Writing the Military History of Pre-Crusade Europe (Variorum Collected Studies) [1 ed.] 9780367547066, 9781003090243, 0367547066

Citation preview

Writing the Military History of Pre-Crusade Europe

Writing the Military History of Pre-Crusade Europe brings together fourteen articles by eminent historians David S. Bachrach and Bernard S. Bachrach. Crucial to the writing of medieval military history is a thorough understanding of the strengths and weaknesses of the available source materials. Just as important is a broad conception of the range of sources which scholars can draw upon to ask and answer questions about the organization and conduct of war. The studies collected in this volume provide insights regarding many of the most important narrative works from pre-Crusade Europe, with a particular emphasis on the ways in which they can be used to write military history, as well as the pitfalls facing historians who read these texts transparently without regard for the authors’ various parti pris and limitations. In addition to their treatment of narrative works, several of the studies in this volume highlight the importance of treating historiographical texts within the broader range of source materials that illuminate the conduct and organization of war in pre-crusade Europe, particularly material sources developed through excavations, as well as contemporary images, most prominently the Bayeux Tapestry. The book will appeal to scholars and students of medieval history, as well as those interested in military history. David S. Bachrach is Professor of Medieval History at the University of New Hampshire. His publications include Religion and the Conduct of War c. 300–c.1215 (2003) and Warfare in Tenth-Century Germany (2012). Bernard S. Bachrach is Professor Emeritus of Medieval History at the University of Minnesota. His publications include Early Carolingian Warfare: Prelude to Empire (2001) and Charlemagne’s Early Campaigns (768–777) (2013).

Also in the Variorum Collected Studies series:

EKMELEDDIN İHSANOGLU Studies on Ottoman Science and Culture (CS1098)

DAVID S. BACHRACH and BERNARD S. BACHRACH Writing the Military History of Pre-Crusade Europe Studies in Sources and Source Criticism (CS1097)

PAMELA M. KING, edited by ALEXANDRA F. JOHNSTON Reading Texts for Performance and Performance as Texts Shifting Paradigms in Early English Drama Studies (CS1096)

FELICE LIFSHITZ Writing Normandy Stories of Saints and Rulers (CS1095)

STEPHEN GERSH Metaphysics and Hermeneutics in the Medieval Platonic Tradition (CS1094)

MICHAEL HESLOP Medieval Greece Encounters Between Latins, Greeks and Others in the Dodecanese and the Mani (CS1093)

T.A BIRRELL, edited by JOS BLOM, FRANS KORSTEN and FRANS BLOM Aspects of Recusant History (CS1092)

PAMELA NIGHTINGALE Mortality, Trade, Money and Credit in Late Medieval England (1285–1531) (CS1091)

SARAH CARPENTER, edited by JOHN J. McGAVIN and GREG WALKER Early Performance: Courts and Audiences Shifting Paradigms in Early English Drama Studies (CS1090)

EVELLEEN RICHARDS Ideology and Evolution in Nineteenth Century Britain Embryos, Monsters, and Racial and Gendered Others in the Making of Evolutionary Theory and Culture (CS1089) For more information about this series, please visit: www.routledge.com/ Variorum-Collected-Studies/book-series/VARIORUM

Writing the Military History of Pre-Crusade Europe

Studies in Sources and Source Criticism David S. Bachrach and Bernard S. Bachrach

VARIORUM COLLECTED STUDIES

First published 2021 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN and by Routledge 52 Vanderbilt Avenue, New York, NY 10017 Routledge is an imprint of the Taylor & Francis Group, an informa business This edition © 2021 David S. Bachrach and Bernard S. Bachrach The right of David S. Bachrach and Bernard S. Bachrach to be identified as authors of this work has been asserted by them in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data A catalog record for this book has been requested ISBN: 978-0-367-54706-6 (hbk) ISBN: 978-1-003-09024-3 (ebk) Typeset in Times New Roman by Apex CoVantage, LLC VARIORUM COLLECTED STUDIES SERIES CS1097

W E D E D I C AT E T H I S V O L U M E TO O U R F R I E N D K A R L F. M O R R I S O N

CONTENTS

Introduction

1

PART 1

Narrative Works

5

1 B. Bachrach, “Gregory of Tours as a Military Historian,” in The World of Gregory of Tours, ed. Kathleen Mitchell and Ian Wood (Leiden, 2002), 351–363, republished with the permission of Brill Press.

