. 52 For details on the careers of Herleva’s relatives, Douglas, WC, 381. 53 Roman de Rou, 164–7, pt. 3, lines 2,823–66. 54 Brevis Relatio, 47; GR, i, 426–7. 55 On this material, Herman Braet, ‘Le songe de l’arbre chez Wace, Benoît et Aimon de Varennes’, Romania, xci (1970), 255–67; Huguette Legros, ‘Naissance d’un héros: de la

30

william the conqueror

notable, however, that none of these stories appears in anything Orderic wrote. Even at the moment of his birth, William seemingly divided opinion. It is, however, surely unlikely that Wace or anyone else had special knowledge of how Robert and Herleva disported themselves in bed. THE RULE OF DUKE ROBERT The initially precarious nature of William’s father’s rule and the consequences of his first years as duke were certainly an important influence on his son’s life. Robert had been involved in a brief rebellion against his elder brother Duke Richard III during his short reign. William of Jumièges also mentioned the possibility that Richard’s death might not have been a natural one and that he might have been poisoned, a suspicion also articulated by a contemporary writing in Aquitaine, Adhemar of Chabannes.56 Jumièges nonetheless presented Robert’s succession as a formal designation by his brother, which presumably represents another manipulation of the language of inheritance by a political elite, since Richard had a son by a concubine, one who might have succeeded him.57 The accusation of fratricide was not made until the early twelfth century when it appears in William of Malmesbury’s Gesta Regum and in the Chronica de Gestis Consulum Andegauorum, with the latter adding that it was to atone for this crime that Robert later went on pilgrimage to Jerusalem.58 Orderic meanwhile, in his interpolations into the Gesta Normannorum Ducum, while making no mention of poisoning, omitted Jumièges’s statement that Richard had made Robert his heir.59 If this evidence is treated as being so allusive or so late as to be unconvincing, there remains the near certainty that Robert’s succession aroused suspicions; it is, for example, very unusual for William of Jumièges to mention poisoning. And, whatever the case, it is surely indicative that suspicions about Robert’s conduct were present throughout his son’s life and, in particular, as he grew to adulthood. These uncertainties must surely

bâtardise à la légende dans la Chronique de Benoît de Sainte-Maure’, in idem (ed.), Guillaume le Conquérant face aux défis (Orléans, 2008), 135–47. A version of the tree sprouting branches also appears in Wulfstan of Winchester’s Vita of St Æthelwold, bishop of Winchester (963– 84), and was repeated by William of Malmesbury, GP, i, 262–3. See also Susan M. Johns, ‘Women and Power in the Roman de Rou of Wace’, ANS, xxxvi (2014), 117–34, at 129–30. 56 GND, ii, 46–7 (ut retulerunt plurimi, ueneno mortem obit); Ademari Cabannensis Chronicon, ed. Pascale Bourgain, assisted by Richard Landes and Georges Pon, CCCM, cxxix (Turnhout, 1999), 184 (Adémar de Chabannes, Chronique, trans. Yves Chauvin and Georges Pon [Turnhout, 2003], 284–5). 57 GND, ii, 46–7 (fratrem suum Robertum relinquens sui ducatus). For Robert’s succession, Garnett, ‘ “Ducal” Succession’, 105–7. 58 GR, i, 308–9; Chronica de Gestis Consulum Andegauorum, in Chroniques des comtes d’Anjou et des seigneurs d’Amboise, ed. Louis Halphen and René Poupardin (Paris, 1913), 50. 59 GND, ii, 46–7.

the first years

31

also have been associated with the quarrel between Robert and his uncle, Robert, Archbishop of Rouen. According to Jumièges, the archbishop went as far as to place the whole of Normandy under an anathema, while for a time seeking refuge outside the duchy with the French king Robert the Pious (996–1031).60 An anathema was one of the most terrible punishments a ruler could incur, with the potential to destabilize rule and give encouragement to enemies and rivals. The existence of a son of Richard III, known to posterity as Nicholas, which was certainly his monastic name, must have been another factor in this uncertainty. Like William the son of a concubine, Nicholas could have been a rival around whom court factions might have gathered. His entry into monastic life, at first as an oblate at the abbey of Fécamp, and then a monk at Saint-Ouen of Rouen, was described by Orderic in terms of the removal by Duke Robert of a political threat; his elliptical observation in the Gesta Normannorum Ducum that Nicholas ‘lacked an earthly inheritance’ became in the later Historia Ecclesiastica a blunt statement that Duke Robert had engineered the forced enclosure of a threatening relative.61 Living until 1092, Nicholas was made abbot of Saint-Ouen of Rouen in 1042. He seems to have accepted his fate without demur and to have lived out a thoroughly conventional political and religious life; he was the only Norman abbot to make a contribution of ships to the expedition of 1066.62 His treatment surely indicates a determination on Robert’s part that he would be succeeded by a son of his own. Jumièges added that the reconciliation with Archbishop Robert brought political stability and enhanced Duke Robert’s respect for the Church. This may well have been an important element in setting the tone at the court in which the young William was growing up. The process of atonement must also have created a religious atmosphere that the young child would have imbibed. Robert’s installation by his father, Richard II, at Falaise introduces two themes that were certainly to be of great importance for his son’s life, namely, the interplay within Normandy of core and periphery and the impact thereof on relationships with the rulers of the principalities and lordships that surrounded Normandy’s fluctuating borders. Falaise was close to an especially contested frontier zone, since beyond Argentan, approximately twenty kilometres further south, were the lands of the Bellême family and the episcopal town of Sées, a bishopric included within the metropolitan jurisdiction of the Norman archbishops of Rouen, but politically only The main source for this episode is ibid., 46–9. GND, ii, 46–7; OV, iv, 306–7. 62 Elisabeth van Houts, ‘The Ship List of William the Conqueror’, ANS, x (1988), 159–83, at 176 (idem, History and Family Traditions, chapter 6). 60 61

32

william the conqueror

intermittently under Norman control. At the time of Robert’s appearance at Falaise, the bishop of Sées was Radbod, a Norman who was clearly a nominee of Duke Richard II with links to Archbishop Robert of Rouen. But his predecessor’s very name, Seginfrid, identifies him as a member of the Bellême family. Radbod’s successor Ivo, whose appointment probably occurred after 1035, was from the same family; he was indeed eventually to combine his bishopric with secular lordship over the family lands.63 If frontiers are thought of as zones of fluctuating loyalties and as discrete social organisms, similar phenomena were observable all around the Norman principality.64 This, however, was an especially difficult place. Robert’s military campaigns consolidated ducal power to the south of Argentan and, further west, on the borders with Brittany. Although in the case of the lords of Bellême the respite from external pressure was only temporary, in that of Count Alan III of Brittany (1008–40), who was Robert’s brother-in-law, a settlement brokered by Archbishop Robert produced a new collaboration that endured into the first years of William’s reign.65 To the north-east of Normandy, a sister of Robert was married to Count Baldwin IV, the ruler of the very important principality of Flanders, to whom Robert also sent military assistance against rebellions, as he also did to the new French king Henry I, whose reign had begun in 1031. As a result, there does not appear by 1035 to have been any substantial external threat to the duchy’s security. Robert’s Normandy also appears to have been well integrated into the political world of northern France and Robert himself a significant player therein. Networks similar to those constructed by Duke Richard II were continued. Robert was present at the new French king’s court at Orléans at 63 For context, Kathleen Thompson, ‘Family and Influence to the South of Normandy in the Eleventh Century: The Lordship of Bellême’, JMH, xi (1985), 215–26, at 215–17; Gérard Louise, La seigneurie de Bellême (Xe–XIIe siècles: dévolution des pouvoirs territoriaux et construction d’une seigneurie de frontières aux confins de la Normandie et du Maine à la charnière de l’an mil (Flers, 1992–3), i, 137–61. Ivo’s first clear appearance as bishop dates to 1046 or early 1047, RADN, no. 107. For the sequence of bishops, David Spear, The Personnel of the Norman Cathedrals during the Ducal Period, 911–1204 (London, 2006), 271–2. 64 There is now a huge literature on frontier societies. For a recent contribution that both reflects on definitions and is especially relevant to this book’s subject, Leonie V. Hicks, ‘The Concept of the Frontier in Norman Chronicles: A Comparative Approach’, in Keith J. Stringer and Andrew Jotischky (eds.), Norman Expansion: Connections, Continuities and Contrasts (Farnham, 2013), 143–64, at 143–6. 65 On these campaigns, David Bates, Normandy before 1066 (London and New York, 1982), 68–72; Cassandra Potts, ‘Normandy or Brittany? A Conflict of Interests at Mont SaintMichel (996–1035)’, ANS, xii (1990), 135–56, at 155–6; Louise, La seigneurie de Bellême, i, 290–5. On the specific case of Mont Saint-Michel, Pierre Bouet, ‘Le Mont-Saint-Michel entre Bretagne et Normandie de 960 à 1060’, in Joëlle Quaghebeur and Bernard Merdrignac (eds.), Bretons et Normands au Moyen Age: Rivalités, malentendus, convergences (Rennes, 2008), 165–200, at 184–5.

the first years

33

Easter (1 April) 1032, a gathering attended by all the great princes of the French kingdom, most notably Count Fulk Nerra of Anjou (987–1040), Count Odo II of Blois-Chartres (996–1037), Duke William the Fat of Aquitaine (1030–8), and the aforementioned Counts Alan III of Brittany and Baldwin of Flanders.66 This assembly occurred while conflict was brewing between the French king and his mother Constance, who wished to depose him in favour of his brother Robert, a war in which the various princes were soon to take sides, with Duke Robert giving significant support to King Henry. Competition and violence were innate in the politics of this northern French world in which William grew up. In terms of his later life, the crucial cultural element was the opportunistic expansionism practised by these princes and, most significantly in relation to his future prospects, by Count Fulk Nerra and his son Count Geoffrey Martel (1040–60), whose ambitions included domination over the county of Maine and the lands up to the south of the territory where the Norman dukes exercised, or claimed to exercise, hegemony. Alongside the counts of Anjou, the great operator during William’s early years was Count Odo II of Blois-Chartres, a man whose ambitions have recently been described as ‘quite breath-taking’, among them fighting Emperor Conrad II (1024–39) in 1031–2 for succession to the kingdom of Burgundy.67 The mixture of instability and resilience involved in these attempts at state-building is shown by their disintegration, albeit not always after one generation, and, as in the case of, for example, Normandy and Anjou, often the long-term survival of the territorial core from which expansion had taken place. Lacking serious bureaucracies and institutions, their ultimate foundation was the domination of a powerful individual, constructed around skilfully deployed violence, ritualized performance, and the fortification of strategically crucial strongpoints.68 William’s father and grandfather had reputations for holding their own in this febrile world, but not for the most dramatic forms of opportunistic aggression characteristic of several of their contemporaries. Although we have no direct information on the subject, it may be that the young William’s role models lay outside Normandy as well as within it. In adulthood, according to William of Malmesbury, he was spurred on 66 For the most recent edition of this charter with a discussion of the problems it poses, Les actes des ducs de Bretagne (944–1148), ed. Hubert Guillotel et al. (Rennes, 2014), no. 21. See further RADN, no. ; O. Guillot, Le comte d’Anjou et son entourage au XIe siècle, 2 vols. (Paris, 1972), ii, no. C49; for a narrative of these times, Bernard S. Bachrach, Fulk Nerra, the NeoRoman Consul, 987–1040 (Berkeley and Los Angeles, CA, and London, 1993), 211–16. 67 For the quotation, West, Reframing the Feudal Revolution, 126. 68 For violence as a legitimate aspect of power, see the succinct statement in Richard E. Barton, Lordship in the County of Maine, c.890–1160 (Woodbridge, 2004), 146–73. For a magisterial general treatment, Thomas N. Bisson, The Crisis of the Twelfth Century: Power, Lordship, and the Origins of European Government (Princeton, NJ, and Oxford, 2009), esp. 128–81.

34

william the conqueror

to emulate the feats of Robert Guiscard (died 1085), one of the great Norman conquerors of southern Italy, a member of the same gens, but a social inferior.69 From earlier times, again from Malmesbury, we have the anecdote about Count Fulk Nerra of Anjou kicking his prostrate son, the future Count Geoffrey Martel, as he lay on the floor begging for forgiveness after a rebellion. Malmesbury wrote that Geoffrey had to walk for several miles, carrying a saddle on his shoulders, a well-known mechanism to demonstrate humiliation, and then added that he regarded Fulk as an upright and honourable man.70 Central to this apparatus was the articulation of ‘anger’, a dramatic expression of will that communicated decisions and shaped the operation of power. Whether ‘anger’ is treated as an aspect of psychology or anthropology – it is of course both, with establishing the boundary line between them being the biographer’s responsibility – it is at the heart of the dominance that William achieved and the methods he used to maintain it.71 There were a lot of people around from whom he could learn. There are references both in narrative sources and charters to Robert abandoning the bad behaviour of his youth after about 1030. His acts take on a religious, even penitential, air.72 The earliest source to mention the pilgrimage to Jerusalem that Robert undertook in 1035, written between c.1036 and c.1041, attributes it to a desire for repentance for past misdemeanours, which in all probability were his conflicts with Archbishop Robert, some arbitrary seizures of church lands, perhaps his relationship with Herleva, and possibly the murder of his elder brother.73 Robert’s foundation in the 1030s of two major religious houses, a monastery in western Normandy at Cerisy and a nunnery to the north of the river Seine at Montivilliers, is indicative of a notably intense concern for his soul. Other signs of piety are his reliance on the counsel of the devout abbot of Verdun, Richard de Sainte-Vanne, and his presence on 12 December 1032 at the translation of the relics of St Nicaise and other saints to the abbey of Saint-Ouen of Rouen.74 These were also all developments that both reinforced and created political links across regions of northern France.75 GR, i, 482–3. Ibid., i, 436–7. 71 For these issues, Stephen D. White, ‘The Politics of Anger’, in Barbara H. Rosenwein (ed.), Anger’s Past: The Social Uses of Emotion in the Middle Ages (Ithaca, NY, 1998), 127–52. 72 See GND, ii, 48–9; Inventio et Miracula sancti Uulfranni, ed. Dom. J. Laporte, Société de l’Histoire de Normandie, Mélanges, xiv (Rouen, 1938), 47; RADN, nos. 74, 85. 73 Rodvlfi Glabri Historiarvm Libri Quinque, 202–5. For the date of writing, ibid., xlv. 74 Hugh of Flavigny, Chronicon, ed. G. Pertz, in MGH Scriptores, viii, 401; Translatio B. Nicasii martyris, PL, clxii, col. 1166. 75 For a detailed study, Samantha Kahn Herrick, Imagining the Sacred Past: Hagiography and Power in Early Normandy (Cambridge, MA, and London, 2007), 38–49. 69 70

the first years

35

It looks as if the young William was growing up not only at a time when his father’s rule was increasingly secure and successful, but in an atmosphere where atonement for sin and religious patronage were a central part of everyday life. The associated influences on the adult William may well have been profound, both in terms of the practice of politics and of the conduct required of him. Growing up as he did with an aristocratic career marked out for him, there was much in what was happening around him that would have shaped his thoughts, especially since his succession to the duchy of Normandy was to an extent fortuitous. It could have shaped a sense of the need for correctness and control. For all this, Duke Robert’s regime cannot be seen as maintaining any sort of perfect peace within Normandy. The later years of his reign are notable for the extension and rise to greater prominence of Humphrey de ‘Vieilles’ and Roger I de Montgommery, both of whose descendants were to play roles of central importance in William’s life. Both were recipients of lands formerly held by the abbeys of Fécamp and Bernay, the latter a recent foundation by Duke Richard II out of the dowry of his first wife Judith. Whether or not the transfers of property were made with the agreement of the two monasteries, the basic point remains that the two men’s gains could be regarded as threatening by others; Humphrey’s acquisitions brought his lands into closer proximity to those of Roger I de Tosny, and Roger de Montgommery’s installation as vicomte of the Hiémois, effectively as Duke Robert’s deputy in the region, created a new presence in southern Normandy that others with power there would have found disturbing.76 Also notable for aggressive behaviour at the expense of others was Count Gilbert of Brionne.77 In the socio-cultural world of competition and feud that characterized the eleventh century, a subject to which I will turn in Chapter 2, there was a potential for explosion here if the umpiring role of an adult duke was removed. NORMANDY, ENGLAND, AND THE NORTH SEA WORLD Ralph Glaber in the later 1030s, and, in somewhat garbled form, Adam of Bremen in his History of the Archbishops of Hamburg-Bremen from about 1080, say that William’s father married, and then rapidly divorced, Estrith, a sister of Cnut. As the Danish conqueror of the English kingdom, Cnut was

76 Kathleen Thompson, ‘The Norman Aristocracy before 1066: The Example of the Montgomerys’, HR, lx (1987), 251–63, at 255–6; Cassandra Potts, Monastic Revival and Regional Identity in Early Normandy (Woodbridge, 1997), 119–27. 77 See, for example, OV, ii, 12–13.

36

william the conqueror

also King of Denmark and, after 1028, King of Norway as well.78 Although scepticism has sometimes been expressed about this story, there are a number of reasons why it should be accepted. In the first place, there are Glaber’s Norman connections already referred to. And, in the second, given the potential consequences of the marriage not just for William, but also for the way in which the Norman rulers were represented by Jumièges and William of Poitiers as the consistent supporters of the future King Edward the Confessor, there were obvious reasons for them not to mention it.79 Dates ranging from 1027 to 1033 have been suggested for the marriage. Estrith had been the wife of Earl Úlfr, a man who had been prominent in England under Cnut and had acted on his behalf in Denmark before participating in a major revolt there. Although the exact date of Úlfr’s death – probably murdered on Cnut’s orders – is unknown, the later 1020s seem likely, suggesting a date of around 1030 for the marriage to Robert.80 The insecurity of Robert’s early years as duke supports this dating; a marriage to a woman of such high status was unlikely to have been considered before his rule had become firmly established. Glaber says that Robert repudiated Estrith because he disliked her. But, with Herleva’s removal from court surely being central to the marriage agreement, the chief cause of its failure could well have been a refusal on Robert’s part entirely to dismiss a partner of possibly as much as five years’ standing. Not only would Estrith have been familiar with such domestic circumstances through her brother Cnut’s behaviour, it is unlikely that a woman of her status and experience would have tolerated Herleva’s continued presence. Despite Herleva’s marriage to Herluin de Conteville, the young William’s very existence guaranteed that contact with his mother would be maintained; not only would a son of a concubine routinely remain within the father’s family circle, his position as a potential heir would have been unaffected until Estrith produced a healthy male child. We might also wonder about Estrith’s feelings, dispatched into this set-up by a brother who was directly responsible for her first husband’s death. William’s position at court and as a potential heir was guaranteed by the 78 Rodvlfi Glabri Historiarvm Libri Quinque, 204–5 (Raoul Glaber, ed. Arnoux, 258); Adam of Bremen, Gesta Hammaburgensis ecclesiae Pontificum, 114–15 (trans. Tschan, 92). While Adam wrongly names Estrith’s husband as Robert’s father, Duke Richard II, the fact that he knew of a marriage between Cnut’s family and the Norman dukes suggests that one took place. 79 Cf. Douglas, ‘Some Problems’, 292–4; GND, ii, 77, note 4. See further M. K. Lawson, Cnut: The Danes in England in the Early Eleventh Century (London and New York, 1993), 109–10; Stafford, Queen Emma and Queen Edith, 235. 80 For what is known about Úlfr, Simon Keynes, ‘Cnut’s Earls’, in Alexander R. Rumble (ed.), The Reign of Cnut, King of England, Denmark and Norway, Studies in the Early History of Britain (London and New York, 1994), 43–88, at 62–4.

the first years

37

marriage’s failure, at least for a foreseeable future, but not necessarily for an indefinite one. There is, however, no evidence that Robert ever contemplated a second marriage. The repudiation of Estrith turned out to have consequences that shaped the rest of William’s life. A complete failure as a piece of diplomacy, it foregrounded the presence at the Norman court of the future King Edward the Confessor, his brother Alfred, and their sister Godgifu (Gode), the three children of Æthelred the Unready, king of the English, and his wife Emma, sister of Duke Richard II. In near-permanent exile from the time of Æthelred and his family’s flight to Normandy into Richard’s protection in 1013, when it looked as if Cnut’s father, King Svein Forkbeard of Denmark, was going to conquer the English kingdom, they became factors of incalculable significance in the politics of north-western Europe once Æthelred’s children had been supplanted in England by Cnut. The repudiation’s impact on the social memory of Estrith’s kin is also a factor to be considered. Her son by Úlfr, Svein, became King of Denmark in 1047 and was to be a vigorous protagonist in events after 1066. And her sister-in-law was none other than Gytha, wife of Cnut’s great protégé Godwine, earl of Wessex, and mother of six known sons, one of whom was Harold, later to become king of the English. If these memories were sufficiently powerful to influence political behaviour, a result was that there were a lot of very powerful people with reasons to dislike the progeny of Robert, duke of Normandy. And, even if they did not think in this way, the episode highlights the vast geographical range of the political and social world within which the rulers of Normandy had to operate. The temptation to make teleological lineal links between these events and later times is one that a biographer of William the Conqueror must resist. The apparently neat processes set out in several of the main narrative sources, and above all by Jumièges and Poitiers, must be simplifications, with the mountain of historiography in existence demonstrating the difficulty of knowing exactly what form this simplification took. It is hard not to think that sentiments encompassing sympathy for three English royal children effectively abandoned by their mother to the charity of her Norman kin and of other northern French rulers when Emma married Cnut in the spring of 1017, alongside an opportunistic awareness of their political potential, must frequently have been present in many minds. However, in the end, the children were no more than members of the considerable body of individuals of kingly stock scattered around the courts of Europe by the recurrent succession crises of the eleventh century. One crucial determinant of their future was going to be their treatment at the hands of Duke Richard II and his sons; the other was the politics of the English kingdom.

38

william the conqueror

In the midst of the turbulent events of 1016–17 that made Cnut secure as king of the English, the actions of the main political players, namely Cnut himself, Duke Richard II, and Emma, were in all likelihood designed to stabilize a potentially volatile situation. Richard, with a long track record of supporting both Cnut and his father Svein Forkbeard while at the same time trying to keep out of any direct involvement in the wars across the Channel through stratagems such as his sister Emma’s marriage to Æthelred, may well have approached the situation pragmatically.81 Although it is possible that Richard sent an embassy to Pope Benedict VIII (1012–24) on the exiles’ behalf, it would be wrong to think of him following a policy that was either pro-Cnut or pro-Edward; the realities of power made accommodation with Cnut a necessity. The latter’s patronage of Richard’s favourite monastery of Fécamp is another sign of cordial relations between them.82 With Cnut in control of a vast North Sea empire, a rational assessment of Edward’s prospects in England in 1016–17 must have been that they were non-existent. Such a calculation would still have applied in the 1020s and early 1030s. Open to rejection as tendentious self-justification, the statement made on Emma’s behalf in the Encomium Emmae Reginae that she was in Normandy when ambassadors were sent to her by Cnut should not be too readily disregarded.83 It may receive support from Ralph Glaber’s report that Richard agreed to Emma’s marriage to Cnut, an opinion that could still be repeated in Normandy in the early 1050s by the anonymous author of the Inventio et Miracula sancti Uulfranni. Later in the same decade, however, Jumièges set out a very different version of events, saying that Emma had stayed in London.84 William of Malmesbury also believed that Richard had consented to the marriage, adding that he thought his conduct despicable.85 The contradictions between the sources notwithstanding, in terms See in general, Bates, Normandy before 1066, 37–8. C. H. Haskins, ‘A Charter of Canute for Fécamp’, EHR, xxxiii (1918), 342–4; P. H. Sawyer, Anglo-Saxon Charters: An Annotated List and Bibliography (London, 1968) (updated at http://www.esawyer.org.uk), nos. 949, 982; Michaël Bloche, Le chartrier de l’abbaye de la Trinité de Fécamp: étude et édition critique, 928/929–1190, Ecole Nationale des Chartes, Thèse pour le diplôme d’archiviste paléographe, 269–72, nos. 14, 15. For the possibility of an embassy, Elisabeth van Houts, ‘Qui étaient les Normands? Quelques observations sur des liens entre la Normandie, l’Angleterre et l’Italie au début du XIe siècle’, in David Bates and Pierre Bauduin (eds.), Penser les mondes normands médiévaux (Caen, 2016), 129–46, at 135–41. 83 Encomium Emmae Reginae, ed. Alistair Campbell, with a supplementary introduction by Simon Keynes, Camden Classic Reprints (Cambridge, 1998), 32–3 (in Normandensi regio). Cf. Simon Keynes, ‘The Æthelings in Normandy’, ANS, xiii (1991), 173–205, at 181–5; Stafford, Queen Emma and Queen Edith, 228. 84 Rodvlfi Glabri Historiarvm Libri Quinque, 204–5 (Raoul Glaber, ed. Arnoux, 258); Inventio et Miracula sancti Uulfranni, 29–30; GND, ii, 20–1; Elisabeth van Houts, ‘Historiography and Hagiography at Saint-Wandrille: The Inventio et Miracula sancti Vulfranni’, ANS, xii (1990), 235–51, at 251 (reprinted in idem, History and Family Traditions in England and the Continent, 1000–1200, chapter 4). On the date of the Inventio, van Houts, ‘Historiography and Hagiography’, 237–8. 85 GR, i, 318–19. 81 82

the first years

39

of the exiles’ prospects and their future impact on the unborn William, Emma’s whereabouts in 1016–17 are ultimately something of a red herring. Given the exigencies of the present, Richard would in all likelihood have given his agreement to the marriage whether Emma had stayed in England or travelled with her family to Normandy. The agreement may well have been given retrospectively once it became clear that Cnut was not going to be removed. As the children of a king, Æthelred’s offspring were nonetheless highly valued in Normandy and were well looked after there. The first overt demonstration of this was Godgifu’s marriage to Drogo, count of the Vexin, probably on 7 April 1024 at Rouen. Elisabeth van Houts has reasonably seen the grants of land in Normandy made to Drogo by Richard II as a dower for Godgifu, given in lieu of what her mother Emma and stepfather Cnut would have been expected to do. An expression of the Norman rulers’ long-standing practice of making marriage alliances that aimed to consolidate alliances with princely rulers across northern France, the marriage also initiated a period of personal collaboration between Richard, his sons, and Count Drogo that was beneficial to the economic and political interests of the Norman rulers and Norman monasteries down to the river Seine towards Paris.86 But it also represents treatment for Godgifu – and by implication her brothers – that is no more than contemporaries would have thought appropriate to their status; the Inventio et Miracula sancti Uulfranni, for example, in addition to reporting Richard’s consent to the marriage of Emma and Cnut, said that he brought up the two brothers Edward and Alfred as if they were his own sons.87 Robert’s marriage to Estrith c.1030 did not necessarily cut across this protectiveness towards them. While, for Cnut, the marriage had the potential to neutralize Robert’s possible support for the English princes, for both it would also at the least have been seen as consolidating a mutually agreed framework of understanding around the present and future status of the exiles. Although Jumièges does not actually mention the marriage between Cnut and Emma, relations between Robert and Cnut in the later years of both of their reigns are presented by him in terms of such hostility that it looks as if they had been soured beyond repair. Thus, Robert is said to 86 On Godgifu’s marriage, David Bates, ‘Lord Sudeley’s Ancestors: The Family of the Counts of Amiens, Valois and the Vexin in France and England during the 11th Century’, in The Sudeleys – Lords of Toddington (London, 1987), 34–48, at 36–7; Pierre Baudin, La première Normandie (Xe–XIe siècles). Sur les frontières de la haute Normandie: identité et construction d’une principauté (Caen, 2004), 255–7; Elisabeth van Houts, ‘Edward and Normandy’, in Richard Mortimer (ed.), Edward the Confessor: The Man and the Legend (Woodbridge, 2009), 63–76, at 63–8. For Drogo’s presence at Rouen in 1024, F. Lot, Etudes critiques sur l’abbaye de Saint-Wandrille (Paris, 1913), recueil des chartes, no. 7. 87 Inventio et Miracula sancti Uulfranni, 29–30 (van Houts, ‘Historiography and Hagiography’, 251).

40

william the conqueror

have threatened to launch an invasion of England and even to have assembled a large fleet to do so, which was driven to the Channel Islands by storms of such ferocity that a crossing of the Channel was impossible. He is also supposed to have made the request of Cnut that Edward and Alfred should be welcomed back to England and restored to what was rightfully theirs, with Cnut supposedly responding with an offer of half of the kingdom.88 Jumièges probably exaggerated the scale of Robert’s support; the dispatch of the embassy might have been bravado and, as we have seen, the diverted fleet proved to be very useful in a campaign against Count Alan III of Brittany. What his narrative certainly does reflect, however, is that, within Normandy, Edward and Alfred were pushed into greater prominence during the later years of Robert’s reign. In addition to Jumièges writing of Robert and Edward as blood-brothers, the two brothers appear for the first time as signa to two charters dating from either 1033 or 1034, charters whose signa were drawn from the duchy’s topmost political elite. They included also, in one case, the French king, Henry I (1031–60), who had sought refuge in Normandy during a revolt and was among those present at a gathering that took place in April 1033.89 While a statement was certainly being made, exactly what it was depends on how one interprets Edward’s apparent appearances with the title rex in two charters dating from Robert’s time.90 One of them is a Fécamp original on which the signum of the young William also appears, and which must date to between 12 September 1033 and early 1035. Three slightly different original versions of this charter existed into the twentieth century, in all of which the names of Duke Robert, William, and Archbishop Robert of Rouen are placed at the start of an extremely prestigious list of signa. However, in the case of the two charters that survive – the third one was seen and edited by Marie Fauroux, but cannot now be found – while the signa were all written by the same scribe, in both of them William’s is placed below his father’s in a way that is not symmetrical with the presentation of the rest of the signa; on one of the charters it is written in a smaller script below his father’s and on the other between the first two of four lines of signa that are otherwise perfectly GND, ii, 76–9. Ibid., ii, 76–7; RADN, nos. 69, 70; Simon Keynes, ‘The Æthelings in Normandy’, ANS, xiii (1991), 173–205, at 186–8. See further Elisabeth van Houts, ‘The Political Relations between Normandy and England before 1066 according to the Gesta Normannorum Ducum’, in Raymonde Foreville (ed.), Les mutations socio-culturelles au tournant des XIe–XIIe siècles (Paris, 1984), 85–97, at 90–3 (reprinted in idem, History and Family Traditions, chapter 5). 90 RADN, no. 85; cf. Keynes, ‘The Æthelings in Normandy’, 186–94. The latest study and edition is by Bloche, Le chartrier de l’abbaye de la Trinité de Fécamp, 281–7, no. 19. For facsimiles of two of the documents, C. H. Haskins, Norman Institutions (Cambridge, MA, 1918), plates 4 and 5. For a partial facsimile showing the signa of the crucial ms. A., Keynes, ‘The Æthelings in Normandy’, plate 2 (Haskins, Norman Institutions, plate 4). 88 89

the first years

41

arranged in four horizontal lines. Edward’s name and cross are also out of symmetry with the others, written at the edge of the parchment. And – crucially for my argument – also out of symmetry is a third interlineal insertion, the signum of Count Gilbert of Brionne, who was the young William’s chief guardian after 1035. There must therefore be a strong case for believing that all three of the signa were insertions into whatever now lost charter the two scribes were copying and that theirs was an updated version of it, produced soon after William’s accession in 1035 to demonstrate continuity into the new reign; that one of the copies contains a further significant interpolation also indicates a text that has been developed from an original base.91 A second apparent appearance by Edward as rex is in a charter that purports to be a grant by him of English properties to the abbey of Mont Saint-Michel. Although arguments have been made in favour of the charter’s authenticity, its uniqueness as a grant made far in advance of any realistic prospect of its being implemented, the implausibility of its contents, and its survival only as a copy in the abbey’s mid-twelfth-century cartulary are causes for concern. It is probably a forgery.92 However, even on an interpretation of these two charters’ significance that rejects there being any reference to Edward’s kingship in Duke Robert’s time, we are still left with Edward and Alfred making formal appearances among the duchy’s elite before Robert’s death. Even if the decision to apply the royal title to Edward was one that was actually taken after Cnut’s death in 1035, either in response to the perceived usurpation of his rights by Harold Harefoot or when his prospects increased after the accession of Cnut’s son Harthacnut, the two exiles’ enhanced importance before 1035 remains obvious and is extremely important. The possibility that Cnut, although only in his late thirties, was in poor health for some time before his death must also have fuelled a revival of interest in the exiles’ claims to the English kingdom.93 What is certain is that William was made well aware of the politics of the English succession while still a young child. THE CHILD WILLIAM William’s appearances in charters before 1035 indicate that he was being pushed into prominence in the 1030s. While the possibility that his name was interpolated later into any one of the seven charters in which he 91 The interpolation appears in Haskins, Norman Institutions, plate 5. For further commentary, ibid., 260–3. 92 RADN, no. 76 (The Cartulary of the Abbey of Mont-Saint-Michel, ed. K. S. B. Keats-Rohan [Donnington, 2006], no. 10); Keynes, ‘The Æthelings in Normandy’, 190–4. 93 The statement that Cnut was in poor health appears in GND, ii, 78–9. It is taken seriously, with supporting arguments, in M. K. Lawson, ‘Cnut [Canute] (d.1035), king of England, of Denmark, and Norway’, ODNB.

42

william the conqueror

apparently appears can never be ruled out, their combined evidence is surely decisive that he was seen as being important and a potential successor to his father. The phrase that specifically identifies William as Robert’s designated successor in a charter for the abbey of La Trinité-du-Mont of Rouen is in all likelihood an insertion made when the document was transcribed into the abbey’s late eleventh-century cartulary, which is its earliest surviving text. On the other hand, it may only be the words successor ipsius that have been interpolated; the confirmation as it stands might still have been a joint one of father and son.94 Three appearances in related confirmations for Rouen cathedral might also be later additions, since the earliest texts are copies dating from the thirteenth, fourteenth, and fifteenth centuries, and because, in one case, the mention of a grant by William himself (+Hec est adfirmacio Vuilelmi de Sota Villa) must be a colophon added to an existing text, a practice common in charters of this period and suggesting that additions may have been made to the documents.95 What is, however, striking is the consistency in the way in which William’s confirmation is presented in these three charters: in two cases as a simple signum and in one, as a confirmation of his father’s gift.96 In addition, such general confirmations in favour of Rouen cathedral are an obvious follow-up to the quarrel between the duke and Archbishop Robert. If none of the charters describes William as being Robert’s designated successor, his inclusion does show that steps were being taken to ensure that, if he did follow his father as duke, he would respect the gifts Robert had made.97 There are close parallels between William’s appearances in charters and those of another only son of a ruler who died young, those of the future Emperor Henry IV (1056–1106) in the charters of his father, Emperor Henry III (1039–56).98 There is also a record of Duke Robert in the year in which he set out on his pilgrimage to Jerusalem (1035) sending his young son (Willelmus adhuc puerulus) to the abbey of Saint-Pierre de Préaux, recently founded by Humphrey de ‘Vieilles’, one of the greatest aristocrats of the duchy, to witness the symbolic transfer of a grant that Robert had made. In this role William received a physical beating in the presence of Humphrey’s three sons, one of whom, Roger (known as Roger de Beaumont), was going to play a major part in William’s later achievements. Although, to modern sensibilities, this is a shocking narrative, it represents the sort of ritualized theatrical play-acting that was a frequent feature of William’s times and of 94 RADN, no. 60 (confirmante hoc Rotberto, inclito principe ac duce Normannorum, et filio eius Willelmo successor ipsius). 95 Ibid., nos. 66, 67, 68. 96 RADN, no. 67 (Willelmus, filius suus, qui et paternum donum dono suo confirmauit). 97 Garnett, ‘ “Ducal” Succession’, 91–3. 98 Ian S. Robinson, Henry IV of Germany (Cambridge, 1999), 23.

the first years

43

which several examples will appear in this book; that narrative passages often appear in the early thirteenth-century cartulary in which this one is preserved means that it can be trusted.99 It was intended as an aide-mémoire, a public performance to which all could refer in the event of any future disputes about the grant and, because William was young, so it was said, the beating would preserve memory for longer.100 Another indication of William’s importance is that, as Jumièges says, provision was publicly made for his education, with a tutor (magister) named Ralph appearing in a charter for the abbey of Saint-Wandrille.101 While we have no specific information about the education William received, there is no reason to be as pessimistic about it as some have been, and a good number of reasons to think that he would have received the normal education for a young boy of princely stock.102 Not only did he have tutors well into his adolescence, his later life shows a consistent awareness of the importance of the written word and of the literate learning of the Church. In all likelihood, therefore, he would have been able to understand Latin, even if he may not have been able to read.103 At the same time William would have been being taught the ethical principles and performative skills required of a member of the aristocratic elite and of a potential ruler. He would also have received the training to prepare him for a world in which violence and its management were culturally embedded, the skills to be acquired being, when necessary, the ability to kill and maim humans and to hunt animals on a regular basis. The life of such a child involved boisterous, and sometimes dangerous, play from a very early age.104 But it would also have meant that he was in regular and relatively close contact with his father, from whom he would be expected to learn, a situation that fits exactly with the evidence we have for William’s early years.105 99 RADN, no. 89 (Le Cartulaire de l’abbaye bénédictine de Saint-Pierre-de-Préaux [1034–1227], ed. Dominique Rouet [Paris, 2005], no. A1 [2]). 100 Emily Zack Tabuteau, Transfers of Property in Eleventh-Century Norman Law (Chapel Hill, NC, and London, 1988), 149. 101 GND, ii, 80–1; RADN, no. 80. 102 Cf. Barlow, William I, 12. 103 David Crouch, The Normans: The History of a Dynasty (London and New York, 2002), 54–61, also takes a positive view of William’s upbringing. For other references to William’s tutors, RADN, nos. 100, 102, 103. 104 Matthew Bennett, ‘Military Masculinity in England and Northern France, c.1050–c.1225’, in D. M. Hadley (ed.), Masculinity in Medieval Europe (London and New York, 1999), 71–88, at 73–6; Régine Le Jan, ‘Frankish Giving of Arms and Rituals of Power: Continuity and Change in the Carolingian Period’, in Frans Theuws and Janet L. Nelson (eds.), Rituals of Power from Late Antiquity to the Early Middle Ages (Leiden, Boston, and Cologne, 2000), 281–309, at 283–4; John Gillingham, William II: The Red King (London, 2015), 16–17. 105 In general, see Jonathan R. Lyon, ‘Fathers and Sons: Preparing Noble Youths to be Lords in Twelfth-Century Germany’, JMH, xxxiv (2008), 291–310, at 297–306.

44

william the conqueror

Paradoxically, a vivid portrayal of the qualities of a child who became a great warrior in adulthood occurs in the life of Hereward, one of William’s most redoubtable foes, which dates from the first half of the twelfth century. He was, so we are told, ready to provoke fights with other children and to quarrel with his elders; these were seen as the attributes of courage and spirit.106 While there is no reason to think that this model applied to the young William, it does sound plausible.107 What was of course also needed was an awareness of behavioural boundaries, something that the evidence we have of his parents’ conduct suggests was in place. The evidence also suggests that for the vast majority of the first eight or so years of his life William grew up in close proximity to his parents and with younger siblings and other young aristocrats around him. Those whom the sources explicitly identify as William’s childhood companions are his cousin Guy, who had been brought from Burgundy to grow up among his Norman kin, and the great associate of William’s adult years, William fitz Osbern; Orderic indeed says that William treated Guy as if he was his brother and William of Poitiers comments on the close relationship between William and William fitz Osbern.108 William’s half-brother Odo must surely have been among the group too. Jumièges reported that William’s father’s decision to go on pilgrimage to Jerusalem caused consternation among the Norman aristocracy.109 While such a reaction to a proposed extended absence on a long and risky journey is eminently plausible, it is obvious that the decision to go was not an impulsive one and that the duke’s departure was carefully prepared; the pilgrimage may indeed have been planned as many as four years prior to his departure.110 He was also following a pattern of behaviour common 106 Gesta Herwardi incliti exulis et militis, ed. T. D. Hardy and C. T. Martin, in Lestoire des Engles solum la Translacion Maistre Geffrei Gaimar, 2 vols., RS (London, 1889), i, 341–2; De Gestis Herwardi, ed. Alberto Meneghetti (Pisa, 2013), 78–9 (trans. Michael Swanton, Three Lives of the Last Englishmen [New York, 1984], 46–7). The evidence for the dating of the Gesta Herwardi is surveyed in Paul Dalton, ‘The Outlaw Hereward “the Wake”: His Companions and Enemies’, in John C. Appleby and Paul Dalton (eds.), Outlaws in Medieval and Early Modern England: Crime, Government and Society, c.1066–c.1600 (Farnham, 2009), 7–36, and especially pp. 31–2, with a mid-twelfth-century date suggested. 107 On the treatment of Hereward’s childhood, Joanna Huntington, ‘ “The Quality of his virtus proved him a perfect man”: Hereward “the Wake” and the Representation of Lay Masculinity’, in P. H. Cullum and Katherine J. Lewis (eds.), Religious Men and Masculine Identity in the Middle Ages (Woodbridge, 2013), 77–93, at 85–6. 108 GND, ii, 120–1; OV, iv, 82–3 (ut unicum fratrem honoraueram); GG, 8–9, 164–5. 109 GND, ii, 80–1. 110 On Robert’s decision to go on pilgrimage, note VIII anno regni nostri, quarto quo et Hierusalem petiturus ibi licentiam eundi a Deo et sanctis eius petii in a charter of January 1035, RADN, no. 90. The matter is not, however, clear-cut, since the key word (quarto) occurs in an eighteenth-century copy of the charter from a lost vidimus of the French King Philip IV, but not in the eighteenth-century copy, which claimed to have used the lost original. For this suggestion, Crouch, The Normans, 52.

the first years

45

among his contemporaries, with Count Fulk Nerra of Anjou’s four pilgrimages to Jerusalem that took place in 1003–5, 1009–11, 1036, and 1039 being the most famous parallel examples.111 In late 1034 and early 1035, the greatest nobles of the duchy were called together and, in proceedings that closely resembled those by which previous Norman rulers had designated their successors, agreed that William should effectively take Robert’s place and that he should become duke if Robert failed to return. What should above all be emphasized is the participation of the great magnates of the duchy and their taking of oaths to William. Jumièges, who had read the accounts of earlier successions in the history of Dudo of Saint-Quentin, produced a record that is worth quoting in full: He presented them with his only son William and earnestly besought them to choose him as their lord in his place and accept him as military leader (militieque sue principem). In spite of the boy’s tender years everyone in the place (oppidum) rejoiced in his encouragement and in accordance with the duke’s decree readily and unanimously acclaimed him their prince and lord and pledged him their fealty with inviolable oaths.112 The framework of consultation and oaths set out here is of profound importance for several central episodes in William’s later life. In the context of the proceedings of 1035, Jumièges stressed the role of Archbishop Robert; William of Malmesbury, who had access to the lost sections of William of Poitiers, locates the assembly at the great ducal palace at Fécamp.113 Ralph Glaber mentions that the French king Henry I, who had recently received crucial support from Duke Robert in 1033 against his mother’s attempt to depose him, confirmed the arrangements, an affirmation which is also mentioned by Malmesbury.114 The evidence of what is in effect the foundation charter of the abbey of Montivilliers that the nonNormans Count Baldwin IV of Flanders and Count Enguerrand I of Ponthieu were at Robert’s court shortly before his departure on pilgrimage, and the presence in the party which set out for Jerusalem of Count Drogo of the Vexin, the husband of Edward and Alfred’s sister Godgifu, along with the role played in Normandy after 1035 by the Breton counts Alan 111 On these, B. S. Bachrach, ‘The Pilgrimages of Fulk Nerra, Count of the Angevins, 987–1040’, in Thomas F. X. Noble and John J. Contreni (eds.), Religion, Culture and Society in the Early Middle Ages: Studies in Honor of Richard E. Sullivan (Kalamazoo, 1987), 205–17; idem, Fulk Nerra, 82, 113–16, 227–8, 253, 278–80. 112 GND, ii, 80–1. See above all Garnett, ‘ “Ducal” Succession’, 89–97. 113 GND, ii, 80–1; GR, i, 426–7. The main elements of the designation also appear in Brevis Relatio, 26. 114 Rodvlfi Glabri Historiarvm Libri Qvinque, 204–5 (ed. Arnoux, 258–9); GR, i, 426–7.

46

william the conqueror

and Eudo, all demonstrate extensive negotiations with neighbouring princely powers to safeguard the duchy during Robert’s absence.115 While Count Baldwin’s death later in 1035 undoubtedly weakened the network, it is hard to think that more could have been done to guarantee Normandy and William’s security during his father’s pilgrimage. Thought must also have been given to the young William’s likely potential before all this ritualizing legitimation was undertaken. Although nothing can be certain in the case of a child aged seven or eight, the fact that William was in adulthood a physically large and strong man suggests that the characteristics essential to medieval rule were already observable and were being nurtured. While William of Poitiers was undeniably setting out to convince his readers of William’s excellence, he did choose his illustrations carefully, with precocity being one of the qualities that he emphasized.116 Precocity was also a characteristic mentioned in the early twelfth-century Brevis Relatio; William was especially loved by his father not only because he was an only son, but also because he was notably strong and handsome for his age.117 And if his parentage was talked about, as suggested both by Glaber’s apologia and the fact that it continued to be referred to in the following decades, it cannot have been thought a significant obstacle to William’s succession. Moreover, in contrast to the contemporary tangled succession politics of the English kingdom, in 1035, in Normandy, there was no real alternative to the young William in sight. The fact that the most powerful and prominent members of the ducal kin, namely, Archbishop Robert of Rouen, Count Gilbert, William’s father’s cousin, who controlled territories around Eu and Brionne, and Bishop Hugh of Bayeux, with extensive lands around Ivry and Breteuil, were all of an older generation and therefore both ageing and not themselves interested in the succession, also favoured a peaceful transition if Robert did not return from his pilgrimage. For them to deny the legitimacy of the offspring of Robert and Herleva’s relationship would have been tantamount to a rejection of the legitimacy of their own careers. It is beyond belief that such people would have wanted to throw away the huge achievements of their lifetimes by supporting a child with no prospects.118 Furthermore, the principality that William was going to inherit, whatever differences there are in historians’ interpretations of the detail, was a well-established and powerful

RADN, no. 90. GG, 6–7 (plus intelligentia rerum honestarum et ui corporis quam aetate asultus). 117 Brevis Relatio, 25. 118 Richard Allen, ‘Praesul praecipue et venerande: The Career of Robert, Archbishop of Rouen, 989–1037’, in Leonie V. Hicks and Elma Brenner (eds.), Society and Culture in Medieval Rouen, 911–1300 (Turnhout, 2013), 153–83, illustrates this point very well. 115 116

the first years

47

one.119 William’s grandfather Richard II had been a notably successful ruler who had made contacts across Europe, had negotiated directly with the papacy, and had established a reputation as an outstanding princely ruler. It was also his reign that saw the completion of the first version of the Historia Normannorum, the foundation from which the written cultural memory of the Normans evolved. Within it are already present powerful elements of imitatio imperii, the notion of ideological continuum from the Carolingian Empire.120 Richard III’s short reign and the ups and downs of Duke Robert’s early years notwithstanding, Richard II’s sons had not dissipated his achievements. The Norman Church was dynamic and evolving rapidly. The establishment of Normans and others from northern France in southern Italy was well under way, a turbulent process that stimulated networks and activity across Europe. As early as the 1050s, it was being suggested that Duke Robert’s death on the return journey from Jerusalem at Nicaea (now Iznik) in modern Turkey in early July 1035 was the result of poisoning. But, as in most eleventh-century cases of a sudden and apparently inexplicable death, it is probably prudent to attribute his demise to natural causes rather than to malevolence.121 The memory in Normandy was that Robert’s progress to Jerusalem and his alms-giving had been remarkably lavish affairs, the message being that his was an example to be followed.122 In accordance with Robert’s wishes, William seems to have been immediately accepted as duke after the news reached Normandy; William of Malmesbury, in language that may echo William of Poitiers’s Gesta Guillelmi, whose lost early sections he may well have been using, presents an image of cooperation at the start of William’s reign.123 In terms of his personal development, William was indisputably exposed to significant risks. In particular, the loss of his father at around the age of eight, whose memory, as we shall see, he seems to have revered, could well have been important. So too may have been his father’s early political 119 Contra Douglas, WC, 7 (‘But it [i.e. the Norman Conquest of England] was not inevitable, and it came from a province which some forty years before the Norman Conquest of England showed few signs of its future achievement.’). For a survey of recent approaches to Normandy’s early history, Mark Hagger, ‘Confrontation and Unification: Approaches to the Political History of Normandy, 911–1035’, History Compass, xi (2013), 429–42. 120 Thus, most recently, Pierre Bauduin, ‘Richard II de Normandie: figure princière et transfers culturels (fin dixième–début onzième siècle)’, ANS, xxxvii (2015), 53–82; Benjamin Pohl, Dudo of Saint-Quentin’s ‘Historia Normannorum’: Tradition, Innovation and Memory (Woodbridge and Rochester, NY, 2015), passim, and esp. pp. 124–9. Both contain references to the voluminous bibliography on the subject. 121 Inventio et Miracula sancti Uulfranni, 47. Bachrach, Fulk Nerra, 228–9, suggests that Count Fulk Nerra of Anjou could have been responsible for Robert’s death. 122 Above all, GND, ii, 80–5. 123 GR, i, 426–7.

48

william the conqueror

insecurity associated with the suspicions that he had murdered his brother, which could have caused great anxiety in the young child. In a different way, the extensive religious patronage must have made an impression on him. But it is every bit as important not to pathologize the impact of the events of William’s childhood as it is not to make assumptions about their significance for the politics of the Norman Conquest of England. In terms of how resilience is understood as a dynamic concept, here were manifestly significant protective psychological factors in place, both intrinsic to William as an individual and in terms of supportive external resources, which could mitigate the disruptive factors. His later life suggests a resilient personality. It is nowadays recognized that responses to all manner of adversities can be extremely diverse. Apparently negative experiences may have either a sensitizing effect or a strengthening one in relation to the response to later stresses and adversity.124 It is surely important that, however wayward Robert’s political and personal behaviour may at times have been, William appears to have been the child of devoted and responsible parents and that he was brought up to expect an aristocratic lifestyle. The dramatic promotion of the cause of the English exiles and the grand religious and ceremonial occasions he would have witnessed would surely have imprinted themselves on his memory. However, for all the careful arrangements that had been made, it was nonetheless well-nigh certain that troubled times lay ahead, it being as good as a rule of nature that a medieval minority was going to be a period of turmoil. This is what the Bible told everyone to expect. Looking back from the perspective of the early twelfth century, William of Malmesbury introduced the next phase of William’s life by platitudinously quoting the famous passage from Ecclesiastes, ‘Woe to the land whose king is a child’.125

124 For a survey of these issues, Michael Rutter, ‘Resilience as a Dynamic Concept’, Development and Psychopathology, xxiv (2012), 335–44. 125 Ecclesiastes 10:16; GR, i, 426–7.

Chapter 2

FROM CH I L D HO O D I N TO AD O LESC ENC E The Battle of Val-ès-Dunes in 1047, in the words of William of Poitiers, ‘brought civil war in Normandy to an end for a long time’.1 With significant assistance from the French king, Henry I, it put William in a position to rule Normandy much more authoritatively than before. Edward the Confessor’s election as king of the English in 1041–2 created circumstances that were in due course to shape the whole of William’s life and the histories of peoples and nations. It is, however, crucial not to view Val-ès-Dunes as a simple end of either a process or of a period in William’s life. It was, in fact, in a lot of ways sui generis because of the specific character of the people involved in the revolt and also because it was a manifestation of a universal political phenomenon, namely, of rivalries for influence at court. Likewise, in relation to Edward the Confessor’s succession, a host of variables, including such factors as births and deaths that did or did not happen, could have interrupted what William of Poitiers more than anyone else wanted the world to believe was an inexorably smooth progression. An assessment of the significance of the years between 1035 to 1047, both in terms of the development of William’s personality and of the ways in which it can be seen as a preparation for the conquest of England, poses profound interpretative problems. It can only be accomplished when issues as basic as straightforward chronology have been tackled. For example, there is considerable debate in the available literature about the date when William began formally to rule.2 It is also crucial to grasp how slender, and in certain respects how prejudiced, are the two main narrative sources, William of Jumièges’s Gesta Normannorum Ducum and Orderic Vitalis’s

GG, 10–11 (bella domestica apud nos in longum sopiuit). David Bates, ‘The Conqueror’s Adolescence’, ANS, xxv (2003), 1–18, at 3–5. Douglas dated the end of William’s minority to 1047, WC, 52. In favour of c.1042 are Michel de Boüard, ‘Sur les origines de la Trêve de Dieu en Normandie’, AN, ix (1959), 169–89, at 174–5; idem, Guillaume le Conquérant, 120–2; Barlow, William I, 11; Mark Hagger, ‘How the West was Won: The Norman Dukes and the Cotentin, c.987–1087’, JMH, xxxviii (2012), 20–55, at 36. For a treatment of this period similar to mine, Dominique Barthélemy, L’an mil et la paix de Dieu. La France chrétienne et féodale, 980–1060 (Paris, 1999), 524–30. 1 2

49

50

william the conqueror

expansions thereof, the latter in many places effectively constituting an independent source (Plate 1).3 CONSTRUCTING A NARRATIVE Jumièges devoted a single short section of one capitulum to the disturbances that occurred within Normandy during this period, mentioning without explanatory narrative a war involving Hugh I de Montfort and Walchelin de Ferrières, in which both were killed, and the violent deaths of Count Gilbert of Brionne, Osbern the steward, and a tutor of William named Turold, all of whom were directly responsible for William’s welfare. In the two following capitula, he wrote a little more extensively about the interventions in the border regions of Normandy by King Henry I.4 The three capitula inserted by Orderic added detail on the circumstances of the killings of Count Gilbert and Osbern the steward, the second of which he said took place within William’s private quarters. He also introduced a completely new narrative of a war between Roger I de Tosny and the family of Humphrey de ‘Vieilles’; and he added some material about Humphrey’s descendants, including Roger de Beaumont and his sons, and some genealogical information about other families.5 Elsewhere, in a speech attributed to Count Alan of Brittany’s son, Count Conan II, Orderic mentions Alan’s death, which in all probability took place in 1040 and which he consistently says was due to poisoning, and in one place indicates that it occurred while he was besieging the castle at Montgommery.6 Most of his other additional information derives from his special knowledge of the Giroie and the Montgommery families. He added further detail in the Historia Ecclesiastica, most notably in the speech that he invented for William on his deathbed, including the statement that the young duke had to be secreted away in poor people’s houses for his safety. But most of what Orderic wrote there is no more than a summary of his interpolations into the Gesta Normannorum Ducum.7 It is possible that he drew on the lost early sections of Poitiers’s Gesta Guillelmi, but there is no way of knowing what the borrowings he made were, if indeed there were any. 3

For Orderic’s interpolations, GND, i, cxxv–cxxvi. Ibid., ii, 92–3, 100–3. 5 Ibid., 92–9. For compelling arguments for abandoning the appellation ‘Humphrey de Vieilles’, Véronique Gazeau, ‘De l’usage des souscriptions et de l’utilité de leur étude: remarques sur l’aristocratie normande dans la première moitié du XIe siècle: L’exemple d’Onfroy (dit ‘de Vieilles’)’, in Crouch and Thompson (eds.), Normandy and its Neighbours, 215–21. 6 GND, ii, 164–5. The story of poisoning is repeated without elaboration in OV, ii, 304–5; iii, 88–9. For the siege of Montgommery, ibid., iv, 76–7. 7 OV, iv, 82–3. For the summaries, ibid., iii, 88–9; iv, 82–3. 4

from childhood into adolescence

51

A second problem lies in the way in which Jumièges and Orderic characterized the years of William’s minority in terms of the construction in many places of earthworks and fortified strongholds, the development of plots, and the lighting of fires, dramatic language that was also used by other contemporaries who wrote about the events of the time.8 To this Orderic added the enormously influential statement that William was despised by the indigenous nobility because he was a bastard, and that he was especially hated by the descendants of the Dukes Richard, a group who are often categorized in francophone historiography as ‘les Richardides’.9 This mishmash of description and imagery shares much with similar language that was commonplace in eleventh- and twelfth-century sources to describe periods of disorder, and it must therefore be viewed in the context of debates about norms of culturally structured violence within eleventh-century society.10 Also relevant is discussion about the role of castle-building in relation to the balance between the public power of what, for the sake of convenience, we can term the state and the private power of the aristocracy, debates that were once located within the framework of what was once described as ‘the feudal revolution’ – and possibly still should be, at least as a means to foreground the social changes that were taking place.11 To enter into a full analysis of all this would require a substantial diversion and is unnecessary. Suffice to say, therefore, that what happened in Normandy was not unique. In terms of ‘the feudal revolution’, the issue in relation to Normandy is the not insoluble one of taking into account circumstances and changes where the state did not fail. And in terms of the events of the 1040s, the situation fits neatly into the framework supplied by other analyses of the times when the ruler of a kingdom or a principality was a child. My general perspective is that the aristocracy existed in a symbiotic relationship with the state in which self-seeking opportunism, self-protection, and honour were influential values, and within which 8 GND, ii, 92–3. See also, Inventio et Miracula sancti Vulfranni, 52 (Preterea discordia principum inter se conflictantium propter predicti principis pueritiam Nortmanorum patriam uehementer atterebat, ferro rapinis et flammis urentibus cuncta longe et lateque incursabat). 9 GND, ii, 96–7 (nobilibus indigenis et maxime ex Ricardorum prosapia natis despectui erat utpote nothus). 10 For warnings similar to mine from a different perspective, John Gillingham, ‘1066 and the Introduction of Chivalry into England’, in Garnett and Hudson (eds.), Law and Government in Medieval England and Normandy, 31–55, at 34–6 (reprinted in idem, The English in the Twelfth Century: Imperialism, National Identity and Political Values [Woodbridge, 2000], 209–29, at 212–14). 11 There is a huge bibliography. For some recent general reflections, West, Reframing the Feudal Revolution, 255–63; Chris Wickham, ‘The “Feudal Revolution” and the Origins of Italian City Communes’, TRHS, 6th ser., xxiv (2014), 29–55.

52

william the conqueror

violence was a part of the repertoire. All needs to be kept within an analytical framework that takes account of some factors tending towards equilibrium and others towards disequilibrium, with every situation needing to be analyzed as both an expression of its specific circumstances and of general norms.12 On this basis, these years should be seen as a disruptive rather than destructive phase within the long-term history of the Norman duchy. Individuals and institutions should be thought of as largely supportive of the young William, while at the same time safeguarding their own interests as much as was deemed necessary. This framework suggests that we are dealing in Normandy with a rapid construction of small fortifications in the countryside additional to existing aristocratic residences, from which to fight aggressive and defensive campaigns and to counter acts of intimidatory, localized devastation of the countryside, observable in other episodes for which the evidence is more plentiful.13 Such violence was strategic and not, as used to be assumed, anarchic.14 More specifically, Orderic’s comments about the descendants of the Dukes Richard, the supposed ‘les Richardides’, can only be tackled through prosopographical and comparative analysis. When this is done, the interpretation common to so much writing collapses.15 It is also worth observing that the categorization ‘les Richardides’ is not a translation of what Orderic wrote; it is a modern invention that misrepresents it.16 In terms of the chronology of disorder, it is crucial to recognize that the wars described in the two versions of the Gesta Normannorum Ducum almost all date to the period between 1040 and 1042. Thus, the war between Hugh I de Montfort and Walchelin de Ferrières is said to have happened before Count 12 For a statement of an approach that seems to me to have a lot in common with mine, Stephen D. White, ‘Preface’, in Feuding and Peace-Making in Eleventh-Century France (Aldershot and Burlington, VT, 2005), viii–ix, with the essays in the volume being extended essays on the theme. 13 In general, Richard E. Barton, Lordship in the County of Maine, c.890–1160 (Woodbridge, 2004), 146–96; David Crouch, The English Aristocracy 1070–1272: A Social Transformation (New Haven, CT, and London, 2011), 99–159, with references to the huge literature. For different perspectives on twelfth-century periods that have a lot in common with this period and which take account of European debates, David Crouch, The Reign of King Stephen, 1135–1154 (London and New York, 2000), 150–3; Hugh M. Thomas, ‘Violent Disorder in King Stephen’s England: A Maximum Argument’, in Paul Dalton and Graeme J. White (eds.), King Stephen’s Reign (1135–1154) (Woodbridge, 2008), 139–70. 14 For a classic statement that epitomizes this approach, François Neveux, La Normandie des ducs aux rois, Xe–XIIe siècle (Rennes, 1998), 112 (‘Mais de nombreux seigneurs normands avaient pris goût à l’anarchie, dont ils tiraient un maximum de profit personnel. Ils voulaient à tout prix éviter le retour en force d’une autorité qui les contrôlerait, en contrariant leurs ambitions et leurs intérêts’). 15 Ibid., 112, for ‘les Richardides’. The usage is so widespread as to make it impossible to list all instances. 16 See above, note 9.

from childhood into adolescence

53

Gilbert of Brionne’s death, which Jumièges explicitly dates to not long afterwards and which probably occurred before February 1041.17 The deaths of Count Alan of Brittany and Osbern the steward, son of Herfast, the nephew of the duchess Gunnor, also occurred during this period; Count Alan’s in all probability on 1 October 1040 and Osbern’s after Count Gilbert’s, but before April 1042.18 The war between the Beaumont and Tosny families, while treated by Orderic in his Historia Ecclesiastica as one of these wars and traditionally dated to c.1040, must, however, have occurred later, perhaps in 1043 or 1044, since a charter shows Roger de Tosny alive after 1042 when William had reached adolescence, at approximately fourteen years of age.19 Taken alongside Jumièges’s use of the life-cycle language of boyhood (pueritia) and adolescence (adolescentia), one consequence of this chronology of the disorders is to show that William was pushed into the foreground from around 1042. Thus, Jumièges says that the young William made his uncle William count of Arques ‘when the duke was flourishing in adolescence’, with a charter showing that William was a count probably by 1042, and certainly by 1044.20 In contrast, Jumièges wrote that Count Gilbert’s death happened while William was still a puer.21 He presented William’s first active military role as the capture of Falaise from Thurstan Goz, an event usually dated to c.1043.22 In contrast again, in dealing with the attack by King Henry I on Tillières-sur-Avre, Jumièges says that the decision not to resist was taken by ‘the Normans’ and consistently describes William as a puer in relation to the episode.23 When it comes to other early sources, notably William of Poitiers and the Inventio et Miracula sancti Vulfranni, both favour a date of c.1042 for the start of William’s formal rule. Thus, Poitiers believed that William began to rule when he was knighted and, although he supplies no date for this, the organization of the Gesta Guillelmi explicitly GND, ii, 92–3. For the date of Alan’s death, André Oheix, ‘La date de la mort d’Alain III, duc de Bretagne’, in Bulletin de la Société d’émulation des Côtes-du-Nord, li (1913), 93–100; Les actes des ducs de Bretagne, 48–9, 272–5 (no. 45). For Osbern’s death, GND, ii, 92–3, note 6. 19 OV, iii, 88–9; iv, 302–3. For Roger de Tosny, RADN, nos. 100, 102. For the date of c.1040, Lucien Musset, ‘Aux origines d’une classe dirigeante: les Tosny, grands barons normands du Xe au XIIIe siècle’, Francia, v (1977), 45–80, at 55. 20 GND, ii, 102–3 (Hic enim Willelmus a duce iam in adolescentia pollente comitatum Talogi percipiens); RADN, no. 100. The charter must date from 1044 at the latest, since it mentions Abbot Robert of Jumièges who accompanied Edward the Confessor to England. While the date when Robert moved to England is not precisely known, it is generally assumed that he was among those who joined Edward there in 1042, H. E. J. Cowdrey, ‘Robert of Jumièges [Robert Champart] (d.1052/1055)’, ODNB; NMProsopographie, 150–1. Ælfweard, his predecessor as bishop of London, died on 25 or 27 July 1044. 21 Gilbert is described as tutor ipsius pueri sed domini, GND, ii, 92–3. 22 Ibid., 100–3. For the suggested date, ibid., 101, note 6. 23 Ibid., 100–1. 17 18

54

william the conqueror

places it some considerable time before the Battle of Val-ès-Dunes.24 The Inventio, largely completed by 1053–4, specifically assigns the time of the conflicts among the aristocracy of Normandy to the period when William was a boy.25 In addition, an account of the miracles of St Catherine, written at Rouen before 1054, says that the period of violence and disruption lasted for seven years up until the time when the duke gained strength, also perhaps confirming a date of 1042, and certainly one shortly afterwards.26 Orderic’s statement that William selected Ralph de Gacé as his guardian (tutor) and the commander of his forces after Count Gilbert’s death also confirms this chronology.27 William of Malmesbury interpreted the early sources available to him as indicating that William was knighted at the earliest possible opportunity.28 The conclusion to draw from this chronology is that the death of the elder statesman Archbishop Robert of Rouen in 1037 undoubtedly contributed to the destabilization of the regime supporting the young William. However, without any evidence of significant violence until approximately three years later, the immediate effect of his passing cannot have been that great. What we know, therefore, suggests that there was a long period when the rule of William’s guardians maintained a relative peace within the duchy, that the years from c.1040 to c.1042 were a time when the previous political equilibrium was compromised rather than overthrown, and that the revival of the norms of ducal rule began when William started formally to act as duke in c.1042. That this was also the canonical age at which marriage could be contemplated also amended his political potential in the eyes of contemporaries. PUTTING WILLIAM’S MINORITY IN PERSPECTIVE A consequence of the relatively short average life expectancy of medieval men and women is that periods when a child was the titular ruler of a kingdom or a principality were a frequent occurrence. There are, therefore, several tenth- and eleventh-century minorities that can be used to bring the events of William’s minority into comparative focus; for example, GG, 6–9. Inventio et Miracula sancti Vulfranni, 52 (propter predicti principis pueritiam). 26 Eodem quoque tempore in potentia ducis sui tota invicem saeviente Normannia . . .; hoc etiam septem annis fieret, dum eorum dux in vires resurgeret, ‘Sanctae Catharinae virginis et martyris Translatio et Miracula Rothomagensia’, ed. A. Poncelet, Analecta Bollandiana, xxii (1903), 423–38, at 438. 27 Rodulfum de Wacceio ex consultu maiorum sibi tutorem eligit, et principem militie Normannorum constituit, GND, ii, 98–9. 28 GR, i, 426–7. 24 25

from childhood into adolescence

55

those of Emperors Otto III (983–1002) and Henry IV (1056–1106), the French king Philip I (1060–1108), and the territorial princes Conan II, duke of Brittany (1040–66), and Herbert II, count of Maine (1051–62). The passage in Ecclesiastes 10:16, ‘Woe to the land whose king is a child’, was of course well known in the eleventh and twelfth centuries. And, for historians of this period, it seems so universally applicable that Gerd Althoff has recently emulated William of Malmesbury by using it as the title of a substantial sub-section of a chapter in a book devoted to the emperor Henry IV.29 While all minorities are naturally to an extent sui generis, the general picture in every instance over many centuries is of the development of factional rivalries around a young ruler, the weakening of the controlling and adjudicating roles at the heart of effective kingly and princely rule, and the strengthening of localized power structures. Child rulers were sometimes placed in great danger, attempts to displace them were quite common, and violence among the aristocratic elite was frequent.30 Yet, in all cases, whatever view is taken of events and whatever the scale of change, the crucial point is that everywhere the core held. The state survived and so did the child ruler, because there was always a sufficiently dominant elite committed to the young ruler’s survival. Even in the case of what is usually judged to have been a relatively calm minority, that of King Philip I of France, Gervase, the former bishop of Le Mans who was by this time archbishop of Rheims, still felt it necessary to write firstly to Pope Nicholas II (1059–61) to express his fears about the potential for disorder and then to Pope Alexander II (1061–73) to say that what he had feared was actually happening.31 An earlier minority, that of Otto III, while also generally seen as a time when the regencies of his mother and grandmother were a success, began with an attempt by his uncle, Duke Henry the Quarrelsome, firstly to seize control of the threeyear-old boy and subsequently to supplant him; and this despite the fact that Otto had been formally designated by his father, Otto II. The period was also characterized by the development of competing factions around both of the possible rulers.32 That Otto III’s mother Theophanu was the niece of a Byzantine emperor shows that the social status of a young ruler’s Gerd Althoff, Heinrich IV. (Darmstadt, 2006), 41–66. In general, Thilo Offergeld, ‘Reges Pueri’: Das Konigtüm Minderjähriger im frühen Mittelalter, MGH Schriften, vol. 50 (Hanover, 2001), 815–34. 31 RHF, xi, 498–9; cited in Andrew W. Lewis, Royal Succession in Capetian France (Cambridge, MA, and London, 1981), 46 (‘such outbreaks of violence, which were frequent in any case, were inevitable during a minority, and they had little to do with Capetian blood-right’). 32 Gerd Althoff, Otto III, trans. Phyllis G. Jestice (University Park, PA, 2003), 29–51; Dominik Waßenhoven, ‘Swaying Bishops and the Succession of Kings’, in Körntgen and Waßenhoven (eds.), Patterns of Episcopal Power, 89–109, at 89–99. 29 30

56

william the conqueror

kindred was no sure protection. The relatively lowly status of William’s mother must not therefore be seen as automatically being a major disadvantage for him, if indeed it was one at all. The narratives of other minorities are full of tales of overbearing relatives and guardians; those involving Count Eudo and Count Conan II of Brittany, and Count Geoffrey Martel and Count Herbert II of Maine, had a big impact on William’s life. In Henry IV’s case, after his father the emperor Henry III had set up a council to support his wife Agnes’s regency, arrangements for the boy king’s early years look to have been as securely made as was reasonably possible. Their disruption over the following decade and the fractiousness of supposedly collaborative relationships is variously described in the remarkably copious contemporary source material, usually written from the perspective of one of the high-ranking churchmen involved.33 To read vivid and detailed sources such as the Annals of Lampert of Hersfeld is a powerful warning that the limited and largely one-dimensional surviving sources for William’s childhood and adolescence must not be taken at face value.34 As important as Lampert for comparative purposes is Adam of Bremen’s extensive narrative of the career of Archbishop Adalbert of Hamburg/Bremen, one of the chief figures in Henry’s regency council. Starting with a well-nigh archetypal piece of medieval misogyny, condemning the rule of a woman and following this by referring to the quarrelsomeness of the aristocracy, Adam then says that a semblance of order was only restored when Archbishop Adalbert and Archbishop Anno of Cologne became co-regents (in 1063), proceeding to praise Anno, but at the same time emphasizing Adalbert’s moral superiority and criticizing Anno’s avarice. Adam finally said, however, that, for all Adalbert’s good qualities, the consequence of his time spent at court for his diocese was economic ruin; he was too eager for worldly glory. Public service of the kind he undertook, although meritorious, came at a price. Minorities damaged subjects as well as child rulers and states.35 From a relatively detached perspective, the decline in the ethical standards of kingly rule during Henry’s minority was enunciated by at least one

33 For a survey of the historians’ and participants’ perspectives, Claudia Zey, ‘Vormünder und Berater Heinrichs IV. Im Urteil der Zeitgenossen (1056–1075)’, Vorträge und Forschungen, lxix Heinrich IV. (Ostfildern, 2009), 87–126. 34 The text for the years 1056 to 1065 covers fifty-one pages in the modern German edition, which has both the Latin text and a German translation, Lampert of Hersfeld, Annales, ed. Adolf Schmidt and Wolfgang Dietrich Fritz (Berlin, 1962), and forty-four pages in the modern English-language translation, The Annals of Lampert of Hersfeld, trans. I. S. Robinson (Manchester, 2015), 65–108. 35 Adam of Bremen, Gesta Hammaburgensis ecclesiae Pontificum, 175–80 (trans. Tschan, 141–3).

from childhood into adolescence

57

writer.36 And the scale of the rivalries and turbulence around him was such that, as a twelve-year-old, he was abducted by a group of princes, led by Archbishop Anno of Cologne, who were at odds with the policies of the Empress Agnes. This abduction, the so-called Coup of Kaiserwerth (April 1062), involved the boy king being kidnapped from a royal palace near Düsseldorf, confined on a boat on the Rhine and taken to Cologne, and him trying to escape by diving into the huge and dangerous river. Yet for all the conflict of the times, the verdict is nonetheless possible for recent historians that royal power was successfully preserved during Henry’s minority. This was partly because, in many respects, the turbulence was an airing of rivalries and feuds from previous times, and the period of child rule a time when they could be more overtly pursued.37 Within this perspective, all the wars described in the Gesta Normannorum Ducum fall into two categories, some being the result of local competitive rivalries, others the result of battles for influence at the ducal court. What have been called the cultures of feud and of socially acceptable honour-based violence, both phenomena constantly present over many centuries in the medieval West, are also relevant to the interpretation of these years of William’s life.38 Within Normandy, a case study from Duke Robert’s time is provided by the story in the early twelfth-century Vita of Herluin, the first abbot of Le Bec, in which Herluin’s lord, Count Gilbert of Brionne, was dissuaded by intermediaries, including emissaries from the duke, from attacking an enemy.39 For all that this indicates a ducal peacekeeping role that was often effective, and acknowledges that the adult William did seek to limit the frequency of feud-motivated behaviour, even at the height of his powers, in 1075, the only restriction that he

36 Thus, Annales Altahenses maiores, ed. Edmund von Oefele, MGH Scriptores (Hanover, 1891), 56 (s.a. 1060: Inicia dolorum haec. Rex enim puer erat, mater uero utpote femina his et illis consiliantibus facile cedebat, reliqui uero palatio praesidentes omnino avariciae inhiabant, et sine pecunia ibi de causis suis nemo iusticiam inueniebat, et ideo fas nefasque confusum erat). 37 Althoff, Heinrich IV., 44–5; cf. Robinson, Henry IV of Germany, 25–62, whose tone is less optimistic. 38 See especially Paul R. Hyams, Rancor and Reconciliation in Medieval England (Ithaca, NY, and London, 2003), 3–68; idem, ‘Afterword. Neither Unnatural nor Wholly Negative: The Future of Medieval Vengeance’, in Susanna A. Throop and Paul R. Hyams (eds.), Vengeance in the Middle Ages: Emotion, Religion and Feud (Farnham, 2010), 203–20, esp. pp. 204–14. See also Matthew Bennett, ‘Violence in Eleventh-Century Normandy: Feud, Warfare and Politics’, in Guy Halsall (ed.), Violence and Society in the Early Medieval West (Woodbridge, 1998), 126–40. 39 ‘Vita Herluini’, in The Works of Gilbert Crispin Abbot of Westminster, ed. Anna Sapir Abulafia and G. R. Evans, Auctores Britannici Medii Aevi, viii (Oxford, 1986), 186–7 (J. Armitage Robinson, Gilbert Crispin, Abbot of Westminster: A Study of the Abbey under Norman Rule [Cambridge, 1911], 88).

58

william the conqueror

could achieve was to secure an agreement that a revenge killing was permitted only when it was a father or a son who had been killed.40 In being judged fit, like Philip I, Henry IV, and Otto III, and many others in earlier times, to participate in rule at around the age of fourteen, William’s life was being shaped by another norm. This criterion was the age at which adolescence was said to begin in the encyclopaedic work known to all the religious advisers of tenth- and eleventh-century rulers, the seventh-century Etymologiae of Isidore of Seville; in this work, the standard terms referring to boyhood, adolescence, and manhood – puer, adolescens, and vir – progressively categorize the advance to maturity.41 At fourteen, an adolescent was deemed old enough to procreate, but not yet fully able to take part in all manly activities. He was ready to be knighted and to undertake military activity, but he might well remain in tutelage, albeit less constrained by guardians than before. But his role from this point on was to be at the heart of a society of warriors. In consequence, like all aristocratic young men of his time, in later childhood and adolescence William would have grown up within a military culture where culturally structured violence was the norm, surrounded by young men of his own age among whom there would have developed camaraderie and lifelong friendships, as well as rivalries and enmities.42 In life-cycle terms, his progress to maturity seems to have been entirely normal. THE RULE OF THE CHILD William of Jumièges names Count Gilbert of Brionne, the son of a halfbrother of Duke Richard II, as the guardian appointed when news of Duke Robert’s death became known in Normandy. William of Malmesbury 40 For the text, Haskins, Norman Institutions, 278, note 9. For wide-ranging commentaries on this whole subject, John G. H. Hudson, ‘Faide, vengeance et violence en Angleterre (ca 900–1200)’, in Dominique Barthélemy, François Bougard, and Régine Le Jan (eds.), La vengeance, 400–1200 (Rome, 2006), 341–82; idem, ‘Feud, Vengeance and Violence in England from the Tenth to the Twelfth Centuries’, and Robert Bartlett, “ ‘Mortal Enmities”: The Legal Aspect of Hostility in the Middle Ages’, in Belle S. Tuten and Tracey L. Billado (eds.), Feud, Violence and Practice: Essays in Medieval Studies in Honor of Stephen D. White (Farnham and Burlington, VT, 2010), 29–53, 197–212. 41 For convenience, see The Etymologies of Isidore of Seville, trans. Stephen A. Barney et al. (Cambridge, 2006), 242, 243. Philip I, born in 1052, began formally to rule in either the last months of 1066 or the first ones of 1067, Henry IV, born 1050, in 1065, and Otto, born in 980, was seen as coming of age in 994. See also Le Jan, ‘Frankish Giving of Arms’, 285–7. 42 In general, Crouch, The English Aristocracy, 20–36, 99–132. The classic essay is Georges Duby, ‘Youth in Aristocratic Society: Northwestern France in the Twelfth Century’, in idem, The Chivalrous Society, trans. Cynthia Postan (London, 1977), 112–22. For the original French version, ‘Au XIIe siècle: les “jeunes” dans la société aristocratique dans la France du Nord-Ouest’, Annales: Economies, sociétés, civilisations, xiv (1964), 835–46.

from childhood into adolescence

59

does the same, but he was probably using Jumièges as his source.43 The location of Gilbert’s signum, along with William’s and the future King Edward’s, on the Fécamp diploma discussed in the previous chapter also points to his exceptional prominence. Orderic, however, twice names Count Alan III of Brittany, the son of Duke Robert’s sister Hawise, as performing the role, while elsewhere describing Gilbert as ‘father of the country’ (pater patrie).44 Gilbert was therefore surely the chief guardian, but, as kinsmen of status and experience, both he and Alan would have been expected to play active roles in protecting William.45 However, Alan, who was married to Bertha, a daughter of Count Odo II of Blois-Chartres, who had met a violent death in 1037, must have been significantly distracted by his involvement in preventing the northwards advance into Maine of the counts of Anjou. The apparent partnership in rule between him and his brother Count Eudo in Brittany nonetheless extended to both of them appearing constructively in Normandy before 1040.46 And Bertha, who was to live on into the mid-1080s, may well have been a much more significant long-term figure in William’s life than is usually realized. An inquiry that must date to before 1037, conducted by Archbishop Robert, Nigel the vicomte, and Count Eudo, which restored possessions to the cathedral church of Bayeux, indicates the continuation of the principles that had animated ducal rule in Duke Robert’s time, and the continued involvement therein of central figures from that period.47 The whole business may well have been a further attempt to stabilize relationships within the duchy’s topmost elite after Bishop Hugh’s quarrel with Duke Robert; that Hugh, who had become bishop of Bayeux in c.1011, was the son of Rodulf, count of Ivry, the half-brother of Richard I, meaning he was very much one of the supposed ‘Richardides’, makes this even more likely.48 Other indications of continuity are a charter containing a promise to restore to the abbey of Saint-Wandrille an estate near Argentan taken GND, ii, 92–3; GR, i, 426–7. OV, iii, 86–7; iv, 76–7, 82–3. Wace, Roman de Rou, pt. 3, lines 2,977–80, assigns the role to Count Alan. 45 For Gilbert, D. C. Douglas, ‘The Earliest Norman Counts’, EHR, lxi (1946), 129–56, at 139–40; Bates, Normandy before 1066, 64, 101. 46 The collaboration between Alan and Eudo can be illustrated in various ways and above all by Eudo’s frequent appearances in his brother’s charters, Les actes des ducs de Bretagne, passim. Note especially, Alanus et Egio Britannorum monarchi in a diploma dating to their minorities, ibid., no. 13. See further, with special reference to Normandy, Stéphane Morin, Trégor, Goëlo, Penthièvre: Le pouvoir des Comtes de Bretagne du XIe au XIIIe siècle (Rennes, 2010), 58–60. 47 Antiquus Cartularius Ecclesie Baiocensis, ed. V. Bourrienne, 2 vols. (Rouen and Paris, 1902– 3), i, no. 21. 48 GND, ii, 52–3. 43 44

60

william the conqueror

away by William’s father, presumably during his time in the Hiémois.49 The only recorded extensive grants to churches also reinforce the impression of an attempt to sustain continuity, since all were to Richard II’s monastic foundation of Fécamp, on whose behalf a life-lease in favour of Hugh, son of Hugh I de Montfort, was sanctioned; the presence among those who confirmed the second of the grants of Roger I de Montgommery and Roger, the son of Humphrey de ‘Vieilles’, the two nobles who had been the recipients of extensive Fécamp property then, further strengthens this argument.50 Those who bore witness to the transaction included the inner group of the ducal kindred and men who had been specifically favoured by Duke Robert. Archbishop Robert of Rouen was succeeded as archbishop by Malger, a son of Duke Richard II’s second marriage and therefore William’s uncle. At Evreux, where he had effectively been count, Robert was followed by his son Richard, who appears as ‘count of Evreux’ in an original charter dated 1038.51 These promotions, like the others that followed in the 1040s, conform to the established conventions whereby close ducal relatives were advanced to high secular or ecclesiastical office, arrangements that once more placed the emphasis on continuity. The possibility that the advent of a new political generation would prompt rivalries and conflicts that had not existed among their elders was, however, always present. When we turn to the wars recorded in the Gesta Normannorum Ducum, it is as good as certain, given the relative proximity of their lands in central Normandy, that the war between Hugh I de Montfort and Walchelin de Ferrières was driven by a local dispute over land and power. Resulting in the deaths of both of the main participants, it looks as if the classic pattern of revenge killings was followed through to its bloody, logical, conclusion.52 Its wider significance must be that by about 1040 those around the young William were no longer capable of containing aristocratic rivalries within the peace-keeping mechanisms that were a norm of effective ducal rule, and that the participants in the conflict did not perceive them as being capable of doing so. But, as far as we know, peaceful relations were restored once the feud had been played out through the two deaths. According to Orderic, Count Gilbert of Brionne’s death occurred when he was ambushed and killed by Ralph de Gacé, a younger son of Archbishop Robert of Rouen, and by Robert fitz Giroie. Ralph’s subsequent appearance as the leader of William’s army suggests that, for him, RADN, no. 95. Ibid., nos. 93, 94. On the Fécamp lands in central Normandy, Potts, Monastic Revival, 119–27. 51 RADN, no. 92. 52 GND, ii, 92–3. 49 50

from childhood into adolescence

61

the ambush was part of a struggle for power at court and that its outcome was a palace coup. Robert fitz Giroie is known to have had a dispute with Gilbert concerning Le Sap in central Normandy that went back to Duke Robert’s time, so he was presumably exploiting the situation to pursue a piece of personal, local, business.53 The likelihood that Count Gilbert had the self-aggrandizing characteristics associated with figures prominent in other minorities is suggested by acts of aggression such as his attack in Duke Robert’s time on Enguerrand I, count of Ponthieu;54 and, in all probability, by his expulsion of the wife and sons of the deceased William, count of Eu, from the castle of Eu, that seems to date from the later 1030s.55 Count Gilbert had made enemies and this led directly to his death. Yet evidence contained in a charter indicating that at least one of Gilbert’s killers was punished shows the continued operation of the coercive disciplinary power associated with the ducal office.56 In the case of Count Alan’s death on, in all probability, 1 October 1040, Orderic says that it happened at Vimoutiers (dép. Orne, chef-lieu du canton) and was attributable to poisoning by William’s partisans.57 However, a charter in the cartulary of the abbey of Saint-Pierre of Préaux says that Alan died at Fécamp.58 That poisoning was often used by Orderic and others to explain a sudden death, or one which seemed to have been politically advantageous to someone, raises the possibility that Alan did not die by violence at all, but was simply taken ill during the siege, subsequently going to Fécamp and dying there. This appears all the more likely when the context of the accusation is taken into account, namely, a tendentious speech attributed by Orderic to Alan’s son Count Conan II, during which he claimed that Alan was Duke Robert’s nominated heir in order to justify 53 GND, ii, 94–5; OV, ii, 24–5. See further Jean-Marie Maillefer, ‘Une famille aristocratique aux confins de la Normandie: les Géré au XIe siècle’, in Lucien Musset, Jean-Michel Bouvris, and Jean-Marie Maillefer (eds.), Autour du pouvoir ducal normand Xe–XIIe siècles, Cahiers des Annales de Normandie, vol. 17 (Caen, 1985), 175–206, at 201–2. On general context, Pierre Bauduin, ‘Une famille châtelaine sur les confins normanno-manceaux: les Géré (Xe–XIIIe s.)’, Archéologie médiévale, xxii (1992), 309–56, at 313–19; Hyams, Rancor and Reconciliation, 281. 54 OV, ii, 12–13. 55 Cartulaire de la Sainte-Trinité-du-Mont de Rouen, ed. A. Deville, in Cartulaire de Saint-Bertin, ed. B. E. C. Guérard (Paris, 1840), no. LXIX. See further, Douglas, ‘The Earliest Norman Counts’, 139–40; Bauduin, La première Normandie, 297. 56 OV, ii, 120–1. 57 GND, ii, 164–5. The story of poisoning is repeated without elaboration in OV, ii, 304–5; iii, 88–9. He also said that the death happened at Montgommery, ibid., iv, 76–7, but this difference is of no great significance, since both places show that what was involved was warfare against the Montgommery family. 58 RADN, no. 97; Cartulaire de Saint-Pierre-de-Préaux, nos. A1[11] and A153, with the contradictions in the sources discussed at no. A153. See also Douglas, WC, 409.

62

william the conqueror

the attack he proposed to make on William.59 Alan’s burial in the abbey of Fécamp in the manner of a member of the Norman ducal family makes it impossible to believe he was murdered by William’s supporters, since, by this act, his support for them was being commemorated.60 The context in which Count Alan’s death occurred was warfare against the family of Roger I de Montgommery, whose son Roger II was later to be one of William’s closest allies, but who were at war with William’s guardians at this point; Vimoutiers was one of their castles.61 This warfare also supplies some of the context for the death of Osbern the steward, killed, so Orderic tells us, in William’s domestic quarters by Roger I’s son William. Its background lay in a battle both for influence at court and for regional power. Roger had been installed by Duke Robert at Falaise and Argentan, effectively as his successor there, and as vicomte of the Hiémois. William’s guardians, for unknown reasons, had expelled him from both places and replaced him with Thurstan Goz. Roger’s exile in Paris at the time of Osbern’s killing shows that those supporting William had both the will and the muscle to enforce his banishment, but, self-evidently, they did not have the power completely to subdue the family.62 The retaliatory slaughter of Osbern’s killer by Barno de Glos, one of Osbern’s followers, seemingly brought the violence to an end. It is an interesting commentary on William of Jumièges’s modus operandi that he does not name Osbern’s killer. But he does insert an aside referring to ‘the madness of certain persons’ and adding that ‘these are the very men who now claim to be most loyal and who have received so many honours from the duke’.63 It suggests that, by the later 1050s when he was writing, a degree of circumlocution was deemed necessary when writing about those whom William had chosen as his favoured companions. Osbern’s killing at court is, above all others, the one incident which suggests that William’s life was in danger during these years, with the evidence of a charter mentioning someone else who was severely injured at the time, presumably showing that we are dealing with a murderous brawl, rather than the assassination within William’s domestic quarters GND, ii, 162–5. Lucien Musset, ‘Les sépultures des souverains normands: un aspect de l’idéologie du pouvoir’, in Musset et al. (eds.), Autour du pouvoir ducal normand, 19–44, at 24. The claim that his tombstone was discovered in 1955 is dismissed in Corpus des inscriptions de la France médiévale: 22. Calvados, Eure, Manche, Orne, Seine-Maritime, ed. Robert Favreau and Jean Michaud (Paris, 2002), 251. 61 Fundamental here is Thompson, ‘The Example of the Montgomerys’, 257–8. 62 Orderic’s choice of positive language (pro perfidia sua tunc exulabat) must be significant, GND, ii, 94–5. 63 eructuante in maius quorumdam rabie and hi qui fideliores se profitentur, et quos nunc maioribus dux cumulauit honoribus, GND, ii, 92–3. 59 60

from childhood into adolescence

63

which Orderic’s narrative might seem to suggest.64 The notion that William’s survival owed something to divine providence is also mentioned by William of Malmesbury and in the Brevis Relatio.65 Orderic further expands on the theme in the highly charged speech that he has William make as he lay dying at the church of Saint-Gervase outside Rouen in 1087.66 While possibly being dramatically exaggerated after the fashion of such speeches, it does identify another factor that contributed to the young William’s security, namely the protective role played by his maternal kin. For all that William’s life was at some point in danger during the period between 1040 and 1042, the continued presence and the support of Herleva and her family were a significant factor in his survival. In terms of his psychological development, this must have been important. It is also notable that the revenge killing by one of Osbern’s followers seems to have brought the feud to an end. Once more the role of culturally structured violence placed a limit on disorder. Moreover, Osbern’s son William (fitz Osbern) and Roger I de Montgommery’s son Roger II were to become exceptionally prominent among Duke William’s supporters before the end of the 1040s. Their collaboration consigned the feud to oblivion. The promotions that followed these deaths conform to convention and demonstrate that the redistribution of lands was controlled from the centre and that they were passed on to the closest ducal kindred, that is, to ‘les Richardides’. Count Gilbert’s vast lands appear to have been split up, with a substantial portion including the castles of Brionne and Vernon, the latter on the frontier close to the lands of the King of France, being granted to a kinsman and near-contemporary of William. This was his childhood companion and cousin Guy, younger son of Richard II’s daughter Adeliza, the wife of Rainald, a count of Burgundy (1026–57). Central to Guy’s installation in Normandy was Adeliza’s purchase from her brother of the castle of Le Homme (now L’Isle-Marie, Manche, cant. Sainte-Mère-Eglise, comm. Picauville) in the northern Cotentin, mentioned in a narrative passage in a charter dated 1075.67 Her return to Burgundy, where she appears in a charter of 1037, shows that Guy was effectively being passed over into the care of his Norman relatives.68 His dependence on the favour

RADN, no. 96. GR, i, 426–7; Brevis Relatio, 27. 66 GND, ii, 94–5; OV, iv, 82–3. 67 AAC, no. 21 (Regesta, no. 58). 68 For Adeliza in Burgundy in 1037, The Cartulary of Flavigny 717–1113, ed. Constance Brittain Bouchard (Cambridge, MA, 1991), no. 15; idem, Sword, Miter, and Cloister: Nobility and the Church in Burgundy, 980–1198 (Ithaca, NY, 1987), 272. Adeliza appears in the charter with 64 65

64

william the conqueror

of his cousin and his guardians was therefore all the greater. With the operation seemingly requiring that the castle pass out of the custody of Nigel the vicomte, another of the central political figures of Richard II and Duke Robert’s time, it is reasonable to think, on the basis of the evident later good personal relations between Guy and Nigel and the latter’s son, another Nigel, that Guy was placed under Nigel’s guardianship.69 For all this, Guy can reasonably be considered to be more of an outsider than any of the others who were emerging into prominence at this time. Count Gilbert’s children, at this point too young to interfere in this dismemberment of their father’s lands, took refuge in Flanders. But this did not mean that they might not later attempt a comeback or be recalled to Normandy. The extreme north-eastern frontier comté of Eu, along with other lands in eastern Normandy, was granted to William ‘Busac’, the son of the William who had held the comté in his brother Richard II’s time, yet another ‘Richardide’.70 Another promotion to comital rank in this period was that of William, the son of Richard II and his second wife Papia, and the brother of Archbishop Malger of Rouen, for whom a comté was specially created in north-east Normandy. This William took the title of count of Arques. Certainly some years older than our William, he was to play a major role in the events of the next ten years. In terms of an elite focused on the reward of the duke’s closest paternal kindred, the appointment made perfect sense. The lands of Osbern the steward passed to his son, William fitz Osbern, who appears to have been under his mother’s guardianship until the mid-1040s.71 In this case, we are dealing with a descendant of Gunnor’s family. Encroachments into the lands over which the Norman dukes had come to expect to exercise control took place between 1040 and 1042. They included King Henry I himself seizing and demolishing the castle of Tillières-sur-Avre, built in 1013–14 by Richard II in a dominant position on the French side of the river Avre in lands over which the French kings, the her mother’s name (Judith), but since the sons named there appear elsewhere as her and Rainald’s children, she must be the same person. 69 For Le Homme, Eric van Torhoudt, ‘Les sièges du pouvoir des Néel, vicomtes dans le Cotentin’, in Anne-Marie Flambard Héricher (ed.), Les lieux du pouvoir au Moyen Age en Normandie et sur ses marges, Publications du CRAHM (Caen, 2006), 7–35, at 11–15, 24–6; Hagger, ‘How the West was Won’, 51–2. 70 For the grant to William Busac, Bauduin, La première Normandie, 297. 71 RADN, nos. 118, 119 (both of which must be in the first half of the dating period suggested by Marie Fauroux); Cartulaire de la Sainte-Trinité-du-Mont de Rouen, no. LIV; Marie-Josephe Le Cacheux, ‘Histoire de l’abbaye de Saint-Amand de Rouen des origines à la fin du XVIe siècle’, BSAN, xliv (1937, for 1936), 5–289, at 246–7, pièces justificatives, no. II/2.

from childhood into adolescence

65

Norman dukes, and the counts of Blois-Chartres claimed power.72 Jumièges pinpoints a confused response among those responsible for William’s protection. While they preferred to appease Henry, the castellan at Tillières, Gilbert Crispin, who had been installed there by William’s father, decided to fight the king, as a result of which the castle was burnt. Henry afterwards rebuilt it in breach of an agreement made with the young William. It is certainly a commentary on William’s psychology that this created a grievance which seems to have festered in him for years to come. The presence at William’s court of Waleran I, count of Meulan, whose principal lands lay on the river Seine between the duchy and Paris, and who was one of the main participants in a revolt against the French king in 1041, further shows that Normandy was not simply the passive victim of turbulence. Having appeared in Norman charters during the time of William’s father, it seems that Waleran found a refuge within the duchy even when its politics were at their most disturbed; his presence there may indeed have been a factor in provoking King Henry’s intervention at Tillières-sur-Avre.73 Troubling as several of the incidents that took place during William’s childhood were, the thread running through all the narratives is that the court remained the centre of politics and that the fundamental issues were factional conflict around the young duke and matters connected to the conduct of those closest to him. All show that norms imposed limits on the violence. EDWARD, KING OF THE ENGLISH A further continuity from the time of Duke Robert’s rule was the active support given by William’s guardians to the two English exiles, the brothers Edward and Alfred. After Cnut’s death in November 1035 had apparently created opportunities for them, and, in the midst of the fraught succession process that had by 1037 resulted in Cnut’s son by Ælfgifu of Northampton, Harold Harefoot, establishing himself as king, Edward crossed to Southampton in 1036 with warriors recruited in Normandy and visited his mother Emma at Winchester; this in all probability their first meeting since Emma had left her children behind in Normandy approximately twenty years earlier. According to Jumièges, the armed force that accompanied Edward was transported in forty ships, indicating that it consisted of several hundred men. However, that he was obliged to fight a battle after he had landed shows that his visit was treated in England as a 72 GND, ii, 100–1. For Tillières-sur-Avre, Bauduin, La première Normandie, 185–91; Astrid Lemoine-Descourtieux, La frontière normande de l’Avre de la fondation de la Normandie à sa réunion au domaine royal (911–1204) (Rouen and Le Havre, 2011), 79–81, 235–40. 73 RADN, nos. 104, 105, 107; Bates, Normandy before 1066, 74; Bauduin, La première Normandie, 261–2.

66

william the conqueror

hostile invasion, with no discernible effort being made either to welcome him as a prospective king or even as the kinsman of previous kings.74 Alfred crossed to England from Wissant in Picardy, apparently aided by his new brother-in-law, Count Eustace of Boulogne, now the husband of Edward and Alfred’s recently widowed sister Godgifu, whose first husband, Count Drogo, had died on the same pilgrimage as Duke Robert. Alfred’s brutal, albeit perhaps accidental, death on orders given by Harold Harefoot, after he had been handed over to him by Earl Godwine of Wessex, became an element in the narratives created in the aftermath of 1066, with everyone apportioning blame as suited their purpose.75 Within the means available to them, William’s guardians look to have promoted Edward’s cause more energetically than had been done before 1035. He appears as ‘king’ among the signa of a charter for Mont SaintMichel about whose authenticity there can be no doubt, the original’s destruction in 1944 notwithstanding.76 But since the military support he had received from Normandy, although substantial, had in no way matched this rhetoric, Edward had to shape his actions accordingly. Having presumably decided that nothing could be accomplished in England, he returned to Normandy, his foray having at most put down a marker. When his mother took refuge in Flanders in 1037, he again visited her, but refused to help her, apart from, so we are told by the Encomium Emmae Reginae, acceding precedence to Harthacnut in the matter of the English succession.77 Whatever the emotions and politics involved, the fact that Harthacnut had succeeded his father as King of Denmark must have been the strategically decisive influence. It nonetheless speaks volumes that Emma chose to base herself in Flanders, where she was better positioned to advance Harthacnut’s cause, rather than to join Edward in Normandy. After Harold Harefoot’s death on 17 March 1040, Harthacnut’s seemingly unopposed succession meant that it would surely still have required exceptional optimism to think that Edward truly had prospects; Harthacnut was a young man in his twenties, in all likelihood with a long reign ahead of him. As far as we know, there was no inkling that he had an illness from which he would die just over two years later. For William’s life it is nonetheless crucial that Edward’s appearance with the title of king within Normandy was the result of a conscious decision by his guardians to promote him as GND, ii, 104–7. The figure of forty is repeated by Poitiers, GG, 2–3. For the main modern accounts, Frank Barlow Edward the Confessor (London, 1970); 2nd ed. (New Haven, CT, and London, 1997), 44–6; Stafford, Queen Emma and Queen Edith, 237–46; Keynes, ‘The Æthelings in Normandy’, 185–95. 76 RADN, no. 111; Keynes, ‘The Æthelings in Normandy’, 196–8, with a reproduction of the facsimile at plate 4. 77 Encomium Emmae Reginae, 48–9. 74 75

from childhood into adolescence

67

such; and that Edward himself had decided he had kingly aspirations which would have been known on both sides of the Channel. A community of interest and aspiration had been forged between Edward and the Norman ducal court that, with the benefit of hindsight, we know was very important. It took on a new form when, in the last weeks of 1041, and possibly already ailing, Harthacnut, along with leading members of the English aristocracy, called Edward to England in the expectation that he would be the next king, something that subsequently happened. Edward’s peaceful succession after Harthacnut’s death at a feast at Lambeth on 8 June 1042 must have been down to a political consensus among the leaders of the English nobility, something that boded well for the stability of his reign.78 Among the party that welcomed Edward to England on the Hurst peninsula near Milford-on-Sea (Hampshire) was Godwine, earl of Wessex. Unquestionably the most powerful man in England, his support for Edward was symbolically demonstrated by the gift of the magnificent ship described in a poem in the Vita Ædwardi Regis.79 On 23 January 1045, Edward married his daughter Edith, the expectation presumably being that Godwine would be the grandfather of the next king. As much as was possible, steps had been taken that would shape the future in the interests of the main players in England. Nonetheless, while Edward was a splendidly authentic representative of the Wessex line of kings that Cnut had supplanted, the king-worthiness that he exemplified as a descendant of previous kings was no more than a basic starting qualification in the eligibility stakes. In installing him, no account that we know of was taken of the claims of his eponymous nephew in exile in Hungary, the son of King Edmund Ironside, Edward’s older halfbrother. In England, the succession had yet again been manipulated to produce a desired result. The malleability of the norms of succession – surely no one could possibly think in terms of rules – was central to the cultural worlds of William and his slightly older contemporary, Harold, second son of Earl Godwine of Wessex. A passage in the Preface to the early twelfth-century Latin translation of the Anglo-Saxon laws known as Quadripartitus has Edward, who must have crossed from Normandy to the Isle of Wight, being met at the headland at the end of the Hurst peninsula and there swearing a most solemn oath to 78 On Edward’s return, Barlow, Edward the Confessor, 48–53; J. R. Maddicott, ‘Edward the Confessor’s Return to England in 1041’, EHR, cxix (2004), 650–66. That Harthacnut’s health was poor is accepted by M. K. Lawson, ‘Harthacnut [Hardecanute] (c.1018–1042), king of England and of Denmark’, ODNB. 79 Vita Ædwardi, 20–1. For the recently discovered full text, Henry Summerson, ‘Tudor Antiquaries and the Vita Ædwardi Regis’, ASE, xxxviii (2009), 157–84, at 170–2. For the ship, Simon Keynes and Rosalind Love, ‘Earl Godwine’s Ship’, the Vita Ædwardi Rejis’, ASE, xxxviii (2009), 185–223, at 186–90.

68

william the conqueror

all the thegns of England that he would respect existing law, that is the laws of Cnut.80 Such an insistence on royal respect for legal continuity had become central to the political culture of the English kingdom from the late tenth century onwards. In association with an emphasis on high religious standards, it was an aspect of wider western European changes and, above all, the means by which the political and religious elites of the English kingdom sought to negotiate recurrent crises and catastrophic military defeats.81 It was to be of central importance to the descendants of the same elite when they came to negotiate with another potential new king after the defeat of 14 October 1066. Given the contemporary troubles in Normandy, the part that the young William and his supporters could have played in these events must have been minimal. Jumièges mentions only that Harthacnut asked Edward back to England, whereas William of Poitiers has William send an embassy to England, adding that the English accepted Edward because they feared that William might intervene on his behalf.82 Whereas it is quite possible that William’s protectors did send emissaries and a bodyguard with Edward, it is impossible to think that they did anything more than this. For Jumièges, whose account echoes exactly the surviving English sources, it was presumably sufficient when he was writing in the mid- to late 1050s to say that the Normans had provided a secure haven for a future king.83 This – it must be presumed – represents the official line of thinking in William’s entourage in 1042 and for a good number of years afterwards. While, in terms of William’s life, Edward’s succession ultimately changed everything, the events of 1041–2 must not be seen as the beginning of an extended line that led directly to the conquest of 1066. There are connecting threads, but such threads could easily have been cut if circumstances had turned out differently; a healthy son born to Edward and Edith, for example, would have placed everything on a new footing. While Edward survived and prospered in England after 1042 through a notable capacity to adapt and to work with those among whom he found himself, an exile of almost thirty years must surely have had a profound psychological and cultural effect on him; he would, for example, have in all On this passage, Maddicott, ‘Edward the Confessor’s Return’, 650, 660–3. Pauline Stafford, ‘The Laws of Cnut and the History of Anglo-Saxon Royal Promises’, ASE, x (1982), 173–90, at 178–90; Wormald, The Making of English Law, 129–34. For a recent summary of the issues, The Political Writings of Archbishop Wulfstan of York, ed. and trans. Andrew Rabin (Manchester, 2015), 31–50. 82 GND, ii, 106–7; GG, 18–19. 83 This comment is a logical consequence of Elisabeth van Houts’s re-dating of the composition of the first version of the Gesta Normannorum Ducum to the later 1050s, GND, i, xxxii–xxxiii. 80 81

from childhood into adolescence

69

likelihood been bilingual, with French perhaps being his preferred language.84 For all the consensus surrounding his succession, it was also certainly not axiomatic that Edward was going to be welcomed or trusted by everyone in England. Edward’s feelings, as opposed to his calculated policy, are probably revealed in the humiliation inflicted on his mother in November 1043, when, on his orders, Emma was stripped of most of her valuable possessions and thereafter kept in comfortable albeit close confinement. The recently discovered conclusion to the Encomium Emmae Reginae might end with an elegant image of mother and son working in harmony, but this, very evidently, was not how Edward saw the situation; all versions of the Chronicle say that Emma was despoiled because her son thought she had failed to help him sufficiently.85 He would also have wanted to reward some of those who had helped him in exile. Among those who migrated to England in his wake were his young nephew Ralph, son of his sister Godgifu’s marriage to Count Drogo, and Robert, abbot of Jumièges, who became bishop of London in 1044. While such Normans and French were in no way a dominant group at a remarkably cosmopolitan court, their presence constituted a link to Edward’s Norman past that events show was never broken.86 There was here a mutuality of sympathy and indebtedness that might be reanimated in the future as circumstances dictated. All, however, depended on Edward’s success as king and on how William shaped up as duke of the Normans. THE RULE OF THE ADOLESCENT Jumièges presents William’s first military campaign, the defeat in c.1043 of Thurstan Goz, as if his emergence as a more independent presence made an immediate difference, by saying that the duke’s army fought with exceptional energy; Orderic amended this only by inserting that William was assisted by his military tutor Ralph de Gacé.87 Both explained the campaign as a reaction to Thurstan’s withdrawal from allegiance to William after introducing troops supplied by King Henry I into Falaise, something he had done because of the feeble support he was receiving from William’s guardians. Thurstan’s actions were therefore presumably a response to a renewal of the turbulence around Sées and Alençon of the kind that had preoccupied William’s father, with such stability as had followed from Barlow, Edward the Confessor, 73–95. Keynes and Love, ‘Earl Godwine’s Ship’, 195–6; ASC, ‘C’, ‘D’, ‘E’, 1043. 86 Barlow, Edward the Confessor, 40–1, 86–7, 93–4; C. P. Lewis, ‘The French in England before the Norman Conquest’, ANS, xvii (1995), 123–44, at 125–9. 87 GND, ii, 100–3. For the situation in general, Louise, La seigneurie de Bellême, i, 341–3. 84 85

70

william the conqueror

Duke Robert’s campaigns having evaporated, partly because of opportunism on the part of the Bellême family and partly because of spin-offs from warfare within that family. In addition to installing Ivo, a younger son of William de Bellême (died c.1027), as bishop of Sées at a date probably after 1035, the Bellême family had also taken control of the two strategic strongpoints of Alençon and Domfront.88 Thurstan’s rivalry for control of Falaise and the Hiémois with the Montgommerys may also have been a factor in persuading him to seek help from outside Normandy. The assistance that King Henry gave to Thurstan was probably also a follow-up to his earlier advance into Normandy to devastate the countryside near Argentan, which is to the north of Alençon and Domfront, but south of Falaise. Viewed as an attempt to stabilize the region, Henry’s interventions may actually have been intended, albeit indirectly, to shore up the young William’s regime by preventing others from profiting from its weakness. Henry’s long-standing collaboration with Count Fulk Nerra of Anjou having come to an end with the latter’s death in 1040, and with the dramatic increase of Angevin power on the river Loire following on from the seizure of the city of Tours by Fulk’s son Count Geoffrey Martel (1040–60) after the defeat of the counts of Blois-Chartres at Nouy (21 August 1044), it was imperative for the king to recruit allies for the likely struggles to come.89 He might disapprove of some recent developments in Normandy, but the re-emergence of effective ducal rule was at this point in his interests. Henry was also morally committed to the young William as he had legitimated his succession in 1035. If Henry was indeed acting to forestall any further advances by the Bellêmes into Normandy, William’s campaign becomes both an attempt at personal assertiveness and a statement that he, not the king of France, was responsible for peace-keeping in this contested region. The way in which Thurstan’s conduct was ultimately deemed by William not to have been too heinous was demonstrated by Thurstan’s family’s relatively rapid return to favour. Orderic commented that his son Richard, who appears in documents in the early 1050s, reconciled William and Thurstan and received more lands from William than his father ever had.90 According to Poitiers, the news that William had been knighted spread fear throughout Francia. Although still an adolescent, he exhibited adult qualities. It was a wondrous and frightening sight to see him armed for war. Like Jumièges, Poitiers also emphasizes William’s early piety and his See ibid., i, 290–5, 343–50; Thompson, ‘The Example of the Montgomerys’, 260. Jan Dhondt, ‘Henri Ier, l’Empire et l’Anjou (1034–1056)’, Revue belge de philologie et d’histoire, xxv (1946–7), 87–109, at 90–101; Guillot, Le comte d’Anjou, i, 67–79. 90 GND, ii, 102–3. 88 89

from childhood into adolescence

71

concern for peace and law. He began to attack external enemies energetically and to remove incompetent counsellors from his entourage.91 Although these eulogies are typical of their authors’ standard rhetoric, it is worth observing that the physiological and psychological complexities of the individual life cycle must have been as difficult for eleventh-century writers as they are for us today. Poitiers may well have been doing no more than expressing the obvious, namely, that someone who was going to be formidable as an adult matured early, it being a facet of his style to select language and classical parallels to make a point. The signa of ducal charters during the 1040s up until 1047 were predominantly individuals with a role at court, rather than locals with a direct interest in the transactions recorded in the charters; the authority of the young duke and his entourage must therefore have been deemed capable of guaranteeing that an agreement made there would hold.92 The names that regularly recur are Archbishop Malger of Rouen, Nigel the vicomte, Ralph Taisson, and Godfrey vicomte of Arques, the son of Goscelin vicomte of Rouen (and possibly Arques), the founder of the two Rouen monasteries of La Trinité-du-Mont and Saint-Amand in c.1030.93 Allowing for changes of generation, this group was also emphatically a continuation of the court circle of William’s father’s time, and indeed of his grandfather’s. In addition to Archbishop Malger, who was a ducal kinsman atypically promoted, Ralph Taisson belonged to a family that had migrated from Anjou to be installed in lands to the south of Caen in the later years of Duke Richard II’s reign.94 Nigel the vicomte had lands in the far west of Normandy and Ralph Taisson’s lands were also well to the west of the duchy’s Upper Norman heartlands. Nigel in particular – or in truth Nigels in the plural, since more than one generation was involved – seems to have been a roving agent of ducal authority with a power base in the west of the

GG, 6–9. On this theme, David Bates, ‘The Prosopographical Study of Anglo-Norman Royal Charters’, in K. S. B. Keats-Rohan (ed.), Medieval Prosopography: The Roots and Branches of Power in France and England from the Tenth to the Twelfth Centuries (Woodbridge, 1997), 89–102. 93 Malger: RADN, nos. 88, 95, 98, 100, 102, 103, 106, 107, 110, 112; Nigel: ibid., nos. 93, 95, 98, 100, 103, 104, 110, 111; Ralph Taisson: ibid., nos. 93, 98, 100, 103, 105, 110; Godfrey: ibid., nos. 93, 98, 100, 102, 103, 110, 111. It is sometimes unclear which of the two Nigels, father or son, appears. The presence of Niellus iuvenis in a list of testes dating from between 1035 and 1041 suggests that the son may have succeeded the father in the later 1030s (ibid., no. 111). For brief comments on Godfrey, Jean-Michel Bouvris, ‘Contribution à une étude de l’institution vicomtale en Normandie au XIe siècle. L’exemple de la partie orientale du Duché: Les vicomtes de Rouen et de Fécamp’, in Musset et al. (ed.), Autour du pouvoir ducal, 149–74, at 155. 94 For the Taissons, Lucien Musset, ‘Actes inédits du XIe siècle: V. Autour des origines de Saint-Etienne de Fontenay’, BSAN, lvi (1963, for 1961–2), 11–41, at 20–2. 91 92

72

william the conqueror

duchy.95 This pattern of signa is persuasive evidence that the vast majority of the Norman aristocracy continued either to support William or at least not to want to remove him; Orderic actually comes close to saying this when he observes that those who wished to maintain justice and religion served William loyally.96 Furthermore, the group at the heart of the ducal court had geographical interests that covered the whole of Normandy. The appearances in these same charters of William fitz Osbern, Roger II de Montgommery, and Roger de Beaumont do arguably indicate that William was asserting his personal preferences. There was, however, nothing revolutionary about the emergence of these individuals, since all were members of the duchy’s wealthiest aristocratic families and were related to the ducal house through kinship to the Duchess Gunnor. William fitz Osbern was certainly an independent presence sometime before the Battle of Val-ès-Dunes.97 Roger de Montgommery looks to have taken over from his father after the latter’s exile, and Roger de Beaumont, although active earlier, first appears in charters in 1046–7, in all probability after his father Humphrey’s death, dateable to between 1045 and 1047.98 For all that feud and consequent violent death were structured into their mental world, their coalescence around the young duke has to be a remarkable testimony to the healing powers of the regime that was starting to emerge, an eloquent demonstration of how, even in late adolescence at the age of either sixteen, seventeen, or eighteen, William could inspire loyalty and trust; Roger de Montgommery’s brother had after all been responsible for the death of William fitz Osbern’s father. Yet at the same time as they become prominent, others appear to have started to become marginalized; for example, the younger Nigel appears not to have had the status that his father had enjoyed.99 These changes presumably lie behind Jumièges and Poitiers’s statements about William choosing his own counsellors. The changes would also have produced tensions. 95 See now Eric van Torhoudt, Centralité et marginalité en Neustrie et dans le duché de Normandie: maîtrise du territoire et pouvoirs locaux dans l’Avranchin, le Bessin et le Cotentin (VIe–XIe siècles), Thèse de Doctorat, Université Paris Diderot (Paris VII) (2008), 538–48, 730–9; Hagger, ‘How the West was Won’, 32–4. 96 GND, ii, 98–9 (Quidam optimatum qui Deum et iustitiam amant libenter duci et domino obtemperant, fidemque seruant). 97 William first appears independently in RADN, nos. 106 and 113, with the limits of the latter being 1043x1047 because Guy of Brionne appears among the signa. 98 For Roger II, Thompson, ‘The Example of the Montgomerys’, 260. For Roger de Beaumont’s first appearance, RADN, no. 106. Humphrey’s death, Cartulaire de Saint-Pierre-dePréaux, xviii–xix, RADN, no. 96, in which William and Roger de Beaumont both appear, does not have to date from the time of Osbern the steward’s death. 99 He is notably absent from documents dating from 1046, RADN, nos. 106, 107, 108. In general, van Torhoudt, Centralité et marginalité, 733–7; Hagger, ‘How the West was Won’, 37–8.

from childhood into adolescence

73

The war between the Beaumont and Tosny families, for which Orderic is the only source, at first sight so resembles the wars of the early 1040s that Orderic treated it as one of them.100 However, while traditionally dated to c.1040, it must have occurred later, perhaps in 1043 or 1044, since a charter shows Roger de Tosny as alive after William had reached adolescence; the statement by Orderic that William was a puer at the time, unless interpreted as a reference to him as a young unmarried man, is arguably misleading.101 In his narrative Orderic also introduces for the first time his theme that William’s illegitimate birth was a political weakness; Roger, he says, found it intolerable to serve a bastard. As Roger had apparently made no objection to doing precisely this since 1035, Orderic’s statement seems rather unlikely as a causative factor; a more likely reason, as Orderic actually indicates, was territorial disputes with Humphrey de ‘Vieilles’ and his family, associated with the proximity of the lands belonging to the Beaumont and Tosny families in central Normandy and to their respective ambitions. In addition, even if we do accept Orderic’s explanation (as I do not think we should), a logical corollary of his narrative must be that Humphrey and his family had no problems with William’s birth at all. As with the earlier wars, we are dealing with a conflict between two of Normandy’s most powerful families. Roger de Tosny may have participated briefly in the wars between Christians and Moslems in Spain, but he was mostly based in Normandy throughout William’s childhood, making it probable that the Roger de Tosny who performed so flamboyantly in Spain was another man, perhaps a nephew.102 The ‘Beaumont’ Humphrey had extended his family’s lands southwards in Duke Robert’s time from a base around Pont-Audemer, taking over Beaumont and Vieilles from the abbey of Bernay founded by Duke Richard II, and acquiring many properties elsewhere in Normandy.103 The two were also pioneers, being almost the first members of the Norman aristocracy to found Benedictine abbeys, in the case of the Tosny at Conches-en-Ouche and of Humphrey at Les Préaux. The Tosny family and their ally Robert de Grandmesnil were badly beaten in the war by Humphrey’s son Roger de Beaumont: Roger de Tosny, two of his sons, and Robert de Grandmesnil were killed. After their OV, iii, 88–9; iv, 302–3. RADN, no. 100. See also ibid., no. 102; Musset, ‘Les Tosny’, 55. 102 For the argument that there were two Rogers, K. S. B. Keats-Rohan, ‘Domesday Book and the Malets: Patrimony and the Private Histories of Public Lives’, Nottingham Medieval Studies, xli (1997), 13–56, at 44, note 178. 103 GND, ii, 94–7; OV, ii, 40–1. For the rise of the Beaumonts, Véronique Gazeau, Monachisme et aristocratie au XIe siècle: l’exemple de la famille de Beaumont, Thèse de Doctorat de troisième cycle, Université de Caen (1986–7), 67–77, 172–4, 256. 100 101

74

william the conqueror

deaths, the response of the ducal court was proactive, with the losers being propped up by marriage alliances to neighbouring families who were powerful at court: Roger de Tosny’s widow was married to Ralph de Gacé’s half-brother Richard, count of Evreux, who also exercised a protectorate over her young son Ralph, while Robert de Grandmesnil’s widow was married to another of Ralph’s half-brothers. The objective, which seems to have been achieved, was to create the equilibrium necessary to pacify the feud. A revenge killing did eventually happen, but not until after 1054, when another of Humphrey’s sons, Robert, was killed by Roger de Clères, one of the major barons of the Tosny honour.104 That the killing occurred at all emphasizes again the presence of the culture of feud and the potential of rivalries for land and office to erupt into revolt and war. The absence of any discernible detrimental impact on Roger de Clères’s fortunes in the 1050s, after William had become much more dominant, surely shows that his level of violence was regarded as legitimate. Hugh of Flavigny, a monk writing in Burgundy in the closing years of the eleventh century, noted that William had a vivid memory of meeting the ecclesiastical reformer Richard, abbot of Sainte-Vanne of Verdun, who died in 1046. This indicates a further continuity from his father’s time, since Duke Robert had received a visit from Abbot Richard in 1032; selfevidently William was keeping in touch with the same religious contacts from whom Robert had obtained spiritual guidance.105 And although diplomatically not a straightforward document, the charter recording William’s grants from 1042 onwards to his father’s monastic foundation at Cerisy-la-Forêt is another clear sign that he was positively continuing a central aspect of Robert’s rule and that he honoured his father’s memory. The charter is also added testimony to a continuing dynamic ducal presence in the western lands of the duchy.106 In all this, there appears to be evidence of an early interest on William’s part in religious matters.

104 For the dating, Gazeau, Monachisme et aristocratie, 138–42; Cartulaire de Saint-Pierre-dePréaux, nos. A98, A99, A165. 105 Hugh of Flavigny, Chronicon, 407. Abbot Richard’s visit to Normandy does not necessarily have to date to 1042, as has sometimes been assumed, Martin Brett, ‘Warfare and its Restraints in England, 1066–1154’, in ‘Militia Christi’ e Crociata nei secoli XI–XIII. Atti della undecima Settimana internazionale di studio Mendola, 28 agosto–1 settembre 1989, Miscellanea del Centro di studi medioevali, vol. xiii (Milan, 1992), 129–44, at p. 131, note 7. 106 RADN, no. 99; van Torhoudt, Centralité et marginalité, 734; Hagger, ‘How the West was Won’, 37–8. Matilda’s appearance among the signa is incompatible with the dating clause of 20 April 1042, but the diplomatic norms of such an extensive confirmation charter indicate that it would have started in 1042 and amended over time, AAC, 25–35; David Bates, ‘Les chartes de confirmation et les pancartes normandes du règne de Guillaume le Conquérant’, in Michel Parisse, Pierre Pégeot, and Benoît-Michel Tock (eds.), Pancartes monastiques des XIe et XIIe siècles (Turnhout, 1998), 95–109.

from childhood into adolescence

75

Episcopal action was a further strength of the regime of William’s adolescence, another continuation from the pre-1035 past. A synodal letter in the name of Archbishop Malger, customarily dated to c.1045, but possibly earlier, is a full and carefully worked condemnation of the evils of the times, in particular of simony.107 Malger reports to his suffragans that he, together with Bishops Hugh of Evreux and Robert of Coutances, has called together a council. The presence of only three bishops, all of whom were based in Upper Normandy since Bishop Robert resided in Rouen at this point, may well be evidence that the Norman Church lacked significantly in cohesion during this period, in other words, a sign of major disruption. Yet the letter is an impressive document, since in the context of conciliar legislation against simony from the French kingdom as a whole, it is the earliest known example outside the great Peace councils of the duchy of Aquitaine.108 Canonical authority is extensively and properly deployed. The preamble to the canons deplores the weakness of recent rulers, in contrast with their predecessors.109 A notable feature, confirming once more the continuing centrality of the court as a focus of business, is the condemnation of those who seek episcopal office through favour at court or through influence with the prince.110 Overall this council demonstrates that the Norman bishops were continuing to play a role during William’s adolescence that was traditional in the post-Carolingian West, criticizing secular authority when it was deficient, and seeking to reinforce it through ecclesiastical means. Since the target of the criticisms must have been the conduct of the adolescent William and those responsible for nurturing him, the bishops’ aim must have been to restore the ethos of ducal rule in earlier times. Archbishop Malger’s anathemas also appear in several diplomas of the period, a novelty in Norman charters and another archetypical feature of a world in which the secular 107 Concilia Rotomagensis Provinciae, ed. G. Bessin (Rouen, 1717), 40–2; reprinted in Sacrorum conciliorum nova et amplissima collectio, ed. J. D. Mansi (Venice, 1774), xix, cols. 751–4. The council must date to before the death of Bishop Hugh of Evreux on 20 April 1046. On the synodal letter and its authorities and the council, see now Richard Allen, ‘Avant Lanfranc. Un réexamen de la carrière de Mauger, archevêque de Rouen (1037–1054/55)’, in Julia Barrow, Fabrice Delivré, and Véronique Gazeau (eds.), Autour de Lanfranc (1010–2010): Réforme et réformateurs dans l’Europe du Nord-Ouest (XIe–XIIe siècles (Caen, 2015), 131–52, at 135–7. 108 See Odette Pontal, Les conciles de la France capétienne jusqu’en 1215 (Paris, 1995), 124, 129–30. At p. 159, Malger’s council is incorrectly dated to after the Council of Rheims of 1049. 109 Quoniam sanctissimi fratres sanctam matrem ecclesiam a primo sui honoris et tranquilitatis statu penitus immutatem videmus, partim sublatis e vita bonis principibus, partim ignaris et imbecillibus successoribus . . ., Sacrorum conciliorum, ed. Mansi, xix, col. 751. 110 Ut illa perniciosa consuetudo et inexplebilis avaritiae corruptelae funditus eradicetur, qua multos perniciosa munera undecumque collegisse audivimus, quibus principem regni et familiares eius corrumpere valeant, ut ad episcopatus honorem valeant pervenire, Sacrorum conciliorum, ed. Mansi, xix, col. 752.

76

william the conqueror

power needed Church support.111 The evidence as a whole once more indicates the continuation of earlier traditions and vigorous efforts by many among the Norman ruling elite to sustain them. The mid-1040s can actually be seen as a time of some religious prosperity in Normandy. The first consecration of the abbey church of Le Bec, founded by the hermit Herluin, a former member of Count Gilbert of Brionne’s entourage, can be assigned to this period; the abbey was attractive enough to recruit into its community another of the great figures of William’s life, the scholar Lanfranc of Pavia.112 The early twelfth-century history of the abbey of Saint-Pierre-sur-Dives identifies 1046 as the year when Ainard, a German cleric and the house’s first abbot, was invested with the ring and staff of office by William.113 And the annals of the abbey of Lyre date its foundation by William fitz Osbern to 1046.114 The appointments to bishoprics made during this time are a good illustration of how those who were dominant at court used their influence to advance clients and to consolidate their own power; the period was indeed one of generational renewal within the episcopacy as well. Thus, Bishop William, appointed to Evreux in 1046, was a member of the Fleitel family connected to both Count William of Arques and Ralph de Gacé, and which had supplied a bishop of Sées in Duke Richard II’s time.115 As already noted, the transfer of Sées to Ivo, a member of the Bellême family, indicates a return there to the family’s early eleventh-century dominance of the bishopric. It is likely, too, that the appointment of Bishop Hugh to Lisieux in 1046 was connected with the power in central and southern Normandy of his family, the counts of Eu. And, after the Battle of Val-èsDunes, Bishop Geoffrey obtained Coutances in 1048 through the local influence of his kinsman Nigel the vicomte and through simony.116 All four 111 See, for example, Allen, ‘Avant Lanfranc’, 149–51, including a complete edition of RADN, no. 112; Lot, Saint-Wandrille, recueil des chartes, no. 26. The anathemas are omitted in the published editions in RADN, nos. 109, 112. Malger’s prohibition is associated with a ducal confirmation in RADN, no. 134. 112 For the re-dating of Le Bec’s early history and Lanfranc’s arrival there, Jean-Hervé Foulon, ‘Le chevalier Herluin et la fondation de l’abbaye du Bec: un dossier complexe entre tentation érémitique et normalisation cénobitique’, Revue Historique, no. 663 (2012), 563–607, at 572–9. 113 GC, xi, instr. col. 155. 114 ‘Breve Chronicon Lyrensis’, in E. Martène and U. Durand (eds.), Thesaurus Novus Anecdotorum, 5 vols. (Paris, 1717), iii, col. 1432. For the Annals of Lyre, Alison Alexander, ‘Annalistic Writing in Normandy, c.1050–c.1225’, University of Cambridge PhD Thesis (2012), 159–65. 115 For the family as tenants of Count William, AD Seine-Maritime, 16 H14, fo. 319r(Bauduin, La première Normandie, 386–7 [Dossier de Textes, no. 16]). According to Robert of Torigni, Ralph de Gacé married a daughter of Gerard Fleitel, GND, ii, 268–9. 116 For the lands of the counts of Eu in central Normandy, Louise, La seigneurie de Bellême, i, 144. For a re-dating of Hugh’s appointment, Foulon, ‘Le chevalier Herluin’, 574, note 41.

from childhood into adolescence

77

cases also serve to illuminate the difficulties of judging these times. They certainly infringed the pronouncements in Archbishop Malger’s synodal letter, and such conditions were subsequently thought intolerable by the adult William. However, just because these bishops were not chosen by the ruler alone does not mean that they were bad appointments; hindsight actually shows that none of them was a poor choice. One, Geoffrey of Coutances, was to become one of the greatest of the makers of the post1066 cross-Channel empire. Those around William during his adolescence would again appear to have been maintaining the standards and practices of earlier times in appointing bishops of respectable character from major aristocratic families. The obvious relationship between influence and appointments may just be the normal business of securing patronage, with the veneer of powerful ducal authority stripped away. Bishops would also have been centrally involved in the proclamations of peace in Normandy. Although these have often been seen in terms of a single proclamation of the Truce of God after the Battle of Val-ès-Dunes, it is certain that peace councils were held before that date and that the fully fledged Truce was a later creation.117 There are three independent references to a council held at Caen at which peace was proclaimed, of which one, from the cartulary of the abbey of Saint-Pierre of Préaux, must date to before 6 March 1047, and therefore to well before the Battle of Val-èsDunes.118 The events as described in the three accounts have all the characteristics of similar councils elsewhere, with, in particular, saints’ relics being brought from various parts of the duchy, oaths being sworn to keep the peace, and the reported occurrence of miracles.119 That the relics which we know were brought to Caen all came from Rouen suggests that Archbishop Malger played a leading role. The story recorded in one of the accounts of the council – that William insisted on delaying the dedication of the church of Rots, near Caen, which was a possession of the abbey of Saint-Ouen of Rouen, until he could arrive to carry a reliquary on his shoulders – may date from this time. If so, it is an early example of the sort of theatrically demonstrative piety that was to be On Geoffrey’s appointment, Marjorie Chibnall, ‘La carrière de Geoffroi de Montbray’, in Pierre Bouet and François Neveux (eds.), Les évêques normands du XIe siècle (Caen, 1995), 279–93, at 280. For its date, Spear, Personnel, 90–1. 117 The classic study is de Boüard, ‘Sur les origines de la Trêve de Dieu’, 169–89. For effective criticisms, Hartmut Hoffmann, Gottesfriede und Treuga Dei, Schriften der MGH, xx (Stuttgart, 1964), 166–7; Brett, ‘Warfare and its Restraints’, 131, note 7; Barthélemy, L’an mil et la paix de Dieu, 524–30. 118 Cartulaire de Saint-Pierre-de-Préaux, no. A1[14]; NMProsopographie, 335. 119 ‘Sanctae Catharinae virginis et martyris Translatio’, 438; ‘Miracula sancti Audoeni’, Acta Sanctorum, August, iv, 836. For the atypicality of the Norman proceedings, Barthélemy, L’an mil et la paix de Dieu, 529–30.

78

william the conqueror

characteristic of the adult William.120 The choice of Caen for the council must also indicate that the town which was going to be so significant in William’s life was already an important place.121 Within the much discussed changes associated with the eleventh-century Peace and Truce of God, the developments in Normandy fit a northern French pattern: they were from the first supported by those who sought to reinforce the authority of kings and territorial princes against excessive disruption by the violence culturally structured into aristocratic society rather than constituting any kind of movement from below.122 The classic case study for the consolidation of local secular power is Count William of Arques. The title ‘William count by the grace of God, son of Richard II, duke of the Normans’, which appears in a charter of his for the abbey of Fécamp, dated 18 July 1047, and which also appears in similar form in documents for other beneficiaries, exudes pretentiousness and a sense of his own importance. It also identifies him as a descendant of Duke Richard II, and therefore as one of ‘les Richardides’.123 His action in the case described in the Fécamp charter resembles a ducal one in style, as indeed do the form and presentation of the document recording it.124 He also appears on occasion in charters acting directly alongside William in a way that is without parallel.125 It is very easy – indeed all too easy – to see this evidence as confirming several aspects of William of Poitiers’s subsequent damning assessment of Count William, and in particular that 120 ‘Miracula sancti Audoeni’, Acta Sanctorum, August, iv, 835. The argument that this event took place at the same time as the peace council is based on the supposition that the monks of Saint-Ouen would not have allowed their relics to travel such a long distance more than once. 121 On Caen’s earlier history, Laurence Jean-Marie, Caen aux XIe et XIIe siècles: espace urbain, pouvoirs et société (Cormeilles-le-Royal, 2000), 27–31; Jacques Le Maho, ‘Nouveaux aperçus sur les origines de la ville de Caen’, BSAN, lxviii (2010, for 2009), 171–99, at 173–86. 122 Jean-François Lemarignier, ‘Paix et réforme monastique en Flandre et en Normandie autour de l’année 1023’, in Robert Aubreton et al. (eds.), Droit privé et institutions régionales: études historiques offertes à Jean Yver (Paris, 1976), 443–68, at 444–6, 455–7 (idem, Recueil d’articles rassemblés par ses disciples [Rouen, 1995], 339–64, at 340–2, 351–3); Barthélemy, L’an mil et la paix de Dieu, 521–55. 123 David Bates and Pierre Bauduin, ‘Autour de l’année 1047: un acte de Guillaume, comte d’Arques, pour Fécamp (18 juillet 1047)’, in Pierre Bouet, Catherine Bougy, Bernard Garnier, and Christophe Maneuvrier (eds.), De part et d’autre de la Normandie médiévale: Recueil d’études en hommage à François Neveux (Caen, 2009), 43–52, at 50; RADN, nos. 100, 103; Lot, Saint-Wandrille, recueil des chartes, no. 15. 124 Bates and Bauduin, ‘Autour de l’année 1047’, 48. 125 . . . auctoritatem suam dederunt comes Nortmannorum Willelmus et patruus eius, idemque comes Archarum Willelmus, RADN, no. 108; per manu Vuillelmi Northmannorum marchionis, et Vuillelmi patrui eius, videlicet Archis tunc temporis comitis, cuius in territorio et comitatu idem alodum habetur, ibid., no. 124. Although this second charter dates from 1051, which is strictly speaking outside the period covered in this chapter, Count William’s continued prominence emphasizes further the point being made.

from childhood into adolescence

79

he attempted both to increase his lands and to exclude young William’s authority from Normandy east of the Seine.126 If we set aside assumptions that can be made on the basis of all of the above being affronts to ducal authority which the adult William would not have tolerated, we see in Count William’s charters a concern for public order, the practice of extensive monastic patronage, the restoration of monastic property claimed to have been lost during the Viking settlements that were the foundation of Normandy, and the righting of recent wrongs; in other words, a consistently positive approach to the wider aristocratic responsibilities of peace-keeping. Even Count William’s marriage to a sister of Enguerrand II, count of Ponthieu, which must have taken place by 1045 at the latest, involved the stabilization of the duchy’s eastern frontier, since the counts of Ponthieu had seized land around Aumale in Duke Robert’s time.127 Such cross-border marriages as a means to stabilize networks around Normandy might even be thought of as official policy, another notable example being the marriage of Duke William’s sister Adelaide to Count Enguerrand.128 A cross-border marriage of these times that fits the same pattern, and which was to have major consequences, was the one between Count William’s future close associate Roger de Beaumont and Adeline, daughter of Waleran I, count of Meulan.129 On the other hand, of course, it is also reasonable to suggest that Count William knew very well how to use the language of legitimacy for personal benefit. If the evidence ultimately conceals the motivations and the thinking of the participants, there is nonetheless a big difference between the selfaggrandizement based on a sense of the status of a ducal kinsman, which seems probable, and the outright disloyalty that Poitiers suggests.130 It is reasonable to think of Count William as being loyal to William, albeit rather on his own terms. Other land seizures unrecorded by Jumièges and Orderic certainly took place during this period, such as, for example, the one involving Bishop 126 GG, 34–5. Duke William confirms numerous charters for the region throughout the 1040s, RADN, nos. 101, 102, 103, 104, 105, 106, 107, 108, 112, 115, 118, 119. 127 For Count William’s marriage, GND, ii, 104–5. His charter for Fécamp of 1047 refers to his sons (pro anima . . . filiorum meorum), Bates and Bauduin, ‘Autour de l’année 1047’, 50. Conditions around Aumale are considered in detail in Bauduin, La première Normandie, 399–405. 128 See now Thompson, ‘Being the Ducal Sister’, 69–71; Bauduin, La première Normandie, 306–7; cf. Searle, Predatory Kinship, 324, note 20. 129 Bauduin, La première Normandie, 261–5. 130 For arguments similar to mine, Olivier Guillot, ‘L’appui apporté par Henri à Guillaume d’Arques dans le conflit de ce dernier avec le duc Guillaume le Bâtard (fin 1053–1054): une motivation juridique?’, in Bernard d’Alteroche, Florence DemoulinAuzuri, Olivier Descamps, and Franck Roumy (eds.), Mélanges en l’honneur d’Anne LefebvreTeillard (Paris, 2009), 473–93, at 482–4, 489–91.

80

william the conqueror

Hugh of Bayeux suffering at the hands of both Count William of Arques and Humphrey de ‘Vieilles’.131 There is also evidence of other erosions of ducal influence around the fringes of the lands that the dukes expected to control. As a result, none of the abbots appointed under the auspices of the then predominant influence in Normandy of the Upper Norman abbey of Fécamp was given an easy ride by the independent-minded monks of Mont Saint-Michel in the far west of the duchy.132 It was also the case that whereas some individuals prospered, others did not; clear evidence, in other words, of factional conflict between those around the young duke. Ralph de Gacé, for instance, did not apparently long enjoy the profits that may have arisen from the killing of Count Gilbert. Not only does he not appear among the signa of ducal charters, the reference to at least one of Gilbert’s killers being punished was a sign that a coup of the kind which Ralph had perpetrated remained outside acceptable limits.133 And Count William of Arques’s 1047 charter for Fécamp, taken alongside other evidence, shows his personal role in the reduction in status of Godfrey vicomte of Rouen, a man who never recovered his earlier eminence.134 By the mid-1040s, the regime that had evolved around the adolescent William seems to have become sufficiently coherent and powerful for its support to be canvassed for an intervention to the south of Normandy. An original charter that must date from either 1046 or the early months of 1047 indicates the presence at a gathering, which probably took place in Rouen, of Bishop Ivo of Sées and Count Hugh IV of Maine.135 Exactly what motivated Bishop Ivo’s appearance is unclear; the fact that his uncle, who was also named Ivo and who was at that time the lord of Bellême, as well as his nephew Arnulf, attest a Norman ducal charter on 30 October 1048 suggests that the family were looking to William for support against developments further south in Maine.136 Count Hugh’s presence has to be seen in the context of the Angevin domination over Maine established by Count Fulk Nerra, and at this point somewhat precariously maintained 131 RADN, no. 100; Cartulaire de Saint-Pierre-de-Préaux, nos. A1[7], A1[14]; Véronique Gazeau, ‘Le patrimoine de l’évêque Hugues de Bayeux (1011–1049)’, in Bouet and Neveux (eds.), Les évêques normands du XIe siècle, 139–47. 132 Bouet, ‘Le Mont-Saint-Michel entre Bretagne et Normandie’, 197–9. 133 OV, ii, 120–1. 134 Bates and Bauduin, ‘Autour de l’année 1047’, 50; Regesta, no. 230. For the narrative, Bates, Normandy before 1066, 100, 104. 135 RADN, no. 107. The earliest date for the charter must be 1046, since it is attested by Bishop William of Evreux, appointed in that year, and because of the death of another of the signa, Abbot Gradulf of Saint-Wandrille on 6 March 1047, NMProsopographie, 335. The presence among the signa of members of the Rouen chapter and the fact that its beneficiary is the abbey of Saint-Ouen suggest that it was confirmed at Rouen. The charter’s importance is emphasized by Louise, La seigneurie de Bellême, i, 355. 136 RADN, no. 115; Thompson, ‘Family and Influence’, 218–19.

from childhood into adolescence

81

by his son Geoffrey Martel. Since Hugh was about to defy Geoffrey – at the time probably absent from Anjou in Rome or on a journey there – he too was undoubtedly on the lookout for allies; Hugh’s marriage at about the same time to Bertha, the widow of Count Alan III of Brittany and the daughter of Count Odo II of Blois-Chartres, was also part of the antiAngevin network-building that made the young William’s fortunes so important for the policies of King Henry I.137 The other signa of this charter, who include the two brothers Archbishop Malger of Rouen and Count William of Arques, William’s young kinsman Count Guy of Brionne, and William’s own protégés Roger de Beaumont and Roger de Montgommery, are evidence of the gathering of the new emergent elite of Norman aristocratic society around its ruler; it is also testimony to the apparent cooperation between individuals who were soon to become winners and losers in the competition for favour. The factional tensions that had accumulated beneath the surface around the adolescent William were about to burst the bounds of normal politics into civil war. But the group’s very business and the subject that they are likely to have been discussing in the charter attest to the achievements and the resilience of the regime that had gathered around the adolescent William. THE YEAR OF VAL-ÈS-DUNES Count Guy of Brionne’s conspiracy to replace William must have been developing during 1046 and early 1047; the above-mentioned charter is Guy’s last recorded appearance in William’s presence. All sources from Jumièges onwards indicate that he was aiming to become duke, with Orderic, as usual, and in this case followed by Wace, adding that Guy’s ambitions derived from William’s perceived illegitimacy, as well as from Guy’s disloyalty.138 While unquestionably a crisis that could have ended William’s rule before it had scarcely begun, the names of Guy’s allies suggest, as Wace actually says, that we are dealing with a quarrel between the duke and a close kinsman and that its origins can be located at court. Guy’s own origins, a younger son passed over by his Burgundian family to his Norman kindred to nurture and promote, give the conflict a special character. It is hard to think that Guy would have attracted a great deal of sympathy or support. On the other hand, he is likely to have been a magnet for those who were discontented with the way William’s rule was taking shape. 137 On Count Hugh, Robert Latouche, Histoire du comté du Maine pendant le Xe et le XIe siècle (Paris, 1910), 27–8; Guillot, Le comte d’Anjou, i, 65–6. 138 GND, ii, 120–3; GG, 8–13; OV, iv, 82–5; GR, i, 426–9; Roman de Rou, pt. 3, lines 3,609–18. For the battle, see recently, Hagger, William: King and Conqueror, 8–11.

82

william the conqueror

Of Guy’s supporters, Jumièges named only one, Nigel, whom he describes as governor (praeses) of the Cotentin, and whose family, as we have seen, had been closely associated with Guy since his mother had sent him to Normandy from Burgundy. Poitiers added the names of Rannulf vicomte of the Bessin and Haimo ‘Dentatus’ (‘Toothy’). Wace further added Ralph Taisson, who he said was an accomplice to the conspiracy but not a wholehearted supporter of it, and Grimoult du Plessis-Grimoult.139 With the exception of Grimoult, who was a lesser figure, all the rebels had been prominent in the ducal charters of William’s adolescence. In various ways, however, all might have had grievances about recent developments around the young ruler. Both Jumièges and Poitiers convey the impression that Guy had a great deal of support, but do so in unspecific terms. Jumièges is as usual less dramatic than Poitiers, who goes as far as to say that the greater part of Normandy was disloyal, a probable exaggeration made for rhetorical effect to give the impression that William overcame greater odds than he actually did; to overstate the scale of opposition would also justify the increasingly authoritarian regime that evolved after the Battle of Val-ès-Dunes.140 Writers in the first half of the twelfth century, such as Baudri of Bourgueil in his poem written for William’s daughter Adela, countess of Blois-Chartres, and Hugh of Fleury, whose history was addressed to William’s granddaughter, the Empress Matilda, nonetheless wrote in terms of the Normans seeking to disinherit William, thereby conveying a sense of extensive opposition, while at the same time emphasizing his right to rule and his legitimacy.141 Since no one before Wace provides any significant detail about the battle, and since his account has something of the character of a celebration of the prowess of the ancestors of the western Norman families of his day, we do not actually know either who fought on William’s side at Val-èsDunes or who might have supported the rebels if they had been given the

139 For Grimoult, Roman de Rou, pt. 3, lines 3,557, 3,777. For Ralph Taisson, ibid., pt. 3, lines 3,873–89. Cf. K. S. B. Keats-Rohan, ‘William I and the Breton Contingent in the NonNorman Conquest, 1060–1087’, ANS, xiii (1991), 157–72, at 162 (‘the quasi-autonomous West Normans of the Cotentin’). 140 plurimos proceres uelut Absalon ab eius fidelitate cepit auertere et in sue perfidie uoragine complicare, GND, ii, 120–1; sequebatur impietatis uexillum pars Normanniae maior, GG, 10–11. 141 Baldricus Burgulensis, Carmina, ed. Karlheinz Hilbert (Heidelberg, 1979), 149 (Qui sibi Normannos inhibentes iura paterna) (‘Baudri of Bourgueil, “To Countess Adela”, trans. Monika Otter, The Journal of Medieval Latin, xi [2001], 60–141, at 66); Hugh of Fleury, Liber qui modernum regum Francorum continent actus, ed. Georg Waitz, MGH Scriptores, xi (1851), 376–95, at 388 (Qui a Normannis exheredatus . . . postmodum meruit heredari); Pierre Bauduin, ‘Hugues de Fleury et l’histoire normande’, in Crouch and Thompson (eds.), Normandy and its Neighbours, 157–74, at 170.

from childhood into adolescence

83

chance to do so.142 In terms of the rebels whose names we know, Guy and Nigel were both young men of approximately the same age as William, who was by this time around twenty years old; they were perhaps already disenchanted with their perceived prospects as William asserted himself. Haimo, Rannulf, and Ralph Taisson were all older and more experienced, but must also have had reasons to think that they were being marginalized. The combination does suggest a formidable opposition. And the way in which the last three, all primarily based around Bayeux and Caen, joined in also suggests that the rebellion was gathering strength as its forces marched eastwards into the heartlands of the duchy. The circumstances were certainly also of a kind that could have encouraged fence-sitting among others. William’s decision immediately to seek the help of King Henry I could indicate that the members of the new and inexperienced regime emerging around William lacked the self-confidence to believe that they could carry their followers into a war in which victory was certain. Equally it is likely that the appeal to the king was intended to give William a decisive advantage in manpower and to reinforce the legitimacy of his cause. The course of the campaign suggests that he went on the attack quickly to prevent his opponents from developing their plans. It appears as though Guy was mustering his forces in the north of the Cotentin in the first half of 1047, perhaps actually at his mother’s castle of Le Homme, which commanded the access route into the north of the peninsula. Wace’s story that William had to flee from Valognes, which is very near to Le Homme, after an assassination attempt by Grimoult du Plessis-Grimoult, suggests either that he was attempting a pre-emptive strike or unsuccessfully trying to calm feelings by his presence; although Wace says that William went to Valognes to hunt and he was unaware that rebellion was brewing, this seems implausible. A rapid retreat took him first to Ryes-en-Bessin, where he was sheltered by Hubert de Ryes, after which William left Normandy to go to Poissy, near Paris, where he sought and obtained King Henry’s help.143 Jumièges’s statement that King Henry marched his army into Normandy through the Hiémois to join up with William suggests that they had made a prearranged rendezvous and that they were deliberately aiming to confine the rebels in Lower Normandy. It also indicates that William had returned from Poissy to Normandy and gathered troops, a sign that he must have had significant support among the Norman aristocracy. Their intention 142 Fundamental here is Elisabeth van Houts, ‘Wace as Historian’, in Keats-Rohan (ed.), Medieval Prosopography, 103–32 (History and Family Traditions, chapter 10). See also Matthew Bennett, ‘Poetry as History? The “Roman de Rou” of Wace as a Source for the Norman Conquest’, ANS, v (1983), 19–39. 143 Roman de Rou, pt. 3, lines 3,685–800.

84

william the conqueror

was presumably to prevent the rebels from gathering additional support as they moved east and, if necessary, to force a battle. They may have decided either that the rebellion was actually quite a weak one and that it should be finished off quickly, or that it was too dangerous to be allowed to escalate. Calculations that the situation was akin to trial by battle and an initiation ritual may also have been in William’s mind; in the face of the first great crisis of his rule, it was obviously important for him to demonstrate a capacity for decisive action. Although the circumstances were different, he arguably took the same approach in 1066. The whole narrative from his presence at Valognes onwards suggests conduct based on confronting problems head-on and seeking rational and realistic solutions. The accounts of the battle indicate that Henry and William were on the defensive to begin with, but that a strong counter-attack scattered their enemies and that panic turned rapidly to a rout. Wace’s account suggests that Guy, Nigel, and their allies had to attack because they were outnumbered. The speed and disorganization of their retreat were such that many were drowned in the river Orne, which, depending on where they tried to cross, is between ten and fifteen kilometres west of Val-ès-Dunes. Wace comments on the savagery of the pursuit and the numbers that were killed. No source mentions the taking of prisoners, therefore we can probably assume that did not happen. In other words, the achievement of supremacy on the battlefield was followed by a massacre. This went against contemporary conventions, since the great majority of usable accounts of battles for the period from 900 onwards do mention the taking of prisoners.144 Jumièges gives much more prominence than Poitiers to the role of Henry I. While it is very probable that it was the French king who was in command, it is also possible that Jumièges was writing in this way in recognition of Henry’s kingly status.145 Another significant factor in determining the result must have been that one of the rebels, Ralph Taisson, changed sides during the battle and rejoined William. In the battle’s aftermath, Guy escaped from the battlefield and shut himself up in his castle at Brionne, a stone hall of an advanced kind then relatively uncommon in Normandy that had presumably been built for Count Gilbert; unlike the late eleventhcentury donjon that can still be seen at Brionne, it was located on an island 144 Roman de Rou, pt. 3, lines 4,147–56 (‘Those who pursued them did not take pity on them. They tore them to pieces and routed them’). For the general point, John Gillingham, ‘Surrender in Medieval Europe – An Indirect Approach’, in Holger Afflerbach and Hew Strachan (eds.), How Fighting Ends: A History of Surrender (Oxford, 2012), 55–72, at 63–4. 145 The point about the French king’s role is well made in John Gillingham, ‘William the Bastard at War’, in Christopher Harper-Bill, Christopher J. Holdsworth, and Janet L. Nelson (eds.), Studies in Medieval History presented to R. Allen Brown, 141–58, at 143 (also printed in Matthew Strickland [ed.], Anglo-Norman Warfare [Woodbridge, 1992], 143–60, at 145).

from childhood into adolescence

85

between two branches of the river Risle at the centre of the modern town.146 Since he is known to have been at Fécamp on 18 July, approximately three weeks before the normally accepted date of the battle, William’s uncle, Count William of Arques, is a possible absentee from the battle.147 However, since the charter already discussed shows him restoring property to one of the greatest of the Norman monasteries, it is possible that what he was actually doing was containing potential rebels in Upper Normandy while his nephew and the king of France were tackling the revolt. Later relations between the count and William do of course raise issues of loyalty, but it seems unlikely that he was actually an opponent of William in 1047. A more likely accomplice to the revolt was William Busac, count of Eu, who, according to Orderic, William at some point besieged at Eu and subsequently exiled.148 Orderic does not date the siege, but the fact that Count William’s last appearance in Duke William’s company is in the charter of 1046/7, in which Count Hugh of Maine and Bishop Ivo of Sées appear, and that his pattern of attestations is similar, if less frequent, than those of other rebels, makes him a likely participant.149 The early twelfthcentury history of the abbey of Saint-Pierre-sur-Dives, which William’s family was in the process of establishing in the 1040s, refers, without naming names, to King Henry giving refuge to some of the defeated of Val-ès-Dunes, thereby supporting this interpretation; it is even possible that Count William of Arques’s staying in Upper Normandy was intended to contain Count William Busac at Eu.150 The latter eventually prospered outside Normandy, obtaining the county of Soissons by marriage. His brother Robert was installed at Eu and became one of Duke William’s most prominent supporters.151 146 GG, 10–13. I am grateful to Jacques Le Maho and Paul Farcy for a walking tour of the site. 147 Bates and Bauduin, ‘Autour de l’année 1047’, 49–51. The traditional date of the battle is 10 August. Douglas, William, 384, proposed to re-date it to the early months of 1047 because the sources indicate that the Orne was in flood. That the Orne is tidal at the likely point of crossing is, however, a sufficient explanation for the fate which befell the retreating rebels. See, in general, De Boüard, Guillaume, 126–7. There is in fact no decisive evidence for the 10 August date, but it is almost certain that the battle took place in the summer. 148 GND, ii, 128–9. 149 RADN, no. 107. For other appearances, often as ‘William son of William’, ibid., nos. 93, 96, 104, 110, 111. 150 GC, xi, instr. col. 155 (ac deuictis fugatisque hostibus ab eodem Guillermo in proximum pagum Vilcassinum accepit). 151 For arguments similar to mine for the date of the siege of Eu, Bauduin, La première Normandie, 297, note 72. Cf. Foulon, ‘Le chevalier Herluin’, 575, note 43, who notes that neither Count William nor his brother Robert occur in charters between 1047 and 1053, and therefore dates the siege to 1053.

86

william the conqueror

THE JOURNEY TO ADULTHOOD The events of the early 1040s and the court-based nature of the rebellion defeated at Val-ès-Dunes in 1047 must surely have made William identify and favour the individuals he could trust and whose abilities he respected. In relation to this process, an inevitable consequence of a strong-willed young ruler advancing to adulthood, there is very little to indicate that Orderic’s belief that William was despised by the indigenous nobility because he was a bastard, and that he was especially hated by the descendants of the Dukes Richard, was the central determinant of events.152 Some of the descendants of Dukes Richard I and II did rebel and oppose William, but many did not. An assessment of the individual life histories demonstrates that these years should be analyzed like all other minorities in terms of jockeying for position and factional struggle. There is no sign that one of Duke Richard I’s descendants, his grandson Count Gilbert, or that a kinsman by marriage, Count Alan, in any way opposed William. Gilbert seems to have been an effective guardian and Alan may even have died fighting for the young duke. Gilbert’s sons Richard and Baldwin who, as children, had been taken for their safety to Flanders when Brionne was handed over to Count Guy, were to be reinstated in the early 1050s, although not at Brionne; in other words, this branch of the descendants of the Dukes Richard prospered.153 The descendants of another son of Richard I, William, count of Eu, also flourished under Duke William, despite the exile of one of them, Count William Busac. William’s uncle, Archbishop Malger of Rouen, was a supportive metropolitan, and perhaps, according to the standards of the times, close to being an exemplary one who worked to preserve peace in troubled times. Duke Richard III’s son Nicholas was appointed abbot of one of the most prestigious and wealthy of the duchy’s monasteries in 1042. Richard, count of Evreux, son of Archbishop Robert of Rouen, caused no difficulties for William that we know of. In the far south-west of the duchy, William Werlenc, count of Mortain, was so marginal that no source mentions him at all. And, while Count William of Arques might fit Orderic’s characterization of the permanently disgruntled kinsman who despised his nephew the duke because he was a bastard, it is very odd that he waited eighteen years until 1053 before entering into conflict with William – especially since, as we have seen, his charters demonstrate a constructive and supportive, if also self-serving, career. To have delayed a rebellion fuelled

GND, ii, 96–7. OV, iv, 208–9; Richard Mortimer, ‘Clare, Richard de [Richard fitz Gilbert] (1030x35–1087x90)’, ODNB. 152 153

from childhood into adolescence

87

by the motives suggested by Orderic, until William was an adult in almost full control of his duchy, seems tactically absurd. It is worth remembering, too, that the social origins of the likes of William fitz Osbern and Roger de Beaumont were every bit as embedded within the heart of Norman aristocratic society as were those of the so-called ‘Richardides’; much more so indeed than those of Nigel the vicomte and Ralph Taisson, who had both been protégés of Duke Richard II, and Count Guy, who had been parachuted into Normandy from Burgundy. In short, there was no co-ordinated opposition to William caused either by his birth or by any group consciousness on the part of the descendants of Dukes Richard I and II. The pack was being reshuffled, as was the norm at such times. Nonetheless, Orderic’s comments on William’s supposed troubles with ‘les Richardides’ ought to be treated as an unconvincing attempt at a generalized assessment of the historical period up until 1060, rather than as an accurate description of the predetermined attitudes of those around the young William. They do seemingly encompass the revolt of 1047 and the conflict with William of Arques in 1053, both of which involved descendants of Dukes Richard I and Richard II, but the categorization as it has since been applied leaves out the many more numerous members of that group who served William well. There were some among these individuals who did have an elevated sense of status and an expectation of reward, as would seem above all to have been the case with Count William of Arques. But this does not explain all the events of William’s adolescence. In a passage in Book I of his Ecclesiastical History completed by 1136, Orderic seems to have changed his views about the whole subject, attributing the disorders to the innate turbulence of the duchy’s aristocracy. His chronology there also accords with that of earlier sources, since he places the wars in William’s childhood, while assigning Guy of Brionne’s revolt and the Battle of Val-ès-Dunes to the time of his adolescence.154 More generally, the use that has been made by modern historians of Orderic’s opinions highlights the one-dimensional nature of the sources for this period in William and Normandy’s history compared with that available for other minorities. The evidence that we have fits much better the context of the kind of factional conflict and feud-based behaviour typical of many medieval minorities. Guy’s conduct also fits in with another phenomenon of the times, namely, the frequency of wars over disputed successions, such as those that occurred in the eleventh-century English kingdom and within eleventh-century northern French territorial principalities. The wars 154 In puericia uero eius Normanni genuine inquietudine concitati rebellauerunt and iuuenis dux (at the time of Val-ès-Dunes), OV, i, 158.

88

william the conqueror

between William’s sons after his death were another manifestation of the same phenomenon. Perhaps Guy actually had to advance a claim in order to give legitimacy to his activities. This, after all, is exactly what his cousin was later to do. The adolescent William surely managed the military rite of passage required of an eleventh-century aristocrat without difficulty. However, if William of Malmesbury is to be believed, he abstained from all sexual activity to the extent that he was thought to be impotent, adding that he had to be persuaded to marry. The statement may just be a colourful version of William of Poitiers’s topos that the young William was moderate in all his indulgences. It could also be an imaginative elaboration of Malmesbury’s theme that William and the Normans were predestined to save the English from religious decadence and his consistent equation of virtue with sexual restraint.155 Yet the absence of any reference to children born to William outside marriage – surely something on which both Orderic and Malmesbury would have commented – supports the probability of sexual abstinence and of behaviour that was rather untypical for a young eleventh-century ruler. If we accept the story, then it is possible that William did find the sexual rite of passage problematic. On the other hand, he might fit a model of masculinity well established in the tenth and eleventh centuries, and one that we know was prevalent at the Norman court; of an older generation, Edward the Confessor is a likely example, while William’s protégé Simon, count of Amiens, Valois and the Vexin, is a certain example from the next generation.156 The attitudes of abstinence could also be a legacy of the strong penitential atmosphere of the last years of William’s father Robert. It is indeed possible that William’s own beginnings influenced his personal behaviour; Herleva not only being alive but also close to her son might even have encouraged his abstinence, as might the opinions of the churchmen who were known to be influential, such as Abbot John of Fécamp. But whatever the explanation, the political implications of sexual inactivity must have prompted thoughts of uncertainty about the succession. In a competitive world within which opportunism and violence were deeply embedded and in which minorities almost automatically produced factional tension, the young William was arguably pretty well served by 155 Preter ceteras uirtutes, precipue in prima adolescentia castitatem suspexit, in tantum ut publice sereretur nichil illum in femina posse, GR, i, 500–1. For Malmesbury and sexual restraint, Kirsten A. Fenton, Gender, Nation and Conquest in the Works of William of Malmesbury (Woodbridge, 2008), 62–75, 103–6. Note, Iuuentutis in primordio moderarissimum uirum agens, GG, 30–1. 156 In general, Janet L. Nelson, ‘Monks, Secular Men and Masculinity, c.900’, in Hadley (ed.), Masculinity in Medieval Europe, 121–42.

from childhood into adolescence

89

those around him, at least as far as circumstances allowed. The traditions and values of his predecessors were certainly sustained and to a large extent respected. What we know suggests that by 1046, William was a formidable young man who was starting to dominate in ways that some were finding irksome. Certain aspects of his behaviour, notably in relation to his religious and sexual attitudes, may already have betrayed a moral austerity that could be a pointer to the future. It is possible, as has been suggested in the case of the emperor Henry IV – whose adult life was both notably turbulent and involved a remarkable doggedness in the face of problems – that the difficulties of these years created a distaste in William for the complexities of politics and fuelled displays of temper and impetuosity.157 But the characteristics nurtured and hardened in the intelligent and physically strong adolescent were surely a sense of entitlement and a remorseless determination to enforce what he believed to be his due. There may have been in his temperament a steeliness of purpose that was exceptional, and present from an early age. With so little evidence to work from, there cannot be any certainties. If the line I have pursued in this chapter has merit, it suggests that these years were more influential in shaping the adult personality and the ruler than those of William’s childhood; they were more turbulent and, for a short time, genuinely disruptive.158 But there is no need to pathologize him, as David Douglas effectively did with his comments about William being ‘bitterly annealed’,159 or to suggest that the troubles of these years made him someone ‘who believed in guile’ and ‘was distrustful and secretive and gave his confidence to few’.160 As a child William had started to learn the military skills required of an adult, had been introduced to a notably religious way of life, and had the matter of the English succession brought to his attention. Many influences illustrating how a successful ruler should conduct himself were present during these years, which indicate continuities stretching back into the times of William’s father and grandfather and sometimes beyond. There were also many examples from elsewhere. Much in his childhood was also culturally predetermined: models of kingship and rule from recent times, Althoff, Heinrich IV., 41–2, 292. As Barlow, William, 12, but without the literary flourishes such as ‘a society riddled with intrigue and deceit’. 159 Douglas, WC, 374 (‘Doubtless his character had been bitterly annealed during his terrible childhood’). Douglas’s choice of the word ‘annealed’ is interesting. A technical term used by a blacksmith, it refers to the process of heating the metal to a specific high temperature at which it becomes workable and malleable, after which it is allowed to cool slowly to harden. I am grateful to Frances Eustace for an informative discussion. 160 Barlow, William, 12. 157 158

90

william the conqueror

and of course from the Old Testament, were profoundly important. Many individuals helped William as a young man, and his maternal kindred were perhaps especially important at crucial times. It is quite easy to see how the events of 1040 through to 1047 could have hardened him and inculcated within him a belief that dominance was the route to quell festering opposition and to reassure the uncertain to trust him, while generating the notion that conduct beyond the accepted norms was sometimes permissible to reach a result. Successful rulers could emerge from minorities that were much more turbulent than William’s had been. England’s Edward III is a good example, with his recent biographer commenting on the clarity and rationality of Edward’s assessment of his circumstances when he assumed power at the age of just seventeen.161 William may not have been very different.

161 W. Mark Ormrod, Edward III (New Haven, CT, and London, 2011), 25 (‘By the time he assumed control of his own regime in 1330, when still just seventeen, Edward III understood only too well that the future of the monarchy would depend on his ability to liberate himself, constitutionally and emotionally, from the thrall of his parents’).

Chapter 3

TH E S HAP I N G O F T HI NG S TO C O ME The six years after the Battle of Val-ès-Dunes in 1047 contain events that were to shape the rest of William’s life and, ultimately, those of thousands of others over many generations. Prestigiously married by mid-1053 to Matilda, daughter of Count Baldwin V of Flanders, he had been involved in 1051–2 in the discussions with Edward the Confessor that were later to be at the heart of the justification for the invasion of England in 1066. William had participated in the establishment of good relations with the series of reforming popes installed after Emperor Henry III’s intervention in Rome in 1046, and he must have been involved in the Norman Church’s positive response to Pope Leo IX’s Council of Rheims of October 1049. In the reversal inflicted on Count Geoffrey Martel at Alençon and Domfront in 1051–2, he gained his first military victory over a ruler of a great northern French territorial principality. Yet, for all that was to follow from these developments, the event that had the greatest immediate political impact was the same Geoffrey Martel’s takeover of the county of Maine in 1051. Probably as a result, by 1053 William was facing the likelihood of the breakdown of good relations with King Henry I and the possibility of fighting a war against him and Geoffrey. Although he must have showed notable tenacity and drive in responding to opportunities and changes during these years, the way in which the future would unfold was extremely unclear. Almost every one of the developments of this period has been the subject of differing opinions and, often, of controversy. Thus, for example, although the date of 1050–1 for William and Matilda’s marriage is deeply embedded in modern historiography, there is actually no evidence at all for it. The later dating that will be proposed here, which was once the generally accepted one, has implications for another central aspect of the narrative, namely, the long-established belief that the couple married in defiance of a papal ban. The capture of Alençon and Domfront has been variously dated to 1049 or 1051–2; my preference – like David Douglas’s – is for the latter dating.1 Above all, there is the problem posed by William of 1 For the arguments for 1049, Guillot, Anjou, i, 71–2, notes 314, 320; Louise, La seigneurie de Bellême, i, 354–8; Bauduin, La première Normandie, 309–10.

91

92

william the conqueror

Malmesbury’s tantalizing admission that he found it impossible to write a convincing narrative of the discussions between Edward the Confessor and William on the basis of the information at his disposal. This, he said, was because his sources disagreed irreconcilably, a problem that he explained in terms of long-standing hostility between the Normans and the English.2 In fact, the uncertainty derived not just from ethnic tensions and conflicts, but because there had never been a narrative on which everyone could agree in the first place. William’s career was irretrievably a subject of controversy from – let’s say – dusk on 14 October 1066! And the discussions with Edward shaped a largely unknowable narrative that lies beneath the surface of events through the 1050s and early 1060s. With all these events and developments interwoven in a bewildering multiplicity of ways – some of which have not been previously fully grasped – a certain amount of technical discussion is indispensable to the construction of this chapter’s narrative. THE AFTERMATH OF VAL-ÈS-DUNES William’s presence at King Henry I’s court at Senlis (modern-day Picardy) at Whitsun (23 May) 1048 signals his resumption of the role within northern French political society that his father and grandfather had played. A narrative passage in the charter that records his attendance describes him proposing that the castle of Vic-sur-Aisne (near Compiègne) be restored to the abbey of Saint-Médard of Soissons over which King Henry had taken on the role of protector; a sign perhaps that William was accorded an honoured position by the French king.3 Also present at the court was a remarkable array of the political elite of northern France: the archbishop of Rheims, ten bishops, and several territorial princes, including Count Baldwin V of Flanders and Count Theobald III of Blois-Chartres, as well as others, some of whom were to play significant roles in William’s life. This journey beyond Normandy’s north-eastern frontier must also have enabled William to transact other business relating to that region. On 30 October he restored a church in Normandy to the possessions of the great Merovingian abbey of Saint-Riquier in Ponthieu, from which he obtained relics of St Vigor for his father’s eponymous religious foundation of SaintVigor of Cerisy.4 A sign surely of reverence for his father’s memory and, yet again, a practical demonstration of continuity from his time, this event GR, i, 354–7. RADN, no. 114. For the assembly at Senlis, Bauduin, La première Normandie, 307–8. 4 RADN, no. 115; Hariulf, Chronique de l’abbaye de Saint-Riquier, ed. Ferdinand Lot (Paris, 1894), 225, 228. 2 3

the shaping of things to come

93

is also indicative of an interest in saints and relics that runs throughout William’s life. The gathering at Senlis was in all likelihood the springboard for the Mouliherne campaign that Henry I led into Anjou in the autumn of 1049, for which William of Poitiers is the only source.5 Involving an army of major princely contingents, including William’s, its purpose must have been to counteract the consequences of the alliance between the emperor Henry III and Geoffrey Martel that threatened the interests of almost everyone present at Senlis, either because they were involved in the complicated ongoing wars between Baldwin of Flanders and the emperor or in the struggle on the river Loire between Geoffrey and Theobald of BloisChartres. The wars between Baldwin and Henry III also involved Edward the Confessor and Earl Godwine of Wessex, albeit in ways that effectively put them on opposite sides: while Edward sent a fleet to blockade the Flemish coast in support of the emperor in 1049, Godwine’s close links with Baldwin made Flanders a refuge for his family in troubled times.6 This situation must also have been a factor in the marriage negotiations between William and Baldwin and the discussions of the English succession involving William, Edward, and Godwine, and – although this is not always appreciated – Baldwin. William of Poitiers used the Mouliherne campaign to insert another of his eulogies of William’s personal qualities into his narrative. Like other similar passages, this was part of the rhetorical apparatus central to the Gesta Guillelmi, in this case to demonstrate to his hoped-for audiences the military skills that were central to the creation of trust and confidence in those whom William was going to have to lead in war. He portrayed William as a young man eager to display prowess, ready to take risks, and physically capable. He tells how William, out riding with four companions, was ambushed by fifteen enemies. William’s response was to incapacitate the strongest of them by throwing his lance to disable him, and then to lead a chase resulting in the capture of seven others. He also stresses the rapidity with which William’s qualities were acknowledged. Geoffrey Martel, a great soldier and therefore, as Poitiers pointed out, the very best of authorities, is said to have thought William peerless as a warrior and a horseman. So important for William of Poitiers was this episode that he authenticated his narrative by mentioning that he was himself at the time at Poitiers, to the south of the county of Anjou, and GG, 14–17. Although Poitiers gives no date, autumn 1049 is generally agreed. For condensed narratives of these wars, Dhondt, ‘Henri Ier, l’Empire et l’Anjou’, 89–101; Heather J. Tanner, Families, Friends and Allies: Boulogne and Politics in Northern France and England, c.879–1160 (Leiden, 2004), 84–8. 5 6

94

william the conqueror

therefore well placed to assemble testimony. Other manifestations of William’s entry into an international noble society that he reported included gifts from Gascony, the Auvergne, and Spain, among them thoroughbred warhorses. Although the earliest reference to a projected marriage between William and Baldwin V’s daughter Matilda is its prohibition in October 1049 at the Council of Rheims, it must surely have been discussed when William and Baldwin were in each other’s company at Senlis in 1048. While Baldwin was certainly in need of allies because of his war with the emperor Henry III and, on this basis would have been proactively seeking support, the fact that, as a ruler with a proven record of competence, he was prepared to contemplate an alliance with William is compelling testimony to the reputation and prestige that William must have acquired by this time. The projected marriage also represents the resumption of the long-term good relations between Flanders and Normandy: Baldwin’s father Baldwin IV had, for example, been in Normandy in 1035 when Duke Robert had secured agreement to William’s succession, and Robert had earlier helped Baldwin defeat the rebellion of his son, the future Baldwin V.7 Poitiers, well aware that Matilda came from socially superior stock to her husband, mentioned she was the descendant of kings and a relative of the emperor.8 That William was prepared to wait until all possible obstacles to the marriage had been cleared away surely indicates a determination on his part to marry at an appropriate social level. Throughout 1048, and during his absences from his duchy, William’s forces kept Count Guy shut up in his castle at Brionne. Although the siege may have lasted into 1049, ending when the garrison was starved out, its long duration is not a sign of any enduring depth of resistance to William. Rather it demonstrates William’s confidence, a statement by him that he was in control of the situation and could devote most of his time to important activities elsewhere; Guy had no allies and was being left to stew.9 The garrison’s supply routes were cut off by the construction of earthwork forts on both sides of the river Risle, one on the site of the castle of which the ruins of the twelfth-century keep can still be seen, and the other, of which some of the earthworks are still visible, in the hills to the west of Brionne, thereby controlling road access and allowing reconnaissance from the higher ground above the encircled castle.10 GND, ii, 52–5. GG, 30–3. 9 For the three-year duration, OV, iv, 201–2; vi, 210–11. On its interpretation, GG, 11, note 3. 10 GG, 10–13. For maps and some discussion, Jimmy Mouchard, ‘Le Vieux Château à Brionne’, in Anne-Marie Flambard Héricher (ed.), Projet collectif de recherches sur les fortifications de terre en Haute-Normandie 2007, SRA Haute-Normandie, 63–96, esp. pp. 76–7, 79–82 7 8

the shaping of things to come

95

According to Jumièges, Guy was placed under house arrest once he realized that his position was hopeless. In Poitiers’s account, we are told that, despite William’s wish to be merciful, Guy had gone home, choosing to go into exile in Burgundy where he tried to take the duchy from his elder brother.11 While past history did indeed indicate that a rebellious ducal kinsman might eventually expect reinstatement, Guy, as William’s companion since childhood, would surely have known him well enough to realize what the future might hold; in all likelihood, his hopes of recovery were at best long term and qualified, if that.12 He was also different from all other previous members of the kindred in that he had been dropped into Norman society. As noted in the Prologue, while scripts and norms certainly existed to deal with a situation of this kind, within the basic framework different people played the game in different ways. With favour, reward, and punishment founded upon the judgements and actions of the individual ruler and the choices that he – and, occasionally, she – made, we begin therefore to have a sense of William’s character as a ruler and how others reacted to him. The date when Guy left Normandy is unknown, but it was presumably soon after the submission of Brionne.13 However, most probably after her husband’s death in 1057, his mother Adeliza did return to the duchy to be reinstated by William at Le Homme.14 The other rebels were treated differently from Guy, since none of them was exiled. Even his closest associate, Nigel the vicomte, seems to have continued in office and appears in charters into the mid-1050s.15 That his kinsman Geoffrey became bishop of Coutances in 1048, possibly because the bishopric was purchased for him by Nigel, suggests that a wish to achieve a modus vivendi was present on both sides. Nigel’s eventual departure from Normandy was therefore an (figs. 1–5). I am grateful to Anne-Marie Flambard Héricher for allowing me to consult this unpublished report. 11 GND, ii, 122–3; GG, 12–13. 12 Lucien Musset, ‘Autour des modalités juridiques de l’expansion normande au XIe siècle: le droit d’exil’, in Musset et al. (eds.), Autour du pouvoir ducal, 45–59, at pp. 46–51; Elisabeth van Houts, ‘The Vocabulary of Exile and Outlawry in the North Sea Area around the First Millennium’, in Laura Napran and Elisabeth van Houts (eds.), Exile in the Middle Ages (Turnhout, 2004), 13–28, at 19–21; Ewan Johnson, ‘The Process of Norman Exile into Southern Italy’, in Napran and van Houts (eds.), Exile in the Middle Ages, 29–38. 13 Of Guy’s three apparent post-1047 appearances in charters, the two for the abbey of Saint-Julien of Tours can be disregarded, RADN, nos. 131, 142, since the eighteenth-century copies of the abbey’s charters seem corrupt (Regesta, no. 279). RADN, no. 194, may be authentic if Matilda’s signum is taken to be a later interpolation, since it confirms grants made to Bishop Herbert, who died in 1046. It therefore dates to within Guy’s career in Normandy. 14 Regesta, no. 58. 15 See, for example, RADN, no. 132.

96

william the conqueror

element in the politics of the mid-1050s, and not, as is usually assumed, of the aftermath of Val-ès-Dunes; Poitiers cryptically mentions a later serious offence on Nigel’s part, and, according to Wace, he went to Brittany.16 His appearance in a charter of Count Eudo, brother of Alan III, count of Brittany, in either 1056 or 1057, confirmed by Geoffrey Martel, shows that Nigel did indeed go to Brittany and that he sought refuge with William’s enemies.17 Although Nigel, or perhaps his son, had regained the family lands by the early 1060s, and, according to Wace, he was present at Hastings, they had lost the status of earlier generations; Wace actually says that his conduct at Hastings was an attempt to win William’s favour. Having previously been one of the great magnates of the Norman duchy, Nigel’s status for the rest of William’s life, like that of Rannulf of the Bessin and the turncoat Ralph Taisson, was that of a major regional magnate and relative outsider. After 1066, whereas some of Nigel’s tenants received lands in England, he acquired none.18 His display of prowess on the field of Hastings had evidently failed to change William’s mind. A commentary on the family’s feelings may be contained in the pancarte, dating to the first half of the 1080s and describing the foundation of the abbey of Saint-Sauveur-le-Vicomte, in which William is said to have acquired the kingdom of England by battle; possibly from within Normandy, a most unusual contradiction of the official case for a legitimate take-over, or perhaps this was just a statement that divine favour had led to William’s victory.19 Of the rebels of Val-ès-Dunes and their families, only the children of Haimo Dentatus, who had been killed in the battle, bucked the general trend by profiting from the conquest of England. But they too, in the shorter term, suffered the loss of lands centred on Creully, between Caen and Bayeux, subsequently recovering only some of them.20 Grimoult du Plessis-Grimoult was singled out for special treatment because 16 GG, 12–13 (Nigellum alio tempore, quoniam improbe offensabat, exilio punitum fuisse comperio); Roman de Rou, 276–7, lines 8,353–62. 17 Cartulaire de l’abbaye de Saint-Aubin d’Angers, ed. A. Bertrand de Broussillon (Paris, 1903), no. 677 (Guillot, Anjou, ii, 127 [Catalogue des actes, no. C176]); Jean-Pierre Brunterc’h, ‘Geffroy Martel, Conan II et les comtes bretons Eudes et Hoël de 1055 à 1060’, in Catherine Laurent, Bernard Merdrignac, and Daniel Pichot (eds.), Mondes de l’Ouest et villes du monde: Regards sur les sociétés médiévales. Mélanges en l’honneur d’André Chédeville (Rennes, 1998), 311–24, at 314–15; Morin, Trégor, Goëlo, Penthièvre, 83–5; Les actes des ducs de Bretagne, 301 (no. 55). 18 DP, 7; Hagger, ‘How the West was Won’, 37–8. This point was clearly made by Léopold Delisle, Histoire du château et des sires de Saint-Sauveur-le-Vicomte (Valognes, 1867), 21. 19 Regesta, no. 260 (qui regnum Anglie bello adquisiuit). 20 The Domesday Monachorum of Christ Church, Canterbury, ed. D. C. Douglas (London, 1944), 55–6; Lucien Musset, ‘Actes inédits du XIe siècle: Les plus anciennes chartes du prieuré de Saint-Gabriel (Calvados)’, BSAN, lii (1952–4), 117–41, at 124–9; K. S. B. Keats-Rohan, ‘The Prosopography of Post-Conquest England: Four Case Studies’, Medieval Prosopography, xiv (1993), 1–52, at 25–8.

the shaping of things to come

97

he had tried to assassinate William, suffering total forfeiture of all his lands and kept imprisoned in chains for the rest of his life.21 Beneath the facade of apparent reconciliation lurks a sense that for William their offences were ultimately unpardonable, a position he seems to have maintained for the rest of his life. Within Normandy, William began to promote individuals who were very obviously his favourites. His half-brother Odo, a son of Herluin and Herleva, was appointed to the bishopric of Bayeux following Bishop Hugh’s death, which occurred soon after his attendance at the Council of Rheims in the autumn of 1049.22 In 1050, William’s stepfather Herluin established a monastery at Grestain, a display of confidence on the part of Herleva’s husband that is very striking. Although the abbey’s subsequent development was a slow one, the foundation of monasteries was at this time very much the unique prerogative of members of the duchy’s most powerful aristocratic families; Herluin and Herleva truly were making a statement that a new order of things was emerging. Although it goes unrecorded, William must surely have played a significant role in this too.23 At around the same time, the vast lands that Bishop Hugh had held in addition to his bishopric, as the son of Duke Richard I’s half-brother, Count Rodulf of Ivry, were mostly passed to his nephew William fitz Osbern; other potential heirs, including fitz Osbern’s brother Osbern, were excluded. It surely cannot be a coincidence that in c.1050 Osbern crossed the Channel, having become a cleric, and became a chaplain to King Edward the Confessor.24 Roger de Beaumont appears as vicomte of Rouen at a date before 1053, thereby sidelining the family of Goscelin and his son Godfrey, the previous vicomtes and the founders in Rouen in c.1030 of the abbeys of La Trinité-du-Mont and Saint-Amand. This may be a comment on the treatment of individuals and communities whose loyalty had been questionable during the 1040s; William of Poitiers does suggest that Rouen had been less than wholehearted in its support of the young William.25 That Roger’s tenure was not long term shows that a statement of authority was being made; he presumably made his, and William’s, 21 For Grimoult, Roman de Rou, pt. 3, lines 4,203–26; Regesta, no. 27; Henri Navel, ‘L’enquête de 1133 sur les fiefs de l’évêché de Bayeux’, BSAN, xlii (1934), 5–80, at 15–16. In general, Emily Zack Tabuteau, ‘Punishments in Eleventh-Century Normandy’, in Warren C. Brown and Piotr Górecki (eds.), Conflict in Medieval Europe: Changing Perspectives on Society and Culture (Aldershot and Burlington, VT, 2003), 131–49, and esp. pp. 137, 143–7; Matthew Strickland, War and Chivalry: The Conduct and Perception of War in England and Normandy, 1066– 1217 (Cambridge, 1996), 231–57. 22 For Odo’s appointment, Spear, Personnel, 32. 23 Regesta, no. 158; Bates and Gazeau, ‘L’abbaye de Grestain’, 11–13, 22–3. 24 Bauduin, La première Normandie, 202–3, 223–31. 25 GG, 12–13.

98

william the conqueror

presence felt, and then moved on.26 The two abbeys were also thereafter treated as ducal foundations. William confirmed several new monastic foundations in Normandy during 1050. William and Robert fitz Giroie and Hugh and Robert de Grandmesnil, the founders of Saint-Evroult, whose beginnings are described in detail by Orderic, approached him to approve their intention and had him confirm the foundation by charter. They also sought his permission to appoint an abbot. When Thierry, a monk of Jumièges, had been chosen, William invested him with his staff of office before he was taken to the new abbey.27 Although both this account and the foundation history of the abbey of Saint-Pierre-sur-Dives (referred to in Chapter 2 in relation to events in 1046) were written in the early twelfth century, together they show that the two authors believed the full paraphernalia of ritualized ducal authority involving abbatial investiture was in place during what were supposedly difficult years.28 The process of regaining and extending influence in the territories around and beyond Normandy’s southern frontier also restarted at this time. Orderic presents a picture of the humiliated William Talvas, sometime lord of Bellême, driven from his lands by his son Arnulf at a date that must have been before 1048, agreeing to the marriage of his daughter Mabel to Roger de Montgommery and using the marriage to obtain Roger’s support. Although – once more – the remarkable results could not have been predicted, the immediate consequence was to give Roger a role within the politics of the region that, for a Norman, was unprecedented. In the short term, however, the forces pulling this area of southern Normandy southwards into the orbit of the hegemonic power of the counts of Anjou were intensified when Ivo, bishop of Sées, succeeded to the lordship of Bellême in the early 1050s and Count Geoffrey Martel occupied the county of Maine in 1051.29 William’s immediate response to this change in the balance of power was a measured proactive one. Having seemingly first been invited to intervene in Maine in the mid-1040s, he provided a refuge for the bishop of Le Mans, Gervase du Château-du-Loir, an inveterate

26 Cartulaire de Saint-Pierre-de-Préaux, no. A162. The notice’s ambiguous dating clause (Eodem anno quo in conjugium sortitus est Normannorum marchio, Willelmus nomine, Balduini comitis filiam), which could refer either to when he decided to get married or to the celebration of the marriage, means that its dating needs to bear in mind my arguments below about the date of the marriage. In general, Bouvris, ‘Contribution à une étude de l’institution vicomtale en Normandie’, 159–61. 27 OV, ii, 16–19. For the charter, RADN, no. 122. 28 See further, NMPrinces, 67–71. 29 GND, ii, 118–19; Louise, La seigneurie de Bellême, i, 358; Thompson, ‘Family and Influence’, 219–21.

the shaping of things to come

99

opponent of Geoffrey’s designs on Maine whom Geoffrey had imprisoned.30 An indication that William was now a power to be feared is Geoffrey’s complaint to Pope Leo IX that Bishop Gervase was inciting both William and Henry I to make war on him.31 William on the other hand could now pose as the protector of a persecuted bishop. The event can be seen as a personal intervention in the politics of northern France that was to have the most profound consequences. In particular, it looks as if William was allying himself with Count Theobald of Blois-Chartres in his difficult struggle against Geoffrey Martel.32 The consequences of his readiness to take on Geoffrey, one of the most formidable political and military figures of his day, were to be remarkable. They suggest an exceptional level of self-confidence. THE COUNCIL OF RHEIMS AND THE PROPOSED MARRIAGE OF WILLIAM AND MATILDA The presence at the Council of Rheims in the autumn of 1049 of five of the seven bishops of the Norman duchy shows a determination on their part and on William’s to engage with the more interventionist papacy that the council epitomized; the only absentees were Archbishop Malger of Rouen and Bishop William of Evreux.33 This council and the other papal councils of 1049–50 were not only the first occasions on which a pope had travelled north of the Alps for a long time, but also opportunities for the dramatic propagation of an agenda whose main targets were simony and the marriage practices of the laity. Under the second of these headings, the council criticized several marriages: the projected marriage between William and Matilda, daughter of Count Baldwin V of Flanders; the marriage of Eustace II, count of Boulogne, to Edward the Confessor’s sister Godgifu; and that of William’s sister Adelaide to Count Enguerrand II of Ponthieu. The reason for the ban on the marriage between William and Matilda is not, however, stated in Anselm of Saint-Rémy’s account of the council’s proceedings, which was written within the following decade, between 1055 and 1060.34 30 Actus Pontificum Cenomannis in urbe degentium, ed. Gustave Busson and Ambroise Ledru (Le Mans, 1902), 366. 31 Briefsammlungen der Zeit Heinrichs IV, ed. Carl Erdmann and Norbert Fickermann (Weimar, 1950), no. 84. 32 For this suggestion, Jean Dunbabin, ‘Geoffrey of Chaumont, Thibaud of Blois and William the Conqueror’, ANS, xvi (1994), 101–16, at 108. 33 Anselme de Saint-Rémy, Histoire de la dédicace de Saint-Rémy, ed. and trans. J. Hourlier, in La Champagne bénédictine: contribution à l’année saint Benoît, Travaux de l’Académie nationale de Reims, vol. clx (Rheims, 1981), 181–297, at 236 (PL, cxlii, col. 1431). Anselm named Herbert as the bishop of Lisieux present, but Jean-Hervé Foulon has argued that it was his successor Bishop Hugh, Foulon, ‘Le chevalier Herluin’, 574, note 41. 34 Anselme de Saint-Rémy, ed. Hourlier, 252 (PL, cxlii, col. 1437).

100

william the conqueror

Because the canon law of marriage at this date was relatively inchoate, various possible explanations for the ban have been suggested, of which the most plausible are the couple’s common descent from Rollo, the first Scandinavian ruler of the lands that became Normandy, that Matilda’s mother Adela had been briefly betrothed to William’s uncle Duke Richard III, and a common descent from Fulk II the Good, count of Anjou (940–62). The descents from Rollo and Count Fulk meant that they were related within the fifth and fourth degrees of consanguinity respectively, the former a similar relationship to the one that led to the separation of William’s sister Adelaide and Count Enguerrand II of Ponthieu, a marriage that was also condemned at the council. Since Adela’s marriage to Duke Richard could not possibly have been consummated because of her young age, the two consanguineous relationships are the more likely explanation.35 Additionally, the council’s unspecified criticism of the marriage of Count Eustace II of Boulogne to the Confessor’s sister Godgifu seems to have resulted in an amicable end to their childless marriage and to her return to England, a land she had probably not seen for over thirty years, where she was treated generously by her brother. A political element was undoubtedly present in the council’s decisions, since all of the marriages condemned involved allies of Count Baldwin and therefore actual or potential enemies of the emperor, Pope Leo IX’s protector. This must have been a factor that influenced both William and Baldwin’s reactions.36 William’s determination at this time to build political bridges with the principalities and lordships to the north-east of Normandy, of which the divorced Adelaide’s marriage to Count Eustace’s nephew, Count Lambert II of Lens, is another symptom, suggests that he would want to overcome the obstacles created by the prohibition.37 That other aspects of the council’s proceedings proved to be controversial in regions beyond Normandy and Flanders also suggests that a process to calm tensions would follow.38 35 Henri Prentout, ‘Etudes sur quelques points de l’histoire de Guillaume le Conquérant. II. Le mariage de Guillaume’, Mémoires de l’Académie de Caen, vi (1931), 29–56, at 24; Ryan Patrick Crisp, ‘Consanguinity and the Saint-Aubin Genealogies’, HSJ, xiv (2005, for 2003), 105–15, at 112–15; Thompson, ‘Being the Ducal Sister’, 70. See further Patrick Corbet, Autour de Burchard de Worms. L’Eglise allemande et les interdits de parenté (IX ème–XII ème siècle) (Frankfurt-am-Main, 2001), 301–5. Douglas described Prentout’s study as ‘a most valuable article’, but was then sceptical of its conclusions, WC, 392, note 8. 36 For a survey of the situation that includes reference to the political aspects of the council, de Boüard, Guillaume, 163–8. On the diplomatic situation, Heather J. Tanner, ‘The Expansion of the Power and Influence of the Counts of Boulogne under Eustace II’, ANS, xiv (1992), 251–86, at 263–8; idem, Families, Friends and Allies, 86–9. 37 For these marriages and the redeployment of Adelaide, Thompson, ‘Being the Ducal Sister’, 70–1. See also Ann Williams, The World before Domesday: The English Aristocracy, 900–1066 (London, 2008), 141. 38 See further Michel Bur, ‘Léon IX et la France (1026–1054)’, in Georges Bischoff and Benoît-Michel Tock (eds.), Léon IX et son temps (Turnhout, 2006), 233–58, at 244–52.

the shaping of things to come

101

The authority accorded to the council’s decisions, not only by the likes of Counts Enguerrand and Eustace who separated from their wives, but within Normandy in relation to matters affecting the clergy, is very striking. Of the bishops from Normandy present at Rheims, two had to clear themselves of serious charges. Geoffrey of Coutances persuaded Pope Leo that, although his bishopric had indeed been purchased for him, he was not guilty of simony because the deed had been done without his knowledge. Ivo of Sées was censured for setting fire to his own cathedral, but convinced the pope that the attack, which was part of the warfare within the Bellême family, had been required to eject enemies who had taken over the church.39 Both followed up their pardons by going to southern Italy – and in Ivo’s case to Constantinople as well – to raise funds to contribute to the construction of new cathedrals. Geoffrey also attended Pope Leo’s Rome synod in April 1050.40 In terms of future developments, the latitude given to the two bishops is extremely significant; it shows that negotiation and conciliation were Leo’s preferred policies. It is also important that England, as well as Normandy, responded positively to the Council of Rheims, which was highlighted in two versions of the Anglo-Saxon Chronicle; ‘E’ also included the information that Bishops Herman of Ramsbury and Ealdred of Worcester – Ealdred would crown William king of the English on 25 December 1066 – were present at the Rome synod.41 Whether Lanfranc, at this point prior of the abbey of Le Bec, attended the Council of Rheims is unclear. Nonetheless, the months that he spent in 1050 travelling around western Europe with Leo IX were to be of the greatest importance to his future and to William’s, because he kept company with the individuals who were to dominate the papacy for the remainder of both of their lives, most notably, Hildebrand (the future Pope Gregory VII) and the polemicist Humbert of Moyenmoutier. This was the foundation for Lanfranc’s later role as a trusted intermediary between the Anglo-Norman lands and the papacy. It also enhanced his credentials as a teacher, since, whatever view is taken of the number and names of his pupils, Lanfranc undoubtedly did attract students to Le Bec in the 1050s and early 1060s, among them protégés of popes.42 But in the immediate Anselme de Saint-Rémy, ed. Hourlier, 248 (PL, cxlii, col. 1436); GND, ii, 116–19. Histoire des conciles d’après les documents originaux, ed. C. J. Hefele and Dom. J. Leclercq, iv, 2 (Paris, 1911), 1,040. 41 ASC, ‘D’, ‘E’, 1049. All versions of the Chronicle mention that the two bishops visited Rome, but only ‘E’ specifically states that they attended the synod, ASC, ‘E’, 1050. 42 For the positive and negative views of the school of Le Bec, Margaret Gibson, Lanfranc of Bec (Oxford, 1978), 34–9; Sally N. Vaughn, ‘Lanfranc at Bec: A Reinterpretation’, Albion, xvii (1985), 135–48, at 136–9; idem, ‘Lanfranc, Anselm and the School of Bec: In Search of the Students of Bec’, in Marc A. Meyer (ed.), The Culture of Christendom (London, 1993), 155–81; H. E. J. Cowdrey, Lanfranc: Scholar, Monk and Archbishop (Oxford, 2003), 19–24. For a judicious assessment, Julie A. Potter, ‘The Friendship Network of the Abbey of Le 39 40

102

william the conqueror

context of the year 1050, it is arguably more significant that Abbot John of Fécamp was entrusted by Leo IX with a mission to the Normans who, with their allies, were taking over large areas of southern Italy. This mission and Abbot John’s response to Leo’s flexibility on matters of apparent principle again show just how seriously the decisions of the Council of Rheims and the pope’s attempts at European peace-keeping were taken in Normandy. Abbot John’s role in 1050 must represent a continuation of the transEuropean engagement that had been a characteristic of the Norman Church’s leaders from Duke Richard II’s time onwards. His letter reporting on his mission’s failure – exactly what he was supposed to be accomplishing is unclear – contains a eulogy of Leo IX, who is compared to his great predecessors, his namesake Leo and Gregory, and a statement of support for the papacy’s campaign of moral reform. It alludes to the Normans’ devotion to the papacy, while warning that recent papal conduct was placing this loyalty in jeopardy and seeking compensation for the treatment of the abbot and his companions.43 However, the concluding sections of the letter, written after John had returned to northern France, make specific reference to the decisions of the Council of Rheims, and in particular to the conduct of Count Theobald III of Champagne, who had left his wife for a concubine. John not only reported to Leo that criticism of his conduct was circulating because the pope was thought to be conniving with Theobald, but went as far as to tell Leo that he was jeopardizing the campaign to reform lay marriage practices.44 If someone of John’s importance was adopting a line that was more stringent than the pope’s in relation to the decisions taken at Rheims, it suggests that the prohibition of William and Matilda’s projected marriage was being taken seriously in Normandy; but that, equally, Leo’s readiness to negotiate with the targets of the prohibitions had not passed unnoticed. The maintenance of peace was after all as much a responsibility of bishops and clergymen as the disciplining of lay morals; there was a long history of compromise in matters of this kind.45 All ultimately depended on William’s – and Count Baldwin’s – readiness to listen to religious advice on the matter of their marriage Bec-Hellouin in the Eleventh and Twelfth Centuries’, University of Cambridge PhD Thesis (2014), 182–90. 43 Note in particular nec erit gens illa Northmannorum bellica in uestra fidelitate adeo prompta et deuota, PL, cxliii, cols. 798–9. 44 PL, cxliii, cols. 799–800. Note in particular quod Tedbaldus etiam vestrum colloquium adierit, nihilque dignum hac culpa actum sit. See further Patrick Corbet, ‘In multis orbis partibus: Léon IX et les interdits de parenté (1049–1054)’, in Bischoff and Tock (eds.), Léon IX et son temps, 343–53, at 348. 45 Laurent Jégou, L’évêque, juge de paix. L’autorité épiscopale et le règlement des conflits (VIIIe–XIe siècle) (Turnhout, 2011), 435–43.

the shaping of things to come

103

negotiations. But, at the least, it raises the possibility that William and Matilda may not have pushed ahead and married in defiance of the prohibition, as has often been thought. Given the seriousness with which the council’s decisions were taken, this is in fact very likely. That William and the Norman Church engaged with the decisions of the papal councils of 1049–50 is also revealed by the way in which they involved themselves in the controversies caused by the views on the Trinity expressed by Berengar of Tours, at the time scholasticus of the cathedral church of Angers. Although the treatise written by Durand, a monk of Fécamp who was later to become the abbot of Roger de Montgommery’s monastic foundation of Saint-Martin of Troarn, dates the events it describes to 1053, the meeting of the Norman churchmen held at Brionne under William’s presidency to consider Berengar’s theories has long, and credibly, been dated to late 1050. It fits neatly in that year among the events that followed Berengar’s re-emergence after his ideas had been condemned in his absence at the papal Council of Vercelli (September 1050).46 After an approach to Ansfrid, abbot of Saint-Pierre of Préaux, Berengar, so Durand tells us, contacted William directly. In what Durand describes as a display of precocious wisdom (Uerum ille, licet aetate adolescentia excederet annos), William then called together a meeting of the great and the good of the Norman Church. That Durand and Abbot John subsequently wrote treatises against Berengar in support of what they believed to be the papally approved line suggests that they took the leading role at Brionne.47 Although present, at least according to Berengar’s account of events, Lanfranc, it would appear, stayed in the background.48 He was not yet remotely close to being the prominent figure in the Norman Church that he was later to become. It was above all Abbot John’s advice that William was listening to at this point, something that is very important in relation to the marriage negotiations with Count Baldwin. William’s role at Brionne appears to have been the first occasion on which he presided in person at such a council. Although we cannot truly believe that he was intimately involved in the intellectual issues under discussion, as an aspect of the norms of rule, his role was in no way remarkable because such direct involvement by a lay ruler had been standard for centuries. It simply showed that William was ready to protect the Church against unorthodox interpretations, the concept of the Trinity being of 46 Durand of Troarn, Liber de corpore et sanguine Christi, in PL, cxlix, cols. 1,373–424, at cols. 1,421–2. 47 For these events, Jean de Montclos, Lanfranc et Bérenger: la controverse eucharistique du XIe siècle (Louvain, 1971), 87–94; Gibson, Lanfranc, 64–6; Cowdrey, Lanfranc, 38–42. 48 Cowdrey, Lanfranc, 62–3. See also Toivo J. Holopainen, ‘ “Lanfranc of Bec” and Berengar of Tours’, ANS, xxxiv (2012), 105–21, at 116–18.