7

2 D. Bachrach and B. Bachrach, “Nithard as a Military Historian of the Carolingian Empire, c. 833–c.843,” Francia 44 (2017), 29–55, republished with the permission of the German Historical Institute of Paris and Jan Thorbecke Verlag.

19

3 B. Bachrach and D. Bachrach, “Saxon Military Revolution, 912–973?: Myth and Reality,” Early Medieval Europe 15 (2007), 186–222, republished with the permission of Wiley Press.

46

4 D. Bachrach, “Early Ottonian Warfare: The Perspective from Corvey,” Journal of Military History 75.2 (2011), 393–410, republished with the permission of the Society for Military History.

81

5 D. Bachrach, “Memory, Epistemology, and the Writing of Early Medieval Military History: The Example of Bishop Thietmar of Merseburg (1009–1018),” Viator 38 (2007), 63–90, republished with the permission of Brepols Press.

98

vii

CONTENTS

6 B. Bachrach, “Dudo of Saint Quentin as a Historian of Military Organization,” The Haskins Society Journal 12 (2002, appeared in 2003), 155–185, republished with the permission of Boydell Press.

132

7 B. Bachrach, “Ademar of Chabannes as a Military Historian,” in Where Heaven and Earth Meet: Essays on Medieval Europe in Honor of Daniel F. Callahan, ed. Michael Frasseto, Matthew Gabriele, and John D. Hosler (Leiden, 2014), 42–62, republished with the permission of Brill Press.

153

8 D. Bachrach and B. Bachrach, “Bruno of Merseburg’s Saxon War: A Study in Eleventh-Century Germany Military History,” Journal of Military History 81.2 (2017), 341–367, republished with the permission of the Society for Military History.

171

9 D. Bachrach, “Feudalism, Romanticism, and Source Criticism: Writing the Military History of Salian Germany,” Journal of Medieval Military History 15 (2015), 1–25, republished with the permission of Boydell Press.

198

10 B. Bachrach and D. Bachrach, “Ralph of Caen as a Military Historian,” in Crusading and Warfare in the Middle Ages: Realities and Representations, Essays in Honour of John France, ed. Nicholas Morton (Aldershot, 2014), 87–99, republished with the permission of Routledge Press.

223

PART 2

Material Sources and Images

237

11 D. Bachrach and B. Bachrach, “Landscapes of Defense: At the Nexus of Archaeology and History in the Early Middle Ages,” Francia 42 (2015), 231–252, republished with the permission of the German Historical Institute of Paris and the Jan Thorbecke Verlag.

239

12 B. Bachrach and D. Bachrach, “The Costs of Fortress Construction in Tenth-Century Germany: The Case of Hildagsburg,” Viator 45.3 (2014), 25–58,” republished with the permission of Brepols Press.

263

viii

CONTENTS

13 B. Bachrach, “Some Observations on the Bayeux Tapestry,” Cithara 27 (1987), 5–28, republished with the permission of Cithara and St. Bonaventure University.

303

14 D. Bachrach, “Henry I of Germany’s 929 Military Campaign in Archaeological Perspective,” Early Medieval Europe 21.3 (2013), 307–337, republished with the permission of Wiley Press.

329

Index

358

ix

INTRODUCTION

We have sought, collectively, for more than seventy years to draw attention to the sophistication, enormous material resources, and considerable intelligence that were required for the conduct of war in the six centuries between the dissolution of imperial power in the West and the rise of the crusading movement. An essential element in this program has been an effort to overturn the generalizations and assertions of generations of scholarship, often developed in and drawing upon the romantic-nationalist traditions of the nineteenth century, that present medieval warfare as primitive, small-scale, and imbued with “dark age” barbarism utterly divorced from the sophisticated military systems of the Roman past. Concomitant with our effort to reconceptualize the conduct of war, we have sought to address both the misuse and, just as importantly, the neglect of crucial source materials that shed light on the scale and complexity of warfare in preCrusade Europe. In numerous published works, we have argued for the signal importance of a thorough understanding of historical epistemology. In our view, without a clear and detailed knowledge of the strength and weaknesses of our sources, as well as a thorough grasp of the kinds of questions that particular types of sources can illuminate, the historian becomes merely a curious type of novelist. The dangers of reading a selective group of sources, particularly narrative sources, as plain text are demonstrated all too frequently in the work of specialists in the military history of the European Middle Ages. The practice of creating a historical pastiche from medieval historiographical works or choosing the putative “best” accounts to achieve a narrative of a battle or description of a socio-military system dates back to the origins of scholarly history writing in the nineteenth century, long before serious source criticism had become de rigueur in many fields of historical inquiry. As R. G. Collingwood, the famed ancient historian, archaeologist, epigrapher, and Britain’s leading philosopher of history, observed in his landmark study The Idea of History, the “scissors and paste” method is incapable of getting the modern scholar into the mind of those whom he seeks to study. Unfortunately, all too often, the practice of attempting to reconcile, or dismiss, narrative texts on the basis of an uncritical reading of these sources still can be seen in the work of military historians, who have refused to adopt the types of sophisticated source analysis seen in other branches of historical inquiry. Even 1

INTRODUCTION

more problematic has been the unwillingness of many medieval military historians to integrate narrative works into a much broader range of source materials, particularly the findings developed by archaeologists on the basis of excavations. Fundamental misconceptions and misrepresentations, often dating back to sociological and political assumptions of the nineteenth century, remain dominant features of current scholarly discussions among numerous military historians, and through them to the broader treatments of medieval history. Significant problems in this vein include the creation from whole cloth of a medieval warrior aristocracy, the putative domination of medieval warfare by mounted forces, the thorough neglect of the central role played by middling and small-scale landowners and landholders in the conduct of military operations, neglect of the centrality of sieges to the conduct of war throughout the medieval millennium, and a minimization of the scale and cost of warfare in medieval Europe with the concomitant impact of war on all aspects of medieval European society. The study of medieval military history lags far behind that of the classical world, despite the immensely greater volume of all types of source materials. This is true, not least, because the study of warfare in the Middle Ages continues to be treated as a marginal topic. It is long past time, therefore, that historians of medieval warfare learn the lessons, techniques, sophisticated analytical tools, and omnivorous approach to the sources of their colleagues investigating the world of classical antiquity. It is only in this way that the field, as a whole, will be able to cast off the image of arm-chair generalship and antiquarianism and assert its proper role as a central and intrinsic element of human affairs throughout the medieval millennium. This volume brings together fourteen essays, which we wrote individually and together over the past twenty years, with the exception of the essay on the Bayeux Tapestry that originally was published in 1987. The volume is divided into two unequal parts. The first of these comprises ten studies that focus on the problems inherent in historical epistemology, with a particular emphasis on how scholars can understand and use source materials for the writing of military history. These include studies of the historiographical works composed by Gregory of Tours (died 594), Nithard (died 844), Widukind of Corvey (died c. 975), Thietmar of Merseburg (died 1018), Ademar of Chabannes (died 1034), Dudo of St. Quentin (died c. 1043), Bruno of Merseburg (died c. 1085), and Ralph of Caen (died c. 1120). The second section brings together four articles that focus on material sources of information that illuminate the conduct of war in early medieval Europe. The first of these is a review article that examines the paradigm of “landscape archaeology,” which provides a holistic approach to “reading” the natural and manmade topography to develop an understanding of decision-making regarding military affairs. The second article in this section examines the costs involved in the construction of fortifications in tenth-century Germany on the basis of archaeological findings and comparisons with building costs in better-documented periods of European history. The third essay analyzes the compositional history of the Bayeux Tapestry and its value for illuminating contemporary military affairs. The 2

INTRODUCTION

final article in this section draws upon the vast body of information developed by archaeologists concerning both fortifications and transportation routes to provide an examination of the famed campaign by King Henry I of Germany (919–936) in the winter of 928–929. All of the essays have been revised, in some cases substantially. We added considerably to the notes to account for scholarship that has been published in the time since these essays first appeared. We also have edited the essays extensively to both improve their clarity and to add material. We thank Michael Greenwood and the staff at Taylor & Francis for the opportunity to publish this volume and for their professionalism at all stages of the production process. It is our great pleasure to dedicate this volume to Karl F. Morrison, who has been a friend to us both, as well as a leader in methodological innovation.