104

william the conqueror

course an especially sensitive area, and that he was ready to take on an expected religious role. Both Henry I and Geoffrey Martel, the latter as Berengar’s patron, were also involved in the controversies that he instigated.49 Within Normandy, the conservative outcome of the debate at Brionne on the Trinity, and the engagement with wider issues preoccupying Christendom, must be seen both as a continuation of a pattern established in Duke Richard II’s time and as setting the tone for the decades to come. The holding of the assembly at Brionne, a place at which no similar gathering was to be held during the rest of William’s life, may have a symbolic secular significance. Normandy’s elite were brought together at the place of victory. Religious orthodoxy was being proclaimed at a site of recent successful secular rule. The deliberate association of the two in a way that demonstrated how determined the Norman Church and William were to please the papacy is extremely important in relation to the ongoing negotiations about William and Matilda’s marriage. With Duke William and the Norman clergy clearly trying to please the papacy in matters of dogma, it seems likely that they would have taken a similar line in relation to his proposed marriage. Although the date of 1050–1 for William and Matilda’s marriage is deeply embedded in modern historiography, there is no evidence whatsoever to support it. The earliest clear-cut mention of Matilda as countess/duchess of the Normans (Mathildis comitisse) is in a charter copied into the late eleventh-century cartulary of the abbey of La Trinité-du-Mont of Rouen that dates from the year 1053.50 Her appearance and also that of the couple’s first child, Robert, in a charter whose textual transmission is very complicated, and which might date from 1053 or early 1054, could push the marriage into 1052, but this piece of evidence is of uncertain quality.51 Of the apparently earlier references, Matilda’s appearance among the signa of the charter 49 For the general point, Guy Lobrichon, ‘The Early Schools, c.900–1100’, in Richard Marsden and E. Ann Matter (eds.), The New Cambridge History of the Bible, volume 2: From 600 to 1450 (Cambridge, 2012), 536–54, at 546–8; Elisabeth van Houts, ‘The Planctus on the Death of William Longsword (943) as a source for Tenth-Century Culture in Normandy and Aquitaine’, ANS, xxxvi (2014), 1–22, at 11; Gibson, Lanfranc, 65–6. 50 RADN, no. 130. For a stimulating re-examination of this charter evidence, Charlotte Cartwright, ‘Matilda of Flanders in Normandy: A Study of Eleventh-Century Female Power’, University of Liverpool PhD Thesis (2012), 56–9. 51 RADN, no. 141. Marie Fauroux dated this charter to 1052–8, but the presence of Matilda, the young Robert Curthose, and Ralph, abbot of Mont Saint-Michel, all point to 1053 or, if later, only slightly so. Although the chronology is not clear-cut, Abbot Ralph seems to have gone on sabbatical in 1053, NMProsopographie, 207, note 101. Likewise, although the transmission of the charter’s text is complicated, it is Robert, not the couple’s second son William, who appears in the twelfth-century ‘copie figurée’ of the diploma (RADN, no. 141, ms. B). Léopold Delisle’s edition has Robert, not William, Saint-Sauveur-leVicomte, 21–2, pièces justificatives, no. 19.

the shaping of things to come

105

that is effectively a record of the foundation of William fitz Osbern’s first monastic foundation, Notre-Dame of Lyre, is dated to c.1050 only on the basis of that year being thought to be the date of the abbey’s dedication; the authority on which this statement is based, Charles Guéry’s early twentieth-century history of the abbey, has at its heart the assumption that the marriage of William and Matilda took place in c.1050.52 Furthermore, not only might Matilda’s signum have been a later addition, but the surviving manuscript of the charter is in an early twelfth-century parchment pancarte that contains at least one forgery. Matilda’s appearances, along with those of the couple’s eldest son Robert, among the signa of two apparently original charters for the abbey of Saint-Wandrille that deal with grants made by a man named Baldwin Filleul, one of which carries the date of 1051, are valueless as evidence for the date of the marriage, since the one that contains the dating clause is a thirteenth-century forgery.53 In the charter that is authentic, Matilda and Robert occur among a group of signa added at the foot of the diploma’s parchment by a second scribe. Even if the presence among the first set of signa of Count William of Arques must indicate that the diploma as first drafted dates from before 1052–3, that Matilda and Robert’s names were added by the second scribe means that their signa do not have to date from that time. And the diploma that contains the dating clause of 1051, although in poor condition, is sufficiently well preserved for its script and appearance to show that it was written in the thirteenth century. The textual parallels with the first diploma certainly show that it was its main source, but they do not show that the authentic diploma dates from 1051 since it has no dating clause. The extensive contacts between Normandy and the papacy at this time make it likely that discussions aimed to produce a settlement agreeable to all sides were taking place in the years after the Council of Rheims; there was no need for William, Baldwin, and Matilda to force the pace because there was good will on all sides. The earliest source to mention the marriage, Jumièges’s original version of the Gesta Normannorum Ducum, describes the eventual marriage ceremony and says that it had been legally conducted, an emphasis that is extremely important; as we have seen, and will see again, Jumièges was ready to criticize William and indeed other Norman 52 RADN, no. 120; C. Guéry, Histoire de l’abbaye de Lyre (Evreux, 1917), 14; NMProsopographie, 183–4. 53 RADN, nos. 124, 126; Lot, Saint-Wandrille, recueil des chartes, nos. 30, 31. For a facsimile of the first of these diplomas, P. Chevreux and J.-J. Vernier (eds.), Les Archives de Normandie et de la Seine-Inférieure. Etat général des fonds. Recueil des fac-similés d’écriture du XIe s. au XVIIIe s. accompagné de transcriptions (Rouen, 1911), plate IV. I am grateful to Tessa Webber for her opinion of the script of this diploma.

106

william the conqueror

rulers when he thought it justifiable.54 According to Poitiers, William was given divergent advice about marriage, presumably a reference to the differences of opinion that can be picked up in later sources.55 Although in both cases the two writers are more than a little allusive, their references to debate and eventual legality surely point to discussions and deliberations that were far more complicated than a simple marriage in defiance of a papal ban. In his interpolations into the Gesta Normannorum Ducum, Orderic says that after the marriage William sent an embassy to the papacy to deal with frequent and continuing criticisms, to which the pope (who is not named) replied by giving absolution and agreeing a penance that was to lead to the foundation of the two monasteries in Caen. His rationale for doing so, according to Orderic, was that conflict between Normandy and Flanders would break out unless the marriage was sanctioned, a statement that not only reinforces the argument for there having been earlier negotiations that produced an amicable agreement about the marriage, but also appears to locate it within a broader process of peace-making; the issues that Orderic was referring to were in fact continued criticism and the need to agree a suitable penance.56 As we shall see, everything in fact falls into place if the marriage process is seen as an element in the general peacemaking that was negotiated in the last months of 1052 after the English crisis of 1051–2, in which William and Baldwin were both significant participants, and after radical changes had occurred within the pattern of political relationships across northern France and northern Europe. Abbot John’s aforementioned letter describing his mission to southern Italy is also important, because it is relevant to the general issue of the influence on William’s life of the shared Norman ethnic identity of the inhabitants of the duchy and the freelance conquerors of southern Italy. While John’s selection as ambassador must have been deliberate, no doubt partly determined by his being an Italian born in Ravenna, but who had been mostly resident in Normandy, his humiliating treatment by members of the indigenous population of the south shows just how much the Normans were hated there; even an Italian monk resident in Normandy could not escape being considered a participant in the military enterprises then ongoing. It might appear likely that relations between Normandy and the papacy could have been disrupted by Pope Leo’s decision to attempt a military intervention in the south, culminating in the disastrous GND, ii, 130–1 (despondit iure coniugali). GG, 30–1 (Consiliis itaque de matrimonio discrepatur, ut solent in diuersum suadere ingenia disparium atque sententiae). 56 GND, ii, 146–9; Jégou, L’évêque, juge de paix, 435–43. 54 55

the shaping of things to come

107

defeat of his army by the Normans and their allies at the Battle of Civitate on 18 June 1053.57 As we shall see, however, relations between Normandy and the papacy remained cordial. But after Leo’s death in 1054, the possibility that the issue of William and Matilda’s marriage could be reopened by a less conciliatory pope remained. It is important that William was apparently being treated primarily as the ruler of a principality rather than as a member of an ethnic group. An incident, first recorded in Gilbert Crispin’s early twelfth-century Vita of Herluin, the first abbot of Le Bec, and further expanded in the midtwelfth-century Vita Lanfranci, that might have something to do with the debates about William and Matilda’s marriage – since the Vita Lanfranci indicates that the cause of the quarrel was Lanfranc’s criticism of William’s proposed marriage – suggests that the route to collaboration and trust between the two men was not an entirely smooth one. William, so we are told, having taken against the abbey of Le Bec and Lanfranc in particular, had devastated some of the abbey’s lands, and had ordered that Lanfranc leave Normandy. As he was setting out into exile, Lanfranc encountered William on the road and joked that he would be able to go more quickly if he was given a decent horse. William is said to have lowered his head as Lanfranc approached – perhaps in embarrassment – and then responded to Lanfranc’s joke and his attempt to plead his case with the words: ‘Whoever asks for gifts from an offended judge while his offence remains unpurged?’ The two men were soon exchanging warm embraces. The incident looks to have been somewhat stage-managed – the meeting cannot have been an accident and the monks of Le Bec could surely have given Lanfranc a better mount – and, like so much else, it had passed through the prisms of distorting memory, something that is especially relevant in this case given the two individuals involved. It does nonetheless show that Gilbert Crispin, who was abbot of Westminster from 1085 and, although of a younger generation, someone who would have known both individuals well, believed William to be approachable and affable and aware of his formal responsibilities as a judge. The anecdote also illustrates the course that flowed out of a theatrical display of anger and the violence that accompanied it – in this case presumably the destruction of the livelihoods of a significant number of peasants – towards the restoration of harmony and favour. Given that the subject of Lanfranc’s possible criticism of William’s marriage is entirely absent from the Vita Herluini and that, as will be suggested later, the Vita Lanfranci’s narrative of the marriage

57 For a summary of events, G. A. Loud, The Age of Robert Guiscard (Harlow, 2000), 110–20.

108

william the conqueror

and Lanfranc’s role is a twelfth-century invention, it is probable that its interpolation into the story should be set aside.58 If, however, there is anything in the Vita Lanfranci’s story, it may make this account a product of the processes associated with negotiations around the marriage rather than evidence that William and Matilda had married in defiance of a papal prohibition. Also, according to William of Malmesbury, another major Norman churchman to express similar concerns had been Archbishop Malger of Rouen, but this story may be based on a misunderstanding by Malmesbury of his source.59 What we do have in all this is evidence of a memory that the issues associated with William and Matilda’s marriage took a while to resolve and that the prohibition was taken seriously. It is unhelpful that, as with so much that happened during these years, it is ultimately impossible to assign either a date or a context to the encounter between William and Lanfranc.60 Whenever it occurred, and whatever form it took, however, the incident could well have been a significant moment in the development of their personal and spiritual relationship. THE ENGLISH CRISIS OF 1051–2 AND THE MARRIAGE OF WILLIAM AND MATILDA William of Malmesbury’s statement that he found it impossible to write a convincing narrative of Edward the Confessor’s supposed promise of the English succession to William on the basis of the information available is the product of the irreconcilable values, constructs, and expectations of the main players.61 The unbridgeable chasm that these created is reflected not just in later writers such as Malmesbury, but in the histories written at the time. As has long been recognized, the two principal contemporary Norman historians of William’s life were putting forward a version of events constructed to justify the conquest of England while it was happening, and developed further after it had happened. Although their moral and literary ‘Vita Herluini’, ed. Abulafia and Evans, 197–8 (Armitage Robinson, Gilbert Crispin, 97–8); ‘Vita Lanfranci’, ed. Margaret Gibson, in G. D’Onofrio (ed.), Lanfranco di Pavia e l’Europa del secolo XI nel IX centenario della morte (Rome, 1993), 661–715, at 675–6. 59 GR, i, 494–5. 60 Gibson, Lanfranc, 31, dated to c.1047. Frank Barlow, The English Church, 1066–1154 (London, 1979), 56, more plausibly argues for 1053, but the incident could equally well date to the earlier 1050s. Most recently, Cowdrey, Lanfranc, 32–4. 61 GR, i, 354–7. For comments similar to mine, Elizabeth Carson Pastan and Stephen D. White, with Kate Gilbert, The Bayeux Tapestry and its Contents: A Reassessment (Woodbridge, 2014), 234–5. 58

the shaping of things to come

109

objectives were elsewhere often more complex, here their straightforward objective was to defend a conquest that many regarded as morally indefensible either in law or in its execution, or in both. The verbal parallels between the texts of Jumièges and Poitiers suggest either that the former was the latter’s starting point or that the two of them worked from a now lost justification of William’s rights written, in all likelihood, in or around 1066, with Jumièges amending what he had written in the 1050s to conform to this official version.62 On balance, it looks as if Jumièges wrote his passage at a single sitting after 1066. The insertion of a reference to Archbishop Robert of Canterbury’s visit to Normandy taking place ‘in the past’ (olim) suggests this, as does the way in which a single section of the Gesta is devoted to the succession, with Harold’s oath sworn when he visited Normandy following on immediately from the account of the visit. The early sources for Edward’s reign written in England were every bit as much political as texts as Poitiers’s Gesta, the political nature of the versions of the Anglo-Saxon Chronicle being nowadays well understood. Their differing perspectives have become all the more relevant to understanding the English crisis of 1051–2 with the revival of the view that ‘D’ was written in the entourage of Ealdred, bishop of Worcester from 1046 until 1062 and archbishop of York from 1060 until his death in 1069.63 Even in a recent subtly nuanced form, this argument makes ‘D’ the product of a circle close to Archbishop Ealdred.64 And like Ealdred, Queen Edith, under whose patronage the Vita Ædwardi Regis was produced, in all probability in c.1067–8, also had an obvious reason to go public to defend both herself and her family: namely, the evident pressure to justify the conduct of an elite and of specific individuals who some must surely have thought had led the English kingdom to catastrophe.65 All had very good reasons to put a particular spin on events. Most of them were also composed before the outcome of the events of 1066 had fully clarified and before, so Compare illum statuens heredem, GND, ii, 158, and eum rata donatione heredem statuere decreuit, GG, 20. 63 Patrick Wormald, How Do We Know So Much about Anglo-Saxon Deerhurst?, Deerhurst Lecture, 1991 (Friends of Deerhurst Church, 1993), 9–17; reprinted in idem, The Times of Bede: Studies in Early English Christian Society and its Historian, ed. Stephen Baxter (Oxford, 2006), 235–40; The Anglo-Saxon Chronicle, a Collaborative Edition, volume 6, ms. D: A SemiDiplomatic Edition with Introduction and Indices, ed. G. P. Cubbin (Cambridge, 1996), liii–lvi, lxxix. 64 Pauline Stafford, ‘Archbishop Ealdred and the D Chronicle’, in Crouch and Thompson (eds.), Normandy and its Neighbours, 135–56, and esp. pp. 142, 145–9, 151–6. Note in particular: ‘On current evidence, some association with Ealdred looks very likely’ (p. 156). 65 On the Vita’s date I follow Stafford, Queen Emma and Queen Edith, 40–8, and Keynes and Love, ‘Earl Godwin’s Ship’, 199, in arguing for composition very soon after 1066; cf. Frank Barlow’s arguments for 1065–7, VEdR, xxix–xxxiii. 62

110

william the conqueror

some might have thought, William’s victory was certain. Hence, perhaps in c.1070, a monk who had migrated to England from the Empire, writing on behalf of a monastery ruled by a French abbot who had sent troops to fight for Harold at Hastings, referred to William’s claim as representing ‘one line of reasoning’ and that ‘many entertained the rumour that King Edward, dear to memory, had named the duke his heir’.66 The basic story that Jumièges and Poitiers tell in their typically different ways is that Edward the Confessor sent Robert of Jumièges, archbishop of Canterbury from 1051 to 1055, to Normandy to inform William that he was designating him as his heir to the English kingdom.67 This visit is generally dated to the spring of 1051 when Robert was on his way to Rome to collect his pallium from Pope Leo; since he was back in England on 27 June, the visit has quite reasonably been thought to have taken place before then, an argument most persuasively set out by David Douglas in an important article.68 To the bare facts set out by William of Jumièges, Poitiers added that Robert brought with him a son and a nepos (in this case probably meaning grandson) of Earl Godwine of Wessex as hostages and that the designation was made with the agreement of the English nobles, whom, in another passage, Poitiers names as Earls Godwine, Leofric of Mercia, Siward of Northumbria, and Archbishop Stigand of Canterbury (1052– 70), respectively the holders of the three largest English earldoms and the man who had displaced Archbishop Robert.69 Along with Edward and his wife Edith, these four were the major players who brought England close to civil war in 1051 and who, after Godwine and his family’s period of exile from October of that year onwards, were the peacemakers whose determination to avoid civil war was of decisive importance in the settlement of 1052. Many turbulent eddies must have swirled beneath the apparent calm of the English 1040s, a tranquillity epitomized by Edward’s collaboration with Godwine, by marriage to his daughter Edith, and by the rapid advancement of Godwine’s two eldest sons Svein and Harold to 66 Herman the Archdeacon and Goscelin of Saint-Bertin, Miracles of St Edmund, ed. and trans. Tom Licence, with the assistance of Linda Lockyer, OMT (Oxford, 2014), 62–3. For the author and the date of writing, see pp. xxxv–xlvi, lvi. 67 GND, ii, 158–9; GG, 20–1. 68 D. C. Douglas, ‘Edward the Confessor, Duke William of Normandy, and the English Succession’, EHR, lxviii (1953), 526–45, at 535–8. Archbishop Robert’s journey is mentioned in all three versions of the Anglo-Saxon Chronicle, but only ‘E’ mentions the date of his return, ASC, ‘C’, ‘D’, ‘E’, 1051. On these events, see most recently, Stephen Baxter, ‘Edward the Confessor and the Succession Question’, in Mortimer (ed.), Edward the Confessor, 77–118, at 86–95. For a reassessment of Archbishop Robert’s career, Tom Licence, ‘Robert of Jumièges, Archbishop in Exile (1052–5)’, ASE, xlii (2013), 311–29. 69 GG, 120–1.

the shaping of things to come

111

earldoms.70 Edward nonetheless certainly continued to harbour grievances against Godwine for his role in the death of his brother Alfred, something that is indicated not just by the Vita Ædwardi’s explicit references to such feelings being the means by which Archbishop Robert supposedly turned Edward’s mind against Godwine, but also by the way in which William of Malmesbury explains Godwine’s sudden physical collapse while feasting with the king as instant divine punishment for a mendacious avowal of innocence.71 Whether the childlessness of Edward and Edith’s marriage was a consequence of Edward’s commitment to chastity, as the Vita may imply, of infertility or sexual incompatibility, or, as William of Malmesbury suggests, of the queen’s adultery, it must surely have been an important source of mounting tension; returning to Jumièges for a moment, it is notable that the absence of children is the sole reason he gives for Edward sending Archbishop Robert to make the promise to William.72 For those who had, over the previous half-century, repeatedly experienced the calamities associated with uncertain successions and invasion from abroad, the prospect of more of the same cannot have been appealing. The solution was not, however, an easy one to find. According to largely uncontested eleventh-century values operating prior to the Council of Rheims, Queen Edith was dispensable and, if Edward did want to father an heir, he may well have decided that she had to be put aside, just as William’s father had ended his short-lived marriage to Estrith. However, if Edward’s intention was to divorce Earl Godwine’s daughter, he was seeking to overturn the political status quo of a decade and a balance of power that went back far into the reign of Cnut. And, if Archbishop Robert did visit William in the spring of 1051 and promise him the English succession, it was an act tantamount to taking the future of the English kingdom out of the hands of the English elite, who had by custom had a major role in deciding it, an act that was about as confrontational to Godwine and his family as was possible. This suggests either that Edward was playing for very high stakes indeed and taking huge risks or that we need to re-date Archbishop Robert’s visit to Normandy and, with it, amend the entire narrative.

70 On the Godwine family, see most recently Frank Barlow, The Godwins: The Rise and Fall of a Noble Dynasty (London, 2002); Emma Mason, The House of Godwine: The History of a Dynasty (London and New York, 2004). For their lands, Robin Fleming, Kings and Lords in Conquest England (Cambridge, 1991), 60–103; Ann Williams, ‘Land and Power in the Eleventh Century: The Estates of Harold Godwineson’, ANS, iii (1981), 171–87. 71 VEdR, 30–1; GR, i, 354–5. 72 GND, ii, 158–9 (prolis carens). For recent discussions of the marriage, Barlow, Edward the Confessor, 81–5; Stafford, Queen Emma and Queen Edith, 260–1, 264–5; idem, ‘Edith, Edward’s Wife and Queen’, in Mortimer (ed.), Edward the Confessor, 119–38, at 133–7.

112

william the conqueror

Edward certainly set out from the later 1040s to promote members of his inner circle and to contact old friends. Thus, in 1050–1 he pushed through Robert’s appointment to the archbishopric of Canterbury, apparently against the claims of a kinsman of Earl Godwine whose candidature was supported by the Canterbury monks.73 He also created an earldom in the East Midlands for his nephew Ralph, a son of his sister Godgifu and Drogo, count of the Vexin, who had accompanied him to England in 1041.74 This promotion too was close to the heartlands of Godwine’s power. The affray at Dover in 1051 involving Edward’s former brother-in-law, Count Eustace of Boulogne, is identified in both the ‘D’ and ‘E’ versions of the Anglo-Saxon Chronicle as the cause of the outbreak of conflict between Edward and Godwine. However, Eustace’s actual visit, although unexplained by them, might not have had any motives more profound than being a journey to confer with a kinsman in difficulty and to exchange opinions on the problems arising for Eustace out of the commercial and maritime implications of Edward’s support for Count Baldwin of Flanders in his war with the emperor Henry III.75 In this fraught situation it is surely a certainty that Edward would have contacted William. Their shared past in Normandy and the continuing dynamism of cross-Channel contact, exemplified by the recent migration of William fitz Osbern’s brother Osbern to become one of Edward’s chaplains, made it inevitable. Also, the one element that is common to William of Poitiers and the later sources written in England is that hostages from Godwine’s family were sent to William. However this is read into the various narratives of possible events, the choice of William as their recipient surely indicates a continuing special level of trust between Edward and the duke that was not only widely recognized, but very specifically so by the main participants in the English crisis of 1051–2.76 The ‘D’ Chronicle famously states that William visited Edward in England in 1051, bringing with him a large entourage, that the king received him, and then sent him on his way. The statement’s location at the end of the annal presumably indicates that the visit occurred late in the year after Godwine and his family had fled. November or December 1051 are Vita Ædwardi, 30–1. Ann Williams, ‘The King’s Nephew: The Family and Career of Ralph, Earl of Hereford’, in Harper-Bill, Holdsworth, and Nelson (eds.), Studies in Medieval History presented to R. Allen Brown, 327–43, at 330–9. 75 ASC, ‘D’, ‘E’, 1051. For the range of issues that might have been discussed, Sally Harvey, ‘Eustace II of Boulogne, the Crises of 1051–2 and the English Coinage’, in David Roffe (ed.), The English and their Legacy, 900–1200: Essays in Honour of Ann Williams (Woodbridge, 2012), 149–57, at 150–5. 76 On hostages in general, Adam J. Kosto, Hostages in the Middle Ages (Oxford, 2012), 24–52, and on this specific case, pp. 24–5, 78–9. 73 74

the shaping of things to come

113

therefore plausible dates. The visit is not mentioned by either Jumièges or Poitiers and, in the early twelfth-century sources it appears only in John of Worcester’s chronicle, which, even if it did not use the ‘D’ Chronicle in the form in which it has been transmitted to us, did base his account on something similar.77 John’s treatment is identical to the manuscript of the ‘D’ Chronicle, except that he clarified that the Chronicle’s reference to Willelm eorl was indeed to William, count of the Normans (Nortmannicus comes Willelmus). Interestingly, the visit is also mentioned in Wace’s Roman de Rou, the work of an author who had read almost everything that it was possible to read and who did not think it incompatible with William’s claim.78 The palaeographical complexity of the ‘D’ Chronicle, written as it was by several scribes, and probably completed in the early 1080s, has over the years made it difficult for everyone to be convinced that William’s visit took place.79 But the view, even in a qualified form, that it was written in the entourage of the future Archbishop Ealdred does mean that its author (or authors) was certain to have been well informed. It is also clear that the concluding section of the 1051 annal was written by a different, and in all likelihood, earlier, scribe than the one who wrote the inserted quire covering the years from 1016 until the part of 1051 that precedes the account of William’s visit.80 At the very least, therefore, there is no decisively compelling reason to reject the statement in the ‘D’ Chronicle that William visited England late in 1051. In addition, that the visit was not mentioned by Poitiers may well argue in favour of it having taken place, since it would have happened at a time when the consent of the chief nobles of England, which he wanted to emphasize, could not possibly have been obtained; the ritual submissions and oaths that might have been expected for a full-scale promise of the succession could not have been performed, because Godwine and some of his family were in exile in Flanders. To have included the visit might also have indicated that there were elements in the background to William and Matilda’s marriage that were more complicated than Poitiers’s purpose

ASC, ‘D’, 1051; JW, ii, 562–3. Roman de Rou, 216–17, pt. 3, lines 5,401–10. 79 Douglas’s arguments against accepting the account of the ‘D’ Chronicle (‘Edward the Confessor, Duke William of Normandy, and the English Succession’, 527–34) were based on a palaeographical assessment of ‘D’, which was out of date soon after his article was published. See N. R. Ker, Catalogue of Manuscripts containing Anglo-Saxon (Oxford, 1957), 254. Douglas ignored Ker’s assessment in WC, 169, and repeated his earlier opinion. 80 See BL, Cott. Tiberius Biv, fo. 74r; Ker, Catalogue, 254; The Anglo-Saxon Chronicle, ed. Cubbin, 71; Timofey V. Guimon, ‘The Writing of Annals in Eleventh-Century England: Palaeography and Textual History’, in Alexander R. Rumble (ed.), Writing and Texts in AngloSaxon England (Cambridge, 2006), 137–45, at 141. 77 78

114

william the conqueror

allowed.81 It is also clear that William’s visit must have been a relatively short one. Moreover, since he would by this point have been deeply absorbed in the problems caused by Geoffrey Martel’s takeover of Maine and his occupation of Domfront and Alençon, places that William was soon going to besiege, he would have had to tell Edward that his capacity to help him was very limited. Paradoxically, therefore, even if Edward did state his personal wish that William succeed him, he would also have learnt that he could expect no significant help from Normandy. So understood, the visit becomes a part of the process that led to the accommodation between Edward and Godwine. Of the early twelfth-century writers who do write about the events of 1051–2, Orderic, who drew his information from Jumièges and Poitiers, is the only one to mention a mission by Archbishop Robert and to associate it with a promise of the succession.82 His brief account in no way differs from theirs. The monk Eadmer, even though he was writing at Canterbury in the late eleventh and early twelfth centuries, and must have had access to memories going back to Archbishop Robert’s time, in his Historia novorum (History of Recent Times) mentions only the dispatch to Normandy of hostages from Godwine’s family; he names them as Godwine’s son Wulfnoth and Hakon, the son of Godwine’s eldest son Svein. However, in contrast to Poitiers, Eadmer says that they were sent to William after the conflict between Edward and Godwine was over. Although he does mention that William had been promised the succession, he places it in Normandy before Edward became king.83 William of Malmesbury omits any reference to Archbishop Robert’s mission. This may be particularly significant, since he had read Jumièges and Poitiers and was presumably either registering disbelief by not following them or, more likely, indicating by his silence that it did not have the significance that they attributed to it. He does mention that Wulfnoth was sent as a hostage, but without providing either a reason or a context. In addition, he places the promise of the succession to William after the death in 1057 of the Confessor’s nephew Edward the Ætheling, the son of the king’s older half-brother Edmund Ironside, whom Malmesbury described as Edward’s first choice.84 Something similar is presumably behind the statement in the mid-twelfthcentury ‘Laws of Edward the Confessor’ (the Leges Edwardi Confessoris),

81 For this argument, T. J. Oleson, ‘Edward the Confessor’s Promise of the Throne to Duke William of Normandy’, EHR, lxxii (1957), 221–8, at 221–2. 82 OV, ii, 134–7. 83 HN, 6. 84 GR, i, 362–3, 416–17.

the shaping of things to come

115

namely, that Edward wished to have Edgar the Ætheling as his heir, but, realizing that it would not work, he turned to William.85 These apparently contradictory narratives can be reconciled into a coherent one if account is taken of the pivotal role of Count Baldwin of Flanders, who, in late 1051 gave refuge to Godwine, his wife Gytha, and their sons Svein, Tostig, and Gyrth, thereby providing them with a base at Bruges on the narrowest sea-crossing back to England. From Bruges, they undertook an obviously preplanned two-pronged naval campaign in collaboration with two of Godwine’s other sons, Harold and Leofwine, who had gone to Ireland to raise forces there.86 With this in mind, if both the date of 1050–1 for the marriage of William and Matilda is accepted and a promise of the succession had been conveyed to William by Archbishop Robert in the spring of 1051, then Baldwin’s behaviour looks very perplexing. If a promise of the succession had been made, what was he doing providing assistance to the family most able to prevent his sonin-law, or possibly, future son-in-law, William, from taking a great prize, the family whose aim must still have been for Edward’s wife Edith to bear him a son? Just as the proposed marriage between William and Matilda had continued a pattern of good relations between Normandy and Flanders, so Baldwin’s support for Godwine and his family maintained a pattern of providing assistance for exiles from England who had links to the kingdom of Denmark. Not only had Baldwin taken in Cnut’s queen, Emma, and supported their son Harthacnut, he had given a refuge to Godwine’s eldest son Svein on two occasions, in 1046–7 and 1049–50, acts of considerable generosity since Svein had on the first occasion abducted an abbess and kept her as a concubine and, on the second, been directly responsible for the murder of a cousin. So profound was Baldwin’s commitment to Godwine’s family that two versions of the Anglo-Saxon Chronicle even say that Svein younger brother Harold was unhappy with the support that Baldwin provided to his unruly sibling.87 In addition to all this, Baldwin had agreed to the marriage of his halfsister Judith to Godwine’s third son Tostig. As a sister, Judith was not as much of an investment in a political alliance as his daughter Matilda was to be, but this was still a notable gesture since Tostig was at that point a landless younger son without clear prospects. The Vita Ædwardi says this marriage took place during the period of exile, whereas the ‘D’ Chronicle, 85 Leges Edwardi Confessoris, ed. Bruce R. O’Brien, in idem, God’s Peace and King’s Peace: The Laws of Edward the Confessor (Philadelphia, PA, 1999), 135–203, at 196–7. 86 ASC, ‘C’, ‘D’, ‘E’, 1051; VEdR, 40–1. 87 ASC, ‘C’, ‘E’, 1049. In general, Philip Grierson, ‘The Relations between England and Flanders before the Norman Conquest’, TRHS, 4th ser., xxiii (1941), 71–112, at 96–100.

116

william the conqueror

followed by John of Worcester, states that Judith accompanied Tostig and the family to Flanders, thereby indicating that both believed the marriage to have taken place before the banishment.88 Since this section of the ‘D’ Chronicle, although included within the same annal as the account of William’s visit to England, is on the quire rewritten in all probability in the 1080s, it may be a later insertion into an earlier annal that was founded in a reasonable deduction made on the basis of the depth of Baldwin’s support for the Godwines. On the other hand, the discrepancy can be reconciled if the marriage had been arranged before 1051 and was celebrated during the exile.89 Whatever the case, all this most certainly highlights an unrecognized strand in the rift between Edward and the Godwine family, since Edward had ordered a naval blockade of the Flemish coast in 1049 in support of the emperor Henry III and against Count Baldwin.90 That Baldwin should have been pursuing alliances with both William and Godwine is entirely unsurprising. He needed powerful support and these two offered the best route to stabilize the North Sea and Channel seaboard. In these circumstances, Edward’s opening discussions about the English succession with William must have thrown a considerable spanner into the works. The exceptional fluidity of the political situation does, however, mean that all this frenetic activity makes sense if the main actors are seen as making and maintaining contacts with friends, clarifying relationships in times of difficulty, and seeking support when the crisis broke. However, once Godwine and his sons had landed triumphantly in England in September 1052, a negotiated settlement that saved everyone’s faces became a necessity. In England, Godwine’s reinstatement and the restoration of Edith as queen could not be presented to the world as a humiliation for Edward, even if this is what they actually were. In terms of relationships across the narrow seas, Baldwin needed to remain on good terms with both Godwine and William. In terms of both Edward’s political standing and Baldwin’s need for allies, William could not be left out. For all their differences, the twelfth-century sources do suggest that events bearing some resemblance to those set out by Jumièges and Poitiers did take place. However, they most definitely also show that the events cannot possibly have had the clear-cut character which Poitiers in particular assigned to them. The crucial point is that the two elements present in most accounts are that some sort of promise was made to William and hostages from Godwine’s family were sent to him. The Anglo-Saxon VEdR, 38–9; ASC, ‘D’, 1051; JW, ii, 560–1. On the marriage, see now Mary Dockray-Miller, The Books and the Life of Judith of Flanders (Farnham and Burlington, VT, 2015), 7–9. For slightly different reasons she also places the marriage before 1051. 90 ASC, ‘C’, ‘D’, 1049; JW, ii, 548–9; Barlow, Edward the Confessor, 97–9. 88 89

the shaping of things to come

117

Chronicle mentions proposals for exchanges of hostages as part of the attempts to reach a settlement between Edward and Godwine; most tellingly, the ‘E’ Chronicle makes an exchange an aspect of the final 1052 peace settlement.91 John of Worcester largely repeats this, even if he amended whatever version of the Anglo-Saxon Chronicle he was using to make an exchange of hostages an aspect of the failed attempt at agreement in the autumn of 1051; it is possible that he was influenced by knowledge of Poitiers’s version.92 If, however, the dispatch of the hostages to Normandy took place in the autumn of 1052, it would be seen by other parties in England, most notably the two other major political players in the kingdom, Earls Leofric and Siward, as a guarantee of good behaviour on the part of the Godwines. It would also be considered logical because of Edward’s special confidence in William, of which all parties on both sides of the Channel would have been aware.93 In the circumstances of late 1052, a reiteration of whatever offer had been made to William when he had visited in late 1051 would have seemed respectful of Edward’s wishes. If this restatement of Edward’s offer is what happened, it explains how Poitiers could write that William’s succession had the agreement of Earls Godwine, Leofric, and Siward, and of Stigand, archbishop of Canterbury. This interpretation also fits with Eadmer’s statement that Edward had promised the succession to William before 1041, because it makes the events of 1051–2 into part of a long-term process with many twists and turns, in which a restatement of Edward’s personal preferences would have seemed unexceptional. It also accommodates William of Malmesbury’s apparently strange statement that Edward the Exile, son of Edmund Ironside, was the Confessor’s first choice on grounds of kinship, since it is possible he meant that any offer of the succession to William could only be valid after the death in 1057 of this representative of the Wessex line of kings; Malmesbury also observed that both Edward the Exile and his son Edgar were morally disqualified from kingship.94 A further result of the consideration given to William in any settlement would also have meant that Baldwin’s commitments to Godwine and to William were no longer irreconcilable, and that the wedding of his daughter to William would consolidate appropriately all that had been agreed and would bring to fruition marriage plans that had long been in the making. Baldwin could henceforth concentrate on the war with the ASC, ‘E’, 1052. JW, ii, 560–1. 93 Cf. Barlow, Edward the Confessor, 305–6; K. E. Cutler, ‘The Godwinist Hostages: The Case for 1051’, Annuale Medievale, xii (1972), 70–7. 94 GR, i, 416–17; Emily A. Winkler, ‘1074 in the Twelfth Century’, ANS, xxxvi (2014), 241–58, at 248–52. 91 92

118

william the conqueror

emperor, with several distractions having been removed. Furthermore, if the outcome of the negotiations in England was communicated to William by Archbishop Robert as he went into exile after Godwine and his family’s return to England, the embassy mentioned by Poitiers dates to late in the year 1052, rather than the spring of 1051. With Archbishop Robert expecting eventually to be reinstated and being consistently treated as an archbishop until his death in 1055, this would have given considerable status in 1052 to the outcome he communicated to William.95 In terms of the actual narrative, as opposed to the version of events conveyed by Jumièges and Poitiers, it also has the consequence of making William’s visit in late 1051 a crucial event; its relative informality and the ambiguities associated with it do, however, mean that few would have seen the future of the English kingdom as having been settled. In the short term, a relative calm had been restored that would seem to have satisfied most interests. Truces of this kind have often performed such a function. They have, however, often failed to maintain peace in the long term, something that becomes all the more likely when the interests of the main parties in this particular situation are taken into account. Whatever offer was actually made to William, he is likely not only to have taken it seriously, but to have seen it as irrevocable. Thus, when later confronted by the dilemma posed by his own designation of his rebellious son Robert as his heir in Normandy, he is said by Orderic to have announced that such a promise could not be abandoned.96 The post obitum gift was also so frequent an occurrence in eleventh-century Norman society that the well-nigh universal perception there could well have been that Edward’s grant ought to be fulfilled. Furthermore, William had himself been designated during his father’s lifetime, as had many of his predecessors as rulers of Normandy; the oaths taken to William in 1034 were said categorically by Jumièges to be ones that were not to be broken.97 The importance of William’s kinship to Edward, a theme powerfully developed in English writs and charters after 1066, also looks to have been consciously emphasized in Normandy from soon after 1051, with the significance of kinship through the female, as well as the male, line being regularly indicated. Soon after 1052 the author of the Inventio et Miraculi sancti Uulfranni wrote that Edward had been formally made king of the English in childhood; we have here perhaps a reflection of the long-term process later picked up by Licence, ‘Robert of Jumièges’, 326–9. Concessus honor nequit abstrahi, OV, iv, 92–3. 97 GND, ii, 80–1 (pangentes illi fidelitatem non uiolandis sacramentis). See above all J. S. Beckermann, ‘Succession in Normandy, 1087, and in England, 1066: The Role of Testamentary Custom’, Speculum, xlvii (1972), 258–60; Garnett, ‘ “Ducal” Succession’, passim. On post obitum gifts, Tabuteau, Transfers of Property, 24–7. 95 96

the shaping of things to come

119

Eadmer.98 That William believed he might one day succeed Edward the Confessor in England must have been publicly known and an aspect from 1051 of western and northern European politics. As we have seen, the pursuit of opportunities of this kind was a deeply entrenched part not only of the political culture of the French kingdom, but of a much wider world. It is, however, extremely unlikely that anyone in England saw matters as having been decisively settled in William’s favour, a point that Poitiers acknowledged.99 With Edward’s offer not ritually substantiated, and with the king’s deathbed wishes always certain to be a significant – if not, indeed, the central – factor in determining the eventual succession, nothing was at all clear-cut. Also, with Edward apparently in good health, and perhaps likely to live for a good number of years, none of the interested parties in England, Scandinavia, and northern France had clear objectives at which to aim. The situation started to change when Godwine’s eldest son Svein died in 1052 on the return journey from a pilgrimage to Jerusalem, opening the way for Harold to succeed to the earldom of Wessex after their father’s death in 1053. Tostig was subsequently appointed to the earldom of Northumbria in 1055, with Earl Siward’s infant son Waltheof pushed to one side. In 1054 a mission led by Bishop Ealdred of Worcester was dispatched to the court of Emperor Henry III. Its purpose was apparently to secure the recall from Hungary of the man with the closest blood-tie to Edward, namely, his nephew Edward the Ætheling, son of King Edmund Ironside. The first known contact with Edward the Ætheling for over thirty years, it was a strong statement by England’s elite of where its prime loyalties lay.100 Through all of this, Wulfnoth, Godwine’s youngest son and one of the hostages sent to Normandy, was kept under house arrest, to be released only by William on his deathbed in 1087. Wulfnoth eventually died as an old man at Salisbury, the victim of the ambitions of his family’s enemies and, we might reasonably suggest, of his so-called friends as well. Orderic considered him to have been a religious man.101 98 Note the statements in Inventio et Miraculi sancti Uulfranni, 29–30, to the effect that Edward the Confessor was Duke Richard II’s nephew (Eguuardus gloriosi principis Richardi nepos). See further van Houts, ‘Historiography and Hagiography’, 247–9, 251; Bauduin, ‘Désigner les parents’, 82–4. 99 Ann Williams, ‘Some Notes and Considerations on Problems Connected with the English Royal Succession, 860–1066’, ANS, i (1979), 144–67, at 166. For the issues in a broader context, John Gillingham, ‘At the Deathbeds of the Kings of England, 1066–1216’, in Brigitte Kasten (ed.), Herrscher- und Fürstentestamente im Westeuropäischen Mittelalter (Cologne, Weimar, and Vienna, 2008), 509–30, at 510–11, 525. 100 Barlow, Edward the Confessor, 214–17. See further Vanessa King, ‘Ealdred, Archbishop of York: The Worcester Years’, ANS, xviii (1996), 123–37, at 127–8; Veronica Ortenberg West-Harling, ‘Un prince évêque anglo-saxon au XIe siècle: l’archevêque d’York Ealdred’, in Alban Gautier and Sébastien Rossignol (eds.), De la mer du Nord à la mer Baltique: Identités, contacts et communications au Moyen Age (Lille, 2012), 143–57, at 148–9. 101 GR, i, 362–3; OV, ii, 178–9.

120

william the conqueror

With considerable justification England’s leaders must have thought that normality was being restored; the failure to reinstate the legitimately consecrated Robert to the archbishopric of Canterbury was perhaps the only cloud on the horizon. Since Stigand was never confirmed in office by a legitimately consecrated pope, he was a blot that always opened the English Church to criticism. In its coded way the Vita Ædwardi drew attention to this by making reference, in the midst of a eulogy of Godwine and a denigration of Archbishop Robert, to a dispute between them about Canterbury lands in which the earl was said to be in the wrong. It later singled out Stigand as mocking Edward’s deathbed vision prophesying invasion and ruin and blaming all on the moral failures of senior members of the English Church.102 In the years after 1052, however, it may well have seemed that the situation could be rectified. One consequence was that the extraordinarily active Ealdred, who became archbishop of York in 1060, was pushed ever more into the forefront of affairs. Jumièges and Poitiers present somewhat differently the marriage of William and Matilda. According to Jumièges, Baldwin brought his daughter to Eu on the Norman frontier, where they were met by William and his military entourage and where the wedding took place. William then escorted his bride with great ceremony and honour to Rouen.103 Poitiers says that Baldwin brought Matilda to Ponthieu, where they were met by William and his mother and stepfather. He makes no mention of an actual wedding, but does say that Matilda’s entry into Rouen as William’s wife was greeted with extravagant celebrations.104 The information about the participation of William’s mother (Herleva) and stepfather (Herluin) in the meeting with their son’s bride-to-be, cryptically supplied without mentioning either of them by name, again underlines the fundamental and enduring importance of William’s maternal kindred to his life. Since these were still times when clergy did not necessarily play a significant part in a marriage ceremony, there is nothing surprising in what was done. Although Poitiers omits any mention of Eu, it is likely that William and Matilda were married there in the presence of their families. Matilda was probably around five years younger than her new husband.105 It is certain from a later reassessment of the findings, when the bones in her Vita Ædwardi, 32–3, 118–21. GND, ii, 128–31. 104 GG, 32–3. 105 Matilda’s date of birth is unknown. Since Jumièges suggests that Baldwin V and his wife had consummated their marriage before his revolt against his father, which came to an end in 1030 (Cuius filius mox, ut nobilis puelle amplexibus cepit frui . . .), it is possible that she was born in the late 1020s, GND, ii, 52–3. The sequence of Baldwin and Adela’s four children is, however, unclear, and Matilda might have been born after 1030. The only certainty is 102 103

the shaping of things to come

121

tomb in the abbey of La Trinité of Caen were examined in 1958, that she was a woman of average height for her times, perhaps around five feet tall – and with a large pelvis. For her to have borne at least nine children in around fifteen years makes it impossible for it to have been otherwise.106 Given calculations to the effect that roughly a third of the children born to English kings from 1150 to 1500 died before the end of their first year, and fewer than half lived into their twenties, it is certain that Matilda was physically very strong, like her husband.107 The continuation of the amicable relations between the ducal and papal courts makes it as good as certain that a penitential solution to any remaining doubts about the legitimacy of the marriage must have been under consideration.108 This fits exactly with Orderic’s statement that William sent an embassy to the papal court and also with the politics of the continued alliance with Count Baldwin.109 In terms of the future, the slow process that had led to a marriage in which the papacy initially acquiesced, and of which it came to approve, was also the basis for the long-term close relations that were so important in 1066. It also means that there was plenty of time for William to make the visit to Bruges that is mentioned in the so-called Chronicle of Tours, which reached its final form in the early thirteenth century. Containing as it does the story of William forcing his way into Matilda’s chamber and inflicting a physical beating of such severity that she decided to marry him, having previously refused to countenance marriage to a bastard, it is most likely yet another one of those legends created after his death, a display of masculinity and authority that was thought acceptable according to contemporary norms.110 On the other hand, the story may at least be evidence that the negotiations for the marriage involved that she was of childbearing age by 1052–3 and continued to have children into the late 1060s. 106 Sir John Dewhurst, ‘A Historical Obstetric Enigma: How Tall was Matilda?’, Journal of Obstetrics and Gynaecology, i (1981), 271–2. I am grateful to Pierre Bouet for making accessible to me the full text of the report of Jean Dastague’s analysis of the bones in Matilda’s tomb conducted in 1959. 107 Nicholas Orme, Medieval Children (New Haven, CT, and London, 2011), 113. 108 For this suggestion, Corbet, Burchard de Worms, 301–5, citing parallel examples. 109 GND, ii, 146–7. 110 ‘Ex Chronico Turonensi auctore Turon. Ecclesiae S. Martini Canonico’, RHF, xi, 346–9, at 348. For a survey of the literature on the Chronicle and comment on its thirteenthcentury sections, Richard Kay, The Council of Bourges, 1225: A Documentary History (Aldershot and Burlington, VT, 2002), 294–7. The treatment of William’s birth and the story’s insertion into the annal for 1056 suggest to me that a story of a visit was greatly expanded later to fit the portrait of William that developed in the twelfth century. For canon law’s relative permissiveness on domestic violence, Hannah Skoda, Medieval Violence: Physical Brutality in Northern France 1270–1330 (Oxford, 2013), 197–210.

122

william the conqueror

William visiting Flanders. It is certainly a stark and necessary reminder of the very different standards of conduct that existed in the medieval period. ALENÇON AND DOMFRONT Over the past one hundred years, William’s capture of the towns of Alençon and Domfront from Geoffrey Martel has been variously dated to 1049 or 1051–2; the accounts of Jumièges and Poitiers, and the interesting additions in Orderic’s interpolations, supply no chronology either of the duration of the sieges or of the dates of William’s victory.111 The one statement in the sources which is apparently useful for dating purposes, Poitiers’s comment that Count William of Arques slipped away from the siege of Domfront without permission, and that this was a prelude to his defiance of Duke William, is not without difficulties.112 Above all, it does not have to be read as indicating that the conflict with Count William, which took place in 1053, broke out either at the siege or immediately after it, since the breakdown in relations between William and his uncle can be shown to have been a gradual one. One element in the argument for 1049 is the suggestion that the Mouliherne campaign was a sequel to William’s victory at Alençon and Domfront; with Geoffrey Martel repulsed on the frontier of Normandy, the way was open for the coalition that King Henry I had organized, probably in the period after the meeting at Senlis in 1048.113 It is more probable, however, that the Mouliherne campaign, which actually achieved little, is a classic example of a large army’s capacity to make apparently impressive inroads into hostile territory, but then find itself unable to sustain the campaign through lack of supplies caused by the shadowing tactics of its opponents; the advance into Normandy by Henry I and Geoffrey Martel, culminating in their defeat by William at Varaville in 1057, is a good example of this phenomenon. It is therefore more likely that the warfare around Alençon and Domfront happened after Geoffrey took control of Maine following the death of Count Hugh IV on 26 March 1051, and drove his widow Bertha and their children, including the nominal Count Herbert II, into exile. By this reckoning, William captured Alençon and Domfront in 1052; Poitiers’s references to William being undeterred by the

GND, ii, 122–5; GG, 22–9. quasi desertoris furtiuo more discessit, nequaquam petita missione, GG, 34–5. 113 For these arguments, Guillot, Anjou, i, 71–2, notes 314, 320; Louise, La seigneurie de Bellême, i, 354–8; Bauduin, La première Normandie, 309–10. 111 112

the shaping of things to come

123

harshness of winter arguably confirm that the blockade of Domfront began in the winter of 1051–2, in all probability after his visit to England.114 Warfare within the Bellême family may well have been an additional factor behind Count Geoffrey’s intervention at Domfront and Alençon.115 It may even be that his seizure of Domfront and his collaboration with the citizens of Alençon were undertaken to protect his dominance of Maine against the Bellême family’s well-established capacity to disrupt conditions in the north of the county, and that his actions were not aimed directly against William at all. Such a challenge was, however, bound to provoke a reaction from William, since Alençon was a town that his ancestors had controlled, and of which he had been deprived, while Domfront, although outside Norman control, was strategically crucial to the maintenance of any semblance of Norman power in this notoriously turbulent region. William established a long-term siege of Domfront by building four siege-castles around the rocky promontory on which the town and castle are sited. Poitiers’s elaborate account mentions deeds of valour on William’s part and an ambush instigated by an unnamed Norman traitor, from which William escaped by turning to face his numerically superior attackers and putting them to flight by himself killing the man leading the attack. Again, therefore, Poitiers is presenting William to his readers as a general who led by example and who tackled situations in the most direct way possible; the frequent references to him overcoming problems presented by the landscape are another facet of his skills as a soldier.116 Poitiers also tells how William filled up the tedious periods of the blockade with hunting. There are close parallels here with his approach to the earlier siege of Brionne, in the form of patience and a thorough application to matters operational. While it is unlikely that the siege-castles could have achieved anything more than containment at a site as formidable as Domfront, they undoubtedly contributed significantly to the decisive turning point of the campaign, the failure of Geoffrey Martel’s relieving force to break the siege. Poitiers’s narrative, the only source for Geoffrey’s intervention, states that he brought a large army and mentions formal preparations for battle with William’s forces. It is possible, however, that the challenges exchanged between the two sides were more an attempt at intimidation on Geoffrey’s part than a

114 For Geoffrey and Maine, Latouche, Histoire du comté du Maine, 29–30; Guillot, Anjou, i, 75–7. 115 Louise, La seigneurie de Bellême, i, 353–63. 116 Hicks, ‘The Concept of the Frontier’, 153–4.

124

william the conqueror

prelude to actual hostilities.117 It is also possible that Geoffrey’s army was not as formidable as Poitiers claimed, since the count of Anjou is unlikely to have seen any point in risking a battle for strategically remote castles at the most northerly point of his expanded territories. In terms of William’s personal reputation, the key point is that William had faced down one of the great warriors of his age. Poitiers singles out William fitz Osbern and Roger de Montgommery for their role in the campaign, a reference intended to highlight their increasingly important role and their significance for the future. That the garrison at Domfront did not surrender after Geoffrey’s withdrawal must be taken as a sign that supplies were holding out and that they expected the count of Anjou to return, another possible indication that his army was perhaps not as large as Poitiers suggested and that the confrontation between William and Geoffrey was seen by the defenders as an episode in a long campaign rather than as a decisive moment. William, however, treated Geoffrey’s withdrawal as a tactical opportunity, striking out eastwards against Alençon, which submitted quickly; it is possible that the town’s commitment to the war had always been less than wholehearted, because it was geographically located in the middle of the region where the Bellême family, the Giroies, and other families were battling for control. The citizens and the defenders are likely to have thought in terms of selfpreservation and submission to the most powerful force around. Jumièges mentions an incident during the siege of Alençon that has become famous as a commentary on William’s birth and character. Apparently the defenders of an outlying fortification so provoked William by their taunts that he ordered his forces to storm the place and burn it down. The mockers, when captured, had their hands and feet chopped off. Orderic expands on Jumièges by adding that the garrison were beating skins to draw attention to William’s parentage, a passage that he wrote over an erasure, replacing his earlier text. Although we cannot know what Orderic originally wrote, it is striking that he also inserted the additional information in the margin, written in the same ink as his main text, that the number of men maimed was thirty-two.118 This statistical interpolation was presumably to illuminate the savagery of William’s reaction to an insult to his dignity. It also shows how focused Orderic was on the subject of William’s parentage. While illustrating William’s sensitivity about his birth, it surely also demonstrates his loyalty to his mother – perhaps the latter more than the former, since it was Herleva and her family who were 117 On this subject, Gillingham, ‘William the Bastard at War’, 149–51 (Anglo-Norman Warfare, ed. Strickland, 151–3); Strickland, War and Chivalry, 43. 118 GND, ii, 124–5.

the shaping of things to come

125

being mocked. Poitiers, typically omitting what he disliked in Jumièges’s work, made no mention of the episode beyond saying that William brought the siege to a rapid conclusion. The mutilation certainly worked as a military tactic, since Alençon rapidly surrendered, followed by Domfront when William returned there shortly afterwards. While extending the area under William’s control to include lands around Domfront that had been beyond his father’s reach, essentially the campaign had done no more than restore conditions to where they had been before 1035. Subsequent events show that William’s authority over the lands around Sées and Alençon continued to be tenuous; the region was to be the subject of further interventions during the 1050s and early 1060s up until his conquest of Maine in 1062–3, and, indeed, in changed forms, for the rest of his life. But the sieges are a major turning point, marking the serious beginning of a consolidation of Normandy’s southern and eastern frontiers that was both aggressive and defensive in intent, and a significant feature of William’s actions during the next decade. The attack on Alençon and Domfront must also have contributed to the change in the political configuration of northern France that was apparently under way by 15 August 1052 when Henry I and Geoffrey Martel were in each other’s company at Orléans.119 That they launched military action against William within little more than a year of this date invites the conclusion that their alliance was directed against him from the start. This may not, however, have been entirely the case. Their resumed good relations can in many respects be seen as the revival of a briefly interrupted accord that dated back to the time of Geoffrey’s father and had only deteriorated into hostilities after 1048. It had been an alliance from which both had profited hugely.120 As the two most experienced rulers in northern France, and the two whose interests were most exclusively concentrated within the region, they may well have joined together to police the region’s stability. The counts of Flanders, while being equally, if not more, formidable than the counts of Anjou, had always to look eastwards and southwards towards the Empire, as well as to the west, while the counts of Blois-Chartres, though still harbouring grudges and showing signs of wishing to fight back, had been eclipsed by Geoffrey’s victories. Henry I and Geoffrey Martel would certainly have seen William as constituting a potential threat. His marriage to Matilda of Flanders marked him out as ambitious, while his protection of Bishop Gervase and his defiance of Geoffrey at Domfront showed him to be both determined and 119 RHF, xi, 590. For a favourable discussion of the diploma’s authenticity, Guillot, Anjou, ii, 105–6 (Catalogue des actes, C 141). 120 For the Capetian–Angevin alliance, Bachrach, Fulk Nerra, 207–41; Guillot, Anjou, i, 56–79.

126

william the conqueror

capable. Even if his prospects in England were remote, their very existence, with the implication that he would instantly become much wealthier and more prestigious than any of his neighbours, was bound to set alarm bells ringing. Henry and Geoffrey may therefore have felt that a display of disapproval was called for, an interpretation supported by Jumièges’s explanation of their behaviour in terms of the Franks’ traditional envy of Norman achievement and of William’s provocative behaviour.121 Poitiers omitted the allusion to William’s conduct, referring only to the French king receiving bad advice; he saw the rift between William and Henry as one that progressively deepened.122 As William’s marriage to Matilda took place in either late 1052 or 1053, it becomes another element in these changes. Not only did the alliance with Count Baldwin become more necessary because William had acquired formidable enemies, it may well also indicate a decision to concentrate on alliances with powers along the Channel littoral. William’s final rupture with King Henry I was of such significance that it was used to date a charter that must have been confirmed before the death of Isembard, abbot of La Trinité-du-Mont of Rouen, which occurred on 1 November 1054. The final trigger for the end of an alliance that had lasted for decades, and that had been a lynchpin of the politics of northern France, may well have been the breakdown of the relationship between William and his uncle, Count William of Arques, along with the aggressive intentions Duke William was displaying in several directions.123 He appears among the signa of a charter for the abbey of La Chaise-Dieu in the Auvergne, at Henry’s court on 20 September 1052 at Vitry-aux-Loges to the east of Orléans, amidst a great assembly including the dukes of Aquitaine and Burgundy, Ralph, count of Amiens and Valois, and numerous highranking ecclesiastics. It suggests that there was still good will at this stage between William and the king. It might even have been a mission to explore the implications of the potential change in their relations, and perhaps to try to reverse it.124 But William’s presence at Vitry-aux-Loges was to turn out to be the last occasion in his lifetime that he attended the French king’s court. In terms of subsequent events, this is arguably either as important – or perhaps almost as important – as anything else that happened during these years which shaped so much of what was to come. GND, ii, 142–3. Note in particular the phrase ob ducis improperium. Inique se alienauit iniquus, diram inimicitiam suscepit rex Henricus, transuersus hominum pessimorum suadela, GG, 18–19. See further GG, 42–3. 123 RADN, no. 123 (tempore quo discordia cepit inter ipsum et Henricum regem Francorum); Guillot, ‘L’appui apporté par Henri 1er’, 492–3. For the date of Isembard’s death, NMProsopographie, 263, with a consequential change to the dating-limits of the charter being required. 124 RADN, no. 127. Anxieties about the charter’s authenticity are answered in Guillot, Anjou, ii, 105–6. 121 122

Chapter 4

TH E M AK I N G O F A R EPU TAT IO N

The absence of any record of direct contact between William and Edward the Confessor after the autumn of 1052 until Harold’s visit to Normandy in 1064 might suggest that nothing of note happened. The possibility that messages passed to and fro informally across the Channel cannot be ruled out, but we can be certain that nothing took place that the writers of the main narrative sources thought worth recording. As far as we can see, all the thought and energy devoted to the succession question in England during these years were directed towards finding an heir from among the direct descendants of previous English kings. But, for all the power that Earl Godwine’s sons accumulated in England during these years, the continued presence in Normandy of the two hostages from his family kept alive memories of whatever had been agreed in 1052. The implications of these memories must, while invisible to us, have been very varied, with some undoubtedly trying to destroy any significance they might have had and William, presumably, retaining the opinion that he was going to be the next king of the English. A reading of the main narrative sources that ignores these undercurrents might even suggest that William’s chief preoccupation during the 1050s was with wars against northern French neighbours, in some of which he was on the defensive, and out of which he apparently emerged with only small gains. Yet these years must have been the time when he made the reputation on which the massive support he received in 1066 was founded; that expedition and the earlier takeover of Maine in 1062–3 could only have been undertaken by someone in whom a lot of people had a great deal of confidence and trust. In 1060 the two northern French rulers most obviously capable of restraining him, Henry I and Count Geoffrey Martel, died on 4 August and 14 November respectively, clearing the decks for a reshaping of alliances. In this, fortune could be said to have smiled on William – as it later did when the brothers Harold and Tostig dramatically fell out in 1065. But by then he had proved himself against some tough opponents and had further developed relations with the papacy in ways that brought him support in the 1060s and beyond.

127

128

william the conqueror

FAMILY MATTERS Although the exact dates of birth of William and Matilda’s sons are unknowable, approximations based on the sequence of their births and their later lives suggest that the first son Robert was born soon after their marriage in 1053–4, the second, Richard, in c.1055, and the future king of England, William Rufus, in c.1060. The dates of the daughters’ births are more obscure, but at least two, Adelida and Cecilia, were certainly born in the 1050s, with two more, Constance and Matilda, also born before 1066.1 Barring quite exceptional misfortune, therefore, the succession to the duchy was assured and a level of stability created that made rebellion against William more difficult to justify. The daughters were available for political marriages. Viewed from a functional perspective, successful procreation also guaranteed the continuation of the marriage. As he passed his thirtieth birthday, William’s status within Normandy was as secure as was possible for a medieval ruler. The reference to Herleva welcoming her son’s bride at Eu is the last mention we have of her. Although no remains of any grave have survived, Robert of Torigni’s statement that Herleva was buried at Grestain can probably be accepted; as a one-time monk of the nearby abbey of Le Bec, he was in a position to know.2 William’s stepfather Herluin then married a second wife named Fredesendis by whom he had at least two children, named Ralph and John. He also remained politically prominent, appearing in a charter that dates from between 1059 and 1066 as vicecomes de Conta Villa, evidently still performing the responsibilities that he had held from the time of Duke Robert, augmented by the more general responsibilities of keeping the peace and managing ducal rights and revenues that went with the office of vicomte. Herluin probably died shortly before 1066.3 After his demise – and perhaps before it – responsibility for Grestain’s development passed to Herluin and Herleva’s second son Robert, who became count of Mortain, probably in the late 1050s. William’s generosity to Grestain in terms of actual land remained as limited as it had been at the 1 Frank Barlow, William Rufus, 2nd edn. (New Haven, CT, and London, 2000), 441–5; Elisabeth van Houts, ‘The Echo of the Conquest in the Latin Sources: Duchess Matilda, her Daughters and the Enigma of the Golden Child’, in Pierre Bouet, Brian Levy, and François Neveux (eds.), The Bayeux Tapestry: Embroidering the Facts of History (Caen, 2004), 135–53, at 135–47. 2 Robert of Torigni, ‘De Immutatione Ordinis Monachorum’, in Chronique de Robert de Torigni, abbé du Mont-Saint-Michel, suivie de divers opuscules historiques de cet auteur et de plusieurs religieux de la même abbaye, ed. Léopold Delisle, 2 vols., Société de l’Histoire de Normandie (Rouen, 1872–3), ii, 202; Bates and Gazeau, ‘L’abbaye de Grestain’, 6–7, 17–19, 22–3; Regesta, no. 158. Cf. Douglas, WC, 382. 3 RADN, no. 218. In general, Mark Hagger, ‘The Norman Vicomte, c.1035–1135: What Did He Do?’, ANS, xxix (2007), 65–83.

t h e m a k i n g o f a r e p u tat i o n

129

time of the foundation in 1050. However, the exemptions granted to it from tolls on cross-Channel transport after 1066 and from the ducal tax known as bernagium are described in the abbey’s pancarte as equivalent to those of William’s own foundation of Saint-Etienne of Caen. Even though the exact chronology of these grants cannot be known, this indicates that in William’s eyes the abbey founded by his parents had the status of one of the greatest abbeys of the Norman duchy.4 WILLIAM OF ARQUES AND THE WARS OF 1053–4 Poitiers dates the start of the conflict between William and Count William of Arques to the latter’s withdrawal from the duke’s army at the siege of Domfront. As already noted, if the earlier date of 1049 that some have proposed for this siege is preferred, Poitiers’s statement must be incorrect because Count William appears in subsequent ducal charters. But if the siege is dated to 1051–2, then Poitiers’s chronology becomes plausible.5 Whatever the date, the breakdown of relations looks to have been a gradual, evolving, one. The significant adjustment in the revisions of the Gesta Normannorum Ducum by both the so-called α-redactor and Orderic of the word rebellandum to resistendum suggests that both believed that what happened was better defined as a quarrel, rather than a revolt, and that Count William was defending himself in a way that had some justification.6 The absence of any reference to Count William as a rebel in the Inventio et Miracula sancti Uulfranni, which was being written in 1053, points in the same direction.7 However, because William had at some stage, as a precaution, taken over Count William’s castle at Arques, he was demonstrating a significant level of distrust, even if Poitiers also says that William had not reduced Count William’s possessions or his dignity in any other way.8 Both Jumièges and Poitiers describe Count William as arrogant, with Poitiers adding a diatribe against his consistent disloyalty over many years. But, while Count William certainly possessed a high opinion of his own importance, as we have seen, there is every likelihood that he had supported the young William through the 1040s, and that the duke had tolerated, if 4 Regesta, no. 158 (sancti Stephani Cadomensis habet ubique sua domina quieta and sicut ecclesia sancti Stephani Cadomensis sua habebat quieta). 5 For the main accounts of the war, GND, ii, 102–5; GG, 32–43. For Count William in charters of the early 1050s, RADN, nos. 124, 125. 6 GND, ii, 103, note 5. For arguments similar to mine, Guillot, ‘L’appui apporté par Henri 1er’, 474, 476–81, 482–4, 489–93. 7 Inventio et Miracula sancti Uulfranni, 45. For the significance of Count William’s treatment in the Inventio, van Houts, ‘Historiography and Hagiography’, 238. 8 GG, 34–5.

130

william the conqueror

not actually looked favourably on, his behaviour. The more assertive William of the years after the Battle of Val-ès-Dunes presumably reached a state of mind in which he was no longer prepared to put up with a relationship that seems to have given Count William extensive autonomy and an exceptional status. Count William must have resented this change. His actions were therefore based on a personal grievance against his nephew that some, mainly outside Normandy, believed to be justified. It had nothing to do with any invented categorization of ‘les Richardides’, but was rather a consequence of the way in which politics and relationships had evolved. Count William’s subversion of William’s garrison at Arques with promises of rewards while the duke was far away in the Cotentin suggests that the outbreak of hostilities came as a surprise to him. It also indicates that Count William chose his moment to attack, with the way in which outside powers came to his aid showing he had been forging alliances in preparation for an opportunistic strike of this kind. Enguerrand II, the one-time husband of Duke William’s sister Adelaide, who had become count of Ponthieu in November 1052, moved swiftly to assist Count William, who was now married to his sister. Enguerrand may well also have had aspirations to profit from the alliance to regain the control over Aumale, which his family had had in the 1030s and 1040s, but which seems to have been treated as Adelaide’s dowry and passed back under William’s direct rule.9 In this case, kindred ties and resentment of William’s achievements created an enemy out of a former ally. Events show that Count William’s allies also included Ralph, count of Valois, the ruler of a substantial group of territories to the east of Normandy and the son of a brother of Count Drogo of the Vexin. Count William’s ability to persuade King Henry I and Count Geoffrey Martel to join him made the situation potentially a very dangerous one for Duke William. But the length of time that the king and Geoffrey took to come to Count William’s aid suggests that their plans were not well co-ordinated. For all the threat that the coalition posed, it is in the end remarkable how little support Count William received from within Normandy. One illustration of this is that, according to Orderic, the relatively minor local magnate Richard de Heugleville tried, before Duke William had even arrived in the region, to impede Count William’s plans by constructing a castle at Saint-Aubin-surScie, some five kilometres to the west of Arques.10 A second illustration is that the commanders whose victory in early 1054 at the Battle of Mortemer decided the outcome of the war comprised the heads of the most powerful 9

Bauduin, La première Normandie, 303–5; Thompson, ‘Being the Ducal Sister’, 71. OV, iii, 254–5.

10

t h e m a k i n g o f a r e p u tat i o n

131

families of eastern Normandy, in other words, the very people who might have turned Count William’s actions to profit if they had been minded to do so. They very evidently decided that their fortunes were best served by supporting Duke William, a very clear indication of how powerful he had become by this time. Another crucial non-Norman, Count Eustace of Boulogne, tended towards neutrality. Although he provided Count William with a refuge after his defeat, his younger brother Count Lambert of Lens was the second husband of Count Enguerrand’s former wife, Duke William’s sister Adelaide, and therefore bound to the duke through kinship.11 Eustace’s continued involvement as the ally of William’s father-in-law Count Baldwin V in his war with the emperor Henry III must also have predisposed him against any further commitment to an active alliance with Count William. It was in the war with Henry III that the aforesaid Lambert died, fighting on Baldwin’s side at Lille in either July or August 1054.12 Poitiers’s dramatic account of Duke William riding across Normandy in the late summer of 1053 at such speed that the horses of most of his companions collapsed with exhaustion reads like another piece of typical exaggeration. But it does convey what was presumably William’s assessment of the situation, namely, that it needed to be contained before his enemies could co-ordinate their resources; Poitiers actually says that his display of decisive action before he had fully assembled an army rallied waverers to William’s cause. The three hundred men known to have set out earlier from Rouen unsuccessfully to try to stop Count William’s men from devastating the countryside around Arques are another sign of the support that Duke William received; they could have been led by Roger de Beaumont, relatively recently installed as vicomte of Rouen. On encountering this troop in retreat, William incorporated it into a rapid assault on Arques, the calculation presumably being that, if allowed to continue, Count William’s policy of devastating the surrounding countryside would prevent a besieging army from establishing itself. While failing to take the citadel high on its hill, William was able to establish a siege-castle on the promontory that was the only access route to the castle, thereby isolating Arques from outside support and the garrison from food supplies, something that would only have been possible if he already had secure supply lines in place.13 William’s departure from Arques after the blockade had been established was undoubtedly motivated by his concern to counteract the coalition that King Henry and Geoffrey of Anjou were organizing. In his absence, the 11 GND, ii, 104–5; Thompson, ‘Being the Ducal Sister’, 69, 71; cf. Bauduin, La première Normandie, 313. 12 For this suggestion, Tanner, ‘Eustace II’, 268–9; Bauduin, La première Normandie, 309–10. 13 For the siege-castle at Arques, Jacques Le Maho, ‘Fortifications de siege et “contrechâteaux en Normandie (XIe–XIIe s.)’, Château Gaillard, xix (2000), 181–9, at 182.

132

william the conqueror

troops in the siege-castle ambushed a relieving force led by King Henry at Saint-Aubin-sur-Scie on 25 October 1053, killing Count Enguerrand and capturing Hugh Bardolf, a major magnate whose lands lay mainly in the Gâtinais to the south of Paris and in Champagne. Although this victory did not prevent the king from getting men and provisions into Arques, it does show that whoever was in charge at Arques on William’s behalf commanded a considerable force and was a good general; the invading force led by the French king was apparently distracted by a breakaway troop that feigned retreat and left them open to attack by the main Norman army. William then returned swiftly to Arques, invested the castle closely and forced the starving garrison to surrender, which they did surprisingly quickly. As in the case of Guy of Brionne, Jumièges and Poitiers treat Count William’s fate differently. The former simply says that he went into exile; the latter that Duke William offered to pardon his uncle, to restore his patrimony and to give him lands, a gesture that, if it was made, was probably a failed ploy to steer a former ally and a kinsman back on side against the formidable threat that the king of France and Geoffrey of Anjou still presented. As with discrepancies about William’s treatment of the defeated rebels after Val-ès-Dunes, the difference may well result from Poitiers’s consistently stylized presentation of William as a righteous and merciful ruler. Since by this time it would have been becoming clear that a pardon from William would not have been an unqualified one, it is probable that Count William decided that, in the light of his social and political standing before 1053, it would have been dishonourable for him to accept what was being offered. Like other exiles from William’s Normandy, he might have been hoping that circumstances would change. But they did not. The date of his death is unknown. As far as we know, he never returned to Normandy. Although perhaps another of his exaggerations, Poitiers’s story of a royal edict summoning troops from all parts of France, including Burgundy, the Auvergne, and Gascony, and of an alliance between King Henry, Duke William VII of Aquitaine, Count Theobald III of Blois, and Count Geoffrey Martel, highlights how ambitious Henry’s plans were. However, the relative ease with which the invading armies were contained indicates that they were not that large. Both Jumièges and Poitiers also suggest that, although King Henry’s closest associates wanted to conquer all or part of Normandy, the allies were far from united. Counts Geoffrey and Theobald and Duke William of Aquitaine are said to have resented the summons, and were interested only in a general way in reducing William’s power. The coalition’s two-pronged attack, with one army advancing into the east of the duchy towards Mortemer, and the other through the Evrecin, suggests that their strategy was a co-ordinated assault on Rouen. That the invasion took place in the middle of winter was presumably a necessary

t h e m a k i n g o f a r e p u tat i o n

133

response to the call for assistance from Count William of Arques while he was still holding out. His defeat, and the consequent loss of a base from which the invaders could live off the countryside, was little short of a disaster for them. This was presumably one of William’s objectives in instantly taking the initiative against his uncle. The invasion through the Evrecin achieved an early success when Guitmund, lord of Moulins-la-Marche, handed the castle there over to King Henry, who in turn gave it to Guy-Geoffrey, brother of the duke of Aquitaine and stepson of Count Geoffrey of Anjou.14 The castle’s location some thirty kilometres north of Bellême, taken alongside the relative absence of Bishop Ivo of Sées from ducal charters in the 1050s, suggests that William’s renewal of Norman power around Alençon and its extension around Domfront had not made much difference to the balance of power in the region.15 William’s strategy was to close in on Henry and Geoffrey’s army, thereby preventing it from devastating the countryside and depriving it of supplies; the implicit invitation to engage in battle was therefore not accepted.16 Meanwhile a second Norman army led by Robert, count of Eu, supported by Hugh de Gournay, Hugh de Montfort, Walter Giffard, and William Crispin caught the eastern invading force unawares near the castle of Mortemer, and defeated it in a hard-fought battle that lasted from early morning to the middle of the day, and which Orderic dates to before the start of Lent (6 February) 1054.17 Having received news of the victory, William ensured that it was transmitted to King Henry in as dramatic a way as possible. Elaborating on Poitiers’s story, Orderic says that Ralph de Tosny, acting as the duke’s herald, woke the French king and his army in the middle of the night to inform them of the disaster their allies had suffered.18 They left Normandy before dawn. According to Poitiers, after a period of unspecified further hostilities, William and Henry finally (tandem) agreed to make peace and to an exchange of captured prisoners. The agreement confirmed William’s takeover of lands from the count of Anjou and, so Poitiers says, he was given carte blanche to take more. This permission to attack Count Geoffrey was not, however, formally part of the settlement, allowing both him and William to interpret it as they wished.19 GG, 42–3. For Bishop Ivo’s absence, Guillot, Le comte d’Anjou, i, 82–5; Bates, Normandy, 79; Louise, La seigneurie de Bellême, i, 362–3. 16 For William’s tactics, Gillingham, ‘William the Bastard at War’, 152–3 (Strickland [ed.], Anglo-Norman Warfare, 155–6). 17 OV, iv, 88–9. 18 For Orderic’s addition, GND, ii, 144–5. 19 GG, 50–1. 14 15

134

william the conqueror

THE DEPOSITION OF ARCHBISHOP MALGER The process that led to the removal – or perhaps retirement – of another of William’s uncles, Archbishop Malger of Rouen, must have begun soon after the victory at Mortemer. According to Poitiers, Malger was a despoiler of his church’s lands who lived extravagantly and who had not obtained a pallium from the pope. While acknowledging that he was sufficiently learned to perform his duties, Poitiers reckoned him in most respects unfit for office.20 Other sources, including the tract known as ‘the Acts of the Archbishops of Rouen’ (Acta Archiepiscopum Rotomagensium), written in c.1070, also contain this critique, with Orderic adding that Malger had fathered a son named Michael.21 As we have seen, according to William of Malmesbury, Malger had opposed William’s marriage on the grounds of consanguinity. It is of course possible that Malger cloaked support for his brother Count William in a show of moral disapproval of their nephew’s conduct.22 Equally, however, he may just have been one of the Norman churchmen who advised William to be cautious; the real problem was probably Malger’s kinship with Count William. Jumièges’s laconic assessment, written in the later 1050s and therefore the earliest commentary, that Malger started to behave foolishly and gave up the archbishopric, raises the possibility that the whole narrative set out in the previous paragraph was a creation of the 1070s, a rationale manufactured after William’s reputation as a supporter of ecclesiastical reform had become fully established.23 If this was the case, Jumièges’s cryptic passage, like his treatment of the fall of Count William, has elements that suggest that what he perceived to be happening was the falling apart of the regime of William’s adolescence, a process whose consequences he could not at the time of writing foresee. Although Malger’s faults were certainly those of many higher clergy in the pre-Gregorian Reform period, much evidence, such as the ecclesiastical council of c.1045, his appearances in charters, and the continuing construction in his time of the new cathedral at Rouen that was to be consecrated in 1063, suggests a conscientious archbishop.24 And, while no one before Orderic suggests that Malger was a member of the alliance of 1053–4, the reason for his removal was surely his association, directly or indirectly, with that conspiracy. Ibid., 86–9. AAR, 38–9, 52; OV, iii, 86–7. 22 GR, i, 494–5. For the possible source of Malmesbury’s confusion, GND, ii, 130–1. 23 GND, ii, 130–1 (desipere cepit et insipientia ductus archiepresulatum reddidit duci). 24 For the Rouen school, Jean-Michel Bouvris, ‘L’école capitulaire de Rouen au XIe siècle’, Etudes normandes, iii (1986), 89–103, with some reference to the period of Malger’s archiepiscopate at 93–7. 20 21

t h e m a k i n g o f a r e p u tat i o n

135

Although unquestionably the outcome that William was seeking, Malger’s removal was formally performed at a synod held at Lisieux; as canon law required, it was carried out by a papal legate, Ermenfrid, bishop of Sion, in Malger and William’s presence.25 The bishop of a diocese in the kingdom of Burgundy on the trans-Alpine routes that were vital to communication between the kingdoms of Germany and Italy and between the pope and the emperor, Ermenfrid subsequently turns up at several crucial moments in William’s life, as well as in pre-1066 England. The whole process of Malger’s removal is a commentary on William’s respect for legal formalities in a matter of this kind. The papacy’s readiness to associate itself with such an obviously political act surely demonstrates a sense of shared interests and values with William. Any residual legacy of tension created by William and Matilda’s marriage, if indeed there was ever any in the first place, must surely have evaporated. Leo IX, pope until his death in 1054, evidently regarded William as a worthy ruler. Uncertainties about the synod of Lisieux’s date have arisen among modern historians for various reasons, with the majority opinion probably preferring 1054 as opposed to 1055.26 The main arguments for a dating of 1054 are that Malger’s successor Maurilius appears among the signa of a charter for Mont Saint-Michel that is dated to 1054 and because the Acta Archiepiscopum Rotomagensium says that Ermenfrid was acting as a legate of Leo IX, who died on 19 April 1054.27 Against this, Malger appears in an original charter for Mont Saint-Michel that is dated to Christmas 1054 and with one exception all the versions of the Norman annals date the deposition to 1055.28 Although the details of the transmission of these annals AAR, 39–40, 52. Jumièges also mentions a synod, GND, ii, 130–1. For Ermenfrid’s career and the date of the synod, see now, ‘Ermenfroi, 1054/1055– 1087/1092’, in Patrick Braun, Brigitte Degler-Spengler. and Elsanne Gilomen-Schenkel (eds.), Helvetia Sacra, i/5: Das Bistum Sitten/Le diocèse de Sion, L’archidiocèse de Tarentaise (Basel, 2001), 149–51, at 149. The synod is dated there to August 1055 and Ermenfrid’s appointment to the bishopric to between 13 July 1054 and June 1055. Previously, the case for dating the synod to 1055 was most effectively set out by Michel de Boüard, ‘Notes et hypothèses sur Maurille, moine de Fécamp’, in L’abbaye bénédictine de Fécamp. Ouvrage scientifique du XIIIe centenaire (658–1958), i (Fécamp, 1959), 81–92, at 89–92. For 1054, Douglas, WC, 69, note 3; H. E. J. Cowdrey, ‘Bishop Ermenfrid of Sion and the Penitential Ordinance following the Battle of Hastings’, JEH, xx (1969), 225–42, at 226–8; Guillot, Anjou, i, 186, note 232; Foreville, ‘The Synod of the Province of Rouen’, 23. 27 RADN, no. 132 (The Cartulary of the Abbey of Mont-Saint-Michel, 126–7, no. 43; Chroniques latines du Mont Saint-Michel [IXe–XIIe siècle], ed. Pierre Bouet and Olivier Desbordes [Caen and Avranches, 2009], 399–400, no. 27); AAR, 39–40, 52. 28 RADN, no. 133 (The Cartulary of the Abbey of Mont-Saint-Michel, 87–8, no. 9; Chroniques latines du Mont Saint-Michel, 400, no. 28). For the references to 1055 in the Norman annals, see ‘Chronicum Rotomagense’, in P. Labbé (ed.), Novae Bibliothecae Manuscript. Librorum, 2 vols. (Paris, 1657), i, 364–90, at 366; ‘Annalis Historia Brevis sive Chronica monasterii S. Stephani Cadomensis’, in J. A. Giles (ed.), Scriptores rerum gestarum Willelmi Conquestoris 25 26

136

william the conqueror

around the several churches of Normandy remain a subject of discussion, the consensus that all versions derive from a text written at Rouen in Maurilius’s time means that they must be treated as a first-hand witness for the dating.29 The statement in the Acta Archiepiscopum Rotomagensium that Ermenfrid was acting as Leo’s legate can therefore be treated either as recording that he had been appointed as legate by Leo, or as having been written with memory of that dynamic pope in mind. That Ermenfrid was not actually appointed bishop of Sion until after 13 July 1054 also supports a date of 1055 for the synod, as does the fact that the whole process must have taken time to set up, involving, as it must have done, the dispatch of a mission from Normandy to the papacy at some time after the victory at Mortemer. The confusion created by the two charters can be explained by treating Maurilius’s apparent appearance in 1054 as an error by a cartulary scribe, who either expanded incorrectly an abbreviated signum that in the case of either of the archbishops had to begin with the capital letter ‘M’, or who copied as a single text a list of signa from an original into which Maurilius’s name had been interpolated. The diploma that has Malger present at Rouen cathedral at Christmas 1054 is an original, a remarkable survivor of the destruction of the Archives de La Manche in 1944. It is possible, as has indeed been suggested, that the charter’s scribe was starting the year at Christmas, which would date it to Christmas 1053. But, even if this was the case, it cannot be treated as being of decisive importance when set against the mass of evidence that exists in favour of the year 1055 for Malger’s removal. After being deposed, Malger was exiled to Guernsey and died there; by drowning, according to the Acta Archiepiscopum Rotomagensium.30 His replacement, Maurilius, a native of Rheims and a former monk of Fécamp, with an unhappy experience as an abbot in Italy behind him, was a man universally praised for his moral and spiritual qualities. His appointment marks an obvious departure from the traditional policy of appointing close ducal relatives to the archbishopric, which had prevailed since 989. It was also in (London, 1845), 161–74, at 165; Les annales de l’abbaye de Saint-Pierre de Jumièges: chronique universelle du XIIIe siècle, ed. Dom. J. Laporte (Rouen, 1954), 56–7; ‘Annales Uticenses’, in Orderici Vitalis Ecclesiasticae Historiae Libri Tredecim, ed. A. Le Prévost, 5 vols. (Paris, 1838–55), v, 139–73, at 157. Only the Annals of Mont Saint-Michel, whose text is otherwise identical to that of the other versions of the annals, place it within the entry for the year 1054, ‘Annales du Mont-Saint-Michel’, in Chronique de Robert de Torigni, ed. Delisle, ii, 207–36, at 220. 29 See, most recently and in my view convincingly, Alexander, ‘Annalistic Writing’, 34–40, 61–71, 78–81; cf. Les Annales de Jumièges, 7–25; Louis Violette, ‘Une étape décisive dans l’éveil des activités historiographiques au service du siège de Rouen’, Tabularia, www.unicaen.fr/ mrsh/craham/revue/tabularia, Etudes, iii (2003), 57–67. 30 GND, ii, 142–3; OV, iv, 84–5; AAR, 39, 52.

t h e m a k i n g o f a r e p u tat i o n

137

contrast to all five of the episcopal appointments that had been made within the metropolitan province of Rouen since the start of William’s reign, all of which had involved individuals from the families of the duchy’s aristocratic elite. Maurilius’s links to Fécamp also suggest that Abbot John continued to be a powerful force within the Norman Church and that he and others, perhaps including Lanfranc, were driving William towards a different perspective on the qualities required of a bishop. FROM MORTEMER TO VARAVILLE (1054–7) The Mont Saint-Michel charter dated to Christmas 1054 that includes Malger’s signum shows William in Rouen cathedral in the presence of a high-powered mission from the far west of Normandy that included tenants of the abbey, the local representatives of religious and secular power, namely, Bishop Hugh of Avranches and Richard vicomte of the Avranchin, and William’s great supporters Roger de Montgommery and Roger de Beaumont. As such, it is good evidence for a display of the kind of ceremonial grandeur that became a frequent feature of later times. Archbishop Malger’s presence and his apparently central role in the proceedings are indicative of another theme that will reappear later, namely, the continuation in functioning office of individuals whose imminent removal was probable. The choice of Ralph de Tosny to be the bearer of bad tidings to King Henry after the Battle of Mortemer illustrates the increasing cohesion of Norman political society around William in a different way. As the head of a family that had suffered in the 1040s and still had a grievance against the Beaumont family, Ralph de Tosny’s central place indicates how interests focused on sustaining ducal authority were taking precedence over specific personal ones. In the aftermath of the victory, William rewarded those who had supported him. Roger de Beaumont, Walter Giffard, Hugh de Montfort, and William de Warenne were all recipients of some of the lands of the fallen William of Arques. At the same time, the comté specially created for Count William was suppressed and dismembered.31 With the exception of Roger de Beaumont, who chose not to receive much land there, all later prospered massively as a result of the conquest of England. On the duchy’s southern frontier William fitz Osbern was made responsible for the castle at Breteuil-sur-Iton in lands which marched with the king of France’s lands contiguous with the long-contested region around Tillières-sur-Avre. While no buildings survive, there is plentiful evidence 31 See Douglas, WC, 92; Jacques Le Maho, ‘L’apparition des seigneuries châtelaines dans le Grand-Caux à l’époque ducale’, Archéologie médiévale, vi (1976), 5–148, at 31, 37.

138

william the conqueror

that he drove forward the dynamic development of what became a massive castle and the community associated with it; the alignment of the castle, the new main street, and the church were probably inspired by the earlier models of Rouen, Fécamp, and Falaise.32 Notwithstanding the much discussed subject of whether William fitz Osbern owed his presence at Breteuil to his descent from Count Rodulf of Ivry through his son Bishop Hugh of Bayeux or through his daughter’s marriage to Osbern the steward, the crucial point is that his role occupies a decisive place in a movement which gathered pace from 1054 onwards, the militarization of Normandy’s border regions.33 Although it has been argued that Guitmund de Moulins-la-Marche would be treated leniently for handing over his castle to King Henry, what we have already seen in this and the previous chapter about William’s treatment of rebels and prisoners suggests otherwise. Hence, either because they knew that William’s acts of mercy came at a price or because they had already been disinherited, Guitmund and his sons seem to have decided that their prospects of holding on to the castle and their lordship were weak. Most of the family therefore departed to join other kindred in southern Italy, where they prospered. In due course, both the lordship and the castle were transferred to a cousin of William whose name was also William, a son of his mother Herleva’s brother Walter. This was another indication of just how important William’s maternal kin continued to be after Herleva’s death, as well as of the increasing influence of William’s closest kindred and family across the regions of Normandy.34 Roger de Mortemer, a major landholder in Upper Normandy near to the duchy’s eastern frontier regions, who had been prominent in the victory at Mortemer, was temporarily exiled for having protected Ralph, count of Amiens and Valois, from capture after the battle, because Roger had performed homage to him. Although William soon restored Roger to his lands, he took the castle of Mortemer away from him and gave it instead to William de Warenne. The story, which Orderic included in the lengthy deathbed speech he attributed to William – inserting the self-justificatory remark ‘acting, I believe, rightly’ – was manifestly intended as a rhetorical demonstration of virtue that was thought likely to earn William a reward in Heaven. William’s response had been a worthy one to the ethical 32 GND, ii, 146–7. For Breteuil, Bauduin, La première Normandie, 229–30; LemoineDescourtieux, La frontière normande de l’Avre, 111–13, 246–56; Bernard Gauthiez, Atlas morphologique des villes de Normandie (Lyon, 1999), 96. 33 For the controversies, with references, Bauduin, La première Normandie, 219–24; cf. D. C. Douglas, ‘The Ancestors of William fitz Osbern’, EHR, lix (1944), 62–79. 34 Emily Zack Tabuteau, ‘The Family of Moulins-la-Marche in the Eleventh Century’, Medieval Prosopography, xiii (1992), 29–65, esp. pp. 32–5, 50–4.

t h e m a k i n g o f a r e p u tat i o n

139

problem of how to treat someone obligated to two lords at war with one another, since Roger might be deemed to have been merciful and correct. Although undoubtedly intended to administer a short, sharp shock to Roger, William’s conduct did acknowledge the conflict of interest involved and the complications attendant on cross-border personal relationships beyond Normandy’s frontiers. Since he was later to take over the castles of individuals far more obviously in favour than Roger de Mortemer, by his own standards William’s conduct in this case possibly leaned towards the generous, especially as Count Ralph remained an adversary into the early 1060s.35 On the other hand, the recognition given to Roger’s dilemma seems to have worked both to his and William’s benefit, since it created a framework for a collaborative relationship. Unlike many who had crossed William before 1066, Roger received substantial rewards in England after the Conquest. In political terms, the most important of the prisoners held by William after the Battle of Mortemer was Guy of Ponthieu, the brother and successor of Count Enguerrand II. According to Orderic, he was held for two years at Bayeux, being released after doing homage to William, agreeing to perpetual fealty, and to provide the service of one hundred warriors every year – terms that Orderic describes as being hard ones.36 If he is the Count Guy who appears among the signa of an important diploma of Count Baldwin V of Flanders for the abbey of Saint-Pierre du MontBlandin of Ghent, dated 13 November 1056 at Saint-Omer, then Orderic’s chronology is probably confirmed.37 It also looks as if Guy abandoned his family’s claims to hold Aumale, since William’s twice-widowed sister Adelaide was installed there and founded a college of secular canons at Auchy-lès-Aumale; she seemingly lived out a life of ‘semi-religious retirement’ there while bringing up two daughters, one by each of her two husbands (Counts Enguerrand II and Lambert II).38 However, as we shall see, the evidence indicates that Guy’s attitude towards William remained at best ambivalent. As we shall also discover later in this chapter, when discussing the diploma that supplies evidence of the meeting at SaintOmer, Guy would have met Harold Godwineson, earl of Wessex, there. In his famous early twelfth-century autobiography, Abbot Guibert of Nogent says that his father was captured during fighting between the 35 OV, iv, 88–9. Note: illi iure ut reor abstuli. For the ransom and the general issues involved, Strickland, War and Chivalry, 231–40. 36 OV, iv, 88–9. 37 For the argument that the Signum Uuidonis comitis is Guy of Ponthieu, Philip Grierson, ‘A Visit of Earl Harold to Flanders in 1056’, EHR, li (1936), 90–7, at 92. 38 Bauduin, La première Normandie, 313; Thompson, ‘Being the Ducal Sister’, 71–2, with the quotation at p. 71.

140

william the conqueror

French and the Normans. He further explained that William did not usually ransom prisoners taken in war, but kept them captive for longer than the accepted norm.39 Since Guibert’s father was in due course released and actually witnessed his son’s difficult birth in c.1055, the abbot’s assessment of William’s conduct may well be an exaggeration. It is also possible that the story is not connected with the Battle of Mortemer. On the other hand, the association with the battle is plausible, since the story is so compatible with the fact that there were lengthy peace negotiations after Mortemer and that William used prisoners captured there to drive a hard bargain.40 Also, given that Guibert was a monk and abbot of a monastery near Normandy’s eastern border, with close links to many people who would have known William well, his comments on William’s methods may indeed have general significance. If so, they draw attention to the ambivalent reactions William must have provoked; he played with the rules in a way that gave offence and, for this reason, he achieved results and attracted support. Guibert’s account of his mother’s psychological suffering and of her persecution by her husband’s relatives during his imprisonment, even if perhaps similarly exaggerated, also provides one of many explicit reminders in the contemporary sources of the human misery that a career so ruthlessly directed as William’s caused. During the lengthy peace negotiations after Mortemer, which may have lasted for two years or more, relations between William and his main enemies unsurprisingly remained tense. The promotion in October 1055 of Bishop Gervase of Le Mans to the archbishopric of Rheims, sanctioned by King Henry I and given papal approval, was presumably partly achieved to consolidate Geoffrey Martel’s grip on Maine. Up to this point still an exile in Normandy, Gervase’s promotion weakened William’s propaganda position as the protector of a persecuted bishop.41 Although the chronology is obscure, it is certain that William went on the offensive in northern Maine at some point after the Battle of Mortemer by constructing a castle near Ambrières, some twenty-three kilometres south of Domfront and some twelve to the north of Mayenne, north of where the river Colmont flowed out of the Mayenne.42 Poitiers’s statement that William announced his 39 Guibert de Nogent, Autobiographie, ed. E.-R. Labande (Paris, 1981), 88–91 (A Monk’s Confession: The Memoirs of Guibert of Nogent, trans. Paul J. Archambault [University Park, PA, 1995], 40). 40 For the many discussions of these issues and of the date of Guibert’s birth, Jay Rubenstein, Guibert of Nogent: Portrait of a Medieval Mind (New York and London, 2002) 17, 66, with the technicalities discussed at p. 222, notes 2 and 3. 41 Guillot, Le comte d’Anjou, i, 91–2. For Gervase’s presence in Normandy, Actus Pontificum Cenomannis in urbe degentium, 366–7; RADN, no. 137. 42 For the castle’s location super ora fluminis quod dicitur Colmunt, Quedam exceptiones de hystoria Normannorum et Anglorum, GND, ii, 300–1.

t h e m a k i n g o f a r e p u tat i o n

141

intention to construct the castle during the peace negotiations makes sense only if these were protracted, so this further encroachment into Maine probably took place during late 1054 or in 1055.43 That the castle was intended as a forward-post to protect territorial advances that William had already made is implied by Poitiers’s statement that the terms agreed after Mortemer confirmed William in possession of lands already taken from Geoffrey Martel. Another aim of the castle was undoubtedly to contain the most powerful aristocrat of northern Maine, Geoffrey de Mayenne, someone who was to trouble William for much of the rest of his life.44 Poitiers’s narrative suggests that the castle, of which nothing now survives, was a substantial enclosure with a strong curtain wall. Summoned by Geoffrey de Mayenne, Geoffrey Martel, together with Duke William VII of Aquitaine and the Breton Count Eudo, attacked the newly constructed castle near Ambrières, but then retreated rapidly when William arrived with a relieving force. With typical extravagance Poitiers describes their withdrawal as a disorderly flight. In truth, as earlier at Domfront, Geoffrey Martel probably decided that it was not worth fighting a battle for what was, from his point of view, a distant outpost at the extreme north of his lands.45 The presence of Count Eudo, who was still taking a major role alongside his now adult nephew Count Conan II, is an important commentary on the balance of power within northern French political society at this time. The fact that Eudo, previously supportive of the young William, had now joined forces with his enemies may indicate how much William was coming to be feared, but, more likely, it shows how dominant Geoffrey Martel had become in north-western France after taking over Maine; Eudo’s alliance with Geoffrey seemingly involved him in receiving land which Geoffrey had removed from the possessions of the abbey of Saint-Aubin of Angers.46 Abandoned by his allies, Geoffrey de Mayenne did fealty to William. Poitiers’s meaning is somewhat obscure, but in all probability all that was involved was an obligation to keep the peace on the frontier.47 Alongside William fitz Osbern’s installation at Breteuil, the Ambrières campaign indicates a co-ordinated approach to the security of lands in the spaces 43 The main source is GG, 50–5. On the date of the castle’s construction, idem, 50, note 1; Guillot, Le comte d’Anjou, i, 80, note 358; Louise, La seigneurie de Bellême, i, 371–2. 44 For Geoffrey de Mayenne’s power base, Annie Renoux, ‘Aux sources du pouvoir châtelain de Geoffroi “seigneur de Mayenne, le plus fort homme du Maine” (c.1040–1098)’, in Dominique Barthélemy and Olivier Bruand (eds.), Les pouvoirs locaux en France du centre et de l’Ouest (VIIIe–XIe siècle). Implantation et moyens d’action (Rennes, 2004), 61–90, at 64–80. 45 For this suggestion, Louise, La seigneurie de Bellême, i, 367–8. 46 Cartulaire de Saint-Aubin d’Angers, no. 677 (p. 172). 47 ut in remotissimis Normanniae partibus sibi manus perdomitas daret, fidelitatem quam satelles domino debet, iurans, GG, 54–5; see further the discussion at p. 54, note 1.

142

william the conqueror

around the southern limits of the duchy. The limitations to the progress made are, however, graphically illustrated by Orderic’s statement that, his marriage alliance with the Bellême family notwithstanding, Roger de Montgommery had to struggle to impose himself in the vicinity of Sées. His behaviour indeed takes on some of the characteristics of a lord with cross-border interests, since he turns up at Angers in 1057 in the company of Bishop Ivo of Sées in a charter of Count Geoffrey Martel.48 Although there is also evidence of a strengthening of influence further west in the territories close to the lands of the dukes of Brittany, relations with Nigel the vicomte had broken down by 1056 at the latest, when he appears in Count Eudo’s entourage in the presence of Geoffrey Martel in either 1056 or early 1057.49 Nigel seems to have continued to play the conventional role of a vicomte up until the mid-1050s, appearing among the signa of several charters that, although in some cases difficult to date closely, can usually be assigned to the years 1053–5. The most complex of them, a confirmation of a grant of the church of Savigny-le-Vieux to the abbey of Marmoutier by Main de Fougères, most of whose extensive lands were in territories controlled by the dukes of Brittany, probably dates from this time, because – almost uniquely for a ducal charter – the signa include William Werlenc, count of Mortain, and because Main died in either 1054 or 1055.50 Nigel also features in the original Mont Saint-Michel diploma, dated 1054, in which Archbishop Maurilius appears and which is problematic in relation to Archbishop Malger’s deposition.51 Of other relevant charters, William’s grant to the abbey of Marmoutier of six churches on the island of Guernsey, churches formerly held by Nigel, in all probability dates to 1053.52 This removal of the Guernsey churches could have been the source of Nigel’s grievance against William, since, while in exile, Nigel confirmed 48 OV, ii, 66–9; Cartulaire de Saint-Aubin d’Angers, no. 941 (Guillot, Anjou, ii, 153–4, no. C230); Thompson, ‘The Example of the Montgomerys’, 261. 49 Cartulaire de Saint-Aubin d’Angers, no. 677 (Guillot, Anjou, ii, 127 [Catalogue des actes, no. C176]); Brunterc’h, ‘Geffroy Martel, Conan II et les comtes bretons’, 314–15; Les actes des ducs de Bretagne, 301 (no. 55). Existing historiography has Nigel exiled after Val-ès-Dunes and returning to Normandy by 1054, Delisle, Saint-Sauveur-le-Vicomte, 19–21; van Torhoudt, ‘Les sièges du pouvoir des Néel’, 14–15. Hagger, ‘How the West was Won’, 38, does not suggest a date for the exile, but rightly emphasizes the severity of Nigel’s treatment. 50 RADN, no. 162. Van Torhoudt, Centralité et Marginalité, 525, dates the diploma to 1054x1056. For Main’s death, Morin, Trégor, Goëlo, Penthièvre, 97–8. After being prominent in Breton comital diplomas up until 1047, he then disappears completely from them, Les actes des ducs de Bretagne, nos. 46–52. 51 RADN, nos. 132 (The Cartulary of the Abbey of Mont Saint-Michel, 126–7, no. 43; Chroniques latines du Mont Saint-Michel, 399–400, no. 27). 52 RADN, no. 141. Marie Fauroux dated this charter to 1052x1058, but the presence of Matilda, the young Robert Curthose, and Ralph, abbot of Mont Saint-Michel, all point to 1053 or if later, only slightly so. Above, 104, note 51.

t h e m a k i n g o f a r e p u tat i o n

143

the gift to Marmoutier, reserving to his own canonical foundation of SaintSauveur-le-Vicomte grants he had already made to it and indicating that monks from Marmoutier might be introduced there. This reads like a negotiating document drawn up by someone who wanted to return to Normandy.53 Another grievance might have been William’s curtailment of Nigel’s regional influence by the marriage of his half-sister Muriel, a daughter of Herleva and Herluin, to a Cotentin rival, Eudo, son of Thurstan Haldup; the date of the actual marriage is, however, unclear.54 Whatever the case, Nigel remained in exile until c.1060. Although Count Eudo’s imprisonment by his nephew Count Conan in 1057 deprived him of a protector, he subsequently turns up in the company of King Henry I in either April or May 1059 among the signa of a second confirmation of the foundation of a priory of Marmoutier by Geoffrey Martel, with the document saying the grant was made while Henry’s forces were besieging the garrison that William had by then installed in the castle of Thimert.55 The first of the many great church dedications in Normandy that took place during William’s lifetime, the consecration in his presence of the cathedral church of Coutances was celebrated on 8 December 1056. The event is also notable as a display of how the power of the secular and religious Norman core was becoming more securely entrenched in the western regions of the duchy. The foundation stone of the new cathedral had been laid by Duchess Gunnor and canons had been installed there in the time of Bishop Robert (c.1025–48), a chronology confirmed by the parts of the church dating from the first half of the eleventh century that survive within the now almost entirely Gothic cathedral. However, all the Norman bishops before Bishop Geoffrey had resided at Rouen and tried to govern their diocese from a distance.56 Geoffrey’s fund-raising in southern Italy and William’s contribution to the endowment allowed the bishop to live at Coutances and to create the full infrastructure of a bishopric. The surviving 53 Delisle, Saint-Sauveur-le-Vicomte, 23–5, pièces justificatives, nos. 20, 21 (ut, si aliquand, fauente Deo, reverteretur in supradictam patriam atque in honorem pristinam). 54 Regesta, no. 175(I); Roman de Rou, pt. 3, lines 6,003–6; van Torhoudt, Centralité et Marginalité, 564–7; Hagger, ‘How the West was Won’, 35–6, 38. 55 Cartulaire de Marmoutier pour le Vendômois, ed. C. A. de Trémault (Paris and Vendôme, 1893), no. CXVII (p. 191) (Catalogue des actes d’Henri Ier, roi de France [Paris, 1907], ed. F. Soehnée, no. 116); cited in Brunterc’h, ‘Geffroy Martel, Conan II et les comtes bretons’, 315; Les actes des ducs de Bretagne, 301 (no. 55). Delisle, Saint-Sauveur-le-Vicomte, 30–1, pièces justificatives, no. 28, published the signa of the diploma. 56 Many of these details are based on the charter confirming the endowment of the cathedral, which dates from 1056x1066, RADN, no. 214 (Le cartulaire du chapitre cathédrale de Coutances, ed. Julie Fontanel [Saint-Lô, 2003], no. 340). For the remains of the Romanesque cathedral, with references to other work, Maylis Baylé, ‘Les évêques et l’architecture normande au XIe siècle’, in Pierre Bouet and François Neveux (eds.), Les évêques normands du XIe siècle (Caen, 1995), 151–72, at 161–7.

144

william the conqueror

sources, namely the charter and the so-called Gesta Gaufridi, written in the early twelfth century, are in some important respects at odds as to who exactly was responsible for these major changes: the charter, for example, has William make grants that included a half of the town of Coutances and its suburbs and numerous churches, some of them on the Channel Islands, while the Gesta states that Geoffrey bought the half of the town from William for £300.57 The latter seems more likely, both in terms of many other aspects of William’s behaviour and of the pattern by which elements of narrative were omitted from charters. But, whatever the case, the crucial point is William’s support for this major change. It also highlights the pre-1066 dynamism of Bishop Geoffrey, another man who was to play a central role in both England and Normandy after 1066. In the summer of 1057, King Henry and Count Geoffrey of Anjou invaded Normandy again. The breakdown of the tenuous peace established after 1054 is not difficult to understand. William’s treatment of prisoners and his incursions into northern Maine were an obvious provocation; charter evidence shows that Henry visited Anjou and that he and Geoffrey were working closely together in the early months of 1057.58 Presumably relying again on at least the acquiescence of Bishop Ivo, they assembled an army that advanced through the Hiémois and into the region around Bayeux.59 Their route took them along the course of the river Dives in central Normandy and then westwards towards the lands between Caen and Bayeux; Wace records what may have been a local tradition that Henry slept in the abbey at Saint-Pierre-sur-Dives and that the invaders occupied Caen, which he believed to have been unfortified at that date. On reaching the coast, they turned eastwards and set out to cross the estuary of the Dives in order to continue to plunder around Lisieux and in the Pays d’Auge. Poitiers notes that the invading force was not as strong as that of 1054. According to both Jumièges and Poitiers, the invasion’s objective was revenge for the defeats of 1054. The tactic of ravaging may have been intended to try to draw William into battle on disadvantageous terms; on the other hand, the motive may just have been to put him in his place and to curtail his ambitions. It is possible that William was taken by surprise at 57 For Geoffrey’s pre-1066 career at Coutances, J. H. Le Patourel, ‘Geoffrey of Montbray, Bishop of Coutances, 1049–1093’, EHR, lix (1944), 129–61, at 134–43; Chibnall, ‘La carrière de Geoffroi de Montbray’, 280–5; Julie Fontanel, ‘La réorganisation religieuse sous Guillaume le Conquérant: le cas de l’église de Coutances’, Revue de l’Avranchin et du pays de Granville, lxxvii (2000), 189–208, at 191–6. 58 Guillot, Le comte d’Anjou, i, 95–7. 59 The main sources are GND, ii, 150–3; GG, 54–7; Roman de Rou, pt. 3, lines 5,115–298. The date of the Battle of Varaville was established by Dhondt, ‘Les relations’, 482–3.

t h e m a k i n g o f a r e p u tat i o n

145

the scale of the invasion.60 He may, however, have been leading on his enemy’s army in order to wait for a moment to pounce; a ravaging army has to keep on the move to feed itself and, if battle was not the objective, then it had either to continue its advance into Normandy or retreat. In trying to cross the marshy estuary of the Dives on a causeway that was probably an old Roman road, Henry and Geoffrey were certainly putting themselves at risk; this was a place where tides rose rapidly and that in no way resembled the town and small river visible nowadays. It is possible that they had fallen into a trap.61 However the campaign is explained, what is certain is that neither William’s nerve nor the discipline of his troops failed. He held off until his opponents were crossing the marshes near Varaville and then commanded his soldiers to pounce on the rearguard that was waiting their turn. Disrupted by this surprise attack and by the rapidly rising tides, Henry and Geoffrey, who had already crossed the river, suffered a defeat that sent them hurrying back to their own lands. According to Poitiers, William’s forces killed almost all of their enemies trapped on the west bank of the river. What was in effect the massacre of a heavily outnumbered force was justified by Poitiers on the grounds that it was essential to the defence of Normandy. The campaigns of the next twenty years were to show that intimidatory tactical violence of this kind was a regular part of William’s armoury. Given how Poitiers set out to present William to the world, it is interesting that he thought it necessary to justify his actions in this case. Jumièges once more makes an oblique reference to the possibility that William’s behaviour after the victory went beyond contemporary norms by mentioning the dispersal of prisoners he had taken to places around Normandy.62 David Douglas’s view that the Battle of Varaville was not as momentous an event as the war of 1053–4 is undoubtedly right.63 In 1057, William’s enemies were less numerous, less determined, and had no support whatsoever that we can determine within the heartlands of the duchy. Nonetheless, Varaville must still be seen as a major landmark, since there was to be no significant invasion of Normandy during the rest of William’s life. The next three years, although superficially less dramatic, would see William take the initiative in all kinds of ways. 60 For this suggestion, Gillingham, ‘William the Bastard at War’, 154–5 (Strickland [ed.], Anglo-Norman Warfare, 157). 61 For recent discussions of the battle, Stéphane Lainé, ‘La bataille de Varaville (1057): examen critique des récits médiévaux’, BSAN, lxvii (2009, for 2008), 257–88, esp. pp. 260–9; Vincent Carpentier, Guillaume le Conquérant et l’estuaire de la Dives, Collection Cahiers d’Auge, 3 (Lisieux, 2011), 28–32, 60–4. 62 GND, ii, 152–3 (hos per Normannica competa dispergeret uiolentia dire captiuitatis). 63 Douglas, WC, 72–3.

146

william the conqueror

ENGLAND IN THE 1050s With Queen Edith restored at court after her banishment by Edward at the start of the crisis of 1051–2, but the marriage remaining childless, the question of who would eventually succeed him remained as open as it had been before 1051. But Edward’s continued good health meant that there was no way anyone could predict when the crisis, which all must have been anticipating, might break. The increase during these years in the dominant place across the entire kingdom of the sons of Earl Godwine, after his sudden death on 15 April 1053, was a major factor that was certain to have a powerful influence on future events. Harold’s succession to his father’s earldom of Wessex (due to the eldest son Svein’s death in 1052), Tostig’s insertion into Northumbria in March 1055 against the claims of the deceased Earl Siward’s young son Waltheof, and their younger brothers Gyrth and Leofwine’s receipt of earldoms in East Anglia and south-east England, apparently tilted the balance of power decisively in the family’s favour. Leofric of Mercia’s son Ælfgar did succeed to his father’s massive earldom after his death in 1057, but his rebellions and periods of exile, and his use of the ruler of Gwynedd, Gruffudd ap Llywelyn, as an ally in England, must be indicative of continuing, sometimes acute, political tensions.64 A tradition has emerged in francophone historiography that the discussions about the succession took place during the visit that Abbot John of Fécamp made to Edward the Confessor’s court in late 1054.65 However, the surviving source, a seventeenth-century printed copy based on an apparently lost manuscript, mentions only that John crossed the Channel to inspect his abbey’s properties, was received by Edward with great honour, and that he obtained a grant of additional lands.66 One very likely political angle to the visit stemmed from John’s being in touch with the exiled archbishop of Canterbury, Robert of Jumièges, with whom he and many others attended the display of the relics of St Denis at Saint-Denis in June 1053, and who consecrated two churches at Fécamp and conferred priestly orders there between the time of his exile and his death in 1055. John might well have been seeking his restoration.67 Given his status in 1054 as 64 On the earldoms, see most recently Stephen Baxter, The Earls of Mercia: Lordship and Power in Late Anglo-Saxon England (Oxford, 2007), esp. chapter 3. 65 Lucien Musset, ‘Les contacts entre l’église normande et l’église d’Angleterre de 911 à 1066’, in Foreville (ed.), Les mutations socio-culturelles, 67–84, at 76; NMProsopographie, 108, note 82. 66 Neustria Pia, ed. A. du Monstier (Rouen, 1663), 223. There is a slightly different shorter text in Jean Mabillon et al., Annales ordinis sancti Benedicti, 6 vols. (Paris, 1703–39), iv, 503. 67 De detectione corporum sanctorum Dionysii, Rustici et Eleutherii, RHF, xi, 467–74, at 473; cited in Licence, ‘Robert of Jumièges’, 314, note 13, with the arguments for dating the detectio to 1053 at 318–23. For the consecrations and the conferment of priestly orders, ‘Fragments

t h e m a k i n g o f a r e p u tat i o n

147

abbot of Saint-Bénigne of Dijon, as well as of Fécamp, and his recent mission as a papal legate, this seems very likely.68 The support that the Norman Church – and presumably William – was giving to two bishops persecuted by the secular power, namely, Archbishop Robert and Bishop Gervase of Le Mans, must have been another distinct advantage for William in the mind of Pope Leo IX and his successors. If the subject of the English kingdom’s future was discussed during Abbot John’s visit to England, he would have learnt that Ealdred, bishop of Worcester, had recently been dispatched on a mission to the imperial court to attempt to make contact with Edward the Exile, Edward the Confessor’s nephew and the son of King Edmund Ironside. The return to England from Hungary in 1057 of the king’s nephew surely provides an insight into the future that many in England were envisaging. Securing Edward the Exile’s return had involved intense efforts. According to John of Worcester’s elaboration of the basic account in the Anglo-Saxon Chronicle, Ealdred’s visit to the Empire involved a reconnaissance trip that lasted for a year.69 Harold’s visit to Flanders in 1056, perhaps en route to or from Rome, may have been part of the same process. Although pure hypothesis, it is quite possible that Harold travelled on from Saint-Omer, where he appears in the diploma of 13 November, to the imperial court at Cologne, and then on to Regensburg and, perhaps, to Hungary.70 If so, he must also have been on the margins of events that shifted the balance of power across western Europe; the emperor Henry III had died on 5 October 1056 and the regents acting for his young son Henry IV rapidly brought the war with Count Baldwin to an end. The discussions at Saint-Omer are likely to have been very interesting since Harold was in the company there not just of Count Baldwin and Count Guy, but of Eustace, count of Boulogne, and Guy, archdeacon (and from 1058, bishop) of Amiens, all of whom had an intimate interest in what William was doing, or would develop one; Bishop Guy was of course the later author of the Carmen de Hastingae Proelio.71 The surviving diploma that provides this information, once condemned as a forgery, is now inédits de chroniques de Fécamp’, in Lucien Musset, ‘Notules fécampoises’, BSAN, liv (1957–8), 584–98, at 596. 68 In general, Licence, ‘Robert of Jumièges’, 323–9. 69 JW, ii, 574–7. 70 Grierson, ‘A Visit of Earl Harold to Flanders’, 96. 71 The diploma is Gent, Rijksarchief te Gent, Sint-Pietersabdij, charter no. 133, and has not been printed since Chartes et documents de l’abbaye de Saint-Pierre au Mont-Blandin à Gand, ed. Auguste van Lokeren (Gand, 1868), 95–6. What follows is based on an extensive report on recent scholarship supplied by Els de Paermentier. I am also grateful to Benjamin Pohl for supplying me with a photograph of the diploma and for advice. The significance of this possible meeting is mentioned by Frank Barlow in Carmen, lii.

148

william the conqueror

generally regarded as a slightly later rewriting of an authentic diploma (Plate 2).72 On the basis of the script and its similarity to an original diploma, dated 1070, of Abbot Folcard of Saint-Bertin, who is named as present at the confirmation of the 1056 diploma (ego quoque Folcardus monachorum ultimus interfui et notavi), it is reasonable to think that the surviving document was written in the 1070s. It is notable, too, that even the scholar who argued it was a forgery accepted the signa as authentic.73 In other words, the meeting of 13 November took place, but the abbey of SaintBertin may have made later changes to the text relating to the advocacy of Harnes, which was the responsibility of Count Eustace of Boulogne.74 The diploma is an interesting prelude to Harold’s 1064 visit to Ponthieu and Normandy, all the more so because the English succession must have been high on Harold’s agenda at Saint-Omer, due to the embassy he was apparently undertaking. The returning Edward’s knowledge of England and the English language must have been rudimentary. And for all that he was the son of a former king, the fact that he belonged to a family excluded in both 1016 and 1041 might have been a consideration present in some minds.75 His death on 19 April 1057 soon after his arrival in England, and before he had even met with King Edward, evoked from the author of the ‘D’ Chronicle the retrospective comment that it ‘was a miserable fate and grievous to all this people that he so speedily ended his life after he came to England, to the misfortune of this poor people’.76 By the early twelfth century, probably either on the basis of this comment or of a similar one in a lost version of the Chronicle, both John of Worcester and William of Malmesbury believed that the king had intended to make Edward the Exile his heir.77 After 1057, his family, including his young son Edgar, whose throne-worthiness was 72 Georges Declercq, ‘Van privaatoorkonde tot vorstelijke oorkonde. De oorkonden van de eerste graven van Vlaanderen, inzonderheid voor de Sint-Pietersabdij te Gent (10de– 11de eeuw)’, in N. Barré, Th. de Hemptinne, and J.-M. Duvosquel (eds.), Chancelleries princières et Scriptoria dans les anciens Pays-Bas Xe–XVe siècles / Vorstelijke kanselarijen en Scriptoria in de Lage Landen 10de–15de eeuw (Brussels, 2010), 41–77, at 58; Jean-François Nieus and Steven Vanderputten, ‘Diplôme princier, matrice de faux, acte modèle. Le règlement d’avouerie du comte Baudouin V pour Saint-Bertin (1042) et ses réappropriations sous l’abbatiat réformateur de Lambert (1095–1123)’, The Medieval Low Countries, i (2014), 1–60, at 23. 73 O. Oppermann, Die älteren Urkunden des Klosters Blandinium und die Anfänge der Stadt Gent (Utrecht, 1928), i, 35–36, 152–3. 74 For Harnes, Tanner, Families, Friends and Allies, 97–8. 75 These events have been much discussed. See in particular, Barlow, Edward the Confessor, 214–19; idem, The Godwins, 52–9; Nicholas Hooper, ‘Edgar the Ætheling: Anglo-Saxon Prince, Rebel and Crusader’, ASE, xiv (1985), 197–214, at 201–3; Stafford, Queen Emma and Queen Edith, 268–9; Baxter, ‘Edward the Confessor and the Succession Question’, 96–105; Emily A. Winkler, ‘1074 in the Twelfth Century’, ANS, xxxvi (2014), 241–58, at 243–4. 76 ASC, ‘D’ 1057. 77 JW, ii, 582–3; GR, i, 416–17.

t h e m a k i n g o f a r e p u tat i o n

149

recognized with the title ‘Ætheling’, remained at the king’s court; he was presumably among the young children on whom Queen Edith is said to have showered maternal love.78 Edgar’s appearance in an entry in the Liber Vitae of New Minster, Winchester, which seems to identify him, King Edward, and Queen Edith as a special royal trio, might be evidence that he was being thought of as the king’s heir.79 In addition, in a somewhat garbled passage, the late eleventh-century chronicler of the history of the abbey of Saint-Riquier in Picardy, Hariulf, wrote that Harold had broken his oath to support the succession of King Edward’s cousin Ælfgar. If Ælfgar was a mistake for Edgar, a possibility heightened by our knowledge that Hariulf ’s text derived only from eighteenth-century copies of an autograph manuscript destroyed by fire, then we may have more evidence for Edgar’s prominence; Hariulf ’s abbot did after all visit the English royal court both before and after 1066.80 In its annal for 1066, the ‘D’ Chronicle referred to kingship as Edgar’s ‘natural right’. For all this, Edgar scarcely appears in the surviving royal diplomas of the last decade of Edward’s reign. Although he may not have been more than five years old in 1057, the contrast with the young William’s prominence in Norman ducal diplomas at the same age is striking. In addition, the title ‘Ætheling’ conferred only the status of a prince of the royal blood, and none of the ritualized paraphernalia of prior designation is observable.81 William of Malmesbury’s belief that Edward’s offer of the succession to William came after Edward the Exile’s death may well be an intelligent deduction, albeit an erroneous one; namely that, because the situation was so unclear, this must have been the point at which Edward despaired of an acceptable successor being found from within the English or the Danish elites.82 Alternatively, the problems may have been that the subject of the succession was distasteful to Edward or that no one in England could agree on a line that was acceptable to all. Picking over the wreckage in the immediate aftermath of 1066, the author of the Vita Ædwardi Regis has the king experience a vision of St Peter in which the Apostle informed him that ‘The kingdom of the English belongs to God; and after you He has already provided a king Vita Ædwardi, 24–5. The Liber Vitae of the New Minster and Hyde Abbey, Winchester, ed. S. D. Keynes, Early English Manuscripts in Facsimile, xxxi (Copenhagen 1996), fo. 29r. See further Baxter, ‘Succession Question’, 98–103; Stafford, Queen Emma and Queen Edith, 269. 80 Hariulf, Chronique de l’abbaye de Saint-Riquier, 241; cited in van Houts, ‘The Norman Conquest through European Eyes’, 845–6 (History and Family Traditions, chapter 8). See also Regesta, no. 159. 81 David N. Dumville, ‘The Ætheling: A Study in Anglo-Saxon Constitutional History’, ASE, viii (1979), 1–33, at 12–13, 33. 82 GR, i, 416–17. 78 79

150

william the conqueror

according to His own will.’83 If – as we must – we are to try to understand the psychology, collective and individual, of the English who had to come to terms with William’s victory after 1066, there must surely have been those who thought it might have been better to have relied less on the Almighty. One likely reaction was that the relative peace of the last decade of Edward’s reign was a reflection on a land well governed by competent people. The uncertainties about the succession were an issue to be resolved when it became necessary. For others, these years might have seemed a time of indecision when a complacent elite sleepwalked into a terrible disaster. NORMANDY AND NORTHERN FRANCE, 1057–60 After the Battle of Varaville, Jumièges and Poitiers become almost silent about the events of the remainder of the decade. The former was in the process of stopping writing. For the latter, perhaps, none of the events fitted his rhetorical and historical purposes; a campaign such as the occupation of the castle of Thimert ultimately led nowhere and involved no remarkable deeds. Because some of the events of these years directly affected his abbey of Saint-Evroult, Orderic’s Historia Ecclesiastica assumes a central role as a source. Writing as he was from the perspective of his and his abbey’s vested interests, rather than as the proclaimed dispassionate commentator on great events that he announced he was going to be at the end of Book III of the Historia Ecclesiastica, the hostility at the heart of his treatment of many of William’s actions introduces a new dimension to the discussion. William was at his palace at Fécamp for Easter 1058, where he received Bishops Lescelin of Paris and Fulk of Amiens on an official peace mission from King Henry I.84 It is possible that the return of the castle of Tillièressur-Avre, which Henry had continued to hold since the early 1040s, took place at this time, since according to Orderic it was at some point handed back as a gesture of reconciliation.85 Although any time between 1058 and 1060 is theoretically possible for this, the fact that William led an army well beyond the Norman frontier in the summer of 1058 and took over the castle at Thimert, close to Châteauneuf-en-Thimerais, some twenty kilometres south of Tillières, suggests that he was by then in control of Tillières, which would otherwise have disrupted his lines of communication. At Vita Ædwardi, 14–15. Musset, ‘Fragments inédits de chroniques de Fécamp’, 596. The mission must date to 1058 because Bishop Fulk II of Amiens died that year, Bauduin, La première Normandie, 257. For the identification of Bishop Fulk, Bates, ‘Lord Sudeley’s Ancestors’, 38. 85 GND, ii, 152–3; Lemoine-Descourtieux, La frontière normande de l’Avre, 81. 83 84

t h e m a k i n g o f a r e p u tat i o n

151

around this time too, Gilbert Crispin, the man who had tried to defy Henry against the wishes of William’s guardians during the duke’s adolescence, was reinstated at Tillières. Gilbert Crispin then proceeded to follow a policy of expanding his interests and influence across the frontier, attitudes that replicated those of William and others at this time.86 Henry I reacted by besieging Thimert, which lay at the crossing of roads linking Chartres and Le Mans to Normandy, at a date between 29 June and 15 August 1058. It was still holding out against him when he died two years later.87 It would therefore appear that whatever peace negotiations had taken place at Fécamp had accomplished little, if anything. Other aggressive campaigns by William along Normandy’s southern frontier were a feature of these final years of the decade. The region between Rouen and Paris, the so-called Norman Vexin, was militarized in a way that had not previously been the case. At some point the castle of Neuf-Marchéen-Lyonnais was transferred into the hands of Hugh de Grandmesnil and Gerald dapifer, and William Crispin, the brother of Gilbert, took on responsibility for the castle at Neaufles. The first transfer involved the removal of the previous holder, and the second, the apparent takeover of land that in the 1030s had been the property of the archbishopric of Rouen.88 The narrative on which the history of William Crispin and Neaufles is based identifies his main enemy as Walter, count of the Vexin, the eldest son of Count Drogo (Duke Robert’s companion on the pilgrimage on which both of them had died) and of Edward the Confessor’s sister Godgifu.89 Walter’s presence among William’s enemies is an eloquent commentary on the way in which the peace-weaving networks of the times of Dukes Richard II and Robert had unravelled in the 1050s. The evidence of the signa of the French king’s charters shows that the Battle of Varaville did not change the basic shape of politics in northern France. Henry’s allies from earlier wars continued to support him; Ralph, count of Valois, Count Walter’s cousin, for example, was with Henry at the siege of Thimert, as was Rotrou, count of Perche. Count Guy of Ponthieu, in spite of his agreement with William, also continued to witness the charters of

Bauduin, La première Normandie, 237–42. For the chronology of events at Thimert, R. Merlet ‘Du lieu où mourut Henri Ier roi de France, le 4 août 1060’, Le Moyen Age, 2nd ser., vii (1903), 203–9, at 206–9. 88 OV, ii, 130–1; ‘Quo B. Maria subvenit Guillelmo Crispino seniori’, PL, cl, cols. 735–44, at col. 737. 89 See further Judith A. Green, ‘Lords of the Norman Vexin’, in John Gillingham and J. C. Holt (eds.), War and Government in the Middle Ages (Woodbridge, 1984), 47–61, at 49–50; Bauduin, La première Normandie, 273–8. 86 87

152

william the conqueror

the French kings.90 Nigel the vicomte was also with Henry’s forces.91 The previous holder of Neuf-Marché-en-Lyonnais is said by Orderic to have been deprived of it for trivial reasons, a statement that must be taken as an indication of William’s ruthlessness in responding to the pressures around this frontier.92 William’s advance to Thimert resembles strategically the earlier establishment of the outpost at Ambrières; the intention therefore must have been to close down a route by which his enemies could gain access to Normandy. This frontier region between the Dreugesin and the Evrecin, like that around Alençon and Domfront immediately to the west, was a notoriously unsettled place where many cross-border lordships were developing. In this particular case, William could easily have seen the recent construction of the castle at Thimert by Guazo, a vassal of both King Henry and Hugh Bardolf, the latter known to be a longstanding enemy of William, as a provocation.93 That Guazo held lands within Normandy in the Evrecin at Croth, in close proximity to Thimert, may well have increased the sense that his activities presented a threat.94 The attack may also have been a logical forward move after the establishment of William fitz Osbern at nearby Breteuil and possibly that of Roger de Beaumont at Ivry, the second of which may have happened at around this time.95 Although we know little about the events of the siege of Thimert, William’s strategy must have been to keep the garrison regularly supplied and to hope that the besiegers would eventually withdraw, as Geoffrey Martel had done from Ambrières. Charter evidence shows William to have been elsewhere in Normandy in 1059; on 13 May he was at Troarn where Roger de Montgommery had founded an abbey, and at another point during the year he may have been at Bonneville-sur-Touques.96 King Henry’s strategy looks to have been more canny than Geoffrey’s had been 90 Recueil des chartes de l’abbaye de Saint-Germain-des-Prés, ed. R. Poupardin (Paris, 1909), i, no. LXII (cum in obsidione cuiusdam castri Timerias nuncupati moraretur); RHF, xi, 559; Recueil des actes de Philippe Ier, no. II. 91 Cartulaire de Marmoutier pour le Vendômois, no. CXVII (Soehnée, Catalogue des actes d’Henri Ier, no. 116). 92 OV, ii, 130–1. 93 For the recent construction of the castle at Thimert, AD Eure-et-Loir H2486 (Soehnée, Catalogue des actes d’Henri Ier, no, 115) (tunc scilicet quando obsidebat castrum Theodemerense nuncupatum in pago Carnoteno noviter a quodam Gazone constructum); Cartulaire blésois de Marmoutier, ed. C. Mètais, 2 vols. (Blois, 1887–91), nos. XXXII, XXXIII. For the first charter and for Guazo, see further André Chédeville, Chartres et ses campagnes (XIe–XIIIe siècles) (Paris, 1973), 270–1. 94 Recueil des actes de Philippe Ier, roi de France (1059–1108), ed. M. Prou (Paris, 1908), no. VIII. 95 On Ivry, Breteuil, and this zone of the frontier, Bauduin, La première Normandie, 210–12, 229–30. Roger de Beaumont’s guardianship of Ivry cannot be dated (OV, iv, 114–15, 204–7). 96 RADN, nos. 142, 144. The list of signa in no. 142, which places William at Bonneville in 1059, is incompatible with that year.

t h e m a k i n g o f a r e p u tat i o n

153

at Ambrières, since he too devoted only intermittent personal attention to the siege.97 While the siege of Thimert looks to have developed into a stalemate in which neither side was willing to back down, William’s support for the garrison must have been notably effective in order to sustain morale against an enemy determined enough to maintain a blockade for two years. Unusually, at least in the context of his later career, he seems, however, to have been unsuccessful in the propaganda war, since the castle’s garrison were placed under an anathema by the diocesan, the bishop of Chartres. In this context, William’s reaction to the unforeseeable event of the commander of his garrison at Thimert, Richard de Reviers, being struck down by what turned out to be a fatal illness, provides a precious insight into his values. Richard had gone to the bishop of Chartres for absolution, obviously crossing enemy lines in the process, and then made a substantial grant to the abbey of Saint-Père at Chartres where he was later buried. William’s confirmation of Richard’s grant, made at Courdemanche, north of Dreux, on 4 August 1060, while the siege was still continuing, and while either on his way to bring relief to Thimert or on the way back from having done so, shows how in his mind conflict had to be sublimated to the need to confirm through a charter that Richard de Reviers had made a good end.98 And who should emerge as an intermediary in all this but a man named Nigel, whom the charter identifies as Richard de Reviers’s brotherin-law. Given their status as near-territorial neighbours in the Cotentin, it is likely that this was Nigel the vicomte.99 If so, his reinstatement in Normandy by William presumably followed soon afterwards. William’s presence at Courdemanche could also indicate that he was about to open peace negotiations with King Henry. Again, however, the unpredictable intervened since Henry died at nearby Dreux on the same day, 4 August.100 Although no source says so, it looks as if William abandoned Thimert when peace was made with the representatives of Henry’s son and successor Philip I, an eight-year-old boy who was placed under the guardianship of Count Baldwin of Flanders, William’s father-in-law. Baldwin’s new responsibilities must surely have been a factor in the 97 He is known to have visited Rheims for the coronation of his son Philip on 23 May 1059 and Poissy on 4 May 1060 (Soehnée, Catalogue, nos. 117, 124). There are also references to visits to Melun, Vitry-aux-Loges, and Paris (ibid., nos. 117, 119, 120, 125). 98 For this evidence, RADN, no. 147. For Richard’s place within the Reviers family, Eric van Torhoudt, ‘L’énigme des origines de l’abbaye de Montebourg: une question de methode?’, in Bouet et al. (eds.), De part et d’autre de la Normandie médiévale, 331–46, at 339. 99 Delisle, Saint-Sauveur-le-Vicomte, 31–2, pièces justificatives, no. 29. For reservations, van Torhoudt, Centralité et marginalité, 559. 100 For this suggestion, Merlet, ‘Du lieu où mourut Henri Ier’, 208.

154

william the conqueror

peace-making. With Count Geoffrey of Anjou dying on 14 November, William may have decided that it was no longer necessary to maintain an isolated outpost against enemies who were now much weaker. Equally, he may have decided to cut his losses. The agreement ushered in a new era of non-confrontational Franco-Norman relations that lasted for over ten years. To the west of the region of Tillières and Thimert, William also besieged Robert fitz Giroie’s castles at Saint-Céneri-le-Géré and La Roche-Mabille, in all probability after 21 June 1059, the date on which he had agreed to the appointment of Robert’s nephew Robert de Grandmesnil as the new abbot of Saint-Evroult.101 It is striking that William was for a time deliberately fighting wars on two fronts, the calculation presumably being that neither siege required a massive deployment of resources. The Giroie family, co-founders in 1050 of Orderic’s abbey of Saint-Evroult, held most of their lands in the turbulent region where the lands of Geoffrey de Mayenne, the lords of Bellême, and the duke of Normandy met. Apparently protégés of the Bellême family, the Giroies had fallen out with them in the 1040s when Robert’s brother, William fitz Giroie, had protected Geoffrey de Mayenne, and had, as a result, been ambushed and savagely mutilated on the orders of the then lord of Bellême, William Talvas. The Giroie family had also been cultivated by Duke Richard II, who had granted Echauffour and Montreuil-l’Argillé in southern Normandy to Robert’s father, the objective presumably being to strengthen his power in the contested region of the Hiémois by forging links with a major crossborder family. William in turn sought to reinforce the alliance with Robert by giving to him in marriage a kinswoman named Adelaide; this would seem to have taken place in the 1050s, since a son of the couple was active at King Henry I’s court in the early twelfth century.102 Strategically, this marriage mirrors William’s earlier attempt to woo Count Enguerrand of Ponthieu with a marriage to his sister Adelaide, and his half-sister Muriel’s marriage to Eudo, son of Thurstan Haldup. It again illustrates that the formation of networks of association was every bit as much a part of William’s socio-political repertoire as violence (or the threat of it). The Giroie family’s central role in the establishment within Normandy of the abbey of Saint-Evroult in 1050 must indicate where their primary loyalties lay at that date. Caught in the middle of the warfare between the dukes of Normandy and the counts of Anjou, Robert fitz Giroie’s sense of how his interests were best served seems to have changed in the aftermath of William’s 101 The basic narrative appears in OV, ii, 28–9, 78–81. On the date of the siege, OV, ii, 79, note 3. 102 OV, ii, 28–9.

t h e m a k i n g o f a r e p u tat i o n

155

attack on Geoffrey de Mayenne and the associated advance to Ambrières; Robert apparently took the Angevin side in the Varaville campaign. An additional, and perhaps decisive, pressure on him could have been Roger de Montgommery’s advance in the late 1050s into the region around Alençon and Sées; Roger’s marriage to William Talvas’s daughter Mabel also made him in certain respects the heir to the feud between the Giroies and the Bellêmes. A further complication was the rapprochement at the same time between Bishop Ivo of Sées and William, which dates from the later 1050s, and would have heightened further Robert’s sense of increased encirclement by potentially, and at times, actually, hostile powers.103 However, the inclusion of Saint-Céneri in the duchy after the settlement that followed the Battle of Mortemer, as Orderic appreciated, converted what had been Robert’s earlier support for a non-Norman lord in a frontier war involving William into an act of rebellion against him, and made it important for William to take action.104 The sieges of Saint-Céneri and La Roche-Mabille were brought to an end by Robert’s death on 6 February 1060. Orderic has a story that he was inadvertently poisoned by his wife; this fatality, one of the least convincing of Orderic’s several tales of supposed poisoning, may not be anything more than a sad demise brought on by eating rotten apples.105 Orderic then tells us that, after a display of fury, accompanied by a threat to deny Robert a Christian burial, William allowed him to be buried at SaintEvroult and his nephew, Arnold d’Echauffour, holder of the family’s Norman lands, to succeed his uncle elsewhere. The striking feature of Orderic’s two accounts of this episode is nonetheless his use of the language of irrational anger to describe William’s behaviour and his statements that William’s conduct was discriminatory and partisan, above all to the benefit of Roger de Montgommery.106 At around this time William also brought about a major change in the south-west of the duchy, replacing William Werlenc, count of Mortain, with Robert, his other half-brother by Herleva and Herluin de Conteville. Orderic tells a story of Count William promising rewards to one of his household warriors, of the tale getting back to the duke, and of William condemning Count William to perpetual exile because he intended to disturb the peace of Normandy by endowing one of his warriors with 103 Thompson, ‘Lordship of Bellême’, 223; ‘The Example of the Montgomerys’, 256–7, 261–2. 104 Note that Orderic says that Robert rebelled against William (contra Willelmum ducem rebellauit), OV, ii, 78–81. On this subject, Louise, Seigneurie de Bellême, i, 316–17; Maillefer, ‘Une famille aristocratique’, 186–8. 105 On poisoning, Douglas, WC, 413–14. 106 Bates, ‘Anger, Emotion and a Biography of William the Conqueror’, 26–8.

156

william the conqueror

plunder.107 In his later version of the episode in the Historia Ecclesiastica, Orderic described the reasons for Count William’s banishment as trivial.108 A son of Archbishop Malger, count of Corbeil on the river Seine to the south of Paris, and therefore a grandson of William’s great-grandfather Duke Richard I, William Werlenc is an almost totally obscure figure in Normandy’s history. Although he probably acquired Mortain and its region in the time of William’s father, perhaps as an aspect of the good relations between Duke Robert and King Henry I, he appears in only one surviving ducal charter, which, since it also includes Nigel the vicomte among its signa, must date to c.1054.109 William Werlenc had taken no known part in the wars of the 1040s and 1050s. He could be said to fall into the category of ‘les Richardides’, but there is no evidence that he despised William or did anything to harm him. Like many before him, William Werlenc departed for southern Italy where he appears to have moved in the most exalted social circles; his daughter became the second wife of Roger, count of Sicily. The earliest certain date at which William’s half-brother Robert appears as count of Mortain is 1063. But since he appears in some charters that can be dated to either the 1050s or the 1060s, and given the re-dating of the one charter in which William Werlenc appears, it is as good as certain that he replaced him in the second half of the 1050s.110 William Werlenc’s removal must thus surely be seen as the last act in that sweeping away of an older generation which had been a characteristic of the 1040s and 1050s. Robert’s promotion also involved the subsequent extension of his landholdings northwards into the Cotentin, making him a dominant figure in western Normandy. At some point he was married to Matilda, a daughter of Roger de Montgommery and Mabel de Bellême, a predictable coda to the tightening of networks joining the frontier periphery to the core of ducal rule and, above all, to the creation of personal networks that connected those with regional influence to William himself. While the religious patronage of previous counts of Mortain had been directed towards the Loire valley, something that did not completely change with Robert, attendance at court and his responsibilities at Grestain pulled him much more closely than his predecessors into the apparatus of ducal rule.111 GND, ii, 126–9. pro minimis occasionibus de Neustria propulerat, OV, iv, 98–9. See also ibid., ii, 312–13. 109 RADN, no. 162. The above discussion of Nigel the vicomte’s exile demonstrates that Marie Fauroux’s suggested dating limits of 1050x1063 need to be drastically revised. 110 For Robert in 1063, RADN, no. 156. On the date of his appointment, Brian Golding, ‘Robert of Mortain’, ANS, xiii (1990), 119–44, at 120; idem, ‘Robert, Count of Mortain (d.1095)’, ODNB. 111 Cassandra Potts, ‘The Earliest Norman Counts Revisited: The Lords of Mortain’, HSJ, iv (1992), 23–35, at 29–31; Golding, ‘Robert of Mortain’, 141–4; idem, ‘Robert, Count of Mortain’, ODNB. 107 108

t h e m a k i n g o f a r e p u tat i o n

157

Robert’s presence also began a southwards push involving the establishment of castles in the lands between the duchy of Brittany and the county of Maine, which replicates the strategy evident along the whole of Normandy’s borders during the 1050s.112 In relation to Brittany, although, as we have seen, at least one Breton, Main de Fougères, had granted lands in Normandy to the abbey of Marmoutier in the first half of the 1050s, his presence there was probably the result of a long-established cross-border landholding and therefore of legally required routine.113 Count Eudo’s alliance with Geoffrey Martel in the warfare around Ambrières in the mid-1050s notwithstanding, the general history of northern Brittany during the 1050s was an introspective one, dominated by the increasingly tense rivalry between Count Conan II and Count Eudo. Conan’s imprisonment of his uncle in 1057 may have given him a new freedom and generated a new dynamic that Count Robert’s appointment was designed to counter. THE CAEN ABBEYS Although the history of William and Matilda’s two monastic foundations at Caen can be traced from a large number of surviving charters, it is only in Orderic’s interpolations into the Gesta Normannorum Ducum that their establishment is described as a penance for consanguinity, with William of Malmesbury a little later telling much the same story.114 In mid-eleventhcentury terms, a princely penance of this kind was quite common, a developing practice of which William and Matilda would have been aware. It was also an act of good rulership, cleansing their marital status in relation to God and enhancing the moral and spiritual welfare of their polity; Orderic’s narrative also makes it clear that papal approval to the foundations was given and that this wiped the slate clean.115 A penitential act along these lines is also not out of keeping with the only strictly contemporary Norman comment on the marriage, William of Jumièges’s statement that it had been entered into legally.116 In assessing this comment, it is important to keep in mind that he was capable of being critical of William. Daniel Pichot, Le Bas-Maine du Xe au XIIIe siècle: étude d’une société (Laval, 1995), 136–7. Van Torhoudt, Centralité et marginalité, 525–31, 756–60; cf. Michel Brand’Honneur, Manoirs et châteaux dans le comté de Rennes: Habitat à motte et société chevaleresque (XI e–XII e siècles) (Rennes, 2001), 109–17, where William’s conduct is seen as having been more aggressive. 114 GND, ii, 146–9; GR, i, 494–5. 115 On this topic Sarah Hamilton, The Practice of Penance, 900–1050, Royal Historical Society Studies in History, new series (Woodbridge, 2001), 174–90. On the practice of founding monasteries as penance for an incestuous marriage, Corbet, Autour de Burchard de Worms, 296–300. 116 . . . illam sibi despondit iure coniugali, GND, ii, 128–31. 112 113

158

william the conqueror

His careful choice of words, written as they were was in the later 1050s, must mean that the marriage had long been regarded by most people as a legitimate one. The couple were now simply taking appropriate measures to safeguard their souls and thereby those of their subjects. Given that the marriage had finally taken place in 1052–3, the amount of time taken to found the monasteries might appear surprising. Although the evidence for the exact dates of the two foundations is somewhat obscure, it points to c.1059 for Matilda’s La Trinité and, as we shall see, less persuasively, to c.1063 for Lanfranc’s appointment as abbot of William’s Saint-Etienne.117 However, since these are only the possible dates at which the first abbess and abbot were appointed, the sites must have been chosen long before. The sheer scale of the work involved in their construction surely suggests that it had begun some years before those dates. On the basis of the narrative contained in the Vita Lanfranci, written in c.1140, to the effect that Lanfranc negotiated the ending of the papal prohibition of William and Matilda’s marriage (in return for their agreement to found two monasteries), when he attended the papal council whose proceedings began in Rome on 13 April 1059, it has often been accepted that this was the crucial moment that regularized the marriage.118 However, in a letter whose tone is notably friendly, from Pope Nicholas II to Lanfranc, possibly sent very soon after the pontiff ’s election in December 1058, Nicholas tells Lanfranc that he is sending two papal chaplains to him at Le Bec for education in dialectic and rhetoric. He suggests that Lanfranc might like to visit him, but acknowledges that this might be difficult. He indicates he has heard that Duke William, whom he describes as a friend (amicum nostrum), follows Lanfranc’s advice in all matters, that he has great confidence in William and in the benefits he derives from Lanfranc’s company and counsel.119 The interpretation of this letter has become an aspect of a controversy about Lanfranc’s role in the continuing proceedings against Berengar of Tours and about whether Lanfranc attended the papal council at all.120 As long as the letter was written in January 1059 at the latest, there was time for Lanfranc to take the decision to go to Rome. However, the verbal similarities between the Vita and the tract written on Lanfranc’s behalf in c.1063, the De corpore et sanguine Domini, supposedly to refute Berengar’s opinions, suggest that the Vita’s account was an For these dates, see AAC, 13–14. ‘Vita Lanfranci’, 677–8. 119 For a modern edition of the letter, R. W. Southern, Saint Anselm: A Portrait in a Landscape (Cambridge, 1990), 32–3. 120 For this controversy, Gibson, Lanfranc, 69; Bates, Normandy before 1066, 199–201; Southern, Saint Anselm, 19–28; Cowdrey, Lanfranc, 36–7, 42. 117 118

t h e m a k i n g o f a r e p u tat i o n

159

imaginative expansion of that text.121 Its statement that Normandy had been under an interdict for ten years should undoubtedly therefore be rejected; its view of ecclesiastical sanctions is very much that of twelfthcentury canon law and the very idea conflicts with what we know of Norman–papal relations in the 1050s. It is also crucially important that Berengar’s response to the council’s decision, a text preserved only within the De corpore et sanguine Domini, clearly implies that Lanfranc was not at the council.122 As far as William’s relations with the papacy are concerned, the foundation of the two Caen abbeys is probably not be the great turning point that was once suggested, the event that secured approval in advance of his bid for the English kingdom. Not only had relations between William and the papacy been consistently good throughout the 1050s, but Lanfranc’s role as William’s adviser, and his network of contacts and pupils, acquired during his long sojourn at the papal court in 1050 and through the school of Le Bec, constituted a secure reinforcement of those relations; the school’s activities clearly expanded greatly in the later 1050s and, as Nicholas II’s letter shows, it was receiving students recommended by the pope and from several regions of western Europe.123 In all likelihood the penance strengthened a continuum and drew a line under any outstanding sense of irregularity; so viewed, the Vita Lanfranci’s narrative may be making an important point, but certainly not on the terms in which the situation would have been understood in the middle of the eleventh century. The dual role of William and Matilda in the two foundations is also the clearest expression so far of her importance as a marriage partner. Her appearances in charters have been interpreted as showing an enhanced prominence that anticipates the role she was to play in 1066 and afterwards.124 In Norman terms – there are of course countless other examples from elsewhere in the contemporary medieval West – for William to build the two abbeys at Caen was also to follow in the footsteps of his father and grandfather: Duke Robert had founded a monastery at Cerisy and a nunnery at 121 On the authorship of the De sanguine et corpore Domini, Toivo J. Holopainen, ‘ “Lanfranc of Bec” and Berengar of Tours’, ANS, xxxiv (2012), 105–21, at 106–8. 122 Beringerius Turonensis Rescriptum contra Lanfrannum, ed. R. B. C. Huygens, 2 vols., CCCM, 84 (Turnhout, 1988), i, 35 (manu, quod mendaciter ad te pervenit). For a clear statement of the argument against Lanfranc being present, de Montclos, Lanfranc et Bérenger, 43, note 5; Gibson, Lanfranc, 69. 123 For the most recent appraisal of Lanfranc’s pupils, Cowdrey, Lanfranc, 20–4. See also Sally N. Vaughn, ‘Lanfranc at Bec: A Reinterpretation’, Albion, xvii (1985), 135–48, at 136–9. 124 For Matilda’s appearances in original diplomas, RADN, nos. 124, 126, 148, 158, 193, 199, 204, 205, 210, 218 (version A); Charlotte Cartwright, ‘Before She Was Queen: Matilda of Flanders and the Use of comitissa in Norman Ducal Charters’, HSJ, xxii (2012, for 2010), 59–81, at 72–81.

160

william the conqueror

Montivilliers, and Duke Richard II had developed the great monastery at Fécamp in close proximity to the ducal palace there. Although Michel de Boüard’s excavations in the 1950s and 1960s revealed no traces of any occupation prior to William’s time, this does not mean either that there had not been an earlier settlement on the rocky outcrop on which the castle was constructed or that it was uninhabited before building work started. Recent arguments in favour of the southern parts of the site, namely, the parish around the church of Saint-Georges, being inhabited, based on a comparative analysis and archaeological and textual analysis of Caen’s history from the early Middle Ages onwards, suggest an evolutionary development of what may well already have been a ducal residence. The confidence of William’s architects in building the huge stone enclosure around the whole of the rocky outcrop on which the castle stands also suggests some prior settlement.125 Although Wace famously said that Caen was not fortified at the time of the Battle of Varaville, this probably means that extensive building in stone had not started at that date; the excavations in 2005 on the north-western section of the ramparts, the area where William’s residence was built, indicate that the stone wall around the site was built at that time.126 It was also, of course, the most convenient place possible for the crossing to England and, specifically, to Winchester. Was this included in the symbolism that the enhancement of the site was intended to convey? On this subject, like many others, it is important not to believe that the eventual result automatically reflects the original intention. Although the scale of Caen’s development over the following decades truly was to be astonishing, the initial works may have been less ambitious. In due course the town’s development did refocus the political life of the Norman court; William’s regular visits to Caen made an enormous difference in terms of the physical immediacy of ducal power, widening the orbit of political life beyond the largely Upper Norman palace-bound itinerary of his predecessors.127 A regional strategy more extensive than just the development of Caen is also observable, in, for example, the association of the foundation of the abbey of Troarn to the east of Caen by Roger de Montgommery with the transfer of property in the region of Dives-sur-Mer from the Upper Norman abbeys of Fécamp and Jumièges.128 Even if, as currently 125 Laurence Jean-Marie, ‘Des maisons au château: le château de Caen, une paroisse urbaine’, in Lalou, Lepeuple, and Roch (eds.), Des châteaux et des sources, 377–88, at 378–80; Le Maho, ‘Nouveaux aperçus sur les origines de la ville de Caen’, 186–97. 126 Roman de Rou, pt. 3, lines 5,171–2; Gaël Carré, ‘Le rempart nord-ouest du château de Caen (XIe–XVIe siècles). L’apport récent de l’archéologie du bâti’, AN, lix (2009), 3–25, at 7–13. 127 For this, Bates, Normandy before 1066, 151–2. 128 Carpentier, Guillaume le Conquérant et l’estuaire de la Dives, 58–60.

t h e m a k i n g o f a r e p u tat i o n

161

understood, the ducal/royal buildings of the castle from William’s time are rather unimpressive, building work must have been continuous at Caen for the rest of his and Matilda’s lives. And their disbursement of funds, especially evident in Matilda’s involvement in the development of La Trinité’s lands, must have fuelled both prosperity and growth. THE MAKING OF A REPUTATION A charter from the great Loire valley abbey of Marmoutier contains the statement that William was ‘the ruler of his whole land, something which is scarcely found anywhere else’. Seemingly dating from late 1055, it underlines just how powerful a prince William was perceived to be by that date, let alone five or ten years later.129 For all that violence was structured into socio-cultural mores, it does suggest a remarkable capacity on his part to keep the peace. Furthermore, William’s wars and political manoeuvring around Normandy’s borders in the 1050s had also been so successful that several references of an unprecedented kind appear in charters in the 1060s and 1070s that demarcate Normandy from neighbouring territories, a sign arguably of an identity that seemed fixed in ways that had not previously existed.130 The sheer force of the strategic impetus behind the consolidation of the lands around the duchy’s borders is a telling commentary on sustained willpower. The defeat of William of Arques and the deposition of Archbishop Malger, both of which occurred at the start of the period covered by this chapter, show how Norman political society and religious opinion as expressed by the papacy were coalescing around their nephew William to shape a formidable apparatus of power. The narrative sources provide us with a portrait of power in operation that often feels both frightening and unscrupulous. If William actually made the threat to deprive Robert fitz Giroie of a Christian burial, it must have been truly shocking to those who heard it. The wait until William relented, as he may well have intended to do all along, must have seemed endless, and not just to friends and relatives. The massacre of Varaville was an act of utter ruthlessness on his part – justifiable in war to achieve the 129 Ut expressius dicatur quod difficile in aliis reperies, totius terre sue regem, RADN, no. 137. Marie Fauroux suggested a date of September or October 1055 on the grounds that the signa include Archbishop Maurilius of Rouen, who she believed was appointed in September 1055, and Bishop Gervase of Le Mans, transferred to the archbishopric of Rheims in October 1055. 130 Regesta, no. 196 (Qui locus habens ad orientem Franciam, ab occidente uero Normanniam); ibid., no. 251 (in confinio Normannie et Britannie sita); Thomas Stapleton, ‘Observations on the History of Adeliza, Sister of William the Conqueror’, Archaeologia, xxvi (1835–6), 349–60, at p. 358 (in externis partbus Normannie super flumen quod Augus dicitur in ea parte que diuidet Ambianensem prouinciam a terra Normannorum).

162

william the conqueror

required result with minimal loss of life, but chilling in its calculation. Nigel the vicomte’s return and reinstatement, after he had presumably fought alongside William’s enemies, were a sign of the magnetic pull of ancestral lands and of conventions applied; as we have seen, an exile had prospects of return, but, with William, the restoration of his favour could so often be qualified that there must have been many whose thoughts about him were distinctly ungenerous. This factor, Orderic’s account of the affairs of Robert fitz Giroie, and the special treatment that William afforded Roger de Montgommery, are remarkable narratives of favour redistributed and withheld, signs of political intelligence at work, and a capacity to assess individuals’ strengths and weaknesses and to act accordingly. William was also someone who seems already to have harboured a special respect for the spiritually minded and the learned religious, and who supported the institutional Church in many ways. The series of popes who ruled during the 1050s were very consistent in their approval of, and assistance to, him. Reading between the lines of the Gesta Guillelmi, it is clear that two of the criticisms Poitiers was setting out to answer were that William’s treatment of prisoners taken in war and the massacre at Varaville constituted breaches of normal standards. So read, these sections of the Gesta Guillelmi anticipate the sections that deal with the Battle of Hastings and the events which followed. Jumièges’s apparent decision to discontinue writing the Gesta Normannorum Ducum in the late 1050s is an altogether more complicated subject. Why this enthusiastic author, who, when he resumed his work after 1066 with the apparent intention of continuing it on into the time of William’s son Robert, should have stopped after the Battle of Varaville, is entirely a matter for speculation.131 Personal circumstances or fresh monastic responsibilities, of which we know nothing, might explain everything. Yet there are signs that he was running out of steam in the mid1050s. After c.9 on William and Matilda’s marriage, the treatment becomes uncharacteristally patchy, with only the battles of Mortemer and Varaville meriting detailed attention, while c.11 on Maine is a muddle. It is possible, too, that the apparently misplaced c.4 on William of Arques was once a eulogy that had to be rewritten after 1053. Steeped as he was in the Norman past, Jumièges might well have wondered where events were leading. He had mostly presented previous dukes as the upholders of legitimate causes, and they had normally been the allies and co-adjutants of the French kings. He had already given vent to his difficulties in writing about aspects of William’s life in his barbed comment that the greatest of the troublemakers were now the duke’s

131

dirigere stilo decreuimus ad Rodbertum eiusdem regis filium, GND, ii, 184–5.

t h e m a k i n g o f a r e p u tat i o n

163

closest associates and that he dare not mention their names.132 In the mid1050s he could not know that the hard labour of these years was laying the foundations of unprecedented triumphs. Jumièges might well have worried that the seemingly errant William, who was making so many enemies, was jeopardizing his self-proclaimed task of writing what was in effect past and contemporary history of the exemplary deeds of the Norman dukes.133 These final speculative observations notwithstanding – and, if they are unacceptable, set aside – it is surely easy to see why William should inspire loyalty and devotion in some and fear and detestation in others. He exhibited a flamboyance and a hardness that made resistance difficult and which encouraged loyalty in its beneficiaries. This was also a time when events worked in William’s favour. The birth of several children provided him with the all-important dynastic security and reinforced Matilda’s status within their marriage and at the family’s core. The emergence of a new generation in the Church, most readily observable during this decade in the lives of Lanfranc and Bishop Geoffrey of Coutances, gave an impetus that continued the work of the previous generation. These years, superficially less spectacular than others, must surely have established William’s reputation, and they were the years that laid the basis for achievements to come. Yet the ambivalence that his actions were already provoking, something much more extensively articulated after 1066, was also emerging.

GND, ii, 92–3. Jumièges’s purpose is set out in a letter of dedication written after 1066 (GND, i, 4–7), but his whole text proclaims that to praise the Norman dukes’ exemplary deeds was always his aim. 132 133

Chapter 5

O N TO T HE AT TAC K

In his mid-thirties and, we can assume, at the height of his powers, the years between 1060 and 1066 were for William the period when possibilities that may have been long contemplated became capable of realization. As always, the unpredictability of events and the invisible history of many people’s thoughts remain a central factor in the construction of a narrative. For quite a number of reasons William may be thought to have been fortunate. But at the same time he displayed a remarkable single-mindedness in seizing opportunities as they arose. His rule within Normandy became demonstrably more expansive in ways that drew attention to his prestige, piety, and achievements. It is quite possible that William was consciously setting out the case to become a king of the English. There are apparently quiet episodes in this period – above all in 1061, 1062, and 1065 – yet the broad picture continues to be one of ruthlessness that went significantly beyond contemporary norms. William’s first conquest, the takeover of the county of Maine in 1063, is an event that, like the subsequent conquest of England, had its roots in the past, in this case going back to the 1040s, and, so William of Poitiers wants us to believe, the early tenth century. It was, however, very different from the conquest of England, a difference that was to have profound consequences for his life and for the lives of his successors. Harold’s visit to Normandy in 1064, and the oath that he swore to support William’s succession to the kingdom of the English after Edward the Confessor’s death, are at the heart of William’s propaganda case against him in 1066. The events of this visit will always remain extremely controversial, ultimately because that is how they were for contemporaries. It was certainly the moment when William’s ambitions towards England were made unambiguously clear to all. Released by the deaths of Geoffrey Martel and Henry I in 1060 from the constraints that these formidable neighbours had imposed on him, William must from this point on have been able to start planning in ways that had never been possible before. The death in 1062 of Count Herbert II of Maine led to a dramatic reshaping of politics across northern France. Like Edward the Confessor’s subsequent death in early January 1066, it was apparently sudden and immediately led to all interested parties calculating 164

o n to t h e at tac k

165

how they must act. In the case of Edward’s demise, however, given the longstanding uncertainties about the succession, it must have been a passing whose potential consequences had long been reflected on and talked about. While William and those around him may well have been speculating on possible scenarios from at least 1051–2 onwards, understanding the evolving events of the years from 1060 until 1066 requires we always keep in mind that others were thinking and planning as well, and that their thoughts would have been moving in very different directions from William’s. The election as pope in 1061 of Anselm, bishop of Lucca (1056–73), who took the name Alexander II, brought onto the stage the man who was to be pontiff throughout all the decisive events of the conquests of Maine and England and who was to play a major part in them. Although the evidence that he had been Lanfranc’s pupil is not very convincing, there were certainly strong personal links between them. Soon after his election, for example, Alexander proposed to send a nephew to Lanfranc for education in the arts; and, in 1071, when Lanfranc visited Rome to collect his pallium as archbishop of Canterbury, the pope rose from his throne to greet him as a man of special eminence and presented him with a second pallium in which he had himself celebrated mass.1 In general, Alexander displayed an interest in regime change as a means to what he regarded as religious improvement. At almost the same time that he supported William’s invasion of England, and for very similar reasons, he effectively sanctioned the deposition of Count Geoffrey the Bearded of Anjou (nephew of Geoffrey Martel). THE PRIVATE AND THE PUBLIC The near-fatal illness that William suffered during these years is mentioned only in a charter that cannot be precisely dated. Since the illness ultimately stands in bewildering isolation from other events, because no source enables us to make any connection with them, we can only speculate on the effect that it had. We may wonder, for example, whether, if the illness occurred in 1062, which is possible, it encouraged Count Walter of the Vexin to intervene in Maine, or whether it persuaded Count Geoffrey the Bearded that an intervention there was not needed since William was incapable of responding to the situation caused by Count Herbert’s death; or, if the illness struck in 1065, which is more likely, whether it gave Harold confidence that William would not be able to cross to England. The narrative we possess certainly says interesting things about William and Matilda’s 1 On the evidence for Alexander as Lanfranc’s pupil, Cowdrey, Lanfranc, 22–4. For his special treatment in Rome and for Alexander’s nephew, HN, 10–11; Lanfranc Letters, no. 3; Alexander II, Epistolae, PL, cxlvi, col. 1,353.

166

william the conqueror

relationship and the illness may well have had an effect on their religious convictions. These were certainly important years in relation to William and Matilda’s marriage and to the lives of their children. This therefore constitutes a suitable point to reflect on the interrelationship of William’s private and public personae, and to reflect on some glimpses of the private man that are available to us during this period. The information about William’s illness occurs within a narrative of early grants to the college of canons of Notre-Dame of Cherbourg, which was compiled over a lengthy period that extended into the reign of William’s son Henry I. This text now survives only in copies made in the nineteenth century from an original charter and a cartulary then extant, both of which were destroyed during the Second World War.2 The description of William’s eldest son Robert as count of Maine in the first section of the document has persuaded its two modern editors to date the section that describes William’s illness to between 1063 and 1066. However, given the way in which the text was constructed, it is possible that Robert’s title is an anachronistic later insertion, a point made more plausible by this being the only occasion on which the title of count is attributed to him in any document with claims to have been composed before 1066. Another argument, albeit a circumstantial one, for dating the illness to 1062 or early 1063 is that William and Matilda are known to have chosen Robert as William’s successor before 29 June 1063, an act surely to be associated with some sort of crisis.3 Other arguments are the apparent delay in launching the invasion of Maine after Count Herbert’s death and a charter recording a prestigious gathering at nearby La Hougue in 1062, which could have taken place during William’s recuperation.4 Late 1063 or early 1064, before the campaign into Brittany on which William was accompanied by Harold, are also possible dates for the illness, since there are gaps in William’s itinerary during these periods. But 1065 is ultimately the most likely date. Not only does it fit in with the title given to Robert, it is the longest period of apparent inactivity on William’s part that we know of during this period. Whatever dating is adopted, it is certain that William was extremely ill; so much so that he was prepared for death.5 In extremis he ordered that the church at Cherbourg be dedicated and he vowed to establish canons there, three of whom were installed soon after his recovery. He granted relics from his personal collection and added further substantial donations. The 2 RADN, no. 224; Le cartulaire du chapitre cathédrale de Coutances, no. 339, with the sad fate of the early archives of the bishopric of Coutances described at pp. 36–7. 3 RADN, no. 158. 4 Ibid., no. 151. 5 cum ab ipsa infirmitate in qua fuit, pene omnino de vita desperatus et ad terram depositus ut iam moriturus.

o n to t h e at tac k

167

narrative associates Matilda with the grants, both as witness and donor, and their two eldest sons Robert and Richard are recorded as present at their father’s sickbed. It also tells us that a distraught Matilda, her hair dishevelled, placed a token of her husband’s gift on the church’s altar so that God and Mary would return ‘her dearest husband’ to her.6 While it can be dismissed as a stereotypical monastic representation of idealized Christian marriage, taken alongside other evidence, the glimpse of Matilda’s emotions is almost certainly an important commentary on her feelings for her husband and on their successful collaborative relationship. It must be associated with her greatly enhanced prominence in the 1060s and after 1066. The fact that she had by this time borne numerous healthy children, and was therefore securely established as the wife and mother of a successful ruler and his heirs, must also have been a significant contributory factor. Whenever it happened, this illness is only one of three periods of serious poor health that William is known to have suffered. A second occurred during the march around Kent to London after the Battle of Hastings.7 And a third, recorded by Orderic, took place at Bonneville-sur-Touques at an unknown date; because this illness involved a repeat of the designation of Robert as his heir in Normandy, it would also appear to have been serious.8 While William of Malmesbury was able to write that William enjoyed good health throughout his life, never suffering from a dangerous illness, this has to be interpreted as an overall reflection on an exceptionally active life and on his basic physical strength.9 A writer close to one of William’s doctors, Abbot Baldwin of Bury St Edmunds, indicates that William took care with his health. After 1066, so we are told, Baldwin was frequently summoned to Normandy to give him advice and treatment.10 A twelfth-century English source, the Historia of the abbey of Abingdon, mentions that William was accustomed to go to the island of Andersey in the Thames to the south of Abingdon for periodic bloodletting and for relaxation and recuperation.11 Given the relatively short periods that he spent in England after 1072, this presumably indicates that similar provision was made for William throughout his lands, and therefore in Normandy before 1066. As periodic bloodlettings prevented ailments through purification and also, within monastic regulations, provided the opportunity for attending 6 Comitissa uero sparsis capillis super altare posuit comitis uagium, ut Deus et Sancta Maria redderent sibi suum carissimum maritum. 7 GG, 144–5. 8 OV, iii, 112–13. 9 GR, i, 508–9. 10 Herman and Goscelin, Miracles of St Edmund, 216–17. 11 Historia Ecclesie Abbendonensis: The History of the Church of Abingdon, 2 vols., ed. and trans. John Hudson, OMT (Oxford, 2002–7), ii, 72–3.

168

william the conqueror

to spiritual as well as physical well-being, it is possible that William used these occasions in this way as well.12 The evidence shows that William did take care of his personal well-being and that – considering the Cherbourg narrative for the moment – he knew how to prepare for a good death. It can be taken alongside William of Malmesbury’s statement that William lived as Christian a life as was possible for a layman, attending mass and hearing vespers and matins every day, thereby indicating he took care of both his physical and his spiritual welfare.13 Malmesbury also presented William’s hunting as being a form of relaxation to escape the pressures of daily business.14 These were years when the patronage of religious foundations increased, often recorded as being done in the joint names of William and Matilda. In 1063 they organized the building of the church at Bonne-Nouvelle near Rouen on behalf of the abbey of Le Bec, which later became the priory of Notre-Dame du Pré.15 Wace also associated their foundation of the college of canons at Cherbourg with a wider programme of foundations to feed the poor, the sick, and the disabled, mentioning similar institutions at Rouen, Bayeux, and Caen. By making these foundations part of the penance for their marriage, which may be a mistake, Wace is, however, explicitly dating this activity to before 1066.16 These years were also ones when William had contact with a range of other religious experiences. His confirmation of the foundation of the tiny Fécamp priory of Saint-Martindu-Bosc near Honfleur was in effect approval of the establishment of a hermitage, since only two monks were involved. In doing this, he also implicitly gave approval to an eremetical tradition for which Abbot John of Fécamp was a known enthusiast; the presence among the signa of Herluin de Conteville vicecomes signifies that William’s stepfather was assigned a local protective responsibility.17 The Cherbourg narrative also mentions the ducal chaplains and their special status. The emergence from within this group of some of the important figures of the post-1066 Church in Normandy and England – including of course William of Poitiers – is 12 For bloodletting, recreation, and well-being, Mary Katherine Keblinge Yearl, ‘The Time of Bloodletting’, Yale University PhD Thesis (2005), 177–204. See specifically, The Monastic Constitutions of Lanfranc, ed. and trans. Dom. David Knowles and Christopher N. L. Brooke, OMT (Oxford, 2002), 138–41. 13 GR, i, 492–3. 14 Ibid., 508–9. 15 Chronicon Beccensis abbatiae ab ipsa fundatione ad annum 1467, ed. L. d’Achery, in Lanfranci Opera Omnia (Paris, 1648), 3 (RADN, 27, note 34; Potter, The Friendship Network of the Abbey of Le Bec-Hellouin, Appendix, 336 [Bec no. 18]). 16 Roman de Rou, pt. 3, 198–201, lines 4,529–36. 17 RADN, no. 218; Tom Licence, Hermits and Recluses in English Society, 950–1200 (Oxford, 2011), 33–4.

o n to t h e at tac k

169

another indication of the religious and cultural ambience that existed around William.18 By far the greatest religious activity of these years is associated with the construction of the abbey of La Trinité at Caen. On the evidence of her mortuary roll, the first abbess, Matilda, was appointed in 1059. She was a nun from the abbey of Notre-Dame of Préaux, founded by Humphrey de ‘Vieilles’ and Roger de Beaumont, who lived until 1113 and was undoubtedly a very capable abbess. The choice of such a young nun from a recently established convent surely shows great imagination on Matilda and William’s part.19 La Trinité’s earliest charter, drawn up at the time of its dedication in 1066, describes the duchess’s extensive activity on the abbey’s behalf. While usually locating her activities within the general framework of her husband’s approval and confirmation, the charter attributes multiple purchases of property to Matilda’s agency alone. It also spells out for her a personal identity distinct from the one that derived from her marriage and position in Normandy. She was ‘the daughter of the most noble Duke Baldwin of Flanders’.20 The question arises, as it also does with Saint-Etienne, of whether La Trinité was from the start intended to be what it became, namely, a mausoleum for Matilda and for the female members of the ducal family. Because Orderic described the location of Matilda and William’s tombs within the two abbeys in identical language and because Matilda’s is marked by its original slab, there is a good chance that this was always the intention (Plate 3).21 If so, it also fits well with contemporary princely and royal culture, and is a considerable affirmation of status and confidence on the couple’s part, one made only by kings and the most successful of territorial RADN, no. 224 (suam liberationem sicut suis capellanis). For Matilda’s early life at Notre-Dame de Préaux, Recueil des rouleaux des morts (VIIIe–vers 1536), ed. Jean Dufour, 4 vols. (Paris, 2005–8), i, no. 114; Monique Goullet, ‘Poésie et mémoire des morts: le rouleau funèbre de Mathilde, abbesse de la Sainte-Trinité de Caen (†1113)’, in Jean-François Cottier, Martin Gravel, and Sébastien Rossignol (eds.), Ad libros! Mélanges d’études médiévales offerts à Denise Angers et Joseph-Claude Poulin (Montreal, 2010), 163–98, at 170–3. For Matilda’s appointment and abbacy, John Walmsley, ‘The Early Abbesses, Nuns and Female Tenants of the Abbey of Holy Trinity, Caen’, JEH, xlviii (1997), 425–44, at 427–33; Catherine Letouzey-Réty, ‘Ecrit et gestion du temporel dans une grande abbaye de femmes anglo-normande: la Sainte-Trinité de Caen (XIe–XIIIe siècle)’, Thèse de Doctorat (co-tutelle), Université de Paris I Panthéon-Sorbonne and University of London (2011), 206–12. 20 For the charter, see AAC, no. 2, an edition that is superior to RADN, no. 231, because it takes account of a twelfth-century copie figurée that was inaccessible to Marie Fauroux and which may reproduce features of the lost original. 21 OV, iv, 44–5, 104–5, describes both William and Matilda’s tombs as being inter chorum et altare. For the mausoleum argument, Eric Gustav Carlson, ‘The Abbey Church of SaintEtienne at Caen in the Eleventh and Early Twelfth Centuries’, Yale University PhD Thesis (1968), 237, note 19. For La Trinité, Letouzey-Réty, ‘Ecrit et gestion’, 123–32. 18 19

170

william the conqueror

princes.22 In terms of their perceived status within Normandy, William and Matilda were placing themselves on the same level as his grandfather Duke Richard II and his foundation at Fécamp, and adding a feminine dimension to it. All this activity of Matilda’s could even be a statement on her part of preparedness to take on queenship in England. The king and queen as partners in monastic prayers was a constant in the great statement of the later Tenth-Century English Reform, the Regularis Concordia, the implication being that they were jointly responsible for the protection of the Church.23 And another image of king and queen working together was available from the Empire, that of Conrad II (1024–39) and Gisela.24 Saint-Etienne certainly progressed more slowly than La Trinité. Although Robert of Torigni says that Lanfranc was appointed abbot in 1063, Orderic dates the appointment to the same day in 1066 as that of Abbot Mainer of Saint-Evroult.25 Since both historians were arguably likely to have special information, because Robert was for many years a monk of Lanfranc’s first monastery of Le Bec and because Orderic was writing about an important matter related to the history of his own monastery, the disagreement may be unresolvable. Although the general scholarly consensus favours 1063, the appearance of Lanfrannus monachus in a diploma dating to 22 September in a year after the conquest of Maine, and his absence from the list of eight abbots drawn from the duchy’s leading monasteries – who are among the signa to the diploma confirmed in 1066 at the time of La Trinité’s dedication – do tend to support Orderic.26 In addition, the only known pre-1066 charter for Saint-Etienne records a grant by Roger de Montgommery that was to be implemented after his death, a probable indication that the endowment was in the early stages of development.27 On the other hand, the evidence for Saint-Etienne’s development after 1066, and for Lanfranc’s activities as abbot in purchasing property, is so extensive as to indicate that, even if his appointment does

22 Elizabeth M. Hallam, ‘Royal Burial and the Cult of Kingship in France and England, 1060–1330’, JMH, viii (1982), 339–80, at 367–77; Letouzey-Réty, ‘Ecrit et gestion’, 113–16. 23 Regularis Concordia Anglicae Nationis Monachorum Sanctimonialiumque: The Monastic Agreement of the Monks and Nuns of the English Nation, ed. and trans. Dom. Thomas Symons (London, 1953), 16, 20, 21, 22; Elisabeth van Houts, ‘Queens in the Anglo-Norman/Angevin realm 1066–1216’, in Claudia Zey, with Sophie Caflisch and Philippe Goridis (eds.), Mächtige Frauen? Königinnen und Fürstinnen im Europäischen Mittelalter (11.–14. Jahrhundert) (Ostfildern, 2015), 199–222, at 200. 24 For references, Herwig Wolfram, Conrad II 990–1039: Emperor of Three Kingdoms, trans. Denise A. Kaiser (University Park, PA, 2006). 25 Chronique de Robert de Torigni, i, 49–50; OV, ii, 144–7. For surveys of the conflicting evidence, OV, ii, 146, note 2; Cowdrey, Lanfranc, 24, note 68; NMProsopographie, 38–41. 26 RADN, no. 159; AAC, no. 2 (RADN, no. 231). 27 AAC, no. 1.

o n to t h e at tac k

171

date from 1066, the construction of the monastery’s endowment was being pushed forward vigorously in the early 1060s.28 The naming of William and Matilda’s eldest son Robert as heir to Normandy is mentioned in an original narrative charter for the abbey of Saint-Ouen at Rouen, which is dated 29 June 1063. The designation, which was presumably a fairly recent one, is presented as a domestic decision, with the use of the plural of the verb in the charter indicating that it was made jointly by William and Matilda. It also has William, Matilda, and Robert confirm the grant as a threesome.29 No reference is made to any public confirmation of the designation. However, although neither of them supplies a specific date for it, both Orderic and John of Worcester say that Robert was granted Normandy before William invaded England, with Orderic adding that William ordered the nobles to do homage and fealty to his son.30 The more precise of his two references to the designation associates it with a time of illness, but in the context of the designations made both before and after 1066.31 We are presumably dealing with a private family decision that became a public one. If William’s serious illness is dated to 1062, then they were following the precedent set by his father of making the designation at a time of danger. However, if it dates to later than this, the designation could have been intended to consolidate the approximately ten-year-old Robert’s position in relation to the then ongoing takeover of Maine. Whatever the case, it fits well with both the private and political circumstances of the time and represents a declaration that the responsibilities of rule were being taken seriously. It also shows that Robert had been prepared for future rule. What we know indicates that William and Matilda took their children’s education seriously. The third and fourth sons, William and Henry, for example, were certainly tutored respectively by Lanfranc and Osmund, bishop of Salisbury.32 Two remarkable charters from these years show William managing a situation by theatrical display and by an apparently dominant presence. 28 For the emptiones Lanfranci, AAC, nos. 14, 20; David Bates, ‘Four Recently Rediscovered Norman Charters’, AN, xxxv (1995), 35–48, at 45–6. 29 RADN, no. 158 (per consensum Guillelmi comitis, domini sui et Mathildis uxoris eius, et Rodberti eorum filii, quem elegerant ad gubernandum regnum post suum obitum). For the use of the word regnum to describe Normandy and other principalities, Bates, Normandy before 1066, 56. On the evidence for Robert’s designations, R. H. C. Davis, ‘William of Jumièges, Robert Curthose and the Norman Succession’, EHR, vc (1980), 597–606, at 599–601. 30 OV, ii, 356–7; iv, 92–3; JW, iii, 30–1. 31 OV, ii, 356–7 (Nam Guillelmus princeps ante Senlacium bellum et post in quadam sua aegritudine). 32 For the sons, Aird, Robert Curthose, 38–40; Barlow, William Rufus, 21–4; C. Warren Hollister, edited and completed by A. Clark Frost, Henry I (New Haven, CT, and London, 2001), 35–7; Judith A. Green, Henry I, King of England and Duke of Normandy (Cambridge, 2006), 22–3.

172

william the conqueror

Both of them may also provide an insight into his attitudes and personality. One, an original charter dating at the earliest from the late 1050s and therefore in all probability to the period covered by this chapter, records as direct speech a public pronouncement made by William during an Easter court held at Fécamp. Questioned by the monks of the Loire valley abbey of Saint-Florent of Saumur about the status of a gift that was being made to them, William responded by saying that he and his followers were indeed Normans, but that they were well aware of what was required of them, and that, God willing, it would be done.33 Plugging into contemporary notions of Norman ethnicity, William’s statement seems to be both aggressive and defensive in character. It appears to indicate that he thought the Normans were distinctive, and, because of their Scandinavian heritage they were somewhat exotic, but at the same time he was saying that they were right-thinking and moral people who would do what was required of them. So viewed, it may suggest a capacity for irony on William’s part. It also indicates a level of ethnic consciousness, something that is also innate in William of Malmesbury’s statement that William was driven to emulate the feats of Robert Guiscard because he was a member of the same gens, all the more so because Robert was a social inferior.34 The second document, the notice written at the abbey of Marmoutier mentioning William’s visit to La Hougue during 1062, describes him there in the company of, among others, Roger de Montgommery and William fitz Osbern. Another Marmoutier notice, although composed after 1066, mentions the same individuals and is likely to be describing a transaction that happened at the same time. It has William attack a ducal forester named Hugh with an animal bone for criticizing a grant. Surely a piece of relaxed, boisterous play-acting at a gathering in the open air, presumably after a successful hunt, it may well portray a moment of informality of which we otherwise know little, an antidote to the excessively religious bias of many of our sources.35 The forester may well have been someone whom William knew well and with whom such humorous exchanges were a commonplace. It is a reminder that William would have been almost permanently surrounded by members of his court and household, some of them soldiers and priests, and some of them the holders of relatively humble offices, such as minor tax collectors and the likes of the one ducal

33 Licet Normannni simus, bene tamen novimus quia sic oportet fieri, et ita, si Deo placuerit, faciemus, RADN, no. 199. The presence of Robert of Mortain among the signa means that Marie Fauroux’s dating limits of 1051x1066 should be narrowed, with the late 1050s becoming the early dating limit. 34 GR, i, 482–3. 35 RADN, nos. 150, 151.

o n to t h e at tac k

173

swineherd to appear in a charter.36 It was after all out of this very mixed and no doubt lively milieu that William’s own mother had emerged. With this glimpse of the rumbustious life that must have been a constant presence around William and in which he must have participated, it is significant that a small group gathered at court soon after his death reflected on his affability and generosity.37 It is moments like this one that they may have been remembering, times when William was away from the formal limelight. It also draws attention to the constant travelling that itinerant rule involved and how even the most important of William’s subjects could be in his company in relatively remote places. RULE IN NORMANDY, 1060–3 John of Ivry’s elevation to the bishopric of Avranches in 1060 was both an important episcopal appointment and another element in the consolidation of the southern and western frontier zones. At first sight an old-style election of a ducal relative, since John was a son of Count Rodulf of Ivry and the much younger brother of Bishop Hugh of Bayeux who had died in 1049 soon after the Council of Rheims, he actually represented the newer type of bishop who appeared in the wake of Archbishop Maurilius. John was an energetic churchman, with strong reforming principles, and a distinguished liturgist. Although considerable changes had taken place at Avranches during the long episcopate of his predecessor, Bishop Hugh, such as, for example, the beginning of the construction of a new cathedral, John’s appointment was rapidly followed by a vigorous restructuring of his diocese.38 His efforts included the settlement in 1061 of the jurisdictional relationship with the monks of Mont Saint-Michel, whereby the monks received the rights of an archdeacon on the Mont but were integrated into the organizational framework of the diocesan synod. John also carried out a survey of estates that resulted in the creation of a pancarte of the kind confirmed by William for the cathedral of Coutances, and he augmented the cathedral’s properties with grants from his own resources.39 These changes reflected general changes in episcopal governance then occurring across Normandy, developments that certainly had William’s approval.40 Ibid., no. 197. See further Bates, Normandy before 1066, 161. Hugh of Flavigny, Chronicon, viii, 407. 38 For John, Richard Allen, ‘ “A Proud and Headstrong Man”: John of Ivry, Bishop of Avranches and Archbishop of Rouen, 1060–79’, HR, lxxxii (2009), 1–39, esp. pp. 8–15. For the cathedral, Baylé, ‘Les évêques et l’architecture normande’, 158–61. 39 Chroniques latines du Mont Saint-Michel, 375–8 (The Cartulary of the Abbey of Mont-SaintMichel, 195–6, no. 5); Allen, ‘ “A Proud and Headstrong Man” ’, 10–11; RADN, no. 229. 40 Bates, Normandy before 1066, 209–18. 36 37

174

william the conqueror

John also acted, when required, beyond Normandy’s borders to stabilize relationships, for example confirming with William’s brother, Count Robert of Mortain, a grant of land in his diocese – that is, within what was technically Normandy – to the Breton abbey of Saint-Sauveur of Redon.41 Orderic’s detailed account of William’s further interventions in the abbey of Saint-Evroult between 1061 and 1063, which are again largely written from the abbey’s perspective, supplies some further very revealing material about his methods of rule. Having allowed Arnold d’Echauffour to succeed to the lands of his uncle Robert fitz Giroie after the latter’s death on 6 February 1060 – an apparent act of reconciliation following events at the end of the 1050s when William had besieged the family’s castles – he subsequently confiscated Arnold’s lands and exiled him from Normandy. During this stage in the history of the founders and patrons of the abbey of Saint-Evroult, Orderic associated Arnold’s banishment with those of Hugh de Grandmesnil, co-founder of Saint-Evroult, and of Ralph de Tosny. He portrayed their punishment in the Gesta Normannorum Ducum as the result of William’s illegitimate display of anger; in the Historia Ecclesiastica he said that all the sentences were unjustified and that they were the result of the machinations of Roger de Montgommery and his wife Mabel.42 They were therefore further manifestations of the type of local competition and rivalry that had flared into violence during William’s adolescence. The apparently rapid restoration of both Hugh and Ralph during the period of the conquest of Maine, and their subsequent receipt of estates in England, suggest that what may have occurred was a preemptive strike on William’s part to remind them of the responsibilities of loyalty. Their treatment could also have been associated with troubles around the abbey linked to Hugh’s brother, Abbot Robert. Alternatively, favouritism towards Roger de Montgommery could have been a cause, one possible objective being William’s desire to consolidate power in a sensitive region at a time when the conquest of Maine was coming on to the agenda. Continuing a pattern in William’s behaviour, it must, however, be significant that neither Hugh nor Ralph obtained the sort of estates in England that might have been expected given their wealth and status in Normandy. Although not excluded in the manner of losers at the Battle of Val-èsDunes and their descendants, the example of Hugh de Grandmesnil and Ralph de Tosny once more emphasizes that restoration to William’s favour was often accompanied by relative long-term disfavour.43 41 Cartulaire de l’abbaye de Redon en Bretagne, ed. M. Aurélian de Courson (Paris, 1863), 277–8, no. CCCXXVI. 42 GND, ii, 152–3 (exasperatus ingenti furia); OV, ii, 90–1, 92–5, 106–7, 122–5. 43 C. P. Lewis, ‘Tosny, Ralph de [Ralph de Conches] (d.1102)’, ODNB.

o n to t h e at tac k

175

The course of Arnold d’Echauffour’s exile was played out within a framework of conventions presupposing a response from William that would lead to some sort of eventual reintegration into Norman aristocratic society. Nigel the vicomte’s perambulations supplied a likely and recent parallel. In Arnold’s case, he was given refuge just beyond Normandy’s southern borders by Rotrou, count of the Perche. He was the holder of an extensive block of lands around Mortagne, who, from the late 1050s and with Henry I’s acquiescence, had taken to calling himself count of the Perche.44 From there, Arnold gave periodic reminders of his presence with plundering raids that included entering the abbey, which he himself had founded, accompanied by a troop of armed men; he was talked out of further violence by a monk named Herman. Arnold then spent some time among kinsmen in southern Italy and raised the funds apparently needed to regain favour with William. Seeking reconciliation through the gift of a precious cloak, he was reinstated, an act that Orderic specifically associates with William’s need for warriors to fight wars in Maine and Brittany. But Arnold died shortly afterwards, in 1063, poisoned according to Orderic by Roger de Montgommery’s wife Mabel, a woman for whom the monk of Saint-Evroult could scarcely manage a good word. Orderic’s narrative, which includes a first blundered attempt at poisoning that caused the death of Roger de Montgommery’s brother instead, may be credible. However, the fact that two other people whom Mabel is supposed to have poisoned actually recovered, may point to a severe case of food poisoning.45 Robert de Grandmesnil, a former soldier appointed with William’s approval as abbot of Saint-Evroult in 1059 – despite a record of disruptive behaviour that at one point required a visit of the monastery by a commission that included Lanfranc – seems to have been a strong character who evoked varying responses. In the Gesta Normannorum Ducum, Orderic says that Robert’s expulsion was motivated solely by his kinship with the Giroie family and therefore, by implication, because of the political support that he was giving them. He also says that Robert was expelled without a trial or without the judgment of a synod; in other words, a monk could be subjected to the arbitrary force of William’s ritualized anger and driven into exile in the same way as a layman.46 Later, presumably drawing on the memory of his fellow monks for his Historia Ecclesiastica, Orderic tells a colourful tale of Robert being denounced by his prior for making jokes 44 Musset, ‘Le droit d’exil’, 45–59, esp. pp. 46–55; van Houts, ‘Vocabulary of Exile’, 19–21; Johnson, ‘The Process of Norman Exile into Southern Italy’, 29–38. For Rotrou’s role, Kathleen Thompson, Power and Border Lordship in Medieval France: The County of Perche, 1000–1226, Royal Historical Society Studies in History (Woodbridge, 2002), 40. 45 OV, ii, 124–31; iii, 134–5. 46 GND, ii, 152–5.

176

william the conqueror

about William and of Robert being summoned before William’s court to answer charges that Orderic said were false; the fact that the prior had previously been a monk of the abbey of Conches founded by the Tosnys may explain Ralph de Tosny’s exile as being a precautionary measure designed to prevent inter-family conflict. Orderic says that Robert left Saint-Evroult on 27 January 1061 to appeal to Pope Nicholas II, after which he returned to northern France seeking to regain his office with the support of papal legates. As they approached Lillebonne where William was staying, they heard stories of his fury and his declaration that, while he was very happy to receive representations from the pope on matters of faith, he would hang any monk from the duchy from a high oak tree who dared to initiate a case before the pope against him. On learning of this, Robert beat a hasty retreat beyond the Norman frontier and tried to entice Osbern, the abbot installed by William as his replacement, into a hearing beyond the borders of Normandy, an invitation that Osbern resisted.47 While Orderic’s account of these affairs is ultimately a sad tale of a community split asunder by conflicting loyalties and the actions of two strong personalities, William’s capacity to frighten off two papal legates by symbolically communicating his disapproval is yet another sign of how feared he now was. Also demonstrative of how dominant William’s power had become was the way in which he pushed through the appointment of Osbern, prior of William fitz Osbern’s foundation of Cormeilles, as the replacement for Robert while the latter was still legally abbot. Although a breach of canon law that also infringed the privilege granted to the monks of Saint-Evroult at the abbey’s foundation, which allowed them to elect their own abbot, Orderic’s narrative shows the leading churchmen of Normandy rallying round to put William’s wishes into operation. Osbern, so we are told, arrived at a synod involving most, if not all, of the bishops of Normandy, and found himself pushed into the job without even being asked whether he wanted it. His investiture at Saint-Evroult with Archbishop Maurilius of Rouen’s staff of office, and his appointment on Lanfranc’s advice, show that William’s intervention had the support of the duchy’s ecclesiastical establishment; as does the fact that Osbern was subsequently met at SaintEvroult by the diocesan, Bishop Hugh of Lisieux. William had evidently consulted with those who had the ecclesiastical authority to confirm the change he wanted, and they had agreed to carry it out. In other words, this was a pre-planned coup. Although legally a usurper, Osbern is consistently presented by Orderic as a good man who laboured to heal wounds. That he was eventually 47

For the exile of Robert de Grandmesnil, OV, ii, 90–7.

o n to t h e at tac k

177

allowed by William to go to the papal court to have his position sanctioned was both a display of confidence in the likely result and a decision which recognized that to allow the situation to continue as it was placed souls at risk. Work behind the scenes may well also have made Abbot Robert recognize that he had become so much persona non grata in Normandy that there was no way back for him. His renunciation of his abbacy at the papal court and Alexander II’s recognition of Osbern as his successor were therefore fundamentally acts of pragmatic appeasement on the pope’s part that reveal how influential William’s power could be and also – perhaps – that he was trusted to produce an outcome in everyone’s interests.48 It may well show that a lot of people could read his intentions and knew that they were neither irrational nor necessarily malign. For all the ruthlessness of an act that had removed someone whom William no longer trusted, everyone influential interpreted his conduct as couched in the expectation that proper standards of religious life would be sustained. Robert thereafter remained in southern Italy, where he was given responsibility for two monasteries by Robert Guiscard. Prayers for his soul continued to be said at Saint-Evroult after his death, pointing perhaps to some long-term ambivalence there about William’s behaviour. Not a great deal else is known about William’s activities in 1061 and early 1062. The Saint-Evroult narrative shows him to have been in Upper Normandy at Lillebonne and Rouen in 1061. A charter for Mont SaintMichel also indicates his presence at Rouen, but Bishop Hugh of Lisieux’s name among the signa may indicate that it was confirmed at the synod at which Abbot Osbern was invested and that it adds nothing to our knowledge of William’s itinerary.49 Otherwise, he is known only to have been at La Hougue at an unknown date in 1062. THE CONQUEST OF MAINE The death of the young and unmarried Count Herbert II of Maine on 9 March 1062, aged just fifteen, created circumstances that led to dramatic changes in the political complexion of northern and western France. William of Poitiers’s narrative of the conquest of Maine, supplemented on some points by Orderic, is effectively the only one with any pretensions to completeness. He says that Herbert contacted William, motivated by weariness of Geoffrey Martel’s overbearing rule. This probably (but not necessarily) occurred before Geoffrey’s death in 1060. Additionally, Poitiers 48 49

For the history of Robert’s replacement by Osbern, OV, ii, 106–15. RADN, no. 148.