3

Part 1 NARRATIVE WORKS

1 G R E G O RY O F TO U R S A S A M I L I TA RY H I S TO R I A N

By what criteria does one classify a particular writer as a military historian? In the Roman world and previously in the Greek-dominated Mediterranean, one of the most important ways by which would-be military commanders gained the competence to serve as officers was by the advice of men who were knowledgeable about war. Thus, it was widely recognized that histories written by such knowledgeable men were fundamental to passing on to posterity valuable information about how to be successful in war. In this context, Julius Caesar’s Commentaries on the Gallic and Civil Wars, Arrian’s Anabasis of Alexander, Ammianus Marcellinus’ Rerum Gestarum Libri, and Procopius’ account of Justinian’s wars are obvious examples of histories from which a great deal of practical knowledge could be gained by someone studying to become an officer.1 It is probably with this teaching of the ancients in mind that Isidore of Seville argued in somewhat broader terms that those men who would seek success in government should study history.2

1 J. Brian Campbell, “Teach Yourself How to Be a General,” Journal of Roman Studies 77 (1987), 13–29, provides a useful introduction to this rather neglected topic and more generally, idem, The Emperor and the Roman Army 31 B.C.–A.D.235 (Oxford, 1984), 325–32, provides a useful introduction to these matters for the period indicated in the title. 2 Isidore, Origines, 1.43 in Isidori Hispalensis episcopi Etymologiarum sive Originum libri xx, ed. W. M. Lindsey, 2 vols. (Oxford, 1911). Regarding Isidore’s view of history, see Jàcques Fontaine, Isidore de Séville et la culture classique dans l’Espagne wisigothique (Paris, 1959), 180–5; Benoît Lacroix, L’Historien au moyen âge (Paris, 1971), particularly Chapter 1, and the review article of this work by Robert W. Hanning, History and Theory 12 (1973), 419–34; A. J. Minnis, Medieval Theory of Authorship: Scholastic Literary Attitudes in the Later Middle Ages (London, 1984), particularly 1–20; D. H. Green, Medieval Listening and Reading: The Primary Reception of German Literature 800–1300 (Cambridge, 1994), 246–8; Gabrielle M. Spiegel, The Past as Text: The Theory and Practice of Medieval Historiography (Baltimore, 1997), 88–90; Almut Sauerbaum, “Accessus ad auctores: Autorkonzeption in mittelalterlichen Kommentartexten,” in Autor und Autorschaft im Mittelalter, ed. Elizabeth Andersen, Jens Haustein, Anne Simon, and Peter Strohschneider (Tübingen, 1998), 29–37; and Sebastian Coxon, ‘Zur Form und Funktion einiger Modelle der Autorenselbstdarstellung, ‘Wolfdietrich’ und ‘Dietrichs Flucht’,’ in ibid., 148–62, particularly, 151 and 162. Also see the translation of the work, The Etymologies of Isidore of Seville, trans. Stephen Barney, J. W. Lewis, J. A. Beach, and Oliver Berghof (Cambridge, 2006). On the use of books in training future officers, see Bernard S. Bachrach, “The Education of the ‘Officer Corps’ in the Fifth and Sixth

7

N A R R AT I V E W O R K S

Gregory of Tours provided his audience with a plethora of information concerning a vast spectrum of military activities in his Ten Books of History.3 Modern scholars have successfully combed the History for information, which they judge to be accurate, concerning a lengthy list of items from armaments to water supplies for garrisons.4 Indeed, Gregory’s picture of Merovingian military organization “resembles Romania rather than Germania.”5 Gregory’s consistent efforts, now well recognized, to deploy accounts of military violence for the purpose of convincing his readers of the grotesque seaminess of secular affairs does not disqualify him as a military historian.6 Gregory’s presentation of Merovingian military operations was so effectively distorted by his skewed emphases that modern scholars on occasion have misunderstood particular situations.7 However, Gregory not only was exceptionally well informed about contemporary military matters, he also understood and described a broad spectrum of basic principles useful to the education of an officer. It might be suggested that Gregory’s ability to project an effectively distorted picture of military matters to his readers stemmed from his detailed knowledge of both small matters and an understanding of the big picture. Although, as will be seen later, Gregory usually is somewhat negative when discussing military matters; on rare occasions he is positive. Thus, for example, Gregory provides a rather detailed description of the background and character of Aetius, the most famous general of the later Roman Empire, to provide a model for his readers: His father Gaudentius was a leading man in the province of Scythia; he began his service as a domesticus, and rose to the highest command as master of the horse. His mother was Italian, a noble and very wealthy lady. Aetius, their son, was made a praetorian as a boy; for three years he was a hostage of Alaric and after that of the Huns. Later, he became sonin-law of Carpilio, former count of the domestics, and was appointed to administer the palace of John. Of middle height, he was manly in appearance and well formed, neither too frail nor too heavy; he was smart and