178

william the conqueror

wrote that Herbert promised the succession to William should he die childless, that he did homage to William, and that he asked for one of his daughters in marriage. Poitiers added that Maine had been subjected to Normandy in the distant past, presumably an allusion to Flodoard of Rheims’s account of the establishment of Normandy in the first decades of the tenth century and the statement therein that the grant made by King Ralph in 924 consisted of Bayeux and Maine.50 Such a constructed framework of legitimation is of course typical of the Gesta Guillelmi. However, with no existing alternative perspective on these events, the criticisms that Poitiers was trying to answer have in this case to be deduced by trying to read between the lines. As in the case of the English succession, a multiplicity of loyalties, norms, scripts, and rules, as well as the impact of recent events, would have made the situation much more complicated than Poitiers’s literary purpose required him to make explicit. Whatever the nature of William’s negotiations with Count Herbert or, more likely, with his advisers, the young man’s death in his mid-teens presumably could not have been foreseen; there is no evidence that he was suffering from any long-term illness or that he was a weakling. Herbert, who cannot have been born before 1046 or 1047, and who had succeeded his father Count Hugh IV in 1051, had been driven out of Maine after his father’s death by the citizens of Le Mans, who had then turned to Geoffrey Martel of Anjou for protection. After spending some time in exile along with his mother Bertha and his siblings, Herbert was subsequently reconciled with Geoffrey, or more likely reinstated by him.51 The young boy then appears in charters, with the count of Anjou remaining the dominant power, although he was presumably technically acting as Herbert’s guardian. The author of the Actus Pontificum Cenomannis wrote that Geoffrey effectively acted as count for nearly ten years.52 One of the small number of surviving charters that mentions Herbert actually goes so far as to spell out explicitly the unequal nature of the relationship, while another accords precedence to Geoffrey.53 In terms of chronology, Herbert

50 GG, 58–61. For the supposed grant of Maine, Les annales de Flodoard, ed. P. Lauer (Paris, 1906), 24 (The Annals of Flodoard of Reims, 919–966, ed. and trans. Steven Fanning and Bernard S. Bachrach [Peterborough, Ontario, 2004], 12). 51 For the exile of Herbert’s family, Latouche, Histoire de comté du Maine, 29, 32. For different emphases, Guillot, Anjou, i, 86–7; Bruno Lemesle, La société aristocratique dans le Haut-Maine (XIe–XIIe siècles) (Rennes, 1999), 33–4; Barton, Lordship in the County of Maine, 112. 52 Actus Pontificum Cenomannis in urbe degentium, 366. 53 Cartulaire de l’abbaye de Saint-Vincent du Mans, ed. R. Charles and le vicomte Menjot d’Elbenne (Mamers and Le Mans, 1886–1913), no. 184 (Actum Cenomannis, curie publice, Gaufrido comite presidente, Herberto puero comite vivente); Cartulaire de Saint-Aubin d’Angers, no. 941 (Guillot, Anjou, ii, no. C230).

o n to t h e at tac k

179

can be shown to have been co-operating with Geoffrey after 31 July 1056 and in all likelihood with King Henry I on 15 November 1058.54 If Herbert’s request, or rather the request made on his behalf, for William’s assistance and the associated homage and succession promise were made before Geoffrey’s death, it must have been at a time when the latter’s power was weakening, that is, after the Battle of Varaville. Hence, the period between 1058 and 1060 has been widely accepted as the time when the contact was made.55 However, the reference to co-operation between Herbert and King Henry late in 1058 makes it possible that the request for William’s assistance actually came only after Geoffrey Martel’s death, something that Poitiers’s ambiguous wording does not rule out; the reference to being wearied by Geoffrey Martel’s rule might mean that Herbert only acted after Geoffrey’s death, his motivation being to secure a protector in the new uncertain world created by the childless Geoffrey’s division of his lands between his two nephews: Geoffrey (Count Geoffrey the Bearded), who became count of Anjou; and his younger brother Fulk (later Count Fulk the Rechin, 1068–1109), for whom a substantial landed provision was made.56 It is likely that Herbert was encouraged to be proactive by his mother Bertha. Having been married first to Count Alan III of Brittany and then to Count Hugh, both enemies of the counts of Anjou, and in full awareness of William’s assistance to Bishop Gervase in 1051 and of the significance of the victory at Varaville, she is likely to have seen William as the obvious person to ask to help her son; the possibility that it was her advice that drove Herbert on is actually suggested by Orderic.57 The contemporary assertiveness of Herbert’s half-brother, Count Conan II of Brittany, the son of Bertha’s first marriage to Count Alan, is also surely not a coincidence; Conan imprisoned his uncle Count Eudo in 1057 and was in Bertha’s company at Chartres in the same year. Although her presence in Brittany, clearly recorded in 1062 or 1063, suggests that she eventually settled there, it would have been easy for her also to advise Herbert.58 She is likely to have been a considerable power behind the scenes in the events that followed. 54 For Herbert and Geoffrey after 1056, Cartulaire de l’abbaye de Saint-Vincent du Mans, no. 303 (Guillot, Anjou, ii, no. C231). For Herbert and King Henry, ibid., no 15 (Actum hoc Cenomannnis, in capitulo sancti Juliani, xviio kal. Xbris, luna xxva, Hainrico rege, Herberto puerulo comite, Vulgrino presule, Ernardo notario). 55 For 1058x1060, Latouche, Histoire de comté du Maine, 32, note 5; Lemesle, La société aristocratique, 33–4. 56 Guillot, Anjou, i, 87, note 391; ii, 153–4. 57 OV, ii, 116–17. 58 Les actes des ducs de Bretagne, nos. 45 (Rédaction I), 55, 56.

180

william the conqueror

On the presumption that Herbert, after having married William’s daughter, would father children, the proposed union was from Herbert’s point of view a replica of the strategy adopted when his father had married his mother in order to obtain the support of the counts of Blois-Chartres, in the 1040s the most effective rivals of the counts of Anjou. The promise of the succession to William if Herbert should die childless was therefore no more than an insurance policy taken out by William against the uncertainties of the future; his primary objective in agreeing to the alliance would have been to stabilize conditions to the south of Normandy that had been so much of a preoccupation for him, and not to set up a potential conquest. In the early 1060s no daughter of William would have been old enough for the marriage to take place, although it was not out of the question that one could have been dispatched to Le Mans to be brought up there; there is, however, no evidence to indicate that this happened. The daughter concerned, in all probability Adelida, was to have a sad life, hawked around to be joined in several proposed unions of varying degrees of suitability or unsuitability. If all this had taken place before his death in 1060, Geoffrey Martel would undoubtedly have been displeased. But it is doubtful that he would have been profoundly disturbed. As Herbert’s overlord, he would have regarded the homage as invalid, because it contravened their agreed relationship, and the marriage to William’s daughter as preventable. But after 1060, Geoffrey the Bearded’s weakness and the war of succession that developed between him and his brother Fulk caused the enlarged territory assembled by Geoffrey Martel to start to disintegrate. In 1062, Fulk was driven out of his predecessors’ southern conquests in the Saintonge by the count of Poitou, receiving neither assistance nor compensation for his loss from Geoffrey. These circumstances left Maine open to intervention following Herbert’s death. Walter, count of the Vexin, rapidly claimed the county as the husband of Biota, the sister of Herbert’s father Count Hugh IV, and seems to have established control over some of it. Whatever the nature of the arrangements that Herbert had made with William, the absence in the Gesta Guillelmi of any reference to either oaths or hostages being exchanged must indicate that none of these conventional formalities took place. In these circumstances, Count Walter may well have believed that the way was legitimately open for him to intervene in Maine and that he would find allies in doing so. In addition, although Walter’s father Count Drogo had died on pilgrimage with William’s father, and his mother was Edward the Confessor’s sister Godgifu, the close relationships of the earlier generation had dissolved, presumably because Godgifu’s return to England had largely severed the link. In the 1050s, while maintaining some contacts with Normandy, Walter mostly appears in charters

o n to t h e at tac k

181

of King Henry I, suggesting that he had moved closer to William’s enemies than he was to William.59 In this, he was following a widespread pattern and apparently aligning himself with the most powerful rulers of those times. His expectation must have been that some of these friends, and, perhaps, specifically Count Geoffrey the Bearded, would support him. If so, he failed to recognize how much circumstances had changed. According to Orderic, Walter’s main adherents within Maine were Geoffrey de Mayenne and Hubert de Sainte-Suzanne, vicomte of Le Mans. With his record of opposition to William, Geoffrey’s appearance alongside Walter is hardly surprising. Based as he was in the north of Maine, Geoffrey’s great fear was presumably an absorption into Normandy that would compromise the relative autonomy he enjoyed because of the weakness of the counts of Maine and the relative remoteness of the counts of Anjou.60 Orderic, whose account is distinctly more sympathetic to William’s opponents than is that of Poitiers, also says that those supporting Walter feared being subjected to ‘the burdensome yoke of Norman rule’ (Normannicum iugum . . . grauissimum), perhaps a piece of stereotypical ethnic labelling informed by knowledge of over fifty years of turbulent relations between the dukes of Normandy and Maine, or perhaps an indirect statement that the one man who was not wanted as count of Maine was William, duke of Normandy.61 The two considerations that determined events thereafter were William’s belief that he had a claim and his commitment to pursuing it, along with the near total failure of any of Walter’s anticipated support to materialize. The possibility that, as a nephew of Edward the Confessor, Walter had the English succession in mind when he intervened in Maine could be a facet of the unknown history of these times, and has sometimes been seen as such.62 His brother Ralph had crossed to England with Edward and had been made an earl by him in either 1049 or 1050. Earning a certain amount of notoriety in 1055 by ordering some English to fight against a Welsh army on horseback, an incident that has a central place in the arguments about whether the English were capable of cavalry warfare and which earned Ralph the epithet ‘the Timid’, he had died in 1057. His marriage to a woman named Gytha and the name given to their son (Harold) suggest integration 59 Bates, ‘Lord Sudeley’s Ancestors’, 41; Carmen, xlix; Bauduin, La première Normandie, 257–8. 60 For this point, W. Scott Jessee, Robert the Burgundian and the Counts of Anjou, c.1025–1098 (Washington DC, 2000), 46–8. 61 OV, ii, 116–19. 62 Thus, Douglas, WC, 171 (‘He [Ralph] passed on his claims to his brother Walter, count of the Vexin’), although he is elsewhere cautious about whether Walter was seriously interested (p. 172). See also Guillaume de Poitiers, Histoire de Guillaume le Conquérant, ed. Raymonde Foreville (Paris, 1952), 92, note 3.

182

william the conqueror

into the increasingly dominant Godwine family of England in the 1050s, rather than independent ambitions of his own. Walter’s only known contact with Edward occurred when, at a date before 1052, he sent a future bishop of Laon to England to become one of the king’s chaplains. Although the chaplain did subsequently act as an ambassador between Edward and King Henry I, this is not enough to indicate serious designs on kingship on Walter’s part or that he followed events in England at all closely.63 Also, although Godgifu lived on until after 1066, neither she nor Walter, nor Ralph’s young son Harold, would have had sufficient weight to take on either William or Harold Godwineson, or indeed a court-backed Edgar the Ætheling. Walter’s actions are best seen as northern French in purpose and context. The date of William’s invasion of Maine is somewhat unclear, with, on balance, 1063 being much preferable to 1062. The best evidence, contained in several versions of the Norman annals, supports 1063.64 Orderic, while showing some confusion by once dating it to 1064, also assigns it to 1063 by writing that it happened in the same year as the dedication of Rouen cathedral; this second statement probably, however, derives from a version of the Norman annals and is not therefore independent evidence.65 The apparently alternative dating of 1062 in a Latin interpolation into the ‘E’ version of the Anglo-Saxon Chronicle should probably be disregarded, since the text was taken unchanged from the Norman annals.66 Its placing in the manuscript within the otherwise blank annal for 1062 might well mean that it should be treated as an insertion at the start of the 1063 annal, rather than as an alternative dating, since all the Latin insertions into ‘E’ are placed at the head of an annal entry and were probably not made until the later eleventh century. The absence of any reference to the capture of Le Mans by John of Worcester is also a relevant consideration, since he was using a version of ‘E’ that existed in c.1080; this indicates that, even though the author was using an early text of the Norman annals, it had probably not reached ‘E’’s author by that date.67 Guibert de Nogent, Autobiographie, 58–61 (A Monk’s Confession, trans. Archambault, 123). Thus, Hoc anno subjugata est Cinomannis comiti Normannorum Willelmo, ‘Annales Uticenses’, in Orderici Vitalis Ecclesiasticae Historiae, v, 157; ‘Chronicum Rotomagense’, 366; ‘Annalis Historia Brevis sive Chronica monasterii S. Stephani Cadomensis’, 165. For minor variants, ‘Annales du Mont-Saint-Michel’, in Chronique de Robert de Torigni, ii, 221; Les annales de l’abbaye de Saint-Pierre de Jumièges, 57. 65 OV, ii, 116–17 (for 1064); iii, 92–3 (for 1063). 66 ASC, ‘E’, 1062 (The Peterborough Chronicle [The Bodleian Manuscript Laud Misc. 636], ed. Dorothy Whitelock, Early English Manuscripts in Facsimile, iv [Copenhagen, 1954], fos. 56b–57a, at pp. 112–13). 67 For comment on the use of the Norman annals in ‘E’ and in England, David Dumville, ‘Some Aspects of Annalistic-Writing at Canterbury in the Eleventh and Twelfth Centuries’, Peritia, ii (1983), 23–57, at 32, 55–7; Alexander, ‘Annalistic Writing’, 12–15, 244–8. See also 63 64

o n to t h e at tac k

183

It is therefore very likely that William spent 1062 raising troops and assessing the situation. The presence among the signa of the diploma confirmed at La Hougue in 1062 of Geoffrey de Chaumont-sur-Loire, the man who regularly appears in the 1060s and 1070s as an intermediary co-ordinating the alliance between the counts of Blois-Chartres and William, and the additional grant to the abbey of Marmoutier made there, must certainly be seen as a reinforcement of the alliance with the counts of Blois-Chartres. They show that William was carefully recruiting allies before going fully on the offensive in 1063, in all probability in the spring.68 A notice from the abbey of Marmoutier dated 14 March 1063, apparently referring to Count Geoffrey the Bearded having already summoned the military service that the abbey owed to him, arguably shows Geoffrey was making preparations in 1062 that he was not able to implement; Poitiers says that his help had been sought from within Maine, but had never been forthcoming.69 It is of course possible that Geoffrey’s initial objective was to act as an umpire in a forthcoming succession dispute and that Walter’s rapid entrance into proceedings and William’s preparations made him realize that this was not going to be possible. William’s campaign, which Poitiers presents as a steady takeover of castles that led up to Le Mans’s isolation and surrender, was quite a long one.70 An obituary from the abbey of Saint-Père of Chartres dates Count Walter’s death to 2 August.71 Since no year is given, its value is confined to indicating that William’s campaign took place in the summer of an unknown year and that it may have continued into the early autumn. But it fits well with a chronology that places the conquest of Maine in 1063 and with what we know of Walter and his wife’s fate. Poitiers presents William’s takeover of Maine as a cautious and pragmatic advance. In explaining why William did not attack Le Mans directly, he offers the rationale that the devastation of the countryside was a more humane way to campaign than an assault on a large city. One intention must have been to build alliances and create collaborative relationships; another, to intimidate. William’s planning must also have been influenced by consideration that a full-scale war with the count of Anjou was not something to be undertaken lightly. This would not have prevented a steady infiltration of Maine and attacks on individual castles, but it surely must have made him think twice The Anglo-Saxon Chronicle. A Collaborative Edition: volume 7, ms. E., ed. Susan Irvine (Cambridge, 2004), xxxvi. 68 RADN, no. 151; Dunbabin, ‘Geoffrey de Chaumont’, 110–11. 69 GG, 60–1; Guillot, Anjou, ii, no. C246. The date of the notice could be amended to 1064 on the grounds that Marmoutier charters sometimes start the year at Easter. 70 GG, 60–3. 71 Obituaires de la province de Sens: 2 (diocèse de Chartres), ed. Auguste Molinier (Paris, 1906), 193.

184

william the conqueror

about a direct attack on Le Mans with the extended lines of communication that would have been necessary. However, once it became clear that Count Geoffrey was not going to move against him, William’s military and political position became much stronger. If he did bide his time in this way, it was not going to be the only major campaign in which he waited for circumstances to shift in his favour. His tactical approach, as described by Poitiers, was very much standard eleventh-century practice, a steady build-up of pressure that preserved the lives of his troops and avoided excessive risk.72 With the support he had anticipated having failed to materialize, Walter seems to have recognized the weakness of his position, since, according to Orderic, he had abandoned his supporters in Maine some time before Le Mans’s surrender.73 According to Poitiers, he was also worried that William would attack his lands in the Vexin. This may be a rhetorical flourish indicating what Poitiers wanted his readers to think William was capable of doing. But it may reflect an acknowledgement of vulnerability on Walter’s part; William’s capacity to fight a war on two fronts was undoubtedly greater than his.74 While Poitiers only mentions Walter’s surrender, Orderic makes two references to him and his wife dying at Falaise while they were enjoying William’s hospitality, and includes an allusion to the possibility of poisoning. If Orderic’s statement that they died while William was still campaigning is correct, they were presumably at Falaise in the expectation of discussing the terms of their surrender with William.75 This mention of poisoning may just be another occasion on which Orderic uses it to explain a convenient sudden death. His language in his main narrative (‘so the rumour goes’ – ut ferunt) is circumspect. However, when he included it within a speech made by one of William’s enemies, he did allow the accusation of murder to be made, although even here the reference to them enjoying William’s hospitality can be read as indicating that William was not present at the time.76 A speech of this kind was in any case the context in which rumour and unsubstantiated accusation could be presented as fact. It does, however, show what Orderic thought William’s enemies could accuse him of doing. David Douglas devoted an Appendix to the subject of the many mentions of poisoning.77 While counselling caution, he did admit that it was necessary to keep an open mind. Others 72 Gillingham, ‘William the Bastard at War’, 148–9 (Strickland [ed.], Anglo-Norman Warfare, 150). 73 OV, ii, 118–19. 74 GG, 62–3. 75 OV, ii, 118–19. 76 Ibid., 312–13 (Falesiae hospitauit). 77 Douglas, WC, 408–15.

o n to t h e at tac k

185

have been more emphatic that poisoning was practised as a means to dispose of enemies in the eleventh century.78 With sudden death often being inexplicable to eleventh-century minds and poisoning being largely undetectable, it is important to share Douglas’s caution. Yet, as we have already seen from Guibert of Nogent’s testimony, William had a reputation for treating prisoners of war in ways that contemporaries thought severe. A letter written by St Anselm, abbot of Le Bec, and after 1093, archbishop of Canterbury, mentions a prisoner, who was not necessarily a prisoner of war, being held on William’s behalf, whose limbs he intended to be crushed.79 If stories of this kind did circulate, they can only have reinforced the awe and the fear in which William was held. At the very least, they confirm a theme that is emerging clearly in this book, namely, William’s consistent readiness to treat enemies and opponents ruthlessly in a way that sometimes went beyond contemporary norms. At the extreme, they make him a murderer who ignored the conventions of fair treatment required by the surrender of an enemy. The capitulation of Le Mans was performed ceremonially with the bishop, clergy, and monks of the city marching out fully robed and carrying Gospel books and crosiers.80 A ritualized surrender performed in this way acknowledged William’s authority and symbolically gave him the support of the city’s religious bodies. It also conferred responsibilities on him since the churches of Le Mans had placed themselves under his protection. A garrison was installed in the castle, and Poitiers emphasized the permanence of its presence.81 But in contrast to England after 1066, where the installation of garrisons in major towns was standard practice from the start, there is no evidence that the members of this garrison were enfeoffed with nearby lands or supported by revenues raised locally. William was effectively assigning a section of his personal military household at his own expense to sustaining power in Maine. The signs are that he was setting out to rule through collaborative relationships with individuals and institutions rather than by a takeover involving dispossessions, disinheritances, and the extensive intrusion of Normans. The one exception to this pattern is the attack on Geoffrey de Mayenne’s castle at Mayenne on William’s return journey to Normandy. It indicates that Geoffrey was being treated as a special case, presumably as punishment for his breach of his fealty after his involvement in the failed attack Thus, Bachrach, Fulk Nerra, 227–9. Sancti Anselmi Opera Omnia, ed. F. S. Schmitt, 6 vols. (Edinburgh, 1946–61), iii, ep. no. 27 (The Letters of Saint Anselm of Canterbury, trans. Walter Fröhlich, 3 vols., Cistercian Studies, nos. 96–8 [Kalamazoo, 1990–4], i, no. 27). 80 OV, ii, 118–19. 81 GG, 62–3 (et urbis firmamentum sua in reliquum custodia occupari). 78 79

186

william the conqueror

on the castle at Ambrières. One of the great fortifications of north-western France, the castle at Mayenne was originally constructed in the tenth century to sustain the power of the then formidable counts of Maine where their territories bordered turbulent zones of Brittany and the region to the north disrupted by the Viking attacks that were shaping the new duchy of Normandy. Subsequently expanded and encircled by a stone curtain wall, it was a daunting place to try to attack.82 Poitiers’s description of the decision to fire the castle as a preliminary to taking it by storm is one of his most powerful narratives of warfare, with William portrayed as personally going among his troops to give encouragement. After the castle had been taken, it was looted and the valuables captured were distributed to the soldiers. This, Poitiers said, was William’s custom, a reward for service and for bravery, an important commentary on the norms of eleventh-century warfare.83 Commanding one of the main southward routes from Normandy to Le Mans, Mayenne was an indispensable cog in the apparatus of rule over William’s newly enlarged territories and a garrison was therefore installed. But, for all the violence of the sack of Mayenne, there is no indication that Geoffrey was deprived of any land or, indeed, of access to the castle. Nevertheless his treatment was surely hardly likely to have made him into a friend for the future. Yet William’s expectation may well have been that Geoffrey would become more collaborative having been taught a lesson. The contrasts between William’s conquests in Maine and England, together with their short-term and long-term consequences, are a major theme in the rest of this book. The presence of individuals from Maine at the Battle of Hastings and in subsequent campaigns in England indicates their active participation in the creation of the new world that came into existence in 1066.84 Indeed, according to Wace, the bishop of Le Mans contributed thirty ships to the expedition and was an enthusiastic supporter of it.85 But the bedrock on which the alliances that shaped the conquest of England were formed usually depended on more long-standing personal relationships than those between William and his new subjects in Maine, and were much more obviously directed towards maintaining control of the Channel. As a possible contemporary commentary, William of Jumièges, upon resuming the Gesta Normannorum Ducum in the later 1060s, only lightly revised his earlier account. Writing at a time when William’s 82 For the castle, Rob Early, ‘Le château de Mayenne: les témoins archéologiques de l’évolution d’un centre de pouvoir entre le Xe et le XIIe siècle’, Château Gaillard, xx (2002), 247–62. 83 GG, 64–9. 84 Ibid., 130–1; Carmen, 16–17; OV, ii, 174–5. 85 Roman de Rou, 232–3, lines 6,165–8.

o n to t h e at tac k

187

hold on Maine had become parlous, he inserted only a brief mention of the conquest, suggesting that he saw the unfolding of events as part of a lengthy continuum of warfare in Maine, and not as a decisive change in the balance of power in northern France.86 But in 1063 many may not have seen the situation as negatively as Jumièges appears to have done. THE CONQUEST OF MAINE AND THE POLITICS OF NORTHERN FRANCE In the immediate aftermath and the short term, changes occurred that suggest the conquest of Maine greatly strengthened William’s position in northern France. For one thing, the long-standing collaboration with Count Theobald of Blois-Chartres, in all probability already deployed in the conquest of Maine, was further reinforced by the marriage of William’s twice-widowed sister Adelaide to Theobald’s nephew Odo. Whatever the role of the union in terms of relieving political pressures on Theobald and of creating further networks of collaboration, it seems to have taken place in 1063 or 1064, the time when Odo disappears from Theobald’s charters.87 A more seismic shift involved William’s rapprochement with Ralph, count of Valois, who had also taken over Amiens and the Vexin after his cousin Count Walter’s death. Described as praepotens comes by Poitiers and portrayed by Guibert of Nogent as a mighty warrior, Ralph had a remarkable record for acquiring lordships and had fought against William’s forces at the Battle of Mortemer, remaining an enemy throughout the 1050s. After his acquisition of Walter’s counties, he controlled lands that, although not a unified territorial block, stretched almost from the Channel coast into regions immediately to the north and east of Paris.88 Although the chronology of the establishment of friendly relations between William and Ralph cannot be precisely established, it must have happened soon after Ralph’s marriage to Anne of Kiev, the widowed queen of King Henry I. This probably took place in 1061 and had temporarily turned him into an exile from the minority court of Henry’s son Philip I. Just how important Ralph became during these years is shown by his being singled out by Poitiers as among those who came to Fécamp to 86 GND, ii, 150–1 (per aliquot annos arma conuertit). For reasons why the passage should be interpreted in this way, ibid., i, p. xxxiii. 87 Dunbabin, ‘Geoffrey de Chaumont’, 110–11; Thompson, ‘Being the Ducal Sister’, 72–3. 88 GG, 178–9; Guibert de Nogent, Autobiographie, 58–60 (A Monk’s Confession, trans. Archambault, 28–9). For the creation of Ralph’s conglomeration of territories, P. Feuchère, ‘Une tentative manquée de concentration territoriale entre Somme et Seine: la principauté d’AmiensValois au XIe siècle’, Le Moyen Age, lx (1954), 1–37, at 11–13, 22.

188

william the conqueror

welcome William back to Normandy at Easter 1067.89 Good relations between Ralph and William were further consolidated before 1066 when William granted him a life-lease of Gisors, an estate of the archbishops of Rouen at the border of their respective territories in the Vexin, the county that Ralph had recently inherited from Count Walter.90 In addition, Ralph’s son Simon was sent to William’s court to be educated. This must also have happened before 1066, since Simon, who must have been born before 1054, was already close to adolescence.91 With bridges soon rebuilt between Ralph and the court of King Philip I, and with William and his father-in-law Count Baldwin of Flanders (the young king’s guardian) also on good terms, conditions in the lands closest to the Channel coast and to the east of Normandy had never previously been so favourable to William’s interests – and especially so with William’s projected kingship in England in mind. These changes must surely have been significant in the calculations of many who joined in the invasion of England and in the immediate policies of the leading figures in the region. Counts Guy of Ponthieu and Eustace of Boulogne, for example, small fry in relation to the bigger powers around them, would always have had to choose their allies carefully. In 1066, many from the region would have realised that participation was possible, since there was now no threat to their lands in France. William’s campaign in Brittany in 1064 must have been a further element in the process of consolidation across the whole of northern France. But since it involved the presence of Harold, earl of Wessex, it will be considered later in the context of his visit to Normandy. RULING MAINE, 1063–6 To reinforce the proclaimed framework of legitimate rule in Maine, William and Matilda’s approximately ten-year-old eldest son Robert was betrothed to Count Herbert’s sister Margaret, who was also a child. Transferred to Normandy to be brought up in proximity to the ducal court, she died at an unknown date before reaching marriageable age and was buried in the abbey of Fécamp like a member of the Norman ducal GG, 178–9. Regesta, no. 229. 91 Vita beati Simonis comitis Crespeiensis, in PL, clvi, cols. 1,211–24, at col. 1,215 (trans. Elisabeth van Houts, The Normans in Europe [Manchester, 2000], 198). On these events, Bates, ‘Lord Sudeley’s Ancestors’, 42–3; Tanner, Families, Friends and Allies, 99–100; H. E. J. Cowdrey, ‘Count Simon of Crépy’s Monastic Conversion’, in Pierre Guichard et al. (eds.), Papauté, monachisme et théories politiques, i. Le pouvoir et l’institution ecclésiale. Etudes d’histoire médiévale offertes à Marcel Pacaut (Lyon, 1994), 253–66, at 257 (idem, The Crusades and Latin Monasticism, 11th– 12th Centuries [Aldershot and Brookfield, VT, 1999], chapter 9). 89 90

o n to t h e at tac k

189

family.92 Robert is said by Orderic to have done homage and fealty to Geoffrey the Bearded at Alençon with William present. Although Orderic does not date the ceremony, it presumably took place very soon after the conquest of Maine, perhaps late in the year 1063.93 That Robert performed the homage, and not William himself, indicates that the betrothal must have been seen as giving Robert rights in Maine, although William may well have regarded entering into such a subordinate relationship with a fellow ruler of a territorial principality as demeaning. All the pre-1066 evidence we have indicates that it was William who acted as count in Maine. Thus, in a charter dating from before 1065 he confirmed the grant of properties within the city of Le Mans by a certain Abelin to the abbey of Marmoutier. The witnesses to the document were drawn from the Le Mans cathedral chapter and the local aristocracy.94 The narrative in a second charter, dated to 22 September in a year that must be before 1066, tells how William presided over an attempt to settle a dispute between the abbeys of Saint-Pierre de la Couture of Le Mans and Marmoutier.95 Taking place at Domfront, that is, in Normandy, but in the lands bordering Maine, it brought together a prestigious assembly that included two Norman bishops, Odo of Bayeux and the recently appointed John of Avranches, Lanfranc, two Norman vicomtes, and four major Manceau lords. Also present was Rivallon de Dol, who, if the inquiry took place in 1063, was soon to incite William’s 1064 intervention into Brittany undertaken in Harold’s company. Rivallon’s presence perhaps indicates a date of 22 September 1063, although not conclusively so; 1064 is also possible, not only because William’s campaign in Brittany was over by then, but because Rivallon’s local and personal interest in the dispute might have meant that the tensions which exploded into war were insufficient to inhibit his attendance. Whatever the case, the list of signa surely represents an attempt to bring major religious and secular figures from western Normandy and Maine into a closer collaborative relationship and to create networks to integrate Maine within a new nexus that could be made to work to everyone’s benefit. Surviving in two early versions with some later additions made in cartulary copies, the charter tells how Guy de Laval, one of the major lords of Maine, had presided over a case in which he was also a plaintiff, with both sides claiming that he had created the conditions which required that a property attached to the castle of Laval should belong to them. Because Guy had failed to resolve the dispute, and because the judges (iudices) GG, 62–5. OV, ii, 304–5; William M. Aird, Robert Curthose, Duke of Normandy, c.1050–1134 (Woodbridge, 2008), 41–7. 94 RADN, no. 165. 95 Ibid., no. 159. 92 93

190

william the conqueror

appointed by William had become so entangled in its complexities that they too had been unable to produce a decision, William imposed one in favour of Marmoutier. It was based on the grounds that the monks of Saint-Pierre could produce neither a witness nor a charter to contradict Guy’s assertion that he had given the property to Marmoutier, something he was prepared to confirm on oath, but not by undergoing an ordeal which he specifically refused to do. The monks of Saint-Pierre had also failed to fulfil the condition that, so they asserted, was attached to Guy’s grant. It is possible that William’s intervention constituted an unaccustomed display of authoritarianism, because pleas of this kind in Maine were customarily subject to a consensus verdict. From William’s perspective, the imposed verdict may have been a demonstration of what he regarded as legitimate proof, which in this case was the absence of written documentation and a willingness to take an oath. The verdict may also have been intended to show William as a vigorous new arbiter who would bring clarity where there had previously been uncertainty; or perhaps he just lost patience.96 Regional interactions of a similar kind were arguably present in the controversies that followed the death of Bishop Wulgrin of Le Mans on either 10 or 11 May 1065. The Actus Pontificum Cenomannis tells us that the clergy and people of Le Mans elected as their next bishop an inhabitant of the Avranchin named Arnold.97 Although not mentioned explicitly, it is hard not to believe that William’s influence played a part in this process, or at the least that the electors decided it would be expedient to consider his interests by choosing someone from the south-westerly region of Normandy that already had significant links with Maine. Seemingly a good choice on personal grounds, Arnold’s suitability was contested by Count Geoffrey the Bearded, who tried to prevent Bartholomew, archbishop of Tours and the metropolitan, from consecrating him. After Bartholomew had protested against the violence of Geoffrey’s behaviour, the dispute passed through a hearing before Philip I at Orléans that upheld the election, and subsequently through appeals by both the archbishop and the Le Mans clergy – for no doubt different reasons – all the way to Pope Alexander II. He proceeded to ratify the election, probably through his legate Stephen when the latter visited northern France in March 1067. It may well be that William did not feel there was any need for him to intervene directly in these proceedings, because correct canonical 96 Barton, Lordship in the County of Maine, 215–19; Tabuteau, Transfers of Property, 48, 201, 215, 296, note 34, 351, note 49. 97 Actus Pontificum Cenomannis in urbe degentium, 374–5. The other main source for this episode is a letter from Archbishop Bartholomew of Tours, Briefsammlungen der Zeit Heinrichs IV, ed. Erdmann and Fickermann, no. 90. See the accounts in Guillot, Anjou, i, 106–9; Yannick Hillion, ‘Arnaud et Hoël, deux évêques du Mans au service de Guillaume le Conquérant’, Annales de Bretagne et des pays de l’Ouest, cx (2003), 49–77, at 61–4.

o n to t h e at tac k

191

procedures had been followed by the Le Mans clergy. Geoffrey the Bearded was also by this time heavily involved and on the defensive in the civil war against his brother. That William was particularly busy during the year 1066 may also have forced him to let events take their course! Whatever the case, the confirmation of Arnold’s election represented the safe route for the papacy to follow, not only because the formalities had been adhered to, but also because Alexander II was in the process of withdrawing papal support from Count Geoffrey. With the future in mind, the episode is nonetheless a prophetic illustration of the problems involved in controlling Maine, and in negotiating with the overlapping secular and religious authorities that had interests there. As comital authority had been seriously weakened during the course of the eleventh century, and as secular and religious jurisdiction over the county were divided between the count of Anjou and the king of France, William’s emphasis on continuity and legitimacy was going to turn out to be a source of both strength and vulnerability. WILLIAM, HAROLD, AND THE ENGLISH SUCCESSION Just as William’s power had dramatically increased in northern France in 1063, so too had the power of Harold and the Godwine family in England. The death in late 1062 or early 1063 of Earl Ælfgar of Mercia, with whom relations had been consistently turbulent, removed the one rival capable of any kind of effective obstruction to them. In contrast to the introduction of Harold’s younger brother Tostig into Northumbria in 1055 at the expense of Earl Siward’s son Waltheof, Ælfgar’s son Edwin was allowed to succeed him in Mercia. However, Gruffudd ap Llewelyn, king of Gwynedd and Deheubarth, who, from 1055 had imposed himself on all the other kingdoms of Wales and had supported Ælfgar against the Godwine family in England (he was the husband of Ælfgar’s daughter Ealdgyth), was attacked by Harold and Tostig. Events in general suggest that Harold’s main reason for the attack on Gruffudd was determined by English politics. Edwin might have been allowed to succeed his father, but his capacity to cause trouble for Harold and Tostig was being greatly reduced. The chronology of these events is somewhat obscure, with – quite remarkably – some modern scholars even maintaining that Harold was in Wales in 1064 at the very time that others say he was in Normandy and Brittany.98 The sequence of events is credible only if Harold’s raid on 98 The main English account is JW, ii, 592–3, 596–7, which expands the compressed accounts in ASC, ‘D’, 1063; ‘E’, 1063; VEdR, 86–7. For a discussion of the full range of the evidence, Benjamin T. Hudson, ‘The Destruction of Gruffudd ap Llywelyn’, Welsh History Review, xv (1991), 331–50, esp. pp. 340–50 (reprinted in idem, Irish Sea Studies, 900–1200 [Dublin, 2006], 81–99, esp. pp. 86–99). For varied approaches to the chronology, T. M. Charles-Edwards,

192

william the conqueror

Gruffudd’s residence at Rhuddlan near the northern Welsh coast took place in mid-winter early in 1063, and if his combined sea and land operation with Tostig followed in May of the same year. Gruffudd, who had probably fled from Rhuddlan to Ireland, was killed there, apparently by Cynan ab Iago, a rival for kingship in Gwynedd. His severed head was then sent back across the Irish Sea to Harold who, in due course, presented it to Edward, an event that, according to John of Worcester, took place after 5 August 1064. This date of 1064 is confirmed by Irish evidence.99 The outcome of the campaign was an imposed settlement that split up the kingdoms Gruffudd had combined under his authority. His half-brothers Bleddyn ap Cynfyn and Rhiwallon ap Cynfyn obtained Gwynedd and Powys respectively, and the descendants of Gruffudd ap Rhydderch were restored to the southern Welsh kingdom of Deheubarth. At some point, too, Harold further strengthened his personal position by marrying Ealdgyth, Edwin’s sister, the widow of Gruffudd ap Llewelyn, for whose gory death he was directly responsible. Although it is possible that the marriage only took place in 1066, it too was part of a scheme to gain allies and remove enemies.100 Harold and Tostig’s campaign also takes its place within the long-term exercise of coercive power within the British Isles, an aspect of a general framework of hegemonic power and influence that was to be extremely important after 1066.101 This chronology leaves space for Harold’s visit to northern France to have taken place in the late spring and early summer of 1064.102 It also reinforces the notion that he crossed the Channel believing himself to be in a strong political position. Deriving ultimately from William of Malmesbury, but surely exceedingly likely on any analysis, the idea that Harold had a secret agenda in visiting northern France – which neither we nor anyone writing after 14 October 1066 can know for certain – has been taken up in different ways in several publications.103 It leads inexorably to the concluWales and the Britons 350–1064 (Oxford, 2013), 566–7; cf. Baxter, The Earls of Mercia, 86–7; Ann Williams, ‘Ælfgar, Earl of Mercia (d.1062?)’, ODNB; David Walker, ‘Gruffudd ap Llywelyn (d.1063)’, ODNB. 99 The Annals of Ulster to A. D. 1131, ed. Seán Mac Airt and Gearóid Mac Niocaill (Dublin, 1983), 502–3. 100 For arguments for 1066, Baxter, The Earls of Mercia, 53, 299–300. 101 For this theme, George Molyneaux, ‘Why Were Some Tenth-Century Kings Presented as Rulers of Britain?’, TRHS, 6th ser., xxi (2011), 59–91, at 75–7; idem, The Formation of the English Kingdom, 33–8, 79–85. 102 Poitiers’ statement that the corn was green in Brittany during the campaign on which Harold accompanied William, GG, 74–5, locates the campaign in spring or early summer. For 1064, Barlow, Edward the Confessor, 220–1; Keats-Rohan, ‘The Breton Contingent’, 164–5. 103 See esp. Jean-Luc Chassel, ‘Le serment de Harold dans la Tapisserie de Bayeux et dans les sources pro-normandes des XIe et XIIe siècles’, in Raymond Verdier (ed.), Le serment, 2 vols. (Paris, 1991), 43–53; N. J. Higham, The Death of Anglo-Saxon England (Stroud,

o n to t h e at tac k

193

sion that, as in the case of the events of 1051–2, no agreed narrative of the significance of Harold’s visit to Normandy ever existed. While it is certain that he took an oath agreeing to support William’s succession to the English kingdom after Edward the Confessor’s death, this does not mean that the two of them understood either the content or the significance of the oath in the same way, even at the moment it was sworn. Also, both Harold and William’s narratives would have been contested by others, most notably by Edgar the Ætheling, Earl Tostig, and King Harold Hardraada of Norway, and probably also by Guy, count of Ponthieu. Some of these individuals would have held that Harold had no moral right to make the oath to William, and William had no right to take it from him. In trying to unravel the narrative, the starting point has to be Jumièges’s brief account and Poitiers’s much fuller one. In both, Harold’s visit to Normandy and his oath are at the heart of the case for William’s right to be king of the English. They are, however, no longer independent of each other, since these sections of Jumièges’s Gesta were written after 1066 and, like Poitiers’s account, must have drawn on what was effectively an official case in favour of William’s claim. Both make Harold’s journey a diplomatic mission authorized by King Edward to promise the succession to William, which was a follow-up to the offer made in 1051–2.104 As for other eleventh-century sources, Harold’s oath is mentioned by implication in the Carmen de Hastingae Proelio, which was almost certainly written by Guy, bishop of Amiens, quite probably before May 1068 and certainly before mid-1070.105 But all other accounts, including the Bayeux Tapestry, offer versions of events that are to a greater or lesser degree ambiguous. In their literary context, Jumièges and Poitiers are setting out the rationale that William wanted everyone to believe. Well characterized as ‘the exposition of a legal case’, Poitiers’s Gesta Guillelmi is actually that and much more.106 But in relation to Harold’s visit to Normandy, this is exactly how it should be treated. The presence of other personal narratives is acknowledged in all the other main accounts with the exception of Orderic, who, for the most part, follows Poitiers. His apparently surprising lack of independence is explicable because of his belief that the Normans’ conquest had brought 1997), 152–62; Ad F. J. van Kempen, ‘The Mercian Connection, Harold Godwineson’s Ambitions, Diplomacy and Channel-Crossing, 1056–1066’, History, xciv (2009), 2–19; Alban Gautier, ‘Comment Harold prêta serment: circonstances et interprétations d’un rituel politique’, Cahiers de civilisation médiévale, lv (2012), 33–56. 104 GND, ii, 158–61; GG, 68–77, 100–1, 120–1, 124–5. 105 Carmen, 16–17. 106 George Garnett, ‘Coronation and Propaganda: some Implications of the Norman Claim to the Throne of England’, TRHS, 5th ser., xxxvi (1986), 91–116, at 110; idem, Conquered England: Kingship, Succession and Tenure, 1066–1166 (Oxford, 2007), 40.

194

william the conqueror

catastrophe to the English people. For this reason, although, as we shall see, he was very critical of some of the things that William did, Orderic ultimately had to place the blame on Harold’s perjury because that had initiated the terrible chain of events.107 In the earliest source of all, the Vita Ædwardi Regis, which was being written by 1067–8 at the latest, the oath must surely be the basis for the enigmatic allusions to Harold as a notable man who knew well how to deal with the schemes of northern French princes – ‘by God’s grace, he came home safely passing through all ambushes with watchful mockery on the way’; hindsight crept in only with the observation that he was ‘alas, too free with oaths’.108 If this second observation was a reference to a general flaw in Harold’s character as opposed to the specific oath to William, it is still pertinent to the subject.109 The passages are arguably best interpreted as referring to the skill that made Harold and his family so capable of ruling a largely peaceful kingdom in the years after 1052, with a nod towards an act that had seemed to be clever, but which had had disastrous consequences. Its basis has to be a belief current in 1064, and perhaps still credible in 1067, that Harold had been skilful. Writing at the abbey of Saint-Riquier in Ponthieu in the late eleventh century, Hariulf ’s comment that the oath broken by Harold was to Edgar the Ætheling could, as we shall see, derive from one of the narratives that are completely different from the ones developed by Harold and William.110 A little later, Eadmer, writing at Canterbury, made Harold’s journey to France a personal one undertaken to secure the release of the two hostages from his family handed over to William in 1052, something that was done against King Edward’s advice. The oath to William was taken under duress.111 Both Hariulf and Eadmer introduce motives and narratives that are very different from those of Poitiers. The portrayal of Harold’s oath on the Bayeux Tapestry arguably makes it the most intriguing account of all (Plate 4).112 Over recent decades, its allusive captions, often ambiguous imagery, and likely audience, its treatment less often as a source that tells the story of William’s triumph and more frequently as one that in significant respects echoes Eadmer and the Vita Ædwardi Regis, together mean it is taken as telling a moral tale relevant to all participants in the story.113 It OV, ii, 134–9. For these passages, VEdR, 50–3, 80–1 (sed ille citius ad sacramenta nimis, proh dolor, prodigus). 109 Cf. VEdR, 80, note 195. 110 Hariulf, Chronique de l’abbaye de Saint-Riquier, 241. 111 HN, 6–7. 112 See Plate 4. 113 Seminal is N. P. Brooks and the late H. E. Walker, ‘The Authority and Interpretation of the Bayeux Tapestry’, ANS, i (1979), 1–34, 191–9, at 11–13 (reprinted in Richard Gameson [ed.], The Study of the Bayeux Tapestry [Woodbridge, 1997], 63–92, at 71–4). 107 108

o n to t h e at tac k

195

is almost universally accepted that the Tapestry was made at Canterbury, and therefore in close proximity to the sources of information that fed into Eadmer’s account and, possibly, the Vita Ædwardi. This in itself reflects Harold’s often favourable treatment in the embroidery’s narrative and his arguable presentation as a noble man – even a tragic hero – destroyed by one mistake.114 So viewed, the Tapestry has been located within genres ranging from the epic Chanson de Geste to monastic pedagogy. In relation to ethnicity, it has been seen as a product of a period of cohabitation, supposedly evident after 1066, which allowed Harold to be favourably presented.115 Whatever view is taken, the Tapestry becomes another of the distorting prisms reinforcing the argument that we cannot know ‘the facts’, only the different interpretations that the participants in the events and those who wrote about them could put on them.116 Arguably the most complex of the accounts is William of Malmesbury’s. After noting the narrative presented in Jumièges and Poitiers, the second of which he had certainly read, Malmesbury teased his readers (as he had done in relation to the events of 1051–2) by referring to a version of events provided by Harold and given credence by those who knew his secret intentions. Writing that he was inclined to believe most of what they said, without indicating what he did and did not believe, Malmesbury’s version of Harold’s visit to Normandy is that it began as a pleasure trip in the Channel that went badly wrong when his ship was blown off course. Captured by Guy of Ponthieu, Harold bribed someone to take a message to William and volunteered to take the oath as a means of persuading William to extricate him from imprisonment by Guy.117 By proclaiming 114 For a survey covering the Tapestry’s treatment of Harold, H. E. J. Cowdrey, ‘King Harold II and the Bayeux Tapestry: A Critical Introduction’, in Gale R. Owen-Crocker (ed.), King Harold II and the Bayeux Tapestry (Woodbridge, 2005), 1–15. On arguments for production at Winchester, Carola Hicks, The Bayeux Tapestry: The Life-Story of a Masterpiece (London, 2006), and ‘The Patronage of Queen Edith’, in Michael J. Lewis, Gale R. OwenCrocker, and Dan Terkla (eds.), The Bayeux Tapestry: New Approaches. Proceedings of a Conference at the British Museum (Oxford, 2011), 5–9. For Normandy, Wolfgang Grape, The Bayeux Tapestry: Monument to a Norman Triumph (Munich and New York, 1994). For Saint-Florent of Saumur, George Beech, Was the Bayeux Tapestry Made in France? The Case for Saint-Florent of Saumur (London and New York, 2005). 115 Compare, e.g., C. R. Dodwell, ‘The Bayeux Tapestry and the French Secular Epic’, Burlington Magazine, cviii (1966), 549–60 (reprinted in Gameson [ed.], The Study of the Bayeux Tapestry, 47–62); Pierre Bouet, ‘Is the Bayeux Tapestry Pro-English?’, in Bouet, Levy, and Neveux (eds.), The Bayeux Tapestry, 197–215, at 213–14; Pierre Bouet and François Neveux, ‘Edward the Confessor’s Succession According to the Bayeux Tapestry’, in Lewis, OwenCrocker, and Terkla (eds.), The Bayeux Tapestry, 59–65; Pastan and White, The Bayeux Tapestry and its Contents, 80–1, 288–92. 116 For an approach close to mine, Pastan and White, The Bayeux Tapestry and its Contents, 228–36. 117 GR, i, 416–19.

196

william the conqueror

that he had assessed the information at his disposal, Malmesbury conveys an analytical authority that is appealing to modern minds. Yet, as becomes apparent in his treatment of the year 1066, we need to remember that he believed it was opportunism of this kind on the part of England’s secular aristocratic elite that had brought the kingdom to ruin. A little later, Henry of Huntingdon modified the basic story slightly by having Harold set out for Flanders, only to be driven off course to Ponthieu by a storm, but concurring that his subsequent oath to William made him a perjurer in 1066.118 Robert of Torigni, as was his customary practice on matters relating to the Historia Anglorum, followed Henry of Huntingdon’s version in preference to anything written in Normandy, and said the same.119 In the 1160s, Wace, who had read most of the earlier accounts and explicitly said that he had done so, referred to ‘whatever business he [i.e. Harold] was pursuing and whatever he intended to do’, adding reflections to the effect that the best-laid plans do not always turn out as intended. His account is closest in tone and content to those of Eadmer and Malmesbury. Wace believed that Harold was tricked into taking the oath at Bayeux.120 One way of making sense of these contradictory narratives is to keep in mind that the presence of the hostages transferred to Normandy was always going to jog memories that some sort of agreement had been made between Edward and William in 1051–2. As a result, everyone would have known that William had an interest in the English succession. Few, however, would surely have had any idea what he intended to do about it, and, before 1060, whether he could possibly have done anything anyway. For this reason, and the ‘official’ version in Jumièges and Poitiers notwithstanding, it is very likely that Harold did not know what he was letting himself in for when he travelled to Normandy; and that Edward, even if he did send Harold to William as his ambassador, would not necessarily have had any preconceived outcome in mind either. The improvement in William’s circumstances after 1060 might well have made a fact-finding mission to the Continent seem desirable to Harold and Edward, since William was manifestly in a position to be actively interested in the succession in a way that had been impossible while he still had to contend with Henry I and Geoffrey Martel. The idea, unique to Malmesbury, that Harold’s visit to France was a pleasure trip, seems to me incredible, for all that it might have been beneficial to Harold’s cause to have people believe that it was. However, it does reflect the fact that, in relation to what we HH, 380–3. Chronique de Robert de Torigni, i, 50. For this aspect of Robert’s use of Henry, David Bates, ‘Robert of Torigni and the Historia Anglorum’, in Roffe (ed.), The English and their Legacy, 175–84, at 178–81. 120 Roman de Rou, 220–3, lines 5,583–724, with the quotation at lines 5,605–12. 118 119

o n to t h e at tac k

197

know for certain, Harold’s landfall in Ponthieu is arguably the most surprising facet of the narrative. The Channel has always been subject to sudden storms, but medieval navigators were not normally so incompetent as to miss their intended destination so dramatically. For all the apparent agreement across all sources, it is hard to believe that Guy of Ponthieu would have incarcerated the second most powerful man in England without having a special reason for doing so, even if, as everyone is suggesting, Harold had arrived unannounced. Guy would have been familiar with the details of the politics of the English succession and who Harold was; they had after all met in 1056 and, when messengers from William arrived, he handed Harold over very quickly, as if Guy knew that this is what he should do. The Tapestry does seem to disagree with most of the written accounts in not portraying a storm blowing Harold off course. The image of ships in full sail, reinforced by the caption uelis uento plenis, suggests a fast, smooth, crossing. But the effect of the storm is in line with the well-nigh universal agreement that Guy detained Harold; the caption Hic Apprehendit Wido Haroldum, the scene below it of Harold being manhandled, and the portrayal of his posturing before Guy as a supplicant, all reinforce this (Plate 5 (a) and (b)).121 Poitiers wrote that Harold and his men were taken into custody.122 In relation to the ‘secret histories’ lost beneath the surface, it may be significant that, as events unfolded, everyone had a motive for making Guy the villain of the piece.123 For William, rescuing Harold reinforced the argument for the generosity with which his eventual rival was treated and for his personal superiority. For Harold, his capture by Guy explained why he ended up in William’s debt and at his mercy. It is possible that Harold was detained by Guy for longer than he might have wished. Or perhaps, for longer than others – most notably William – might have wished. It is possible, too, that there was an element of collusion on Harold’s part. Knowing that he was going to have to go to Normandy, he may have decided it was necessary to arrive there on terms that were apparently not of his choosing. An additional unknowable is whether Harold was at this point conspiring to be the next king of the English. All that can be said is that in 1064 there was still a long way to go before that could happen. Procedures that involved a designation by Edward and/or the agreement of the main nobles and clergy of the kingdom still had to be negotiated, and there was no certainty about the outcome; Edgar the Ætheling must Arnold J. Taylor, ‘ “Belrem” ’, ANS, xiv (1993), 1–24, at 6–9. GG, 68–9 (Capti in custodiam traduntur). 123 For a shared scepticism about the ‘official’ version of Guy’s role, K. S. B. Keats-Rohan, ‘Through the Eye of the Needle: Stigand, the Bayeux Tapestry and the Beginnings of the Historia Anglorum’, in Roffe (ed.), The English and their Legacy, 159–74, at 164–5. 121 122

198

william the conqueror

surely still have been considered Edward’s most likely successor. Harold had to behave like a devoted supporter of the king. If Harold actually was blown off course to Ponthieu, then a stay with Guy might have suited both of them. Distrust of William, at whose hands he and his family had suffered considerably, could well have been a factor with Guy. He would surely have wanted to discuss the possibilities that might open up in relation to the English succession. If William’s interest was publicly known – as it surely was – then the impact on Guy of the potential union of Normandy and England would be considerable. Harold would have been interested in the positions that Guy and his ilk were likely to adopt in the event of William taking action to succeed Edward. If possible, it would also have been desirable for Harold to have discussions with Eustace of Boulogne. But once the embassy from Normandy arrived, force majeure would have dictated that Harold had to go there. For all that elements of fiction may have been involved in the story that circulated about this, for Harold to have refused the hospitality of his (probably) selfstyled rescuer would have been unthinkable. And it would have been dishonourable for him not to look into the welfare of the two hostages, who were after members of his own family. After Guy had delivered Harold to the Norman frontier at Eu, William escorted him to Rouen. He then looked after him for some considerable time, perhaps as much as three months, taking him round Normandy and on a campaign against Count Conan II of Brittany. The agreement they reached is described in its fullest detail by Poitiers. Harold swore fealty to William, promised to help him succeed to the English kingdom, and agreed to hand Dover and other fortified places over to him. In return, William confirmed Harold in all his possessions and offices, promised him a daughter in marriage, and allowed him to take home one of the two hostages.124 The daughter concerned is named by Orderic as Adelida, who had already perhaps been the putative bride of Count Herbert of Maine. The unlikely prospect was of a girl less than ten years old being betrothed to a man in his early forties who was already in a long-established relationship that had produced several children.125 The enigmatic and much-discussed Ælfgyva scene on the Bayeux Tapestry may well represent another strand in their discussions – perhaps about a marriage alliance involving an otherwise scarcely known sister of Harold, who had died by 1066, and a member of GG, 70–1, 156–7. GND, ii, 160–1. In the Historia Ecclesiastica, Orderic names the daughter betrothed to Harold as Agatha, OV, iii, 114–15. It is generally assumed that Adelida and Agatha were the same person, as OV, iii, 114, note 1; GND, ii, 262, note 2; Barlow, William Rufus, 442–3. Robert of Torigni believed that Adelida was the third eldest daughter born to William and Matilda, GND, ii, 262–3. 124 125

o n to t h e at tac k

199

William’s family, who might have been Robert, count of Mortain. This forgotten element in the narrative is actually mentioned by Eadmer.126 The hostage who returned to England with Harold was his nephew, Hakon. The apparent contradiction between Poitiers and the Bayeux Tapestry over the location of the oath, with the former placing it at Bonneville-surTouques before the Breton campaign and the latter apparently at Bayeux after it, might be explained by a double ceremonial endorsement of the agreement, a secular oath confirmed by a religious affirmation on relics, or by a very likely reading of the Tapestry’s caption that argues it is not locating the oath at Bayeux at all.127 Orderic’s statement that the oath was made at Rouen may therefore be a reference to earlier discussions and to a first stage of the agreement; because Rouen was the duchy’s chief town, it was logical for Orderic, and for others as well, to believe that the oath must have been sworn there.128 Whatever the sequence of events, given the amount of time that they probably spent in each other’s company, and for all that a language barrier must have restricted communication, Harold and William must have come to know each other pretty well during this period. Apart from anything else, these dramatic months served to make William’s intentions clear to everyone with an interest in the future of the English kingdom. Whether he could fulfil them was of course another matter, but securing Harold’s support must have seemed a good route to try to ensure a peaceful transfer of the kingdom to him. Although there is no specific evidence for how the news of what was in effect a declaration of intent was received in England, or anywhere else, later events presumably indicate that the idea of William as king of the English was not greeted with rapturous enthusiasm anywhere. The main outcome of these events in Normandy may well therefore have been to concentrate minds, not just in England, but in Norway and Denmark as well. Harold, however, had seemingly passed through many snares. It was entirely logical for him to think that circumstances would arise that would invalidate his oath.129 From his perspective, the mission may not have gone as planned or hoped. But the situation was redeemable. And Edward the Confessor’s continued good health made impenetrably obscure the future course of events. 126 HN, 8. Most recently on this scene, Keith Moore, ‘Ælfgyva and the Cleric of the Bayeux Tapestry’, Foundations, vii (2015), 79–117, with the conclusions at pp. 111–17. 127 Pastan and White, The Bayeux Tapestry and its Contents, 107–21, with references to the many other studies. 128 Cf. OV, ii, 135, note 5. The comment there that Orderic’s naming of Rouen resulted from a misreading of Poitiers seems to me unlikely; although he actually contradicted himself on aspects of the agreement between William and Harold, Orderic did not normally make mistakes of the kind suggested on important matters. For a summary of the arguments for Rouen, Pastan, and White, The Bayeux Tapestry and its Contents, 113–15. 129 Gautier, ‘Comment Harold prêta serment’, 51–3.

200

william the conqueror

As a final commentary on Harold’s mission and the oath extracted from him, and on the invisible histories that never happened, the ‘D’ Chronicle’s statement in its 1066 annal that kingship was Edgar the Ætheling’s ‘natural right’ must be significant.130 And then there is Hariulf ’s comment that the oath which Harold broke was in fact to Edgar the Ætheling.131 Was Hariulf saying that, because of this, the oath Harold made to William was invalid because he was already committed and that William’s claim was therefore invalid, too? Hariulf ’s abbey of Saint-Riquier was in the lands of Guy, count of Ponthieu. Are we in the presence of unspoken thoughts that could be euphemistically translated as something like ‘a plague on both your houses’? As we have seen, the notion that William’s behaviour was dishonourable and that he was a deceiver is present in many of the sources. It may be that Hariulf is quietly saying that Guy had been manipulated by two unscrupulous people. THE BRETON CAMPAIGN William of Poitiers, the Bayeux Tapestry, Orderic, and Wace are the only narrative sources for the campaign against Count Conan II of Brittany on which Harold accompanied William. Poitiers justifies it on the grounds of Conan’s provocative behaviour and his defiance of Norman lordship over Brittany which, so he says, had been created by King Charles the Simple’s grant to Rollo. Poitiers thereby manufactures yet again a historical and justificatory context for William’s intervention, which also included his close kinship to the Breton count.132 He also laces his narrative with an ethnic denigration of the Bretons, borrowing language from Sallust and Caesar, and condemning them for polygamy, agricultural backwardness, and lack of respect for chivalrous customs in war. Rarely in the Gesta Guillelmi does he parade the classical vocabulary of barbarism so extravagantly. But in much that he wrote Poitiers was basing his account on Dudo of Saint-Quentin, with similar themes appearing in Jumièges’s Gesta Normannorum Ducum.133 The campaign is probably best understood in the context of the strengthening of alliance systems to the west of Normandy that was the equivalent of what had already been accomplished to the east and south. ASC, ‘D’, 1066. Hariulf, Chronique de l’abbaye de Saint-Riquier, 241. 132 GG, 70–7. 133 Dudo of Saint-Quentin, De Gestis Normannie, 184, has Dedit itaque rex filiam suam, Gislam nomine, uxorem illi duci . . . totamque Britanniam de qua posset vivere (trans. Christiansen, 62); GND, i, 64–5. For discussion and references, GND, i, 65, note 7; Dudo, De Gestis Normannie, trans. Christiansen, 195 note 204. 130 131

o n to t h e at tac k

201

It has its specific background in the close, often amicable, but sometimes tense and fractious, long-term relations between Normandy and Brittany, with William’s father having campaigned in the border regions and in all probability consolidated Norman power in the lands to the south of Mont Saint-Michel. Within William’s lifetime, Conan’s father Count Alan had died fighting in Normandy on the young William’s behalf in 1040. Alan’s brother Eudo had also intervened constructively during William’s childhood, but Eudo had later fought alongside Geoffrey Martel at Ambrières. In this context, William’s advance into Brittany was probably a response to issues connected to short-term loyalties and territorial control, rather than a consequence of any long-term tension. The campaign must also be located within the internal politics of the duchy of Brittany, from Conan’s accession as a minor in 1040 onwards. And it must have a place within the processes that produced massive support from within sections of Brittany’s aristocracy for the invasion of England in 1066.134 Identified by Poitiers as the agent provocateur who had called on William to campaign in Brittany, Rivallon de Dol was a major landholder in eastern Brittany, with interests that stretched across the regions to the north of the Loire. He had seemingly gravitated into William’s orbit, perhaps in the aftermath of the appointment of William’s brother Robert as count of Mortain, and more decisively after the conquest of Maine, but he seems also to have remained on good terms with Count Conan. Rivallon’s only appearance in William’s charters is at Domfront at the settlement of the dispute between the abbeys of Saint-Pierre de la Couture of Le Mans and Marmoutier, which followed the conquest of Maine. He had a considerable local interest in the affair because he was a benefactor of the great abbey of Marmoutier; its political significance is therefore questionable.135 Although Rivallon’s appearances in Conan’s charters are difficult to date with any precision, his major grant to the abbey of Marmoutier, which Conan confirmed, seems to date to not long before 1063.136 William had, however, seemingly worked to attract his support, transferring to Rivallon an estate from the lands of the abbey of Mont Saint-Michel

134 For the most recent studies, François Neveux, ‘L’expédition de Guillaume le Bâtard en Bretagne (vers 1064)’, in Joëlle Quaghebeur and Sylvain Soleil (eds.), Le pouvoir et la foi au Moyen Age en Bretagne et dans l’Europe de l’Ouest: Mélanges en mémoire du professeur Hubert Guillotel (Rennes, 2010), 619–37; Katharine S. B. Keats-Rohan, ‘L’expédition de Guillaume, duc de Normandie, et du comte Harold en Bretagne (1064): le témoignage de la tapisserie de Bayeux et des chroniqueurs anglo-normands’, MSHAB, xl (2013), 203–24. 135 RADN, no. 159. 136 Les actes des ducs de Bretagne, no. 57. Although the limits are 1047x1064, arguments are given there for a date ‘largement postérieur à 1047’. See also ibid., no. 65.

202

william the conqueror

at Céaux.137 William had also reinforced his power in the sector by having a castle built at Saint-James-de-Beuvron within Normandy, to the south of Avranches and just to the east of the river Couesnon.138 Poitiers identifies the purpose of the campaign as being to help Rivallon defend Dol against Conan, who was treating him as a rebel. While this may have been an element that was present, a campaign on such a scale must surely have had more profound causes, the most likely being to send a warning shot across Conan’s bows and to re-establish good relations with Count Eudo and his family. With all of them disappearing from the charters of the counts of Anjou soon after Geoffrey Martel’s death in 1060, and presumably thereby abandoning the alliance that had seen Eudo support Geoffrey at Ambrières, it looks as if Eudo and his sons decided to throw in their lot with William. How otherwise to explain the massive profits that accrued to Eudo’s sons Brian and Alan after 1066? And although Eudo’s imprisonment by Conan in 1057 may not have lasted long, its tensions undoubtedly persisted, creating a rationale for closer ties with William and for William to show how much he appreciated them. Although he must have been quite elderly by this time, Eudo visited Normandy soon after 1066, as if to secure arrangements that had been made beforehand.139 William’s advance to Dol caused Conan to retreat. According to the Tapestry, William then advanced southwards and attacked Rennes, before turning north and attacking Dinan, where Conan surrendered. However, in Poitiers’s account, Conan, presumably in anticipation of William’s intervention, had called in Count Geoffrey the Bearded who, this time, did respond. His arrival created a situation in which a decisive outcome, if anyone wanted one, could only be achieved by battle. William therefore chose to leave, neither the first nor the last time that he and a count of Anjou would draw back from a major conflict. Rivallon, so Poitiers tells us, had also started to complain about his lands being ravaged by all these armies. As a result, he came to terms with Conan, as shown not only by Conan confirming a grant to Marmoutier that Rivallon had made, but then, after Rivallon’s death, reconfirming it for his son William, his widow, and William’s other brothers; both diplomas must date to after 20 May 1064 and to before Conan’s own death on 11 December 1066, with the 137 For the main evidence, The Cartulary of the Abbey of Mont-Saint-Michel, no. 115; Regesta, no. 269. See further Keats-Rohan, ‘The Breton Contingent’, 166–7; van Torhoudt, Centralité et marginalité, 529–30. 138 GG, 72–3. 139 For this line of argument, Morin, Trégor, Goëlo, Penthièvre, 98–9; Keats-Rohan, ‘L’expédition de Guillaume, duc de Normandie’, 216–20. For Eudo’s visit to Normandy, Judith A. Everard, ‘The Foundation of an Alien Priory at Linton, Cambridgeshire’, Proceedings of the Cambridge Antiquarian Society, lxxxvi (1997), 169–74, at 173, no. 4.

o n to t h e at tac k

203

second diploma being confirmed at Dol, showing that Rivallon was effectively in possession of all that he had previously held.140 The conclusion that this Breton campaign was largely a piece of theatre is inescapable. Yet it probably achieved what William wanted. Through Harold’s presence, it publicized William’s claim to England; Wace illustrates this by mentioning that a relatively humble aristocrat, William de La Lande-Patrie, later recalled Harold passing through his village on his way to Brittany to join William who was already at Avranches.141 The campaign also recreated the alliance with one of the ruling families of Brittany that had existed in the days of William’s grandfather, this time to many people’s ultimate profit. Showing off Harold would have been an excellent recruiting device. Also, for all the instability of the political situation in Brittany, the campaign seems to have brought Conan round to William’s side, since in 1066 he set out to make gains at the expense of his former ally Geoffrey the Bearded by invading Anjou and attacking the major castle of ChâteauGontier, one of the strongpoints on Anjou’s western boundaries. Although this could have been another element within the long-standing territorial competition in northern France, it might also have been a device to develop larger ambitions. Conan’s death at Château-Gontier is attributed by Orderic to poison administered on William’s behalf.142 Since this would have happened after William had invaded England, this, more than almost any other of Orderic’s assertions of poisoning reads as if it is a reflex reaction to a politically advantageous death. The possibility of an agent acting on William’s behalf cannot, however, be ruled out. Whatever the truth of the matter, the story indicates that William was thought to be capable of exercising power far beyond the lands that he controlled directly; Conan’s death removed a dangerous loose cannon who might have proved troublesome.143 Although Conan’s ambitions, whatever they were, died with him, they may make sense if we think for a moment about the possibilities that might have opened up if William had been defeated in England. In this scenario, Orderic’s apparently bizarre statement, that Conan claimed he was the rightful ruler of Normandy as the son of Count Alan, could have been a way to stake a claim to Maine and

Les actes des ducs de Bretagne, nos. 66, 67. Roman de Rou, 280–1, lines 8,585–94. 142 GND, ii, 162–5. 143 For elements of this argument, W. Scott Jessee, ‘The Angevin Civil War and the Norman Conquest of 1066’, HSJ, iii (1991), 101–10, at 106–7. One element in this argument that must be abandoned, however, is the possibility of a visit to Tours and Blois in the company of Geoffrey de Mayenne. For a re-dating of the relevant charter, Les actes des ducs de Bretagne, no. 54. 140 141

204

william the conqueror

Normandy if William had failed in England. Here again it is important to think about the secret histories that lie beneath the surface. RULE IN NORMANDY, 1063–6 The dedication of the metropolitan cathedral of Rouen on 1 October 1063 was a great religious event, the culmination of decades of work that had started in Archbishop Robert’s time and of a major building campaign in the 1050s. Although nothing of the church now survives above ground, excavations of the former crypt and in the vicinity of the cathedral suggest that it was on a monumental scale and had numerous innovative features, bearing architectural and stylistic parallels with several of the great contemporary churches of the French kingdom.144 The timing of the dedication fits a pattern that was repeated in and after 1066, and also after William’s significant military setback in 1076, of holding a great ecclesiastical ceremony as a preparation for a campaign, as a gesture of thanksgiving for a victory, and in atonement for a defeat. The dedication could also have been associated with William’s recovery from illness, if the 1062 date for that is accepted. Whatever the case, in William’s mind and the minds of his advisers, the divine and the worldly aspects of rule were being co-ordinated. Another aspect of this period of intense religious activity was the holding of an ecclesiastical council at Lisieux in 1064. Held, so we are told, under William’s authority, with Archbishop Maurilius of Rouen presiding, and bishops and abbots present whose names are not given, the legislative record is the first to survive since that of Archbishop Malger’s council during William’s adolescence.145 The council looks to have been a wideranging statement of considerable importance. Indeed, it may well have been a significant political moment in the preparations for becoming king of the English. Although we do not know the date on which the council was held within the year 1064, its chronological proximity to the public declaration of interest in the English succession and its ideological parallels with the very latest papal policies cannot be a coincidence. Whether it took place before or after Harold’s oath is immaterial. The council surely set out to confirm William’s worthiness as a ruler who protected the Church and who sought to improve its condition. In reaffirming the Truce of God and recording that a discussion was held concerning the Eucharist, the council continued traditions established 144 Baylé, ‘Les évêques et l’architecture normande’, 153–8. For the date, OV, iii, 92–3; AAR, 39–40, 53–4, with a discussion of the chronological inconsistences in the passage at pp. 15–17. 145 Léopold Delisle, ‘Canons du concile tenu à Lisieux en 1064’, Journal des Savants (1901), 516–21, with the canons printed at p. 517.

o n to t h e at tac k

205

since the 1040s. The second of these also shows that distancing Normandy from Berengar’s theories was deemed to be as important as ever and, perhaps all the more urgent in the early 1060s in order to show solidarity with the papacy after Berengar’s condemnation at the 1059 Rome synod. The composition in the early 1060s of the rhetorical De corpore et sanguine Domini, attributed to Lanfranc, shows that this responsibility was being taken seriously by those with the duty of maintaining orthodoxy within the Norman Church.146 In addition, the prohibition of priests in rural parishes, and of deacons having either wives or concubines and the specific order that they should separate from them, also shows how the recent papal programme was taken up; Nicholas II (1059–61) had been the first pope to legislate generally for all Christians, although local prohibitions and papal admonitions had been made during previous decades. While the canon’s severity was moderated by an exemption granted to clergy in minor orders, the legislation is above all important because of the way in which it placed Normandy in the forefront of the programme that was going to bring about a social revolution. By referring back to an earlier Rouen council, which had presumably been held in 1063 at the time of the cathedral’s consecration, the Lisieux council showed that such strictness was being incorporated into the traditions of the Church in Normandy. It also became the benchmark to which the next Norman council from which legislation survives, the Council of Rouen of 1072, made reference.147 It is easy, and perhaps justifiable, to associate this display of moral austerity with William’s personal influence. William of Malmesbury’s comments on his chaste adolescence, his apparent fidelity in marriage, and his protectiveness of his mother’s memory might suggest as much. There is, however, a huge difference between legislating and acting and, as we shall see, after the 1072 Rouen council William intervened against the overzealous implementation of its almost identical decrees. But even if the 1064 Lisieux council was to an extent window-dressing, it is still important. William evidently wished to present himself and the duchy of Normandy as in line with the most up-to-date aspirations of a vigorous papacy. The council’s legislation also set out to protect clerical property in various ways. False 146 Holopainen, ‘ “Lanfranc of Bec” and Berengar of Tours’, 105–8, 120–1; idem, ‘L’héritage intellectuel de Lanfranc revisité’, in Barrow, Delivré, and Gazeau (eds.), Autour de Lanfranc, 107–15. 147 OV, ii, 290–1. For this subject, see the pioneering essay by C. N. L. Brooke, ‘Gregorian Reform in Action: Clerical Marriage in England, 1050–1200’, Cambridge Historical Journal, xii (1956), 1–21 (reprinted in idem, Medieval Church and Society: Collected Essays [London, 1971], 69–99) and, most recently, with translations of the legislation, Elisabeth van Houts, ‘The Fate of Priests’ Sons in Normandy with Special Reference to Serlo of Bayeux’, HSJ, xxv (2014, for 2013), 57–105, at 58–63, 79.

206

william the conqueror

religious confraternities that actually were fronts for drinking clubs were banned. A commentary on the ever-present culture of violence embedded in society is provided by the statements forbidding the clergy to bear arms and prohibiting attacks on them, unless they were justified; in which case the bishop’s permission should be sought before resorting to violence. ON THE EVE OF 1066 The years since the mid-1040s had shown that William possessed the attributes of a medieval warrior ruler: a high level of skill in the practice of war; a capacity to sustain morale through reward; and, ultimately, the ability to command obedience and loyalty through charismatic qualities that mixed the power both to instil fear and to give reassurance. The young elite that had emerged in the 1040s was now mostly in control of everything. Moreover, the years since 1060 had witnessed a remarkable change in political circumstances in northern France and the cultivation of close relations with Pope Alexander II that were going to be extremely important. The development of Caen represented a massive statement of William and Matilda’s standing as rulers. Their readiness to adapt the landscape through large building projects was also evident in the construction of new walls around Rouen, which remained the duchy’s chief city.148 Another instance of this type of social engineering is the development of Lillebonne, already an earthwork enclosure, but only first mentioned as a ducal residence in 1061.149 The nearby forest made it an exceptionally good place for hunting. It was to be the site of the great consultation that, at least in theory, ended with an undertaking to embark on the invasion of England in 1066. This position of strength within Normandy was based not just on William’s achievements and reputation, but on the duchy’s evolution from its tenth-century beginnings under a succession of rulers. This process had made it, in contemporary terms, a notably resilient eleventh-century state, something that, as we have seen, had already been a subject of comment almost a decade earlier.150 There were strong links between centre and localities, an awareness of hierarchy, a culture of legal documentation with a capacity on the part of the dukes to legislate for the entire duchy, and control over a coinage which, although not as sophisticated as the English 148 Bernard Gauthiez, ‘The Urban Development of Rouen, 989–1345’, in Hicks and Brenner (eds.), Society and Culture in Medieval Rouen, 17–64, at 28–30. 149 Le Maho, ‘L’apparition des seigneuries châtelaines’, 16–17. 150 Ut expressius dicatur quod difficile in aliis reperies, totius terre sue regem, RADN, no. 137. See above, chapter 4.

o n to t h e at tac k

207

system, was still a ducal monopoly that allowed re-coinages.151 The economy was prosperous and expanding. Towns were growing and many small markets towns had been founded since the early eleventh century.152 The Church, both secular and regular, had evolved steadily since the great creative period of Duke Richard II’s reign. Within the seven bishoprics of the duchy, cathedral chapters were being established around large new cathedrals and jurisdiction was being defined and developed. The re-founded Carolingian abbeys of Fécamp, Mont Saint-Michel, SaintOuen of Rouen, Jumièges, and Saint-Wandrille remained the main powerhouses of monasticism, but new houses – and of course, above all, Le Bec – were also evolving within the landscape. For all that there had been many changes, an awareness of the Scandinavian past must still have been present in many minds; there were, for example, many place names, personal names, and much nautical language that derived from those times. This sense of distinctiveness was ultimately a powerful reinforcement for a strong political, social, and cultural identity that William’s thirty years as duke had further nurtured.153 One certain departure from the tenth-century beginnings that was to prove important was that, over the generations, there had been a complete assimilation to Frankish methods of warfare. Leaving aside until the next chapter the debates about the possible technological superiority of the cavalry and infantry force that won the Battle of Hastings by defeating an exclusively English infantry army, there is no doubt that the warriors who rode and fought with William conformed to the highest standards of their times. For all this, however, the fissiparous nature of eleventh-century society must never be forgotten. While Chapter 2 argued that the disruptions of William’s adolescent years were contained within the structures of ducal 151 In addition to the works already cited, important recent studies are Gilduin Davy, Le duc et la loi: Héritages, images et expressions du pouvoir normative dans le duché de Normandie, des origines à la mort du Conquérant (fin du IXe siècle–1087) (Paris, 2004); Mark Hagger, ‘Secular Law and Custom in Ducal Normandy, c.1000–1144’, Speculum, lxxxv (2010), 827–67; Jens Christian Moesgaard, ‘Renovatio Monetae et la chronologie des monnaies de Richard Ier, duc de Normandie 942–945–996’, Bulletin de la Société française de numismatique, lxvi (2011), 125–33; idem, ‘Saints, Dukes and Bishops: Coinage in Ducal Normandy, c.930–c.1150’, in Giles E. M. Gaspar and Svein H. Gullbeckk (eds.), Money and the Church in Medieval Europe, 1000–1200 (Farnham and Burlington, VT, 2015), 196–207; Benjamin Pohl, Dudo of Saint-Quentin’s ‘Historia Normannorum’: Tradition, Innovation and Memory (Woodbridge, 2015). 152 For a survey, Bernard Gauthiez, ‘The Evolution of Towns in the Duchy of Normandy in the 11th–15th Centuries: An Essay in Historical Geography Using Material and Quantitative Data’, in Andrzej Buko and Mike McCarthy (eds.), Making a Medieval Town: Patterns of Early Medieval Urbanization (Warsaw, 2010), 81–103, at 85–8. 153 For stimulating ways of thinking about tenth-century Normandy, see Lesley Abrams, ‘Early Normandy’, ANS, xxxv (2013), 45–64. For significant aspects of the subject, see the essays in Sylvette Lemagnen (ed.), La Tapisserie de Bayeux. Une chronique des temps Vikings? Actes du colloque international de Bayeux, 29 et 30 mars 2007 (Bonsecours, 2009).

208

william the conqueror

rule and that ducal authority remained resilient, the cultures of feud, honour-based violence, and the expectations of reward for performance remained of central importance. In choosing to cross the Channel to try to remove Harold, William was also entering new sociological territory in his relations with those who followed him and those he claimed to rule. Removed from the realms of the institutional and the religious, it is also extremely important that the way in which the very foundations of William’s personal power had been developed had been offensive to many. The history of the years from 1047 to 1066 is littered with histories of those who had rebelled, who had been destroyed, or who were only partially reintegrated. Some of Orderic’s narratives dealing with the histories of families of the founders and patrons of Saint-Evroult show power exercised in a partisan way to secure domination through the agency of favoured individuals or to intimidate. Orderic’s references to poisoning are a reminder of just how ruthless was William’s exercise of power. This remains so even if none of his examples are actually true and if, as is often the case, Orderic judiciously distances himself from explicitly saying that William was guilty of poisoning his enemies by using phrases such as ‘so they say’. Throughout the entire Middle Ages, poisoning was generally seen as the most heinous of all ways of taking a life by violence, an underhand method of killing that gave the victim no opportunity to defend him- or herself and, if the poison acted quickly, no chance to make confession before death.154 Orderic’s references are ultimately therefore metaphors for power amorally exercised, the product usually of discussions with the many people on whom he relied for information that reflected widespread opinion. In reaching a verdict on William for the years before 1066, it can sometimes feel that the sources we possess tilt the narrative too far towards examples of peace-making, charity, and security. Just how ruthless a ruler William was can never truly be known. But the extreme possibilities must never be forgotten. Once more we move beyond old controversies by recognizing the difficulties of resolving them. Yet what is surely certain is that the years covered by this chapter, like those of the previous decade, are characterized by his harddriven tactical manoeuvring towards clearly defined goals. We can see why, when William set off to take the English kingdom in 1066, so many people trusted in his capacity to succeed. And why the final verdict of so many historians of his life was so deeply ambivalent.

154 For the basis of these observations, Franck Collard, Le crime de poison au Moyen Age (Paris, 2003), 9–48, 137–45 (The Crime of Poison in the Middle Ages, trans. Deborah Nelson-Campbell [Westport, CT, 2008], 1–38, 127–34).

o n to t h e at tac k

209

Meanwhile, in England, probably in 1065, Waltheof, son of Earl Siward, who had been replaced as earl in Northumbria by Harold’s brother Tostig, was made earl in several shires in the East Midlands (possibly Northamptonshire, Huntingdonshire, Cambridgeshire, and Bedfordshire).155 An acknowledgement of his family’s standing, it can be seen as recognition of Waltheof ’s entitlement and also as a promotion made at a suitable time: he was around fifteen years old. Given Harold’s extraordinary pre-eminence in affairs by this time, Waltheof ’s advancement must have been made with his approval, the presumed intention being to stabilize relationships across the families of the greatest earls. However, the English kingdom’s political stability was destroyed when an uprising in the north in late 1065 brought about Tostig’s expulsion from the earldom of Northumbria. Occurring in October, while Tostig was hunting with King Edward in Wiltshire, it seems to have been caused by resentment at what was regarded as his oppressive rule and heavy taxation. There had also been signs of a build-up of resentment in the murder at the royal court on Queen Edith’s orders of the Northumbrian noble Gospatric on 28 December 1064, and of two thegns at York in Tostig’s own chamber during the previous year. It is yet another indication of the rivalries among England’s elite that periodically, as in 1051–2, brought the kingdom close to civil war, but which were usually resolved by compromises. It is also a reminder that, despite modern historiography’s emphasis on the sophistication and resilience of the later tenthand early eleventh-century English kingdom, it was far from being unified. Tostig had been the first earl from the south appointed to exercise power north of the river Tees and he had not been popular there.156 This situation was to have profound and immediate implications when William became king of the English. The rebels marched south, rendezvousing with Edward on 28 October 1065. The king seems to have advocated an invasion of the north to restore Tostig by force, but the idea was not popular. Harold counselled negotiation and appeasement, and the result was that Earl Edwin’s brother Morcar was installed in Northumbria, a compromise candidate apparently acceptable to all. Tostig, however, accused his brother Harold of betraying him and left the kingdom with his family for Flanders, where he was welcomed by Count Baldwin. The author of the Vita Ædwardi Regis, very reasonably, but very obscurely in his actual treatment, saw the quarrel among brothers 155

See C. P. Lewis, ‘Waltheof, Earl of Northumbria (c.1050–1076)’, ODNB. On the importance of seeing the situation in the long term, W. E. Kapelle, The Norman Conquest of the North: The Region and its Transformation, 1000–1135 (London, 1979), 87–94; William M. Aird, ‘Northumbria and the Making of the Kingdom of the English’, in Hirokazu Tsurushima (ed.), Nations in Medieval Britain (Donnington, 2010), 45–60, at 52–5; Molyneaux, The Formation of the English Kingdom, 6–9. 156

210

william the conqueror

as the prelude to disaster.157 As noted, such situations had been managed before, above all in 1051–2. It is possible that a similar managed process was envisaged to be going into operation in late 1065. At the end of December of that year most of the great nobles of the English kingdom would have been at court for Christmas and for the dedication on 28 December of Edward’s new church at Westminster. During the festivities, the king fell ill and declined rapidly, dying on 4 or 5 January 1066. The suddenness of his death created a crisis. Within days, Harold had formally succeeded him (Plate 6 (a) and (b)). On the Bayeux Tapestry, Harold’s coronation is represented as an almost wholly correct ceremony. The only discordant element is the appearance in the scene of Archbishop Stigand of Canterbury, the canonical irregularity of whose position might have been intended as a reference to Harold’s succession lacking legitimacy.158 The Tapestry then follows on with a representation of Halley’s Comet, whose appearance portended that kingdoms would change hands, and of a contorted Harold, the manner of whose depiction foretells his coming doom. But this is history written with the benefit of hindsight. A very respectable case can be made out for Harold as a king who was likely to succeed and who enjoyed extensive support.159 However, not all would have agreed with the opinion that ‘the omens were good’.160 But, very evidently, some would.

For the main events, Barlow, The Godwins, 83–8. Barbara English, ‘The Coronation of Harold in the Bayeux Tapestry’, in Bouet, Levy, and Neveux (eds.), The Bayeux Tapestry, 347–82, at 375–8; Pastan and White, The Bayeux Tapestry and its Contents, 227–8. 159 Thus, e.g., N. J. Higham, ‘Harold Godwinesson: The Construction of Kingship’, in Owen-Crocker (ed.), King Harold II and the Bayeux Tapestry, 19–34; Ian Howard, ‘Harold II, a Throne-Worthy King’, in ibid., 35–52. 160 Thus, Barlow, Edward the Confessor, 255. 157 158

Chapter 6

T H E YE A R O F V I CTO RY

The literary creativity of the years immediately after the Battle of Hastings is a commentary on its multiple consequences and on their turbulence. William of Jumièges started writing again. The Carmen de Hastingae Proelio and the Vita Ædwardi Regis were both produced by 1067–8 at the latest, for individuals or groups with a strong personal interest in putting their version of events on the record. The section of the ‘D’ Chronicle that deals with these years has long been seen as possessing a contemporaneity that makes it likely that it was based on a text written close to events and subsequently revised.1 Poetry was written in recognition of William’s new status. The commissioning of William of Poitiers must be a little later, but the Gesta Guillelmi’s conception, as opposed to its execution, could well belong to this period. Although the subject has been much debated, the same could apply to the Bayeux Tapestry; and, if not, it still belongs within the framework of the unresolved personal, moral, and political uncertainties of the times. This almost frenetic productivity makes the interpretation of the year 1066 and the years that followed different from that of any other period of William’s life. Much that we rely on to provide our evidence was forged in the crucibles of victory and defeat and must be read with this in mind.2 For William, this was the year of the great victory and, in William of Malmesbury’s famous words, ‘the day of destiny for England, a fatal disaster for our dear country’.3 The shaping of the invasion and its consequences for the decisions and events that followed are central to William’s historical reputation.

1 Note Dorothy Whitelock’s comment in The Peterborough Chronicle (The Bodleian Manuscript Laud Misc. 636), 30 (‘an impression of contemporary writing rather than late compilation’). On ‘D’ as an evolving text post-1070, Stafford, ‘Archbishop Ealdred and the D Chronicle’, 145–6, 149–52. See also Dumville, ‘Some Aspects of Annalistic-Writing at Canterbury’, 34–5; Guimon, ‘The Writing of Annals’, 144–5. 2 For a similar approach, Florian Mazel, ‘Qu’est-ce qu’une bataille decisive? Jugement de Dieu et legitimation dans les premiers récits de la bataille d’Hastings (v.1066–1087)’, in Arianne Boltanski, Yann Lagadec, and Franck Mercier (eds.), La Bataille. Du fait d’armes au combat idéologique, XIe–XIXe siècle (Rennes, 2015), 15–30, esp. pp. 17–18. 3 GR, i, 456–7 (Illa fuit dies fatalis Angliae, funestum excidium dulcis patriae).

211

212

william the conqueror

THE RULE OF KING HAROLD For all the variations and ambiguities in the sources’ treatment of Edward the Confessor’s deathbed and of his apparent designation of Harold as his successor, it is as good as certain that such a designation did happen; or, at the very least, that the events which took place made it possible for the vast majority of contemporaries in England to believe that something sufficiently resembling a designation had occurred, so that Harold could be regarded as a legitimate king. As already observed, since all the accounts of the events of the year 1066 were shaped by ethical didacticism, hindsight, and straightforward self-interest, to try to reduce the situation that confronts us to a simple analysis of the relative legal rights and wrongs of Harold and William’s claims is to attempt the impossible. What mattered is that both Harold and William had manufactured a legal, moral, and popular case that their supporters could accept as convincing. In short, and as ever when there were alternative claimants to a kingdom, it was almost inevitable that the issue was going to be resolved by war. From the perspective of events in England, one of the centrepieces of analysis is the passage in the Vita Ædwardi Regis in which the author has the dying Edward say to his brother-in-law Harold: ‘I commend this woman with all the kingdom to your protection’. In writing this, the anonymous author could either have been drawing on the eyewitness memory of his patron, Harold’s sister Queen Edith, or presenting events in the form that he thought she wanted; or, of course, both. Since this instruction followed immediately after the king had recounted a prophetic vision in which he said he had been told that the English kingdom would be devastated by war and that, within a year and a day, God would hand it over to its enemies, the passage might be read as an instruction to Harold to carry out the agreement with William.4 Alternatively – and much more likely – it could be the circumlocution by which the author tried to put distance between the saintly king he was in the process of creating and the bloodbath he knew was going to occur as a result of what he had done. When it comes to the sources with a claim to contemporaneity, the three versions of the Anglo-Saxon Chronicle all effectively say that Edward designated Harold as his successor; ‘E’ is the most explicit, but the ‘C’ and ‘D’ versions, whose texts are at this point identical, say pretty much the same thing.5 Poitiers’s mention of a deathbed designation as one of Harold’s arguments in the exchange of messages that preceded the Battle VEdR, 122–5, with the prophecies at pp. 116–23. ASC, ‘E’, 1066 (‘And Earl Harold succeeded to the realm of England, just as the king had granted it to him’); ibid., ‘C’, ‘D’, 1065 (‘Yet the wise ruler entrusted the realm to a man of high rank, to Harold himself ’). 4 5

t h e y e a r o f v i c t o ry

213

of Hastings, although for him one of the flawed arguments which Harold used, was also surely as good as an acknowledgement that there was a designation.6 The Bayeux Tapestry may well be portraying the scene described in the Vita, but doing so even more ambiguously (Plate 6 (b)).7 Among the later sources, both John of Worcester and Eadmer wrote that Edward established Harold as his heir.8 The only apparently dissenting voices, the early twelfth-century accounts in a history written with the express purpose of praising William, namely the Brevis Relatio, and in the poem of approximately the same date by Baudri of Bourgueil composed for William’s daughter Adela, both say that the dying Edward wanted to stand by his promise to William; the poem even has the king send a mission to William late in his life to reaffirm the earlier promise. The Brevis Relatio does, however, say that there was strong support for Harold. And both texts read as if they are glossing over Harold’s designation rather than denying that it happened.9 Viewed in terms of the simple facts of the situation, it is certain that Edward did designate Harold as his successor. The complexities present in the sources are explicable in ways analogous to the shift present in the Vita Ædwardi’s account and the hindsight it exemplifies. The unavoidable perspective for its author and its patron was that, however Harold had become king, it had all gone terribly wrong. In the same way, when the texts of the ‘C’ and ‘D’ Chronicles bifurcate in the middle of the 1066 annal, ‘D’, whose unique character for this and the following years is more fully discussed in the next chapter, refers to kingship as Edgar the Ætheling’s ‘proper due’.10 In so writing, the author was making a legal and a moral point. But later in the annal, in critisizing the attempt after Hastings to promote Edgar’s claims to kingship by saying that they should have been abandoned sooner, he was conveying the opinion that acting in defiance of God’s will had already brought sufficient punishment and that the only route to atonement was to accept William, whose armies were by this time ravaging south-eastern England. The statement does not GG, 118–19. Brooks and Walker, ‘The Authority and Interpretation of the Bayeux Tapestry’, 11–13 (Gameson [ed.], The Study of the Bayeux Tapestry, 71–4); H. E. J. Cowdrey, ‘Death-Bed Testaments’, in Fälschungen im Mittelalter, iv. Diplomatische Fälschungen II, MGH Schriften, 33 (Hanover, 1988), 703–24, at 716–20 (reprinted in idem, Popes and Church Reform in the 11th Century [Aldershot and Burlington, VT, 2000], chapter 9); Pastan and White, The Bayeux Tapestry and its Contents, 285. 8 JW, ii, 600–1 (quem rex ante suam decessionem regni successorem elegerat); HN, 8 (obit Edwardus, et juxta quod ipse ante mortem statuerat in regnum ei successit Haroldus). Other sources are surveyed in GR, ii, 214–15. 9 Brevis Relatio, 28–9 (note, multis aliis insanie eius fauentibus); Baldricus Burgulensis, Carmina, 156 (lines 271–7) (‘Baudri of Bourgueil, “To Countess Adela” ’, trans. Otter, 72). 10 ASC, ‘D’, 1066. 6 7

214

william the conqueror

therefore contradict the earlier one about Edgar, as at first sight might seem to be the case. It is reflecting the complex emotions and anxieties of the times and the period that followed. Finally, by adding that, at the time of writing, he could see no end to the sufferings of the English, the annalist was saying that, with the benefit of hindsight, he could acknowledge that accepting Harold as king must have been a terrible mistake – but only because of the misery it had brought, and not in terms of the process followed after Edward’s death. If, as is likely, this is history written by someone close to Archbishop Ealdred, it is an account that tries to chronicle what happened and to find a way out of present troubles. In Herman’s Miracles of St. Edmund, which might have been written before 1070, although perhaps composed or amended later, we are told that Harold had succeeded by guile and force and that this was ultimately his undoing, but without there being any direct denial of his legitimacy. Indeed, Herman demonstrably thought him a good man who would have prospered had not luck turned against him.11 Another dimension to this version of the story was that Herman had drawn on the memory of his abbot, Baldwin, the monk of Saint-Denis appointed to Bury St Edmunds by Edward the Confessor in 1065. Since Baldwin was Edward’s and subsequently William’s doctor, and possibly therefore present at Edward’s end, he is someone who might have been able to provide the dispassionate narrative of events that we crave. But, in a different context, he is said to have remarked that ‘the wise man is known for his few words’, a direct quotation from the Rule of St Benedict and advice that Baldwin followed so well that he took any secrets he knew to the grave in 1097.12 Whatever his personal opinion, however, he certainly sent troops to fight for Harold at Hastings.13 His life exemplifies a capacity to adapt to cataclysmic change and indeed to exploit it to his own and his community’s advantage. When it comes to Malmesbury and Orderic, we are dealing with later elaborations of a narrative that introduced the possibility that Edward had designated Harold under duress; or, alternatively, with the other-worldly perspective that God’s support for William inevitably meant Harold had to be a liar. In his earlier additions to the Gesta Normannorum Ducum, Orderic had expressed the opinion that Harold was a brave and honourable man.14 However, in the Historia Ecclesiastica, his line was essentially undiluted 11 Herman and Goscelin, Miracles of St Edmund, 62–3 (Haroldus . . . callida ui ueniens ad regnum). See further Tom Licence, ‘History and Hagiography in the Late Eleventh Century: The Life and Works of Herman the Archdeacon, Monk of Bury St Edmunds’, EHR, cxxiv (2009), 516–44, at 522–3. 12 Herman and Goscelin, Miracles of St Edmund, 70–1. 13 This is the only possible conclusion to draw from Regesta, no. 37. 14 GND, ii, 166–9.

t h e y e a r o f v i c t o ry

215

Poitiers, combined with a profound sympathy for the English people.15 Malmesbury in contrast supplies two slightly different versions of Harold’s conduct. In the Gesta Regum he says that the tale of designation was an invention of Harold’s partisans, adding that he could not believe that Edward would have designated someone he did not trust and whom he believed to be blatantly ambitious. In his Life of the saintly Wulfstan, the one bishop of English birth whom William kept in office, Malmesbury wrote that Harold either won the crown by favour or extorted it by force. This is a difference of emphasis that is likely to be his way of adjusting information derived from his Old English source in order both to mention Wulfstan’s role in major events and to avoid implying too strong a commitment to Harold on the part of his hero.16 Both because of the philosophical complexities present for those who chose to write the histories of these events and because the matter was beyond objective appraisal for contemporaries, it is inevitable that diametrically opposing views about Harold’s actions in 1066 have been held ever since. Those actions may well have been premeditated. The redistribution of his brother Tostig’s estates in northern England among himself, Edwin, and Morcar, and his marriage to their sister Ealdgyth, suggest that Harold was making alliances in advance of the king’s death. But this need not be decisive evidence of preparations to take the crown; it could equally have been a reaction to circumstances. With his visit to Normandy having made it crystal-clear to everyone that William was interested in the succession, minds would have focused on what was likely to happen when Edward died, without perhaps a clear collective view evolving on how to react.17 In this context, the sidelining of Edgar the Ætheling could have reflected the certain knowledge in England that William of Normandy and, in all probability, King Harold Hardraada of Norway, were going to invade. Malmesbury, with his usual facility for telling tales that seem helpful and interesting, included one that Hardraada, while serving in the armies of the Byzantine emperor, had a socially inappropriate sexual relationship with a woman of high social status; he was subsequently thrown to a lion, but managed to kill the animal with his bare hands.18 Although Hardraada was now into his fifties, as everyone would have known, he and OV, ii, 134–9. GR, i, 418–21; Vita Wlstani, in William of Malmesbury: Saints’ Lives, ed. and trans. Michael Winterbottom and Rodney M. Thomson, OMT (Oxford, 2002), 9–155, at 56–7 (uel fauore impetrate uel vi extorta corona). 17 For these points, see, e.g, Barlow, Godwines, 61, 89; Mason, The House of Godwine, 121; Fleming, ‘Harold II [Harold Godwineson] (1022/3–1066)’, 360–1; Baxter, The Earls of Mercia, 51–2. 18 GR, i, 480–1. 15 16

216

william the conqueror

William were formidable people. Given that every royal succession since 975 had presented problems, English leaders would have been aware of how fraught the situation was and that there were difficult times to come. In this situation the proven military competence of an adult might well have been judged a necessity. On the other hand, Edgar, who was approximately thirteen years of age in 1066, was much older than William had been in 1035 when he became duke of Normandy. It would surely have been possible for the English elite to have rallied around Edgar. It is just as easy to create an argument for Harold’s succession in terms of political pragmatism as it is to see it as a ruthless coup d’état. As the months passed, Harold’s position within England certainly became increasingly secure. The general impression from the full range of evidence is that he was shaping up well and, given different circumstances, that he would have proved a good king. If, as has been argued, his coronation was delayed until either the feast day of the Epiphany or Easter, then the case for a consensual succession is strengthened.19 The Bayeux Tapestry’s representation of Archbishop Stigand alongside Harold at his coronation can be read as metaphorically conveying a lack of legitimacy. However, whereas Stigand is named by Poitiers and Orderic as the archbishop who crowned Harold, sources deriving from the memories of those close to Archbishop Ealdred of York, namely John of Worcester and an anonymous early twelfth-century historian of the church of York, say that the ceremony was performed by Ealdred. The Tapestry might therefore be representing Stigand as merely present rather than performing the coronation, and its message may not be as subversive of Harold as has often been thought. It looks as if Harold had the full support of England’s religious elite.20 In addition to his massive wealth and landed power, Harold was also likely to respect existing interests; William and any of the Scandinavian claimants would bring their own supporters with them and redistribute patronage in new and unpredictable directions. Residual tension with Edwin and Morcar was tackled through a conciliatory visit by Harold to northern England; it is reasonable to think that his major concession to them was to renounce support for his brother Tostig, who might have been expected to be brought back from his banishment at Morcar’s expense. Whether Harold was trusted by them is another matter. As the military situation became clearer, Harold made extensive preparations to counter 19 English, ‘The Coronation of Harold’, 347–82, at 374–8; Pastan and White, The Bayeux Tapestry and its Contents, 227–8. 20 GG, 100–1; OV, ii, 136–9; vi, 320–1; JW, ii, 600–1; Chronicon pontificum ecclesiae Eboracensis, in Historians of the Church of York and its Archbishops, ed. James Raine, 3 vols., RS (London, 1879–94), ii, 312–87, at 348; English, ‘The Coronation of Harold’, 377–8.

t h e y e a r o f v i c t o ry

217

an invasion from Normandy. A fleet was assembled at Sandwich and then moved to the Isle of Wight. Troops were stationed along the south coast of England and they remained at the ready throughout the summer.21 Harold’s success in doing all this again indicates that he was as close to being the popular choice as king as was possible. As the aristocracy’s apparent choice, the apparent nominee of the previous king, and with some Danish royal blood on his mother Gytha’s side, Harold’s claims to kingship were as good as those of many who had been, and were to be, successful medieval kings. Poitiers acknowledged this, by writing that in English custom a deathbed bequest by the current ruler took precedence over any previous designation, while of course also arguing that because of his oath to William, Harold could not accept the bequest.22 Poitiers also admitted that the English did not wish to be ruled by a foreigner; misguided though they were, it was for this reason that they fought so determinedly against William.23 Those around Harold are unlikely to have seen the oath to William as much of an obstacle to his legitimacy. The version of events proposed by Eadmer and William of Malmesbury shows how easy it would have been to dismiss the taking of the oath as an inconvenient, but ultimately irrelevant, episode. Malmesbury indeed gives what may have been the arguments used by Harold: the oath to William lacked the necessary sanction of the English aristocracy, and William failed to fulfil his part of the bargain because he had withheld the daughter he had betrothed to Harold.24 PREPARATIONS FOR INVASION According to Poitiers, William gave himself a year to remove Harold. Although this might well be another piece of rhetorical invention, the reasoning underlying it is realistic: the longer the delay, the stronger Harold would become, and the less convincing William’s cause would appear.25 The first half of his year was therefore dominated by extensive military, diplomatic, and intellectual preparations. At the heart of the narrative of this crucial time are three sources: William of Poitiers, the poem known as the Carmen de Hastingae Proelio, and the Bayeux Tapestry. Other sources are also valuable, notably charters. When it comes to the Battle of Hastings itself, the crucial starting points are that none of the authors was in any ASC, ‘C’, 1066; GG, 106–7. GG, 118–19. 23 Ibid., 126–7. See also Carmen, 12–13 (lines 193–4: Bella magis cupimus quam sub iuga colla reponi/Alterius regis, uel magis inde mori). 24 GR, i, 446–7. 25 GG, 102–3. 21 22

218

william the conqueror

way an eyewitness to most of the events and that none tries to provide us with a systematic narrative. Although there is something resembling a broad modern consensus on many aspects of the course of the battle, in terms of detail there are significant differences. In particular, reinterpretation of the relationship between the narratives of the Carmen de Hastingae Proelio and Poitiers’s Gesta Guillelmi and of the purpose of the Bayeux Tapestry has led to some thought-provoking perspectives.26 The subject of a controversy during the 1980s and 1990s, but now restored to its rightful place, the Carmen de Hastingae Proelio is above all a major source for the events of the final three months of the year 1066. As long as account is taken of its poetic genre and of the weaknesses on some matters of its author’s knowledge, it can be used with confidence. Almost certainly written by Guy, bishop of Amiens, quite probably before May 1068 and certainly before mid-1070, it sets out, like Poitiers, to praise William’s exploits, but it does so from a standpoint that is distinctly less pro-Norman and which has different emphases in relation to William’s conduct.27 Because, as was first suggested by L. V. Engels and then reaffirmed by Frank Barlow in his edition of the Carmen, Poitiers’s text at times seems to be in dialogue with the Carmen, we can often believe that we have different versions of William’s conduct set out before our eyes.28 There 26 M. K. Lawson, The Battle of Hastings 1066 (3rd pdf edition, 2016) (https://archive.org/ details/LawsonBattleofHastings3rdedn), 155–90, supplies the fullest and most thoughtful modern account that also draws attention to much other literature. R. Allen Brown, ‘The Battle of Hastings’, ANS, iii (1981), 1–21, at 10–21 (Strickland [ed.], Anglo-Norman Warfare, 161–81, at 169–81), is a classic essay. On the different emphases of the Carmen and Poitiers, Frank Barlow’s Introduction to the OMT edition is crucial, Carmen, lv–lx; and see further John Gillingham, ‘ “Holding to the Rules of War (Bellica Iura Tenentes)”: Right Conduct before, during, and after Battle in North-Western Europe in the Eleventh Century’, ANS, xxix (2007), 1–15. Pastan and White, The Bayeux Tapestry and its Contents, 237–59, supply a radical reappraisal of the Tapestry’s portrayal that attacks the consensus that it was a long and closely fought battle (n.b.: ‘On the contrary, the embroidery depicts a terrible mismatch – a rout, a slaughter – in which the English were doomed to defeat from the very start of the battle’). Brown, ‘The Battle of Hastings’, 10–21 (Strickland [ed.], Anglo-Norman Warfare, 169–81). For other valuable general treatments, Stephen Morillo, ‘Hastings: An Unusual Battle’, HSJ, ii (1990), 95–103; Carmen, lxxvii–lxxxv; Jim Bradbury, The Battle of Hastings (Stroud, 1998), 158–221; Pierre Bouet, Hastings: 14 octobre 1066 (Paris, 2010). Translated sources and the main interpretative essays then published are collected in Stephen Morillo (ed.), The Battle of Hastings: Sources and Interpretations (Woodbridge, 1996). Many other important publications are referred to in the notes to this chapter. 27 Frank Barlow’s Introduction to Carmen sets out all the arguments and also does justice to the contributions of other scholars; see esp. pp. xxiv–xci. Papers relating to the controversy, including those of Ralph Davis, that were passed to Frank Barlow, are now stored in Exeter University Archives. 28 R. H. C. Davis, L. V. Engels, et al., ‘The Carmen de Hastingae Proelio: A Discussion’, ANS, ii (1980), 1–20, at 6 (‘there is a sort of dialogue between the Carmen and William of Poitiers perceptible in what seem to be reactions of William of Poitiers to passages in the Carmen’); Carmen, lv–lx.

t h e y e a r o f v i c t o ry

219

are also significant differences of emphasis and detail between Poitiers and the very brief resumed text of William of Jumièges’s Gesta Normannorum Ducum.29 Interpreting Poitiers’s text becomes a major problem at this point.30 In trying to create an image of a perfect prince within the long-established traditions of the medieval West, and seeking to answer criticisms of William made by both English and Normans, at times its author seems almost to tie himself in knots in ways that can be revealing.31 Moreover, by turning at several points in relation to Harold to the doctrine of tyrannicide, and by characterizing the English who fought at Hastings as impious, Poitiers may either be demonstrating thorough intellectual preparation for the invasion on the part of those advising William or creating a post facto framework for what had turned out to be a very violent campaign. The Bayeux Tapestry supplies all kinds of insights into matters such as armour and gesture, but if, as has recently been proposed, it was intended as primarily a warning for victors and vanquished to avoid the horrors of war and of the terrible price that a people had been made to pay for their sins, then its interpretation can often become a matter of spotting allegories rather than extracting facts.32 In addition to being crucial as the time of preparation that resulted in military victory, these months were important in shaping the way in which it was implemented. Poitiers writes briefly of discussions between William and Normandy’s aristocracy, a passage expanded by Orderic to include debate between optimists and pessimists. William of Malmesbury independently added the information that the assembly took place at Lillebonne, and Wace created a great set-piece debate in which William fitz Osbern took the lead to argue the case for invasion.33 All indicate a process of consultation and persuasion that revealed the confidence of those who chose to contribute and to participate. Orderic and Malmesbury’s accounts also demonstrate how rapidly William took the diplomatic initiative. In particular, both write of a surge of enthusiasm after the receipt of a banner from Pope Alexander II: Orderic names Gilbert, archdeacon of Lisieux (from 1071 bishop of Evreux), as head of the mission, and both he and Malmesbury place the discussions with the aristocracy after the receipt 29

GND, i, liii–liv. For some aspects of these problems, George Garnett, ‘Coronation and Propaganda’, 110; idem, Conquered England, 40. 31 For brief comments on Poitiers’s place within the traditions of writing on kingship, David Bates, ‘William the Conqueror’s Wider Western European World’, HSJ, xv (2006, for 2004), 73–87, at 80–2. 32 For this argument, Pastan and White, The Bayeux Tapestry and its Contents, 237–55. 33 GG, 100–3; OV, ii, 142–3; GR, i, 448–9; Roman de Rou, 230–1, lines 6,095–122. 30

220

william the conqueror

of the banner, saying that William’s plans were received enthusiastically and that he encouraged support with generous promises.34 On this chronology, preparations for the invasion would have moved into top gear in the late spring and summer. The estuary of the river Dives was chosen as the place for the army and the fleet to assemble. Charters show that William visited Fécamp, Rouen, Bonneville-surTouques, Caen, and Bayeux during the year, indicating that he travelled extensively around Normandy, overseeing preparations, and consulting and exhorting widely.35 Since not all the charters contain dating-clauses, it is not possible to construct a precise itinerary; his presence at Caen on 18 June and Bonneville-sur-Touques between 27 May and 16 July does, however, suggest that he started out in Upper Normandy and moved westwards to Lower Normandy over the course of the first six months of the year, and therefore that he spent the summer in close proximity to the forces assembling at Dives. Abbot Bartholomew of Marmoutier’s dispatch to Rouen of one of his monks to seek William’s confirmation of his gifts to the abbey is an obvious indication of a widespread awareness that momentous events were afoot. William’s agreement that his son Robert also confirm the gifts shows him taking appropriate precautions. Although it might be thought to betray more signs of nerves on the duke’s part than Poitiers’s portrayal of a man confident in divine support would suggest, this is fundamentally a commentary on the thoroughness with which preparations were being made.36 Several disputes were settled during the year, again giving an impression that many were setting their houses in order, and that reconciliation and respect for justice were deemed by William and those around him to be the right messages to transmit.37 Routine business was not neglected, with Orderic’s narrative showing that William appointed a new abbot of SaintEvroult after 27 May while at Bonneville-sur-Touques.38 Matilda and their eldest son Robert attest every one of the charters of the year and even the third son, the future King William Rufus, who must have been at this time around six years old, was recorded in his father’s company at Bonnevillesur-Touques, Bayeux, and Caen. Others frequently in William’s company were Roger de Montgommery, William fitz Osbern, and Roger de Beaumont, as well as his two half-brothers Bishop Odo and Robert of Mortain.39 OV, ii, 142–3; GR, i, 448–9. RADN, nos. 227, 228, 230, 231, 232. 36 Ibid., no. 228. 37 Ibid., nos. 229, 230, 232, 233. 38 OV, ii, 144–7. 39 For Roger de Montgommery: RADN, nos. 228, 229, 230, 231, 232; William fitz Osbern, ibid., nos. 228, 229, 230; Roger de Beaumont, ibid., nos. 228, 229, 232; Bishop Odo, ibid., nos. 227, 229, 231; Count Robert of Mortain, ibid., nos. 229, 230. 34 35

t h e y e a r o f v i c t o ry

221

Arrangements must also have been settled for the government of Normandy during William’s absence in England. These involved Matilda, who was to be supported, according to Poitiers, by Roger de Beaumont and others, and, according to Orderic, by Roger de Montgommery.40 Since both writers had special knowledge of the individuals concerned, it is certain that the two of them stayed behind in Normandy.41 The two men’s standing at the very heart of William’s regime and their massive wealth are again indicative of careful preparation, as also is the presence of the two people who were to represent ducal authority during the hazardous enterprise, namely Matilda and Robert in so many charters. While Poitiers’s brief account of the embassy to Alexander II which secured the papal banner suggests that William’s cause carried the day without difficulty, other sources suggest that the business was less straightforward. Orderic’s emphasis, for example, is slightly different in indicating that William sought advice from the pope rather than approval, while Malmesbury says that the pope weighed the arguments before deciding.42 Although this may just reflect a different perception of papal authority, it could suggest that William’s ambassador had needed to work harder than Poitiers wants us to believe to persuade Alexander of the rightness of his cause. A later letter of Pope Gregory VII (1073–85), referring to the support he had given William and the criticism he had received for sanctioning so much killing, while undoubtedly supporting the story of the banner, may indicate that there had been disagreement; it is ultimately impossible to decide whether Gregory was referring to debates about violence at the time or later concern that the Conquest had been so long, drawn out, and brutal.43 This range of references, to which can be added the apparent near-contemporary representation in the so-called Ramsey Benedictional (Plate 7), is sufficient to dismiss any doubts about whether a banner was given.44 William’s and his advisers’ initiative in approaching Alexander tapped into recent developments whereby the papacy gave its approval to ‘just’ wars as a way supposedly to increase its moral and political authority and to support what it regarded as good rule, themes that have been extensively GG, 178–9; OV, ii, 230–1. Attempts to argue for Roger de Montgommery’s presence at Hastings are authoritatively dismissed in OV, ii, 230, note 1. 42 GG, 104–5; OV, ii, 142–3; GR, i, 448–9. 43 Registrum, ii, no. VII. 23 (Register, 353). 44 BNFr, ms. lat. 987, fo. 111r. Cf. Catherine Morton, ‘Pope Alexander II and the Norman Conquest’, Latomus: Revue d’études latines, xxxiv (1975), 362–82; Ian W. Walker, Harold: The Last Anglo-Saxon King (Stroud, 1997), 148–9. Sandy Heslop has pointed out to me that the hand of this section of the Benedictional resembles the one that wrote the text of the decrees of the Winchester Council of 1070 (BL, ms. Cott. Tiberius, C I). 40 41

222

william the conqueror

explored as part of the prehistory of the First Crusade. This espousal of violence, it must be added, was extremely controversial.45 The fact that the great majority of cases of the granting of banners and of overt support for war by the papacy at this time involved warfare against the Moslems makes the award of the banner to William all the more remarkable and, by implication therefore, the severity of the condemnation of Harold’s supposed perjury that must have carried the case in Rome.46 The ideas that are likely to have been involved are those which emerge much more explicitly with Gregory VII to the effect that bad kings and princes were the very body of the Devil.47 The almost contemporary decision by Pope Alexander II that Count Geoffrey the Bearded was unfit to rule and that support should be transferred to Fulk Rechin demonstrates that north-western Europe remained every bit as much on the papal agenda as it had been since the time of the Council of Rheims.48 According to Malmesbury, Harold decided not to put his case to the papacy. This may be indicative of overconfidence; alternatively, as a consecrated king, he may have felt that his position needed no further reinforcement. That he even considered sending a mission is one of many signs that throughout the year the two rivals were well informed of each other’s actions. Harold’s decision not to approach the pope allowed William to pull off a propaganda coup, although not one that persuaded anyone we know of to abandon Harold. The invasion acquired an enhanced moral character, evidenced by the references to it as publicum bellum, that is, a justified war against a tyrant who was subverting the Christian peace. The term’s appearance in the so-called Penitential Ordinance that defined the penances to be performed by members of William’s victorious army, and which was confirmed by the papal legate Bishop Ermenfrid of Sion, in all probability c.1067, shows that the invasion had acquired a status beyond that of a mere war of conquest.49 The approach to Rome fits logically with recent developments in Normandy, such as the extravagant penance involved in the construction of the two Caen abbeys and the propagation of up-to-date papal legislation against married clergy at the Council of 45 The classic study is Carl Erdmann, Die Entstehung des Kreuzzugsgedankens (Stuttgart, 1935) (English translation: The Origin of the Idea of Crusade, trans. Marshall W. Baldwin and Walter Goffart (Princeton, NJ, 1977). For an appreciation, Norman Housley, ‘Crusades against Christians: Their Origins and Early Development, c.1000–1216’, in Thomas F. Madden (ed.), The Crusades: The Essential Readings (Oxford and Malden, MA, 2002), 69–97, at 69–77. 46 For a survey, Ian S. Robinson, The Papacy 1073–1198: Continuity and Innovation (Cambridge, 1990), 312–13, 323–5. 47 Thus, H. E. J. Cowdrey, Pope Gregory VII 1073–1085 (Oxford, 1998), 529–34. 48 For the Anjou case, Guillot, Anjou, i, 108–11. 49 On the invasion as publicum bellum, Garnett, ‘Coronation and Propaganda’, 95–9; idem, Conquered England, 5–10.

t h e y e a r o f v i c t o ry

223

Lisieux. Papal involvement also took the invasion of England into the sort of moral territory that is a controversial ever-present in human history, namely, the direct linking of violence to a religion or a set of ideas.50 Papal support did not, however, come without obligations. It created a relationship that was to have its ups and downs during the decades that followed. In the context of divided opinion among contemporaries about the justification for papal involvement in warfare and about the suitability or otherwise of particular rulers, the events of 1066 may well have forged a reciprocal bond between William and the papacy. As a favoured son, he was even more in a position to turn to the popes for support, since the validity of their acts was henceforth associated with the success and probity of his kingship. His actions were also to become linked to Gregory VII’s increasing and controversial use of war to promote what he regarded as in the interests of Christianity.51 Events were in due course to show that the implications of all this were more complex than was anticipated in 1066. But in the context of that year, like so much else that evolved around William then, his actions were built on carefully cultivated long-term relationships. As also with Poitiers’s espousal of tyrannicide – a topic to which I will turn later – the level of engagement with ideologies that legitimated violence as an unqualified means to an end has to be seen as ruthlessly efficient preparation. It is all a part of the apparatus of trust that persuaded warriors to join the invasion force. Poitiers mentions William’s agreements with the German ruler Henry IV and Svein Estrithsson, King of Denmark.52 Since Henry had only recently assumed authority and Svein was later to invade England and had claims on the kingdom, it is unlikely that either agreement amounted to much; in both cases, nothing more than a mission to declare William’s intention can have been involved. It is indeed possible to suspect Svein of duplicity; unable at this point to attack himself, he may well have calculated that it was in his interests to encourage turmoil to destabilize Harold. If so, he was probably among many to underestimate what William might achieve. As the son of William’s father’s rejected wife, he may well have had a personal grudge against William. A potential ally who may have sought William out was Harold’s banished brother Tostig. A visit by him to Normandy is mentioned only by Orderic in rather garbled terms and is not included by Poitiers. If, as is suggested in the Gesta Normannorum Ducum version of events, William sent Tostig to 50 For a recent essay on this very theme, Philippe Buc, Holy War, Martyrdom, and Terror: Christianity, Violence, and the West (Philadelphia, PA, 2015), with full reference to the eleventh century in this context. 51 For a survey, Cowdrey, Pope Gregory VII, 650–8. 52 GG, 104–7.

224

william the conqueror

England where he was unable to land, it looks as if he was an embarrassment to William.53 William may well have calculated that to ally with a man who had made himself so intensely disliked in northern England was to saddle himself with a liability. Also, whatever Edward’s personal feelings for Tostig might have been, for William to have given credence to someone whom the king had so recently exiled was illogical in the context of William’s claim to be his designated successor. William needed to put on an appearance of detachment from the recent quarrels among the English elite which, he hoped, would react to his advantage by displaying a commitment to justice rather than taking on an ally who would offend those in England whom he hoped might support him. Having been sent on his way, Tostig appeared off the Isle of Wight in May, then ravaged the east coast of England, visited Malcolm III, King of Scots, and finally threw in his lot with Harold Hardraada. It is ironic that Harold’s reasons for not wanting to welcome Tostig back to England might well have been rather similar to William’s. The approach to William by the three non-Norman abbeys of Marmoutier, Coulombs, and Beaumont-lès-Tours for confirmations or restitutions of property in 1066 underlines the extensive non-Norman interest in the preparations for the invasion of England. Among nonNorman individuals who appear in charters during the year are Count Eustace of Boulogne, who was in the duke’s company at Bayeux; Geoffrey de Chaumont-sur-Loire, who was at Caen with William, in all likelihood present as the representative of Count Theobald III of Blois-Chartres; and Aimeri, vicomte of Thouars, from Poitou, for whose presence in Normandy the evidence is persuasive in general terms, if less than clear-cut in detail.54 All three, in different ways, were to play a prominent role in the months and, in two cases, the years ahead. Along with that of William’s Breton cousins Counts Brian, Alan the Red, and Alan the Black, their presence demonstrates the scale of support that the expedition had throughout much of northern and western France at the level of the territorial princes and of the most powerful members of the aristocracy. In Eustace’s case, however, old tensions were not entirely forgotten, since he was obliged to give up a son to William as a hostage.55 The collaboration nonetheless made sense for both of them: for Eustace, it represented a way to consolidate and expand long-standing interests in England, while for William it 53 GND, ii, 162–3; OV, ii, 140–1, 142–5. For a critical assessment of Orderic’s account, OV, ii, 140, note 3; 142, note 3. 54 RADN, nos. 227, 231. On Geoffrey, Dunbabin, ‘Geoffrey of Chaumont’, 111–12. For Aimeri in Normandy, Jane Martindale, ‘Aimeri of Thouars and the Poitevin Connection’, ANS, vii (1985), 224–45, esp. pp. 226–7, 233–4. 55 GG, 182–3.

t h e y e a r o f v i c t o ry

225

made an ally out of a potentially disruptive power on Normandy’s eastern flank that controlled the lands on the French side of the shortest Channel crossing. Much of this response must ultimately be seen as a consequence of the alliance-building of previous years, and, most importantly, of the years since 1060, with the great majority of non-Normans likely to have come from Aquitaine, Brittany, the regions to the east of Normandy, and, to a lesser extent, from Maine.56 Since all of the above were drawn from the aristocratic elites of northern and western France, it also demonstrates that William’s army was far from consisting primarily of adventurers on the make, as is sometimes suggested. The subject of the recruitment of mercenaries is an obscure one; the references to troops being paid off after 1066 must, however, show that it was done, and probably in large numbers.57 In an apparently simple statement, Poitiers remarked that these foreign knights were attracted because all were fully confident of the justice of William’s cause, but that they were also partly attracted by his renowned generosity.58 The complexities inherent in this single sentence are apparent after no more than a moment’s thought. Although we have few details of any of the negotiations that took place or of what was specifically offered, interpreted as an apologia and as an answer to criticisms, they surely lead directly to the disequilibrium apparent later between the expectations of William’s followers and his promise to treat the English justly; for Orderic, as we shall see, this very situation was a central cause of all the problems that followed. Alongside the approach to the pope and the rhetoric of tyrannicide, we are, as set out in this book’s Prologue, in the presence of the exigencies of medieval warrior kingship, with rulers needing to reward to retain loyalty as well as to create an ambience of legitimacy and confidence. When it comes to the negotiations that must have taken place, William of Malmesbury remarked that lavish promises were made.59 Baudri of Bourgueil went as far as to insert the prospect of reward into a speech he 56 GG, 130–1, describes William’s army as consisting of ‘men of Maine, Frenchmen, Bretons, Aquitanians’, as well as Normans. For recent surveys of non-Norman participation, DP, 13–75; Elijas Oksanen, Flanders and the Anglo-Norman World 1066–1216 (Cambridge, 2012), 13–15. 57 Note the rewards given to warriors who returned to France in early 1067, GG, 166–7; OV, ii, 196–7. 58 GG, 102–3 (Conuenit etiam externus miles in auxilium copiosus, quos ex parte notissima ducis liberalitas, uerum omnes iustae causae fiducia contraxit). 59 GR, i, 446–7 (largis sumptibus milites suos continebat, alienos inuitabat), 448–9 (omnes eius uoluntatem plausibus excipientes magnificis promissis animasset). For comment on financial incentives offered by William, J. O. Prestwich, The Place of War in English History, 1066–1214, ed. Michael Prestwich (Woodbridge, 2004), 112–14.

226

william the conqueror

puts into William’s mouth, writing that ‘This is the chance for you to increase your estates. For that land, as you know, abounds in all earthly riches.’60 The text known as the ‘Ship List of William the Conqueror’, while listing ships supplied by individuals, adds that others supplied ships according to their means, again suggesting that participation involved negotiation and the potential for reward. It also contains the remarkable statement that William had granted the earldom of Kent to his wife Matilda in return for her gift of his flagship, the Mora. Since she did not in fact receive the earldom, it also shows that such promises could be reconsidered after the victory.61 Writing a century later, Wace’s debate included a discussion in which many resisted the notion that their military contribution might be regarded as a precedent to determine future service. Once reassured, many agreed to make generous contributions, with William then promising land to them and also recruiting mercenaries whose livelihoods depended on the expectation of reward.62 A record from the abbey of Fécamp, like the account of Abbot John’s visit to England in 1054 (now preserved only in a seventeenth-century printed summary that looks as if it may derive from a pancarte), suggests that William promised to restore Steyning (Sussex) to the abbey should he gain victory, symbolically communicating his guarantee by the gift of a knife.63 Poitiers elsewhere appears to have tried to make a virtue out of the necessity of reward by having William inform Harold that the victor will be the one who can distribute more extensively what he does not own.64 The dedication of Matilda’s foundation of La Trinité of Caen on 18 June was an event of great symbolic importance, an act that a poem written in c.1075 evocatively described as having a parallel with the sacrifice of the biblical king Jepthah. Since Caen is only some twenty-five kilometres from the mouth of the Dives where the fleet and army were assembling, the ceremony must have been particularly dramatic and must have been closely associated in people’s minds with the forthcoming invasion. On the 60 Baldricus Burgulensis, Carmina, 157 (lines 319–21) (Nomen ego regis, uos diuitias habeatis./ Dumque uacat, uestras amplificate domos./Omnibus illa bonis, ut nostis, terra repletur) (‘Baudri of Bourgueil, “To Countess Adela” ’, trans. Otter, 73). 61 Van Houts, ‘The Ship List’, 176 (History and Family Traditions, chapter 6). Note especially habuit dux a quibusdam suis hominibus secundum possibilitatem multas alias naues. 62 Roman de Rou, 230–3, pt. 3, lines 6,033–180. Note especially A cels qui voldrent pramist terre/ se Engleterre poet conquerre (lines 6,175–6). 63 Neustria Pia, ed. du Monstier, 223; Bloche, Le chartrier de l’abbaye de la Trinité de Fécamp, 341, no. 42 (Regesta Regum Anglo-Normannorum, volume I: Regesta Willelmi Conquestoris et Willelmi Rufi 1066–1100, ed. H. W. C. Davis, with the assistance of R. J. Whitwell [Oxford, 1913], no. 1). For a discussion, Martindale, ‘Aimeri of Thouars’, 226, note 11. 64 GG, 108–9 (Hostem haud dubie superabit qui non minus quae hostis possidet quampropria largiri ualet).

t h e y e a r o f v i c t o ry

227

very day of the dedication William and Matilda offered their daughter Cecilia, then approximately seven years of age, to the abbey as a child oblate.65 In doing so they followed a centuries-old practice by which, through an act bearing some of the characteristics of a sacrifice and some of a gift, the parents showed the strength of their support for the abbey and sought the benefits of the nuns’ prayers.66 The uniqueness of the act – this was the only one of the couple’s many children to enter a religious house at so young an age – underlines just how important Cecilia’s claustration was to them at this particular moment. Also, although she was formally subject to the Rule of St Benedict, as we shall see, the evidence shows that she remained a figure in whom contemporaries retained an interest as the daughter of William and Matilda, a constant reminder therefore of the gift to God made at the time of William’s greatest adventure. The charter drawn up at the dedication has a long preamble on the importance of good works and the prayers of religious to individual salvation. It also drew directly on St Augustine’s sermon that set out the idea of making Christ one of the donors’ heirs, a traditional text that appears nowhere else in any surviving Norman charter. Its appearance here surely draws attention to the importance of the occasion within the preparations for the invasion of England.67 It was confirmed by a large number of bishops and abbots. The document also shows that women from several prominent aristocratic families had already become nuns at the abbey, creating a spiritual network binding them the more closely to ducal power. The grant confirmed by the diocesan bishop, William’s brother Odo, whereby any inhabitant of Caen who had chosen the abbey as their burial place during their lifetime would be buried there without paying any dues to their parish priest, consolidated the monastery’s special place within the urban community that was growing up around it. Although we cannot know exactly what state the church and the conventual buildings were in at the time of the dedication, we can deduce something and perhaps thereby have an atmospheric sense of the occasion. The charter’s mention of transactions undertaken by the first abbess, the long-lived Matilda, and the 65 AAC, no. 2 (RADN, no. 231); ‘Fulcoii Belvacensis epistolae’, ed. M. Colker, Traditio, x (1954), 191–273, at 245–6. 66 On this subject, see Janet L. Nelson, ‘Parents, Children and the Church’, in The Church and Childhood, ed. Diana Wood, Studies in Church History, xxxi (1994), 81–114, at 106–11; Mayke De Jong, In Samuel’s Image: Child Oblation in the Early Medieval West (Leiden, New York, and Cologne, 1996), esp. chapter 7. 67 Letouzey-Réty, Ecrit et gestion, 144–5. On the possibilities explicit in the sermon, Susan Wood, The Proprietary Church in the Medieval West (Oxford, 2006), 734–5; Janet L. Nelson, ‘Review-Article: Church Properties and the Propertied Church: Donors, the Clergy and the Church in Medieval Western Europe from the Fourth Century to the Twelfth’, EHR, cxxiv (2009), 355–74, at 357.

228

william the conqueror

clear statement there that a church had been built by 1066, at least indicate that something significant existed; in all probability this would have amounted to a choir (much smaller than the one extended in the later eleventh century that we can now see), of which the main survival consists of the walls around where Matilda’s tomb is located, some of the lower parts of the crossing, and perhaps the walls of the nave aisles (Plates 3 and 8).68 Poitiers makes no comment on the logistics of the process of assembling an army and the invasion fleet at the mouth of the Dives and in neighbouring harbours, beyond praising William for provisioning his forces without resorting to ravaging the surrounding lands. His omission, justified by Poitiers on the grounds that it would take too long to recount, is regrettable, since supplying so large a force and maintaining its morale and discipline must have been a considerable achievement.69 Understanding the scale of it is further hampered by several other factors. One problem is that early estimates of the size of William’s army vary hugely: a Poitevin chronicle indicates 14,000 men, Poitiers 50,000 in one place and 60,000 in another, while the Carmen de Hastingae Proelio mentions 150,000 for the force that eventually left Saint-Valéry to cross the Channel.70 A second problem is that Poitiers’s reference to neighbouring harbours rules out most types of cliometric calculation. While ingenious attempts to estimate the quantities of hay and grain required, as well as the quantities of excreta and urine from the horses that had to be coped with, show ultimately how complicated and uncertain such calculations are, they do, whatever conclusions are adopted, convincingly demonstrate the scale of the enterprise.71 The modern tendency to increase the numerical size of the armies at Hastings based on studies of the battlefield underlines further William’s achievement as well as the enthusiasm that the enterprise engendered. In the context of the renowned ability of medieval writers to exaggerate the size of armies, the figure of 14,000, the only one at first sight to appear 68 Maylis Baylé, La Trinité de Caen: sa place dans l’histoire de l’architecture et du décor romans (Paris, 1979), 13, 54–8. For a short summary of the architecture of La Trinité, Maylis Baylé, L’architecture normande au Moyen Age, 2 vols. (Caen, 1997), ii, 50–5. Note the phrase sanctimonialibus construxit basilicam, AAC, no. 2. 69 GG, 102–5. 70 La chronique de Saint-Maixent, 751–1140, ed. and trans. Jean Verdon (Paris, 1979), 136–7; GG, 102–3, 116–17; Carmen, 8–9. 71 Bernard S. Bachrach, ‘Some Observations on the Military Administration of the Norman Conquest’, ANS, viii (1986), 1–25, at 11–19 (reprinted in Warfare and Military Organization in Pre-Crusade Europe [Aldershot and Burlington, VT, 2002], chapter 14); R. H. C. Davis, ‘The Warhorses of the Normans’, ANS, x (1988), 67–82, at 80 (reprinted in idem, From Alfred the Great to Stephen [London, 1991], 63–78, at 76); Carmen, lxi–lxvii; Lawson, The Battle of Hastings, 144–50; Bernard S. Bachrach, ‘The Norman Conquest, Countess Adela, and Abbot Baudri’, ANS, xxxv (2013), 65–78, at 65–9.

t h e y e a r o f v i c t o ry

229

plausible, and which perhaps has claims to draw on the memory of Aimeri, vicomte of Thouars, who was present, may be acceptable. However, arguably the only true certainty is that William would not have embarked without knowing that he had an army large enough to match the force that it was probably going to have to fight on the other side. The Bayeux Tapestry has some marvellous scenes of ships being built. However, given how easily a substantial fleet could be raised around the North Sea and Channel coasts in the eleventh century, and how long it would have taken to build new ships in any numbers, the construction of ships can only have been a small part of the operation. An indication of what was possible in terms of the rapid recruitment of a flotilla is provided by Tostig’s procurement of sixty ships in Flanders during the winter of 1065–6; and earlier, by his family’s political recovery in 1051–2, which had been dependent on ships and sailors recruited in Flanders and Ireland.72 Attempts at a statistical analysis of the number of ships that transported William’s army across the Channel are subject to exactly the same problems as attempts to estimate the number of effective troops. William of Jumièges’s seemingly remarkable total of 3,000 ships for the size of the fleet that left Saint-Valéry, which Orderic repeated, may not be that far out if the figure of 14,000 for the army is accepted. On the basis of each ship transporting fifty men as crew and soldiers, and taking account of the number of horses that would have been carried, the number may be about right.73 The lower figures that appear in other places all pose problems. The figure of thousand ships recorded as being supplied by individual magnates in the ‘Ship List’, which is probably based on a document in circulation in 1066, is problematic because the list’s apparent total is 776. It is also manifestly not all-inclusive, since it mentions ships that others supplied according to their means.74 And Wace’s very precise figure of 696, based so he said on a story told by his father, also presents difficulties, since Wace himself admitted that his figure was a contested one, adding that he had read elsewhere the figure of 3,000.75 Orderic’s additional comment made in the same passage of the Gesta Normannorum Ducum that provides the 3,000 statistic, that William assembled ‘an immense army’ (ingentem quoque exercitum), supports the idea that William’s forces were exceptionally large.76 There seem to me to be good reasons for suggesting as high a number as available maritime resources would justify. For a survey, Oksanen, Flanders and the Anglo-Norman World, 159–61. GND, ii, 164–5; Lawson, The Battle of Hastings, 147. 74 Van Houts, ‘The Ship List’, 169–70. 75 Roman de Rou, 238–9, pt. 3, lines 6,423–31. 76 GND, ii, 164–5. 72 73

230

william the conqueror

The fleet appears to have been assembled by the effective use of systems operating around the coasts of northern France and Flanders. This applies to the transport of horses as well as troops.77 While the extent to which this process drew on Normandy’s Scandinavian past remains a subject of debate, the sheer scale of maritime activity around the North Sea and Atlantic worlds makes the assembling of the fleet part of a wider cultural world rather than a specifically Norman one, with the Tapestry’s design and manufacture in England also being a factor to take into account. The absence of any significant naval activity by a Norman duke in the eleventh century, with the exception of Duke Robert’s rather feeble attempt to cross to England in support of the claims of the two brothers Edward and Alfred, must surely indicate that seafaring activity in the Scandinavian tradition had died away in Normandy.78 In the end, it is the size of the 1066 operation that is remarkable. For all the massive industry displayed in that year, the invasion first of all underlines yet again the importance of the networks of support that William had long been cultivating. Furthermore – something that may be extremely important – whatever figure is eventually adopted for the number of ships and troops, in the context of the number of ships recorded for Harold Hardraada’s invasion and Danish attacks after 1066, William’s forces were by any count numerically much larger than anyone else could readily muster. It is often, and entirely rightly, pointed out that after 1066 England’s conquerors were a small minority of the kingdom’s population. But the preparations for the invasion do seem to indicate that William was from the start aware of the need to deploy, if not overwhelming military power, then at the least forces that could hope to be significantly superior to those they were going to have to fight. THE CHANNEL CROSSING In July 1066, William decided to move his fleet up the coast to the large natural harbour of Saint-Valéry-sur-Somme in Ponthieu. The dedication of La Trinité of Caen may well be the time when this decision was taken. Both the Carmen and Poitiers say that it represented a change of plan and

77 For general treatments of the subject, C. M. Gillmor, ‘Naval Logistics of the CrossChannel Operation, 1066’, ANS, vii (1985), 105–31; Elisabeth Ridel, Les navires de la Conquête (Cully, 2010), 7–23. 78 GND, ii, 76–9. For current debates, François-Xavier Dillmann, ‘Les figures de proue de la Tapisserie de Bayeux et les traditions nautiques de la Scandinavie ancienne’, Eric Rieth, ‘Construire les navires au temps des Vikings: une relecture archéologique d’une scène de la Tapisserie de Bayeux’, Elisabeth Ridel, ‘Les navires de la Tapisserie de Bayeux à la lumière du vocabulaire nautique normand’, in Lemagnen (ed.), La Tapisserie de Bayeux, 177–93, 195–207, 209–27.

t h e y e a r o f v i c t o ry

231

that there was a long wait for the necessary southerly wind; for Poitiers, William was anxious to cross as quickly as possible. If they are right, the original intention must have been to sail to the Isle of Wight and the Solent, advancing on from there to Winchester.79 Although we have no detailed evidence for the weather that summer, a series of low-pressure systems in the Atlantic could certainly have made a crossing hazardous throughout the whole of July and August, with the loss of ships in storms on the coastal voyage to Saint-Valéry highlighting the difficulties.80 Harold’s positioning of a fleet off the Isle of Wight in May and the billeting of troops along the south coast would also have made William very wary.81 Poitiers mentions the capture of an English spy, showing that both sides were trying to keep abreast of each other’s preparations.82 The ‘E’ Chronicle and a stray reference in Little Domesday Book indicate that there may have been a sea battle, suggesting skirmishing of which we otherwise know nothing.83 The garbled account of a sea battle involving Aquitanians and Anglo-Saxons in a set of Bavarian annals written in 1075 is probably referring to this incident as well, although it may show that troops from Poitou and Aquitaine were travelling to Normandy by sea to reinforce the invasion fleet.84 On the basis of his record to this point, we can conclude that William is unlikely to have taken risks with either unreliable weather or potentially strong coastal defences. It is possible, therefore, that the delay in crossing was to an extent tactical. And, although no source says this, another point is that the move to Saint-Valéry would have been a convenient way to bring all of William’s forces and ships together. For the Bretons and western Normans, the Dives would have been a good place to muster, while Saint-Valéry would have been ideal for Eustace of Boulogne, Flemings, and forces from Upper Normandy. From the time of the arrival at Saint-Valéry, circumstances changed dramatically. Both Poitiers and the Carmen describe William’s devotions at SaintValéry, including his moving in procession behind the relics of St Valéry and making a vow that, if a late thirteenth-century source is to be believed, he fulfilled soon after his victory by granting land in Essex to the church GG, 108–9; Carmen, 4–5. GG, 108–9. For this and other long-delayed Channel crossings, Susan Alice Raich, ‘The Sea in the Anglo-Norman Realm, c.1050 to c.1180’, University of Cambridge PhD thesis (2014), 61–4. 81 For Harold’s use of a fleet and troops, ASC, ‘C’, 1066. 82 GG, 106–7. 83 ASC, ‘E’ 1066; DB, ii, fo. 14v. 84 Annales Altahenses Maiores, ed. Wilhelm de Giesebrecht and Edmund L. B. ab Oefele, MGH Scriptores, xx (Hanover, 1868), 772– 824, at 817–18; van Houts, ‘The Norman Conquest through European Eyes’, 841. The writer’s reference to the victors subsequently subjugating the defeated supports the first interpretation. 79 80

232

william the conqueror

of Saint-Valéry; the grant is certainly mentioned in Domesday Book.85 Viewed in entirely materialistic terms, this is of course another example of promises of landed reward being made on credit. In terms of the medieval culture of warrior leadership, it was intended to maintain morale in a large force that could easily have lost heart during the long delay before embarkation. With the same objective in mind, so Poitiers tells us, the casualties at sea on the voyage from the Dives were as far as possible hushed up and the captured spy sent back to England with a resounding justification of William’s cause ringing in his ears, the audience presumably having been all the possible listeners within earshot. Whether all this is true in detail scarcely matters. The crux – and what Poitiers seeks above all to convey – is the image of a general leading from the front, alternately inspiring and reassuring with displays of valour and piety against a background of demanding conditions. But his mention of many desertions indicates that William’s efforts at maintaining morale were not entirely successful.86 Both Poitiers and the Carmen identify a change in the direction of the wind as the decisive moment. On the evening of either 27 or 28 September – the English and Norman sources are ultimately irreconcilable on this detail – William’s fleet embarked on an overnight crossing.87 By this time, he would have known that Harold had marched to northern England to confront Harold Hardraada, with whom Tostig had joined forces. Caution could now be abandoned. Harold’s victory on 25 September at the Battle of Stamford Bridge, after Edwin and Morcar had been defeated five days before at Fulford on the river Ouse to the south of York, was unquestionably an act of epic heroism; the march north, the tactical surprise that allowed Harold to take his enemies off guard, and the deaths in battle of Harold Hardraada and Tostig are the stuff on which great military reputations are built.88 Yet in terms of the bigger picture, it was William who profited without having had to do anything. Harold’s coastal defences in the Channel disintegrated as soldiers were withdrawn to fight in the north. His fleet’s provisions had run out around 25 August before he left for the north. The ships had been withdrawn in some disarray and, having sustained losses, moved to London.89 The defeat of Edwin and Morcar at Fulford had left them seriously weakened and arguably unprepared to join the march south to 85 Carmen, 6–7 (note Pro uotoque tibi suppeditauit opus, line 71); GG, 110–11; Facsimiles of Early Charters in Oxford Muniment Rooms, ed. H. E. Salter (Oxford, 1929), no. 27; DB, ii, fo. 21r. 86 neque pauida fuga multorum, GG, 108–9. 87 On the conflicting evidence, GND, ii, 166, note 1; Carmen, lxv. 88 These campaigns have recently been comprehensively analyzed in Kelly DeVries, The Norwegian Invasion of England in 1066 (Woodbridge, 1999), 230–91. 89 ASC, ‘C’, 1066.

t h e y e a r o f v i c t o ry

233

confront William. There may also be here an element of the lack of coordination that characterized the English elite’s dealings with William after 1066. Edwin and Morcar’s decision to fight when they must have known that their king was marching north with a substantial army has often seemed perplexing. It is possible that they were over-confident or, alternatively, that the invasion created conditions in which they could see their earldoms crumbling before their eyes. Harold too must have had similar fears. Support in the north for his brother Tostig could have been sufficient to dismember the English kingdom, transferring Northumbria into the power of the king of Norway; York’s submission to them on 24 September was certainly a step in this direction.90 Here, as after 1066, it is important to keep in mind the power of memory. The links between northern England and Norway and Denmark remained strong. It must never be forgotten that less than a quarter of a century had passed since the kingdoms of Denmark and England had been ruled by the same king.91 While the early accounts of these two battles at Fulford and Stamford Bridge are very sparse, the statistic in the ‘D’ Chronicle recording that the survivors of an army which had arrived in 300 ships left in twenty-four adds a significant dimension to the Battle of Stamford Bridge. John of Worcester’s figure of 520 ships arriving is even more chilling.92 Another aspect of the battle, if Orderic is to be believed, is that the dead were left unburied, at least presumably the defeated army’s dead; he recorded that in his day the site of the battle was still recognizable because of the mountain of bones still to be seen there, testimony to a great slaughter.93 If William did not already know the quality of the opponent he was up against, after Stamford Bridge he certainly did. Additionally, cultural values may also have been involved that worked to William’s benefit. William of Malmesbury mentions that Harold did not distribute booty after his victory and, as a result, he had to travel south with a depleted army; in other words, looting was something that sustained morale, a device that, as we have seen, William employed when he allowed his troops to sack Mayenne.94 William and his fleet crossed the sea unobstructed on a calm night. Poitiers tells us that the duke’s ship outstripped the others and lost touch with them. Unperturbed, he ate a hearty breakfast and waited until they caught up.95 Although Poitiers is categorical that William landed at For arguments along these lines, DeVries, The Norwegian Invasion, 256–61. Timothy Bolton, The Empire of Cnut the Great: Conquest and the Consolidation of Power in Northern Europe in the Early Eleventh Century (Leiden, 2009), 312–16. 92 ASC, ‘D’, 1066; JW, ii, 602–5. 93 OV, ii, 168–9. 94 GR, i, 422–3. 95 GG, 112–13. 90 91

234

william the conqueror

Pevensey in Sussex, there must in practice be some doubt whether so large a fleet could all have disembarked there.96 Some ships are known to have strayed off course and landed instead some thirty miles away at Romney in Kent, where the soldiers they had carried were attacked and killed.97 THE BATTLE OF HASTINGS While the rhetorical purpose of the main surviving sources puts obstacles in the way of writing a factual narrative of what was arguably the most dramatic day of William’s life, certain themes do emerge in relation to his conduct as a general and a soldier. His victory was arguably the result of superior generalship, both at a tactical and strategic level. Even allowing for the exaggerations of both the Carmen and Poitiers about his fortitude and bravery, there can be no doubt that both were crucial factors in inspiring his troops; the effect of the false rumour of his death, the fact that three horses (or possibly only two) were killed under him, and his role at the ‘Malfosse’ are but a selection from the examples underlining the point. All show that he was ready to take personal risks at the times when it was necessary to inspire his army. On the other hand, the incident in which he rallies retreating troops shows that he was at other times in the rear of his army giving direction and assessing strategy. Likewise, while medieval armies engaged in battle must have been very difficult to control, the persistent determination to take the attack to the enemy indicates a drive that must ultimately have been unnerving to his foes. As we shall see, William’s capacity to control his army is a point made by both Poitiers and Bishop Guy. His ruthlessness and single-mindedness are illustrated by the savagery of the pre-battle ravaging, the decisiveness of the advance from Hastings to the battlefield, and the way in which Harold was killed. After a landing that seems to have been unopposed, William concentrated his troops at Hastings, where an Iron Age fort provided some defence and, most probably, a temporary command headquarters. The Bayeux Tapestry shows the construction of a motte, presumably a symbolic representation of military activity intended to enhance security within the immediate locality of the landing.98 Given that the army would have crossed the Channel without sufficient supplies for a campaign, the ravaging of the 96 Ibid., 112–15. For recent discussions of William’s landing, ibid., 112, note 3; Lawson, The Battle of Hastings, 147–8. 97 GG, 142–3. 98 Most recently on the representation of the motte at Hastings, Pierre Bouet, ‘Châteaux et residences princières dans la Tapisserie de Bayeux’, in John A. Davies, Angela Riley, JeanMarie Levesque, and Charlotte Lapiche (eds.), Castles and the Anglo-Norman World: Proceedings

t h e y e a r o f v i c t o ry

235

countryside that followed was arguably a necessity in order to feed the troops. It was, however, also a military and political tactic, its scale another commentary on William’s conduct in relation to contemporary norms. The Tapestry apparently shows a woman and a child either meeting their deaths in a burning house or escaping from it.99 The Carmen, describing a report sent to Harold, says that William’s forces took captive boys and girls, and even widows, and also took all the cattle.100 Poitiers, who mentions none of this detail, despite writing regularly elsewhere of William’s protection of the weak, does say it was William’s ravaging that convinced Harold he must come south very quickly. This suggests that it was especially destructive.101 In terms of norms, while the Carmen may have been exaggerating, or indeed inventing, what happened, it is very likely that Bishop Guy was describing what he regarded as standard practice. Harrying of this kind had been the norm for centuries in the medieval West, the purpose being to show that a region’s protector was failing in his responsibilities and to draw him into battle.102 The Annales Regni Francorum, which have something of the status of an ‘official history’ of the Carolingian kings and can therefore be treated as an expression of an official ideology, show how such deliberate destruction was frequent practice with the ruler whom many sought to emulate, namely Charlemagne. The annalist’s choice of language also draws attention to apparent differentials by sometimes referring to wasting and sometimes to depopulation.103 The systematic nature and the scale of the destruction that was possible are shown by the emperor Louis the Pious devastating Brittany for forty days on account of a breach of faith.104 Similar state-sponsored violence was practised in England before 1066. In 1041, for example, after the murder of two tax collectors at Worcester, Harthacnut had sent a large force to burn the city and ravage the region.105 of a Conference Held at Norwich Castle in 2012 (Oxford and Philadelphia, PA, 2016), 135–46, at 45–8. See also Arnold J. Taylor, ‘ “Belrem” ’, ANS, xiv (1992), 1–23, at 19. 99 Bayeux Tapestry, plate 47. 100 Carmen, 12–13 (Captiuos ducit pueros, captasque puellas/Insuper et uiduas, et simul omne peccus, lines 165–6). 101 GG, 124–5. On William as a protector of women and children, see, e.g., pp. 162–3, 180–1. 102 For a survey, Halsall, Warfare and Society, 137–40. 103 For wasting with the connotation of political punishment, Annales Regni Francorum unde ab. a. 741 usque ad a. 829, qui dicuntur Annales laurissenses maiores et Einhardi, ed. Friedrich Kurze, MGH Scriptores (Hanover, 1895), 4–5, 21, 85, 96, 97, 103, 117, 142, 158, 156, 165, with references to wasting and depopulation at 22, 102, 120, 122, 125, 128–9, and for a decision not to ravage, 74. On the Annales, Rosamond McKitterick, Charlemagne: The Formation of a European Identity (Cambridge, 2008), 31–49. 104 Annales Regni Francorum, 165. 105 JW, ii, 532–3.

236

william the conqueror

Given that Sussex was one of the heartlands of Harold’s family’s power, a perhaps ungenerous view of William’s conduct is that non-combatants may have been used as emotional blackmail against fathers who had marched with Harold’s army, or that women were being forced into nonconsensual sexual relations or taken as domestic slaves to help supply the army. Permissive violence against women is a tactic known from elsewhere; Roger of Howden, for example, records that the emperor Henry VI put wives and children up for sale to his soldiers at Salerno in 1191 and enslaved others because they had betrayed his wife to his rival Tancred of Lecce.106 There are many cases of enslavement of women and children from across medieval Europe.107 Perceived disloyalty was also the motivating factor behind the emperor Frederick Barbarossa’s destruction of the city of Milan in 1162.108 The taking of cattle was to an extent necessary to feed William’s army. But the removal of livestock was also one of the most effective ways to achieve the long-term impoverishment of agricultural land, since replacements take several years to install and breed. This is a point that is also very important in relation to the ‘Harrying of the North’ and which I will discuss further in that part of this book. Because of the long-term destruction of livelihoods, it was another form of direct assault on the wealth and honour of William’s enemy. The issue of how far Domesday Book evidence can be used to estimate deliberate destruction is one that will have to be considered several times in this and the following chapters. The case of the area around Hastings is, however, relatively straightforward; Domesday shows that in 1086 the area known as the Rape of Hastings had lost approximately 75 per cent of its value since 1066 and that the three coastal hundreds where the army was mostly based had lost 80 per cent, a level of loss much more severe than anywhere else in the county.109 The message being transmitted to Harold was loud and clear and would have been understood. The intended audience would also have been the people of England and, above all, their 106 Chronica Rogeri de Houedene, ed. William Stubbs, 4 vols., RS (London, 1868–71), iii, 268–9; cited in Thomas Foerster, ‘Imperial Tradition and Norman Heritage: Cultures of Violence and Cruelty’, in Stefan Burkhardt and Thomas Foerster (eds.), Norman Tradition and Transcultural Heritage: Exchanges of Cultures in the ‘Norman’ Peripheries of Medieval Europe (Farnham and Burlington, VT, 2013), 161–88, at 167–8. 107 For a consideration of some of the issues, John Gillingham, ‘Women, Children and the Profits of War’, in Nelson, Reynolds, and Johns (eds.), Gender and Historiography, 61–74. For this specific case interpreted as the taking of slaves, idem, ‘ “Holding to the Rules of War” ’, 7–8. 108 Foerster, ‘Imperial Tradition and Norman Heritage’, 168. 109 For the general issues, J. J. N. Palmer, ‘The Conqueror’s Footprints in Domesday Book’, in Andrew Ayton and J. L. Price (eds.), The Medieval Military Revolution (London, 1995), 23–44, with this specific case at p. 26.

t h e y e a r o f v i c t o ry

237

other political leaders. While one aim of this type of intimidation was to draw an opponent into negotiation, if that negotiation did not happen it became a means of demonstrating disapproval of the people who had agreed to, and supported, Harold’s kingship. William’s strategy would also have been to maintain communications with Normandy and to enable fresh troops to land in England. Since it would have been some days before he had reliable information about the outcome of the Battle of Stamford Bridge, he would also have been waiting for news. According to Poitiers, William returned from a demanding reconnaissance expedition carrying William fitz Osbern’s hauberk on his shoulders. Poitiers treats the story as a joke, and it was probably intended as an illustration of camaraderie. Since it is easier to wear than to carry a hauberk, it is inconceivable that William fitz Osbern was suffering from exhaustion.110 The serious purpose of this excursion was to learn more about the lie of the land in anticipation of the arrival of Harold’s army. In terms of William’s qualities as a leader, Poitiers added the observation that generals did not normally undertake reconnaissance missions of this kind because of the risks involved. The message coming across, which is verifiable from the evidence of different sources for other events of his career, is that William was directly involved in operational matters with his troops. Harold would have reached London in early October and, after a short break, he set off again to confront William. His journey from Yorkshire had been so remarkably fast that it could have been accomplished only by elite troops with horses. Many good foot soldiers must therefore have been left behind. All three of the early French written sources say that Harold tried to take William by surprise, a tactic that had worked for him at Stamford Bridge.111 However, later writers in England offer the opinion that, with hindsight, Harold’s generalship was impulsive, with both John of Worcester and William of Malmesbury suggesting that he fought before much of his army had arrived, the former attributing it to bravery and the latter to rashness and folly.112 Malmesbury and Orderic even say that Harold rejected his brother Gyrth’s suggestion that he, rather than Harold, should lead the army against William, since Gyrth had not taken an oath and was therefore not obligated to him. Orderic adds his mother Gytha to those who pleaded with Harold to allow this.113 The situation was in fact very different from the one he had encountered at Stamford Bridge, since William’s army was larger and not exhausted by a recent battle. If William’s GG, 114–17, with the suggested interpretation at p. 116, note 1. GND, ii, 166–7; GG, 124–5; Carmen, 18–19. 112 GR, i, 422–3, 452–3; JW, ii, 604–5. 113 GR, i, 450–1; GND, ii, 164–7. 110 111

238

william the conqueror

devastation of the countryside and the taking of women and children in the region where Harold’s ancestral lands lay were meant to provoke Harold prematurely into battle by showing that he was not able to protect his own people, they succeeded. Although we do not have such specific evidence of the nature of the ravaging undertaken by William’s enemies during the campaigns of 1054 and 1057, it is arguable that William would not have risen to this particular bait in the way that Harold evidently did. In retrospect, it would probably have been in Harold’s interests to have waited, perhaps to have secured overwhelming numerical superiority, or at least to have assembled more effective troops. Although the view is not universally accepted that the shrunken pathetic figure of a single English archer portrayed on the Bayeux Tapestry should be taken literally, it may be indicative of an army outweighed in a crucial area.114 What would have happened if Harold had held back is of course yet another of the great ‘what ifs?’ of the subject. Since William required a decisive result, it was imperative that he force an engagement.115 However, that both he and Harold faced the task of holding together very large armies with late autumn and winter rapidly approaching certainly was a strong influence pushing both of them to fight the battle as soon as possible. Harold must have been concentrating his forces close to modern-day Battle by 12 October at the latest, approximately seven miles from where the invading army was based at Hastings, thereby providing William with his opportunity.116 The ‘D’ Chronicle’s statement that William was able to attack before Harold’s troops were fully organized indicates that it was ultimately William who achieved surprise.117 Once more, therefore, we encounter his readiness to take the initiative, albeit on the basis of thorough preparation through reconnaissance. As the two armies moved within range of one another, emissaries were exchanged through whom each leader invited the other to withdraw and set out why his own cause was the just one. Both William and Harold sent a monk as their envoy. Harold’s embassy was sent first, and William’s involved an unnamed monk of Fécamp. William kept Harold’s ambassador overnight, so the main exchanges took place on Friday, 13 October. Poitiers and the Carmen tell approximately the same story about these preliminaries, although Poitiers, as so often, is more elaborate. In particular, he mentions the preliminary embassy to William of Robert fitz Cf. Lawson, The Battle of Hastings, 69–70, 135, note 4. A point well made by Gillingham, ‘William the Bastard at War’, 157–8 (Strickland [ed.], Anglo-Norman Warfare, 159–60). 116 For the site of the battle, Roy Porter, ‘ “On the very spot”: In Defence of Battle’, English Heritage Historical Review, vii (2012), 5–17. 117 ASC, ‘D’, 1066. 114 115

t h e y e a r o f v i c t o ry

239

Wymarc, a man probably of Norman origin established in England before 1066, presumably chosen to communicate in French, and his insolent speech advising William to pull out of a battle he would inevitably lose; William’s deception of Harold’s monastic messenger by a pretence that he was actually the ducal steward; and William’s challenge to Harold to settle the matter by single combat.118 Such diplomatic exchanges were central to the procedures of formal defiance in a world where battle was equated with a judicial duel in which God would decide. Religious preparations were similarly directed both towards securing divine favour and lifting morale. William heard mass before going into battle and in the conflict wore around his neck the relics on which Harold had sworn his oath; these were subsequently bequeathed to Battle Abbey, which would be founded on the site of the great victory.119 Poitiers also makes much of the presence of the two bishops, William’s half-brother Odo of Bayeux and Geoffrey of Coutances, and other clergy at the battle. There was a similar religious presence on Harold’s side.120 The speeches and the prayers were also intended as a final assertion of right. The challenge to single combat, something frequently done in such circumstances and always rejected by the party challenged, was intended superficially to show an intention to settle the issue by law and spare lives. In practice, it was a way of seizing the moral high ground and, in this case, very probably a device used to force the pace; Harold’s reply that he was preparing for a battle in which God would decide whose cause was just is that of someone drawn on psychologically into a battle when it might have been in his interests to delay.121 William’s treatment of Harold’s messenger may well have been another example of Poitiers telling a story of the duke’s light-hearted humour at a time of high tension; an anecdote to show William at ease with his cause. Poitiers seeks to show similar inward calm with the story of William putting on his hauberk back to front; the superstitious would have regarded this as a bad omen, but the duke, with GG, 116–23. On Robert fitz Wymarc’s origins, DP, 19. GG, 124–5. For the relics’ complicated later history, Elisabeth van Houts, ‘The Memory of 1066 in Written and Oral Traditions’, ANS, xix (1997), 167–79, at 167–8. 120 GG, 124–5. 121 Denique iudicium, quod iura gentium definirent, accipere praesto fuit, GG, 122–3. On the context for this episode, Stephen D. White, ‘Proposing the Ordeal and Avoiding It: Strategy and Power in Western Frankish Litigation, 1050–1110’, in Thomas N. Bisson (ed.), Cultures of Power: Lordship, Status, and Process in Twelfth-Century Europe (Philadelphia, PA, 1995), 89–123 (reprinted in White, Feuding and Peace-Making in Eleventh-Century France, chapter 7). On the specific issue of offers of single combat before battle, Matthew Strickland, ‘Provoking or Avoiding Battle? Challenge, Duel and Single Combat in the Warfare of the High Middle Ages’, in idem (ed.), Armies, Chivalry and Warfare in Medieval Britain and France: Proceedings of the 1995 Harlaxton Symposium (Stamford, 1998), 317–43, at 323–7. 118 119

240

william the conqueror

his mind on higher things, laughed it off.122 While Poitiers’s account of the speeches is more legalistically precise than the Carmen’s, the clarity with which they are set out in the Carmen shows how well William’s arguments had been disseminated beyond Normandy; as with so much else that happened during 1066 and the immediately preceding years, the preparations had been very thorough.123 The passage in which William is made to offer Harold peace and his father’s lands if he abandons the kingship is not repeated by Poitiers. Thoroughly logical in the light of the 1064 agreement – if utterly impractical in the circumstances of 1066 – it is an offer that William might well have made. If he did, it would not have been the first time he had made a gesture that an enemy could not honourably accept.124 The accounts of the battle in Poitiers and the Carmen, which are by far the most detailed, indicate that William drew up his forces in three lines: archers, including crossbowmen, in front, heavier infantry behind, and cavalry to the rear. The Carmen conveyed the impression that in the haste the cavalry may not have been fully organized when battle was first joined. William was himself in the centre of his forces. The left wing consisted of Bretons and the right wing of French. Harold had taken up a position on high ground, forcing William’s troop to advance up a steep slope, thereby presumably largely negating the mobility of his cavalry. Well placed as Harold’s army seemed to be, there were two significant weaknesses to its seemingly formidable location: the troops were so cramped as to deny room for manoeuvre to individual soldiers, and access to the site was along a narrow isthmus, thereby making orderly retreat difficult. Even if we accept arguments for the English line being deployed over approximately a little more than half a mile, that is, significantly longer than usually thought, these same points still apply.125 In placing cavalry across the whole of his army rather than only in the wings, as was frequent practice, William was presumably accepting that the English army was not going to be outflanked. The strategy was therefore to prepare for a direct confrontation on a relatively narrow front. On the basis of the main narratives, it looks as if fighting began at around nine o’clock in the morning and continued until dusk; by medieval standards, an exceptionally long battle. We know more about the early stages of the battle than about the later ones, perhaps because conditions became very confused as the day wore on, but more probably because the action became for a long while close to a stalemate as the two sides strove GG, 124–5. See also Breuis Relatio, 31. Carmen, 14–19. 124 Ibid., 16–17 (lines 243–5). There are obvious parallels with his treatment of Guy of Brionne and William of Arques. 125 Lawson, Battle of Hastings, 163–8. 122 123

t h e y e a r o f v i c t o ry

241

for tactical advantage, and thus was difficult to describe. Poitiers regarded the battle’s later period as extraordinary, with one side mobile and on the attack and the other static in a defensive position.126 This characterization should not, however, be applied to the whole battle since it belies the English capacity to launch an attack, decisive at Stamford Bridge and evident in the early part of this battle when a general advance seems to have failed to materialize. Over a whole day fighting must have been intermittent and concentrated into bursts of intense activity. It must also have been exceptionally brutal. There are no mentions of the prisoners and ransoms that were a normal feature of almost all eleventh-century warfare; this was presumably because it was in the interests of both sides to kill as many of their opponents as possible.127 The skill of William and his commanders was to adopt tactics that gradually wore their opponents down and denied Harold’s troops the opportunity to launch the mass attack which their position and methods of warfare logically required. Hence, when it was deemed advisable, William’s army advanced up the slope. The potentially devastating impact of the archers on a compressed infantry force must have been limited. The possible lack of arrows fired by the opposition and the fact that a good supply of them was kept in reserve for later would both have reduced their effect. The curtailment of the archers’ assault and their later crucial role does, however, suggest that William or someone else took what proved to be a good tactical decision. When the infantry failed to break through, the cavalry had to join in handto-hand fighting. In time this initial assault faltered and the Bretons on the left fell back in disarray, with a demoralizing effect on the whole army. Poitiers’s account of these setbacks may be overstated in order to highlight William’s achievement in reversing them. The Carmen presents the affair rather differently, placing this chaotic phase after the failure of a feigned retreat and blaming it on the Normans. What both agree on, however, is that there was a moment in the battle when William’s army came near to defeat. This reversal may have been associated with a rumour that William had been killed. Poitiers, the Carmen, and the Tapestry all identify this as a decisive moment. It was here that the two literary sources see William as truly coming into his own, with both attributing to him a furious harangue and describing him as rushing back into the fray at the head of his men, laying about him. As a result, the French rallied, surrounded the English and killed large numbers. For this to have happened as described, the English advance must have been somewhat uncoordinated; it is possible that they believed themselves to be so close to victory 126 127

GG, 132–3. See Strickland, War and Chivalry, 4.

242

william the conqueror

that discipline was abandoned. But William is portrayed both as being located behind his troops to stop the rot and then leading them from the front, after insulting them for being ineffective; Poitiers has him rush ahead of his men brandishing his sword, mowing down his enemies.128 It must have been conduct of this kind that led to three (only two in the Carmen) horses being killed under him. The way in which the Carmen says that Count Eustace of Boulogne and a large troop of warriors rushed to William’s aid after he had lost his second horse, while intended to praise Eustace, also shows that, for all the bravado, William would never have been truly isolated. He certainly put his life at risk, but in a calculated way.129 Other nobles would have played similar roles, and Poitiers named some of them. It was in the role of preventing indiscipline and sustaining morale that the Tapestry portrays Bishop Odo, protected by chain mail, rallying the younger warriors. In Poitiers’s account the episode of the real retreat is followed by two feigned retreats that disrupted the English lines and led to the slaughter of those who had been tricked into an unwise pursuit. This stratagem, whose credibility has occasionally been doubted, was quite feasible for welltrained horsemen; it was indeed part of the common stock of western cavalry warfare. It cannot, however, have involved more than a section of the army and has to be seen as part of a plan gradually to wear down the English. In spite of the relative weakness of the accounts of later stages of the battle, it is obvious that the English force was faltering, but still capable of determined defensive resistance. The place of certain key events in the battle is very unclear. The Carmen and the Tapestry are the only sources to make much of the deaths of Harold’s brothers Gyrth and Leofwine, yet neither version of events is really credible. The Carmen’s statement that William himself killed Gyrth was manifestly designed to flatter, and the Tapestry’s location of their deaths before even the episode when rumours spread of William’s death is very unlikely. The removal of the two brothers must nonetheless have significantly undermined the command structure of the English army. The manner of Harold’s death is a perpetual mystery for historians. Jumièges’s statement that it occurred during ‘the first assault’ is usually dismissed, although it does depend on what he meant by ‘first’; his clear association of the king’s death with the final English flight might just indicate that he was referring to the first decisive attack.130 Harold’s time of death in both the

GG, 130–1. Ibid., 134–5; Carmen, 30–1. 130 GND, ii, 168–9 (in primo militum congressu). 128 129

t h e y e a r o f v i c t o ry

243

Carmen and the Tapestry as late in the day makes much more sense, but in detail neither version is persuasive. Interpretation of the Tapestry’s portrayal of Harold’s death is bedevilled by uncertainty about the consequences of imperfect modern restoration of the figure who is either holding a spear or clutching an arrow that has inflicted a head wound.131 The Carmen’s tale of Harold’s death at the hands of a group consisting of William, Count Eustace of Boulogne, Hugh, heir to the county of Ponthieu, and (Robert) Gilfard has a strong feeling of poetic licence in praise of the great.132 Even if Bishop Guy did personally know the last three (and probably all four), it is hard to think of them taking such personal risks; in practice all four, as the poet admits, would certainly have been to the fore with others of the charge, which effectively settled the battle, and would very probably have reached the scene very rapidly.133 Poitiers’s studied reticence on the manner of Harold’s death has finally muddied the waters beyond recovery.134 Either he did not believe the other early accounts, which is likely, or he did not know what had happened, which is beyond belief. It was of course central to the image that Poitiers was seeking to create that William was the merciful and forgiving king; in pursuit of this theme he goes as far as to say that William did not want Harold killed, a statement that is a logical continuation of the theme established by his account of Harold’s oath, namely that if only Harold had accepted the offer, then all the bloodshed would have been avoided.135 Poitiers’s silence is, however, also likely to be explicable by something discreditable to William having occurred that did not fit with his rhetorical purposes. Whereas Guy of Amiens might have seen Harold’s killing as something to be praised, especially with reference to ‘the noble heir of Ponthieu’, for Poitiers it did not fit with the merciful image he was portraying. In all probability, the ultimate strategic objective must surely always have been to kill Harold. This accords with William’s long-term record as a soldier and a general. Moreover, Poitiers does devote considerable space not just to justifying tyrannicide, but to praising it.136 Even if the Tapestry is discounted as unreliable, the story starts to appear in sources from c.1080 that Harold had been killed by an arrow which penetrated his brain. It first appears in Amatus of Montecassino’s History of the Normans, written in southern Italy at the abbey of Montecassino On this, see Lawson, The Battle of Hastings, 181–6, 204–13. Carmen, 32, lines 531–50. 133 Note Ast alii plures; aliis sunt hi meliores, ibid., line 541. 134 regem ipsum et fratres eius, regnique primates nonullos occubuisse, GG, 136–7. 135 Ibid., 156–7 (Hic ne Heraldum uellet occubuisse). 136 Ibid., 26–7, 138–9, 156–7; Garnett, Conquered England, 36–40. The first of the three passages is directed against Geoffrey Martel. 131 132

244

william the conqueror

in c.1080, and then in the late eleventh- or early twelfth-century poem that Baudri of Bourgueil addressed to the Conqueror’s daughter Adela, countess of Blois-Chartres.137 It was later accepted by William of Malmesbury, Henry of Huntingdon, and Wace.138 Such a wound would have been instantly fatal. Malmesbury adds a story of a soldier who hacked at the king’s fallen body and was punished by William for unchivalrous conduct.139 Harold could therefore have been killed by a stray arrow, which in a few seconds decisively and irrevocably turned the battle in William’s favour. Tactically, however, there is a lot to be said for interpreting the events in terms of Henry of Huntingdon’s belief that Harold’s death occurred as a result of William ordering his archers to fire arrows into the air; such an assault would not have involved indiscriminate firing, but would have targeted the English command centre. As the English forces were now hemmed in, such a tactic accords well with the ruthlessness with which William had conducted the entire campaign. Harold’s death turned the battle into a rout. From this point on the sources tell of a relentless pursuit, interrupted only by isolated moments of determined, but futile, resistance. William is portrayed by Poitiers as unremitting, sparing no one who stood in his way.140 A group of English who did make a stand at a strong defensive site, the skirmish that Orderic developed into the ‘Malfosse’ incident, were overwhelmed despite the difficulties of the terrain and the significant casualties suffered by William’s army. Count Eustace of Boulogne’s advice to retreat was brushed aside.141 The pursuit undoubtedly went on into the night as William’s army hunted down the demoralized English in the difficult countryside in order to make their victory as decisive as possible. These last hours were in all probability a massacre. Poitiers’s account does acknowledge the scale of the slaughter and suggests that William felt pity. However, the way in which Poitiers branded all who were killed as ‘impious’ indicates that he sees them as guilty by association in having chosen to fight for the oath-breaking tyrant Harold.142 On the following day, William gave orders that his army’s dead be buried. As far as his treatment of the English is concerned, the main 137 Aimé du Mont-Cassin, Ystoire de li Normant, ed. Michèle Guéret-Laferté (Paris, 2011), 244 (The History of the Normans by Amatus of Montecassino, trans. Prescott N. Dunbar, revised with an Introduction and Notes by Graham A. Loud [Woodbridge, 2004], 46); Baldricus Burgulensis, Carmina, 161, line 463 (‘Baudri of Bourgueil, “To Countess Adela” ’, 77; van Houts, Normans in Europe, 127). 138 GR, i, 456–7; HH, 394–5; Roman de Rou, 272–3, pt. 3, lines 8,139–74. 139 GR, i, 456–7. 140 GG, 136–7 (ducis eam saeuitiam quae nulli contra stanto parceret). 141 Ibid., 136–7; OV, ii, 176–9. 142 GG, 138–9 (tametsi factam in impios).

t h e y e a r o f v i c t o ry

245

sources tell slightly different stories. According to the Carmen, the English corpses were left where they had fallen to be eaten by animals and birds; Poitiers, however, says that William allowed the English to bury their dead. Since Poitiers remarked elsewhere that many English were indeed left unburied, it is the Carmen that was probably closer to the mark, especially as the intimidatory tactics pursued by William in the weeks after the battle suggest his general instinct was to overawe with a display of power, rather than to conciliate. For William to have allowed the enemy dead proper burial would have been a notably generous act by contemporary standards.143 His distribution of alms to the poor recorded in the Carmen was most likely a gesture of thanksgiving for victory.144 As far as Harold’s body was concerned, William is said in the early sources to have rejected the entreaties of the dead king’s mother and entrusted the burial to the half-Norman, half-English William Malet. Both the Carmen and Poitiers say that Harold was buried on the seashore, and Poitiers adds that this was considered an ironic reflection on his defence of England’s shores.145 William of Malmesbury, however, says that William acceded to the request of Harold’s mother and allowed him to be buried at Holy Cross, Waltham, the church which the dead king had founded and endowed. This story was elaborated in the late twelfth-century chronicle of Waltham Abbey in which two canons successfully plead with William for the corpse.146 Harold’s final resting place is likely to have been a difficult problem for William and his companions to resolve. Whatever was done might well make Harold a figure of veneration among an English people who were largely hostile to William, and a standard Christian burial for a man whom William and his followers wished to denigrate might have been difficult to countenance. In all probability it was eventually decided that the least unusual solution was the one least likely to cause disturbances and initiate a cult; Harold was therefore sent for burial at Waltham. There may have been a delay while the decision was taken. Both Poitiers implicitly and the Waltham Chronicle explicitly say that the corpse was difficult to identify. The Carmen and Poitiers most likely enjoyed recording the early deliberations about Harold’s final resting place; their utterly orthodox outcome 143 Note cuius ob nimiam cupiditatem insepulti remanerent innumerabiles, GG, 140–1. On this subject, Jennie Hooper, ‘The “Rows of the Battle-Swan”: The Aftermath of Battle in Anglo-Saxon Art’, in Strickland (ed.), Armies, Chivalry and Warfare, 82–99, at 96–8. 144 Carmen, 34, line 594. 145 Ibid., 34–5; GG, 140–1 (Dictum est illudendo, . . .). 146 GR, i, 460–1; The Waltham Chronicle: An Account of the Discovery of our Holy Cross at Montacute and its Conveyance to Waltham, ed. Leslie Watkiss and Marjorie Chibnall, OMT (Oxford, 1994), 50–6, with the whole problem discussed at pp. xliii–xlvi.

246

william the conqueror

was a topic in which neither would have been interested. Poitiers may also have had classical-referencing in mind in the way he has William refuse an offer of money from Harold’s mother Gytha, because this made him more noble than Achilles, who had accepted King Priam’s offer for the body of his son Hector, whom Achilles had killed in single combat and to whom Poitiers had compared Harold.147 In using the analogy of Achilles and Hector for William and Harold, Poitiers, who was exceptionally well read in the Latin Classics, may also have been making coded criticisms of William.148 Achilles was in no way a straightforward model of knightly chivalry; among other things, for example, as Poitiers and his readers would have known, he had dragged Hector’s body nine times around the walls of Troy behind his chariot. Poitiers also specifically has Gytha refer to Harold as ‘her beloved child’ (pro corpore dilectae prolis).149 There was presumably no need for him to include the adjective. Even in the moment of victory, and in an apparently unlikely place, Poitiers may well be expressing the ambivalence about William that surfaced so often. The sheer length of the battle does surely show that the two armies were evenly matched, at least numerically. It is, therefore, probably wrong to accept the opinion that the Battle of Hastings was a rout from the very beginning. On the other hand, the Bayeux Tapestry’s presentation may well be drawing attention to the massacre that the battle became in its later stages and asking its audience to ponder its morality. The battle may have become all the more disturbing because the later stages took place in darkness. Once Harold had been killed, why was it necessary to continue the slaughter of an obviously defeated army?150 In terms of the overall course of the battle, while modern discussion has abandoned the notion that French cavalry possessed decisive technical superiority over English infantry, and that it was incapable of breaking through with a shock cavalry charge, the innate advantage of greater mobility must be important. Harold could have won only if he had worn his enemy down to such an extent that his infantry could then advance like a steam-roller down the slope, or if William’s troops had become so exhausted that they had been forced to withdraw.151 By wheedling away with tactical retreats and using 147 For this subject, GG, 140, note 3, with a crucial point being that Poitiers would have known the Iliad at second-hand through the Latin Ilias Latina. 148 GG, 134–5. 149 Ibid., 140–1. 150 Cf. Pastan and White, The Bayeux Tapestry and its Contents, 237–59 (‘On the contrary, the embroidery depicts a terrible mismatch – a rout, a slaughter – in which the English were doomed to defeat from the very start of the battle’). 151 There is again a vast bibliography on issues of this kind. See, with emphases similar to those here, Morillo, ‘Hastings: An Unusual Battle’, 101–2; Matthew Strickland, ‘Military Technology and Conquest: The Anomaly of Anglo-Saxon England’, ANS, xix (1997),

t h e y e a r o f v i c t o ry

247

his archers skilfully, William ultimately ensured that his troops slowly gained the initiative. The great drama of the Battle of Hastings does show off the brute physicality that made William so effective in a society in which military prowess was culturally central. But it also reveals his intelligence and thorough preparation which meant it was not until 1076, when he was approaching his fiftieth birthday, that William suffered any significant military setback. FROM HASTINGS TO CORONATION William and his army remained in the vicinity of Hastings for a few days after the battle, according to the ‘D’ Chronicle, in the hope of receiving the first English submissions, as well as, it must be assumed, to recover and regroup.152 When the submissions did not materialize, William set out on a progress around south-eastern England, which both the Carmen and the ‘D’ Chronicle present as intimidation, with the Chronicle specifically stating that the land was ravaged up until the final English submission at Great Berkhamsted (Hertfordshire).153 John of Worcester later said that the army ravaged Sussex, Kent, Hampshire, Middlesex, and Hertfordshire.154 The Carmen also mentions tribute-taking.155 Poitiers, who is the only historian to describe in detail William’s itinerary around the south-east, was presumably mostly silent about all this because it did not fit with his purpose. Malmesbury’s portrayal of a stately progress that befitted a victor also omits the ravaging and does not convince.156 At Romney, William avenged the killing of the troops who had landed there, an act of strategic retaliation that not only fitted a general perception that such revenge was justified, but, on these terms, would also have been good for his troops’ morale.157 When the army reached Dover, some English who may have been intending to defy him lost confidence and surrendered. While the negotiations with them were going on, some lower 353–82, at 359–69; Matthew Bennett, ‘The Myth of the Military Supremacy of Knightly Cavalry’, in Strickland (ed.), Armies, Chivalry and Warfare, 304–16; Halsall, Warfare and Society, 180–8. 152 ASC, ‘D’, 1066. 153 Carmen, 38–9, lines 657–8; ASC, ‘D’, 1066. 154 JW, ii, 604–7. 155 Carmen, 36–7, lines 625–6. 156 GR, i, 460–1. On the period from Hastings to the coronation, see Paul Dalton, ‘William the Peacemaker: The Submission of the English to the Duke of Normandy, October 1066– January 1067’, in Paul Dalton and David Luscombe (eds.), Rulership and Rebellion in the AngloNorman World, c.1066–c.1216: Essays in Honour of Professor Edmund King (Farnham and Burlington, VT, 2015), 21–44, at 24–37. 157 Knowledge of William’s march through Kent depends mainly on GG, 142–5, with valuable additional detail in Carmen, 36–7.

248

william the conqueror

ranks in William’s army started a fire, apparently while in search of booty. Poitiers excuses William’s failure to punish them on grounds of their number and humble social status. If perhaps understandable in the circumstances of a campaign that was in all likelihood turning out to be more arduous than most had expected, what looks like a serious breach of discipline was presumably tolerated because William could not risk an action likely to be unpopular; as we have seen, troops expected to profit from looting. During the eight days spent at Dover organizing the repair of the fortifications, many of William’s men were stricken with dysentery. When the decision was taken to move on, those who were still suffering were left behind with the garrison that had been installed there. Soon afterwards, the citizens of Canterbury came out of their city to surrender. However, at an unidentified place that Poitiers calls ‘the Broken Tower’, William was taken ill. The nature of his affliction is not specified and Poitiers presents him as pushing on regardless in order to sustain momentum. The Carmen, however, while not mentioning the illness, says William remained in the vicinity of Canterbury for a period of a month.158 News of this must have reached London and encouraged a group led by Archbishops Ealdred and Stigand, and supported by Earls Edwin and Morcar, to put Edgar the Ætheling forward as king.159 All, it might have seemed to them, was not lost. Just how widespread was the expectation that resistance would continue is demonstrated by Peterborough abbey’s request to Edgar that he confirm the appointment of their new abbot, who was named Brand and whose predecessor had been present at Hastings and had died soon afterwards.160 The Carmen even seems to indicate that a coronation ceremony was performed, but with Poitiers saying nothing and all the later sources silent, it is surely certain that Bishop Guy was misinformed.161 Explaining Edgar’s promotion in terms of the English people’s wish to be ruled by one of their own, Poitiers probably rhetorically diverts the depth of hostility towards William into a patriotic cause whose motivation his readers would comprehend as trumping even William’s supposed fine qualities. The failure to crown Edgar may suggest that the allies lacked commitment. Alternatively, William’s recovery and the destructive advance of his troops may well have undermined their confidence. In the ‘D’ Chronicle’s words: ‘the more it Carmen, 36–7, line 623. Ibid., 146–7. Ealdred’s role is mentioned in JW, ii, 604–7. Poitiers stresses Stigand’s wealth and influence, on which see Mary Frances Smith, ‘Archbishop Stigand and the Eye of the Needle’, ANS, xvi (1994), 199–219; H. E. J. Cowdrey, ‘Stigand (d.1072)’, ODNB. 160 ASC, ‘E’, 1066. 161 Scilicet ut puerum natum de traduce Regis/In regem sacrent, ne sine rege forent, Carmen, 38–9, lines 647–8. 158 159

t h e y e a r o f v i c t o ry

249

ought to have been taken forward, the more it got behind, and the worse it grew from day to day’. William of Malmesbury suggests that either Edwin or Morcar wanted to be king and John of Worcester, that the two of them abandoned Edgar.162 In both cases, these comments reflect their authors’ perception that God’s will had been defied for too long and that the disunity and self-seeking opportunism of the English elite had brought the kingdom to ruin. William’s actions after his recovery from illness were designed to accomplish the weakening of any English will that might have existed to raise troops for a second battle. The mistaken notion that drastic falls in Domesday Book values at the moment when an estate was acquired indicate the line of the march of William’s army around the south of London has long been deeply embedded in the historiography.163 However, while the army would undoubtedly have moved forward en bloc, sending out foraging and reconnaissance parties as required, and would have made for strategic centres in order to secure their submission, the distribution of low values indicates a continuation of the campaign of intimidation that had begun on the south coast before the Battle of Hastings. The main army most likely crossed Surrey along the ridgeway to Guildford and Farnham, before turning northwards towards Wallingford via Basing. According to Poitiers, a detachment sent northwards towards London defeated some English who had sallied forth from the city across London Bridge, and set fire to houses situated on the south bank of the Thames. The, in relative terms, exceptionally low Domesday Book values along the south bank of the Thames must also indicate that deliberate destruction was employed there to show William’s disapproval and to undermine the morale of his opponents. Incidents of this kind may well explain other instances where the land values cannot be made to fit a preconceived line of march. With Poitiers writing that ‘the fierce pride of his enemies [was] subdued by a twofold disaster’, the evidence of a narrative source and Domesday Book dovetails into a coherent story of continuing intimidation.164 All this must have confirmed William in the view that London was too large and well defended to be taken by a direct assault.

ASC, ‘D’, 1066; GR, i, 460–3; JW, ii, 606–7. Thus, e.g., John Blair, ‘An Introduction to the Surrey Domesday’, in Ann Williams and R. W. H. Erskine (eds.), The Surrey Domesday (London, 1989), 1–17, at 5. Palmer, ‘The Conqueror’s Footprints’, surveys the large earlier literature and emphasizes the difficulties of the evidence. For the evidence on which this paragraph is based, the detail in F. H. Baring, Domesday Tables for the Counties of Surrey, Berkshire, Middlesex, Hertford, Buckingham and Bedford (London, 1909), 20–3, 207–16, remains useful. 164 GG, 146–7 (ut malo duplici superba ferocia contundatur). 162 163

250

william the conqueror

From the Carmen we learn that William sent a detachment to Winchester to secure the city’s surrender and to obtain the payment of tribute. While there, he reached an agreement with Queen Edith that allowed her to keep her dower lands, thereby permitting her to live as befitted a dowager queen. The Vita Ædwardi Regis on its own is sufficient to refute Poitiers’s statement that Edith consistently supported William. Its regretful tone suggests that, in the immediate aftermath of Hastings, with her family decimated, she may have resigned herself to her fate and decided to come to terms.165 Unlike several of Harold’s sons by Edith the Fair and, in all probability, her mother Gytha, she apparently did not later move into direct confrontation with the new regime. But, as we shall see, Queen Edith was never passively acquiescent, and the Vita’s much-discussed allusive references that distinguish between the conduct of her brothers are arguably indicative of the theme also picked up by William of Malmesbury, of fraternal strife as the cause of disaster. In its complex multi-layered way, the Vita’s anonymous author found a path through the wreckage that enabled him scarcely to mention William’s victory at all.166 The reciprocity created by this settlement with one of the most important figures of the old regime, along with the gift exchanges implied by tribute-taking, can be seen as the start of a process in which William was trying to set up working relationships with his new subjects. It was of course in his interests to treat well the widow of the king to whom he supposedly owed his kingdom, as well as its second most important town and a place of central importance to the landscape of regality. When William had crossed the Thames at Wallingford, with the intention presumably being to complete the encirclement of London from the north, other submissions started to come in. First was Archbishop Stigand of Canterbury who, according to Poitiers, submitted at Wallingford. What was in effect a general submission was made shortly afterwards at Berkhamsted, with the ‘D’ Chronicle and John of Worcester naming Archbishop Ealdred, Edwin and Morcar, Edgar the Ætheling, Bishops Wulfstan of Worcester and Walter of Hereford, and the chief citizens of London as present. ‘D’ grimly added that they should have given up long before so as to spare the kingdom from William’s ravaging, and recorded a settlement by which William took hostages, secured oaths, and promised to be a good lord to them all. It also mentioned that he nonetheless continued to destroy the countryside; this must to an extent presumably have been to 165 Carmen, 36–7; GG, 114–15. The idea that Winchester was taken by fresh reinforcements that had crossed from Normandy to the Solent is based on a dubious interpretation of Domesday Book statistics, on which see Palmer, ‘Conqueror’s Footprints’, 31–2. 166 Monika Otter, ‘1066: The Moment of Transition in Two Narratives of the Norman Conquest’, Speculum, lxxiv (1999), 565–86, at 579–85.

t h e y e a r o f v i c t o ry

251

feed his troops, but probably also to continue to intimidate those who might still be inclined to oppose him.167 The accounts vary in the detail they give on William’s advance to London. The Carmen refers to preparations for a siege and some devious negotiating, while Jumièges mentions fighting within the city between an advance party and some of the citizens.168 Although none of this is recorded elsewhere, the accounts ring true in the context of the uncertainties of the times. Poitiers’s silence on these matters is once more probably a consequence of it not fitting his rhetorical purpose to include them. However, the notion that a major battle took place within the city near Ludgate is a mistaken one. The unpublished report on the burials discovered in the Ludgate/Fleet Valley excavation during the construction of the Thameslink line between Holborn and Blackfriars is most unlikely to be indicative of a battle; quite apart from the problems of dating the finds, only three skulls and eleven bodies were discovered.169 It is reasonable to think that there must have been disorder and posturing as William took over the greatest city of the English kingdom. But there is no sign that there was significant resistance. Once London and Westminster had been occupied, Poitiers describes a debate between William and his followers as to whether he should be crowned or should delay his coronation until the kingdom was fully subjugated and Matilda could be crowned with him. His followers, led by the Poitevin Aimeri, vicomte of Thouars, persuaded him to be crowned; most abnormally, so Poitiers says, even ordinary warriors (milites) were asked their opinions.170 When dealing with the same subject, Orderic, from this point producing an amended version of the Gesta Guillelmi with notably different emphases, omitted all reference to William’s reluctance and simply said that he accepted the advice of the Normans and the English and went ahead with his coronation.171 These narratives and the associated possibility of cross-referencing between Poitiers and the Pseudo-Isidorean Decretals is a facet of our understanding of the cultural and intellectual preparations for the Conquest. As already noted, well known in Normandy long before 1066, the Pseudo-Isidorean Decretals included in its text Canon 75 of the Fourth Council of Toledo of 633, which stipulated that ‘the bishops and the head men of the people would decide who should 167 GG, 146–7; ASC, ‘D’, 1066; JW, ii, 606–7. For the identification of Berkhamsted as ‘Great Berkhamsted’, Palmer, ‘Conqueror’s Footprints’, 33–4. 168 Carmen, 40–5; GND, ii, 170–1. 169 Cf. Peter Mills, ‘The Battle of London 1066’, London Archaeologist, viii, no. 3 (1996), 59–62. I am grateful to Derek Keene, John Clark, Francis Grew, and Bob Cowie for their advice on the Ludgate/Fleet Valley burials. 170 GG, 146–51. 171 OV, ii, 182–3.

252

william the conqueror

succeed to a kingdom’, therefore broadly speaking setting out the procedure that Poitiers was describing.172 The version of Pseudo-Isidore known as the Collectio Lanfranci, now preserved at Trinity College, Cambridge, was in all likelihood written in Normandy a little before 1066. It contains a marginal mark alongside this canon, suggesting that someone wanted to draw attention to its specific importance (Plate 9).173 As emphasized in relation to the preparations for the invasion, with consultation being a norm within Norman aristocratic society, albeit probably one that was mostly used to ratify decisions that William wanted taken in order to ensure participation, or at most to modify them, such debates as took place were mostly staged and theatrical. Pseudo-Isidore may have had little to do with the matter, although, as we shall soon see, it was probably extremely important for other aspects of these last weeks of 1066. A rapid resolution and an early coronation were presumably expected by the thousands who had put their lives at risk by supporting William. Likewise, Poitiers’s reference to the English asking William to be crowned must indicate that a new stability was being sought to bring an end to the killings and the destruction.174 A display of humility may well have been thought a necessary adjunct of pre-coronation theatre by contemporaries. With so many influences coming into England and Normandy from the Empire, it is likely that the traditions associated with the theatre of Emperor Conrad’s coronation in 1024 were known; it was perhaps no coincidence that Conrad’s accession also took place after a disputed succession. The account also put great emphasis on the importance of the coronation of Conrad’s queen, mentioning that she was descended from Charlemagne.175

172 Decretales Pseudi-Isidoriannae et Capitula Angilranni, ed. Paul Hinschius (Leipzig, 1863), 373–4 (primatus totius gentis cum sacerdotibus successorum regni concilio communi constituent). For this argument, Garnett, ‘Coronation and Propaganda’, 108–10. Also Neil Strevett, ‘The AngloNorman Civil War of 1101’, ANS, xxvi (2004), 159–75, at 171. 173 The manuscript’s Le Bec origin arguably dates it to before Lanfranc’s transfer to Saint-Etienne of Caen, although it cannot be conclusively shown that it was composed under Lanfranc’s supervision. The scribe’s insertion of a letter of Alexander II to Lanfranc about Saint-Etienne does seemingly link it to Lanfranc in Normandy. See further Michael Gullick, ‘The English-Owned Manuscripts of the Collectio Lanfranci (s.xi/xii)’, in Lynda Dennison (ed.), The Legacy of M. R. James: Papers from the 1995 Cambridge Symposium (Donnington, 2001), 99–107, at 100–2; idem, ‘Lanfranc and the Oldest Manuscript of the Collectio Lanfranci’, in Bruce C. Brassington and Kathleen G. Cushing (eds.), Bishops, Texts, and the Use of Canon Law around 1100: Essays in Honour of Martin Brett (Aldershot, 2008), 79–89, and esp. pp. 83–4. 174 GG, 146–9. 175 Wiponis Gesta Chuonradi II Imperatoris, ed. Harry Bresslau, MGH Scriptores (Hanover and Leipzig, 1915; reprinted Hanover, 1977), 20–5 (translated, Imperial Lives and Letters of the Eleventh Century, trans. T. E. Mommsen and K. F. Morrison [New York and London, 1962], 66–9).

t h e y e a r o f v i c t o ry

253

With Poitiers making it explicit that some believed further pacification was a desirable preliminary to coronation, there must have been a widespread awareness that, by taking a coronation oath, William was agreeing to undertake responsibilities for all of his subjects in ways that might obstruct the imposition of power. Archbishop Ealdred, as we shall see, must have been making his own preparations with exactly the same eventuality in mind. On the basis of the way in which William and Matilda’s political relationship had come to be defined, it is also quite possible that William’s personal preference was for a joint coronation with Matilda, although this may well have been for formulaic reasons, a necessity required by scripts long defined in England and elsewhere. As so often, in setting out these debates, Poitiers was probably trying to defend William against criticisms, but in this case from both victors and vanquished; the carnage of the following years must have been very fresh in many minds when he was writing.176 With hindsight, there might have been better ways of doing things. Or so it was thought. The extensive prior preparation for the assumption of kingship must also have been a factor in pushing William to a coronation. As we shall see, this was very likely to have been the case with the seal made for William at this time. Also, if, as the Carmen suggests, William had new regalia made, it is impossible that all the work could have been completed in the short period between London’s submission and the coronation. In particular, the new crown described by Bishop Guy has often been compared to imperial exemplars, among them the crown of one of the most successful medieval emperors, Otto I the Great (king of the Germans, 936–73; emperor, 962–73).177 And, if the story of the new crown is set aside as intrinsically incredible, it still remains as early evidence of a powerful momentum that emphasized William’s new dignity. In the Carmen’s words, he ‘discarded the title of duke to be made a king’, as it happens a technical inaccuracy, but a statement that concretely describes his move upwards in the hierarchy of earthly dignitaries.178 Although usually impossible to date precisely, the Caesar motif appears in several poems written in praise of the new king.179 On the basis of the narrative set out in this chapter, William’s conduct at Hastings and up until his coronation was both extremely effective and GG, 148–9. For the crown, Carmen, 44–7, lines 757–82. For the Otto I parallel, Karl Leyser, ‘England and the Empire in the Twelfth Century’, TRHS, 5th ser., x (1960), 61–83, at 65; cited in Carmen, 45, note 6. 178 Carmen, 44–5, line 756 (Et ducis abiecto nomine, rex fieri). 179 Elisabeth van Houts, ‘Latin Poetry and the Anglo-Norman Court 1066–1135: The Carmen de Hastingae Proelio’, JMH, xv (1989), 39–62, at 41–2, 56–7 (History and Family Traditions, chapter 9). 176 177

254

william the conqueror

notably ruthless, a point confirmed by the emphasis on the extreme violence of William’s campaign that is present in many contemporary sources written outside Normandy and England.180 In terms of what was to follow, Poitiers’s deployment of the doctrine of tyrannicide throughout the Gesta Guillelmi, even citing Cicero to the effect that to kill a tyrant was ‘the finest of all glorious deeds’, supplies an intellectual framework that may either have been in place in 1066 or which was developed afterwards to justify what was done.181 The major issue both in the present and for the future was going to be how far the English who had supported Harold could be judged as accomplices and treated accordingly. Notwithstanding George Garnett’s observation that Poitiers ‘remains revealingly unclear about the reason why the English shared in their new king’s [i.e. Harold’s] guilt’, it is certain that, where William was personally affected, they definitely did share in it. Thus, in the one place where Poitiers seemingly slips up by describing William as a king before 25 December 1066, he is said at the Battle of Hastings to have ‘rushed forward at their head, brandishing his sword, and mowed down the hostile people who deserved death for rebelling against him, their king’.182 In other words, we are dealing with an affront to the personal dignity of a man who deemed himself to be rightfully a king and who, in this situation, was showing it. Poitiers’s readiness also to use the language of barbarism in relation to the English further reinforces the point he wanted to make, which was that their cultural backwardness and their complicity in Harold’s cause justified William’s extension of the targeted violence that had been a feature of his entire career. This violence could legitimately be extended to an entire people until unrest and rebellion had come to an end. The fact that the passage differentiating the people of Kent from the savage people who inhabited the rest of the island was drawn directly from Caesar’s De bello Gallico only gave greater authority to Poitiers’s rhetoric.183 It is important, too, that none of this language of impiety and barbarism appears in the Carmen. There Harold is every bit as much a perjurer as in Poitiers’s account, but the English are simply described as foolish for supporting him. For this they deserve to die, but there is none of the religious and ethnic inferiority that appears in Poitiers.184 The presence in both of them of praise for the patriotism of the English, and their wish to defend their land against invaders, was in part because this patriotism emphasized Van Houts, ‘The Norman Conquest through European Eyes’, 852–3. GG, 26–7. 182 Ibid., 130–1. 183 Ibid., 164–5. 184 Carmen, 10–11, lines 147–8 (Nec mirum, regem quia te plebs stulta negabat;/Ergo perit iuste, audit et ad nichilum). 180 181

t h e y e a r o f v i c t o ry

255

sincerity and devotion, albeit misguided; but the praise also carried with it, for any English man or woman who read either the Gesta Guillelmi or the Carmen, the message that such patriotism should be transferred to William.185 It may be important that all this can be fitted into the framework set out in passages in Pseudo-Isidore highlighted by Lanfranc. The extensive consultation brought all the English who had opposed William within the framework of his discretionary favour and disfavour. Whether conceptually defined in terms of ira and misericordia or a variant of lèse-majesté, it made William’s will all-encompassing. Additionally, in the context of 1066, Harold’s perjury shaped everything. Forgiveness for supporting him was possible, as with the bishops who had participated in Harold’s coronation or with Edgar the Ætheling, whose submission is dramatically described in the Carmen.186 In the case of the bishops, they could be forgiven probably because they were special people, the Lord’s anointed, although, as is well known, this was in several instances to be no protection. In the case of Edgar, he was presumably deemed never to have participated in Harold’s takeover of the kingdom. He might even have been viewed as a victim of it. Or perhaps he was accorded special treatment as the closest living kinsman of King Edward. The link that was made between Harold’s supposed oath-breaking and doctrines of tyranny also goes a long way towards explaining the differences between the conquests of England and Maine. In the latter case, no oaths had been sworn. It is notoriously difficult to name all but a tiny number of the thousands on either side who died at Hastings or in the warfare of 1066.187 Creating the life histories that would illuminate what must surely have been the most extraordinarily fraught of times is therefore mostly impossible, although, as we shall see, not always so. The disruption of families and to property management in the south and east of England would have been huge. A general idea of the tensions that must have existed and of the culture that could have determined behaviour can surely be gained by bringing into the discussion the honour-based and feud-orientated values set out in Chapter 2, which were equally present in societies on both sides of the Channel. Of course, the decision as to whether to pursue a feud involved individual choice. Resort to violence was not inevitable. No outcomes were predetermined and mediation might be imposed. There were certainly forces at work that made for accommodation. Whatever the case, all contributed to a chemistry of uncertainties, fears, and expectations that created an GG, 126–7; Carmen, 22–3, lines 371–2. Carmen, 44–5, lines 740–52. 187 For English names, Lawson, The Battle of Hastings, 127–8, 189. 185 186

256

william the conqueror

extremely combustible compound which has to be seen as central to the new world that started to emerge after 1066. William was crowned in Westminster Abbey on Christmas Day 1066. Neither William of Jumièges nor the Carmen name the archbishop who crowned the new king, but both Poitiers and the ‘D’ Chronicle cite Archbishop Ealdred of York.188 The Carmen, in indicating that both archbishops were present, and Jumièges, in referring to the collective role of the bishops of the realm, could be more accurate. It is inconceivable that someone with Stigand’s influence would have been left out, and the possibility that, if events had turned out otherwise after Bishop Guy had completed his poem, William might have worked to regularize Stigand’s position, cannot be ruled out. Both ‘D’ and Poitiers had reasons to single out Ealdred. This was because the likely author of ‘D’ was in his entourage and because of the way in which the case for William’s legitimacy was being made after Stigand had been deposed in 1070. Whereas Poitiers mostly presents the coronation as the grand culmination of William’s fight for his just rights, the ‘D’ Chronicle portrays it in terms of a concession to William by Archbishop Ealdred, saying that Ealdred would crown William only when he agreed to swear on the Bible that he would rule as well as the best kings before him, if the people would be loyal to him. John of Worcester amplifies the statement by saying that William agreed to rule each of the peoples subject to him justly while also having Ealdred take the initiative.189 This version of events complements Poitiers’s statement that the ceremony began with an address by Ealdred; the two accounts are best seen as reflecting Poitiers’s emphasis on William’s fortitude and the ‘D’ Chronicle’s focus on the history of the English kingdom. The whole story also fits with William’s often displayed willingness – or perhaps it is better to say apparent willingness – to prostrate himself before ecclesiastical authority in order to legitimate his position. In this case, appearing as a supplicant would strengthen his hand – in the eyes of both God and man – if the English did not keep their side of the bargain. According to Poitiers, after the English had acclaimed William as their king and after Bishop Geoffrey of Coutances had addressed the French, who then did the same, the noise of the acclamation so alarmed the soldiers guarding Westminster Abbey that they set fire to some houses in the city. Nothing could illuminate better the true character of William’s accession to kingship than this tale of heavy security, raw nerves, and, ultimately, panic. Orderic, while relying largely on Poitiers, elaborated the story with 188 GND, ii, 170–3; Carmen, 46–7; ASC, ‘D’, 1066; GG, 150–1. For Ealdred, see also JW, ii, 606–7. 189 JW, ii, 606–7.

t h e y e a r o f v i c t o ry

257

an account of mass flight and a badly shaken king being crowned by the bishops and the small number of clergy and monks who remained.190 As a piece of theatre, this fits exactly with Orderic’s general perception of the Norman Conquest. The events were the work of the Devil, a cause of the long-term discord between Normans and English and portents of the troubles to come.191 If Orderic’s portrayal of a trembling king is not pure invention, it shows a man fearful of divine wrath and dismayed at the ruin of what should have been the greatest ceremonial event of his life. If, as is more likely, Orderic overplays the story to make his point, it is still impossible not to think that William would have been shaken by the turn of events. In the end, he had achieved all he wanted after the coronation, namely legitimacy according to English tradition and a guarantee of loyalty from his new subjects – it matters not whether the so-called Second or Third ordines were used.192 Yet some must surely have thought this a shabby beginning to a new era.

GG, 150–1; OV, ii, 182–5. Doubts about the accuracy of Orderic’s story were expressed by Marjorie Chibnall: see Janet L. Nelson, ‘The Rites of the Conqueror’, ANS, iv (1982), 117–32, 210–23, at 214, note 48; (reprinted in idem, Politics and Rituals in Early Medieval Europe [London, 1986], 375–401, at 386, note 48). 192 For discussion, Nelson, ‘The Rites of the Conqueror’, 119–28 (Politics and Rituals, 380–94); George Garnett, ‘The Third Recension of the English Coronation ordo: The Manuscripts’, HSJ, xi (1998), 43–71, esp. pp, 65–71. 190 191

Chapter 7

K I N G O F T HE ENG LISH

INTERPRETING THE FIRST YEARS OF WILLIAM’S ENGLISH KINGSHIP William’s coronation was certain to mark a new beginning: a new beginning in the history of England, the British Isles, and, although it is much less studied as such, in the histories of Normandy, France, Scandinavia, and Europe. Those alive on Christmas Day 1066 who speculated on the likely future would have been well aware of all this, but naturally quite unable to see the course that events were going to take. They certainly would not have known they were at the start of a process that can be traced fully, but not comprehensively, through Domesday Book’s testimony, namely, the decimation of the fortunes of the native English aristocracy. To refer back to the Prologue for a moment, we are dealing with what John Gillingham has characterized as ‘an event unparalleled in European history’.1 Debatable as such a statement must be, staying a little longer with another theme of the Prologue, it is only through biography that we can understand why events should have happened that could cause one of the most influential historians of the subject to write in this way. In the end, it was decisions taken by William that were of prime importance in shaping events. These decisions can, however, only be comprehended through an understanding of the context in which they were taken. The coronation oath and the supplementary oath sworn to Archbishop Ealdred gave ideological shape to William’s rule in England and seemingly protected the English. However, the arbitrary powers vested in William as king were to be of much greater immediate significance to the shaping of the English kingdom in the years after 1066 than an abstract framework of legal and ethical responsibility. As the channel through which he distributed power and favour, they supplied the mechanisms through which he dealt with the multiple pressures present in the aftermath of victory. For those who lived through these events, the legal continuity from King Edward’s day enshrined in writs, charters, and Domesday Book must often 1

Gillingham, ‘Problems of Integration’, 85.

258

king of the english

259

have seemed close to meaningless. One of the central themes of the years from 1066 to 1087 is how this proclaimed legal continuity did become a central organizing principle. While always present in the background, and sometimes moving into the foreground, it is, however, a grave mistake to think that this continuity alone determined William’s acts and decisions during his years as king. Modern studies of the fate of the English after 1066 have rightly emphasized that the results of the survey that became Domesday Book underestimate the survival of English landholders. A comprehensive analysis that takes account of the methodological problems has produced the names of approximately 600 landholders and suggested that there are many hundreds who go unrecorded.2 The Conquest was undeniably a catastrophe for English landholders, which becomes all the more shocking when the four major ones who do appear in Domesday are discounted from a statistical analysis. The rationale that shaped such processes can only be comprehended by taking full account of William’s reaction to the pressures on him dating back to his decision to claim the English kingdom and then to invade it, and by trying to understand what it was about William that made him proceed in the way he did. Such an analysis must focus not just on the English, but on the Normans and all the other peoples involved, and especially on the norms of eleventhcentury rule and how William interpreted and put them into practice.3 It is very important to keep in mind that the conceptual framework of power and culture was in many respects very different from one that we would recognize. For example, in the succinct words of Hugh Thomas: ‘though there was no official ideology of assimilation, there clearly was a theoretical policy, stemming from the king, that the conquered English had rights and should be treated accordingly’.4 Equally important are the same historian’s words that ‘it would be a mistake to underestimate the hatred engendered by the violence of the conquest’, the only reservations being that it is possible to hate and to co-operate, and to wonder whether the notion of ‘two sides’ does not simplify matters.5 2 The classic and indispensable book is Ann Williams, The English and the Norman Conquest (Woodbridge, 1995), 71–125. See also C. P. Lewis, ‘The Domesday Jurors’, HSJ, v (1993), 17–44. For a full study, see DP, with conclusions set out at 23–4. For methodology, see C. P. Lewis, ‘Joining the Dots: A Methodology for Identifying the English in Domesday Book’, in Keats-Rohan (ed.), Medieval Prosopography, 69–87. 3 For an important study that takes account of some of the factors that are central to my argument, Hugh M. Thomas, ‘The Significance and Fate of the Native English Landholders of 1086’, EHR, cxviii (2003), 303–33. 4 Hugh M. Thomas, The English and the Normans: Ethnic Hostility, Assimilation, and Identity 1066–c.1220 (Oxford, 2003), 87. This book contains numerous stimulating observations on the context in which power was exercised. 5 Ibid., 61.

260

william the conqueror

Superficially, the period from Christmas 1066 until Whitsun 1068 might appear to have been one of relative calm. But the seeds of the conflicts that broke out after Whitsun 1068 were either already in the soil or were sown during this short period. With the Battle of Hastings having the appearance of being a single decisive victory that transformed everything – as, indeed, it ultimately turned out to be – it is essential to grasp that the normal expectations of reward that had been central to the conduct of war for centuries before 1066 created obligations among his followers that William had to meet. Moreover, continued participation in the pacification of the English kingdom was bound to require more of the same. Alongside this, the impact of defeat on the population of England must have created something like a violent whirlpool in which all struggled to swim, albeit some much better than others. This approach fits well with Orderic’s statement that the events of William’s coronation were the work of the Devil that began a time of irreconcilable discords, in modern parlance presumably a perception that this was a situation where forces were present that it was beyond human agency to control.6 While the decisions taken by William and those around him, though calculated and rational, were not the only ones they could have taken, it is also significant that, however well he prepared himself for kingship, he was in new territory from 25 December 1066 onwards. Unfolding events exacerbated the situation further. The tendency to place special reliance on trusted individuals and the apparent inability truly to forgive those who had opposed him were, however, surely going to be part of William’s personal approach to rule, as they had been in Normandy before 1066. For this reason, a very important conclusion to emerge from what follows is an expression of doubt about the validity of any references to ‘William’s early mildness’ and ‘that political dream, briefly cherished by King William, of Englishmen and Normans joining hands to forge an “Anglo-Norman” kingdom’, which appear in some of the most distinguished of publications.7 William’s treatment of the English was certainly determined by the loyalty and utility of individuals and communities, but it was also shaped – and probably decisively shaped – by the security and rewards demanded by those who had participated in the invasion of 1066 and who then committed themselves to sustaining the new political organism that it had brought into being; and by an application 6 OV, ii, 184–5 (Interea instigante Sathana qui bonis omnibus contrarius est, importuna res utrique populo et portentum future calamitatis ex improuiso exortum est). 7 See, e.g., Williams, The English, 23; Bouet, ‘Is the Bayeux Tapestry pro-English?’, 214. Both scholars, it must be emphasized, see the period of cohabitation as only lasting until 1068.

king of the english

261

of the arbitrary powers embodied in ira and misericordia that had long been ruthlessly exerted.8 Reading the narrative sources for this period provides another reminder that what David Douglas referred to as ‘the controversies of the past’ were present from the very first day of William’s kingship.9 Even more than the pre-1066 sections of the Gesta Guillelmi, Poitiers’s narrative of the year 1067 must have been written to answer criticisms and should be interpreted as such. Thus, in writing that William set a limit on taxation and tribute so that they did not become burdensome and that he distributed rich fiefs to those on whose loyalty he relied for the kingdom’s security, Poitiers implicitly acknowledged that, even if all was done as equitably and compassionately as he wants us to believe, both were certain to provoke ambivalence in many and, in many others, immense resentment.10 In contrast, the passage in the ‘D’ Chronicle, which seems to be in constant dialogue with Poitiers, says, immediately after the mention of the supplementary coronation oath, that, despite the agreement with Ealdred, William imposed severe taxes on the people. In other words, the moderation that Poitiers wished to present to his audience did not exist for the Chronicle’s anonymous author. The regime was oppressive from the start. The annal for the year 1066 ends with the prayer ‘May the end be good when God wills.’11 When its powerful contemporary feel is taken into account, it becomes tantamount to a statement that William had almost immediately broken his oath to Ealdred and was failing to treat his new English subjects equitably. This, of course, was a verdict that William could not accept and as a result of which the commissioning of William of Poitiers becomes logical. If, as we have seen, on any interpretation, ‘D’ was written by someone close to Archbishop Ealdred, then it must reflect opinions held by those English who remained influential and who were trying to guide William. They, of course, had every bit as much of a need to justify their actions and to counter criticisms as William did. In terms of interpreting the ‘D’ Chronicle, for all the much-discussed complexities of the palaeography of the manuscript, the fact that the section dealing with events up until Archbishop Ealdred’s death in late 1069 was written by a single scribe, who was either copying or adapting a text that had already been written, makes it very likely that it had its origins 8 For the emphasis on loyalty and utility, Thomas, ‘The Significance and Fate’, 313–19. See also John Blair, Anglo-Saxon Oxfordshire (Stroud, 1994), 174–7. 9 Douglas, WC, xiii. See above for a discussion of this in the Prologue. 10 GG, 160–1, 162–5. 11 ASC, ‘D’, 1066.

262

william the conqueror

in a contemporary account.12 Textual and palaeographical evidence indicate that the Chronicle reached its final form in the 1080s, above all because the insertion of the eulogy of Edgar the Ætheling’s sister, (St) Margaret, Queen of Scots, must be later than the 1060s since her marriage to King Malcolm only took place in 1070. However, her reputation for exceptional piety having been established during her lifetime, and the point that the poem praising her forms an integral part of the text, must mean that the final version was written as part of a composite single whole.13 In addition, the confused annalistic rubrication of ‘D’, with the start of the year 1067 located at William’s return from Normandy in December 1067, the beginning of the 1068 annal being placed towards the end of that year and its end at Easter 1070, and the absence of entries under the rubrics 1070 and 1071, is another reason to think that it has its basis in an earlier tract which a later writer reorganized into annalistic form. So viewed, it becomes a commentary of exceptional importance on William’s first years as king of the English. In having the Vita Ædwardi Regis written for her at this time, Queen Edith was also making a statement, one whose motivations and intended audience are not entirely clear. However, even with a proposed dating of 1067–8 for its completion, the decision to produce it must have been taken in early 1067; it therefore represents a very early response from someone at the heart of events.14 Its eventual impact, which was probably not the one Edith was seeking, was to be profound, the very foundation on which the twelfth-century case for the Confessor’s sanctity was built. But, as a way to understanding her frame of mind in the first weeks of 1067, it may be significant that her one post-1066 charter, which, although dating from some years later, was written in Old English and certainly reached only a small audience, announced that she was ‘the Lady, King Edward’s widow’, whereas Matilda was William’s ‘bedfellow’; in other words, the new regime was illegitimate and long-standing doubts about William and Matilda’s marriage were still relevant.15 An extraordinarily large number of coins in 12 Ker, Catalogue of Manuscripts Containing Anglo-Saxon, 254. The Anglo-Saxon Chronicle, ed. Cubbin, xiv, 79, is more tentative, suggesting that the scribe started work a little after the beginning of the 1066 annal, but he still identifies the whole section after this as the work of his fourteenth-century hand. 13 Thomas A. Bredehoft, ‘Malcolm and Margaret: The Poem in Annal 1067D’, in Alice Jorgensen (ed.), Reading the Anglo-Saxon Chronicle: Language, Literature and History (Turnhout, 2010), 31–48, esp. pp. 40–2. 14 As noted in Chapter 3, on the Vita’s date I follow Stafford, Queen Emma and Queen Edith, 40–8, and Keynes and Love, ‘Earl Godwin’s Ship’, 199, in arguing for composition very soon after 1066; cf. Frank Barlow’s arguments for 1065–7, VEdR, xxix–xxxiii. 15 F. H. Dickinson, ‘The Sale of Combe’, Somerset Archaeological and Natural History Society Proceedings, xxii (1876), 106–13, at 107 (David Pelteret, Catalogue of English Post-Conquest

king of the english

263

her dead brother’s name minted at Wilton, where Edith often resided, may represent another statement by her; this evidence has even been used to argue that she tried to co-ordinate a fightback soon after 1066 which came to nothing.16 There is enough in all this to point up an imagined alternative world very different from the one that is likely to have existed around William. A further problem is that, after the end of the surviving text of the Gesta Guillelmi, we are dependent on second-hand knowledge of what Poitiers wrote for the period from 1067 to 1071 on Orderic’s adaptation of his text. While often reproducing extensive passages verbatim, Orderic supplies us with a direct ethical dialogue between the eleventh and the twelfth centuries. Thus, for example, while Orderic could reproduce Poitiers’s statements about William ‘reigning well and justly in prosperity and adversity’ and of ‘his virtue and high character’ as a king, he omitted all of Poitiers’s eulogies of his moderation and justice, and, most tellingly of all, he left out the statement that William took nothing unjustly from the English to give to his French followers.17 For Orderic, the knowledge that William had successfully subdued the English created the paradox that, while his triumph must have been pleasing to God, it entailed events that Orderic ultimately viewed as being a human catastrophe, ‘a mournful theme of ruin for the pen of true historians’.18 This also means that we can only attempt to reconstruct what Poitiers might have written about the years between 1067 and 1071 by trying to read between the lines of Orderic’s Historia.19 Awareness of the likelihood of invasion from Denmark must have been an omnipresent destabilizing factor during these years. For William and those associated with him, it was a major threat. It put the onus on security and on effective defence organized by reliable people. In writing that messengers were sent by the English to encourage intervention from Denmark, Poitiers makes it clear just how superficial was the constructed facade of normality in the early weeks after the Conquest.20 For some of Vernacular Documents [Woodbridge, 1990], no. 56); cited in Stafford, Queen Emma and Queen Edith, 276. 16 Gareth Williams, ‘Was the Last Anglo-Saxon King of England a Queen? A Possible Posthumous Coinage of Harold II’, The Yorkshire Numismatist, iv (2012), 159–70, at 163–9. 17 Nulli tamen Gallo datum est quod Anglo cuiquam iniuste fuerit ablatum, GG, 164–5. For Orderic’s revision of Poitiers, ibid., xxxii–xxxv, 162, note 4. From different perspectives, Roger D. Ray, ‘Orderic Vitalis and William of Poitiers’, Revue belge de philologie et d’histoire, l (1972), 1,116–27; Pierre Bouet, ‘Orderic Vital lecteur critique de Guillaume de Poitiers’, in C. E. Viola (ed.), Mediaevalia Christiana XIe–XIIIe siècles: Hommage à Raymonde Foreville (Paris, 1989), 32–43. For Orderic’s estimate of William, OV, ii, 184–5 (rex in aduersis et prosperis strenue utiliterque rexit and De cuius probitate et eximiis moribus . . .). 18 OV, ii, 190–1 (unde flebile tema de sua ruina piis historiographis ad dictandum tribuit). 19 GG, xxxviii–xxxix. 20 Ibid., 182–3.

264

william the conqueror

the English, a Danish invasion must have offered an appealing alternative. But for many – probably the great majority for whom self-preservation was the dominant instinct – it must have been a complication in a situation that at times probably seemed close to being out of control. There were of course many still residing in England who owed their fortunes to Cnut, as well as many English in Denmark who had travelled in the opposite direction during the time of Cnut’s North Sea empire.21 According to Adam of Bremen, the King of Denmark Svein Estrithsson, the son of Cnut’s sister Estrith who had briefly been married to William’s father Duke Robert, believed that he had a claim to the English kingdom based on a visit he had made to Edward the Confessor soon after 1042 that had involved a promise of the succession should Edward die without children. He added that this agreement was accompanied by an oath and an exchange of hostages. The inability of the English to organize themselves to fight a second battle was something that William of Poitiers commented on, putting it down to cowardice.22 This is another piece of typical Poitiers rhetoric, in this case one that grossly simplifies a complicated situation. The disjointedness of the opposition to William, already evident in the weeks between Hastings and the coronation, was certainly an aspect of it. But, in addition, although the little we know about the identities of those who had died at Hastings, and in other conflicts, makes the writing of life histories and the creation of specific scenarios well-nigh impossible, the disruptive impact on many lives and livelihoods must have caused great upset that fed into attitudes across the whole of English society.23 The effect of so much violence and so many deaths on the psyche and morale of the victors is also a factor for which we have relatively little direct evidence. Other episodes in history of a similar kind suggest that in such febrile conditions responses among the defeated like self-serving opportunism, co-operation based on the principle that it was the best route to individual and collective self-preservation, and actual and latent opposition, would all rapidly come to the surface. Bitter resentment and a huge level of anxiety 21 Adam of Bremen, Gesta Hammaburgensis ecclesiae Pontificum, 136 (trans. Tschan, 108). For the general case, Bolton, The Empire of Cnut the Great, 308–16. See also idem, ‘English Political Refugees at the Court of King Sveinn Ástriðarson’, Mediaeval Scandinavia, xv (2005), 17–36, at 17–19; Paul Gazzoli, ‘Anglo-Danish Relations in the Later Eleventh Century’, University of Cambridge PhD Thesis (2010), 78–103. 22 GG, 182–3. 23 For the few names mentioned in Domesday Book, Lawson, Battle of Hastings, 216–17; Garnett, Conquered England, 20, note 155 (DB, i, fos. 50r, 208r; ii, 275v, 409v, 449r). See also Historia Ecclesie Abbendonensis, i, 224–5.

king of the english

265

must have been directed towards the new rulers. But, equally, emotional responses would have taken the form of sullen resignation and of anger against those deemed responsible for what in time evolved into a truly terrible defeat. In such a situation, the local and sometimes the seemingly insignificant event could have a major impact. As we shall see, there are times when these forces come to the surface and leave their mark in the surviving sources. But they are only a tiny fraction of the multiple hidden histories of which we know nothing. The balance of forces across northern France also started to change after 1066. Baldwin V’s death on 1 September 1067 was the beginning of a sequence of events that led to a civil war in Flanders and to the new count of Flanders, Robert the Frisian, becoming William’s enemy after 1071, thereby tilting the balance of power around the North Sea littoral. The end of King Philip I’s minority in the last months of 1066 and, in Anjou, the usurpation and imprisonment in 1067–8 of Geoffrey the Bearded by his brother Fulk Rechin, with both new rulers subsequently drifting into hostility towards William, eroded the conditions that had facilitated the victories of 1063 and 1066. The coincidence of the overthrow of Norman rule in Maine in early 1069 and the arrival of a Danish fleet on the east coast of England soon afterwards created a threat with the potential to destroy all that William had achieved since 1063. Although Hastings had made William a king, the fundamentals of the political and cultural worlds surrounding Normandy and England on the continental mainland and around the North Sea remained largely unchanged. THE FIRST WEEKS OF KINGSHIP The unbridgeable abyss between the ‘D’ Chronicle’s statement that William’s first recorded action was to impose a severe tax and Poitiers’s idealized image of law-giving and moderation shows that the new king’s methods of rule were controversial from the start. For all that he was often to be extremely critical of William, in reproducing Poitiers’s text verbatim, Orderic was presumably indicating he believed that William’s initial intention was to try to govern in accordance with his promises. All three sources show that William had straightaway grasped the potential of the economic and organizational resources of the English kingdom. The tax taken must have been a geld, a system of national taxation that went back to the time of Æthelred the Unready’s efforts to buy off the Danes. Its continuation by Cnut had involved the taking of apparently massive sums. The abandonment by Edward the Confessor in 1051 of the heregeld, a national land tax to support the king’s army, was treated as a source of great relief in the

266

william the conqueror

Anglo-Saxon Chronicle.24 William, it appears, had almost instantly brought it back, presumably using the money to pay off troops and to impress his authority on his new subjects.25 Poitiers, still followed by Orderic, says that William made a progress around his new kingdom before his departure for Normandy in March 1067. Constraints of time, however, mean that he can only have made his personal presence felt in the south-east.26 Their statement that William received the submissions of Earls Edwin and Morcar at Barking near London after Christmas 1066 appears to contradict the ‘D’ Chronicle’s that they had submitted earlier at Berkhamsted.27 Although some – including David Douglas – have doubted that the meeting at Barking happened, a meeting with the two earls at which agreement was reached about their new roles seems likely; decisions taken at this time, as opposed to earlier, would have had a solemnity and authority that had not been present before the coronation. According to Orderic, but not Poitiers, William at the time that he made peace with Edwin, the elder of the brothers, promised him one of his daughters in marriage, an offer made earlier to Harold, and perhaps involving the same daughter, namely, Adelida.28 Others who submitted at Barking are named by Orderic, at this point again expanding Poitiers’s text, as the man conventionally known as Thorkell of Warwick, Siward son of Æthelgar, his brother Ealdred, and his cousin Eadric, who came to be known as ‘the Wild’.29 Since these last three names are associated with Shropshire, where Orderic was born in 1075, it is likely that his local knowledge has skewed his presentation at this point. However, if we think of what he wrote 24 For 1051, ASC, ‘D’, 1052 (recte 1051). In general, Judith A. Green, ‘The Last Century of Danegeld’, EHR, xcvi (1981), 241–58; Ann Williams, Kingship and Government in pre-Conquest England (Basingstoke and London, 1999), 115–16, 144–6. For the beginnings, M. K. Lawson, ‘The Collection of Danegeld and Heregeld in the Reigns of Aethelred II and Cnut’, EHR, xcix (1984), 721–38; idem, ‘ “Those Stories Look True”: Levels of Taxation in the Reigns of Aethelred II and Cnut’, EHR, civ (1989), 385–406; John Gillingham, ‘ “The Most Precious Jewel in the English Crown: Levels of Danegeld and Heregeld in the Early Eleventh Century’, EHR, civ (1989), 373–84; idem, ‘Chronicles and Coins as Evidence of Levels of Tribute and Taxation in Late Tenth- and Early Eleventh-Century England’, EHR, cv (1990), 939–50. 25 For recent arguments in favour of William reviving the tax, David Pratt, ‘Demesne Exemption from Royal Taxation in Anglo-Saxon and Anglo-Norman England’, EHR, cxxiii (2013), 1–34, at 8. 26 GG, 162–3; OV, ii, 192–5. 27 GG, 162–3; Williams, The English, 8, note 4; Baxter, The Earls of Mercia, 270–1; cf. Edward Impey, ‘London’s Early Castles and the Context of their Creation’, in Edward Impey (ed.), The White Tower (New Haven, CT, and London, 2008), 13–26, at 17–19. 28 OV, ii, 214–15 (Guillelmus etenim rex quando Eduinus comes cum eo concordiam fecerat . . . filiam suam se in coniugem ei daturum spoponderat); van Houts, ‘The Echo of the Conquest in the Latin Sources’, 142. 29 OV, ii, 194–5.

king of the english

267

as being a local perspective on a mass national submission, it may have created an impression on William and those around him that he could rely on loyalty in many of England’s regions, and even the more distant ones. If so, as we shall see, while such an impression does have some elements of reality, it was also to a large extent illusory. William probably also saw himself as restoring and repositioning the territorial earldoms of the Confessor’s rule during this period, while at the same time making strategic decisions about how this was best done.30 Waltheof, son of Earl Siward, was confirmed in several shires in the East Midlands (possibly Northamptonshire, Huntingdonshire, Cambridgeshire, and Bedfordshire) to which, as we have seen, he had been appointed in 1065. Copsig, who had been the deputy of Harold’s brother Tostig, was sent to Northumbria beyond the Tyne as earl.31 The East Anglian earldom of another of Harold’s brothers, Gyrth, was probably granted at this time to Ralph, who had migrated from Brittany to England in the Confessor’s reign, and who was possibly the son of a mixed Anglo-Breton marriage.32 The last two of these three appointments smack of haste and a rather ill-considered pragmatism. Ralph’s background and pre-1066 status certainly made him worthy of appointment as an earl. But he was an elderly man whose active career in Brittany can be traced back to almost forty years earlier; the grant of an earldom surely cannot have been thought of as anything other than a temporary solution. As for Copsig, because he was the close associate of Tostig, who had been thrown out of Northumbria just over a year earlier, and because he may well have fought against Edwin, Morcar, and numerous others of his new charges at Fulford Gate and been on the losing side at the Battle of Stamford Bridge, he must have been thought suspect by many. His appointment seemingly sidelined Morcar, who had been titular earl in the north, and his deputy Osulf, son of Eadulf of Bamburgh, who had been the actual power on the ground there. On the other hand, both Copsig and Ralph possessed the requisite status and may well have seemed the best people available at a time when decisions had to be taken quickly. According to Poitiers, William also endowed Edgar the Ætheling with considerable estates to compensate for his not having

30 See, in general, C. P. Lewis, ‘The Early Earls of Norman England’, ANS, xiii (1991), 207–23, at 215–18. 31 For Copsig, Kapelle, The Norman Conquest of the North, 106–8; LDE, 181, note 56; William M. Aird, St Cuthbert and the Normans: The Church of Durham, 1071–1153 (Woodbridge, 1998), 63–8; Richard Fletcher, Bloodfeud: Murder and Revenge in Anglo-Saxon England (London, 2002), 169–70; William M. Aird, ‘Copsi [Coxo], Earl of Northumbria (d.1067)’, ODNB. 32 Ann Williams, ‘Ralph the Staller, Earl of East Anglia (d.1068x1070)’, ODNB.

268

william the conqueror

become king. No trace of where these might have been has survived and Orderic omitted the passage.33 The earldoms of Wessex (or, more likely, parts of it) and Kent, both vacant because of the deaths of Harold and his brother Leofwine, were granted to William fitz Osbern and Bishop Odo of Bayeux. While respectively based at Winchester and Dover, they were manifestly endowed with wider responsibilities. Poitiers inserts a remarkable eulogy of fitz Osbern at this point, saying that William had loved him above all others since they had played together as boys. He added that fitz Osbern’s loyalty was a continuation of his father’s, a flashback to the events of William’s minority.34 If Poitiers was once more answering criticisms when he wrote this passage and also reflecting William’s thoughts, it foregrounds how much long-term loyalty mattered to the new king. It might even suggest that his trust was not easily given. The way in which these two appointments functioned is another commentary on the limits of William’s actual power. William fitz Osbern’s base at Winchester was identified by Poitiers as a suitable place to govern the kingdom to the north, as a place where neighbours were untrustworthy, and which might receive help from the Danes. Perplexing as it seems, the passage must be read as representing a perception of the scale of territorial control as Poitiers wanted to define it at this point; it makes sense if Winchester is seen as a base for military expeditions north of the Thames and from which to organize the defence of the south coast.35 Eustace of Boulogne’s appearance as an addressee in two writs relating in one case to Wiltshire and Gloucestershire and, in the other, to Surrey, completes the picture of a regime operating across southern England into which newcomers had been inserted at the top, but which appears to have been intended to be a variant on the old.36 The central role assigned immediately to individuals from Normandy and France in the lands over which William had meaningful control cannot have passed unnoticed. According to Poitiers, castles were built in London, and his reference to William’s need to impose his power there makes it clear that security was the major consideration.37 The castles referred to were located on the site of the White Tower and on the city walls to the west of St Paul’s, the latter GG, 162–3; Nicholas Hooper, ‘Edgar Ætheling (b.1052?, d. in or after 1125)’, ODNB. GG, 164–5. 35 For Poitiers’s Guenta as Winchester, and not Norwich, as Freeman believed, Frank Barlow, ‘Guenta’ (appendix to Martin Biddle, ‘Excavations at Winchester 1962–3’, Antiquaries Journal, xliv [1964], 188–219, at 217–19). 36 For Eustace, Regesta, nos. 223, 291. Eustace’s return to northern France by Easter 1067 supplies a terminus, Tanner, ‘Eustace II’, 272–3. 37 GG, 160–3. 33 34

king of the english

269

a huge enclosure transformed after William’s death into Montfichet’s Castle and Baynard’s Castle; it may well have been a fortification on the same scale as the Tower would later be.38 The future White Tower site commanded the river approach to the city and was located in the corner at which the surviving Roman walls turned northwards. Its positioning in relation to both city and river replicated the arrangements at Rouen.39 The symbolic messages drawn from the Roman past and from rule in Normandy would have been addressed both to William’s followers and to a wider audience that would appreciate the abstract reference to the long-term historical past. That William is said by Poitiers to have made laws for London’s benefit suggests that his famous writ for its citizens dates from this time.40 Surviving as an original, it confirmed to the citizens all the laws of which they were worthy in King Edward’s day and that every child should be his father’s heir (Plate 10). It shows that documents were continuing to be written in a traditional way in the English language and that William’s rule was being ideologically integrated into a continuous legal thread going back through and beyond Edward the Confessor’s 1041 promise.41 Edward’s former priest and chancellor Regenbald, who was rewarded with two of Harold’s former estates in Wiltshire, seemingly before William’s departure for Normandy in March, could have been a significant contributor to this process of articulating continuity.42 However, one of the surviving writs in Regenbald’s favour refers to Harold as ‘king’, showing that, whatever the scale of the intellectual preparations for the Conquest had been, the full logic of William’s proclaimed status as the Confessor’s heir had not yet been fully grasped. The grant made by William to Wulfstan, bishop of Worcester, who had submitted at Berkhamsted, can be viewed both as a sign that William’s authority was being widely acknowledged across his new kingdom and that he was using his generosity to consolidate links with chosen individuals. 38 For the early London castles, Derek Keene, ‘From Conquest to Capital: St Paul’s c.1100–1300’, in Derek Keene, Arthur Burns, and Andrew Saint (eds.), St Paul’s: The Cathedral Church of London, 604–2004 (New Haven, CT, and London, 2004), 17–32, at 18–20; Impey, ‘London’s Early Castles’, 16–26. 39 Bernard Gauthiez, ‘Hypothèses sur la fortification de Rouen au onzième siècle: le Donjon, la Tour de Richard II, et l’enceinte de Guillaume’, ANS, xiv (1992), 61–76, at 63–8; idem, ‘The Urban Development of Rouen, 989–1345’, in Hicks and Brenner (eds.), Society and Culture in Medieval Rouen, 17–64, at 21–3. 40 GG, 158–9. 41 Regesta, no. 180; B&C, plate XIV; Maddicott, ‘Edward the Confessor’s Return’, passim. 42 Regesta, no. 223; S. D. Keynes, ‘Regenbald the Chancellor (sic)’, ANS, x (1988), 185–222, at 211–13; Garnett, Conquered England, 10–11, with some comments on the possibly corrected version, Regesta, no. 224.

270

william the conqueror

The surviving form of the diploma recording the transaction, albeit only a cartulary copy, suggests that Wulfstan obtained a second confirmation in 1068 after Normans had been appointed to local and regional offices in the West Midlands.43 Although, according to Malmesbury, Wulfstan had been actively supportive of Harold, William, so he said, never caused him any problems.44 Wulfstan possibly appealed to William because he was a notably holy man, although Archbishop Ealdred of York’s support for Wulfstan must also have been a factor working in his favour. More generally, the way Wulftsan was treated probably reflects William’s respect for the authority of bishops and his belief that they were central to a kingdom’s well-being, an attitude already demonstrated in Normandy and fully in accord with the English traditions that Ealdred and others would have been setting before him.45 It is nonetheless a commentary on the instability of the times that Wulfstan deemed it politic to make such regular reference to William’s authority that he obtained a second confirmation so soon after the first one. The confirmation in office of Brand, abbot of Peterborough, also took place during the short period before William left for Normandy. His predecessor Leofric had been with Harold’s army at Hastings, and had died a fortnight afterwards. Brand had then secured agreement to his appointment from Edgar the Ætheling when the former believed that Edger was going to be king. His conscientiousness in seeking William’s early confirmation is, like Wulfstan’s diploma, another sign of just how much kingly confirmation mattered in these uncertain times. As such, and as appears to have been the case with Wulfstan, it illuminates the powerful forces that were working in William’s favour; whoever you were, only kingship could legitimate authority at the level of a bishop or an abbot. However, there is another perspective in the narratives written later at Peterborough, including insertions made into the ‘E’ Chronicle after it started to be written there post-1116, that a negotiation had to be conducted through intermediaries who placated William’s anger by paying the huge sum of forty marks of gold (£240) in order to guarantee the estates Brand had granted to the monastery. The surviving charter naming the estates, including among its signa the diocesan, Wulfwig, bishop of Dorchester-on-Thames, who died in December 1067, dates the transaction to this early period.46 The large payment extorted shows that the processes associated with arbitrary royal Regesta, no. 345. Vita Wlstani, 60–1. 45 The Political Writings of Archbishop Wulfstan of York, 37–8, with references to other important work. 46 For Brand, ASC, ‘E’, 1066; The Chronicle of Hugh Candidus, a Monk of Peterborough, ed. W. T. Mellows (London, 1949), 76–7; Regesta, no. 216. See also The Anglo-Saxon Chronicle. A Collaborative Edition: volume 7, ms. E., xcvii. 43 44

king of the english

271

anger were in place from the very beginning. A display was being put on to parade Abbot Brand as having stepped out of line and an example was being made of him. The large payment exacted must have struck many as an ominous portent. If someone so obviously ready to acknowledge the new regime was treated like this, what could others expect? It is hard not to think of Brand as an over-conscientious unfortunate and, as such, as supplying a remarkable illustration of the uncertainties of the times. Specimens of William’s first issue of coinage, generally assumed to date from 1066 to c.1068, survive from forty mints, that is from eight fewer than Harold’s single coinage. But, as for Harold, the list of surviving coins includes pieces from all of the great distant centres of the kingdom, such as Exeter, Gloucester, Worcester, Norwich, and York, and there is considerable continuity in the names of the moneyers who struck the coins.47 Overall, just as after Harold’s accession, the kingdom’s organizational structure showed a remarkable capacity to adjust to dramatic change at the top; the continuation of what can pass as normality also meant that the coins resemble closely the single issue in Harold’s name, indicating that the new dies were rapidly dispatched to the provinces from London (Plate 11 (a), (b) and (c)).48 As a result of this issue of coinage, the medium through which most people in England encountered the new king’s image in their daily lives was in circulation almost immediately. That it was so successfully accomplished may also reflect on the taking of the geld mentioned by the ‘D’ Chronicle. The tentacular reach of the administration of the kingdom’s money may well demonstrate that the instructions to pay the tax would also have travelled to distant regions. The reference in the London writ’s address clause to the French and the English (frencisce > englisce) as its recipients should probably be interpreted as an English draftsman’s response to a situation in which a multiplicity of peoples were now responsible for upholding the law.49 Its only occasional 47 Martin Allen, ‘Mints and Moneyers of England and Wales, 1066–1158’, British Numismatic Journal, lxxxii (2012), 54–120, at 57–9; idem, ‘Coinage and Currency under William I and William II’, in Rory Naismith, Martin Allen, and Elina Screen (eds.), Early Medieval Monetary History: Studies in Memory of Mark Blackburn (Farnham, 2014), 85–112, at 86, 90–4. I am grateful to Martin Allen for the information that a coin from the Ipswich mint, the only major mint seemingly unrepresented, has now been discovered. 48 Most recently, Marion Archibald, ‘The German Connection: German Influences on the Later Anglo-Saxon and Norman Coinages in their English Context (10th and 11th centuries)’, in Rainer Polley and Andreas Röpcke (eds.), Fundamenta Historiae. Geschichte im Spiegel der Numismatik und ihrer Nachbarwissenschaften. Festschrift für Niklot Klüßendorf zum 60. Geburtstag am 10. Februar 2004 (Hanover, 2004), 131–50, at 143. 49 Most recently, Richard Sharpe, ‘Peoples and Languages in Eleventh- and TwelfthCentury Britain and Ireland: Reading the Charter Evidence’, in Dauvit Broun et al., The Reality behind Charter Diplomatic in Anglo-Norman Britain (Glasgow, 2011), 1–119, at 3–16. Also Thomas, The English and the Normans, 33–4; Garnett, Conquered England, 12–13.

272

william the conqueror

use in similar documents during William’s lifetime probably makes its invention an instinctively traditional one that was in no way legally motivated. Its reference point may well have been the supplementary oath that William had taken to treat peoples equally; its appearance in a writ obtained by Archbishop Ealdred suggests that this was so.50 The surviving seal, although badly damaged and no longer attached to the writ – one of only six authentic specimens of William’s seal still in existence, albeit in most cases fragmentary and damaged – is a demonstration that a seal had been designed for the new king within days of his coronation.51 A double-sided seal that portrays William on horseback in armour and carrying a lance and a shield on its obverse side and as a king enthroned in majesty on its reverse, it is the first such equestrian seal to survive from the medieval West and – in all probability – the first one ever created (Plate 12). It is therefore notably innovatory, a statement of the nature of William’s kingship of the most important kind.52 The seal’s legend, beginning on the obverse side and running onto the reverse is + HOC NORMANNORVM VVILLELMUM NOSCE PATRONUM S/I/ HOC ANGLIS REGEM SIGNO FATEARIS EVNDEM (‘With this seal recognize William, patron of the Normans; as with this you acknowledge him as king of the English’).53 Such a legend, running on two sides of the seal, written in two hexameters, represents a continuation of a tradition going back to Carolingian times that includes the seal of Charles the Bald (840–77). In relation to influences likely to have been operational in 1066, a legend running across both sides appears on the seal of Emperor Conrad II and, probably even more relevantly, on those of Popes Nicholas II and Alexander II.54 The way in which the hexameter gives precedence to the Norman side of William’s dignity, and the use in the legend of the unusual word patronus, are arguably both very significant ideological statements. Although patronum Regesta, no. 31. For the documents bearing seals, all of which are now fragments, Regesta, 102–5; nos. 68, 144, 180, 254, 326, 334. For a drawing, now lost, based on a complete specimen in the archives of the abbey of Saint-Etienne of Caen that gives the complete legend, A. L. Léchaudé d’Anisy, Recueil des sceaux normands et anglo-normands, précédé de l’extrait du cartulaire des chartes, diplômes et autres actes qui existent encore dans les archives du Calvados ou d’autres fonds (Caen and London, 1834), plate 1. 52 On its being the earliest such seal, see Jean-François Nieus, ‘Early Aristocratic Seals: An Anglo-Norman Success Story’, ANS, xxxviii (2016), 97–123, at 100–3. I am very grateful to Jean-Luc Chassel and Sandy Heslop for sharing thoughts and unpublished work with me on William’s seal, and to Anna Gannon and Benjamin Pohl for discussions. 53 For the translation, Barlow, William Rufus, 59. 54 Toni Diederich, ‘Zum Gebrauch lateinischer Verskunst in Siegelumschriften’, in idem, Siegelkunde, Beiträge zu ihrer Vertiefung und Weiter führing (Vienna, Cologne, and Weimar, 2012), 146–77, at 154–5. 50 51

king of the english

273

might have been chosen to fit the metre, as ducem, for example, would not, and although the word had acquired a multiplicity of meanings since classical times, the usage in the sense of ‘protector’ implies a general protective responsibility that is greater than that of a mere ruler. It creates an inclusive status that bestrides and unites the distinct cross-Channel territories of Normandy and England under the umbrella of a responsibility for a people. In this sense, it may mirror a constant theme of Poitiers’s Gesta of William as the war leader who cared for the welfare of his soldiers. It also has the sense of someone who is the founder and protector of churches and, as such, was sometimes used in relation to Emperor Constantine, themes that were already present within the historiography of the Norman rulers since Dudo of Saint-Quentin had deployed them in relation to Rollo.55 The ideology associated with the seal apparently remained dynamic throughout William’s life, since the title patronus Normannorum appears in two charters for the abbey of Fécamp; one dates from the year 1085 and was apparently drafted within the abbey in advance of a second version written by a scribe of William’s chancery, who reproduced the formula verbatim.56 The equestrian side must surely reflect the image of God’s chosen warrior, of membership of the militia Christi, a representation made all the more likely by the probable papal influence on the seal. Byzantine influences have been detected in the representation and it is notable that the Laudes sung on religious feast days in Normandy feature the three Byzantine warrior-saints, Maurice, George, and Sebastian.57 Given its early creation, this side of the seal can also be seen as an exemplification of the duties conferred by the papal banner. The reverse side of William’s seal is an obvious adaption of Edward the Confessor’s seal, itself profoundly influenced by imperial traditions, both Western and Byzantine.58 The only significant differences between William’s seal and Edward’s are that William’s is larger and that the sword is more prominently displayed. Two final comments on this remarkable seal are that a similar double-sided Pohl, Dudo of Saint-Quentin’s ‘Historia Normannorum’, 199–214. For the charters, Regesta, nos. 141, 144. For the second of these and the two surviving originals, Pierre Chaplais, ‘Une charte originale de Guillaume le Conquérant pour l’abbaye de Fécamp: la donation de Steyning et de Bury (1085)’, in L’abbaye bénédictine de Fécamp, i, 93–104, 335–7 (reprinted in idem, Essays in Medieval Diplomacy and Administration [London, 1981], chapter 16). 57 H. E. J. Cowdrey, ‘The Anglo-Norman Laudes Regiae’, Viator, xii (1981), 37–78, at 68, 73, 77 (reprinted in idem, Popes, Monks and Crusaders [London, 1984], chapter 8). 58 Brigitte Bedos-Rézak, ‘The King Enthroned, a New Theme in Anglo-Saxon Royal Iconography: The Seal of Edward the Confessor and its Political Implications’, in Joel T. Rosenthal (ed.), Kings and Kingship, Acta, xi (New York, 1986), 53–88, at 59–66 (reprinted in idem, Form and Order in Medieval France [Aldershot, 1993], chapter 4). 55 56

274

william the conqueror

equestrian seal was created for William’s half-brother Bishop Odo, and that the first two rulers known to have adopted similar seals are Robert the Frisian, count of Flanders, and Cnut IV, King of Denmark (1080–6).59 Odo’s seal surely illustrates the sense of partnership shared by the ruling family and also their unique standing in their own eyes. The seals of Robert and Cnut either demonstrate that imitation is the sincerest form of flattery or are examples of clever and possibly co-ordinated sarcasm, since these two men were among William’s most determined enemies. With the seal’s themes well known among both the English and Norman intellectuals who would have been gathered around William during these weeks, it is ultimately impossible to decide where the inspiration for the seal came from. There is actually no reason why it should not have emerged out of discussions taking place during these frantic weeks. Yet because the seal was such a crucial authenticating element in kingly rule, it is surely certain that the message conveyed would have been checked with William. On these terms, it becomes an insight into his thoughts about the responsibilities of power. If this is so, it is the Norman heritage that takes precedence, a prevailing sense of himself as the protector of the people he had already ruled for over thirty-one years. The attitudes made explicit by the seal are also expressed in the famous murdrum fine that set a high, discriminatory, penalty to be imposed on anyone who killed a Norman or a Frenchman. As formulated in a record that may date from William’s reign, a murderer’s lord would have to pay a penalty of forty-six shillings of silver if the killer had not been caught within five days, a fine so heavy that provision was also made for the hundred in which the murder had taken place to contribute the outstanding balance if the lord could not pay.60 In the second half of the twelfth century, Richard fitz Nigel in the Dialogus de Scaccario attributed the fine’s creation to guerrilla activities by the English and the secret slaying of the newcomers.61 Although discussions of the origins and development of the fine are complicated by later legal changes and by the possibility of an origin under Cnut, whether it is an innovation or a revival the crucial point has to be its integration into the framework of William as patronus. A relatively early date for its creation is also likely because these first years of his reign were so disturbed.62 Nieus, ‘Early Aristocratic Seals’, 103–5. Ten Articles of William I, in Die Gesetze der Angelsachsen, ed. Liebermann, i, 486–8, art. 3. 61 Richard fitz Nigel, Dialogus de Scaccario: The Dialogue of the Exchequer, and Constitutio Domus Regis, The Disposition of the King’s Household, ed. and trans. Emilie Amt and S. D. Church, OMT (Oxford, 2007), 80. 62 For a full discussion of modern assessments, Hudson, The Oxford History of the Laws of England, 405–9. See further George Garnett, ‘ “Franci et Angli”: The Legal Distinctions 59 60

king of the english

275

Another commentary on these times is that, according to William of Malmesbury, Bishop Wulfstan of Worcester was in the habit of preaching to his flock that the defeat at Hastings was God’s judgment on the sinful English and that they should therefore submit to William. A classic statement in a direct ideological line from the writings of Wulfstan’s eponymous early eleventh-century predecessor, the archbishop of York who had dealt with Cnut, it would seem to indicate the presence of powerful currents driving the English to submission.63 Its influence may well be articulated in the last statement in the surviving text of the Gesta Guillelmi, which records the importance of the support that several bishops had given to William.64 If so, a sense of reciprocal responsibility founded in an awareness of the duties of their respective offices of king and bishop could have been involved here, since, as we shall see, several diplomas confirming grants to bishops survive from this early period, a sign that they were people who, albeit selectively, William was seeking to cultivate. It also suggests that a dialogue deeply entrenched within the traditions of the English kingdom was a part of the attempted pacification taking place. THE TRIUMPHANT RETURN TO NORMANDY Imperatives connected with the need to reassure and thank those in Normandy who had supported the invasion of England without crossing the Channel, and to show off William’s triumph to Normandy’s neighbours, were probably involved in the decision to return to Normandy so soon after his coronation. A wish to see Matilda and perhaps to celebrate the birth of their newborn daughter Adela may also have been a reason. Although the exact date of Adela’s birth is, as always, not entirely clear, it is as good as certain that Matilda was pregnant when William set off to invade England, since Adela was certainly born during 1067. Poitiers presents the visit as a triumphant homecoming and has William greeted in Rouen by cheering crowds.65 The Channel crossing must also have been connected with paying off soldiers who decided not to stay in England. Whatever the reason for the early return to Normandy, however, it may have been of crucial significance to the unfolding of events for unintended between Peoples after the Conquest’, ANS, viii (1986), 109–37, at 116–28; Bruce O’Brien, ‘From morðor to murdrum: The Pre-Conquest Origin and Norman Revival of the Murder Fine’, Speculum, lxxi (1996), 321–57, at 331–8, 350–1. 63 GG, 58–9; The Political Writings of Archbishop Wulfstan of York, 41–4, 131 (‘If it comes about that a great misfortune befalls the realm because of the people’s deeds . . . then amends is to be made to God himself ’); Giandrea, Episcopal Culture in Late Anglo-Saxon England, 35–69. 64 GG, 186–7. 65 Ibid., 176–7.

276

william the conqueror

reasons. Some in England may well have drawn the conclusion that their new king was going to be an absentee and set their sights accordingly. Others, notably the English leaders taken by William to Normandy in his entourage, may well have felt dishonoured and disenfranchised by their treatment. Poitiers foregrounds the uncertainty of the times and the political tightrope William was walking by acknowledging that the rationale behind taking so many of the English leaders to Normandy was to prevent them from fomenting rebellion in his absence.66 Those taken included Archbishop Stigand, Edgar the Ætheling, Earls Edwin, Morcar, and Waltheof, Æthelnoth, abbot of Glastonbury, the wealthiest monastery in England, and Æthelnoth of Canterbury. Instead of trying to whitewash the decision as an act to honour William’s new English subjects, as he could easily have done, Poitiers says that they were being treated ‘as if they were hostages (uelut obsides)’, that is, as sureties against their kinsmen’s and compatriots’ good behaviour. Since all seem to have returned to England with William later in the year and to have lived freely afterwards, they turned out not to be hostages in the technical meaning of the term. But Poitiers, who used language carefully and is in this case followed by Orderic, was presumably indicating the possibility that they might have been retained as prisoners.67 In terms of the aforementioned tightrope, what was presumably foremost in Poitiers’s mind was William’s need to reassure his followers that he was protecting their interests by keeping the most likely leaders of revolt under guard. But, from an early twelfth-century perspective, the author of the Brevis Relatio, who was trying to write favourably about William, commented on the situation by saying simply that William did not trust the English.68 This wholesale denial to so many of England’s topmost elite of their accustomed roles in the kingdom must surely have profoundly influenced their future behaviour. Although they might have grudgingly understood the reasoning behind the new king’s decision, their treatment must surely also have been the basis of festering resentment. But to William’s associates his conduct may well have seemed appropriately circumspect. Poitiers presents the court held at Fécamp at Easter (8 April 1067) as a splendid occasion, attended not only by Normans and William’s English ‘guests’, but by many French, including Ralph, count of Amiens, Valois and the Vexin.69 Both as a religious occasion and an event at which the new For the names, ASC, ‘D’, 1066; GG, 166–9; JW, iii, 4–5. Kosto, Hostages, 25–6. 68 Brevis Relatio, 33. 69 GG, 178–81; OV, ii, 196–7. 66 67

king of the english

277

king’s wealth and status were on display, Poitiers’s statements, all reproduced by Orderic, emphasize a magnificence of a kind that the spectators had not previously seen; William himself was dressed in garments woven from, and encrusted with, gold. To be a king was different from being a duke, or so both of them wanted to say. The choice of Fécamp as the venue was presumably based on its long history as a ducal palace and because it was the burial church of William’s grandfather and greatgrandfather. The festivities must surely also be seen as thanksgiving offerings for divine favour, as already noted, thanks that by this stage may well have included those for the successful birth of another healthy child. A poem written by Geoffrey of Rheims when Adela was betrothed at around fourteen years old, c.1081, suggests this; the child’s birth was presented there as the reason why William had become a king, not wealth, or victory in battle, or any such thing.70 Literary extravagance this undoubtedly was, but the poet evidently thought that the drama he invested in this healthy birth represented some kind of reality and a reason for William and Matilda to celebrate. William’s visit to Le Vaudreuil to the south of Rouen in April, his presence at the dedications of the new abbey churches at Saint-Pierre-surDives and at Jumièges (Plate 13) on 1 May and 1 July respectively, and a visit to Lyons-la-Forêt on 24 June, indicate that he moved south to Rouen immediately after Easter, then turned away to central and western Normandy for a visit that almost certainly took in Caen, before returning to Upper Normandy and the Seine valley for the summer.71 The church dedications must be seen as further statements of thanksgiving for divine favour. The essentially local witness-lists to the charters confirmed at Le Vaudreuil and Lyons-la-Forêt suggest that William travelled accompanied by Matilda and a household, and that their two eldest sons, Robert and Richard, were present at Le Vaudreuil. None of the English appear among the signa and of the most powerful members of the aristocratic elite who had conquered England, only the Breton Alan, the son of Count Eudo of Penthièvre, appears in any of the charters.72 However, they may well not be a good guide to the party that travelled with William. William of Malmesbury believed that Archbishop Stigand was warmly welcomed in every diocese and abbey in Normandy he visited, suggesting that the English were indeed kept with William and shown respect. However, Malmesbury’s added comment to the effect that William was devious and 70 Van Houts, ‘The Echo of the Conquest in the Latin Sources’, 145–6 (Quid causa fuerit? Filia causa fuit). 71 For this itinerary, GND, ii, 172–3; OV, ii, 198–9; Regesta, nos. 196, 251; Les Annales de SaintPierre de Jumièges, 57. 72 Regesta, no. 243.

278

william the conqueror

insincere may offer a different perspective.73 It picks up the thread of fundamental distrust that we have already seen expressed by the author of the Brevis Relatio. Poitiers writes of William’s extensive religious benefactions during this time. His gifts to Saint-Etienne of Caen were said by Poitiers to have been particularly lavish and to have included embroideries and other richly worked objects from England.74 In stating at this point that William gave only from his own property and that compensation was given when objects were removed from English churches, Poitiers is once more seeking to defend him, this time against the charge of looting; Orderic’s omission of the passage is a telling one.75 Given that Saint-Etienne was still in an early stage of development, these gifts must represent an important point in its history. It is actually possible that the decision to build the church on such a large and original scale was taken at this very time. Although the part of the building that would have been constructed first, the choir, was replaced in the late twelfth century, it is generally considered to have had an apsidal east end that, while being on a large scale, was of a relatively traditional kind. The grand elevation of the transepts and nave and the extraordinary tribune passages are, however, indicative of great ambition and convey a remarkable sense of unity in the church’s design.76 If, as seems likely, William had already decided that he was going to be buried at SaintEtienne, then the church as constructed becomes not just an emblem of victory but also a statement of personal achievement (Plate 14).77 Its development was to be a crucial element in determining William’s itinerary as a cross-Channel ruler, with the crossing from Caen or Ouistreham to Portsmouth or Southampton, and then on to Winchester, being the route of choice. GP, i, 48–9. GG, 176–7. 75 Ibid., 178–9; OV, ii, 198–9. On English works of art transferred to Normandy and elsewhere in France, C. R. Dodwell, Anglo-Saxon Art: A New Perspective (Manchester, 1982), 216–19; David N. Dumville, ‘Anglo-Saxon Books: Treasure in Norman Hands?’, ANS, xvi (1994), 83–99. 76 Maylis Baylé, ‘Caen: Abbatiale Saint-Étienne (Abbaye-aux-Hommes)’, in Baylé and Bouet (eds.), L’architecture normande au Moyen Age, ii, 56–61; Eric Fernie, The Architecture of Norman England (Oxford, 2000), 101–2. The suggested chronology of Saint-Etienne’s construction goes back to Carlson, The Abbey Church of Saint-Etienne at Caen, 162–335; idem, ‘Excavations at Saint Etienne, Caen (1969)’, Gesta, x (1971), 23–30. For the rebuilding of the choir, Lindy Grant, Architecture and Society in Normandy 1120–1270 (New Haven, CT, and London, 2005), 97–106. 77 For Saint-Etienne being intended as a mausoleum from the beginning, Carlson, The Abbey Church of Saint-Etienne at Caen, 237. The site of the original tomb is uncertain, but Orderic describes it as being inter chorum et altare as he also described Matilda’s tomb, OV, ii, 104–5. 73 74

king of the english

279

It must have been during this period that the Norman bishops drew up the document now known as the Penitential Ordinance, fixing a range of graduated penances for killing and acts of violence during the recent and ongoing campaigns. Its confirmation by the papal legate Ermenfrid of Sion, who was almost certainly in Normandy late in the year, was a logical corollary to of the grant of the banner and represents a continuation of the collaboration between William and the papacy.78 Although the mention of the Norman bishops indicates that the Penitential Ordinance was initially drafted for the troops who were being disbanded, its subsequent deployment in England is revealed by the fact that the three surviving manuscripts are preserved there. Like so much else about the Hastings campaign, it follows patterns long traditional in medieval western Europe. The issue of killing in a supposedly legitimate cause had been the subject of much thought and of differing opinions. However, before Hastings, actual post-battle penitentials survive only from the Battles of Fontenoy (843) and Soissons (923), with all three of them being the product of battles that involved conflicts between consecrated kings. The resemblances between the Soissons and Hastings texts indicate how much the Hastings one lies with in the European mainstream. An interest in penance on the part of Archbishop John of Rouen (1067–79), recorded in a manuscript now in the Bodleian Library (ms. Barlow 37, fos. 50rv), reinforces the same conclusion.79 Nothing is known of the Penitential Ordinance’s implementation and the possibility that it is simple window-dressing cannot be ruled out; it would certainly have involved a long-term process on both sides of the Channel. In fixing penances for those who killed opponents who continued to resist William after his coronation, it catered for the prospect of continuing hostilities. It also draws attention to potentially contradictory values of reward and equity that had been present throughout the whole enterprise. While Poitiers might stress the importance of William’s reputation for generosity, the ordinance differentiated between those whose motivation was founded in the justice of the cause and those who participated for gain, with the latter to be treated as murderers. Making the distinction was, however, presumably usually a matter for an individual’s conscience.80 Business of a kind that may well have been a regular feature of this period appears in a charter for the abbey of La Trinité-du-Mont of Rouen, 78 Alexander II’s letter about the Rouen appointment says that he had received information from Ermenfrid, implying that he had been in Normandy, AAR, 40, 54 (conperimus Sedunensis episcopi et Lanfranci abbatis relatione). 79 For the text, CS, 582–4. The main modern discussions are Cowdrey, ‘Penitential Ordinance’, 234–8, 241–2; Hamilton, The Practice of Penance, 87–8, 194–6. 80 CS, 582–4 (c.5 and c.9 for reward and for killings after the coronation).

280

william the conqueror

with William confirming a grant that someone who had died on the 1066 campaign had intended to make, but had been prevented by death from doing.81 The long-term theme of frontier consolidation that had been a constant element of William’s life is also present. The charter for the abbey of Saint-Benoît-sur-Loire confirmed at Le Vaudreuil compensated the abbey for the detrimental impact of the construction of a castle on its lands just within the duchy’s south-western frontier at Saint-Jamesde-Beuvron during the 1064 campaign into Brittany. It is, however, also indicative of a continuing campaign on William’s part to secure spiritual benefits. In return for the grants, the monks agreed to say masses for the souls of William’s close family, his ancestors and his successors, to maintain a monk in the abbey as a surrogate for William, to feed a pauper, and to say prayers for William during his lifetime and after his death.82 After the apparently sudden death of Archbishop Maurilius of Rouen in the summer of 1067, the Rouen clergy chose Lanfranc as their next archbishop, a promotion said to have been unacceptable to Lanfranc, and perhaps (although Orderic says the opposite) to William as well if he already had Canterbury in mind for the abbot of Saint-Etienne of Caen.83 The decision to transfer Bishop John of Avranches to Rouen instead, which emerged out of the undoubtedly complicated discussions that ensued, was secured by Lanfranc travelling to Rome and by a visit to Normandy by the papal legate, Bishop Ermenfrid of Sion. Alexander II’s letter preserved in the Acta Archiepiscopum Rotomagensium specifically says that John had been elected by William, another illustration of a continuing spirit of co-operation; the legal norm was of course for the cathedral clergy to be involved too. However, William’s respect for proper canonical procedure is still evident, since the transfer of a bishop from one diocese to another required papal agreement.84 These journeys also produced a papal privilege in favour of SaintEtienne of Caen. Dated 14 January 1068, it placed the abbey under the special protection of the papacy, while keeping responsibility for its supervision within the structures of the metropolitan province of Rouen and, at a pastoral level, the bishop of Bayeux.85 Another landmark in SaintEtienne’s development, it shows a monastery that was functioning as a community, as well as illustrating yet again the level of trust that evidently Regesta, no. 231. Ibid., no. 251. 83 For the conflicting evidence for the date of Maurilius’s death, Spear, Personnel, 197. 84 The main sources for this episode are AAR, 40–1, 54; OV, ii, 200–1; ‘Vita Lanfranci’, 682. See further Cowdrey, Lanfranc, 37–8; Allen, ‘ “A Proud and Headstrong Man” ’, 14–15. 85 PL, cxlvi, cols. 1,339–41; Lemarignier, Etudes, 144, note 43. For discussion, Lemarignier, Etudes, 143–4, 16–76; Cowdrey, Lanfranc, 27–8. 81 82

king of the english

281

existed between William, Alexander, Lanfranc, and – in all probability – the diocesan, Bishop Odo. It is also an episode which shows that Normandy and Caen specifically remained in the forefront of William’s thoughts, no matter what was happening on the other side of the Channel. The death on 13 December 1067 of Richard, count of Evreux, one of the most powerful members of the Norman aristocracy, exactly a week after William had left Normandy for England, illustrates the problems that rule over vastly expanded territories now posed. While Richard, the successor in 1037 to the lands of his father, Archbishop Robert of Rouen, had apparently played no part worthy of mention in the politics or government of Normandy during a long life, he must have performed his responsibilities on a difficult frontier competently. His loyalty to William seems a certainty and he had made a contribution of ships in 1066. Richard’s son William, who had fought at Hastings and who, since he lived on until 1118, must have been young in 1066, succeeded him. This major shift within Normandy’s aristocracy, with potential implications for the equilibrium of the frontier in a sensitive area, apparently passed off smoothly. Count William was to play a major role in William’s military campaigns in Normandy, but he never received the rewards in terms of land in England that might have been expected for someone of his status. His life history brings home another facet of the complexities of the new cross-Channel world that had come into being.86 THE RETURN TO ENGLAND William was undoubtedly receiving reports of turbulence in England while he was in Normandy. Orderic mentioned that he dispatched troops across the Channel and that their ships were blown off course and forced to take shelter at Exeter, where the men they had transported were rough-handled; this expedition certainly happened before the siege of early 1068.87 The accounts in the three main sources of Odo and William fitz Osbern’s stewardships in England during William’s absence are consistent with their general treatment of this whole period; Poitiers saw them and their associates as praiseworthily preserving the peace, while Orderic and the ‘D’ Chronicle portrayed their actions as oppressive and inflammatory. The Chronicle mentions extensive castle-building.88 A document dealing with English affairs that certainly dates to the period of William’s absence, a writ in Old English, apparently in his and Bauduin, La première Normandie, 332–4. OV, ii, 212–15. 88 GG, 180–3; OV, ii, 202–3; ASC, ‘D’, 1066 (recte 1067). 86 87

282

william the conqueror

William fitz Osbern’s joint names, confirming Charlcombe (Somerset) to the abbey of Bath, might be thought to suggest calm conditions and normal communication within Wessex. But it must surely be treated as evidence of nothing more than another individual churchman coming to terms with William.89 The document’s formulation in the two men’s names is an important indication of how what was effectively a partnership in rule was already being deployed to sustain cross-Channel rule. The king’s ultimate authority was paraded, as indeed was the temporary (that is, revocable) nature of the arrangement, but fitz Osbern’s capacity to act for William is also clear; exactly, that is, the situation described by William of Poitiers.90 What a document of this kind does not do, however, is illuminate the personal relationships that underpinned it. The levels of trust and loyalty involved were huge. Their resilience and the times when they broke down are major themes of the next twenty years. A jockeying for position, with William’s absence perhaps fuelling expectations that much was up for grabs, probably explains many of the events of this period. In addition to the events at Exeter, which persuaded William that he had to set out to subdue the city immediately after his return to England, the narrative sources record outbreaks of violence in the south-east, the far north, and the West Midlands. The devastation of parts of Herefordshire by Eadric the Wild, assisted by Bleddyn ap Cynfyn, the ruler of Gwynedd after Harold’s destruction of Gruffudd ap Llewelyn, suggests that, his possible submission at Barking notwithstanding, Eadric was responding to political uncertainty by trying to increase his local power because of a quarrel with the castellan of Hereford, who was one of the Frenchmen installed by Edward the Confessor. Eadric’s collaboration with Bleddyn is also a continuation of pre-1066 behaviour, an alliance for which Bleddyn would presumably have expected to be paid with the booty that John of Worcester says was taken.91 That Eadric went unpunished suggests an uncertain world in which everyone, including William, was trying to find their feet. Apparently, according to the values of these unusual times, Eadric’s conduct was deemed excusable.

Regesta, no. 11, where the complexities of the document are discussed. GG, 164–5 (ut in uice sua interim toti regno Aquilonem uersus praeesset). See also the early twelfth-century tract known as Quaedam exceptiones, GND, ii, 304 (. . . qui totius Anglie et Normannie dominatum sub rege Willelmo habebat). 91 The main source is JW, iii, 4–5. See further Susan Reynolds, ‘Eadric silvaticus and the English Resistance’, BIHR, liv (1981), 102–5; K. L. Maund, ‘The Welsh Alliances of Earl Ælfgar of Mercia and his Family in the mid-Eleventh Century’, ANS, xi (1989), 181–90; idem, The Welsh Kings (Stroud, 2002), 71–2; Williams, The English, 14–15; idem, ‘Eadric [Edric] the Wild [Eadric Cild] (fl.1067–1072)’, ODNB. 89 90

king of the english

283

William’s appointee in the far north, Copsig, was killed north of the Tyne on 12 March 1067, having survived in office for less than five weeks. His killer, Osulf, son of Eadulf of Bamburgh, Morcar’s former deputy, took over from him and lasted until the autumn before he too met a violent end. Copsig’s close association with Tostig must surely always have made his appointment problematic.92 On the other hand, given William’s limited knowledge of people and of regions that were far beyond the lands where he could aspire to have any direct control, Copsig might have looked like a reasonable option, especially since Osulf ’s and later appointments seem to indicate that William was for a long time ready to give this particular job to anyone who would take it on. Also during William’s absence Eustace of Boulogne attacked Dover.93 Given the tensions that possibly existed between him and William, the origins of this attack may lie deep in their respective pasts, or it may have been associated with whatever agreement they had reached in 1066; possibly Eustace had some expectations as a result of his dissolved marriage to the Confessor’s sister Godgifu, who was still living in England. His presence, already noted, in the address clauses of writs dealing with Wiltshire, Gloucestershire, and Surrey could indicate that a role similar to those of William fitz Osbern and Bishop Odo may well have been set up for him. But, if this was the case, he over-reached himself. The surviving accounts suggest that the English in Dover tried to submit to Eustace as their lord and that he then attempted to impose himself there by force with an attack involving a maritime expedition, launched presumably from Boulogne.94 While Eustace’s conduct in 1051 may have left a bad memory at Dover, he was at least the devil the townsmen knew; with the short Channel crossing to Boulogne in mind, he may well also have seemed a plausible protector in the midst of the new uncertainties. However, the attack was easily repulsed by William’s garrison and Eustace was sentenced to banishment at the 1067 Christmas court, a sentence approved, so Poitiers says, by William’s English and French subjects. His conduct draws attention to the opening fault lines within the network of alliances that had achieved the victory of 1066. That Eustace was later restored to his lands in England, in all probability as early as 1070–1, shows that his control of

92 See especially Kapelle, Norman Conquest of the North, 106–8. On the other hand, Aird, St Cuthbert and the Normans, 67, describes his appointment as ‘a logical expedient’. 93 GND, ii, 176–9; GG, 182–5; OV, ii, 204–10. On the episode, see Tanner, ‘Counts of Boulogne’, 272–4; idem, Families, Friends and Allies, 101–2. 94 Note . . . quorumdam Anglorum intra Cantuariensem comitatum consistentium uersutiis deprauuatus, Dorobernie castellum inuolare est nisus, GND, ii, 176–7, and Si erat seruiendum non compatriotae, noto seruire atque uicino satius putabant, GG, 182–3.

284

william the conqueror

the French side of the narrowest sea crossing to England made him too valuable an ally to be dismissed.95 Crossing the Channel from Dieppe to Winchelsea on the night of 6 December 1067, William again left Matilda and their eldest son Robert to act on his behalf in Normandy; Matilda was probably in the early stages of another pregnancy, this time with the future King Henry I. William is said by Orderic to have celebrated Christmas in London with a substantial group of laity and clergy in attendance. There, so Orderic tells us, he welcomed the English very warmly and sought to instruct his followers to behave similarly, since kindness was often the best way to persuade the hesitant and uncommitted.96 With the surviving text of Poitiers’s Gesta Guillelmi coming to an end after its account of William’s visit to Normandy and a eulogy of the murdered Copsig, which is followed by a truncated statement about the value of the support William was receiving from Ealdred and other English bishops, we are no longer able to check what Orderic was borrowing from it and what he was contributing himself.97 While it is reasonable to think that this passage does derive from Poitiers’s lost text, it is important that Orderic retained it. It suggests that he still believed William was trying to uphold the requirements of his coronation oath. But ‘D’’s reference to another heavy tax indicates that a second geld was levied, with the author retaining the moral tone of earlier passages.98 Gospatric, son of Maldred, the cousin of the murdered Osulf, was appointed as earl north of the Tyne at this Christmas court, a choice that continued William’s attempt to control the far north from a distance through reliance on locally powerful individuals. Since Gospatric survived as Copsig and Osulf had not, he was clearly thought a more acceptable choice by those subjected to him. But it is an indication of the circumstances of late 1067 that the Historia Regum, written in the 1120s at Durham, said that he obtained his position by offering a large sum of money, which William accepted.99 William’s calculation must have been that the down For the dating, Tanner, Families, Friends and Allies, 106, note 147. OV, ii, 210–13 (Desertores huiusmodi arte aliquoties reducuntur. Pari sedulitate et sollertia Gallos nunc instruebat . . .). 97 GG, 184–7. 98 ASC, ‘D’, 1067 (‘And here the king set a great tax on the wretched people’). 99 HRA, ii, 199 (multa emptum pecunia adepts est comitatum Northymbrensium); Kapelle, Norman Conquest of the North, 108–9; Aird, St Cuthbert and the Normans, 69; Fletcher, Bloodfeud, 171. On the much-discussed authorship and method of composition of the Historia Regum, David Rollason, ‘Symeon of Durham’s Contribution to Historical Writing in Northern England’, in idem (ed.), Symeon of Durham, 1–13, at 10, summarizes what was then the consensus. Much depends on the reception at Durham of a text of John of Worcester’s Chronicle and on the date of Symeon’s death, with 1122 and soon after 1128 being plausible limits. See Martin Brett, ‘John of Worcester and his Contemporaries’, in R. H. C. Davis and J. M. WallaceHadrill (eds.), The Writing of History in the Middle Ages: Essays Presented to Richard William 95 96

king of the english

285

payment would encourage commitment. It also illuminates again the urgent need to raise money, which was a feature of these times. The appointment at the 1067 Christmas court of Remigius, monk and almoner of the abbey of Fécamp, as bishop of the huge diocese of Dorchester-on-Thames after the death of Bishop Wulfwig, may have been disquieting to William’s new subjects – not only because any English hopefuls were disappointed, but because Remigius had contributed ships to the 1066 expedition. So obviously was he considered to be being rewarded for his support for the invasion of England that he was later accused of simony and forced in 1071 to answer charges at the papal court. Given the strategic importance of the diocese, which stretched from the Thames to the Humber and included many regions where a Danish army might land, it is arguably unsurprising that Remigius was chosen; indeed, with hindsight he proved to be a good appointment. However, Eadmer and William of Malmesbury write of him having made a bargain with William that he would receive the first vacant bishopric in England, an accusation to which he had no answer in Rome; his continuation in office, so we are told, owed everything to Lanfranc’s special pleading.100 That Remigius made his required profession of obedience to Stigand does, however, indicate that, at this stage, the archbishop of Canterbury must have enjoyed William’s positive support at a level not given to him by either the Confessor or Harold.101 Malmesbury, who admired Remigius, describing him as a small man with a great spirit, believed that he initially had great plans for Dorchester. A surviving memorandum shows the new bishop assessing and organizing his diocese’s resources around its traditional episcopal centre, an indication of the rapidity with which the newcomers took on what they identified as their new duties.102 The ‘E’ version of the Chronicle cryptically records that William gave away every man’s land, presumably dating it to the Christmas court.103 Immediately raising the subject of William’s treatment of the English, this much-discussed passage may well also be referring to further grants of land to Frenchmen; at this time Chichester and Arundel are known to have been given to Roger de Montgommery, who had been deliberately left Southern (Oxford, 1981), 101–26, at 120–1; JW, ii, lxxi–lxxiii. For a recent discussion of the full complexities, David Rollason, ‘Symeon of Durham’s Historia de Regibus Anglorum et Dacorum as a Product of Twelfth-Century Historical Workshops’, in Brett and Woodman (eds.), The Long Twelfth-Century View of the Anglo-Saxon Past, 95–111. 100 HN, 11; GP, i, 472–3. 101 David Bates, Remigius, Bishop of Lincoln (1067–1092) (Lincoln, 1992), 2–3, 4–7. 102 John Blair, ‘Estate Memoranda of c.1070 from the See of Dorchester-on-Thames’, EHR, cxvi (2001), 114–23. 103 ASC, ‘E’, 1067.

286

william the conqueror

behind in Normandy in 1066.104 The creation of three of the other so-called Sussex rapes – Pevensey, Lewes, and Hastings – under the lordship of Robert, count of Mortain, at Pevensey, William de Warenne at Lewes, and Count Robert of Eu at Hastings, suggests that the phenomenon of compact territorial blocs centred on a castle, in this case stretching back from the coast to the Weald, was being replicated in England with the strategic purpose of safeguarding the Channel crossing to Upper Normandy.105 The transformation from the initial reliance on a garrisoned castle within the town of Hastings, which had been made the responsibility of Humphrey du Tilleul soon after the Battle of Hastings, to one on large lordships that required an elaborate tenurial infrastructure, represented a major change.106 The installation of individuals drawn from the highest elite of pre-1066 Norman aristocratic society, a pattern only slightly modified when a fifth rape of Bramber was created for the less socially elevated William de Briouze in the early 1070s, was accompanied by the extensive introduction of the sub-tenants, many of whom appear in Domesday Book. Sussex had also been one of the heartlands of Godwine and his family’s power.107 The dispossessions must have been extensive, and the message that the very core of Godwine and Harold’s power was being ripped up must have been obvious to everyone. As far as the English were concerned, the process associated with the ‘giving away of everyone’s land’ was a logical continuation of the decision that those who had supported Harold fell within the king’s discretionary power to punish or pardon. The scale of the changes involved cannot, however, be truly verified because, as already noted, we know so little about the thousands of mostly nameless individuals who had died in the warfare of 1066. The one apparently specific commentary on what was involved for the English may be supplied by a writ from the remarkably full archives of the abbey of Bury St Edmund’s, ordering the transfer into the king’s hands of all land formerly held of the abbey by those who had died at Hastings fighting against William; in other words, the act of going into battle on Harold’s behalf identified them as traitors and their families were therefore disinherited. At the same time, the writ confirmed that others OV, ii, 210–11. J. F. A Mason, William the First and the Sussex Rapes (Hastings and Bexhill Branch of the Historical Association, 1966), 5–16; Robin Fleming, Kings and Lords in Conquest England (Cambridge, 1991), 146–7. 106 For Humphrey, OV, ii, 220–1. See further Brian Golding, ‘Robert, Count of Mortain (d.1095)’, ODNB. 107 Ann Williams, ‘Land and Power in the Eleventh Century: The Estates of Harold Godwineson’, ANS, iii (1981), 171–87, at 172–3, 185. 104 105

king of the english

287

who held of the abbot should continue to do so, a provision which presumably shows that those who had fought and survived were confirmed in their lands alongside all those tenants who had not been at Hastings.108 That someone as influential as Abbot Baldwin of Bury, a former monk of SaintDenis, should have felt obliged to move so swiftly to protect tenants who had fought for Harold suggests that the situation must have been more difficult for less obviously favoured individuals and institutions.109 On this basis, those who had to buy back their lands would be the kindred of those who had lost their lives and of the men who had fought against William in 1066, and those whose loyalty was in some way in doubt. Applied universally, any widespread redemption of lands that might have followed on from William’s order would have been an immensely complex process, since individuals had to be identified and decisions then had to be taken on the level of acceptable payment. Thousands of estates, most of them small, could have been taken over on William’s behalf.110 Evidence that local inquiries actually were undertaken again exists at Bury St Edmund’s in a reference to inquests involving Ingelric, a French priest who had come to England in Edward the Confessor’s day, and who was one of those who almost instantly profited from the Conquest.111 Two references in Little Domesday Book, both of which also mention Ingelric, also concern Bury estates. As it has now been argued that the abbey heavily influenced the Domesday text, all of the most specific evidence for whatever process of redemption took place ultimately comes from one source, which may mean that Bury is an exceptional case, or perhaps that the documentary thoroughness of Baldwin and his colleagues provides detail that is elsewhere mostly lost to us.112 For all Baldwin’s formidable abilities and presence, an extensive, albeit less cleanly articulated, process elsewhere is surely likely. That it was at its most intense in Wessex, Kent, and East Anglia is a logical corollary of these being the shires where the power of Harold and his brothers was mainly based, and where the lands of most of 108 Regesta, no. 37. Technically the dating limits of the writ are 1066xApril 1070, but it is hard to believe that it does not date from the first part of that period. 109 David Bates, ‘The Abbey and the Norman Conquest: An Unusual Case?’, in Tom Licence (ed.), Bury St Edmunds and the Norman Conquest (Woodbridge, 2014), 5–21, at 15–16. 110 Fundamental for this subject are Williams, The English, 8–10, 18–19; Garnett, Conquered England, 20–4. 111 Feudal Documents from the Abbey of Bury St Edmunds, ed. D. C. Douglas, British Academy Records of the Social and Economic History of England and Wales, viii (London, 1932), 8, with discussion at pp. xcvii–xcix; Pamela Taylor, ‘Ingelric, Count Eustace and the Foundation of St Martin-le-Grand’, ANS, xxiv (2002), 215–37, at 228–30. 112 DB, ii, fos. 360v, 367v. For Bury and the Domesday text, Stephen Baxter, ‘The Representation of Lordship and Land Tenure in Domesday Book’, in Elizabeth Hallam and David Bates (eds.), Domesday Book (Stroud, 2001), 73–102, at 93.

288

william the conqueror

those who had died would have been located.113 Whatever form the process took, it cannot possibly have applied soothing balm to many still open wounds. CAMPAIGNS AND CEREMONY (JANUARY–MARCH 1068) William’s readiness to campaign in the depths of winter, indicated by his leading an army to Exeter and on into Cornwall, before being back at Winchester for Easter, shows that he thought the problems facing him were very serious ones.114 His conduct on this campaign is located by Orderic within a framework of legitimate anger (ira). While the fate of the troops blown off course in the Channel and mistreated at Exeter could have been one reason for this act of retaliation, viewed as a continuation of the main themes of his conduct of 1066 that had prioritized the welfare of his troops, there are likely to have been other reasons for his rapid move against the city. The presence at Exeter, recorded by John of Worcester, of Harold’s mother Gytha must have been another consideration, with the wives (presumably widows) of other English nobles being with her. The city’s formal defiance may also have been motivated by resentment of taxation. This and the citizens’ efforts to bring in support from elsewhere in the south-west could indicate that Gytha and others were trying to co-ordinate a fightback in the West Country while awaiting the arrival of reinforcements from Ireland recruited by some of Harold’s sons, the offspring of his union with Edith ‘Swan-Neck’. Even if not on a scale that was going to reverse the English defeat at Hastings, the revolt is at least indicative of organized discontent sufficient to create a regional power base from which those who wanted to remove William could operate. Such was the rapidity of William’s advance that Gytha’s escape from the encircled city during the siege, presumably via the river Exe, suggests he had not called up ships to establish a blockade by river and sea. Rather than being seen as a strategic oversight on his part, this shows that he had once more taken the tactical option of going on to the offensive to isolate potentially dangerous opponents before they had completed their preparations. Another important development, reported by Orderic, is that William for the first time called English troops into his army. Gytha’s flight while the siege was ongoing, and references in the sources to divisions within the For this suggestion, Williams, The English, 8–10. For this campaign, ASC, ‘D’, 1067 (recte 1068); OV, ii, 210–15; JW, iii, 4–7; GR, i, 462–3; Robert Higham, ‘William the Conqueror’s Siege of Exeter in 1068’, Report of the Transactions of the Devonshire Association for the Advancement of Science, cxlv (2013), 93–132; J. O. Prestwich, ‘Military Intelligence under the Norman and Angevin Kings’, in Garnett and Hudson (eds.), Law and Government, 1–30, at 4–8; Prestwich, The Place of War, 27–31. 113 114

king of the english

289

city, suggest that any grand plans which had been formed rapidly came to nothing. When William’s forces were encamped four miles from Exeter, probably at Clyst Honiton, some citizens are said to have favoured submission and, as a result, hostages were handed over, presumably as a surety to guarantee their willingness to negotiate. However, when William asked the citizens to swear fealty and they refused, he attacked, ravaging the countryside as he went, and levying what the ‘D’ Chronicle again describes as a heavy tax. The citizens’ counter-proposal to pay the customary tribute, but not to do fealty, may have been a delaying tactic to test William’s resolve, with Gytha and her associates perhaps further arguing in favour of holding out until reinforcements arrived. William’s rejection of the proposal reflects his determination to rule as he believed his predecessors had done, as well as perhaps being a recognition that his enemies were playing for time. He then had a hostage blinded in full view of the defenders, a symbolic communication that the behaviour of both the citizens and their allies constituted an offence to his royal dignity, a message that the defenders of Exeter presumably understood.115 After this, his forces encircled the city. According to ‘D’, the subsequent siege lasted for eighteen days and William’s army suffered heavy losses; although, as we have seen, ‘D’ set out consistently to portray William in unfavourable terms, the comment is not unreasonable, since the strategy of direct assault on the city’s high ramparts was bound to be costly in human terms.116 All sources indicate that the terms on which Exeter surrendered were generous ones, with the ‘D’ Chronicle adding the caveat that William did not keep his promises; that he probably did accede to the citizens’ request to continue to pay dues at the pre-1066 level does not of course rule out other exactions being taken.117 It must also be important that William deliberately prevented his soldiers from looting the city. Manifestly copying Poitiers at this point, Orderic describes trustworthy guards being posted on the gates to deny entry to the rank and file of the army, showing that the customary rewards were denied them and a traditional punishment for rebellion was not allowed to happen.118 After the victory, a castle was built

115 For the symbolic significance of blinding, Geneviève Bührer-Thierry, ‘ “Just Anger” or “Vengeful Anger”? The Punishment of Blinding in the Early Medieval West’, in Rosenwein (ed.), Anger’s Past, 75–91, at 79–91. 116 Robert Higham, Making Anglo-Saxon Devon (Exeter, 2008), 167–74; idem, ‘William the Conqueror’s Siege of Exeter’, 112–13. 117 DB, i, fo. 100r. On this point, J. H. Round, ‘The Conqueror at Exeter’, in idem, Feudal England (London, 1895), 431–55, at 446–50; Higham, ‘William the Conqueror’s Siege of Exeter’, 104–5. 118 OV, ii, 124–5.

290

william the conqueror

within the walls of Exeter and the garrison initially placed under the command of William de Vauville, who was subsequently replaced by Baldwin de Meulles, a son of Count Gilbert of Brionne. Brian, son of Count Eudo of Penthièvre, was installed as earl in the south-west, an appointment that also puts William’s 1064 Breton campaign into perspective by affirming in the strongest terms his link with that family.119 Although many English remained important in the south-western counties at this stage, Norman and Breton landowners were installed throughout the region in the months after the siege; a diploma for Wells Cathedral confirmed by William at Whitsun 1068 already shows a multi-ethnic amalgam in the south-west and how the English, although still present, were being pushed aside.120 As with the Sussex rapes, the combined need to provide security and reward was driving through huge changes in the distribution of wealth and power. The siege of Exeter and the possibility that Gytha was planning to establish a base for Harold’s sons to make a landing also introduced William directly to the world of the politics of the Irish Sea and to the disruptive potential of any military intervention from there. However, when Harold’s sons Godwine, Edmund, and Magnus – the sources disagree slightly on their number and their names – arrived in the Severn estuary later in the year, they were beaten off at Bristol, where the citizens closed the gates, and then defeated in battle near Bleadon (Somerset) by a force led by Eadnoth the Staller, a prominent member of Edward the Confessor’s household, who was killed in the battle.121 His decision to fight for William must reflect the way in which some among the English chose to throw in their lot with him, a theme that runs through these years, and perhaps a perception of the changed balance of power in the south-west after the Exeter campaign. That Eadnoth’s descendants prospered under the new regime is also a comment on William’s treatment of the English.122 It is possible, too, that Harold’s sons may have been perceived as lacking the legitimacy to lead a large-scale rebellion.123 After Exeter’s submission, William marched his army into Cornwall, before returning to Winchester for Easter (23 March 1068) for one of the For Brian, Keats-Rohan, ‘Breton Contingent’, 160–1. Williams, The English, 21–2; Regesta, no. 286. 121 ASC, ‘D’, 1067 (recte 1068). For Bristol, Roger H. Leech, ‘The Medieval Defences of Bristol Revisited’, in Laurence Keene (ed.), ‘Almost the Richest City’: Bristol in the Middle Ages, British Archaeological Association Conference Transactions, xix (London, 1997), 18–30, at 18–23. 122 Ann Williams, ‘Eadnoth the Staller (d.1068)’, ODNB. 123 Barlow, The Godwines, 120–1, comments that they ‘were generally regarded as freebooters’. Other views are of course possible; see Higham, ‘William the Conqueror’s Siege of Exeter’, 102–3. 119 120

king of the english

291

first of the great gatherings held around royal crown-wearings that became central to his rule in England. While there he probably received a visit from Gervin, abbot of Saint-Riquier in Ponthieu. Although the abbey’s chronicle says that Gervin crossed the Channel from Wissant in February to secure a charter confirming his abbey’s English possessions, he surely did not travel into the far south-west to find William; a meeting at Winchester is much more likely. The chronicle also says that Gervin was granted almost instant access to the king, something that was unusual.124 Its tale of Gervin initially setting out to cross the Channel in the company of more than one hundred abbots and monks, as well as many military men and merchants, who were then deterred from crossing by severe weather, also illuminates the expectations that the conquest of England had raised. A truly great gathering was presumably anticipated, an indication that messages were circulating, both formally and informally, about the planned grandeur of William’s new kingship. Apart from this, little is otherwise known about William’s activities either at Winchester or afterwards between Easter and Whitsun. The one certainty is that arrangements to bring Matilda from Normandy for her coronation were being made. MATILDA’S CORONATION Matilda’s coronation on Whit Sunday (11 May) at Westminster, so soon after William’s, indicates that it was thought to be a very important step in the consolidation of rule. As we have seen, there were a host of reasons for this, both within the traditions of English and western European queenship, and arising out of her standing in Normandy in the years immediately before 1066. It may also be connected to a perceived need to affirm her position as the bearer of William’s heirs, a necessity made all the more compelling by the likelihood that she was going to undergo yet again the risks of childbirth; although the exact date of birth of the couple’s ninth child, the future King Henry I, is somewhat unclear, it is generally thought to have occurred in the autumn of 1068, and most probably between the coronation and early September.125 The political calculations involved would probably also have taken into account the need to emphasize a relationship between William and Matilda that was tantamount to a partnership in rule, at least on the terms that the eleventh-century mind could understand one. From this point onwards, Hariulf, Chronique de l’abbaye de Saint-Riquier, 241; Regesta, no. 259. For Henry’s birth, Hollister, Henry I, 30–1; Green, Henry I, 20–1. Given Matilda’s fertility, conception during William’s 1067 visit to Normandy seems likely and makes me prefer Hollister’s earlier dating. 124

125

292

william the conqueror

although she spent most of her time in Normandy, she is consistently present among the signa of the English diplomas, with, where necessary, arrangements seemingly being made to add her approbation when it could not initially be obtained.126 She also acquired extensive English lands scattered across eight counties of southern England.127 Although it is not known which ordo was used to crown her, the overwhelming probability is that it was a traditional English one. As with William in 1066, the coronation was performed by Archbishop Ealdred.128 Attended by the great figures of English and Norman aristocratic society, with visitors from France crossing the Channel especially to be there, it seemingly illuminates the joining together of disparate groups at what was intended to be a great celebration of the new kingship. The Laudes chanted were composed especially for the occasion. They elevated the newly crowned queen to third place in the celestial hierarchy, below only Pope Alexander and her husband, and above all the other clergy and lay magnates. William is described as pacificus et magnus rex and the adjective serenissimus is used in relation both to him and to Matilda. Their mixed ethnicity notwithstanding, the magnates are collectively categorized as the principes of the English. The language of these Laudes, unknown in the English and French kingdoms, is of a kind used in imperial Laudes in Germany during the time of Henry III.129 It enshrines an unprecedentedly elevated view of the royal and queenly dignity and, in the emphasis on peace and the absence of ethnic distinction among the principes, it puts the accent on the reconciliation of peoples. The inclusion of a chant calling for prayers for the health and long life of Archbishop Ealdred raises the possibility that he was influential in their composition and certainly underlines his role in trying to bring peoples together; it also fits well with a great deal of other evidence for his activities in bringing ideas and practices into England from the Empire.130 Archbishop Stigand’s recorded presence enhances further just how remarkable this emphasis on the centrality of Ealdred’s role was. Like so much else, it shows the depth of thought that went into the presentation of William’s kingship. 126 Regesta, 93–4; David Bates, ‘The Representation of Queens and Queenship in AngloNorman Royal Charters’, in Paul Fouracre and David Ganz (eds.), Frankland: Essays in Honour of Dame Jinty Nelson (Manchester, 2008), 285–303, at 287–91. 127 For Matilda’s lands, DP, 296–7. 128 ASC, ‘D’, 1067 (recte 1068); OV, ii, 214–15; JW, iii, 6–7. 129 Above all, Cowdrey, ‘The Anglo-Norman Laudes Regiae’, 50–61, 70–1. 130 Nelson, ‘Rites of the Conqueror’, 129–30 (Politics and Ritual, 397–9); Stafford, Queen Emma and Queen Edith, 183–5; Laura L. Gathagan, ‘The Trappings of Power: The Coronation of Matilda of Flanders’, HSJ, xiii (2004, for 1999), 21–39, at 26–31. These arguments are further strengthened by C. R. J. Currie, ‘Odda’s Chapel, Ealdred’s Inscriptions? The Deerhurst Inscriptions in Some Continental Contexts’, HR, lxxxiii (2010), 1–45, esp. pp. 6–7, 9–19; Ortenberg West-Harling, ‘Un prince évêque anglo-saxon’, 155–7.

king of the english

293

Two diplomas confirmed at this time, although in neither case surviving as an original, are also at first sight redolent of an apparent assimilation of the newcomers into an English world.131 The form of both is that of the traditional Anglo-Saxon diploma. William’s kingship is described in language derived directly from the English past, which stresses an imperial role across the British Isles: in one of them he is basileus, in the other, ‘monarch of all of Britain’ (tocius Brittanie monarches). The beneficiaries of the diplomas are both figures from Edward the Confessor’s time, but both were also individuals of continental origin. Ingelric, the French priest who had turned the Conquest to profit in eastern England, had William confirm his and his brother’s foundation of the collegiate church of St Martinle-Grand outside the walls of London. And the Lotharingian Giso, bishop of Wells, had William restore an estate taken from his church by Harold, whose despoliation of that church is vilified in the text of the diploma but who is nonetheless referred to as ‘king’ (rex). Giso’s efforts to assimilate William to English kingship, therefore, did not rewrite the recent past in a way that was soon to become universal, but William was probably pleased to listen to a tale of his predecessor’s supposed villainy.132 Like Bishop Wulfstan, Giso was a bishop whom William wished to encourage. It is also hard to think that Giso would have been successful in his suit if Harold had won the Battle of Hastings. The absorption of such individuals into a closer collaboration with the new regime was every bit as much a strategic move as the ongoing construction of castles in towns. The grandeur of the occasion is further underlined by the presence as witnesses to the diplomas of a remarkable mixture of English and French, with, according to Orderic, Matilda’s entourage also including the author of the Carmen de Hastingae Proelio, Bishop Guy of Amiens.133 Not unconnected with this paraphernalia of elevated dignity and status must surely be the fact that William and Matilda’s fourth son, the soon-to-be-born Henry, did not receive a name from the naming stock of the Norman dukes, but from that of the Capetian kings and the Salian emperors. The propagation of harmony and peace at Matilda’s coronation was shattered within a few weeks. Very little imagination is required to grasp 131 Regesta, nos. 181, 286. For further discussion, Keynes, ‘Regenbald’, 218–20; idem, ‘Giso, Bishop of Wells’, ANS, xix (1997), 203–71, at 242–3; Taylor, ‘Ingelric, Count Eustace and the Foundation of St Martin-le-Grand’, 215–37 (with amendments to the edition in Regesta, no. 181, at p. 215, note 3). On the diplomas’ pre-1066 antecedents, Molyneaux, The Formation of the English Kingdom, 206–9. 132 Since Bishop Giso’s diploma survives only in a fifteenth-century copy, there is a possibility that the word rex is a later copyist’s interpolation. This seems to me unlikely, given that Harold’s kingship was subsequently defamed and that the surviving text has every appearance of being an accurate reproduction of a documentary form soon to become obsolete. 133 OV, ii, 214–15.

294

william the conqueror

the undercurrents that must have been present at the great gathering. Not only had the traditional leaders of the English been denied their accustomed roles by being taken to Normandy for much of 1067, but, beginning in the immediate aftermath of the Battle of Hastings, the intrusion of Norman/French newcomers into well-established and dynamic local and regional societies created an overawing military presence that must have been a constant factor in exacerbating strained relationships wherever it occurred. The taxation and castle-building so soon after William’s coronation would have been regarded as indispensable components of peace and security by the victors and, in all likelihood, by some of the defeated as well. But for others they involved the costs of paying for and billeting what must have felt like an occupying army, and the cause of the destruction of their livelihoods and those of their descendants. The situation was highly volatile and in some places explosive. And, as we shall see in the next chapter, its immediate implications in terms of living conditions were very great. With William’s seal also giving a very clear indication of priorities that favoured protection of those who had served him before and during 1066, under his personal direction the balance from the start tilted heavily in favour of reward and security. Exceptional wealth and power were given to newcomers from France from a very early date. The events that followed are a strong warning about the limitations of charter and liturgical evidence. Since both forms of evidence are ultimately designed to register agreement and harmony, they reveal only the thoughts and intentions of those with an interest in maintaining each of these two states. They are illustrative of the powerful dynamic with which William’s kingship in England was being articulated, both by the new king and the continental newcomers and by the English who were trying to effect assimilation and secure peace. But, in the end, they illustrate neither an ‘early mildness’ nor a ‘political dream, briefly cherished by King William, of Englishmen and Normans joining hands to forge an “AngloNorman” kingdom’ referred to earlier in this chapter.134 They state an ideal that many might have hoped to move towards, not the reality of 1068. The fact that William had spent approximately half of the first eighteen months of his reign as king of the English in Normandy must also have sent mixed messages to his new subjects. For all its articulation of another ideal, the continuation of the kingdom of the English under William’s rule, the evidence shows that the cross-Channel empire was going to be a central factor in its future history.

134

Above, 260.

Chapter 8

VI C TO RY T R AN S F O RMED

THE FIRST ENGLISH REVOLTS Both the ‘D’ and the ‘E’ Chronicles say that Edgar the Ætheling, his mother, his sisters, and other English fled to Mael Coluim (Malcolm) III, King of Scots, soon after Matilda’s coronation; Orderic that, at around the same time, the two brothers Earls Edwin and Morcar left William’s court to start a rebellion that was assisted by King Bleddyn ap Cynfyn.1 According to Orderic, Edwin was disenchanted with William because the promised marriage to one of the king’s daughters had not taken place. Since the Chronicle makes no mention of Edwin and Morcar, and Orderic, none of Edgar, it is as good as certain that their actions were not co-ordinated. There is indeed nothing to show that Edgar was at this point involved in the fighting at all. With the recent experience of 1066 in mind, Edgar and, since he was after all still in his mid-teens, presumably those advising him, probably reckoned that it was essential to organize significant support before taking on William; subsequent events do indeed show that he was using his time in Scotland for just this purpose.2 Possibly influential too on his thinking was the fact that Edwin and Morcar had done nothing to help him in 1066; indeed, they may have done their best to obstruct him. Whatever the values and the psychology involved, these high-level English defections are eminently understandable as a reaction to the course of events initiated after Hastings. None of them had been treated as their status would have led them to expect. But there is nothing in any of it to suggest the existence of a unified coherent attempt to remove William. In addition to the promised bride being withheld from Edwin, both he and Morcar were being deprived of the status and authority that had gone

1 ASC, ‘D’, ‘E’ (1068), OV, ii, 214–21; JW, iii, 6–9. ‘D’ places Edgar’s departure before Matilda’s coronation, but its statement that it happened in the summer requires that the departure be placed afterwards, as in John of Worcester. 2 For the main modern accounts, Kapelle, Norman Conquest of the North, 108–19; Williams, The English, 24–7; Fletcher, Bloodfeud, 171–3; Paul Dalton, Conquest, Anarchy and Lordship: Yorkshire 1066–1154 (Cambridge, 1994), 8–11; Baxter, The Earls of Mercia, 273–4.

295

296

william the conqueror

with the rank of an earl before 1066.3 Although it is difficult to date all the compendious evidence for them being marginalized to before 1068, their eight-month sojourn in Normandy must have reinforced feelings of exclusion, with the intrusion of others into their area of jurisdiction contributing further to it. One good piece of evidence is contained in a writ for Westminster Abbey, dating from approximately this time, which concerned an estate in Staffordshire. Correctly addressed to Edwin as the earl within whose jurisdiction it lay, it assigns an executive role in implementing the royal order to Æthelwig, abbot of Evesham. Known from an Evesham Gesta Abbatum of the early twelfth century and from the thirteenth-century chronicle of his abbey to have been given special authority in the West Midlands by William, Æthelwig was one of the Englishmen whose loyalty and utility were recognized from an early date and who most definitely prospered under William’s rule.4 Although not occurring in 1066 as the Evesham Chronicle States, the abbey of Winchcombe was also placed under his authority when its abbot, Godric, was deposed, probably in 1074.5 Æthelwig’s prominence continued until his death in 1077. The way in which events unfolded shows that William, as he had done before Val-ès-Dunes against William of Arques, and very recently against Exeter, and on many other occasions, went on the offensive before his opponents’ plans could be brought to fruition. Although he had disbanded troops, he must still have had a considerable army at the ready, whereas his opponents had to gather theirs. As a result, he was able to place forces between Edwin and Morcar and their allies from Wales located in Mercia and the army that was forming further north, but which had remained beyond the Humber. A march to Warwick, where William is said to have built a castle, was enough to persuade Edwin and Morcar to sue for peace. They were then pardoned. A march further into the north intimidated the northern rebels, who also made peace and gave hostages. More castles were built at Nottingham and York, the latter on the site of what is now known as Clifford’s Tower at the junction of the Ouse and Foss rivers.6 Reverberations were felt much further north, with Bishop Æthelwine of Durham making his peace with William before acting as an intermediary Baxter, The Earls of Mercia, 281–97. Regesta, no. 292; Thomas of Marlborough, History of the Abbey of Evesham, ed. Jane Sayers and Leslie Watkiss, OMT (Oxford, 2003), 162–3. On the Gesta Abbatum, J. C. Jennings, ‘The Writings of Prior Dominic of Evesham’, EHR, lxxvii (1962), 298–304, at 302–3. 5 Thomas of Marlborough, History of the Abbey of Evesham, 164–7. For the chronology, The Winchcombe and Coventry Chronicles: Hitherto Unnoticed Witnesses to the Work of John of Worcester, ed. and trans. Paul Antony Hayward, 2 vols., Arizona Center for Medieval and Renaissance Studies (Tempe, AZ, 2010), i, 286–8. 6 For York, David Rollason with Derek Gore and Gillian Fellows-Jensen, Sources for York History to AD 1100 (York, 1998), 180–3. 3 4

v i c t o ry t r a n s f o r m e d

297

with the King of Scots. Mael Coluim (Malcolm), having seemingly not moved in any way to support Edgar, agreed terms with William and swore fealty and obedience to him, presumably through intermediaries. On his journey south, William organized the construction of further castles at Lincoln, Huntingdon, and Cambridge. According to Orderic, Warwick was placed under the control of Henry, son of Roger de Beaumont; Nottingham, of William Peverell; and York, of Robert fitz Richard. Other Normans and French were installed throughout the kingdom. Also introduced into Yorkshire were William Malet, Gilbert de Ghent, and probably William de Percy, all with the manifest responsibility of defending York. It is possible, too, that William de Warenne was introduced into southern Yorkshire in a compact lordship at Conisbrough at this time.7 An equally dramatic shift in the relative power of the old elite and the newcomers was the probable creation at around this time of the compact territorial earldoms of Shrewsbury and Chester for Roger de Montgommery and Gerbod of Oosterzele-Scheldewindeke, the brotherin-law of William de Warenne.8 This overall chronology of castle-building and land transfer ultimately relies on a pervasive modern trend to bring to earlier dates the creation of compact lordships in the north and west of England. Roger de Montgommery’s installation in Sussex and Shropshire and William de Warenne’s in Sussex and Yorkshire are also striking manifestations of the deployment across the whole of Normandy and England of a small elite with concomitantly massive imperial responsibilities. The construction of these castles within English towns has recently been described in terms of ‘a whole series of fluid and rapidly evolving “borderlands” ’. A statement that harks back to John Le Patourel’s introduction of the concept of colonization into the analysis of the Norman Conquest, but which refines it, it brings into focus the dynamic and the disruption that are absolutely indispensable to understanding the processes involved during these times.9 At Exeter, the castle was built within the city walls. While houses were destroyed to make space for it, there was no enforced takeover of land because it was built on royal land. The surviving gate-house of the castle was a unitary structure that combined English and continental 7 For Yorkshire, Dalton, Conquest, Anarchy and Lordship, 10–11, 65–6, 101–2; Fletcher, Bloodfeud, 174–5. 8 For the probability that Roger acquired Shrewsbury in 1068, Lewis, ‘Early Earls’, 219–20. For Chester, C. P. Lewis, ‘The Formation of the Honor of Chester, 1066–1100’, in A. T. Thacker (ed.), The Earldom of Chester and its Charters: A Tribute to Geoffrey Barraclough, Journal of the Chester Archaeological Society, lxxi (1991), 37–68, at 38–41. 9 Richard Eales, ‘Castles and Borders in England after 1066’, Château Gaillard, xxvi (2014), 149–57, at 155; cf. John Le Patourel, ‘The Norman Colonization of Britain’, in I Normanni e la loro espansione in Europa sull-alto medioevo, Settimane di Studi del Centro Italiano sull’Alto Medioevo, xvi (Spoleto, 1968), 409–38.

298

william the conqueror

stylistic features, indicating conquerors and defeated working together (Plate 15).10 It also shows that, as with coins, the life histories of buildings can be every bit as eloquent an illustration of the new societies and relationships that were being created after 1066 as histories and documents. At York, the castle was placed within the existing city walls at the point where they reached the river Ouse on a site that had almost certainly performed defensive functions before 1066. With the city easily accessible by water, as Harold Hardraada had shown as recently as 1066, the purpose was to create what was effectively barracks for a garrison, placed there to do what they could to obstruct a hostile fleet sailing in up the river and protected by the earthworks from any disloyalty on the part of the citizens.11 While each local case of castle-building has to be studied individually and the chronology of construction on every site is often difficult to pin down, what was initially involved, as at Exeter, York, Norwich, and Wallingford, was the creation of a simple, but large, new earthwork fortification within the existing urban space or, occasionally, as at Nottingham, and the second castle at York, just outside it.12 Even at Winchester, which had surrendered very early and which was in no way a centre of resistance or rebellion, the destruction within the walls was considerable and an area of approximately 5.7 acres was taken over for it.13 The same applies to Oxford, likewise a town with no history of opposition to the new regime.14 Influenced by a famous passage in which Orderic, probably copying Poitiers, said that castles were scarcely known to the English before the arrival of the Normans and that they had as a result of their absence been able to put up only weak resistance to their enemies, the placing of these castles within towns has too often been seen as more innovative than it actually was.15 As long as Poitiers’s authorship is accepted, the passage becomes another piece of rhetoric, a proclamation that the kingdom’s Higham, ‘William the Conqueror’s Siege of Exeter’, 119–23. Most recently on York, Sarah Rees Jones, York: The Making of a City 1068–1350 (Oxford, 2013), 62–6; David M. Palliser, Medieval York 600–1540 (Oxford, 2014), 87–8. 12 In general, Barbara English, ‘Towns, Mottes and Ring-Works of the Conquest’, in Ayton and Price (eds.), The Medieval Military Revolution, 45–61, at 52–3. See further, e.g., Elizabeth Popescu, ‘Norwich Castle’, in Davies, Riley, Levesque, and Lapiche (eds.), Castles and the Anglo-Norman World, 1–29, at 12–13; Oliver Creighton, ‘Castle, Burh and Borough: Unravelling an Urban Landscape of Power at Wallingford, Oxfordshire’, Château Gaillard, xxvi (2014), 113–24, at 114–15, 119; David Roffe, ‘Anglo-Saxon Nottingham and the Norman Conquest’, in John Beckett (ed.), A Centenary History of Nottingham (Manchester, 1997), 24–42, at 36–7, 38–9; Pamela Marshall and Trevor Foulds, ‘The Royal Castle’, ibid., 43–55, at 44–7; Rollason et al., Sources for York History, 182–3. 13 Martin Biddle (ed.), Winchester in the Early Middle Ages: An Edition and Discussion of the Winton Domesday, Winchester Studies, i (Oxford, 1976), 302–3. 14 Blair, Anglo-Saxon Oxfordshire, 177. 15 OV, ii, 218–19. 10 11

v i c t o ry t r a n s f o r m e d

299

defences were being improved so that there would not be another conquest. It is not a statement of massive technological innovation. In the context of these years, their significance lay elsewhere.16 Viewed sociologically, the construction of these fortified enclosures is merely a development within a continuum whereby the duties of fortressbuilding and defence that were an aspect of pre-1066 English society were adjusted to new circumstances. Viewed in human terms, however, the whole process was deeply destructive and intrusive, since what was involved in the short term was a well-nigh instantaneous creation of a new physical and military presence. While William’s methods were pretty much standard in terms of the practices of eleventh-century rulers who had to overcome rebellion, it is essential to keep in mind just how deeply the results cut into society and how much resentment and misery they could cause. As Thomas Bisson, using the expansive narrative sources that the German kingdom supplies, and also taking full account of their rhetorical exaggerations, has put it: ‘the castles were generators of violence’; and that is without the ethnic factors that were present in England.17 This side of their role was soon to become very evident. The disruptive factors present were almost always the requisitioning of land and, when it was needed, the destruction of houses and the movement of populations; and the installation of a garrison who were soon rewarded with land and who, as happened later at York, could seize valuables from the local population. What can be known of the lands that William Malet acquired in Yorkshire between 1068 and 1070 – the latter being the date when he left York for southern England and, for our purposes, an exceptionally useful terminus – shows that the objective in creating an extensive landed estate for him formed out of the lands of multiple English landholders was the defence of York and the Humber.18 The speed with which it was put together was both a remarkable feat of organization and an awesome display of coercive power to achieve a political end. Another case study from this early period, William fitz Osbern’s installation of followers and his grants to Norman monasteries in the 16 In general, Robert Liddiard, ‘Landscapes of Lordship’: Norman Castles and the Countryside in Medieval Norfolk, 1066–1200, British Archaeological Reports, British Series, no. 309 (Oxford, 2000), 3–5; Charles Coulson, Castles in Medieval Society: Fortresses in England, France, and Ireland in the Central Middle Ages (Oxford, 2003), 30–41; John Goodall, The English Castle, 1066–1650 (New Haven, CT, and London, 2011), 49–65; Robert Higham, ‘Castle Studies in Transition: A Forty-Year Reflection’, The Archaeological Journal, clxvii (2010), 1–13; McNeill, ‘Davison versus Brown’, 48–9; Eales, ‘Castles and Borders’, 152–5. 17 Bisson, The Crisis of the Twelfth Century, 215–21. 18 P. R. Newman, ‘The Yorkshire Domesday Clamores and the “Lost Fee” of William Malet’, ANS, xxii (2000), 261–77, at 264–6, 270–2; Fleming, Kings and Lords in Conquest England, 159–60.

300

william the conqueror

vicinity of Chepstow, all of which must have happened before his death in early 1071, also shows just how intrusive this process was.19 And another, Norwich, where the earthworks were steadily extended after a probable start date of c.1068, suggests the destruction of ninety-eight properties and the creation of a fortification that was held by 300 armed men when it was besieged in 1075. The Little Domesday Book entry for the town refers to the construction of the new ‘French borough’ to the east of the castle and the miserable conditions endured by many in the old town.20 In all probability this period also witnessed the creation of what came to be called castle-guard obligations from the newly granted estates to support garrisons. In later times, these were commuted as it became necessary to have only small peacetime garrisons. But the first phase involved new infrastructures and obligations.21 Although not concerned with castlebuilding, a narrative that describes the sort of incident that must have occurred – perhaps quite often – is supplied by the Gesta Herwardi’s account of Hereward’s return to Bourne (Lincolnshire) to find that his brother had been murdered, with the beheadings, revenge, and the rallying of the English of the locality perhaps being representative of many similar incidents elsewhere.22 All of this, combined with the levying of taxation on a scale unknown since the time of King Æthelred’s wars and Cnut’s early years, and extensive enforced military service, were certain to stoke the fires of resentment just as they also secured political dominance. Events were subsequently to show that it was the seizure of the strategic centres at the heart of the political power of the English state in 1068 that made the Conquest. A new domination was imposed. The short and militarily successful campaign of the year 1068 and its consequences ultimately changed everything. In the short term, its outcome was to persuade William that a visit to Normandy could be made.

19 For Chepstow, Matthew Strickland, ‘Status, Display and Defence on the Welsh March, circa 1067 to 1087: Reflections on the Context and Functions of the Great Tower at Chepstow, Gwent’, in Erik Beck et al. (eds.), Burgen im Breisgau: Aspekte von Burg und Herrschaft im überregionalen Vergleich (Osfildern, 2012), 89–109, with references to other literature. For the tenants and grants to Norman abbeys, David Bates, ‘William the Conqueror, William fitz Osbern, and Chepstow Castle’, in Rick Turner and Andy Johnson (eds.), Chepstow Castle, its History and Buildings (Logaston, 2006), 15–22; Baxter, The Earls of Mercia, 285–6. 20 DB, ii, fo. 116v–17r; Brian Ayers, ‘The Urban Landscape’, in Medieval Norwich, ed. Carole Rawcliffe and Richard Wilson (London and New York, 2004), 1–28, at 13–14; James Campbell, ‘Norwich before 1300’, in idem, 29–48, at 39–40; Popescu, ‘Norwich Castle’, 12–13. 21 John S. Moore, ‘Anglo-Norman Garrisons’, ANS, xxii (2000), 205–59, esp. pp. 209–10, 218–21. 22 Gesta Herwardi, 364–8; De Gestis Herwardi, ed. Meneghetti, 112–19 (trans. Swanton, Three Lives of the Last Englishmen, 62–4).

v i c t o ry t r a n s f o r m e d

301

A SHORT VISIT TO NORMANDY Although it is not mentioned in the literary sources, William and Matilda must have crossed to Normandy in the autumn of 1068 and spent Christmas there. Given the certainty that the future Henry I was born in England, the journey presumably took place after Matilda had recovered from childbirth. References to visits to Bonneville-sur-Touques and, in a charter dated to 1069, to Valognes, with other charters also being confirmed for the abbeys of Saint-Désir of Lisieux and Troarn, suggest an itinerary that took them westwards from the Seine valley to the north of the Cotentin, with a crossing to England, perhaps from Barfleur to Portsmouth, taking place in January.23 This journey would surely have taken in a visit to Caen, with William and Matilda and their court perhaps spending Christmas there. This itinerary, followed in the midst of so much turbulence, is an important commentary on priorities that kept visits to Normandy as a central feature of rule. In addition to the confirmation of the charters, the one piece of business known to have been conducted during this visit is recorded in a remarkable personal narrative composed by Rainald, at the time a royal chaplain and later a monk at Jumièges, and, from 1084, abbot of Abingdon. It tells how William and Matilda, along with Archbishop John of Rouen, Roger de Beaumont, and unnamed others, heard a case involving an attempt by a woman to reclaim a child that she said she had sold to the wife of a certain Stephen after he and his wife had lost their child, the sale having been made to the wife without Stephen’s knowledge. William and the others decided that the woman should have the child back if she could demonstrate that she was the mother by passing unscathed through the ordeal by hot iron, which she duly did. She therefore regained her son, but, with impeccably ruthless logic, was then denied the property he would have inherited, which was granted by the court to Matilda and, later, by her, to Rainald.24 While Archbishop John’s role as the duchy’s senior churchman is self-explanatory, Roger de Beaumont’s central role is significant as a commentary on the politics of favour and on the distribution of responsibilities within cross-Channel rule. If William of Malmesbury’s explanation is true that Roger acquired few lands in England because he regarded the Conquest as an act of plunder, it indicates that diversity of opinion was no barrier to high favour with 23 Regesta, nos. 162, 179, 256, 280, with a discussion of their dates at p. 78. No. 256 might indicate a second visit to Normandy in the summer of 1069, but the serious deterioration of the military situation in England makes this unlikely. It therefore surely belongs in this period. 24 Regesta, no. 162.

302

william the conqueror

William; it was loyalty and ability that presumably counted, and for this reason Roger remained of central importance in Normandy. His principles did not prevent his sons, one of whom had fought at Hastings, in due course receiving huge lands in England.25 In contrast, Roger de Montgommery, equally favoured by William and in the process of becoming massively endowed with land in England, presented a complete record of the abbey of Troarn’s endowment to William for confirmation in the last weeks of 1068. The timing, which cannot have been accidental, must be associated with the imminence of the start of a new life involving wide-ranging cross-Channel responsibilities; it includes mention of a grant made to the abbey when Roger was about to cross the sea with William, which was presumably intended as an offering to secure divine approval for enterprises to come.26 These charters and some other evidence also suggest that, the more William was preoccupied by English troubles, the more prominent his eldest son Robert became in Normandy. Thus, when Matilda crossed to England for her coronation, it was Robert who was left behind to mind the duchy.27 The phraseology of the charters confers on him a special status and authority. One, a confirmation for the nuns of Saint-Désir of Lisieux, refers to William as ruling over the English, and Robert as exercising princely power over the Normans. A second, the one for Troarn, was the subject of a special confirmation by Robert alongside that of his father. Although in both of them William’s overall authority is also referred to, the emphasis on Robert’s importance is striking.28 It is perhaps no wonder that, from the perspective of 1070–1, when he finally stopped writing, William of Jumièges could refer to Robert as ‘duke and advocate’ and ‘a most admirable and handsome youth’.29 When Robert’s status is placed alongside the special eminence conferred on Matilda, it looks as if William was deliberately enhancing the roles of individual members of his family to deal with the new circumstances of ruling territories on both sides of the Channel. Likewise, a confirmation, 25 GR, i, 736–7. For an argument that Roger’s cross-Channel lands were more extensive than is immediately apparent, David Crouch, The Beaumont Twins: The Roots and Branches of Power in the Twelfth Century (Cambridge, 1986), 10, note 33, 116–19. 26 Regesta, no. 280 (Et quia mare erat transiturus cum rege addidit ipse Rogerius). 27 GND, ii, 178–9. 28 Note . . . Guillelmo regum nobilissimo apud Anglos, Rotberto filio eius principante apud Normannos et Cenomannos and concedente etiam Rotberto Normannorum Cenomannorumque principe et eius patre Guillelmo Anglorum rege, Regesta, no. 179; Aprobamus (sic) Roberti fillii (sic) regis, ibid., no. 280. See further, GND, i, 137–8; Regesta, 95; Davis, ‘William of Jumièges’, 597–606; Aird, Robert Curthose, 62–4; Katherine Lack, ‘Robert Curthose: Ineffectual Duke or Victim of Spin?’, HSJ, xx (2009, for 2008), 110–40, at 120–2. 29 . . . quo in presentiarum duce et aduocato gaudemus, GND, ii, 184–5.

v i c t o ry t r a n s f o r m e d

303

in all probability made in England, of Norman property held by the abbey of Saint-Denis, shows William fitz Osbern also operating within this small group.30 What the charters do not tell us about of course is the nature of the personal relationship between William and his maturing adolescent eldest son. THE GREAT CRISIS OF 1069–70 William’s crossing to England, with Matilda again travelling with him, must have been made in response to news that a serious crisis was developing in the north. With Earl Gospatric having joined Edgar the Ætheling in Scotland during the last months of 1068, a new earl, the Fleming Robert de Comines, was appointed to replace him, presumably on instructions sent from William in Normandy during the autumn. Robert’s journey to Durham, the furthest point north that any French follower of William had travelled, was in all likelihood deemed logical as a follow-up to Bishop Æthelwine’s submission and also as a way of signalling a presence. But on 31 January 1069, Robert and the several hundred men with him were ambushed in Durham and killed. Symeon of Durham says that the slaughter was provoked by the earl’s forces having ravaged the countryside and killed peasants.31 Jumièges’s account, although somewhat garbled, implies that Durham had become the centre from which a co-ordinated attack on William’s power in England was being launched. He also said that the rebels sent emissaries to Svein Estrithsson, a step taken in undoubted awareness that the massacre had raised the stakes.32 Given the timescale of events in the first months of 1069, it is also clear that Edgar the Ætheling must have moved south by this point. This and the killings at Durham must have prompted an uprising in York, as a result of which the commander of the garrison, Robert fitz Richard, and many of his soldiers were killed by local insurgents, who quickly formed an alliance with Edgar. It surely cannot have been a coincidence that a revolt broke out in Maine early in 1069 at almost the same time that all this was happening in England. The rebellion, which attempted to replace William with a descendant of Count Herbert ‘Wake-Dog’ (1017–35), was in full swing by April, by which time William had already been on a campaign that had taken him north to York and then south to Winchester for Easter (13 April). The revolt was in all 30 Rolf Grosse, Saint-Denis zwischen Adel und König: Die Zeit vor Suger (1053–1122) (Stuttgart, 2002), 239–40 (Urkundenanhang, no. 2); Bauduin, La première Normandie, 382–3 (Annexe II: Dossier de Textes, no. 13). 31 OV, ii, 220–3; LDE, 182–5; HRA, ii, 186–7. See further Aird, St Cuthbert and the Normans, 70–1. On Robert, Oksanen, Flanders and the Anglo-Norman World, 197. 32 GND, ii, 178–81.

304

william the conqueror

probability masterminded by William’s old adversary, Geoffrey de Mayenne. With Herbert’s daughter Gersendis, her husband Azzo, lord of Este, and their son Hugh, having been brought from Italy, the plotting must have started during 1068, which is also indicative of an awareness that William was becoming bogged down in England. It attracted widespread, if indeed not total, support from the Manceau aristocracy, and the castellans holding Le Mans on William’s behalf were chased out and sought refuge in Normandy.33 Although the rebels’ initial unity rapidly fractured, neither William nor anyone from Normandy was at this time in a position to intervene. Bishop Arnold of Le Mans, William’s one overt ally in Maine at this point, crossed to England and was at Winchester at Easter.34 William’s decision to stay in England to tackle the rebellion there was presumably based on the calculation that a Danish intervention could lead to the irreversible collapse of all that he had achieved since 1066. Although the numbers were unpredictable, the likelihood that, with a direct invitation to intervene having been sent to King Svein, many English would join him constituted a massive threat. He would also have known that supporters in Maine would chip away at the consortium that was trying to replace him. At some point he sent Æthelsige, abbot of St Augustine’s, Canterbury, to King Svein’s court to try to persuade him against launching an invasion.35 The responsibility entrusted to an abbot of English birth is another illustration of the surviving religious elite trying to work with William. The mission’s failure indicates both Svein’s determination and, presumably, an assumption on his part that he would be welcomed by many when he reached England. Orderic’s account of the English rebellion indicates that Edgar had attracted additional support, including Maerleswein, the pre-1066 sheriff of Lincolnshire and the holder of extensive lands stretching from the south-west of England to Yorkshire, and several members of the house of Bamburgh, including Arnketel, son of Ecgfrith, and the sons of Karl, as well as Gospatric.36 That this group brought together individuals who had previously been on opposite sides in a feud shows just how determined they were to remove William’s power from northern England.37 Jumièges, uncorrected by Orderic in his interpolations, has them proclaim Edgar 33 The main sources are Actus Pontificum Cenomannis in urbe degentium, 376–80; OV, ii, 306–7. See further Latouche, Maine, 35–8. Note the statement in the Actus Pontificum, 376 (Cenomannensium proceres, una cum populo, ab ipsius regis fideliter unanimiter defecerunt). 34 Actus Pontificum Cenomannis in urbe degentium, 377; Regesta, no. 254. 35 Miracula Sanctae Virginis Mariae, ed. Elise F. Dexter (Madison, WI, 1927), 37—8; R. W. Southern, ‘The English Origin of the “Miracles of the Virgin” ’, Journal of Medieval and Renaissance Studies, iv (1958), 176–216, at 178, note 3. For recent comment, Bolton, ‘English Political Refugees at the Court of King Sveinn Ástriðarson’, 25–7. 36 OV, ii, 222–3. 37 Williams, The English, 29–32.

v i c t o ry t r a n s f o r m e d

305

king.38 William’s new commander at York, William Malet, shut himself up in the castle and sent messengers to the king asking for help. Since William moved north in February, it looks as if he had been anticipating a crisis, possibly even in advance of the massacre at Durham. The rapidity of his march was such that he appears to have taken the besiegers by surprise outside York and to have overwhelmed them in a battle near the city. The sources tell us nothing about the execution of this campaign or about the events of the battle. All that can therefore be said is that the rapidity of William’s intervention was typical of his methods, and that a successful land-based attack on York was a considerable achievement. Although the city’s vulnerability was a regular feature of these years, all the successful attacks had been along the Ouse. William’s forces presumably came from the west and forced their way in through the Mickelgate. Although the sources disagree on the scale of the destruction, it also appears that William sacked York and in some way desecrated the cathedral, perhaps by infringing rights of sanctuary in pursuit of his enemies.39 The violence inflicted within the city probably occurred because the troops were allowed to loot and take booty as a way of showing William’s disapproval. The contrast with the earlier protection afforded to Exeter shows both a perception on William’s part that the citizens of York had to be intimidated and punished, and, perhaps, also a change of attitude towards the English in general. William then left for the south after a stay of eight days, leaving William fitz Osbern in charge at York. A second castle was established to the west of the Ouse on the site later known as the Old Baile. Orderic mentions a further attack on the castles while William fitz Osbern was there and that it was beaten off. It did, however, also involve violence within the city.40 Edgar the Ætheling then returned to Scotland.41 An attempt to avenge the killing of Robert de Comines, most likely undertaken on William’s orders, failed when the force sent towards Durham encountered a thick fog near Northallerton that Symeon of Durham attributed to a miraculous intervention by St Cuthbert. He presumably thought the saint’s protectiveness a necessary shield against the likely violence of the revenge that might have been exacted.42 The GND, ii, 180–1. For the suggestion of desecration, Fletcher, Bloodfeud, 177. 40 OV, ii, 222–3. Jumièges and the ‘D’ Chronicle mention violence by William within the city, GND, ii, 180–1; ASC, ‘D’, 1069. The former may conflate the events of this campaign and William’s second one later in the year. 41 ASC, ‘E’, 1069. 42 LDE, 184–5. The Libellus is the only source for this abortive expedition, which its editor suggests may have been invented by Symeon, 185, note 59. An attempt at revenge on William’s part seems likely, however, as, in the circumstances, does the prudent retreat. 38 39

306

william the conqueror

contrast between the execution of this campaign and the Exeter campaign of the year before shows an increased severity and ruthlessness on William’s part that were to be continued over a year until the threat of a successful Danish invasion had been countered. William was back at Winchester for Easter (13 April), as indeed was William fitz Osbern. Two diplomas confirmed there in favour of Bishop Leofric of Exeter and the abbey of Saint-Denis show again the continuation of English diplomatic form. That both were produced by scribes supplied by their beneficiaries, and therefore presumably independently of any direct influence from a royal chancery, is an important demonstration of how widely disseminated the ideology of cultural continuity was; the somewhat eccentric attempts of the scribe from Saint-Denis to write a charter in terms of the norms of English charter production, no doubt guided in his endeavours by the abbey’s former monk, Abbot Baldwin of Bury St Edmunds, are a spectacular illustration of this.43 In showing favour to Leofric and Baldwin, William was again cultivating two churchmen whose support was presumably deemed essential; in Leofric’s case only a little over a year since Exeter’s rebellion. The signa continue the patterns of 1068 diplomas. Matilda is in second place after William, and Archbishop Stigand heads the list of clergy, with the couple’s second son Richard, the only one of their offspring recorded as present, occupying a pre-eminent position in the Saint-Denis diploma. Frenchmen are more in evidence than in diplomas from 1068, but the two English earls Edwin and Morcar, apparently restored to favour, appear in one of them. William was also making grants of English lands to Norman monasteries during this period. Hayling Island (Hampshire) had been given to Jumièges at about the time of its dedication in 1067; and in 1069, in all probability at the Easter court, Harmondsworth (Middlesex) was granted to the abbey of La Trinité-du-Mont of Rouen at the suggestion of William fitz Osbern. When the abbot expressed anxieties about the permanence of the gift, William – in jest, so we are told – made to pierce the abbot’s hand with a knife to show how firmly the grant would stick. This dramatic playacting involving investiture by a knife had a long history going back at least to Carolingian times; it is therefore yet another example of how wellversed William was in the culture of kingly performance. It aimed to

43 Regesta, nos. 138, 254. It is normally assumed that the first of these dates, like the second, dates to Easter 1069 because of similarities between the witnesses. For the Saint-Denis diploma, Thomas Waldman, ‘Charters and Influences from Saint-Denis c.1000–1070’, in Licence (ed.), Bury St Edmunds and the Norman Conquest, 22–30, at 23, 28–9.

v i c t o ry t r a n s f o r m e d

307

demonstrate confidence and certainty. A steady nerve must nonetheless have been required on both sides to pull it off.44 A remarkable episode in the dialogue between William and Archbishop Ealdred most likely also dates from this time. After the new sheriff of Yorkshire had plundered a convoy transporting some of the archbishop’s goods to feed the castle garrison and had abused Ealdred’s representatives who were seeking restitution, the archbishop, so we are told, travelled to London, before approaching the king directly at Westminster.45 Dressed in full archiepiscopal vestments, he reproached him for breaking oaths and cursed him. In response, William instantly fell at Ealdred’s feet and begged to be told his faults. Ignoring his courtiers’ demands that Ealdred be exiled, he listened sympathetically and then ordered that the goods be returned. Ealdred, arguably showing considerable diplomacy by depersonalizing the situation, advised all in attendance that William was prostrate, not before him, but before St Peter, to whom the cathedral at York is dedicated, and that it was right for William to feel St Peter’s power since he had not feared to harm him. This act of dramatic self-abasement has many parallels in the history of medieval kingship and fits well with the ambience of Whitsun 1069, the violence of the previous weeks having made relations between newcomers and natives around York very tense.46 The incident must also be an attempt by Ealdred to demonstrate to William just how volatile the situation in the north had become and that he should respond and do something about it. It was to be the great peacemaker’s last intervention that we know about. Matilda’s return to Normandy soon after the Winchester assembly is explained by Orderic as intended to enable her to pray in tranquillity away from the violence in England. More likely, however, is that her presence in Normandy was needed, and intended to calm fears about events in 44 Regesta, nos. 159, 232. On the performance confirming the Harmondsworth grant, Benjamin Pohl, ‘Translatio imperii Constantini ad Normannos: Constantine the Great as a Possible Model for the Depiction of Rollo in Dudo of St Quentin’s Historia Normannorum’, Millennium: Yearbook on the Culture and History of the First Millennium C. E., ix (2012), 299–341, at 333–4. 45 Chronicon Pontificum Ecclesiae Eboracensis, ii, 350–3 (English Lawsuits from William I to Richard I, ed. R. C. van Caenegem, 2 vols., Selden Society, vols. cvi–cvii [London, 1990–1], i, 1–2, no. 1). On the source, the so-called ‘Digby Chronicle’, see HCY, lvii; Elisabeth Austin, ‘Thomas of Bayeux, Archbishop of York, 1070–1100’, University of St Andrews PhD Thesis (1996), 12–13. The incident could also date to 1068, but 1069 is more likely in the aftermath of the serious warfare of that year. There is no other evidence that William visited Westminster at Whitsun 1069, but it does fit with the famous sequence of crownwearings described in the ‘E’ Chronicle. See further Regesta, 76, 78–9. Freeman, Norman Conquest, iv, 261–2, also preferred 1069. 46 In general, Hamilton, Practice of Penance, 173–90. For some discussion of the incident, Frank Barlow, The English Church, 1000–1066, 2nd edn. (London, 1979), 86–90.

308

william the conqueror

England among families with men on military duties there, a theme that Orderic develops at exactly this point in his narrative.47 In England, the summer was a time of widespread, but ultimately unco-ordinated, risings. A second invasion launched from Ireland by Harold’s sons in command of some sixty ships landed at north Devon in June in the estuary of the Taw, but was defeated by Count Brian and William de Vauville, in all probability at Northam.48 A revolt in Dorset and Somerset was overcome by Bishop Geoffrey of Coutances, leading men from London, Winchester, and Salisbury, and an uprising in Devon succumbed to William’s former opponents, the citizens of Exeter. The disturbances on the Welsh border look to have been the most serious of these local uprisings, with an alliance of Welsh, men from Chester, and Eadric the Wild, besieging and burning Shrewsbury, before fleeing at the approach of a relieving force.49 All these revolts might in some way have been stirred up by awareness of the imminence of the Danish fleet’s arrival. Although no details are available about how widely news of King Svein’s coming intervention travelled, events such as Abbot Æthelsige’s mission would surely have ensured that a lot of people knew about it. In any case, however, in circumstances such as these, emotions, opinions, and behaviour would be swayed by events and would veer wildly. Thus, even in the apparently pacified Thames valley, although Abbot Ealdred of Abingdon had earlier negotiated a favourable settlement in a land dispute with William, the abbey’s tenants subsequently joined in revolts, in all probability in the summer of 1069; their participation was deemed to reflect on Ealdred’s ability to keep his tenants in line and was probably the cause of his deposition in 1071.50 Non-written evidence can sometimes speak loudly too about conditions. The second issue of William’s coinage in c.1068 is another instance of the new regime seemingly fulfilling the normal responsibilities expected of it; the pattern of surviving specimens indicates the same kingdom-wide production as with the first issue, with notably more moneyers who had occurred on Harold’s coins reappearing. But there are variations in the quality of the coin that are not evident in the surviving specimens of the first issue, such as evidence of malpractice by four London moneyers involving the use of altered dies and the production of lightweight coins at London, Lincoln, and some other mints. It is as if a demoralized workforce

OV, ii, 218–21, 222–3. ASC, ‘D’, 1069; GND, ii, 180–3; OV, ii, 224–5; JW, iii, 6–9. See now Nick Arnold, ‘The Defeat of the Sons of Harold in 1069’, Report of the Transactions of the Devonshire Association for the Advancement of Science, cxlvi (2014), 33–56, at 39–49. 49 OV, ii, 228–9. 50 Historia Ecclesie Abbendonensis, i, 226–7. 47 48

v i c t o ry t r a n s f o r m e d

309

was failing to sustain quality.51 In that the rebellions of this time drew on assistance from Ireland and Wales, they also show the wider dimensions of the politics of the British Isles that William would in due course have to take into account. But their defeat again points to how deeply entrenched the power of William’s regime was in many localities of the kingdom. Although specific details are lacking, there is no doubt that some English must have fought on the king’s side on these campaigns. The instability of these times affected victors as well as vanquished. Thus, major allies from the time of the invasion, including Aimeri de Thouars, who had taken the lead in persuading William to be crowned, and Geoffrey de Chaumont-sur-Loire, had returned home to Aquitaine and the Loire valley respectively by 1069 at the latest, taking no part in the takeover of English land. Geoffrey, so we are told, left England richly rewarded. That he continued to play a significant political role in sustaining good personal relations between Count Theobald of BloisChartres and William shows that this must have been the case.52 It is hard to believe that Aimeri, whom Poitiers deliberately set out to praise, was not treated in the same way. Their withdrawal illustrates the multiple costs that William incurred and which, so it seems, he honoured.53 The number of violent deaths among the victors in the campaigns after 1066 would also have made it essential to provide incentives to others to stay. While couching his narrative in terms of standard medieval clichés about insatiable female sexuality, in mentioning several Normans who abandoned England at the behest of their wives who were still in Normandy and did not wish to cross the sea, Orderic is illustrating other demands that incessant campaigns amidst a hostile population were making. In this case, too, William’s first response was to offer enhanced rewards. It was only after these inducements had failed that he resorted to punishment and confiscated their lands.54 Commanded by King Svein Estrithsson’s brother Earl Osbjorn, two (or perhaps three) of his sons, and by others, the Danish fleet travelled up the east coast of England during the summer of 1069, with attempts to land at Dover, Sandwich, Ipswich, and Norwich all being beaten off by local 51 Allen, ‘Mints and Moneyers’, 57, note 30, 63–5; idem, ‘Coinage and Currency’, 91–3. See further, idem, Medieval European Coinage. 9(a). The British Isles, 1066–1279 (Cambridge, forthcoming). 52 Dunbabin, ‘Geoffrey of Chaumont’, 111–13. For Geoffrey leaving with great wealth, see the mid-twelfth-century Gesta Ambaziensium dominorum, in Chroniques des comtes d’Anjou et des seigneurs d’Amboise, 74–132, at 89, which includes a general statement about the lavish rewards given by William. 53 Martindale, ‘Aimeri of Thouars’, 232–4. 54 OV, ii, 218–21.

310

william the conqueror

defenders. Its initial strategy was presumably based on the calculation that there would have been sufficient English support to guarantee a landing. But the way in which they were driven back to their ships when they did land indicates that the coastal defenders were both sufficiently numerous and committed to the new regime to prevent them doing so. Effectively, therefore, forced to make a landing beyond William’s reach, they finally weighed anchor in the Humber, where they were joined by Edgar the Ætheling, Earls Waltheof and Gospatric, Maerleswein, and others, with Waltheof being the most prominent addition to the alliance since the debacle earlier in the year.55 An attempted raid southwards by Edgar the Ætheling, which was presumably intended to expand the territory under his and his allies’ control, was defeated by the garrison at Lincoln, further evidence of the scale of the military presence installed on William’s orders across the English kingdom. But the main Danish army marched north from the Humber and, with local support, occupied York. The Normans stationed in the castles took them on in a pitched battle and were defeated. The ‘D’ Chronicle says that the fleet’s arrival was greeted with rejoicing in the north of England. However, if the estimates in the two versions of the Chronicle of either 240 or 300 ships are anywhere near accurate, then, while formidable, it was most certainly not an invasion force on anything like the scale of William’s in 1066. Closer in size to Harold Hardraada’s army, it was most probably an advance party to establish a base for King Svein’s follow-up.56 The decision by the commanders at York to risk a battle with it, albeit mistaken, must be a sign that they did not think they were facing a numerically overwhelming force. ‘D’ blamed the French defenders of the castles for destruction in the city and in particular for the burning of the cathedral, a statement which John of Worcester amplified by explaining that the fires were intended to clear away houses in order to improve the castles’ defences. John also supplies the clearest chronology of events: the fire took place on 19 September and the victory of the combined English and Danish army over the garrisons shortly afterwards. References to huge amounts of treasure captured in the castles must indicate that William’s commanders had been plundering the local communities, thereby supplying further context for Archbishop Ealdred’s clash with William. In the midst of these unfolding events, Archbishop Ealdred died on 11 September. William of Malmesbury reports that he had formally cursed William for his excessive taxes and unjust seizures of property, that he had 55 The main sources for events in northern England are ASC, ‘D’ and ‘E’; OV, ii, 224–35; JW, iii, 8–11; LDE, 184–5; HRA, 187–8; GR, i, 464–5. There are slight discrepancies between the accounts on the names of the Danish army’s leaders. 56 Cf. Lawson, Battle of Hastings, 147–50.

v i c t o ry t r a n s f o r m e d

311

ceased to regard the king as an honourable man, and that his death was brought on by distress.57 John of Worcester says that he died broken in spirit on learning of the arrival of the Danish fleet.58 Whether it had actually arrived by 11 September – the sources disagree – is an irrelevant detail; the archbishop would have known that it was coming. Ealdred, who had been at the heart of affairs since the 1040s and who had toiled mightily to avoid the very disaster that he must have thought was now taking place, can be viewed as a heroic peacemaker, albeit perhaps a failed one, as a foolish appeaser, or as someone who had worked long and hard for the English kingdom’s benefit. Lives such as his will always be controversial. Malmesbury believed him to have been a brave man, an opinion surely reflected by the range of business with which he was entrusted. His life certainly demolishes any notion that the pre-1066 English Church was isolated from the European mainstream. According to Orderic, William was hunting in the Forest of Dean when he heard the news of the Danish army’s arrival. Since he was still drawing on Poitiers’s narrative, it is likely that this apparent display of sang-froid is another portrayal of William as a cool customer in a crisis, borrowed from Poitiers. The visit to the Forest, the first time he is known to have travelled to the borders of Wales, must have been designed to impress his authority there; the hunting, an expedient use of local facilities. His rapid move north after the fall of York to the Danes and the English rebels indicates that preparations had already been made and that forces had already been mustered; he was presumably waiting to see whether, and where, the Danish army would make a landing. According to Orderic, the rebels abandoned York in alarm at the news of William’s possible arrival and sought refuge in their ships on the Humber and in Lindsey. This reads like another piece of exaggeration taken from Poitiers. More likely, deeming York to be secure, with the garrison massacred, its commander a hostage, and the citizens supportive, their strategy was to deny William access to Yorkshire by holding the line of the rivers Humber, Aire, and Ouse, all of them undoubtedly swollen by late autumnal and winter rains. This would enable them to recruit more English support and to await reinforcements from Denmark, with serious campaigning following in the spring. On the other hand, William had started a process that kept the Danish army confined in the north of England, which was soon to prove decisive. As often, William’s strategy was to try to catch his opponents off balance, thereby to break their hold on the north before consolidation could take place and to prevent an advance south. Some Danes who had advanced 57 58

GP, i, 384–5. JW, iii, 8–9.

312

william the conqueror

into Lindsey were hunted down and killed. Most of their companions then retreated back across the Humber, with some who remained south of the estuary being surprised and slaughtered by a blockading force led by William’s brother Count Robert of Mortain and by Count Robert of Eu. On his own march north, William put down a revolt at Stafford, before moving to Nottingham and trying to cross the river Aire near Pontefract. This proved difficult and he was delayed for three weeks, vetoing the idea of trying to build a bridge on the grounds that this would involve concentrating his forces and making them vulnerable to a surprise attack. After one of his followers had undertaken a reconnaissance, a crossing was eventually forced at an unknown point, which was presumably well to the north of Pontefract; Orderic’s description of the subsequent march to York across difficult terrain suggests that it may have been made at Leeds or even further north. William’s men then held the bridgehead that had been established until the main body of the army arrived and forded the river unimpeded.59 The army then fanned out over the countryside and sought out his enemies, killing them wherever they were found. His victory enabled him to spend Christmas in the burnt city of York. According to Orderic, William sent to Winchester for his crown and ceremonially wore it in York Minster on Christmas Day, a symbolic display of kingship amidst the havoc wrought by war in the chief town of northern England, the seat of the kingdom’s second archbishopric and a place with an outstandingly distinguished religious past. If the York historian Hugh the Chanter’s account of Thomas, the first Norman archbishop, having to organize the re-roofing of the cathedral reflects the state of the church at the time of the crown-wearing, then it was performed in circumstances where stark reminders of the destruction that had brought the solemn occasion into being were all around.60 It certainly took place at the heart of a society, much of which, at least in spirit, was likely still to have been bitterly hostile to William’s rule. York seems to have suffered terribly from these events. Although archaeological excavations have found no evidence of fire in the Coppergate district between the castles and the cathedral, the evidence of Domesday Book indicates that two-thirds of the housing in the city had been made uninhabitable, with major long-term economic and social consequences.61 But, if William took his coronation oath seriously, the crown-wearing also signified that York would be rebuilt. For this suggestion, Fletcher, Bloodfeud, 179. HCY, 18–19. 61 DB, i, 298r; Rollason et al., Sources for York History, 183–4. For strikingly different emphases in the impact on York, Rees Jones, York: The Making of a City, 89–90; Palliser, Medieval York, 89–91. 59 60

v i c t o ry t r a n s f o r m e d

313

By pressing home the attack as he had done, William had prevented the Danes and the English rebels from consolidating their grip on York and its environs in advance of Svein Estrithsson’s anticipated arrival. Their reluctance to fight a battle meant, however, that they were confined to their ships on the Humber, where they must have deemed themselves safe because William had no fleet at his disposal.62 While Orderic was probably copying Poitiers in describing a decimated army hemmed in by a superior opponent and scarcely able to feed itself, this assessment surely does capture very well the central aspects of the strategic situation.63 Having probably started with the aim of taking control of most of the north of England, the Danes found themselves in very different circumstances. They had been driven into a proverbial corner. According to John of Worcester, Earl Osbjorn agreed to leave England in the spring of 1070 when William paid him a large sum of money and allowed him to feed his troops by ravaging the coastal lands.64 If so, William was taking the most expedient route to send the fleet on its way as quickly as possible. John criticized Osbjorn’s greed and thought his action a disgrace. THE ‘HARRYING OF THE NORTH’ The most detailed narrative of the first weeks of 1070, namely Orderic’s, says that William’s army resumed the systematic devastation of the countryside after Christmas and that he led it to the Tees where he stayed for fifteen days. There Waltheof and Gospatric submitted; the former in person, the latter through intermediaries. Edgar the Ætheling returned to Scotland and the protection of Malcolm (Mael Coluim) III. Although the account is confused on the geography of the far north of England, it is possible that Orderic’s statement that William marched his army as far as Hexham in late January, before returning south, should be accepted.65 Whatever the case, the march north from York completed the defeat of the English revolt. Orderic also makes it clear that throughout this time William ordered his men to destroy everything. How far Orderic was using Poitiers for any of this is a moot point. He certainly departed from him to

62 ASC, ‘D’, comments that ‘the fleet lay all the winter in the Humber where the king could not get at them’. 63 JW, iii, 10–11; OV, ii, 232–5. 64 JW, iii, 10–11. 65 OV, ii, 230–5. For arguments in favour of William’s army travelling to Hexham, David Palliser, ‘Domesday Book and the “Harrying of the North” ’, Northern History, xxix (1993), 1–23, at 5. I am grateful to George Hope and his colleagues at the Scottish Environment Agency for an extremely helpful discussion of the Harrying of the North that has reinforced my opinion that a new kind of discussion is needed.

314

william the conqueror

write of helpless children, the young and the old, dying of starvation. God, he thought, would surely not allow such slaughter to go unpunished. In relation to these much discussed events, William of Malmesbury once more showed himself to be the worldliest of the twelfth-century historians in his appreciation that both political and military factors were involved. Hence, he explained them in terms both of William’s anger at persistent revolts and as an attempt to deny access to the region to the anticipated further Danish invasions.66 But in terms of the values that he believed appropriate to interpreting past events, what happened was so terrible that he decided in his Gesta Pontificum that he must correct a prophecy of Charlemagne’s spiritual adviser, Alcuin of York, who had believed that a terrible disaster would afflict the people of the north of England during his lifetime because of their sins. The disaster had indeed happened, so Malmesbury opined, but more than two hundred and fifty years later than the great man had predicted.67 He went on to say that the effects of the devastation were still evident in his own day, that is, half a century after it had taken place. The many modern discussions of these events have until recently relied on trying to bring the literary sources and Domesday Book into alignment with one another. It is important therefore that, even within an argument that the literary sources exaggerate the scale of the destruction and Domesday’s description of an estate as ‘waste’ is not always a reference to destruction inflicted by William’s armies, what happened has been described as ‘catastrophic’.68 A subsequent restatement of the view that ‘waste’ did describe land devastated by William’s armies must presumably make the destruction even more terrible.69 When material culture, viewed through the evidence of the architecture of parish churches and other objects, is taken into account, a more nuanced picture does start to emerge, one that puts the emphasis on long-term process, human resilience, and the determination of the surviving indigenous inhabitants and the newcomers to rebuild a disrupted economy.70 Such an assessment also brings into play the necessity of undertaking an environmental analysis that shifts the focus of the debate.71 GR, i, 462–5. GP, i, 326–7. 68 Palliser, ‘Domesday Book and the “Harrying of the North” ’, 20. 69 John Palmer, ‘War and Domesday Waste’, in Matthew Strickland (ed.), Armies, Chivalry and Warfare in Medieval Britain and France, Harlaxton Medieval Studies, new series vii (Stamford, 1998), 256–75. 70 Aleksandra McClain, ‘Rewriting the Narrative: Regional Dimensions of the Norman Conquest’, in Dawn Hadley and Christopher Dyer (eds.), Transformations and Continuities in the Eleventh Century: the Archaeology of the Norman Conquest (London, forthcoming). 71 This point is well made in David Roffe, Decoding Domesday (Woodbridge, 2007), 251–5. 66 67

v i c t o ry t r a n s f o r m e d

315

The verdict on the events in the literary sources is clear-cut. The ‘E’ Chronicle says simply ‘And King William went into the shire and ruined it completely’.72 Of other writers, John of Worcester wrote of the surviving inhabitants resorting to desperate methods to survive, including cannibalism.73 Symeon of Durham described the Durham clergy as being so fearful that they fled on 11 December 1069 to Lindisfarne, taking with them the body of St Cuthbert, returning only on 25 March 1070.74 The Durham Historia Regum of the 1120s states that land remained uncultivated for nine years and that no village between York and Durham was inhabited; it also mentioned decomposing unburied corpses.75 Other historians writing in Yorkshire tell a similar story. Thus, Hugh the Chanter tells how William’s armies destroyed the whole district around York, and later, Alfred of Beverley, writing in the 1140s, states that William’s forces had ravaged as far as the coast, but had left Beverley untouched.76 And at Evesham, the late eleventh-century chronicle, which was the basis for Thomas of Marlborough’s early thirteenth-century one, describes the arrival at the abbey of refugees, including large numbers of children, seeking food and hospitality. It adds that many died after their arrival because of trying to eat after long-term food deprivation.77 The Harrying of the North also attracted notice beyond Normandy and England, with Marianus Scotus, the author of a Universal Chronicle written at Mainz before 1076, apparently inserting a marginal note alongside his annal for the year 1070 referring to famine and cannibalism in England caused by the devastations of the French and the Scots.78 These writers’ collective testimony is so unanimous that it must surely be accepted in its broad outlines, albeit not in every detail. Thus, when Orderic writes of 100,000 people dying of starvation, it can reasonably be treated as another inflated statistic typical of medieval writers. And, likewise, even if his account is devoted entirely to Abbot Æthelwig’s works of charity, Thomas of Marlborough might be thought to have had a supplementary purpose of using the arrival of the refugees as the basis for his account of his monastery’s role in the subsequent reconstruction of monastic life in the north. It is, however, totally unacceptable to think ASC, ‘E’, 1069. JW, iii, 10–11. For the literary sources in general, Palliser, ‘Domesday Book and the “Harrying of the North” ’, 3–5. For a conveniently assembled translation of some of them, Rollason et al., Sources for York History, 188–90, 195–6. 74 LDE, 184–9. The episode is discussed in ibid., 184–7, notes 60–3. 75 HRA, ii, 188. 76 HCY, 2–3; Aluredi Beverlacensis Annales, ed. Thomas Hearne (Oxford, 1716) 129. 77 Thomas of Marlborough, History of the Abbey of Evesham, 166–7. 78 Marianus Scotus, Chronicon, ed. G. Waitz, MGH Scriptores, v (1884), 481–62, at 559–60; cited in van Houts, ‘The Norman Conquest through European Eyes’, 839, note 2. 72 73

316

william the conqueror

either that exaggerations or special interests such as these undermine their essential veracity or to brand them as ‘late’. The insertion of social memory into the analysis means that they become in many respects as authoritative as something written contemporary with events.79 They must be treated as snapshots that focus on what was done in the winter of 1069–70 and not as analytical assessments of the region’s later eleventh- and early twelfthcentury history. And it has to be very important that the theme of famine and starvation appears in almost all of them.80 For all the reservations that can be made about them, when it comes to the raw statistics that can be extrapolated from Domesday, many are truly chilling. Thus, to select two examples from many: ‘the county had a deficit of roughly 75% of the plough teams and population it might be expected to possess in normal times’; and ‘at a conservative estimate, more than 80,000 oxen and 150,000 people were needed to make good that deficit’.81 These – be it noted – are figures that describe the situation sixteen years later. In addition, while the word wasta that occurs so often in the Yorkshire folios of Domesday does not always have to refer to land destroyed by war, since etymologically wasta means deserted land and the causes of desertion need not be ravaging, the coincidence of the word’s use with areas known to have been devastated by armies cannot be accidental, especially as this is so not just in Yorkshire (where the great majority of references do occur) but elsewhere. As such, the usage is probably a convincing demonstration of what the mind that oversaw the writing of the monumental text wanted to convey.82 The argument that wasta was used in relation to estates that had not been surveyed, or where the material available was judged inadequate as the final text was compiled, does have the great merit of highlighting that the survey and its results were far from an administratively perfect operation.83 But in the context of the Harrying of the North, it is also a deceptively circular argument. It does not rule out the strong likelihood that the population and resources of many estates could not be recorded because there was little, or often nothing, there to record. It can also almost go without saying that other armies devastated Yorkshire and the north of England between 1066 and 79 Thus Palliser, ‘Domesday Book and the “Harrying of the North” ’, 4 (‘There is furthermore the difficulty that all or almost all the chroniclers’ accounts are sub-contemporary rather than contemporary’). 80 For this point and several arguments similar to mine, Sean McGlynn, By Sword and Fire: Cruelty and Atrocity in Medieval Warfare (London, 2008), 200–8. 81 Palmer, ‘War and Domesday Waste’, 273. 82 Ibid., passim. 83 David Roffe, ‘Domesday Book and Northern Society: A Reassessment’, EHR, cv (1990), 310–36, at 334; cf. Harvey, Domesday, 217–18.

v i c t o ry t r a n s f o r m e d

317

1086 – this chapter has included mention of two invasions from the north, three interventions by William, the brief Danish occupation of the Humber estuary and of York, and the installation of garrisons that needed to be provisioned – but none behaved with the systematic ruthlessness characteristic of William’s campaign. The arrival of more armies in the fifteen years after the winter of 1069–70 further perpetuated its impact. Crucial to any interpretation of the Harrying is to avoid treating it simply as a military campaign. Not only Malmesbury, but also Orderic, insert it into a framework of royal anger, that is, of a political act which aimed to demonstrate that certain behaviour and attitudes lay beyond the pale and had to be punished. Even in describing William’s actions as shameful, Orderic consistently uses the technical word ira as the determinant of his behaviour.84 On these terms, from William’s perspective, it was a continuation of his treatment of the south-east of England after the victory at Hastings, and also conduct that had long been a part of the armoury of kingly rule and was to remain so for a long time afterwards. For all this, a comparative study of such violence has observed that ‘Only very rarely, as in the case of William the Bastard’s “harrying of the north”, was the ravaging of a region taken to the point of starving non-combatants to death’.85 That this assessment is based on the evidence of a literary source is mostly by the by. Depopulation was an identified objective of other ravaging, notably Charlemagne’s, and, since what we know of William’s ravaging of Yorkshire not only created a refugee problem, but also conditions of long-term deprivation, it must have been so in that case as well.86 Once more, so it seems, we are dealing with conduct that both conformed to contemporary norms in a general way, but went beyond them in its specific modus operandi. It is also significant that the ravaging was carried out in the winter. Even a small number of horses ridden over muddy fields at that time would have destroyed their productivity for much more than one year. The Domesday evidence for the drastic reduction in the number of plough teams and the deficit of 80,000 oxen in 1086 indicates just how profound the long-term problem was, because oxen are hard to replace, have to be bred, and then have to grow before they can be harnessed to a plough. They were also crucial to the provision of manure. Thus, on the basis that very few crops would be harvested in 1070 and that an ox takes three years to grow before 84 Thus, in the crucial passage (dum iram suam regere contempsit, et reos innocuosque pari animaduersione peremit. Iussit enim ira stimulante segetibus et pecoribus cum uasis et omni genere alimentorum repleri, OV, ii, 230–3), the translation of the two words ira in the first case as ‘fury’ and then as ‘anger’ is misleading. 85 Gillingham, ‘Women, Children and the Profits of War’, 61, note 4. 86 Thus, for example, Annales Regni Francorum, 21, 22.

318

william the conqueror

it can start to be trained to pull a plough, no recovery of any kind could possibly have begun before 1075.87 Studies of other periods for which more material is available have shown just how devastating in the long term the destruction of the landscape by ravaging armies could be. Even where a relatively minimal case is being made, land that has been ravaged can always be distinguished from that which has not. And in any case, a distinction always needs to be made between land affected by warfare and land deliberately devastated to destroy livelihoods.88 The long-term impact on the population’s health of homelessness and the consequent poor nutrition would also have prevented recovery, especially as the statistics we possess surely indicate the removal of the very resources that were indispensable to it. With the rhythms and sensitivities of the agricultural cycle being very familiar to almost everyone, it is important to spell out that William and those responsible for the Harrying of the North would have known very well what the consequences of their actions were going to be. The location of so many wasted estates either on the flank of the Pennines, in the uplands of the Pennines, in the far north of the county, on the coast, or along the Humber, has over the years seemed to be a perplexing interpretative problem; in particular the devastation of the upland estates.89 Armies did not need to go there. Arguments as ingenious as the one that these vills had been depopulated in order to restock the more productive lands in the Vale of York have even been developed.90 A more likely explanation, however, is that the lowlands were devastated, whereupon many of the peasant population then fled into the uplands where they were hunted down, thereby destroying their plans to recolonize the lowlands. This fits with much in Orderic’s narrative, especially the statements that William travelled to mountainous places, cut down many, and burned homes. The destruction in the North Riding, on the coast, and on the Humber, is readily explicable in military as well as political terms, the specific strategic purpose being to deter further interventions by the Scots and the Danes, interventions by both of whom were with good reason expected to happen again soon. 87 Debby Banham and Rosamond Faith, Anglo-Saxon Farms and Farming (Oxford, 2014), 108–10. 88 Thus, e.g., Richard Lomas, ‘The Impact of Border Warfare: The Scots and South Tweedside, c.1290–c.1520’, Scottish Historical Review, lxxv (1996), 143–67; Philip Slavin, ‘Warfare and Ecological Destruction in the Early Fourteenth-Century British Isles’, Environmental History, xix (2014), 528–50, at 539–45. 89 H. C. Darby and I. S. Maxwell (eds.), The Domesday Geography of Northern England (Cambridge, 1962), 65–7, 144–5, 217. 90 Thus T. A. M. Bishop, ‘The Norman Settlement of Yorkshire’, in R. W. Hunt et al. (eds.), Studies in Medieval History presented to Frederick Maurice Powicke (Oxford, 1948), 1–14, at 2, 3.

v i c t o ry t r a n s f o r m e d

319

The signs of strong recovery on many estates by 1086 and the prosperity of some English landholders are not in fact much of an argument against the destruction having been a very severe one.91 What they do demonstrate is the speed and the determination with which the conquerors and the remaining inhabitants set about reorganizing the north. Although most of the subject is outside the scope of this biography, it is important to note observations that directly link subinfeudation to the recovery of estates, the presence of stylistic forms typical of the ‘Saxo-Norman overlap’ in surviving churches, and the considerable number of English tenants on the most northerly of the territorial blocs created in the 1070s, Count Alan’s honour of Richmondshire.92 The tendency among scholars to date the new lordships earlier than previously logically creates the expectation of an earlier start to the process of recovery.93 Although its chronology cannot be pinned down very precisely, an indicator of large-scale investment is the construction of the region’s first major monastic site at Lastingham in c.1080, a place that began to be abandoned as a monastery before 1086 when the community moved to St Mary’s, York, in part apparently because of the continued lawlessness in the region consequent on the Harrying.94 The foundation of St Mary’s, in which William was a significant participant, was part of a major personal intervention in the region that came late in his reign.95 These developments are indicative of a determination that York and Yorkshire would be rebuilt, and that the region should be made defensible against future rebellions and invasions. The slowness of the recovery is, however, a clear sign of the ruthlessness of the destruction wrought in the winter of 1069–70. Implausible as it may at first sight seem, William may well have been involved at this time in the beginnings of what became the monastery at Selby, on the river Ouse to the south of York. Although no source says so, it is possible that penitential atonement was an influence on his thoughts. The narrative in the abbey’s Historia, completed by 1174, tells of the sheriff of York, Hugh fitz Baldric, sailing on the river with an armed patrol, and 91 Dalton, Conquest, Anarchy and Lordship, 19–78; Hugh M. Thomas, ‘A Yorkshire Thegn and His Descendants after the Conquest’, Medieval Prosopography, viii (1987), 1–22, at 4–6. 92 McClain, ‘Rewriting the Narrative’, forthcoming; K. S. B. Keats-Rohan, ‘A Question of Identity: Domesday Prosopography and the Formation of the Honour of Richmond’, in David Roffe and K. S. B. Keats-Rohan (eds.), Domesday Now: New Approaches to the Inquest and the Book (Woodbridge, 2016), 169–96, at 190–3, 196. 93 Dalton, Conquest, Anarchy and Lordship, 19–78, with a succinct statement of conclusions at 77–8 (‘And if the speed of the conquest has been underestimated, so has the level of control exercised over the process by the Conqueror’). 94 Stuart Harrison and Christopher Norton, ‘Lastingham and the Architecture of the Benedictine Revival in Northumbria’, ANS, xxxiv (2012), 63–103, at 66–71. 95 Rees Jones, York: The Making of a City, 98–100.

320

william the conqueror

discovering a small community established at Selby by a monk of the abbey of Saint-Germain of Auxerre named Benedict. Benedict, so we are told, had travelled by sea along England’s east coast and had found the site he was seeking on the Ouse; he was at prayer when Hugh found him. Hugh then introduced him to William, who gave permission for a monastery to be founded.96 With Selby tradition dating the abbey’s foundation to 1069–70 and the Historia setting it into a context of the violence disturbing the north of England, it is possible that William did give his consent to the establishment of a small community in the midst of the military campaigns of these times. The role of Hugh fitz Baldric becomes plausible if it is accepted that he took over from William Malet as sheriff around Christmas 1069 after the latter’s harsh experiences. And Benedict had presumably arrived in south Yorkshire before the Danish fleet had established itself in the Humber. Although the Selby Historia does concertina events to produce a tale of unbelievable early prosperity, an image that is reinforced by a forged foundation charter that also dates from the mid-twelfth century, but which is contradicted by Domesday Book that shows only a tiny endowment in existence in 1086, the slow steady development that actually took place was typical of such institutions.97 The legends that later developed which have Queen Matilda present and the future Henry I born at Selby, neither of them in the Historia, must be beyond the bounds of belief.98 It is possible that some of William’s involvement with the Selby community dates to the second of his two visits to northern England in 1072. But the story ultimately fits into the pattern of attention to the minutiae of rule that is observable in the evidence for his times in England. The equally wellknown history of Reinfrid, the soldier in William’s army who gazed at the monastic remains at Whitby and, sickened by what he had witnessed during the Harrying, vowed to become a monk and subsequently played a central part in the revival of monasticism in the north, is one of the great life histories of these times.99 It shows how native and newcomer mixed from almost the earliest times. Reinfrid might also be said to epitomize the central paradox of William the Conqueror’s biography. Would Reinfrid 96 For William’s and Hugh fitz Baldric’s roles, Historia Selebiensis Monasterii: The History of the Monastery of Selby, ed. and trans. Janet Burton with Lynda Lockyer, OMT (Oxford, 2013), 38–43. 97 Selby’s early history is critically examined in ibid., xxix–xxxviii. For the forged charter, Regesta, no. 272; Historia Seleberiensis Monasterii, 50–3. 98 Green, Henry I, 20, is decisive on this. Hollister, Henry I, 32, does not entirely rule it out (‘a possibility, albeit a remote one’). 99 Cartularium Abbathiae de Whiteby, ed. J. C. Atkinson, 2 vols., Surtees Society, lxix, lxii (1879–81), i, 1–2. For Reinfrid’s life and significance, Janet Burton, The Monastic Order in Yorkshire, 1069–1215 (Cambridge, 1999), 13–14, 32–9.

v i c t o ry t r a n s f o r m e d

321

have prospered if William had not restored order after the revolts of 1069 in the way he did? William’s return to the south of England is presented by Orderic as a heroic march through dreadful weather. Crossing the Pennines, he led his army to Chester, suppressing revolts in Mercia on the way. He built castles at Chester and Stafford and finally disbanded his army at Salisbury. Very obviously copying Poitiers at this point, Orderic portrays the Angevins, Bretons, and Manceaux in William’s army as quarrelsome and cowardly, and William as refusing to be distracted by conduct verging on the mutinous, except to promise special rewards to those who followed him loyally. He is said to have marched with the foot soldiers on the most treacherous parts of the journey.100 The Domesday Book evidence for values, people, and oxen in Cheshire indicates that the destruction there was not on the scale of that inflicted on Yorkshire.101 Given, however, the arguments about the length of the recovery period of devastated lands set out above for Yorkshire, the evidence for a strong revival in Cheshire does not rule out the Evesham Chronicle’s statement that refugees from there travelled to Evesham and from other shires through which William’s army had marched.102 But, viewed comparatively, it surely does demonstrate just how terrible was the destruction inflicted on Yorkshire. The decisive point was undoubtedly that William had made his presence felt there as well and had given a display of symbolic and actual disapproval that, according to contemporary norms, was intended to intimidate and punish. At Salisbury, William paid off the loyal soldiers first and kept back those who had threatened to desert for a further month. In making what is in effect a distinction between the Normans and other ethnic groups, with the possible exception of the Flemings, the text that Orderic surely copied from Poitiers is also underlining how William’s choices came increasingly to focus on the Normans as a group he could rely on and therefore reward. 1068–70: AN ASSESSMENT Viewed solely as a military achievement, what William accomplished in the last months of the year 1069 and the first months of 1070 was truly remarkable. All his opponents in England had either been outmanoeuvred or defeated. He had held armies together in circumstances where OV, ii, 234–7. J. S. Matthews, ‘William the Conqueror’s Campaign in Cheshire in 1069–70: Ravaging and Resistance in the North-West’, Northern History, xl (2003), 53–70, at 59–63. 102 Thomas of Marlborough, History of the Abbey of Evesham, 166–7. 100 101

322

william the conqueror

demoralization could easily have set in. He had retained the psychological advantages of someone who kept winning, who rewarded generously, and who paid his troops. He showed concern for their material and spiritual welfare and this stilled much opposition and anxiety. He quickly took over places of strategic importance and ensured that they strengthened his hold on his kingdom. His continuing attention to Normandy since 1066 must also have reassured, since many people were now, like him, living crossChannel lives. Out of his achievements and the ideology of continuity a peaceful and prosperous kingdom and a cross-Channel empire emerged. It was also a vehicle enabling William to claim to occupy the moral high ground, abstractions that could be used to justify the coercive enforcement of rule. In terms of the ongoing debate about the ethics of William’s methods, however, the scale of the state-sponsored violence that all this involved is arguably deeply shocking. The policy of taking over strategic centres and granting them to followers from Normandy and France meant that there was never any possibility of French and English having equal roles within the English kingdom under William’s rule. The creation of new infrastructures around blocs of territorial power that was under way following William’s return from Normandy in December 1067 led immediately to changes in the societies of well-nigh every locality in the English kingdom. The taking of tax and tribute to support armies brought in a level of state exploitation unknown since the days of Æthelred and Cnut, if then. This and the destruction of houses for castle sites meant economic decline for many, and, for some, ruin. The violence against women that had been a feature of the 1066 campaign looks to have continued into this period. The nature of this phase is highlighted by the contrast between the Penitential Ordinance including rape as one of the offences for which penance could be done and the statement in the famous obituary of William in the ‘E’ Chronicle that rape was punished by castration. The latter stated what looks to have been English law before and after 1066; the former deals with conditions in a time of war.103 In one of the most telling of Orderic’s amendments to Poitiers, he replaced the latter’s statement that women were safe from violence by writing that ‘noble maidens were exposed to the insults of lowborn soldiers and lamented their dishonouring by the scum of the earth’.104 As we have seen, there were circumstances in which the society of these CS, 584; ASC, ‘E’, 1087; Hudson, The Oxford History of the Laws of England, 401, 722. GG, 158–9; OV, ii, 268–9. For this whole subject, Elisabeth van Houts, ‘Intermarriage in Eleventh-Century England’, in Crouch and Thompson (eds.), Normandy and its Neighbours, 237–70, at 251–4. 103 104

v i c t o ry t r a n s f o r m e d

323

times could regard violence against women as justified. The issue of English women being forced into marriages by newcomers and taking refuge in nunneries was one that preoccupied William and Lanfranc long after these years.105 The ethics of state-sponsored violence as a means to an end is a thread that runs through these first years and many other periods of medieval warfare. What was illegal in normal times must sometimes have been permitted in this early period of William’s reign. In drawing attention specifically to the effect of the redistribution of lands as a cause of English resentment, Orderic sensitively, if rather naively, laid the blame on the greed of William’s Norman and French followers and censured the king for giving in to them; it is actually a sign of how the culture of reward and the expectation thereof could not be reconciled with giving adequate protection to the English.106 Malmesbury’s assessment that William’s treatment of the English followed a policy of ‘divide and rule’ articulates a similar version of the same phenomenon.107 Both are effectively saying that the exercise of power in ways that aimed to complete the establishment of William’s kingship through conquest took precedence over all other objectives. When it comes to the twelfth-century histories, the theme that runs through them, from the Brevis Relatio through Malmesbury to the Dialogus de Scaccario, is that William never trusted the English.108 This did not mean that some English could not be treated favourably – as indeed some were – but it did mean that factors were at work that limited how far such favour would go. And some who were retained, most obviously Edwin, Morcar, Waltheof, and Edgar the Ætheling, were also greatly diminished. The operation of discretionary favour and disfavour according to a framework established before the coronation meant that many among the English were at William’s mercy. The case of the hapless Abbot Brand of Peterborough shows what could be involved. The balance of favour weighed heavily towards those whom William trusted or, sometimes, towards those to whom he felt a responsibility to reward. Wars notwithstanding, the creation of a new infrastructure of kingly rule proceeded apace. William showed generosity towards Edward the Confessor’s religious foundation at Westminster with an exchange that made Windsor into a royal possession, but which compensated the abbey with properties of much greater value; a narrative inserted into a twelfthcentury forged charter indicates that this was done so that William could

Lanfranc Letters, no. 53. OV, 214–17; 222–3. 107 GR, i, 470–1. 108 Brevis Relatio, 33; GR, i, 470–1; Richard fitz Nigel, Dialogus de Scaccario, 82–3. 105 106

324

william the conqueror

hunt in the woods nearby.109 With Windsor on the route between the crown-wearing centres of Winchester and Westminster, it looks as if the landscape was being adjusted to suit the new regime’s needs at the same time that the proclaimed continuity from King Edward’s time was being foregrounded. In relation to the needs of security and domination, the building within the castle site in London established in 1067 of the White Tower, now securely dated to this early period, was started. Modelled as it was on the great Norman tower of Ivry-la-Bataille and therefore a direct importation of a Norman model, it was a massive demonstration of a new presence in a crucial place (Plate 16 (a) and (b)).110 The same processes were also at work in the other great city of southern England, Winchester, where in c.1070 the enclosure of the royal palace was doubled in size by taking in the cemetery of one of the two great Winchester monasteries, the New Minster.111 Also, although the contemporary evidence for this is almost non-existent, the extensive restitutions of Winchester cathedral estates said to have been taken away by William, accomplished in 1136 by Bishop Henry of Winchester from his recently crowned brother King Stephen, may well indicate other means by which the churches of the city were forced to contribute to its development as a centre of royal power.112 New Minster also claimed in a twelfth-century forged charter that William had inflicted extensive damage on the monastery.113 And another major building that is now dated to this time is the keep of Colchester castle. Modelled on the White Tower and Ivry-la-Bataille, its strategic significance is presumably because it occupied a site on the east coast threatened by invasions from Denmark along the shortest sea route. That it seems never to 109 Regesta, nos. 290, 299; Barbara Harvey, Westminster Abbey and its Estates in the Middle Ages (Oxford, 1977), 27–8; Emma Mason, Westminster Abbey and its People, c.1050–c.1216 (Woodbridge, 1996), 16. 110 The latest arguments for an early dating for the White Tower based on dendrochronology are summarized in Roland B. Harris, ‘Recent Research on the White Tower: Reconstructing and Dating the Norman Building’, in Davies, Riley, Levesque, and Lapiche (eds.), Castles and the Anglo-Norman World, 177–89. See also Impey, ‘London’s Early Castles’, 18–22. 111 Biddle (ed.), Winchester in the Early Middle Ages, 292–305; idem, ‘Early Norman Winchester’, in J. C. Holt (ed.), Domesday Studies: Papers Read at the Novocentenary Conference of the Royal Historical Society and the Institute of British Geographers, Winchester, 1986 (Woodbridge, 1987), 311–31, at 315–16; Regesta, no. 344. I am very grateful to Sophie Rixon for a careful examination of this Winchester material and a discussion. 112 Regesta Regum Anglo-Normannorum: vol. iii, Regesta Regis Stephani ac Mathildis Imperatricis ac Gaufridi et Henrici Ducum Normannorum, 1135–1154, ed. H. A. Cronne and R. H. C. Davis (Oxford, 1968), nos. 945–9; Alexander R. Rumble, ‘The Purposes of the Codex Wintoniensis’, ANS, iv (1982), 153–66, 224–32, at 158–60; Edmund King, ‘Henry of Winchester: The Bishop, the City, and the Wider World’, ANS, xxxvii (2015), 1–23, at 7–8. 113 Alexander R. Rumble, Property and Piety in Early Medieval Winchester (Oxford, 2002), 178–9.

v i c t o ry t r a n s f o r m e d

325

have been completed probably reflects the changing circumstances from the 1070s onwards once it started to become clear that the threat posed by the Danish kings was not as great as initially thought.114 But Colchester, like London, had extremely powerful Roman connotations, with the castle being built on the site of the Temple of Claudius.115 A final major project under way by this time was the Noua Foresta, the New Forest, a dating confirmed by the fact that William and Matilda’s second son Richard died there, perhaps in c.1070, and certainly not much later. With the forest as a legal concept that defined land specifically reserved for hunting now seen not as an innovation brought in from Normandy, but rather as an evolution out of pre-1066 English practices adjusted by the royal will, the New Forest becomes not just a special place for hunting created by William, but an element in the infrastructure of royal and crossChannel rule.116 The Noua Foresta – in so describing it, Domesday Book is identifying innovation not as a concept, but as an expansion of the forest area so great as to need to be taken notice of – has recently been seen as the equivalent of the Sussex rapes, that is as an area where previous patterns of landholding were overridden for strategic purposes by the royal will.117 On these terms, it was not just a convenient place for William to hunt as soon as he had disembarked, but was an integral element in the English triangle of Winchester, Abingdon/Windsor, and Westminster/London on which William’s English rule was founded; and also on the cross-Channel routes linking the Solent to Caen and Rouen. William, his family, and his court could hunt there either almost immediately on landing from France or when waiting for good weather to cross. As far as the Church in England was concerned, the approach to St Hugh, abbot of Cluny, for – dependent on which source is accepted – either six or twelve monks to assist in the reform of the English Church, for whom William offered to pay £100 of silver for each monk, looks to date 114 Peter Berridge, ‘Colchester Castle: “Some Tyme Stronge and Statelye, as the Ruynes Do Shewe” ’, in Davies, Riley, Levesque, and Lapiche (eds.), Castles and the Anglo-Norman World, 55–68, at 55–6, 67; Impey, ‘The Ancestry of the White Towers’, in idem (ed.), The White Tower, 227–41, at 239–42. 115 For this idea fully explored, T. A. Heslop, ‘Constantine and Helena: The Roman in English Romanesque’, in Jill A. Franklin, T. A. Heslop, and Christine Stevenson (eds.), Architecture and Interpretation: Essays for Eric Fernie (Woodbridge 2012), 163–75, at 170–2. 116 For the general arguments, Dolly Jørgensen, ‘The Roots of the English Royal Forest’, ANS, xxxii (2010), 114–28, at 123–8; Graham Jones, ‘A “Common of Hunting”? Forests, Lordship and Community before and after the Conquest’, in John Langton and Graham Jones (eds.), Forests and Chases of England and Wales c.1000–1500 (Oxford, 2010), 36–67, at 48–61; Judith A. Green, ‘Forest Laws in England and Normandy in the Twelfth Century’, HR, lxxxvi (2013), 416–31, at 416–25. 117 DB, i, fo. 51rv; Karin Mew, ‘The Dynamics of Lordship and Landscape as Revealed in the Domesday Study of the Nova Foresta’, ANS, xxiii (2001), 155–66, at 162–5.

326

william the conqueror

from this period. Although this story has been the subject of different datings and interpretations, it is logical to assign it to the early years of the reign because William’s purpose then would have been to have monks to help shape the Church in England; at a late date, such as the later 1070s, this would not have been necessary. So viewed, it also fits with William’s general dealings with Pope Alexander II.118 Hugh’s response that his monks were not for sale and his pastoral advice to William to the effect that in gaining the world he should not lose his soul (Matthew 16:26) are said to have angered William at first, but to have been seen subsequently by him as wise.119 The writs of these times, with the exception of one for the small abbey of Abbotsbury, were for institutions associated with William’s supporters, namely, the abbeys of St Augustine’s, Canterbury, Bury St Edmunds, and Westminster, and the cathedral of York.120 Although there are issues relating to possible distortions resulting from arbitrary documentary survival involved in saying this, the picture created is of a regime operating selectively, partly on the basis of limited, albeit powerful, support and favour, and partly through making strategic choices that were potentially to its profit. The corollary is that they represent a regime still lacking kingdomwide reach and unconfident of its capacities. That the initiative for the confirmation of many such documents came from their beneficiaries points to a mutually advantageous relationship that was also significant in shaping the kingdom’s legal future.121 Their consistent representation of William as King Edward’s kinsman suggests that this was how the case for William’s succession was viewed at that time by those with a vested interest in sustaining it.122 One of Poitiers’s contributions to the debates about William’s conduct encompassed a historic perspective by comparing it to Cnut’s.123 William, so he argued, had tried to rule with his English subjects, whereas Cnut had 118 For the approach to Cluny, H. E. J. Cowdrey, ‘William I’s Relations with Cluny Further Considered’, in Judith Loades (ed.), Monastic Studies: The Continuity of Tradition (Bangor, 1990), 1–14, at 6–9 (reprinted in idem, The Crusades and Latin Monasticism, 11th–12th Centuries (Aldershot and Brookfield, VT, chapter 8); cf. Frank Barlow, ‘William I’s Relations with Cluny’, JEH, xxxii (1981), 131–41, at 137–41 (reprinted in idem, The Norman Conquest and Beyond [London, 1983], 245–56, at 252–5), where the request for monks is associated with Abbot Hugh’s visit of 1078x1080. 119 For Hugh’s letter, ‘Memorials of Abbot Hugh of Cluny (1049–1109)’, ed. H. E. J. Cowdrey, Studi Gregoriani, xi (1978), 11–175, at 143–4. 120 The writs that can certainly be assigned to this period are Regesta, nos. 1, 11, 12, 31, 32, 34–8, 80, 91, 292, 295, 299, 351. 121 See especially, ibid., nos. 1, 12, 31, 34–8, 80, 98, 351; Richard Sharpe, ‘The Use of Writs in the Eleventh Century’, ASE, xxxii (2003), 247–91, at 283–9. 122 Regesta, nos. 1, 34, 36, 38, 80, 292, 351. 123 GG, 156–7.

v i c t o ry t r a n s f o r m e d

327

delayed his coronation, and therefore the responsibilities that it entailed, and killed people. For all that the argument is factually inaccurate – the date of Cnut’s coronation is unclear – it did contain elements of accuracy: Cnut had certainly had English aristocrats killed.124 William of Malmesbury also picked up the theme and pointed out that Cnut’s conduct had led to a collaborative relationship between victors and vanquished.125 The two writers’ contrasting opinions are not entirely valid analytically, since Malmesbury had the benefit of hindsight as Poitiers did not, and because there were other factors involved: Cnut had been campaigning in England alongside his father for some time and knew the people he was dealing with, whereas William was a stranger to most of his new subjects, as was he to them. One element of common ground between the two is that both had taxed heavily at the start of their reigns.126 The most interesting point, however, is surely that Poitiers introduced a historical perspective, a reflection perhaps of the education that William was theoretically receiving from Ealdred and others. With this in mind, it may be important that he also says that William was establishing his rule for the benefit of his heirs, something that Cnut had ultimately failed to achieve.127 Was he saying that William was learning from what some might have thought Cnut’s mistakes? The great difference is that, after an initial period of installing Scandinavian allies in key positions, Cnut moved within a decade to a much more extensive reliance on the English. His main ecclesiastical advisers were always English.128 After 1066 this just did not happen. The period from Whitsun 1068 until early 1070 witnessed an intensification of the processes that had been started before the invasion of England and further developed between 1066 and 1068. The violence of these years was clearly debated much more widely than in England and Normandy. Gregory VII’s reference in 1080 to the criticism he had received for sanctioning so much killing by supporting William is just one sign of this.129 Another, written at approximately the same time, admittedly by one of the pope’s enemies, is the description of Gregory’s support for William as being given to a tyrant who had usurped a kingdom by violence, shedding a great deal of blood in the process.130 Finally, Lawson, Cnut, 82–9. GR, i, 470–1. 126 For comment on this parallel, Harvey, Domesday, 211–12. 127 GG, 156–7. 128 Bolton, The Empire of Cnut the Great, 77–106. 129 Registrum, ii, no. VII. 23 (Register, 353). 130 Wenric of Trier, Epistola sub Theoderici episcope Virdunensis nominee composite, ed. Kuno Francke, MGH Libelli de Lite (Hanover, 1891), i, 280–99, at 294 (trans. van Houts, ‘The Norman Conquest through European Eyes’, 852). 124 125

328

william the conqueror

however William’s military and political achievements between 1066 and early 1070 are viewed, it was now blatantly clear that the attempt to hold down the north of England with smallish forces garrisoned in castles supported by episodic interventions from the south had failed. For this problem, an extremely drastic solution had been found. However, in Maine, the reliance on a garrison in Le Mans and the support of some churchmen had been shown to be insufficient against determined opponents. Furthermore, Edgar was back in the kingdom of Scots and still plotting, Svein Estrithsson was about to invade, and Maine was controlled by Geoffrey de Mayenne. Leaving to one side for now the ethics of how he had done it, for all that he had achieved, William had still not defeated his main enemies.

Chapter 9

F RO M C R I S I S TO TRIU MPH By late 1073, William had established a position in England and across the British Isles that replicated that of the strongest of his pre-1066 predecessors and he had regained control over Maine. The subsequent period until early 1075 is probably the longest period of sustained peace that he experienced between 1066 and his death and is arguably the high-water mark of his whole career. The balance of relations between William and the surviving English elite was, however, tilted even more decisively than before in favour of the newcomers, and also, among the newcomers, quite significantly in favour of Normans. Major changes were also effected to the governance and personnel of the English Church. In terms of how contemporaries assessed him, William and Matilda’s introduction into the confraternity of the abbey of Cluny in 1074 must be very important. The year 1074 was also when Edgar the Ætheling finally submitted, an event that some twelfth-century writers saw as a great turning point. The last representative of the Wessex line of kings had abandoned any claim to rule; William’s legitimacy might seem to have been established. As a reflection of change, it is interesting that by 1075, with a rebellion breaking out in England, Lanfranc, from 1070 Archbishop of Canterbury, could counsel William that there was no need for him to leave Normandy since the situation in England was under control; whereas, a little over two years earlier, he had entreated Pope Alexander II to pray that William should have a long life, since while he lived all enjoyed peace of a kind, but after his death no one expected either peace or any good.1 This apparent new-found confidence was, as we shall see, intermittently punctured. And, for all the enhancement of legitimacy created by Edgar’s capitulation, in 1101, the ‘E’ Chronicle described Edith/Matilda, daughter of Malcolm and Edgar’s sister Margaret, and Malcolm, King of Scots, as being ‘of the rightful royal family of England’.2 A sense of the illegitimacy of William’s rule in England must have remained pervasive, something of which, as we shall see, Orderic was well aware. Across northern France, the balance of Lanfranc Letters, nos. 1, 34. ASC, ‘E’, 1100. See in general Emily A. Winkler, ‘1074 in the Twelfth Century’, ANS, xxxvi (2014), 241–58. 1 2

329

330

william the conqueror

power was significantly changed by Robert the Frisian’s coup in Flanders in 1070–1. PERSONAL AND FAMILY MATTERS These years were a time of personal and family tragedy. William fitz Osbern’s death in early 1071, during the unsuccessful intervention in the civil war in Flanders, deprived William of the man on whom he had depended more than on anyone else all of his adult life. And William and Matilda’s second son Richard died in a hunting accident in the New Forest during this period, the combination of Orderic and Malmesbury’s rather odd accounts suggesting that he rode into an overhanging branch, perhaps disoriented by a mist or strong smell. With Richard’s last appearance in a charter dating from April 1069 and it being known that he died before being knighted, his death can probably be dated to c.1070 or a little after. He was generally considered to be a promising young man.3 The one indication of William commemorating him is the grant of an estate in Hertfordshire to Healfdene, one of King Edward’s thegns, who was still holding the land in 1086 as a tenant of Peter de Valognes, the sheriff of the county. This was presumably a charitable act in favour of an Englishman who might have rendered William or his deceased son service at some point.4 The Geld Rolls preserved in Exon Domesday mention that Matilda had granted exemption from geld to ‘a certain widow’ (quidam uidua) for Richard’s soul. Both are references to the kind of small transactions that normally fall below the radar of charters and histories, and also do so because their beneficiaries were laypersons.5 As such they may well indicate extensive charitable activity of which we otherwise know nothing. At some point during this period, the couple’s eldest daughter Adelida retired to a religious community, probably the abbey of Saint-Léger of Préaux under the guardianship of Roger de Beaumont.6 Although the evidence relating to the names and lives of some of William and Matilda’s daughters is extremely confused, especially Orderic’s, the final list of Adelida’s proposed husbands might consist of Herbert II, count of Maine, Harold, earl of Wessex, Edwin, earl of Mercia, and, perhaps, a Spanish king who was probably the future King Alfonso VI of Galicia. If this is so, 3 OV, iii, 114–15; GR, i, 502–5; Regesta, no. 254. Barlow, William Rufus, 445, on the basis of a similar dating to mine for William and Matilda’s marriage, dated Richard’s birth to c.1055. 4 DB, i, fo. 141rv (Sed W. rex ded[it] hoc M. huic Aldene et matri eius pro anima Ricardi filii sui ad ipsemet dicit et pro breue suum ostend[it]). 5 ‘The Dorset Geld Rolls’, ed. and trans. Ann Williams, VCH, Dorset, iii, 115–49, at 128. 6 OV, iii, 114–15.

from crisis to triumph

331

then the most likely sequence would have the Spanish betrothal follow after the unfulfilled offer to Edwin and a chronology that places four proposed marriages within a period from 1062 to c.1070. One of Orderic’s stories even has a daughter, whom he names as Agatha, set out to travel to Spain, but die on the journey, detesting the proposed marriage because she continued to love Harold. Her body was brought back to Bayeux for burial.7 Even if, as is possible, we are dealing with two daughters rather than one, the reference to the betrothal to Harold makes it very unlikely the Spanish betrothal occurred before 1066, as is sometimes suggested.8 William would presumably not have dispatched her to Spain when he wanted to demonstrate that it was Harold who was not keeping the bargain agreed in 1064. A letter to Adelida from St Anselm, then prior of Le Bec, which has been dated to c.1072, but may date to a little later, and was clearly sent at Adelida’s request, offering spiritual comfort, suggests that she was not very happy; some of the prayers that Anselm recommended dwelt on the theme of the transformation of hate into love and through it the route to God.9 Orderic added that Adelida was still of marriageable age when she retired from the world. Among many things, her life epitomizes the obedience that medieval parents could expect. It is perhaps doubtful that William and Matilda were untypical in this respect. The way in which their daughter was used as a pawn in political manoeuvring is nonetheless shocking. She predeceased her father on 7 December of an unknown year, with him financing the construction of a bell-tower for the cathedral at Chartres for the salvation of her soul. So lavish a gift might indicate some sort of wish to atone for the treatment she had received. It might date from the final years of his reign when Adelida’s sister Adela had become countess of Blois-Chartres by marriage.10 AFTER THE HARRYING OF THE NORTH The Harrying of the North is just one sign that the arrival of the Danish army and the great revolt in the north had brought William to the conclusion 7

Ibid. For Adelida, GG, 94, note 2; van Houts, ‘The Echo of the Conquest in the Latin Sources’, 141–4; Barlow, William Rufus, 444. See also Felipe Ramon Cordero Carrete, ‘Los esponsales de una hija de Guillermo el Conquistador con un “Rey de Galicia” ’, Cuadernos de Estudios Gallegos, vii (1952), 55–78, who names Alfonso as the probable putative husband, but prefers a date of 1064–5. 9 Sancti Anselmi Opera Omnia, iii, ep. no. 10 (The Letters of Saint Anselm, trans. Frölich, iii, no. 10); Southern, Saint Anselm, 92–3; van Houts, ‘The Echo of the Conquest in the Latin Sources’, 144. 10 Cartulaire de Notre-Dame de Chartres, ed. E. de Lepinois and Lucien Merlet, 3 vols. (Chartres, 1862–5), iii, 218; Obituaires de la province de Sens. ii. Diocèse de Chartres, ed. Auguste Molinier and Auguste Longnon (Paris, 1906), 25. 8

332

william the conqueror

that more drastic action was required to consolidate his hold on England. Another, recorded in both versions of the Chronicle, is the statement that ‘in the following spring [i.e. in 1070] the king had all the monasteries that were in England plundered’. John of Worcester later explained that what was involved was the seizure of goods deposited in monasteries for safekeeping by anxious English and their transfer into the royal treasury. He says that this was done during Lent, that is between 17 February and 30 March, on the advice of William fitz Osbern; the order must therefore have been transmitted as his army marched across the Pennines or south through the west Midlands.11 This operation, if it took place on anything approaching the scale that the sources imply, must have been logistically a huge one involving most of the king’s sheriffs. By any standard, it was another landmark in the process of political punishment inaugurated before the Battle of Hastings. Svein Estrithsson’s arrival, which finally occurred in the spring of 1070, merited a mention only in the ‘E’ version of the Chronicle.12 After joining the remainder of the fleet that William had paid to leave the Humber, he divided his forces between the Humber estuary and the Isle of Ely, presumably intending to use the water-surrounded fortress as a base from which to build up numbers; ‘E’ says that the local peoples both around the Humber and in the Fenland flocked to him in the expectation that he would carry on and conquer the kingdom. The only recorded instance of how this support operated is, however, the sack of Peterborough abbey, to which William had just appointed a Norman, a former monk of Fécamp, as abbot after the death of Abbot Brand on 27 or 29 November 1069. The new abbot, Turold, after a brief spell as abbot of Malmesbury (notable, according to William of Malmesbury, for his tyrannical rule), was transferred to Peterborough by William on the grounds of his military reputation. The king, so Malmesbury said, joked that someone of his temperament deserved to have a worthy opponent to fight.13 Turold did indeed arrive at Peterborough with a large band of armed men, but only after the abbey had been sacked by the Lincolnshire thegn Hereward and the bulk of its treasures transferred to the Danes on the Isle of Ely. All this is described in a lengthy interpolation into the ‘E’ Chronicle made at Peterborough in the early twelfth century. Hereward – subsequently and popularly known as Hereward the Wake – could well have been acting to protect treasures given to Peterborough by his lord, Earl Leofric of Mercia, and his descendants Edwin and Morcar from the new abbot,

JW, iii, 10–11. ASC, ‘E’, 1070. 13 GP, i, 628–9. 11 12

from crisis to triumph

333

whose reputation might well have preceded him.14 If such it was, the ‘E’ Chronicle and other later Peterborough writers were nonetheless somewhat ambivalent about Hereward’s idealism. ‘E’, without any of the moralizing that could have been included, noted that the Danes carried off the treasures and that most were lost when their ships sank in a storm in the North Sea. Such moralizing was, however, in all likelihood redundant for a writer who saw the rule of Abbot Leofric, who had died soon after Hastings, as a golden age that was followed by a catalogue of disasters for the abbey that went on well into the twelfth century.15 The course of events over the spring and the summer suggests that Svein must have found the response to his arrival disappointing. Hence, at a date that must be after 2 June, he reached an agreement with William, whose details are not known, after which he and his forces returned to Denmark. His attempts to claim the English kingdom at first sight look unconvincing, perhaps even feeble. But Svein did not have the luxury that William had had in 1066, namely the weakening, albeit temporary, of neighbouring enemies. In 1066 he must have feared what Harold Hardraada might do. And in 1069–70, although released from anxiety about the threat from Norway, serious threats from the Empire founded in Svein’s kinship and alliances with the rebels against Henry IV expedited his return to Denmark.16 Many among the English must by this stage have been coming to the conclusion that the cost of further resistance was too high. Some may also have not thought Svein Estrithsson likely to be any better as a king than William. There is no record of William travelling to negotiate directly with Svein, although there would certainly have been time for this to have happened after the great ecclesiastical councils of the Easter period that are the subject of the next section. If he decided that the negotiations could be conducted through intermediaries, it suggests that he believed the threat from Denmark to have been contained and that it would not ignite a large rebellion in England. Events were to show that, although undoubtedly chastened, the rulers of the Danes evidently saw what had happened as a setback rather than as a conclusive defeat. With Hereward and a small 14 Baxter, The Earls of Mercia, 261–6, with references there to the copious previous bibliography. 15 Malasree Home, The Peterborough Version of the Anglo-Saxon Chronicle: Rewriting Post-Conquest History (Woodbridge, 2015), 44–60. 16 Laurence M. Larson, ‘The Efforts of the Danish Kings to Recover the English Crown after the Death of Harthacnut’, Annual Report of the American Historical Association, for the Year 1910 (Washington, 1912), 71–81, at 76–7; Erich Hoffman, ‘Dänemark und England zur Zeit König Sven Estridsons’, in Horst Furhmann, Hans Eberhard Mayer, and Klaus Wriedt (eds.), Aus Reichsgeschichte und Nordischer Geschichte, Keiler Historische Studien 16 (Stuttgart, 1972), 92–111, at 109–11; Gazzoli, Anglo-Danish Relations, 119–22.