3 4

5 6

7

Centuries,” in La noblesse romaine et les chefs barbares du IIIe au VIIIe Siècles, ed. Françoise Vallet and Michel Kazanksi (Paris, 1995), 7–13. Gregory of Tours, Historiarum libri X, ed. Bruno Krusch and Wilhelm Levison MGH SRM I.1 (Hanover, 1951). See, for example, Bernard S. Bachrach, Merovingian Military Organization (Minneapolis, 1972), and Margarete Weidemann, Kulturgeschichte der Merowingerzeit nach den Werken Gregors von Tours, 2 vols. (Mainz, 1982), II: 238–82. Bachrach, Merovingian Military Organization, 128. Walter Goffart, The Narrators of Barbarian History (A.D. 550–800): Jordanes, Gregory of Tours, Bede, and Paul the Deacon (Princeton, 1988), 219–22 with the review article by Bernard S. Bachrach, in Francia 17.1 (1990), 250–6. See, for example, the discussion by Bernard S. Bachrach, The Anatomy of a Little War: A Diplomatic and Military History of the Gundovald Affair (568–586) (Denver, 1995), 83–4, 246 n. 47.

8

G R E G O RY O F T O U R S A S A M I L I TA RY H I S T O R I A N

agile, a very well trained horseman and a skillful archer; he was tireless in the use of the spear. Although exceptionally talented in war, he was renowned for the arts of peace. He was without avarice and little moved by lust. He was graced with intellectual gifts and did not alter his goals because of any kind of evil instigation. He suffered wrongs with great patience, and loved labor. He was unafraid of danger and surpassed all in the endurance of hunger, thirst, and the watch.8 Gregory’s interest in having his readers accept this model of a great general is made clear by the claim that he found the information regarding Aetius in the History written by Renatus Frigeridus.9 The intended audience for Gregory’s work undoubtedly was assumed to have regarded Renatus as a reliable authority concerning matters which took place almost a century and a half earlier. The decision to quote, an unusual tactical device for Gregory, rather than either to paraphrase or to summarize Renatus’ views in this instance would seem to highlight the importance attributed by the bishop of Tours to having this information accepted. It is doubtful that Gregory’s readers were aware that Renatus likely acquired some of these data from Meroabaudes’s panegyric on Aetius.10 However, some may have been aware that many of the values that were expressed by Renatus regarding the character of a good general were found detailed in classical works such as the writings of Caesar and Cicero.11 Gregory has provided a traditional set of values for the good general. For example, Onasander’s work on the art of generalship (1.1) written in the mid-first century, held that “the general should show integrity and self-restraint; be sober, frugal, hardworking, alert, in fact a father of children if possible; he should be a good speaker and have a distinguished reputation.” Thus, the reader of Gregory’s Histories, like the reader of a substantial number of Cicero’s works, would be aware of the role played by “birth, upbringing, and inherited ability” in making a good general. In this tradition, moral excellence was very important, as was “doing one’s duty, good judgment, energy, concern, personal reputation and dignity.”12 Gregory also provides some very useful and traditional insights regarding the strategy of conquest. Gregory places these views in the mouth of Aridius, a clever Gallo-Roman advisor to the Burgundian king Gundobad. As part of a stratagem,

8 Gregory, His. 2.8. 9 Ibid. 10 Panegyricus, 2.2 106ff; Flavius Merobaudes reliquiae, ed. F. Volmer MGH, AA 14 (Berlin, 1905), 1–20. See also F. M. Clover, Flavius Merobaudes: A Translation and Historical Commentary (Philadelphia, 1971); and Idem, “Toward and Understanding of Merobaudes’ ‘Panegyric I’,” Historische Zeitschrift für alte Geschichte 20 (1971), 354–67. 11 See, for example, Campbell, “How to Be a General,” 13–29. 12 Ibid., 23 for these quotations and a discussion of them.