334

william the conqueror

group of rebels left isolated at Ely, William must have calculated that the threat he posed was a local one that could be left to others to contain. The two English earls Edwin and Morcar, whose earldoms had been ravaged and the livelihoods of many of whose tenants and dependants would have been destroyed, had to this point been entirely absent from the narrative of events. Neither seems to have joined William’s opponents and neither apparently fought with him against them. They had apparently either been totally marginalized or had effectively marginalized themselves.17 THE PAPAL LEGATES’ VISIT AND THE RESHAPING OF THE ENGLISH CHURCH IN 1070 The arrival in England of three papal legates sent by Alexander II, the Anglo-Norman ‘expert’, Bishop Ermenfrid of Sion, John Minutus, cardinal-priest of S. Maria in Trastevere, and Peter, probably the cardinal priest of St Chrysogonus, represents an effort on William’s part to reshape the English Church and also one on Alexander’s part to implement a programme of reform that he must have anticipated was going to take place when the papal banner was granted in 1066. The legates must have arrived in England some considerable time before Easter, because the text survives of a letter of summons from them to Bishop Wulfstan of Worcester to attend the council held at Winchester on 7 or 11 April.18 Orderic says that after the council the legates remained with William for a year and travelled around the country conducting business necessary to the reform of the English Church. Although the time period must be an exaggeration, they were certainly extensively involved over a period of several months; according to John of Worcester, the cardinals John and Peter left England at Whitsun, with John known to have stayed on in Normandy after leaving England. Ermenfrid held a second council, at Windsor at Whitsun, before he too crossed to Normandy.19 Orderic and the author of the Vita Lanfranci, both of them probably relying on Poitiers’s lost text, say that the initiative to call in the legates came from William.20 William of Malmesbury comes close to confirming this by saying that the Easter council was called on William’s initiative.21 In practice, however, the background to the legates’ arrival may not have been Williams, English and the Norman Conquest, 47–8; Baxter, The Earls of Mercia, 293–6. CS, 568. One of the cardinals, Peter, was still in Rome on 13 January 1070, ibid., 564, note 3. 19 The sources for the councils are printed together in ibid., 568–81. For John in Normandy, Regesta, no. 237. 20 OV, ii, 236–7 (ex petitione ipsius Alexander papa tres idoneos ei ut karissimo filio legauerat uicarios); ‘Vita Lanfranci’, 683 (missi ad petitionem ipsius a papa Alexandro). 21 GP, i, 48–9. 17 18

from crisis to triumph

335

as straightforward as they indicate and, even if it was William who instigated the mission, its agenda may well not have been the same as his. A surviving fragment of a letter of Pope Alexander II suggests a stricter attitude towards the leading English clergy than William’s had so far been.22 If in the letter, which surely must date from before 1070, Alexander’s comment on ‘members of an evil head’ who ‘have turned the English from the path of truth’ refers to Stigand, then it would appear that the pope was anticipating a programme of correction in England that William had not been carrying out. The letter also mentions that the annual payment from the English kingdom known as Peter’s Pence, a unique payment to the papacy that had been paid for many years before 1066, had gone unpaid. payment defines Alexander’s responsibilities in terms of a special protection that went back to the Gregorian mission of 597. The only source to present the legates’ perspective on the visit, the letter of summons to Bishop Wulfstan, stresses their responsibilities for the Church’s welfare, attitudes in conformity with Alexander II’s opinion that the Conquest had restored the English Church to its normal relationship with Rome.23 Whoever was responsible for initiating this visitation, its dispatch must have been agreed by William and Alexander before the expedition to northern England. Ealdred’s death, and the removal thereby of the one person deemed capable of guiding the Church in England in the desired direction, was possibly the catalyst. Although rather confused in its details, the Liber Eliensis’s account of Stigand depositing his treasure at Ely suggests that he had advance knowledge of what was likely to happen.24 This presumably means that William and those around him had been sending out signals. According to John of Worcester, Stigand, his brother Bishop Æthelmær of East Anglia, and three unnamed abbots were all deposed at the Easter council of Winchester. Other sources indicate that the married Bishop Leofwine of Lichfield was summoned to answer charges there, and that when he did not turn up, he was excommunicated in the expectation that he would be deposed afterwards. Æthelric, the former bishop of Durham, was imprisoned and his brother Æthelwine, to whom he had passed his see and who had fled with Edgar the Ætheling to the King of Scots in early 1070, was outlawed. It is possible that Bishop Remigius’s plans

22 PL, cxlvi, col. 1413; H. E. J. Cowdrey, ‘Lanfranc, the Papacy and Canterbury’, in d’Onofrio (ed.), Lanfranco di Pavia, 439–500, at 449–56 (Popes and Church Reform in the 11th Century, chapter 10); Cowdrey, Lanfranc, 83. 23 CS, 568. 24 Liber Eliensis, ed. E. O. Blake, Camden, 3rd ser., xcii (1962), 176–7 (Liber Eliensis: A History of the Isle of Ely from the Seventh Century to the Twelfth, trans. Janet Fairweather [Woodbridge, 2005], 207–8).

336

william the conqueror

to move his see from Dorchester-on-Thames to Lincoln were also discussed.25 At some point, possibly at the Windsor council, Bishop Æthelric of Selsey was also deposed by a procedure that was thought suspect by the pope.26 The purge of the episcopate left four bishops with Lotharingian backgrounds appointed by Edward the Confessor and just two Englishmen in office. Of the Englishmen, one, Siward, bishop of Rochester, was an elderly man holding the smallest diocese in the kingdom that was closely supervised from Canterbury, and the other was Bishop Wulfstan of Worcester. A further Lotharingian was added to the list with the appointment of Walcher to Durham. That he was not a Norman may have been intended to be as non-provocative as was possible in the circumstances. The Durham Historia Regum of the 1120s describes him travelling north with an armed guard to be installed in Durham by Earl Gospatric. What we know of his episcopate indicates a man who tried hard to engage with the sensibilities of the Durham clergy; he famously welcomed monks from Evesham and Winchcombe who had read Bede’s Historia Ecclesiastica and had decided to refound the monasteries of Jarrow and Wearmouth.27 According to Orderic and the Vita Lanfranci, both again to an extent probably reliant on Poitiers’s lost text, the legates also performed a ceremony that had some elements of a second coronation; the Vita Lanfranci goes further than Orderic and says that the ceremony confirmed William as king.28 Such an obviously constitutive role is unlikely; the statement’s absence in Orderic’s version surely demonstrates that it was not in Poitiers’s Gesta. William’s wish to have the ceremony could well have stemmed from a desire to reaffirm his kingship after the buffeting of the previous years. But there is no mention anywhere of any oaths. The coronation of 1066 remained the basis of his kingly authority. But the legates’ presence supplied an opportunity for a display of approval by representatives of the highest religious authority on Earth. Additional coronations had their place in English history, most spectacularly with King Edgar’s first and second coronations in 946 and 973. The near-contemporary representation of the ceremony in the manuscript known as the Ramsey Benedictional

Lanfranc Letters, no. 2; ASC, ‘D’ 1070; CS, 565–6, 569. CS, 566. For Alexander’s anxieties, Lanfranc Letters, no. 7. 27 HRA, ii, 195; Henrietta Leyser, ‘Walcher, Earl of Northumbria (d.1080)’, ODNB. 28 OV, ii, 236–7 (ubi cardinales Romanae aecclesiae coronam ei solenniter imposuerunt); ‘Vita Lanfranci’, 683 (in regem Anglicum confirmauerunt). For a discussion of these texts, OV, ii, xxxii–xxxiv. H. E. J. Cowdrey’s comment that ‘it was the legates who officiated at King William’s Easter crownwearing’ (‘Lanfranc, the Papacy and Canterbury’, in Lanfranco di Pavia, ed. D’Onofrio, 452–3, reprinted in Popes and Church Reform in the 11th Century [Aldershot, 2000], chapter 10) seems to me to downplay excessively the significance of the accounts. 25 26

from crisis to triumph

337

confirms the long-standing collaboration between William and the papacy by showing the king, the two legates, and the papal banner (Plate 7). Like Malmesbury, John of Worcester and the Vita Lanfranci emphasize William’s dominant role in the councils held at Easter and Whitsun; in the former, both councils took place at his command and in his presence, with Alexander II’s agreement, while in the latter, the king presided at the single synod that it mentions. Orderic, however, differs in referring to the jointpresidency of the king and the legates. Aspects of the proceedings’ rigour may well have stemmed from the legates’ brief. Thus, the grounds for Stigand’s deposition as set out by John of Worcester and in Remigius’s subsequent profession of obedience to Archbishop Lanfranc, having previously professed obedience to Stigand, were ones that would have pleased the legates, namely that he had held Canterbury and Winchester in plurality, that he had usurped Archbishop Robert of Jumièges’s position during his lifetime, and that he had obtained his pallium from the anti-pope Benedict X.29 While these multiple irregularities in his position made Stigand an easy target, as indeed did the advanced age of a man with the remarkable record of somehow surviving every political crisis since the time of Cnut, Malmesbury’s comments that William had accorded him all the dignity appropriate to his position, and that Stigand protested his loyalty as he was removed from office, are largely confirmed by his appearance in the traditional place of the archbishop of Canterbury in the witness lists of all but one of William’s surviving pre-1070 authentic diplomas.30 On this basis, Malmesbury described William’s treatment of Stigand as duplicitous. The former archbishop spent the rest of his life imprisoned at Winchester, dying on either 21 or 22 February 1072. According to Malmesbury, he refused food and lived austerely, suggesting that his confinement was intended to be a comfortable one.31 The canons of the two councils over which the legates presided set out a thorough-going programme. Although only the headings of the capitula survive, they covered all crucial aspects of up-to-date papal reform.32 Bishops were forbidden to hold more than one see, a necessary statement that dealt with a well-established papal complaint about the pre-1066 English Church. Simony and clerical marriage were condemned. No cleric was to bear arms. Diocesan synods were to be held annually. Archdeacons JW, iii, 10–13; CS, 573–4. GP, i, 48–9; Regesta, nos. 138, 181, 254, 286. He is omitted only from no. 345(I). For Stigand’s status before 1066 and his deposition, N. P. Brooks, The Early History of the Church of Canterbury (Leicester, 1984), 304–13; Barlow, The English Church, 1000–1066, 304–7; Smith, ‘Archbishop Stigand and the Eye of the Needle’, 199–219; H. E. J. Cowdrey, ‘Stigand (d.1072), Archbishop of Canterbury’, ODNB. 31 GP, i, 48–9. 32 For the canons, CS, 574–6, 580–1. 29 30

338

william the conqueror

were to be appointed. Tithes were to be paid. And so on. They were actually much more wide-ranging and ambitious than the sketchily preserved canons of the councils of the pre-1066 Norman Church. As such, they must be to a considerable extent symbolic, a display of zeal required by the alliance of papacy and conqueror that had given a sort of legitimacy to William’s subjugation of the English kingdom. New bishops were soon appointed. The vacant archbishopric of York went to Thomas, an educated man and a protégé of Odo of Bayeux; the bishopric of Winchester went to Walchelin, a canon of Rouen and a royal chaplain, a bishop praised to the skies by Malmesbury.33 Bishop Ermenfrid meanwhile crossed to Normandy and, in the company of yet another legate, Hubert, implemented William’s and the legates’ decision that Lanfranc should be persuaded to become archbishop of Canterbury.34 He was appointed by William on 15 August 1070 and consecrated at Canterbury on 29 August. All the other vacant bishoprics and abbacies went to Frenchmen, the great majority of whom were Normans. Thus, Lichfield went to a royal chaplain named Peter, East Anglia to Herfast, who had briefly served as the king’s chancellor after Regenbald, and Selsey to Stigand. William was still in England for Lanfranc’s appointment on 15 August and was then further detained by a dispute that broke out between the two new archbishops. When Thomas, as he was obliged to do, went to Canterbury for consecration on 29 August, Lanfranc asked him for a written profession of obedience and an oath of loyalty.35 Thomas then asked for evidence that they were required. According to Lanfranc’s version of events, this was duly supplied, but Thomas, unconvinced, departed unconsecrated.36 On hearing what had happened, William is said to have been much displeased. Lanfranc’s personal account of the proceedings has the angry king react against Lanfranc, who he thought was making excessive demands and being clever rather than truthful, before being pacified by the strength of the archbishop of Canterbury’s arguments. The York version, supplied by Hugh the Chanter, written later, and completed in the 1120s, has the angry king browbeat Thomas into professing obedience. His anger is portrayed as truly frightening; not only Thomas, but all of his kin, were threatened with exile and disinheritance.37

33 GP, i, 270–3. For Thomas, Christopher Norton, Archbishop Thomas of Bayeux and the Norman Cathedral at York, Borthwick Paper, no. 100 (York, 2001). 34 For the procedure by which Lanfranc was appointed, Lanfranc Letters, no. 1. 35 JW, iii, 14–15; The Anglo-Saxon Chronicle. A Collaborative Edition: 3. ms. A, ed. Janet M. Bateley (Cambridge, 1986), 84–5. 36 Lanfranc Letters, no. 3. 37 Ibid.; HCY, 6–7.

from crisis to triumph

339

That William’s anger was directed towards both parties presumably indicates that it was being used to push them towards a settlement. It was also in the interests of writers on both sides to put William’s anger in the foreground of their accounts, since neither finally obtained the outcome they wanted; a display of royal anger constituted the perfect excuse for their champions’ failure to carry the day at the time. Having expended so much time on the affairs of the English Church, it is actually quite likely that William was genuinely irritated by this quarrel over a procedural issue. It is also significant that Lanfranc’s case was partly constructed on the basis of the pre-1066 past. The proclaimed continuity of William’s rule meant that such proceedings would increasingly arise, with, as Lanfranc’s narrative makes claim, the testimony of the English being crucial. In due course, Thomas, having been asked to think about it, did profess obedience, but for Lanfranc’s lifetime only, and not to his successors, and was then consecrated. The formula saved face for everyone and satisfied neither side.38 Although William’s role in producing this modus vivendi is not known in detail, the fact that one was achieved at all is surely testimony to his personal authority. That Lanfranc and Thomas always remained on good terms is arguably even more impressive as evidence of how his dominance could subdue argument. While Thomas had been Lanfranc’s pupil, this of itself can hardly have guaranteed long-term goodwill. Also remarkable is the speed with which Lanfranc appears to have assembled his case. His profession to Alexander II that he was ‘a novice Englishman’ notwithstanding, he showed himself remarkably well briefed very quickly. He was soon to show the same alacrity when it came to claiming back lands lost by earlier archbishops. It is also notable that William did not throw all his weight behind Lanfranc, given the potential advantages that might have arisen from the existence of a single source of authority within the Church to parallel his own. At the least he may well have had reservations about Lanfranc’s arguments. Recent experience of Ealdred’s supportive role would have shown him just how important the archbishop of York was. And given that his control over northern England remained precarious, William may have decided that the direct relationship with Thomas was to be valued above his personal intimacy with Lanfranc. He may also have recognized that this was a situation where final power of decision lay with the pope; given his manifest trust in Alexander II, William may well have decided that a compromise in England was the way to hold things together in anticipation of the dispute being opened up in Rome, as it was surely going to be when the two archbishops went there for their pallia.

38

On this episode, see most recently, Cowdrey, Lanfranc, 87–93.

340

william the conqueror

NORMANDY AND NORTHERN FRANCE, 1070–1 According to Orderic, Matilda and Robert Curthose had been formally responsible for the duchy during William’s absence in England of almost two years until his return in either the late summer or the early autumn of 1070, the longest time he ever spent away from Normandy during his entire life. His visit lasted several months.39 With nothing of note known to have happened during his absence, the period had presumably been an untroubled one, although also one when no attempt was made to recover the lost ground in Maine. During the visit William was involved in further discussions with the legates John and Peter about the endowment of Saint-Etienne of Caen that were apparently aimed to ensure that all parties were content with what was being set up.40 John’s appearance in a diploma in favour of the nunnery of Saint-Amand of Rouen may indicate that he visited Rouen.41 An element in the discussions that took place must also have been the appointment of a new abbot for Saint-Etienne, William, known as ‘Bona Anima’, a former archdeacon of Rouen who had become a monk of Le Bec and then transferred to Caen with Lanfranc, becoming claustral prior before 1070.42 His appointment, effectively an internal promotion, albeit of someone with excellent connections within Normandy’s secular and religious elites, suggests a purposeful effort to sustain the culture already established at the monastery, which included a significant external school that continued to attract some very impressive pupils.43 It is a commentary on how much Saint-Etienne meant to William that this business was done at the time when a political crisis was developing in Flanders involving Matilda’s closest kindred. The death of Count Baldwin V’s successor, Matilda’s brother Count Baldwin VI, on 17 July 1070, was followed by his son and successor, the adolescent Count Arnulf III (1070–1), being attacked by his uncle Robert ‘the Frisian’. William’s response was to send William fitz Osbern to Flanders to help him, but both he and Arnulf were killed while fighting against Robert at the Battle of Cassel on 22 February 1071.44 We are once more dealing with the seemingly perennial eleventh-century problem of a young ruler confronted by a member of his kindred seeking to supplant him. In this case, however, the rebel won. And, from William’s perspective, OV, ii, 222–3. Registrum, i, no. 1.70 (Register, 74). 41 Regesta, no. 237. 42 David Spear, ‘William Bona Anima, Abbot of St Stephen’s of Caen’, HSJ, i (1989), 51–60, at 51–2; NMProsopographie, 41–3. 43 David Spear, ‘The School of Caen Revisited’, HSJ, iv (1992), 55–66, at 56–60. 44 OV, ii, 280–1. 39 40

from crisis to triumph

341

the loss of William fitz Osbern must have been an extremely serious blow, both personally and politically. While Orderic and William of Malmesbury disagree on some of the details of fitz Osbern’s disastrous intervention, they concur in the opinion that he was accompanied only by a small military force; Orderic says that it consisted of just ten men.45 Orderic also saw the contribution as an expedition in support of the French king, with whom William was still on friendly terms, rather than one uniquely sponsored by William himself. Malmesbury inserted a tale of the widower fitz Osbern seeking marriage to Arnulf ’s widowed mother, but, if this was the case, it is inconceivable that he would have so acted without William’s approval, with the proposed marriage being an obvious route to consolidate further influence along the North Sea littoral. Given the scale of William’s preoccupations in England, it is arguably doubtful whether more could have been done in Flanders at this time. With either support or acquiescent neutrality from the counts of Flanders having been a favourable force throughout William’s life, this failed campaign was to turn out to be a significant setback. The subsequent settlement between Philip I and Count Robert left William on the sidelines and, from 1071 onwards, he had to cope with a ruler who was always unhelpful and, on occasion, overtly hostile. According to Malmesbury, William continued to give support to Arnulf ’s younger brother Baldwin, count of Hainault, but his actual influence must have been limited. This was perhaps the first time in his life that William had so badly failed to steer events in the way he wanted. Flanders once more became a potential stopping-off point for an invasion of England from Denmark. Awareness of this possibility surely led to the marriage of Robert the Frisian’s daughter Adela to Svein Estrithsson’s son, King Cnut IV, in c.1080. The war in Flanders had another major impact on the composition of England’s new aristocratic elite. Gerbod of Oosterzele-Scheldewindeke, installed as earl of Chester in c.1068, who had also travelled to Flanders to support Arnulf, was entirely accidentally responsible for the young count’s death at Cassel. Taking advice, he abandoned his earldom and possessions, becoming a monk at Cluny to atone for his part in the tragedy.46 His replacement at Chester was a member of Normandy’s aristocracy, Hugh, the son of Richard Goz, vicomte of Avranches, a notable promotion for the 45 Ibid., ii, 280–3; GR, i, 472–5; GND, ii, 146–7, 224–7; Charles Verlinden, Robert Ier Le Frison, comte de Flandre (Antwerp and Paris, 1935), 61–4. 46 Van Houts, ‘Hereward’, 219; The Warenne (Hyde) Chronicle, ed. Elisabeth M. C. van Houts and Rosalind C. Love, OMT (Oxford, 2012), 97–9 (citing La Chronique de Saint-Hubert dite Cantatorium, ed. K. Hanquet [Brussels, 1906], 65–9; Gilbert of Mons, La Chronique, ed. L. Vanderkindere [Brussels, 1904], 8–10; Gilbert of Mons, Chronicle of Hainault, trans. Laura Napran [Woodbridge, 2005], 6–7).

342

william the conqueror

descendants of someone who had seemingly gone against William’s interests during his adolescence. Another dilution of the multi-ethnicity of the new elite took place at around this time when Brian, earl of Cornwall, returned to Brittany, perhaps disabled in warfare. His replacement was William’s half-brother, Robert, count of Mortain.47 These changes highlight another trend in the shifting dynamic of these years. Just as some of the participants in the invasion of 1066 had earlier abandoned England, others subsequently found the demands of controlling networks that stretched beyond the frontiers of England and Normandy impossible to sustain. Although Flemings continued to settle in England and the British Isles in considerable numbers throughout the eleventh century, there was a clear tendency for individuals from the heartlands of the county of Flanders to abandon England after 1070–1. Furthermore, the financial arrangements that had probably been negotiated between William and Baldwin V, whereby Baldwin received cash in return for supplying soldiers, must also have been disrupted.48 The composition of the topmost imperial elite almost instantly became much more Norman at the start of the 1070s. However, the motivations behind this change should not be seen as ethnic in either inspiration or purpose. The stability of the new crossChannel empire depended on maintaining a balance between a core that consisted of Normandy and England and peripheries that stretched for vast distances around the British Isles and northern France and, arguably, into the Empire as well. Orderic described William fitz Osbern as ‘the bravest of the Normans, renowned for his generosity, ready wit and outstanding integrity’. Elsewhere, however, he called him ‘the first and greatest oppressor of the English’, a judgement that undoubtedly stems more from Orderic’s moral perception of the Conquest than from an objective assessment of fitz Osbern’s actual conduct; his statement that the ruin of fitz Osbern’s descendants was God’s judgment on his acts must be seen in the same light, a teleological comment based on his view of the operation of divine providence. Malmesbury described him as brave, determined, and generous, and the equal, if not the superior, of the best of princes.49 For a man who was often with William on the military campaigns of 1066–70, the range of fitz Osbern’s activities in Herefordshire as a castle-builder and a religious patron is truly remarkable, with the magnificent great hall of

For Brian’s departure, DP, 48. Oksanen, Flanders and the Anglo-Norman World, 195–7. 49 For the verdicts on William fitz Osbern, GND, ii, 146–7; OV, ii, 282–3, 318–21; GR, i, 472–5. 47 48

from crisis to triumph

343

Chepstow Castle probably owing its inspiration to him.50 He had even found time to campaign in Wales, launching what may well have been a major campaign into Gwent, Brycheiniog, and southern Powys in alliance with Caradog ap Gruffydd, king of Upper Gwent.51 In Normandy, fitz Osbern had founded two monasteries at Lyre and Cormeilles, in the latter of which he was buried. His premature death left William and his family with the problem of how to distribute the huge cross-Channel estate he had assembled. According to Orderic, it was the king’s decision that the paternal estates in Normandy should go to the elder son William (de Breteuil) and the English lands to the younger son, Roger. William’s intervention may well have been necessary because William fitz Osbern had made no provision for succession to his lands. It is also a commentary on how the relationship between family, territory, and politics was perceived, a product of a situation in which acquired land was considered more readily transmittable around members of the male kindred than inherited. The issues it embodied were to recur many times during the years that followed and, most significantly of all, in the provisions that William himself made on his deathbed. In the end, however, there were no rigid rules. William’s intervention should be seen as a response to circumstances that lay within acceptable social and legal norms.52 THE CONQUEST COMPLETED William was back in England by Easter (24 April) 1071 for an Easter court at which Bishop Leofwine of Lichfield, who had been excommunicated in 1070, came to the king and resigned his bishopric.53 If what was already becoming William’s established itinerary was followed, this would have been at Winchester. From here it looks as if William and his court then travelled westwards into Somerset, spending time at the royal manor of either North or South Petherton, where Bishop Wulfstan of Worcester successfully contested Archbishop Thomas’s claim to continue to hold twelve estates that had 50 Most recently, Strickland, ‘Status, Display and Defence on the Welsh March’, passim; cf. Rick Turner, Chepstow Castle (Cardiff, 2002), 5–8. 51 OV, ii, 260–1; JW, iii, 4–5. For the campaign, see most recently, David Crouch, ‘The Transformation of Medieval Gwent’, in Ralph A. Griffiths, Tony Hopkins, and Ray Howell, The Gwent County History, volume 2: The Age of the Marcher Lords, c.1070–1536 (Cardiff, 2008), 1–45, at 2–3. 52 OV, ii, 284–7; J. C. Holt, ‘Politics and Property in Early Medieval England’, Past and Present, no. 57 (1972), 3–52, at 12–17 (reprinted in idem, Colonial England, 1066–1215 [London and Rio Grande, 1997], 113–59, at 122–5); Emily Zack Tabuteau, ‘The Role of Law in the Succession to Normandy and England, 1087’, HSJ, iii (1991), 141–69, at 157–68; Green, The Aristocracy of Norman England, 337–40; Garnett, Conquered England, 177, note 354. 53 Lanfranc Letters, no. 2.

344

william the conqueror

formerly been held by Archbishop Ealdred. William of Malmesbury’s Vita of Wulfstan, although surely incredible in some of its details – such as having Wulfstan sleep through most of the proceedings, then sing psalms with his monks, before eventually pleading his case with his predecessors Saints Dunstan and Oswald before his eyes – portrays an assembly of many of the most powerful members of the new elite in full debate. Thomas’s patron Bishop Odo made what Malmesbury calls a fine speech, before Wulfstan, so we are told, supported only by Lanfranc, not only won his case but also ensured that his diocese was subject to Canterbury rather than York.54 The proceedings were thought so remarkable by Malmesbury that he cites Bishop Walchelin of Winchester, whom he held in high regard, as his source. At the heart of Malmesbury’s narrative is Wulfstan’s capacity to weigh in with the full authority of the English past behind him and to indicate his trust in the king’s judgement, citing Proverbs (‘My counsel is in you. The king’s heart is in the hand of God’).55 As we have seen and will see again, spectacular religious theatre of this kind was certainly an element in the life of William’s court. In all likelihood, the assembly’s chief purpose was the resolution of some of the many outstanding disputes between Canterbury and York in advance of Lanfranc and Thomas travelling to Rome for their pallia later in the year. It once more shows William presiding over a hearing regarding a major church matter and, as so often, imposing a decision. He did, however, supposedly compensate Thomas and his archbishopric for their losses.56 He also enhanced further Wulfstan’s status by giving him responsibility for the shires of Cheshire, Shropshire, and Staffordshire within the temporarily vacant bishopric of Lichfield. In terms of the evolutionary processes initiated in 1066, the way in which the arguments and the decision are anchored in the English past is a striking illustration of how the newcomers were adapting. But it is only an aspect of the politics of the new world of post-Conquest England, a facet of the shift in which William’s role changed from that of conqueror to that of ruler. Both versions of the Anglo-Saxon Chronicle say that at some point in 1071 Edwin and Morcar fled from court and, in their words, ‘wandered aimlessly’. They also introduce a chronological confusion that runs through into the later sources, with ‘D’ having Edwin killed by his own men and Morcar subsequently taking refuge with the rebels in Ely, and ‘E’ reversing Vita Wlstani, 62–5. On the assembly and its date, ibid., 64, note 2. Proverbs 21:1. On the assembly, Ann Williams, ‘The Cunning of the Dove: Wulfstan and the Politics of Accommodation’, in Julia S. Barrow and N. P. Brooks (eds.), St Wulfstan and his World (Aldershot and Burlington, VT, 2005), 23–38, at 27–8. 56 Vita Wlstani, 64–5. HCY, 2–3, 18–19, mentions only the loss of the estates, presumably because, from a York perspective, no compensation could be sufficient. 54 55

from crisis to triumph

345

the sequence.57 The brevity of their accounts and the apparent dismissiveness of ‘wandered aimlessly’ may reflect despair at what was thought another failure of the survivors of the pre-1066 English elite to organize effective resistance properly. Alternatively, it may indicate the futility by this stage of such attempts at resistance. Of later writers, John of Worcester believed that Morcar and Edwin’s second revolt occurred because William intended to imprison them, while Orderic, who said that they were unjustly treated by William in a way which damaged his reputation, thought the king misled by bad counsel in his treatment of them. Malmesbury on the other hand opined that Morcar and his brother had been persistently disruptive, and presumably believed that they deserved their fate.58 Orderic and John’s narratives replicate some of the chronological confusion of the Chronicle’s accounts, with Orderic saying that Edwin set off for Scotland after his brother had been captured at Ely and John seemingly suggesting that Edwin was already travelling north before the Isle of Ely fell to William.59 Orderic’s statement that Edwin spent six months trying to rally support and John’s that Morcar took refuge in Ely in advance of the winter do, however, indicate that the most likely interpretation of the muddle is that the campaigning associated with their flight from court and the presence of rebels securely ensconced on the Isle of Ely lasted throughout the summer months. The information common to both ‘D’ and ‘E’ that Morcar travelled to Ely by ship must place his arrival there before William had established a complete blockade by land and sea, that is, in the late spring or early summer. The brothers’ collaboration, therefore, looks to have involved Morcar in bolstering the rebels in Ely while Edwin tried over a lengthy period to raise support for a major campaign. Although the sources disagree on whether Edwin was betrayed by, or actually killed by, his own men, whichever is accurate is a graphic commentary on the destruction of his status that had taken place since 1066 and on how powerful William was perceived to have become; the explanation can only be that those bound to support Edwin by oaths had decided that his capacity to give them the protection expected of a lord was no longer sufficient for their oaths to remain valid. The summer campaigns of 1071 bring into play the Gesta Herwardi and the Liber Eliensis, the former in all likelihood written by a monk of Ely named Richard between 1109 and 1131, some of it on the basis of an earlier Old English text, and the latter in the 1160s. For all the Gesta’s literary ASC, ‘D’, ‘E’, 1071. GR, i, 468–9. 59 OV, ii, 256–9; JW, iii, 18–21. For the main modern accounts, Williams, The English, 51–7; Baxter, The Earls of Mercia, 278–80. 57 58

346

william the conqueror

extravagances and sometimes arguably fantastic contents, its picture of localized discontent and disputes between newcomers and natives over possession of land usually rings true.60 The description of William’s siege of the well-nigh impregnable Isle of Ely, in those days surrounded by swamps, placed alongside earlier narratives of warfare in Poitiers’s Gesta Guillelmi, supplies the two fullest accounts we have of William as a soldier. Present in both is the methodical and thorough conduct of the campaign and the hands-on personal involvement in the fighting; after the second failed attack on the Isle at Aldreth, for example, William is portrayed returning to base with an arrow in his shield and angry at the failure of his tactic. And, as at the siege of Domfront, he is said to have hunted during lulls in the warfare. The Gesta and the Liber Eliensis show very well the fluctuations of fortune as each side manoeuvred for advantage. At Ely, Morcar, and possibly Edwin, joined with Æthelwine, the deposed bishop of Durham, Siward Bearn, a landholder with extensive holdings across eastern England and the east Midlands, Hereward, and others, some of them very wealthy. The local background to this assemblage was one of disappointed hopes, local turbulence that had continued after the sack of Peterborough, and fears about William’s intentions. Hereward’s personal links to the two earls and the military prowess that had made him an attractive proposition to rulers around much of north-western Europe was mixed into the culture of feud; namely, a desire to avenge the killing of his brother, and a corresponding aim on the part of William de Warenne, whose Flemish half-brother Frederick of Oosterzele-Scheldewindeke had been killed by Hereward, probably in 1070.61 The monks of Ely are said to have given the rebels access to the ISLE because they feared that their English abbot, Thurstan, would be replaced 60 My own view is that the Gesta Herwardi conforms to many of the basic conventions of an early twelfth-century narrative and should be treated as such. This makes it especially useful for the siege of Ely for which the author drew on local knowledge. For a similar assessment, Elisabeth van Houts, ‘Hereward and Flanders’, ASE, xxviii (1999), 201–23, at 202–4; De Gestis Herwardi, ed. Meneghetti, 13–21. Liber Eliensis is largely derivative, with some additional local detail and some corrections. On its relationship to the Gesta, Liber Eliensis, xxxvi–xxxvii, 173–6, 177–88; van Houts, ‘Hereward’, 204. On the episode in general, see C. R. Hart, ‘Hereward “the Wake” and his Companions’, in idem, The Danelaw (London and Rio Grande, 1992), 625–48, at 633–6; Williams, The English, 49–51; Simon Keynes, ‘Ely Abbey 672–1109’, in Peter Meadows and Nigel Ramsay (eds.), A History of Ely Cathedral (Woodbridge, 2003), 3–58, at 41–5; Dalton, ‘The Outlaw Hereward “the Wake” ’, 15–35. See also Peter Rex, Hereward: The Last Englishman (Stroud, 2005), 92–140. These and the footnotes that follow contain references to relevant items from the voluminous bibliography devoted to Hereward. 61 Gesta Herwardi, 369; De Gestis Herwardi, ed. Meneghetti, 120–1 (trans. Swanton, 65); Liber Eliensis, 180 (trans. Fairbrother, 212); The Warenne (Hyde) Chronicle, 26–7. For Hereward and Earls Edwin and Morcar, Baxter, The Earls of Mercia, 261–6.

from crisis to triumph

347

by a French one, a not unreasonable belief since Thurstan had been appointed by King Harold.62 It is, however, possible that some of the community were from the start reluctant participants in events, since the Gesta Herwardi’s passage about their fear is not reproduced in Liber Eliensis. A letter from Lanfranc to an Abbot Thurstan, whose monastery is not specified, expresses sympathy in present troubles and advises him to make contact with the king to make peace. If this is Thurstan of Ely, as opposed to the Abbot Thurstan of Glastonbury appointed in 1082 who had been a monk of Saint-Etienne of Caen, then the letter may be no more than an expression of routine pastoral responsibilities. Lanfranc also felt obliged to tell whichever Thurstan it was that the king might not listen to him, another sign that it was William who took the decisions.63 However, although the abbey was subsequently punished for its participation in the revolt, Thurstan was not apparently deposed. William’s basic tactic was to blockade the isle. This operation involved the use of land- and sea-based forces on a large scale in order to maintain guard at all points where a crossing of the fen might be possible along a circumference of approximately one hundred miles, and to prevent access from the sea.64 Aldreth, to the south-west of Ely, where attempts were made to build a causeway, was chosen as the place from which to launch attacks. William’s headquarters were probably at Brampton near Huntingdon, although he is also known to have retreated at one point to Cambridge to regroup.65 References to skirmishes at Burwell and Reach, as well as Aldreth, show the extent of his troop deployments. The ships would have had to patrol the river Ouse and the Wellstream. Although the Gesta Herwardi and Liber Eliensis tell a story that consists principally of Hereward’s heroic exploits and William’s persistent discomfiture by them, they also pay testimony to the king’s fortitude; it is clear from them that he succeeded in his basic objective of keeping the isle’s defenders penned in. On the other hand, two attempts at direct assault across hastily constructed causeways at Aldreth failed.66 For the second of them, according to the Gesta, William was persuaded by Ivo Taillebois to employ the services of a witch in a futile attempt to cast spells on the Gesta Herwardi, 374; De Gestis Herwardi, ed. Meneghetti, 126–7 (trans. Swanton, 68). Lanfranc Letters, no. 57. Cowdrey, Lanfranc, 172–3, prefers the abbot of Glastonbury. On the date of Thurstan’s appointment, see chapter 11. 64 ASC, ‘D’, ‘E’, 1071. 65 For reasons why this identification should be preferred to Brandon (Suffolk), Liber Eliensis, lvii (trans. Fairbrother, 210, note 484); Rex, Hereward, 106. 66 Gesta Herwardi, 376–8, 388–90; De Gestis Herwardi, ed. Meneghetti, 130–3, 146–9 (trans. Swanton, 69–70, 77). Matthew Bennett, Campaigns of the Norman Conquest (Oxford, 2001), 58–60, restates the argument that the first causeway was at Stuntney, where the swamp crossing was shorter; cf. Hart, ‘Hereward’, 633–4. 62 63

348

william the conqueror

defenders. The tactic having gone farcically wrong, William is then made to castigate his troops on the folly of using methods displeasing to God, another indication of how the Gesta was portraying him. In the Gesta, the two heroes of the business are Hereward, the great warrior, and William, the just and chivalrous king. This is yet another part of the way in which later values required the imposition of a reputation that may not reflect a contemporary reality. Their duet was a dramatic requirement of a narrative that wished to present an Englishman as the equal of the best of the Normans and as superior to most of them.67 The orthodox tactics of constructing siege-works and installing engines to batter the defenders were also deployed by William. His forces suffered casualties, in all probability most notably William Malet, previously made responsible for Harold’s burial and later chosen to defend York, and presumably in 1071 an important member of William’s military household.68 The breakthrough onto the Isle of Ely was finally accomplished through the conventional intimidatory tactic of ravaging the abbey’s lands outside the isle and threatening that these lands would be granted to Frenchmen, a procedure that split the rebels and persuaded the monks to sue for terms.69 William then forced an entry onto the isle by a route that has been much discussed. Although what happened is ultimately unclear, he certainly abandoned the attempt to break through at Aldreth; there is much to favour the view that he followed the line of the river Cam, perhaps crossing the fen near Little Thetford.70 On entering the isle, he placed guards around the abbey church so that he could pray undisturbed. But he was afraid to approach St Ætheldreda’s tomb. According to Liber Eliensis, this was because he feared punishment for the harm he had done to her church; all he could bring himself to do was toss a gold mark onto the altar. This happened on 27 October.71 He also ordered the construction of a castle next to the abbey church and the installation of a garrison on the isle. After this, Hereward fled towards Northamptonshire with William’s forces in pursuit. The story of Hereward’s life thereafter is notoriously obscure; the Gesta Herwardi famously says that he became William’s faithful subject, a denouement that must be part of the relationship between the

67 See Hugh M. Thomas, ‘The Gesta Herwardi, the English, and their Conquerors’, ANS, xxi (1999), 213–32, at 229–30. 68 On William Malet’s death, see most recently, Keats-Rohan, ‘Domesday Book and the Malets’, 14. 69 Gesta Herwardi, 390–1; De Gestis Herwardi, ed. Meneghetti, 150–3 (trans. Swanton, 79); Liber Eliensis, 189–90 (trans. Fairbrother, 223). 70 Rex, Hereward, 132–40, surveys this vexed issue. 71 Liber Eliensis, 194 (trans. Fairbrother, 228).

from crisis to triumph

349

two men that the Gesta’s author creates.72 Others who had sought refuge in Ely were severely dealt with. Siward Bearn was imprisoned and released only when William was dying. Bishop Æthelwine of Durham was sent to the abbey of Abingdon where he died soon afterwards.73 According to John of Worcester, William had the less socially exalted of the defenders mutilated by having them blinded and their hands cut off.74 All this is consistent with the severity he had shown in 1069–70 and on many similar occasions. Liber Eliensis tells of how the monks handed over treasure to the value of 700 marks of silver at Cambridge where William, having discovered that the payment was one eighth of an ounce light, threatened reprisals and then took a further 300 marks. Even if the melodramatic elements in the story are a product of the passage of time, the Liber’s details on images and ornaments removed or dismantled show that its factual core is sound. It is possible that the king at this time also discovered a large hoard of gold and silver secreted by the monks and removed it to his treasury at Winchester.75 The monks’ final submission was made at Witchford (Cambridgeshire). After the fall of Ely, Morcar spent the rest of his life in an honourable captivity as a prisoner in Normandy in the custody of Roger de Beaumont; his confinement cannot have been too strict since he witnesses a charter of Roger’s in 1086, but confined he nonetheless was.76 Released by William on his deathbed, Morcar was brought to England, only to be immediately reincarcerated by William Rufus. According to Orderic, William negotiated with Morcar in bad faith during the siege of Ely and treated him dishonourably.77 As we have seen, the sources disagree on whether Edwin was betrayed by, or actually killed by, his own men. There is, however, agreement between Orderic and Malmesbury that, when his killers brought Edwin’s severed head into William’s presence, the king wept at the death of a handsome and pious man. According to Orderic he then ordered that the killers be exiled.78 This terrible scene epitomizes as much as any can the ethical problems of judging William, because it posed exactly the same complexities for Orderic and Malmesbury as it still does. The only occasion when William is known to have cried in public, it still 72 Gesta Herwardi, 404; De Gestis Herwardi, ed. Meneghetti, 170–1 (trans. Swanton, 87–8). If it happened, there was certainly a substantial cooling-off period first, Thomas, ‘The Gesta Herwardi’, 230–1. 73 JW, iii, 20–1, 46–7. For these and other participants, Williams, The English, 50–7. 74 JW, iii, 20–1. 75 Liber Eliensis, 194–5 (trans. Fairbrother, 230–1). 76 Lot, Saint-Wandrille, recueil des chartes, no. 41. 77 OV, ii, 256–7. 78 Ibid., ii, 258–9; GR, i, 468–9.

350

william the conqueror

needs to be seen as an aspect of kingly display and theatre, while at the same time recognizing that William’s tears were indicative of his awareness that the circumstances required very special behaviour, an appropriate show of disapproval for the killing of a nobleman to whom he had at one point promised a daughter in marriage. But the performance of course also extricated him from any direct personal responsibility he might have felt for Edwin’s death and for possibly being an accomplice to the processes that had undermined Edwin’s status.79 It managed the situation in a way that forestalled discussion. By the end of 1071, Waltheof, son of Earl Siward, with his responsibilities lying in the east Midlands, was the only surviving earl of English birth. Following his submission in early 1070 after having briefly joined William’s opponents, the king arranged for him to marry Judith, the daughter of his sister Adelaide and Lambert of Lens.80 The purpose of the marriage, one of the kind dangled in front of both Harold and Edwin but later withheld, could have been to contain English resistance in the regions around Ely and hold on to the loyalty of the greatest Englishman of that region. Alternatively – and the two possibilities are not mutually exclusive – William may well have had long-term plans for a young man who was, by 1070, maturing into adulthood. In comparison to Edwin and Morcar and others, Waltheof had played no part in the events of 1066, being too young to do so. For all that we have here arguably the first truly convincing attempt to build trans-ethnic networks at the highest social level, the redistribution of English lands to newcomers continued. It is likely, for example, that many Yorkshire lordships were established at around this time.81 The beneficiaries included Count Alan of Brittany, with a compact lordship in the north of the shire around Richmond, and Count Robert of Mortain with a large lordship in the Vale of York. With William de Warenne installed in the south of the county at Conisbrough, perhaps as early as 1068, and Ilbert de Lacy, who was probably a protégé of Bishop Odo of Bayeux, at an unclear date appearing at Pontefract, the region that had once been the domain of Tostig, Morcar, and Edwin had been handed over to members of an imperial aristocracy with interests across northwestern France and all the regions of England. Another important development that appears to date from around this time is the change in the language of royal writs and writ-charters, with Latin almost completely replacing Old English. Although the argument is Baxter, Earls of Mercia, 281–97. For the marriage, OV, ii, 262–3. 81 Dalton, Conquest, Anarchy and Lordship, 19–78; Fleming, Kings and Lords in Conquest England, 153–61. 79 80

from crisis to triumph

351

not entirely straightforward because of the difficulty of dating writs and writ-charters, the simple fact that almost all of them which can be dated to before the end of 1070 are in Old English, and that almost all that are demonstrably after that date are in Latin, suggests an abrupt change. And since the representation of the royal will was involved, the decision to change must have been one in which William was involved.82 The clear implication is that the shire courts to which the writs were sent were much less often than before overseen by officials familiar with Old English, a conclusion reinforced by what we know of the personnel of William’s sheriffs by this date, with almost all English having been replaced by newcomers.83 It is possible that the change is also associated with the story told by Orderic, but surely reproducing Poitiers, that William had tried to learn the English language so that he could listen to the petitions of his new subjects, but that his advancing age and his many other preoccupations prevented him from doing so.84 Both explanations are plausible; William was by this date well into his forties and it becomes harder to learn languages at that age, especially with a lifestyle in which privacy was restricted. But we might stop to wonder whether he ever truly had these good intentions; the passage has the appearance of yet another attempt to answer criticisms on Poitiers’s part. The result was that communication with the king’s new subjects would henceforth have to be conducted through interpreters, an example that could have been influential. WRITING CONQUEST After the passage devoted to Morcar’s imprisonment and Edwin’s death, Orderic indicates that William of Poitiers’s text has come to an end and that he can no longer use it to inform his account. However, in the sections that immediately precede it, before his account of the fate of the two earls, he inserted a description of conquerors and conquered living in harmony that must surely derive from Poitiers: English and Normans were living peacefully together in boroughs, towns and cities, and were intermarrying with each other. You could see many villages or town markets filled with displays of French wares and merchandise, and observe the English, who had previously seemed Regesta, 50; Sharpe, ‘The Use of Writs’, 248–53. Judith A. Green, ‘The Sheriffs of William the Conqueror’, ANS, v (1983), 129–43, at 131–2. 84 OV, ii, 256–7. 82 83

352

william the conqueror

contemptible to the French in their native dress, completely transformed by foreign fashions.85 In the context of the four years after 1066, harmony, intermarriage, and cultural assimilation on such a widespread scale are well-nigh incredible. The reference to dress fashions feels like another of Poitiers’s references to the civilizing mission that he wanted to argue justified the Conquest. The passage acquires some plausibility from the rapidity with which Normans and French were introduced right across the kingdom; extensive interaction between peoples must have happened as a consequence of the normal routines of daily life. Orderic’s retention of the passage may also draw on his childhood memories of Shropshire in the 1080s before he left for Normandy to become an oblate at Saint-Evroult; he was, after all, the son of a union between a French priest in the entourage of Roger de Montgommery and an English woman. But a little later – ten pages further on in Marjorie Chibnall’s edition – he embarks on a lengthy diatribe against England’s conquerors that lasts for a further ten pages and whose first sentences are: So when the Normans had become over-wealthy on the spoils garnered by others, they arrogantly abused their authority and mercilessly slaughtered the native people like the scourge of God smiting them for their sins . . . Noble maidens were exposed to the insults of low-born soldiers, and lamented their dishonouring by the scum of the earth.86 And so on. How, we must wonder, could he write such apparently contradictory statements in such proximity? The only plausible route to reconciliation is to accept that both represented some form of reality for Orderic. In approaching what this reality might be, the near-total absence of usable life histories and quantitative data has to be all the more deeply felt. All that can be said is that we are back with John Gillingham’s comment about ‘an event unparalleled in European history’, mentioned in the Prologue and again in Chapter 7, and, alongside this, Orderic’s comment in his Prologue to Book IV of the Historia Ecclesiastica, the book in which all these passages appear, about ‘a theme of ruin for devoted historians’.87 The multiplicity of human reactions to catastrophe must always defy a simple definition. To attempt one 85

Ibid. Ibid., 268–9. 87 Gillingham, ‘Problems of Integration’, 85; OV, ii, 190–1 (flebile tema de sua ruina piis historiographis). 86

from crisis to triumph

353

is therefore a mistake. Immense variations are to be expected. Orderic’s method of compilation is also important. Book IV does seem to be continuing the theme for which Orderic took responsibility in his Prologue to Book III, namely to write about ‘the acts of the Normans’, a passage that follows a call to meditate on the deeds of saints.88 He was respectful of his authorities – in this case William of Poitiers – where he thought he should be. But he was also conveying moral messages for meditation to his readers and listeners, primarily his monks at Saint-Evroult, but also all who wished to learn from those who absorbed his tales.89 We must not, therefore, expect the logical consistency thought mandatory for a modern historian. There has been much in this and the two previous chapters to indicate that, among the many acts of violence which undoubtedly did take place between 1066 and 1071, there were some against non-combatants that included looting, violence against women, and the destruction of livelihoods that were, according to circumstances, thought legitimate or excusable. The actual evidence of inter-marriage is weak, at least at the elite level for which we have clear-cut evidence. And even if it did take place at social levels below those recorded in the sources, as it surely did, there is evidence for forced marriages and of women trying to protect themselves that must indicate it was widespread.90 Not only is there a letter of Lanfranc’s referring to English women taking refuge in nunneries to avoid forced marriage – a letter which incidentally says that the policy to be followed was not only his but William’s as well – the subject was commented on outside Normandy and England; the theme of forced marriages appears in the almost contemporary chronicle of Frutolf of Michelsberg, written at Bamberg, this time with William being blamed for it happening.91 The imposition of penance for rape appears in Bishop Ermenfrid’s Penitential Ordinance of 1067.92 In the famous homily on William’s virtues and faults that appears in the ‘E’ Chronicle, it is mentioned as a good quality that he ordered any common man who committed rape to be castrated.93 The kindest assessment of William is that he did take seriously his responsibilities to protect and to OV, ii, 2–3. On Orderic’s English beginnings, Marjorie Chibnall, The World of Orderic Vitalis (Oxford, 1984), 3–11. 90 For this whole subject, fundamental now is Elisabeth van Houts, ‘Intermarriage in Eleventh-Century England’, in Crouch and Thompson (eds.), Normandy and its Neighbours, 237–70, with rape and forced marriage discussed at pp. 250–6. 91 Lanfranc Letters, no. 53; Frutolf of Michelsberg, Chronicon, ed. Franz-Josef Schmale and Irene Schmale-Ott, in Frutolfs und Ekkehardi Chroniken und die Anonyme Kaiserchronik (Darmstadt, 1972), 78–9 (uxores indigenarum universorum aduenis in matrimonium subiugauit); cited in van Houts, ‘The Norman Conquest through European Eyes’, 841. 92 CS, 584. 93 ASC ‘E’, 1086 [1087]. 88 89

354

william the conqueror

keep the peace while simultaneously sanctioning the violence necessary to achieve his political aims. Orderic follows this indictment with an attack on churchmen who used the conquest of England for personal profit, illustrating his theme with a sermon that he says was delivered in William’s presence by Guitmund, a monk of the Norman abbey of La Croix Saint-Leuffroy, when he was invited to England to be offered a bishopric.94 A pupil of Lanfranc, Guitmund was a theologian who wrote a tract on the Trinity against Berengar of Tours, a man, in other words, within the conservative traditions of the Norman Church that stretched back to Abbot John of Fécamp and which, of course, included Lanfranc. Turning down the offer of the bishopric, Guitmund delivered a speech whose essence was that the conquest of England was illegitimate and that it represented ‘the spoils of robbery’. Although the invention of such speeches was a device to which Orderic frequently resorted throughout the Historia Ecclesiastica, the belief that Guitmund had turned down an English bishopric on grounds resembling those set out by Orderic must surely have had some basis in fact. As we have seen too, William did on occasion listen to sermons critical of his conduct. The whole drama must therefore at the very least represent behaviour that Orderic believed William to have been capable of. Guitmund’s departure for Italy, where he joined the entourage of Pope Gregory VII, may well have been because he felt uncomfortable in Normandy, but it does not have to follow that it was William who drove him away. The location of the idealized portrayal of ethnic assimilation and the fall of Edwin and Morcar at the apparent end of Poitiers’s Gesta, with its author seemingly in full flow, comes as a surprise. A major difficulty in reaching any conclusion about why this should have happened is that no manuscript of the Gesta Guillelmi survives; the text reproduced in modern editions depends on André Duchesne’s seventeenth-century transcription, which was probably based on a manuscript destroyed in the Cottonian fire of 1731. Another is that the Gesta seems not to have been widely disseminated. Although known to Orderic, Malmesbury, Wace, Robert of Torigni, Benoît de Saint-Maure, and the authors of the Vita Lanfranci and Liber Eliensis, it has no significant textual history that we know of.95 On the basis of current knowledge, therefore, we have to accept that it ended where Orderic said it did. According to Orderic, Poitiers abandoned the Gesta due to unfavourable circumstances whose nature he does not specify, and that, at the end of his life, Poitiers gave himself up to silence and prayer. He also says that Poitiers served as archdeacon of Lisieux in the time of 94 95

OV, ii, 268–81. For this detail, GG, xliii–xlvii.

from crisis to triumph

355

bishops Hugh and Gilbert, the latter appointed in 1077, and that it was his original intention to continue his history up until the Conqueror’s death, thereby indicating that he died after 1087.96 Poitiers was probably the witness to a transaction involving Abbess Ansfrida of Saint-Léger of Préaux, who took office in c.1075 and who was probably his sister, and he – again probably – appears holding a prebend of the church of St Martin of Dover in Domesday Book.97 That Poitiers wrote of Bishop Hugh of Lisieux (died 1077) as still living suggests that he stopped writing some time before he was stricken by ill-fortune. Poitiers is not just one of our main sources for William’s life, he is one of its leading dramatis personae. As William’s chaplain, Poitiers may have known him very well indeed. The choice of so skilled a Latinist to justify William’s deeds must reflect the king’s aspirations and be the product of a learned court – a subject to which I will return – as well as being a commentary on a man with a very elevated perception of his place in the world. It is possible Poitiers thought the Conquest complete with the fall of Edwin and Morcar and that, in combination with the idealized image of harmony, it was an appropriate signing-off point, albeit one that Orderic did not attribute to him. It is, however, also possible that his personal convictions persuaded Poitiers that to continue the task of producing a rhetorical apologia for William the Conqueror had become a moral impossibility and that this explains his misfortunes. The Gesta’s intended audience must surely have been the bishops and cathedral chapters of Normandy, England, and, if possible, further afield; the material for sermons on the qualities required to be a good prince. Apparently achieving no significant popularity, perhaps because it was so much out of tune with the more reflective character of other writing, it was on these terms a failure. It is unfortunate that so little of his text covering the period after William’s coronation survives, since this represents a much more difficult subject for Poitiers to write about than Normandy before 1066. It is quite possible that he eventually found his allotted task of answering criticisms impossible; the use of the Achilles analogy is certainly one instance where he made coded criticisms of William. Without him, our knowledge of William’s deeds becomes thinner. He and Orderic’s dialogue with him exemplify the problems that confront any serious historian writing about William the Conqueror. It is also the case that Poitiers’s text ceases to be available at the moment when the subject is on the cusp of the transformation from the phase of conquest to that of rule over cross-Channel lands. 96 OV, ii, 184–5, 258–61. For a summary of all the known facts, GG, xv–xvii; Spear, Personnel, 174. 97 Regesta, no. 217 (p. 695); DB, i, fo. 1v.

356

william the conqueror

THE CONSOLIDATION OF EMPIRE Orderic suggests that William crossed to Normandy at the very end of 1071 and that he stayed there for Christmas, thereby replicating what was becoming a pattern of spending Christmas in the duchy when commitments allowed; he had also spent the Christmases of 1068 and 1070 in Normandy. According to Orderic, on encountering turmoil, he called together an assembly of Normans and Manceaux – the latter a group presumably restricted to those who still acknowledged his rule – to re-establish peace. This may well have been a first step towards a campaign to regain authority in Maine. During 1072, a synod of the Norman Church met at Rouen under the presidency of Archbishop John, in all probability at Easter, since two of the participants, Bishops Odo of Bayeux and Geoffrey of Coutances, had returned to England by Whitsun. The only surviving account, Orderic’s, does not mention William’s presence, so it is likely that he had already crossed back to England by Easter; that Orderic prefaces his list of the canons with a statement that William encouraged the bishops to take action for the good of the Church does, however, indicate that the council had William’s approval.98 The council’s legislation is notably wide-ranging and rigorous, especially in comparison to that of previous Norman councils, of which the most recent was that of Lisieux in 1064, and in some respects to the 1070 English legislation as well.99 It indicates that the spirit which had animated the legates’ visitation in 1070 also permeated Normandy. Its continuation was to an extent guaranteed by ties of mutual understanding between Archbishops John and Lanfranc, who are known to have kept in regular contact.100 Thus, on the issue of clerical marriage and concubinage, the Lisieux legislation was considerably elaborated at Rouen, with the earlier rather permissive formulation being developed to prohibit cohabiting priests from governing churches and to enforce absolute chastity at the levels of archdeacon and dean. Just how controversial this was is indicated by the story that a riot broke out and that Archbishop John was stoned and driven out of the synod.101 98 OV, ii, 284–5. For the synod, ibid., 284–93. For Odo and Geoffrey at Windsor at Whitsun, CS, 603; Regesta, no. 68. 99 For some comment, Allen, ‘ “A Proud and Headstrong Man” ’, 19–20. For the stricter character of the Norman canons, Brooke, ‘Gregorian Reform in Action’, 11, note 32 (reprinted, Medieval Church and Society, 84, note 32). For translations of the relevant canons, van Houts, ‘The Fate of Priests’ Sons’, 79–80. 100 Lanfranc Letters, nos. 14–17, 41, contain references to other letters and to contact through messengers. See further Cowdrey, Lanfranc, 213–15. 101 OV, ii, 200–1; AAR, 41–4, 55–9; Alison Alexander, ‘Riots, Reform and Rivalry: Religious Life in Rouen, c.1073–c.1092’, ANS, xxxiii (2011), 23–40, at 30–2.

from crisis to triumph

357

Although Archbishop John’s reputation for being over-zealous may have fuelled the fire, disturbances of this kind were widespread throughout western Europe at this time, because there were many cohabiting priests. William’s half-brother Bishop Odo may have been one of them, since he had a son who was still alive in 1131 and was prominent at Henry I’s court.102 Many members of the duchy’s cathedral chapters certainly either kept concubines or were married.103 As often with such legislation, it is an ideal that was being stated, with some certainly recognizing that change involving the break-up of households and the destruction of the assumed prospects of the children of the unions was not going to happen overnight and might even have been inhumane. A further contemporary indication of William’s achievements in Normandy is the appointment to the bishopric of Sées of Robert, son of Hubert de Ryes, a member of a Norman family notable for its loyalty to William, to replace Bishop Ivo (of Bellême) who had died, in all probability on 12 April 1071. In association with Roger de Montgommery and his wife Mabel’s succession to the Bellême lands, this represents a significant extension southwards of influence and authority. The ambassador to the papal court in 1066, Gilbert, archdeacon of Lisieux, was also appointed to the bishopric of Evreux at around the same time.104 During the latter stages of the defeat of the revolt based at Ely, Archbishops Lanfranc and Thomas, accompanied by Bishop Remigius of Dorchester-on-Thames and Abbot Baldwin of Bury St Edmunds, set off for Rome in October 1071, ostensibly for the two archbishops to obtain their pallia. The timing of the mission in relation to the ending of the siege of Ely may not have been accidental; Remigius was Ely’s diocesan bishop and Baldwin’s abbey a near-neighbour to the fighting. Lanfranc, so we are told, was warmly greeted by Alexander, who rose to greet him and presented him with a second pallium, which he had himself worn. While there, however, Thomas reopened the issue of Lanfranc’s request for a profession of obedience with Pope Alexander II.105 Alexander in turn immediately wrote William a letter that is a notable illustration of how the pope regarded him and of the extent to which he saw him as responsible

OV, iv, 116–17. For a general statement on the subject, van Houts, ‘The Fate of Priests’ Sons’, 59–65. 104 For the dates, Spear, Personnel, 134, 272. On Ivo, see also Jean-Michel Bouvris, ‘En marge de l’année des abbayes normandes: la date de la restauration de l’abbaye d’Almenênches’, Bulletin de la Société Historique et Archéologique de l’Orne, xcviii (1980), 113–41, at 124, note 52. 105 Lanfranc Letters, no. 3. For the visit to Rome and the 1072 settlement reached in England, most recently, Cowdrey, Lanfranc, 93–103. 102 103

358

william the conqueror

for church business in his lands.106 It praises William’s kingship and exhorts him to continue in the path of righteousness and to follow Lanfranc’s advice; the archbishop is described as ‘one of the leading sons of the Roman Church’. With recent events perhaps in mind, it reminds him of his responsibilities to protect widows, orphans, and the oppressed. William is then asked to look again at the case of Æthelric, bishop of Selsey, who, it was thought, had been uncanonically deposed, and is told that Lanfranc has been given power to decide the boundary dispute between Archbishop Thomas and Bishop Remigius.107 During the visit to Rome, Lanfranc’s personal authority was sufficient to protect Remigius against the charge of simony and have him restored to office. Baldwin used the occasion to secure an exemption from the authority of the bishop of Elmham (soon to be Thetford and later Norwich) that paralleled the one that Lanfranc had earlier secured for Saint-Etienne of Caen. It is an interesting example of how an English monastery under leadership approved by the kingdom’s new rulers could already acquire a status analogous to that of one of Normandy’s main monasteries.108 These two outcomes again demonstrate the level of personal trust that had consistently run through the relationship between William, Alexander, and Lanfranc. And the stylistic influences on the aspects of the new churches that Remigius and Baldwin were soon to start building at Lincoln and Bury St Edmunds respectively raise the intriguing prospect of the two of them gazing up at the Arch of Constantine during their visit. If so, this supplies more imagery that was thought appropriate for the new regime in England, locating it within a continuum going back to the most distinguished elements in the nation’s past (Plate 17).109 And Remigius, like Baldwin, seems to have seen continued direct contact with the papacy as an important constituent of ecclesiastical change. The discovery at Lincoln in 1985 of a seal matrix with the incomplete legend SIGNO SIGILLATUR LEGATIO has led to the suggestion that the city was visited by the legate Hubert in 1072.110 As a result of Alexander’s referral of the dispute between the two archbishops back to England, William and Matilda, along with the pope’s legate Hubert and a group of English bishops, gathered at Easter (8 April) Lanfranc Letters, no. 7. On this dispute, Bates, Bishop Remigius, 20–1. 108 Herman and Goscelin, Miracles of St Edmund, 68–71; Bates, ‘The Abbey and the Norman Conquest’, 16–17. 109 For this suggestion, Heslop, ‘Constantine and Helena’, 170–2. 110 T. A. Heslop, ‘A Walrus Ivory Seal Matrix from Lincoln’, Antiquaries Journal, lxvi (1986), 371–2; Bates, Remigius, 7–8. The year 1080 is also possible for the visit, but the level of collaboration in 1071–2 makes the early date more likely. 106 107

from crisis to triumph

359

at Winchester, and then with a much larger group of bishops and abbots at Windsor at Whitsun.111 According to the two diplomas that record the settlement, the hearing took place at Alexander’s command and with William’s agreement. Its terms confirmed Canterbury’s status as primate of the whole of Britain (primas totius Britanniae) and that the archbishop of York should profess obedience. York was awarded jurisdiction over only one English bishopric, Durham, and over the whole of Scotland. The archbishop of York and his suffragans were required to attend councils summoned by Canterbury. Lanfranc’s apparent victory was, however, qualified in two significant ways. In the first place, William intervened to stop Thomas having to swear an oath of obedience. While this was portrayed as a personal favour to Thomas and it was accepted that future archbishops of York could be asked for an oath, the precedent was one that Thomas’s successors were to be readily able to exploit. The diploma’s final clause to the effect that future archbishops of York should be ordained according to the canonical custom (et ab ipso ordinationem canonico more suscipiet) was also hardly watertight from Lanfranc’s point of view. The issues raised by this settlement rumbled on fairly unobtrusively during the rest of William’s life, but, since both archbishops outlived him, they were not raised again in acute form until after his death. Lanfranc’s dissatisfaction was, however, manifested by his immediate lobbying of the papacy for a formal confirmation of Canterbury’s permanent primacy, something that he failed to achieve either with Alexander or his successor Gregory VII.112 Lanfranc’s impotence suggests that William was at heart either sceptical or unconcerned about him securing permanent superiority. He is presented in Lanfranc’s account of the 1072 proceedings as intervening to make a settlement and to restore peace (pacem diligerem, concordiam facerem) and, in Lanfranc’s case, of him backing down on his claim to an oath from Thomas ‘out of love for the king’ (ob amorem regis).113 Since the subject of Canterbury’s Primacy had first been raised in 1070, William’s main interventions are often expressed in terms of anger, but, as seems always to have been recognized, with the objective always being to produce a settlement. At some point, the settlement looks to have been followed by a further, albeit undated, council of the English Church. Its legislation was less far-reaching than that of 1070, perhaps because it was

111 For a diplomatic analysis of the two diplomas that record the settlement of the Primacy Dispute, CS, 591–6, 601–4; Regesta, nos. 67, 68. See also Lanfranc Letters, no. 3. 112 Lanfranc Letters, nos. 5, 6. 113 Ibid., no. 4, lines 91–2; no. 3, line 124.

360

william the conqueror

thought that serious reform was now in train. Some of the canons parallel those of the contemporary Norman council of Rouen.114 Another facet of the 1072 settlement was the clarification of the English archbishops’ religious responsibilities across the entire British Isles, an aspect of the hegemonic powers claimed by the English kings, something that diplomas drawn up in William’s name before 1070 had already alluded to, and which was another part of his perceived inheritance from the pre1066 past. In pursuance of this hegemony, preparations to assemble an army and a fleet for an invasion of Scotland which, according to John of Worcester, took place after 15 August, would have been ongoing at the very time of the ecclesiastical councils.115 Alongside the longer-term objectives of consolidating what William would have seen as legitimately due to him, the campaign was also the logical resumption of unfinished business. King Malcolm (Mael Coluim) was still providing a refuge for English rebels and, in either 1070 or 1071, he had overtly publicized his links with them by marrying Edgar the Ætheling’s sister Margaret. According to the Durham Historia Regum, Malcolm had in 1070 followed up on William’s destruction of northern England by devastating lands in Stainmore and Teesdale to the north of the area of the Harrying, seizing large numbers of slaves in the process. This raid, apparently representative of different cultural values because the Normans no longer enslaved in this way, should probably also be seen as a provocative affront to William’s claim to rule in these regions; according to the Historia Regum, it was the savagery of Malcolm’s incursion that provoked the invasion of Scotland.116 It was also a piece of transparent opportunism on Malcolm’s part; William’s destruction of northern England would have created conditions that, for some, might even have made slavery in the kingdom of Scots seem like a preferable alternative to eking out a livelihood in a devastated land. Gospatric, at this point still William’s earl in the north, had made some attempt at resistance by devastating Cumbria, at that time under Malcolm’s rule, an act that was if anything a provocation rather than a deterrent to Malcolm’s attacks.117 The sources for the invasion of Scotland, the two versions of the Chronicle and John of Worcester, whose text was reproduced verbatim in 114 Martin Brett, ‘A Collection of Anglo-Norman Councils’, JEH, xxvi (1975), 301–8, with a date of 1072 suggested at p. 308, note 1. 115 JW, iii, 20–1. 116 HRA, ii, 191–2, 195–6 (Grauiter namque rex Scottorum Malcolmus eum offenderat, quia, ut supradictum est anno praeterito, regni suo terminus atrociter depopulates fuerat). On this Durham passage, John Gillingham, ‘Conquering the Barbarians: War and Chivalry in Britain and Ireland’, HSJ, iv (1992), 67–84, at 71, 73 (idem, The English in the Twelfth Century, 41–58, at 45, 47); David Wyatt, Slaves and Warriors in Medieval Britain and Ireland. 800–1200 (Leiden and Boston, MA, 2009), 367–8. 117 HRA, ii, 191; William M. Aird, ‘Gospatric, Earl of Northumbria (d.1073x5), ODNB.

from crisis to triumph

361

the Durham Historia Regum, say simply that the two kings met at the great religious site of Abernethy on the Tay and that Malcolm submitted to William and became his man.118 The Annals of Ulster add that one of Malcolm’s sons was handed over as a hostage; we know from other sources that this was Duncan (Donnchad), a son by his first marriage to Ingibjorg, either the widow or daughter of Thorfinn, earl of Orkney, later to be briefly King of Scots.119 Such a submission had presumably always been William’s objective. And if it was not, Malcolm’s refusal to risk a battle soon ruled out any more ambitious schemes William might have had; he certainly would not have wanted to become bogged down in a long war of attrition. By crossing the Forth, he had moved into the heartlands of Malcolm’s kingdom, but an advance to Abernethy did not constitute much of an inroad. William had, however, shown that he was capable of invading Scotland. Given William’s recent record and the likelihood that to have joined battle with him would have led to defeat, the outcome was probably also a desirable one for Malcolm.120 With Duncan (Donnchad) placed in the custody of Count Robert of Mortain and his status so clearly recognized that he appears as a landholder in Domesday Book, (albeit a very minor one, and was treated as a member of the aristocratic elite), a mutually understood collaborative framework looks to have been established.121 Since William is known to have been on his way south in late October, the short duration of the campaign probably shows that both he and Malcolm had a clear idea of what they had always wanted out of the situation. The settlement freed Malcolm to concentrate on problems within his kingdom that had a history going back to the supplanting of Macbeth (died 1057) and his stepson Lulach (died 1058), and conferred legitimacy in relation to the new regime that he would certainly have grasped was now becoming securely established in England.122 No source mentions whether Edgar the Ætheling was discussed at Abernethy. As we shall see, he next appears in 1074 in Flanders, presumably trying to recruit support for his cause among ASC, ‘D’ and ‘E’; JW, iii, 20–1. The Annals of Ulster to A. D. 1131, 508–9. For Duncan (Donnchad), Stephen Marritt, ‘Coincidences of Names, Anglo-Scottish Connections and Anglo-Saxon Society in the Late Eleventh Century’, Scottish Historical Review, lxxxiii (2004), 150–70, at 151–9. 120 For differing perspectives, A. A. M. Duncan, The Kingship of Scots, 842–1292: Succession and Independence (Edinburgh, 2002), 44–6; Dauvit Broun, Scottish Independence and the Idea of Britain from the Picts to Alexander III (Woodbridge, 2007), 101–3; Alexander Grant, ‘At the Northern Edge: Alba and its Normans’, in Stringer and Jotischky (eds.), Norman Expansion, 49–85, at 49–50. 121 DB, i, fo. 93r; Marritt, ‘Coincidences’, 155–9. 122 G. W. S. Barrow, ‘Malcolm III [Mael Coluim Ceann Mór, Malcolm Canmore] (d.1093)’, ODNB, supplies the basic material. 118 119

362

william the conqueror

William’s many potential enemies, a sign perhaps of an ambivalence on Malcolm’s part that had not been present in 1069–70.123 With the scholarship of the past thirty years having decisively demolished the old perspective that the Abernethy agreement established some form of ‘feudal’ dependence of the Scottish kings on the English, we can be certain that William would have seen Malcolm’s submission as restoring the generalized form of superiority of status that William’s English predecessors had enjoyed.124 He and Malcolm would have thought in terms of a modus vivendi that served both of their interests. Taken alongside the creation of the compact lordships along the borders with the Welsh kingdoms, William would most likely have seen this campaign as stabilizing his authority over the British Isles. An indication of changing times is that John of Worcester mentions that Eadric the Wild was a member of William’s army on this campaign.125 On his way south William again, and this time finally, removed Gospatric from the earldom of Northumbria, replacing him with Waltheof. The Durham Historia Regum says that this was because Gospatric had participated in the massacre of Robert de Commines and his men and because of his earlier rebellion, adding that the first of these charges was false.126 In choosing the son of the Confessor’s most successful earl in the far north who was also his nephew-in-law, William may have thought that he had at last appointed someone as likely as anyone to appease the multiple resentments felt across the northern British Isles, as well as someone whose link to him personally, combined with his power base in the more southerly regions of England, gave him both the prestige to carry weight and the incentive to remain loyal. The Historia Regum presents Waltheof as working well with Bishop Walcher. But events, specifically the massacre of Settrington (Yorkshire) of 1073–4, were to show that Waltheof became tangled up in the legacy of past feud-related histories and could not be politically neutral in his relations with his new charges.127 Gospatric in the meantime went north again, then visited Flanders, and finally established himself in the kingdom of Scots receiving Dunbar from Malcolm. Also on his journey south, around 1 November, William visited Durham where, according to Symeon, he expressed scepticism about St Cuthbert’s sanctity and ordered that the tomb be opened. Immediately stricken with a burning fever, he rode south as fast as he could, leaving behind an See, e.g., Hooper, ‘Edgar the Ætheling’, 205. For the long-term picture, George Molyneaux, ‘Why Were Some Tenth-Century Kings Presented as Rulers of Britain?’, TRHS, 6th ser., xxi (2011), 59–91, at 75–9. 125 JW, iii, 20–1. 126 HRA, ii, 196. 127 Ibid., ii, 200; Fletcher, Bloodfeud, 187–93; Lewis, ‘Waltheof, Earl of Northumbria’. 123 124

from crisis to triumph

363

unfinished feast.128 This version of events is, however, contradicted by what is thought to be an earlier source, which says that William listened reverently to the histories of the Durham saints and was so moved that he placed a mark of gold and a valuable cloak on Cuthbert’s tomb. What are said to be his words are then recorded as he announced that he was confirming in perpetuity the church’s property and privileges. He then granted Waltham (Essex), the place of Harold’s burial, to Bishop Walcher to be used as a residence when he visited London.129 There is no other record of William being suspicious of saints, particularly ones with pedigrees as good as Cuthbert’s. The church of Durham’s involvement in earlier revolts might have persuaded him to treat it to a display of severity along the lines of, for example, Ely. Alternatively, since so many sources see events from William’s perspective, this tale of a rare moment of discomfiture could be an act of regality that went wrong, a piece of tactlessness that offended ecclesiastical sensibilities. Ever ready to recount and amplify St Cuthbert’s powers, Symeon probably either exaggerated or misrepresented a trivial incident. In all probability after the Windsor Whitsun council, and perhaps while William was campaigning in Scotland, proceedings began concerning disputes over property and jurisdiction involving the archbishopric of Canterbury, William’s half-brother Odo, the earl of Kent, and the king himself. Although there are discrepancies in the surviving accounts, most of which date from the early twelfth century and reflect the way in which pre-1066 and post-1066 material was ever more imaginatively reconstructed, it is probable that the starting point was an enlarged gathering of the shire court of Kent, held at its traditional meeting place on Penenden Heath near Maidstone.130 William’s delegation of royal power to Bishop Geoffrey of Coutances is yet another sign of the trust placed in specific chosen individuals, while also showing that he acted under the umbrella of royal authority; the reports state William did at the end give his approval to the verdict. The way in which Lanfranc and Odo also accepted that Bishop Geoffrey should hear the plea is a further indication of how LDE, 196. H. H. E. Craster, ‘The Red Book of Durham’, EHR, xl (1925), 504–32, at 528; The Early Charters of the Augustinian Canons of Waltham Abbey, Essex, 1062–1230, ed. Rosalind Ransford (Woodbridge, 1989), xxxv, no. 10. For recent discussion of the Durham sources, LDE, lxxiv, 197, note 72. On the incident, Aird, St Cuthbert and the Normans, 85–91. 130 For the reports of the trial, Regesta, no. 69 (English Lawsuits from William I to Richard I, ed. van Caenegem, i, nos. 5, 5A, 5B, 5C). On the texts and the disputes, with references to earlier literature, David Bates, ‘The Land Pleas of William I’s reign: Penenden Heath revisited’, BIHR, li (1978), 1–19, with the date discussed at pp. 12–13. See further Alan Cooper, ‘Extraordinary Privilege: The Trial of Penenden Heath and the Domesday Inquest’, EHR, cxvi (2001), 1,167–92; Paul A. Hayward, ‘Some Reflections on the Historical Value of the So-Called Acta Lanfranci’, HR, lxxvii (2004), 141–60, at 152–3. 128 129

364

william the conqueror

William’s personal authority was channelled and respected, with such delegation of responsibility becoming even more a central feature of rule over his extensive lands in the 1070s. The plea’s concern with title to jurisdictional rights and lands, many of which had been lost to the church of Canterbury long before 1066, shows how the newcomers were inserting themselves into institutional structures that had existed for a long time; some aspects of the business indeed harked back to the time of Archbishop Æthelnoth, who had died in 1040. The trial was conducted according to existing English procedure at the traditional meeting place of the shire court of Kent, with Englishmen present and the aged and deposed bishop of Selsey Æthelric brought along in a cart to pronounce on English law. It is possible to see the dispute between Lanfranc and Odo as a fissure in the political elite on which William depended, but this is debatable because the landholdings and tenurial links in Kent of Odo and Lanfranc’s men were so entangled that the plea cannot be seen as a straightforward confrontation between two groups. The two men’s wider interests, as well as their partnership in Kent as earl and diocesan, also put a premium on collaboration. Even though the Penenden plea lasted for three days, it did not conclusively resolve many of the matters in dispute. It is, however, another facet of how the political and economic interests of the newcomers were becoming embedded in pre-existing conditions and into the complexities of the local societies they had moved into. New social relationships were being formed, not just by the violence of conquest, but around the existing nexuses of the traditional courts of shire and hundred. The abstract ideology of conquest and domination is seen transformed into a framework for conflict resolution, the newcomers participants in disputes of which they would have known nothing only a decade previously. Another land plea, the first inquest into the territorial losses suffered by the abbey of Ely as a result of the siege of 1071, although dateable only to between 1071 and 1075, had a similar character to the Penenden plea; Geoffrey of Coutances again oversaw it, supported by the appropriate regional authorities, the diocesan Bishop Remigius and the relevant earl, Waltheof, as well as two sheriffs.131 It is not known where William spent Christmas 1072 or whether he crossed to France before or after Christmas. The next certain recorded date in his itinerary is 30 March 1073, when he was at Bonneville-surTouques to add his confirmation to the two foundation charters of the abbey of Saint-Pierre of Solesmes in Maine. Although not without their 131 Regesta, no. 117. For recent comment on the Ely land pleas, Keynes, ‘Ely Abbey 672–1109’, 48–50.

from crisis to triumph

365

problems and surviving only in late copies, these confirmations, along with a similar one for the abbey of Saint-Vincent of Le Mans, which may be of the same date, were probably a prelude to the reconquest; one shows that Bishop Arnold of Le Mans visited William.132 Although they have often been used to argue that William completed the reconquest of Maine in the winter and spring of 1073 – since, so it was thought, they are unlikely to have been confirmed before William’s authority was restored – when taken together with a statement in the Acta Archiepiscopum Rotomagensium that William was in Maine on 24 August 1073 and that Abbot Nicholas of Saint-Ouen of Rouen was there with him, they become part of preliminary negotiations to a campaign that took place in the summer.133 So viewed, the charters also fit in with the statement in the Actus of the bishops of Le Mans that in victory William agreed to respect the existing customs, and with the suggestion that the Actus was written in a way intended to conceal the political role of Bishop Arnold, whose appointment had been sponsored by William.134 The meeting with unnamed Manceaux in the winter of 1071–2 reported by Orderic might well have been an even earlier stage in these contacts. A summer dating also fits with the fact that we know nothing about any possible alternative activities by William at that time. It is also possible that a dedication of the partly completed abbey church of Saint-Etienne of Caen took place before the campaign, with mention of a dedication appearing in the version of the Norman annals compiled at Saint-Etienne – good evidence that one happened. If so, it probably marked the completion of the choir.135 William’s presence in Lower Normandy at Bonneville-sur-Touques at Easter does make a ceremony at Caen at this time likely. On the other hand, the pattern of dedications that occurred in 1063 and 1066 suggests that holding one either to 132 William’s confirmation on 30 March 1073 is specifically recorded on the two Solesmes charters, Regesta, nos. 274, 275. One difficulty is that he is described in the second of them as ruling Maine (qui Willermus Cenomannensis ciuitatis tunc principatum tenebat). The fact that we are dealing with a fifteenth-century copy of an early eleventh-century diploma onto which William’s confirmation was added at the foot allows for the statement to have been added by a scribe who wished to give authority to William’s confirmation. The Saint-Vincent charter (Regesta, no. 173), while confirmed by William at Bonneville, need not necessarily date from 1073. In this case, we are dealing with a seventeenth-century copy of a lost twelfth-century cartulary. 133 AAR, 41, 55; cf. Latouche, Maine, 35–8; Douglas, WC, 229, note 3; Aird, Robert Curthose, 68–9; Lemesle, La société aristocratique, 35. 134 Actus Pontificum Cenomannis in urbe degentium, 381; Bruno Lemesle, ‘Le discours de l’Eglise aux temps grégoriens: évêques et laïcs dans le Maine aux XIe–XIIe siècles d’après les Actus Pontificum’, Annales de Bretagne et des pays de l’Ouest, cii (1995), 17–32, at 19–24. 135 ‘Annalis Historia Brevis sive Chronica monasterii S. Stephani Cadomensis’, 166 (Hoc anno dedicata est basilica S. Stephani Cadomi); Carlson, The Abbey Church of Saint-Etienne at Caen, 104–5.

366

william the conqueror

prepare for a campaign or as thanksgiving for success in one was a norm. A date in the late summer or autumn is therefore also possible. The winter and spring of 1072–3 would therefore have been spent assembling an army for the invasion of Maine and in preparatory diplomacy. The presence of English troops in this army is a sign that William was by this stage mobilizing the resources of his cross-Channel lands to suit his wider purposes. The initial apparent unity of the rebels had by this point proved to be very superficial; it is likely that some of the chief powers in Maine had come to believe that the political instability since the revolt of 1069 made William’s intervention desirable.136 Azzo d’Este quickly returned to Liguria, apparently short of money, leaving behind his wife Gersendis and their son Hugh to rule in collaboration with Geoffrey de Mayenne. Gersendis soon became Geoffrey’s mistress. They were ejected from Le Mans in 1070 by a sworn commune of the citizens, but had regained control of the city by the end of the year. In 1072 the citizens rose again against them and invited Count Fulk Rechin of Anjou to intervene, which he did, but only briefly, possibly deciding that to resolve the situation was beyond his means so soon after he had deposed his brother.137 If negotiations had taken place with William at the turn of the year, Count Fulk may also have decided to avoid a direct confrontation with him. The main sources indicate that William intervened with an overwhelmingly large army. He marched south along the Sarthe, first besieging Fresnay-sur-Sarthe, where the castellan, Hubert vicomte of Le Mans, who had, along with Geoffrey de Mayenne, been William’s main opponent in 1062–3, surrendered quickly and also handed over the castle of Beaumontsur-Sarthe. At Fresnay, William knighted Robert de Bellême, the eldest son of Roger de Montgommery, a sign of the emergence of the younger generation of the Norman aristocracy and a reminder also that the generation which had accomplished the conquest of England was now an ageing one. The castle of Sillé also capitulated quickly and, after William had surrounded Le Mans, the city sought terms the next day. William, so we are told, agreed to an amnesty and to respect existing customs. He also took steps to repair war-damage to the cathedral church.138 During the campaign, he had followed normal military practice and devastated the countryside, the writers of both versions of the Anglo-Saxon Chronicle taking particular pride in the damage which the English troops in his army inflicted, a passage written either as irony or as a sign that this sort of 136 The main sources for William’s reconquest of Maine are Actus Pontificum Cenomannis in urbe degentium, 377–81; OV, ii, 306–9. 137 Lemesle, La société aristocratique, 35–6. 138 Regesta, no. 169; Nécrologe-obituaire de la cathédrale du Mans, ed. Gustave Busson and Ambroise Ledru (Le Mans, 1906), 238.

from crisis to triumph

367

destruction was accepted as a norm.139 Orderic believed that the victory was an easy one.140 But no sanctions that we know of were imposed by William against those who had opposed him. On his return to Normandy, William went to Rouen to adjudicate a dispute that had broken out between Archbishop John and the abbey of Saint-Ouen, the cause of contention being a brawl that had occurred when the archbishop arrived late to celebrate mass in the abbey on St Ouen’s feast-day (24 August); he had apparently scurried back from Maine, where he had been advising William, to take his customary place, only to find that the monks had asked Robert, abbot of Saint-Martin of Sées, to deputize in his absence. The fight, which was caused either by the monks or by John’s abrasive behaviour, dependent on which source is consulted, was extensively written about, with the most detailed account being in the Acta Archiepiscopum Rotomagensium.141 It was so serious that it was only quelled by the intervention of the vicomte. When William, still in Maine, heard about what had happened from an emissary sent by Archbishop John, his first reaction was to reinforce episcopal authority by dispatching the four monks who were judged to be culpable to four different abbeys. Later, however, he presided over a synod at Rouen that reversed this decision and fined the archbishop. John seems to have been deeply affected by this incident, with five letters from Lanfranc offering him advice and pastoral counsel over a considerable period. It is quite likely that the escalation of the riot also involved canons of the cathedral and was effectively a continuation of the disturbances of the previous year.142 The impact of the event may also explain John’s observation during a later dispute heard before William and Matilda that there were churches in his diocese over which he had no rights at all.143 In terms of what all this tells us about William, it indicates once more a tendency to impose a rapid judgement that favoured existing authority – perhaps in this case an impulsive one – that was later followed by a formal hearing and a considered verdict after he had listened to advice. If he needed to learn a lesson from it, it was that the rigorous application of the reforms being sponsored by the papacy was an inflammatory route to follow; hence, perhaps, his turning away from his initial instinctive support for Archbishop John. 139

ASC, ‘D’, ‘E’. OV, ii, 308–9 (Pacatis itaque sine magno discrimine Cenomannensibus). 141 For the main sources, see AAR, 41–5, 55–9; ‘Chronicum Rotomagense’, in Labbé (ed.), Novae Bibliothecae Manuscript. Librorum, i, 367; ‘Annalis Historia Brevis sive Chronica monasterii S. Stephani Cadomensis’, 166. 142 On this episode, Alison Alexander, ‘Riots, Reform, and Rivalry: Religious Life in Rouen, c.1073–c.1092’, ANS, xxxiii (2011), 23–40, at 24–30. 143 Regesta, nos. 29(I), 29(II). 140

368

william the conqueror

Pope Alexander II’s death on 21 April 1073, followed by the well-nigh immediate election of Archdeacon Hildebrand, who took the name Gregory VII, ushered in what turned out to be a time of significant change. In the short term, however, since Gregory had been a long-standing supporter of William’s cause, his election must have been seen as a portent of continued good relations. A letter in William’s name was sent to congratulate him, while the new pope in turn wrote to William, Matilda, and Lanfranc. Within weeks of his consecration Gregory was referring to William as ‘that most dear and special son of the holy Roman Church’ in his letter to Lanfranc.144 To William, Gregory supplied a lengthy homily on the need to persevere in good works and said that he believed him ‘to stand alone among kings as the one whom, above all others, cherished the things that are written about above’.145 This statement is about as eloquent an assessment of the reputation that William had created for himself as we have. In response to William’s enquiries about the condition of the papacy, Gregory wrote back describing an institution assailed by ferocious storms and tossed on the sea by huge waves, harassed by hypocrites and avoiding with difficulty traps laid by heretics. To Matilda, he wrote more briefly as to a ‘beloved daughter’ and emphasized the standard wifely duties – even a religious husband, such as he obviously believed William to be, would benefit from a supportive, religious wife. Lanfranc was addressed as ‘dearest brother’, asked for his and his monks’ prayers, and, like William, given a rather pessimistic appraisal of present troubles.146 Other papal letters absolved Bishop Remigius of Lincoln of his sins and instructed Lanfranc to protect the abbey of Bury St Edmunds against the new bishop of Thetford Herfast, about whom Gregory was very rude. The clear emphasis is on continuity from Alexander II’s time.147 As we shall see, Gregory’s opinion of William either became more complex over time, or perhaps already was, but, from William’s perspective, all this must have been encouraging and reassuring after the successful completion of some hard campaigning and the defeat of many enemies. His status in Gregory’s eyes is eloquent testimony to the context that brought support to the conquest of England. THE YEAR 1074 We know that William crossed to England at some point late in 1073, since the first sentence in the Anglo-Saxon Chronicle’s entry for 1074 records that Registrum, i, 52, no. 1.31 (Register, 37). Registrum, i, 101, no. 1.70 (Register, 74). 146 Registrum, i, 100–3, nos. 1.70, 1.71 (Register, 73–4); Lanfranc Letters, no. 8. 147 Registrum, 51–2, 55, nos. 1.31, 1.34 (Register, 36–7, 39–40). 144 145

from crisis to triumph

369

he left England for Normandy.148 He presumably spent Christmas somewhere in the south of England, although there is no record to indicate where it might have been or what he did there. William was at Rouen in May 1074 and at some point between April and September he visited Lillebonne.149 On 30 November 1074 he was again at Rouen where he confirmed a purchase made for his cathedral by his half-brother Bishop Odo in the hall of the ducal palace.150 Although it is just possible that William crossed to England for Christmas, the fact that all the events mentioned below either occurred in Normandy, or in all probability did so, makes this unlikely. In addition to confirming Odo’s purchase at the end of November, William also made a substantial grant to his cathedral, in doing so disinheriting the heirs of the rebel of Val-ès-Dunes, Grimoult du PlessisGrimoult. In the other diplomas of the year, he confirmed the possessions of the abbey of Saint-Wandrille and intervened to prevent trial by battle between the champions of Saint-Wandrille and William, count of Evreux.151 The signa of the four Norman diplomas definitely dateable to this year show many of the leading members of the Anglo-Norman aristocracy in the duchy. Roger de Montgommery, Roger de Beaumont, Count Robert of Mortain, and Count Robert of Eu all appear alongside William and Matilda, their two sons Robert and William, and Bishop Odo.152 Although there are dangers present in arguing from the non-appearance of particular individuals, the absence of Bishop Geoffrey of Coutances is striking, especially since all six of the other Norman bishops were present at Rouen in May 1074.153 Richard fitz Gilbert and William de Warenne are also absent, suggesting that all three and probably Earl Hugh of Chester remained in England. The two Saint-Wandrille diplomas contain triumphalist language. In one, William is ‘the undefeated’ (inuictus), and in the other there is a dating clause that refers to him as the ‘glorious’ (gloriosus) king and locates the proceedings to the eighth year after William had taken over the kingdom of the English.154 There are signs here of a family and an aristocratic elite gathered together to celebrate triumph. This was also the year when Edgar the Ætheling finally submitted to William. The main source for these events, the ‘D’ Chronicle, records that Edgar returned to King Malcolm’s court from Flanders on 8 July 1074. 148

ASC, ‘D’, ‘E’. Regesta, nos. 261, 262. 150 Ibid., no. 26 (in aula turris). 151 Ibid., nos. 26, 27, 261, 262. 152 For Roger de Montgommery, ibid., nos. 26, 27, 262; Roger de Beaumont, ibid., nos. 26, 27, 262; Robert of Mortain, ibid., nos. 26, 27; Robert of Eu, ibid., no. 26. 153 Regesta, no. 261. 154 Ibid., nos. 261, 262. 149

370

william the conqueror

Since Robert, count of Flanders, was among William’s enemies by this point, it is likely that Edgar had embarked on a campaign to recruit allies on both sides of the Channel and around the North Sea. The offer from King Philip I of the castle of Montreuil in Picardy within reach of the Norman frontier must have been both part of this and of Philip’s plans to assert himself around northern France. The regular presence of several of the territorial rulers with lands to the north and east of Normandy among the signa of Philip’s charters is striking testimony to how all of them tried to maintain a balancing act between rivals. William’s enhanced power had evidently not increased his prestige and influence so much that he could rely on loyalties beyond Normandy.155 Provided, so we are told, with great treasure by the Scottish king, Edgar set sail for France only to suffer shipwreck in the North Sea and to return in a miserable state to Scotland. This time Malcolm and Queen Margaret dispatched him south to submit to William. Received by the sheriff of York at Durham, he was escorted from there to William in Normandy, where, so the ‘D’ Chronicle says, the king received him with great honour and gave him a place at court.156 The ‘E’ Chronicle was, however, less impressed by how Edgar was treated, saying he received such rights as the king would grant him; his discernible rewards in Domesday Book amounted to two estates in Hertfordshire.157 Whatever the authors of the ‘D’ and ‘E’ Chronicles thought, however, William’s treatment of Edgar worked. He took no part in any further rebellion or invasion during William’s reign, was described by Orderic as virtually the foster-brother of William’s oldest son Robert, and remained at court until 1086 when he set off for a spell in southern Italy, a habitué of the Norman world, or so it appears. It cannot, however, be an accident that he never married. This was surely deemed impossible by William. Malmesbury, in line with his opinion that the Ætheling, like his father, was morally unsuited to kingship, records that he was still alive in the 1120s, living out an inglorious old age in the countryside.158 This is arguably rather unfair on someone who had led a remarkably active life. But it is notable that his efforts never attracted truly convincing support, before, in, or after 1066. Orderic, while praising his eloquence, deportment, and activity, also thought him to be lazy; meaning presumably that he did not have the mental toughness that distinguished

Tanner, Families, Friends and Allies, 126–7. ASC, ‘D’, 1074. 157 DB, i, fo. 142r; OV, v, 272–3 (quasi collactaneum fratrem). In general, Hooper, ‘Edgar the Ætheling’, ODNB. 158 GR, i, 416–17, 466–7. 155 156

from crisis to triumph

371

two of the other kings of the year 1066, Harold and William, and made them convincing rulers.159 In 1074, William was also invited to intervene in the civil war that had broken out in Germany. Bruno of Merseburg, who wrote in 1082, says that William was one among several princes from whom King Henry IV sought support, and that William turned him down on the grounds that he feared that if he left England, he would lose it.160 A different story, told by the contemporary chronicler Lampert of Hersfeld, recorded that one of Henry’s main opponents, Archbishop Anno of Cologne, had persuaded William to help him. The rumour that he was coming was sufficiently strong to persuade King Henry to move from Regensburg to the Rhineland to spend Easter at Mainz.161 With a later reworking of the story describing the rumour as being a false one, it is clear that William did not undertake military preparations; the same source, the Vita of Archbishop Anno does, however, mention in an unspecific way that there was contact between Anno and William.162 William’s reasons for not becoming involved are an illuminating comment on how he perceived his standing in his new kingdom, and a warning against thinking with the benefit of hindsight that he would by this stage have felt fully secure. The way in which William was hereafter sought out by both sides in the forthcoming conflict between the King of Germany and the pope nonetheless indicates that William was by this stage thought of as a major player in European politics. They are eloquent commentaries on how he was now perceived. Another sign of William’s by now immense prestige is his entry into the confraternity of the great abbey of Cluny, which happened in either 1074 or 1075, apparently at his own request. The scene of his admission, described in the life of Abbot Hugh of Cluny known as ‘Anonymous II’, is another demonstration of William’s theatrical humility in the presence of an especially religious individual (Plate 18). In Abbot Hugh’s absence, the monastery was represented by Warmund, abbot of Déols, who subsequently became archbishop of Vienne in 1076. Receiving the abbot, William removed his crown and bowed his head as if he was being invested by an angel. To his amazed nobles, he is said to have explained that the gift of confraternity was greater even than the kingship he had gained, because it OV, v, 272–3. Bruno, ‘De Bello Saxonico’, ed. Franz-Josef Schmale, in Quellen zur Geschichte Kaiser Heinrich IV (Darmstadt, 1963), 192–405, at 242–3. 161 Lampert of Hersfeld, Annales, ed. Schmidt and Fritz, 252–3 (The Annals of Lampert of Hersfeld, trans. Robinson, 235). Both episodes are discussed in van Houts, ‘The Norman Conquest through European Eyes’, 840. 162 Vita Annonis archiepiscopi Coloniensis, ed. G. H. Pertz, MGH Scriptores, xi (Hanover, 1854), 462–514, at 495. 159 160

372

william the conqueror

was a means to heavenly glory.163 He and Matilda then made lavish gifts to the abbey. Other members of Cluny’s confraternity included the Salian emperors, kings and queens from Spain, and many French princes.164 All this was truly remarkable for a man who, within the kingdom of France, had not apparently ventured beyond Normandy’s borders since his visit to Henry I’s court at Vitry-aux-Loges in September 1052 except to make war, and whose monastic patronage had apparently not stretched beyond Marmoutier on the Loire. The changed situation is reflected by William of Malmesbury’s comment on the cosmopolitan nature of his court.165 Evidently there was no need for William to travel. The world would come to him. There are obvious hints in all these developments of the changing times that were to cloud William’s last years. As he aged, new personalities were coming on to the scene. Philip I’s gesture on behalf of Edgar the Ætheling was indicative that the modus vivendi between him and William of the previous decade was breaking down and the hostilities of pre-1060 set to be resumed. Count Fulk Rechin’s intervention in Maine presaged a revival of conflict between Normandy and Anjou. In general, as both rulers ceased to be preoccupied with other business, their policies reverted to those of Henry I and Geoffrey Martel. It may well be that William’s reassertion of his power over Maine, traditionally an area of royal and Angevin influence, was an intolerable provocation for both of them. The death on 8 September 1074 of Count Ralph of Amiens, Valois, and the Vexin removed someone who in his last years had been a firm ally on Normandy’s eastern border. His successor, his son Simon, who had been brought up in William’s court, was immediately assailed by King Philip, but, holding his own, suggested at this point that he might well maintain the status quo. The turbulence of Pope Gregory’s thoughts and his already strained relations with King Henry IV were the first signs of the great conflict to come. And revolt in England was about to erupt again, albeit this time organized by a combination of members of the new elite whom William had himself either installed or promoted.

163 Vita sancti Hugonis abbatis Cluniacensis, PL, clix, cols. 923–4. On this episode and its date, Barlow, ‘William I’s Relations with Cluny’, 132–3; Cowdrey, ‘William I’s Relations with Cluny Further Considered’, 12–13 (idem, Crusades and Latin Monasticism, 11–12th Centuries, chapter 8). 164 H. E. J. Cowdrey, ‘Unions and Confraternity with Cluny’, JEH, xvi (1965), 152–62, at 161. 165 GR, i, 506–7.

Chapter 10

C RO S S - C HAN N E L RU LE Between the end of the year 1074 and the late summer of 1080, William spent the great majority of his time in Normandy and northern France. And the one visit to his kingdom that certainly took place, which was between the autumn of 1075 and the summer of 1076, was provoked by a crisis, the threat of another invasion from Denmark and the need to clear up after a rebellion.1 A second visit possibly took place at the end of 1077, presumably around Christmas, because the famous letter refusing fealty written on William’s behalf to Gregory VII, which must date to the second half of the year 1080, includes the statement that he had then been absent from his kingdom for almost three years.2 Even allowing for the fact that in terms of the precise dates needed truly to chart an itinerary, 1078 is one of the most obscure of the post-1066 years, it looks as if William spent at most approximately twelve months out of a possible sixty-eight in England, that is, around 17.5 per cent of the time available. One determinant of this remarkable imbalance was certainly William’s quarrel with his eldest son Robert, which in all probability began during 1078 and was not finally resolved until well into 1080.3 It was deeply destabilizing both within his family and within the Norman aristocratic elite, and it gave opportunities to William’s northern French opponents that some set out to exploit. Another was the erosion of alliances and networks of friendship that had stabilized Normandy’s frontiers during the previous decade and which encouraged King Philip I and Count Fulk Rechin to become militarily active against William. Also, as he advanced towards and moved beyond fifty years of age, William had to confront the ambitions and aspirations of the next generation. For all this, these years were a time of significant developments in the infrastructure and practice of rule in England, as also, to an extent, they 1 This calculation is based on William’s crossing to England to confront the possible Danish invasion taking place in September 1075 and him leaving England soon after the end of June 1076. Unrecorded short visits to England are possible, but none is known. See further the narrative below and Regesta, 77, 80–1. 2 Lanfranc Letters, no. 39 (tribus ferme annis in Gallia me agente). 3 For the chronology of Robert’s first revolt, Lack, ‘Robert Curthose’, 126–8; Kathleen Thompson, ‘Robert [called Robert Curthose], Duke of Normandy’, ODNB.

373

374

william the conqueror

were in Normandy. The year 1077 and the final months of 1080 have much of the same aura of magisterial display about them as the year 1074 had had, with 1077 being notable for several spectacular church dedications, including those of the abbey church of Saint-Etienne of Caen and Bayeux cathedral. War, or more often the threat of war, might well have been a frequent presence during this period, but most of the confrontations that occurred were conducted in terms of a kind of point-scoring stasis, with no enemy or potential enemy of William having the resources to do anything likely to bring about major change. 1075: ESTABLISHING PRIORITIES Christmas 1074 is one of the great religious festivals after 1066 when William’s whereabouts are not known and cannot even be surmised. It is possible that he crossed to England for a crown-wearing, but he was certainly back in Normandy – if indeed he ever left it – by Easter (5 April), when he was at Fécamp. There he prohibited private vengeance in cases of murder except to the father or a son of the victim, and his and Matilda’s daughter Cecilia was ordained as a nun of La Trinité of Caen by Archbishop John of Rouen.4 The former, an addition to ducal peace-keeping powers that limited legitimate vengeance to close kindred, shows how a dominant figure such as William could achieve a measure of social control within a society in which honour-based violence was culturally accepted; it is nonetheless salutary to reflect that it would not have made illegal many of the killings which had taken place during William’s late childhood. Cecilia’s entry into religion fulfilled the vow made in 1066 when she was handed over as a child oblate. A poem by Fulcoius of Beauvais, which compared William and Matilda’s act to the sacrifice of the Old Testament king Jephthah, was in all likelihood written at this time.5 Although the stories are not strictly comparable – William and Matilda had consciously committed their daughter to a religious life whereas Jephthah was trapped by a vow to sacrifice the life of whoever came out first to meet him after a military victory – the poem’s inclusion of a reference to emotional loss may be significant. If its sentiments are taken seriously, William and Matilda must be seen as parting from a dearly loved daughter who was henceforth going to be an enclosed nun. However, since Cecilia’s claustration would have been widely, if not indeed universally, viewed as an act of public duty 4 ‘Annales Uticenses’, in Orderici Vitalis Ecclesiasticae Historiae, v, 158; ‘Annalis Historia Brevis sive Chronica monasterii S. Stephani Cadomensis’, 166–7; OV, iii, 8–11. 5 ‘Fulcoii Belvacensis epistolae’, ed. Colker, 245–6 (trans. van Houts, The Normans in Europe, 132–3).

cross-channel rule

375

that both parents thought an indispensable element in securing divine favour for the conquest of England, an emotional aspect was only a part of the story. Fulcoius’s reference to ‘my holy wretched mother’, thereby adjusting the Jephthah story to focus on Matilda’s distress, may well be very significant. As well as drawing attention to her feelings and, by its inclusion, to her political standing, it suggests a likely role as the emotional heart of her large family. As Orderic noted, the best possible provision was made for Cecilia’s education; she is known to have reached a level at which she could read and appreciate sophisticated Latin poetry. According to Guibert of Nogent, her teacher was the Fleming and, after the First Crusade, patriarch of Jerusalem, Arnulf de Chocques.6 Her long and apparently dedicated career as a nun and her eventual election as second abbess of La Trinité on 6 July 1113 suggest that Cecilia was fully accepting of her fate.7 But in terms of how she perceived her situation, it must be significant that, though she had withdrawn from the world, she continued to proclaim publicly her status within the royal kindred throughout her life; she described herself in charters not only as a nun and subsequently as an abbess, but also often as ‘the king’s daughter’ (filia regis), a practice that sustained memory of La Trinité as her parents’ foundation and its special place within their personal landscape of rule.8 Also in 1075, in Matilda’s presence, William’s aunt Adeliza gave to La Trinité the castle of Le Homme, the source of so much difficulty around the time of her son Guy of Brionne’s revolt in 1047.9 The date of Adeliza’s return to Normandy and her reinstallation at the place she had bought from William’s father before 1035 is unknown; it presumably happened after her husband Count Rainald’s death in 1057. Her gift to La Trinité illustrates how the abbey had come to function as a focus for the piety of women in the ducal house, as well as for members of aristocratic families.10 As an exemplification of this role and of the female solidarities that added to the social cohesion of Norman society, the castle did not immediately transfer to La Trinité, but was instead passed to William’s sister Adelaide,

6 OV, iii, 10–11; Guibertus abbas S. Mariae Nogenti, Gesta Dei per Francos, ed. R. B. C. Huygens, CCCM, no. 127A (Turnhout, 1996), 290–1 (The Deeds of God through the Franks: A Translation of Guibert of Nogent’s ‘Gesta Dei per Francos’, trans. Robert T. Levine [Woodbridge, 1997], 135). 7 For Cecilia’s education and career, Walmsley, ‘The Early Abbesses’, 427–33; LetouzeyRéty, Ecrit et gestion, 138–41, 217–50. 8 AAC, nos. 25, 27; Charters and Custumals of the Abbey of Holy Trinity Caen, Part 2: The French Estates, ed. John Walmsley, Records of Social and Economic History, new ser., no. 22 (Oxford, 1994), 131–2 (Cartulary Documents, no. 20). 9 AAC, no. 21; Regesta, no. 58. 10 AAC, 47–8.

376

william the conqueror

countess of Aumale, on a life-lease.11 It is likely, too, that the unfortunate Adelida, although she had retired to the abbey of Saint-Léger of Préaux, was buried there at some point during this period, to be joined later by her mother and her sister Cecilia in the mausoleum for the female members of the family.12 The record of a council of the English Church held in London in 1075 says that William was abroad at the time and that he was engaged in a war. Although the surviving account of the council is dated only by its year, it is a reasonable surmise that it was held at Easter, since two of the attendees, Lanfranc and Bishop Wulfstan of Worcester, were directly involved in combatting the uprising known as the Revolt of the Three Earls in the summer. Lanfranc’s responsibility to co-ordinate operations in William’s absence and Wulfstan’s to be one of the leaders of an army in the West Midlands must surely rule out their attendance at an ecclesiastical council at Whitsun.13 It is not, however, in any way clear against whom William was fighting, especially given that we know he spent Easter at Fécamp. While Orderic says that the Revolt of the Three Earls was breaking out concurrently with William’s wars against Count Fulk Rechin at La Flèche in the south of Maine that precede it in his narrative, a chronology based on annals written in Anjou that date Fulk’s two attacks on La Flèche to late 1076 or early 1077 and to 1081 shows that this cannot possibly have been the case.14 It is therefore as good as certain that the record of the council is referring to the threat of war rather than to an actual conflict. William might have feared that Fulk Rechin would intervene in Maine. Also of concern would have been developments to the east and south of Normandy, where William’s foster-son Simon was attacked by King Philip I after the death in late 1074 of Simon’s father and William’s ally, Count Ralph of Amiens, Valois, and the Vexin. It would certainly have made sense for Fulk to try to dislodge William as soon as possible after his reoccupation of the county in 1073 or, at the least, to seek to re-establish the relationship involving homage set up in 1063. But there is nothing to suggest that he attempted to do this until late 1076 at the earliest. The one indication that there were disturbances in Maine at this 11 Regesta, no. 59. For an analysis of this narrative, Elisabeth van Houts, ‘Les femmes dans l’histoire du duché de Normandie’, Tabularia. Etudes, ii (2002), 19–34, at 21–3. 12 For the argument that the now lost plaque above the tomb of Adelia filia regis, known only from a seventeenth-century drawing, is Adelida’s, Letouzey-Réty, Ecrit et gestion, 123–8. 13 . . . dilatum est usque ad regis audientiam, qui in transmarinis partibus tunc temporis bella gerebat, Lanfranc Letters, no. 11; CS, 613. For the likely date, CS, 607–8. The broad dating limits are 25 December 1074x28 August 1075. 14 OV, ii, 308–9 (Uerum eodem tempore alia tempestas grauissima orta est), with Marjorie Chibnall describing Orderic’s chronology as ‘very confused’ (p. 310, note 2). The accepted chronology of William’s campaigns in Maine was established by Halphen, Le comté d’Anjou, 182–4. See further Douglas, WC, 401–7; Guillot, Le comte d’Anjou, i, 119–21.

cross-channel rule

377

time is a letter written in 1079 by Archbishop Gebuin of Lyons, acting on the basis of a primacy over other archbishoprics such as Tours and Rouen that had recently been created by Gregory VII, which says that Abbot Rainald of Saint-Pierre de La Couture in Le Mans had suffered persecution at William’s hands for five years. Although we do not know why this was done, steps must have been taken to remove him from the effective conduct of affairs during either 1074 or 1075, with the objective being to reinforce a regime that depended heavily on the support of churchmen.15 Even so, William’s general approach to the affairs of Maine seems to have remained a cautious one, since the sidelining of Abbot Rainald did not lead to his immediate deposition; his eventual replacement, the abbey’s prior Juhel, appears as holding the office of prior in a charter dating from 1076 that resolved a dispute with the abbey of Marmoutier.16 The matter seemingly only blew up in 1078 when the metropolitan, Archbishop Ralph of Tours, wrote to Bishop Arnold of Le Mans, who had by this time consecrated Juhel as abbot, advising that it would be preferable to prevent Rainald appealing to the pope.17 With Count Ralph’s territories having been expediently assembled through a mixture of inheritance and force, Simon was legitimately open to be attacked, an opportunity that Philip and others seem to have taken up almost immediately. A chronology based on Simon’s early twelfth-century Vita, various charters, and the testimony of several writers fascinated by his decision to become a monk – something that had happened by 23 May 1077 – indicates that Simon was able to fight back with considerable success.18 Although there is nothing to indicate that William intervened directly to support him, he would certainly have been concerned about the shift in the balance of power around Normandy’s eastern frontiers that these changes threatened to bring about. It is understandable that he decided to maintain a watching brief. Whatever the case, William’s reasons for staying in France must have been serious ones, with a rebellion developing in England during the spring and summer of 1075. The letters exchanged between William in Normandy and Lanfranc in England, in which Lanfranc wrote confidently of military success and 15 RHF, xiv, 668–9 (a rege Anglorum quinquennio cruciatum). For a suggested dating of 19 April 1074x24 April 1075, Latouche, Maine, 85, notes 4 and 5. In general, H. E. J. Cowdrey, ‘Pope Gregory VII and the Anglo-Norman Church and Kingdom’, Studi Gregoriani, ix (1972), 77–114, at 99–100. 16 Cartulaire manceau de Marmoutier, ed. Ernest Laurain, 2 vols. (Laval, 1911–45), ii, 73 (Sablé, no. 5). 17 For the letter, RHF, xiv, 667–8; Cowdrey, Gregory VII, 391. 18 For the chronology of events, Augustin Fliche, Le règne de Philippe Ier, roi de France (1060– 1108) (Paris, 1912), 147–9. See also Feuchère, ‘Une tentative manquée’, 13–15.

378

william the conqueror

advised William that there was no need for him to cross the Channel, show that a strategic decision had been taken that the revolt in England constituted less of a threat than the developments in northern France. Also, for all that William did not cross the Channel, several of his closest lieutenants did, among them Bishops Odo of Bayeux and Geoffrey of Coutances, both of whom had been in Normandy in 1074. With William de Warenne and Richard fitz Gilbert seemingly having remained in England in 1074 and certainly participating in the suppression of the revolt, we can see very clearly how the cross-Channel regime functioned in a crisis and how the responsibilities of governance and loyalty operated. ENGLAND, 1075–6: THE REVOLT OF THE THREE EARLS AND OTHER BUSINESS The revolt’s ringleaders were Ralph, earl of East Anglia since his father’s death in 1068, and Roger, son of William fitz Osbern and earl of Hereford following his father’s death in 1071.19 Also involved in ways that were not entirely clear either to contemporaries or to subsequent generations, and which the manner of his death and the subsequent cult of sanctity have obscured further, was Waltheof, earl of Northumbria, the holder of extensive lands in the east Midlands and husband of William’s niece Judith. Although the different sources supply differing details about the development of events, there is little that is fundamentally incompatible between them. The plot was hatched at the wedding-feast of Earl Ralph and Roger’s sister Emma, which, according to the two versions of the Anglo-Saxon Chronicle, took place at Norwich, and according to John of Worcester, at Exning (Suffolk). Since the choice of Exning, although an obscure place, would have as good as guaranteed the attendance of Waltheof as Cambridgeshire’s earl, something that the two others may have been trying to achieve, it is possible that we are dealing with two events.20 According to the two almost identical versions of the Chronicle, the marriage had been sanctioned by William, but John, although reliant on a version of the Chronicle, directly contradicted both of them and said that the king had not agreed. Since an alliance between one of the greatest of the new Norman families in England and the descendant of an Anglo-Breton family 19 All of the main sources identify these two as the chief rebels. For the main accounts, ASC, ‘D’ and ‘E’, 1075; OV, ii, 310–23; GR, i, 468–71, 472–3; GP, i, 486–9; JW, iii, 24–9. For an assessment similar to mine, Ann Williams, ‘Ralph [called Ralph de Gael, Ralph Guader], Earl (d.1097x9)’, ODNB. 20 For this suggestion, Lucy Marten, ‘The Rebellion of 1075 and its Impact in East Anglia’, in Christopher Harper-Bill (ed.), Medieval East Anglia (Woodbridge, 2005), 168–82, at 174.

cross-channel rule

379

favoured by Edward the Confessor and raised still further by William was the sort of liaison which he is likely to have wanted to encourage, the truth may well be that William’s agreement might not have been as clear-cut as the Chronicle implied, and that later versions of events transformed this into a not entirely accurate view that he had prohibited it. Malmesbury, as often more down to earth than others, paints a picture of an increasingly drunken banquet, at which talk got out of control, and out of which emerged what he thought to be a pretty hare-brained conspiracy. The plot hatched, Ralph and Roger departed to their earldoms at opposite ends of the kingdom and started to raise troops. Emissaries were sent to Denmark to request assistance, with in due course – but, as it turned out, too late – a fleet of 200 ships setting out across the North Sea, one of whose commanders was Svein Estrithsson’s son Cnut. Roger presumably sought support from his tenants, most of whom would have been introduced into Herefordshire and others counties by his father, and from South Wales where he seems to have further expanded the area of influence established by William fitz Osbern, and also to have established collaborative relationships with Welsh rulers. A sense of both self-importance and achievement is indicated by his appearance with the title ‘earl of Hereford and lord of Gwent’ in a charter for Llandaff Cathedral.21 This would have fuelled the discontents that he felt and which Lanfranc tried to address. While the fact that Lanfranc took it upon himself to negotiate with some of the chief rebels is a certain sign of personal trust between him and William, as we shall see, his powers of decision had their limits. During the period when the rebels were organizing themselves, Lanfranc wrote to Roger as ‘his dearest son and friend’, implying thereby that he was giving spiritual guidance, to remind him that his responsibility was to defend castles against the king’s enemies and to remember his father’s distinguished career. As far as Roger’s complaint that the king’s sheriffs were holding pleas in his lands was concerned, he said William had ordered that they should desist until he returned to England to hear the dispute.22 A second letter, albeit with the same cordial address, was more admonitory, advising Roger either to come to his senses or to set out to clear himself of the accusations against him. It promised a safe-conduct.23 A third letter, whose composition was apparently interrupted by a conciliatory message from the earl, refers to Roger as Lanfranc’s former dearest son and friend, censures his obduracy, and excommunicates him until he seeks the king’s forgiveness. Its 21 The Text of the Book of Llan Dâv, ed. J. Gwenogvryn Evans (Oxford, 1893), 274 (et uerbo comitis Herefordie et domini Guenti Rogerii filii Willelmi filii Osberni); Crouch, ‘The Transformation of Medieval Gwent’, 4. 22 Lanfranc Letters, no. 31. 23 Ibid., no. 32.

380

william the conqueror

conclusion, while mentioning that Roger had asked to see Lanfranc, advises that William is now so angry that he should remain where he is and wait on events. Lanfranc offers all help to him compatible with his allegiance to the king.24 Nothing is known of any negotiations that might have taken place with Earl Ralph, whom Lanfranc describes as a perjured robber and a traitor and his followers as ‘Breton dung’ in a letter to William.25 It looks as if a distinction was being made between the two rebels, with Ralph seen as something like the evil genius leading others astray and Roger as a member of the inner circle whose behaviour was initially regarded as a source for regret. While all sources agree that Waltheof was present when the revolt was plotted, even at this point in the proceedings they assess his conduct differently. William of Malmesbury in the Gesta Regum wrote that the Normans and English differed in their views of his complicity: for the former, he was a traitor, while for the latter he was a reluctant conspirator dragged in by circumstances, adding that his English informants were individuals of impeccable integrity.26 In the later Gesta Pontificum, although continuing to write that he was still not entirely convinced, Malmesbury announced that he leant towards the English version and said that he believed Waltheof to have been coerced into joining the revolt and that he confessed as such to Lanfranc.27 John of Worcester, on the other hand, explicitly states that Lanfranc believed Waltheof to have been innocent. The accounts of the feast also differ slightly, with Waltheof either a reluctant plotter from the start in some, or forced by an oath sworn to the conspirators not to reveal what he had heard, or involved at the beginning, but soon repenting of his actions. No source suggests that Waltheof took any military role in the revolt. In due course, in all probability, advised like Earl Roger by Lanfranc, he sought out William, crossing to Normandy to do so, and threw himself on the king’s mercy. He was kept in custody awaiting judgment. Of the near-contemporary sources, only Lanfranc’s letters suggest a cause for the revolt, namely Earl Roger’s complaints about uncustomary interference in his lands by royal sheriffs, indicating that he felt he was not being accorded the same dignity as his father. It is likely that Ralph also felt downgraded in comparison with his father.28 Although his case is somewhat different because of his eventual fate, it looks as if Waltheof was also pushed to the margins of power. For all that he was superficially favoured by William, it appears his actual treatment was similar to that of Edwin Ibid., no. 33. Ibid., nos. 34, 35. 26 GR, i, 468–71, 472–3. 27 GP, i, 486–9. 28 For this suggestion, Williams, ‘Ralph [called Ralph de Gael, Ralph Guader]’. 24 25

cross-channel rule

381

and Morcar in the 1060s; he attests no known charter in the 1070s, for example, and when Edgar the Ætheling handed himself in in 1074, it was the sheriff of York, rather than Waltheof, who escorted him to William’s court in Normandy.29 Orderic, as he often did, included a lengthy speech in which he sets out what he believed were the general complaints about William’s rule. The main charges – maltreatment of his enemies, oppression of the English, and failure to reward his followers adequately – read like a repetition of his general critique of William’s rule rather than as a specific set of grievances for 1075, and should be treated as such. They in fact describe the social and psychological stresses that had been generated from the moment the decision to invade England had been taken and which had swirled around ever since. The readiness of the Danes to become involved and Waltheof ’s willingness to flirt with revolt after his earlier defeats at William’s hands both nonetheless suggest that those who took part, or who thought about taking part, saw the events as a continuation of the campaigns of the late 1060s and early 1070s, with, however ineffectually, Anglo-Danish resentment this time fusing with an individual Norman’s sense of injustice. Both Orderic and Malmesbury elaborate on statements that appear in both versions of the Chronicle and say that the rebels were aiming to take the kingdom from William. Malmesbury suggested that Earl Ralph thought his kinship to the Norman ducal line created by his marriage to a daughter of William fitz Osbern made him a suitable candidate, and Orderic, that the plan was to divide the kingdom into three parts and make one of them king. It therefore looks as if they were creating a rationale for personal grievances out of circumstances – kinship as a situational construct again – to try to create a legitimacy that might persuade others to join them. It must also be relevant that the two leaders, and indeed Waltheof, were all young men who felt themselves less valued than their fathers and predecessors had been. The revolt can therefore arguably be seen as a first manifestation of the sort of cross-generational conflict that becomes something of a feature of the last decade of William’s life. An army commanded by Bishop Wulfstan of Worcester, Abbot Æthelwig of Evesham, Walter de Lacy, and Urse d’Abetôt, sheriff of Worcester, kept such forces as Roger assembled hemmed in west of the Severn. It looks as if he had not been much supported by his Herefordshire tenants; Walter

29 Andrew Wareham, ‘Two Models of Marriage: Kinship and Social Order in England and Normandy’, in Arnoud-Jan A. Bijsterveld, Henk Teunis, and Andrew Wareham (eds.), Negotiating Secular and Ecclesiastical Power: Western Europe in the Central Middle Ages (Turnhout, 1999), 107–32, at 128–9.

382

william the conqueror

de Lacy was of course a major landholder in the shire.30 Meanwhile, Ralph’s attempt to march west to join Roger was halted when he was defeated in a battle at Fawdon in Whaddon, to the south-west of Cambridge.31 The sources differ on the names of William’s commanders, with Bishop Odo of Bayeux, Bishop Geoffrey of Coutances, Richard fitz Gilbert, and William de Warenne all being named. Ralph fled, firstly to his castle at Norwich, then by sea to Denmark, and thence to Brittany where he had inherited lands, leaving his wife to hold Norwich against a besieging force led by Bishop Geoffrey, William de Warenne, and Robert Malet. In due course, after a siege that may have lasted for as much as three months, she surrendered on terms that allowed her to rejoin her husband. Earl Roger in contrast accepted the summons to the king’s court and stood trial. Lanfranc’s second letter to William makes it clear that all this happened while the king was still in Normandy.32 It looks as if the revolt had been quite easily defeated. William’s eventual Channel crossing is dated to the autumn by John of Worcester, with the imminence of invasion from Denmark presumably being the cause. Another of Lanfranc’s letters, this time to Bishop Walcher of Durham, advises him that, although the rebels have been defeated, it was certain that the Danes were coming and that he should put his castle and men on a war footing. He informs Walcher that he has been in touch with William about this.33 When the Danes did arrive, which was after the castle at Norwich had surrendered, they sailed north, ravaging the Lincolnshire coastline, and devastated York, breaking into the cathedral, before leaving for Flanders. As in 1069–70 and 1071, their leaders must have quickly grasped that the rebellion in England they had come to support had no chance whatsoever of defeating William and that their own force, which had been transported in 200 ships, although large, was inadequate for a sustained campaign. The sack of York once more demonstrated how easy it was for a hostile army to enter the city along the Humber estuary and the river Ouse, and how difficult it was for William to defend the northern parts of his kingdom effectively. There is again a sense here of a replay of former times and, therefore, from both his and the Danes’ perspective, of unfinished business.

30 Christopher Lewis, ‘The Norman Settlement of Herefordshire under William I’, ANS, vii (1985), 195–213, at 205–8, 212. 31 OV, ii, 316–17; JW, iii, 24–5. For the site, Bruce Dickins, ‘Fagaduna in Orderic (A.D. 1075)’, in Folke Sandgren (ed.), ‘Otium et negotium’: Studies in Onomatology and Library Science Presented to Olof von Feilitzen (Stockholm, 1973), 44–5. 32 Lanfranc Letters, no. 35. 33 Ibid., no. 36.

cross-channel rule

383

Once he was back in England, William treated the rebels with a severity that is predictable in the context of his past record. That it was arguably more severe than anything before could be a reflection of special circumstances associated either with the sense of security that came with the completion of conquest or because of the particular personnel who were involved. Earl Roger was condemned to perpetual imprisonment despite seeking the king’s mercy, with Orderic writing that he remained angry about his treatment by William for the rest of his life, for example setting fire to a set of valuable clothes that the king sent him one Easter. This was surely the behaviour of a man who believed himself personally affronted by punishment beyond what he thought he deserved; the exchange of messages with Lanfranc had, after all, given him cause to hope for mercy and the friendship of William and his father might have made him think that William would treat him indulgently. Roger was seemingly never released and, according to Orderic, his sons were still unpardoned in Henry I’s reign.34 While the garrison that had defended Norwich left for Brittany under safe conduct, the prisoners taken at Fawdon in Whaddon were mutilated, with, according to Orderic, each of them having their right foot chopped off.35 At William’s Christmas court, held at Westminster, others were exiled, blinded, or had their hands removed.36 Once more we see how ruthlessly William could treat defeated rebels. For all that decisions about the punishment of rebels were a major subject at the Christmas court, Queen Edith’s funeral and burial must surely have dominated the proceedings; William of Malmesbury wrote that William treated her funeral with the same reverence that he was later to display at his wife Matilda’s funeral.37 Edith had died at Winchester on 19 December. According to both versions of the Chronicle, she was buried on William’s orders ‘with great honour’ in Westminster Abbey next to her husband.38 Although Edith’s own wishes are ultimately unknown and unknowable, her burial next to Edward may have been calculated to cut her off from any links to her brothers’ surviving kinsmen and allies and to associate her solely with the framework of ideas which was claimed to legitimize William’s rule in England. Equally, she may have decided herself that she wished to be remembered as England’s queen and not as one of Earl Godwine’s children. The burial together of a king and a queen was unusual in England and was not – of course – what William and Matilda were OV, ii, 318–19. Ibid., 316–17. 36 ASC, ‘D’, ‘E’, 1075; JW, iii, 26–7. 37 GR, i, 502–3. 38 ASC, ‘D’, ‘E’, 1075. 34 35

384

william the conqueror

themselves planning.39 With Edith’s final illness having, so tradition records, been a long one, it would have been possible for William to have visited her at Winchester if he had crossed the Channel to Portsmouth or Southampton during the autumn of 1075. But no source mentions such a visit. The surviving record of the ecclesiastical council that met at Winchester after Easter 1076 on 1 April does not actually mention the names of anyone who was present.40 The fact that it was held at the by now traditional site of the Easter crown-wearing and that it was William’s normal practice, when it was possible for him, to be present on such occasions makes his involvement highly probable. The council’s legislation followed up on the business of earlier councils and, on some important issues, brought England into line with earlier Norman legislation. The deposition of Bishop Æthelric of Selsey, about which Alexander II had raised queries in a letter to William in 1071, was finally confirmed.41 Clear guidance was given on the topic of clerical celibacy. No canon was to have a wife, but those who resided in castles and in other communities were not to be forced to give them up. Bishops were not in future to ordain priests or deacons who had wives. This gradualist approach to changes being pushed by the papacy mirrored that taken in Normandy. Lessons may also have been learnt from the physical assault on Archbishop John of Rouen after the 1072 Council of Rouen, with the still unsettled conditions in England being another factor that was borne in mind. In another piece of the council’s legislation, it was decreed that laymen who refused to obey the bishops’ jurisdiction were, if they ignored three summonses to an episcopal court, to be excommunicated. The parallels between this text and the famous royal writ which, in addition to requiring that ecclesiastical cases should not be heard in the hundred court, also made royal officials available to coerce the recalcitrant laity who refused the summons to the bishops’ courts, make it likely that both are part of a drive that was being undertaken to clarify and reinforce the jurisdictional role of the bishops. This, again a feature of William’s rule in Normandy from before 1066 and of his early reliance on bishops in England, fits well with the general emphasis on authority and discipline.42 The development of a more precise approach to authority and jurisdiction is a feature of a changing approach to rule in England that becomes more evident in 1077. However, an English background to the instruction in the writ suggests that its objective was to clarify rather than initiate, the issue being the On Edith’s final illness and burial, Stafford, Queen Emma and Queen Edith, 94–6. For the text and a commentary, CS, 616–20; Cowdrey, Lanfranc, 127–8. 41 Lanfranc Letters, no. 7. 42 For the latest editions of the writ, CS, 620–4; Regesta, no. 128, with comments on possible dating. 39 40

cross-channel rule

385

location of pleas rather than the absence of a specific ecclesiastical jurisdiction in England before 1066.43 After the death at a date between 1072 and 1075 of Bishop Siward of Rochester, two appointments to the bishopric had to be made in rapid succession, with Arnost, who died in July 1076 approximately six months after taking office, being followed by Gundulf, who was sent by Lanfranc to Normandy for William’s approval and then consecrated on 19 March 1077. Both had been monks of Le Bec and subsequently of Saint-Etienne of Caen, a very clear sign of Lanfranc’s preferences. Gundulf ’s episcopate was to be a distinguished one; he was also responsible as an architect for the building of the Tower of London.44 The first abbot of Battle was also appointed in 1076, with the time taken to advance to this point perhaps suggesting that progress on William’s foundation on the battlefield of Hastings had been slow. But, quite apart from the impact of the disaster in which the first nominee as abbot was drowned at sea, the history of the two Caen abbeys indicates that a measured approach which brought in an abbot and monks only when substantial progress had been made on the buildings and the endowment was going to be taken. The monastery is also likely to have developed in relation to William’s confidence in the long-term security of the Conquest. The writers of the so-called Battle Chronicle would later say that William did not visit the site as often as he would have wished, and that the endowment would have been richer had he lived longer.45 Their complaints could well have some basis in that archaeology has shown that the church was significantly smaller than many buildings of comparable prestige.46 In addition, Domesday statistics show that its wealth did not come anywhere near rivalling that of England’s great pre-Conquest abbeys.47 Another significant point is the monastery’s failure to attract extensive patronage from anyone other than local lords, something that is in sharp contrast to the two Caen abbeys. The monastery was incomplete at William’s death and was not dedicated until 1094. All this said, the chroniclers’ complaints may well reflect more the pretensions of the twelfth-century monks than any lack of interest on William’s part. Albeit known only through a set of forged charters of the second half of the twelfth century, privileges such as geld exemption were presumably already in place and, by analogy with the 43 Barlow, The English Church 1000–1066, 146–53; Colin Morris, ‘William I and the Church Courts’, EHR, lxxxii (1967), 449–63, at 453–5; Hudson, The Oxford History of the Laws of England, 63–4. 44 Martin Brett, ‘Gundulf (1023/4–1108), Bishop of Rochester’, ODNB. 45 The Chronicle of Battle Abbey, ed. and trans. Eleanor Searle, OMT (Oxford, 1980), 44–9. 46 See the comments in Fernie, The Architecture of Norman England, 104. 47 Dom. David Knowles, The Monastic Order in England, 2nd edn. (Cambridge, 1966), 702–3.

386

william the conqueror

Caen abbeys, early exemption from episcopal jurisdiction is very probable.48 Integration into William’s English itinerary is indicated by the grant of a lodging for the monks at Windsor.49 Sussex, with so much land granted out to the lords of the rapes, may not have been an ideal place for an extensive endowment. Battle’s location was also much more rural than the Caen abbeys’, which unquestionably prohibited the financial investments that contributed to the prosperity of those two churches. It is notable that Saint-Etienne and Battle both received extravagant gifts from the king on his deathbed; it was, however, Saint-Etienne that was given the king’s regalia. It is possible that the story that William made a vow on the battlefield of Hastings to found a great abbey was probably a twelfth-century elaboration of a simpler initial act of atonement. It certainly looks as though the religious patronage he hoped would save his soul focused primarily on Caen and on French monasteries. Waltheof ’s execution on St Giles Hill outside Winchester on 31 May 1076 was in due course to become a very controversial event. After his beheading and his initial burial at Winchester, his body was taken to the abbey of Crowland for interment, where, from the late eleventh century onwards, he began to be venerated as a saint and, when his tomb was later moved into the abbey church, miracles allegedly started to happen. Stories therefore circulated that he had spent his last weeks in prayer and penance and that his decapitated head continued to say the Lord’s Prayer after the axe had severed it from his neck. As already noted, both Malmesbury and Orderic found it hard to assess Waltheof ’s level of guilt and, according to John of Worcester, Lanfranc believed him innocent of the charges brought against him.50 The situation was one in which later events profoundly influenced what was written from the early twelfth century onwards. Orderic presents the process that led up to the execution as a time when Waltheof ’s enemies, coveting his lands, plotted against him, with William finally giving in to their arguments.51 It is, however, hard to believe that William capitulated to pressure in this way and went against his own inclinations, since every other situation of this kind that we know about suggests that he would have been thinking what to do and would then have imposed his personal decision. It is also a standard theme with Orderic to say that 48 Regesta, nos. 13, 15, 19, 20, 22, 23; DB, i, 17v. For a judicious assessment of the sources written at Battle, see Brevis Relatio, 17–19. 49 Regesta Regum Anglo-Normmanorum vol. ii, Regesta Regis Henrici Primi, 1100–1135, ed. Charles Johnson and H. A. Cronne (Oxford, 1956), ii, no. 1807. 50 JW, iii, 28–9. For the development of Waltheof ’s cult, Carl Watkins, ‘The Cult of Earl Waltheof at Crowland’, Hagiographica, iii (1996), 95–111; Joanna Huntington, ‘The Taming of the Laity: Writing Waltheof and Rebellion in the Twelfth Century’, ANS, xxxii (2010), 79–95. 51 JW, ii, 320–3.

cross-channel rule

387

William’s followers’ greed was always pushing him into attitudes that were against the interests of the English. Ultimate responsibility for what happened must, however, surely rest with William. With the death penalty being permitted within the customary law in both Normandy and England when a crime was judged sufficiently severe, the case for capital punishment would have been that favour had been lavished on a man who had persistently associated himself with revolts and had used his position to prosecute feuds in northern England.52 While arguably excessive in relation to Waltheof ’s faults in 1075–6, the sentence can be viewed as logical in relation to the history of the English kingdom since 1066. It is therefore another of those situations in which very different judgements on William’s conduct are possible; as we have seen many times, his record of forgiving persistent opposition was one where initial displays of mercy were followed by exclusion or greater severity. Orderic’s narrative is therefore probably a misrepresentation of extensive consultations on William’s part before doing something as serious as executing one of the leading aristocrats of the English kingdom. It is also surely relevant that when the story of the saintly Waltheof started to be written up from the late eleventh century onwards, the emphasis was on Waltheof having died as a saint. There is no trace in it of his being thought of as an English patriot.53 The defeat of the 1075 revolt brought about significant changes in the organization of local rule and landholding. The responsibilities of the earl in the far north were transferred to Bishop Walcher of Durham, an apparently extraordinary appointment that is probably explicable as recourse to someone who could be deemed to have acquired knowledge of the north. Having apparently begun well as a bishop, he may have seemed to be a promising prospect for a position in which four men had already failed since 1066.54 Neither Roger nor Ralph was replaced with a new earl, the suppression of the two earldoms and the reliance instead in the localities on officials of lesser status being conceptually a break with English tradition. In Herefordshire, it would appear that so few of the men introduced by William fitz Osbern had supported his son that adjustments were relatively minor; some new landholders were introduced, but, mostly, existing ones were upgraded into tenants-in-chief.55 King William’s second surviving son, William, was entrusted with a punitive raid against Roger’s main 52 John Hudson, ‘The Fate of Earl Waltheof and the Idea of Personal Law in England after 1066’, in Crouch and Thompson (eds.), Normandy and its Neighbours, 223–35; idem, The Oxford History of the Laws of England, 399–401. 53 Watkins, ‘The Cult of Earl Waltheof ’, 103, 111; Huntington, ‘The Taming of the Laity’, 88–91. 54 Aird, St Cuthbert and the Normans, 90–1. 55 Lewis, ‘Norman Settlement’, 208–10.

388

william the conqueror

supporter in Wales, King Caradog ap Gruffudd, with the intention being presumably to give him a warning not to get involved in rebellions in England again. Glamorgan was devastated and burnt and, so we are told, deprived of all of its wealth.56 The future King William II, who cannot yet have been out of his teens, was presumably being given his first experience of military command in relatively undemanding circumstances. The changes in East Anglia primarily involved advancing those on whose loyalty William knew he could rely. Thus, Richard fitz Gilbert and Count Alan benefited hugely, the former effectively obtaining a new English base to add to his lands around Tonbridge that became what was later referred to as the Honour of Clare, and the latter adding forfeited lands to the lands of Edith (Eadgifu) the Fair, Harold’s partner, whose daughter he had married and whose East Anglian lands he had taken over. At the same time several individuals of diverse ethnic origins were awarded forfeited lands with responsibilities for oiling the wheels of local governance. This suggests that, whatever ethnic elements may have influenced the execution of the last earl of English birth, by this time opposition and disloyalty were withering away and competence was becoming the rationale for advancement.57 1075 TO 1077: NORMANDY AND NORTHERN FRANCE While William was in England, at a date that may have been in the last months of 1075, or, more probably, was between Christmas 1075 and 25 March 1076, Count Simon of Amiens, Valois, and the Vexin returned Gisors, on the border of Normandy and his county of the Vexin, to the cathedral church of Rouen in return for a payment of £300 of Rouen money. Having initially tried to retain the life-interest that his father had been granted as part of the process by which William constructed alliances around Normandy’s frontiers before 1066, he seems to have changed his mind. The ceremony was performed before the high altar of the cathedral on which Simon laid a knife in the presence of Matilda, Archbishop John of Rouen, Roger de Beaumont and others.58 The splendour of the occasion might indicate a political rapprochement between Simon and William or a gesture of thanks from Simon for assistance rendered. Alternatively it could just be a consequence of the close personal relations between Simon, Matilda, and William, or an aspect of Simon’s deepening piety. This final 56 Vita sancti Gundlei, in Vitae sanctorum Britanniae et Genealogiae, ed. A. W. Wade-Evans (Cardiff, 1944), 172–93, at 190–1. 57 On these changes, see especially Marten, ‘The Rebellion of 1075’, 179–82; KeatsRohan, ‘A Question of Identity’, 179–81. For Count Alan and Eadgifu’s lands and daughter, Richard Sharpe, ‘King Harold’s Daughter’, HSJ, xix (2008, for 2007), 1–27, at 7–22. 58 Regesta, no. 229.

cross-channel rule

389

suggestion is arguably the most convincing. Simon’s personal closeness to William and Matilda suggests that no rapprochement was needed. The large sum of money could have been to help Simon visit Pope Gregory VII for spiritual advice about the reburial of his father’s body, a subject that vexed Simon deeply, because his father had died excommunicate, or to help him further resist King Philip.59 The presence on the original charter in which this narrative is recorded of the autograph crosses of William’s son Robert and of Roger de Beaumont underlines again Robert’s status in Normandy in his father’s absence, as well as how important Roger de Beaumont continued to be (Plate 19).60 However, since Robert is not mentioned in the narrative, his confirmation may well be an expression of his symbolic standing; it is as if Matilda and Roger had done all of the practical business and that he was needed only to sign it off. A similar philosophy may well be present behind the occasional references to Robert as ‘count of Maine’ in charters dating from after 1073, with one charter being jointly confirmed by William and Robert and one specifically referring to him as exercising authority in Maine in August 1076, again presumably reflecting arrangements set up while William was in England.61 Taken alongside Lanfranc’s co-ordinating role in the defeat of the rebellion in England and the future King William Rufus’s first known campaign, this evidence again underlines the importance of a consortium of family and close associates to the maintenance of authority across the great distances from south of Le Mans to the Scottish borders. Lanfranc’s constant references to William’s ultimate power of decision during 1075 do, however, indicate that it was the king’s power which was ubiquitous. Perhaps it was only Matilda, Lanfranc, and Roger de Beaumont who were truly trusted. After returning to Normandy in the summer of 1076, William took an army into Brittany to join in the siege of the castle of Dol. Treated by David Douglas as an extension of the Revolt of the Three Earls because Earl Ralph is named among those who defended Dol against William by John of Worcester, it may not have been that at all.62 In the first place, John’s statement that Dol was Ralph’s castle is a mistake; it was the castle 59 For Simon’s spiritual development, Cowdrey, ‘Count Simon of Crépy’s Monastic Conversion’, 256–9. 60 For this theme, David Bates, ‘The Origins of the Justiciarship’, ANS, iv (1982), 1–12, 167–71, at 6–8. 61 Regesta, nos. 27, 172; Cartulaire de l’abbaye de Saint-Vincent du Mans, no. 589 (. . . Robertoque, Willelmi regis Anglorum filio, Cenomannicum urbem gubernante). See further Aird, Robert Curthose, 68–9. 62 JW, iii, 28–9; cf. Douglas, WC, 234, 402 (‘the campaign followed indeed logically upon the revolt of Earl Ralph in England’). The main Anglo-Norman sources for the siege of Dol are ASC, ‘D’, ‘E’, 1076; OV, ii, 350–3; JW, iii, 28–9.

390

william the conqueror

of John, son of Rivallon de Dol who had been so prominent in William and Harold’s expedition into Brittany in 1064. And, in the second, William’s intervention at Dol in 1076 is treated by all the Breton and Angevin sources that describe it either as a war between Count Hoel of Brittany and Geoffrey Grennonat, an illegitimate son of Count Alan III installed at Rennes by Hoel after the death of Geoffrey’s half-brother Conan II in 1066, or as a setback for William caused by the strength of the Angevin contribution to its defence.63 It is therefore much more likely that Ralph inserted himself into a situation already existing within Brittany, if indeed he was involved in events at all. In other words, John of Worcester’s version of events may be an intelligent deduction made several decades later that is either completely wrong or substantially so. As in 1064, William’s interest in Dol in 1076 is likely to have derived from concern that a strongpoint close to Normandy’s western frontier was going to pass into the hands of an enemy. Geoffrey Grennonat’s interest in Dol might have installed a potentially hostile influence there. Another aspect must have been the attempt to replace William’s long-standing ally, Juhel, bishop of Dol, who had been appointed by Count Alan III as long ago as 1039. His removal was clearly sought by several interested parties in Brittany, and it was sanctioned by Gregory VII on 27 September 1076 in letters to the clergy and people of Dol and the bishops of Brittany after he had received an embassy from Brittany that had travelled to Rome to seek Juhel’s replacement by a young man named Gilduin. Gregory had overruled this proposal in favour of the appointment of Evenus, abbot of Saint-Melaine of Rennes, a former monk of the abbey of Saint-Florent of Saumur.64 Since Gilduin was the third son and Juhel the brother of the late Rivallon de Dol, the mission looks like an attempt to sustain the family’s traditional domination of the bishopric, a motive to which Gregory was not entirely insensitive since the abbot of Evenus’s first monastery of SaintFlorent was another of Rivallon’s sons. There does also look to have been a reform agenda in this, since Juhel was both a simoniac and cohabiting with a partner. The only snag with this neat plan was that Juhel declined to play along with it, installing himself at Mont Saint-Michel and presumably calling on William to support him.65 Gregory VII showed awareness of 63 Thus, in excerpt form, Chronicon Britannicum, RHF, xi, 412–13, at 413; Chronicon Briocensi, in RHF, xii, 565–7, at 566. For Anjou, Annales qui dicuntur Rainaldi archidiaconi sancti Mauricii Andegavensis, in Recueil d’annales angevines et vendômoises, ed. Louis Halphen (Paris, 1903), 88 (defendentibus illud fortibus Andecauorum militibus); Annales sancti Sergii Andegavensis, in idem, 91–9, at 92. 64 Registrum, i, nos. IV, 4 and 5 (Register, 213–15). 65 The various relationships involved are set out in Hubert Guillotel, ‘Bretagne et papauté au XIe siècle’, in Rolf Grosse (ed.), L’Eglise de France et la papauté (Xe–XIIIe siècle). Actes du XXVIe colloque historique franco-allemand organisé en coopération avec l’Ecole nationale des chartes par l’Institut

cross-channel rule

391

William’s interests by sending him a letter that outlined Juhel’s faults, but it may not have arrived before William had launched his campaign.66 Gregory’s decision to make Dol an archbishopric for Brittany may also have been intended to flatter all involved in the scheme to replace Juhel.67 The potential in all this for William’s discomfiture must also explain the introduction into Dol of troops and supplies from Anjou, which was also probably another attempt to exploit William’s preoccupations on the other side of the Channel at a time when he seemed to be tied down.68 In September, William invested the castle with some support from Hoel, duke of Brittany (1066–84), who held the southern and western Breton counties of Cornouaille, Nantes, and Vannes. At this point, Philip I was also prompted to intervene. On 14 October he was in the company of Duke William VIII of Aquitaine at Poitiers, assembling troops for an attack, having travelled to Poitou in such haste that he left his seal behind.69 The point of concern for both him and Count Fulk now that William had taken the initiative was presumably that, if William took Dol, a large slice of north-eastern Brittany would fall under his control. In addition, his relations with the sons of Count Eudo and their cross-Channel interests would greatly extend his influence along much of Brittany’s northern coast. Following his usual practice, William tried to induce the garrison to surrender by ravaging the surrounding countryside. However, he was outmanoeuvred probably because the situation of his army was an exposed one since he could not command the roads leading south from Dol towards Anjou and Aquitaine and because he did not have many effective allies in Brittany, with his obvious supporters, the sons of Count Eudo, playing no known part. In November, King Philip surprised William’s besieging army and forced him to abandon the siege with great losses of men, horses, and treasure, valued, according to Orderic, at around £15,000 in English money.70 With none of the surviving accounts of the debacle being particularly detailed, an explanation for the defeat has to be surmised from other similar situations. At Domfront in 1051–2, and, in the analogous, if somewhat historique allemande de Paris (Bonn, 1993), 265–86, at 276–84. On the Breton context, see most recently Katharine S. B. Keats-Rohan, ‘Raoul Anglicus et Raoul de Gaël: Un réexamen des données anglaises et bretonnes’, MSHAB, xciv (2016), 61–91, and see especially, 77. 66 The Epistolae Vagantes of Pope Gregory VII, ed. and trans. H. E. J. Cowdrey, OMT (Oxford, 1972), no. 16. 67 For the affair of the archbishopric of Dol, see also Cowdrey, Gregory VII, 460–2; Paula de Fougerolles, ‘Pope Gregory VII, the Archbishopric of Dol and the Normans’, ANS, xxi (1999), 47–66. 68 For the presence of Angevin troops, Annales qui dicuntur Rainaldi and Annales sancti Sergii Andegavensis, in Recueil d’annales angevine et vendômoises, 88, 92. 69 Recueil des actes de Philippe Ier, no. 84. 70 OV, ii, 352–3.

392

william the conqueror

different, circumstances at Hastings in 1066, scouts must have given him warning of the advance of a hostile enemy and he had moved his army to confront the threat. Such advanced warning systems seemingly failed at Dol, presumably because Geoffrey Grennonat controlled the roads to the south. Although this defeat must have given encouragement to many who had previously thought William invulnerable, in immediate terms it had no significant repercussions. This was presumably because the removal of the threat that William might either take over parts of Brittany or substantially increase his hegemonic power there by taking Dol meant that neither Philip I nor Fulk Rechin had any further reason to become involved; their respective priorities lay close to the royal lands around Paris, where the possibility of Count Simon’s withdrawal from the world made the situation very uncertain, and in Maine. Fulk’s attack on the castle of La Flèche in southern Maine in either late 1076 or early 1077 might well have been an attempt to profit from William’s discomfiture at Dol and to chip away at a vulnerable frontier position. But, assisted by reinforcements from Normandy, it was beaten off by the castle’s defenders, led by John de La Flèche, a consistent supporter of William among the aristocracy of Maine. Fulk himself was badly wounded.71 It is an interesting commentary on the dynamics of politics and influence that John de La Flèche, whose lordship was much closer to Anjou than to Normandy, chose to favour William. Also, William’s defeat at Dol notwithstanding, any threat from the direction of Brittany was minimal since the internal politics of the region remained as turbulent as before, with Hoel being captured and imprisoned in 1078, after which his son Count Alan IV eventually overcame Geoffrey Grennonat in 1084 to acquire hegemonic power over most of Brittany. In addition, the complexities created by Gregory VII’s short-lived acknowledgement in September 1076 of Dol’s claims to be an archbishopric and the appointment of Bishop Evenus continued to be handled by him in a way that took account of William’s interests, while at the same time not abandoning his own principles. In response to William’s diplomatic offensive in support of Juhel, Gregory replied by setting out Juhel’s faults, but indicating in his letter to William dated 21 March 1077 that he would have his legates, Bishop Hugh of Die, Hubert, and the monk Teuzo, travel to Dol to hear all sides of the matter, something that was not actually done until 22 May 1078, and then with the provision that William should be asked to send a representative.72 71 This attack on La Flèche has been deduced from Annales sancti Albinii Andegavensis, in Recueil d’annales angevine et vendômoises, 1–49, at 5 (Exercitus de Fissa), and a notice for the abbey of Saint-Nicholas of Angers that mentions that Fulk suffered a serious wound, Halphen, Le comté d’Anjou, 182–3 (Catalogue d’Actes), no. 233: Guillot, Anjou, ii, 199–200 (Catalogue des actes, no. C316). 72 Registrum, i, no. IV, 17; ii, no. V. 22 (Register, 227–8, 271–2).

cross-channel rule

393

For all that Juhel’s position was a weak one, since he had already been condemned as a simoniac at three papal councils, Gregory continued to temporize, not forcing a resolution in Evenus’s favour until January 1081. Count Simon finally gave way to his ascetic impulses between March and May 1077 and became a monk at the monastery of Saint-Oyend in the Jura. His journey to a religious life attracted much contemporary comment and was very publicly played out. He visited Gregory VII in Rome in 1076 and was advised by him that he should not withdraw from the world until the peace of his lands was assured. The level of papal involvement in this peacemaking is striking, since both Gregory’s legate Bishop Hugh of Die and Abbot Hugh of Cluny were party to it. As a result, according to Simon’s early twelfth-century Vita, he returned to northern France and made peace with King Philip I against whom he had been battling since 1074.73 William is also known to have tried to dissuade Simon from withdrawal from the world during a visit to him and Matilda that is likely to have occurred early in 1077. Their having been responsible for a crucial stage in Simon’s upbringing, such a visit was probably inevitable for personal and political reasons; events suggest that he and Matilda were probably very close.74 Despite Simon having already persuaded one fiancée, Judith, daughter of Count Robert II of the Auvergne, to become a nun rather than go through with marriage, William offered him an unnamed daughter as his wife along with advice on how penance might allay divine and ecclesiastical disapproval because Simon and Matilda were related within the prohibited degrees; according to Simon’s Vita, the daughter had already been offered to King Alfonso VI of Galicia and Robert Guiscard.75 If this is in any way near to being accurate – the reference to Robert Guiscard is exceedingly implausible – then it is possible that Adelida was going to be deployed again. Orderic’s comment about her retiring from the world after the failure of earlier marriage projects does not necessarily mean that she became a nun. And, if she had achieved a level of religious contentment, then she might well have been thought suitable for Simon in terms of both attitudes and age.76 When Simon sought permission to visit Pope Gregory VII for advice, William concurred, only for Simon to turn aside on the journey and enter 73 Vita Beati Simonis, cols. 1213–14; Cowdrey, ‘Count Simon of Crépy’s Monastic Conversion’, 261–2. 74 The personal closeness of Matilda and Simon is indicated by her decision to pay for his tomb, Vita Beati Simonis, col. 1222 (trans. van Houts, Normans in Europe, 199) (quia eum nutrierat . . . Ob uiri Dei memoriam ad sepulturam componenda, Romae auri quamplurimum et argenti monacho deferente transmisit). 75 Vita Beati Simonis, cols. 1215–16 (trans. van Houts, Normans in Europe, 198–9). 76 OV, iii, 114–15.

394

william the conqueror

religion. William’s comments to Simon about penance, involving as they did advice that he should consult wise churchmen with a view to giving alms or founding a monastery to legalize the marriage, very likely reflect his perceptions of his own experience. As such, they presumably again encapsulate an inner confidence that he and Matilda had done the right thing in founding the two abbeys at Caen. The consequence of Simon’s withdrawal from the world was that his large principality was dismembered, with the region known as the French Vexin, which lay between William’s lands and the French king’s, passing to King Philip as a vassal of the abbey of Saint-Denis.77 Although this might have been the cause of hostilities between William and Philip, it looks as if it persuaded them to make peace; Philip also confirmed Simon’s grants to the abbey of Le Bec.78 The one source to mention an actual treaty between the two kings, the ‘E’ version of the Chronicle, does, however, ruefully add the observation that ‘it held good only for a little while’.79 It did, however, seemingly last for all of the year 1077 and it is very probable that collaborative relations lasted for longer than this. From both perspectives, there was surely little to be gained from a continuation of hostilities; on William’s side, it was desirable to lick wounds and reassess after being pulled to and fro across the Channel during 1075 and 1076, while for Philip, it was probably good policy to consolidate his gains. For both, it would probably have been unseemly to seek to profit too obviously from the acts of a holy man and to go against peace-making that had been sponsored by Gregory VII. An unprecedented number of religious events took place in Normandy during the summer and autumn of 1077, with, in total, five dedications of major churches being known to us. Although no source explicitly says so, the consecration of so many abbeys and cathedrals in such a short space of time must have been deliberate; the only parallel is the dedication of the abbeys of La Trinité of Caen, Jumièges, and Saint-Pierre-sur-Dives in 1066 and 1067, one taking place immediately before, and two very soon after, the expedition of 1066. It is possible that a time of relative peace was chosen as being the right moment to give appropriate thanks for achievement and, perhaps, to atone, with recent turmoil possibly reflecting divine disapproval. Although William’s date of birth is unknown, he must have been around fifty years of age in 1077, an anniversary that might have been influential.

77

37. 78 79

On the dismemberment of Simon’s principality, Feuchère, ‘Une tentative manquée’, Recueil des actes de Philippe Ier, no. 90. ASC, ‘E’, 1077.

cross-channel rule

395

Bishop Odo’s new cathedral at Bayeux and William’s abbey of SaintEtienne of Caen were consecrated on 14 July and 13 September respectively, both in the king’s presence.80 If a second dedication of Saint-Etienne was involved, following a first one in 1073, it was in all likelihood performed in a significantly extended church, including the transepts and some bays of the nave.81 The cathedral of Evreux was consecrated at an unknown date, the abbey of Saint-Désir of Lisieux on 17 July, and the abbey church of Le Bec on 23 October; Orderic’s account implies that William was present at Evreux, but explicitly says that he was unable to attend Le Bec’s consecration. The two English archbishops, Lanfranc and Thomas, both crossed the Channel for at least some of these great events, with both having compelling personal reasons to attend certain of them; in Lanfranc’s case, the dedications at Caen and Le Bec in particular, and in Thomas’s, the cathedral at Bayeux. Orderic also says that Lanfranc was present at the consecration of Bayeux cathedral, something that is very likely since he and Odo had had to work closely together over many years in Normandy and south-east England.82 Although the charters surviving from the time of the consecrations are unhelpful, many of the greatest Anglo-Norman magnates must surely also have been present. The evidence that the same craftsmen had worked on parts of the churches at Bayeux and Saint-Etienne of Caen suggests that the two building campaigns had to an extent been co-ordinated by the two brothers.83 There is also evidence of collaboration between their two schools.84 The holding of the two consecrations so close together should probably, therefore, be seen as a display of co-ordinated family piety; William also made a generous grant to the cathedral of Bayeux at the time of its dedication, symbolically transferred by him placing on the cathedral’s 80 OV, iii, 10–13, is the only source for all of the dedications. On the year, Jean-Michel Bouvris, ‘La dédicace de l’église cathédrale Notre-Dame de Bayeux (14 juillet 1077)’, Société des sciences, arts et belles-lettres de Bayeux, xxviii (1982), 3–16. For the date of the dedication of Saint-Etienne, Musset, AAC, 14–15. For an edition of the charter giving the date, which is in the abbey’s cartulary rediscovered in 1996, Tamiko Fujimoto, ‘Recherche sur l’écrit documentaire au Moyen Age. Edition et commentaire du cartulaire de l’abbaye SaintEtienne de Caen (XIIe siècle)’, 2 vols., Thèse de Doctorat, Université de Caen BasseNormandie (2012), ii (Edition), 80–2, no. 20. 81 Carlson, The Abbey Church of Saint-Etienne at Caen, 105–7. 82 OV, iii, 12–13. Lanfranc’s presence at Bayeux has been doubted by Cowdrey, Lanfranc, 207, note 6, on the grounds that a writ dated to the dedication and concerned with English affairs was addressed to him (Regesta, no. 83). However, since the subject matter of the writ concerned the shire court of Kent, the insertion of his name in the address clause would have been formulaic and would not reflect on his immediate availability as a recipient. For the date of the dedication of Saint-Désir, recorded in a charter destroyed in 1944, Bouvris, ‘La dédicace’, 13. 83 Baylé, ‘Les évêques et l’architecture normande’, 169–71. 84 Spear, ‘The School of Caen’, 62–5.

396

william the conqueror

high altar a copper helmet surmounted by a golden crown that was still preserved at Bayeux in the fifteenth century.85 William appears to have travelled to Upper Normandy between the dedications at Bayeux and Caen. Again according to Orderic, he settled a dispute at Rouen between the canons of Lisieux cathedral and the nuns of Saint-Désir of Lisieux in late July about where the body of Bishop Hugh of Lisieux, who had died on 17 July, should be buried. William decided the case in favour of the nuns, apparently on the grounds that the bishop had already promised them that he would be buried at their church.86 Like so many other episcopal appointments after 1066 on both sides of the Channel, Hugh’s replacement, Gilbert Maminot, was a member of the royal chapel while also being someone whose family had territorial interests in the diocese.87 He was also one of William’s doctors. At around this time, apparently while on his way to bury Bishop Hugh, Archbishop John of Rouen was incapacitated by a severe stroke. He lingered on for another two years, unable to perform his ministry, but making sporadic and disturbing attempts to do so.88 This became an element in relations with Gregory VII and also in the governance of the Church in Normandy that had to be managed during the complex years between 1078 and 1080. A short visit to England of a few weeks would explain William’s otherwise surprising absence from the dedication of the abbey church of Le Bec at the end of October.89 But after his return the quarrel with his son Robert that in all probability broke out during the summer campaigning season of 1078 dominated politics for more than a year and ensured that William had to remain in Normandy. Nonetheless, this was also a period of significant developments in the ethos and practice of governance on both sides of the Channel. Changes in the relationship with Gregory VII also became more overt. THE QUARREL WITH ROBERT CURTHOSE While supplying by far the most detailed and informative of the accounts of William and Robert’s quarrel, Orderic is as chronologically muddled on it as 85 Regesta, no. 52; Ordinaire et Coutumier de l’église cathédrale de Bayeux, ed. chanoine Ulysse Chevalier (Paris, 1902), 418; E. Deslandes, Le Trésor de l’église Notre-Dame de Bayeux, d’après les inventaires manuscrits de 1476, 1480, 1498 conservés à la Bibliothèque du chapitre de Bayeux (Paris, 1898), 378, 380–1; Bouvris, ‘La dédicace’, 10; François Neveux, ‘Les reliques de la cathédrale de Bayeux’, in Pierre Bouet and François Neveux (eds.), Les Saints dans la Normandie médiévale (Caen, 2000), 109–33, at 115. 86 OV, iii, 16–19. 87 For Gilbert, Spear, Personnel, 170. 88 OV, iii, 18–19; AAR, 45, 60; Allen, ‘ “A Proud and Headstrong Man” ’, 27–9. 89 Lanfranc Letters, no. 39, suggests that a visit took place.

cross-channel rule

397

he is on William’s interventions in Maine. However, the proximity of L’Aigle, where the incident that brought underlying tensions into the open took place, to the abbey of Saint-Evroult does suggest that he had special knowledge of its opening events.90 In his account, the spark that ignited the quarrel was an argument between Robert and his younger brothers when they and their followers made a lot of noise and poured water – presumably a euphemism for urinated – onto Robert and his men on the lower storey of the house where he was lodging at L’Aigle on Normandy’s southern frontier. Accepted at face value, Orderic’s narrative suggests that the row that followed brought into the open jealousies about status and expectation that may have been festering for a long time, with, among other things, Robert being accused of aiming to acquire the whole inheritance of Normandy and England and behaving as if he was his father’s equal. Robert’s followers egged him on to respond to the taunts and told him not to tolerate such insults. William arrived and patched up a truce, but Robert and his associates slipped away during the next night and tried to take the tower at Rouen by a trick. Thwarted by the castellan, the king’s butler Roger d’Ivry, and alarmed at news of William’s fury and his order that all of them be seized, they fled into exile.91 The names of some of the individuals involved in the attack may appear in a list of excommunicates written at Rouen at around this time.92 If, as seems likely, the quarrel started during the campaigning season of 1078, then, although Orderic does not explain its rationale, it looks as if William may have been setting out to conduct tidying-up operations into the region known as the Corbonnais, which lies to the east of the lordship of Bellême, where, as we have seen, Rotrou, count of the Perche, had consolidated regional power with the support of the French kings.93 The projected campaign may also have had something to do with another disruptive event, the murder of Roger de Montgommery’s wife Mabel at Bures-sur-Dives by Hugh Bunel (also referred to as de Ialgeio), which is dated by Orderic to a December that is customarily thought to be December 1077. Her status as heiress to the lordship of Bellême had enabled her husband to operate beyond the lands directly subject to William’s jurisdiction. If, as Orderic’s 90 For Orderic’s two accounts of the quarrels, OV, ii, 356–61; iii, 96–113. Indicative of his confused chronology is that in his second longer account he recounts events that he said took place after the council of Lillebonne of 1080 which include the siege of Gerberoi of January 1079. The weaknesses of Orderic’s chronology have been analyzed in Lack, ‘Robert Curthose’, 126–9. For treatments of the quarrel, Barlow, William Rufus, 32–7; Aird, Robert Curthose, 78–90; Katherine Lack, Conqueror’s Son: Duke Robert Curthose, Thwarted King (Stroud, 2007), 20–7; idem, ‘Robert Curthose’, 117–31. 91 OV, ii, 356–9. 92 Richard Allen, ‘The Earliest Known List of Excommunicates from Ducal Normandy’, JMH, xxxix (2013), 394–415, at 408–9, 413. 93 Thompson, Power and Border Lordship, 35–42.

398

william the conqueror

account suggests, the murder was motivated by a classic honour-based feud, with Mabel having deprived Hugh’s family of their castle, the potential for destabilization that might follow on from her death was considerable, all the more so because she is said to have deprived others of lands and castles. Hugh and his accomplices fled Normandy and settled in Apulia.94 Some aspects of Orderic’s narrative are hard to believe. It is difficult, for example, to think that William’s youngest son Henry, who was around ten years old at the time, was much involved. But, in general terms, the story he tells rings true as an almost stereotypical quarrel between a ruler of mature years unwilling to relinquish any of his hold on power and an expectant eldest son, a phenomenon with which the eleventh and twelfth centuries abound. It also fits with the sort of boisterous masculinity that underpinned the military societies of these times. Robert by this date was approaching his mid-twenties. Among the names that Orderic associates with his rebellion are the potential leaders of the next generation of the Norman aristocratic elite, Robert de Bellême, who was the son of Roger de Montgommery, William de Breteuil, son of the late William fitz Osbern and the successor to his father’s immense Norman lands, Roger, son of Richard fitz Gilbert, and Robert de Montbrai (Mowbray), a young kinsman of Bishop Geoffrey of Coutances.95 The scenario created takes us back again to the social climate of the 1040s, a time when a younger generation started to push an older one to one side, except that William was now a member of the older generation, something that makes a huge difference to the way in which events unfolded. Both Orderic and Malmesbury suggest that significant temperamental differences between father and son were an aspect of the quarrel. For the former, Robert was dissolute and inconstant, a great contrast to his stern and puritanical father.96 Malmesbury has William ridicule his son’s efforts in the early stages of the revolt in a patronizing dismissal of a young man playing at politics.97 Such assessments must owe a lot to knowledge of Robert’s later failures, even if William’s oath reported as direct speech by Malmesbury that ‘By God’s resurrection, this little Robert [or Robin – the word used is Robelinus] Curthose will be a hero!’ does smack of the domineering, and arguably unfeeling, figure that the Conqueror may well have been; the same theme, that Robert was short, and probably plump, is indicated by his nickname ‘Curthose’. What the quarrel most certainly was not was a conflict about personal morality. Orderic might refer to the men and OV, iii, 136–7, with the suggested date discussed at 136, note 1. Ibid., 100–3. 96 See in particular ibid., 98–9. 97 GR, i, 700–1. 94 95

cross-channel rule

399

women of William’s time as more restrained in their personal demeanour than those of his own times, but, puritanical as William undoubtedly was in sexual and religious matters, his court was in all likelihood as unlicensed as any other; both of his two younger sons William and Henry were far from being paragons of virtue and neither had any difficulty remaining on good terms with their father. The contrast between the circumstances of Robert’s mid- and late teens in the late 1060s and early 1070s, when William was preoccupied with wars in England and the two can scarcely have seen one another, and the period after 1073 when the family seem to have been almost constantly in each other’s company, may well have generated friction.98 As we know, promises of the succession to Normandy had been made to Robert and his homage to Count Geoffrey the Bearded for the county of Maine in 1063 gave him standing there, as a result of which his status had been formally deployed to uphold William’s rule in both places.99 Yet, as we have seen, it was Matilda’s authority to act in William’s absence that was regularly emphasized and, when William was in Normandy, everything passed through him. Most of Orderic’s two accounts of the quarrel focus on rivalry between William’s sons and on Robert’s demands that his father give him power in Normandy and Maine. A further theme in them is the uncertainty about the English succession: on one occasion during the quarrel Robert is portrayed by Orderic as perhaps coveting the whole of William’s inheritance; on another, as asking for a share of England.100 Orderic tells us that Robert had already repeatedly asked his father for more power and that William had not only refused his request, but had indicated that he intended to retain authority over Normandy and England until his death.101 All this points to a very well-known theme in medieval royal and aristocratic families, namely, the heir pushing for responsibility against a father determinedly hanging on to it. It also highlights again the corrosive impact of succession politics. Robert was, after all, not aware of exactly what he was the heir to. Robert’s flight from Normandy took him first to the very region into which his father had been proposing to campaign. Welcomed there by Hugh de Châteauneuf-en-Thimerais, Robert was allowed to set himself up in the castle of Rémalard, twenty-five miles due south of L’Aigle and the same distance due east of Alençon, with the objective of harrying the 98 Thus, all four Norman diplomas that date to 1074 have William, Matilda, and their sons Robert and William among the signa, Regesta, nos. 26, 27, 261, 262, with the third surviving son Henry appearing in no. 261. 99 For the evidence assembled, Lack, ‘Robert Curthose’, 120–6. 100 indignum ducebat quod frater eorum solus habere patrium ius ambiebat, OV, ii, 358–9; a genitore tuo partem regni Albionis exige, ibid., iii, 96–7. 101 Ibid., ii, 356–7; iii, 98–9.

400

william the conqueror

Norman frontier. Hugh’s lands between Mortagne and Dreux made him into a neighbour and rival of Count Rotrou, as well as someone else with aspirations to dominate the lands to the south of Normandy. William’s response was to buy Rotrou’s support by granting him a subsidy, a piece of business typical of William’s management of the politically fluid lands around Normandy’s borders by which individuals were drawn temporarily or permanently into collaborative relationships with him; in this case, an old alliance was resurrected, since Rotrou’s son Geoffrey had fought at Hastings. William then encircled Rémalard with four siege-castles.102 Although Orderic’s narrative then breaks off, it looks as if Robert must have travelled to seek help from Robert the Frisian, count of Flanders, eventually establishing himself in the castle of Gerberoi, close to Normandy’s southeastern frontier on the border between the French and the Norman Vexin; presumably he had assistance from Philip I, although not necessarily his actual support.103 William’s decision to move against Gerberoi in mid-winter may well have been taken because Robert’s troops were using the location to launch raids across the Norman frontier. Although he had garrisoned the neighbouring Norman castles, William’s inability to contain an enemy established so close to the duchy would have been embarrassing.104 A charter for the collegiate church of Saint-Quentin of Beauvais, which explicitly states that it was enacted while William and King Philip I were besieging Gerberoi, probably indicates that the French king had always been ambivalent towards Robert, while at the same time thinking that his cause was sufficiently justified to merit being helped. Once warfare had started that was going to damage his lands as much as William’s, he may well have decided that it was in his best interests to collaborate with an attempt to reach a settlement. The presence at Gerberoi of Geoffrey de Chaumontsur-Loire, the frequent intermediary between William and other northern French rulers since before 1066 and a participant in the 1066 invasion, and of Anselm, recently appointed as abbot of Le Bec, a monastery with extensive links to the lands of the French king, also suggests that negotiations took place.105 The diploma, of which the original does not survive, was drawn up in the diplomatic form of a Capetian royal charter. The text, preserved in a twelfth-century cartulary, describes the presence of the two Ibid., ii, 358–61; Thompson, Power and Border Lordship, 41–3. ASC, ‘D’, ‘E’, 1079; OV, iii, 108–9; JW, iii, 30–1. 104 OV, iii, 108–11. 105 Regesta, no. 28 (Actum publice in obsidione predictorum regum . . . circa Gerborredum); Dunbabin, ‘Geoffrey of Chaumont’, 112–13; Sally N. Vaughn, Anselm of Bec and Robert of Meulan: The Innocence of the Dove and the Wisdom of the Serpent (Berkeley and Los Angeles, CA, and London, 1987), 55–6. 102 103

cross-channel rule

401

kings’ seals and that their crosses were placed alongside each other above all the other signa. This document, the only surviving record of the two kings in each other’s company after 1066, displays their equal status as kings; no reference at all is made to William as duke of Normandy or count of Maine. The earliest narratives of the events at Gerberoi, the two versions of the Anglo-Saxon Chronicle, tell in slightly different ways of a conflict in which Robert wounded his father and the king had a horse shot from under him. William’s second surviving son William was also wounded in the affray. ‘D’ inserts an anecdote about an Englishman, Toki son of Wigod of Wallingford, who was killed by a crossbow bolt immediately after supplying William with a new horse.106 Interesting because it once more shows William using English troops in France, the story suggests that Toki was a member of William’s personal elite force, his military household, since the movement of a large troop of shire levies to France in winter seems unlikely. It also suggests that William was in considerable personal danger. Of the early twelfth-century sources, William of Malmesbury tells much the same story as the two versions of the Chronicle, while John of Worcester embellishes events by having Robert recognize his father by his voice and order him to remount his horse, thereby presumably sparing his life.107 Orderic, the fullest source for many of these events, surprisingly omits the incident. Not only was it highly unusual for William to be caught out in battle in this way, but the story that Robert did not immediately recognize his father is incredible; while personal heraldic devices had not yet been adopted, William’s physique and his place among his troops must have made him stand out. It looks as if Robert launched a direct personal assault on his father, but then either drew back from administering the coup de grâce or, much more likely, set out to show him that he was vulnerable to attack. The incident also suggests that William did not have much of an army with him; he would normally have embarked on a long, cautious blockade rather than engaging in risky hand-to-hand fighting of the kind that seems to have occurred. It is likely once more that personal honour and reputation were deemed to be at stake; he may well have thought it essential to be seen to confront Robert as soon as he could. From William’s point of view the whole episode must ultimately have been a humiliation. After a three-week siege he retreated discomfited to Rouen. Robert meanwhile withdrew to Flanders.108

ASC, ‘D’, ‘E’, 1079; Barlow, William Rufus, 35–6; Aird, Robert Curthose, 87–8. JW, iii, 30–3; GR, i, 476–7. 108 OV, iii, 110–11; ASC, ‘D’, 1079. 106 107

402

william the conqueror

The quarrel is known to have caused conflict between William and Matilda, who, apparently distraught, took to sending money and valuables to her eldest son. When William found out, he accused her of disloyalty and forbade her to continue the gifts. When she continued, a stand-up row followed with neither apparently backing down. William’s argument was based on political security; Matilda’s, on a mother’s love for her son. William furiously ordered the blinding of one of her servants who had acted as a messenger, a Breton named Samson who fled the court and became a monk at Orderic’s abbey of Saint-Evroult where he lived out the rest of his life. Matilda is also said to have sought guidance from a hermit in Germany who was reputed to have the gift of prophecy.109 This very public domestic disagreement also shows beyond doubt that right was far from being the monopoly of either of the two sides in the quarrel. The events at Gerberoi having created an impasse, although we know little about the detail, most of 1079 must have been taken up with attempts to repair relations. The earliest secure date by which we know an agreement had been reached is Easter (12 April) 1080 when William and Robert were together with Matilda, William Rufus, and the royal court on the island of Oissel to the south of Rouen. Present with them were two French archbishops, Richard of Bourges and Warmund of Vienne, who are likely to have been mediators acting on Gregory VII’s behalf.110 There is, however, good reason to think that a settlement was reached quite a lot earlier. Orderic records the distress of the greatest Norman magnates at the discord and the split in families that had occurred. He specifically names Earl Roger de Montgommery, Earl Hugh of Chester, Hugh de Gournay, Hugh de Grandmesnil, and Roger de Beaumont and his two sons as those who took the lead in trying to bring about reconciliation. Even in the face of this formidable coalition, William was unrelenting, refusing to forgive or come to terms with a son guilty of such disloyalty. Orderic attributes to him the statement that Robert deserved death, an observation possibly pertinent to William’s approving the execution of Waltheof.111 Count Simon also paid a visit to William and Matilda that must date to the spring of 1079. Simon and Matilda were in all probability in each other’s company and that of King Philip I at Compiègne on 7 March 1079 for the translation of a set of relics; Matilda is recorded as having financed the extremely expensive new reliquary, so was presumably present.112 Although now a hermit, Simon had been sent by Gregory VII on a diplomatic mission OV, iii, 102–5. Regesta, no. 235. 111 For the negotiations, OV, iii, 110–13. On the chronology of the settlement, Lack, ‘Robert Curthose’, 128–9. 112 Recueil des actes de Philippe Ier, no. 126. 109 110

cross-channel rule

403

to the French king. His diversion to Normandy must have been motivated by concern for the family that had nurtured him at an important period in his life, as well as being part of the search to reconcile father and son. He, William, and Matilda are said to have had long and deep religious conversations, with the king making generous gifts to those who sought Simon’s blessing.113 Although no source mentions it, it may also be relevant that he and Robert were almost exact contemporaries; they must have known each other well as teenagers. Simon is also likely to have been involved in the early stages of the negotiations that led to the marriage in the early 1080s of William and Matilda’s youngest daughter Adela to Stephen-Henry, the son of Count Theobald of Blois-Chartres; the regular intermediary Geoffrey de Chaumont-sur-Loire, who had been present at Gerberoi, was also likely to have been involved. The proposed marriage made sense as a reinforcement of the two rulers’ long-standing collaboration against the counts of Anjou and the kings of France that went back to the later 1040s. It also contributed to the stabilization of Simon’s dismembered territories, from which Theobald had profited because he was Simon’s brother-in-law, a situation that would enable him to act in William’s interests in the lands to the east of Normandy.114 While the presence of Archbishops Richard and Warmund in Normandy in 1080 was certainly connected with a range of issues that concerned Gregory, there is no doubt that the quarrel between William and Robert was one of them. Both were regularly used as emissaries by Gregory, and Warmund had earlier been Abbot Hugh of Cluny’s ambassador to William in the mid-1070s.115 One of Gregory’s letters shows that his regular AngloNorman legate Hubert had kept him informed of events.116 The pope’s interest in the settlement was expressed by the dispatch of three separate letters, all dated 8 May. The two to William and Matilda make no mention of the quarrel. The king is counselled to persevere in the ways of righteousness. Matilda is thanked for a letter she had sent and for the offer of generous gifts, and urged to live a good life and distribute alms. Robert is advised to obey his father with quotations from some standard biblical texts and to put aside evil counsellors. He is reminded that his father will one day

113 Vita Beati Simonis, PL, clvi, col. 1219 (trans. van Houts, The Normans in Europe, 199). For Simon’s role in search of peace, Cowdrey, ‘Count Simon of Crépy’s Monastic Conversion’, 262–3. 114 On the early negotiations for the marriage, LoPrete, Adela of Blois, 34–7. The sources for the negotiations are OV, iii, 116–17; Gesta Ambaziensium dominorum, 97–8. 115 Cowdrey, Gregory VII, 372, 382, 395. 116 Registrum., ii, no. VII. 27 (Register, 358).

404

william the conqueror

be succeeded by an heir and that he should be patient. The tone of the letter is that of a spiritual adviser concerned for its recipient’s welfare.117 Gregory’s letter to Robert probably indicates the general lines of the settlement; Robert was to mend his ways and return to his expected role at court. According to Orderic, his status as heir to Normandy was confirmed.118 From charter and other evidence it seems likely that, at least on the surface, Robert’s role was enhanced. One charter for the abbey of Lessay, dated to July 1080, includes him in his normal place among the signa, but then has him separately confirm it four days later, having travelled from Bonneville-sur-Touques to Caen.119 He appears consistently in other charters dated to 1080, a year that is unusually replete with dated documents.120 And, as we shall see, when William subsequently went to England, Robert went with him for what appears to have been his first ever extended visit to the kingdom – perhaps indeed his first visit. But it is surely unlikely that the rift was ever truly healed. According to Orderic, Robert soon returned to open defiance and William was critical in public of his son’s failings.121 Going right back to the treatment of the rebels of Val-èsDunes, we have seen plenty of evidence that William’s readiness to forgive anyone was very limited. RULING ENGLAND AS AN ABSENTEE, 1078–80 A writ specifically dated to the time of the dedication of Bayeux cathedral ordered Archbishop Lanfranc, Bishop Geoffrey of Coutances, Count Robert of Eu, and Hugh de Montfort to return to the abbey of St Augustine’s, Canterbury, property that Haimo, the sheriff of Kent, was holding, and to restore any property which the abbot had wrongly alienated or which had been taken away violently.122 Its transmission from Normandy to England indicates yet again – and superbly – that kingdom and duchy cannot be considered as self-contained entities when William’s personal authority was concerned. The presence of the place-date clause, something that is extremely unusual in William’s surviving writs, underlines the importance of the Bayeux dedication as an occasion. The names in the address clause comprise the sort of consortium to which we have already seen authority delegated, as in the inquiry into the abbey of Ely’s

Ibid., nos. VII. 25, 26, 27 (Register, 356–9). OV, iii, 112–13. 119 Regesta, no. 175(I), (II). 120 Ibid., 80–1. 121 OV, iii, 112–13. 122 Regesta, no. 83. 117 118

cross-channel rule

405

losses in the first half of the 1070s, with Geoffrey of Coutances once more in the forefront.123 Textual similarities between this writ and the famous writ by which the king ordered all his sheriffs throughout England to return lands to bishoprics and abbeys that they had wrongfully taken points to a wide-ranging campaign of inquiry and restitution which was probably ongoing at this time.124 The presence of the same individuals in the address clauses of several of the writs suggests a commission with wide-ranging responsibilities. Although there is plentiful evidence of inquiries into English land tenure and attempted restoration of lost estates from before 1077, there are signs here, as with the clarification of ecclesiastical jurisdiction around the time of the 1076 Council of Winchester, of a more concerted and generalized effort to come to grips with the problems that the conquest of England and the subsequent takeover of lands had created. It cannot be ruled out that this attempt at the restoration of illegally appropriated church lands was another part of the general programme of pious display associated with the dedication of so many Norman churches. It could even be suggested that it is an important landmark in the transition from rule shaped primarily by the exigencies of conquest and security to one that foregrounded the requirements of the supplementary oath that William had sworn on 25 December 1066. Whether or not William made a short visit to England during the winter of 1077–8, his lengthy sojourn in France raised in acute form the problems of maintaining rule in his absence. William’s half-brother Bishop Odo became centrally important in England during this period in a way that he had not previously been. Presumably crossing to England in the last months of 1077 after the period of church dedications, he was involved in several major pieces of business that reinforce the theme of justice and judicial inquiry set out in the writ about encroachments by sheriffs. One narrative, a late eleventh-century memorandum composed at Rochester, tells how the recently appointed Bishop Gundulf, who was not only a former monk of Le Bec but also an architect who worked on the Tower of London, claimed that land belonging to Freckenham (Suffolk) in Isleham (Cambridgeshire) belonged to his church against the claims made by Picot, sheriff of Cambridgeshire, who maintained that it was royal land. The process, with the shire court of Cambridgeshire initially finding for Picot, out of fear, so the narrative says, and Odo intervening and summoning representatives of the shire to London and threatening them with the ordeal by hot iron to

For this earlier Ely plea, ibid., no. 117. For the argument about textual similarities and the writ, ibid., no. 129. For writs that may belong to the same process, ibid., nos. 42, 43, 310, 330. 123 124

406

william the conqueror

prove that they were not perjurers, can be read as both an exemplar of the sort of tenurial problem present pretty well everywhere and of just rule tenaciously enforced against a local Norman tyrant.125 Life is rarely so straightforward, however. One wonders, for example, whether this is the outcome that would have been reached if Gundulf had not been so influential a personage and whether, if, as we know from Little Domesday Book, Lanfranc had not intervened earlier on Rochester’s behalf; Picot, it should be added, on the basis of the Domesday evidence, looks to have had more of a case than the Rochester narrative allows to him.126 Odo’s involvement in the affairs of the abbeys of Evesham and Ely is fully illustrative of the complexities of these inquests. In the case of Evesham, he presided over a land plea at Agildeburga (Evenlode or Four-Shire Stone, near Moreton-in-Marsh, Warwickshire) that inquired into the state of the abbey’s lands after the death of the redoubtable Abbot Æthelwig on 16 or 17 February 1077, and the appointment of his successor, Abbot Walter, a monk of Cerisy and one of Lanfranc’s chaplains, approximately three months later. The outcome was to strip the abbey of acquisitions that Æthelwig had made and which were presumably reflective of just how valued he was after 1066; the latest study has pointed out that the abbey lost forty-two out of an astonishing total of fifty-one estates that he had acquired.127 In the case of Ely, a land plea held at Kentford (Suffolk) on 2 April 1080 to inquire into the abbey’s liberty was called together through Odo’s agency.128 This dealt with the massive issues created by the aftermath of the abbey’s support for Hereward and the rebels who had installed themselves on the isle. Evesham’s record of events accuses Odo of removing estates from the abbey for his own profit, while also creating a vivid picture of the swirling currents of change and contradictory values present in the years after 1066; Abbot Walter, so it tells us, was young and inexperienced, refusing homage from some of the abbey’s tenants and, so we learn from another source, even subjecting relics of English saints whom he distrusted to ordeal by fire. But whatever the rights and wrongs of the situation after Æthelwig’s death, subsequent events initiated processes of such complexity that they were still ongoing at the time of the Domesday inquiry and led to the creation of extensive documentation well into the twelfth century.129 Regesta, no. 225. DB, ii, 318r. 127 Clarke, ‘ “Those Five Knights” ’, 4–8. 128 Regesta, no. 118 (per Baiocensem episcopum precepit hec discutere). For the course of the Ely land pleas, Liber Eliensis, 426–32; Keynes, ‘Ely Abbey 672–1109’, 48–50. 129 Thomas of Marlborough, History of the Abbey of Evesham, 172–81; Vita sancti Wistani, in Chronicon Abbatiae de Evesham, ed. W. D. Macray, RS, no. 29 (London, 1863), 325–37, at 335–7. 125 126

cross-channel rule

407

Like the earlier Penenden Heath inquest, the Evesham and Ely inquiries started with grand set-piece occasions. There is disagreement in the sources as to how many shires’ representatives actually attended at Four-Shire Stone, albeit this discrepancy is resolvable when account is taken of the lost shire of Winchcombeshire. The Ely memorandum that describes the Kentford inquest is also internally contradictory, mentioning both three and four shires. The key point is that, with the exception of the Penenden Heath inquiry, whose driving force was Lanfranc, and the earlier Ely inquiry, these occasions represent a new phase in the history of conquered England. Conceptually and thematically they lead on directly to the multiplicity of placita that appear in Domesday Book. Odo’s personal status in all this has been much discussed. Was he effectively acting as William’s alter ego exercising both de facto and de iure regal powers? The answer has to be a qualified yes, but a distinctly qualified one. Thus, while he confirmed the outcome of the Four-Shire Stone inquiry in a writ, the king also confirmed it, albeit with less detail about what had been decided in the writ in his name. Anyone who consulted Evesham’s archive would therefore discover the specifics of the settlement in Odo’s writ and learn that it had been confirmed by the king; a documentary relationship of a kind combining a narrative in a writ in the name of the person responsible for an action and a précis in a royal confirmation is one that appears frequently in similar contexts during the more voluminously documented reign of Henry I.130 And, while the Kentford inquest was set up by Odo, it was the king who ordered it to happen and who confirmed its outcome. Odo’s special eminence at this time is confirmed by some spectacular references in the literary sources, by references in Domesday Book to transactions performed on his authority, and also by the way in which Domesday uniquely treats a judicial verdict of his as having the same authority as the king’s; no one else had the same status.131 In all this, there are elements that resemble very closely William fitz Osbern’s special status in 1066–70. But William’s presence is never far away. Thus, in relation to another aspect of these inquiries, albeit probably at a later date in the 1080s, a dispute between the abbey of Evesham and Bishop Wulfstan of Worcester about jurisdictional rights, heard by Bishop Geoffrey of Coutances, was initiated by a writ sent from Normandy.132

For context, Patrick Wormald, ‘Lordship and Justice in the Early English Kingdom: Oswaldslow Revisited’, in idem, Legal Culture in the Early Medieval West: Law as Text, Image and Experience (London, 1999), 313–32, at 318–24. 130 Regesta, nos. 134, 135; Bates, ‘The Representation of Queens and Queenship’, 294–8. 131 Bates, ‘Origins of the Justiciarship’, 2–4. 132 Regesta, no. 349.

408

william the conqueror

Mael Coluim (Malcolm) III’s raid as far south as the Tyne in August 1079 must have revived memories of 1069–70, since, so we are told, he and his forces killed many people and took much plunder and many slaves, while apparently meeting no effective resistance.133 A breach of the 1072 Abernethy agreement, this invasion looks very much like an opportunistic attempt to exploit William’s preoccupations with his son and in northern France in general. Bishop Walcher’s failure to respond must have been among the factors contributing to his murder on 14 May 1080 at Gateshead. The tale of his death, as told in several sources, indicates that feuds within his household and an inability to reconcile tensions between newcomers and natives were further reasons why the violence occurred.134 Praised in the late eleventh- and early twelfth-century Durham histories for his part in the beginnings of the revival of monasticism in northern England, Walcher was the tenth man since 993 exercising the power of an earl in Northumbria to meet a violent end.135 Malcolm’s raid and the killing of Walcher were events that were certain to provoke a response from William once he was able to return to England. William’s absences notwithstanding, the infrastructure of kingship continued to be developed, above all at the crown-wearing sites of Westminster and Winchester. At Winchester, in addition to the new hall that William had had started in c.1070, Bishop Walchelin began the construction of a new cathedral in 1079 that was from the start intended completely to replace its English predecessor.136 A vast building on a scale unrivalled in either England or Normandy, and almost double the length of the abbey church of SaintEtienne of Caen, it was paralleled in the whole of western Christendom only by the new abbey church at Cluny that St Hugh was having built at the same time. It also has significant features that drew inspiration from the Empire and, specifically, from the imperial burial church at Speyer.137 Symbolically, it again made the link to Constantine and may well also in its final form have been designed to provide facilities for royal display at crown-wearings which included a royal pew in a gallery at the west end that looked over the

ASC, ‘D’, ‘E’, 1079; HRA, 208. JW, iii, 32–7; HRA, 208–11; GR, i, 498–501; GP, i, 412–15. For the main modern accounts of Walcher’s death, Aird, St Cuthbert and the Normans, 94–9; Henrietta Leyser, ‘Walcher, Earl of Northumbria (d.1080)’, ODNB. 135 For the statistic, Leyser, ‘Walcher, Earl of Northumbria’. 136 For the date, Annales monasterii de Wintonia, in Annales Monastici, 5 vols., RS (London, 1864–9), ii, 1–125, at 32. For the site, Biddle (ed.), Winchester in the Early Middle Ages, 309–11. 137 Fernie, The Architecture of Norman England, 117–21; idem, ‘Three Romanesque Great Churches in Germany, France and England, and the Discipline of Architectural History’, Architectural History, liv (2011), 1–22, at 11–17. 133 134

cross-channel rule

409

space between the cathedral and the palace.138 In 1086 William personally contributed to the work with a grant of timber from Hempage Wood.139 For all that Bishop Walchelin was ultimately responsible for the cathedral’s design, it was most assuredly William and his conception of rule that he had in mind when he embarked on such a huge project. At Westminster, major work on the abbey established by Edward the Confessor was certainly going on at this time, with significant, but unspecified, building taking place during the abbacy of the former monk of Fécamp, Vitalis (1076–85).140 It has recently been suggested that substantial sections of the church and the monastic buildings were completed in this period.141 And alongside ecclesiastical construction, as at Winchester, there was also building on the royal palace, with a miracle attributed to St Augustine included in the collection compiled by Goscelin of SaintBertin mentioning fifteen ships carrying stone from Caen to Westminster for work on the royal palace.142 In this case, however, we know nothing about the work that was being done, since nothing survives above ground and all is dwarfed by the vast building attributed to William Rufus. All that can be said is that fifteen is a large number of ships and that the use of Caen stone is indicative of the construction of a prestigious building also intended to impress. NORMANDY, MAINE, AND NORTHERN FRANCE, 1078–80 A charter from the abbey of Saint-Vincent of Le Mans refers to a peace made by William and Fulk Rechin at an unidentified place named Castellum Vallium. Since William (of Saint-Calais), who became abbot of SaintVincent in 1078 and bishop of Durham on 9 November 1080, is mentioned in the charter, the agreement, which might have followed a military campaign, presumably dates to either the summer of 1078 or to 1079. It 138 Martin Biddle, ‘Seasonal Festivals and Residence: Winchester, Westminster and Gloucester in the Tenth to Twelfth Centuries’, ANS, viii (1986), 51–72, at 60–3. 139 Annales monasterii de Wintonia, 34–5; Biddle (ed.), Winchester in the Early Middle Ages, 308, note 4. 140 B. W. Scholz, ‘Sulcard of Westminster: Prologus de constructione Westmonasterii’, Traditio, xx (1964), 59–91, at 81 (quod regitis et construitis monasterio, referring to Abbot Vitalis). 141 Stuart Harrison and John McNeill, ‘The Romanesque Buildings at Westminster Abbey’, in Warwick Rodwell and Tim Tatton-Brown (eds.), Westminster: II. The Art, Architecture and Archaeology of the Royal Palace, British Archaeological Association Conference Transactions, xxxix, Part II (Leeds, 2015), 69–103, at 70–3. 142 Richard Gem, ‘Canterbury and the Cushion Capital: A Commentary on Passages from Goscelin’s De Miraculis Sancti Augustini’, in Neil Stratford (ed.), Romanesque and Gothic: Essays for George Zarnecki, 2 vols. (Woodbridge, 1987), i, 83–101, at 83, 88–93, 98–9 (sub primo ex Normannis Anglorum rege Willelmo and conuehendis lapidibus ad regis palatium Westmonasterii).

410

william the conqueror

also certainly involved a visit to Le Mans. The party that accompanied William, including as it did Bishop Geoffrey of Coutances and Roger de Montgommery, was a prestigious one.143 The visit could also have been the time when Abbot Rainald of Saint-Pierre de la Couture was released from confinement. On whatever terms this was done, it did not, however, prevent Rainald asking Bishop Arnold of Le Mans to reinstate him, thereby unleashing a controversy that involved Arnold firstly seeking help and advice from his metropolitan Archbishop Ralph of Tours; the case was subsequently taken to Gregory VII by Archbishop Gebuin of Lyons under the auspices of his primacy over Tours and other metropolitans.144 William and those advising him nonetheless attempted to be diplomatic and to construct networks to reinforce the change; the new abbot imposed on the community, the prior Juhel, was one of its members and was a Norman from a family whose lands lay in several parts of the duchy, but most notably in the Avranchin. Juhel’s kinsmen subsequently made a grant to the abbey of lands in western Normandy.145 As in the case of Bishop Arnold’s appointment in 1065, an attempt was being made to acknowledge sensitivities and to build networks between western Normandy and Maine. The same strategy must have underlain the appointment of Abbot William of Saint-Vincent in 1078. Although for a time a protégé of Odo of Bayeux, he seems to have been originally from Maine and always to have felt that his prime loyalties lay there, leaving Odo’s circle to become a monk in Maine at the abbey of Saint-Calais, an abbey of which William’s father had also been a monk and from which William was recruited to Saint-Vincent.146 But the tendency of local interests to dispute appointments of this kind nonetheless made William very dependent on Gregory VII’s support, as he previously had been on Alexander II’s. 143 Regesta, no. 174. Possible identifications are within Le Lude (Sarthe, chef-lieu du canton) or Château-L’Hermitage (Sarthe, cant. La Fléche). I am grateful to Daniel Pichot and Florian Mazel for a discussion of the problems of identifying the site of the agreement. 144 RHF, xiv, 668–9. See further François Villard, ‘Primatie des Gaules et réforme grégorienne’, Bibliothèque de l’Ecole des Chartes, cil (1991), 421–34, at 424–8; Cowdrey, Gregory VII, 391–3. 145 Cartulaire des abbayes de Saint-Pierre de La Couture et de Saint-Pierre de Solesmes, ed. Les Bénédictins de Solesmes (Le Mans, 1881); Jean-Michel Bouvris, ‘Pour une étude prosopographique des familles nobles d’importance moyenne en Normandie au XIe siècle: l’exemple du lignage des Dastin’, Revue de l’Avranchin et du Pays de Granville, lxi (1984), 65–101, at 68–75; Hillion, ‘Arnaud et Hoël’, 64–8. 146 LDE, 222–5, with the evidence for his appointment and pre-Durham career surveyed at p.222, notes 2 and 3. See also William M. Aird, ‘An Absent Friend: The Career of William of St Calais’, in David Rollason, Margaret Harvey, and Michael Prestwich (eds.), AngloNorman Durham, 1093–1193 (Woodbridge, 1994), 283–97, at 287–9.

cross-channel rule

411

Two figures in the Norman Church whose careers went back to the time of William’s father died during 1078, Abbot John of Fécamp on 22 February and Abbot Herluin of Le Bec on 26 August, both of them after long and very influential lives.147 At Fécamp, we are simply told that the new abbot, William de Rots, was appointed on William’s orders.148 A former dean, chanter, and archdeacon of Bayeux cathedral and, for one year only before his appointment to Fécamp, a monk of Saint-Etienne of Caen, his rapid promotion reflects the new sources of patronage that emerged after 1066 with Lanfranc, Odo of Bayeux, and the royal chapel at their heart. Living on until 1107, William de Rots has the reputation of having been an excellent abbot. There are, however, some signs of tensions developing between Abbot John and William during the former’s last years, with John’s letter to the king agreeing to the appointment of Abbot Vitalis of Bernay to Westminster in 1076 also pleading that long-standing requests for the settlement of some claims to property be dealt with as the venerable abbot was very old and near to death. Despite the many decades of collaboration and Fécamp’s distinguished place in the duchy’s history, William’s reign is largely ignored in the abbey’s liturgy with the exception of a mention of Matilda.149 This may reflect a drift away from Fécamp on William’s part. Some of the abbey’s losses were only restored after William’s death on 7 July 1088 at the start of his son Robert Curthose’s reign, with William said to have taken property away in anger (ira commotus).150 The later years of William’s reign also seem to have witnessed the start of a reduction in the importance of the ducal palace, which became less central than previously to William’s rule.151 At Le Bec, the history of Anselm’s appointment as Herluin’s successor is set out in the De libertate Beccensis monasterii, written between c.1136 and 1140 to justify the abbey’s special status. The Le Bec story is that the monks unanimously elected the abbey’s prior, the learned Anselm, as Herluin’s successor and then informed William of what they had done. There was then an inspection of the procedure on the king’s behalf by a group consisting of Roger de Montgommery, William de Breteuil, and Roger de Beaumont, after which the new abbot was invested by William. This rings For the main material on both successions, NMProsopographie, 7–11, 105–11. Catalogus abbatum Fiscannensium, ed. Labbé, Novae Bibliothecae Manuscript. Librorum, i, 328–9, at 328 (abbas fuit electione Guillelmi regis). 149 Regesta, no. 139, Appendix no. 1; The Ordinal of the Abbey of the Holy Trinity Fécamp (Fécamp, Musée de la Bénédictine, Ms 186), ed. David Chadd, 2 vols., Henry Bradshaw Society, vols. cxi–cxii (Woodbridge, 1999–2002), i, 13–15. 150 For the charter, Haskins, Norman Institutions, 288–9; Regesta Regum Anglo-Normannorum: vol. i, Regesta Willelmi Conquestoris et Willelmi Rufi 1066–1100, no. 297. 151 Annie Renoux, Fécamp: du palais ducal au palais de Dieu (Paris, 1991), 481–9. 147

148

412

william the conqueror

true, as also, because of the special circumstances of Archbishop John’s incapacity, does the statement that Anselm made no profession of obedience to Bishop Gilbert of Evreux who, because Archbishop John, the diocesan, was unavailable, consecrated him as abbot on 22 February 1079. It is, however, most unlikely that William did not demand homage from Anselm.152 In the end, the author of the De libertate may well have utilized the absence of strictly contemporary evidence in order to mislead, as well as cashing in on Anselm’s later reputation and conduct.153 The procedure otherwise follows the lines that were normal in Normandy before the Investiture settlement of 1107 between Pope Paschal II and William’s son Henry I. Archbishop John’s inability to function did provoke a concerned letter from Gregory VII dated 4 April 1078 and the dispatch to Normandy of the legate Hubert, whom Gregory characterized as being faithful to both William and the pope, to investigate whether anything needed to be done. The mission does not appear to have had any impact since John remained technically in office into 1079, in part presumably because Gregory was satisfied that the situation was being managed, and also because he recognized that to interfere further would not have produced any results.154 In July 1079, however, two months before Archbishop John’s death on 9 September, the abbot of Saint-Etienne of Caen, William Bona Anima, who had been a canon and archdeacon of Rouen before his profession as a monk of Le Bec, was enthroned as his successor.155 According to Orderic, he was not William’s first choice, the preferred candidate being Guitmund, the monk of La Croix Saint-Leuffroy who had lectured the king on the immorality of the conquest

152 De libertate Beccensis monasterii (On the Liberty of the Monastery of Bec), in Three Treatises from Bec on the Nature of the Monastic Life, ed. Giles Constable and trans. Bernard S. Smith (Toronto, 2008), 135–67, at 138–41. 153 For the most recent discussion of De libertate and Anselm’s election, Jean-Hervé Foulon, ‘Les investitures abbatiales en Normandie: quelques réflexions autour du cas de l’abbaye du Bec-Hellouin (1034–1136)’, ANS, xxxv (2013), 181–212, at 185–7, 207–11; idem, ‘La liberté de l’abbaye du Bec entre image et réalité. Réflexions autour d’un modèle réformateur normand aux XIe et XIIe siècles’, in Barrow, Delivré, and Gazeau (eds.), Autour de Lanfranc, 57–83, at 76–8. See also Julie Potter, ‘Monastic Freedom vs. Episcopal and Aristocratic Power in the Twelfth Century: Context and Analysis of the De libertate Beccensis’, in Arnoud-Jan A. Bijsterveld, Henk Teunis, and Andrew Wareham (eds.), Negotiating Secular and Ecclesiastical Power: Western Europe in the Central Middle Ages (Turnhout, 1999), 73–85; Giles Constable, ‘Abbatial Profession in Normandy and England in the Eleventh and Twelfth Century, with Particular Attention to Bec’, in Mario Ascheri (ed.), Ins Wasser geworfen und Ozeane durchquert: Festschrift für Knut Wolfgang Nörr (Cologne, 2003), 105–20; NMPrinces, 67–87. 154 Registrum, ii, no. V. 19 (Register, 268–9). 155 Spear, Personnel, 197.

cross-channel rule

413

of England and had left Normandy to join Gregory VII’s entourage. Guitmund, so we are told, turned the offer down.156 Another phenomenon of these times is the confirmation by William of charters of a special kind for Norman monasteries whose founders and main patrons were usually members of the aristocratic elite and, in a small number of cases, the dukes themselves. Of a kind sometimes referred to as pancartes, they are effectively extensive narratives of the creation of the monastery’s endowment since its establishment. Surviving for the abbeys of Montivilliers, Fontenay, Le Bec, Notre-Dame and Saint-Léger of Préaux, Saint-Evroult, Saint-Martin of Sées, Jumièges (albeit in unfinished form), Lessay, Troarn, Saint-Sauveur-le-Vicomte, Grestain, and for the collegiate church of Saint-Evroult of Mortain, and all dating from the 1070s or early 1080s, they clearly reflect a general trend.157 The existence of documents in similar form from the two Caen abbeys and, in incomplete form from Saint-Wandrille, and their deployment at the abbeys of Le Bec and Saint-Evroult to record the English lands further illuminate a wellnigh universal process in Normandy (Plate 20).158 The only significant exceptions to the pattern are the longer established abbeys of Mont SaintMichel, Saint-Ouen of Rouen, and Fécamp, which already had extensive archives and earlier confirmations, abbeys such as Saint-Amand of Rouen and Cerisy that delayed producing a pancarte into the early twelfth century, and La Trinité-du-Mont of Rouen, which had produced a cartulary before the end of the eleventh century.159 156 OV, ii, 278–81. Orderic’s chronology is again muddled since, whereas he indicates that Guitmund had not left Normandy when the suggestion of his appointment was made, other evidence suggests that he had departed for the south earlier than this. However, his presence in Gregory VII’s entourage does not mean that an approach to him could not have been made, especially given the protracted discussions during Archbishop John’s paralysis. 157 Regesta, no. 212 (Montivilliers, 1068x76), no. 149 (Fontenay, 1070–9), no. 166 (Le Bec, 1077), no. 217 (Notre-Dame and Saint-Léger of Préaux, 1077x78 or 1080x81), no. 271 (SaintMartin of Sées, 1078x82), no. 164 (Jumièges, 1079–87), nos. 175(I) and (II) (Lessay, 1080), no. 281(I) (Troarn, 1080x2), no. 260 (Saint-Sauveur-le-Vicomte, 1080/1x85), no. 158 (Grestain, 1082), no. 215 (Saint-Evroult of Mortain, 1082). These editions contain references to the many studies that have contributed to the interpretation of these important documents. For a second modern edition of the Jumièges charter, Michel Parisse, ‘Les pancartes. Etude d’un type d’acte diplomatique’, in Parisse, Pégeot, and Tock (eds.), Pancartes monastiques, 11–62, at 48–58. 158 For Saint-Etienne, AAC, nos. 4, 6, 7 (Regesta, nos. 45, 49, 50, 54); La Trinité, AAC, nos. 8, 11, 12 (Regesta, nos. 59, 61, 62). For Saint-Wandrille, Regesta, no. 261; RADN, no. 234. For Le Bec and Saint-Evroult, Regesta, nos. 167, 255. 159 Lucien Musset’s study supplies the foundations on which all study of these documents are based, AAC, 25–41. For a general survey, David Bates, ‘Les chartes de confirmation et les pancartes normandes du règne de Guillaume le Conquérant’, in Parisse, Pégeot, and Tock (eds.), Pancartes monastiques, 95–109.

414

william the conqueror

In several cases, the processes by which these documents were compiled out of existing records can be illustrated in some detail. The Fontenay charter, for example, passed through three recensions, with only the third one being confirmed by William; the Saint-Martin of Sées through four, with again only the last one being confirmed by William. The Fontenay text is unique in describing the processes that brought it to its final confirmation, namely, an inquiry conducted on oath at Caen on William’s orders by Richard vicomte of Avranches and Abbot William of Saint-Etienne of Caen, which required representatives of the tenants of the abbey’s estates to swear to its veracity.160 Possibly dating from either 1081 or 1086, a similar solemnity was present when Abbot Anselm of Le Bec tried to secure William’s confirmation in England of a pancarte for his abbey, something that the king refused to do until all the donors whose gifts were recorded in the charter were present, thereby causing a delay from Easter to Whitsun.161 If this process dates from 1086, then what happened could have been influenced by the rigours of the Domesday survey, There are, of course, parallels between the Fontenay procedure and that which later animated the Domesday survey in England. These similarities should not, however, be used to argue that the mechanisms were transferred from Normandy into England, but rather, as will be further argued in the next chapter, that the duchy and the kingdom were participants in a common western European documentary culture; for all its uniqueness, Domesday Book did not come out of nowhere. The crucial point about all this activity is the drive within Norman society to obtain William’s confirmation for the extensive transfer of property into monastic hands that had mostly taken place from the 1030s onwards and to obtain the stamp of his authority for it. This was not innovatory; the group of charters for Fécamp, Jumièges, Saint-Wandrille, and Bernay that dates from the final years of Richard II’s reign has an identical function.162 But it is evidence of the respect due to a long-established and effective regime, and of those who had profited from it consolidating their achievements in preparation for the possible uncertainties of the future. At some point, so Orderic tells us, William Pantulf, on returning to Normandy from Apulia at an unspecified later date that may well belong in this period, was accused of being complicit in the murder of Mabel de Bellême because she had also taken away one of his castles. Roger de Montgommery immediately seized his lands and he took refuge in the Regesta, no. 149. Sancti Anselmi Opera Omnia, i, ep. no. 118 (The Letters of Saint Anselm, trans. Frölich, i, no. 118). 162 RADN, nos. 34, 35, 36, 53. 160 161

cross-channel rule

415

abbey of Saint-Evroult, eventually agreeing to clear himself on the orders of William’s court by undertaking the ordeal by hot iron at Rouen, supervised by appropriate clergy. Orderic’s extremely vivid account tells of Roger and his sons, fully armed, standing around ready to inflict vengeance on William should he fail at the ordeal. Successfully carrying the hot iron bar without any injury, he was cleared and went on to continue a prosperous career, appearing in Domesday Book as a major tenant of Earl Roger.163 That William Pantulf was accorded a trial by due process can surely be seen as testimony to the order consequent on William’s rule in Normandy. William’s presence at Caen on 7 January 1080 suggests that he had spent Christmas there or in the near vicinity; the narrative that records his whereabouts says that the monks of Saint-Florent of Saumur deliberately set out from the Loire valley to seek an adjudication because they knew that they would find him there, implying, in other words, that the visit was a long one and that his stays at Caen were often predictably lengthy.164 The dispute itself, which concerned the status of the church of Saint-Gervase and Saint-Prothaise of Briouze in the southern border lands of Normandy, whose grant to Saint-Florent by William de Briouze was contested by the monks of the abbey of Lonlay, was tackled in a typically systematic way by William, with the final verdict being in favour of Saint-Florent. The detailed narratives of the plea show that it was decided by a significant group of churchmen, including the newly appointed archbishop of Rouen. This also indicates that a major gathering took place at the religious festival. A decision having been reached in Saint-Florent’s favour, William asked the monk of the abbey who was present to set before him a charter so that he could authorize it with his signum, presumably, that is, to draw his autograph cross on the parchment. The monk then admitted that he did not have the required charter, sped back to Briouze to collect one, presumably having it written during his time there, and then caught up with William again at Boscherville on the river Seine on 31 January where everything was formally confirmed in the abbey church. Illuminating for several reasons, the story above all shows William’s trust in a document produced by the beneficiary of the judgment, while, at the same time, indicating that care would have been taken to verify its contents.165 William de Briouze’s grant to the abbey of Saumur is also again indicative of how powerful OV, iii, 160–3. For William’s English lands, DP, 493–4. Regesta, no. 267(I) (ibique eum morari predictus abbas et monachi et Guillelmus de Braiosa audissent). 165 The narrative is set out in Regesta, nos. 267(I) and 267(II), with the second version, written in Robert Curthose’s reign after the monks of Lonlay had reopened the dispute, containing many of the arguments used. 163 164

416

william the conqueror

individuals in the south and west of Normandy continued proactively to develop links with monasteries on the Loire. The settlement was part of a process whereby a network connected Saint-Florent with William de Briouze’s lands in southern Normandy and his holdings in the Rape of Bramber in Sussex.166 From Caen, as we have seen, William moved into Upper Normandy, reaching Boscherville by 31 January. The itinerary has the appearance of being a perambulation of the duchy undertaken with the purpose of displaying publicly the reconciliation with his son; although no names are actually given, the confirmation at Boscherville is said to have been made by William, Matilda, and their sons.167 The grand occasion at Oissel on 12 April at which Gregory VII’s representatives were present indicates that Easter was spent at Rouen.168 William was then at Lillebonne for Whitsun (31 May), where a great council of the Norman Church was held whose extensive legislation has been preserved by Orderic, with a charter dated 2 June at Lillebonne confirming that Robert Curthose was also present.169 At the heart of the council’s legislation is the theme that runs through earlier legislation on both sides of the Channel, but which had never been set out in as much detail for us as at Lillebonne, namely, the availability of ducal officials to enforce the moral authority of the Church. Thus, for example, any difficulties encountered by bishops in enforcing the Truce of God were to be supplemented both by the offender’s lord and the duke’s vicomte; ducal officials were made available to assist in the examination of married priests; and the ducal court was made into a sort of supervisory tribunal to oversee the relationship between secular and ecclesiastical jurisdiction.170 A complex and truly remarkable diploma, the so-called pancarte of the abbey of Lessay shows that William then returned to Lower Normandy by placing him at Bonneville-sur-Touques on 14 July, again in the company of Matilda and their sons, with Robert also independently confirming the same grants to Lessay four days later at Caen.171 Although presenting many problems of interpretation that are further compounded by the earliest text of the diploma surviving only in a nineteenth-century facsimile of an original that was already damaged before its destruction in 1944, the Ibid., nos. 266(I), (II), (III). Ibid., no. 266(II) (confirmauit Willelmus rex Anglorum hanc meam elemosinam et Mahildis regina uxor sua eorumque filii) no. 267 (II) (confirmauit Willelmus rex Anglorum et Mahild(is) regina eorumque filii). 168 Ibid., no 235. 169 OV, iii, 24–35. For the grant, Regesta, no. 30, where the arguments for dating it to the council of Lillebonne are set out. 170 OV, iii, 26–7, 34–5. 171 Regesta, nos. 175(I), (II). On the interpretation of the three surviving versions of this text, van Torhoudt, Centralité et Marginalité, 953–9. 166 167

cross-channel rule

417

central problem is that the signa were written by more than one scribe and, in a small number of cases, were added around half a century later than the creation of the original. These complexities notwithstanding, the document is certainly evidence of the gathering of the duchy’s and the cross-Channel realm’s topmost elite. It also indicates a family gathering. Eudo Haldup, the son of Lessay’s founder Thurstan, had been married to Muriel, daughter of Herluin de Conteville and Herleva, probably since the 1040s; she and her two brothers, Bishop Odo and Count Robert of Mortain, the latter of whom had been endowed with extensive property interests in the Cotentin by William, are all among the signa. Also among the signa are the crosses and names of the two English archbishops, as also are those of Bishop Walchelin of Winchester and Bishop Osmund of Salisbury, albeit written in a different hand from the majority of the signa. The grouping together of the signa of Archbishop Thomas and Bishops Walchelin and Osmund on the left side of the parchment, all of them written by the same scribe, suggests that an insertion was made at a single time. Lanfranc’s, although positioned elsewhere, was also written by a different scribe from the majority of the others. If these four bishops had crossed the Channel from England, then there must have been extremely serious reasons for them to do so; bishops from the kingdom and the duchy normally did not cross from one to the other, and certainly not four of them as a group. THE WIDER WESTERN WORLD The involvement in the wider politics of Christendom that had fuelled the invitations to intervene in the Empire in 1074 intensified during the years that followed, driven by both the elevated sense of status that went with kingship and by the developing and increasingly bitter conflict between Gregory VII and Henry IV. In 1075 a partisan of Gregory, Bishop Hugh of Langres, prophesied in a poem that William would become emperor, albeit perhaps more to flatter than to express a realistic political aspiration because William had subsidized his pilgrimage to Jerusalem.172 And, in 1081, a letter of Gregory to his legates in France indicates that William had been approached by the pope’s enemies, meaning presumably supporters of Henry IV’s initiatives against Gregory from the synod of Brixen (25 June 1080) onwards and his military intervention in Italy.173 But, on the 172 Chronicon Sancti Huberti Andaginensis, ed. Ludwig C. Bethmann and Wilhelm Wattenbach, MGH Scriptores, viii, 565–630, at 577 (Si quis in ante uidet qui te circumspicit ex te/ colligit, ante comes, rex modo, Caesar erit). See further van Houts, ‘Latin Poetry’, 42; Cowdrey, Gregory VII, 345–6, 373, 379, 399, 412. 173 Registrum., ii, no. IX. 5 (Register, 405). For this period, Cowdrey, Gregory VII, 201–15.

418

william the conqueror

basis of material surveyed in the next chapter, it looks as if William did not reply. While all contact between William and Gregory – as far as we know, they never met – was conducted through intermediaries, usually the pope’s legates, at a personal level, it was contact redolent of well-informed interaction.174 During the 1070s William and Matilda must have worked hard and developed extensive contacts to assemble the collections of relics with which they endowed the two Caen abbeys.175 They also became extremely generous religious patrons on a scale that had only limited pre-1066 precedents. Although the specific dates of most of the donations are unknown, many must belong to the period covered by this chapter, with Matilda’s death in 1083 making it unlikely that many are later than 1081. Some of the grants continue patterns of patronage followed by William’s predecessors, who had, for example, been notable patrons of Marmoutier and elsewhere. That England’s earlier eleventh-century conqueror, Cnut, as well as his wife Emma, had also been religious patrons on a European scale may have supplied a model.176 But there were certainly other influences pushing the two of them in that direction. Beyond the frontiers of Normandy, England, and Maine, either singly or together, William and Matilda contributed a bell-tower to Chartres cathedral for the soul of their deceased daughter Adelida, a tower to the abbey of Saint-Denis, and a dormitory to Marmoutier.177 The Saint-Denis tower was so magnificent that its construction posed insuperable difficulties and it collapsed during William’s lifetime before it could be completed. Reflecting William’s integration into some aspects of the cross-Channel lands he ruled, it contained an altar dedicated to St Edmund.178 Matilda paid for the construction of a large refectory at Marmoutier.179 Gifts of 174 For an earlier version of the themes in this section, Bates, ‘William the Conqueror’s Wider Western European World’, 74–9. 175 AAC, 15–17, nos. 29, 30. 176 Lawson, Cnut, 133–8; Stafford, Queen Emma and Queen Edith, 146–9. 177 For Chartres: Obituaires de la province de Sens. ii. Diocèse de Chartres, Cartulaire de Notre-Dame de Chartres, iii, 218. For Saint-Denis: Guibert de Nogent, Autobiographie, 466–9 (A Monk’s Confession, trans. Archambault, 211–12). For Marmoutier, ‘Chronicon abbatum Majoris Monasterii’, in Recueil des chroniques de Touraine, ed. André Salmon (Tours, 1854), 318–37, at 318. On the whole subject, Lucien Musset, ‘Le mécenat des princes normands au XIe siècle’, in X. Barret I Altet (ed.), Artistes, artisans et production artistique au Moyen Age, 3 vols. (Paris, 1986–90), ii, 121–33, at 125–8. 178 For archaeological confirmation of the tower’s existence, which locates it at the northwest angle of the transept and nave of the eighth-century church that was replaced by Abbot Suger’s famous building, S. McK. Crosby, The Royal Abbey of Saint-Denis from its Beginnings to the Death of Suger, 475–1151, ed. Pamela Z. Blum (New Haven, CT, and London, 1987), 96–100. 179 ‘Chronicon abbatum Majoris Monasterii’, 318.

cross-channel rule

419

precious objects are known to have been made by William to Cluny and Marmoutier and by Matilda to Cluny, Marmoutier, Saint-Corneille of Compiègne, Saint-Florent of Saumur, and, perhaps, to La Chaise-Dieu in the Auvergne.180 Matilda also paid for the tomb in Rome of Simon of Amiens, Valois, and the Vexin, and gave valuable gifts to the church where he was buried.181 And William funded part of Bishop Hugh of Langres’s pilgrimage to Jerusalem.182 The couple’s reputation for generosity was a subject for positive comment in an obituary written in Anjou and in a letter that their son-in-law Stephen-Henry, count of Blois-Chartres, wrote to his wife Adela as he contemplated the wonders of Constantinople; he said the experience showed him riches more wondrous than those of William and Matilda, which he had previously thought unrivalled.183 A visit to William by Abbot Hugh of Cluny, although undated in the sources that mention it, probably took place in 1078, or perhaps during 1079 or in early 1080.184 On the basis of what most modern commentators are prepared to accept as a version of a narrative foundation history of the abbey of St Pancras, Lewes, inserted into an early thirteenth-century forged charter, one subject discussed was William de Warenne and his wife Gundrada’s proposal to establish a Cluniac abbey at Lewes (Sussex), the first in either England or Normandy.185 The consequences of Count 180 For William and Matilda’s gifts to Cluny and Marmoutier: Gilo, Vita sancti Hugonis abbatis, in ‘Memorials of Abbot Hugh of Cluny’, 64–5; Vita sancti Hugonis abbatis Cluniacensis, PL, clix, cols. 923–4; ‘Chronicon abbatum Majoris Monasterii’, 318. For Matilda’s gifts: Recueil des actes de Philippe Ier, no. CXXVI; Paul Marchegay, ‘Chartes angevines des XIe et XIIe siècles’, Bibliothèque de l’Ecole des Chartes, xxxvi (1875), 381–441, at 413. For La ChaiseDieu, George T. Beech, ‘Queen Mathilda of England (1066–1083) and the Abbey of La Chaise-Dieu in the Auvergne’, Frühmittelalteriche Studien, xxvii (1993), 350–74; but see Baxter, Earls of Mercia, 300, for the intriguing suggestion that this last grant was made by Queen Ealdgyth, widow of King Harold. 181 Vita beati Simonis, PL, col. 1222 (trans. van Houts, Normans in Europe, 199). 182 Chronicon Sancti Huberti Andaginensis, 577. 183 Annales qui dicuntur Rainaldi, in Recueil d’annales angevines et vendômoises, 88–9; Die Kreuzzugsbriefe aus dem Jahren 1088–1100, ed. Heinrich Hagenmeyer (Innsbruck, 1901), 139 (ut nobis uidetur, non fuit princeps uniuersa morum honestate adeo praeclarus, pater, mi dilecta, tuus multa et magna dedit, sed ad hunc paene nihil fuit); Lo Prete, Adela of Blois, 448, Appendix 1, no. 14. 184 Vita sancti Hugonis abbatis Cluniacensis, PL, clix, col. 923; Gilo, Vita sancti Hugonis abbatis, in ‘Memorials of Abbot Hugh of Cluny’, 64–5. 185 In general, Brian Golding, ‘The Coming of the Cluniacs’, ANS, iii (1981), 65–77, 208–12, at 65, 71–2. For the charter, Monasticon Anglicanum, ed. William Dugdale; new edition by J. Caley, H. Ellis, and B. Bandinel, 6 vols. (London, 1817–30), v, 12–13; Recueil des chartes de l’abbaye de Cluny, ed. Auguste Bernard and Alexandre Bruel, 6 vols. (Paris, 1876–1903), no. 3651. For the date of the forgery, Early Yorkshire Charters, volume viii: The Honour of Warenne, ed. C. T. Clay, Yorkshire Archaeological Society Record Series, ed. C. T. Clay (Wakefield, 1949), viii, 59–62. For Abbot Hugh’s visit, Barlow, ‘William I’s Relations with Cluny’, 137–41 (idem, The Norman Conquest and Beyond, 249–52); Cowdrey, ‘William I’s Relations with Cluny Further Considered’, 2, 13–14.

420

william the conqueror

Simon’s withdrawal from the world and the quarrel with Robert Curthose could have been other topics. But the likelihood that Matilda had an interest in building links with Cluny, which were associated with the Warenne initiative, created an impetus that would have made Abbot Hugh as good as obliged to visit Normandy. With Gundrada’s brother Gerbod – the accidental killer of Matilda’s nephew Count Arnulf of Flanders – by this point well established as a monk at Cluny, Matilda may well have wanted to help Gundrada, who may have been her kinswoman, to assist the family’s atonement for the terrible deed.186 Having visited Cluny in 1076 on a journey to Rome, William and Gundrada’s plans were apparently floundering until Hugh’s visit put everything back and the more conventional alternative of establishing a priory of Marmoutier was abandoned. Informed of the election of William Bona Anima to the archbishopric of Rouen, Gregory VII wrote to his legate in northern France, the subdeacon Hubert, on 23 September 1079 to say that he had heard a story that William was the son of a priest; as indeed he was – he was the son of Radbod, who had been Ivo of Bellême’s predecessor as bishop of Sées in the 1020s and 1030s.187 Having indicated to Hubert that there was no way in which he would agree to this promotion, the pope subsequently became more flexible, writing to the new archbishop in 1081 to censure him severely for not coming to Rome to collect his pallium, and his suffragans for not attending papal synods. He thanked the archbishop for a letter – no longer in existence – that he described as well written, but entirely lacking in either sincerity or substance. His threat of sanctions was, however, only of a general kind.188 As far as we know, Archbishop William continued in post untroubled. Similar criticisms about deliberate absence from Rome were also directed at Lanfranc on 25 March 1079. In pointing out that Lanfranc had not visited Rome at any time during his pontificate, Gregory was much more expansive on the subject than he had been to Archbishop William, suggesting that the cause of the absences might have been William himself and that Lanfranc’s behaviour was determined by fear of the king.189 In a letter to his legate the sub-deacon Hubert written approximately six months later, Gregory explicitly says that William had prohibited bishops from visiting Rome and that his behaviour was worse than that of pagan kings. At the same time, however, he told Hubert that the apparently confrontational approach of another of his legates, who had supposedly For this suggestion, van Houts, The Warenne (Hyde) Chronicle, 96–102. Registrum, ii, no. VII. 1 (Register, 324–5). 188 Ibid., no. IX. 1 (Register, 398). 189 Ibid., no. VI. 30 (Register, 312–13). 186 187

cross-channel rule

421

criticized William publicly, was not his own.190 In parallel to all this, Gregory’s attempt in 1079 to resuscitate a primacy of the archbishop of Lyons over the archbishoprics of Rouen, Tours, and Sens does not seem to have had any actual impact in Normandy.191 And in 1080 he wrote two succinct letters to Bishop Arnold of Le Mans and William to say that Arnold’s suspension had been rescinded, and that Abbot Rainald of Saint-Pierre de la Couture had been deposed and Abbot Juhel reinstated, thereby overturning without explanation decisions taken by his legates.192 So much was this whole situation a consequence of the violent intrusion of the secular power into ecclesiastical affairs that Archbishop Ralph of Tours and Bishop Arnold had earlier tried to prevent the issue coming before Gregory.193 In Gregory’s letters, William was, among other things, ‘that most dear and special son of the holy Roman Church’, ‘a beloved son’, and, later on in his pontificate, when Gregory wanted to be critical, ‘a good deal worthier of approval and more to be honoured than other kings’.194 Elsewhere he could write that ‘we embrace you amongst kings with a special love as well as for the good character and virtue . . .’ and ‘I have noticed you to be devoted to the apostolic see beyond others of your order’.195 Matilda was seen as sharing in the same qualities and was often referred to as a ‘beloved’ or ‘dearest daughter’ and encouraged to nurture further her husband’s religious life.196 She was, in short, regarded as William’s partner in the pursuit of just rule and salvation after the fashion of Carolingian and Ottonian queens. For all this, the change that had occurred since Alexander II’s time, when bishops and abbots from Normandy and England had regularly visited the pope, was a considerable one. Alongside the fact that Gregory and Henry IV had effectively been at war since 1075, the sheer intensity and attention to detail that Gregory brought to the papal office can reasonably be seen as causing William to put up the shutters and to instruct the clergy in the lands he ruled to keep a respectful distance. For all that his treatment of the Church in Normandy before and after 1066 had always been authoritarian, it is very likely that Eadmer’s famous statement about the restrictive control that William had exercised over relations between the English Church and kingdom refers to the last years of his reign in England

Ibid., no. VII. 1 (Register, 324). Ibid., no. VI. 35 (Register, 317–18). 192 Ibid., nos. VII, 22 and 23 (Register, 352–5). 193 RHF, xiv, 667–9, nos. 1–3; Cowdrey, Gregory VII, 391–3. 194 Registrum, i, nos. I. 31, I. 70; ii, no. IX. 5 (Register, 37, 73, 405–6). 195 Registrum, ii, nos. V. 19, IX. 37 (Register, 268, 441–2). 196 Registrum, i, no. I. 71; ii, VII. 26 (Register, 75, 357–8). 190 191

422

william the conqueror

rather than to all of it.197 And the weighty matter that probably explains the presence of four English bishops in Normandy was Gregory VII’s proposal that William do fealty to him. The letter that answered it indicates that replying to it had been left until after William’s crossing to England, which must have occurred soon after the gatherings at Bonneville-sur-Touques and Caen.198

197 198

HN, 9–10. Lanfranc Letters, no. 39 (Nunc uero diuina misericordia me in regnum meum reuerso).

Chapter 11

POL I T I C A L C O N S O LIDAT IO N A N D P E R S O NAL LO SS Although approaching his mid-fifties at the start of the 1080s, and certainly very corpulent in the later years of his life, William does not appear to have suffered any significant associated physical decline.1 As far as we know, he remained as prodigiously active as he had always been. His death on 9 September 1087 followed on from a sudden collapse, rather than at the end of a gradual enfeeblement. The period from the last months of 1080 until Matilda’s death on 2 November 1083 was to be a time of familial and political consolidation on both sides of the Channel, with the quarrel with Robert Curthose seemingly resolved and peace treaties made with King Malcolm (Mael Coluim) and Count Fulk Rechin which remained in place for the rest of William’s life. Visits to England were more frequent during these years, although never especially long. They did, however, all have a very specific and necessary purpose. These achievements notwithstanding, some new cracks appeared and there was a lot of underlying uncertainty. Late in the year 1082 William quarrelled with his half-brother Odo and imprisoned him. In terms of the hidden histories of these times, although we know nothing directly about them, there must surely have been speculation about what arrangements William was going to make for the succession to Normandy, Maine, and England. Direct contact with Gregory VII became less regular as the pope was increasingly assailed and undermined by Henry IV. Again, a relationship that had reinforced William’s rule since the time of the Council of Rheims of 1049 began to slip away. Finally, as we shall see, William of Malmesbury with his interest in biography did believe that Matilda’s death made a difference. We might indeed believe that William was somewhat diminished thereafter. For that reason her death is adopted as the dividing line between the two chapters that deal with the last seven years of William’s life. GREGORY VII’S REQUEST FOR FEALTY A Marmoutier notice recounting the journey of the monks’ representative across the Channel to William in England, only for him to be dispatched 1

For William’s corpulence, OV, iv, 78–9; GR, i, 508–9, 510–11.

423

424

william the conqueror

back to Normandy with orders to seek out Matilda so that she could hear and settle the dispute there, must indicate that William remained in the south of England in the final months of 1080 while his half-brother Odo and his son Robert were sent separately with armies to the north of England and to Scotland. The dispute’s resolution before Matilda and others at Cherbourg on 27 December 1080 suggests that the monk made two fairly rapid transfretations.2 A continuation of the discussions started in Normandy of Gregory VII’s request that William do fealty to him must have taken up a good amount of William’s time. The famous letter written on William’s behalf to refuse Gregory’s request that he do fealty indicates that the pope’s messenger had been the subdeacon Hubert who had played an important role in the Canterbury–York settlement of 1072, and who had been sent to Normandy with the three separate papal letters dated 8 May 1080 at Rome to William, Matilda, and Robert Curthose.3 The care taken to send such a familiar and noninflammatory personage is a sign that the request for fealty was not an attempt to impose authority, but rather to draw William into a closer relationship; its context an endeavour to increase the number of papal fideles and the events of the year 1080 after Gregory’s second excommunication of Henry IV.4 The actual request was in all likelihood made orally, or perhaps in a letter whose text has been lost, since neither the preparatory overtures evident in the letter Gregory sent to William dated 24 April 1080 nor the letter addressed to William dated 8 May are fully explicit about it. The first of these letters contains statements like ‘the more the powerful of this world being blinded by their pride and ungodly deeds rush into the abyss, the more it beseems you, who have been found very dear to God before the rest, by dutifully humbling yourself to be raised up, and by obeying to be exalted’, while the second is a general exhortation to Christian kingship.5 William’s refusal was made on the grounds that the request was unprecedented and therefore unacceptable. Written in crisp, but amicable, terms, the letter acknowledges that Gregory’s request for Peter’s Pence was just and promises to do better now that William has crossed to England after his long absence. In a separate letter to Gregory, Lanfranc wrote that he had done his best to persuade the king, suggesting that the discussions had been long and far from one-sided.6 If, as Lanfranc suggests, William had been the great 2 Regesta, no. 200 (pro hac re mare transgrediens, clamorem fecit regi, qui iratus remisit predictum monachum . . . regine). 3 Lanfranc Letters, no. 39. The mission is mentioned in Gregory’s letter to Matilda, Registrum, ii, no. VII. 26 (Register, 358). 4 Cowdrey, Gregory VII, 646–7. 5 Registrum, ii, 23, no. VII (Register, 354). 6 Lanfranc Letters, no. 38.

political consolidation and personal loss

425

obstacle, then it is likely that his clear sense that his dignity as king of the English meant no fealty was owed to the pope was the crucial influence on the decision. The refusal notwithstanding, relations between William and Gregory remained cordial and respectful, while clearly being the subject of debate among those around the pope. In 1081, for example, Gregory had to write to his legates in France to remind them that, while William might not be perfect, he had many virtues and should be treated more gently than other kings. In the same letter he added that the legates’ suspension of the entire Norman episcopate, with the exception of the archbishop of Rouen, for failing to attend a papal council, should be rescinded; it was William’s prohibition, rather than the bishops’ reluctance, which explained their absence.7 Such pragmatism also fits with Gregory’s general approach to the French kingdom that has recently been characterized as ‘moderation in practice’.8 His statement that it was William who was determining policy is nonetheless very important. For all the continued good relations, Lanfranc did attract some sharp words from Gregory, with the letter he wrote to the pope in the summer of 1080 responding once more to criticism for not having visited Rome by firmly pointing out that affection and loyalty did not require his personal presence. In adding that if there had been a falling-away from earlier standards, it was Gregory, and not Lanfranc, who was at fault, he came close to being very rude.9 It is surely likely that all these exchanges should not be taken at face value; Gregory may well have attempted to discipline Lanfranc to work round William and also to have to demonstrate that he had the support of a very distinguished churchman by using threats to force him to visit Rome. Not to be drawn in, Lanfranc’s legalistic response sidestepped any issues of conflicting loyalties that might have arisen and showed where his own loyalties ultimately lay. All this is a continuation of the defensiveness towards Gregory and his legates that emerged in the later 1070s. After 1080–1, relations with Gregory were to be increasingly affected by Henry IV’s attacks on the pope and his direct interventions at Rome. STABILIZING THE BRITISH ISLES (1080–1) According to Symeon of Durham, the army led by Odo of Bayeux devastated the countryside between the Tees and the Tyne in the summer of 1080, indiscriminately killing innocent and guilty alike. Those who had trusted in their innocence were beheaded or had their limbs amputated as if they had been participants in the murder. Others, although innocent, Registrum, ii, no. IX. 5 (Register, 405–6). Cowdrey, Gregory VII, 398–422, with the quotation at p. 420. 9 Lanfranc Letters, no. 38; Cowdrey, Lanfranc, 197–202.

7 8

426

william the conqueror

were forced to make money payments to protect themselves. In addition, Odo removed a valuable staff and other ornaments from the cathedral treasury.10 The area chosen for ravaging shows that its purpose – at least in theory – was to punish those held responsible for Walcher’s death. Robert Curthose later, probably in the autumn, led another army further north with the deliberate intention of confronting the King of Scots. As in 1072, Malcolm avoided taking on a potentially superior enemy in battle and, when Robert entered the kingdom of the Scots proper near Falkirk, terms similar to the earlier Peace of Abernethy were reaffirmed and Malcolm did homage to Robert. The pattern of earlier times of savage statesponsored violence inflicted as a punishment for disobedience within the kingdom of the English and of negotiated settlement with a neighbouring power is of course repeated here. The responsibility entrusted to Robert makes it look as if the agreement between him and his father, which involved Robert having a role in England as well as Normandy, was being implemented, something that is all the more remarkable because, as far as we know, this was his first ever visit to England. According to the Durham Historia Regum, Robert established a castle at Newcastle-on-Tyne as he returned south. He was back with his father at Salisbury by 2 February 1081.11 Although the author of the Historia Regum thought Robert’s expedition had achieved little, the writer of the twelfth-century Abingdon Historia, presumably drawing on the memory of Abbot Adelelm who had been with the army, described it as a success, adding that William was pleased with Robert’s efforts.12 The typically pessimistic Durham assessment was almost certainly grounded in a local long-term perspective on the always fraught relations with the kings of Scots, to which the campaign did not produce a long-term solution, if indeed one were possible. Robert had probably accomplished all that could reasonably have been expected of him. His campaign also established spiritual bonds between the two royal families when he became godfather to Malcolm and Margaret’s fifth child, their first daughter Edith/Matilda, who was later to become the wife of Roberts’ younger brother King Henry I. According to a letter written in the 1140s by Gilbert Foliot, at the time abbot of St Peter’s, Gloucester, relations were further consolidated during 1081 when Matilda became Edith/Matilda’s godmother.13 Conventional network-building practices had reinforced a relationship that was shaped LDE, 220–1; Aird, St Cuthbert and the Normans, 100–1. HRA, 211; Regesta, no. 154. On Robert’s campaign, Aird, Robert Curthose, 91–3. 12 HRA, 211; Historia Ecclesie Abbendonensis, ii, 14–15. 13 The Letters and Charters of Gilbert Foliot, ed. Z. N. Brooke, Dom. Adrian Morey, and C. N. L. Brooke (Cambridge, 1967), 66 (no. 26); Lois L. Huneycutt, Matilda of Scotland: A Study in Medieval Queenship (Woodbridge, 2003), 9–10. 10 11

political consolidation and personal loss

427

mostly by both of them needing to concentrate on problems elsewhere. As far as we can see, consensual good relations were re-established that lasted for the rest of William’s life.14 Replacement appointments were also made to the earldom of Northumbria and the bishopric of Durham, with Aubrey de Coucy-le-Château-Auffrique being named as earl and William (usually known as William of Saint-Calais), abbot of Saint-Vincent of Le Mans, as bishop on 9 November 1080. Although the date of Aubrey’s appointment is unknown, he was presumably installed as soon as was possible, in all probability by either Odo or Robert. His origins in southern Picardy and the distribution of his English lands across nine English counties, of which, with the exception of two estates in south Yorkshire, the most northerly were in Leicestershire, suggest that he was yet another example of someone being appointed who was ready to volunteer to take on a very risky job; self-evidently he lacked both a personal base and any previous influence in the far north of England.15 Although at one point supported by Odo of Bayeux’s patronage, Bishop William’s origins and prime loyalties seem to have lain in Maine, since his father had also been a monk of Saint-Calais before him. For a short time abbot of Saint-Vincent of Le Mans, again according to Symeon, he commended himself to William because he had already shown an aptitude for dealing with difficult business.16 Bishop William’s career could be seen in terms of the recruitment of a Manceau into the Anglo-Norman elite. But, as a celibate monk, he brought little or nothing in the way of social networks to the cross-Channel state; the promotion was probably therefore in recognition of his abilities rather than a piece of imperial bridge-building. On the basis of very limited information, Aubrey’s career seems to have been uneventful, and perhaps undistinguished. Bishop William’s, about which we are well informed, although somewhat up and down, was to be an outstanding success. The narrative of the Marmoutier monk’s audience with William while his brother and son were campaigning in the north includes the story of one of William’s more anodyne displays of ritualized anger, although whether it was directed against the monk for crossing the Channel when he could just have gone to Matilda in Normandy or against the vicomte who

14 For this view of the relationship between the kings of the English and the Scots, Marritt, ‘Coincidences of Names’, 157–9, 163–5; Duncan, The Kingship of the Scots, 45–6. 15 For Aubrey, see now Aird, St Cuthbert and the Normans, 101–2; idem, ‘Mowbray, Robert de, Earl of Northumbria (d.1115/1125)’, ODNB. 16 LDE, 222–5, with the evidence for his appointment and pre-Durham career surveyed at p. 222, notes 2 and 3. See also Aird, ‘An Absent Friend’, 287–9; Frank Barlow, ‘St Calais [St Carilef], William of (c.1030–1096)’, ODNB.

428

william the conqueror

had made the encroachment is not entirely clear.17 The monk’s apparent confusion as to whom he should be approaching again demonstrates the centrality of William’s family relationships to the formal exercise of power, while at the same time again indicating that no one could entirely predict how the partnerships that operated around William’s ultimate personal authority were going to function. A unique personal letter describing a grant to Marmoutier and freeing the abbey from a specific tax (grauarium) and forest dues from William to Matilda, which probably dates from this time and resembles, but does not slavishly imitate, writ-form, addresses Matilda as ‘beloved queen and wife’ (. . . Mathildi regine dilectarie coniugi perpetuam salutem).18 But the record of her presidency of the court that decided the case for whose prosecution the monk had crossed the Channel nonetheless says that she acted in obedience to the king (regi obediens).19 William spent Christmas 1080 at Gloucester, staying on into early January 1081.20 The record of the consecration there of Bishop William of Durham in the Acta Lanfranci, albeit composed from a Canterbury perspective, illuminates the successful working of the compromises involved in operating the Primacy Agreement of 1072. Since Durham was York’s only suffragan within the kingdom of the English and assistance from York’s suffragans in the kingdom of Scots was unobtainable, Lanfranc and his suffragans helped out.21 Also, despite the famous statement in the ‘E’ version of the Chronicle that William was accustomed to spend Christmas at Gloucester, this is the first occasion on which he can actually be shown to have been there at that time of year; the only known previous visit to the vicinity of the town had been in the late summer of 1069, when he was hunting in the Forest of Dean and news reached him of the Danish fleet’s arrival. This extension of William’s formal itinerary to Gloucester is likely to have been another manifestation of the way in which he deliberately structured his rule in England. Edward the Confessor had sometimes held court at Gloucester and facilities there appropriate for royal visits had been significantly developed by Ealdred during his years as bishop of Worcester and archbishop of York, so William can yet again be seen as consciously setting out to emulate Edward. Further influences must have been that Gloucester was a thriving town and the revival of the abbey of St Peter’s under Abbot Serlo, appointed in 1072 and previously a monk of Mont Saint-Michel. William of Malmesbury, undoubtedly exaggerating, 17 Regesta, no. 200 (pro hac re mare transgrediens, clamorem fecit regi, qui iratus remisit predictum monachum . . . regine). 18 Ibid., no. 202. 19 Ibid., no. 200. 20 LDE, 222–3; CS, 629–31. 21 The Anglo-Saxon Chronicle. A Collaborative Edition: 3. ms. A, 86; CS, 629–31.

political consolidation and personal loss

429

describes an impoverished community reduced to three monks before Serlo’s arrival. His achievement in attracting patronage – presumably with William’s encouragement – was nonetheless remarkable.22 After this visit to Gloucester, William travelled eastwards, being at Salisbury on 2 and 4 February and in London before the end of the month.23 He was back at Winchester for Whitsun (31 May), presumably after that moving westwards into south and west Wales.24 This itinerary indicates that the campaign in Wales in the summer was a response to a crisis, and not a planned follow-on from the Christmas visit to Gloucester. The charters of this period also illustrate the increasing diversity of business that was coming before William as the new society of the Conquest evolved; the charters at Salisbury confirmed grants by Arnulf de Hesdin, a wealthy Domesday landholder originally from the western area of the county of Flanders, to St Peter’s, Gloucester, and by the Breton Wihenoc of some of his holdings at Monmouth (Gwent) to the abbey of SaintFlorent of Saumur. The first charter illuminates the developing engagement of the newcomers with English religious institutions and, specifically with the development of Gloucester as a place for crown-wearings. The second charter extended further new networks that linked the British Isles to north-western France, while also reinforcing traditional ones between Brittany and Anjou.25 The diploma confirmed in London described a gift of an estate that Matilda had made to the abbey of Malmesbury. Written in a format that resembles a pre-1066 English diploma, albeit entirely in Latin, it mentions William’s rule over peoples who were neighbours of the English, and also describes Matilda as William’s legal wife (legalis coniunx), a piece of phrasing that suggests that someone still thought it necessary to affirm the legitimacy of the marriage. The visit to Winchester for the Whitsun court included a great set-piece occasion at which Abbot Baldwin of Bury St Edmunds quashed the ambition of Herfast, bishop of the East Anglian see at that time located at Thetford, to move his cathedral church to Bury. Surviving in its earliest 22 GP, i, 446–7; Emma Cownie, Religious Patronage in Anglo-Norman England, 1066–1135, Royal Historical Society Studies in History (Woodbridge, 1998), 56–7, 61–2; Michael Hare, ‘Kings, Crowns, and Festivals: The Origins of Gloucester as a Royal Ceremonial Centre’, Transactions of the Bristol and Gloucestershire Archaeological Society, cxv (1997), 41–78, at 52–65; Nigel Baker and Richard Holt, Urban Growth and the Medieval Church: Gloucester and Worcester (Aldershot, 2004), 24–5. 23 Regesta, nos. 154, 268. For the London visit, ibid., no. 193. 24 Ibid., nos. 39, 255. 25 For Wihenoc’s grants and the associated networks, Jane Martindale, ‘Monasteries and Castles: The Priories of St-Florent de Saumur in England after 1066’, in Carola Hicks (ed.), England in the Eleventh Century: Proceedings of the 1990 Harlaxton Symposium (Stamford, 1992), 135–56, at 145–8; Crouch, ‘The Transformation of Medieval Gwent’, 12, 16.

430

william the conqueror

incomplete form in Latin and English, the diploma that describes the occasion portrays a great gathering of major churchmen and magnates. As had by this time become standard, Baldwin’s victory depended on citing pre1066 precedents and authorities, one of which was reliant on a recently forged charter, albeit one that did make use of authentic information.26 The fact that no use that we know of was made of the papal confirmation that Baldwin had obtained in 1071, and which would seemingly have been extremely useful because it supported Bury’s case against the bishop, is arguably another sign of the cross-Channel ecclesiastical elite reacting against Gregory VII’s assertiveness; royal charters were deemed preferable to papal authority as evidence in a major ecclesiastical issue. As a probable indicator of change, Baldwin’s choice is all the more significant because of his known political astuteness and because he would have been well aware of the extensive forging of papal confirmations at his former abbey of Saint-Denis.27 Another diploma in all likelihood forged at the same time is the first full-scale confirmation to survive of the lands that a specific Norman abbey had acquired in England, in this case, the abbey of SaintEvroult, at which Orderic was to become an oblate in 1085.28 The culture of pancartes and confirmation charters so prominent in Normandy at this very time was being extended to cover these monasteries’ English estates, another sign of how conquest and the cross-Channel empire that it had created were influencing evolving routines of governance and rule. William’s journey to St David’s in the summer of 1081 must have been a response to the outcome of the recent battle that had taken place at Mynydd Carn, an unidentified site somewhere to the north of St David’s. The result having radically changed the political landscape of southern and western Wales – above all to the benefit of Rhys ap Tewdwr, who had emerged as king in Deheubarth in 1079, and who after Mynydd Carn became the dominant power in west and, to an extent, south Wales – probably made an intervention by William a necessity. The battle also created something of a power vacuum to the west of the lands held by William’s followers in Herefordshire and Gwent, since Caradog ap Gruffudd, who had previously dominated much of Gwent and Morgannw ˆ g, was also killed in the battle. Also slain was Trahaern ap Caradog, whose death enabled his rival, Rhys’s ally Gruffudd ap Cynan, who had crossed from Ireland to establish himself as king of Gwynedd in 1075, finally to

26 Regesta, no. 39. For the Cnut diploma, Facsimiles of Anglo-Saxon Charters, ed. Simon Keynes (Oxford, 1991), no. 33. 27 Bates, ‘The Abbey and the Norman Conquest’, 16–18. 28 Regesta, no. 255.

political consolidation and personal loss

431

consolidate his power in North Wales.29 This radical redistribution of power across almost the whole of Wales drastically distorted both political relationships within Wales and the new ones between Welsh rulers and the newcomers that had developed since 1066. In addition to the advances into south Wales accomplished by William fitz Osbern and his son, there had also been raids launched from the earldoms of Shrewsbury and Chester that had resulted in territorial acquisition and destruction. But collaborative relationships had also formed, with these perhaps becoming so entwined that the release of hundreds of prisoners mentioned in the ‘E’ Chronicle as a result of William’s march to St David’s may well indicate that Normans and French had fought at Mynydd Carn on the losing side.30 William’s intervention has much of the character of a progress whose purpose was to stabilize relationships between the Welsh and the newcomers and at the same time to pay respect to Welsh saints. There is nothing to suggest that any fighting took place. Alongside the ‘E’ Chronicle’s mention of the release of prisoners, both versions of the Brut y Tywysogyon describe it as a pilgrimage; while undoubtedly a simplification, this illustrates William once more showing deference to a prestigious saint, a phenomenon that we have already observed in the cases of St Edmund, St Cuthbert, and St Ætheldreda.31 The march to St David’s does, however, mean that William had travelled much more deeply into Wales than any of his predecessors had ever done. The great irony – again – is that the only comparable recent penetration had been Harold and Tostig’s in 1063. The Wales that William and his successors encountered had been drastically reshaped by their extremely violent campaign. The final outcome of William’s intervention is likely to have been the agreement whereby Rhys paid William the tribute of £40 that is recorded in Domesday Book, an arrangement that recognized the power and status of both parties, and stabilized their

29 J. E. Lloyd, rev. David E. Thornton, ‘Rhys ap Tewdwr (d.1093)’, ODNB. For Gruffudd ap Cynan, Vita Griffini filii Conani: The Medieval Life of Gruffudd ap Cynan, ed. Paul Russell (Cardiff, 2005), 60–73; C. P. Lewis, ‘Gruffudd ap Cynan and the Normans’, in K. L. Maund (ed.), Gruffudd ap Cynan: A Collaborative Biography (Woodbridge, 1996), 61–77, at 68–70; Huw Pryce, ‘Gruffudd ap Cynan (1054/5–1137)’, ODNB. 30 In general on Mynydd Carn’s place in the history of Wales, R. R. Davies, The Age of Conquest: Wales, 1063–1415 (Oxford, 1991), 33–4; A. G. Williams, ‘Norman Lordship in South-East Wales during the Reign of William I’, Welsh History Review, xvi (1993), 445–66, at 452–66. 31 ASC, ‘E’, 1081; Brut y Tywysogyon or the Chronicle of the Princes, Peniarth MS 20 Version, trans. Thomas Jones (Cardiff, 1952), 17; Brut y Tywysogyon or the Chronicle of the Princes, Red Book of Hergest Version, trans. Thomas Jones (Cardiff, 1955), 31. In terms of establishing the chronology of the year, it is important that William’s journey to St David’s is placed after Mynydd Carn in both versions of the Brut.

432

william the conqueror

relationship.32 Another result of the expedition was the construction of a castle at Cardiff, as so often within the walls of a ruined Roman fort. The discovery of coins minted there in William’s name further indicates an extension of influence and, to an extent, territorial control.33 The followup in North Wales to William’s expedition was different, with Earl Hugh of Chester luring Gruffudd into a trap and then holding him prisoner for approximately twelve years, and Robert of Rhuddlan and Hugh taking over much of North Wales. Robert’s status as controller of a land beyond the kingdom of the English was, however, also accompanied by him paying a tribute of £40 to William, with Robert presumably extracting this considerable sum from the Welsh who had fallen under his control.34 Although conceptually different, since tribute, and therefore clientage, was involved, the march into Wales had something of the character of a tidying-up operation similar to the one that Robert Curthose had accomplished in Scotland. As with Scotland, the progress to St David’s seems to have created a stability that endured for the rest of William’s life. In the context of the longer-term history of Wales and all of the lands around the Irish Sea, both re-established the hegemonic superiority across the British Isles that William believed his pre-1066 predecessors to have held. So viewed, it explains the ‘E’ Chronicle’s apparently extraordinary statement that, having subdued Wales and Scotland, William would, had he not died, have done the same to Ireland within two years by peaceful means.35 After the early period when Harold’s sons had made their landings in western England in 1068 and 1069 supported by King Diarmait of Leinster, more amicable relationships had been established between Irish kings and the new rulers of England, constructed for the most part around Lanfranc consecrating Irish bishops and trading links that brought coins minted at Dublin into stylistic and weight ratio synchronicity with English issues.36 According to the Vita Wlstani, these commercial links involved among other things the sale of slaves to Ireland through Bristol and frequently their sexual abuse, practices against which Wulfstan, maintaining a tradition associated with his eponymous predecessor Archbishop Wulfstan of 32 DB, i, fo. 179r; Robert S. Babcock, ‘Rhys ap Tewdwr, King of Deheubarth’, ANS, xvi (1994), 21–35, at 26–7. 33 Williams, ‘Norman Lordship’, 457–61; Crouch, ‘The Transformation of Medieval Gwent’, 12. 34 DB, i, fo. 269r. 35 ASC, ‘E’, 1086 (recte 1087). 36 Benjamin T. Hudson, ‘William the Conqueror and Ireland’, Irish Historical Studies, xxix (1994–5), 145–58, at 149–55; Mark Philpott, ‘Some Interactions between the English and Irish Churches’, ANS, xx (1998), 187–204, at 189–200; Clare Downham, ‘England and the Irish-Sea Zone in the Eleventh Century’, ANS, xxvi (2004), 55–73, at 69–71.

political consolidation and personal loss

433

York, had preached.37 William is said to have deplored the trade and, according to the legal text known as the Willelmi Articuli, to have forbidden it.38 But, according to William of Malmesbury in the Gesta Regum, Lanfranc had to work hard to persuade him to do this because he received a share of the profits of the trade through taxation.39 In legislating at all, he was theoretically bringing England into line with Normandy, since Rouen had ceased to be a slave-trading port in the early years of the eleventh century.40 According to the Vita Wlstani again, fear of William’s wrath had no impact on those who practised the trade, with a probable reference in Domesday Book to the sale of slaves at Lewes (Sussex) indicating that they were still being exported through many ports.41 It looks as if William did espouse a humanitarian cause, but did not make a great deal of effort to act on it. There are parallels here with his reaction to the riots against the Norman legislation prohibiting priests from marrying, namely, a preference for not disturbing a status quo when serious disruption was likely to result if he did so. For all that compromises were made, in terms of the British Isles, these settlements gave a recognition and support to those who exercised local, regional, and national dominance around them. William’s personal presence in Scotland in 1072 and in Wales in 1081, and Robert’s in Scotland in 1080, provided a necessary personal authenticating guarantee. After William’s return to Normandy, in all probability during the autumn of 1081, in a reversal of the normal pattern Matilda was left behind in England. In addition to a record of her doing justice at Windsor in William’s absence, Domesday Book includes a reference to her presiding at a gathering of four shires that found in favour of the bishopric of Worcester’s rights at Alveston (Warwickshire), which may well date from this time; if so, it would have involved her in a visit to the West Midlands.42 A writ addressed to Lanfranc, Count Robert of Mortain, and Bishop Geoffrey of Coutances, ordering them to reassemble the representatives of the shires who had been present at a plea concerning the lands of the abbey of Ely that had taken place before Matilda’s most recent crossing to

37 Vita Wlstani, 100–3; David A. E. Pelteret, Slavery in Early Medieval England from the Reign of Alfred until the Twelfth Century (Woodbridge, 1995), 76–8; Wyatt, Slaves and Warriors, 149–50, 219–20. 38 Ten articles of William I, in Die Gesetze der Angelsachsen, ed. Liebermann, i, 486–8, art. 9; Hudson, The Oxford History of the Laws of England, 212, 424–5. 39 GR, i, 496–9. 40 Lucien Musset, ‘La Seine normande et le commerce maritime du IIIe au XIe siècle’, Revue de la société savante de la Haute-Normandie, no. 53 (1970, for 1969), 3–14, at 9–11. 41 DB, i, fo, 26r. 42 Historia Ecclesie Abbendonensis, ii, 14–15; DB, i, fo. 238v; Ann Williams, ‘A Vice-Comital Family in pre-Conquest Warwickshire’, ANS, xi (1989), 279–95, at 282–3.

434

william the conqueror

Normandy, suggests that she may have been involved in that too.43 In both cases, if they do date to this period, she was continuing major inquiries into landholding that had started during the time that Bishop Odo was acting in England on William’s behalf in the later 1070s. If the story in Gilbert Foliot’s letter of the 1140s to Brian fitz Count, that King Malcolm and Queen Margaret’s baby daughter Edith had tugged at her godmother’s veil and pulled it over her own head, has any veracity, then either the Scottish royal family travelled south or Matilda went north.44 Given its context in a letter setting out the claims of the Empress Matilda against those of her cousin King Stephen, the tale is one of those portents that might have been invented, just like those associated with William’s birth. It was meant to enhance further awareness of the empress’s links to the rulers whose lands she was claiming to rule. For all this, it is hard not to think that the story has some factual basis, at least to the extent of there actually having been a meeting. It is possible, too, that Matilda would have been consulted during this time by Bishop William of Durham about seeking Gregory VII’s approval for the project to replace secular canons at Durham with monks. Since, according to Symeon of Durham, Bishop William was advised by William, Matilda, and Lanfranc, he is likely to have consulted Matilda and Lanfranc in England, before crossing to Normandy to consult William.45 His presence in William’s company in Normandy on 15 September 1082 and subsequently in England at Downton (Wiltshire) before the end of the year suggests that he was then on his return journey from Rome, having obtained what he wanted from Gregory during the summer and then crossed back to England in the king’s company.46 Although our knowledge of these events is distorted by the creation at Durham in the twelfth century of an array of forged charters and papal bulls, there is no doubt that the mission took place. If this chronology is an acceptable one, then it shows once more how important Matilda was. It is also indicative of continued consultation with Gregory on subjects of great importance. Changing the way in which the cult of St Cuthbert was supported and celebrated was no small matter. Matilda would appear to have got through a formidable amount of work during this period when she was acting on William’s behalf in England. The extremely unusual absence of her signum and cross from all of the diplomas confirmed in Normandy during 1082 makes it possible that she acted on William’s behalf in England for almost a year Regesta, no. 120 (antequam mea conniux in Normanniam nouissime ueniret). The Letters and Charters of Gilbert Foliot, 66 (no. 26). See further Elisabeth van Houts, Memory and Gender in Medieval Europe, 900–1200 (Basingstoke and London, 1999), 73. 45 LDE, 226–7. 46 Regesta, nos. 253, 264. 43 44

political consolidation and personal loss

435

and that she did not cross to Normandy to join him until shortly before he decided to return to England in the autumn.47 All this is to an extent a commentary on the settlement that had been reached with their son Robert in 1080. Although not a royal charter, a record of a grant to the abbey of Ramsey of the church of Burnham Deepdale (Norfolk) dated 1081, describing the establishment of a confraternity, names only William, Matilda, and Robert. Arguably this was a reflection of what were anticipated to be the likely arrangements for the succession in England at that date, especially since Robert’s brothers were also in England.48 The continuation of the type of arrangement of the times that had preceded his rebellion, which foregrounded Matilda as William’s preferred representative during his absences on the other side of the Channel on major military campaigns, might well have seemed disparaging to Robert. On the other hand, since the events that took William to France were an encroachment by Count Fulk Rechin into Maine, it was probably judged that Robert’s presence was required there in his role as titular count of Maine. Both Robert and his younger brother William do appear in the 1082 diplomas confirmed in Normandy, probably indicating that the two of them accompanied their father on campaign.49 The apparently unusual selection of Matilda to rule on William’s behalf in England may have been intended to placate and elevate Robert. She, after all, had been his supporter and to move her ahead of all of their sons was uncontroversial given her status within the royal/ducal kindred. It is notable that the mention of a triumvirate, which appears in the Ramsey notice referred to above, replicates a formula that had been used before Robert’s revolt where reference to his authority was required.50 NORMANDY, MAINE, AND NORTHERN FRANCE, 1081–2 The reason why William left England was undoubtedly to tackle another intervention in Maine by Fulk Rechin, with two sets of Angevin annals mentioning a second attack by him on the castle of La Flèche in 1081

Regesta, nos. 158, 204, 205, 215, 264. Cartularium monasterii de Rameseia, ed. W. H. Hart and P. A. Lyons, RS, 3 vols. (London, 1884–93), i, 127–8 (ab omnibus fratribus plenam fraternitatem pro rege Willelmo, et pro regina Matilda, et pro comite Roberto); Lack, Conqueror’s Son, 27. Although the abbot of Ramsey named in the cartulary copy is incorrect, the reference to Matilda and the presence of the three of them in England arguably confirm the 1081 date. 49 Regesta, nos. 158, 204, 205. 50 Ibid., no 169 (laudantibus et consentientibus Mathildi Anglorum regina uxore mea et Roberto comite filio meo), a notice in favour of the cathedral church of Saint-Julien of Le Mans, dateable to between 1073 and Robert’s revolt. 47 48

436

william the conqueror

which, in contrast to the earlier one in 1076, was successful.51 Like King Malcolm’s 1079 raid and the breakdown of stability in Wales in 1081, this created a situation where an intervention was required. However, before leading an army southwards, having crossed from Southampton or Portsmouth to either the estuary of the Dives or the Orne, William is known to have confirmed a judgment made on his behalf in his absence, as well as a further grant to the abbey of Marmoutier, at Bénouville on the east bank of the estuary of the river Orne, sitting, so we are told, on his carpet between the forester’s house and the church; it is almost as if the monks were waiting for him to land in order to pounce to confirm their rights, a nice illustration of how informal rule could sometimes be and of how his itinerary must have been announced in advance.52 His immediate destination must have been Caen before leading an army south. Orderic mentions that Fulk had the support of Hoel, count of Nantes and Cornouaille, indicating yet another shift of alliances within Brittany, this one detrimental to William’s interests. While aspects of Orderic’s narrative are certainly over-dramatized, its essence is that these interventions by Fulk and William were on a different scale from previous ones. Not only does he say that William brought a 60,000-strong cosmopolitan army of Normans, English, and other peoples subject to his rule into Maine, he recounts how the two sides prepared for battle. Fulk, he says, was so determined that he ordered the boats on the river Loir be burnt to deter faint-hearted members of his army from retreat. However, before battle was joined, a cardinal priest, whom Orderic does not name, and some especially holy monks persuaded the two leaders to make peace. He adds that William, count of Evreux, and Earl Roger (presumably Roger de Montgommery) worked towards achieving the same end. With both having extensive lands and connections along Normandy’s southern borders and, in Count William’s case, almost no lands in England, an affirmation of the status quo was very much in both of their interests. A final peace was made at Blanchelande (or Landes de la Bruère) (Sarthe, cant. Le Lude).53 If this battle had happened, it would have been the first time that William had fought an engagement on this scale against a northern French territorial ruler. Since 1057, few of the actions of William’s northern French rivals 51 Annales sancti Albinii Andegavensis and Annales qui dicuntur Rainaldi, in Recueil d’annales angevines et vendômoises, 5, 88. 52 Regesta, no. 201. 53 Annales sancti Albinii Andegavensis and Annales qui dicuntur Rainaldi, in Recueil d’annales angevines et vendômoises, 5, 88; OV, ii, 308–11. I am grateful to Florian Mazel and Daniel Pichot for advice on the identification of this place name. The geographical proximity of the river Loir to the site suggests that Marjorie Chibnall’s translation of sed potius Ligerim fluuium audacter pertransierunt as the river Loire should be amended.

political consolidation and personal loss

437

look as if they were either intended or expected to be decisively significant. Hence, that the two sides came so near to a battle is extraordinary. It suggests that both of them saw the situation as so intolerable as to require something far beyond what had become conventional, either, from William’s perspective, because of Fulk’s meddling, or, from Fulk’s, because William had simply become too powerful. The peace that was made restored many aspects of the status quo ante bellum. William’s son Robert again did homage to Fulk, just as he had done to Geoffrey the Bearded almost twenty years before, and it is likely that he and William’s brother Count Robert of Mortain were effectively guarantors of the treaty, with the assumption being that either or both would be handed over as hostages if it was broken.54 A general amnesty was extended to the supporters of both sides, with Orderic finally adding that this agreement created a peace between William and Fulk that lasted for the rest of William’s life. The remarkable status of the guarantors is testimony to an extraordinary situation having arisen. The peace treaty shows a determination on both sides to put relations on a new footing. After finalizing the agreement, William may well have passed through Le Mans on his journey north, perhaps spending Christmas there and staying on into the New Year. A notice describing his confirmation of a transaction in favour of the palace chapel of the counts of Maine in Le Mans, the collegiate church of Saint-Pierre-de-la-Cour, says that it took place in the year of Bishop Arnold’s death and in the time of Bishop Hoel.55 Since Arnold died on 29 November 1081 and William was certainly in Normandy during the summer of 1082, it is quite likely that he stayed on in Maine for some weeks after the settlement with Fulk. As Hoel was a member of the Le Mans chapter, his appointment conforms to the pattern established in Maine under William’s rule, namely, the choice of someone with local connections, but also with a record of being associated with William’s court; while being Bishop Arnold’s disciple, according to Orderic, he had also done service as one of William’s chaplains.56 Orderic’s account of the debates that led to Hoel’s appointment suggests that William may have wanted to push a Norman into the see, but that he was talked out of it. His preferred candidate was another of his chaplains, Samson, the brother of Archbishop Thomas of York and one of Bishop 54 Annales qui dicuntur Rainaldi, in Recueil d’annales angevines et vendômoises, 88 (obsidibus pacis pro fide datis fratre suo, consule uidelicet Mauritanie, et filio suo, et multis aliis). 55 Regesta, no. 170. 56 OV, ii, 300–3; Actus Pontificum Cenomannis in urbe degentium, 382–3; Hillion, ‘Arnaud et Hoël, deux évêques du Mans au service de Guillaume le Conquérant’, 57. For Hoel as decanus and a likely member of the Le Mans cathedral chapter, Regesta, nos. 173, 174; Cartulaire de l’abbaye de Saint-Vincent du Mans, nos. 28, 29.

438

william the conqueror

Odo’s protégés. Samson, however, persuaded William that the humble and pious Hoel was a much worthier person to be appointed than he was. Since Samson did eventually become a bishop, succeeding Wulfstan at Worcester in 1095, he was probably being disingenuous. Portrayed by William of Malmesbury as a worldly man who enjoyed good living and was married, and with several years of a long career as a royal administrator behind him, he may well have decided that Le Mans was not the safest or most congenial place to hold episcopal office.57 If the discussion that Orderic describes did take place in William’s presence, it would not be the first or last example of him preferring an appointee like Samson. The placatory attitude to local interests and Hoel’s good personal qualities notwithstanding, Fulk still did all he could to obstruct his appointment, as his brother Geoffrey the Bearded had Bishop Arnold’s in 1065, so that Hoel was not consecrated until 21 April 1085 when the ceremony was performed in Rouen by Archbishop William.58 The Saint-Pierre-de-la-Cour notice mentioned above also refers to the requirement that the church serve William in England, Normandy, and elsewhere, one of the few surviving pieces of evidence for the Le Mans churches’ being integrated into any type of cross-Channel rule network.59 Other developments of these times, some the consequences of births and deaths, and some the outcome of conscious alliance-building, also consolidated William’s power and influence across northern France. In late 1080, Roger de Beaumont’s son Robert became count of Meulan in the French Vexin by right of his wife and, probably in association with his new status, a cross-Channel estate that included lands mainly in Warwickshire and Leicestershire was developed for him. This created an imperial assemblage of lands and networks that extended beyond Normandy and England and that was to remain influential throughout the twelfth century.60 Also, perhaps at almost the same time that William and Fulk Rechin made peace, William and Matilda’s youngest daughter Adela was betrothed to Stephen-Henry, the eldest son of Theobald, count of Blois-Chartres. The betrothal took place at William fitz Osbern’s former stronghold at Breteuil in close proximity to the lands of the counts of Blois-Chartres. Given the likelihood of Matilda’s presence, it may well have taken place in the late summer of 1082 after she had returned from England. The marriage was later celebrated at Chartres,

GP, i, 394–5, 438–41. Actus Pontificum Cenomannis in urbe degentium, 383; Latouche, Histoire du comté de Maine, 79; Guillot, Anjou, i, 120–1. 59 Regesta, no. 170. 60 David Crouch, ‘Beaumont, Robert de, Count of Meulan and First Earl of Leicester (d.1118)’, ODNB. 57 58

political consolidation and personal loss

439

perhaps in 1083.61 Adela would have been approximately fourteen years old at the time of her betrothal, and, at her marriage, approximately sixteen. Her husband was in his thirties. She was probably married ahead of Constance, the other available daughter, because, unlike her, she had been born when William was a king, a status celebrated by the poet Geoffrey of Rheims.62 Constance was herself married in 1086 to Alan IV Fergeant, count of Brittany, the son of Count Hoel and his wife Hawise, daughter of Count Alan III and Bertha. Alan IV had succeeded Hoel as count of Brittany in 1084. Just as Adela’s marriage consolidated William and his family’s relationship with the one ruler of a northern French territorial principality to whom he had been consistently allied, so Constance’s replicated earlier alliances, this time going back into the early eleventh century and to Duke Richard II’s time.63 Taken together, the two marriages surely had the purpose of building links with the two most powerful rulers in northern France to the east and west of the lands of the count of Anjou. After the grand marriages earlier mooted for William’s daughters, it must be significant that the only two marriages that actually ever took place involved northern French territorial princes, a reflection surely of where William identified the most serious threat to his achievement as lying. Of the two, Constance died childless in 1090. Adela went on to have a formidably successful career, effectively ruling the county of Blois-Chartres while her son Theobald was a minor, and living until 1137. By that time another of her sons, Stephen, was ruling Normandy and England, and another, Henry, was bishop of Winchester. After returning from Maine, William seems to have stayed in Normandy at least into the very late summer of 1082, and perhaps on into the autumn. Three dated diplomas show that he was at Oissel near Rouen on 24 June and 15 September.64 The final version of a diploma confirming all the possessions of the abbey of La Trinité of Caen also dates from this year, but it has neither a record of the place of confirmation nor a specific date within the year.65 If Matilda did spend most of 1082 in England, then the most likely time for the visit to Caen would have been before the two of them crossed to England in the autumn. All that we know of William’s itinerary after his arrival in England places him in the vicinity of Winchester, to where he could have OV, iii, 116–17. On the betrothal and marriage, Lo Prete, Adela of Blois, 52–60; van Houts, ‘The Echo of the Conquest in the Latin Sources’, 145–6. 63 For the marriage, OV, ii, 352–3; iii, 114–15; Joëlle Quaghebeur, ‘Havoise, Constance et Mathilde, princesses de Normandie et duchesses de Bretagne’, in Quaghebeur and Merdrignac (eds.), Bretons et Normands au Moyen Age, 145–63, at 152–4; Les actes des ducs de Bretagne, 371–2 (no. 90). 64 Regesta, nos. 204, 205, 264. 65 Ibid., no. 59. 61 62

440

william the conqueror

travelled either from the Seine or from the vicinity of Caen. One of the diplomas, the 15 September diploma for the abbey of Saint-Wandrille, looks like a follow-on from the council of Lillebonne and its clarification of a bishop’s jurisdiction over the administration of the ordeal by hot iron. Archbishop William of Rouen, so we are told, had unsuccessfully challenged the abbey’s right to administer the ordeal in four parishes over which it claimed to have full rights of jurisdiction after a monk had, from ignorance, converted the abbey’s ordeal-iron to other unspecified purposes and the archbishop had been asked to bless a new one for the abbey to use. The dispute was settled in William’s presence and that of a large number of bishops and abbots, a sign of sustained commitment to making the Lillebonne decisions work. Four of the diplomas of this year are confirmations of grants by William’s half-brother, Count Robert of Mortain; two of them were pancartes and a third a record of gifts to the abbey of Marmoutier that included the church of Notre-Dame of Mortain. The two pancartes, for, respectively, the abbey of Grestain and the collegiate church of Saint-Evroult of Mortain, are effectively full records of their entire endowments; Grestain of course was the foundation of Robert’s parents, and therefore of William’s mother and stepfather, and Saint-Evroult of Mortain, Robert’s own foundation at the heart of his comté. The Saint-Evroult of Mortain pancarte also mentions the church’s consecration by Archbishop William of Rouen and four out of his five suffragans, an event that must have taken place in 1082 or shortly before.66 It is ultimately impossible not to think of all this as a personal statement made when Robert’s wife Matilda, the daughter of Roger de Montgommery, was ailing; her death is mentioned in sections of the Grestain pancarte, which must date it to after 24 June, a date at which one of the Marmoutier diplomas mentions her as being alive. Her death seems to have had a considerable impact on William and Robert’s family, since a later record of another grant to Marmoutier associates both of them together in sadness.67 The sheer scale of Robert of Mortain’s endowment of the Church largely speaks for itself. Although described by William of Malmesbury as slowwitted, Robert’s responsibilities and achievements completely belie this cruel verdict. The holder of vast lands valued at more than £2,000 in over twenty English counties, with the largest concentrations being in Cornwall and the south-west, Sussex, what we now call the Home Counties, and Yorkshire, he was clearly one of the major figures at the heart of William’s rule; statistically the Domesday value of Robert’s lands makes him the wealthiest lay landowner after William and Roger de Montgommery, and all Ibid., nos. 158, 204, 205, 215. Ibid., no. 207 (Qua de re Willelmus rex eiusque frater Robertus, predicte Mathildis maritus, tristes effecti). 66 67

political consolidation and personal loss

441

the indications are that they were vigorously administered.68 Unquestionably one of the great figures of the cross-Channel elite, the only possible justification for Malmesbury’s assessment of Robert is that the absence of references to his having exceptional responsibilities either in military campaigns or in any of the great judicial inquiries may show that he lacked some of the intelligence of other members of William’s inner circle. The anonymous De Obitu Willelmi included the statement that William trusted Robert in everything, as befitted a kinsman.69 The appearance of this statement in relation to William’s distribution of his treasures on his deathbed surely confirms what the charter evidence of 1082 suggests, namely, that the two brothers were close. ENGLAND, 1082–3, AND THE ARREST OF BISHOP ODO OF BAYEUX According to both Orderic and Malmesbury, Bishop Odo was aiming to buy the papacy for himself, a scheme that he was expediting by sending money and letters to the citizens of Rome with pilgrims who visited the city. Orderic added that he had bought a palace there and that he had recruited Earl Hugh of Chester to help him.70 The mid-twelfth-century Warenne Chronicle, whose author knew Orderic’s text, repeated essentially the same story.71 Other evidence that Odo was plotting something grand may occur in a letter written to him by Berengar of Tours that borrowed some of its language directly from a letter that Berengar had written to Alexander II. In acknowledging Odo’s goodwill and his readiness to help him, Berengar may even be hinting at Odo’s ambitions; if so, it suggests that Odo was ready to go against the anti-Berengar line that had been sustained in Normandy since the meeting at Brionne in 1050.72 The letter cannot, however, be precisely dated, and it may be that Berengar was at some other date simply giving due acknowledgement to Odo’s extraordinary standing and wealth. A completely different story appears in the early twelfth-century Gesta Dei per Francos of Guibert of Nogent to the effect

68 GR, i, 506–7 (crassi et hebetis ingenii hominem); cf. Golding ‘Robert of Mortain’, 124–39; idem, ‘Robert, Count of Mortain (d.1095)’, ODNB. 69 GND, ii, 186–7 (quem quanto sibi proprinquiorem nouerat, tanto ei familiarius sua omnia credebat). 70 OV, iv, 38–41; GR, i, 506–7. 71 The Warenne (Hyde) Chronicle, 26–9. 72 Die Hannoversche Briefsammlung: iii. Briefe Berengars von Tours, in Briefsammlungen der Zeit Heinrichs IV, ed. Erdmann and Fickermann, 169, no. 101; H. E. J. Cowdrey, ‘The Papacy and the Berengarian Controversy’, in R. B. C. Huygens and F. Niewöhner (eds.), Auctoritas und Ratio. Studien zu Berengar von Tours (Wiesbaden, 1990), 109–38, at 137–8 (Cowdrey, Popes and Church Reform in the 11th Century, chapter 6).

442

william the conqueror

that Odo was plotting to take the kingdom from his half-brother. Guibert also emphasizes Odo’s great wealth and restless ambition.73 Odo’s absence from among the signa of the diplomas confirmed at Oissel in June and September may indicate that he had crossed to England to make preparations, a possibility increased by several of the other Norman bishops having been present on both occasions.74 But whatever he was doing cannot have immediately offended William, since Odo was among a substantial group of signa who confirmed a grant by Bishop William of Durham to his former abbey of Saint-Calais in Maine at Bishop Walchelin of Winchester’s great estate at Downton (Wiltshire) after William had crossed to England in late 1082.75 He was, however, absent from an equally prestigious list that confirmed English estates to La Trinité of Caen.76 Like the diplomas from the period of William’s return to England in 1080–1, the list of signa to these documents of later 1082 indicates William meeting with members of the group that were his closest associates. Lanfranc appears in both of them, a presence that might suggest that he, William, and others were working out how to deal with Odo. The SaintCalais diploma also shows that Matilda and their son Robert had again crossed the Channel with William; it looks, in other words, as if the general outlines of the 1080 settlement with Robert were being sustained. Another of the set-piece hearings into disputes between major ecclesiastics, which took place in William’s presence, also probably dates from this time. Bishop Giso of Wells claimed jurisdiction against the abbey of Glastonbury over the abbeys of Mulcheney and Athelney before William, Lanfranc, and others. At the assembly, as described by William of Malmesbury in the De Antiquitate Glastonie Ecclesie, Lanfranc’s first instinct was to support the bishop against the abbeys. At one point he even proposed to take away the abbot of Athelney’s pastoral staff, provoking from him a very curt refusal to yield it up. The recently appointed Abbot Thurstan then pleaded Glastonbury’s cause and drew attention to the abbey’s privileges, persuading William and Lanfranc to be more sympathetic. The final hearing was then transferred to the abbot’s court, where Bishop Giso accepted defeat. The arguments that are said to have proved decisive with William and Lanfranc were the abbey’s ancient records and also the reference to the authority of St Dunstan.77 In 73 Guibertus abbas S. Mariae Nogenti, Gesta Dei per Francos, 291 (The Deeds of God through the Franks, 136). 74 Regesta, nos. 204, 205, 264. 75 Ibid., no. 253. 76 Ibid., no. 60. 77 Early History of Glastonbury: An Edition, Translation and Study of William of Malmesbury’s ‘De Antiquitate Glastonie Ecclesie’, ed. and trans. John Scott (Woodbridge, 1981), 154–7; CS, 631–2; Regesta, no. 151. The arguments here for a 1082x83 date are based on Malmesbury’s

political consolidation and personal loss

443

addition, William confirmed properties to Glastonbury with the notice that records this showing that Archbishop Thomas was also present.78 According to Orderic, Odo was arrested on the Isle of Wight, just as he was preparing to depart. He was then formally tried at one of William’s palaces. Malmesbury on the other hand is entirely unspecific about both the place of arrest and the trial. With, as we have seen, charter evidence showing William and Odo together in southern England in late 1082, Orderic’s account must be inaccurate in its statement that William crossed the sea to arrest Odo. It looks therefore as if Odo was taken as he prepared to leave England late in the year and then transported for trial to a palace, with Winchester being the obvious venue. If it took place at the Christmas crown-wearing, then Gloucester cannot be ruled out, but there is no specific reason to locate it there. Both Orderic and Malmesbury say that Odo was condemned as an earl, and therefore as a layman, rather than as a bishop, on the grounds that by leaving England without William’s permission, he was deserting his post and abandoning his responsibility to defend the kingdom.79 Both also produce some typical moralizing on the price paid for excessive worldly ambition, but neither has William condemn his half-brother for aiming to become pope. It is Malmesbury who says that the distinction between Odo’s two roles was attributed to Lanfranc, a creative legalism for which Odo is said to have come to detest the archbishop.80 The necessity for proceeding in this way was not to deal with Odo as a rebel, something that could have been done within existing law on which Lanfranc’s version of the PseudoIsidorean Decretals provided clear guidance, but to place him outside the norms associated with the punishment of a cleric; the objective was not to send him to a monastery as had been done with English rebel ecclesiastics in the aftermath of 1066, and as Pseudo-Isidore recommended, but to imprison him.81 This was duly done and Odo was kept closely confined in statement that Thurstan was appointed in 1082 and the presence of his predecessor Abbot Æthelnoth at Winchester at Whitsun 1081, Regesta, no. 39. For a dating of Christmas 1080 at Gloucester, CS, 630; Cowdrey, Lanfranc, 125–6, based on the statement in the so-called Acta Lanfranci that Æthelnoth was deposed in 1078 (The Anglo-Saxon Chronicle. A Collaborative Edition: 3. ms. A, 86), but its authority when set against the Glastonbury evidence is relatively weak, especially since its source cannot be identified. See further Hayward, ‘Some Reflections’, 147. Æthelnoth also made a grant as abbot in a Glastonbury charter dated 15 August 1079, The Great Chartulary of Glastonbury, ed. Dom. Aelred Watkin, 3 vols. Somerset Record Society, vols. lix, lxiii, lxiv (1947–56), iii, 701–2. 78 Regesta, no. 151. 79 Orderic typically creates a speech for William in which he sets out these arguments, OV, iv, 40–5. 80 GR, i, 544–5. 81 Daniel Gerard, ‘Ermenfroi de Sion, l’archevêque Lanfranc et le problème des ecclésiastiques rebelles’, in Barrow, Delivré, and Gazeau (eds.), Autour de Lanfranc, 305–11, at 307–8, 310.

444

william the conqueror

the castle at Rouen for the rest of William’s life. Gregory VII wrote a ferocious letter to his legate Archbishop Hugh of Lyons, of which only a fragment survives, condemning William’s action. His letter to William himself, however, did not rise above mild reproach, even if it did censure him for not showing sufficient respect for the clergy; he began by mentioning their long-standing friendship and then cited authorities such as Constantine to point out that there were higher ideals involved than mere worldly politics in matters of this kind.82 While a degree of incredulity must always attend the notion that Odo was aiming for the papacy, the possibility must not be entirely ruled out; the history of the sixteenth century does, after all, provide the parallel case of Cardinal Wolsey proceeding along very similar lines to Odo at a time when relations between an English king and a pope were also fraught. The idea that, in the midst of the conflict between Gregory VII and Henry IV, an experienced outsider might have been appealed to as a compromise is not total fantasy. That Odo had fathered a son, and that the Bayeux chapter contained many clergy cohabiting with women, does, however, make him an unlikely prospect in the context of attitudes that had prevailed at Rome for several decades.83 While he might have thought that the Norman settlers in southern Italy could intervene to help him, if true, the story must surely involve a considerable capacity for self-delusion. It looks like a tale of ambition that in the next generation grew in the telling. It is possible that Odo had decided on an exploratory visit to southern Italy to assess his prospects there. Disenchantment caused by the way in which Matilda took over responsibilities in England in 1081–2 that he had previously exercised could have been a factor. He was perhaps being pushed to one side within the ruling kindred. Whatever Earl Hugh of Chester’s involvement, it must have been superficial, since there is no sign that he was in any way punished. Bishop Odo has always provoked widely differing assessments. A more flamboyant figure than his brother Robert, and perhaps even than his halfbrother William, with whom he nonetheless shared a capacity for show and self-aggrandizement, Odo was characterized by Orderic as someone in whom vices and virtues were strangely mixed.84 His huge wealth and his lavish patronage made him an easy target for moralizing commentators. The value of his English lands listed in Domesday Book, lands that stretched across eastern England from Kent to just south of the Humber across twenty-two counties, vastly exceeded that of anyone other than the 82 The Epistolae Vagantes of Pope Gregory VII, ed. H. E. J. Cowdrey, OMT (Oxford, 1972), no. 53; Registrum, ii, no. IX. 37 (Register, 441–2). 83 For a recent discussion along these lines, van Houts, ‘The Fate of Priests’ Sons’, 74. 84 OV, iv, 114–19.

political consolidation and personal loss

445

king himself. Marbod, who later became bishop of Rennes and who benefited from Odo’s patronage, commented that the bishopric of Bayeux was wealthy enough to support three bishops.85 Yet, when examined in detail, Odo’s activities encapsulate on a grand scale the ethical dilemmas associated with writing about the conquest of England and the eleventh-century Church. Exploitative and oppressive he certainly was, yet there are also clear signs in his judicial activities of a concern for equity and, within his earldom of Kent and elsewhere in England, of a positive contribution to the integration of newcomers and natives. He was, for example, remembered as a patron at St Augustine’s, Canterbury, and St Albans and Odo advised the first of these monasteries on the translation of St Augustine’s relics. Several of his tenants, most notably Turold, Wadard, and the Caen stone entrepreneur Vitalis of Canterbury, were also exceptionally well integrated into their new societies.86 The bishop who was also a warrior was hardly an alien figure in eleventh-century western Europe. Indeed, as Gregory VII put a greater emphasis on violence sometimes being a Christian duty, Odo was in certain respects in tune with his times.87 In this overall context, to marginalize him excessively as a patron of the Bayeux Tapestry seems to me unjustified, although it is entirely right to remove him from any interfering role in it.88 His patronage of learning was also remarkably extensive.89 THE YEAR 1083 AND MATILDA’S DEATH In all probability early in the year 1083, William and his family crossed the Channel to Normandy. As there is nothing in the narrative sources to explain their movements, it is likely that they were determined by routine, the relatively short visit to England being adjudged to have dealt with all 85 PL, clxxi, col. 1658. For a survey of Odo’s lands, Rowley, The Man behind the Bayeux Tapestry, 116–19. 86 For the basic points about Odo’s patronage in England, David Bates, ‘The Character and Career of Odo, Bishop of Bayeux (1049/50–1097)’, Speculum, l (1975), 1–20, at 9–10; Heslop, ‘Regarding the Spectators of the Bayeux Tapestry’, 229–32; Pastan and White, The Bayeux Tapestry and its Contents, 82–104. 87 Most recently on the subject, J. R. Webb, ‘Representations of the Warrior-Bishop in Eleventh-Century Lotharingia’, EME, xxiv (2016), 103–30. 88 Elizabeth Carson Pastan and Stephen D. White, ‘Problematizing Patronage: Odo of Bayeux and the Bayeux Tapestry’, in Martin K. Foys, Karen Eileen Overbey, and Dan Terkla (eds.), The Bayeux Tapestry: New Interpretations (Woodbridge, 2009), 1–24; Pastan and White, The Bayeux Tapestry and its Contents, 64–81. 89 For a survey, David Bates, ‘Le patronage clérical et intellectuel de l’évêque Odon de Bayeux 1049/50–1097’, in Sylvette Lemagnen (ed.), Chapitres et cathédrales en Normandie, Annales de Normandie, série des congrès des sociétés historiques et archéologiques de Normandie, vol. 2 (Caen, 1997), 105–14; Heslop, ‘Regarding the Spectators of the Bayeux Tapestry’, 232–5.

446

william the conqueror

necessary business. The adjudication in April at Fécamp in favour of Archbishop William of Rouen’s reclamation of an estate alienated from his cathedral’s possessions by Archbishop Robert at least fifty years previously suggests that William spent Easter (9 April) there,90 after which he moved westwards across Normandy to be at Caen on 18 July.91 There he judged several disputes about the demarcation of the zones of influence within and around Caen of the two abbeys. The implication of these cases is that both of the abbeys had become very prosperous, since the records of the settlement refer to the expansion of the urban areas (bourgs) controlled by the monasteries.92 Both of these charters also indicate that Matilda and their eldest son Robert were present, with the second son William also being with them at both Fécamp and Caen. The judgment reached at Fécamp is described as having been a collective one. It would seem that Robert was still remaining, or being kept, close to his parents on both sides of the Channel, continuing the arrangements initiated in 1080 that involved association in rule. In contrast to the years between 1077 and 1082, no one appears to have been left behind in England with specific responsibilities to represent William. This may well point to a degree of confidence that much had been successfully settled. The events of the year indicate that, as had happened before, action could still be taken by the regime when serious situations arose. An occurrence of this kind, the murder of several monks of Glastonbury after Abbot Thurstan had introduced soldiers into the monastery to discipline brethren with whom he was effectively at war, is described in several sources. All in one way or another imply that the problem stemmed from Thurstan’s attempts to modernize, as he saw it, certain practices; John of Worcester, probably followed in this by Orderic as a result of the visit he made to Worcester, specifically refers to Thurstan trying to replace the monks’ traditional chant.93 The sources differ a little on the number of monks killed and on details of the fighting, with the ‘E’ Chronicle mentioning three deaths and eighteen wounded and John two deaths and fourteen injuries, but all say that the outcome was Thurstan’s effective suspension from office and his transfer back to the abbey of Saint-Etienne of Caen from whence he had come. With William in Normandy, all the crucial decisions must have been taken by others, albeit, as John and Malmesbury’s version of the De Antiquitate make clear, in anticipation of the approval by William that was in due course given. Thurstan’s Regesta, no. 230. AAC, no. 17 (with facsimiles); Regesta, no. 64. 92 On this, Jean-Marie, Caen aux XIe et XIIe siècles, 37–57. 93 JW, iii, 38–41; OV, ii, 270–1. For the other main accounts, ASC, ‘E’, 1083; GP, i, 308–11; Early History of Glastonbury, 156–9; Regesta, no. 146; DB, iv, 491. 90 91

political consolidation and personal loss

447

return to England by 1086 does suggest that his full reinstatement was under consideration during William’s lifetime. As had happened after the disorders in the abbey of Saint-Ouen of Rouen in 1072, William ordered that some of the Glastonbury monks be distributed around other monasteries. The presence of Abbot Baldwin of Bury St Edmunds in Caen in July may indicate that he had crossed the Channel because there were concerns about Matilda’s health. As we have already seen, the monk Herman, writing about Baldwin, said that he was frequently summoned to Normandy because of his medical expertise.94 Orderic’s reference to an illness that gave increasing cause for concern suggests a steady and worrying deterioration. Malmesbury, on the other hand, somewhat disturbingly, mentions a rumour that Matilda died as a result of a beating from William after she had discovered that he had had a relationship with a priest’s daughter. He does, however, dismiss the story as lunacy.95 While, as mentioned in Chapter 3, physical discipline was considered legitimate correction in certain circumstances for perceived wifely transgressions, Baldwin’s presence in Normandy and Matilda’s continuing appearances in charters do indicate that she was thought strong enough to maintain a public role and, perhaps, that she was thought likely to recover, a probable confirmation of Orderic’s commentary.96 She nonetheless died on 2 November. Both Orderic and Malmesbury describe her funeral as an exceptionally splendid one, with Malmesbury adding that William then remained in mourning for many days and that he was so affected by the loss that he renounced every sensual pleasure for the rest of his life, without specifically saying what this involved. Orderic, in typical fashion, concentrates on Matilda’s works of charity and says that many poor people who had been beneficiaries attended the funeral; the reference noted earlier to a grant of geld exemption to a widow living in Dorset may be one example of this.97 Shortly before her death, Matilda bequeathed her crown and her sceptre to La Trinité.98 As planned, she was buried in the choir of the abbey, where her tombstone can still be seen (Plate 3). The inscription drew attention to her virtues, her foundation of the abbey, her marriage to William, and her illustrious ancestry as the descendant of kings.99

Regesta, no. 64; Herman and Goscelin, Miracles of St Edmund, 216–17. GR, i, 500–3. 96 OV, iv, 44–5. 97 ‘The Dorset Geld Rolls’, 128. 98 AAC, no. 16; Regesta, no. 63. 99 For the text of the inscription, OV, iv, 44–7; Corpus des inscriptions, ed. Favreau and Michaud, 51–4. 94 95

448

william the conqueror

While we know a lot about Matilda’s acts – she is so often presented in the sources in terms of being the embodiment of what eleventh- and twelfth-century writers regarded as the medieval stereotypical wifely and female virtues – we can feel as if we know very little about her. A physically strong woman – to repeat the point about her stature and physique made in Chapter 3 – her range of activity was certainly prodigious.100 While charters often conventionally frame her donations and decisions in relation to William’s authority, it is clear that she did possess considerable autonomy and that her religious patronage was very often personally directed. And, for all that she is often associated with sustaining rule in Normandy when William was in England, she also appears in all of the major surviving English diplomas of William’s reign and, as we have seen, acted there on her husband’s behalf in 1081–2. Her personification of pro-active queenship within the framework of her husband’s rule was therefore an essential part of the apparatus of rule on both sides of the Channel. The laudatio clauses in Norman and French charters often mention her consent to a grant alongside William’s and sometimes place it on the same level as his.101 The judicial decisions taken by her in England recorded in Domesday Book and her role in presiding at a gathering of four shires in the Alveston (Warwickshire) case mentioned above, like some of the evidence for her religious patronage, show that Matilda travelled more extensively in England than the charters would suggest.102 She also influenced the distribution of landed wealth to the foremost members of the lay elite, with later traditions and a record of Count Alan’s grant of Swavesey (Cambridgeshire) including the information that his receipt of Earl Edwin’s lands in Yorkshire and of estates in Cambridgeshire came about in part as a consequence of her favour and support.103 The main focus of Matilda’s religious patronage was ultimately on Normandy where, in addition to the establishment of La Trinité of Caen, gifts to several houses are recorded, with Orderic mentioning a visit to his abbey of Saint-Evroult and her financial contribution to the construction of a refectory; while grants and her generosity are recorded at Le Bec, Saint-Amand of Rouen, Fécamp, and in versions of the Norman annals

100

Elisabeth van Houts, ‘Matilda [Matilda of Flanders] (d.1083)’, ODNB. The material is summarized in Regesta, 92–4. See further Bates, ‘The Representation of Queens and Queenship’, 287–91. 102 DB, i, fos. 48v, 238v. 103 Early Yorkshire Charters, 12 vols., ed. W. Farrer and C. T. Clay, Yorkshire Archaeological Society Record Series (Wakefield, 1935–65), volume iv, The Honour of Richmond, ed. C. T. Clay, no. 1; Sharpe, ‘King Harold’s Daughter’, 22; Keats-Rohan, ‘A Question of Identity’, 179–81. 101

political consolidation and personal loss

449

from Rouen and Jumièges.104 In France and continental Europe, she paid for the construction of a large refectory at Marmoutier, made gifts of valuables to Cluny, Marmoutier, Saint-Corneille of Compiègne, and SaintFlorent of Saumur, and paid for the tomb of Count Simon of Amiens-Valois-Vexin in Rome, as well as making gifts to the church where he was buried.105 In England, she was a benefactor of the abbeys of Malmesbury and, very generously, of Bury St Edmunds. Little Domesday Book mentions at least two visits to Bury with William and her special affection for the abbey and for St Edmund that caused William to make a gift to the abbey specifically for her soul after her death.106 She used her influence to secure a grant for Wells Cathedral.107 However, the Abingdon Historia also mentions how Matilda joined in the despoliation of English monasteries, arriving at the abbey and demanding gifts, only to reject them and extort much more valuable ones, namely a chasuble, an alb with a stole, and a Gospel book, all of which the monks regarded as being among their greatest treasures.108 This could have happened when she and William visited the island of Andersey in the Thames near to the monastery for one of their rest cures. Her devotion to her son Robert, her tears when her daughter Cecilia was confirmed as a nun, and her special relationship with her and William’s foster-son Count Simon, all point to Matilda being an emotionally dedicated mother.109 With his unique interest in biography, William of Malmesbury identifies some personal characteristics and even says something about the marriage. In addition to Matilda’s religious patronage, the quality he emphasizes is her loyalty to her eldest son. He goes on to say that, although her sending money and troops to Robert during his quarrel with William led to tension between the couple, it did not in any way reduce William’s devotion to her. He actually writes of love being present in the marriage.110 Although there is Malmesbury’s possible evidence for the infliction of physical violence 104 OV, ii, 148–51; Regesta, nos. 167, 239; ‘Chronicum Rotomagense’, 366, and Les Annales de Saint-Pierre de Jumièges, 59, with de cuius nobilitate largiflua multa apud nos habentur beneficia appearing in both; The Ordinal of the Abbey of the Holy Trinity Fécamp, i, 13–15. 105 ‘Chronicon abbatum Majoris Monasterii’, 318; Gilo, Vita sancti Hugonis abbatis, in ‘Memorials of Abbot Hugh of Cluny’, 64–5; Vita sancti Hugonis abbatis Cluniacensis, PL, clix, cols. 923–4; Recueil des actes de Philippe Ier, no. CXXVI; Marchegay, ‘Chartes angevines des XIe et XIIe siècles’, 413–14; Vita beati Simonis, PL, col. 1222 (trans. van Houts, Normans in Europe, 199). 106 For Matilda’s monastic patronage in England, Cownie, Religious Patronage, 71; Bates, ‘The Abbey and the Norman Conquest’, 12. For a reference to a visit which makes it clear that there were others, DB, ii, 210r. 107 Regesta, no. 289. 108 Historia Ecclesie Abbendonensis, i, 224–5. 109 GR, i, 500–3. 110 Ibid., 502–3.

450

william the conqueror

and of William possibly having sexual relations with other women, by the standards of the eleventh and early twelfth centuries, Malmesbury’s verdict on William’s devotion has to stand. William was certainly unusual among aristocratic men of his time in not fathering illegitimate children. In terms of rule as represented through the evidence mainly of charters and Domesday Book and, in all probability of emotional support, of which we know much less, her death was a major loss. For all William’s achievements in his final years, he does in some ways seem reduced. And the breach between him and their eldest son that was so profoundly to influence events after his death way well not have occurred if Matilda had lived. Her role in the political consolidation of these years looks to have been very important.

Chapter 12

T H E FI NAL Y EARS

In what turned out to be the final years of his life, William had to continue to juggle cross-Channel priorities as he had done ever since 1066. Another military intervention in Maine became necessary in either 1084 or 1085. And in the autumn of 1085 he went to England to confront an expected invasion from Denmark, having assembled a large mercenary army in Normandy in advance of the crossing.1 Although the invasion did not materialize, it was potentially more dangerous than the previous ones. The year spent in England in 1085–6 also involved the ‘deep speech’ at Gloucester at Christmas 1085 and the subsequent survey that produced what later ages have called Domesday Book, processes in which William was a centrally active participant.2 Relations with his eldest son broke down again during this period. And having crossed to France in the autumn of 1086, William was taken ill in the summer of 1087 while leading a campaign into the French Vexin. The weaknesses of the available chronological information for the second quarrel with Robert and for events in Maine pose problems for the construction of a narrative. It is, however, at least clear that these were years of crisis of a familiar kind and that steps were taken to deal with it, which, partly through good fortune and partly through efficient organization, enabled the crisis to be weathered. The rationale behind the survey that produced Domesday Book has been the subject of what can feel like an infinite number of discussions among scholars. At this point, all that needs to be said is that it is essential to look beyond the iconic status it has acquired in English national life and identity, and to place it in its contemporary context, while at the same time not losing sight of just how remarkable a record it is. And that the emphasis has to be on its traditional, multi-faceted, and potentially multi-functional, characteristics. Although seemingly the product of a moment, so many of the themes of what had happened since 1066 intersect within it that Domesday Book must be viewed as a product, both conscious and unconscious, of the conquest of England. 1 2

For the date of William’s crossing, JW, iii, 42–3. ASC, ‘E’, 1085. Regesta, nos. 9 and 156, were both confirmed at Gloucester.

451

452

william the conqueror

1084 AND 1085 IN FRANCE AND ENGLAND As far as we know, William spent all of 1084 and more than half of 1085 in northern France. He was certainly in Normandy on 9 January 1084 and at Easter (31 March) when, according to the Abingdon Historia, he and his sons Robert and William were there while the third son Henry was at Abingdon.3 He was at Rouen on 19 June when he appointed his and Matilda’s former chaplain and the monk of Jumièges, Rainald, as abbot of Abingdon.4 Two diplomas for the abbeys of Jumièges and Saint-Ouen of Rouen also indicate time spent at Rouen, with the second of them being confirmed at Saint-Ouen itself.5 Another diploma for the abbey of Lessay is further evidence of his presence in Normandy and, because of the location of its beneficiary, suggests that a journey was made to Lower Normandy.6 The Anglo-Saxon Chronicle’s reference to the imposition after the mid-winter of 1083–4 of a ‘great and heavy tax’ of seventy-two pence (six shillings) on the hide might indicate a very short visit to England between 9 January and Easter but, in all likelihood, the instruction to levy the geld was transmitted from Normandy.7 Its imposition could be connected either with a very advanced awareness of the threat of the invasion from Denmark or with the possibility of needing to finance more warfare in northern France. Its severity means that a visit to England in the late summer or autumn of 1084 and perhaps a Christmas crown-wearing there, motivated by the need to display his presence, cannot be ruled out, but there is no evidence for either of them. A Fécamp diploma devoted entirely to Norman business and almost certainly written by a Fécamp scribe suggests a visit to Upper Normandy at some point during 1085. We know from John of Worcester that it was not until the autumn of that year that William crossed to England.8 The combination of this absence from England and the chronological confusions of the surviving evidence for some crucial episodes actually makes it possible that the year 1084 was one of the few peaceful years of William’s entire life. But the levying of the heavy geld set in train at the start of the year must indicate that a crisis was already anticipated and that preparations to confront it began then. Regesta, no. 252; Historia Ecclesie Abbendonensis, ii, 16–17. Historia Ecclesie Abbendonensis, ii, 18–19. For evidence that Rainald was Matilda’s chaplain, Bloche, Le chartrier de l’abbaye de la Trinité de Fécamp, 323–4, no. 38. For Matilda described as Rainald’s domina, Regesta, no. 162. 5 Regesta, nos. 163, 248. 6 David Bates, ‘A Charter of William the Conqueror and Two of His Sons’, Tabularia. Documents, v (2005), 17–27, at 24–6 (www.unicaen.fr/mrsh/craham/revue/tabularia). 7 ASC, ‘E’, 1083. 8 Regesta, no. 145; JW, iii, 42–3. 3 4

the final years

453

While helpfully dating the outbreak of the conflict with the vicomte of Maine, Hubert de Sainte-Suzanne, to after Matilda’s death, Orderic nonetheless not only manages to have the siege last for three years in one place in the Historia Ecclesiastica and for four in another, but also does not spot that the longer of these two periods would have had William organizing it from the grave.9 However, his statement that Hubert eventually crossed to England to reach a settlement with William shows that the siege ended either late in 1085 or in 1086, with a charter for the abbey of SaintVincent of Le Mans supplying a second set of dating-limits of between 23 May 1085 and 20 April 1086 for Hubert’s Channel crossing.10 It is therefore at least clear that the siege was a long one. And also that it was a divisive factor in the politics of Maine, since Orderic says that William’s intervention was in part a response to the citizens of Le Mans’ complaints about Hubert’s destructive raids southwards from Sainte-Suzanne. Once more, the efficacy of his rule in Maine was being called into question in ways that were beyond the capacity of local supporters to control, making an intervention essential. Orderic’s narrative is the only written account we have of the siege. An exceptionally detailed one that is a goldmine of information for the strategy, tactics, and values that such a military operation involved, it is a little obscure on the causes of the quarrel. It indicates that Hubert defied William because he had offended him, firstly in an insignificant way, but later more seriously; in other words, he believed that he was fighting for his personal honour rather than rebelling to overthrow William’s rule in Maine. Orderic’s statement that the final settlement involved William guaranteeing the security of Hubert’s inheritance suggests that he had been threatening a withdrawal of favour and perhaps reprisals against a powerful man who had been one of William’s main opponents in both 1062–3 and 1069–73.11 Hubert was supported by his kinsman Robert the Burgundian, lord of the castles of Sablé and Craon, which lay to the south and south-west respectively of Sainte-Suzanne. Their landholdings meant that, working together, they placed a substantial barrier between William and the city of Le Mans and also threatened turbulence in regions close to Brittany where regime change was in process when Count Alan IV replaced his father Hoel as duke of Brittany in 1084. But Hubert seems not to have received any significant assistance from within Maine, or from anywhere else; while Orderic mentions the arrival at Sainte-Suzanne of warriors OV, iv, 46–53. Ibid., 52–3; Cartulaire de l’abbaye de Saint-Vincent du Mans, no. 492 (et in illo eodem tempore Hubertus, uicecomes de Sancta Susanne, fecit pacem et concordiam cum Anglico rege). 11 OV, iv, 46–7, 52–3 (paternum ius honorifice recepit). For some comment on the quarrel, Lemesle, La société aristocratique, 36–7; Jessee, Robert the Burgundian, 131–4. 9

10

454

william the conqueror

from Aquitaine, Burgundy, and elsewhere in France, he puts their interest down to making their reputations and acquiring rewards. There is no reference at all to any direct intervention by Fulk Rechin or Philip I, or indeed even by Geoffrey de Mayenne, although it may not be a coincidence that Robert the Burgundian and Geoffrey de Mayenne had attended Fulk’s court at Christmas in 1083 and in early January 1084.12 Their presence there may well have alerted William to the situation’s potential to explode into another major conflict in Maine and across northern France, and explain the levying of the heavy tax in England. Although not mentioning where it was confirmed, a charter of 9 January 1084 describing a grant by Count Robert of Mortain to the abbey of SaintBenoît-sur-Loire could supply a date for William and his army’s march south. Mortain’s location must have made Robert central to the campaign and the charter’s largely western Norman group of signa could support the argument for it being confirmed there. William’s two eldest sons Robert and William were present with him, another indication perhaps that the campaign was being organized.13 On the other hand, any time during the spring, summer, or even the autumn of the year is also possible. Given the timing of Robert the Burgundian and Geoffrey de Mayenne’s visit to Anjou, the charter may be evidence of a preliminary marshalling of forces, with the hope being that the threat of an intervention would produce negotiations. On balance, mid-summer after the 19 June visit to Rouen is the most plausible time for the actual advance to Sainte-Suzanne. Whenever it happened, the campaign resulted in a blockading force being installed in the earthwork known as ‘Le Camp de Beugy’ a short distance to the north of Sainte-Suzanne (Plate 21). It appears that an unfortified earthwork enclosure that was an aristocratic dwelling existed on the site in the tenth century. But recent archaeological work indicates that major changes which have contributed significantly to its present remarkable appearance were made in the twelfth century. It is therefore possible that what William and his forces did was to enlarge an existing earthwork to create a siege-castle, which served as a relatively secure place for the troops to live and from which to launch what turned out to be fruitless attacks. Exactly what he might have done in the 1080s is, however, very unclear and awaits further investigation.14 No attempt that we know of was made to surround Sainte-Suzanne with siege-castles to the south or to For this point, Jessee, Robert the Burgundian, 131. Regesta, no. 252. 14 Pierre-Yves Laffont and Lucie Jeanneret, ‘Le site de la Motte à Sainte-Suzanne (Mayenne) est-il une fortification de siege de Guillaume le Conquérant?’, in Signes de pouvoir laic et religieux au Moyen Age. Les fortifications laïques et religieuses: actes du Xe colloque biennal de Pommiers-en-Forez, 12 juin 2015, Association culturelle de Pommiers (2016), 77–94. I am 12 13

the final years

455

attack either Sablé or Craon. This was presumably to keep the quarrel at a personal level between William and Hubert and to avoid provoking others to join in. It indicates that a strategic decision was taken to contain the conflict rather than to go over to an all-out attack. According to Orderic, William was eventually obliged to leave SainteSuzanne because of urgent business elsewhere, with the garrison left behind at ‘Le Camp de Beugy’ commanded by Count Alan. The choice of Alan as commander was probably determined by Sainte-Suzanne’s proximity to Brittany and his capacity to maintain regional loyalties in a way that a Norman might not have done. However, his attestation of a diploma that must have been confirmed in England in the last months of 1085 either indicates that he was not continuously present at Sainte-Suzanne or that Orderic greatly exaggerated the duration of the siege and that Hubert’s submission, which took place between 23 May 1085 and 20 April 1086, had happened by then.15 The presence at one time or another at the siege of William de Warenne and Count William of Evreux, as well as of Count Alan, suggests that substantial military weight was deployed against Hubert and his allies. Given Alan and William de Warenne’s responsibilities throughout the cross-Channel territorial complex created by William, we also see from this episode how important their local deputies left in England must have been to its functioning, just as William’s were to his cross-Channel rule. The narrative tells of attempts to take the castle of Sainte-Suzanne on its high rock by storm and of the capture and ransoming of prisoners by both sides after skirmishes. Eventually, given that the situation had effectively become a stalemate, Hubert crossed to England and sought, either in late 1085 or in 1086, to make peace with William. This was done and, as always with rebels in Maine, he suffered no penalty that we know of. Orderic says that peace between him and William endured for the rest of William’s life. Viewed in a broader perspective, the siege of Sainte-Suzanne had involved a great expenditure of time and resources. But the way in which Hubert and Robert the Burgundian remained isolated is surely testimony to the awe that William inspired. The urgent business that Orderic referred to could be either the outbreak of the second major quarrel with Robert Curthose or reports of a threatened invasion from Denmark and Flanders. However, the vagueness of the chronological information we possess once more makes a fully informed narrative of the quarrel with Robert difficult to construct. He attests only grateful to Pierre-Yves Laffont for discussions of the site of Sainte-Suzanne and for sharing with me the reports of the excavations of 2014 and 2015. 15 Regesta, no. 144. The argument for a confirmation in England is based on Lanfranc’s presence and on what we know of William’s itinerary.

456

william the conqueror

two charters during this period, one the diploma of 9 January 1084 confirming Robert of Mortain’s grant to Saint-Benoît-sur-Loire, and the second, a grant to the abbey of St Peter’s, Gloucester, made at Gloucester at Christmas in 1085.16 On the basis of the second of these and of other more circumstantial arguments, the long-held view that Robert was in exile for upwards of three years has recently been contested in favour of an estrangement that may either have been episodic or have lasted for not much more than a year. Attempts to achieve a reconciliation that failed, of which we know nothing, may have been made. Alternatively, if the period of estrangement was confined to the years 1086 and 1087, then no negotiations may have been needed; it is even possible that they quarrelled only after William had returned to Normandy in the autumn of 1086.17 The one chronological certainty is that Robert was an exile during his father’s final illness in 1087, with, according to Robert of Torigni, him being at Abbeville in Ponthieu.18 Only one diploma, a Fécamp original confirmed in England in late 1085, features his younger brother William on his own in a way that might indicate Robert was an absent exile.19 His presence at Gloucester at Christmas 1085 could be a sign that the quarrel was one in which father and son periodically put distance between themselves to allow feelings to calm down, with the appearance at Gloucester being an occasion on which a reconciliation was attempted. The gathering was in any case of such great importance that Robert’s acknowledged interest in succession to the kingdom surely required that he be present when the survey that preceded Domesday Book was initiated. Matilda’s death must have contributed to the new breakdown of relations because it took away her role as mediator. Odo of Bayeux’s support for Robert in the wars that followed William’s death suggests that his imprisonment removed a powerful ally.20 Given that there are, also, clear signs that the youngest son Henry was being promoted to a more prominent status in England where he had been brought up, this could have been a cause of friction; although he was not knighted until 1086, in the context of what appears to have been agreed in 1080–1, his more elevated status was surely provocative.21 Orderic’s account conveys very clearly the impression that neither William nor Robert was blameless; for Regesta, nos. 156, 252. For the case for a short estrangement that began in 1086, Lack, Conqueror’s Son, 38–40; ‘Robert Curthose’, 129–31. 18 GND, ii, 202–03. 19 Regesta, no. 144. The case for confirmation in England is strongly indicated by Lanfranc’s presence. 20 Aird, Robert Curthose, 95–6. 21 Hollister, Henry I, 35–7; Green, Henry I, 23. 16 17

the final years

457

him, Robert may have left for trivial reasons and have been stubborn and disobedient, but his father was guilty of regularly criticizing him in public and his behaviour was emotional rather than rational.22 But whatever view is taken of the estrangement’s causes and its length, it never developed into a conflict of the kind that had produced the warfare around Gerberoi in 1078–9 and had divided aristocratic families. Both Orderic and Malmesbury mention Robert leaving Normandy with only a few followers, and Orderic refers to ‘long and aimless wanderings’ and to the reasons for the disagreement as ‘trivial’. It is Malmesbury who tells the story of Robert’s journey at some point to Italy and his unsuccessful proposal to marry Matilda, countess of Tuscany (c.1046–1115), a formidable character and a great supporter of Gregory VII; given her status as an heiress and how involved she was in the politics of the popes and the emperor Henry IV, she would not have considered moving to Normandy. Her second matrimony shows that she married politically and scarcely cohabited at all. For the union to have been a serious one, Robert would have been forced to move to Italy.23 Whatever the state of relations between father and son, William crossed to England in the autumn of 1085 to confront the threat of an invasion from Denmark without seemingly having any qualms about Normandy’s security. If the estrangement had already begun, Robert apparently either made no attempt we know of to profit from his father’s difficulties or received no support to enable him to do so. The large army that William took across the Channel was presumably in part financed by the early 1084 geld.24 On the basis of William of Malmesbury’s statistics, the invasion planned by Cnut IV in alliance with his father-in-law, Count Robert the Frisian of Flanders, was intended to be on a much larger scale than any of the previous ones. While the armies of 1069 and 1075 had been transported in 300 and 200 ships respectively, this one was going to require more than 1,000 ships, with Robert the Frisian contributing a further 600, and the king of Norway also being involved.25 In the Vita Wlstani, Malmesbury says that William was frightened, a comment that should probably be translated as meaning that he recognized the scale of the potential threat and made appropriately thorough preparations to counter

For the crucial passages, OV, iii, 112–13. Ibid., 102–3, 108–9 (post multas et inutiles circumitiones), 112–13; iv, 80–1; GR, i, 502–3. For differing commentaries on whether Robert did seek to marry Matilda, Aird, Robert Curthose, 96, note 195. For a survey of Matilda’s relations with Gregory and his successors, Cowdrey, Gregory VII, 296–303. 24 ASC, ‘E’, 1085. 25 GR, i, 480–1. 22 23

458

william the conqueror

it.26 The quartering of mercenary troops with lay magnates, bishops, and at monasteries referred to in general terms in the ‘E’ Chronicle, John of Worcester, and the Abingdon Historia, and also alluded to in Ailnoth of Canterbury’s Vita of King Cnut, is described in an anecdote telling how Bishop Wulfstan successfully countered the drunkenness of the soldiers assigned to him at Worcester.27 The crucial piece of information that troops were billeted in the West Midlands presumably indicates that what was feared was an uprising throughout England and interventions from Wales and across the Irish Sea. Memories of the period from 1066 to 1072 remained strong, it appears. And not just with William and those around him. According to Ailnoth, a former monk of St Augustine’s, Canterbury, who had taken refuge in Denmark, and who was therefore as much parti pris as anyone else in relation to William, Cnut was responding to a desperate appeal from the English to be freed from the tyranny of William and his followers.28 If this has any significant basis in fact, it confirms a fear on William’s part that the victory of Hastings could still be overthrown. The Vita Wlstani says that William called together a meeting of magnates at which Lanfranc proposed the extensive billeting of troops. In this we see again the part that consultation with his supporters played in William’s regime and how important Lanfranc, now well into his seventies, remained as an adviser on secular as well as ecclesiastical matters.29 The ‘E’ Chronicle says that coastal lands were ravaged to obstruct a landing that was presumably expected to occur on the east coast after a crossing from Flanders. But if this was so, then the impact measurable from the evidence of Domesday Book was minor. Falling values in Essex coastal hundreds have been cited as a possible consequence, but, if they were, the extent of the destruction was small in comparison to what was done in south-eastern England in 1066 and in Yorkshire in 1069–70. This evidence can even be used to launch an argument about the possibly huge differences between ravaging as an aspect of warfare and as a form of political punishment.30 At the Christmas meeting at Gloucester, the elderly abbots of the two monasteries of Crowland and Thorney, both of them located in 26 Vita Wlstani, 130–1 (rex timore percitus). For the crisis of 1085, see now, John Maddicott, ‘Responses to the Threat of Invasion, 1085’, EHR, cxxii (2007), 986–97. 27 Historia Ecclesie Abbendonensis, ii, 16–17; Gesta Swenomagni regis et filiorum eius et Passio gloriosissimi Canuti regis et martyris, in Vitae Sanctorum Danorum, ed. M. C. Gertz (Copenhagen, 1908–12), 77–136, at 98–9; Vita Wlstani, 130–1. 28 Gesta Swenomagni regis et filiorum eius et Passio gloriosissimi Canuti regis et martyris, 96–8 (van Houts, ‘The Norman Conquest through European Eyes’, 837 and note 2). 29 Vita Wlstani, 130–1. On the importance of this intervention, Maddicott, ‘Responses to the Threat of Invasion’, 987–9. 30 For the statistics, R. Welldon Finn, The Norman Conquest and its Effects on the Economy: 1066–1086 (London, 1971), 251.

the final years

459

obviously sensitive military areas, were dismissed and replaced, even though neither of their abbeys formally owed military service. Three new bishops were appointed to vacant sees, one of them being the East Anglian bishopric of Thetford, and the others London and Chester. In every case the appointees were individuals of obvious secular competence, with all three of them having been royal chaplains.31 By early 1086, it must have been becoming clear that the invasion from Denmark was not going to happen. William therefore dismissed some of his army, but kept other troops with him throughout the winter, that is, presumably on until towards Easter. Although Malmesbury puts the invasion’s failure down to adverse sea conditions, the internal politics of the Danish kingdom and of the Empire had to take precedence for Cnut. After Gregory VII’s death in 1085 the German anti-king Hermann of Salm, whom Cnut had supported against Henry IV, had taken refuge in Denmark and Henry had moved north into Saxony to pursue him. Cnut also had to defeat a rebellion by his brother Olaf and try to counter opposition to the taxes he was levying, presumably in part to support the planned expedition. In addition, Cnut’s coinage reforms and his now lost equestrian seal suggest an unwisely ambitious programme of changes that may also have stirred up enmities. He was murdered in the cathedral of Odense on 10 July 1086, and subsequently canonized in 1099 as a saint and martyr.32 With the rebels of Sainte-Suzanne having also submitted by this date, William had escaped unscathed. But the scale of the demands on his English subjects during these times is made very explicit in the Chronicle, comments that lead on into its subsequent general assessment of William and his rule in his famous obituary that will be discussed in the next chapter. However, at the same time that these problems were being negotiated, big changes in the politics of papacy and empire were introducing other sources of concern. GREGORY VII’S LAST YEARS AND THE WIDER EUROPEAN WORLD While Bishop William of Durham’s visit to Gregory VII in 1082 had shown that some business continued as normal, Gregory’s increasing difficulties after his second excommunication of Henry IV in 1080 and the setting-up 31 ASC, ‘E’, 1085; CS, 632–4; The Anglo-Saxon Chronicle. A Collaborative Edition: 3. ms. A, 86–7; Maddicott, ‘Responses to the Threat of Invasion’, 991–5. 32 On the failure of Cnut’s invasion and the politics of the Danish kingdom, see most recently, Paul Gazzoli, ‘Anglo-Danish Connections and the Origins of the Cult of Knud’, Selected Papers from the Inaugural St Magnus Conference 2011, Journal of the North Atlantic, special vol. iv (2013), 69–76, at 70–1. See also Larson, ‘The Efforts of the Danish Kings’, 79–80; Bolton, The Empire of Cnut the Great, 313–14.

460

william the conqueror

of Archbishop Guibert of Ravenna as an anti-pope contributed to a lessening of contacts of a routine kind. As part of this normality Gregory returned more forcefully in the summer of 1082 to the matter of Lanfranc’s refusal to visit Rome, ordering him to be there by 1 November or be suspended.33 Since nothing further is known to have happened, Lanfranc, with William’s undoubted agreement, did not go. And Gregory took no action. As we have already seen, Gregory also remonstrated with William about the imprisonment of Bishop Odo. However, an entirely new situation was created after Henry’s armies had again entered Rome in 1084, with Archbishop Guibert’s enthronement as Pope Clement III on 24 March and Henry’s imperial coronation on 31 March creating a papal schism. Gregory fled to Salerno about three months later and died there on 25 May 1085. Although Lanfranc was approached by Clement and his supporters before Gregory’s death, as far as we know, none of them made contact directly with William. Gregory’s friends, on the other hand, did seek his aid and Bishop Anselm of Lucca, Gregory’s legate in Lombardy, wrote to him in what must seem astonishingly extravagant terms to ask that he come to Gregory’s help; William, so he wrote, was not just renowned in the western parts of the world, but among the peoples living in its most remote regions, truly remarkable statements from a papal supporter given how controversial William’s violent career was. The whole tone of the letter and what is known of Anselm’s thinking express the opinion that it is the responsibility of the material sword to repress the impious, an interesting restatement of arguments deployed by William of Poitiers to justify William’s conduct in 1066.34 Presumably at around the same time, Lanfranc replied to a letter written on Clement’s behalf from ‘Hu’, who is normally identified as Gregory’s long-standing enemy Hugh Candidus, cardinal priest of San Clemente. Responding to an appeal to support Clement, Lanfranc continued to refer to Gregory as pope and refused to become involved in anything that might seem to give support to Clement. He did, however, acknowledge that God must in some ways be favouring Henry because his armies were being so successful and he referred to Henry in eulogistic terms (gloriosus imperator). But Lanfranc concluded by saying that he did not favour ‘Hu’ visiting England without having first obtained William’s consent, something that presumably was never given.35

Registrum, ii, IX, 20 (Register, 419–20). Die Hannoversche Briefsammlung: i. Die Hildesheimer Briefe, no. 1, in Briefsammlungen der Zeit Heinrichs IV, ed. Erdmann and Fickermann, 15–17. On this episode, Cowdrey, Lanfranc, 202–5. On Anselm of Lucca and this letter, Kathleen G. Cushing, Papacy and Law in the Gregorian Revolution: The Canonistic Work of Anselm of Lucca (Oxford, 1998), 132. 35 Lanfranc Letters, no. 52. 33 34

the final years

461

Clement’s two surviving personal letters to Lanfranc asked him to visit Rome and to organize the payment of Peter’s Pence, repetitions of requests that Gregory had made and therefore tactical ploys designed to create legitimacy and to obtain recognition. Since no reply to either of them is known and given that their texts were deliberately inserted into Lanfranc’s personal canon law collection, if Lanfranc had replied to either he would surely have kept it as a record.36 William and Lanfranc’s policy was undoubtedly to stand back and wait on events, confident that they could oversee the affairs of the Church in England as they had always done. The development of a papal schism and Gregory’s death nonetheless made the relationship with the papacy more complex than it had been for decades. As is well known, the situation became much more fraught after William and Lanfranc had died. Although the undercurrents of diplomacy and the subtexts of letters indicate much turbulence, their deaths and Gregory’s earlier passing brought to an end almost forty years of largely friendly and collaborative relations that had mostly been conducted on William’s terms, but in ways and according to values that a succession of popes had generally felt able to support and sustain. These years do not contain much evidence of the development of wider European links that exists for the 1070s. If the story is true, as its author admits it might not be, the one incident that must date from this time is of Diego Paláez, bishop of Santiago, being deposed in 1088 for plotting to hand the kingdom of Galicia over to ‘the king of the English and the Normans’, a title that must refer to William and not to his son.37 Whether true or not, the story and the invitations from Gregory VII’s supporters indicate that William’s prestige remained as great as ever. The possible Spanish intervention is also, like the earlier possible marriage alliance for William’s daughter Adelida, a reference to contacts of a kind of which we know little, but which must have been more frequent than they appear. There may nonetheless have been a conscious narrowing of dynamic European-wide activity on William’s part.

36 For the texts, Felix Liebermann, ‘Lanfranc and the Antipope’, EHR, xvi (1901), 328–32, at 330–1. The third letter in the collection dates to after William’s death. 37 Historia Compostellana, ed. Emma Falque Rey, CCCM, vol. lxx (Turnhout, 1988), 221 (quod Gallicie regnum prodere regi Anglorum et Normannorum); Manuel Castiñeiras, ‘Didacus Gelmiras, Patron of the Arts. Compostela’s Long Journey: From the Periphery to the Center of Romanesque Art’, in Manuel Castiñeiras (ed.), Compostela and Europe: The Story of Diego Gelmirez (Santiago de Compostela and Milan, 2010), 32–97, at 38; cited in Fernie, ‘Three Romanesque Great Churches’, 16–17, 22, note 17.

462

william the conqueror

THE DOMESDAY SURVEY AND ENGLAND IN 1086 The Anglo-Saxon Chronicle famously records that at Gloucester in 1085 William ‘had great thought and very deep conversation with his council about this land, how it was occupied, or with which men’. Commissioners were sent out into every shire of the kingdom. And ‘not one ox, one cow, not one pig was left out’, an exaggeration of the survey’s comprehensiveness, but an understandable reaction to its thoroughness.38 The survey was concluded before William left for Normandy in the late summer or early autumn of 1086 (‘And all the records were brought to him afterwards’). For the writer of the Chronicle the scale of the operation was shameful. But William, he added, was untouched by any such sentiments. This comment is the first one of a series that this anonymous author made that are deeply critical of William’s rule and which culminate in the famous obituary that summed up his kingship. Their location within a litany of natural disasters, famines, fires, and disease suggests that the author is linking these manifestations of divine displeasure to the perceived failures of earthly authority and that they cannot be taken entirely at face value. They are nonetheless very specific. And they do place the great enterprise of what turned out to be William’s last visit to his kingdom within the same framework of ambivalence that characterizes so much that others wrote about him. The Chronicle further mentions that the survey was intended to discover how many hundreds of hides there were in each shire, how much land and livestock the king had, and what revenues he ought to receive every year from the shire. It then adds that the land of every landowner in the kingdom was surveyed, with a record made of its livestock and its value. The one other strictly contemporary account of the survey by Bishop Robert of Hereford, like the Chronicle, says that the land was much troubled by the collection of tax. His statement that a second set of commissioners was sent out to check on the findings of the first ones is, however, original and further indicative of an extremely thorough operation.39 His comments about tax suggest that another geld was taken while the survey was in progress. Many of the technical discussions associated with the conduct of the survey and the making of the Domesday manuscripts lie beyond the scope ASC, ‘E’, 1085. W. H. Stevenson, ‘A Contemporary Description of the Domesday Survey’, EHR, xxii (1907), 72–84, at 74. For convenient translations and comment, English Historical Documents 1042–1189, ed. David C. Douglas and George W. Greenaway, 2nd edn. (London and New York, 1981), 912; Harvey, Domesday, 90; Frank Thorn, ‘Non Pascua sed Pastura: The Changing Choice of Terms in Domesday’, in Roffe and Keats-Rohan (eds.), Domesday Now, 109–36, at 113–14. 38 39

the final years

463

of this biography, but their relationship to the questions of how and why something so huge and ambitious was thought possible and deemed to be necessary are central to it. In addition, while maintaining this book’s principle of standing back from historiography, some awareness of the range of the opinions that have been advanced, either over the thirty years since the novocentenary of 1086 or before it, is indispensable, since they have produced radically different perceptions of the politics of the times and of William’s kingship. Thus, for example, the survey has been seen as a product of political bargaining between William and the aristocracy, as a sign of a regime in difficulties, and as a mixture of an act of atonement, a publicity exercise, and a bid for ‘hearts and minds’.40 Interestingly, a recent commentary has managed to reject the first and retain the second and then to invest the whole exercise with immense significance for the present and the future.41 An area of common ground to all of these opinions is that the survey and Domesday Book were political and multi-functional, with the meaning of both of these words then being subject to different definitions.42 All also take the subject out of the domain of financial, legal, and administrative history where it was once thought to belong, while not necessarily rejecting that it may belong in some, or perhaps all, of these categories.43 They also see the so-called Salisbury Oath of 1 August 1086 as an essential part of the same processes. The way in which the early twelfth-century narrative sources offer different explanations of the survey’s purpose and content reinforces this emphasis on multi-functionality. Just as we do, they found both hard to pin down. Orderic, for example, while noting that the whole of England was surveyed, in one place focused on its retrospective character, writing that it included a very exact statement of dues owed to the king in King Edward’s day.44 In another, in the middle of his account of the siege of 40 For the first, J. C. Holt, ‘1086’, in idem (ed.), Domesday Studies, 41–64, at 64 (‘It indicates hard bargaining between William and his men, that the “deep speech” at Gloucester embodied a deal’). For the second, N. J. Higham, ‘The Domesday Survey: Context and Purpose’, History, lxxviii (1993), 7–21, at 11–16, 18 (‘an embattled regime’). For the third, Harvey, Domesday, 308–22. 41 David Roffe, ‘Domesday Now: A View from the Stage’, in Roffe and Keats-Rohan (eds.), Domesday Now, 7–60, at 28 (‘The crisis of 1085 was a threat to the common weal and its organisation and communal action was needed to meet it. The outcome was effectively a new social contract which was to define relations between the king and his subjects for the next hundred years or so’). 42 For the classic essay on Domesday’s multi-functionality that has shaped a lot of what has followed, Holt, ‘1086’, 41–64. For its versatility and relative ease of use, H. R. Loyn, ‘A General Introduction to Domesday Book’, in Ann Williams and R. W. H. Erskine (eds.), Domesday Book Studies (London, 1987), 1–21, at 7–14. 43 For an indispensable forward-looking assessment of scholarship up until the 1986 novocentenary, Elizabeth M. Hallam, Domesday Book through Nine Centuries (London, 1986), 11–31. 44 OV, ii, 266–7.

464

william the conqueror

Sainte-Suzanne and of the 1085 invasion scare, he wrote that William had a record of the kingdom’s military strength made and commanded that all of the 60,000 warriors he believed that it revealed should be ready to serve at his command.45 John of Worcester’s account said that it was a survey of what each of William’s barons possessed, of enfeoffed knights, as well as of peasants, animals, and dues.46 He also echoed the ‘E’ Chronicle by saying that the kingdom was much troubled by disasters that followed from the survey. From all this, it is, I think, quite reasonable to deduce that the uses of Domesday Book, both potential and actual, were probably not either fully defined or necessarily grasped by those who initiated and carried out the project. It is arguably one of the great ironies of the subject that it is Domesday Book which enables us to explore the charges made against its creator. It is, for example, central to any analysis of the destructive campaigns of 1066 and 1069–70. But it is also the unique foundation for an understanding of William’s practice of kingship and the expectations that his subjects had of him which goes beyond anything that any other source can provide. The references to royal writs as confirmations of landholding and the number of times when, even when one was not recorded, there was an expectation that there might be one, show just how deeply William’s authority as king penetrated into English local society. Thus, for example, at Sharnbrook in Bedfordshire, Almær, a burgess of Bedford, had had his father’s former tiny holding of one half of a virgate returned to him by William’s writ (per breuem suum reddidit), or so he claimed (Plate 22 (a)).47 If as many writs were provided as is implied by Domesday Book, then William must have been constantly besieged by petitioners, to many of whom he must have responded. Other evidence of this kind, selected from a multitude of examples, is a reference in Berkshire to the wife of Godric ‘who cared for the king’s dogs’; a priest named Alwine in Bedfordshire who performed a mass every Monday for William and Matilda’s souls (Plate 22 (b)); and, again in Berkshire, the case of a hundred court whose members sent a case to William to judge because they did not wish to comment on it before the commissioners.48 An illustration of how directly interventionist William’s presence was, either at first- or second-hand, is that thirteen of his chaplains appear scattered throughout the folios of Domesday Book, usually with quite small landholdings, as well Ibid., iv, 52–3. JW, iii, 44–5. 47 DB, i, fo. 218r. In general, Fleming, Domesday Book and the Law: Society and Legal Custom in Early Medieval England (Cambridge, 1998), 28–33, and for this specific case, 96, no. 62. 48 DB, i, fos. 57v, 58v, 218v (Fleming, Domesday Book and the Law, nos. 62 [p. 96], 71 [p. 97], 95 [p. 100]). 45 46

the final years

465

as many of his more humble lay servants.49 Occasionally a record of these transactions does find its way into charters, as in the case of William’s grant of churches in Cornwall at Launceston and Bodmin, and at Plympton in Devon, to his chaplain, the later bishop of Exeter, William Warelwast, but usually it does not.50 For all the considerable changes in personnel and roles that had occurred since 1066, this is at one level a manifestation of the well-known organizational penetration of English local societies by royal agents and their interaction with the king’s authority that had existed before 1066 and continued after it.51 But it also shows that, although William was so much an absentee, he was deeply absorbed into a culture of kingship that made any king the focus of so many personal expectations and a practitioner of what can look like micro-management. The same conclusion also follows from what I have already written about the work at crown-wearing centres, the king’s buildings, and the adjustment of landscape to suit his purposes. Without knowing, as we wish we did, any of the details of ‘the deep speech’ at Gloucester, we can surmise that there were some certainties that influenced the decision to proceed. One is the massively influential presence of those who were there. We can take for granted the political weight of the attendance at such a gathering. It is nonetheless worth emphasizing that one of the two charters that dates from the time names as present William’s two eldest sons, both English archbishops, Bishop Geoffrey of Coutances, Bishops Walchelin of Winchester, Osmund of Salisbury, and William of Durham, Count Robert of Mortain, and Count Alan.52 The second one also names Bishop William and adds Bishop Remigius of Lincoln to the list.53 While conjuring up an image of several sexagenarians or near-sexagenarians sitting down to decide on a great project, it is also worth dwelling for a moment on just how massively powerful this group was. William’s son Robert might have broken a spell of enforced or selfimposed exile to be there. Archbishops Lanfranc and Thomas and Bishops Walchelin and Osmund were precisely the group of bishops who crossed to Normandy in 1080 to discuss Gregory VII’s request for fealty.54 And Bishop William was going to play a central part in the proceedings that Lack, ‘De Obitu Willelmi’, 1,428–30. The Cartae Antiquae Rolls 11–20, ed. J. C. Davies, Pipe Roll Society, lxxi, new series, xxxiii (London, 1960, for 1957), no. 537 (Regesta Regum Anglo-Normmanorum, vol. ii, Regesta Regis Henrici Primi, nos. 841, 1391). 51 Campbell, ‘Some Agents and Agencies of the Late Anglo-Saxon State’, in Holt (ed.), Domesday Studies, 201–18, remains central. 52 Regesta, no. 156. 53 Ibid., no. 9. 54 Ibid., no. 175(I). 49 50

466

william the conqueror

were to follow. Robert of Mortain was William’s trusted brother and Count Alan had been in command at Sainte-Suzanne. This truly was the innermost circle. To suggest that much hard bargaining between king and subjects in advance of the survey was needed when such a group was in agreement strains belief. This was most certainly ‘communal action’, but whether it was the crisis of 1085 that pushed them into their big decision strikes me as being very doubtful. By Christmas they would have been beginning to know that Cnut’s plans were not going smoothly and that Hubert, if he had not submitted already, did not constitute the threat that had once been thought. And in any case, they had all seen this sort of situation before. The coincidence of cross-Channel and trans-North Sea problems may have focused minds. But there must have been other factors at work and these are likely to have been deeply embedded in the history of the conquest of England since 1066 and in an historical awareness of a past that stretched back at least to the birth of Christ. The sheer scale of the enterprise would have had to be discussed at Gloucester. The prior existence of documentation relating to tax and landholding and the interest of many religious communities in creating more of it must have been an influence. However the extensive Domesdayrelated documents that were produced are analyzed, a certainty that follows from their creation is that a cultural world existed that understood very well what was happening. It also shows that what was proposed could be done; it is once more a tribute to the resources inherited from the pre1066 English kingdom as well as to the inner group’s grasp of the possibilities that they provided.55 When statistical calculations are attempted, the figures for the number of people likely to have been involved in the collection and verification of information are truly astonishing. Thus, for example, it has been suggested that over 60,000 people were probably heard in court during the inquiry of 1086 and that between 20,000 and 40,000 people contributed local knowledge at some point.56 Such figures also indicate that thought must have been given to what was going to happen in the shires and the hundreds. It would have been well known that there were ‘bad boys’ out there, with some of the sheriffs normally identified as the villains. It was surely also recognized that voices were going to be heard that normally either went unheard or whose audience never transcended the local. For all that the basic accounts of the purpose of the 55 For a recent discussion of many Domesday Book issues, David Roffe, ‘Domesday Now: A View from the Stage’, in Roffe and Keats-Rohan (eds.), Domesday Now, 7–60. 56 David Roffe, Domesday: The Inquest and the Book (Oxford, 2000), 123; Harvey, Domesday, 312–13.

the final years

467

survey emphasize an inventory of resources, with it sometimes feeling as if people were being treated as such, the indissoluble link between the payment of tax and title to land would have been taken for granted. The latter required the former. For this reason, it is wrong to attempt to distinguish between tax and title as a motivation for the survey. It was allinclusive. It took place within a culture of reciprocity, of give and take. It was also being conducted within communities in which recovery from the traumas and human disasters that had followed 1066 would have been fresh in people’s minds. Although debated in detail and ultimately dependent for its formulation on the internal evidence of Domesday Book, it is widely accepted that the kingdom was divided into circuits and that commissioners were dispatched, usually consisting of a bishop and major barons, who, so Bishop Robert of Hereford says, were not major landholders in the shires for which they were responsible.57 In the case of the West Midlands circuit that included Worcestershire, they are known to have been Bishop Remigius of Lincoln, Walter Giffard, Henry de Ferrières, and Adam fitz Hubert, with the first two laymen being exceptionally wealthy tenants-in-chief, and the third, a well-endowed former tenant of Bishop Odo of Bayeux and the brother of a prominent member of William’s household, Eudo dapifer.58 Landholders certainly prepared material in advance and testimony on oath was taken throughout the kingdom from local assemblies of jurors from every shire, hundred, and, in the counties of the former Danelaw, wapentake. Whether all this was done at the first stage of the inquiry or the second is the subject of debate.59 The personnel known to have been responsible confirm yet again the centrality of bishops and a small trusted elite to William’s rule in England. The use of the word descriptio in the documents produced by the survey identifies it as a record of a kind with which contemporaries and previous generations would have been familiar. Meaning ‘a writing down’, it is the 57 Seven circuits is the most common number, but nine have recently been proposed, Harvey, Domesday, 87–90. For new ways of looking at the circuits, J. J. N. Palmer, ‘A Digital Latin Domesday’, in Roffe and Keats-Rohan, Domesday Now, 61–80, at 63–75. 58 For a convenient text and translation of the Worcester text that records the commissioners’ names, English Lawsuits from William I to Richard I, ed. van Caenegem, i, 40. For a summary of the evidence, Roffe, Decoding Domesday, 71–4; idem, ‘Domesday Now: A View from the Stage’, 24. 59 For a recent statement of the argument that the first stage was a local inquiry, with the commissioners coming later, David Roffe, ‘A Profession of Ignorance: An Insight into Domesday Procedure in an Early Reference to the Inquest’, in Paul Dalton and David Luscombe (eds.), Rulership and Rebellion in the Anglo-Norman World, c.1066–c.1216: Essays in Honour of Professor Edmund King (Farnham and Burlington, VT, 2015), 45–60, at 55–6.

468

william the conqueror

term used in the colophon of what is called Little Domesday Book, the written version of the shires of Norfolk, Suffolk, and Essex, and remained current as the term often used to describe Domesday Book until the 1120s (Plate 22 (c)). It had a long history behind it in 1086.60 This conservative language makes the point that the survey and Domesday Book have a place within the documentary traditions of western Europe, and in the clerical and biblical heritage of which the religious men around William must often have spoken. While it continues the patterns of thought and the grandiloquent expressiveness that had been a feature of William’s rule since before 1066, the use made of a cultural world inherited from the past and very much alive in the eleventh-century present is extremely important. Not only are inquests known from England before 1066, inquests and surveys of an analogous kind had their place in Roman and Carolingian times in a tradition of descriptiones going back to Caesar Augustus and passing through the surveys associated with the Treaty of Verdun (843). Charlemagne had commissioned a survey of conquered lands in Istria that included reference to conditions under the previous rule of the Byzantine emperors.61 All of these were made at times of crisis, but the purpose of the operation went much further than immediate crisis-management. Their essential aim was to record the present in order to stabilize the future. Anyone who tries to understand the making of Domesday Book surely has to remember the opening verses of Luke 2:1 (‘In those days Caesar Augustus issued a decree that a census should be taken of the entire Roman world. And everyone went to their own town to register’) and then note modern developments that have increasingly closed the so-called tenthcentury gap between the Carolingian surveys mentioned above and the late eleventh century.62 Also, the Carolingian surveys known as polyptychs would have been familiar to those around William, perhaps through their many links into Lotharingia. One of them, the ninth-century Polyptych of 60 Most recently, Sally Harvey, ‘ “A Deed without a Name” ’, in Roffe and Keats-Rohan (eds.), Domesday Now, 277–88, at 277–8; David Roffe, ‘Talking to Others and Talking to Itself: Government and the Changing Role of the Records of the Domesday Inquest’, ibid., 289–303, at 302, for a different emphasis. 61 John Percival, ‘The Precursors of Domesday: Roman and Carolingian Land Registers’, in P. H. Sawyer (ed.), Domesday Book: A Reassessment (London, 1985), 5–27; H. R. Loyn, ‘The Beyond of Domesday Book’, in Holt (ed.), Domesday Studies, 1–13, at 9–11; R. H. C. Davis, ‘Domesday Book: Continental Parallels’, in Holt (ed.), Domesday Studies, 15–39; Stefan Esders, ‘Regionale Selbstbehauptung zwischen Byzanz und dem Frankenreich: Die inquisitio der Rechtsgewohnheiten Istriens durch die Sendboten Karls der Großen und Pippins von Italien’, in Stefan Esders and Thomas Scharff (eds.), Eid und Wahrheitssuche. Studien zu rechttlichen Befragungspraktiken im Mittelalter und früher Neuzeit, Gesellschaft, Kultur und Schrift Mediävistiche Beiträge, vii (1999), 49–112, at 77–8. 62 Janet L. Nelson, ‘Henry Loyn and the Context of Anglo-Saxon England’, The Henry Loyn Memorial Lecture for 2006, HSJ, xix (2007), 154–70, at 164–9.

the final years

469

Irminon, was still in use in the late eleventh century as a working document at the Parisian abbey of Saint-Germain-des-Prés; this was within an easy day’s journey of Saint-Denis with which Abbot Baldwin of Bury St Edmunds as a former monk of the abbey certainly had links, and as William and others around him must have also done since, as we have seen, he had commissioned the construction of a bell-tower at SaintDenis.63 Much more detailed than Domesday Book in its treatment of individual estates, the Polyptych of Irminon exemplifies the traditions on which the makings of the inquest and the survey drew. The well-known point that Harold’s kingship is only once referred to, and this obliquely and seemingly accidently in the entry for Soberton (Hampshire), is just one factor that locates Domesday within the statement of legitimacy that was starting to be created well before the Battle of Hastings.64 The continuum that informs the entire Domesday text is the one from King Edward’s day (tempore regis Edwardi). Harold is therefore consistently referred to as an earl (comes), his status at that time, and his time as king of the English is written out of history as if it had never existed. There is in this a ruthlessness every bit the equal of anything else that either William himself or those acting in his name had ever accomplished.65 Thus, for all its appearance of objectivity, Domesday has as much of a contemporary character as the so-called literary sources that have featured throughout this book. But by 1086, with victory long assured, it is doubtful that this was much more than formulaic. It comes at the end of a long process of imposing domination. It is the social interactions and exchanges around the survey that were transferred into Domesday Book by the unnamed scribe who wrote it that surely constitute its rationale. For all that the narratives concentrate on resources and dues owed to the king, it was the multiple human investments in these that matter as well. ‘Hard bargaining’ there may not have been. But the dialogues around property, tax, and mutual obligation were constant sources of debate, and were certain to continue to be. It is this that must lie behind the Anglo-Saxon Chronicler’s statement that William wanted to know how the land was occupied and with what men. When the performance of the survey is translated into human experience and the mostly unknowable life histories of many among those who gave sworn testimony, and into the multiplicity of situations that the survey 63 Das Polyptychon von Saint-Germain-des-Prés, ed. Dieter Hägermann (Cologne, 1993), 28, 36–9, 138, 148; Harvey, Domesday, 38–41; Bates, ‘The Abbey and the Norman Conquest’, 18–20; Robert F. Berkhofer III, Day of Reckoning: Power and Accountability in Medieval France (Philadelphia, PA, 2004), 112. 64 DB, i, fo. 38r (quando regnabat abstulit ei). 65 For this argument, fundamental is Garnett, Conquered England, 1–44.

470

william the conqueror

created, many must have had to affirm the legitimacy of events that had often involved their own impoverishment, or the deaths, ruin, or disparagement of those they knew, or had known.66 The pressures on these individuals were multiple and sometimes huge, and there must have been many occasions when they decided that discretion was the better part of valour.67 In Sally Harvey’s words, ‘other cases could be heart-rending if visualised’, citing the cases of two women who offered to go to the ordeal to get back relatively small areas of land.68 It was perhaps all this that most shocked the writer of the ‘E’ Chronicle. Yet the record set out in Domesday can also be humane. There is the shock recorded at the condition of Æthelric, who held at Marsh Gibbon in Buckinghamshire of William fitz Ansculf. Æthelric had held it in King Edward’s time and in 1086 he held it at rent, overburdened and miserably.69 And then there is the reference to an unnamed Breton at North or South Pickenham in Norfolk who had married an Englishwoman because he loved her, an entry which does, however, then add that he had occupied the land without royal authority.70 These very experiences of a world where many had suffered terribly and where opportunism had been at a premium were at the heart of the process that was setting out to transform conquest into government. From a top-down perspective, and probably from a good number of bottom-up ones as well, the carrying out of the survey could be viewed as bringing communities together in an acknowledgement that a regime imposed by violence had secured its victory and was going to last, a landmark in the creation of a new world and a launching-pad for what it was hoped was going to be a better one. Viewed idealistically, the traumas of the past were being absorbed into a new and peaceful present symbolized by a record of resources and rights to which all of the free population had in theory agreed by taking oaths, or having their representatives do so. Defining the survey and the making of Domesday Book in these terms means that we are moving into scenarios that are very familiar in modern times, where societies try to move out of conflicts and civil wars and where reconciliation is an unpredictable evolutionary process, a situation that exactly parallels the one that those who had lived through the conquest of England and

66 For effective statements of this point, J. C. Holt, ‘Introduction’, in idem, Colonial England, xiii–xiv; Sally Harvey, ‘ “A Deed without a Name” ’, in Roffe and Keats-Rohan (eds.), Domesday Now, 277–88, at 288. 67 Alan Cooper, ‘Protestations of Ignorance in Domesday Book’, in Robert F. Berkhofer III, Alan Cooper, and Adam J. Kosto (eds.), The Experience of Power in Medieval Europe, 950–1350 (Aldershot, 2005), 169–81. 68 Harvey, ‘ “A Deed without a Name” ’, 284–5. 69 DB, i, fo. 148v. 70 Ibid., ii, fo. 232r.

the final years

471

their heirs had experienced, and were experiencing. We also know from modern times that such processes are never fully successful, and sometimes do not work at all. Among many avenues through which this subject can be explored, the lists of jurors inserted into the Domesday-related Inquisitio Comitatus Cantabrigiensis and Inquisitio Eliensis is one, with a remarkable number of those with French Christian names having assumed English toponyms, thus pronouncing themselves as being in some respects integrated into new societies on the English side of the Channel. Another is the likely extent to which lords could manipulate local testimony or intimidate jurors, just as Picot the sheriff had once done in the case of the Bishop of Rochester’s claim to Isleham before Odo of Bayeux intervened.71 In these terms, the survey and the making of Domesday Book become truly multi-functional and versatile and, from the perspective of William and those around him, a product not just of the claim to legitimacy, but of an act whose comprehensiveness and theatrical solemnity made all the free classes of society accomplices and participants in the new world. This argument also explains several of the processes that were involved in the conduct of the survey and the making of Domesday Book. Thus, the modern trend demonstrating how particular individuals and institutions influenced the Domesday record through their own written submissions places it within the culture of pancartes and confirmation charters and the politics of securing William’s confirmation of property rights that might be in dispute.72 In every case, the individuals and the institutions were seeking to obtain a formal confirmation most advantageous to themselves, while William and those advising him were aiming to protect him and themselves against further claims and to resolve disputes. Illustrative of this culture is the already mentioned story of how, when Anselm, abbot of Le Bec, crossed the Channel to ask William to confirm a charter listing the English possessions of the abbey of Le Bec – perhaps, although not necessarily, in 1086 – the king, so we are told, refused to confirm the charter at Easter because not all the donors were present to testify to their grants, with the result that the confirmation was postponed until Whitsun.73 The final outcome, as with Domesday Book, was a document to which all Lewis, ‘The Domesday Jurors’, 20–1, 23–5; Fleming, Domesday Book and the Law, 17–28. Stephen Baxter, ‘The Representation of Lordship and Land Tenure in Domesday Book’, in Elizabeth Hallam and David Bates (eds.), Domesday Book (Stroud, 2001), 73–102, at 82–93; Bates, ‘The Abbey and the Norman Conquest’, 18–19; Pamela Taylor, ‘The Episcopal Returns in Domesday’, in Roffe and Keats-Rohan (eds.), Domesday Now, 197–217. 73 Sancti Anselmi Opera Omnia, i, ep. no. 118 (The Letters of Saint Anselm, trans. Fröhlich, i, no. 118). For the charter and the problems of dating, Regesta, no. 167. From what we know of William’s itinerary, 1081 and 1086 are possible years. 71 72

472

william the conqueror

interested parties had agreed – or at least had given the impression of having agreed. More business would inevitably follow, but from a basis that all could acknowledge. This one example inserts Domesday Book very firmly into the cultural world of the pancarte. It was not a register of title as such, but it was a record from which discussion of title could proceed.74 What makes Domesday Book so remarkable, of course, is that the process involved an entire kingdom, not just the estates of a single monastery or church. But given that monasteries had lay magnates as patrons and protectors, they too would have been intimately familiar with what was being done; it was a product of a world that they would have understood very well and the next stage of whose politics those with savvy would have been ready to tackle. It was also, in a different format, a continuation of the processes epitomized by the many pancartes confirmed in Normandy in the late 1070s and early 1080s, and of the writ dispatched to England at the time of the dedication of Bayeux cathedral in 1077 setting up a commission to inquire into the conduct of sheriffs.75 So viewed, it also answers the often asked question as to why there was no Domesday Book made for Normandy. To which the answer is that in a relatively settled society, there was no need for an inquest and a formal record to set out what social and sometimes written memory could deal with. Domesday Book’s role in recording the transfer from an English predecessor to a Norman or French successor also has to be a crucial element in the rationale for its creation. This does not, however, mean that it was a record of the terms on which transfer had been made. Rather it was a record of an almost infinite number of transmissions of property that could have been accomplished by a multiplicity of means. Above all, and crucially in terms of the strategy behind the creation of the new power structures, a proven royal grant could override any other procedure. Equally crucially, a transfer from an English landholder to a newcomer would frequently have involved passing on the disputes associated with it. For this reason, Domesday Book seems to repeat endlessly cases where the identity of an individual’s predecessor (antecessor) was unclear or disputed. Placing such cases on the record might be deemed therapeutic within local societies, though at the same time the many references to the use of violence may suggest that the process sometimes had no such effect. All 74 For this point, see also Paul R. Hyams, ‘ “No Register of Title”: The Domesday Inquest and Land Adjudication’, ANS, ix (1987), 127–41, at 139 (‘Exchequer Domesday may indeed have been intended as the “register of title” that Maitland said it was not, or perhaps rather as a giant Pancarte or general royal confirmation’), citing an opinion expressed by the author of this book. 75 Regesta, no. 129.

the final years

473

would certainly fit with a world of worldly-wise newcomers settling into new local, regional, and national societies and trying to turn situations to their advantage. They also belong within a culture that wanted to define the process of transfer as orderly, even if it sometimes was not. The often alluded to cultures of feud and honour-based violence did after all have their place here as well. Hence, Domesday Book’s frequent references to violence or to landholding for which there was no apparent warrant.76 The sheer complexity of the lands and rights that had been taken over, although no more complex sociologically than the multi-layered ones with which the newcomers were familiar in their homelands, meant that considerable ingenuity was required to define them. On this subject, recent work on the nature of English society has significantly adjusted perspectives.77 The creation of a literary version of the survey returns involved the scribe who wrote Domesday Book and his advisers in considerable creativity. The term antecessor applied to the transmission of lands and rights was imported from canon law and took on meanings that it had not previously had.78 Similarly the word ‘manor’, in Domesday Book manerium or less often mansio, relatively recently invented to try to supply a word to define the phenomena involved in English landholding, was brought in to try to categorize for administrative purposes.79 Neither had been used in either Normandy or England before 1066. For all that Domesday Book was intended to be a conservative record that reinforced a very fluid status quo, transforming the survey returns sent to William into a kingdom-wide record unleashed imaginations in ways that were sometimes to have important consequences. But we must be careful to dissociate such flights of fancy from William. Whatever opinion is taken of the writing-up of the returns to make Domesday Book, it was certainly done either after he had crossed to France or after his death. The evidence of William’s itinerary in 1086 has him move around the south of England while the survey was in progress; this is in fact the only year in which he actually fulfilled the pattern of crown-wearings described 76 For a general approach similar to mine, Green, The Aristocracy of Norman England, 48–99. Roffe, ‘Domesday Now: A View from the Stage’, 37–41, provides a good critical assessment of the debates of the last thirty years. See further Ann Williams, ‘Hunting the Snark and Finding the Boojum: The Tenurial Revolution Revisited’, in Roffe and Keats-Rohan (eds.), Domesday Now, 155–68. 77 Thus, e.g., Stephen Baxter, ‘Lordship and Justice in Late Anglo-Saxon England: The Judicial Functions of Soke and Commendation Revisited’, in Stephen Baxter, Catherine Karkov, Janet L. Nelson, and David Pelteret (eds.), Early Medieval Studies in Memory of Patrick Wormald (Farnham, 2009), 383–420. 78 Garnett, Conquered England, 24–33. 79 C. P. Lewis, ‘The Invention of the Manor in Norman England’, ANS, xxxiv (2012), 123–50.

474

william the conqueror

in the ‘E’ Chronicle. After spending Christmas 1085 at Gloucester, he was at Winchester for Easter (5 April) and Westminster for Whitsun (24 May). This created an availability that must have been crucial to the whole enterprise. His travels also took him at some point in or after April to Lacock in Wiltshire, perhaps soon after the Winchester crown-wearing, or perhaps as the court was gathering for the great assembly that took place at Salisbury on 1 August. This visit to Lacock, an occasion at which a dispute between the abbey of Fécamp and William de Briouze about their respective rights and possessions at Steyning in Sussex was settled, shows William visiting an estate of one of his great lords, William, the son of Robert, count of Eu. It was also a gathering with a remarkably prestigious attendance that included William’s sons William and Henry, both English archbishops, eight bishops, one of whom was Geoffrey of Coutances and another the known Domesday commissioner Remigius of Lincoln, and the likes of Robert, count of Mortain, Roger de Montgommery, and Count Alan; in other words, the same group that had been present at Gloucester at Christmas.80 The presence together of so many powerful people as William travelled around the south of his kingdom again shows that the work on the survey that made Domesday Book was a massive collaborative exercise which kept the political class in regular contact with each other. Given the role of bishops in the survey, it suggests that the Lacock meeting took place either just before several of them set out to do their work in the shires and hundreds or as they were coming together to discuss the results. It is another one of numerous possible illustrations of how many people must have been involved at the centre and in the localities in the making of the survey and with associated projects, of which another pertinent illustration is the involvement of Salisbury scribes in the writing of the pre-Domesday text known as Exon Domesday.81 But the Lacock assembly and the great gathering at Old Sarum for the so-called Salisbury Oath on 1 August are, most importantly, a demonstration of the collaborative working of the elite that had sustained William throughout his life. Collaborating with bishops had been a top priority in Normandy before 1066 and in England from immediately after his coronation. The whole episode is a landmark in the collective biography of the often long-lived but nonetheless new – by this stage ‘newish’ may be a better word – elite of the English kingdom and of the cross-Channel empire. Regesta, no. 146. For the scribes, see Teresa Webber, Scribes and Scholars at Salisbury Cathedral c.1075–c.1125 (Oxford, 1992), 16, 17. A major re-evaluation of Exon is currently being made by the AHRC project on which Professor Julia Crick is the Principal Investigator. 80 81

the final years

475

It was perhaps at the Winchester crown-wearing that William is supposed to have confronted Bishop Walchelin of Winchester about the way in which he had interpreted William’s grant of as much timber as Walchelin could have collected from Hempage Wood (Hampshire) in three days and nights to contribute to the building of his new cathedral. Assembling, so we are told by the author of the Annals of Winchester, a huge number of carpenters (Carpentariis igitur innumerabilis congregatis), Walchelin had managed to have the whole wood cut down in the allotted time. Unfortunately, William’s bewildered and angry reaction recorded in some marvellous direct speech in the annals is probably a late twelfthcentury invention most likely attributable to Richard of Devizes.82 At the Westminster crown-wearing, William had his youngest son Henry knighted, thereby marking his entry into adulthood at approximately eighteen years of age.83 And at Salisbury, so both Domesday and Exon tell us, he ordered that dues and services belonging to the bishop of Winchester’s manor of Taunton (Somerset) should be confirmed to Bishop Walchelin, with Bishop William of Durham ordered to make sure that they were written into the returns (in breuibus).84 The way in which they are entered may well have important implications for the making of Domesday. Their importance here, however, is to underline again how deeply William was personally involved in the whole process. At Salisbury, that is, at the place we now know as Old Sarum, according to the ‘E’ Chronicle, William took oaths from all the chief men of England, including everyone of any account and not only those who held land of him, but also many who held land of others, having all of them swear that they would be loyal to him against anyone else. John of Worcester and Henry of Huntingdon, both dependent on a version of the Chronicle, also stressed the size of the gathering, but paraphrased its account to have it comprise an oath of allegiance against all men and an act of homage by all the landholders of England, of whomever they held their land, and the performance of oaths of fealty.85 However it is interpreted, the sheer scale of this operation as well requires emphasis, since the numbers present must have run into multiple hundreds. And it took place at a site that owed 82 Annales de Wintonia, 34–5. For the composition of the annals and their use of direct speech, John T. Appleby, ‘Richard of Devizes and the Annals of Winchester’, BIHR, xxxvi (1963), 70–7, at 73. For a consideration of the issues of authorship, Antonia Gransden, Historical Writing in England, c.550 to c.1307 (London, 1974), 252–3. 83 ASC, ‘E’, 1086; GR, i, 710–11. 84 DB, i, fo. 87v; iv, fo. 175v. For discussion and facsimiles of these passages, Pierre Chaplais, ‘William of Saint-Calais and the Domesday Survey’, in Holt (ed.), Domesday Studies, 65–77, at 75–6, plates III (b), (c), (d). See also Harvey, Domesday,113. 85 ASC, ‘E’, 1086; JW, iii, 44–5; HH, 402–3.

476

william the conqueror

its existence to the new regime and which was only in the early stages of development, with the new cathedral started there in 1075 within a huge Iron Age hill-fort after the site of the cathedral church had been moved there from Sherborne.86 The Salisbury Oath’s significance in relation to the Domesday survey has been much discussed. It is actually possible to believe that it was there the returns were brought to William. It is certainly not possible to believe that there was not some sort of relationship between the making of Domesday and the oath, since both, for all their roots in documentary and social culture of their times, were without precedent in terms of scale. The key must surely lie in their roles as landmarks in the consolidation and transformation of conquest. Domesday’s multi-functionality resided in its status as a record of landholding resources and royal rights and as a record of landholding that looked backwards and forwards. On these terms, it was not formally confirming everyone’s landholdings since, in a very high proportion of cases, the question of who held what was uncontroversial. No such thing was necessary. But it was creating a record against which all could be assessed if required. Hence, if there was a dispute, then Domesday was in no way an obstacle to it being pursued by anyone with the will to do so; it is possible to cite instances of disputes recorded in Domesday that were wrapped up very soon after the survey was completed, as if it had been the catalyst for their resolution.87 It is also, however, possible to cite cases, such as the abbey of Glastonbury’s claim to Uffculme in Devon, where a dispute was not registered in Domesday, even though William may himself have played a part. With its roots in the tenth century, this one was still ongoing in the mid-twelfth century.88 In a similar way to the survey and the making of Domesday Book, while the oath did no more than confirm formally the status quo, the fact of it happening was a dramatic demonstration of majesty. It has its affirmation, and perhaps even a record of its content, in the second capitulum of the Articuli Willelmi (‘We lay down that all free men shall affirm by fealty and oath that they are willing to be faithful to King William and preserve his lands and honour in all fidelity and defend him against his enemies both

86 Richard Gem, ‘The First Romanesque Cathedral at Salisbury’, in Eric Fernie and Paul Crossley (eds.), Medieval Architecture and its Intellectual Context (London, 1990), 9–18; Fernie, The Architecture of Norman England, 152–3. 87 David Bates, ‘Two Ramsey Writs and the Domesday Survey’, HR, lxiii, (1990), 337–9. 88 H. P. R. Finberg, ‘Uffculme’, in W. G. Hoskins and H. P. R. Finberg (eds.), Devonshire Studies (London, 1952), 59–77; Lesley Abrams, Anglo-Saxon Glastonbury: Church and Endowment (Woodbridge, 1996), 236–8.

the final years

477

within and without England’).89 Not only were many of the sub-tenants named in Domesday Book immensely wealthy and therefore effectively responsible for the governance of their localities, but the way in which the new regime had been constructed, with a small elite having massive crossChannel estates, made them into indispensable components of rule. The oath did not just control them – although, of course, it did – because it affirmed a solemn relationship which encapsulated responsibility that had to be reciprocal, it also flattered everyone by bringing them into such a close association with kingly power. Grand claims have been made for the Salisbury Oath in terms of creating new relationships and new forms of rule. From a biographer’s perspective, the Domesday survey and the Oath confirm William’s capacity for performance, a theme that started to come to prominence during the 1050s and which was sustained for the rest of his life. However, in 1086 it was the theatre that mattered more than anything. In the immediate term, I would doubt whether legally either Domesday or the Salisbury Oath did much at all.90 In relation to customary rights in free property and the relationships that affirmed them, Domesday’s language is conservative; the crucial usage of the verb tenere continued practice in Normandy before 1066.91 And, although Domesday can give the impression that title in property only went back to ‘the time of King Edward’, it was not creating an absolute prescription that this was so. As we have already seen, for example, in relation to the Penenden plea and the activities of Abbot Baldwin of Bury St Edmunds, raising disputes whose origins lay before King Edward’s day was certainly practised when it was needed to make a case. The kingdom surveyed in 1086 was being located within a much longer historical continuum than the period between 1066 and 1086. At the heart of all this was the notion of William as patronus and rex present from 1066, a projection that encapsulated immense personal responsibility and power. It and Domesday were demonstrations of rule that consolidated and sought to lay a new foundation for social and personal relationships including the inter-ethnic. Although I cannot share David Roffe’s argument that the manuscript we know as Great Domesday Book was a later conceptually independent product written up after the survey in William II’s reign, Roffe’s readiness 89 Ten articles of William I, in Die Gesetze der Angelsachsen, ed. Liebermann, i, 486 (Statuimus etiam ut omnis liber homo foedere et sacramento affirmet, quod infra et extra Angliam Willelmo regi fideles esse uolunt, terras et honorem illius omni fidelitate cum eo seruare et ante eum contra inimicos defendere). 90 For different perspectives, Holt, ‘1086’, 57–64; Roffe, Decoding Domesday, 179–82. 91 For these complex debates, Susan Reynolds, ‘Tenure and Property in Medieval England’, HR, lxxxviii (2015), 563–76, at 546–6. I return to some of these issues in Chapter 13.

478

william the conqueror

to distinguish between the Inquest and the Book is exceedingly helpful.92 The first was what William ordered to be done at Gloucester and the second was undoubtedly what he decided should be done either then or during the intense discussions that must have been going on over the period of the approximately nine months that lie between Christmas 1085 and his departure for Normandy in the late summer or autumn.93 The detail of the writing did not concern him and the end product took on a life of its own. But what was done in 1086 does belong within the themes of protection, confidence, reward, and trust, to which I will return in the next chapter and the Epilogue. Much was summed up in Sir James Holt’s concluding words in his lecture to the 1986 novocentenary conference: ‘As such it was not intended to “do” anything. It simply “was”.’94 But it must be added that the making of Domesday Book did a huge amount for many people. It is the ultimate monument to William’s belief that he was a responsible ruler and to the attempts of Lanfranc and others to shape him as one. And also to Orderic’s opinion that what happened after 1066 was a human catastrophe. It cannot be a coincidence that William’s intervention in the redevelopment of York seems to have taken place in 1086 or shortly before 1086. Although he certainly did not travel to northern England, the encouragement given to Count Alan to establish monks just outside the city in the church of St Olaf and, afterwards, to create the basis that led to the foundation of St Mary’s abbey, dates from around this time. As also does the development of what became the hospital of St Leonard’s.95 All this, taken together with other developments in and around York, has recently been used as the grounds for the argument that William began a process intended ‘to turn York into a monumental royal capital in the north’, a continuation, in other words, of processes initiated earlier at Winchester and Westminster and, later at Gloucester, with analogous Christian and Roman elements present.96 In this case, he was also contributing to the renewal of a city for whose misfortunes he was partly responsible.

92 Roffe, The Inquest and the Book, 242–8. His arguments have recently been restated in answer to criticism in ‘Domesday Now: A View from the Stage’, 28–9. 93 For the writing-up of the returns into Great Domesday Book as a single process, Frank and Caroline Thorn, ‘The Writing of Great Domesday Book’, in Hallam and Bates (eds.), Domesday Book, 37–72, at 69–72. 94 Holt, ‘1086’, 64. 95 For the sources conveniently assembled, Rollason et al., Sources for York History, 200–4. For St Leonard’s, Regesta, no. 8. For some major aspects of the rebuilding, Harrison and Norton, ‘Lastingham and the Architecture of the Benedictine Revival in Northumbria’, 68–9. 96 Rees Jones, York, 98–100, with the quotation at 98.

the final years

479

Another change that may well have taken place during this visit was the replacement of Earl Aubrey in Northumbria after he had resigned his office and returned to northern France. As was customary for its demanding author, the writer of the Durham Historia Regum dismissed him as being of little use in difficult matters.97 While we know almost nothing about what he might have done during his years as earl beyond a reference to Aubrey confirming Durham’s possession of Tynemouth after the monks had been introduced into Durham by Bishop William, there is no evidence that anything went seriously wrong.98 It is possible that Bishop Geoffrey of Coutances was then given general oversight of Northumbria, although, given the scale of his commitments, it is likely that the actual exercise of authority was delegated to his nephew Robert de Montbrai (‘Mowbray’), who eventually appears as the next earl. Robert had been one of the young men who had joined Robert Curthose during his first exile and, according to Orderic’s character sketch, was distinguished by great physical strength and military prowess, an insubordinate nature, and the absence of any sense of humour.99 He held office on into William Rufus’s reign and within ten years had been responsible for the killing of the King of Scots and the organization of a major rebellion against Rufus. After describing the Salisbury Oath, the ‘E’ Chronicle says that William travelled to Normandy via the Isle of Wight, taking with him as much money as he could, as was his custom. It adds a comment that it was his practice to obtain money by whatever means he was able, whether just or unjust. This is followed in the annal for the next year by a savage indictment of William’s regime: The king and the principal men greatly loved, and over-greatly, greed in gold and silver, and did not care how sinfully it was got as long as it came to them. The king granted his land on such hard terms, the hardest he could. Then a second came and offered more than the other earlier gave, and the king let it go to the man who offered him more. Then a third came and offered yet more, and the king let it go into the hands of the man who offered him most of all, and did not care how sinfully the reeves got it from wretched men, nor how many unlawful things they did; but the greater the talk about just law, the more unlawful things were done. They levied unjust tolls and they did many other unjust things that are difficult to relate.100 HRA, 199 (Quo in rebus difficilibus parum ualente). Ibid., 261; Aird, St Cuthbert and the Normans, 101–2, 136–7, note 160. 99 For this suggested succession, Aird, ‘Mowbray, Robert de, Earl of Northumbria (d.1115/1125)’, ODNB. For Orderic and Robert, OV, iv, 278–81. 100 ASC, ‘E’, 1086. I have used Michael Swanton’s translation. 97 98

480

william the conqueror

Should we accept this in every detail? As we have already seen, the writer was seeking an earthly explanation for the miseries of his times. The association of financial exactions and the crossing to Normandy is nonetheless important. It suggests that the costs of defending Normandy were being transferred on to England, a statement that anticipates William of Malmesbury’s terrible interpretation of the prodigy of the conjoined twins, one of them a corpse (Normandy) sucking the living body (England) dry, their very existence being an affront to nature.101 This and much else constitute some of the themes of the next chapter. It must also be significant that the Chronicle then added that Edgar the Ætheling left England for Apulia because he was not being treated with the honour due to him. John of Worcester is, however, less damning of William, saying that the Ætheling went with William’s permission and that he took 200 knights with him. But that he went to Apulia is of course an indication of his personal initiation into the new Norman world in which he had lived for almost twenty years. Both the Chronicle and John also record that his sister Christina became a nun.102 The interest in Edgar and his sister continues their interest in the fate of the descendants of the former line of English kings who would, in the annal for 1100, be referred to by the then writer of the Chronicle, who was not necessarily the same person, as ‘the rightful royal family of England’.103 THE RETURN TO NORMANDY (1086–7) The date at which William crossed the Channel to Normandy is not known. Although he of course did not know it, he was leaving England for the last time. It being likely that the processing of the Domesday returns was lengthy, he must surely have stayed on for a while. There is a record of a reduction of the geld liability of an estate belonging to the abbey of Westminster that is specifically dated to after the survey.104 The crossing cannot, however, have been much delayed into the autumn, since he is known to have been present at either Caen or Bayeux – Orderic’s two references contradict themselves on the venue – for the wedding of his daughter Constance to Count Alan IV of Brittany. We know that the ceremony had taken place before 8 December, when Constance is recorded as being Alan’s wife.105 This suggests that William would have crossed, as so GR, i, 384–7; Bates, The Normans and Empire, 58. ASC, ‘E’, 1086; JW, iii, 44–5. 103 ASC, ‘E’, 1100. 104 Regesta, no. 326 (post descriptionem totius Anglie). 105 OV, ii, 352–3; iii, 114–15; Les actes des ducs de Bretagne, 371–2 (no. 90), citing AD Maineet-Loire H3713, fo. 86rv. 101 102

the final years

481

often, from Portsmouth or Southampton to Caen. The confirmation of a grant by Maurice, the new bishop of London, to the nuns of Saint-Amand of Rouen that may date from either 1086 or 1087 suggests that William moved eastwards and spent some time at Rouen.106 His presence in Upper Normandy in the first months of 1087 would also have been required by the preparations for what would turn out to be his last military campaign. Orderic’s account of William’s campaign makes it a response to raids into Normandy by the garrison of Mantes-la-Jolie. According to him, these were especially destructive, since people were captured and herds of cattle driven away in the vicinity of Evreux and Pacy-sur-Eure.107 Malmesbury in contrast makes it a reaction to King Philip I’s obscene jokes about his corpulence. He adds that William’s response was to let Philip know that, after hearing mass, he would have 100,000 candles lit, presumably in mocking pretence that so many would be necessary to obtain divine forgiveness for Philip’s rudeness. William then swore an oath in a loud voice, something that Malmesbury says was his typical means to frighten those who were listening.108 In most respects, Orderic’s more detailed account seems preferable since the abbey of Saint-Evroult had priories in the French Vexin and Orderic could draw on informants, as we know he often did. And so too does his chronology, which has William advance to Mantes in the last week of July. He does, however, also have William make threats against King Philip and demand that he hand over the French Vexin, justifying the claim on the grounds that Philip’s father Henry I had promised it to William’s father when Duke Robert had helped defeat the rebellion which had threatened to overthrow Henry in 1033. This claim at first sight seems far-fetched, partly because there is no mention of such a grant in Jumièges, Poitiers, or any other earlier source, and partly because William had made no earlier attempt to assert himself in the Vexin in 1077 after Count Simon’s retirement. On the other hand, William’s behaviour may have been pragmatic. To have tried to take over the Vexin before 1060 would have involved a direct attack on Henry I that would undoubtedly have led to war against him and Count Geoffrey Martes. And after 1063, he would have had to take on the formidable Ralph, count of Amiens, Valois, and the Vexin, who was in any case an ally; and in 1077 and after he may have decided that he had enough difficulties without adding another one. The claim therefore becomes feasible, albeit probably an extension of the truth on the basis of the good relations his father had had with Count Drogo; and Regesta, no. 242. For Orderic’s account of the attack on Mantes, OV, iv, 74–9. 108 GR, i, 510–11. 106 107

482

william the conqueror

also as a means to indicate disapproval to King Philip of his apparent inability to keep soldiers responsible to him under control. Although events moved quickly, perhaps catching the French king unaware, there is nothing to suggest that Philip moved to protect Mantes, something which possibly indicated that he thought the garrison’s behaviour unacceptable. Orderic’s account has William take the garrison at Mantes by surprise and his army gain entry because they were unprepared. After this, his troops started to burn and to sack the town, with the death of a hermit and the burning of the churches of the town being mentioned in English sources as well as by Orderic.109 The troops going on the rampage may not be due to a failure of discipline, since William’s record shows that similar things had happened before, at Mayenne in 1063, at Romney and Dover in 1066, and York in 1069. It may well have been another manifestation of the punishment of flagrantly disrespectful behaviour, an argument that was actually made in relation to the sacks of Mayenne and Romney by William of Poitiers. If so, it is possible that the looting got out of hand, although, as we have also seen on several occasions, the innocent had suffered before when violence of this kind had been inflicted in order to make a point. Tactically, if events had followed their normal course, peace would have been made with King Philip, and the leaders of the raids into Normandy would have been deemed to have been sufficiently chastized. However, while the town was being burnt, according to both Orderic and Malmesbury, William was taken ill, with Orderic putting it down to the heat of the flames and Malmesbury writing that some attributed it to an injury sustained when his horse failed to clear a ditch and the pommel of the saddle ripped into his intestines.110 The De Obitu Willelmi’s version differs slightly by saying that William’s illness came on as he returned from Mantes to Rouen, a difference presumably explicable in terms of the illness worsening as he returned from the campaign.111 Clinical diagnosis is hazardous at this distance in time, but one reasonable suggestion is of cardiovascular failure after a long and very stressful life, with obesity and a life-long heavy meat-orientated diet being further factors. The latter could have led to liver problems or late-onset diabetes. Malmesbury’s information may indicate a contributory factor, but, in all likelihood, the layers of fat would have acted as protection.112 ASC, ‘E’, 1087; JW, iii, 46–7. See also OV, iv, 80–1. OV, iv, 78–9; GR, i, 510–11. 111 GND, ii, 184–5. 112 For these suggestions, Robert Soulignac, ‘La dernière maladie de Guillaume le Conquérant’, AN, xxxvii (1987), 389–90; Crouch, The Normans, 122. 109 110

Chapter 13

D E AT H A N D L E G AC Y

DEATH AND BURIAL The minor chronological difference between Orderic, who places the attack on Mantes in the last week of July, and the Anglo-Saxon Chronicle and John of Worcester, both of which sources place it before 15 August, does not affect the conclusion that William’s final illness was a lengthy one that lasted for several weeks.1 Both Orderic and Malmesbury say that he remained lucid to the end, talking and explaining his wishes to those around him.2 It is therefore possible, as Eadmer suggests, drawing on the memory of St Anselm, who, as we shall see, had good reason to remember this period very well, that there was an expectation that he might recover.3 If this were so, it would explain why he left announcing his wishes for the succession until his last days. William was, however, so weakened that he was at some point moved to the priory of Saint-Gervase in the hills to the north-west of Rouen because he was finding the noise of the city wearisome. According to Orderic, he woke in the morning of 9 September, announced to those around him that he was commending his soul to the Virgin Mary, and then passed away.4 We know that he made a further grant to Saint-Etienne of Caen on what turned out to be his deathbed.5 And, although the sources on the subject are at first sight contradictory, with De Obitu Willelmi saying that he left his crown, sword, and sceptre to his son William, and a charter recording that he bequeathed his royal regalia to the abbey of Saint-Etienne from which William obtained them ten years later by the grant of a manor in England, it is certain that he left the regalia to Saint-Etienne.6 With the De Obitu reasonably being seen as OV, iv, 78–9; ASC, ‘E’, 1087; JW, iii, 46–7. OV, iv, 78–9; GR, i, 510–13. 3 HN, 24 (eo quod infirmitatem suam paulum leuigari sentiret). 4 OV, iv, 100–1. 5 Jean-Michel Bouvris, ‘Aux origines du prieuré de Vains: Une version inédite de la confirmation par le Duc Robert-Courte-Heuse d’une donation faite en 1087 par Guillaume Le Conquérant à l’abbaye de Saint-Etienne de Caen’, Revue de l’Avranchin et du Pays de Granville, lxiv (1987), 3–21, 67–90, at 84–5. 6 GND, ii, 186–7; AAC, no. 24. For the rediscovery of the fifteenth-century vidimus on which knowledge of Rufus’s grant to Saint-Etienne is based, Barbara English, ‘William the 1 2

483

484

william the conqueror

propaganda on Rufus’s behalf, it looks as if its author simply missed out the interim stage before Rufus effectively bought the regalia from the abbey. It has to be significant that several of those who confirmed the charter that described this transaction are known to have been present at William’s deathbed.7 Occupying as it does over thirty pages in the modern edition of the Historia Ecclesiastica, Orderic’s account of William’s death and funeral is a magisterial set-piece, a commentary on the transitory nature of worldly power and a vehicle for Orderic’s moral views on the conquest of England in which he has William express regret for the terrible bloodshed (multa effusione humani cruoris) as a part of his deathbed confession.8 Although the speech is, as far as we know, entirely Orderic’s invention and an idealized version of what he thought ought to have happened, what he was certainly saying is that he believed William made a good end, as also did Eadmer and William of Malmesbury.9 It was also built around ‘the facts’ as Orderic knew them. In their essentials, Orderic’s and the other briefer accounts of William’s final weeks are in substantial agreement. His decline and passing were very much played out in public, with many of the great figures of the cross-Channel empire visiting him. They record the presence of several senior churchmen and physicians, and of William’s two sons William and Henry and his brother Count Robert of Mortain, whose devotion to William is singled out in the De Obitu Willelmi.10 The obvious absentee from all this was Robert Curthose. There are different explanations for why this should have been so: Orderic and Robert of Torigni say that William did not ask him to come, and the De Obitu Willelmi censures him for failing to come to apologize to his father. According to Torigni, Robert only returned from Abbeville in Ponthieu when he heard the news of his father’s death through a messenger.11 Orderic, however, says that William did summon him so that he could take possession of the duchy of Normandy, sending Aubrey, the former earl of Northumbria, to do so, a statement that is not necessarily incompatible Conqueror and the Anglo-Norman Succession’, HR, lxiv (1991), 221–36, at 232, note 45. The seminal study of how the earlier sections of the De Obitu Willelmi on the Vita Hludouici imperatoris is L. J. Engels, ‘De Obitu Willelmi ducis Normannorum regisque Anglorum: Texte, modèles, valeur et origine’, in Mélanges Christine Mohrmann: Nouveau recueil offert par ses anciens élèves (Utrecht and Antwerp, 1973), 209–55. 7 GND, i, lxiii–lxv; Lack, ‘The De Obitu Willelmi’, 1,425–32, 1,441. For the idealization of death in this way, David Crouch, ‘The Culture of Death in the Anglo-Norman World’, in C. Warren Hollister (ed.), Anglo-Norman Political Culture and the Twelfth-Century Renaissance (Woodbridge, 1997), 157–80, at 158–62. 8 OV, iv, 94–5. 9 HN, 24 (non tamen, ut dicitur, inconfessus); GR, i, 510–11 (quae Christiani sunt exsecutus est in confessione et uiatico). 10 OV, iv, 80–1; GR, i, 512–13; GND, ii, 186–7. 11 GND, ii, 202–3.

death and legacy

485

with Torigni, since William may have died while the messenger was on his journey.12 Since William’s public pronouncement about the future of his lands was made towards the end of his life, presumably as he felt his strength ebbing away, it may be that he did not want Robert to come to him until the formalities required that he do so. Alternatively, if Eadmer is correct in saying that, having first summoned Abbot Anselm of Le Bec to come to Rouen, William subsequently informed the abbot that there was no need for him to visit because he felt that he was improving, it is possible either that Robert was not told how ill his father was or that he felt that a return to face his wrath would have been difficult.13 Bearing in mind again both the recent argument that the De Obitu was written in England as propaganda for William Rufus and the fact that its crucial phrase – which I translate rather loosely – that Robert either did not wish to come to apologize or did not have the good manners to do so was inserted by the author and is not present in his literary template, the common ground to all accounts, for all the differences between them, surely demonstrates just how badly relations between William and his eldest son had broken down.14 If Robert was not told how ill his father was, it could have been to exclude him from deliberations about the succession or as an act of denial on the part of William and others that the king was mortally ill. Several of the narratives mention the release of prisoners and the distribution of treasure to churches. Those released included Harold’s brother Wulfnoth, who had been held since 1052, Earl Morcar, and Roger, earl of Hereford. Both they and others who were liberated were required to take oaths not to disturb the peace. Orderic’s account is the only one to mention William’s refusal to release his half-brother Odo, and William eventually being talked round by their brother Robert of Mortain; the basis for the refusal was the depth of Odo’s betrayal and his future capacity to make trouble.15 The distribution of treasure in England is described in some detail by William of Malmesbury and John of Worcester. According to the former, William Rufus distributed gold to monasteries, five silver shillings to churches in the countryside, and 100 shillings to every shire for distribution to the individual poor.16 John gives slightly different figures, but confirms the same broad pattern of distribution to churches and to the poor, adding that Robert did the same in Normandy. His account also indicates that OV, iii, 112–13. HN, 24. 14 GND, ii, 186–7 (‘Quia ipse’, inquit, ‘uenire satisfacturus non uult aut dedignatur’ ); Lack, ‘The De Obitu Willelmi’, 1,422, 1,436. 15 JW, iii, 46–7; OV, iv, 96–101. 16 GR, i, 512–13. 12 13

486

william the conqueror

William’s sons seem to have respected all of his wishes, with one exception being that Rufus re-incarcerated Wulfnoth and Morcar at Winchester after his crossing to England, and a second that, according to another part of Orderic’s narrative, Earl Roger died a prisoner, so was also not released.17 In the midst of all this, Anselm, abbot of Le Bec, having initially been put off by William, arrived on the south bank of the Seine opposite Rouen, presumably at the church of Bonne-Nouvelle that William and Matilda had given to his monastery before 1066, where he was taken ill. Eadmer tells a touching story of William sending delicate foods across the river to Anselm to assist his recovery and of the abbot never being well enough to cross the Seine to see him.18 With different emphases of detail, all sources concur in having William confirm the earlier designation of Robert as his successor in Normandy and name his second son William as king of the English. Maine is largely left to one side, but events show that Robert’s earlier role there was continued. Quite apart from his having done homage for the county to successive counts of Anjou, Maine’s history since 1063 and the frequent military interventions that had been needed must surely have shown that any move which separated rule there from Normandy would have disastrous consequences. Orderic stresses that William felt obliged to honour his earlier promise to Robert, as does the De Obitu Willelmi, with Malmesbury saying that he did so reluctantly and under pressure.19 Orderic presents the grant of England as being made to God (‘I dare not transmit the government of this kingdom, won with so many sins, to any man, but entrust it to God alone, for fear that after my death my evil deeds should become the cause of even worse things’), but follows it up by having William express the sentiment that he hopes that his son William will bring lustre to the kingdom, if God wills it.20 At first sight, apparently a statement that succession to the English kingdom was left open, this passage should be treated as yet another of Orderic’s moral comments on the violence of the conquest of England. When the moralizing is stripped away, William’s words become a formal designation of the succession to the English kingdom to his second surviving son that would be confirmed when Lanfranc crowned him.21 If William had made a good end, as Orderic believed he had, then this acknowledgement of sinfulness was a literary necessity in the type of speech that Orderic was composing. To ensure that his wishes were carried out, William is said by Orderic to have given his JW, iii, 46–7; OV, ii, 318–19. HN, 24. 19 GR, i, 510–11 (Normanniam inuitus et coactus Roberto). 20 OV, iv, 94–5. 21 Cf. English, ‘William the Conqueror and the Anglo-Norman Succession’, 227–31. 17

18

death and legacy

487

son a sealed letter to present to the archbishop and sent him on his way to England. William Rufus had reached a Channel port – in all probability Touques for the crossing into the Solent – when news reached him of his father’s death.22 The Breuis Relatio, Orderic, and Malmesbury all record a grant of treasure to the third son Henry, with, as the flavour of his account again required, Orderic seemingly endowing William with the gift of prophecy. The apparently disappointed Henry was told that he would one day be greater than either of his brothers.23 As far as we know, William had kept his intentions to himself until his last days.24 It is, however, surely impossible that he did not show something of his preferences, given the bitterness of his quarrels with Robert and the ruthless, unforgiving nature that has featured so often in this book. His older sons had after all often been in his company over many years, as also had Henry at the end of William’s life. The insistence that earlier promises to Robert had to be respected indicate a proclaimed – and probably sincere – sense of integrity that went right back to William’s childhood and his own succession to the duchy of Normandy and which was central to his belief in the rectitude of his cause in England. In delaying his decision until almost the moment of death, he was following the precedent of Edward the Confessor, except that William would of course argue that, unlike Edward, he had respected previous commitments. This separation of Normandy and England under the rule of two of his sons does seem a considerable distance from what appears to have been envisaged in 1080 at the time of the settlement of William and Robert’s first quarrel, when Robert was for the first time launched on a full-blown cross-Channel life. It is also surprising in the context of the powerful unitary personal control over Normandy and England that William had consistently exercised. The notion that Robert lost England is mentioned by both Malmesbury and Robert of Torigni, with Torigni’s version being especially remarkable; Robert, so he says, believed that his younger brother would step aside in his favour and would not usurp him, albeit Torigni then added that this was an example of his simple-mindedness.25 The scale 22 In one place Orderic has William receive the news while waiting to cross from Wissant (OV, iv, 96–7) and in another at Touques (OV, v, 202–3), while Robert of Torigni, who knew Orderic’s Historia Ecclesiastica, locates the event at Touques (GND, ii, 202–3). Touques is preferable because of the ease of the crossing to the Solent and Winchester and for strategic reasons, Barlow, William Rufus, 53–5. 23 OV, iv, 94–7; GR, i, 510–13. 24 For the main modern analyses of William’s deathbed bequest in addition to the ones already cited, Holt, ‘Politics and Property’, 45–8 (idem, Colonial England, 149–51); Barlow, William Rufus, 40–51. 25 GR, i, 502–3 (Anglia post mortem caruit), 542–3 (maxime post abdicationem fratris maioris); GND, ii, 204–5.

488

william the conqueror

of the support for Robert’s invasion of England in 1088 would also indicate that he was thought hard done by in the eyes of some of the most powerful members of the cross-Channel elite, such as Bishops Odo of Bayeux and Geoffrey of Coutances, Roger de Montgommery, and Gilbert, the son of Richard fitz Gilbert. It is notable that two of these four and the still-living father of a third had been utterly devoted in their support for Robert’s father for more than three decades. There are therefore compelling reasons to think that they disagreed with William’s final act. In terms of the situational constructionist approach to succession that was set out in Chapter 1, the norm seems to have been that the dying ruler designate, but that consultation and consensus were also expected. As we have seen very often, there might be objections and, quite often, war followed when a consensus could not be reached; the contrast between Edward the Confessor’s accession in 1041–2 and the carnage of 1066 is pretty much all that needs to be said on the matter.26 In relation to William’s conduct, Orderic’s use of metaphor may well again be important, since after William’s death he has the cross-Channel aristocracy seek to mitigate a division of the kind that had fallen on the Israelites after Solomon’s death with his designated successor Rehoboam being supplanted, a punishment inflicted because Solomon himself had broken God’s law.27 Whatever the case, there is certainly a huge contrast between the thoroughness of the arrangements that William’s father had made to secure his young son’s succession and the muddle of William’s last years, when, as far as we can see, any advanced consultation about the succession was minimal. A distinction between inherited and acquired land may have had some influence in persuading William towards division, as it had certainly done in relation to the succession to some cross-Channel estates, most notably those of William fitz Osbern in 1071, but the predominant pattern in such situations was one of flexibility with family members retaining an interest in all lands.28 Both Malmesbury and the Durham Historia Regum suggest that William Rufus crossed to England in great haste, another sign that little had been done to prepare for the future.29 A considerable onus was therefore placed on Lanfranc to manage the situation in England on William’s behalf, something that, as so often, he did successfully, with Rufus being crowned king 26 September. Frank Barlow’s verdict on what happened was of ‘the dismemberment of a great achievement done in anger, bitterness and sorrow. Human passions See also Gillingham, ‘At the Deathbeds of the Kings of England’, 510–16. OV, iv, 122–3; Strevett, ‘The Anglo-Norman Civil War of 1101 Reconsidered’, 172–4. 28 Tabuteau, ‘The Role of Law’, 155–69. 29 GR, i, 502–3, 542–3; JW, iii, 46–7; HRA, 214. 26 27

death and legacy

489

had torn an empire apart.’30 This is possible, but the depths of William’s stubbornness and inflexibility have featured so often in this book that the scale of passion involved seems to me to be doubtful. It is also, given events since Matilda’s death, hard to think that a unified succession in favour of either Robert or William would have gone uncontested. For all the cogency of recent publications attempting partially or totally to rehabilitate Robert, it is still difficult to believe on the basis of his later record that he had the hardness of character which successful rule required. But given his popularity among the leaders of the cross-Channel aristocracy, a unified succession for William would certainly have provoked a great uprising in Robert’s favour. Since there was probably going to be war anyway, a generous assessment of King William’s final act would be that it was a skilful piece of damage limitation that created a situation in which the brothers were at least given a responsibility to collaborate. A less generous one would be that he passed on to others a situation that his bewildering mix of flaws and strengths of character had done a great deal to create. Orderic and, much more briefly, Eadmer present the hours after William’s death as a shambolic time, as all the nobles abandoned his body and the citizens of Rouen reeled round in the streets as if in a daze wondering what to do. A great ruler had died and the certainties had been taken out of their lives. Some of the nobility went away both to fortify their castles in anticipation of coming troubles and, in some instances, to remove the garrisons that William had installed in them. In the case of the second of these, Orderic mentions Robert de Bellême (son of Roger de Montgommery), Count William of Evreux, and Ralph de Tosny as among those who ejected garrisons installed by William. Robert de Bellême, whom Orderic detested, had apparently reached Brionne en route to Rouen from Bellême, but then turned back on learning that William was dead. Meanwhile, at the priory of Saint-Gervase, the attendants plundered the room in which William had died and his body was left lying on the floor. The canons and monks of the city eventually recovered from their state of shock and organized a procession. And with none of the servants remaining around to do it, a country knight named Herluin took it upon himself to prepare William’s body for burial. Archbishop William then took control of the situation and had the body transported to the Seine, where it was placed in a boat that carried it to Caen for burial.31 Some version of the equally shambolic events at Caen during William’s funeral appears in most of the sources. As the body was carried to the church of Saint-Etienne, there was first of all a fire in the town which severely 30 31

Barlow, William Rufus, 49. OV, iv, 102–5, 106–9, 112–15; HN, 24.

490

william the conqueror

disrupted the procession. After mass, Bishop Gilbert of Evreux preached what Orderic describes as an eloquent sermon to a congregation consisting mainly of monks and clergy and a relatively small number of laity. All the bishops of Normandy, including Odo of Bayeux, were present as were many abbots. According to Malmesbury and Robert of Torigni, of William’s sons, only Henry was present; William was by this point in England, but Robert’s absence is perhaps surprising given that he had only to travel into Normandy from Abbeville in Picardy, although a very rapid burial might have made it impossible.32 Then, before the body could be committed to the tomb, a man named Ascelin, son of Arthur, stepped out from the crowd and claimed that the earth in which William was buried was his. The king had taken it from him when the church was built and given no compensation. His claim being judged to have validity, he was paid off by the clergy or, according to William of Malmesbury, by Henry. William’s body was then lowered into a grave that was too small for the corpse. It ruptured and let off a terrible stench. The ceremony was eventually completed and the distressed clergy left the church, although a story recorded by the near-contemporary chronicler Hugh of Flavigny, which mentions that he was relying on the testimony of Walter, abbot of La Trinité-du-Mont of Rouen from 1078 to 1120, does describe a gathering of the clergy immediately after the funeral at which the king’s qualities were discussed.33 At his death William must have been approaching sixty years of age. His sons seem to have honoured his memory well and respected his wishes. According to the De Obitu Willelmi and Orderic, William Rufus commissioned Otto the Goldsmith to build a tomb over the grave of gold, silver, and precious stones.34 KING, DUKE, AND THE MAKER OF EMPIRE Although subject to significant mutations, most notably in and after 1154, the union of Normandy and England created by William’s victory at Hastings endured until King Philip Augustus’s conquest of Normandy in 1204. Its survival in subsequent generations arguably depended on the continued commitment of a small, extremely wealthy, self-perpetuating elite that dominated economic resources and strategic policy-making, and which was capable of absorbing and encouraging the ambitious and upwardly mobile. For Henry’s presence, GR, i, 512–13; GND, ii, 194–5. OV, iv, 106–9; GR, i, 512–13; HN, 24. For Ascelin’s claim, AAC, 45–6. For the gathering of clergy, Hugh of Flavigny, Chronicon, 407. For the identification of Abbot Walter, rather than the otherwise unknown Abbot Hugh who is named in Hugh’s text, NMProsopographie, 263. Walter is known to have been present at William’s funeral, OV, iv, 104–5. 34 GND, ii, 188–9; OV, iv, 110–11. 32 33

death and legacy

491

It also relied on maintaining trust and resilience, a power over minds that might fracture if threats from outside or within became overwhelmingly strong or if leadership became unconvincing. It seemingly had such a powerful hold on peoples’ imaginations that its continuation could feel like a moral necessity.35 Why else would William’s sons have fought a stop-go civil war during the nineteen years after his death to re-establish a single ruler over the cross-Channel empire? And, likewise, why did King Stephen and the Empress Matilda, supported by her husband and her son, fight a long civil war that led to the reunification of Normandy and England under King Henry II? Born plural, the cross-Channel empire remained plural. The continuing distinctiveness of its three parts, Normandy, Maine, and England, being the ultimate foundation of its proclaimed legitimacy, even in an apparent expression of unity, such as William’s seal, plurality was articulated alongside singularity. With its two sides presenting William as the rex of the English and the patronus of the Normans, the recipient of any document sealed in his name was made aware that he personally ruled everywhere, and that different identities, histories, and traditions determined the way in which these responsibilities were exercised in each place. As a result, since the kingdom of the English under William’s rule was conceptually the continuation of the kingdom ruled by Edward the Confessor and his predecessors, and Normandy remained a constituent part of the regnum Francorum, the short- and long-term histories of both have to be seen as a continuum. But neither can be written without reference to the dynamic of the cross-Channel empire and to William’s contribution as its creator. In England, awareness of a continuing national history within an empire dominated by an elite with interests on both sides of the Channel animated William of Malmesbury, Henry of Huntingdon, and others in the twelfth century, with their profound ambivalence influencing almost everything that has been written about England’s history ever since.36 It is because of these factors that England after 1066 has been analyzed in modern times as both a colony and as a state with a continuous history. It is possible to write credible histories within both formats.37 For similar reasons, Normandy can be analyzed as an imperial core, with the commitments that have to follow as well as the obvious profits needing to be taken For these arguments, Bates, The Normans and Empire, 7–15, 185–90. On this theme, Eric Christiansen, ‘The Difficulty of Composing Patriotic History in the Twelfth Century: An Essay on Essays’, in John Bergsagel, David Hiley, and Thomas Riis (eds.), Of Chronicles and Kings: National Saints and the Emergence of Nation States in the High Middle Ages (Copenhagen, 2015), 11–21, at 14–17. 37 For a critique of these approaches, Francis J. West, ‘The Colonial History of the Norman Conquest’, History, lxxxiv (1999), 219–36. 35 36

492

william the conqueror

into account. After 1087 it became the battleground for the wars between William’s sons, with what were sometimes perceived as terrible consequences. Thus, to select one example from many, the poet Serlo of Bayeux, after describing the sack of the city in 1105 by Henry I’s multi-ethnic army as he advanced to take the duchy from his brother Robert, opined that the Normans, who had once honourably subdued England, Apulia, Calabria, and Le Mans, had become trash, unworthy of the great achievements of their ancestors.38 Of the two surviving early obituaries of William, one, the De Obitu Willelmi, was copied out almost verbatim from Einhard’s Life of Charlemagne, which was itself modelled on Suetonius’s Lives of the Caesars. This, so the anonymous author presumably thought, was the way to convey William’s historical importance; there must therefore be elements in what he wrote that he judged were appropriate because they represented a great king, rather than necessarily being an accurate personal portrait.39 The details borrowed included that William would never allow himself to be distracted from the task in hand, neither by its difficulty nor out of fear, that he was tall and physically strong, that he ate and drank moderately and abhorred drunkenness, that he spoke fluently and eloquently, that he had been a devout Christian since childhood and, health permitting, attended religious services every morning and evening. But the anonymous author chose to leave out vivid and personal details from Einhard’s portrait, including Charlemagne’s intellectual curiosity. Thus, for example, we are told that, although he tried, Charlemagne could not write. He did, however, speak several languages fluently, including Latin, and he had works of history and theology read to him.40 By implication, none of this presumably applied to William. The notion that William could be placed on the same level as Charlemagne appears in historical writing in England in the twelfth century. Henry of Huntingdon wrote that William was greater than all the kings who had preceded him and put him on the ultimate pedestal sitting on the right hand of God, a place he reserved among kings otherwise only for Charlemagne, and later, in his Contemptu mundi, for Henry I. For Henry of Huntingdon, as for the anonymous author of the De Obitu, William had to be among the greatest of kings. But ambivalence was central to what he wrote since, employing a literary topos that does not actually refer to geno-

38 ‘Serlo of Bayeux, The Capture of Bayeux’, ed. and trans. Moreed Arbabzadah, in van Houts, ‘The Fate of Priests’ Sons’, 86–105, at 90–3. 39 Lack, ‘The De Obitu Willelmi’, 1,420. 40 GND, ii, 188–9. Neither of the published parallel texts gives the full text of Einhard, Engels, ‘De Obitu Willelmi’, 228–30; Lack, ‘The De Obitu Willelmi’, 1,450–5.

death and legacy

493

cide as we would define it, he opined that God had chosen the Normans to exterminate the English.41 The other early obituary, the homily in the ‘E’ Chronicle, was written by an author who says that he had lived at court. It describes William as more powerful than any of the kings before him, a very wise man, and one who was stern in his treatment of those who opposed him. He had an exceptional sense of his personal dignity, which was expressed, when he was in England, by the crown-wearings at Winchester at Easter, Westminster at Whitsun, and Gloucester at Christmas. He brought peace and maintained such firm justice that people carrying valuables could travel unmolested. Men who raped women were castrated. Monasteries and the Church in general flourished and were protected by him. However, returning to the theme of sternness and severity, the writer also mentioned William’s violence without developing the statement. He might have been referring only to his conduct at court, but perhaps he had his behaviour in general in mind. William was so fierce and austere that he kept earls imprisoned and could even lock up his own brother. He had subdued Wales and Scotland and, had he lived, he would have done the same to Ireland within two years, by peaceful means. He had faults that the author advised his readers to avoid, namely, greed, cruelty, and an excessive love of hunting. His castle-building oppressed the poor. His financial extortions were such that the author thought him avaricious. He ‘loved the stags as much as if he were their father’. And, finally, before the conventional prayer that the Lord should show mercy: ‘Alas! Woe that any man so proud should go and exalt himself, and reckon himself above all men.’42 Since he was writing in Old English, the author was presumably a monk of English birth. For all the complexities associated with compilation of the text of the ‘E’ Chronicle, if Canterbury is accepted as the place of authorship, he was writing in a community where the interaction of the indigenous and the imported peoples and cultures had begun soon after 1066. The Chronicle’s author might even have known the librettist of the Bayeux Tapestry and to have been schooled in, and have absorbed, similar values. Whatever the case, given that he knew the court and had seen William in 41 HH, 402–3 (Elegerat enim Deus Normannos ad Anglorum gentem exterminandam). In general, Bates, The Normans and Empire, 21–2, with references to essential work by others. For Henry and William, HH, 394–5, 410–11, 594–5; Wendy Marie Hoofnagle, ‘Charlemagne’s Legacy and Anglo-Norman Imperium in Henry of Huntingdon’s Historia Anglorum’, in Matthew Gabriele and Jane Stuckey (ed.), The Legend of Charlemagne in the Middle Ages: Power, Faith and Crusade (New York, 2008), 77–94, at 81, 87. 42 ASC, ‘E’, 1086 (recte 1087). The Anglo-Saxon Chronicle: A Collaborative Edition. Volume 7: ms. E, 96–8. My paraphrase mostly follows the translation by Susie I. Tucker in English Historical Documents 1042–1189, 170–2.

494

william the conqueror

action there, he must be credited with the capacity to set William’s life into a context that took in the pre-1066 English past and the late eleventh-century present.43 As such, he was maintaining the tradition of historical writing focused on the patria that was continued in the twelfth century by the likes of Malmesbury and Huntingdon, although much more expansively.44 After making the statement that it was God’s will that William had conquered England, the author added nothing, presumably because his subject was how William had ruled, a topic suitable for his audience to reflect on; God’s will was undoubtedly beyond his and their ken. Some of his commentary was an obviously direct response to events that had shown divine disapproval of William to the world. That two of his sons, and also a grandson, died accidentally while hunting in the New Forest must have been a commentary both on a dangerous sport and on excess of a wellknown kind in which William must have been thought to have overindulged since God punished his family so terribly.45 But other sections, for all their concision, focus on themes that have been central to this book’s narrative. The references to castle-building, greed, and cruelty draw attention to the distress and disruption that the first of these had caused and to the mutilation and killing of rebels. They also continue the diatribe that featured in the passage quoted at the end of the previous chapter. The final sentence must show that he thought William arrogant. Even if William was sincerely devout and did many things for the good of the Church and to keep the peace, it was all insufficient compensation for a sinful absence of humility; the displays of charisma and dominance went beyond what was acceptable. The author could not avoid being anxious about him as a model to imitate.46 Resolving the ethical dilemmas presented by worldly power so manifestly grounded in violence, yet subsequently creative of what might be thought peace, was not something that could be easily done. Among William’s children, as is often the case, it was the women who were the main guardians of memory. Thus, one of the two long-lived daughters, 43 Dumville, ‘Some Aspects of Annalistic-Writing at Canterbury’, 31–5; The Anglo-Saxon Chronicle: A Collaborative Edition. Volume 7: ms. E, xxxv, lxxxv. 44 Pauline Stafford, ‘The Anglo-Saxon Chronicles, Identity and the Making of England’, HSJ, xix (2008, for 2007), 28–50, at 40–2; Nicholas Brooks, ‘ “Anglo-Saxon Chronicles” or “Old English Royal Annals” ’, in Nelson, Reynolds, and Johns (eds.), Gender and Historiography, 35–48, at 43–5. 45 See Fabrice Guizard, ‘Les accidents de chasse dans les récits du premier Moyen Age’, in Laurent Jégou, Sylvie Joye, Thomas Lienhard, and Jens Schneider (eds.), Faire lien: Aristocratie, réseaux et échanges compétitifs. Mélanges en l’honneur de Régine Le Jan (Paris, 2015), 289–97. 46 For a recent essay on this theme, Sigbjørn Olsen Sønnesyn, ‘ “Only through Time is Time Conquered”: Liturgy, History and the Timeless Aspirations of the Temporal’, in Bergsagel, Hiley, and Riis (eds.), Of Chronicles and Kings, 23–50, at 36–8.

death and legacy

495

Cecilia, although a nun and eventually an abbess, as already noted, often referred to herself as ‘the king’s daughter’ (filia regis); her place in the world still mattered.47 And whether we accept as factually accurate or not Baudri of Bourgueil’s statement that Adela, countess of Blois-Chartres (died 1137), had an embroidery of her father’s achievements close to her bed in her private chamber, it remains significant that a poet seeking to flatter her considered this an appropriate thing to write. Again presumably writing what he believed Adela wanted to hear, in the midst of imperial language and comparisons, he announced that ‘In the end his power was so great that his name alone made the entire earth tremble and bow to his rule.’48 Adela’s charters could also refer to her parentage and often request prayers for William and Matilda.49 She provided consistent support to her brother Henry.50 Although buildings cannot speak as directly to us as the written word, they are very informative about the attitudes of the sons who were directly responsible for their father’s legacy. If it is accepted that William Rufus was responsible not only for the great hall at Westminster, built within a stone’s throw of where his father had been crowned, but also for the Salle de l’Echiquier at Caen, a hall built in similar style (albeit only approximately a quarter of the size and from which the towers of his father’s burial church were visible, as they are today), then his role is made manifest.51 Henry I financed further building at Caen as well as the donjon at Falaise, his father’s birthplace. The two Caen abbeys remained a focus for the family with grants to Saint-Etienne in Normandy and England being the subject of double confirmations by Robert Curthose and William Rufus, and later Robert Curthose and Henry I; La Trinité became a mausoleum for Matilda’s female descendants.52 Returning for a moment to the way in which legends developed around William’s birth, it is clear that his achievements and reputation remained live topics for his family long after his death. 47 AAC, nos. 25, 27; Charters and Custumals of the Abbey of Holy Trinity Caen, Part 2, 131–2 (Cartulary Documents, no. 20). 48 Baldricus Burgulensis, Carmina, 149, 164 (lines 17–18, 581–2) (‘Baudri of Bourgueil, “To Countess Adela” ’, trans. Otter, 66, 80). 49 LoPrete, Adela of Blois, 150, 152, 156, 167, 434. 50 Green, Henry I, 85, 161–2; Kimberley LoPrete, ‘The Anglo-Norman Card of Adela of Blois’, Albion, xxii (1990), 569–89. 51 For the general argument, Edward Impey, ‘The Great Hall at Caen and its Affinities with Westminster’, in Warwick Rodwell and Tim Tatton-Brown (eds.), Westminster: II. The Art, Architecture and Archaeology of the Royal Palace, British Archaeological Association Conference Transactions, xxxix, Part II (Leeds, 2015), 72–88, at 81–4. For the stylistic arguments in detail, Edward Impey, ‘The Great Hall of the Dukes of Normandy in the Castle at Caen’, in Davies, Riley, Levesque, and Lapiche (eds.), Castles and the Anglo-Norman World, 101–32, at 104–20. 52 David Bates, ‘A Neglected English Charter of Robert Curthose, Duke of Normandy’, BIHR, lix (1986), 121–4; Letouzey-Réty, Ecrit et gestion, 123–32.

496

william the conqueror

William of Malmesbury’s comment that William Rufus was a peerless prince overshadowed by his father may actually reflect more than its author’s personal opinion.53 It is possible to see Rufus as so enthusiastically devoted to his father’s memory that he fought wars his father would never have fought. How otherwise can we explain the conquests of parts of Wales and the dispatch of armies into Scotland that, unlike those sent north in his father’s reign, did fight battles? In his determination to attack where his father would have negotiated and in his recruitment of followers from Maine, Rufus can almost be seen as a critic of his – in his eyes – over-cautious father.54 In not releasing prisoners whom William had freed on his deathbed, he might have thought he was continuing his father’s security policies or reacting against his excessive piety. One of Rufus’s first acts after his coronation was to go north to York in early 1088 to be present at the very beginnings of the abbey of St Mary’s. If so, he was ceremonially celebrating in person the rebirth of the city as a great royal and religious centre that William had sponsored right at the end of his reign. In doing so, Rufus was performing an act of atonement for the blood his father had shed there at a time when he must have known that the widely supported revolt in favour of his brother was developing.55 Rufus’s rule also saw extensive investment in Gloucester, specifically through patronage of the abbey of St Peter’s, again a continuation of the kingship of William’s later years.56 What was not done may also illustrate how venerated William’s memory was. As we have seen, it was only in the second generation after his death that the subject of the substantial estates he was claimed to have removed from the landholdings of English monasteries, notably from the two Winchester monasteries and Glastonbury, was raised by his grandson, Bishop Henry of Winchester, previously the abbot of Glastonbury. Stephen’s ‘Oxford’ Charter of Liberties can be viewed as a commentary on practices going back to William’s day. The genesis of empire resides in the fusion of William’s increasingly dominant rule in Normandy before 1066 with the political resilience and organizational sophistication of the English kingdom. In both cases, the political, military, and intellectual foundations go back into the early tenth century and, in England’s case, in certain respects, to earlier times. Both GR, i, 542–3. Bates, The Normans and Empire, 99–100, 146–7. 55 For William II and York, Stephen of Whitby, De Fundatione Abbatiae Sanctae Mariae Virginis Eboraci, in Monasticon Anglicanum, ed. Dugdale, iii, 544–6; Christopher Norton, ‘The Buildings of St Mary’s Abbey, York’, Antiquaries Journal, lxxiv (1994), 256–88, at 280–2; Burton, The Monastic Order in Yorkshire, 40–1; Harrison and Norton, ‘Lastingham’, 68–9. 56 Cownie, Religious Patronage in Anglo-Norman England, 57–60. 53 54

death and legacy

497

states were also constructed through the efforts of capable practitioners of early medieval warrior kingship. Both can in certain ways be regarded as legatees of the Carolingian Empire. As we have already seen, in whatever ways the arguments for the integration of the duchy of Normandy and its settlers into the post-Carolingian societies of the French kingdom are played out, its rulers in many fundamental respects have to be seen as the cultural descendants of the Carolingian Empire, albeit with variations involving influences and the active social memory of the Scandinavian past. Variety is present in the huge literature on the tenth- and eleventh-century English kingdom, but Carolingian influence has effectively become an orthodoxy there too, so much so that, as noted in the Prologue, Chris Wickham has gone so far as to entitle the chapter of a book ‘Carolingian England, 800–1000’, while the late Patrick Wormald described the achievements of the English kings as ‘plus Carolingien que les Carolingiens’.57 The great irony is that, as will be explored later, this very efficiency – even precocity – may have made it easier to conquer and take over the English kingdom.58 For all this emphasis on organization and sophistication in Normandy and England, the cross-Channel empire was also one forged in a crucible of instability. To be counter-factual, it is worth asking whether the invasion of 1066 would have happened if Geoffrey Martel and King Henry I had still been alive; certainly, the endemic competitiveness of northern French politics that they epitomize had revived within a decade of 1066. Also, while Pope Alexander II had been so powerfully committed to William’s cause in and after 1066, and Gregory VII likewise until his death, the moral support they provided could have been withdrawn; given the scale of criticism of the violence of the Conquest, it is perhaps remarkable that it was not. There is, as already mentioned in this chapter and Chapter 1, also the matter of succession viewed as a situational construct. The history of the English kingdom throughout the tenth, eleventh, and twelfth centuries is from one perspective a history of one succession dispute after another. It was therefore certainly not inevitable that the cross-Channel empire would endure until 1204. Nonetheless, even if William’s children did proceed to fight among themselves, he was arguably fortunate to have produced longlived male offspring, as some of his English predecessors had not. 57 Wickham, Inheritance of Rome, 453–71; Patrick Wormald, ‘Pre-Modern “State” and “Nation” ’, in Stuart Airlie, Walter Pohl, and Helmut Reimitz (eds.), Staat im frühen Mittelalter (Vienna, 2006), 179–89, at 184. For an admirable summary of these discussions setting them in a European context, Molyneaux, The Formation of the English Kingdom, 231–48. 58 Stephen Baxter, ‘The Limits of the Late Anglo-Saxon State’, in Walter Pohl and Veronika Wieser (eds.), Der frühmittelalterliche Staat – europäische Perspektiven, Forschungen zur Geschichte des Mittelalters, xvi (Vienna, 2009), 503–13, at 512–13 (‘the strengths of the late Anglo-Saxon state contributed to its fall’).

498

william the conqueror

In relation to William as a maker of empire, the contrasts between the conquests of Maine and England have to be more important than the similarities. For all that England’s identity as a kingdom was preserved, so overwhelming was the destruction of the Old English aristocracy that, concretely defined, its conquest was effectively the takeover of one state by another. Maine, in contrast, always remained marginal to William’s authority, a society within which potential rebels could look for serious support from other powers and into which almost no Normans were inserted. Its integration into the cross-Channel empire was always limited, with only two abbeys, Saint-Calais and Saint-Pierre de la Couture, recorded as obtaining any land in England after 1066, but neither apparently as the result of a grant by William.59 And, while churchmen such as William, bishop of Durham, were recruited from there, as Katharine Keats-Rohan has pointed out, there is ‘no indisputable documentary evidence of any landholding by a Manceau layman (in England) before 1087’.60 The networks that connected the county of Maine to Normandy and England were mostly only the regional ones that linked it to western Normandy. The decision not to create strong networks linking Maine to England through Normandy is perhaps surprising, the more so because William Rufus and Henry I very consciously tried to compensate for what their father had not done.61 Why this should have been so is ultimately a matter for speculation. But it is speculation of the greatest importance in terms of understanding both William as a maker of empire and the history of the empire after his death. In all probability, for William to have attempted a conquest in Maine along the lines of England would have involved taking on enemies who were in military terms his equals and who could readily muster numerically superior resources. The near miss of late 1081 notwithstanding, a decisive battle that would resolve all the issues was not possible and, so it would appear, no one actually wanted to try to fight such a battle because they knew this. It was also the case that the justifications for confiscations and banishments in Maine were arguably never as persuasive as they were in England. Moreover, viewed pragmatically, they might not have been effective because any major aristocrat who might have been exiled could recruit allies from neighbouring regions and use them to stir up trouble. Even when a churchman was removed, lengthy legal wrangling followed. The argument for an ecclesiastical takeover never existed. The absence of any equivalent of King Harold, against whom a case for perjury

DB, i, fos. 147v, 340v; DP, 522–3. DP, 33. 61 Bates, The Normans and Empire, 99–100, 146–7. 59 60

death and legacy

499

and tyranny could be made, meant that there was no moral high ground to try to occupy. It is also possible that William thought too many close links between Maine and England likely to be disruptive because they would give outsiders a stake in the Anglo-Norman nexus with the potential to destabilize it; the consequences of his son Henry’s decision to marry his daughter and sole heir, the Empress Matilda, to Geoffrey, count of Anjou, can be used as an argument that William had been right. Tendentious as this last part of the argument might be, it does take the discussion into the more certain territory of no empire being able to escape its past. The problems that had beset William in France before 1066 remained with him. They were still with him at the end. And he bequeathed them to his successors. For all the approximations of necessity involved, we can calculate that William spent about 130 out of a possible 170 months in Normandy and northern France between the reconquest of Maine in 1073 and his death in 1087; that is, between 23 per cent and 24 per cent of his time in England and between 76 per cent and 77 per cent in Normandy and northern France. Geographically, his travelling was also limited, since, after the end of 1072, he only occasionally ventured out of the southern English triangle of Westminster, Winchester, and, after 1080, Gloucester; the journey into west Wales as far as St David’s in 1081 was the only recorded major excursion out of a southern English regional itinerary. So gross an imbalance of time spent in northern France and England might at first sight seem surprising. However, warfare or at least the threat of warfare and family quarrels obviously necessitated his long-term presence in Normandy and made his itinerary to an extent unpredictable. The visits to England, although quite often prompted by crisis, were always purposeful. A fullblown strategy of kingly and imperial rule existed from the beginning. Moreover, it evolved and was continuing to evolve at the time of William’s death, as illustrated by the survey that led to the making of Domesday Book and the plans for York that were being developed at the end of his reign. Of William’s known crown-wearings, Easter in 1068, 1069, 1070, 1072, and 1086 were spent at Winchester. It is likely that he was there in 1071 and 1076 as well and that 1081 is also a possibility; in other words, whenever he was in England, he usually, if not invariably, aimed to be at Winchester for Easter.62 When Normandy is built into this itinerary, the principle that this great festival was spent at a major residence was also a firmly established one. Three Easters were spent at Fécamp (1067, 1075, 1083), one at Bonneville-sur-Touques (1073), and one at Rouen (1080). Although the 62

These and the following comments on itinerary are based on Regesta, 82–4.

500

william the conqueror

location of the Whitsun gatherings is both less fixed and less often referred to in the surviving sources, two were spent at Westminster (1068, 1086), two at Windsor (1070, 1072), one at Lillebonne (1080), one at Winchester (1081), and possibly three at Rouen (1067, 1074, 1082). Christmas is also not very clear-cut, with Gloucester being the location only in 1080 and 1085. It is also notable that at least eleven Christmases – that is, a majority – were spent in Normandy. Since these statistics depend on variable charter survival and erratic referencing in histories, there is no reason why many among the nine Easters that cannot be conclusively accounted for should not have been spent at major residences either. In relation to William’s role as the maker of empire, it is the grandeur and display of the crown-wearings that matter and the expectation that they created. They were maintained and then changed by his children; William of Malmesbury says that William Rufus continued the practice but that Henry I allowed it to decline.63 The building works at Westminster and Winchester show that every bit as much attention was given to English royal centres as to the ducal centres in Normandy. They were huge operations begun very soon after 1066, demonstrating a personal aim to show off majesty and power. Windsor was developed early, too, because it supplemented movement between these places. And Gloucester after 1080, because it fitted into the preconceived notion that the great royal centres were those frequented by Edward the Confessor. Malmesbury’s belief that William innovated in holding such lavish crown-wearings, while probably reductive of pre-1006 English practice, should be treated as evidence for the association between grand theatre and ideology that characterized William’s kingship.64 The exception that effectively proves this particular rule is Caen. William often visited Caen, but, for all the grandeur of the ecclesiastical building and the development of the castle within the town, its residential buildings that we know about were mediocre, despite the Channel crossing between Caen, Winchester, and the New Forest being the axis on which the cross-Channel empire turned.65 In William’s time, Caen did not achieve the palatial status of Rouen, Fécamp, Winchester, and Westminster. Also at the heart of this mixture of conservatism and grandeur was William’s respect for the great English saints and his patronage of some English monasteries and churches, notably Bury St Edmunds and Westminster. St David was also dutifully added to the list of honoured GR, i, 508–9. Biddle, ‘Seasonal Festivals’, 52–7. Some aspects of this very good article require amendment in the light of amendments to William’s itinerary set out in this book. 65 See the latest survey that covers this point, Jean-Marie Levesque and Pascal Leroux, ‘Le château de Caen’, in Davies, Riley, Levesque, and Lapiche (eds.), Castles and the AngloNorman World, 89–100, at 89, 94. 63 64

death and legacy

501

saints in 1081. His veneration of well-established and authenticated cults may also explain the statement in both versions of the Anglo-Saxon Chronicle that on his march into Scotland in 1072 he went beyond the Forth and ‘found nothing there that he was any better for’.66 As a measure of his perspectives, after 1083 he granted Scaldwell in Northamptonshire to Bury St Edmunds for Matilda’s soul.67 It is possible, too, that this conservatism lies behind his treatment of Edgar the Ætheling. Although his record of rebellion outstripped that of anyone else, Edgar was always protected, as if his status placed him on a different plane from everyone else. In 1086 he was given a force of 200 soldiers to travel to southern Italy. William II allowed him to lead an army into Scotland and Henry I famously married Edgar’s niece in 1101. It is as if Edgar’s place within the royal kindred of which William was also a member was being given special recognition, something that his children continued. It was also an easy way for them to make themselves appear magnanimous. Even where we think we detect innovation, as in the case of the Laudes and William’s seal, what was ultimately involved was an upgrading with, as we have seen, the Empire having provided significant inspiration in both cases. Yet while imperial responsibilities were encompassed by their very existence, the conceptual separation of Normandy and England was also sustained. Thus, as far as we can see, William would never have been the subject of the same liturgical acclamation in Normandy and England.68 As a result, although he lived out almost four decades in a state of actual conflict or varying levels of tension with Henry I and Philip I, the Laudes chanted in Normandy included prayers for the well-being of the French kings.69 And, as a statement of legitimacy, the ‘D’ Chronicle in 1079 recorded that William had had his designation of his son Robert as the next duke of Normandy confirmed by Philip I; in other words, just as William’s father Duke Robert had had his own future rule confirmed by Henry I before he set off on his fatal pilgrimage in 1035.70 In terms of those two other public demonstrations of rule, royal charters and coins, no significant changes were made to the basic patterns existing before 1066, except that the weight of the coins was brought into a defined ratio to ASC, ‘D’, ‘E’, 1072. DB, i, fo. 222r. 68 Cowdrey, ‘The Anglo-Norman Laudes Regiae’, 48 (‘the Norman king-dukes were never acclaimed as rex in texts that were sung within the regnum Francorum) (Popes, Monks and Crusaders [London, 1984], chapter 8). 69 Thus, e.g., the Fécamp Laudes Regiae of c.1100 (Illi Francorum regi in Christi pace uita et uictoria), Cowdrey, ‘The Anglo-Norman Laudes Regiae’, 68 (Popes, Monks and Crusaders, chapter 8). 70 ASC, ‘D’, 1079. 66 67

502

william the conqueror

facilitate cross-Channel exchange. As far as charters were concerned, the revolution was going to come, but there were only stirrings during William’s lifetime.71 The so-called introduction of knight service into England also belongs within the framework of caution and conservatism that surrounds William. Its implementation was a central aspect of the culture of reward and security that had shaped the takeover of the English kingdom from the beginning. The defence of conquest required a reliable military force and the squadrons of warriors who had accompanied their lords expected reward and had to be given the wherewithal to live in a new land. The exercise of lordship required a retinue and arrangements to support one. Much of the manpower recruited under the previous systems by English kings was either intrinsically untrustworthy or had been decimated in the warfare of 1066. However, although a different type of personal relationship to provide military service was introduced after 1066 whose workings had important long-term implications for England’s history, it was a variant on a sociological theme, not a new species.72 Aristocrats had had every bit as much a responsibility to recruit armies for the king before 1066 as they did afterwards, even if they usually did so before 1066 on the basis of land assessments described as hides, carucates, or sulungs, depending on which area of the kingdom was involved.73 So-called incidents such as succession dues, whether described as heriots or reliefs, were present in both systems; to make a payment instead of actually performing the service, likewise. To misuse the terms as John Gillingham has provocatively done in order to make an entirely legitimate point, England before 1066 was every bit as much a society of ‘gentlemen and gentry’ as England after 1066.74 They 71 For general comment on charters, David Bates, Re-ordering the Past and Negotiating the Present in Stenton’s ‘First Century’, The Stenton Lecture for 1999 (University of Reading, 2000), 6–7 (‘the apparent paradox of almost no change in the midst of great change’), with references cited there; Richard Mortimer, ‘Anglo-Norman Lay Charters, 1066–c.1100: A Diplomatic Approach’, ANS, xxv (2003), 153–75. The subject of the weight-ratio of English and Norman coinages depends on whether the PAXs issue is assigned to William I or William II. See further Pamela Nightingale, ‘The Evolution of Weight-Standards and the Creation of New Monetary and Commercial Links in Northern Europe from the Tenth to the Twelfth Century’, Economic History Review, 2nd. ser., xxxviii (1985), 192–209, at 200–1 (reprinted in Pamela Nightingale, Trade, Money, and Power in Medieval England [Aldershot, Burlington, VT, 2007], chapter 4); Martin Allen, Mints and Money in Medieval England (Cambridge, 2012), 25–6. 72 For a recent statement on the subject, Nicholas Brooks, ‘The Archbishopric of Canterbury and the So-called Introduction of Knight-Service into England’, ANS, xxxiv (2012), 41–62, with copious references to the previous literature. 73 Richard P. Abels, Lordship and Military Obligation in Anglo-Saxon England (London, 1988), 79–184. 74 John Gillingham, ‘Thegns and Knights in Eleventh-Century England: Who Then Was the Gentleman?’, TRHS, 6th ser., v (1995), 129–53. See also David Bates, ‘England and the

death and legacy

503

were a reflex conditioned by norms deeply embedded in a centuries-long past, another manifestation of the symbiotic relationship between ruler and nobles that existed throughout the medieval West and was naturally present in Normandy.75 Although the specific evidence we have for the early years of William’s reign in England is thin, the processes of the so-called introduction of knight service must have begun early. The imposition of especially heavy burdens of service on the existing English monasteries was a source of complaint for twelfth- and thirteenth-century writers and constitutes a special case because it was so obviously exploitative in practice. But again it was a sociological norm; not only did monasteries provide service in Normandy and many other places, William’s exploitation of the monasteries of Normandy is exemplified by the two abbeys of Mont Saint-Michel and Fécamp complaining that he had taken lands away from them.76 All rulers in this and previous centuries regarded the Church’s wealth as being to an extent at their disposal.77 The existing early post-1066 lists of knights and their service, such as the interpolated Abingdon list originally compiled before 1083 and the list of c.1090 in the Domesday Monachorum of Christ Church, Canterbury, show that service was being defined against available resources from the very beginning.78 The processes of change that led up to this are well described in the Abingdon Historia: newcomers took over estates previously held by English killed in 1066, service from an early date was required on William’s campaigns in Britain and France and to garrison the castle at Windsor, and the situation evolved over time.79 Where there is sufficiently extensive documentation to analyze the process, as there is at the abbey of Evesham, the granting out of lands in return for service can be shown to have taken place over decades, often with the purpose of making rewards to an abbot’s relatives.80 The early twelfth-century account in the Liber Eliensis emphasizes exactly the same points, namely, the need to negotiate with the new

“Feudal Revolution” ’, in Il Feudalesimo nell’Alto Medioevo: Settimane di Studi del Centro Italiano di Studi sull’Alto Medioevo, xlvii (2000), 611–49, at 619–25. 75 The general point is well made in John Gillingham, ‘The Introduction of Knight Service into England’, ANS, iv (1982), 53–64, at 62–4. For Normandy, Bates, Normandy before 1066, 168–9. 76 The Cartulary of the Abbey of Mont-Saint-Michel, no. 115; Regesta, no. 139, Appendix no. 1. 77 Thus, e.g., Robin Fleming, ‘Monastic Lands and England’s Defence in the Viking Age’, EHR, c (1985), 247–65. 78 Historia Ecclesie Abbendonensis, ii, 322–5; Domesday Monachorum of Christ Church, Canterbury, ed. D. C. Douglas (London, 1944), 105. On these lists, Clarke, ‘ “Those Five Knights” ’, 25–9. 79 Historia Ecclesie Abbendonensis, ii, 4–9. 80 Clarke, ‘ “Those Five Knights” ’, 8–15.

504

william the conqueror

powers introduced into local society and to grant out lands on terms to secure service.81 What these changes could involve in human terms takes us back to the story of Hereward’s return to Bourne to find that his brother had been murdered, and to situations where, even if what occurred was less dramatic and terrible, massive upheavals still took place within local and regional societies. This was the transformation on the ground that followed on from the comprehensive takeover that began very soon after the victory at Hastings. For this reason, the way in which Orderic and John of Worcester associate the making of Domesday Book with a survey of the kingdom’s knights is not a reference to fundamental changes in arrangements that were already in place or to any far-reaching national project, but rather to it once more potentially supplying evidence of the resources that supported the king’s rule and allowed negotiations about service to take place.82 Its essence was the transformative process from the violence of the first phase of conquest into the routines of governance. While most of the direct evidence for the taking of tax comes from either the beginning or the end of William’s reign, the assumption normally made on the basis of one reference in Domesday Book and of practice under Henry I is that the taxes that might be called geld were taken in some form on an annual basis, although not necessarily always on a nationwide one in William’s time.83 As we have seen, the geld of 1084 and the probable geld of 1086 were the subject of the ‘E’ Chronicle’s criticisms of William’s rule, just as those at the start of the reign had been of the ‘D’ Chronicle. The striking feature of the Domesday evidence for the assessments in hides, carucates, and sulungs on which geld and other burdens were assessed is the scale of the exemptions or reductions that had come into existence. The scale of these exemptions was such that tax could never have been collected in the massive quantities that had been done in the reigns of Kings Æthelred and Cnut; a sum of approximately £9,000 for the six-shilling geld of 1084 has been suggested, whereas, according to the Anglo-Saxon Chronicle, £83,500 had been collected in 1018.84 The deficiencies of the evidence notwithstanding, the central issues are whether William had effectively disabled the tax in order to reward his followers and whether the survey that led to the making of Domesday Book was part of a plan to recover some of the lost fiscal ground. From the document known as the Northamptonshire Geld Roll, which dates from the Liber Eliensis, 216–17 (trans. Fairbrother, 258–60). OV, iv, 52–3; JW, iii, 44–5. 83 Green, ‘Last Century of Danegeld’, 242; Harvey, Domesday, 212–13. 84 Harvey, Domesday, 211. 81 82

death and legacy

505

1070s, and the Geld Rolls incorporated into Exon Domesday, it is clear that manorial demesne, that is the part of an estate worked directly on the lord’s behalf, was normally exempt from payment. The returns to Circuit I of the Domesday survey, and especially those for the counties of Berkshire and Surrey – as the only circuit where the hidages on which geld was assessed and paid are recorded for King Edward’s reign, the time an estate was received, and in 1086 – are the only ones that supply fully specific evidence. As such, they show that the beneficiaries were many of the wealthiest members of the cross-Channel elite, major churches such as the archbishopric of Canterbury, and sometimes Englishmen listed as king’s thegns. Thus, for example, the lands in Surrey held by Bishop Odo of Bayeux before his imprisonment seem to have paid scarcely any geld at all. In some cases, it looks as if the holder of the property pocketed the tax; in others, it appears that it was not collected. The outcome of a detailed study of these and other shires has stated that ‘Whilst tenants-in-chief and royal lessees were exacting much higher returns from freeholders and small tenants in William’s reign than previously, they themselves paid much less tax’.85 Such exemptions had certainly existed before 1066, but the paucity of information means that it is impossible to define precisely the extent of their expansion after 1066. The author of the twelfth-century Laws of Edward the Confessor certainly believed that the demesne of all churches, even parish churches, had been exempt before 1066.86 It is even possible to argue that exemption on a national scale was an innovation that took place during William’s reign.87 Conceivably, if Edward the Confessor’s abolition of the heregeld in 1051 had held firm for the rest of his reign, then William and his associates would in 1067 have effectively been reviving a defunct tax, leaving open the mechanism for new ways of operating. This would not have ruled out those who had enjoyed exemptions or reductions before 1066 from claiming them again. It would, however, have opened up the possibility of extensive rewards to those who supported William, some of whom could have been English. The directed nature of exemptions towards wealthy supporters and the general exemption of demesne that would benefit tenants owing services suggest that this is what happened. Although it has been argued that the Domesday record of ploughlands, that is of the number of ploughs normally required to plough a particular estate, constituted an attempt at reassessment, based on perceived agrarian 85 Sally Harvey, ‘Taxation and the Economy’, in Holt (ed.), Domesday Studies, 249–64, at 256–63; idem, Domesday, 215–21. More generally, Green, ‘The Last Century of Danegeld’, 242–52. 86 Leges Edwardi Confessoris, ed. O’Brien, 168–9. 87 Pratt, ‘Demesne Exemption’, 2–8.

506

william the conqueror

realities, this seems intrinsically unlikely.88 Most likely, the insertion of ploughlands into the Domesday record was just another means of describing the resources of an estate; one where the surprising statistic that estates sometimes had more ploughs working on them than were apparently necessary might be explained by specific circumstances relating to the estate, such as dispersed settlement.89 Whatever the case, if a reform was ever proposed, it was never implemented, since the basis for the taking of geld in the one full surviving Pipe Roll of Henry I’s reign shows that it was being taken annually on the hidage of King Edward’s time. According to Orderic, William Rufus’s financial agent Rannulf Flambard did propose a remeasurement of ‘all the ploughlands, which in English are called hides, with a rope, and to make a record of them’.90 This constitutes evidence of the possibilities that Domesday Book had opened up, albeit ones that were excessively ambitious. It is also another sign that Rufus was trying to amend policies followed by his father. A further one was clause 11 of the document known as Henry I’s Coronation Charter, by which Henry confirmed that the manorial demesne of all knights who owed military service was exempt from danegeld and labour services.91 The Laws of Edward the Confessor indicate that Rufus had tried to override this in 1096 when he raised a tax to take over Normandy during his brother Robert’s participation in the First Crusade.92 All this draws attention to the avarice and money-grabbing so forcefully criticized by the author of the ‘E’ Chronicle and quoted in Chapter 12 (‘The king and the principal men greatly loved, and over-greatly, greed in gold and silver, and did not care how sinfully it was got as long as it came to them . . .). When exemption from tax is considered alongside the transfer of burdens on to free and dependant peasants and the many mentions in Domesday Book of manors rented out at a level above their value, then

88 Sally Harvey, ‘Taxation and the Ploughland’, in Sawyer (ed.), Domesday Book, 86–103; idem, Domesday, 223–6. 89 Nick Higham, ‘Settlement, Land Use and Domesday Ploughlands’, Landscape History, xii (1990), 33–44. See also Roffe, ‘Domesday Now: A View from the Stage’, in Roffe and Keats-Rohan (eds.), Domesday Now, 44–6. 90 OV, iv, 172–3 (Annuente rege omnes carrucatas quas Anglice hidas uocant funiculo mensus est descripsit . . .). On this passage, see most recently Howard B. Clarke, ‘Condensing and Abbreviating the Data: Evesham C, Evesham M, and the Breviate’, in Roffe and KeatsRohan (eds.), Domesday Now, 247–75, at 259–62. 91 Judith A. Green, ‘ “A Lasting Memorial”: The Charter of Liberties of Henry I’, in Marie Therese Flanagan and Judith A. Green (eds.), Charters and Charter Scholarship in Britain and Ireland (Basingstoke and New York, 2005), 53–69, at p. 55; Green, Henry I, 45–9. 92 Leges Edwardi Confessoris, ed. O’Brien, 168–9.

death and legacy

507

it is easy to see what the author had in mind.93 Townsmen were also penalized through the innovation of the monetagium tax, an imposition that involved a collective payment to produce coin. Here, too, there was a massive increase on what had been taken before 1066. Since the costs could be passed on, the vast majority of those penalized would have been of English birth. So onerous was this tax that Henry I in his Coronation Charter agreed to abolish it.94 The only conclusion that can possibly be drawn from this material is that William deliberately set out to make conquest and the cross-Channel empire that followed on from it rewarding in ways designed to encourage those who had supported him to want to sustain it. The interplay between the obligation to provide security and rewards was crucial. In relation to the longer term, the early redistribution of land in England not only destroyed the fortunes of the vast majority of the topmost portion of the pre-1066 English elite, it radically adjusted the balance of power between king and individual magnates. No individual was anywhere near as wealthy as Godwine had been. Also, not only were substantial blocks of landholding created around the periphery, their holders were all linked by their other landholdings in England to their ancestral lands in Normandy and elsewhere in northern France.95 This gave a large number of powerful people an interest on both sides of the Channel, with many of these interests being focused on Normandy but not confined to it. Yet it was the same process that led to the creation of changed local and regional societies within which the resilience of thousands of human beings forged the future. It was all this that made the empire. Thinking of William as a conservative and cautious ruler but a ruthlessly violent one who rewarded generously and selectively is to identify two sides of the same coin. The mixture of fear, awe, trust, and acquiescence that he inspired was centrally important. THE LONG TERM An assessment of the significance of William’s career requires a perspective that encompasses centuries, arguably from c.900 to c.1250.96 The trend

93 Interestingly and provocatively on this subject, Stephen Baxter, ‘Lordship and Labour’, in Julia Crick and Elisabeth van Houts (eds.), A Social History of England, 900–1200 (Cambridge, 2011), 98–114, at 104–7; Harvey, Domesday, 200–9. 94 On this subject, fundamental is D. M. Metcalf, ‘The Taxation of Moneyers under Edward the Confessor and in 1086’, in Holt (ed.), Domesday Studies, 279–93. 95 For this general assessment, Fleming, Kings and Lords in Conquest England, 215–31. See also Clarke, The English Nobility under Edward the Confessor, 157–63. 96 Thus, Judith Green, ‘Kingship, Lordship, and Community in Eleventh-Century England’, ANS, xxxi (2009), 1–16, at 7–11.

508

william the conqueror

in recent publications to modify the thesis of the sophistication and power of the English kingdom before 1066 – central to the publications of the late James Campbell and the late Patrick Wormald – and instead stressing the multi-layered nature of this unity and sometimes its weaknesses, makes William’s conquest a landmark in a long history and not a phenomenon that can be analyzed in a narrow isolation from it.97 The invasion of 1066 was William’s personal response, and the collective response of those who chose to support him, to an inherent instability caused by recurrent dynastic crises. In England’s and western Europe’s history, it did not come out of nowhere; it was a product of the culture.98 After 1066, as was inevitable, the crises continued; the only unpredictable issue was the form that they would take. The claims of Cnut’s and Harold’s heirs were obliterated and those of Edgar the Ætheling’s kin integrated into the rule of William’s descendants by Henry I’s marriage to Edgar’s niece Edith/Matilda, daughter of Malcolm III and St Margaret. The way in which the crises were shifted into the political context of the French kingdom as well as the British Isles was, however, very important. Accidents were also always certain to happen. William of Malmesbury, now writing of events of which he was a contemporary, described the day on which Henry I’s only legitimate son William died in the wreck of the White Ship on 25 November 1120 as the time when everything seemed to change.99 The complexities that led to Henry I’s decision to try to produce peace through a marriage alliance with the counts of Anjou, the traditional enemy, led to a long cross-Channel civil war that had massive consequences for the histories of England, Britain, and France. This was arguably a consequence of the legacy left by William’s intervention in Maine in 1062–3. The ambivalence about the decisions that were taken was ever present. The author of the ‘E’ Chronicle wrote that Henry I’s decision to marry his widowed daughter the Empress Matilda to Geoffrey IV, count of Anjou, ‘offended all the French and the English’.100 Viewed in this way, the civil war between Stephen and Matilda was the product of crisis; or, rather, of yet another crisis in a very long series of crises in England’s history between the tenth and the thirteenth centuries. In terms of what has been called ‘English exceptionalism’, these recurrent crises can be seen as having had a decisively important influence. Counter-factuality must always be present in any estimate of the long-term historical significance of William the Conqueror’s life. What 97 Thus, for summaries of the state of play and new interpretations, Roach, Kingship and Consent, 226–7; Molyneaux, The Formation of the English Kingdom, 231–4. 98 For a crisp comparative assessment, Molyneaux, The Formation of the English Kingdom, 246–8. 99 GR, i, 762–3 (Iuuenculi ergo morte cognita res mirum in modum mutatae). 100 ASC, ‘E’, 1127.

death and legacy

509

happened because of what he did that would not otherwise have happened? Succession politics notwithstanding, we cannot, for example, say without multiple qualifications that his conquest brought England closer to Europe or that it destroyed a civilization in England. While being notably distinctive, both Normandy and England were polities that were already deeply integrated into the political and religious cultures of the medieval West. In terms of England’s relationship to Europe, the reorientation caused by 1066 was relative. Churchmen in England had always looked in that direction and not just because of the universal authority claimed by the papacy; and so, too, to a degree, had the laity. The consensus of the past forty years that the evolution of the pre-1066 English kingdom was profoundly influenced by the legacy of the Carolingian Empire makes the point very well. Nor can we say that changes would not have taken place if William had not conquered. Romanesque architecture would have come to England and Britain, as indeed it was already doing before 1066. Some variant of the castle as a royal and aristocratic residence would surely have done so as well; it is the form and the suddenness in which they came that were determined by the Conquest. The trend towards the construction of extremely large cathedrals and monastic churches drawing on inspirations that were often not in Normandy was present from the earliest period of conquest. Here William’s personal insistence on grandeur and display must have been a powerful influencing factor. The initiatives in relation to the introduction of knight service changed the relationship between kings and subjects. But it embedded them within a social and legal context that was different from Normandy and was a product of ‘English exceptionalism’. The Domesday survey would surely not have been required in a settled English kingdom, but we must not totally reject the assumption that its later iconic status was in any way anticipated in 1085–6; in context it was a remarkable, but traditional, response to unique circumstances. The ultimate counter-factual task is to try to imagine what the history of England and the British Isles would have been like if Harold had won the Battle of Hastings; and also how Normandy and France’s histories would have evolved if that had happened. It is even possible that the cultural conservatism of William’s conquest either fossilized or perpetuated aspects of the pre-1066 English state that might otherwise have evolved very differently. While raising a host of ethical humanitarian issues, an approach to the English kingdom’s long-term history that foregrounds crisis opens up an analysis connected to the role of war as a stimulus to change, and as a creative force. Put provocatively, as it has been: ‘War Makes the State, and the

510

william the conqueror

State Makes Peace.’101 One way of proceeding is to argue that a central strand in the creation of English and England’s identity, namely, the remarkable centralization of the medieval English state and the parliamentary institutions that eventually emerged from it, were a product of recurrent military and political catastrophe. Significant elements in the phenomena that became central to ‘English exceptionalism’ were forged in a cauldron of national disaster. It was the later twelfth-century crisis associated with the appearance after 1199 in King John of a ruler who lacked the personal qualities to command loyalty on the scale required that led to defeat in France and the rebellion that produced Magna Carta in 1215. Although Magna Carta can be seen as being in lineal succession to the promise of equitable treatment exemplified by promises of counsel and consent in King Æthelred’s time, by Edward the Confessor’s undertaking in 1041 to uphold the law of Cnut, by William’s coronation oath, and by the promises made by Henry I and by Stephen in their Coronation Charters and the latter’s ‘Oxford’ Charter of Liberties, it was another disputed succession and the collapse of the cross-Channel empire that produced the crisis.102 The rebels acted within deeply embedded English traditions, moving from a mixture of personal grievances and resentment at excessive taxation to demands that the Laws of Edward the Confessor be upheld and the Coronation Charter of Henry I be re-issued, to a statement that produced the charters of 1215 and their subsequent confirmations.103 For all that ‘English exceptionalism’ may not be that exceptional, with Magna Carta having a European context of its own it nevertheless remains a remarkably articulate expression of the dynamics of relations between a king and his subjects.104 It does exemplify something intrinsically English. But would the English kingdom have reached that point without its exceptional integration into a cross-Channel empire that was the direct result of decisions taken by William the Conqueror? William’s role as a maker of empire also sustained the hegemonic framework of power across the British Isles that had existed with fluctuating intensity from the time of King Æthelstan onwards. For this reason, ‘The First English Empire’ of Rees Davies’s Ford Lectures has to be seen as a 101 Thus, Ian Morris, War: What Is It Good For? The Role of Conflict in Civilisation, from Primates to Robots (London, 2015), 15–19, with the quotation at p. 15. 102 Regesta Regum Anglo-Normmanorum vol. iii, Regesta Regis Stephani ac Matildis Imperatricis ac Gaufridi et Henrici Ducum Normannorum, nos. 270, 271. 103 Green, ‘ “A Lasting Memorial” ’, 65; J. C. Holt, Magna Carta, 2nd edn. (Cambridge, 1992), 222–5; Maddicott, The Origins of the English Parliament, 118–19; David Carpenter, Magna Carta, with a New Commentary (London, 2015), 310–13; Nicholas Vincent, Magna Carta: The Foundation of Freedom (London, 2015), 32–4. 104 For observations on the European context, Holt, Magna Carta, 272–3; Carpenter, Magna Carta, 273; Vincent, Magna Carta, 34–5.

death and legacy

511

creation of the mid-tenth century and William as another strong English king who was perpetuating it. The annus terribilis of 1093 that was central to Rees Davies’s argument in this scenario becomes an episode in a long history of occasional violence and threats and of interaction and reward; Æthelstan had after all ravaged the kingdom of Scots and Harold had destroyed Gruffudd ap Llewelyn, claiming to act on Edward the Confessor’s behalf. William Rufus’s violent interventions in both Wales and Scotland look like attempts of a similar kind to coerce in Wales and to shape a disputed succession in Scotland in a way favourable to his and the English kingdom’s interests.105 This long-term analysis must be conducted in full recognition of the diversity and different cultures and identities that were always involved.106 William’s life ultimately creates an irresistible case for a book to be written on the English longue durée from c.900 to c.1250. Its starting point would be Alfred the Great’s victories that created a situation in which a modus vivendi had to be found in subsequent generations between the mixed peoples inhabiting England and the British Isles and the Scandinavian chieftains who had led invading armies and the peoples who had as a result settled. It is crucial that this whole process is now seen in terms of multiple interactions of peoples with multiple identities. In the same way that it is now argued that we should forget the stereotypical Viking, so too should we forget the stereotypical Norman.107 Such a book must emphasize diversity and multiculturalism. It must place English and British history within wider processes of European and global change. And it must take full account of human resilience in the face of crisis, violence, and oppression. As a terminus, c.1250 can be justified because it is close in time to the Treaty of Paris of 1259 by which Henry III renounced his claim to rule Normandy, Anjou and most of Aquitaine. It is also in the midst of transformations out of which parliamentary institutions and different mechanisms for the interactions of kings and subjects evolved.108 The continued rule of the English kings in Gascony perhaps makes this cut-off date a slightly unconvincing one since the legacy of the cross-Channel empire was an element in the causes of the Hundred Years War.

105 R. R. Davies, The First English Empire (Oxford, 2000), 4–9; Molyneaux, ‘Why Were Some Tenth-Century Kings?’, 77–82. 106 Green, ‘Kingship, Lordship, and Community’, 7–11. 107 For powerful clarion calls in this direction, Janet L. Nelson, ‘England and the Continent in the Ninth Century’, TRHS, 5th ser., xiii (2003), 1–28, at 28; Dawn M. Hadley, The Vikings in England (Manchester, 2006), 278–9; Lesley Abrams, ‘Diaspora and Identity in the Viking Age’, EME, xx (2012), 17–38, especially at pp. 34–8. 108 For a survey that confirms the importance of taking ‘the long view’, J. R. Maddicott, The Origins of the English Parliament, 924–1327 (Oxford, 2010). The quotation is taken from the book’s dedication. Laura Ashe is working on a book that will take this general approach.

512

william the conqueror

William’s life and his achievements were the products of a multicultural world in which there were multiple diversities and many common factors. All this also means that the last legacies of so-called ‘Whig history’, which still find their way into major general books on the subject, should be abolished.109 William the Conqueror’s life exemplifies forms of statesponsored violence deployed in what was proclaimed to be a legitimate cause. It deserves close attention and to be understood on these terms. But it is only part of a much larger story.

109 For an example of the continuation of Whig history alive and kicking in a major book intended for a general audience, Robert Tombs, The English and their History (London, 2014), 47 (‘Though the Conquest preserved many institutions, it yet changed the way England was ruled. Before, government had depended largely on consent; now it was dominated by force’). This quotation alone shows the importance of inserting the debates to which this book has drawn attention into public discourse. The crucial point, of course, is to think what might be meant by consultation and consent. Contrast the sensitive appraisal that takes the long view and looks forward to parliamentary institutions, Maddicott, The Origins of the English Parliament, 32–67.

E P I LO GUE

Auferre trucidare rapere falsis nominibus imperium, atque ubi solicitudinem faciunt, pacem appellant.1

William the Conqueror imposed his personal and political domination on the duchy of Normandy and after 1066 he and those associated with him transformed a violent conquest into peace. The cross-Channel empire they created endured for more than a century. The quotation from Tacitus encapsulates all the contradictions and controversies inherent in these processes. William’s life is ultimately a parable on the eternal moral conundrum of the legitimacy of violence used to achieve what its perpetrators believe to be a justifiable end. His contemporaries knew this. And so did those who wrote about him afterwards; Orderic and William of Malmesbury above all wrote aptly and poignantly on the subject. At the time that this book has been written, the issues are every bit as much with us as they have always been. A response to François-Théodore Licquet’s unhappiness as quoted in the Prologue (‘cold seizes us. William frightens us, and, we are, in sum, almost reduced to lamenting the good that he did’) is that, in the face of William’s life, we cannot afford to freeze.2 Nor can we go along with Douglas’s paradox (‘There was in fact an element of paradox in his character’) or posit the issues, as he did, in terms of the extremes of ‘a crude ruffian, or a sanguinary brute’ and ‘he won the respect of many of his most illustrious contemporaries’.3 Cultural history and biography working in tandem can provide an escape route from all this and a road to clarity. In the end William’s life and his actions have a specific place within the 1 Tacitus, Agricola, ed. A. J. Woodman, with contributions from C. S. Kraus (Cambridge, 2014), 55, ch. 30, with commentaries on this famous passage at pp. 15–25, 243–4 (‘To plunder, slaughter, and theft, they falsely give the name rule; they make a desolation and they call it peace’). I am grateful to Tom Licence for a discussion of the translation of this passage. 2 Depping, Histoire de la Normandie, ii, 173–4 (‘le froid nous saisit; Guillaume nous fait peur, et nous sommes presque réduits à déplorer le bien qu’il a fait’). 3 Douglas, WC, 375.

513

514

william the conqueror

infinitely long history of violence’s place in human life and human history.4 The rejection of Douglas’s sub-title and of any sub-title at all as justified in the Prologue is a part of this. This book’s basis was always the proposition that, by locating its subject within cultures and context, it was possible to proceed from the individual to an understanding of his or her life and its historical significance. Two certainties about William the Conqueror are that he was physically a large man and that he possessed a commanding presence in war, at the hunt, and at court. William of Malmesbury, as we have seen many times, a writer more interested in biography than any of the others we rely on, paints a portrait of a powerful man with exceptional skills on horseback, who could draw a bow while controlling the horse only with his legs, who dressed magnificently, entertained lavishly, and at court terrified all around him with oaths and a loud voice.5 In the section of the De Obitu Willelmi that describes his personal qualities that is plagiarized almost word for word from Einhard’s Life of Charlemagne, the word describing Charlemagne’s voice as high-pitched was changed to say that William’s was harsh and rough.6 This physicality is confirmed by the left thigh-bone examined when the tomb in the abbey of Saint-Etienne of Caen was opened in 1983. It belonged to a large, strong, man, approximately 173–4 cm in height (5ft 10 in).7 One aspect of Malmesbury’s portrait that is at first reading perhaps surprising is the statement that William lived as religious a life as was possible for a layman. Another, one possible story to the contrary notwithstanding, is that he thought him to have been devoted to his wife. It was probably characteristics such as these that were at the heart of Douglas’s paradox. Yet Malmesbury, who knew the snares and the temptations of worldly life and that kings could never measure up to the standards that were ideally required, even if some performed better than others, was, as we have seen, at times ambivalent about him.8 This ambivalence resides within the complexities of the interplay between personal interpretations of legitimacy and rule and the means employed to put them into action, along with the historian’s own ethical standards and the religious and secular cultures of the later eleventh and earlier twelfth centuries.

4 Thought-provokingly relevant is Steven Pinker, The Better Angels of Our Nature: A History of Violence and Humanity (London, 2011). 5 GR, i, 492–3, 508–11. 6 GND, ii, 188–9 (uox clara is changed to uox rauca). 7 Jean Dastugue, ‘Le femur de Guillaume le Conquérant’, AN, xxxvii (1987), 5–10. 8 For an informative essay on this theme, Björn Weiler, ‘William of Malmesbury on Kingship’, History, xc (2005), 3–22, and, for William, see esp. pp. 13–14.

epilogue

515

From a modern perspective, the expression of apparently sincere religious belief, the performance of extensive charitable works, and the marital fidelity are just the more obviously laudable parts – at least for most – of a spectrum of norms and values that is recognizable to us, and yet is in many aspects of their performance different from the expectations of the second decade of the twenty-first century. In terms of eleventh-century values, they are certainly compatible with the violence that characterizes William’s exercise of power, even if this violence was at times taken beyond the norms that many who wrote about him thought acceptable. William in fact embodies many of the complexities of his era. He lived at a time when the moral reform of lay and clerical lives that often resulted in the disruption of well-established relationships and aggressive violence in the service of Christianity were being driven forward by some at the highest levels in the Church. And also in times when the papacy claimed the right to sanction the removal of rulers it thought inadequate. And in times when honourbased violence and the culture of feud were widely prevalent as values in the secular world. But he also lived in a period when the role of the state as the preserver of peace and justice was fully and persuasively articulated and when the more extreme pronouncements of the pre-Gregorian and Gregorian papacy were being forcefully contested. While William fits well into this turbulent cultural world, it is ultimately only through biography that we can truly understand his place within it. Frank Barlow’s observation noted in the Prologue that, because of the relative absence of colourful anecdotes, William’s life is less accessible to a biographer than those of his sons Robert and William, must to an extent be true.9 But in many respects, it is not. The differences of emphasis in the sources can be explained and other ways of understanding the man have to be found. While the preponderance in them of stories of dispute settlement, confirmations of land grants, and ecclesiastical matters in general is significant, alongside – of course – the central narrative of the conquest of England, it is also a major distortion. As also is the tendency of medieval writers to attribute currently fashionable qualities to subjects who personified success. For authors schooled in the classical authors, the issues of human agency and good fortune, alongside divine providence, were all part of the essential discourse. Thus – to cite one example – Hugh of Fleury, in all likelihood writing between 1114 and 1125 in a work addressed to William’s granddaughter the Empress Matilda, described William as a king who exceeded all others in good fortune and wisdom and whose generosity of spirit and magnificence were truly remarkable, an assessment that focuses on norms that might be described in the emerging cultural 9

Thus, in the Preface to the 1999 Yale edition of Douglas’s WC, p. x.

516

william the conqueror

language of the twelfth century as both chivalric and courteous.10 It is arguably better to put such material to one side in the same way as the remarkable stories that developed around William’s birth. They were ultimately products of later generations’ awareness of his remarkable life. Care is also essential in interpreting the many anecdotes mentioned throughout this book. The danger within them is that they are mainly showing that performance feeds performance and that reputation demands that expectations be fulfilled. The story of William jokingly threatening to drive a knife through an abbot’s hand when confirming a grant at Winchester in 1069 may be nothing more than the type of theatre that was required of him.11 It shows that William knew how to dominate his court and that his way of doing so must demonstrate a type of political intelligence, but it may not say much about his personality. The significance of the debates that have followed on after Gerd Althoff ’s formulation of the concept of Spielregeln and discussions of ‘anger’ have among other things shown that, even if there were ‘rules of the game’, not everyone necessarily understood them and that clever people knew how to manipulate them. William and others were playing roles that their social positions assigned to them. It is what we can see beneath the surface that is crucial.12 So what can we confidently say that we know about William? In writing as he did, Frank Barlow perhaps had in mind that we know so little about the homosocial camaraderie of William’s military household and the life of the hunt in William’s time, even though both of them were central to his life. This is obviously regrettable when we reflect on the sobering thought that it is a reasonable estimate that he must have spent over 50 per cent of his adult life either at war or preparing for war. Most of this time would also have been spent almost exclusively in the company of men. It is even more sobering to reflect that the same statistic would have applied to many of his aristocratic contemporaries and to the many less socially eminent members of William’s military household. That we know much less about William’s military household than we do about those of his sons is partly because Rufus’s household was portrayed as scandalous by some of the ecclesiastics who wrote about him in order to present him as an immoral oppressor of the Church, and partly because,

10 Hugh of Fleury, Liber qui modernorum regum Francorum continent actus, 376 (Nullus rex nostrorum temporum hoc Guillelmo fuit felicior ac moderacior). For discussion of the date of composition, Bauduin, ‘Hugues de Fleury et l’histoire normande’, 157–74, at 162–3. 11 Regesta, no. 232. 12 Bates, ‘Anger, Emotion, and a Biography’, 32–3.

epilogue

517

through Orderic, we know so much more about Henry I’s reign.13 It is, however, hard to believe that the mores of this male-dominated community were in any way significantly different in William’s reign. That two of his sons, Robert and Henry, fathered illegitimate children and that the sexual proclivities of William Rufus have been endlessly discussed, suggests that William’s apparent sexual puritanism had a limited influence on those around him; if indeed it had any influence at all. What would have mattered in an environment in which physical prowess and a domineering presence would have been centrally influential were his military record, the multiple opportunities to distribute rewards available to him, and the physicality to which William of Malmesbury drew attention. It all indicates a mastery of the culture of military masculinity and great skill as a soldier and leader of men. His willingness to authorize looting shows that William recognized current norms in relation to reward and morale. Although we have to wish that we knew more of this aspect of his life, much can reasonably be surmised. While some of his personal values may have made him distinctive, this was an environment in which he was very much at home and whose boisterousness he tolerated and participated in. The characteristic strikingly present almost from the beginning of William’s life is his sense of entitlement. The conquest of England would not have taken place without it. It was the ultimate demonstration of his ability to convince others that he could obtain what he wanted; and often, therefore, what they wanted, or at least hoped for. This capacity to project onto others a sense of confidence and capability must have been present from an early time. Arguably it was operating before Val-ès-Dunes. It is certainly notable how isolated Count William of Arques was within Normandy in 1053–4. Aside from the possible support of Count William’s brother Archbishop Malger, no Norman of any political significance allied with him. The social importance of the allies Count William recruited from outside Normandy shows that William was already frightening people by his mid-twenties. The invasion of Normandy in 1053–4 was surely an attempt to rein him in. Whatever form the offer made by Edward the Confessor took, it played into this sense of entitlement and would have had consequences for the way William behaved. The pursuit of a succession to which one believed oneself entitled was a matter of honour and a facet of the culture of feud. It was an essential aspect of rule and the ethos and practice expected of an 13 Fundamental for the military households is J. O. Prestwich, ‘The Military Household of the Norman Kings’, EHR, xcvi (1981), 1–37 (reprinted in Strickland [ed.], Anglo-Norman Warfare, 93–127). Several astute points about these male-dominated worlds are made in Gillingham, William II, 49–58.

518

william the conqueror

eleventh-century ruler, and therefore something that William would have unreservedly believed himself obliged to pursue and which others would have expected of him. Since his aspirations would have been publicly known and his potential capacity to pursue and achieve them widely recognized, the chemical mixture of the personal and the cultural must have been central to the politics of northern Europe throughout the 1050s and early 1060s. It is reasonable to think that William knew how to comport himself in a way that kept the associated possibilities in the forefront of people’s minds. When it comes to the specific, William consistently displays the ability to assess situations and to decide when it was necessary to act quickly, or, at other times, not to act at all. This is especially evident in his conduct of warfare. But it is surely certain to have been the case in other situations as well; William of Malmesbury did, after all, write of him as being quickwitted and a good judge of character.14 At Mayenne in 1063 and Exeter in 1068 he had his forces storm well-fortified sites despite the difficulties that this presented. Both the campaign to Exeter and the subsequent ones to the north of England in 1068 and 1069 were very rapidly organized and caught William’s opponents off balance. He decided that the north of England needed to be destroyed in 1069–70 because enemies beyond his reach, namely, the Danes and the Scots, had to be shown the consequences of trying to use it as a base, and its inhabitants the price of welcoming them. On the other hand, after Val-ès-Dunes, Guy of Brionne could be left at Brionne to submit at his leisure because he was completely isolated. In 1057 William held off from confronting Henry I and Geoffrey Martel until the opportunity to massacre a part of their army at Varaville gave him a decisive advantage. In the 1080s the vicomte Hubert could be left at SainteSuzanne, albeit, like Guy, under surveillance. Throughout William’s life, he acted quickly in ways that disrupted opponents while at the same time consistently marshalling the necessary resources in advance. Numerous episodes illustrate that he allowed no one who showed signs of defying him to be left in peace for long. We can, I think, take William’s political intelligence for granted. Indeed, there is much that testifies to supreme political intelligence. He certainly possessed the ability to extricate himself from difficult situations, albeit, as, for example, in the cases of the contretemps with Archbishop Ealdred over the plundering of his goods and the arrival of the killers of Earl Edwin, we might think he did so in ways that passed responsibility on to others when he should have shouldered at least some of it himself. He was undoubtedly fortunate in his marriage, with what we know suggesting a considerable 14

GP, i, 50–1; GR, i, 416–17, 492–3.

epilogue

519

degree of compatibility between the partners. Matilda must also have possessed considerable political know-how. While the details are mostly hidden from us, she must have been a strong personality. Her frequent association in charters as a partner in decision-making with William goes far beyond the formulaic. It surely shows that she was often centrally involved in the practice of rule in areas beyond the exclusively military. As a ruler William was hugely successful in harnessing the secular and the religious to his advantage. A succession of popes supported him, Alexander II seemingly unreservedly and Gregory VII consistently, despite the frequently expressed qualifications about William’s conduct by those who were both the pope’s friends and enemies. William’s qualities were also of a kind that retained the loyalty of some very able people, some of whose longevity was a great bonus for him. Of the individuals who emerged during his adolescence and early adulthood in Normandy, among the clergy, Lanfranc and Bishop Geoffrey of Coutances, and, among the laity, Count Robert of Mortain, Roger de Montgommery, Roger de Beaumont, Richard fitz Gilbert, and William de Warenne all outlived him. Their dramatic falling out notwithstanding, his brother Bishop Odo must be added to the list; they do, after all, seem to have worked well together for over thirty years, with the period of Odo’s greatest responsibility coming just before their quarrel. William’s court and his chapel also nurtured the early careers of many individuals who were to be at the heart of events in the next generation. All this takes us out of the world of the court and the hunt into the wider western European world. There is evidence that William was good at listening and that he could change his mind when faced with strong arguments, as in the extensive narratives of the first stages of the Primacy Dispute or in the less weighty matter of the dispute between Bishop Giso of Wells and Abbot Thurstan of Glastonbury. He was capable of apparent acts of kindness, as when he insisted that the abbey of Abingdon make provision for a soldier mutilated by pirates on a Channel crossing, albeit that it was the abbey on William’s orders that made the provision rather than the king himself.15 The many narratives we have of him doing justice describe verdicts reached logically, usually in consultation with others, and with a concern for equity. He showed a liking for clear-cut decisions and was sometimes ready to impose a verdict on those who were vacillating. He could be witty, as when joking about concepts of Norman identity to the discomfiture of the monks of Saint-Florent of Saumur. He respected authority and used it to his advantage. Archbishop Lanfranc’s letters, to cite from many examples, show William involved in determining the religious status of English women 15

Historia Ecclesie Abbendonensis, ii, 8–9.

520

william the conqueror

who had taken refuge in nunneries on the basis of canon law.16 The decision on Canterbury’s Primacy over York was an issue on which a decision was reached within existing scripts and rules with regard to English custom as it was believed to have existed before 1066 and in relation to canon law. To fit William’s life and early development into any simple framework of child and adolescent psychology is impossible. We can – and should – speculate and try to assess what might have been the prime factors, but we should not be dogmatic. William was an only male child who must have showed promise early. His parents’ relationship seems to have been a stable one; they may even have been devoted to each other. He was brought up in an environment in which overt religious practice was important. He lost his father on a pilgrimage while he was young and he seems to have revered his memory; according to William of Malmesbury, he tried in later life to have Duke Robert’s remains brought back to Normandy.17 He showed exceptional favour to the children of his mother’s marriage to Herluin de Conteville and he looked after the couple’s foundation of the abbey of Grestain. It was, after all, Herleva and Herluin who went with him to welcome his bride to Normandy. The act of savagery at Alençon may have been motivated by filial devotion and affection, rather than a reference to the circumstances of William’s birth. The stories that he would have been told about the insecurities of his father’s early years would surely have influenced him towards ways of operating that would protect him against such troubles. Equally, Robert’s successes as a ruler must have been influential, notably the pacification and extension of Normandy’s frontiers and his extensive grants to the Church, both of which William continued. We underestimate Duke Robert’s importance in the development of the Norman duchy and on his son at our peril. When the general nature of periods of medieval underage rulership are taken into account, it is reasonable to think that William was well looked after by his guardians and tutors. But he was favouring his own friends to the exclusion even of those who had treated him well some time before his eighteenth birthday. The turbulence of his adolescence may well have been of great importance in shaping his later behaviour, but not as a passive victim of it.18 The breakdown of the alliance with the kings of Lanfranc Letters, no. 53. GR, i, 504–05 (Patris memoriam quantis poterat occasionibus extollens). 18 Although I am largely forsaking references to earlier historians of William’s life, I cannot resist inserting that Barlow, William I and the Norman Conquest, 12, has a similar emphasis to the one here on this aspect of William’s early years (‘Probably even more influential on William’s character than his illegitimate birth were the perils and treacheries of his boyhood . . . At the age of fifteen he was ready to fight for his inheritance, to recover the rights lost since his father’s death.’). 16 17

epilogue

521

France around 1052 was probably an event of great importance. It ended a mutually supportive alliance that had lasted for well over half a century, which was central to William of Jumièges’s Gesta Normannorum Ducum, and, with it, the role of the rulers of Normandy as the upholders of legitimate authority as they had been in the days of William’s highly regarded grandfather Duke Richard II. It left William isolated in northern France and facing formidable enemies. It is surely no accident that William of Poitiers chose Count Geoffrey Martel as the man to praise the young William.19 Martel, too, had been a political pariah before William came on the scene, a man whose acts attracted and terrified, the winner of battles and the conqueror of territories. What William would have learnt in his twenties and thirties is that northern France mattered hugely, that political control was indispensable to rule in such a world, and therefore that England had to be secured quickly and decisively. It is possible that William’s early years ultimately influenced him into doing things with an almost obsessive correctness and care. His approach to all the big life-shaping situations was built on meticulous preparation and then on thorough application. This applies to the politics and legitimacy of his marriage, to the invasion and conquest of England, and to the projects like the one that produced the Domesday survey. William possibly had an over-developed sense of responsibility, with his very thoroughness being transformed into ruthlessness in certain circumstances, and into terrible violence in some of them. He must be seen as a religious man who took the round of church services seriously, who showed respect for saints, who listened to monks and priests, and who built magnificently at Caen, at Winchester, and elsewhere. He conversed attentively with religious men, St Anselm being one of the most frequent visitors; according to Eadmer, William became a changed man in Anselm’s company. The subjects on which he listened to Anselm may well have been on the need to keep the peace for the good of humanity and to protect the Church, and the importance of prayer and charity.20 If Orderic’s story of the monk Guitmund lecturing William on the conquest of England as an act of robbery is true, then William did listen to painful truths about his conduct.21 And, likewise, if he absorbed the moral messages of the Bayeux Tapestry fables and zthe passage of Augustine’s City of God to the effect that ‘with justice left out, what are kingdoms (regna) except great robber bands GG, 16–17. The Life of St Anselm, Archbishop of Canterbury by Eadmer, ed. R. W. Southern, OMT (Oxford, 1962), 56; Walter Fröhlich, ‘St Anselm’s Special Relationship with William the Conqueror’, ANS, x (1988), 101–10, at 105–9. 21 OV, ii, 268–81. 19 20

522

william the conqueror

(magna latrocinia)’, then he may also have been given pause for thought by them.22 His level of theological understanding may, however, have been a limited one. It was precisely in relation to this subject that the author of the De Obitu Willelmi chose to omit the passages from Einhard’s Life of Charlemagne that refer to Charlemagne listening to readings from works of theology. Deployed comparatively, Henry Mayr-Harting’s essay on Charlemagne’s religion suggests that we should probably see William’s as propitiatory and ethical and characterized by extensive works of charity which, so he and others hoped, would mitigate the many sins which he did commit, and on occasion must have believed he had to commit.23 It is notable that while he espoused reforming causes such as the abolition of clerical marriage and the selling of people into slavery, he did not push them, drawing back from taking action when both threatened to create difficulties for him. The almost consistent rejection by Norman churchmen of the ideas of Berengar of Tours on the Trinity from c.1050 onwards points to a conservatism that runs through William’s whole life. And, in relation to England after 1066, as we have seen, Eadmer’s famous passage about William controlling contact with the papacy does not demonstrate that he was opposed to visits by legates to the lands he ruled or of senior churchmen from Normandy and England to Rome. It does show a kingly paternalism in which most of the clergy shared and which was starting to become somewhat outdated during his life. It is ultimately unfortunate that we know so little about William’s private world and inner thoughts in comparison, for example, with Charlemagne’s.24 As is often asserted, the commissioning of William of Poitiers must indicate a concern to defend himself against criticisms that were being made. Whether this was out of conscience or for propaganda are both further subjects that are open for debate. The absence of reflective writings from William’s ecclesiastical intimates of the kind produced by Alcuin leaves us less well informed than we would like to be on all of these subjects. If Charlemagne had indeed repented of his destruction of the Saxons, then it is possible that William was influenced in the direction of doing the same and those around him wrestled with the moral problem of his excessive

22 City of God against the Pagans, vol. ii, Books 4–7, trans. William M. Green (London, 1963), ii, 16–17. The potential importance of this passage alongside the Tapestry fables is emphasized in Pastan and White, The Bayeux Tapestry and its Contents, 182. 23 Henry Mayr-Harting, ‘Charlemagne’s Religion’, in Peter Godman, Jörg Jarnut, and Peter Jonanek (eds.), Am Vorabend der Kaiser Krönung (Berlin, 2002), 113–24. 24 Janet L. Nelson, ‘Did Charlemagne have a Private Life?’, in Bates, Crick, and Hamilton (eds.), Writing Medieval Biography,15–28, at 18–25.

epilogue

523

violence.25 In the end, William arguably belongs in a pre-twelfth-century world that had not fully rediscovered Augustine’s distinction between actions and intentions. His and Matilda’s extraordinarily western European-wide religious patronage after 1066 and the sponsorship of the redevelopment of York towards the end of his life do suggest that William did take positive steps to atone for his most terrible sins. His son William Rufus’s visit to York in the first months of his reign must indicate that those close to his father believed that this is what William would have wanted to happen. The anecdote in the Vita Lanfranci of Lanfranc sitting at the king’s side at a great banquet and witnessing a court jester lauding the king’s magnificence with the words ‘Lo! I see God’ opens up several perspectives on all this. Lanfranc telling the king firmly that this was blasphemy, and, by implication, that the terrible fate of Herod Agrippa awaited William unless the jester was punished (as he subsequently was), shows William as responsive to religious advice, yet needing to have both the wrong he had committed and what was required of him spelt out.26 Since the jester was providing a performance that he thought would please William, it must demonstrate that William viewed himself as a special king and expected to be treated publicly as one; it may well indeed indicate that the maintenance of this image was uppermost in William’s mind, as well as the limited theological understanding already referred to. It could suggest that William’s predominant tendency was to follow his own instincts and, if necessary, fall into line later. This incident, or others like it, could lie behind the lengthy passage that William of Malmesbury excised from his later version of the Gesta Pontificum. Lanfranc, so he said, had no choice but to tolerate William’s extraordinary arrogance. But he had the skill to chip away and sometimes to make him change his mind, with good results since monasticism flourished and the king would listen to those whom he regarded as especially devout.27 If Lanfranc, on whose counsel William was manifestly very reliant, struggled to get his way, then how difficult must it have been for others to make headway against William, for example, for the laity, over whom, as

25 Henry Mayr-Harting, ‘Alcuin, Charlemagne and the Problem of Sanctions’, in Baxter, Karkov, Nelson, and Pelteret (eds.), Early Medieval Studies in Memory of Patrick Wormald, 207–18, at 215–18. 26 Vita Lanfranci, 708–9. On the incident, Margaret Gibson, ‘The Image of Lanfranc’, in Lanfranco di Pavia, 21–8, at 25–6; Cowdrey, Lanfranc, 187. 27 For the whole passage, GP, i, 90–7. Note especially sunt modo predicanda magnanimitatis exempla quae quondam dicebantur presumptionis facinora (p. 96). For commentary, Michael Winterbottom, ‘A New Passage of William of Malmesbury’s Gesta Pontificum’, Journal of Medieval Latin, xi (2001), 50–9, with Malmesbury’s text at pp. 52–5.

524

william the conqueror

Malmesbury believed, William had established a total dominance.28 If a similar intransigence as displayed towards Lanfranc determined his relations with his eldest son Robert, as it surely did, then their estrangement at the end of William’s life is readily explicable; as also is the messy situation that unravelled as soon as his sons decided to fight each other. Matilda, it must be thought, probably had the same capacity to chip away at her husband’s stubbornness that Lanfranc possessed. It may well be that her death left William somewhat rudderless in his last years, particularly in the maintenance of personal relationships. Similar characteristics may well have shaped Gregory VII’s treatment of William. Although his attitude to him has something of the faute de mieux about it, it is possible that even this zealous man was overawed by William’s reputation and achievements and drew back from confrontations because of this. Like others, he certainly found William’s mixture of qualities baffling and feared very greatly the consequences of William transferring his support to his enemies. Although William of Malmesbury wrote that William was spurred on by the achievements of Robert Guiscard, a member of the same gens, regarded by William as his social inferior and therefore as someone whose deeds he had to exceed, William manifestly did not see himself as a fully signed-up participant in the European-wide diaspora of the Normans. He was conscious of the remarkable events that were taking place in southern Europe, but his priorities lay elsewhere.29 His personal identity and his conceptual ethnicity were ultimately defined politically, with his prime responsibilities being those of a ruler. William was dux or comes – to mention his most used titles – Normannorum. In other words, while he belonged to the people who lived in the multi-ethnic society of Normandy, most of whose inhabitants would define themselves as Normans, but many of whom would not, he positioned himself in relation to centuries-old ideals and practices of rulership that encompassed and transcended the ethnic. William’s adoption of the title patronus Normannorum after 1066 redefined a role that continued to have elements of the ethnic, but positioned his responsibilities in the context of empire. Stories such as those involving the likes of Arnold d’Echauffour, Abbot Robert of Saint-Evroult, Count William Werlenc of Mortain, the monk Guitmund, and Edgar the Ætheling, show that all of them made their way to southern Italy because they got on the wrong side of William. After the fund-raising missions of Bishops Geoffrey of Coutances and Ivo of Sées GP, i, 90–1 (Nam preter secularia quae ad libitum suum omnino torquebat). GR, i, 482–3. On the use of the term ‘diaspora’ in relation to the Normans, Bates, The Normans and Empire, 29, 42–4. It is crucial to my arguments here that the term be used with its technical meaning, rather than as a general description of a migration. I hope to return to this subject in a separate publication. 28 29

epilogue

525

and the papal legation of Abbot John of Fécamp in c.1050, there is no evidence that he positively encouraged anyone to go to southern Italy. These journeys and the dramatic case of Bishop Odo show that the prime influence on William as a ruler was the demand for loyalty. To try to desert one’s allocated post, as Odo was doing, was a terrible crime. All this is the obverse side of William’s reluctance to forgive. More generally, all this shows that William’s life and achievements belong conceptually more within the analytical framework of the evolution of the medieval state and the cultures of rule that animated it rather than within the diaspora of the Normans. In terms of the relationship between William’s sense of his personal role and identity, the conquest of England, and the broader processes of the making of Europe, it is crucial to think primarily in terms of ‘regnal solidarity’, a phrase coined by Susan Reynolds.30 While what might constitute such ‘regnal solidarity’ is almost endlessly variable, in modern parlance what happened after 1066 was effectively the takeover of one state by another one on terms that emphasized plurality and legitimacy. It is this that makes the link between William’s personality and the implications of those two phrases that appeared in the Prologue – ‘an event unparalleled in European history’ and ‘small-scale elite transfer’ – and their overt reference to the remarkable comprehensiveness of the takeover of England after 1066.31 It also means that the phrase ‘the Norman Conquest’ works as a political definition; but that, as a cultural one, it does not. In terms of ‘the image of the conqueror’ as influentially explored by Robert Bartlett within the framework of the making of Europe, William’s life sits within the history of state-building rather than of adventure. His personal non-participation in the diaspora of the Normans means that it must be thus. Although Anna Comnena’s famous portrait of Robert Guiscard’s son Bohemond – one of the great opportunists of the southern European world and of the First Crusade, which is at the heart of ‘the image of the conqueror’ – shows that not dissimilar characteristics were required for both roles, there were major differences.32 For all that multiple literary devices are at work in this portrait of Bohemond, whom Anna was seeking to present as the very antithesis of the civilized values that her father represented, the same physicality and capacity to manipulate

30 Susan Reynolds, Kingdoms and Communities in Western Europe, 2nd edn. (Oxford, 1997), 256–66, with the actual phrase at p. 262. 31 Gillingham, ‘Problems of Integration’, 85; Heather, Empires and Barbarians, 23, 298–9, 327, 350–1, 614. 32 On Bohemond, Robert Bartlett, The Making of Europe: Conquest, Colonization and Cultural Change 950–1350 (London, 1993), 90.

526

william the conqueror

situations were required of William.33 Both are representative of the cultural phenomenon of medieval warrior kingship/leadership. But the context was different, as are the thought-worlds and the norms, scripts, and rules that govern them. Something resembling the ‘War Makes the State, and the State Makes Peace’ slogan mentioned at the end of the previous chapter was deployed during William’s lifetime by Fulcoius of Beauvais, a poet who wanted to flatter him: ‘That man is David, “strong in hand”, as the English bear witness, the same a Solomon, “peacemaker” as the same bear witness. He beats back, he withdraws, he heals where he wounds; both war and peace obey him sympathetically.’34 But for others, as we have seen many times during this book, William was the destroyer of lives, the bringer of a disaster of unnecessary proportions to the English. As Augustine would have surely recognized, justice had been brought to the world of the robber bands. But what sort of justice? And how much justice? The chilling passage in the Liber Eliensis that ‘with ferocious resolve, he unrelentingly kept laying waste to the country everywhere, kept slaughtering people and doing many wrongs, prompted merely, as we have said, by suspicion of his new kingdom’ has to be set alongside the lines of the poet. It underlines the undeniable violence of the first years of William’s rule in England.35 The scale of human tragedy is beyond our knowing. For all his wobbles between worldliness and censoriousness and apparent changes of mind, across the span of his writings, William of Malmesbury probably thought that William had gone to Hell. In the Gesta Regum, he expressed the opinion that William’s only fault was his greed. He was ready to suggest that even this might be excused on the grounds that William had many enemies. However, in the draft that he later removed from the Gesta Pontificum referred to earlier, he wrote that he now doubted whether God would be able to forgive William. In a lengthy passage greatly influenced by Eadmer, he dwelt on William’s excessive interference in the affairs of the Church and on his arrogance. He even went as far as saying that he regretted what he had written in the Gesta Regum. In the final version of the Gesta Pontificum, he replaced this with a simple statement that William treated Lanfranc with respect, and added nothing more.36 When he took a longer-term perspective, he could certainly say that William’s achieve33 For Anna’s treatment of Bohemond, ‘doppelgänger to Alexius’, Penelope Buckley, The ‘Alexiad’ of Anna Komnene (Cambridge, 2014), 208–14, with the quotation at p. 211. 34 ‘Fulcoii Belvacensis epistolae’, ed. Colker, 245 (trans. van Houts, The Normans in Europe, 132). 35 Liber Eliensis, 172 (trans. Fairbrother, 204). I have followed Janet Fairweather’s translation. 36 GR, i, 508–9; GP, i, 90–7. For a commentary, GP, ii, 44–50.

epilogue

527

ments had brought ruin to the English. He was the inferior of Alfred, Æthelstan, and Edgar, kings in a late ninth- and tenth-century Golden Age. Towards the end of the 1130s and for all that he was of mixed ethnic birth, William of Malmesbury gave vent to an anger that seethed above and below the surface in all of his works, writing that the Normans as a people were boorish and violent, and inferior to the English.37 As we have already seen, in his interpretation of the prodigy of the conjoined twins, he could describe the union of Normandy and England as an offence against Nature.38 Orderic’s treatment of William’s funeral and burial is arguably a metaphor that says the same. There was a centuries-old hagiographical tradition that tombs expanded to accommodate saints. When this did not happen, it was presumably an allegorical statement that the deceased had been exceptionally sinful.39 In the coda to his symphonic treatment of William’s life and death in the deathbed speech he composed for him, Orderic inserted the word ‘false’ into the famous passage from the Psalms, ‘Put not your trust in princes’. Whether the passage should be read as a commentary on all princes whose conduct did not conform to Orderic’s religious ideals or on William personally is in fact not entirely clear. But it must be significant that the specific faults he tells his readers to avoid are putting faith in iniquity and becoming rich through robbery, and then aiming to become even richer, the same sentiments as had been expressed by the writer of the ‘E’ Chronicle.40 These words also straddle the two worlds of monastic morality and the exercise of power in the secular world. We know that Orderic thought carefully about power.41 For all that he might have believed monastic ideals and secular power irreconcilable, he was also saying that they might not be and that they must interact. He must surely be telling us that others had ruled with much more respect for humane principles than William had done. 37 Willelmi Meldunensis monachi liber super explanationum Lamentationum Ieremiae prophetae, ed. Michael Winterbottom and R. M. Thomson, CCCM (Turnhout, 2011), 84, lines 2,194–212. For the development of Malmesbury’s opinions, Michael Winterbottom, ‘William of Malmesbury and the Normans’, Journal of Medieval Latin, xx (2010), 70–7; R. M. Thomson, ‘William of Malmesbury’s Diatribe against the Normans’, in Brett and Woodman (eds.), The Long Twelfth-Century View of the Anglo-Saxon Past, 113–21. 38 GR, i, 384–7. 39 For references, David Bates, ‘William the Conqueror’s Earliest Historians and the Writing of His Biography’, in Bates, Crick, and Hamilton (eds.), Writing Medieval Biography, 129–41, at 139. 40 OV, iv, 108–09 (Nolite ergo confidere in principibus falsis O filii hominum . . . Nolite sperare in iniquitate, et rapinas nolite concupiscere. Diuitiae si affluent, nolite cor apponere); Matthew Kempshall, Rhetoric and the Writing of History, 400–1500 (Manchester and New York, 2011), 404. 41 Richard E. Barton, ‘Emotions and Power in Orderic Vitalis’, ANS, xxxiii (2011), 41–59, at 46–51.

528

william the conqueror

Of course, we ultimately do not know where William’s soul resides. However, both of the two authors who had reflected so much on his life and its consequences had no doubt that the outcome of his conquest of England was a human catastrophe. They wrestled with their own and others’ ambivalence about what William had done and what he motivated others to do. Too many innocent people had suffered. Yet William lived in times when biblical exegesis led popes, especially Gregory VII, to fill their writings with calls to violence, sometimes metaphorically and sometimes literally, in what they believed to be legitimate causes that locate them within a continuous long history.42 The new approach to William’s life in this book means that more must be done to locate him within this framework. Another history that needs to be written is one of medieval state-sponsored violence. Like the debate on English exceptionalism and its potential consequences for world history, this would also globalize a subject that is too often thought of as being of national interest at most and, often, as insignificant. But, in the end, the greatest priority must surely be to try to find a way to listen to the silent voices of the thousands whose lives William the Conqueror ruined. Only thus can the cause of humanity truly be served.

42 Buc, Holy War, Martyrdom, and Terror, is an essay on this theme, with full reference to the eleventh century in this context.

BI BL I O GR AP H Y

PRIMARY SOURCES ‘Acta Archiepiscopum Rotomagensium: Study and Edition’, ed. Richard Allen, Tabularia www. unicaen.fr/mrsh/craham/revue/tabularia, ix (2009), 1–66 Les actes de Guillaume le Conquérant et de la reine Mathilde pour les abbayes caennaises, ed. Lucien Musset, Mémoires de la Société des Antiquaires de Normandie, vol. xxxvii (Caen, 1967) Les actes des ducs de Bretagne (944–1148), ed. Hubert Guillotel et al. (Rennes, 2014) Actus Pontificum Cenomannis in urbe degentium, ed. Gustave Busson and Ambroise Ledru (Le Mans, 1902) Adam of Bremen, Gesta Hammaburgensis ecclesiae Pontificum, ed. Bernhard Schmeidler, MGH Scriptores (Hanover and Leipzig, 1917) (Adam of Bremen: History of the Archbishops of Hamburg-Bremen, trans. Francis J. Tschan, with a new Introduction and Notes by Timothy Reuter [New York, 2002]) Ademari Cabannensis Chronicon, ed. Pascale Bourgain, assisted by Richard Landes and Georges Pon, CCCM, vol. cxxix (Turnhout, 1999) (Adémar de Chabannes, Chronique, trans. Yves Chauvin and Georges Pon [Turnhout, 2003]) Aimé du Mont-Cassin, Ystoire de li Normant, ed. Michèle Guéret-Laferté (Paris, 2011) (The History of the Normans by Amatus of Montecassino, trans. Prescott N. Dunbar, revised with an Introduction and Notes by Graham A. Loud [Woodbridge, 2004]) Anglo-Saxon Charters: An Annotated List and Bibliography, ed. P. H. Sawyer (London, 1968) (updated at http://www.esawyer.org.uk) Anglo-Saxon Chronicle: A Collaborative Edition, volume 3, ms. A, ed. Janet M. Bateley (Cambridge, 1986) Anglo-Saxon Chronicle: A Collaborative Edition, volume 6, ms. D: A Semi-Diplomatic Edition with Introduction and Indices, ed. G. P. Cubbin (Cambridge, 1996) Anglo-Saxon Chronicle: A Collaborative Edition, volume 7, ms. E: A Semi-Diplomatic Edition with Introduction and Indices, ed. Susan Irvine (Cambridge, 2004) Anglo-Saxon Chronicles, trans. and ed. Michael Swanton (London, 2000) Annales Altahenses Maiores, ed. Wilhelm de Giesebrecht and Edmund L. B. ab Oefele, MGH Scriptores (Hanover, 1868), 772–824 Les annales de Flodoard, ed. P. Lauer (Paris, 1906) (The Annals of Flodoard of Reims, 919–966, ed. and trans. Steven Fanning and Bernard S. Bachrach [Peterborough, Ontario, 2004]) Les annales de l’abbaye de Saint-Pierre de Jumièges: chronique universelle du XIIIe siècle, ed. Dom. J. Laporte (Rouen, 1954) ‘Annales du Mont-Saint-Michel’, in Chronique de Robert de Torigni, ed. Delisle, ii, 207–36 Annales monasterii de Wintonia, in Annales Monastici, 5 vols., RS (London, 1864–9), ii, 1–125 Annales ordinis sancti Benedicti, ed. Jean Mabillon, 6 vols. (Lucca, 1703–39) Annales qui dicuntur Rainaldi archidiaconi sancti Mauricii Andegavensis, in Recueil d’annales angevines et vendômoises, ed. Louis Halphen (Paris, 1903), 80–90 Annales Regni Francorum unde ab. a. 741 usque ad a. 829, qui dicuntur Annales laurissenses maiores et Einhardi, ed. Friedrich Kurze, MGH Scriptores (Hanover, 1895) Annales sancti Albinii Andegavensis, in Recueil d’annales angevines et vendômoises, ed. Louis Halphen (Paris, 1903), 1–49

529

530

bibliography

Annales sancti Sergii Andegavensis, in Recueil d’annales angevines et vendômoises, ed. Louis Halphen (Paris, 1903), 91–9 ‘Annales Uticenses’, in Orderici Vitalis Ecclesiasticae Historiae Libri Tredecim, ed. Auguste Le Prévost, 5 vols. (Paris, 1838–55), v, 139–73 ‘Annalis Historia Brevis sive Chronica monasterii S. Stephani Cadomensis’, in Scriptores rerum gestarum Willelmi Conquestoris, ed. J. A. Giles (London, 1845), 161–74 Annals of Ulster to A. D. 1131, ed. Seán Mac Airt and Gearóid Mac Niocaill (Dublin, 1983) Anselme de Saint-Rémy, Histoire de la dédicace de Saint-Rémy, ed. and trans. J. Hourlier, in La Champagne bénédictine: contribution à l’année saint Benoît, Travaux de l’Académie nationale de Reims, vol. clx (Rheims, 1981), 181–297 Antiquus Cartularius Ecclesie Baiocensis, ed. V. Bourrienne, 2 vols. (Rouen and Paris, 1902–3) Les Archives de Normandie et de la Seine-Inférieure. Etat général des fonds. Recueil des facsimilés d’écriture du XIes. au XVIIIes. accompagné de transcriptions, ed. P. Chevreux and J.-J. Vernier (Rouen, 1911) Baldricus Burgulensis, Carmina, ed. Karlheinz Hilbert (Heidelberg, 1979) (‘Baudri of Bourgueil, “To Countess Adela” ’, trans. Monika Otter, Journal of Medieval Latin, xi [2001], 60–141) Beringerius Turonensis Rescriptum contra Lanfrannum, ed. R. B. C. Huygens, 2 vols., CCCM, vol. lxxxiv (Turnhout, 1988) ‘Breve Chronicon Lyrensis’, in Thesaurus Novus Anecdotorum, ed. E. Martène and U. Durand, 5 vols. (Paris, 1717), iii, cols. 1,432–4 ‘The Breuis Relatio de Guillelmo nobilissimo comite Normannorum, written by a Monk of Battle Abbey’, ed. and trans. Elisabeth van Houts, in idem, History and Family Traditions in England and the Continent, 1000–1200 (Aldershot and Burlington, VT, 1999), chapter 7. Reprinted, with a translation added, from Camden Miscellany, 5th ser., x (Cambridge, 1997) Briefsammlungen der Zeit Heinrichs IV, ed. Carl Erdmann and Norbert Fickermann (Weimar, 1950) Bruno, ‘De Bello Saxonico’, ed. Franz-Josef Schmale, in Quellen zur Geschichte Kaiser Heinrich IV (Darmstadt, 1963), 192–405 Brut y Tywysogyon or the Chronicle of the Princes, Peniarth MS 20 Version, trans. Thomas Jones (Cardiff, 1952) Brut y Tywysogyon or the Chronicle of the Princes, Red Book of Hergest Version, trans. Thomas Jones (Cardiff, 1955) The Carmen de Hastingae Proelio of Guy Bishop of Amiens, ed. Frank Barlow, OMT (Oxford, 1999) The Cartae Antiquae Rolls 11–20, ed. J. C. Davies, Pipe Roll Society, lxxi (new series, xxxiii) (London, 1960, for 1957) Cartulaire blésois de Marmoutier, ed. C. Métais, 2 vols. (Blois, 1887–91) Le Cartulaire de l’abbaye bénédictine de Saint-Pierre-de-Préaux (1034–1227), ed. Dominique Rouet (Paris, 2005) Cartulaire de l’abbaye de Redon en Bretagne, ed. M. Aurélian de Courson (Paris, 1863) Cartulaire de l’abbaye de Saint-Vincent du Mans, ed. R. Charles and le vicomte Menjot d’Elbenne (Mamers and Le Mans, 1886–1913) Cartulaire de la Sainte-Trinité-du-Mont de Rouen, ed. A. Deville, in Cartulaire de Saint-Bertin, ed. B. E. C. Guérard (Paris, 1840) Cartulaire de Notre-Dame de Chartres, ed. E. de Lepinois and Lucien Merlet, 3 vols. (Chartres, 1862–5) Cartulaire de Saint-Aubin d’Angers, ed. A. Bertrand de Broussillon (Paris, 1903) Cartulaire des abbayes de Saint-Pierre de La Couture et de Saint-Pierre de Solesmes, ed. Les Bénédictins de Solesmes (Le Mans, 1881) Le cartulaire du chapitre cathédral de Coutances, ed. Julie Fontanel (Saint-Lô, 2003) Cartulaire manceau de Marmoutier, ed. Ernest Laurain, 2 vols. (Laval, 1911–45)

bibliography

531

Cartularium monasterii de Rameseia, ed. W. H. Hart and P. A. Lyons, RS, 3 vols. (London, 1884–93) Cartulary of Flavigny 717–1113, ed. Constance Brittain Bouchard (Cambridge, MA, 1991) The Cartulary of the Abbey of Mont-Saint-Michel, ed. Katharine S. B. Keats-Rohan (Donnington, 2006) Catalogue des actes d’Henri Ier, roi de France, ed. F. Soehnée (Paris, 1907) Catalogus abbatum Fiscannensium, in Novae Bibliothecae Manuscriptorum Librorum, ed. P. Labbé, 2 vols. (Paris, 1657), i, 328–9 Charters and Custumals of the Abbey of Holy Trinity Caen, Part 2: The French Estates, ed. John Walmsley, Records of Social and Economic History, new series, no. 22 (Oxford, 1994) Chartes et documents de l’abbaye de Saint-Pierre au Mont-Blandin à Gand, ed. Auguste van Lokeren (Gand, 1868) Chronica de Gestis Consulum Andegauorum, in Chroniques des comtes d’Anjou et des seigneurs d’Amboise, ed. Halphen and Poupardin, 25–73 Chronica Rogeri de Houedene, ed. William Stubbs, 4 vols., RS (London, 1868–71) The Chronicle of Battle Abbey, ed. and trans. Eleanor Searle, OMT (Oxford, 1980) The Chronicle of Hugh Candidus, a Monk of Peterborough, ed. W. T. Mellows (London, 1949) Chronicle of John of Worcester, ed. R. R. Darlington and Patrick McGurk, vols. ii-iii, OMT (Oxford, 1995–8) Chronicon abbatum Majoris Monasterii, in Recueil des chroniques de Touraine, ed. André Salmon (Tours, 1854), 318–37 Chronicon Beccensis abbatiae ab ipsa fundatione ad annum 1467, in Lanfranci Opera Omnia ed. L. d’Achery (Paris, 1648) Chronicon Briocensi, RHF, xii, 565–7 Chronicon Britannicum, RHF, xi, 412–13 Chronicon pontificum ecclesiae Eboracensis, in Historians of the Church of York and its Archbishops, ed. James Raine, 3 vols., RS (London, 1879–94) Chronicon Sancti Huberti Andaginensis, ed. Ludwig C. Bethmann and Wilhelm Wattenbach, MGH Scriptores, viii (Hanover, 1848), 565–630 ‘Chronicum Rotomagense’, in Novae Bibliothecae Manuscript. Librorum, ed. P. Labbé, 2 vols. (Paris, 1657), i, 364–90 Chronique de Robert de Torigni, abbé du Mont-Saint-Michel, suivie de divers opuscules historiques de cet auteur et de plusieurs religieux de la même abbaye, ed. Léopold Delisle, 2 vols., Société de l’Histoire de Normandie (Rouen, 1872–3) La Chronique de Saint-Hubert dite Cantatorium, ed. K. Hanquet (Brussels, 1906) La Chronique de Saint-Maixent, 751–1140, ed. and trans. Jean Verdon (Paris, 1979) Chroniques des comtes d’Anjou et des seigneurs d’Amboise, ed. Louis Halphen and René Poupardin (Paris, 1913) Chroniques latines du Mont Saint-Michel (IXe–XIIe siècle), ed. Pierre Bouet and Olivier Desbordes (Caen and Avranches, 2009) City of God against the Pagans, vol. II, Books 4–7, trans. William M. Green (London, 1963) Concilia Rotomagensis Provinciae, ed. G. Bessin (Rouen, 1717) Corpus des inscriptions de la France médiévale: 22. Calvados, Eure, Manche, Orne, Seine-Maritime, ed. Robert Favreau and Jean Michaud (Paris, 2002) Das Register Gregors. VII., ed. Erich Caspar, 2 vols., MGH Epistolae Selectae (Berlin, 1920–3) (The Register of Pope Gregory VII: An English Translation, ed. H. E. J. Cowdrey [Oxford, 2002]) Decretales Pseudo-Isidoriannae et Capitula Angilranni, ed. Paul Hinschius (Leipzig, 1863) De detectione corporum sanctorum Dionysii, Rustici et Eleutherii, RHF, xi, 467–74 De Gestis Herwardi, ed. Alberto Meneghetti (Pisa, 2013) De libertate Beccensis monasterii (On the Liberty of the Monastery of Bec), in Three Treatises from Bec on the Nature of the Monastic Life, ed. Giles Constable and trans. Bernard S. Smith (Toronto, 2008), 135–67 Die Gesetze der Angelsachsen, ed. Felix Liebermann, 3 vols. (Halle, 1903–16)

532

bibliography

Die Kreuzzugsbriefe aus dem Jahren 1088–1100, ed. Heinrich Hagenmeyer (Innsbruck, 1901) Domesday Book, seu Liber Censualis Willelmi Primi, ed. Abraham Farley, 2 vols. (London 1783); Great Domesday Book: Library Edition, ed. Ann Williams and R. W. H. Erskine, Alecto Historical Editions (London 1986–92) Domesday Monachorum of Christ Church, Canterbury, ed. D. C. Douglas (London, 1944) ‘The Dorset Geld Rolls’, ed. and trans. Ann Williams, VCH, Dorset, iii, 115–49 Dudo of Saint-Quentin, De Gestis Normannie Ducum seu de moribus et actis primorum Normanniae Ducum, ed. Jules Lair, Mémoires de la Société des Antiquaires de Normandie, vol. xxiii (Caen, 1865) (translated, Dudo of Saint-Quentin, History of the Normans, trans. Eric Christiansen [Woodbridge, 1998]) Durand of Troarn, Liber de corpore et sanguine Christi, in PL, cxlix, cols. 1,375–422 Eadmer, Historia novorum in Anglia, ed. Martin Rule, RS 81 (London, 1884) The Early Charters of the Augustinian Canons of Waltham Abbey, Essex, 1062–1230, ed. Rosalind Ransford (Woodbridge, 1989) Early History of Glastonbury: An Edition, Translation and Study of William of Malmesbury’s ‘De Antiquitate Glastonie Ecclesie’, ed. and trans. John Scott (Woodbridge, 1981) Early Yorkshire Charters, 12 vols., ed. W. Farrer and C. T. Clay, Yorkshire Archaeological Society Record Series (Wakefield, 1935–65) Ecclesiastical History of Orderic Vitalis, ed. and trans. Marjorie Chibnall, 6 vols., OMT (Oxford 1969–80) Encomium Emmae Reginae, ed. Alistair Campbell, with a supplementary introduction by Simon Keynes, Camden Classic Reprints (Cambridge, 1998) English Historical Documents 1042–1189, ed. David C. Douglas and George W. Greenaway, 2nd edn. (London and New York, 1981) English Lawsuits from William I to Richard I, ed. R. C. van Caenegem, 2 vols., Selden Society, vols. cvi–cvii (London, 1990–1) The Epistolae Vagantes of Pope Gregory VII, ed. H. E. J. Cowdrey, OMT (Oxford, 1972) Etymologies of Isidore of Seville, trans. Stephen A. Barney et al. (Cambridge, 2006) ‘Ex Chronico Turonensi auctore Turon. Ecclesiae S. Martini Canonico’, RHF, xi, 346–9 Facsimiles of Anglo-Saxon Charters, ed. Simon Keynes (Oxford, 1991) Facsimiles of Early Charters in Oxford Muniment Rooms, ed. H. E. Salter (Oxford, 1929) Facsimiles of English Royal Writs to A.D. 1100 presented to V. H. Galbraith, ed. T. A. M. Bishop and P. Chaplais (Oxford, 1957) Feudal Documents from the Abbey of Bury St Edmunds, ed. D. C. Douglas, British Academy Records of the Social and Economic History of England and Wales, viii (London, 1932) Frutolf of Michelsberg, Chronicon, ed. Franz-Josef Schmale and Irene Schmale-Ott, in Frutolfs und Ekkehardi Chroniken und die Anonyme Kaiserchronik (Darmstadt, 1972) ‘Fulcoii Belvacensis epistolae’, ed. M. Colker, Traditio, x (1954), 191–273 Gallia Christiana in Provincias Ecclesiasticas distributas, ed. D. Samarthanus et al., 16 vols. (Paris, 1715–1865) Gesta Ambaziensium dominorum, in Chroniques des comtes d’Anjou et des seigneurs d’Amboise, ed. Halphen and Poupardin, 74–132 Gesta Guillelmi of William of Poitiers, ed. and trans. R. H. C. Davis and Marjorie Chibnall, OMT (Oxford, 1998) Gesta Herwardi incliti exulis et militis, ed. T. D. Hardy and C. T. Martin, in Lestoire des Engles solum la Translacion Maistre Geffrei Gaimar, 2 vols., RS (London, 1889) (trans. Michael Swanton, Three Lives of the Last Englishmen [New York, 1984]) Gesta Normannorum Ducum of William of Jumièges, Orderic Vitalis, and Robert of Torigni, ed. and trans. Elisabeth M. C. van Houts, 2 vols., OMT (Oxford 1992–5) Gesta Stephani, ed. and trans. K. R. Potter and R. H. C. Davis, OMT (Oxford, 1976) Gesta Swenomagni regis et filiorum eius et Passio gloriosissimi Canuti regis et martyris, in Vitae Sanctorum Danorum, ed. M. C. Gertz (Copenhagen, 1908–12), 77–136 Gilbert of Mons, La Chronique, ed. L. Vanderkindere (Brussels, 1904) (translated, Gilbert of Mons, Chronicle of Hainault, trans. Laura Napran [Woodbridge, 2005])

bibliography

533

Gilo, Vita sancti Hugonis abbatis, ed. H. E. J. Cowdrey, in ‘Memorials of Abbot Hugh of Cluny’, 43–109 The Great Chartulary of Glastonbury, ed. Dom. Aelred Watkin, 3 vols. Somerset Record Society, vols. lix, lxiii, lxiv (1947–56) Guibert de Nogent, Autobiographie, ed. E.-R. Labande (Paris, 1981) (A Monk’s Confession: the Memoirs of Guibert of Nogent, trans. Paul J. Archambault [University Park, PA, 1995]) Guibertus abbas S. Mariae Nogenti, Gesta Dei per Francos, ed. R. B. C. Huygens, CCCM, no. 127A (Turnhout, 1996), 290–1 (The Deeds of God through the Franks: A Translation of Guibert of Nogent’s ‘Gesta Dei per Francos’, trans. Robert T. Levine [Woodbridge, 1997]) Guillaume de Poitiers, Histoire de Guillaume le Conquérant, ed. Raymonde Foreville (Paris, 1952) Hariulf, Chronique de l’abbaye de Saint-Riquier, ed. F. Lot (Paris, 1894) Henry of Huntingdon, Historia Anglorum: The History of the English People, ed. and trans. Diana Greenway, OMT (Oxford, 1996) Herman the Archdeacon and Goscelin of Saint-Bertin, Miracles of St Edmund, ed. and trans. Tom Licence, with the assistance of Linda Lockyer, OMT (Oxford, 2014) Histoire des conciles d’après les documents originaux, ed. C. J. Hefele and Dom. J. Leclercq (Paris, 1911) Historia Compostellana, ed. Emma Falque Rey, CCCM, vol. lxx (Turnhout, 1988) Historia Ecclesie Abbendonensis: The History of the Church of Abingdon, 2 vols., ed. and trans. John Hudson, OMT (Oxford, 2002–7) Historia Regum Anglorum, in Symeonis Monachi Opera Omnia, ed. T. Arnold, 2 vols., RS (London, 1882–5) Historia Selebiensis Monasterii: The History of the Monastery of Selby, ed. and trans. Janet Burton with Lynda Lockyer, OMT (Oxford, 2013) History of the Norman People: Wace’s Roman de Rou, trans. Glyn S. Burgess (Woodbridge, 2004) Hugh of Flavigny, Chronicon, ed. Georg Pertz, in MGH Scriptores, viii (Hanover, 1848), 280–503 Hugh of Fleury, Liber qui modernum regum Francorum continent actus, ed. Georg Waitz, MGH Scriptores, xi (Hanover, 1851), 376–95 Hugh the Chanter, The History of the Church of York, 1066–1127, ed. and trans. Charles Johnson, revised by M. Brett, C. N. L. Brooke, and M. Winterbottom, OMT (Oxford, 1990) Inventio et Miracula sancti Uulfranni, ed. Dom. J. Laporte, Société de l’Histoire de Normandie, Mélanges, xiv (Rouen, 1938) Lambert of Saint-Omer, Liber Floridus, ed. Albert Derolez (Gent, 1968) Lampert of Hersfeld, Annales, ed. Adolf Schmidt and Wolfgang Dietrich Fritz (Berlin, 1962) (The Annals of Lampert of Hersfeld, trans. I. S. Robinson [Manchester, 2015]) Leges Edwardi Confessoris, ed. Bruce R. O’Brien, in idem, God’s Peace and King’s Peace: the Laws of Edward the Confessor (Philadelphia, PA, 1999), 135–203 Letters of Lanfranc, Archbishop of Canterbury, ed. and trans. Helen Clover and Margaret Gibson, OMT (Oxford, 1979) The Letters and Charters of Gilbert Foliot, ed. Z. N. Brooke, Dom. Adrian Morey, and C. N. L. Brooke (Cambridge, 1967) Liber Eliensis, ed. E. O. Blake, Camden 3rd ser., xcii (1962) (translated Liber Eliensis: A History of the Isle of Ely from the Seventh Century to the Twelfth, trans. Janet Fairweather [Woodbridge, 2005]) Liber Vitae of the New Minster and Hyde Abbey, Winchester, ed. S. D. Keynes, Early English Manuscripts in Facsimile, vol. xxxi (Copenhagen, 1996) The Life of King Edward who Rests at Westminter, Attributed to a Monk of Saint-Bertin, ed. and trans. Frank Barlow, 2nd edn., OMT (Oxford 1992) The Life of St Anselm, Archbishop of Canterbury, by Eadmer, ed. R. W. Southern, OMT (Oxford, 1962) ‘Memorials of Abbot Hugh of Cluny (1049–1109)’, ed. H. E. J. Cowdrey, Studi Gregoriani, xi (1978), 11–175 Miracula Sanctae Virginis Mariae, ed. Elise F. Dexter (Madison, WI, 1927) ‘Miracula sancti Audoeni’, Acta Sanctorum, August, iv, 825–37

534

bibliography

Monastic Constitutions of Lanfranc, ed. and trans. Dom. David Knowles and Christopher N. L. Brooke, OMT (Oxford, 2002) Monasticon Anglicanum, ed. William Dugdale; new edn. by J. Caley, H. Ellis, and B. Bandinel, 6 vols. (London, 1817–30) Nécrologe-obituaire de la cathédrale du Mans, ed. Gustave Busson and Ambroise Ledru (Le Mans, 1906) Neustria Pia, ed. A. du Monstier (Rouen, 1663) The Normans in Europe, ed. Elisabeth van Houts (Manchester, 2000) Obituaires de la province de Sens: 2 (diocèse de Chartres), ed. A. Molinier (Paris, 1906) Ordinaire et Coutumier de l’église cathédrale de Bayeux, ed. chanoine Ulysse Chevalier (Paris, 1902) The Ordinal of the Abbey of the Holy Trinity Fécamp (Fécamp, Musée de la Bénédictine, Ms 186), ed. David Chadd, 2 vols., Henry Bradshaw Society (Woodbridge, 1999–2002) Patrologia cursus completus, series Latina, ed. J.-P. Migne, 221 vols., Paris 1844–65 The Peterborough Chronicle (The Bodleian Manuscript Laud Misc. 636), ed. Dorothy Whitelock, Early English Manuscripts in Facsimile, iv (Copenhagen, 1954) The Political Writings of Archbishop Wulfstan of York, ed. and trans. Andrew Rabin (Manchester, 2015) Das Polyptychon von Saint-Germain-des-Prés, ed. Dieter Hägermann (Cologne, 1993) ‘Quo B. Maria subvenit Guillelmo Crispino seniori’, PL, cl, cols. 735–44 Raoul Glaber: Histoires, ed. Mathieu Arnoux (Turnhout, 1996) Recueil des actes de Philippe I er, roi de France (1059–1108), ed. M. Prou, (Paris, 1908) Recueil des actes des ducs de Normandie de 911 à 1066, ed. Marie Fauroux, Mémoires de la Société des Antiquaires de Normandie, vol. xxxvi (Caen 1961) Recueil des chartes de l’abbaye de Cluny, ed. Auguste Bernard and Alexandre Bruel, 6 vols. (Paris, 1876–1903) Recueil des chartes de l’abbaye de Saint-Germain-des-Prés, ed. R. Poupardin (Paris, 1909) Recueil des Historiens des Gaules et de la France, ed. Martin Bouquet (Paris, 1738–1876) Recueil des rouleaux des morts (VIIIe-vers 1536), ed. Jean Dufour, 4 vols. (Paris, 2005–08) Regesta Regum Anglo-Normannorum: vol. i, Regesta Willelmi Conquestoris et Willelmi Rufi 1066–1100, ed. H. W. C. Davis, with the assistance of R. J. Whitwell (Oxford, 1913) Regesta Regum Anglo-Normmanorum: vol. ii, Regesta Regis Henrici Primi, 1100–1135, ed. Charles Johnson and H. A. Cronne (Oxford, 1956) Regesta Regum Anglo-Normannorum: vol. iii, Regesta Regis Stephani ac Mathildis Imperatricis ac Gaufridi et Henrici Ducum Normannorum, 1135–1154, ed. H. A. Cronne and R. H. C. Davis (Oxford, 1968) Regesta Regum Anglo-Normannorum: The Acta of William I (1066–1087), ed. David Bates (Oxford, 1998) Regularis Concordia Anglicae Nationis Monachorum Sanctimonialiumque: The Monastic Agreement of the Monks and Nuns of the English Nation, ed. and trans. Dom. Thomas Symons (London, 1953) Richard fitz Nigel, Dialogus de Scaccario: The Dialogue of the Exchequer, and Constitutio Domus Regis, The Disposition of the King’s Household, ed. and trans. Emilie Amt and S. D. Church, OMT (Oxford, 2007) Robert of Torigni, ‘De Immutatione Ordinis Monachorum’, in Chronique de Robert de Torigni, abbé du Mont-Saint-Michel, ed. Delisle, ii, 181–206 Rodulfus Glaber Opera, ed. John France, Neithard Bulst, and Paul Reynolds, OMT (Oxford, 1989) Sacrorum conciliorum nova et amplissima collectio, ed. J. D. Mansi (Venice, 1774) ‘Sanctae Catharinae virginis et martyris Translatio et Miracula Rothomagensia’, ed. A. Poncelet, Analecta Bollandiana, xxii (1903), 423–38 Sancti Anselmi Opera Omnia, ed. F. S. Schmitt, 6 vols. (Edinburgh, 1946–61) (The Letters of Saint Anselm of Canterbury, trans. Walter Fröhlich, 3 vols., Cistercian Studies [Kalamazoo, 1990–4]) ‘Serlo of Bayeux, The Capture of Bayeux’, ed. and trans. Moreed Arbabzadah, in Elisabeth van Houts, ‘The Fate of Priests’ Sons in Normandy with Special Reference to Serlo of Bayeux’, HSJ, xxv (2014, for 2013), 57–105, at 86–105 (Appendix 3)

bibliography

535

‘Sulcard of Westminster: Prologus de constructione Westmonasterii’, ed. B. W. Scholz, Traditio, xx (1964), 59–91 Symeon of Durham, Libellus de Exordio atque Procursu istius hoc est Dunhelmensis Ecclesie, ed. and trans. David Rollason, OMT (Oxford, 2000) Tacitus, Agricola, ed. A. J. Woodman, with contributions from C. S. Kraus (Cambridge, 2014) The Text of the Book of Llan Dâv, ed. J. Gwenogvryn Evans (Oxford, 1893) Thomas of Marlborough, History of the Abbey of Evesham, ed. Jane Sayers and Leslie Watkiss, OMT (Oxford, 2003) ‘Translatio B. Nicasii martyris’, PL, clxii, cols. 1,163–6 Vita Annonis archiepiscopi Coloniensis, ed. G. H. Pertz, MGH Scriptores, xi (Hanover, 1854), 462–514 Vita beati Simonis comitis Crespeiensis, in PL, clvi, cols. 1,211–24 Vita Griffini filii Conani: The Medieval Life of Gruffudd ap Cynan, ed. Paul Russell (Cardiff, 2005) ‘Vita Herluini’, in The Works of Gilbert Crispin Abbot of Westminster, ed. Anna Sapir Abulafia and G. R. Evans, Auctores Britannici Medii Aevi, viii (Oxford, 1986) (J. Armitage Robinson, Gilbert Crispin, Abbot of Westminster: A Study of the Abbey under Norman Rule [Cambridge, 1911]) Vita Lanfranci, ed. Margaret Gibson, in Lanfranco di Pavia e l’Europa del secolo XI nel IX centenario della morte, ed. G. D’Onofrio (Rome, 1993), 661–715 Vita sancti Gundlei, in Vitae sanctorum Britanniae et Genealogiae, ed. A. W. Wade-Evans (Cardiff, 1944), 172–93 Vita sancti Hugonis abbatis Cluniacensis, PL, clix, cols. 918–27 Vita sancti Wistani, in Chronicon Abbatiae de Evesham, ed. W. D. Macray, RS, no. 29 (London, 1863), 325–37 Vita Wlstani, in William of Malmesbury: Saints’ Lives, ed. and trans. Michael Winterbottom and Rodney M. Thomson, OMT (Oxford, 2002) Waltham Chronicle: An Account of the Discovery of our Holy Cross at Montacute and its Conveyance to Waltham, ed. Leslie Watkiss and Marjorie Chibnall, OMT (Oxford, 1994) The Warenne (Hyde) Chronicle, ed. Elisabeth van Houts and Rosalind C. Love, OMT (Oxford, 2012) Wenric of Trier, Epistola sub Theoderici episcope Virdunensis nominee composite, ed. Kuno Francke, MGH Libelli de Lite (Hanover, 1891), i, 280–99 Willelmi Meldunensis monachi liber super explanationum Lamentationum Ieremiae prophetae, ed. Michael Winterbottom and R. M. Thomson, CCCM (Turnhout, 2011) William of Malmesbury, Gesta pontificum Anglorum: The History of the English Bishops, ed. and trans. Michael Winterbottom, and R. M. Thomson, OMT (Oxford, 2007) William of Malmesbury, Gesta regum Anglorum: The History of the English Kings, ed. and trans. R. A. B. Mynors, M. Winterbottom, and R. M. Thomson, 2 vols., OMT (Oxford 1998–9) The Winchcombe and Coventry Chronicles: Hitherto Unnoticed Witnesses to the Work of John of Worcester, ed. and trans. Paul Antony Hayward, 2 vols., Arizona Center for Medieval and Renaissance Studies (Tempe, AZ, 2010) Wiponis Gesta Chuonradi II Imperatoris, ed. Harry Bresslau, MGH Scriptores (Hanover and Leipzig, 1915; reprinted, Hanover, 1977) (translated, Imperial Lives and Letters of the Eleventh Century, trans. T. E. Mommsen and K. F. Morrison [New York and London, 1962])

SECONDARY SOURCES In the case of essays published in edited collections, when the essay is the only one referred to from the volume, I have provided a single reference that includes both the article and the collection. Where there is more than one article from a collection, I have given a full reference to the collection under the name of the editors and with an abbreviated reference accompanying each article refered to.

536

bibliography

Abels, Richard P., Lordship and Military Obligation in Anglo-Saxon England (London, 1988) Abrams, Lesley, Anglo-Saxon Glastonbury: Church and Endowment (Woodbridge, 1996) —— ‘Diaspora and Identity in the Viking Age’, EME, xx (2012), 17–38 —— ‘Early Normandy’, ANS, xxxv (2013), 45–64 Aird, William M., ‘An Absent Friend: The Career of William of St Calais’, in David Rollason, Margaret Harvey, and Michael Prestwich (eds.), Anglo-Norman Durham, 1093– 1193 (Woodbridge, 1994), 283–97 —— St Cuthbert and the Normans: The Church of Durham, 1071–1153 (Woodbridge, 1998) —— ‘Copsi [Coxo], Earl of Northumbria (d.1067)’, ODNB —— ‘Gospatric, Earl of Northumbria (d.1073x5)’, ODNB —— ‘Mowbray, Robert de, Earl of Northumbria (d.1115/1125)’, ODNB —— Robert Curthose, Duke of Normandy, c.1050–1134 (Woodbridge, 2008) —— ‘Northumbria and the Making of the Kingdom of the English’, in Hirokazu Tsurushima (ed.), Nations in Medieval Britain (Donnington, 2010), 45–60 Airlie, Stuart, ‘Private Bodies and the Body Politic in the Divorce Case of Lothar II’, Past and Present 161 (1998), 3–38 (reprinted in Stuart Airlie, Power and its Problems in Carolingian Europe [Farnham and Burlington, VT, 2012], chapter 10) Alexander, Alison, ‘Riots, Reform, and Rivalry: Religious Life in Rouen, c.1073–c.1092’, ANS, xxxiii (2011), 23–40 Allen, Martin, ‘Mints and Moneyers of England and Wales, 1066–1158’, British Numismatic Journal, lxxxii (2012), 54–120 —— Mints and Money in Medieval England (Cambridge, 2012) —— ‘Coinage and Currency under William I and William II’, in Rory Naismith, Martin Allen, and Elina Screen (eds.), Early Medieval Monetary History: Studies in Memory of Mark Blackburn (Farnham, 2014), 85–112 —— Medieval European Coinage. 9(a). The British Isles, 1066–1279 (Cambridge, forthcoming) Allen, Richard, ‘ “A Proud and Headstrong Man”: John of Ivry, Bishop of Avranches and Archbishop of Rouen, 1060–79’, HR, lxxxii (2009), 1–39 —— ‘Praesul praecipue et venerande: The Career of Robert, Archbishop of Rouen, 989–1037’, in Hicks and Brenner (eds.), Society and Culture in Medieval Rouen, 911–1300, 153–83 —— ‘The Earliest Known List of Excommunicates from Ducal Normandy’, JMH, xxxix (2013), 394–415 —— ‘Avant Lanfranc. Un réexamen de la carrière de Mauger, archevêque de Rouen (1037–1054/55)’, in Barrow, Delivré, and Gazeau (eds.), Autour de Lanfranc (1010–2010), 131–52 Althoff, Gerd, Spielregeln der Politik im Mittelalter: Kommunikation in Frieden und Fehde (Darmstadt, 1997) —— Otto III, trans. Phyllis G. Jestice (University Park, PA, 2003) —— Heinrich IV (Darmstadt, 2006) Archibald, Marion, ‘The German Connection: German Influences on the Later AngloSaxon and Norman Coinages in their English Context (10th and 11th centuries)’, in Rainer Polley and Andrea Röpcke (eds.), Fundamenta Historiae. Geschichte im Spiegel der Numismatik und ihrer Nachbarwissenschaften. Festschrift für Niklot Klüßendorf zum 60. Geburtstag am 10. Februar 2004 (Hanover, 2004), 131–50 Arnold, Nick, ‘The Defeat of the Sons of Harold in 1069’, Report of the Transactions of the Devonshire Association for the Advancement of Science, cxlvi (2014), 33–56 Arnoux, Mathieu, ‘Un observateur indulgent: Raoul le Glabre, les ducs de Normandie, leurs concubines et leurs enfants’, in François de Beaurepaire and Jean-Pierre Chaline (eds.), La Normandie vers l’An Mil, Société de l’Histoire de Normandie (Rouen, 2000), 88–92 Ayers, Brian, ‘The Urban Landscape’, in Rawcliffe and Wilson (eds.), Medieval Norwich, 1–28 Ayton, Andrew, and Price, J. L. (eds.), The Medieval Military Revolution (London, 1995) Babcock, Robert S., ‘Rhys ap Tewdwr, King of Deheubarth’, ANS, xvi (1994), 21–35

bibliography

537

Bachrach, Bernard S., ‘Some Observations on the Military Administration of the Norman Conquest’, ANS, viii (1986), 1–25 (reprinted in Bernard S. Bachrach, Warfare and Military Organization in Pre-Crusade Europe [Aldershot and Burlington, VT, 2002], chapter 14) —— ‘The Pilgrimages of Fulk Nerra, Count of the Angevins, 987–1040’, in Thomas F. X. Noble and John J. Contreni (eds.), Religion, Culture and Society in the Early Middle Ages: Studies in Honor of Richard E. Sullivan (Kalamazoo, 1987), 205–17 —— Fulk Nerra, the Neo-Roman Consul, 987–1040 (Berkeley and Los Angeles, CA, and London, 1993) —— ‘The Norman Conquest, Countess Adela, and Abbot Baudri’, ANS, xxxv (2013), 65–78 Baker, Nigel, and Richard Holt, Urban Growth and the Medieval Church: Gloucester and Worcester (Aldershot, 2004) Banham, Debby, and Rosamond Faith, Anglo-Saxon Farms and Farming (Oxford, 2014) Baring, F. H., Domesday Tables for the Counties of Surrey, Berkshire, Middlesex, Hertford, Buckingham and Bedford (London, 1909) Barlow, Frank, ‘Guenta’, appendix to Martin Biddle, ‘Excavations at Winchester 1962–3’, Antiquaries Journal, xliv (1964), 188–219 —— William I and the Norman Conquest (London, 1965) —— Edward the Confessor (London, 1970); 2nd edn. (New Haven, CT, and London, 1997) —— The English Church, 1000–1066, 2nd edn. (London, 1979) —— The English Church, 1066–1154 (London, 1979) —— ‘William I’s Relations with Cluny’, JEH, xxxii (1981), 131–41 (reprinted in Barlow, The Norman Conquest and Beyond, 245–56) —— The Norman Conquest and Beyond (London, 1983) —— ‘Foreword’, D. C. Douglas, William the Conqueror, 2nd edn. (New Haven, CT, and London, 1999) —— William Rufus, 2nd edn. (New Haven, CT, and London, 2000) —— The Godwins: The Rise and Fall of a Noble Dynasty (London, 2002) —— ‘St Calais [St Carilef], William of (c.1030–1096)’, ODNB Barrow, G. W. S., ‘Malcolm III [Mael Coluim Ceann Mór, Malcolm Canmore] (d.1093)’, ODNB Barrow, Julia S., and N. P. Brooks (eds.), St Wulfstan and his World (Aldershot and Burlington, VT, 2005) —— , Fabrice Delivré, and Véronique Gazeau (eds.), Autour de Lanfranc (1010–2010): Réforme et réformateurs dans l’Europe du Nord-Ouest (XIe–XIIe siècles) (Caen, 2015) Barthélemy, Dominique, L’an mil et la paix de Dieu. La France chrétienne et féodale, 980–1060 (Paris, 1999) —— , François Bougard, and Régine Le Jan (eds.), La vengeance, 400–1200 (Rome, 2006) Bartlett, Robert, The Making of Europe: Conquest, Colonization and Cultural Change 950–1350 (London, 1993) —— ‘ “Mortal Enmities”: The Legal Aspect of Hostility in the Middle Ages’, in Tuten and Billado (eds.), Feud, Violence and Practice, 197–212 Barton, Richard E., Lordship in the County of Maine, c.890–1160 (Woodbridge, 2004) —— ‘Emotions and Power in Orderic Vitalis’, ANS, xxxiii (2011), 41–59 Bates, David, ‘The Character and Career of Odo, Bishop of Bayeux (1049/50–1097)’, Speculum, l (1975), 1–20 —— ‘The Land Pleas of William I’s Reign: Penenden Heath revisited’, BIHR, li (1978), 1–19 —— Normandy before 1066 (London and New York, 1982) —— ‘The Origins of the Justiciarship’, ANS, iv (1982), 1–12, 167–71 —— ‘A Neglected English Charter of Robert Curthose, Duke of Normandy’, BIHR, lix (1986), 121–4

538

bibliography

—— ‘Lord Sudeley’s Ancestors: The Family of the Counts of Amiens, Valois and the Vexin in France and England during the 11th Century’, in The Sudeleys – Lords of Toddington (London, 1987), 34–48 —— ‘Two Ramsey Writs and the Domesday Survey’, HR, lxiii (1990), 337–9 —— William the Conqueror (London, 1989; repr. Stroud, 2001) —— Remigius, Bishop of Lincoln (1067–1092) (Lincoln, 1992) —— ‘Four Recently Rediscovered Norman Charters’, AN, xxxv (1995), 35–48 —— ‘The Prosopographical Study of Anglo-Norman Royal Charters’, in Keats-Rohan (ed.), Medieval Prosopography, 89–102 —— ‘Le patronage clérical et intellectuel de l’évêque Odon de Bayeux 1049/50–1097’, in Sylvette Lemagnen (ed.), Chapitres et cathédrales en Normandie, Annales de Normandie, série des congrès des sociétés historiques et archéologiques de Normandie, vol. 2 (Caen, 1997), 105–14 —— ‘Les chartes de confirmation et les pancartes normandes du règne de Guillaume le Conquérant’, in M. Parisse, P. Pégeot, and B.-M. Tock (eds.), Pancartes monastiques (Turnhout, 1998), 95–109 —— Re-ordering the Past and Negotiating the Present in Stenton’s ‘First Century’, The Stenton Lecture for 1999 (University of Reading, 2000) —— ‘England and the “Feudal Revolution” ’, in Il Feudalesimo nell’Alto Medioevo: Settimane di Studi del Centro Italiano di Studi sull’Alto Medioevo, xlvii (2000), 611–49 —— ‘The Conqueror’s Adolescence’, ANS, xxv (2003), 1–18 —— ‘William the Conqueror’s Earliest Historians and the Writing of His Biography’ (2006), in Bates, Crick, and Hamilton (eds.), Writing Medieval Biography, 129–41 —— ‘William the Conqueror’s Wider Western European World’, The Henry Loyn Memorial Lecture, HSJ, xv (2006, for 2004), 73–87 —— ‘William the Conqueror, William fitz Osbern, and Chepstow Castle’, in Rick Turner and Andy Johnson (eds.), Chepstow Castle, its History and Buildings (Logaston, 2006), 15–22 —— ‘The Representation of Queens and Queenship in Anglo-Norman Royal Charters’, in Paul Fouracre and David Ganz (eds.), Frankland: Essays in Honour of Dame Jinty Nelson (Manchester, 2008), 285–303 —— ‘Frank Barlow, 1911–2009’, PBA, clxxii (2011), 3–24 —— ‘Anger, Emotion and a Biography of William the Conqueror’, in Nelson, Reynolds, and Johns (eds.), Gender and Historiography, 21–33 —— ‘Robert of Torigni and the Historia Anglorum’, in Roffe (ed.), The English and their Legacy, 175–84 —— The Normans and Empire (Oxford, 2013) —— ‘The Abbey and the Norman Conquest: An Unusual Case?’, in Tom Licence (ed.), Bury St Edmunds and the Norman Conquest (Woodbridge, 2014), 5–21 —— , and Pierre Bauduin, ‘Autour de l’année 1047: un acte de Guillaume, comte d’Arques, pour Fécamp (18 juillet 1047)’, in Bouet, Bougy, Garnier, and Maneuvrier (eds.), De part et d’autre de la Normandie médiévale, 43–52 —— , and Véronique Gazeau, ‘L’abbaye de Grestain et la famille d’Herluin de Conteville’, AN, xl (1990), 5–30 —— , Julia Crick, and Sarah Hamilton (eds.), Writing Medieval Biography, 750–1250: Essays in Honour of Professor Frank Barlow (Woodbridge, 2006) Bauduin, Pierre, ‘Une famille châtelaine sur les confins normanno-manceaux: les Géré (Xe–XIIIe s.)’, Archéologie médiévale, xxii (1992), 309–56 —— ‘Désigner les parents: le champ de la parenté dans l’œuvre des premiers chroniquers normands’, ANS, xxiv (2002), 71–84 —— ‘En bon usage de la dos dans la Normandie ducale (Xe–début du XIIe siècle)’, in François Bougard, Laurent Feller, and Régine Le Jan (eds.), Dots et douaires dans le haut Moyen Age (Rome, 2002), 429–65 —— ‘L’insertion des normands dans le monde franc, fin IXe–Xe siècles: l’exemple des pratiques matrimoniales’, in Anne-Marie Flambard Héricher (ed.), La progression des Vikings, des raids à la colonisation (Rouen, 2003), 105–17

bibliography

539

—— La première Normandie (Xe–XIe siècles). Sur les frontières de la haute Normandie: identité et construction d’une principauté (Caen, 2004) —— ‘Hugues de Fleury et l’histoire normande’, in Crouch and Thompson (eds.), Normandy and its Neighbours, 900–1250, 157–74 —— ‘Richard II de Normandie: figure princière et transfers culturels (fin dixième–début onzième siècle)’, ANS, xxxvii (2015), 53–82 Baxter, Stephen, ‘The Representation of Lordship and Land Tenure in Domesday Book’, in Hallam and Bates (eds.), Domesday Book, 73–102 —— The Earls of Mercia: Lordship and Power in Late Anglo-Saxon England (Oxford, 2007) —— ‘Edward the Confessor and the Succession Question’, in Mortimer (ed.), Edward the Confessor, 77–118 —— ‘Lordship and Justice in Late Anglo-Saxon England: The Judicial Functions of Soke and Commendation Revisited’, in Baxter, Karkov, Nelson, and Pelteret (eds.), Early Medieval Studies in Memory of Patrick Wormald, 383–420 —— ‘The Limits of the Late Anglo-Saxon State’, in Walter Pohl and Veronika Wieser (eds.), Der frühmittelalterliche Staat – europäische Perspektiven, Forschungen zur Geschichte des Mittelalters, xvi (Vienna, 2009), 503–13 —— , Catherine Karkov, Janet L. Nelson, and David Pelteret (eds.), Early Medieval Studies in Memory of Patrick Wormald (Farnham, 2009) Baylé, Maylis, La Trinité de Caen: sa place dans l’histoire de l’architecture et du décor Romans (Paris, 1979) —— ‘Les évêques et l’architecture normande au XIe siècle’, in Bouet and Neveux (eds.), Les évêques normands du XIe siècle, 151–72 —— (ed.), L’architecture normande au Moyen Age, 2 vols. (Caen, 1997) Beckermann, J. S., ‘Succession in Normandy, 1087, and in England, 1066: The Role of Testamentary Custom’, Speculum, xlvii (1972), 258–60 Beckett, John (ed.), A Centenary History of Nottingham (Manchester, 1997) Bedos-Rézak, Brigitte, ‘The King Enthroned, a New Theme in Anglo-Saxon Royal Iconography: The Seal of Edward the Confessor and its Political Implications’, in Joel T. Rosenthal (ed.), Kings and Kingship, Acta, xi (New York, 1986), 53–88 (reprinted in Brigitte Bedos-Rézak, Form and Order in Medieval France [Aldershot, 1993], chapter 4) Beech, George T., ‘Queen Mathilda of England (1066–1083) and the Abbey of La ChaiseDieu in the Auvergne’, Frühmittelalteriche Studien, xxvii (1993), 350–74 —— Was the Bayeux Tapestry Made in France? The Case for Saint-Florent of Saumur (London and New York, 2005) Bennett, Matthew, ‘Poetry as History? The “Roman de Rou” of Wace as a Source for the Norman Conquest’, ANS, v (1983), 19–39 —— ‘The Myth of the Military Supremacy of Knightly Cavalry’, in Strickland (ed.), Armies, Chivalry and Warfare in Medieval Britain and France, 304–16 —— ‘Violence in Eleventh-Century Normandy: Feud, Warfare and Politics’, in Guy Halsall (ed.), Violence and Society in the Early Medieval West (Woodbridge, 1998), 126–40 —— ‘Military Masculinity in England and Northern France, c.1050–c.1225’, in Hadley (ed.), Masculinity in Medieval Europe, 71–88 —— Campaigns of the Norman Conquest (Oxford, 2001) Bergsagel, John, David Hiley, and Thomas Riis (eds.), Of Chronicles and Kings: National Saints and the Emergence of Nation States in the High Middle Ages (Copenhagen, 2015) Berkhofer, Robert F. III, Day of Reckoning: Power and Accountability in Medieval France (Philadelphia, PA, 2004) Berridge, Peter, ‘Colchester Castle: “Some Tyme Stronge and Statelye, as the Ruynes Do Shewe” ’, in Davies, Riley, Levesque, and Lapiche (eds.), Castles and the Anglo-Norman World, 55–68 Biddle, Martin, ‘Excavations at Winchester 1962–3’, Antiquaries Journal, xliv (1964), 188–219 —— Winchester in the Early Middle Ages: An Edition and Discussion of the Winton Domesday, Winchester Studies, 1 (Oxford, 1976)

540

bibliography

—— ‘Seasonal Festivals and Residence: Winchester, Westminster and Gloucester in the Tenth to Twelfth Centuries’, ANS, viii (1986), 51–72 —— ‘Early Norman Winchester’, in Holt (ed.), Domesday Studies, 311–31 Bijsterveld, Arnoud-Jan A., Henk Teunis, and Andrew Wareham (eds.), Negotiating Secular and Ecclesiastical Power: Western Europe in the Central Middle Ages (Turnhout, 1999) Bischoff, Georges, and Benoît-Michel Tock (eds.), Léon IX et son temps (Turnhout, 2006) Bishop, T. A. M., ‘The Norman Settlement of Yorkshire’, in R. W. Hunt et al. (eds.), Studies in Medieval History Presented to Frederick Maurice Powicke (Oxford, 1948), 1–14 Bisson, Thomas N. (ed.), Cultures of Power: Lordship, Status, and Process in Twelfth-Century Europe (Philadelphia, PA, 1995) —— The Crisis of the Twelfth Century: Power, Lordship, and the Origins of European Government (Princeton, NJ, and Oxford, 2009) Blair, John, ‘An Introduction to the Surrey Domesday’, in Ann Williams and R. W. H. Erskine (eds.), The Surrey Domesday (London, 1989), 1–17 —— Anglo-Saxon Oxfordshire (Stroud, 1994) —— ‘Estate Memoranda of c.1070 from the See of Dorchester-on-Thames’, EHR, cxvi (2001), 114–23 Bolton, Timothy, ‘English Political Refugees at the Court of King Sveinn Ástriðarson’, Medieval Scandinavia, xv (2005), 17–36 —— ‘Ælfgifu of Northampton: Cnut the Great’s Other Woman’, Nottingham Medieval Studies, li (2007), 247–68 —— The Empire of Cnut the Great: Conquest and the Consolidation of Power in Northern Europe in the Early Eleventh Century (Leiden and Boston, MA, 2009) Bouchard, Constance Brittain, Sword, Miter, and Cloister: Nobility and the Church in Burgundy, 980–1198 (Ithaca, NY, 1987) Bouet, Pierre, ‘Orderic Vital lecteur critique de Guillaume de Poitiers’, in C. E. Viola (ed.), Mediaevalia Christiana XIe–XIIIe siècles: Hommage à Raymonde Foreville (Paris, 1989), 32–43 —— Guillaume le Conquérant et les Normands au XIe siècle (Condé-sur-Noireau, 2000) —— ‘Is the Bayeux Tapestry pro-English?’, in Bouet, Levy, and Neveux (eds.), The Bayeux Tapestry, 197–215 —— ‘Le Mont-Saint-Michel entre Bretagne et Normandie de 960 à 1060’, in Quaghebeur and Merdrignac (eds.), Bretons et Normands au Moyen Age, 165–200 —— Hastings: 14 octobre 1066 (Paris, 2010) —— ‘Châteaux et residences princières dans la Tapisserie de Bayeux’, in Davies, Riley, Levesque, and Lapiche (eds.), Castles and the Anglo-Norman World, 135–46 —— , and François Neveux, ‘Edward the Confessor’s Succession according to the Bayeux Tapestry’, in Lewis, Owen-Crocker, and Terkla (eds.), The Bayeux Tapestry, 59–65 —— , and François Neveux (eds.), Les évêques normands du XIe siècle (Caen, 1995) —— , Brian Levy, and François Neveux (eds.), The Bayeux Tapestry: Embroidering the Facts of History (Caen, 2004) —— , Catherine Bougy, Bernard Garnier, and Christophe Maneuvrier (eds.), De part et d’autre de la Normandie médiévale: Recueil d’études en hommage à François Neveux (Caen, 2009) Bouvris, Jean-Michel, ‘En marge de l’année des abbayes normandes: la date de la restauration de l’abbaye d’Almenênches’, Bulletin de la Société Historique et Archéologique de l’Orne, xcviii (1980), 113–41 —— ‘La dédicace de l’église cathédrale Notre-Dame de Bayeux (14 juillet 1077)’, Société des sciences, arts et belles-lettres de Bayeux, xxviii (1982), 3–16 —— ‘Pour une étude prosopographique des familles nobles d’importance moyenne en Normandie au XIe siècle: l’exemple du lignage des Dastin’, Revue de l’Avranchin et du Pays de Granville, lxi (1984), 65–101 —— ‘Contribution à une étude de l’institution vicomtale en Normandie au XIe siècle. L’exemple de la partie orientale du Duché: Les vicomtes de Rouen et de Fécamp’, in Musset, Bouvris, and Maillefer (eds.), Autour du pouvoir ducal, 149–74

bibliography

541

—— ‘L’école capitulaire de Rouen au XIe siècle’, Etudes normandes, iii (1986), 89–103 —— ‘Aux origines du prieuré de Vains: Une version inédite de la confirmation par le Duc Robert-Courte-Heuse d’une donation faite en 1087 par Guillaume le Conquérant à l’abbaye de Saint-Etienne de Caen’, Revue de l’Avranchin et du Pays de Granville, lxiv (1987), 3–21, 67–90 Bradbury, Jim, The Battle of Hastings (Stroud, 1998) Braet, Herman, ‘Le songe de l’arbre chez Wace, Benoît et Aimon de Varennes’, Romania, xci (1970), 255–67 Braun, Patrick, Brigitte Degler-Spengler, and Elsanne Gilomen-Schenkel (eds.), Helvetia Sacra, i/5: Das Bistum Sitten/Le diocèse de Sion, L’archidiocèse de Tarentaise (Basel, 2001) Bredehoft, Thomas A., ‘Malcolm and Margaret: The Poem in Annal 1067D’, in Alice Jorgensen (ed.), Reading the Anglo-Saxon Chronicle: Language, Literature and History (Turnhout, 2010), 31–48 Brett, Martin, ‘A Collection of Anglo-Norman Councils’, JEH, xxvi (1975), 301–8 —— ‘John of Worcester and his Contemporaries’, in R. H. C. Davis and J. M. WallaceHadrill (eds.), The Writing of History in the Middle Ages: Essays Presented to Richard William Southern (Oxford, 1981), 101–26 —— ‘Warfare and its Restraints in England, 1066–1154’, in ‘Militia Christi’ e Crociata nei secoli XI-XIII. Atti della undecima Settimana internazionale di studio Mendola, 28 agosto–1 settembre 1989, Miscellanea del Centro di studi medioevali, vol. xiii (Milan, 1992), 129–44 —— ‘Gundulf (1023/4–1108), Bishop of Rochester’, ODNB —— and David A. Woodman (eds.), The Long Twelfth-Century View of the Anglo-Saxon Past (Farnham and Burlington, VT, 2015) Brooke, C. N. L., ‘Gregorian Reform in Action: Clerical Marriage in England, 1050–1200’, Cambridge Historical Journal, xii (1956), 1–21 (reprinted in C. N. L. Brooke, Medieval Church and Society: Collected Essays [London, 1971], 69–99) —— et al. (eds.), Church and Government in the Middle Ages: Essays Presented to C. R. Cheney on his 70th Birthday (Cambridge, 1976) Brooks, N. P., The Early History of the Church of Canterbury (Leicester, 1984) —— ‘ “Anglo-Saxon Chronicles” or “Old English Royal Annals” ’, in Nelson, Reynolds, and Johns (eds.), Gender and Historiography, 35–48 —— ‘The Archbishopric of Canterbury and the So-called Introduction of Knight-Service into England’, ANS, xxxiv (2012), 41–62 —— , and H. E. Walker, ‘The Authority and Interpretation of the Bayeux Tapestry’, ANS, i (1979), 1–34, 191–9 (reprinted in Gameson [ed.], The Study of the Bayeux Tapestry, 63–92) Broun, Dauvit, Scottish Independence and the Idea of Britain from the Picts to Alexander III (Woodbridge, 2007) Brown, R. Allen, ‘The Battle of Hastings’, ANS, iii (1981), 1–21 (reprinted in Strickland [ed.], Anglo-Norman Warfare, 161–81) Brown, Shirley Ann, The Bayeux Tapestry: Bayeux, Médiathèque Municipale: MS. 1. A Sourcebook, Publications of the Journal of Medieval Latin, no. 9 (Turnhout, 2013) Brown, Warren C., and Piotr Górecki (eds.), Conflict in Medieval Europe: Changing Perspectives on Society and Culture (Aldershot and Burlington, VT, 2003) Brunterc’h, Jean-Pierre, ‘Geffroy Martel, Conan II et les comtes bretons Eudes et Hoël de 1055 à 1060’, in Catherine Laurent, Bernard Merdrignac, and Daniel Pichot (eds.), Mondes de l’Ouest et villes du monde: Regards sur les sociétés médiévales. Mélanges en l’honneur d’André Chédeville (Rennes, 1998), 311–24 Buc, Philippe, Holy War, Martyrdom, and Terror: Christianity, Violence, and the West (Philadelphia, PA, 2015) Buckley, Penelope, The ‘Alexiad’ of Anna Komnene (Cambridge, 2014) Bührer-Thierry, Geneviève, ‘ “Just Anger” or “Vengeful Anger”? The Punishment of Blinding in the Early Medieval West’, in Rosenwein (ed.), Anger’s Past, 75–91

542

bibliography

Buko, Andrzej, and Mike McCarthy (eds.), Making a Medieval Town: Patterns of Early Medieval Urbanization (Warsaw, 2010) Bur, Michel, ‘Léon IX et la France (1026–1054)’, in Bischoff and Tock (eds.), Léon IX et son temps, 233–58 Burkhardt, Stefan, and Thomas Foerster (eds.), Norman Tradition and Transcultural Heritage: Exchanges of Cultures in the ‘Norman’ Peripheries of Medieval Europe (Farnham and Burlington, VT, 2013) Burton, Janet, The Monastic Order in Yorkshire, 1069–1215 (Cambridge, 1999) Campbell, James, ‘Observations on English Government from the Tenth to the Twelfth Century’, TRHS, 5th ser., xxv (1975), 39–54 (reprinted in James Campbell, Essays in AngloSaxon History [London and Ronceverte, 1986], 155–70) —— ‘Some Agents and Agencies of the Late Anglo-Saxon State’, in Holt (ed.), Domesday Studies, 201–18 (reprinted in James Campbell, The Anglo-Saxon State [London and New York, 2000], 201–25) —— ‘Norwich before 1300’, in Rawcliffe and Wilson (eds.), Medieval Norwich, 29–48 Carlson, Eric Gustav, ‘Excavations at Saint Etienne, Caen (1969)’, Gesta, x (1971), 23–30 Carpenter, David, The Minority of Henry III (London, 1990) Carpentier, Vincent, Guillaume le Conquérant et l’estuaire de la Dives, Collection Cahiers d’Auge, no. 3 (Lisieux, 2011) Carré, Gaël, ‘Le rempart nord-ouest du château de Caen (XIe–XVIe siècles). L’apport récent de l’archéologie du bâti’, AN, lix (2009), 3–25 Cartwright, Charlotte, ‘Before She Was Queen: Matilda of Flanders and the Use of comitissa in Norman Ducal Charters’, HSJ, xxii (2012, for 2010), 59–81 Castiñeiras, Manuel, ‘Didacus Gelmiras, Patron of the Arts. Compostela’s Long Journey: from the Periphery to the Center of Romanesque Art’, in Manuel Castiñeiras (ed.), Compostela and Europe: the Story of Diego Gelmirez (Santiago de Compostela and Milan, 2010), 32–97 Chaplais, Pierre, ‘Une charte originale de Guillaume le Conquérant pour l’abbaye de Fécamp’, in L’abbaye bénédictine de Fécamp: ouvrage scientifique du XIIIe centenaire, 658–1958 (Fécamp, 1959), i, 93–104, 335–7 (reprinted in Pierre Chaplais, Essays in Medieval Diplomacy and Administration [London, 1981], chapter 16) —— ‘William of Saint-Calais and the Domesday Survey’, in Holt (ed.), Domesday Studies, 65–77 Charles-Edwards, T. M., Wales and the Britons 350–1064 (Oxford, 2013) Chassel, Jean-Luc, ‘Le serment de Harold dans la Tapisserie de Bayeux et dans les sources pro-normandes des XIe et XIIe siècles’, in Raymond Verdier (ed.), Le serment, 2 vols. (Paris, 1991), 43–53 Chédeville, André, Chartres et ses campagnes (XIe–XIIIe siècles) (Paris, 1973) Chibnall, Marjorie, The World of Orderic Vitalis (Oxford, 1984) —— ‘La carrière de Geoffroi de Montbray’, in Pierre Bouet and François Neveux (eds.), Les évêques normands du XIe siècle (Caen, 1995), 279–93 Christiansen, Eric, ‘The Difficulty of Composing Patriotic History in the Twelfth Century: An Essay on Essays’, in Bergsagel, Hiley, and Riis (eds.), Of Chronicles and Kings, 11–21 Clanchy, M. T., ‘Introduction’, in David C. Douglas, The Normans (London, 2002), xi–xix Clarke, Howard, ‘The Identity of the Designer of the Bayeux Tapestry’, ANS, xxxv (2013), 119–39 Clarke, Peter A., The English Nobility under Edward the Confessor (Oxford, 1994) Collard, Franck, Le Crime de poison au Moyen Age (Paris, 2003) (The Crime of Poison in the Middle Ages, trans. Deborah Nelson-Campbell [Westport, CT], 2008) Constable, Giles, ‘Abbatial Profession in Normandy and England in the Eleventh and Twelfth Century, with Particular Attention to Bec’, in Mario Ascheri (ed.), Ins Wasser geworfen und Ozeane durchquert: Festschrift für Knut Wolfgang Nörr (Cologne, 2003), 105–20

bibliography

543

Cooper, Alan, ‘Extraordinary Privilege: The Trial of Penenden Heath and the Domesday Inquest’, EHR, cxvi (2001), 1,167–92 —— ‘Protestations of Ignorance in Domesday Book’, in Robert F. Berkhofer III, Alan Cooper, and Adam J. Kosto (eds.), The Experience of Power in Medieval Europe, 950–1350 (Aldershot, 2005), 169–81 Corbet, Patrick, Autour de Burchard de Worms. L’Église allemande et les interdits de parenté (IXème–XIIème siècle) (Frankfurt am Main, 2001) —— ‘In multis orbis partibus: Léon IX et les interdits de parenté (1049–1054)’, in Bischoff and Tock (eds.), Léon IX et son temps, 343–53 Cordero Carrete, Felipe Ramon, ‘Los esponsales de una hija de Guillermo el Conquistador con un “Rey de Galicia” ’, Cuaderno de Estudios Gallegos, vii (1952), 55–78 Coulson, Charles, Castles in Medieval Society: Fortresses in England, France, and Ireland in the Central Middle Ages (Oxford, 2003) Cowdrey, H. E. J., ‘Unions and Confraternity with Cluny’, JEH, xvi (1965), 152–62 —— ‘Bishop Ermenfrid of Sion and the Penitential Ordinance following the Battle of Hastings’, JEH, xx (1969), 225–42 —— ‘Pope Gregory VII and the Anglo-Norman Church and Kingdom’, Studi Gregoriani, ix (1972), 77–114 —— ‘The Anglo-Norman Laudes Regiae’, Viator, xii (1981), 37–78 (reprinted in H. E. J. Cowdrey, Popes, Monks and Crusaders [London, 1984], chapter 8) —— ‘Death-Bed Testaments’, in Fälschungen im Mittelalter, iv. Diplomatische Fälschungen II, MGH Schriften, 33 (Hanover, 1988), 703–24 (reprinted in Cowdrey, Popes and Church Reform in the 11th Century, chapter 9) —— ‘William I’s Relations with Cluny Further Considered’, in Judith Loades (ed.), Monastic Studies: The Continuity of Tradition (Bangor, 1990), 1–14 (reprinted in Cowdrey, The Crusades and Latin Monasticism, 11th–12th Centuries, chapter 8) —— ‘The Papacy and the Berengarian Controversy’, in R. B. C. Huygens and F. Niewöhner (eds.), Auctoritas und Ratio. Studien zu Berengar von Tours (Wiesbaden, 1990), 109–38 (Cowdrey, Popes and Church Reform in the 11th Century, chapter 6) —— ‘Lanfranc, the Papacy and Canterbury’, in G. d’Onofrio (ed.), Lanfranco di Pavia, 439–500 (Cowdrey, Popes and Church Reform in the 11th Century, chapter 10) —— ‘Count Simon of Crépy’s Monastic Conversion’, in Pierre Guichard et al. (eds.), Papauté, monachisme et théories politiques, i. Le pouvoir et l’institution ecclésiale: Etudes d’histoire médiévale offertes à Marcel Pacaut (Lyon, 1994), 253–66 (Cowdrey, The Crusades and Latin Monasticism, 11th–12th Centuries, chapter 11) —— Pope Gregory VII, 1073–1085 (Oxford, 1998) —— The Crusades and Latin Monasticism, 11th–12th Centuries (Aldershot and Brookfield, VT, 1999) —— Popes and Church Reform in the 11th Century (Aldershot and Burlington, VT, 2000) —— Lanfranc: Scholar, Monk and Archbishop (Oxford, 2003) —— ‘Robert of Jumièges [Robert Champart] (d.1052/1055)’, ODNB —— ‘Stigand (d.1072), Archbishop of Canterbury’, ODNB —— ‘King Harold II and the Bayeux Tapestry: A Critical Introduction’, in Owen-Crocker (ed.), King Harold II and the Bayeux Tapestry, 1–15 Cownie, Emma, Religious Patronage in Anglo-Norman England, 1066–1135, Royal Historical Society Studies in History (Woodbridge, 1998) Coz, Yann, Guillaume le Conquérant (Paris, 2011) Craster, H. H. E., ‘The Red Book of Durham’, EHR, xl (1925), 504–32 Creighton, Oliver, ‘Castle, Burh and Borough: Unravelling an Urban Landscape of Power at Wallingford, Oxfordshire’, Château Gaillard, xxvi (2014), 113–24 Crisp, Ryan Patrick, ‘Consanguinity and the Saint-Aubin Genealogies’, HSJ, xiv (2005, for 2003), 105–15 Crosby, S. McK., The Royal Abbey of Saint-Denis from its Beginnings to the Death of Suger, 475–1151, ed. Pamela Z. Blum (New Haven, CT, and London, 1987)

544

bibliography

Crouch, David, The Beaumont Twins: The Roots and Branches of Power in the Twelfth Century (Cambridge, 1986) —— ‘The Culture of Death in the Anglo-Norman World’, in C. Warren Hollister (ed.), Anglo-Norman Political Culture and the Twelfth-Century Renaissance (Woodbridge, 1997), 157–80 —— The Reign of King Stephen, 1135–1154 (London and New York, 2000) —— The Normans: The History of a Dynasty (London and New York, 2002) —— ‘Beaumont, Robert de, Count of Meulan and First Earl of Leicester (d.1118)’, ODNB —— The Birth of Nobility: Constructing Aristocracy in England and France, 900–1300 (Harlow, 2005) —— ‘The Transformation of Medieval Gwent’, in Ralph A. Griffiths, Tony Hopkins, and Ray Howell (eds.), The Gwent County History, volume 2: The Age of the Marcher Lords, c.1070– 1536 (Cardiff, 2008), 1–45 —— The English Aristocracy 1070–1272: A Social Transformation (New Haven, CT, and London, 2011) —— , and Kathleen Thompson (eds.), Normandy and its Neighbours, 900–1250: Essays for David Bates (Turnhout, 2011) Cubbitt, Catherine, ‘Bishops and Succession Crises in Tenth- and Eleventh-Century England’, in Körntgen and Waßenhoven (eds.), Patterns of Episcopal Power, 111–26 Currie, C. R. J., ‘Odda’s Chapel, Ealdred’s Inscriptions? The Deerhurst Inscriptions in Some Continental Contexts’, HR, lxxxiii (2010), 1–45 Cushing, Kathleen G., Papacy and Law in the Gregorian Revolution: The Canonistic Work of Anselm of Lucca (Oxford, 1998) Cutler, K. E., ‘The Godwinist Hostages: The Case for 1051’, Annuale Medievale, xii (1972), 70–7 Dalton, Paul, Conquest, Anarchy and Lordship: Yorkshire 1066–1154 (Cambridge, 1994) —— ‘The Outlaw Hereward “the Wake”: His Companions and Enemies’, in John C. Appleby and Paul Dalton (eds.), Outlaws in Medieval and Early Modern England: Crime, Government and Society, c.1066–c.1600 (Farnham, 2009), 7–36 —— ‘William the Peacemaker: The Submission of the English to the Duke of Normandy, October 1066–January 1067’, in Dalton and Luscombe (eds.), Rulership and Rebellion in the Anglo-Norman World, c.1066–c.1216, 21–44 —— , and David Luscombe (eds.), Rulership and Rebellion in the Anglo-Norman World, c.1066–c.1216: Essays in Honour of Professor Edmund King (Farnham and Burlington, VT, 2015) Darby, H. C., and I. S. Maxwell (eds.), The Domesday Geography of Northern England (Cambridge, 1962) Dastugue, Jean, ‘Le femur de Guillaume le Conquérant’, AN, xxxvii (1987), 5–10 Davies, John A., Angela Riley, Jean-Marie Levesque, and Charlotte Lapiche (eds.), Castles and the Anglo-Norman World: Proceedings of a Conference Held at Norwich Castle in 2012 (Oxford and Philadelphia, PA, 2016) Davies, R. R., The Age of Conquest: Wales, 1063–1415 (Oxford, 1991) —— The First English Empire: Power and Identities in the British Isles, 1093–1343 (Oxford, 2000) Davis, Jennifer R., Charlemagne’s Practice of Empire (Cambridge, 2015) Davis, R. H. C., ‘William of Jumièges, Robert Curthose and the Norman Succession’, EHR, vc (1980), 597–606 —— ‘David Charles Douglas, 1898–1982’, PBA, lxix (1983), 513–42 —— ‘Domesday Book: Continental Parallels’, in Holt (ed.), Domesday Studies, 15–39 —— ‘The Warhorses of the Normans’, ANS, x (1988), 67–82 (reprinted in R. H. C. Davis, From Alfred the Great to Stephen [London, 1991], 63–78) Davy, Gilduin, Le duc et la loi: Héritages, images et expressions du pouvoir normative dans le duché de Normandie, des origines à la mort du Conquérant (fin du IXe siècle–1087) (Paris, 2004) —— Guillaume le Conquérant. Le bâtard de Normandie (Paris, 2014) De Boüard, Michel, ‘Notes et hypothèses sur Maurille, moine de Fécamp’, in L’abbaye bénédictine de Fécamp. Ouvrage scientifique du XIIIe centenaire (658–1958), i (Fécamp, 1959), 81–92

bibliography

545

—— ‘Sur les origines de la Trêve de Dieu en Normandie’, AN, ix (1959), 169–89 —— Guillaume le Conquérant (Paris, 1984) De Fougerolles, Paula, ‘Pope Gregory VII, the Archbishopric of Dol and the Normans’, ANS, xxi (1999), 47–66 De Jong, Mayke, In Samuel’s Image: Child Oblation in the Early Medieval West (Leiden, New York, and Cologne, 1996) De Montclos, Jean, Lanfranc et Bérenger: la controverse eucharistique du XIe siècle (Louvain, 1971) Declercq, Georges, ‘Van privaatoorkonde tot vorstelijke oorkonde. De oorkonden van de eerste graven van Vlaanderen, inzonderheid voor de Sint-Pietersabdij te Gent (10de– 11de eeuw)’, in N. Barré, Th. de Hemptinne, and J.-M. Duvosquel (eds.), Chancelleries princières et Scriptoria dans les anciens Pays-Bas Xe–XVe siècles / Vorstelijke kanselarijen en Scriptoria in de Lage Landen 10de-15de eeuw (Brussels, 2010), 41–77 Delisle, Léopold, Histoire du château et des sires de Saint-Sauveur-le-Vicomte (Valognes, 1867) —— ‘Canons du concile tenu à Lisieux en 1064’, Journal des Savants (1901), 516–21 Depping, Georges-Bernard, Histoire de la Normandie, depuis les temps les plus reculés jusqu’à la conquête de l’Angleterre en 1066, 2 vols. (Rouen, 1835) Deslandes, E., Le Trésor de l’église Notre-Dame de Bayeux, d’après les inventaires manuscrits de 1476, 1480, 1498 conservés à la Bibliothèque du chapitre de Bayeux (Paris, 1898) DeVries, Kelly, The Norwegian Invasion of England in 1066 (Woodbridge, 1999) Dewhurst, Sir John, ‘A Historical Obstetric Enigma: How Tall Was Matilda?’, Journal of Obstetrics and Gynaecology, i (1981), 271–2 Dhondt, J., ‘Henri Ier, l’Empire et l’Anjou (1034–1056)’, Revue belge de philologie et d’histoire, xxv (1946–7), 87–109 Dickins, Bruce, ‘Fagaduna in Orderic (A.D. 1075)’, in Folke Sandgren (ed.), ‘Otium et negotium’. Studies in Onomatology and Library Science Presented to Olof von Feilitzen (Stockholm, 1973), 44–5 Dickinson, F. H., ‘The Sale of Combe’, Somerset Archaeological and Natural History Society Proceedings, xxii (1876), 106–13 Diederich, Toni, ‘Zum Gebrauch lateinischer Verskunst in Siegelumschriften’, in Toni Diederich, Siegelkunde, Beiträge zu ihrer Vertiefung und Weiter führing (Weimar, 2012), 146–77 Dillmann, François-Xavier, ‘Les figures de proue de la Tapisserie de Bayeux et les traditions nautiques de la Scandinavie ancienne’, in Sylvette Lemagnen, La Tapisserie de Bayeux. Une chronique des temps Vikings? Actes du colloque international de Bayeux, 29 et 30 mars 2007, 177–93 Dockray-Miller, Mary, The Books and the Life of Judith of Flanders (Farnham and Burlington, VT, 2015) Dodwell, C. R., ‘The Bayeux Tapestry and the French Secular Epic’, Burlington Magazine, cviii (1966), 549–60 (reprinted in Gameson [ed.], The Study of the Bayeux Tapestry, 47–62) —— Anglo-Saxon Art: A New Perspective (Manchester, 1982) D’Onofrio, G. (ed.), Lanfranco di Pavia e l’Europa del secolo XI nel IX centenario della morte (Rome, 1993) Douglas, David C., ‘The Ancestors of William fitz Osbern’, EHR, lix (1944), 62–79 —— ‘The Earliest Norman Counts’, EHR, lxi (1946), 129–56 —— The Norman Conquest and British Historians (Glasgow, 1946) —— ‘Some Problems of Early Norman Chronology’, EHR, lxv (1950), 289–303 —— English Scholars (London, 1951) —— ‘Edward the Confessor, Duke William of Normandy, and the English Succession’, EHR, lxviii (1953), 526–45 —— William the Conqueror: The Norman Impact on England (London, 1964; new edn., New Haven, CT, and London, 1999) —— The Norman Achievement 1050–1100 (London, 1969) —— The Norman Fate, 1100–1154 (London, 1976) —— The Time and the Hour (London, 1977) —— The Normans (London, 2002)

546

bibliography

Downham, Clare, ‘England and the Irish-Sea Zone in the Eleventh Century’, ANS, xxvi (2004), 55–73 Duby, Georges, ‘Youth in Aristocratic Society: Northwestern France in the Twelfth Century’, in Georges Duby, The Chivalrous Society, trans. Cynthia Postan (London, 1977), 112–22 (original French version published as ‘Au XIIe siècle: les “jeunes” dans la société aristocratique dans la France du Nord-Ouest’, Annales Economies, sociétés, civilisations, xiv [1964], 835–46) Dumville, David N., ‘The Ætheling: A Study in Anglo-Saxon Constitutional History’, ASE, viii (1979), 1–33 —— ‘Some Aspects of Annalistic-Writing at Canterbury in the Eleventh and Twelfth Centuries’, Peritia, ii (1983), 23–57 —— ‘Anglo-Saxon Books: Treasure in Norman Hands?’, ANS, xvi (1994), 83–99 Dunbabin, Jean, ‘Geoffrey of Chaumont, Thibaud of Blois and William the Conqueror’, ANS, xvi (1994), 101–16 Duncan, A. A. M., The Kingship of Scots, 842–1292: Succession and Independence (Edinburgh, 2002) Eales, Richard, ‘Castles and Borders in England after 1066’, Château Gaillard, xxvi (2014), 149–57 —— , and Richard Sharpe (eds.), Canterbury and the Norman Conquest: Churches, Saints and Scholars, 1066–1109 (London and Rio Grande, 1995) Early, Rob, ‘Le château de Mayenne: les témoins archéologiques de l’évolution d’un centre de pouvoir entre le Xe et le XIIe siècle’, Château Gaillard, xx (2002), 247–62 Engels, L. J., ‘De Obitu Willelmi ducis Normannorum regisque Anglorum: Texte, modèles, valeur et origine’, in Mélanges Christine Mohrmann: Nouveau recueil offert par ses anciens élèves (Utrecht and Antwerp, 1973), 209–55 English, Barbara, ‘William the Conqueror and the Anglo-Norman Succession’, HR, lxiv (1991), 221–36 —— ‘Towns, Mottes and Ring-Works of the Conquest’, in Ayton and Price (eds.), The Medieval Military Revolution, 45–61 —— ‘The Coronation of Harold in the Bayeux Tapestry’, in Pierre Bouet, Brian Levy, and François Neveux (eds.), The Bayeux Tapestry: Embroidering the Facts of History (Caen, 2004), 347–82 Erdmann, Carl, Die Entstehung des Kreuzzugsgedankens (Stuttgart, 1935) (English translation: The Origin of the Idea of Crusade, trans. Marshall W. Baldwin and Walter Goffart [Princeton, NJ, 1977]) Esders, Stefan, ‘Regionale Selbstbehauptung zwischen Byzanz und dem Frankenreich: Die inquisitio der Rechtsgewohnheiten Istriens durch die Sendboten Karls der Großen und Pippins von Italien’, in Stefan Esders and Thomas Scharff (eds.), Eid und Wahrheitssuche. Studien zu rechttlichen Befragungspraktiken im Mittelalter und früher Neuzeit, Gesellschaft, Kultur und Schrift Mediävistiche Beiträge, vii (1999), 49–112 Everard, Judith A., ‘The Foundation of an Alien Priory at Linton, Cambridgeshire’, Proceedings of the Cambridge Antiquarian Society, lxxxvi (1997), 169–74 Fenton, Kirsten A., Gender, Nation and Conquest in the Works of William of Malmesbury (Woodbridge, 2008) Fentress, James, and Chris Wickham, Social Memory (Oxford, 1992) Fernie, Eric, The Architecture of Norman England (Oxford, 2000) —— ‘Three Romanesque Great Churches in Germany, France and England, and the Discipline of Architectural History’, Architectural History, liv (2011), 1–22 Feuchère, P., ‘Une tentative manquée de concentration territoriale entre Somme et Seine: la principauté d’Amiens-Valois au XIe siècle’, Moyen Age, lx (1954), 1–37 Finberg, H. P. R., ‘Uffculme’, in W. G. Hoskins and H. P. R. Finberg (eds.), Devonshire Studies (London, 1952), 59–77 Finn, R. Welldon, The Norman Conquest and its Effects on the Economy: 1066–1086 (London, 1971) Flambard Héricher, Anne-Marie (ed.), La progression des Vikings, des raids à la colonisation, Cahiers du GRHIS, 14 (Rouen, 2003)

bibliography

547

—— (ed.), Les lieux du pouvoir au Moyen Age en Normandie et sur ses marges, Publications du CRAHM (Caen, 2006) Fleming, Robin, ‘Monastic Lands and England’s Defence in the Viking Age’, EHR, c (1985), 247–65 —— Kings and Lords in Conquest England (Cambridge, 1991) —— Domesday Book and the Law: Society and Legal Custom in Early Medieval England (Cambridge, 1998) —— ‘Harold II [Harold Godwineson] (1022/3–1066), King of England’, ODNB Fletcher, Christopher, Jean-Philippe Genet, and John Watts (eds.), Government and Political Life in England and France, c.1300–c.1500 (Cambridge, 2015) Fletcher, Richard, Bloodfeud: Murder and Revenge in Anglo-Saxon England (London, 2002) Fliche, Augustin, Le règne de Philippe Ier, roi de France (1060–1108) (Paris, 1912) Foerster, Thomas, ‘Imperial Tradition and Norman Heritage: Cultures of Violence and Cruelty’, in Stefan Burkhardt and Thomas Foerster (eds.), Norman Tradition and Transcultural Heritage: Exchanges of Cultures in the ‘Norman’ Peripheries of Medieval Europe (Farnham and Burlington, VT, 2013), 161–88 Fontanel, Julie, ‘La réorganisation religieuse sous Guillaume le Conquérant: le cas de l’église de Coutances’, Revue de l’Avranchin et du pays de Granville, lxxvii (2000), 189–208 Foreville, R., ‘The Synod of the Province of Rouen in the Eleventh and Twelfth Centuries’, in C. N. L. Brooke et al. (eds.), Church and Government in the Middle Ages: Essays Presented to C. R. Cheney on his 70th Birthday (Cambridge, 1976), 19–39 Foreville, Raymonde (ed.), Les mutations socio-culturelles au tournant des XIe–XIIe siècles (Paris, 1984) Foulon, Jean-Hervé, ‘Le chevalier Herluin et la fondation de l’abbaye du Bec: un dossier complexe entre tentation érémitique et normalisation cénobitique’, Revue Historique, 663 (2012), 563–607 —— ‘Les investitures abbatiales en Normandie: quelques réflexions autour du cas de l’abbaye du Bec-Hellouin (1034–1136)’, ANS, xxxv (2013), 181–212 —— ‘La liberté de l’abbaye du Bec entre image et réalité. Réflexions autour d’un modèle réformateur normand aux XIe et XIIe siècles’, in Barrow, Delivré, and Gazeau (eds.), Autour de Lanfranc (1010–2010), 57–83 Foys, Martin K., Karen Eileen Overbey, and Dan Terkla (eds.), The Bayeux Tapestry: New Interpretations (Woodbridge, 2009) Freeman, E. A., The History of the Norman Conquest of England: Its Causes and Its Results, 2nd revised edn., 6 vols. (Oxford, 1867–79) Fröhlich, Walter, ‘St Anselm’s Special Relationship with William the Conqueror’, ANS, x (1988), 101–10 Gameson, Richard (ed.), The Study of the Bayeux Tapestry (Woodbridge, 1997) Garnett, George, ‘Coronation and Propaganda: Some Implications of the Norman Claim to the Throne of England’, TRHS, 5th ser., xxxvi (1986), 91–116 —— ‘ “Franci et Angli”: The Legal Distinctions between Peoples after the Conquest’, ANS, viii (1986), 109–37 —— ‘ “Ducal” Succession in Early Normandy’, in Garnett and Hudson (eds.), Law and Government in Medieval England and Normandy, 80–110 —— ‘The Third Recension of the English Coronation ordo: The Manuscripts’, HSJ, xi (1998), 43–71 —— Conquered England: Kingship, Succession and Tenure, 1066–1166 (Oxford, 2007) —— , and John Hudson (eds.), Law and Government in Medieval England and Normandy: Essays in Honour of Sir James Holt (Cambridge, 1994) Gathagan, Laura L., ‘The Trappings of Power: The Coronation of Matilda of Flanders’, HSJ, xiii (2004, for 1999), 21–39 Gauthiez, Bernard, ‘Hypothèses sur la fortification de Rouen au onzième siècle: Le Donjon, la Tour de Richard II, et l’enceinte de Guillaume’, ANS, xiv (1992), 61–76 —— Atlas morphologique des villes de Normandie (Lyon, 1999)

548

bibliography

—— ‘Verneuil-sur-Avre, Falaise, Pont-Audemer et Lisieux en Normandie’, in Bernard Gauthiez, Elisabeth Zadora-Rio, and Henri Galinié (eds.), Village et ville au Moyen Age: les dynamiques morphologiques, 2 vols. (Tours, 2003), 35–95 —— ‘The Evolution of Towns in the Duchy of Normandy in the 11th–15th Centuries: An Essay in Historical Geography Using Material and Quantitative Data’, in Buko and McCarthy (eds.), Making a Medieval Town, 81–103 —— ‘The Urban Development of Rouen, 989–1345’, in Hicks and Brenner (eds.), Society and Culture in Medieval Rouen, 17–64 Gautier, Alban, ‘Comment Harold prêta serment: circonstances et interprétations d’un rituel politique’, Cahiers de civilisation médiévale, lv (2012), 33–56 Gazeau, Véronique, ‘Le patrimoine de l’évêque Hugues de Bayeux (1011–1049)’, in Bouet and Neveux (eds.), Les évêques normands du XIe siècle, 139–47 —— Normannia Monastica (Xe–XIIe siècle): Princes normands et abbés bénédictins; Prosopographie des abbés bénédictins, 2 vols. (Caen, 2007) —— ‘De l’usage des souscriptions et de l’utilité de leur étude: remarques sur l’aristocratie normande dans la première moitié du XIe siècle: L’exemple d’Onfroy (dit “de Vieilles”)’, in Crouch and Thompson (eds.), Normandy and its Neighbours, 900–1250, 215–21 —— ‘Imagining the Conqueror: The Changing Image of William the Conqueror, 1830– 1945’, HSJ, xxv (2014, for 2013), 245–64 Gazzoli, Paul, ‘Anglo-Danish Connections and the Origins of the Cult of Knud’, Selected Papers from the Inaugural St Magnus Conference 2011, Journal of the North Atlantic, special vol. iv (2013), 69–76 Gem, Richard, ‘Canterbury and the Cushion Capital: A Commentary on Passages from Goscelin’s De Miraculis Sancti Augustini’, in Neil Stratford (ed.), Romanesque and Gothic: Essays for George Zarnecki, 2 vols. (Woodbridge, 1987), i, 83–101 —— ‘The First Romanesque Cathedral at Salisbury’, in Eric Fernie and Paul Crossley (eds.), Medieval Architecture and its Intellectual Context (London, 1990), 9–18 Gerard, Daniel, ‘Ermenfroi de Sion, l’archevêque Lanfranc et le problème des ecclésiastiques rebelles’, in Barrow, Delivré, and Gazeau (eds.), Autour de Lanfranc, 305–11 Giandrea, Mary Frances, Episcopal Culture in Late Anglo-Saxon England (Woodbridge, 2007) Gibson, Margaret, Lanfranc of Bec (Oxford, 1978) —— ‘The Image of Lanfranc’ (1993), in d’Onofrio (ed.), Lanfranco di Pavia, 21–8 Gillingham, John, ‘ “The Most Precious Jewel in the English Crown”: Levels of Danegeld and Heregeld in the Early Eleventh Century’, EHR, civ (1989), 373–84 —— ‘Chronicles and Coins as Evidence of Levels of Tribute and Taxation in Late Tenthand Early Eleventh-Century England’, EHR, cv (1990), 939–50 —— ‘Conquering the Barbarians: War and Chivalry in Britain and Ireland’, HSJ, iv (1992), 67–84 (reprinted in Gillingham, The English in the Twelfth Century, 41–58) —— ‘William the Bastard at War’, in Harper-Bill, Holdsworth, and Nelson (eds.), Studies in Medieval History Presented to R. Allen Brown, 141–58 (also printed in Strickland [ed.], AngloNorman Warfare, 143–60) —— ‘1066 and the Introduction of Chivalry into England’, in Garnett and Hudson (eds.), Law and Government in Medieval England and Normandy, 31–55 (reprinted in Gillingham, The English in the Twelfth Century, 202–29) —— The English in the Twelfth Century: Imperialism, National Identity and Political Values (Woodbridge, 2000) —— ‘Problems of Integration within the Lands Ruled by the Norman and Angevin Kings of England’, in Werner Maleczek (ed.), Fragen der Politischen Integration in Mittelalterlichen Europa, Vorträge und Forschungen, lxiii (Ostfildern, 2005), 85–135 —— ‘ “Holding to the Rules of War (Bellica Iura Tenentes)”: Right Conduct Before, During, and After Battle in North-Western Europe in the Eleventh Century’, ANS, xxix (2007), 1–15 —— ‘At the Deathbeds of the Kings of England, 1066–1216’, in Brigitte Kasten (ed.), Herrscher-und Fürstentestamente im Westeuropäischen Mittelalter (Cologne, Weimar, and Vienna, 2008), 509–30

bibliography

549

—— ‘Women, Children and the Profits of War’, in Nelson, Reynolds, and Johns (eds.), Gender and Historiography, 61–74 —— ‘Surrender in Medieval Europe – An Indirect Approach’, in Holger Afflerbach and Hew Strachan (eds.), How Fighting Ends: A History of Surrender (Oxford, 2012), 55–72 —— William II: The Red King (London, 2015) Gillmor, C. M., ‘Naval Logistics of the Cross-Channel Operation, 1066’, ANS, vii (1985), 105–31 Golding, Brian, ‘The Coming of the Cluniacs’, ANS, iii (1981), 65–77, 208–12 —— ‘Robert of Mortain’, ANS, xiii (1990), 119–44 —— ‘Robert, Count of Mortain (d.1095)’, ODNB Gonduin, Stéphane W., Guillaume Roi (Condé-sur-Noireau, 2002) —— Guillaume le Conquérant et les femmes (Cahors, 2012) Goodall, John, The English Castle, 1066–1650 (New Haven, CT, and London, 2011) Goullet, Monique, ‘Poésie et mémoire des morts: le rouleau funèbre de Mathilde, abbesse de la Sainte-Trinité de Caen (†1113)’, in Jean-François Cottier, Martin Gravel, and Sébastien Rossignol (eds.), Ad libros! Mélanges d’études médiévales offerts à Denise Angers et Joseph-Claude Poulin (Montreal, 2010), 163–98 Gransden, Antonia, Historical Writing in England, c.550 to c.1307 (London, 1974) Grant, Alexander, ‘At the Northern Edge: Alba and its Normans’, in Stringer and Jotischky (eds.), Norman Expansion, 49–85 Grant, Lindy, Architecture and Society in Normandy 1120–1270 (New Haven, CT, and London, 2005) Grape, Wolfgang, The Bayeux Tapestry: Monument to a Norman Triumph (Munich and New York, 1994) Green, Judith A., ‘The Last Century of Danegeld’, EHR, xcvi (1981), 241–58 —— ‘The Sheriffs of William the Conqueror’, ANS, v (1983), 129–43 —— ‘Lords of the Norman Vexin’, in John Gillingham and J. C. Holt (eds.), War and Government in the Middle Ages (Woodbridge, 1984), 47–61 —— The Aristocracy of Norman England (Cambridge, 1997) —— Henry I, King of England, Duke of Normandy (Cambridge, 2006) —— ‘Kingship, Lordship, and Community in Eleventh-Century England’, ANS, xxxi (2009), 1–16 —— ‘Forest Laws in England and Normandy in the Twelfth Century’, HR, lxxxvi (2013), 416–31 Grierson, Philip, ‘A Visit of Earl Harold to Flanders in 1056’, EHR, li (1936), 90–7 —— ‘The Relations between England and Flanders before the Norman Conquest’, TRHS, 4th ser., xxiii (1941), 71–112 Grosse, Rolf, Saint-Denis zwischen Adel und König: Die Zeit vor Suger (1053–1122) (Stuttgart, 2002) Guéry, C., Histoire de l’abbaye de Lyre (Evreux, 1917) Guillot, Olivier, Le comte d’Anjou et son entourage au XIe siècle, 2 vols. (Paris, 1972) —— ‘L’appui apporté par Henrier à Guillaume d’Arques dans le conflit de ce dernier avec le duc Guillaume le Bâtard (fin 1053–1054): une motivation juridique?’, in Bernard d’Alteroche, Florence Demoulin-Auzuri, Olivier Descamps, and Franck Roumy (eds.), Mélanges en l’honneur d’Anne Lefebvre-Teillard (Paris, 2009), 473–93 Guillotel, Hubert, ‘Bretagne et papauté au XIe siècle’, in Rolf Grosse (ed.), L’Eglise de France et la papauté (Xe–XIIIe siècle). Actes du XXVIe colloque historique franco-allemand organisé en coopération avec l’Ecole nationale des chartes par l’Institut historique allemande de Paris (Bonn, 1993), 265–86 Guimon, Timofey V., ‘The Writing of Annals in Eleventh-Century England: Palaeography and Textual History’, in Alexander R. Rumble (ed.), Writing and Texts in Anglo-Saxon England (Cambridge, 2006), 137–45 Guizard, Fabrice, ‘Les accidents de chasse dans les récits du premier Moyen Age’, in Laurent Jégou, Sylvie Joye, Thomas Lienhard, and Jens Schneider (eds.), Faire lien: Aristocratie, réseaux et échanges compétitifs. Mélanges en l’honneur de Régine Le Jan (Paris, 2015), 289–97

550

bibliography

Gullick, Michael, ‘The Hand of Symeon of Durham: Further Observations on the Durham Martyrology Scribe’, in Rollason (ed.), Symeon of Durham, 14–31 —— ‘The English-Owned Manuscripts of the Collectio Lanfranci (s.xi/xii)’, in Lynda Dennison (ed.), The Legacy of M. R. James: Papers from the 1995 Cambridge Symposium (Donnington, 2001), 99–107 —— Lanfranc and the Oldest Manuscript of the Collectio Lanfranci’, in Bruce C. Brassington and Kathleen G. Cushing (eds.), Bishops, Texts, and the Use of Canon Law around 1100: Essays in Honour of Martin Brett (Aldershot, 2008) Hadley, Dawn M. (ed.), Masculinity in Medieval Europe (London and New York, 1999) —— The Vikings in England (Manchester, 2006) Hagger, Mark, ‘The Norman Vicomte, c.1035–1135: What Did He Do?’, ANS, xxix (2007), 65–83 —— ‘Secular Law and Custom in Ducal Normandy, c.1000–1144’, Speculum, lxxxv (2010), 827–67 —— ‘How the West Was Won: The Norman Dukes and the Cotentin, c.987–1087’, JMH, xxxviii (2012), 20–55 —— William: King and Conqueror (London and New York, 2012) —— ‘Confrontation and Unification: Approaches to the Political History of Normandy, 911–1035’, History Compass, xi (2013), 429–42 Hallam, Elizabeth M., ‘Royal Burial and the Cult of Kingship in France and England, 1060–1330’, JMH, viii (1982), 339–80 —— Domesday Book through Nine Centuries (London, 1986) ——, and David Bates (eds.), Domesday Book (Stroud, 2001) Halphen, Louis, Le comté d’Anjou au XIe siècle (Paris, 1906) Halsall, Guy (ed.), Warfare and Society in the Barbarian West, 450–900 (London, 2003) Hamilton, Sarah, The Practice of Penance, 900–1050, Royal Historical Society Studies in History, new series (Woodbridge, 2001) Hare, Michael, ‘Kings, Crowns, and Festivals: The Origins of Gloucester as a Royal Ceremonial Centre’, Transactions of the Bristol and Gloucestershire Archaeological Society, cxv (1997), 41–78 Harper-Bill, Christopher, Christopher J. Holdsworth, and Janet L. Nelson (eds.), Studies in Medieval History Presented to R. Allen Brown (Woodbridge, 1989) Harris, Roland B., ‘Recent Research on the White Tower: Reconstructing and Dating the Norman Building’, in Davies, Riley, Levesque, and Lapiche (eds.), Castles and the AngloNorman World, 177–89 Harrison, Stuart, and Christopher Norton, ‘Lastingham and the Architecture of the Benedictine Revival in Northumbria’, ANS, xxxiv (2012), 63–103 ——, and John McNeill, ‘The Romanesque Buildings at Westminster Abbey’, in Warwick Rodwell and Tim Tatton-Brown (eds.), Westminster: I. The Art, Architecture and Archaeology of the Royal Abbey, British Archaeological Association Conference Transactions, xxxix, Part I (Leeds, 2015), 69–103 Hart, C. R., ‘Hereward “the Wake” and his Companions’, in C. R. Hart, The Danelaw (London and Rio Grande, 1992), 625–48 Harvey, Barbara, Westminster Abbey and its Estates in the Middle Ages (Oxford, 1977) Harvey, Sally, ‘Eustace II of Boulogne, the Crises of 1051–2 and the English Coinage’, in Roffe (ed.), The English and their Legacy, 149–57 —— Domesday: Book of Judgement (Oxford, 2014) —— ‘ “A Deed without a Name” ’, in Roffe and Keats-Rohan (eds.), Domesday Now, 277–88 Haskins, C. H., ‘A Charter of Canute for Fécamp’, EHR, xxxiii (1918), 342–4 —— Norman Institutions (Cambridge, MA, 1918) Hayward, Paul A., ‘Some Reflections on the Historical Value of the So-Called Acta Lanfranci’, HR, lxxvii (2004), 141–60 Henry, Gilles, Guillaume le Conquérant (Paris, 1996)

bibliography

551

Herrick, Samantha Kahn, Imagining the Sacred Past: Hagiography and Power in Early Normandy (Cambridge, MA, and London, 2007) Hervé, Alain, Guillaume le Conquérant (Paris, 2003) Heslop, T. A., ‘A Walrus Ivory Seal Matrix from Lincoln’, Antiquaries Journal, lxvi (1986), 371–2 —— ‘Regarding the Spectators of the Bayeux Tapestry: Bishop Odo and His Circle’, Art History, xxxii (2009), 223–49 —— ‘Constantine and Helena: The Roman in English Romanesque’, in Jill A. Franklin, T. A. Heslop, and Christine Stevenson (eds.), Architecture and Interpretation: Essays for Eric Fernie (Woodbridge, 2012), 163–75 Hicks, Carola, The Bayeux Tapestry: The Life-Story of a Masterpiece (London, 2006) —— ‘The Patronage of Queen Edith’, in Lewis, Owen-Crocker, and Terkla (eds.), The Bayeux Tapestry, 5–9 Hicks, Leonie V., ‘The Concept of the Frontier in Norman Chronicles: A Comparative Approach’, in Stringer and Andrew Jotischky (eds.), Norman Expansion, 143–64 ——, and Elma Brenner (eds.), Society and Culture in Medieval Rouen, 911–1300 (Turnhout, 2013) Higham, N. J., ‘The Domesday Survey: Context and Purpose’, History, lxxviii (1993), 7–21 —— The Death of Anglo-Saxon England (Stroud, 1997) —— ‘Harold Godwinesson: The Construction of Kingship’, in Owen-Crocker (ed.), King Harold II and the Bayeux Tapestry, 19–34 Higham, Robert, Making Anglo-Saxon Devon (Exeter, 2008) —— ‘Castle Studies in Transition: A Forty-Year Reflection’, Archaeological Journal, clxvii (2010) —— ‘William the Conqueror’s Siege of Exeter in 1068’, Report of the Transactions of the Devonshire Association for the Advancement of Science, cxlv (2013), 93–132 Hillion, Yannick, ‘Arnaud et Hoël, deux évêques du Mans au service de Guillaume le Conquérant’, Annales de Bretagne et des pays de l’Ouest, cx (2003), 49–77 Hoffman, Erich, ‘Dänemark und England zur Zeit König Sven Estridsons’, in Horst Furhmann, Hans Eberhard Mayer, and Klaus Wriedt (eds.), Aus Reichsgeschichte und Nordischer Geschichte, Keiler Historische Studien 16 (Stuttgart, 1972), 92–111 Hoffmann, Hartmut, Gottesfriede und Treuga Dei, Schriften der MGH, xx (Stuttgart, 1964) Hollister, Warren C., edited and completed by A. Clark Frost, Henry I (New Haven, CT, and London, 2001) Holopainen, Toivo J., ‘ “Lanfranc of Bec” and Berengar of Tours’, ANS, xxxiv (2012), 105–21 —— ‘L’héritage intellectuel de Lanfranc revisité’, in Barrow, Delivré, and Gazeau (eds.), Autour de Lanfranc (1010–2010), 107–15 Holt, J. C., ‘Politics and Property in Early Medieval England’, Past and Present, no. 57 (1972), 3–52 (reprinted in Holt (ed.), Colonial England, 1066–1215, 113–59) —— ‘1086’, in Holt (ed.), Domesday Studies, 41–64 —— (ed.), Domesday Studies: Papers Read at the Novocentenary Conference of the Royal Historical Society and the Institute of British Geographers, Winchester, 1986 (Woodbridge, 1987) —— (ed.), Colonial England, 1066–1215 (London and Rio Grande, 1997) Home, Malasree, The Peterborough Version of the Anglo-Saxon Chronicle: Rewriting Post-Conquest History (Woodbridge, 2015) Hoofnagle, Wendy Marie, ‘Charlemagne’s Legacy and Anglo-Norman Imperium in Henry of Huntingdon’s Historia Anglorum’, in Matthew Gabriele and Jane Stuckey (ed.), The Legend of Charlemagne in the Middle Ages: Power, Faith, and Crusade (New York, 2008), 77–94 Hooper, Jennie, ‘The “Rows of the Battle-Swan”: The Aftermath of Battle in Anglo-Saxon Art’, in Matthew Strickland (ed.), Armies, Chivalry and Warfare in Medieval Britain and France: Proceedings of the 1995 Harlaxton Symposium (Stamford, 1998), 82–99 Hooper, Nicholas, ‘Edgar the Ætheling: Anglo-Saxon Prince, Rebel and Crusader’, ASE, xiv (1985), 197–214

552

bibliography

—— ‘Edgar Ætheling (b.1052?, d. in or after 1125)’, ODNB Hosking, Geoffrey, Trust: A History (Oxford, 2014) Housley, Norman, ‘Crusades against Christians: Their Origins and Early Development, c.1000–1216’, in Thomas F. Madden (ed.), The Crusades: The Essential Readings (Oxford and Malden, MA, 2002), 69–97 Howard, Ian, ‘Harold II, a Throne-Worthy King’, in Owen-Crocker (ed.), King Harold II and the Bayeux Tapestry, 35–52 Hudson, Benjamin T., ‘The Destruction of Gruffudd ap Llywelyn’, Welsh History Review, xv (1991), 331–50 (reprinted in Benjamin T. Hudson, Irish Sea Studies, 900–1200 [Dublin, 2006], 81–99) —— ‘William the Conqueror and Ireland’, Irish Historical Studies, xxix (1994–5), 145–58 Hudson, John G. H., ‘Faide, vengeance et violence en Angleterre (ca 900–1200)’, in Dominique Barthélemy, François Bougard, and Régine Le Jan (eds.), La vengeance, 400–1200 (Rome, 2006), 341–82 —— ‘Feud, Vengeance and Violence in England from the Tenth to the Twelfth Centuries’, in Tuten and Billado (eds.) Feud, Violence and Practice, 29–53 —— ‘The Fate of Earl Waltheof and the Idea of Personal Law in England after 1066’, in Crouch and Thompson (eds.), Normandy and its Neighbours, 223–35 —— The Oxford History of the Laws of England, volume II: 871–1216 (Oxford, 2012) Huneycutt, Lois L., Matilda of Scotland: A Study in Medieval Queenship (Woodbridge, 2003) Huntington, Joanna, ‘The Taming of the Laity: Writing Waltheof and Rebellion in the Twelfth Century’, ANS, xxxii (2010), 79–95 —— ‘ “The Quality of his virtus proved him a perfect man”: Hereward “the Wake” and the Representation of Lay Masculinity’, in P. H. Cullum and Katherine J. Lewis (eds.), Religious Men and Masculine Identity in the Middle Ages (Woodbridge, 2013), 77–93 Hyams, Paul R., ‘ “No Register of Title”: The Domesday Inquest and Land Adjudication’, ANS, ix (1987), 127–41 —— Rancor and Reconciliation in Medieval England (Ithaca, NY, and London, 2003) —— ‘Afterword. Neither Unnatural nor Wholly Negative: The Future of Medieval Vengeance’, in Susanna A. Throop and Paul R. Hyams (eds.), Vengeance in the Middle Ages: Emotion, Religion and Feud (Farnham and Burlington, VT, 2010), 203–20 Impey, Edward, ‘London’s Early Castles and the Context of their Creation’, in Edward Impey (ed.), The White Tower (New Haven, CT, and London, 2008), 13–26 —— ‘The Great Hall at Caen and its Affinities with Westminster’, in Warwick Rodwell and Tim Tatton-Brown (eds.), Westminster: II. The Art, Architecture and Archaeology of the Royal Palace, British Archaeological Association Conference Transactions, xxxix, Part II (Leeds, 2015), 72–88 —— ‘The Great Hall of the Dukes of Normandy in the Castle at Caen’, in Davies, Riley, Levesque, and Lapiche (eds.), Castles and the Anglo-Norman World, 101–32 Jégou, Laurent, L’évêque, juge de paix. L’autorité épiscopale et le règlement des conflits (VIIIe–XIe siècle) (Turnhout, 2011) Jennings, J. C., ‘The Writings of Prior Dominic of Evesham’, EHR, lxxvii (1962), 298–304 Jessee, W. Scott, ‘The Angevin Civil War and the Norman Conquest of 1066’, HSJ, iii (1991), 101–10 —— Robert the Burgundian and the Counts of Anjou, ca.1025–1098 (Washington, DC, 2000) Johns, Susan M., ‘Women and Power in the Roman de Rou of Wace’, ANS, xxxvi (2014), 117–34 Johnson, Ewan, ‘The Process of Norman Exile into Southern Italy’, in Napran and van Houts (eds.), Exile in the Middle Ages, 29–38 Jones, Graham, ‘A “Common of Hunting”? Forests, Lordship and Community before and after the Conquest’, in John Langton and Graham Jones (eds.), Forests and Chases of England and Wales c.1000–1500 (Oxford, 2010), 36–67 Jorgensen, Alice (ed.), Reading the Anglo-Saxon Chronicle: Language, Literature and History (Turnhout, 2010)

bibliography

553

Jørgensen, Dolly, ‘The Roots of the English Royal Forest’, ANS, xxxii (2010), 114–28 Kapelle, W. E., The Norman Conquest of the North: The Region and its Transformation, 1000–1135 (London, 1979) Karras, Ruth Mazo, ‘The History of Marriage and the Myth of Friedelehe’, EME, xiv (2006), 119–51 —— Sexuality in Medieval Europe: Doing unto Others, 2nd edn. (London and New York, 2012) —— Unmarriages: Women, Men and Sexual Unions in the Middle Ages (Philadelphia, PA, 2012) Kay, Richard, The Council of Bourges, 1225: A Documentary History (Aldershot and Burlington, VT, 2002) Keats-Rohan, Katharine S. B., ‘William I and the Breton Contingent in the Non-Norman Conquest, 1060–1087’, ANS, xiii (1991), 157–72 —— ‘The Prosopography of Post-Conquest England: Four Case Studies’, Medieval Prosopography, xiv (1993), 1–52 —— (ed.), Medieval Prosopography: The Roots and Branches of Power in France and England from the Tenth to the Twelfth Centuries (Woodbridge, 1997) —— ‘Domesday Book and the Malets: Patrimony and the Private Histories of Public Lives’, Nottingham Medieval Studies, xli (1997), 13–56 —— Domesday People: A Prosopography of Persons Occurring in English Documents, 1066–1166. 1. Domesday Book (Woodbridge, 1999) —— ‘Through the Eye of the Needle: Stigand, the Bayeux Tapestry and the Beginnings of the Historia Anglorum’, in Roffe (ed.), The English and their Legacy, 159–74 —— ‘L’expédition de Guillaume, duc de Normandie, et du comte Harold en Bretagne (1064): le témoignage de la tapisserie de Bayeux et des chroniqueurs anglo-normands’, MSHAB, xl (2013), 203–24 —— ‘A Question of Identity: Domesday Prosopography and the Formation of the Honour of Richmond’, in Roffe and Keats-Rohan (eds.), Domesday Now, 169–96 Keene, Derek, ‘From Conquest to Capital: St Paul’s c.1100–1300’, in Derek Keene, Arthur Burns, and Andrew Saint (eds.), St Paul’s: The Cathedral Church of London, 604–2004 (New Haven, CT, and London, 2004), 17–32 Keene, Laurence (ed.), ‘Almost the Richest City’, Bristol in the Middle Ages, British Archaeological Association Conference Transactions, xix (London, 1997) Kempshall, Matthew, Rhetoric and the Writing of History, 400–1500 (Manchester and New York, 2011) Ker, N. R., Catalogue of Manuscripts Containing Anglo-Saxon (Oxford, 1957) Keynes, Simon D., ‘Regenbald the Chancellor (sic)’, ANS, x (1988), 185–222 —— ‘The Æthelings in Normandy’, ANS, xiii (1991), 173–205 —— ‘Cnut’s Earls’, in Alexander R. Rumble (ed.), The Reign of Cnut, King of England, Denmark and Norway, Studies in the Early History of Britain (London and New York, 1994), 43–88 —— ‘Giso, Bishop of Wells’, ANS, xix (1997), 203–71, at 242–3 —— ‘Ely Abbey 672–1109’, in Peter Meadows and Nigel Ramsay (eds.), A History of Ely Cathedral (Woodbridge, 2003), 3–58 ——, and Rosalind Love, ‘Earl Godwine’s Ship’, ASE, xxxviii (2009), 185–223 King, Edmund, ‘Henry of Winchester: The Bishop, the City, and the Wider World’, ANS, xxxvii (2015), 1–23 King, Vanessa, ‘Ealdred, Archbishop of York: The Worcester Years’, ANS, xviii (1996), 123–37 Knowles, Dom. David, The Monastic Order in England, 2nd edn. (Cambridge, 1966) Körntgen, Ludger, and Dominik Waßenhoven (eds.), Patterns of Episcopal Power: Bishops in Tenth- and Eleventh-Century Western Europe (Bayreuth, 2011) Kosto, Adam J., Hostages in the Middle Ages (Oxford, 2012) Lack, Katherine, Conqueror’s Son: Duke Robert Curthose, Thwarted King (Stroud, 2007) —— ‘The De Obitu Willelmi: Propaganda for the Anglo-Norman Succession, 1087–88?’, EHR, cxxiii (2008), 1,417–56

554

bibliography

—— ‘Robert Curthose: Ineffectual Duke or Victim of Spin?’, HSJ, xx (2009, for 2008), 110–40 Laffont, Pierre-Yves, and Lucie Jeanneret, ‘Le site de la Motte à Sainte-Suzanne (Mayenne) est-il une fortification de siège de Guillaume le Conquérant?’, in Signes de pouvoir laic et religieux au Moyen Age. Les fortifications laïques et religieuses: actes du Xe colloque biennal de Pommiersen-Forez, 12 juin 2015, Association culturelle de Pommiers (2016), 77–94 Lainé, Stéphane, ‘La bataille de Varaville (1057): examen critique des récits médiévaux’, BSAN, lxvii (2009, for 2008), 257–88 Lalou, Elisabeth, Bruno Lepeuple and Jean-Louis Roch (eds.), Des châteaux et des sources: Archéologie et histoire dans la Normandie médiévale. Mélanges en l’honneur d’Anne-Marie Flambard Héricher (Rouen, 2008) Larson, Laurence M., ‘The Efforts of the Danish Kings to Recover the English Crown after the Death of Harthacnut’, Annual Report of the American Historical Association, for the Year 1910 (Washington, DC, 1912), 71–81 Latouche, Robert, Histoire du comté du Maine pendant le Xe et le XIe siècle (Paris, 1910) Laurence, Jean-Marie, Caen aux XIe et XIIe siècles: espace urbain, pouvoirs et société (Cormeilles-leRoyal, 2000) —— ‘Des maisons au château: le château de Caen, une paroisse urbaine’, in Lalou, Lepeuple and Roch (eds.), Des châteaux et des sources, 377–88 Lawson, M. K., ‘The Collection of Danegeld and Heregeld in the Reigns of Aethelred II and Cnut’, EHR, xcix (1984), 721–38 —— ‘ “Those Stories Look True”: Levels of Taxation in the Reigns of Aethelred II and Cnut’, EHR, civ (1989), 385–406 —— Cnut: The Danes in England in the Early Eleventh Century (London and New York, 1993) —— The Battle of Hastings 1066 (3rd pdf edition, 2016) (https://archive.org/details/ LawsonBattleofHastings3rdedn) —— ‘Cnut [Canute] (d.1035), King of England, of Denmark, and Norway’, ODNB —— ‘Harthacnut [Hardecanute] (c.1018–1042), King of England and of Denmark’, ODNB Le Cacheux, Marie-Josephe, ‘Histoire de l’abbaye de Saint-Amand de Rouen des origines à la fin du XVIe siècle’, BSAN, xliv (1937, for 1936), 5–289 Léchaudé d’Anisy, A. L., Recueil des sceaux normands et anglo-normands, précédé de l’extrait du cartulaire des chartes, diplômes et autres actes qui existent encore dans les archives du Calvados ou d’autres fonds (Caen and London, 1834) Le Jan, Régine, Famille et pouvoir dans le monde franc (VIIe–Xe siècle): étude d’anthropologie sociale (Paris, 1995) —— ‘Frankish Giving of Arms and Rituals of Power: Continuity and Change in the Carolingian Period’, in Frans Theuws and Janet L. Nelson (eds.), Rituals of Power from Late Antiquity to the Early Middle Ages (Leiden, Boston, MA, and Cologne, 2000), 281–309 Le Maho, Jacques, ‘L’apparition des seigneuries châtelaines dans le Grand-Caux à l’époque ducale’, Archéologie médiévale, vi (1976), 5–148 —— ‘Fortifications de siège et “contre-châteaux en Normandie (XIe–XIIe s.)” ’, Château Gaillard, xix (2000), 181–9 —— ‘Nouveaux aperçus sur les origines de la ville de Caen’, BSAN, lxviii (2010, for 2009), 171– 99 Le Patourel, John, ‘Geoffrey of Montbray, Bishop of Coutances, 1049–1093’, EHR, lix (1944), 129–61 —— ‘The Norman Colonization of Britain’, in I Normanni e la loro espansione in Europa sullalto medioevo, Settimane di Studi del Centro Italiano sull’Alto Medioevo, xvi (Spoleto, 1968), 409–38 Leech, Roger H., ‘The Medieval Defences of Bristol Revisited’, in Laurence Keene (ed.), ‘Almost the Richest City’, Bristol in the Middle Ages, British Archaeological Association Conference Transactions, xix (London, 1997), 18–30 Legros, Huguette, ‘Naissance d’un héros: de la bâtardise à la légende dans la Chronique de Benoît de Sainte-Maure’, in Huguette Legros (ed.), Guillaume le Conquérant face aux défis (Orléans, 2008), 135–47

bibliography

555

Lemagnen, Sylvette (ed.), La Tapisserie de Bayeux. Une chronique des temps Vikings? Actes du colloque international de Bayeux, 29 et 30 mars 2007 (Bonsecours, 2009) Lemarignier, Jean-François, ‘Paix et réforme monastique en Flandre et en Normandie autour de l’année 1023’, in Robert Aubreton et al. (eds.), Droit privé et institutions régionales: Etudes historiques offertes à Jean Yver (Paris, 1976), 443–68 ——, Jean-François Lemarignier : Recueil d’articles rassemblés par ses disciples (Rouen, 1995) Lemesle, Bruno, ‘Le discours de l’Eglise aux temps grégoriens: évêques et laïcs dans le Maine aux XIe–XIIe siècles d’après les Actus Pontificum’, Annales de Bretagne et des pays de l’Ouest, cii (1995), 17–32 —— La société aristocratique dans le Haut-Maine (XIe–XIIe siècles) (Rennes, 1999) Lemoine-Descourtieux, Astrid, La frontière normande de l’Avre de la fondation de la Normandie à sa réunion au domaine royal (911–1204) (Rouen and Le Havre, 2011) Levesque, Jean-Marie, and Leroux, Pascal, ‘Le château de Caen’, in Davies, Riley, Levesque, and Lapiche (eds.), Castles and the Anglo-Norman World, 89–100 Lewis, Andrew W., Royal Succession in Capetian France (Cambridge, MA, and London, 1981) Lewis, C. P., ‘The Norman Settlement of Herefordshire under William I’, ANS, vii (1985), 195–213 —— ‘The Early Earls of Norman England’, ANS, xiii (1991), 207–23 —— ‘The Formation of the Honor of Chester, 1066–1100’, in A. T. Thacker (ed.), The Earldom of Chester and its Charters: A Tribute to Geoffrey Barraclough, Journal of the Chester Archaeological Society, lxxi (1991), 37–68 —— ‘The Domesday Jurors’, HSJ, v (1993), 17–44 —— ‘The French in England before the Norman Conquest’, ANS, xvii (1995), 123–44 —— ‘Gruffudd ap Cynan and the Normans’, in K. L. Maund (ed.), Gruffudd ap Cynan: A Collaborative Biography (Woodbridge, 1996), 61–77 —— ‘Joining the Dots: A Methodology for Identifying the English in Domesday Book’, in Keats-Rohan (ed.), Medieval Prosopography, 69–87 —— ‘Tosny, Ralph de [Ralph de Conches] (d.1102)’, ODNB —— ‘Waltheof, Earl of Northumbria (c.1050–1076)’, ODNB —— ‘The Invention of the Manor in Norman England’, ANS, xxxiv (2012), 123–50 Lewis, Michael J., Gale R. Owen-Crocker, and Dan Terkla (eds.), The Bayeux Tapestry: New Approaches. Proceedings of a Conference at the British Museum (Oxford, 2011) Leyser, Conrad, ‘Introduction: England and the Continent’, in David Rollason, Conrad Leyser, and Hannah Williams (eds.), England and the Continent in the Tenth Century: Studies in Honour of Wilhelm Levison (1876–1947) (Turnhout, 2010), 1–13 Leyser, Henrietta, ‘Walcher, Earl of Northumbria (d.1080)’, ODNB Leyser, Karl, ‘England and Empire in the Twelfth Century’, TRHS, 5th ser., x (1960), 61–83 Licence, Tom, ‘History and Hagiography in the Late Eleventh Century: The Life and Works of Herman the Archdeacon, Monk of Bury St Edmunds’, EHR, cxxiv (2009), 516–44 —— Hermits and Recluses in English Society, 950–1200 (Oxford, 2011) —— ‘Robert of Jumièges, Archbishop in Exile (1052–5)’, ASE, xlii (2013), 311–29 —— (ed.), Bury St Edmunds and the Norman Conquest (Woodbridge, 2014) Liddiard, Robert, ‘Landscapes of Lordship’: Norman Castles and the Countryside in Medieval Norfolk, 1066–1200, British Archaeological Reports, British Series, no. 309 (Oxford, 2000) Liebermann, Felix, ‘Lanfranc and the Antipope’, EHR, xvi (1901), 328–32 Livingstone, Amy, Out of Love for my Kin: Aristocratic Family Life in the Lands of the Loire, 1000– 1200 (Ithaca, NY, and London, 2010) Lloyd, J. E., rev. David E. Thornton, ‘Rhys ap Tewdwr (d.1093)’, ODNB Lobrichon, Guy, ‘The Early Schools, c.900–1100’, in Richard Marsden and E. Ann Matter (eds.), The New Cambridge History of the Bible, vol. 2 from 600 to 1450 (Cambridge, 2012), 536–54 Lomas, Richard, ‘The Impact of Border Warfare: The Scots and South Tweedside, c.1290–c.1520’, Scottish Historical Review, lxxv (1996), 143–67

556

bibliography

LoPrete, Kimberley A., ‘The Anglo-Norman Card of Adela of Blois’, Albion, xxii (1990), 569–89 —— Adela of Blois: Countess and Lord (c.1067–1137) (Dublin, 2007) Lot, F., Études critiques sur l’abbaye de Saint-Wandrille (Paris, 1913) Loud, G. A., The Age of Robert Guiscard (Harlow, 2000) —— ‘The Kingdom of Sicily and the Kingdom of England, 1066–1266’, History, lxxxviii (2003), 540–67 Louise, Gérard, La seigneurie de Bellême (Xe–XIIe siècles): dévolution des pouvoirs territoriaux et construction d’une seigneurie de frontières aux confins de la Normandie et du Maine à la charnière de l’an mil (Flers, 1992–3) Loyn, H. R., ‘A General Introduction to Domesday Book’, in Ann Williams and R. W. H. Erskine (eds.), Domesday Book Studies (London, 1987), 1–21 —— ‘The Beyond of Domesday Book’, in Holt (ed.), Domesday Studies, 1–13 Lyon, Jonathan R., ‘Fathers and Sons: Preparing Noble Youths to be Lords in TwelfthCentury Germany’, JMH, xxxiv (2008), 291–310 McClain, Aleksandra, ‘Rewriting the Narrative: Regional Dimensions of the Norman Conquest’, in Dawn Hadley and Christopher Dyer (eds.), Transformations and Continuities in the Eleventh Century: The Archaeology of the Norman Conquest (forthcoming) McGlynn, Sean, By Sword and Fire: Cruelty and Atrocity in Medieval Warfare (London, 2008) McKitterick, Rosamond, Charlemagne: The Formation of a European Identity (Cambridge, 2008) McLaughlin, Megan, Sex, Gender, and Episcopal Authority in an Age of Reform, 1000–1122 (Cambridge, 2010) McNeill, Tom, ‘Davison versus Brown, quarante ans après’, in Anne-Marie Flambard Héricher and Jacques Le Maho (eds.), Château, ville et pouvoir au Moyen Age, Publications du CRAHM (Caen, 2012), 41–50 Maddicott, J. R., ‘Edward the Confessor’s Return to England in 1041’, EHR, cxix (2004), 650–66 —— ‘Responses to the Threat of Invasion, 1085’, EHR, cxxii (2007), 986–97 —— The Origins of the English Parliament, 924–1327 (Oxford, 2010) Maillefer, Jean-Marie, ‘Une famille aristocratique aux confins de la Normandie: les Géré au XIe siècle’, in Lucien Musset, Jean-Michel Bouvris, and Jean-Marie Maillefer (eds.), Autour du pouvoir ducal (Caen, 1985), 175–206 Marchegay, Paul, ‘Chartes angevines des XIe et XIIe siècles’, Bibliothèque de l’Ecole des Chartes, xxxvi (1875), 381–441 Marritt, Stephen, ‘Coincidences of Names, Anglo-Scottish Connections and Anglo-Saxon Society in the Late Eleventh Century’, Scottish Historical Review, lxxxiii (2004), 150–70 Marshall, Pamela, and Trevor Foulds, ‘The Royal Castle’, in Beckett (ed.), A Centenary History of Nottingham, 43–55 Marten, Lucy, ‘The Rebellion of 1075 and its Impact in East Anglia’, in Christopher Harper-Bill (ed.), Medieval East Anglia (Woodbridge, 2005), 168–82 Martindale, Jane, ‘Aimeri of Thouars and the Poitevin connection’, ANS, vii (1985), 224–45 —— ‘Monasteries and Castles: The Priories of St-Florent de Saumur in England after 1066’, in Carola Hicks (ed.), England in the Eleventh Century: Proceedings of the 1990 Harlaxton Symposium (Stamford, 1992), 135–56 Mason, Emma, Westminster Abbey and its People, c.1050–c.1216 (Woodbridge, 1996) —— The House of Godwine: The History of a Dynasty (London and New York, 2004) Mason, J. F. A., William the First and the Sussex Rapes (Hastings and Bexhill Branch of the Historical Association, 1966) Matthews, J. S., ‘William the Conqueror’s Campaign in Cheshire in 1069–70: Ravaging and Resistance in the North-West’, Northern History, xl (2003), 53–70 Maund, K. L., ‘The Welsh Alliances of Earl Ælfgar of Mercia and his Family in the Mid-Eleventh Century’, ANS, xi (1989), 181–90 —— The Welsh Kings (Stroud, 2002) Maurice, Philippe, Guillaume le Conquérant (Paris, 2002)

bibliography

557

Mayr-Harting, Henry, ‘Charlemagne’s Religion’, in Peter Godman, Jörg Jarnut, and Peter Jonanek (eds.), Am Vorabend der Kaiser Krönung (Berlin, 2002), 113–24 —— ‘Alcuin, Charlemagne and the Problem of Sanctions’, in Stephen Baxter, Nelson Karkov, and David Pelteret (eds.), Early Medieval Studies in Memory of Patrick Wormald (Farnham, 2009), 207–18 Mazel, Florian, ‘Qu’est-ce qu’une bataille decisive? Jugement de Dieu et legitimation dans les premiers récits de la bataille d’Hastings (v.1066–1087)’, in Arianne Boltanski, Yann Lagadec, and Franck Mercier (eds.), La Bataille. Du fait d’armes au combat idéologique, XIe– XIXe siècle (Rennes, 2015), 15–30 Merlet, R., ‘Du lieu où mourut Henri Ier roi de France, le 4 août 1060’, Moyen Age, 2nd ser., vii (1903), 203–9 Mew, Karin, ‘The Dynamics of Lordship and Landscape as Revealed in the Domesday Study of the Nova Foresta’, ANS, xxiii (2001), 155–66 Mills, Peter, ‘The Battle of London 1066’, London Archaeologist, viii, no. 3 (1996), 59–62 Moesgaard, Jens Christian, ‘Renovatio Monetae et la chronologie des monnaies de Richard Ier, duc de Normandie 942–945–996’, Bulletin de la Société française de numismatique, lxvi (2011), 125–33 —— ‘Saints, Dukes and Bishops: Coinage in Ducal Normandy, c.930–c.1150’, in Giles E. M. Gaspar and Svein H. Gullbeckk (eds.), Money and the Church in Medieval Europe, 1000– 1200 (Farnham and Burlington, VT, 2015), 196–207 Molyneaux, George, ‘Why Were Some Tenth-Century Kings Presented as Rulers of Britain?’, TRHS, 6th ser., xxi (2011), 59–91 —— The Formation of the English Kingdom in the Tenth Century (Oxford, 2015) Moore, John S., ‘Anglo-Norman Garrisons’, ANS, xxii (2000), 205–59 Moore, Keith, ‘Ælfgyva and the Cleric of the Bayeux Tapestry’, Foundations, vii (2015), 79–117 Morillo, Stephen, ‘Hastings: An Unusual Battle’, HSJ, ii (1990), 95–103 —— (ed.), The Battle of Hastings: Sources and Interpretations (Woodbridge, 1996) Morin, Stéphane, Trégor, Goëlo, Penthièvre: Le pouvoir des Comtes de Bretagne du XIe au XIIIe siècle (Rennes, 2010) Morris, Colin, ‘William I and the Church Courts’, EHR, lxxxii (1967), 449–63 Morris, Ian, War. What Is It Good For? The Role of Conflict in Civilisation, from Primates to Robots (London, 2015) Mortimer, Richard, ‘Anglo-Norman Lay Charters, 1066–c.1100: A Diplomatic Approach’, ANS, xxv (2003), 153–75 —— ‘Clare, Richard de [Richard fitz Gilbert] (1030x35–1087x90)’, ODNB —— Edward the Confessor: The Man and the Legend (Woodbridge, 2009) Morton, Catherine, ‘Pope Alexander II and the Norman Conquest’, Latomus: Revue d’études latines, xxxiv (1975), 362–82 Mouchard, Jimmy, ‘De la voie navigable aux sites portuaires en basse vallée de Seine: maîtrise et gestion des accès (Antiquité-époque moderne)’, in Lalou, Lepeuple, and Roch (eds.), Des châteaux et des sources, 103–27 —— ‘Le Vieux Château à Brionne’, in Anne-Marie Flambard Héricher (ed.), Projet collectif de recherches sur les fortifications de terre en Haute-Normandie 2007, 63–96 Musset, Lucien, ‘Actes inédits du XIe siècle: Les plus anciennes chartes du prieuré de SaintGabriel (Calvados)’, BSAN, lii (1952–4), 117–41 —— ‘Fragments inédits de chroniques de Fécamp’, BSAN, liv (1957–8), in ‘Notules fécampoises’, BSAN, liv (1959, for 1957–8), 584–98 —— ‘A-t-il existé en Normandie au XIe siècle une aristocratie d’argent?’, AN, ix (1959), 285–99 —— ‘Actes inédits du XIe siècle: V. Autour des origines de Saint-Etienne de Fontenay’, BSAN, lvi (1963, for 1961–2), 11–41 —— ‘La Seine normande et le commerce maritime du IIIe au XIe siècle’, Revue de la société savante de la Haute-Normandie, 53 (1970, for 1969), 3–14

558

bibliography

—— ‘Gouvernés et gouvernants dans le monde scandinave et dans le monde normand (XIe–XIIe siècle)’, in Nordica et Normannica. Recueil d’études sur la Scandinavie ancienne et médiévale, les expéditions des Vikings et la fondation de la Normandie (Paris, 1976), 415–37 —— ‘Aux origines d’une classe dirigeante: les Tosny, grands barons normands du Xe au XIIIe siècle’, Francia, v (1977), 45–80 —— ‘Les contacts entre l’église normande et l’église d’Angleterre de 911 à 1066’, in Foreville (ed.), Les mutations socio-culturelles au tournant des XIe–XIIe siècles, 67–84 —— ‘Autour des modalités juridiques de l’expansion normande au XIe siècle: le droit d’exil’, in Musset, Bouvris, and Maillefer (eds.), Autour du pouvoir ducal, 45–59 —— ‘Les sépultures des souverains normands: un aspect de l’idéologie du pouvoir’, in Musset, Bouvris, and Maillefer (eds.), Autour du pouvoir ducal, 19–44 —— ‘Le mécenat des princes normands au XIe siècle’, in X. Barret I Altet (ed.), Artistes, artisans et production artistique au Moyen Age, 3 vols. (Paris, 1986–90), ii, 121–33 ——, Jean-Michel Bouvris, and Jean-Marie Maillefer (eds.), Autour du pouvoir ducal normand Xe–XIIe siècles, Cahiers des Annales de Normandie, vol. 17 (Caen, 1985) Napran, Laura, and Elisabeth van Houts (eds.), Exile in the Middle Ages (Turnhout, 2004) Navel, Henri, ‘L’enquête de 1133 sur les fiefs de l’évêché de Bayeux’, BSAN, xlii (1934), 5–80 Nelson, Janet L., ‘The Rites of the Conqueror’, ANS, iv (1982), 117–32, 210–23 (reprinted in Janet L. Nelson, Politics and Rituals in Early Medieval Europe [London, 1986], 375–401) —— ‘Parents, Children and the Church’, in Diana Wood (ed.), The Church and Childhood, Studies in Church History, xxxi (1994), 81–114 —— ‘Monks, Secular Men and Masculinity, c.900’, in Hadley (ed.), Masculinity in Medieval Europe, 121–42 —— ‘England and the Continent in the Ninth Century’, TRHS, 5th ser., xiii (2003), 1–28 —— Charlemagne and the Paradoxes of Power, The Reuter Lecture 2005 (Southampton, 2006) —— ‘Did Charlemagne Have a Private Life?’ (2006), in Bates, Crick, and Hamilton (eds.), Writing Medieval Biography, 15–28 —— ‘Henry Loyn and the Context of Anglo-Saxon England’, The Henry Loyn Memorial Lecture for 2006, HSJ, xix (2007), 154–70 —— ‘Review-Article: Church Properties and the Propertied Church: Donors, the Clergy and the Church in Medieval Western Europe from the Fourth Century to the Twelfth’, EHR, cxxiv (2009), 355–74 ——, Susan Reynolds, and Susan M. Johns (eds.), Gender and Historiography: Studies in the Earlier Middle Ages in Honour of Pauline Stafford (London, 2012) Neveux, François, La Normandie des ducs aux rois Xe–XIIe siècle (Rennes, 1998) —— ‘Les reliques de la cathédrale de Bayeux’, in Pierre Bouet and François Neveux (eds.), Les Saints dans la Normandie médiévale (Caen, 2000), 109–33 —— ‘L’expédition de Guillaume le Bâtard en Bretagne (vers 1064)’, in Joëlle Quaghebeur and Sylvain Soleil (eds.), Le pouvoir et la foi au Moyen Age en Bretagne et dans l’Europe de l’Ouest: Mélanges en mémoire du professeur Hubert Guillotel (Rennes, 2010), 619–37 ——, and Claire Ruelle, Guillaume le Conquérant. Le bâtard qui s’empara de l’Angleterre (Rennes, 2013) Newman, P. R., ‘The Yorkshire Domesday Clamores and the “Lost Fee” of William Malet’, ANS, xxii (2000), 261–77 Nieus, Jean-François, ‘Early Aristocratic Seals: An Anglo-Norman Success Story’, ANS, xxxviii (2016), 97–123 ——, and Steven Vanderputten, ‘Diplôme princier, matrice de faux, acte modèle. Le règlement d’avouerie du comte Baudouin V pour Saint-Bertin (1042) et ses réappropriations sous l’abbatiat réformateur de Lambert (1095–1123)’, Medieval Low Countries, i (2014), 1–60 Nightingale, Pamela, ‘The Evolution of Weight-Standards and the Creation of New Monetary and Commercial Links in Northern Europe from the Tenth to the Twelfth Century’, Economic History Review, 2nd. ser., xxxviii (1985), 192–209, at 200–1 (reprinted in Pamela Nightingale, Trade, Money, and Power in Medieval England [Aldershot and Burlington, VT, 2007], chapter 4)

bibliography

559

Norton, Christopher, Archbishop Thomas of Bayeux and the Norman Cathedral at York, Borthwick Paper, no. 100 (York, 2001) O’Brien, Bruce R., ‘From morðor to murdrum: The Pre-Conquest Origin and Norman Revival of the Murder Fine’, Speculum, lxxi (1996), 321–57 —— God’s Peace and King’s Peace: The Laws of Edward the Confessor (Philadelphia, PA, 1999) Offergeld, Thilo, ‘Reges Pueri’: Das Konigtüm Minderjähriger im frühen Mittelalter, MGH Schriften, vol. 50 (Hanover, 2001) Oheix, André, ‘La date de la mort d’Alain III, duc de Bretagne’, Bulletin de la Société d’émulation des Côtes-du-Nord, li (1913), 93–100 Oksanen, Elijas, Flanders and the Anglo-Norman World 1066–1216 (Cambridge, 2012) Oleson, T. J., ‘Edward the Confessor’s Promise of the Throne to Duke William of Normandy’, EHR, lxxii (1957), 221–8 Oppermann, O., Die älteren Urkunden des Klosters Blandinium und die Anfänge der Stadt Gent (Utrecht, 1928) Orme, Nicholas, Medieval Children (New Haven, CT, and London, 2011) Ormrod, W. Mark, Edward III (New Haven, CT, and London, 2011) Ortenberg West-Harling, Veronica, ‘Un prince évêque anglo-saxon au XIe siècle: l’archevêque d’York Ealdred’, in Alban Gautier and Sébastien Rossignol (eds.), De la mer du Nord à la mer Baltique: Identités, contacts et communications au Moyen Age (Lille, 2012), 143–57 Otter, Monika, ‘1066: The Moment of Transition in Two Narratives of the Norman Conquest’, Speculum, lxxiv (1999), 565–86 Owen-Crocker, Gale R. (ed.), King Harold II and the Bayeux Tapestry (Woodbridge, 2005) Palliser, David M., ‘Domesday Book and the “Harrying of the North” ’, Northern History, xxix (1993), 1–23 —— Medieval York 600–1540 (Oxford, 2014) Palmer, John J. N., ‘The Conqueror’s Footprints in Domesday Book’, in Ayton and Price (eds.), The Medieval Military Revolution, 23–44 —— ‘War and Domesday Waste’, in Strickland (ed.), Armies, Chivalry and Warfare in Medieval Britain and France, 256–75 —— ‘A Digital Latin Domesday’, in Roffe and Keats-Rohan, Domesday Now, 61–80 Parisse, Michel, ‘Les pancartes. Etude d’un type d’acte diplomatique’, in Parisse, Pégeot, and Tock (eds.), Pancartes monastiques, 11–62 ——, P. Pégeot, and B.-M. Tock (eds.), Pancartes monastiques des XIe et XIIe siècles (Turnhout, 1998) Pastan, Elizabeth Carson, and Stephen D. White, ‘Problematizing Patronage: Odo of Bayeux and the Bayeux Tapestry’, in Foys, Overbey, and Terkla (eds.), The Bayeux Tapestry, 1–24 —— with Kate Gilbert, The Bayeux Tapestry and its Contents: A Reassessment (Woodbridge, 2014) Pelteret, David, Catalogue of English Post-Conquest Vernacular Documents (Woodbridge, 1990) —— Slavery in Early Medieval England from the Reign of Alfred until the Twelfth Century (Woodbridge, 1995) Percival, John, ‘The Precursors of Domesday: Roman and Carolingian Land Registers’, in P. H. Sawyer (ed.), Domesday Book: A Reassessment (London, 1985), 5–27 Philpott, Mark, ‘Some Interactions between the English and Irish Churches’, ANS, xx (1998), 187–204 Pichot, Daniel, Le Bas-Maine du Xe au XIIIe siècle: étude d’une société (Laval, 1995) Pinker, Steven, The Better Angels of Our Nature: A History of Violence and Humanity (London, 2011) Pohl, Benjamin, ‘Translatio imperii Constantini ad Normannos: Constantine the Great as a Possible Model for the Depiction of Rollo in Dudo of St Quentin’s Historia Normannorum’, Millennium: Yearbook on the Culture and History of the First Millennium C.E., ix (2012), 299–341

560

bibliography

—— Dudo of Saint-Quentin’s ‘Historia Normannorum’: Tradition, Innovation and Memory (Woodbridge and Rochester, NY, 2015) Pontal, Odette, Les conciles de la France capétienne jusqu’en 1215 (Paris, 1995) Popescu, Elizabeth, ‘Norwich Castle’, in Davies, Riley, Levesque, and Lapiche (eds.), Castles and the Anglo-Norman World, 1–29 Porter, Roy, ‘ “On the very spot”: In Defence of Battle’, English Heritage Historical Review, vii (2012), 5–17 Potter, Julie A., ‘Monastic Freedom vs. Episcopal and Aristocratic Power in the Twelfth Century: Context and Analysis of the De libertate Beccensis’, in Bijsterveld, Teunis, and Wareham (eds.), Negotiating Secular and Ecclesiastical Power, 73–85 Potts, Cassandra, ‘Normandy or Brittany? A Conflict of Interests at Mont Saint-Michel (996–1035)’, ANS, xii (1990), 135–56 —— ‘The Earliest Norman Counts Revisited: The Lords of Mortain’, HSJ, iv (1992), 23–35 —— Monastic Revival and Regional Identity in Early Normandy (Woodbridge, 1997) Pratt, David, ‘Demesne Exemption from Royal Taxation in Anglo-Saxon and AngloNorman England’, EHR, cxxiii (2013), 1–34 Prentout, Henri, ‘Études sur quelques points de l’histoire de Guillaume le Conquérant. II. Le mariage de Guillaume’, Mémoires de l’Académie de Caen, vi (1931), 29–56 Prestwich, J. O., ‘Military Intelligence under the Norman and Angevin Kings’, in Garnett and Hudson (eds.), Law and Government in Medieval England and Normandy, 1–30 —— ‘The Military Household of the Norman Kings’, EHR, xcvi (1981), 1–37 (reprinted in Strickland [ed.], Anglo-Norman Warfare, 93–127) —— The Place of War in English History, 1066–1214, ed. Michael Prestwich (Woodbridge, 2004) Pryce, Huw, ‘Gruffudd ap Cynan (1054/5–1137)’, ODNB Quaghebeur, Joëlle, ‘Havoise, Constance et Mathilde, princesses de Normandie et duchesses de Bretagne’, in Quaghebeur and Merdrignac (eds.), Bretons et Normand au Moyen Age, 145–63 ——, and Bernard Merdrignac (eds.), Bretons et Normands au Moyen Age: Rivalités, malentendus, convergences (Rennes, 2008) Rawcliffe, Carole, and Richard Wilson (eds.), Medieval Norwich (London and New York, 2004) Ray, Roger D., ‘Orderic Vitalis and William of Poitiers’, Revue belge de philologie et d’histoire, l (1972), 1,116–27 Rees Jones, Sarah, York: The Making of a City 1068–1350 (Oxford, 2013) Renoux, Annie, Fécamp: du palais ducal au palais de Dieu (Paris, 1991) —— ‘Aux sources du pouvoir châtelain de Geoffroi “seigneur de Mayenne, le plus fort homme du Maine” (c.1040–1098)’, in Dominique Barthélemy and Olivier Bruand (eds.), Les pouvoirs locaux en France du centre et de l’Ouest (VIIIe–XIe siècle). Implantation et moyens d’action (Rennes, 2004), 61–90 Reuter, Timothy, ‘Plunder and Tribute in the Carolingian Empire’, TRHS, 5th ser., xxxv (1985), 75–94 (reprinted in Janet L. Nelson [ed.], Medieval Polities and Modern Mentalities [Cambridge, 2006], chapter 13) —— ‘Assembly Politics in Western Europe from the 8th Century to the 12th’, in Peter Linehan and Janet L. Nelson (eds.), The Medieval World (London, 2001), 423–50 —— review of Gerd Althoff, Spielregeln der Politik im Mittelalter: Kommunikation in Frieden und Fehde (Darmstadt, 1997), in Bulletin of the German Historical Institute, xxiii (2001), 407 Rex, Peter, Hereward: The Last Englishman (Stroud, 2005) —— William the Conqueror: The Bastard of Normandy (Stroud, 2011) Reynolds, Susan, ‘Eadric silvaticus and the English Resistance’, BIHR, liv (1981), 102–5 —— Kingdoms and Communities in Western Europe, 2nd edn. (Oxford, 1997) —— ‘Assembly Government and Assembly Law’, in Nelson, Reynolds, and Johns (eds.), Gender and Historiography, 191–9

bibliography

561

—— ‘Trust in Medieval Society and Politics’, in Susan Reynolds, The Middle Ages without Feudalism: Essays in Criticism and Comparison on the Medieval West (Farnham and Burlington, VT, 2012), chapter 13 —— ‘Tenure and Property in Medieval England’, HR, lxxxviii (2015), 563–76 Ridel, Elisabeth, ‘Les navires de la Tapisserie de Bayeux à la lumière du vocabulaire nautique normand’, in Lemagnen (ed.), La Tapisserie de Bayeux, 209–27 —— Les navires de la Conquête (Cully, 2010) Rieth, Eric, ‘Construire les navires au temps des Vikings: une relecture archéologique d’une scène de la Tapisserie de Bayeux’, in Lemagnen (ed.), La Tapisserie de Bayeux, 195–207 Roach, Levi, Kingship and Consent in Anglo-Saxon England, 871–978: Assemblies and the State in the Early Middle Ages (Cambridge, 2013) Robinson, Ian S., The Papacy 1073–1198: Continuity and Innovation (Cambridge, 1990) —— Henry IV of Germany (Cambridge, 1999) Roffe, David, ‘Domesday Book and Northern Society: A Reassessment’, EHR, cv (1990), 310–36 —— ‘Anglo-Saxon Nottingham and the Norman Conquest’, in Beckett (ed.), A Centenary History of Nottingham, 24–42 —— Domesday: The Inquest and the Book (Oxford, 2000) —— Decoding Domesday (Woodbridge, 2007) —— ‘A Profession of Ignorance: An Insight into Domesday Procedure in an Early Reference to the Inquest’, in Dalton and Luscombe (eds.), Rulership and Rebellion in the Anglo-Norman World, 45–60 —— ‘Domesday Now: A View from the Stage’, in Roffe and Keats-Rohan (eds.), Domesday Now, 7–60 —— ‘Talking to Others and Talking to Itself: Government and the Changing Role of the Records of the Domesday Inquest’, in Roffe and Keats-Rohan (eds.), Domesday Now, 289–303 ——, and K. S. B. Keats-Rohan (eds.), Domesday Now: New Approaches to the Inquest and the Book (Woodbridge, 2016) —— (ed.), The English and their Legacy, 900–1200: Essays in Honour of Ann Williams (Woodbridge, 2012) Rollason, David, ‘Symeon of Durham’s Contribution to Historical Writing in Northern England’, in Rollason (ed.), Symeon of Durham, 1–13 —— ‘Symeon of Durham’s Historia de Regibus Anglorum et Dacorum as a Product of TwelfthCentury Historical Workshops’, in Brett and Woodman (eds.), The Long Twelfth-Century View of the Anglo-Saxon Past, 95–111 ——, with Derek Gore and Gillian Fellows-Jensen, Sources for York History to AD 1100 (York, 1998) ——, Conrad Leyser, and Hannah Williams (eds.), England and the Continent in the Tenth Century: Studies in Honour of Wilhelm Levison (1876–1947) (Turnhout, 2010) ——, (ed.), Symeon of Durham: Historian of Durham and the North (Stamford, 1998) Rosenwein, Barbara H. (ed.), Anger’s Past: The Social Uses of Emotion in the Middle Ages (Ithaca, NY, 1998) Round, J. H., ‘The Conqueror at Exeter’, in J. H. Round, Feudal England (London, 1895), 431–55 Rowley, Trevor, The Man behind the Bayeux Tapestry: Odo, William the Conqueror’s Half-Brother (Stroud, 2013) Rubinstein, Jay, ‘The Life and Writings of Osbern of Canterbury’, in Eales and Sharpe (eds.), Canterbury and the Norman Conquest, 27–40 —— Guibert of Nogent: Portrait of a Medieval Mind (New York and London, 2002) Rumble, Alexander R., ‘The Purposes of the Codex Wintoniensis’, ANS, iv (1982), 153–66, 224–32 —— Property and Piety in Early Medieval Winchester (Oxford, 2002)

562

bibliography

—— Writing and Texts in Anglo-Saxon England (Cambridge, 2006) Rutter, Michael, ‘Resilience as a Dynamic Concept’, Development and Psychopathology, xxiv (2012), 335–44 Searle, Eleanor, ‘Possible History’, Speculum, lxi (1986), 779–86 —— Predatory Kinship and the Growth of Norman Power, 840–1066 (Berkeley and Los Angeles, CA, and London, 1988) Sharpe, Richard, ‘The Setting of St Augustine’s Translation, 1091’ in Eales and Sharpe (eds.), Canterbury and the Norman Conquest, 1–13 —— ‘The Use of Writs in the Eleventh Century’, ASE, xxxii (2003), 247–91 —— ‘King Harold’s Daughter’, HSJ, xix (2008, for 2007), 1–27 —— ‘Peoples and Languages in Eleventh- and Twelfth-Century Britain and Ireland: Reading the Charter Evidence’, in Dauvit Broun et al., The Reality behind Charter Diplomatic in Anglo-Norman Britain (Glasgow, 2011), 1–119 Skoda, Hannah, Medieval Violence: Physical Brutality in Northern France 1270–1330 (Oxford, 2013) Slavin, Philip, ‘Warfare and Ecological Destruction in the Early Fourteenth-Century British Isles’, Environmental History, xix (2014), 528–50 Smith, Mary Frances, ‘Archbishop Stigand and the Eye of the Needle’, ANS, xvi (1994), 199–219 Sønnesyn, Sigbjørn Olsen, William of Malmesbury and the Ethics of History (Woodbridge, 2012) —— ‘In vinea Sorech laborare: The Cultivation of Unity in Twelfth-Century Monastic Historiography’, ANS, xxxvi (2014), 167–87 —— ‘ “Only through Time is Time Conquered”: Liturgy, History and the Timeless Aspirations of the Temporal’, in Bergsagel, Hiley, and Riis (eds.), Of Chronicles and Kings, 23–50 Soulignac, Robert, ‘La dernière maladie de Guillaume le Conquérant’, AN, xxxvii (1987), 389–90 Southern, R. W., ‘The English Origin of the “Miracles of the Virgin” ’, Journal of Medieval and Renaissance Studies, iv (1958), 176–216 —— ‘England’s First Entry into Europe’, in Medieval Humanism and Other Studies (Oxford, 1970) —— Saint Anselm: A Portrait in a Landscape (Cambridge, 1990) Spear, David, ‘William Bona Anima, Abbot of St Stephen’s of Caen’, HSJ, i (1989), 51–60 —— ‘The School of Caen Revisited’, HSJ, iv (1992), 55–66 —— The Personnel of the Norman Cathedrals during the Ducal Period, 911–1204 (London, 2006) Stafford, Pauline, ‘The Laws of Cnut and the History of Anglo-Saxon Royal Promises’, ASE, x (1982), 173–90 —— Queen Emma and Queen Edith (Oxford, 1997) —— ‘Writing the Biography of Eleventh-Century Queens’, in Bates, Crick, and Hamilton (eds.), Writing Medieval Biography, 750–1250, 99–109 —— ‘The Anglo-Saxon Chronicles, Identity and the Making of England’, HSJ, xix (2008, for 2007), 28–50 —— ‘Edith, Edward’s Wife and Queen’, in Mortimer (ed.), Edward the Confessor, 119–38 —— ‘Archbishop Ealdred and the D Chronicle’, in Crouch and Thompson (eds.), Normandy and its Neighbours, 135–56 —— ‘Royal Women and Transitions. Emma and Ælfgifu in 1035–1042/3’, in Körntgen and Waßenhoven (eds.), Patterns of Episcopal Power, 127–44 Stapleton, Thomas, ‘Observations on the History of Adeliza, Sister of William the Conqueror’, Archaeologia, xxvi (1836), 349–60 Stevenson, W. H., ‘A Contemporary Description of the Domesday Survey’, EHR, xxii (1907), 72–84 Strevett, Neil, ‘The Anglo-Norman Civil War of 1101’, ANS, xxvi (2004), 159–75 Strickland, Matthew, War and Chivalry: The Conduct and Perception of War in England and Normandy, 1066–1217 (Cambridge, 1996)

bibliography

563

—— ‘Military Technology and Conquest: The Anomaly of Anglo-Saxon England’, ANS, xix (1997), 353–82 —— ‘Provoking or Avoiding Battle? Challenge, Duel and Single Combat in the Warfare of the High Middle Ages’, in Matthew Strickland (ed.), Armies, Chivalry and Warfare in Medieval Britain and France, 317–43 —— ‘Status, Display and Defence on the Welsh March, circa 1067 to 1087: Reflections on the Context and Functions of the Great Tower at Chepstow, Gwent’, in Erik Beck et al. (eds.), —— Burgen im Breisgau: Aspekte von Burg und Herrschaft im interdisziplinären und überregionalen Vergleich (Ostfildern, 2012), 89–109 —— (ed.), Anglo-Norman Warfare (Woodbridge, 1992) —— (ed.), Armies, Chivalry and Warfare in Medieval Britain and France: Proceedings of the 1995 Harlaxton Symposium (Stamford, 1998) Stringer, Keith J., and Andrew Jotischky (eds.), Norman Expansion: Connections, Continuities and Contrasts (Farnham and Burlington, VT, 2013) Suhamy, Henri, Guillaume le Conquérant (Paris, 2008) Summerson, Henry, ‘Tudor Antiquaries and the Vita Ædwardi Regis’, ASE, xxxviii (2009), 157–84 Tabuteau, Emily Zack, Transfers of Property in Eleventh-Century Norman Law (Chapel Hill, NC, and London, 1988) —— ‘The Role of Law in the Succession to Normandy and England, 1087’, HSJ, iii (1991), 141–69 —— ‘The Family of Moulins-la-Marche in the Eleventh Century’, Medieval Prosopography, xiii (1992), 29–65 —— ‘Punishments in Eleventh-Century Normandy’, in Warren C. Brown and Piotr Górecki (eds.), Conflict in Medieval Europe: Changing Perspectives on Society and Culture (Aldershot and Burlington, VT, 2003), 131–49 Tanner, Heather J., ‘The Expansion of the Power and Influence of the Counts of Boulogne under Eustace II’, ANS, xiv (1992), 251–86 —— Families, Friends and Allies: Boulogne and Politics in Northern France and England, c.879–1160 (Leiden, 2004) Taylor, Arnold J., ‘ “Belrem” ’, ANS, xiv (1993), 1–23 Taylor, Pamela, ‘Ingelric, Count Eustace and the Foundation of St Martin-le-Grand’, ANS, xxiv (2002), 215–37 —— ‘The Episcopal Returns in Domesday’, in Roffe and Keats-Rohan (eds.), Domesday Now, 197–217 Thomas, Hugh M., ‘A Yorkshire Thegn and His Descendants after the Conquest’, Medieval Prosopography, viii (1987), 1–22 —— ‘The Gesta Herwardi, the English, and their Conquerors’, ANS, xxi (1999), 213–32 —— The English and the Normans: Ethnic Hostility, Assimilation, and Identity 1066–c.1220 (Oxford, 2003) —— ‘The Significance and Fate of the Native English Landholders of 1086’, EHR, cxviii (2003), 303–33 —— ‘Violent Disorder in King Stephen’s England: A Maximum Argument’, in Paul Dalton and Graeme J. White (eds.), King Stephen’s Reign (1135–1154) (Woodbridge, 2008), 139–70 Thompson, Kathleen, ‘Family and Influence to the South of Normandy in the Eleventh Century: The Lordship of Bellême’, JMH, xi (1985), 215–26 —— ‘The Norman Aristocracy before 1066: The Example of the Montgomerys’, HR, lx (1987), 251–63 —— Power and Border Lordship in Medieval France: The County of Perche, 1000–1226, Royal Historical Society Studies in History (Woodbridge, 2002) —— ‘Robert [called Robert Curthose), Duke of Normandy’, ODNB —— ‘Being the Ducal Sister: The Role of Adelaide of Aumale’, in Crouch and Thompson (eds.), Normandy and its Neighbours, 900–1250, 61–76

564

bibliography

Thomson, R. M., ‘William of Malmesbury’s Diatribe against the Normans’, in Brett and Woodman (eds.), The Long Twelfth-Century View of the Anglo-Saxon Past, 113–21 Thorn, Frank, ‘Non Pascua sed Pastura: The Changing Choice of Terms in Domesday’, in Roffe and Keats-Rohan (eds.), Domesday Now, 109–36 —— , and Caroline Thorn, ‘The Writing of Great Domesday Book’, in Hallam and Bates (eds.), Domesday Book, 37–72 Todière, Louis-Phocion, Guillaume le Conquérant (Tours, 1856; Paris, 1993) Tombs, Robert, The English and their History (London, 2014) Turner, Rick, Chepstow Castle (Cardiff, 2002) Tuten, Belle S., and Tracey L. Billado (eds.), Feud, Violence and Practice: Essays in Medieval Studies in Honor of Stephen D. White (Farnham and Burlington, VT, 2010) Urbanski, Charity, ‘Apology, Protest and Suppression: Interpreting the Surrender of Caen (1105)’, HSJ, xix (2008, for 2007), 137–53 Van Eickels, Klaus, Vom inszenierten Konsens zum systematisierten Konflikt (Stuttgart, 2002) Van Houts, Elisabeth, ‘The Political Relations between Normandy and England before 1066 according to the Gesta Normannorum Ducum’, in Foreville (ed.), Les mutations socioculturelles au tournant des XIe–XIIe siècles, 85–97 (reprinted in Elisabeth van Houts, History and Family Traditions [Aldershot 1999], chapter 5) —— ‘The Origins of Herleva, Mother of William the Conqueror’, EHR, ci (1986), 389–404 —— ‘The Ship List of William the Conqueror’, ANS, x (1988), 159–83 (reprinted in van Houts, History and Family Traditions, chapter 6) —— ‘Latin Poetry and the Anglo-Norman Court 1066–1135: The Carmen de Hastingae Proelio’, JMH, xv (1989), 39–62 (reprinted in van Houts, History and Family Traditions, chapter 9) —— ‘The Norman Conquest through European Eyes’, EHR, cx (1995), 832–53 (reprinted in van Houts, History and Family Traditions, chapter 8) —— ‘The Memory of 1066 in Written and Oral Traditions’, ANS, xix (1997), 167–79 —— ‘Wace as Historian’, in Keats-Rohan (ed.), Medieval Prosopography, 103–32 (reprinted in van Houts, History and Family Traditions, chapter 10) —— ‘Historiography and Hagiography at Saint-Wandrille: The Inventio et Miracula sancti Vulfranni’, ANS, xii (1990), 235–51 (reprinted in van Houts, History and Family Traditions, chapter 4) —— History and Family Traditions in England and the Continent 1000–1200 (Aldershot, 1999) —— ‘Countess Gunnor of Normandy (c.950 – 1031)’, Collegium Medievale, xii (1999), 7–24 —— ‘Hereward and Flanders’, ASE, xxviii (1999), 201–23 —— Memory and Gender in Medieval Europe, 900–1200 (Basingstoke and London, 1999) —— ‘Les Femmes dans l’histoire du duché de Normandie’, Tabularia. Etudes www.unicaen. fr/mrsh/craham/revue/tabularia, ii (2002), 19–34 —— ‘The Echo of the Conquest in the Latin Sources: Duchess Matilda, her Daughters and the Enigma of the Golden Child’, in Bouet, Levy and Neveux (eds.), The Bayeux Tapestry, 135–53 —— ‘The Vocabulary of Exile and Outlawry in the North Sea Area around the First Millennium’, in Napran and van Houts (eds.), Exile in the Middle Ages, 13–28 —— ‘Edward and Normandy’, in Mortimer (ed.), Edward the Confessor, 63–76 —— ‘Intermarriage in Eleventh-Century England’, in Crouch and Thompson (eds.), Normandy and its Neighbours, 900–1250, 237–70 —— ‘The Fate of Priests’ Sons in Normandy with Special Reference to Serlo of Bayeux’, HSJ, xxv (2014, for 2013), 57–105 —— ‘The Planctus on the Death of William Longsword (943) as a Source for Tenth-Century Culture in Normandy and Aquitaine’, ANS, xxxvi (2014), 1–22 —— ‘Queens in the Anglo-Norman/Angevin Realm 1066–1216’, in Claudia Zey, Sophie Caflisch, and Philippe Goridis (eds.), Mächtige Frauen? Königinnen und Fürstinnen im Europäischen Mittelalter (11.–14. Jahrhundert) (Ostfildern, 2015), 199–222

bibliography

565

—— ‘Qui étaient les Normands? Quelques observations sur des liens entre la Normandie, l’Angleterre, et l’Italie au début du XIe siècle’, in David Bates and Pierre Bauduin (eds.), 911–2011: Penser les mondes normands médiévaux (Caen, 2016), 129–46 Van Kempen, Ad F. J., ‘The Mercian Connection, Harold Godwineson’s Ambitions, Diplomacy and Channel-Crossing, 1056–1066’, History, xciv (2009), 2–19 Van Torhoudt, Eric, ‘Les sièges du pouvoir des Néel, vicomtes dans le Cotentin’, in Flambard Héricher (ed.), Les lieux du pouvoir au Moyen Age en Normandie et sur ses marges, 7–35 —— ‘L’ “énigme” des origines de l’abbaye de Montebourg: une question de methode?’, in Bouet, Bougy, Garnier, and Maneuvrier (eds.), De part et d’autre de la Normandie médiévale, 331–46 Vaughn, Sally N., ‘Lanfranc at Bec: A Reinterpretation’, Albion, xvii (1985), 135–48 —— Anselm of Bec and Robert of Meulan: The Innocence of the Dove and the Wisdom of the Serpent (Berkeley and Los Angeles, CA, and London, 1987) —— ‘Lanfranc, Anselm and the School of Bec: In Search of the Students of Bec’, in Marc A. Meyer (ed.), The Culture of Christendom (London, 1993), 155–81 Verlinden, Charles, Robert Ier Le Frison, comte de Flandre (Antwerp and Paris, 1935) Villard, François, ‘Primatie des Gaules et réforme grégorienne’, Bibliothèque de l’Ecole des Chartes, cil (1991), 421–34 Violette, Louis, ‘Une étape décisive dans l’éveil des activités historiographiques au service du siège de Rouen’, Tabularia, iii (2003), 57–67 Waldman, Thomas, ‘Charters and Influences from Saint-Denis c.1000–1070’, in Licence (ed.), Bury St Edmunds and the Norman Conquest, 22–30 Walker, David, ‘Gruffudd ap Llywelyn (d.1063)’, ODNB Walker, Ian W., Harold: The Last Anglo-Saxon King (Stroud, 1997) Walmsley, John, ‘The Early Abbesses, Nuns and Female Tenants of the Abbey of Holy Trinity, Caen’, JEH, xlviii (1997), 425–44 Wareham, Andrew, ‘Two Models of Marriage: Kinship and Social Order in England and Normandy’, in Bijsterveld, Teunis, and Wareham (eds.), Negotiating Secular and Ecclesiastical Power, 107–32 Waßenhoven, Dominik, ‘Swaying Bishops and the Succession of Kings’, in Körntgen and Waßenhoven (eds.), Patterns of Episcopal Power, 89–109 Watkins, Carl, ‘The Cult of Earl Waltheof at Crowland’, Hagiographica, iii (1996), 95–111 Webb, J. R., ‘Representations of the Warrior-Bishop in Eleventh-Century Lotharingia’, EME, xxiv (2016), 103–30 Webber, Teresa, Scribes and Scholars at Salisbury Cathedral c.1075–c.1125 (Oxford, 1992) Weiler, Björn, ‘William of Malmesbury on Kingship’, History, xc (2005), 3–22 West, Charles, Reframing the Feudal Revolution: Political and Social Transformation between Marne and Moselle, c.800–c.1100 (Cambridge, 2013) West, Francis J., ‘The Colonial History of the Norman Conquest’, History, lxxxiv (1999), 219–36 White, Stephen D., ‘Proposing the Ordeal and Avoiding It: Strategy and Power in Western Frankish Litigation, 1050–1110’, in Bisson (ed.), Cultures of Power, 89–123 (reprinted in White, Feuding and Peace-Making in Eleventh-Century France, chapter 7) —— ‘The Politics of Anger’, in Rosenwein (ed.), Anger’s Past, 127–52 (reprinted in White, Feuding and Peace-Making in Eleventh-Century France, chapter 4) —— Feuding and Peace-Making in Eleventh-Century France (Aldershot and Burlington, VT, 2005) —— ‘The Beasts who Talk on the Bayeux Embroidery: The Fables Revisited’, ANS, xxxiv (2012), 209–36 Wickham, Chris, The Inheritance of Rome: A History of Europe from 400 to 1000 (London, 2009) —— ‘The “Feudal Revolution” and the Origins of Italian City Communes’, TRHS, 6th ser., xxiv (2014), 29–55 Williams, A. G., ‘Norman Lordship in South-East Wales during the Reign of William I’, Welsh History Review, xvi (1993), 445–66

566

bibliography

Williams, Ann, ‘Some Notes and Considerations on Problems Connected with the English Royal Succession, 860–1066’, ANS, i (1979), 144–67 —— ‘Land and Power in the Eleventh Century: The Estates of Harold Godwineson’, ANS, iii (1981), 171–87 —— ‘The King’s Nephew: The Family and Career of Ralph, Earl of Hereford’, in HarperBill, Holdsworth, and Nelson (eds.), Studies in Medieval History Presented to R. Allen Brown, 327–43 —— ‘A Vice-Comital Family in Pre-Conquest Warwickshire’, ANS, xi (1989), 279–95 —— The English and the Norman Conquest (Woodbridge, 1995) —— Kingship and Government in Pre-Conquest England (Basingstoke and London, 1999) —— ‘Ælfgar, Earl of Mercia (d.1062?)’, ODNB —— ‘Eadric [Edric] the Wild [Eadric Cild] (fl.1067–1072)’, ODNB —— ‘Eadnoth the Staller (d.1068)’, ODNB —— ‘Ralph [called Ralph de Gael, Ralph Guader], Earl (d.1097x9)’, ODNB —— ‘Ralph the Staller, Earl of East Anglia (d.1068x1070)’, ODNB —— ‘The Cunning of the Dove: Wulfstan and the Politics of Accommodation’, in Barrow and Brooks (eds.), St Wulfstan and his World, 23–38 —— The World before Domesday: The English Aristocracy, 900–1066 (London, 2008) —— ‘Hunting the Snark and Finding the Boojum: The Tenurial Revolution Revisited’, in Roffe and Keats-Rohan (eds.), Domesday Now, 155–68 Williams, Gareth, ‘Was the Last Anglo-Saxon King of England a Queen? A Possible Posthumous Coinage of Harold II’, Yorkshire Numismatist, iv (2012), 159–70 Winkler, Emily A., ‘1074 in the Twelfth Century’, ANS, xxxvi (2014), 241–58 —— ‘The Norman Conquest of the Classical Past: William of Poitiers, Language and History’, JMH, xlii (2016), 456–78 Winterbottom, Michael, ‘A New Passage of William of Malmesbury’s Gesta Pontificum’, Journal of Medieval Latin, xi (2001), 50–9 —— ‘William of Malmesbury and the Normans’, Journal of Medieval Latin, xx (2010), 70–7 Wolfram, Herwig, Conrad II 990–1039: Emperor of Three Kingdoms, trans. Denise A. Kaiser (University Park, PA, 2006) Wood, Susan, The Proprietary Church in the Medieval West (Oxford, 2006) Wormald, Patrick, How Do We Know So Much about Anglo-Saxon Deerhurst?, Deerhurst Lecture, 1991 (Friends of Deerhurst Church, 1993) (reprinted in Patrick Wormald, The Times of Bede: Studies in Early English Christian Society and its Historian, ed. Stephen Baxter [Oxford, 2006]) —— The Making of English Law: King Alfred to the Twelfth Century, volume 1: Legislation and its Limits (Oxford, 1999) —— ‘Lordship and Justice in the Early English Kingdom: Oswaldslow Revisited’, in idem, Legal Culture in the Early Medieval West: Law as Text, Image and Experience (London, 1999), 313–32 Wyatt, David, Slaves and Warriors in Medieval Britain and Ireland, 800–1200 (Leiden and Boston, MA, 2009) Zey, Claudia, ‘Vormünder und Berater Heinrichs IV. Im Urteil der Zeitgenossen (1056– 1075)’, Vorträge und Forschungen, lxix, Heinrich IV. (Ostfildern, 2009), 87–126 Zumthor, Paul, Guillaume le Conquérant (Paris, 1964; reprinted, 2003)

UNPUBLISHED REPORT Flambard Héricher, Anne-Marie (ed.), ‘Projet collectif de recherches sur les fortifications de terre en Haute-Normandie 2007’

bibliography

567

UNPUBLISHED THESES Alexander, Alison, ‘Annalistic Writing in Normandy, c.1050–c.1225’, University of Cambridge PhD Thesis (2012) Austin, Elisabeth, ‘Thomas of Bayeux, Archbishop of York, 1070–1100’, University of St Andrews PhD Thesis (1996) Bloche, Michaël, ‘Le chartrier de l’abbaye de la Trinité de Fécamp: étude et édition critique, 928/929–1190’, École Nationale des Chartes, Thèse pour le diplôme d’archiviste paléographe (2011) Carlson, Eric Gustav, ‘The Abbey Church of Saint-Etienne at Caen in the Eleventh and Early Twelfth Centuries’, Yale University PhD Thesis, 1968 Cartwright, Charlotte, ‘Matilda of Flanders in Normandy: A Study of Eleventh-Century Female Power’, University of Liverpool PhD Thesis, 2012 Fujimoto, Tamiko, ‘Recherche sur l’écrit documentaire au Moyen Age. Edition et commentaire du cartulaire de l’abbaye Saint-Etienne de Caen (XIIe siècle)’, 2 vols., Thèse de Doctorat, Université de Caen Basse-Normandie (2012) Gazeau, Véronique, ‘Monachisme et aristocratie au XIe siècle: l’exemple de la famille de Beaumont’, Thèse de Doctorat de troisième cycle, Université de Caen (1986–7) Gazzoli, Paul, ‘Anglo-Danish Relations in the Later Eleventh Century’, University of Cambridge PhD Thesis (2010) Letouzey-Réty, Catherine, ‘Ecrit et gestion du temporel dans une grande abbaye de femmes anglo-normande: la Sainte-Trinité de Caen (XIe–XIIIe siècle)’, Thèse de Doctorat, Université de Paris I Panthéon-Sorbonne and University of London (co-tutelle, 2011) Potter, Julie A., ‘The Friendship Network of the Abbey of Le Bec-Hellouin in the Eleventh and Twelfth Centuries’, University of Cambridge PhD Thesis (2014) Raich, Susan Alice, ‘The Sea in the Anglo-Norman Realm, c.1050 to c.1180’, University of Cambridge PhD thesis (2014) Van Torhoudt, Eric, ‘Centralité et marginalité en Neustrie et dans le duché de Normandie: maîtrise du territoire et pouvoirs locaux dans l’Avranchin, le Bessin et le Cotentin (VIe– XIe siècles)’, Thèse de Doctorat, Université Paris Diderot (Paris VII, 2008) Yearl, Mary Katherine Keblinger, ‘The Time of Bloodletting’, Yale University PhD Thesis (2005)

INDEX Abbeville (Somme), 456, 484, 490 Abbotsbury (Dorset), abbey of, 326 Abelin, donor to the abbey of Marmoutier, 189 Abernethy (Perth and Kinross), agreement of (1072), 361, 362, 408, 426 Abingdon (Oxfordshire), abbey of, 167, 325, 349, 449, 452, 503, 519 Historia, 167, 426, 449, 452, 458, 503 See Adelelm, Ealdred, Rainald, abbots of Achilles, 246, 355 Acta Archiepiscopum Rotomagensium, 134–6, 280, 365, 367 Acta Lanfranci, 428 Actus Pontificum Cenomannis, 178, 190, 365 Adalbert, archbishop of Hamburg/ Bremen, 56 Adam fitz Hubert, tenant of Odo, bishop of Bayeux and Domesday commissioner, 467 Adam of Bremen, 22, 35, 56, 264 Adela, daughter of Count Robert the Frisian of Flanders, wife of Cnut IV, king of Denmark, 341 Adela, daughter of King Robert the Pious, wife of Duke Richard III and Count Baldwin V of Flanders, 18 and note Adela, daughter of William and Matilda, countess of Blois-Chartres, 82, 213, 244, 275, 277, 331, 403, 419, 438–9, 439, 495 Adela, wife of Count Baldwin V of Flanders and mother of Matilda, 100, 120 note Adelaide, William’s sister or half-sister, countess of Aumale, 20, 26, 28, 79, 99–100, 130–1, 139, 154, 187, 350, 375–6. See Euguerrand II, count of Ponthieu, Odo, Lambert II, count of Lens, husbands of

Adelaide, kinswoman of William, wife of Robert fitz Giroie, 154 Adelelm, abbot of Abingdon, 426 Adelida, daughter of William and Matilda, 128, 330–1, 376, 393, 418, 461 betrothals to Count Herbert II of Maine, 180, 198; Earl/King Harold, 198, 266; Earl Edwin (possibly), 266; King Alfonso VI of Galicia, 330–1, 461; Count Simon of Amiens, Valois, and the Vexin (possibly), 393; Robert Guiscard (unlikely), 393 Adeline, daughter of Waleran I, count of Meulan, wife of Roger de Beaumont, 79 Adeliza, daughter of Duke Richard II, wife of Rainald, a count of Burgundy, 63, 63–4 note, 95, 375 Adhemar of Chabannes, 30 Ælfgar, earl of Mercia, son of Earl Leofric, 146, 191 Ælfgifu of Northampton, concubine of King Cnut, 20, 65 Ælfgyva, woman on the Bayeux Tapestry, 198, 199 note Ætheldreda, St, 348, 431 Æthelmær, bishop of East Anglia, 335 Æthelnoth, abbot of Glastonbury, 276, 442 note Æthelnoth, archbishop of Canterbury, 364 Æthelnoth of Canterbury, 276 Æthelred the Unready, king of England, 24, 37, 38, 39, 265, 300, 322, 504, 510. See Emma, wife of; Edmund Ironside, Alfred, Edward the Confessor, Godgifu, children of Æthelric, bishop of Durham, 335 Æthelric, bishop of Selsey, 336, 358, 364, 384 Æthelric, tenant of Marsh Gibbon (Buckinghamshire), 470

568

index Æthelsige, abbot of St Augustine’s, Canterbury, 304, 308 Æthelstan, king of the English, 510, 511, 527 Æthelwig, abbot of Evesham, 296, 315, 381, 406–07 Æthelwine, bishop of Durham, 296–7, 303, 335, 346, 349 Agatha, possible daughter of William and Matilda, 331. See Adelida Agildeburga (Evenlode or Four-Shire Stone, near Moreton-in-Marsh, Warwickshire), 406–7 Agnes of Poitou, Empress, wife of Emperor Henry III, 56–7 Ailnoth of Canterbury, 458 Aimeri vicomte of Thouars (Deux-Sèvres), 309 Ainard, abbot of Saint-Pierre-sur-Dives, 76 Aire, river (Yorkshire), 311, 312, Alan III, count of Brittany, 32, 33, 40, 45–6, 50, 53, 59, 61–2, 81, 86, 96, 179, 201, 203, 390, 439, 448. See also Bertha, wife of Alan IV Fergeant, count of Brittany, 392, 439, 453, 480 Alan Niger, son of Count of Eudo, 202, 224, 277, 319, 350, 388, 448, 455, 465–6, 474, 478, 455, 465–6, 474, 478 Alan Rufus, son of Count Eudo, 224 Alcuin of York, 314, 522 Aldreth (Cambridgeshire), 346, 347, 348 Alençon (Orne), 26–7, 28, 69, 70, 91, 114, 122–6, 133, 152, 155, 189, 399, 520 Alexander II, pope, 55, 165, 177, 272, 190–1, 206, 219, 221–2, 252 note, 280–1, 292, 326, 329, 334, 335, 337, 339, 357–9, 368, 384, 411, 421, 441, 497, 519 See also papal banner, grant of Alfonso VI, king of Galicia, 330, 393 Alfred the Great, king of Wessex, 511, 527 Alfred, brother of King Edward the Confessor, 37, 39–41, 45, 65–6, 111 Alfred of Beverley, 315 Almær, burgess of Bedford, 464 Althoff, Gerd, 2, 516 Alveston (Warwickshire), 433, 448 Alwine, a priest, 464 Amatus of Montecassino, 243

569

Ambrières (Mayenne), castle of, 140–1, 152–3, 155, 157, 186, 201–2 Amiens (Somme), 187 Andersey, island of (Oxfordshire), 167, 449 ‘anger’ (ira) and rule , 34. See ‘anger’ (ira), William’s and as a facet of rule Angers (Maine-et-Loire), 142 abbey of Saint-Aubin, 141 cathedral church of, 103 Angevins, people, 321 Anglo-Saxon Chronicle, 69, 109, 110 note, 115, 117, 147, 265–6, 332, 344–5, 360, 366–7, 368–9, 378–9, 384, 401, 452, 462, 469, 483, 501 ‘C’ version, 212–13 ‘D’ version, 109, 112–13, 115–16, 148–9, 159, 200, 211, 212–14, 233, 238, 247, 248–50, 256, 261–2, 265, 266, 271, 281, 289, 295, 310, 344–5, 369–70, 401, 501 ethics and legitimacy of William’s actions and rule, 4–5, 14, 159, 200, 211, 212–14, 256, 261–2, 265, 281, 289, 305 note, 504 on the significance of his birth, 22 ‘E’ version, 101, 112, 117, 182, 212, 231, 270, 285, 295, 315, 322, 329, 332, 344–5, 353, 370, 394, 428, 431, 432, 446, 458, 459, 464, 470, 473–4, 475, 479–80, 493–4, 506, 508, 527 ethics and legitimacy of William’s actions and rule, 5, 462, 470, 493–4, 504, 506–7, 527 Anjou, counts and county of, 59, 70, 71, 81, 93, 98, 125, 144, 154, 179–81, 183, 191, 202, 203, 265, 372, 391, 392, 403, 419, 429, 439, 486, 508, 511. See Fulk II the Good, Fulk III Nerra, Fulk IV Rechin, Geoffrey II Martel, Geoffrey III the Bearded, Geoffrey IV Plantagenet, counts of Anna Comnena, 525 Annales Regni Francorum, 235 Annals of Lyre, 76. See also Norman annals Annals of Ulster, 361 Annals of Winchester, 475 Anne of Kiev, queen of France, 187 Anno, archbishop of Cologne, 56–7, 371 Anselm I, bishop of Lucca (Pope Alexander II), 165 Anselm II, bishop of Lucca and papal legate, 460 Anselm of Saint-Rémy, 99

570

index

Anselm, St, prior and abbot of Le Bec, later archbishop of Canterbury, 185, 331, 400, 411, 414, 471, 483, 485–6, 521 Ansfrid, abbot of Saint-Pierre of Préaux, 103 Ansfrida, abbess of Saint-Léger of Préaux, 355 antecessor, 472–3 Apulia (Italy), 398, 414, 480, 492 Aquitaine, 30, 225, 231, 309, 391, 454, 511 duchy and dukes of, 33, 75, 126. See William VI the Fat, William VII, William VIII, dukes of archers at the Battle of Hastings, 240–1, 244, 246 Archives Départementales de La Manche, 136 Argentan (Orne), 31, 32, 59–60, 62, 70 Arnketel, son of Ecgfrith, 304 Arnold, bishop of Le Mans, 186 (probably), 190–1, 304, 365, 377, 410, 421, 437 Arnold d’Echauffour, 155, 174–5, 524. See also Robert fitz Giroie, uncle of Arnost, monk of Le Bec, bishop of Rochester, 385 Arnulf III, count of Flanders, 340–1, 420 Arnulf de Hesdin (Pas-de-Calais), 429 Arnulf of Chocques, 375 Arnulf, son of William Talvas and nephew of bishop Ivo of Sées, 80, 98 Arques (Seine-Maritime), castle of, 129–32. See William, count of Arundel (Sussex), 285 Ascelin, son of Arthur, citizen of Caen, 490 Athelney (Somerset), abbey of, 442 Atlantic Ocean, 230, 231 Aubrey de Coucy-le-Château-Auffrique, earl of Northumbria, 427, 479, 484 Auchy-lès-Aumale (Seine-Maritime), college of secular canons, 139 Augustine, St, prayers and City of God, 227, 521–2, 523, 526 Augustine of Canterbury, St, 409, 445 Aumale (Seine-Maritime), 79, 130, 139. See also Adelaide, countess of Auvergne, the, 94, 126, 132, 419 Auxerre (Yonne), abbey of Saint-Germain, 320 Avranches (Manche), 173, 202–03. See Hugh, John, bishops of Avranchin, the, 137, 190, 410

Avre, river, 64 Azzo, lord of Este, 304, 366 Baldwin, abbot of Bury St Edmunds, 167, 214, 287, 306, 357–8, 429–30, 447, 469, 477 Baldwin IV, count of Flanders, 32, 45, 94 Baldwin V, count of Flanders, 18, 91, 92–4, 99–100, 102–3, 105–6, 112, 115–17, 120–1, 126, 131, 139, 147, 169, 188, 209, 265, 340, 342 guardianship of King Philip I of France, 153–4 Saint-Omer diploma of (1056), 147–8 Baldwin VI, count of Flanders, 340 Baldwin, count of Hainault, 341 Baldwin de Meulles, son of Gilbert, count of Brionne, 86, 290 Baldwin Filleul, 105 Bamberg (Upper Franconia), 353 Bamburgh, house of, 304 See also Arnketel, son of Ecgfrith; Gospatric son of Maldred; Karl, Osulf, son of Eadulf of Bamburgh Barfleur (Manche), 301 Barking (Essex), 266, 282 Barlow, Frank, 1, 4 note, 7–8, 10, 218, 488–9, 515, 516, 520 note Barno de Glos, 62 Bartholomew, abbot of Marmoutier, 220 Bartholomew, archbishop of Tours, 190 Bartlett, Robert, 525 Basing (Hampshire), 239 Bath (Somerset), abbey of, 282 Battle (East Sussex), 238 abbey of, 29, 239, 385–6 Baudri of Bourgueil, 82, 213, 225, 244, 495 Bavarian annals, 231 Bayeux (Calvados), 83, 96, 97, 139, 144, 168, 178, 196, 199, 220, 224, 331, 472, 480 cathedral and chapter, 28, 59, 374, 395–6, 404, 411, 444–5. See Hugh, Odo, bishops of Bayeux Tapestry, 193–5, 197, 198–9, 200, 202, 210, 211, 213, 216, 217–19, 229–30, 234–5, 238, 241–3, 246, 445, 493, 521 Beaumont (-le-Roger) (Eure), 73 family of, 53, 73, 137. See Humphrey ‘de Vieilles’, Roger de Beaumont Beaumont-lès-Tours (Indre-et-Loire), abbey of, 224

index Beaumont-sur-Sarthe (Sarthe), 366 Beauvais (Oise), church of Saint-Quentin, 400 Bede, Historia Ecclesiastica, 336 Bedford, 464 Bedfordshire, 209, 267, 464 Bellême (Orne), 98, 133, 489 lords and lordship of, 31, 32, 70, 76, 101, 123–4, 142, 154–5, 397 See also Arnulf; Ivo, bishop of Sées; Ivo de Bellême; Mabel; Roger de Montgommery; Robert de Bellême; Seginfrid, bishop of Sées Benedict VIII, pope, 38 Benedict X, anti-pope, 337 Benedict, monk, founder of the abbey of Selby, 320 Benoît de Sainte-Maure, 354 Bénouville (Calvados), 436 Berengar of Tours, 103–4, 158–9, 205, 354, 441, 522 Berkshire, 464, 505 bernagium (ducal tax in Normandy), 129 Bernay (Eure), abbey of, 35, 73, 411, 414 Bertha, wife of Alan III, count of Brittany, and Hugh IV, count of Maine, 59, 81, 122, 178–80, 439 Beverley (East Riding of Yorkshire), 315 biography, writing, 1–5, 9–14 Biota, sister of Count Hugh IV of Maine and wife of Count Walter of the Vexin, 180 Bisson, Thomas, 299 Blanchelande (Sarthe), peace of, 436 Bleadon (Somerset), 290 Bleddyn ap Cynfyn, king of Gwynedd and Powys, 192, 282, 295 Blois-Chartres, counts of, 65, 70, 125, 180, 183 See Odo II, Stephen-Henry, Theobald III, Theobald IV, Bodmin (Cornwall), 465 Bohemond, son of Robert Guiscard, 9, 525–6 Bonne-Nouvelle (Seine-Maritime), church of, 168, 486 Bonneville-sur-Touques (Calvados), 152, 167, 199, 220, 301, 364, 499, 364, 365, 404, 422, 499 Boscherville (Seine-Maritime), 415, 416 Boulogne (Pas-de-Calais), 283 Bourne (Lincolnshire), 300, 504 Bramber (Sussex), 286, 416 Brampton (Cambridgeshire), 346

571

Brand, abbot of Peterborough, 248, 270–1, 323, 332 Breteuil-sur-Iton (Eure), castle of, 46, 137–8, 141, 152, 438. Bretons, people, 200, 231, 240–1, 321, 380 Brevis Relatio, 29, 46, 63, 213, 276, 278, 323, 487 Brian, earl of Cornwall, son of Count Eudo, 202, 224, 290, 308, 342 Brian fitz Count, 434 Brionne (Eure), castle and lordship of, 46, 63, 84, 86, 94–5, 123, 489, 518. See also Gilbert, Guy, counts of ecclesiastical council of (1050), 103–4 Briouze (Orne), 415 church of Saint-Gervase and SaintProthaise, 415 Bristol, 290, 432 University of, 6 British Isles, 192, 293, 309, 329, 342, 360, 362, 425–7, 429, 432–3, 508, 510–11 history of, 258, 509 Brittany, duchy and dukes of, 96, 142, 157, 166, 175, 179, 186, 188–9, 191, 192 note, 200–5, 235, 267, 342, 382, 383, 389–92, 429, 436, 453. See also Alan III, Alan IV, Conan II, Eudo, Geoffrey, Hoel, counts/ dukes of William’s 1064 campaign into, 280, 389. William’s 1076 campaign into, 389–92 Brixen (South Tyrol, Italy), synod of (25 June 1080), 417 ‘Broken Tower’, the, 248 Bruges (West Flanders), 115, 121 Bruno of Merseburg, 371 Brut y Twysogyon, 431 Brycheiniog, 343 Bures-sur-Dives (Calvados), 397 Burgundy, 44, 63, 74, 81–2, 87, 95, 132, 135, 454 dukes of, 126. See Robert I kingdom of, 33 See also Rainald, a count of Burnham Deepdale (Norfolk), 435 Burwell (Cambridgeshire), 347 Bury St Edmunds (Suffolk), abbey, 214, 286–7, 326, 358, 368, 429–30, 449, 500–01. See also Baldwin, abbot of Byzantine Empire and emperors, 55, 215, 468 Byzantium, influences of, 273

572

index

Caen (Calvados), 71, 83, 96, 106, 144, 157–61, 168, 206, 220, 222, 224, 226–7, 277, 278, 281, 301, 325, 404, 409, 414, 415, 416, 422, 436, 440, 445, 446, 447, 480, 481, 489–90, 500, 521 abbey of La Trinité, 121, 157–61, 169, 226–7, 230, 374–6, 385–6, 394, 413, 439, 442, 445, 447, 448, 495. See Cecilia, Matilda, abbesses of abbey of Saint-Etienne, 129, 157–61, 222, 278, 280–1, 340, 347, 358, 365, 373, 385–6, 394, 395–6, 408, 411, 412, 413, 445, 483–4, 489–90, 495, 514. See Lanfranc, William ‘Bona Anima’, abbots of church of Saint-Georges, 160 council at (before 6 March 1047), 77–8 Salle de l’Echiquier, 495 stone, 409, 445 University of, 9 Caesar Augustus, 468 Caesar (Julius), 200 Caesar, poetic motif, 253, 254 Calabria (Italy), 492 Cam, river, 348 Cambridge, 297, 347, 349, 382 Cambridgeshire, 209, 267, 378, 405–06, 448 Campbell, James, 508 canon law, 28, 54, 75, 100, 121 note, 135, 159, 176, 190, 210, 280 Canterbury (Kent), 114, 194, 195, 248, 493 abbey of St Augustine’s, 326, 404, 445, 458 archbishopric of, 112, 120, 7/32, 336, 337, 363–4, 424, 428, 505. See Æthelnoth, Dunstan, St, Lanfranc, Robert of Jumièges, Stigand, archbishops of Caradog ap Gruffydd, king of Upper Gwent, 343, 387, 430 Cardiff, 432 Carmen de Hastingae Proelio, 147, 193, 211, 217–18, 228, 230–2, 234–5, 238, 240–3, 245–6, 247–8, 250–1, 253–6, 293 Carolingian period and Empire, 8, 13, 15, 47, 207, 245, 282, 306, 421, 468, 497, 509 Cassel (Nord), Battle of, 340, 341 Castellum Vallium, peace of, 409–10

castles, 51, 62–3, 94, 123–4, 139, 154, 157, 174, 183, 268–9, 289–90, 293, 294, 297–300, 322, 493, 494, 509 siege-, 123, 131–2, 454–5 social impact of construction in England, 297–300, 324–5 Catherine, St, 54 cavalry, 181, 207, 240–2, 246 Céaux (Manche), 201–2 Cecilia, nun and later abbess of La Trinité of Caen, daughter of William and Matilda, 128, 227, 374–5, 449, 494–5 Cerisy-(la Forêt) (Manche), abbey of, 34, 74, 92, 159, 406, 413 Champagne, county of, 132 Channel, English, 2, 38, 40, 67, 77, 97, 112, 116–17, 126, 127, 129, 146, 186, 187–8, 192, 195, 197, 208, 225, 228–9, 230–34, 255, 275, 279, 281, 283, 284, 286, 288, 291, 292, 302, 325, 370, 384, 391, 395, 396, 416, 423, 428, 435, 442, 445, 446, 447, 453, 457, 471, 480, 487, 500, 507, 519 Channel Islands, 40, 144. See also Guernsey Chanson de Geste, 195 Charlcombe (Somerset), 282 Charlemagne, emperor, 13, 235, 252, 314, 317, 468. 492, 514, 522 Charles the Bald, emperor, 272 Charles the Simple, king of the Franks, 200 Chartres (Eure-et-Loir), 151, 179 abbey of Saint-Père, 153, 183 bishops and cathedral, 153, 331, 418, 438 Château-Gontier (Mayenne), 203 Châteauneuf-en-Thimerais (Eure-et-Loir), 150 Chepstow (Gwent), 300, 342–3 Cherbourg (Manche), 166, 168, 424 church of Notre-Dame, 166 Cheshire, 321, 344 Chester, 308, 321, 459 earldom of, 297, 431 Chibnall, Marjorie, 352 Chichester (West Sussex), 285 child rulership, 51, 54–7 Christendom, 104 Christina, sister of Edgar the Ætheling, 480 Chronica de Gestis Consulum Andegauorum, 30 Chronicle of Tours, 121

index Church, English, 120, 168–9, 170, 221, 325–6, 329, 334–9, 357–60, 376, 421–2, 460, 461, 493, 494, 509 Church, Norman, 47, 75–6, 91, 102–4, 137, 147, 162–3, 168–9, 174, 204–5, 207, 338, 354, 356, 411, 416, 421–2, 520 Church, Roman, 358, 368, 421 Cicero, 254 Civitate (Apulia, Italy), Battle of (1053), 107 Clare (Essex), 388 Claudius, emperor, Temple of (Colchester), 325 Clement III, anti-pope, 460–1 clerical celibacy/marriage, 205, 337, 356–7, 384, 433, 522 Cluny (Saône-et-Loire), abbey of, 329, 341, 371, 408, 419–20, 449, See also Hugh, St, abbot of Clyst Honiton (Devon), 289 Cnut, king of Denmark, England, and Norway, 20, 24, 28, 35–41, 65, 67–8, 111, 115, 264, 265, 278, 300, 322, 326–7, 337, 418, 504, 508. See also Estrith, sister of Cnut IV, St, king of Denmark, 274, 379, 457–9, 466. See also Olaf, brother of coinage, English and Norman, 206–7, 263, 271, 308–09, 501–02 Colchester (Essex), castle, 324–5 Collectio Lanfranci, 252 Colmont, river, 140 Cologne (North Rhine-Westphalia), 57, 147 Compiègne (Oise), 92, 402 abbey of Saint-Corneille, 419, 449 Conan II, count of Brittany, 50, 55, 61, 141, 143, 157, 179, 198, 200–3, 390 Conches-en-Ouche (Eure), abbey of, 73, 176 concubinage, 15–23, 30, 31, 36, 102, 115, 205, 356–7 Conisbrough (South Yorkshire), 297, 350 Conrad II, emperor, 33, 170, 252, 272. See also Gisela, wife of Constance, daughter of William and Matilda, wife of Alan IV, count of Brittany, 128, 439, 480 Constance, queen, wife of King Robert II of France, 33, 45 Constantine, emperor, 273, 409, 444

573

Arch of (Rome), 358 Constantinople, 101, 419 continuity, legal, of William’s regime in England, 68, 225, 258–9, 322, 323–4, 325, 339, 344, 364, 386–7, 430, 463, 469, 477, 491, 505 Copsig, earl of Northumbria, 267, 283, 284 Corbeil (Essonne), 156. See Malger, count of Corbonnais (Orne), 397 Cormeilles (Eure), abbey of, 176, 343 Cornouaille (Brittany), county of, 391 Cornwall, 288, 290, 440, 465. See Brian, Robert, count of Mortain, earls of Cotentin, the (Manche), 63, 82–3, 130, 143, 153, 156, 301, 417 Cottonian fire (1731), 354 Couesnon, river, 202 Coulombs (Eure-et-Loir), abbey of, 224 Courdemanche (Eure), 153 Coutances (Manche), 143–4 cathedral, 86, 143, 173. See Geoffrey, Robert, bishops of Coventry (Warwickshire), 4 Craon (Mayenne), 453, 455 Creully (Calvados), 96 crossbowmen, 240 Croth (Eure), 152 Crowland (Linolnshire), abbey of, 458 Cumbria, 360 Cuthbert, St, 305, 315, 362–3, 431, 434 Cynan ab Iago, claimant to the kingdom of Gwynedd, 192 Danegeld/geld, 265–6, 452, 463, 480, 504–8. See taxation Danelaw, 467 Danes, the, 265, 268, 311, 313, 318, 332, 333, 381, 382, 518 David, St, 431, 500–01 David, king of the Jews, 526 Davies, Rees, 510–11 de Boüard, Michel, 11, 160 De Corpore et Sanguine Domini, 158–9, 205 De libertate Beccensis monasterii, 411–12 De Obitu Willelmi, 25, 441, 482, 483–90, 492, 514, 522 Dean, Forest of (Gloucestershire), 311, 428 Deheubarth, kingdom of, 191–2. See Gruffudd ap Llewelyn, Rhys ap Tewdwr, kings of Denis, St, 146

574

index

Denmark, 36, 115, 199, 233, 263, 264, 324, 333, 341, 379, 382, 458, 459. See Cnut, Cnut IV, Harold Bluetooth, Harthacnut, Svein Estrithsson, Svein Forkbeard, kings of; Hamlet, prince of invasions of England from, 263, 303–04, 308, 309–13, 324–5, 332–4, 341, 373, 379, 382, 428, 451, 452, 455, 457–9 Devil, the, 222, 257, 260 Devon, 308 Dialogus de Scaccario, 274, 323 Diarmait mac Maél na mBó, king of Leinster, 432 Diego Paláez, bishop of Santiago, 461 Dieppe (Seine-Maritime), 284 Dijon (Côte-d’Or), abbey of SaintBénigne, 147 Dinan (Côtes-d’Armor), 202 Dives, river, 144–5, 220, 226, 228, 231–2, 436 Dives-sur-Mer (Calvados), 160 Dol (Ille-et-Vilaine), 202–3, 389–92 bishopric/archbishopric of, 390–2. See Evenus, Juhel, bishops of; Gilduin, candidate for bishop Domesday Book, 4, 5, 232, 236, 249, 258, 258–9, 286, 287, 312, 314, 316, 320, 321, 325, 355, 361, 370, 385, 407, 414, 415, 431, 433, 444, 448, 450, 456, 458, 462–78, 499, 504–06, 509, 521 making of, 465–73, 504 Little Domesday Book, 231, 287, 300, 406, 449, 468 Domesday Monachorum of Christ Church, Canterbury, 503 Domfront (Orne), 69, 91, 114, 133, 140, 141, 152, 189, 201, 346, 391 siege of, 122–6, 129 Dorchester-on-Thames (Oxfordshire), 285, 336 See Remigius, Wulfwig, bishops of Dorset, 308, 447 Douglas, David, 1–2, 4, 5–11, 14, 16 note, 28, 89, 91, 110, 145, 184–5, 261, 266, 389, 513, 514 Dover (Kent), 112, 198, 247–8, 268, 283, 309, 482 church of St Martin, 355 Downton (Wiltshire), 434, 442 Dreugesin, the, 152 Dreux (Eure-et-Loir), 153, 400

Drogo, count of the Vexin, 39, 45, 66, 69, 112, 130, 151, 180, 481. See also Godgifu, wife of; Ralph, Walter, sons of Dublin, 432 Duchesne, André, 354 Dudo of Saint-Quentin, 17, 45, 200, 273 Dunbar (East Lothian), 362 Duncan (Donnchad), king of Scots, 361 Dunstan, St, abbot of Glastonbury and archbishop of Canterbury, 344, 442. Durand, abbot of Saint-Martin of Troarn, 103 Durham, 25, 284, 303, 305, 315, 336, 362–3 bishopric of, 358, 427, 428 cathedral clergy, 315, 336 historical writing, 408 See Æthelric, Æthelwine, Walcher, William of St Calais, bishops of; Historia Regum; Symeon of Durham Düsseldorf (North Rhine-Westphalia), 57 Eadmer, 114, 117, 119, 194–6, 199, 213, 217, 285, 421–2, 483–90, 521, 522 Eadnoth the staller, 290 Eadric ‘the Wild’, 266, 282, 308, 362 Ealdred, abbot of Abingdon, 308 Ealdgyth, daughter of Earl Ælfgar and wife of Kings Gruffudd ap Llewelyn and Harold, 20, 191–2, 215, 419 note Ealdred, bishop of Worcester and archbishop of York, 7, 101, 109, 113, 119–20, 147, 214, 216, 248, 250, 253, 256, 258, 261, 270, 272, 284, 292, 307, 310–11, 327, 335, 339, 344, 518 Ealdred, brother of Siward son of Æthelgar, 266 earldoms, English, 110, 146, 233, 267, 268, 297, 334, 379, 396, 431 East Anglia, 287, 388 bishopric of, 338, 429. See Æthelmær, Herfast, bishops of; Elmham, Thetford, Norwich earldom, 146, 267. See Gyrth, Ralph the Staller, Ralph, earls of East Midlands, 209, 346, 350, 378 earldom in, 112, 209, 267 Ecclesiastes 10:16, 48, 55 Echauffour (Orne), 154 Edgar, king of the English, 326, 527

index Edgar the Ætheling, 3, 115, 117, 148–9, 182, 193–4, 197, 200, 213–14, 215–16, 248–9, 250, 255, 262, 267–8, 270, 276, 295, 297, 303–04, 310, 313, 323, 328, 329, 335, 360, 361–2, 369–71, 372, 381, 480, 501, 508, 524. See Christina and Margaret, St, sisters of Edith, queen of England, wife of King Edward the Confessor, 67, 68, 109, 110–11, 115–16, 146, 149, 209, 212, 250, 262–3, 383–4 Edith/Matilda, daughter of King Malcolm III and St Margaret, wife of King Henry I, 329, 426, 434, 508 Edith (Eadgifu) the Fair (‘Edith SwanNeck’), 20, 250, 288, 388 Edmund, St, 431 Edmund Ironside, king of England, 24, 67, 114, 117, 119, 147 Edmund, son of King Harold, 290 Edward the Confessor, king of England, 7, 7–8, 11, 24, 36, 45, 59, 88, 91, 92, 93, 97, 99–100, 118, 127, 146–50, 151, 180, 181–2, 192–4, 196, 199, 209–10, 214–15, 224, 230, 255, 258–9, 262, 264, 265, 269, 270, 282, 287, 290, 293, 323, 336, 362, 379, 491, 500, 505, 511 attitude to the English succession, 264, 517 crisis of 1051–2, 106, 108–22, 196, 210 deathbed arrangements for the succession, 212–15, 217, 487 exile in Normandy, 37–41, 65, 66, 68–9 relationship with William, 7–8, 22, 36, 41, 108–22, seal, 273–4 succession to the English kingdom in 1041–2, 49, 65–9, 270 See also, Edith, Queen, wife of Edward III, king of England, 90 Edward the Ætheling, ‘the Exile’, son of Edmund Ironside, 67, 114, 117, 119, 147–9. See also Edgar the Ætheling, son of; Christina, Margaret, St, daughters of Edwin, earl of Mercia, 20, 191, 209, 215, 216, 232–3, 248–50, 266, 267, 276, 295–6, 297, 306, 323, 330–1, 332, 334, 344–5, 346, 349–50, 350, 351, 354, 355, 380–1, 448, 518 Einhard, Life of Charlemagne, 492, 514, 522–4

575

Elmham (Norfolk), bishopric of, 358. See also East Anglia, bishopric of Ely (Cambridgeshire), 334, 344, 345, 349, 350, 357 abbey of, 335, 363, 364, 405, 406, 407, 433–4. See Thurstan, abbot of, 346–7 monks of, 346 See also Isle of Ely Emma, queen of England and mother of Edward the Confessor, 20, 24, 37, 38–9, 65, 66, 69, 115 Emma, daughter of William fitz Osbern, wife of Ralph, earl of the East Angles, 378 empire, cross-Channel, 77, 282, 304, 322, 342, 490–2, 496–9 Empire, the, 110, 125, 147, 170, 252, 342, 371, 408, 417, 459 Encomium Emmae Reginae,38, 66, 69 Engels, L. V., 218 English, the as a conquered people, 92, 217, 245, 248, 254–6, 258–9, 260–1, 262, 264–5, 274–5, 276–8, 285–8, 293–4, 294–6, 300, 303, 304, 321–3, 329, 332, 333, 351–4, 458, 493, 527 ‘exceptionalism’, 15, 258, 491–2, 509–11 language, 148, 269, 350–1, 430 William’s use of English troops, 288, 366, 401, 436 Enguerrand I, count of Ponthieu, 45, 61 Enguerrand II, count of Ponthieu, 79, 99–101, 130–2, 139, 154. See also Adelaide, daughter of Duke Robert, wife of Ermenfrid, bishop of Sion, papal legate, 135–6, 222, 279, 280, 334, 338, 353 Essex, 231, 458, 468 Estrith, wife of Duke Robert of Normandy, sister of King Cnut, 28, 35–7, 39, 111 Etymologiae of Isidore of Seville, 58 Eu (Seine-Maritime), county of, 46, 64, 198 castle of, 61, 85, 120, 128 comital family, 76. See Robert, Willkim, William Busac, counts of Eucharist, the, 204 Eudo dapifer, 467. See Adam fitz Hubert, brother of

576

index

Eudo, count of Brittany, 46, 56, 59, 96, 141–3, 157, 179, 201–2, 277, 290, 391. See Alan III, count, brother of; Alan Niger, Alan Rufus, Brian, sons of Eudo, son of Thurstan Haldup, 143, 154, 417. See also Muriel, wife of Europe, 17, 19, 37, 47, 68, 101, 102, 106, 119, 147, 159, 222, 236, 279, 291, 311, 346, 357, 371, 414, 418, 445, 449, 461, 468, 518, 519, 523 European history, 8, 9, 12, 14–15, 258, 352, 508, 509, 510, 511, 524–5 Eustace, count of Boulogne, 66, 99–101, 112, 131, 147–8, 188, 198, 224, 231, 242–4, 268, 283–4 Eustace de Breteuil, 22 Evenlode. See Agildeburga. Evenus, abbot of Saint-Melaine of Rennes and bishop of Dol, 390, 392–3 Evesham (Worcestershire), abbey of, 296, 315, 321, 336, 406–08, 503 See also Æthelwig, Walter, abbots of; Thomas of Marlborough Evrecin, the, 132–3, 152 Evreux (Eure), 18, 60, 481 cathedral, 395. See Gilbert, Hugh, William, bishops of counts of. See Richard, Robert, archbishop of Rouen, William Exe, river, 288 Exeter (Devon), 271, 281, 282, 288–90, 296, 297–8, 305, 306, 308, 518 See Leofric, William Warelwast, bishops of Exning (Suffolk), 378 Exon Domesday Book, 330, 475 See also Geld Rolls Falaise (Calvados), 26, 27–8, 31–2, 53, 62, 69–70, 138, 184, 495 Falkirk (Stirlingshire), agreement of (1080), 426 Farnham (Surrey), 249 Fauroux, Marie, 6, 40 Fawdon in Whaddon (Cambridgeshire), battle of (1075), 382, 383 Fécamp (Seine-Maritime), 26, 28, 45, 61, 85, 138, 146–7, 150–1, 167–8, 172, 187, 220, 276–7, 374, 376, 411, 446, 499, 500 abbey of, 19, 31, 35, 38, 40–1, 59, 60, 62, 78, 79 note, 80, 103, 136–7, 160, 168, 170, 188, 207, 226, 238, 273,

276–7, 285, 332, 409, 411, 413, 414, 448, 452, 456, 474, 503. See John, William de Rots, William of Volpiano, abbots of Fenland (Cambridgshire), 332 feud, 21, 35, 57, 60, 72, 74, 87, 155, 208, 255, 304, 346, 362, 387, 398, 408, 473, 515 ‘feudal revolution’, the, 15, 51 First Crusade, 222, 506, 525 ‘First English Empire’, 510–11 Flanders, 22, 24, 32, 64, 66, 86, 93–4, 100, 106, 113, 115–16, 122, 147, 196, 209, 229–30, 330, 340–2, 361–2, 362, 369, 382, 401, 429, 455, 458 counts of, 125. See Arnulf III, Baldwin IV, Baldwin V, Baldwin VI, Robert the Frisian, counts of Fleitel, family of, 76 Flemings, 321, 342 Flodoard of Rheims, 178 Folcard, abbot of Saint-Bertin, 148 Fontenay (Calvados), abbey of, 413, 414 Fontenoy (Yonne), Battle of (843), 7/30 Forth, river, 361, 501 Foss, river (Yorkshire), 296 Fossa Herluini, 29. See Herluin de Conteville Four-Shire Stone. See Agildeburga. France, 7, 9, 268, 292, 302, 322, 362, 370, 377, 401, 405, 417, 425, 433, 449, 451, 454, 473, 499, 503, 510 history of, 258, 508, 509 kingdom and kings of (Capetians), 20, 63, 70, 75, 119, 137, 204, 292, 293, 372, 397, 403, 435, 520–1. See Henry I, Philip I, Philip II Augustus, Robert II the Pious, kings of northern (usually referring to the political setting in which the dukes of Normandy operated), 18, 32, 34, 39, 47, 78, 87, 91–2, 99, 102, 106, 119, 125–7, 141, 150–7, 164, 176, 177, 182, 186, 187–91, 192, 194, 203, 206, 224–5, 230, 265, 329–30, 340, 342, 350, 370, 373, 378, 388, 393, 408, 409, 420, 429, 435, 438, 439, 452, 454, 479, 507, 521 Freckenham (Suffolk), 405 Frederick Barbarossa, emperor, 236 Frederick of Oosterzele-Scheldewindeke, 346

index Fredesendis, second wife of Herluin de Conteville, 128 French, the, 69, 140, 239, 256, 315 Fresnay-sur-Sarthe (Sarthe), 366 Frutolf of Michelsberg, 353 Fulcoius of Beauvais, 374–5, 526 Fulford (North Yorkshire), battle of (1066), 232–3, 267 Fulbert, father of Herleva, 27–8 Fulk II, bishop of Amiens, 150 Fulk II the Good, count of Anjou, 100 Fulk III Nerra, count of Anjou, 33, 34, 45, 70, 80 Fulk IV Rechin, count of Anjou, 179–80, 222, 265, 366, 372, 373, 376, 391, 392, 409, 423, 435, 435–8, 454 Galicia, kingdom of, 461. See also Alfonso VI, king of Gallic Wars, 254 Garnett, George, 254 Gascony, 94, 132, 511 Gateshead (Tyne and Wear), 408 Gâtinais, the, 132 Gebuin, archbishop of Lyons, 377, 410 geld. See danegeld. Geld Rolls, 330, 505 Geoffrey, bishop of Coutances, 76–7, 95, 101, 143–4, 163, 239, 256, 308, 356, 363–4, 364, 378, 382, 398, 404–05, 408, 410, 433, 465, 474, 488, 519, 524–5 Geoffrey, count of Brittany, 18 Geoffrey II Martel, count of Anjou, 56, 70, 81, 91, 93, 96, 98–9, 104, 114, 122–6, 127, 130–3, 140–5, 152–4, 157, 164–5, 177–80, 196, 201–2, 243 note, 481, 497, 518, 521 Geoffrey III the Bearded, count of Anjou, 165, 179–81, 183–4, 189–91, 202–3, 222, 265, 399, 437, 438 Geoffrey IV Plantagenet, count of Anjou, 499, 508 Geoffrey Grennonat, count of Rennes, 390, 392 Geoffrey de Chaumont-sur-Loire, 183, 224, 309, 400, 403 Geoffrey de Mayenne, 141, 154–5, 181, 185–6, 203 note, 304, 328, 366, 454 Geoffrey of Rheims, poet, 277, 439 Geoffrey, son of Rotrou, count of the Perche, 400

577

George, St, 273 Gerald dapifer, castellan of Neufmarché-enLyonnais, 151 Gerard Fleitel, 76 note Gerbod of Oosterzele-Scheldewindeke, earl of Chester, 297, 341, 420 Gerberoi (Oise), 400–1, 402, 403, 457 Germany, 371, 402 kingdom and kings of (Ottonian and Salian emperors), 135, 292, 293, 299, 371, 372, 402. See Otto I, Otto II, Otto III; Henry II, Henry III, Henry IV Gersendis, daughter of Herbert ‘Wake Dog’, count of Maine, 304, 366 See also Azzo d’Este, husband of; Hugh, son of Gervase du Château-du-Loir, bishop of Le Mans and archbishop of Rheims, 55, 98–9, 125, 140, 147, 161 note, 179 Gervin, abbot of Saint-Riquier, 291 Gesta Gaufridi, 144. See Geoffrey, bishop of Coutances Gesta Guillelmi. See William of Poitiers Gesta Herwardi, 44 note, 300, 345–9 Gesta Normannorum Ducum, 17, 25, 27, 49–50, 52, 57, 60, 129. See also De Obitu Willelmi, Orderic Vitalis, Robert of Torigni, William of Jumièges Ghent/Gent (East Flanders), abbey of Saint-Pierre du Mont-Blandin, 139 Gilbert, archdeacon of Lisieux and bishop of Evreux, 219, 357, 412, 490 Gilbert, count of Brionne, 35, 41, 46, 50, 53–4, 57, 58–9, 60–1, 63, 76, 80, 84, 86 children of, 64. See also Baldwin de Meulles, Richard fitz Gilbert, sons of Gilbert Crispin, abbot of Westminster, 107 Gilbert Crispin, castellan of Tillières-surAvre, 65, 151 Gilbert Foliot, abbot of St Peter’s, Gloucester, 426, 434 Gilbert Maminot, bishop of Lisieux, 396 Gilbert de Ghent, 297 Gilbert, son of Richard fitz Gilbert, 488 Gilduin, son of Rivallon de Dol and candidate for the bishopric of Dol, 390 Gillingham, John, 12, 258, 352, 502

578

index

Giroie, family of, 50, 124, 154–5, 175. See also Arnold d’Echauffour, Robert fitz Giroie Gisela, empress, wife of Conrad II, 150 Giso, bishop of Wells, 293, 442, 519 Gisors (Eure), 188, 388 Glamorgan, 387–8. See also Morgannwˆg Glasgow, University of, 5 Glastonbury (Somerset), abbey of, 442–3, 446–7, 476, 496. See Æthelnoth, Thurstan, abbots of Gloucester, 271, 428–9, 443, 456, 458, 462, 465, 466, 474, 478, 496, 499 abbey of St Peter, 428–9, 429, 496. See Serlo, abbot of crown-wearings, 429, 456, 458, 462, 465, 466, 474, 478, 493, 500. See also Robert, earl of Gloucestershire, 268 Godfrey, vicomte of Arques and Rouen (possibly), 71, 80, 97 Godgifu, sister of Edward the Confessor, 37, 39, 45, 66, 99–100, 112, 151, 180, 182. See also Drogo, count of the Vexin, Eustace, count of Boulogne, husbands of Godric, abbot of Winchcombe, 296 Godric, wife of, 464 Godwine, earl of Wessex, 37, 66–7, 93, 110–20, 127, 146, 286, 383, 507 family of, 111–18, 127, 182, 191. See Gytha, wife of; Gyrth, Harold, Leofwine, Svein, Tostig, Wulfnoth, sons of; Edith, queen, daughter of Godwine, son of King Harold, 290 Gorton, Neville, bishop of Coventry, 4 Goscelin of Saint-Bertin, 409 Goscelin, vicomte of Rouen, 71, 97 Gospatric, Northumbrian noble, 209 Gospatric son of Maldred, earl of Northumbria, 284–5, 303, 304, 310, 313, 336, 360, 362 Great Berkhamsted (Hertfordshire), 247, 250, 251 note, 266, 269 Gregory I, the Great, pope, 102 Gregory VII, pope, 101, 221–3, 354, 359, 388–9, 393–4, 412–13, 417–18, 420, 430, 444, 445, 457, 459, 459–61, 497, 515, 517, 524, 528 and Brittany, 390–1, 392–3, 421 and Maine, 377, 410–11, 421 and the conquest of 1066, 221–3, 327 relationship with Matilda, 368, 403

relationship with William, 221–3, 368, 372, 396, 402, 403–04, 416, 417–18, 420–2, 424–5, 430, 434, 444–5, 459–61, 497, 524 request to William for fealty, 373, 422, 423–5, 465 Gregorian mission to England (597), 335 Grestain (Eure), abbey of, 97, 128–9, 156, 413, 440, 520 Grimoult du Plessis-Grimoult, 82–3, 96, 369 Gruffudd ap Cynan, king of Gwynedd, 430–1, 432 Gruffudd ap Llewelyn, king of Gwynedd and Deheubarth, 146, 191–2, 282, 511 Gruffudd ap Rhydderch, king of Deheubarth, 192 Guazo, vassal of King Henry I of France and Hugh Bardolf, 152 Guernsey, island of, 136, 142 Guéry, Charles, 105 Guibert, archbishop of Ravenna. See Clement III, anti-pope Guibert of Nogent, 139–40, 185, 187, 375, 441–2 Guildford (Surrey), 249 Guitmund, lord of Moulins-la-Marche, 133, 138 Guitmund, monk of La Croix SaintLeuffroy, bishop of Aversa, 354, 412–13, 521, 524 Gundrada, wife of William de Warenne, 419, 420 Gundulf, monk of Le Bec, bishop of Rochester, 385, 405–06 Gunnor, concubine and wife of Count/ Duke Richard I, duchess of Normandy, 17–18, 19, 53, 143 family of, 64, 72 Guy, archdeacon and bishop of Amiens, 147, 193, 218, 234–5, 243, 248, 253, 256, 293. See also Carmen de Hastingae Proelio Guy, count of Brionne, 44, 63–4, 72 note, 81–4, 86–8, 94–5, 132, 240 note, 375, 518. See also Adeliza, mother of Guy, count of Ponthieu, 139, 147, 151, 188, 193, 195, 197–8, 200 Guy de Laval, 189–90 Guy-Geoffrey, brother of William VII of Aquitaine, 133

index Gwent, 343, 379, 430. See Caradog ap Gruffudd, king of Gwynedd, kingdom of, 146, 191–2, 282. See Bleddyn ap Cynfyn, Gruffudd ap Cynan, Gruffudd ap Llewelyn, Trahaern ap Caradog, kings of; Cynan ab Iago, claimant to Gyrth, earl of the East Angles, brother of King Harold, 115, 146, 237, 242, 277 Gytha, wife of Godwine, earl of Wessex, 37, 115, 217, 237, 245–6, 250, 288–90 Haimo ‘Dentatus’, 82–3, 96 Haimo, sheriff of Kent, 404 Hakon, son of Svein, son of Godwine, earl of Wessex, 114, 199 Halley’s Comet, 210 Hamlet, prince of Denmark, 7 Hampshire, 247 Hariulf of Saint-Riquier, 149, 194, 200 Harmondsworth (Middlesex), 306 Harnes (Pas-de-Calais), 148 Harold Bluetooth, king of Denmark, grandfather of King Cnut, 20 Harold I Harefoot, king of England, 20, 24, 41, 55, 65–6 Harold II, king of England, earl of East Anglia, earl of Wessex, 3, 37, 67, 110, 115, 119, 139, 146, 165, 182, 208, 209, 219, 224, 226, 231–3, 254, 266, 267, 268, 270, 282, 285, 286–7, 288, 290, 308, 330–1, 347, 348, 350, 371, 388, 431, 485, 498–9, 508, 509, 511 Battle of Hastings (1066), 234–42 Breton campaign (1064), 166, 188–9, 192 note, 198–9, 200–4, 389 coinage, 271, 308 concubine/partner of. See Edith (Eadgifu) the Fair coronation, 210, 216, 255 death and burial, 242–6, 348, 373 deathbed designation by Edward, 212–15, 217 English succession, 191–200, 210, 216 marriage to Ealdgyth, daughter of Earl Ælfgar of Mercia, 192 oath to Edgar the Ætheling, 149, 194 oath to William, 109, 164, 193–6, 198–200, 204, 217, 222, 239, 243–4, 254–5 rule as king, 212–17, 223, 235–7

579

sons of, 290, 308. See also Edmund, Godwine, Magnus treatment of reign after 1066, 269, 293, 469 visit to Flanders (1056), 139, 147–8 visit to Normandy (1064), 109, 127, 148, 164, 166, 188–9, 192–200, 215 See also Godwine, earl, father of; Gytha, mother of; Tostig, brother of Harold Hardraada, king of Norway, 193, 215, 224, 230, 232, 298, 310, 333 Harold, son of Earl Ralph and Gytha, 181–2 ‘Harrying of the North, the’, 313–22, 331, 360, 518 Harthacnut, king of England and Denmark, 20, 24, 51, 66–8, 115, 235 Harvey, Sally, 470 Hastings (East Sussex), 234, 236, 238, 247, 286, 391–2 battle of (1066), 29, 68, 96, 110, 162, 167, 186, 207, 211, 212–14, 217, 219, 221 note, 228, 234–47, 248–50, 253–5, 260, 265, 270, 275, 279, 281, 286, 286–7, 288, 293, 294, 295, 317, 332, 333, 385, 400, 458, 469, 490, 504, 509 Hawise, countess of Rennes, duchess of Brittany, daughter of count/duke Richard I, 59 Hawise, wife of count Hoel of Brittany, 439 Hayling Island (Hampshire), 306 Healfdene, thegn of King Edward, 330 Heather, Peter, 12 Hector, son of King Priam, 246 Hempage Wood (Hampshire), 409, 475 Henry, bishop of Winchester, 325, 439, 496 Henry III, emperor, 42, 56, 91, 93–4, 100, 112, 116, 118, 119, 131, 147, 292 Henry IV, emperor, 42, 55–6, 58, 89, 147, 223, 333, 371, 372, 417, 421, 423, 424, 425, 444, 457, 459, 459–60 Henry VI, emperor, 236 Henry I, king of England and duke of Normandy, son of William and Matilda, 21, 166, 171, 284, 291, 293, 301, 320, 383, 398, 399, 426, 452, 456, 474, 475, 484, 487, 490, 492, 495, 498, 499, 500, 501, 508, 517 Coronation Charter, 506 Pipe Roll, 506

580

index

Henry II, king of England, duke of Normandy and Aquitaine, and count of Anjou, 491 Henry III, king of England, 511 Henry I, king of France, 32, 40, 45, 49, 50, 53, 63–5, 69–70, 81, 83–4, 91, 92–3, 99, 104, 122, 125–6, 127, 130–3, 137–8, 140, 143–5, 150–3, 154, 156, 164, 175, 179, 181, 182, 187, 196, 372, 481, 497, 501, 518 court at Senlis (1048), 92–3 court at Vitry-aux-Loges (1052), 126 Henry II ‘the Quarrelsome’, duke of Bavaria and Carinthia, 55 Henry de Ferrières, 467 Henry of Huntingdon, 196, 244, 475, 491, 492–3, 494 Henry, son of Roger de Beaumont, 297 Herbert, bishop of Lisieux, 95 note, 99 note Herbert I ‘Wake Dog’, count of Maine, 303 Herbert II, count of Maine, 55, 56, 122, 164, 165–6, 177–81, 188, 198, 330. See Margaret, sister of Hereford, 282, 379. See Robert, Walter, bishops of; Ralph, Roger, William fitz Osbern, earls of Herefordshire, 282, 342, 379, 387 heregeld, 265–6, 505. See also taxation Hereward ‘the Wake’, 44, 300, 332–3, 346, 347, 348, 406, 504 Herfast, brother of duchess Gunnor, 53 Herfast, royal chancellor and bishop of East Anglia, 338, 368, 429–30 Herleva, William’s mother, 16, 17, 18, 25–30, 34, 36, 46, 56, 63, 88, 97, 120, 124, 128, 138, 143, 155, 173, 520. See also Herluin de Conteville, husband of; Odo, bishop of Bayeux, Robert, count of Mortain, sons of; Adelaide, Muriel, daughters of; Walter, brother of Herluin, abbot of Le Bec, 57, 76, 107, 411 Herluin de Conteville, William’s stepfather, 26, 28, 29, 36, 97, 120, 128, 143, 155, 168, 520 Herluin, the knight who organised William’s body for burial, 489 Herman, bishop of Ramsbury, 101 Herman, monk of Saint-Evroult, 175 Herman the Archdeacon, 214 Hermann of Salm, German anti-king, 459 Herod Agrippa, 523

Hertfordshire, 247, 330, 370 Hexham (Northumberland), 313 Hiémois (Orne), the, 26, 35, 60, 62, 69, 83, 144, 154 Hildebrand, archdeacon, 368. See Gregory VII, pope Historia Regum (Durham), 284, 315, 336, 360, 361, 362, 426, 488 Hitler, Adolf, 5 Hoel, bishop of Le Mans, 437–8 Hoel, count of Brittany, 390, 391, 392, 436, 439, 453. See also Hawise, wife of Holt, Sir James, 478 Home Counties, 440 Honfleur (Calvados), 168 hostages, 110, 112, 114, 116–17, 119, 127, 180, 194, 196, 198–9, 224, 250, 264, 276, 289, 296, 437 Hubert, papal legate, 338, 358–9, 392, 403, 412, 420, 424 Hubert de Ryes, 83, 357. See Robert, bishop of Sées, son of Hubert de Sainte-Suzanne, vicomte of Le Mans, 181, 366, 453–5, 466, 518 Hugh, St, abbot of Cluny, 325–6, 371, 393, 403, 408, 419–20 Hugh, bishop of Avranches, 137, 173 Hugh, bishop of Bayeux, 46, 59, 80, 97, 138, 173 Hugh, bishop of Die and archbishop of Lyons, papal legate, 392, 393, 444 Hugh, bishop of Evreux, 75 Hugh, bishop of Langres, 417, 419 Hugh, bishop of Lisieux, 76, 99 note, 176, 177, 355, 396 Hugh IV, count of Maine, 80–1, 85, 122, 178–80 Hugh, earl of Chester, son of Richard Goz, 341–2, 369, 402, 432, 441, 444 Hugh, ducal forester, 172 Hugh, heir to the county of Ponthieu, 243 Hugh Bardolf, 132, 152 Hugh Bunel (also de Ialgeio), 397–8 Hugh Candidus, cardinal priest of San Clemente, 460 Hugh de Châteauneuf-en-Thimerais, 399–400 Hugh de Gournay, 133, 402 Hugh de Grandmesnil, 98, 151, 174, 402. See Robert de Grandmesnil, abbot of Saint-Evroult, brother of

index Hugh I de Montfort (-sur-Risle), 50, 52–3, 60 Hugh II de Montfort (-sur-Risle), son of Hugh I, 60, 133,137, 404 Hugh fitz Baldric, sheriff of Yorkshire, 319–20 Hugh of Flavigny, 74, 490 Hugh of Fleury, 82, 515–16 Hugh, son of Azzo, lord of Este, and Gersendis, claimant to the county of Maine, 304, 366 Hugh the Chanter, 312, 315, 338 Humber, river, 285, 296, 299, 310–13, 317, 318, 320, 332, 382, 444 Humbert of Moyenmoutier, 101 Humphrey de ‘Vieilles’, 50, 60, 72, 73–4, 80, 169 Humphrey du Tilleul, 286 Hundred Years War, 511 Hungary, 67, 119, 147 Huntingdon (Cambridgeshire), 297, 347 Huntingdonshire, 209, 267 Hurst peninsula (Hampshire), 67 Ilbert de Lacy, 350 Ingelric the priest, 287, 293 Ingibjorg, first wife of Malcolm III, king of Scots, 361 Inquisitio Comitatus Cantabrigiensis, 471 Inquisitio Eliensis, 471 Institute of Historical Research, 6 Inventio et Miraculi sancti Uulfranni, 38–9, 53–4, 118, 129 Ipswich (Suffolk), 271 note, 309 Ireland, 115, 192, 229, 288, 308, 309, 430, 432, 493 Irish Sea, 192, 290, 432, 458 Isembard, abbot of La Trinité-du-Mont of Rouen, 126 Isidore of Seville, 58 Isle of Ely (Cambridgeshire), 332, 345–8 Isle of Wight, 67, 217, 224, 231, 443, 479 Isleham (Cambridgeshire), 405, 471 Israelites, 488 Istria, 468 Italy, 135, 136, 304, 354, 417, 457 southern, Norman settlers in, 34, 47, 101, 102, 106, 138, 143, 156, 175, 177, 243, 370, 444, 501, 524–5 Ivo (of Bellême), bishop of Sées, 70, 76, 80, 85,98, 101, 133, 142, 144, 155, 357, 420, 524–5 Ivo, lord of Bellême, 80 Ivo Taillebois, 347

581

Ivry-la-Bataille (Eure), 46, 152, 324 Iznik (Turkey). See Nicaea. Jarrow (Tyne and Wear), monastery, 336 Jephthah, biblical king, 226, 374–5 Jerusalem, 25 note, 34, 42, 45, 47, 119, 375, 417, 419 jester, court, 523 John, abbot of Fécamp, 19, 23, 88, 102–3, 106, 137, 146–7, 168, 226, 354, 411, 525 John, bishop of Avranches and archbishop of Rouen, 173–4, 189, 279, 290, 301, 356, 357, 367, 374, 384, 388, 396, 412 John, king of England, duke of Normandy and Aquitaine, and count of Anjou, 510 John de La Flèche, 392 John Minutus, cardinal-priest of S. Maria in Trastavere, 334, 340 John of Worcester, 113, 116, 117, 147, 148, 171, 182, 192, 213, 216, 233, 237, 247, 249–50, 256, 282, 288, 310, 313, 332, 334, 335, 337, 345, 349, 360, 362, 378, 381, 382, 386, 389–90, 401, 446, 452, 458, 464, 475, 480, 483, 485–6, 504 John, son of Herluin de Conteville and Fredesendis, 128 John, son of Rivallon de Dol, 389 Judith, daughter of Count Robert II of the Auvergne, 393 Judith, half-sister of Baldwin V of Flanders and wife of Earl Tostig, 115–16 Judith, niece of William and Matilda, wife of Earl Waltheof, 350, 378 Judith, wife of Duke Richard II, 18, 19, 35 Juhel, bishop of Dol, 390–1, 392–3 Juhel, prior and abbot of Saint-Pierre de La Couture of Le Mans, 377, 410, 421 Jumièges (Seine-Maritime), abbey of, 98, 160, 207, 277, 301, 306, 394, 413, 414, 449. See Robert of, archbishop of Canterbury; William of Kaiserwerth (North Rhine-Westphalia), coup of, 57 Karl, member of the house of Bamburgh, 304 Keats-Rohan, Katharine, 498 Keble College, Oxford, 6 note

582

index

Kent, 167, 226, 234, 247, 254, 287, 363, 444 earldom of, 268, 445. See also Leofwine, Odo, Bishop of Bayeux, earls of shire court, 363–4; see Penenden Heath Kentford (Suffolk), 406, 407 King Lear, 7 L’Aigle (Orne), 397, 399 La Croix-Saint-Leuffroy (Eure), abbey of, 354, 412. See Guitmund, monk of La Chaise-Dieu (Haute-Loire), abbey of, 126, 419 La Flèche (Sarthe), 376, 392, 435–6. See John de La Hougue (Manche), 166, 172, 177, 183 La Roche-Mabille (Orne), castle of, 154–5 Lacock (Wiltshire), 474 Lambert II, count of Lens, 100, 131, 139, 350 Lambert of Saint-Omer, 22 Lambeth (London), 67 Lampert of Hersfeld, 56, 371 Lanfranc, monk of Le Bec, abbot of Saint-Etienne of Caen, archbishop of Canterbury, 76, 101, 103, 137, 163, 171, 175–6, 189, 205, 252 note, 255, 285, 323, 329, 337, 344, 347, 353, 354, 356, 367, 376, 385, 395, 404, 406, 407, 411, 417, 433, 435, 442, 443, 458, 465, 474 abbot of Saint-Etienne of Caen, 158, 170–1, 280–1, 340 archbishop of Canterbury, 338–9, 344, 353, 357–9, 363–4, 428, 432, 434, 442–3, on Englishwomen who had taken refuge in nunneries, 323, 353 personal relationship with Archbishop John of Rouen, 356, 367 personal relationship with William, 107–8, 158–9, 175–6, 280–1, 329, 338–9, 347, 353, 357–9, 377–83, 389, 417, 433, 458, 460–1, 465, 478, 519, 523–4, 526 relations with popes, 101, 158–9, 165, 280–1, 339, 344, 357–9, 368, 417, 420, 424–5, 460–1 refuses the archbishopric of Rouen, 280 Revolt of the Three Earls, 377–83, 386, 389 William II’s succession, 486–7, 488 Laon (Aisne), 182

Lastingham (North Yorkshire), 319 Latin, language, 351, 429, 430 Laudes Regiae, 273, 292, 501 Launceston (Cornwall), 465 Laval (Mayenne), castle of, 189 Laws of Edward the Confessor, 114, 505, 506 Le Bec (Eure), abbey of, 76, 101, 107, 128, 158–9, 168, 170, 207, 252 note, 331, 340, 385, 394, 395, 400, 405, 411, 412, 413, 414, 448, 471. See Anselm, Herluin, abbots of; Lanfranc, monk and prior of; Arnost, Gundulf, William ‘Bona Anima’, monks of ‘Le Camp de Beugy’, 454–5 Le Homme (now, L’Isle Marie, Manche), castle of, 63, 83, 95, 375–6 Le Mans (Sarthe), 151, 178, 180, 182–6, 189–91, 304, 328, 366, 389, 410, 437, 453, 492 abbey of Saint-Pierre de la Couture, 189–90, 201, 498 abbey of Saint-Vincent, 365, 409, 427 453. See also William of SaintCalais, abbot of cathedral and chapter, 437–8. See Arnold, Hoel, Wulgrin, bishops of chapel of Saint-Pierre-de-la-Cour, 437, 438 Le Patourel, John, 297 Le Sap (Orne), 61 Le Vaudreuil (Seine-Maritime), 277, 280 Leeds (South Yorkshire), 312 University of, 5 legates, papal, 135–6, 147, 176, 190, 212, 279, 280, 334–8, 340, 356, 358, 392–3, 403, 412, 417–18, 420–1, 425, 444, 460, 522 Leges Edwardi Confessoris. See Laws of Edward the Confessor Leicestershire. 427, 438 Leo I, pope, 102 Leo IX, pope, 23, 91, 99–102, 106–7, 110, 135–6, 147 Leofric, abbot of Peterborough, 270, 333 Leofric, bishop of Exeter, 306 Leofric, earl of Mercia, 110, 117, 146, 332 Leofwine, bishop of Lichfield, 335, 343 Leofwine, earl of Kent, brother of King Harold, 115, 146, 242, 268 Les Préaux (Eure), 73 abbey of Notre-Dame and Saint-Léger, 169, 330, 376, 413. See Ansfrida, abbess of

index abbey of Saint-Pierre, 42, 61, 77, 103, 413. See Ansfrid, abbot of Lescelin, bishop of Paris, 150 Lessay (Manche), abbey of, 404, 413, 416–17, 452 Lewes (East Sussex), 286, 419, 433 abbey of St Pancras, 419 Liber Eliensis, 335, 345, 346, 347, 348, 354, 503–04, 526 Liber Vitae of New Minster, Winchester, 149 Lichfield (Staffordshire), bishopric of, 338, 344 See Leofwine, Peter, bishops of Licquet, François-Théodore, 3, 513 Liguria (Italy), 366 Lille (Nord), 131 Lillebonne (Seine-Maritime), 176, 177, 206, 219, 369, 416, 500 ecclesiastical council of (1080), 416, 440 Lincoln, 297, 308, 310, 336, 358 See Remigius, bishop of Lincolnshire, 304, 382 Lindisfarne (Northumberland), 315 Lindsey, 311, 312 Lisieux (Calvados), 76, 144 abbey of Saint-Désir, 301, 302, 395, 396 cathedral, 354, 396. See Herbert, Hugh, Gilbert, bishops of council of (1064), 204–5, 222–3, 356 synod of (1055), 135–6 Little Thetford (Cambridgeshire), 348 Loir, river, 436 Loire, river, 70, 93, 201, 309, 372, 415, 416 valley, 156, 161, 172 Lombardy, 460 London, 38, 167, 232, 237, 248, 249–51, 253, 266, 268–9, 269, 271, 284, 308, 308–09, 325, 363, 405 Baynard’s Castle, 269 Blackfriars, 251 bishopric of, 459. See Maurice, bishop of collegiate church of St Martin-le-Grand, 293 council of (1075), 376 Holborn, 251 London Bridge, 249 Ludgate, 251 Montfichet’s Castle, 269 St Paul’s cathedral, 269 White Tower, 268–9, 324, 385, 405 William’s writ for (1067), 269, 271–2 Lonlay (Orne), abbey of, 415 Lotharingia, 336, 468

583

Louis the Pious, emperor, 235 Ludgate/Fleet Valley excavation, 251 Lulach, king of Scots, 361 Lyons (Rhône), archbishopric of, 421. See Gebuin, Hugh, archbishops of Lyons-la-Forêt (Eure), 277 Lyre (Eure) abbey of Notre-Dame, 105, 343 annals of, 76 Mabel (of Bellême), wife of Roger de Montgommery, 98, 155, 156, 174–5, 357, 397–8, 414. See William Talvas, father of Macbeth, king of Scots, 361 Maerleswein, English aristocrat, 304, 310 Magna Carta, 510 Magnus, son of King Harold, 290 Main de Fougères, 142, 157 Maine, county of, 14, 24, 33, 59, 80, 91, 98–9, 114, 122–3, 125, 140–1, 144, 157, 162, 165, 175, 178, 181, 183–6, 203, 225, 265, 303–04, 328, 329, 340, 356, 364–7, 372, 376–7, 389, 392, 396, 399, 409–11, 418, 423, 427, 435, 435–8, 439, 442, 453–5, 486, 491, 496, 498–9, 508. See Herbert I ‘Wake Dog’, Herbert II, Hugh IV, Robert, William, counts of. See also Manceaux Mainer, abbot of Saint-Evroult, 170 Mainz (Rhineland-Palatinate), 315, 371 Malcolm (Mael Coluim) III, king of Scots, 224, 262, 295, 297, 299, 313, 329, 335, 360–2, 369–70, 408, 423, 426–7, 434, 436, 508 See Ingibjorg, Margaret, St, wives of Malger, archbishop of Rouen, 19 note, 60, 64, 71, 75, 76 note, 77, 81, 86, 99, 108, 137, 156, 204, 517 deposition of, 134–7, 142, 161 Malmesbury (Wiltshire), abbey of, 429, 449 Manceaux, 189, 321, 356, 365, 427 manor (manerium), 473 Mantes-la-Jolie (Yvelines), 481–2, 483 Marbod, bishop of Rennes, 445 Margaret, St, queen of Scots, 262, 329, 360, 426, 434, 508 Margaret, sister of count Herbert of Maine, 188 Marianus Scotus, 315 Marmoutier (Indre-et-Loire), abbey of, 142–3, 157, 161, 172, 183, 189–90, 201–2, 224

584

index

372, 377, 418–19, 420, 423–4, 427–8, 436, 440, 449 Marsh Gibbon (Buckinghamshire), 470 Matilda, abbess of La Trinité, Caen, 169, 227–8 Matilda, countess of Tuscany, 457 Matilda, daughter of Roger de Montgommery and Mabel de Bellême, 156, 440. See Robert, count of Mortain, husband of Matilda, daughter of William and Matilda, 128 Matilda, duchess of Normandy and queen of England, 11 note, 91, 94, 161, 165–6, 166–7, 171, 188, 206, 251, 253, 275, 277, 293, 301, 302, 303, 306, 307–8, 320, 325, 329, 330, 331, 340, 358, 367, 368, 369, 374–5, 383–4, 388–9, 393–4, 399, 402, 402–03, 416, 421, 423, 424, 426, 439, 442, 444, 446, 447–50, 453, 456, 486, 489, 495, 501, 518–19, 523, 524 birth of children, 128, 163, 275, 277, 293, 320 coronation, 253, 291–2, 295 final illness and death, 423, 447 gift of the ‘Mora’ in 1066, 226 letters from Gregory VII, 368, 403 marriage to William, 94, 99–108, 113, 115–22, 125–6, 135, 157–9, 162, 165, 167–9, 262–3, 439 partnership in rule with William in Normandy and England, 159, 163, 168, 169–70, 171, 206, 220–1, 251, 253, 284, 291–2, 301–2, 307–8, 340, 358, 367, 388–9, 399, 424, 428, 433–5, 448 personal relationship with William, 166–7, 251, 275, 428, 429 relations with children, 171, 227, 301–2, 330, 331, 374–5, 402, 449 religious patronage, 157–8, 168, 169, 226–7, 329, 372, 411, 418–19, 429, 447–9 tomb, 169, 228, 447 Matilda, Empress, daughter of Henry I, 82, 434, 491, 499, 508, 509, 515 Maurice, St, 273 Maurice, bishop of London, 481 Maurilius, archbishop of Rouen, 135–7, 142, 173, 176, 204, 280

Mayenne (Mayenne), 140, 233, 482, 518 castle of, 185–6 river, 140 Mayr-Harting, Henry, 522 medieval warrior kingship, 12, 225, 497, 526 Mercia, earldom, 191, 296, 321. See Ælfgar, Edwin, Leofric, earls of ‘mercy’ (misericordia), of a ruler, 13, 255, 261 Michael, son of Archbishop Malger of Rouen, 134 Middlesex, 247 Milan (Lombardy), 236 Milford-on-Sea (Hampshire), 67 Miracles of St Edmund, 214 monetagium, 507 Monmouth (Gwent), 429 Mont Saint-Michel (Manche), abbey of, 41, 66, 80, 135, 136–7, 142, 173, 177, 201, 207, 390, 413, 428, 503 Montecassino (Lazio), abbey of, 243 Montgommery (Sainte-Foy and SaintGermain de Montgommery, Calvados), castle of, 50 family of, 50, 61 note, 70. See Roger I de, Roger II de Montgommery Montivilliers (Seine-Maritime), abbey of, 34, 45, 160, 413 Montreuil (Pas-de-Calais), 370 Montreuil-l’Argillé (Eure), 154 ‘Mora’, William’s flagship, 226 Morcar, earl of Northumbria, 20, 209, 215, 216, 232–3, 248–50, 266, 267, 276, 283, 295–6, 306, 323, 332, 334, 344–5, 346, 349, 350, 351, 354, 355, 380–1, 485–6 Morgannwˆg, 430. See also Glamorgan Mortagne (Orne), 175, 400 Mortain (Manche) 440, 454. See Robert, William Werlenc, counts of church of Notre-Dame, 440 collegiate church of Saint-Evroult, 413, 440 Mortemer (Seine-Maritime), 132–3 battle of (1054), 130, 134, 136, 137, 138–41, 155, 162, 187 castle of, 138 See also Roger de Moslems, 73, 222 Mouliherne (Maine-et-Loire), campaign, 93, 122 Mulcheney (Somerset), abbey of, 442 murdrum fine, 274

index Muriel, daughter of Herluin de Conteville and Herleva, 143, 154, 417. See also Eudo Haldup, husband of Mynydd Carn, battle of (1081), 430, 431 Nantes (Loire-Atlantique), county of, 391 Neaufles-(Saint-Martin) (Eure), castle of, 151 Neuf-Marché-en-Lyonnais (SeineMaritime), castle of, 151–2 New Forest (Hampshire), 325, 330, 494, 500 Newcastle-on-Tyne (Tyne and Wear), 426 Nicaea (now Iznik), 47 Nicaise, St, relics of, 34 Nicholas, abbot of Saint-Ouen of Rouen, son of Duke Richard III, 31, 86, 365 Nicholas II, pope, 55, 158–9, 176, 205, 272 Nigel, the vicomte, 59, 64, 71, 76, 82–4, 87, 95–6, 142–3, 152–3, 156, 162, 175 Nigel, son of Nigel the vicomte, 64, 71 note, 72 Norfolk, 468 Norman annals, 135–6, 182, 448–9 Normandy, 6, 17, 18, 19, 21, 22, 24, 25, 29, 32, 33, 35, 59–61, 63–5, 65–6, 67, 68, 70–2, 73–5, 79, 80, 82–5, 87, 92, 94–7, 100, 102, 104, 105, 106–7, 109, 110, 111, 112, 114–15, 117–19, 127, 128, 130–3, 137, 143–5, 148, 151, 152, 153, 157, 159, 164, 167, 169, 171, 174, 177–8, 180, 181, 185–6, 189, 190, 191, 193, 195–9, 203–4, 215–17, 220, 222–5, 230, 231, 237, 240, 251–2, 254, 260, 266, 268, 269, 273, 284, 286, 291, 292, 297, 303, 304, 308, 315, 322, 325, 327, 329, 334, 338, 340, 342, 343, 356, 472, 480, 492, 496–7, 503, 511 aristocratic elite, 20, 97, arrangements for rule during William’s absences from 1066 onwards, 221, 284, 285–6, 307–8, 340–1, borders/frontiers, 26, 27–8, 31–2, 64, 79, 80–1, 98–9, 122–5, 137–40, 150–2, 154–7, 180, 188, 200–2, 280, 389–94, 397–8, 438–9, 480–2 church dedications in, 143, 204, 226–8, 277, 395–6, 399–401 episcopate, 75–8, 99, 101, 143, 173–4, 176, 189, 270, 357

585

history and historiography of the duchy, 8–9, 15, 18, 49–54, 57, 87–8, 161–3, 169–70, 178, 206–8, 230, 258, 355, 509 monasticism, 19, 34, 80, 98, William’s visits to after 1066, 188, 262, 275–81, 294, 296, 301–3, 340–3, 365–8, Normans, people, 5–6, 19, 47, 53, 68–9, 81–2,88, 92, 102, 104, 106–7, 140, 172, 185, 218–19, 224–5, 231, 239, 241, 251, 257, 259, 260, 270, 272–4, 276, 294, 297, 298, 309, 310, 321, 329, 338, 342, 348, 351–2, 353 356, 492, 493, 527 diaspora of, 524–5 North or South Pickenham (Norfolk), 470 North Sea, 20, 38, 116, 229, 230, 264, 265, 333, 341, 370, 379, 466 North Wales, 430–1, 432 Northallerton (North Yorkshire), 305 Northam (Devon), 308 Northamptonshire, 209, 267, 348 Northamptonshire Geld Roll, 504–05 Northumbria, 408, 479 earldom, 119, 146, 191, 209, 233, 427 Norway, kingdom of, 199, 233, 333 king of (Olaf III Haroldson), 457 Norwich, 271, 298, 300, 309, 378, 382, 383 bishopric of, 358. See also East Anglia, bishopric of Notre-Dame du Pré (Seine-Maritime), priory of, 168. See also BonneNouvelle, church of Nottingham, 296, 297, 312 Nouy (Indre-et-Loire), battle of, 70 nunneries, 323 Odense (Denmark), 459 Odo, bishop of Bayeux and earl of Kent, 26, 28, 44, 97, 189, 220, 227, 239, 242, 268, 274, 281, 283, 338, 344, 350, 356, 357, 363–4, 369, 378, 382, 395, 405–08, 410, 411, 417, 423, 424, 425–6, 427, 434, 441–2, 456, 460, 467, 471, 485, 488, 490, 505, 519, 525 Odo II, count of Blois-Chartres, 33, 59, 81 Odo, nephew of Count Theobald III of Blois-Chartres, third husband of William’s sister Adelaide, 187 Oissel (Seine-Maritime), island of, 402, 416, 439, 442

586

index

Olaf, brother of Cnut IV, king of Denmark, 459 Old Sarum. See Salisbury Orderic Vitalis, 49–50, 51–4, 59–63, 69, 70, 72, 73, 79, 81, 85, 86–8, 98, 106, 114, 118–19, 121, 122, 124, 129–30, 133, 134, 138, 139, 142, 150, 152, 154–6, 157, 162, 167, 169–71, 174–6, 177, 179, 181–2, 184, 188–9, 193–4, 198–9, 200, 203, 208, 214, 216, 219–21, 223, 224 note, 225, 229, 233, 237, 251, 256–7, 260, 263, 265, 266, 268, 276, 277, 278, 280, 281, 284, 288, 289, 293, 295, 297, 298–9, 304–05, 307–08, 309, 311–13, 313, 315, 317, 318, 321, 322, 323, 329, 330, 331, 334, 336, 340, 341, 342, 343, 345, 349, 351–3, 354, 355, 365, 367, 370, 381, 383, 386–7, 391, 393, 395, 396, 396–402, 412, 414–15, 416, 430, 436–7, 441, 443, 444, 447, 448, 453–5, 456–7, 463, 478,479, 480, 481–2, 483–90, 504, 513, 517, 521, 527 moral and historical perspective on William’s acts and rule, 3, 4–5, 29–30, 150, 152, 154–6, 174–6, 184–5, 208, 256–7, 260, 263, 265, 268, 276–7, 278, 281, 284, 308–9, 313–14, 315, 318, 322, 323, 342, 345, 349–50, 351–3, 354, 355, 381, 484, 527 on William’s adolescence, 49–54, 59–63, 69–73, 79, 86–8 on William’s birth and parentage, 4, 17, 21, 22, 29–30, 81 Orléans (Loiret), 32–3, 125, 126, 190 Orne, river, 84, 85 note, 436 Osbern, abbot of Saint-Evroult, 176–7 Osbern, brother of William fitz Osbern and chaplain of Edward the Confessor, bishop of Exeter, 97, 112 Osbern the steward, 50, 53, 62, 63, 64, 72 note, 138. See Osbern, William fitz Osbern, sons of Osbjorn, earl, brother of Svein Estrithsson, king of Denmark, 309–10, 313 Osmund, bishop of Salisbury, 171, 417, 465 Osulf, son of Eadulf of Bamburgh, earl of Northumbria, 267, 283, 284

Oswald, St, archbishop of York and bishop of Worcester, 344 Otto I ‘the Great’, emperor, 253 Otto II, emperor, 55 Otto III, emperor, 55, 58 Otto the Goldsmith, 490 Ottonian period, 421 Ouen, St, 367 Ouistreham (Calvados), 278 Ouse, river (Cambridgeshire), 347 Ouse, river (Yorkshire), 232, 296, 298, 305, 311, 319–20, 382 Oxford, 298, Pacy-sur-Eure (Eure), 481 pancartes, Norman, 96, 105, 129, 173, 226, 413–14, 416–17, 430, 440, 471–2 papacy, the, 99, 101–2, 104–7, 121, 127, 135–6, 158–9, 161, 191, 205, 221–3, 279, 280, 335, 337, 338, 358, 359, 367, 368, 394, 441, 444, 459, 461, 509, 515, 522 papal banner, 219–22, 334, 337 papal legates. See legates Papia, wife and possible concubine of Duke Richard II, 19, 64 Paris, 39, 62, 65, 83, 132, 151, 153 note, 156, 187, 392 Pays d’Auge, 144 Penenden Heath, near Maidstone (Kent), land plea, 363–4, 407, 477 Penitential Ordinance, the, 222, 279, 322, 353 Pennines, the, 318, 321, 332 Peter, St, 149, 307 Peter, bishop of Lichfield, 338 Peter, probably cardinal-priest of St Chrysogonus, 334, 340 Peter de Valognes, sheriff of Hertfordshire, 330 Peter’s Pence, 335, 424, 461 Peterborough (Northamptonshire), abbey of, 248, 332–3, 346 See Brand, Leofric, Turold, abbots of Petherton, North or South (Somerset), 343 Pevensey (East Sussex), 234, 286 Philip I, king of France, 55, 58, 153, 188, 190, 265, 341, 370, 372, 373, 376, 377, 389, 391, 392, 393, 394, 400–01, 402–03, 454, 481–2, 501 Philip II Augustus, king of France, 490 Picardy, 66, 149, 370, 427, 490 Picot, sheriff of Cambridgeshire, 405–06, 471

index Plympton (Devon), 465 poisoning, 47, 50, 61, 155, 175, 184–5, 203, 208 Poissy (Yvelines), 83, 153 note Poitiers (Vienne), 391 Poitou, province of, 224, 231, 391 counts of, 180 Polyptych of Irminon, 468–9 polyptychs, 468–9 Pont-Audemer (Eure), 73 Pontefract (South Yorkshire), 312, 350. Ponthieu, county of, 120, 148, 194, 196–8, 230, 243, 456, 484. See Enguerrand I, Enguerrand II, Guy, counts of; Hugh, heir to Portsmouth (Hampshire), 301, 384, 436, 481 Powys, kingdom of, 192, 343 Priam, king, 246 Primacy Dispute/Agreement, 338–9, 344, 357–9, 428, 519, 520 Pseudo-Isidorean Decretals, 251–2, 255, 443 publicum bellum, 222 Quadripartitus, 67 Quaedam Exceptiones, 26 Rabel, commander of Duke Robert’s fleet, 27 Radbod, bishop of Sées, 32, 420 Rainald, abbot of Saint-Pierre de La Couture of Le Mans, 377, 410, 421 Rainald, a count in Burgundy, 63, 63–4 note, 375. See Adeliza, wife of; Guy, son of Rainald, royal chaplain, monk of Jumièges, and abbot of Abingdon, 301, 452 Ralph, abbot of Mont Saint-Michel, 104 note, 142 note Ralph, archbishop of Tours, 377, 410, 421 Ralph, count of Amiens, Valois, and the Vexin, 126, 130, 138–9, 151, 187–8, 276, 372, 376, 377, 481 Ralph, earl of Hereford, son of Count Drogo and Godgifu, 69, 112, 181–2 Ralph, earl of the East Angles, son of Ralph the staller, 378–81, 387. 389–90. See Emma, wife of Ralph the Staller, earl of the East Angles, 7/14, 378 Ralph, king of the Franks, 178

587

Ralph, son of Herluin de Conteville and Fredesendis, 128 Ralph, tutor to the young William, 43 Ralph de Gacé, 54, 60, 69, 74, 76, 80 Ralph de Tosny, son of Roger de Tosny, 74, 133, 137, 174, 176, 489 Ralph Taisson, 71, 82–4, 87, 96 Ramsey (Cambridgeshire), abbey of, 435 Ramsey Benedictional, 221, 336–7 Rannulf, vicomte of the Bessin, 82–3, 96 ravaging as a military and political strategy, 144–5, 213, 228, 234–5, 238, 247, 250, 289, 313, 316–18, 348, 382, 391, 426, 458. See also ‘Harrying of the North’ Ravenna (Emilia-Romagna, Italy), 106 Reach (Cambridgeshire), 347 Redon (Ille-et-Vilaine), abbey of SaintSauveur, 174 Regenbald the chancellor, 269, 338, Regensburg (Bavaria), 147, 371 Regularis Concordia, 170 Rehoboam, son of King Solomon, 488 Reinfrid, monk, 320–1 Rémelard (Orne), 399–400 Remigius, bishop of Dorchester-onThames and Lincoln, 285, 335–6, 337, 357–8, 364, 368, 465, 474 Rennes (Ille-et-Vilaine), 202 Revolt of the Three Earls (1075), 376, 378–83, 386–8, 389 Reynolds, Susan, 525 Rheims (Marne), 136, 153 note archbishop of (unnamed), 92. See Gervase du Château-du-Loir council of (1049), 75 note, 91, 94, 97, 99–108, 111, 173, 222, 423 Rhine, river, 57 Rhineland, 371 Rhiwallon ap Cynfyn, king of Gwynedd and Powys, 192 Rhuddlan (Flintshire), 192 Rhys ap Tewdwr, king of Deheubarth, 430–1 Richard I, count/duke of the Normans, 18, 19, 45, 59, 86–7, 97, 156 Richard II, duke of the Normans, 18, 19, 20, 26, 31–2, 35, 37–9, 58, 60, 63, 64, 71, 73, 76, 78, 86–7, 102, 104, 119 note, 151, 154, 160, 170, 203, 207, 414, 439, 521. See Judith, Papia, wives of; Robert, archbishop of Rouen, brother of; Richard III and Robert I, dukes of Normandy,

588

index

William, count of Arques, Malger, archbishop of Rouen, sons of; Adeliza, daughter of Richard III, duke of the Normans, 18, 26, 30, 31, 47, 86, 100. See Adela, wife of; Nicholas, son of Richard, abbot of Sainte-Vanne of Verdun, 34, 74 Richard, archbishop of Bourges, 402, 403 Richard, count of Evreux, 60, 74, 86, 281 Richard de Heugleville, 130 Richard de Reviers, 153 Richard fitz Gilbert, son of Gilbert, count of Brionne, 86, 369, 378, 382, 388, 398, 488, 519. See also Baldwin de Meullles, brother of; Gilbert, Roger, sons of Richard Goz, vicomte of Avranches, son of Thurstan Goz, 70, 137, 341–2, 414. See also Hugh, earl of Chester, son of Richard of Devizes, 475 Richard, monk of Ely, possible author of the Gesta Herwardi, 345 Richard, second son of William and Matilda, 128, 167, 277, 306, 325, 330 ‘les Richardides’, 51–2, 59, 63, 64, 78, 87, 130, 156 Richmond (North Yorkshire), 350 Richmondshire (North Yorkshire), 319 Risle, river, 85, 94 Rivallon de Dol, 189, 201–3, 389, 390. See Gilduin, John, William, sons of Robert II the Pious, king of France, 18, 31 Robert I, duke of Normandy, William’s father, 16, 18–20, 21, 25–30, 30–7, 39–41, 42, 57, 59, 60, 61, 62, 64, 65, 66, 69–70, 73, 74, 79, 88, 94, 128, 151, 156, 159, 180, 201, 230, 481, 501, 520 Robert II ‘Curthose’, duke of Normandy, oldest son of William and Matilda, 10, 104–5, 118, 128, 142 note, 162, 167, 220–1, 277, 284, 302, 369, 370, 389, 411, 416, 420, 423, 424, 426, 427, 432, 433, 442, 446, 449, 454, 465, 479, 484–90, 495, 506, 515, 517, 524 as William’s heir, 166–7, 171, 397, 399, 404, 435, 442, 456, 486–9, 501 first quarrel with his father, 373, 396–404

role in the government of Normandy and Maine, 166, 188–9, 302, 340, 389, 435, 437 second quarrel with his father, 455–7, 484–90 Robert Guiscard, duke of Apulia, 9, 34, 172, 177, 393, 524, 525. See Bohemond, son of Robert, abbot of Jumièges, bishop of London, archbishop of Canterbury, 53 note, 69, 109–15, 118, 120, 146–7, 337 Robert, abbot of Saint-Martin of Sées, 367 Robert de Grandmesnil, abbot of Saint-Evroult, 73–4, 98, 154, 174–7, 524 Robert, archbishop of Rouen, 18, 31, 32, 34, 40, 42, 45, 46, 54, 59, 60, 86, 204, 281, 446 Robert, bishop of Coutances, 75, 143 Robert, bishop of Hereford, 462, 467 Robert, bishop of Sées, son of Hubert de Ryes, 357 Robert, count of Eu, 85, 133, 286, 312, 369, 404, 474. See William ‘Busac’, brother of; William, son of Robert, count of Meulan, son of Roger de Beaumont, 438 Robert, count of Mortain, half-brother of William, 26, 128, 155–7, 172 note, 174, 199, 201, 220, 286, 312, 342, 350, 361, 369, 417, 433, 437, 440–1, 454, 456, 465–6, 474, 484, 485, 519. See also Matilda, wife of Robert, earl of Gloucester, son of Henry I, 21 Robert de Bellême, son of Roger de Montgommery, 366, 398, 489 Robert de Comines, earl of Northumbria, 303, 305, 362 Robert de Montbrai, earl of Northumbria (Manche), 398, 479 Robert fitz Giroie, 60–1, 98, 154–5, 161–2, 174 Robert fitz Richard, castellan of York, 297, 303 Robert fitz Wymarc, 238 (Robert) Gilfard, 243 Robert Malet, 382 Robert of Rhuddlan, 432 Robert of Torigni, 25, 28, 128, 170, 196, 198 note, 354, 456, 484–5, 487–8, 490

index Robert, son of Humphrey de ‘Vieilles’, 74. See Roger de Beaumont, brother of Robert the Burgundian, 453–5 Robert the Frisian, count of Flanders, 265, 274, 330, 340–1, 370, 400, 457 Rochester (Kent), bishopric, 405–06, 471. See Arnost, Gundulf, Siward, bishops of Rodulf, count of Ivry, brother of Count/ Duke Richard I, 59, 97, 138, 173 Roffe, David, 477–8 Roger, count of Sicily, 9, 156 Roger, earl of Hereford, son of William fitz Osbern, 343, 378–83, 387, 485–6 Roger de Beaumont, 42, 50, 60, 72, 73, 79, 81, 87, 97, 131, 137, 152, 169, 220–1, 297, 301, 301–02, 330, 349, 369, 388–9, 402, 412, 438, 519. See also Beaumont family, Adeline, wife of; Robert, brother of; Henry, Robert, count of Meulan, sons of Roger de Clères, baron of the Tosny honour, 74 Roger d’Ivry, 397 Roger I de Montgommery, 35, 60, 62, 63 Roger II de Montgommery, 62, 63, 72, 81, 98, 103, 124, 137, 141, 152, 155, 156, 160, 162, 170, 172, 174–5, 220–1, 285–6, 302, 352, 357, 366, 369, 397, 402, 410, 412, 414–15, 436, 474, 488, 489, 519. See also Mabel, wife of; Robert de Bellême, son of; Matilda, daughter of Roger de Mortemer, 138–9 Roger I de Tosny, 35, 50, 53, 73–4. See Ralph de, son of; Tosny honour and family widow of, 74 Roger of Howden, 236 Roger, son of Richard fitz Gilbert, 398 Rollo, count of Rouen, first ruler of the lands that became Normandy, 100, 200, 273 Roman de Rou, 113. See Wace Romanesque architecture, 509 Rome, 5, 29, 81, 91, 110, 147, 165, 222, 280, 285, 335, 339, 358, 390, 393, 419, 420, 424, 425, 434, 441, 444, 449, 460, 461, 522 influence of, 325, 468, 478 papal council (1059), 158–9, 205 synod (1050), 101 See also Arch of Constantine

589

Romney (Kent), 234, 247, 482 Rotrou, count of the Perche, 151, 175, 397, 400. See also Geoffrey, son of Rots (Calvados), church of, 77 Rotterdam (Netherlands), 5 Rouen (Seine-Maritime), 39, 54, 63, 75, 77, 80, 97, 104, 120, 131, 132, 136–7, 138, 143, 151, 168, 177, 198–9, 205, 206, 220, 269, 275, 277, 7/16, 7/25, 7/27–8, 325, 340, 367, 369, 397, 401, 402, 415, 416, 433, 438, 439, 444, 448, 452, 454, 481, 482, 483, 485–6, 489, 499, 500 abbey of La Trinité-du-Mont, 42, 71, 97, 104, 126, 279–80, 306, 413, 490. See Isembard, Walter, abbots of abbey of Saint-Amand, 71, 97, 340, 413, 481 abbey of Saint-Ouen, 21, 34, 77, 78, 171, 207, 365, 367, 413, 447, 452. See Nicholas, abbot of archbishopric and cathedral, 31, 42, 134, 136, 137, 182, 188, 204, 280, 338, 340, 377, 388, 412, 448. See John, Malger, Maurilius, Robert, William, archbishops of citizens of, 489 ecclesiastical council of (1063), 205 ecclesiastical council of (1072), 205, 356–7, 360, 384 money of, 388 priory of Saint-Gervase, 63, 483, 489 riots of 1072 and 1073, 356, 367 vicomte of, 367. See also, Godfrey, Goscelin, Roger de Beaumont Rule of St Benedict, 214, 227 Ryes-en-Bessin (Calvados), 83 Sablé (Sarthe), 453, 455 St Albans (Hertfordshire), abbey of, 445 Saint-Aubin-sur-Scie (Seine-Maritime), 130, 132 Saint-Benoît-sur-Loire (Loiret), abbey of, 280, 454, 456 Saint-Bertin (Pas-de-Calais), abbey of, 148 Saint-Calais (Sarthe), abbey of, 410, 427, 442, 498. See William of Saint-Calais Saint-Céneri-le-Géré (Orne), castle of, 154–5 St David’s (Pembrokeshire), 430–2, 499

590

index

Saint-Denis (Seine-Saint-Denis), abbey of, 146, 214, 287, 302–03, 306, 394, 418, 430, 469 Saint-Evroult (Orne), abbey of, 98, 150, 154–5, 174–7, 208, 220, 352, 353, 397, 402, 413, 415, 430, 448, 481. See Mainer, Osbern, Robert de Grandmesnil, Thierry, abbots of; Orderic Vitalis, monk and historian of Saint-Germain-des-Prés (Paris), abbey of, 469 Saint-James-de-Beuvron (Manche), 202, 280 Saint-Martin-du-Bosc (Seine-Maritime), priory of, 168 Saint-Omer (Pas-de-Calais), 139, 147–8 Saint-Oyend (Jura), abbey of, 393 Saint-Pierre-sur-Dives (Calvados), abbey of, 76, 85, 98, 144, 277, 394. See also Ainard, abbot of Saint-Riquier (Somme), abbey of, 92, 149, 194, 200, 291 See also Gervin, abbot of; Hariulf of Saint-Sauveur-le-Vicomte (Manche), abbey of, 96, 143, 413 Sainte-Suzanne (Sarthe), 453–5, 459, 463–4, 466, 518. See also Hubert de Saint-Valéry-sur-Somme (Somme), 228–9, 230–2 Saint-Wandrille (Seine-Maritime), abbey of, 43, 59, 105, 207, 369, 413, 414, 440 Saintonge, the, 180 Salerno (Campania), 236, 460 Salian emperors. See Germany, kings and kingdom Salisbury (Old Sarum) (Wiltshire), 119, 308, 321, 426, 429, 474, 475, 475–6 bishopric, 474. See Osmund, bishop of Salisbury Oath (1 August 1086), 463, 475–7 Sallust, 200 Samson, royal chaplain, later bishop of Worcester, 438. See Thomas, archbishop of York, brother of Samson, servant of Matilda and monk of Saint-Evroult, 402 Sandwich (Kent), 217, 309 Sarthe, river, 366 Savigny-le-Vieux (Manche), church of, 142 Saumur (Maine-et-Loire), abbey of Saint-Florent, 172, 390, 415–16, 419, 429, 449, 519

‘Saxo-Norman overlap’, 319 Saxons, people, 522 Scaldwell (Northamptonshire), 501 Scandinavia and Scandinavians, 20, 119, 172, 207, 230, 327, 497, 511 history of, 258 Scotland, 295, 305, 313, 345, 359, 360–2, 363, 370, 389, 424, 432, 433, 493, 496, 501, 511 Scots, 315, 318, 518 kings and kingdom of, 328, 426–7, 428, 479, 511 Sebastian, St, 273 Second World War, 4, 166 Sées (Orne), 69, 76, 125, 141, 155 abbey of Saint-Martin, 367, 413, 414 bishopric, 31–2, 357. See Ivo (of Bellême), Radbod, Robert, Seginfrid, bishops of Seginfrid, bishop of Sées, 32 Seine, river, 18, 29, 34, 39, 65, 156, 277, 415, 440, 486, 489 Selby (North Yorkshire), abbey of, 319–20 Selsey (West Sussex), bishopric of, 338 See Æthelric, Stigand, bishops of Senlis (Oise), 92–3 Sens (Yonne), archbishopric, 421 Serlo, abbot of St Peter’s, Gloucester, 428–9 Serlo of Bayeux, 492 Settrington (North Yorkshire), 362 Severn, river, 290 sexuality, female, 309 Shakespeare, William, 7 Sharnbrook (Bedfordshire), 464 Sherborne (Dorset), bishopric of, 476 sheriffs, 351, 405 Ship List of William the Conqueror, 226, 229 Shrewsbury, 308 earldom of, 297, 431 Shropshire, 266, 297, 344, 352 Sillé-(le Guillaume) (Sarthe), 366 Simon, count of Amiens, Valois, and the Vexin, 88, 188, 372, 376, 377, 388–9, 392, 393–4, 402–03, 419, 419–20, 449, 481 simony, 75, 76, 99, 101, 337 Siward, bishop of Rochester, 336, 385 Siward, earl of Northumbria, 110, 117, 119, 146, 191, 209, 267, 350 Siward Bearn, 346, 349 Siward son of Æthelgar, 266. See also Ealdred, brother of

index slave-trade, 432–3, 522 Soberton (Hampshire), 469 Soissons (Aisne) abbey of Saint-Médard, 92 Battle of (923), 279 county of, 85 Solent, the, 231, 325 Solesmes (Sarthe), abbey of Saint-Pierre, 364–5 Solomon, king, 488, 526 Somerset, 308, 343 South Wales, 379, 431 Southampton (Hampshire), 65, 278, 384, 436, 481 Spain, 73, 94, 331, 372 Speyer (Rhineland-Palatinate), cathedral, 408 Stafford, 312, 321 Staffordshire, 296, 344 Stainmore (Cumbria, Durham, and North Yorkshire), 360 Stamford Bridge (East Riding of Yorkshire), battle of (1066), 232–3, 237, 241, 267 Stephen, king of England and duke of Normandy, 325, 434, 439, 491, 508 ‘Oxford’ Charter of Liberties. 496 Stephen, deceived father of a purchased child, 301 Stephen, legate of Pope Alexander II, 190 Stephen-Henry, count of Blois-Chartres, 403, 419, 438–9 Steyning (West Sussex), 226, 474 Stigand, archbishop of Canterbury and bishop of Winchester, 110, 117, 120, 210, 216, 248, 250, 256, 276, 277–8, 285, 292, 306, 335, 337 Stigand, bishop of Selsey, 338 succession disputes, 10, 23–5, 33, 37, 87–8, 180, 216, 488–9, 497, 517–18 English, 41, 46, 49, 65–9, 89, 102, 108–19, 126, 146–8, 149–50, 164, 178, 180–1, 190, 216, 264, 326 in Maine, 177–80, 183 Suetonius, Lives of the Caesars, 492 Suffolk, 468 Surrey, 249, 268, 505 Sussex, 234, 236, 248, 286, 290, 297, 386, 416, 419, 440, rapes, 286, 290, 325, 386 Swavesey (Cambridgeshire), 448 Svein Estrithsson, king of Denmark, 22, 37, 223, 264, 303, 304, 308–10, 313, 328, 332–3, 341, 379

591

Svein Forkbeard, king of Denmark, 37, 38. See Cnut, king, son of Svein, earl, son of Earl Godwine, 110 114–15, 119, 146 Svein, son of King Cnut by Ælfgifu of Northampton, 20 Symeon of Durham, 25, 303, 305, 315, 362–3, 425, 427, 434 Tacitus, 513 Tancred of Lecce, 236 Taunton (Somerset), 475 Taw, river (Devon), 308 Tay, river, 361 taxation, in England, 261, 265–6, 271, 288, 294, 310–11, 322, 327, 433, 452, 454, 457, 462, 493, 504–08. See also, danegeld, heregeld Tees, river, 209, 313, 425 Teesdale (Durham), 360 Teuzo, monk and papal legate, 392 Thames, river, 167, 249–50, 268, 285, 308, 449 Theobald III, count of Blois-Chartres, 92–3, 99, 102, 132, 187, 224, 309, 438 Theobald IV, count of Blois-Chartres, 439 Theophanu, empress, 55 Thetford (Norfolk), bishopric of, 358, 429, 459. See also, East Anglia, bishopric of; Æthelmar, Herfast, bishops of Thierry, abbot of Saint-Evroult, 98 Thimert (Eure-et-Loir), castle of, 143, 150–4 Thomas, archbishop of York, 312, 338, 338–9, 343–4, 357–9, 395, 417, 443, 465, 474. See Samson, royal chaplain, brother of Thomas, Hugh, 259 Thomas of Marlborough, 315, Thorfinn, earl of Orkney, 361 Thorkell of Warwick, 267 Thorney (Cambridgeshire), abbey of, 458 Thurstan, abbot of Ely, 346–7 Thurstan, abbot of Glastonbury, 347, 442, 446–7, 519 Thurstan Goz, 53, 62, 69–70. See Richard Goz, son of Thurstan Haldup, 143, 417. See Eudo Haldup, son of Tillières-sur-Avre (Eure), 53, 137, 154 castle of, 64–5, 150–1 Toki, son of Wigod of Wallingford, 401 Toledo, Fourth Council of (633), 251

592

index

Tonbridge (Kent), 388 Tosny (Eure), honour of, 74 family of, 53, 73, 176. See also Ralph de Tosny, Roger de Tosny Tostig, earl of Northumbria, brother of King Harold, 115–16, 119, 127, 146, 191–3, 209, 215, 216, 223–4, 229, 232–3, 267, 283, 350, 431 Touques (Calvados), 487 Tours (Indre-et-Loire), 70 abbey of Saint-Julien, 95 note archbishopric, 377, 410, 421. See Bartholomew, Ralph, archbishops of; Berengar of Trahaern ap Caradog, king of Gwynedd, 430 Treaty of Paris (1259), 511 tribute, in England, 261, 322 from Wales, 431–2 Trinity College, Cambridge, 252 Troarn (Calvados), abbey of, 103, 152, 160, 301, 302, 413. See also Durand, abbot of Troy, 246 Truce of God, 77–8, 204, 416 Turkey, 47 Turold, abbot of Malmesbury and Peterborough, 332 Turold, tenant of Odo, bishop of Bayeux, 445 Turold, tutor to the young William, Tyne, river, 267, 283, 284, 408, 425 Tynemouth (Tyne and Wear), 479 Uffculme (Devon), 476 Úlfr, earl, first husband of Cnut’s sister Estrith, 36 Urse d’Abetôt, sheriff of Worcestershire, 381 Val-ès-Dunes, battle of (1047), 49, 54, 72, 76, 77, 81–5, 86–7, 91, 92, 95, 96, 130, 132, 174, 296, 369, 404, 517, 518 Valéry, St, 231 Valognes (Manche), 83–4, 301 Vannes (Morbihan), county of, 391 Varaville (Calvados), battle of (1057), 122, 145, 150, 151, 155, 160, 161, 162, 179, 518 Vercelli (Piedmont), council of (1050), 103 Verdun, Treaty of (843), 468 Vernon (Eure), castle of, 63

Vexin, county of, 388, 394, 400, 438, 481–2. See Drogo, Ralph, Simon, Walter, counts of Vic-sur-Aisne (Aisne), castle of, 92 violence, 10, 33, 61, 145, 154, 174–5, 186, 190, 206, 217, 235, 254–5, 259, 264, 279, 282, 305, 307, 317, 320, 353–4, 364, 408, 445, 470, 482, 493, 494, 504, 511, 512, 513–14, 522–3, 526, 528 against women, 121, 235–6, 238, 322–3, 449–50 contemporary debates about the violence of the conquest of England, 221–3, 253–4, 327, 353–4, 486, 497, 513–14, 526 culturally structured, 10, 12, 14, 33, 43, 51–5, 57–8, 61–3, 65, 74, 78, 88–9, 107, 161, 207–8, 222–3, 299, 305, 374, 445, 472–3, 494, 515 honour-based, 57, 208, 255, 374, 472–3, 515 ravaging of land, 144–5, 202, 213, 224, 228, 234–5, 238, 247, 250–1, 482 state-sponsored, 235–6, 322–3, 425–6, 493, 512, 528 See also ‘anger’; feud; medieval warrior kingship Vigor, St, 92 Vieilles (Eure), 73 Vimoutiers (Orne), 61–2 Virgin Mary, 483 Vita Ædwardi Regis, 67, 109, 111, 115, 120, 149, 194–5, 209–10, 211, 212–13, 250, 262 Vita Herluini, 57, 107 Vita Lanfranci, 107–8, 158–9, 334, 336, 337, 354, 523 Vitalis, monk of Fécamp, abbot of Bernay, and abbot of Westminster, 409, 411 Vitalis of Canterbury, tenant of Odo, bishop of Bayeux, 445 Vitry-aux-Loges (Loiret), 126, 372 Wace, 29, 30, 81–4, 96, 113, 144, 160, 168, 186, 196, 200, 203, 219, 226, 229, 244, 354 on William’s birth and parentage, 4, 81 Wadard, tenant of Odo, bishop of Bayeux, 445 Walchelin, bishop of Winchester, 338, 344, 408, 417, 442, 465, 475

index Walchelin de Ferrières, 50, 52, 60 Waleran I, count of Meulan, 65, 79. See Adeline, daughter of Walcher, bishop of Durham, 336, 362, 363, 382, 387, 408, 426 Wales, 296, 309, 311, 343, 379, 429, 430–2, 433, 436, 458, 493, 496, 511 kingdoms of, 191, 362 Wallingford (Oxfordshire), 249–50, 298, 401 Walter, abbot of Evesham, 406 Walter, abbot of La Trinité-du-Mont of Rouen, 490 Walter, bishop of Hereford, 250 Walter, count of the Vexin, son of Drogo and Godgifu, 151, 165, 180–4, 187–8 Walter de Lacy, 381–2 Walter, brother of Herleva, 63, 138. See also William, son of Walter Giffard, 133, 137, 467 Waltham (Essex), abbey of, 245, 363 Waltham Chronicle, 245 Waltheof, earl, son of Earl Siward, 119, 146, 191, 209, 267, 276, 310, 313, 323, 350, 362, 364, 378–81, 386–7, 402, Warmund, abbot of Déols and archbishop of Vienne, 371, 402, 403 Warwick, 296, 297 Warwickshire, 438 Weald, the, 286 Wearmouth (Tyne and Wear), monastery of, 336 Wells (Somerset), cathedral, 290, 449 See also Giso, bishop of Wellstream (Cambridgeshire), 347 Wessex, 282, 287 earldom of, 119, 146, 268. See also, Godwine, Harold, earls of line of kings, 67, 117, 329 West Midlands, 270, 282, 296, 332, 346, 376, 433, 458, 467 Westminster, 107, 210, 251, 291, 307, 325, 383, 408, 474, 478, 499 abbey, 256, 296, 323, 326, 383–4, 409, 411, 480, 500. See Gilbert Crispin, Vitalis, abbots of crown-wearings at, 324, 383, 408, 474, 475, 493, 500 royal palace, 409, 495 ‘Whig history’, 512 White Ship, wreck of (1120), 508

593

White Tower of London, 268–9, 324, 385, 405 Wickham, Chris, 497 Wihenoc, landholder at Monmouth, 429 Willelmi Articuli, 433, 476–7 William, duke of Normandy, count of Maine, and king of England adolescence, 49–50, 69–81, 86–90, 520 aftermath of Val-ès-Dunes, 92–9 ‘anger’ (ira), William’s, as a facet of rule, 13, 34, 107, 155, 174–6, 260–1, 270–1, 288, 314, 317, 323, 338–9, 380, 411, 427–8. See also theatrical performance attempt to learn the English language, 351 births of children, 128, 163, 275, 277, 293, 320 Channel crossing (1066), 230–4 chapel and chaplains, 168–9, 301, 338, 355, 396, 437–8, 452, 459, 464–5 childhood, 21, 33–4, 41–4, 47–8, 53, 56, 58, 63, 65, 73, 87–90, 95, 201, 374, 487, 492 comparison with Cnut, 327 conduct of the Hastings campaign, 234–47, 247–51, 254 consolidation of Normandy’s borders, 79, 80–1, 98–9, 122–5, 137–40, 150–2, 154–7, 180, 188, 200–2, 280, 389–94, 397–8, 438–9, 480–2 conquest of Maine, 14, 125, 127, 164–5, 166, 170–1, 174, 177–87, 189, 201, 255 coronation (1066), 253, 256–7, 260, 261, 266, 294, 312; coronation (1070), 336 coronation by papal legates (1070), 336–7 date and place of birth, 25–6 daughters, marriages of, actual and proposed, 180, 198, 266, 330–1, 331, 376, 393, 438–9, 461, 480 deathbed, 483–8 English and British saints, attitude to, 362–3 first months of kingship in England, 258–75 foundation of Battle Abbey, 385–6 funeral and burial, 489–90 health and illnesses, 165–8, 482, 483–90 historical significance, 509–12, 513–14 humour, 306–07 hunting, 325, 493

594

index

‘introduction of knight service into England’, 502–04, 509 justice, the maintenance of order, and the settlement of disputes,190, 220, 333–4, 344, 363–4, 379, 384–5,405, 406, 411, 415, 433–4,442–3, 446, 471, 472, 476, 477, 493, 519 legends about, 29–30, 121–2 marriage to Matilda, 91, 94, 99–100, 104–8, 115, 120–1 maternal kin, 63, 90, 120, 138 military household, 348 organization of Queen Edith’s funeral and burial, 383–4 patronus Normannorum, 272–4, 477, 491, 524 personal characteristics, 9–10, 492, 493, 514–28 personal relationship with Matilda, 427–8 piety, 70, 77–8, 89, 164, 168, 232, 522–3 possibility of military intervention in Germany (1074), 371 pre-1066 rule in Normandy, 161–3, 173–7, 204–6 preparations for the invasion of England, 217–30, 252, prisoners, treatment of, 138, 185 not taken at Val-ès-Dunes and Hastings, 84, 241 ransomed, 455 release of in Wales (1081), 431 release of on deathbed, 485, 495–6 taken in war, 133, 139, 140, 144, 145, 162, 185 See also rebels, treatment of prostrations, 307, 371–2 rebels, treatment of, 95–6, 132, 138, 174–5, 380, 383, 404, 455, 494, 498 relationship with Edward the Confessor, 37–41, 65, 66–7, 68–9, 112–13 religious patronage, 128–9, 157–61, 168–71, 226–7, 231–2, 278, 280, 331, 371–2 rule in England, 499–507 rule over Maine, 188–91, 265, 303–04, 328, 329, 340, 356, 364–7, 372, 376–7, 389, 392, 396, 399, 409–11, 418, 423, 427, 435, 435–8, 439, 442, 453–5, 486, 491, 496, 498–9, 508 rulership and the Church, 103–4, 134–7, 162, 170, 204–5, 273, 334–8, 344, 356–7, 358–60, 365, 384–5, 416, 421–2, 520–1, 526. See also papacy, the

seal, 272–4, 400–01, 491, 501 sexual behaviour, 88–9, 205, 399, 450, 517 significance of his birth and parentage, 4, 16–30 soldier and war-leader, 54, 69, 83–4, 93–4, 123–5, 131–3, 144–5, 183–4, 185–6, 202–3, 234, 243, 273, 288–9, 296, 305, 311–13, 321–2, 345–8, 366–7, 391–2, 401, 435–7, 482, 516–17. See also conduct of the Hastings campaign; medieval warrior kingship theatrical performance, 42, 77, 107, 171–2, 252, 257, 306–7, 371–2, 349–50 (tears), 477, 500–16. See also ‘anger’, prostrations visits to England after 1066: late 1067 to late 1068, 281–94, 295–300; early 1069–late 1070, 303–39; early-late 1071, 343–51; 1072 (and Scotland), 358–64; late 1073 to early 1074, 368–9; autumn 1075 to summer 1076, 382–8; possible short visit, late 1077 to early 1078, 396; autumn 1080 to autumn 1081 (and Wales), 425–33; autumn 1082 to early 1083, 441–4; possible visit in 1084, 462; autumn 1085 to autumn 1086, 457–9, 462–80 visits to Normandy after 1066: 1067, 275–81; late 1068 to early 1069, 301–3, late 1070 to early 1071, 340–3; 1071–2, 356–7; (and Maine) late 1072 to late 1073, 365–8; early 1074 to autumn 1075, 369–72, 374–8, 382; summer 1076 to autumn 1080 (and Maine, and with a possible short visit to England in 1077–8), 389–96, 396–404, 409–17; autumn 1081 to autumn 1082 (and Maine), 435–41; early 1083 to autumn 1085 (and Maine, with a possible visit to England in 1084), 445–6, 452–5; autumn 1086 to 9 September 1087, 480–3 William II (‘Rufus’), son of William and Matilda, king of the English, 10, 128, 171, 220, 349, 369, 387–8, 389, 399, 401, 402, 409, 435, 446, 454, 456, 465, 474, 477, 479, 483–90, 495, 496, 498, 500, 501, 506, 511, 515, 516–17, 523

index William Ætheling, son of King Henry I, 508 William Crispin, 133, 151 William ‘Bona Anima’, abbot of SaintEtienne of Caen and archbishop of Rouen, 340, 412, 414, 415, 420, 425, 438, 440, 446, 489 William, bishop of Evreux, 76, 99 William, count of Arques, 53, 64, 76, 78–80, 81, 85, 86–7, 105, 122, 126, 129–33, 134, 137, 161–2, 240 note, 296, 517 William, count of Eu, son of count/duke Richard I, 61, 64, 86. See Robert, William Busac, sons of William ‘Busac’, count of Eu, 64, 85, 86 William, count of Evreux, 281, 369, 436, 455, 489 William Longsword, count of Rouen, 20 William VI the Fat, duke of Aquitaine, 33 William VII, duke of Aquitaine, 132,141, 391 William VIII, duke of Aquitaine, 391 William de Bellême, 70. See Ivo, bishop of Sées, son of William de Breteuil, son of William fitz Osbern, 343, 398, 412, William de Briouze (Orne), 286, 415–16, 474 William de La Lande-Patrie, 203 William de Percy, 297 William de Rots, abbot of Fécamp, 411 William de Vauville, 290, 308 William de Warenne, 137, 138, 286, 297, 346, 350, 369, 378, 382, 419, 420, 455, 519. See also Gundrada, wife of William fitz Ansculf, 470 William fitz Giroie, 98, 154 William fitz Osbern, earl of Hereford, 44, 63, 64, 72, 76, 87, 97, 105, 112, 124, 137–8, 141, 152, 172, 176, 219, 220, 237, 268, 281–2, 283, 299–300, 303, 305–06, 306, 330, 332, 340–3, 378, 379, 381, 387, 398, 407, 431, 438, 488. See William de Breteuil, Roger, earl of Hereford, sons of; Emma, daughter of William Malet, 245, 297, 299, 320, 348 William of Jumièges, 17, 22, 23, 25, 30, 31, 36, 37, 38, 39, 40, 43, 44, 45, 49, 50, 51, 53, 58–9, 62, 65, 68, 69, 70, 72,

595

79, 81–4, 95, 105, 109–11, 113–14, 116, 118, 120, 122, 124–6, 129, 132, 134, 135 note, 144–5, 150, 157, 162–3, 186–7, 193, 195–6, 200, 211, 219, 229, 242, 251, 256, 302, 303, 304–05, 481, 521 moral and historical perspective on William’s acts and rule, 62, 145, 162–3 on William’s birth and parentage, 17, 22–3 William of Malmesbury, 4, 5, 19 note, 20, 22, 25, 26, 27, 28, 29, 30, 33–4, 38, 45, 47, 48, 54, 55, 59, 63, 88, 91, 108, 111, 114, 117, 134, 148, 149, 157, 167–8, 172, 192, 195–6, 205, 211, 214–15, 217, 219, 221–2, 225, 233, 237, 244–5, 247, 249–50, 270, 275, 277–8, 285, 301, 310–11, 314, 317, 323, 327, 330, 332, 334, 337, 338, 341, 342, 345, 349, 370, 372, 379, 380, 381, 383, 386, 398, 401, 423, 428–9, 440–1, 441, 442, 443, 447, 449–50, 457, 459, 480, 481–2, 483–90, 491, 494, 496, 500, 508, 513, 514, 517, 518, 526–7 moral and historical perspective on William, 4–5, 22, 33–4, 48, 55, 168, 172, 225, 277–8, 301–2, 310–11, 314, 323, 327, 337, 345, 372, 380, 433, 447, 449–50, 480, 481, 484, 493–4, 500, 514, 520, 523–4, 526–7 on William’s birth and parentage, 4, 20, 22, 24, 27, 29 De Antiquitate Glastonie Ecclesie, 442, 446 Gesta Regum, 380, 433, 526 Gesta Pontificum, 314, 380, 523–4, 526 Vita Wlstani, 215, 344, 432–3, 457, 458 William of Poitiers, 4, 8, 17, 44, 45, 47, 50, 53–4, 68, 70–1, 72, 78–9, 82, 84, 88, 93–6, 97,106, 109–10, 112–14, 116–20, 122–6, 129, 131–3, 134, 140–1, 144–5,150, 162, 164, 169, 177–9, 180, 181, 183–6, 187, 193–9, 200–2, 211, 212, 215–17, 218–21, 223, 225–6, 228, 230–3, 234–5, 237, 238–46, 247–54, 256, 263, 264, 266, 267, 268, 269, 276, 276–7, 278, 279, 281, 282, 283, 284, 289, 298–9, 309, 311, 313, 313–14, 321, 322, 327, 334, 336, 346, 351, 351–2, 353, 461, 481, 482, 521, 522

596

index

moral and historical perspective on William, 4–5, 23, 46, 93–4, 162, 178, 200, 254–5, 261, 275, 354–5 on William’s birth and parentage, 17, 22–3 treatment of the English succession, 36, 37, 193 William of Saint-Calais, abbot of Saint-Vincent of Le Mans and bishop of Durham, 409–10, 427, 428, 434, 442, 459, 465–6, 475 William of Volpiano, abbot of Féamp, 19 William Pantulf, 414–15 William Peverell, 297 William Talvas, lord of Bellême, 98, 155–6. See Arnulf, son of; Mabel, daughter of William Warelwast, bishop of Exeter, 465 William Werlenc, count of Mortain, 86, 142, 155–6, 524 William, son of Rivallon de Dol, 202 William, son of Robert, count of Eu, 474 William, son of Roger I de Montgommery, 62 William, son of Herleva’s brother Walter, 138 Wilton (Wiltshire), 263 Wiltshire, 209, 268, 269, 283, 474 Winchcombe (Gloucestershire), abbey of, 296, 336 See Godric, abbot of Winchcombeshire, 407 Winchelsea (East Sussex), 284 Winchester (Hampshire), 65, 160, 231, 250, 268, 279, 288, 290–1, 298, 308, 312, 324, 325, 337, 343, 383, 384, 386, 408, 409, 439–40, 443, 478, 486, 499, 500, 516, 521 New Minster, 149, 324, 496 Old Minster, 496 bishopric of, 337 cathedral, 408–09 crown-wearings at, 290–1, 303, 304, 306, 307, 358–9, 384, 408–9, 429–30, 474, 475, 493, 499–500 ecclesiastical council of (1070), 334, 335–8 ecclesiastical council of (1072), 358–9 ecclesiastical council of (1076), 384, 405 royal palace, 324

royal treasury, 349 St Giles Hill, 386 See Henry, Stigand, Walchelin, bishops of Windsor (Berkshire), 323–4, 325, 386, 433, 500 castle, 503 ecclesiastical council of (1070), 334, 336, 337, 337–8 ecclesiastical council of (1072), 359, 363 Wissant (Pas-de-Calais), 66 Witchford (Cambridgeshire), 349 Wolsey, Cardinal, 444 Worcester, 109, 235, 271, 381, 458 bishopric, 433. See Ealdred, Wulfstan, bishops of; Urse d’Abetôt, sheriff of Worcestershire, 467 Wormald, Patrick, 497, 508 writs, English, 269, 270, 303, 350–1, 464. See also, charters Wulfnoth, brother of King Harold, hostage in Normandy, 114, 119, 485–6 Wulfstan, archbishop of York, 275, 432–3 Wulfstan, bishop of Worcester, 215, 250, 275, 293, 303, 334, 335, 343–4, 376, 381, 408, 432–3, 458 Wulfwig, bishop of Dorchester-onThames, 270, 285 Wulgrin, bishop of Le Mans, 190 York, 109, 209, 232–3, 271, 296, 297–9, 303, 305, 307, 310–13, 313, 315, 317, 319, 348, 370, 382, 478, 482, 496, 499, 523 abbey of St Mary’s, 319, 478, 496 archbishopric of, 338, 344, 358–9, 424, 428 cathedral of St Peter (York Minster), 307, 312, 326, 382 church of St Olaf, 478 Clifford’s Tower, 296 Coppergate, 312 hospital of St Leonards, 478 Mickelgate, 305 Old Baile, 305 Vale of, 318, 350 See Ealdred, Oswald, St,Thomas, Wulfstan, archbishops of Yorkshire, 237, 297, 299, 304, 307, 311, 315–21, 350, 427, 440, 458 North Riding, 318