9

N A R R AT I V E W O R K S

Aridius is seen to advise Clovis while the latter was laying siege to the massively fortified urbs of Avignon: O king, if the glory of your highness (Gloria celsitudinis) deigns to hear from me a few words of humble advice, though, indeed, you have little need of counsel (consilium), I will offer them in complete faith; and it shall be useful to you, in general, and to the districts (civitates) through which you intend to pass. Why, he said, “do you keep this army (exercitus) in the field when your enemy (inimicus) sits in this exceptionally strong place (loco firmissimo)? You ravage the fields, you consume the meadows, you hack down the vines, you fell the olive trees, and all the fruits of this region, you completely destroy. Yet you do not prevail against your enemy. Rather, send a legatio to him and impose on him a yearly tribute (tributum . . . annis singulis) that he will pay to you so that this region may be saved. You will be lord (domineris) with the tribute (tributa) being paid in perpetuum.13 In effect, this advice given by Aridius encapsulates rather clearly the differences between the putative actuality of “barbarian” warfare and the teachings of ancient military science. The former is presumed to have seen an emphasis on taking booty, spreading terror, and wreaking destruction on both the human and material resources of the enemy. By comparison, the fundamental goals of classical strategy required the preservation of both human and material resources, whenever possible, and to obtain tribute or taxes from one’s adversaries on a regular basis. Corollary with the effort to develop secure and continuing sources of income was the imposition of an institutional structure through which the tribute payers recognized the dominium of those to whom they were subject. Consistent with this view, Alexander the Great kept his soldiers from ravaging Asia and is reported to have told them “they ought not to destroy what they were fighting to possess.”14 Closer in time to the events at hand and more relevant to the Salian Franks, in particular, is a policy attributed by Eunapius to the Emperor Julian. The latter is reported to have: Ordered the Romans to harm none of the Salii nor to ravage or plunder their territory. He told them [the Romans] that they should understand that all the land they had won without fighting and toil was theirs; thus while they must regard as enemy territory that which belonged to those at war with them, they must treat as their own that which belonged to those who had submitted to them.15 13 Gregory, Hist. 2.32. 14 J. F. C. Fuller, The Generalship of Alexander the Great (London, 1958), 285 where the work of Justin is quoted and discussed. 15 Eunapius, fr. 18 in The Fragmentary Classicising Historians of the Later Roman Empire: Eunapius, Olympiodorus, Priscus, and Malchus, ed. and trans. R. C. Blockley, 2 vols. (Liverpool, 1981).

10

G R E G O RY O F T O U R S A S A M I L I TA RY H I S T O R I A N

Indeed, according to Gregory, Clovis himself made clear that he had accepted this “ancient” way of doing things that had been advised by Aridius. In addition, when Clovis invaded Aquitaine in 507, he issued an order (edictum) that his troops should take only water and grass from the lands which they traversed. When some of his soldiers violated this order, he had them executed.16 It is not without interest that when Pippin and Charlemagne campaigned in Aquitaine, they followed the same strategy.17 Readers of Gregory’s Histories also had the opportunity to learn a great deal about sound campaign strategy. For example, prior to the beginning of the campaigning season of 584, Chilperic learned that Childebert II and King Guntram had formed an alliance. Information on this pact was followed by intelligence that Childebert had mustered an army, ostensibly for a campaign in Spain, but which the Neustrian ruler would appear to have believed was intended to attack him. Indeed, any Austrasian invasion of Spain very likely would have to pass through Neusterian territory. Thus, Chilperic sent a nuntius to his duces and to the comites in each of the civitates of his regnum ordering them to repair the walls of the urbes and to bring their troops and resources inside the defenses in order to withstand a siege. Chilperic himself mustered a field army (exercitus) and kept it in the field under his direct command.18 Despite Gregory’s meager description, it is clear that Chilperic was employing a defense-in-depth strategy. His urbes were to be well defended, in that the duces and comites not only were commanded to put their walls in good repair but also to bring their troops into the cities. These would then serve as strategic “hardpoints” which the enemy could not easily overwhelm. Chilperic himself stayed in the field with an army that could be brought up quickly against any force that was besieging one of his urbes. Should an enemy force refuse to raise its siege, the fortified hardpoint served as the potential anvil and Chilperic’s field army as the hammer. Consistent with the strategy of defense-in-depth, any besieging force would be caught between the two and seriously discomforted. Because field forces had been brought into the urbes, the cities would be very difficult to take by storm. In addition, these regular troops could sally from the mural defenses and in coordination with Chilperic’s army, catch the besieges in a pincer movement. Finally, troops

16 Gregory, Hist. 2.37. These views on the preservation of resources within the framework of proper religious observance and respect for God exercised by Clovis are echoed in a speech attributed by Gregory to King Guntram (8.30) with the observation, “How . . . are we to win victory in our day when we no longer keep the observances of our fathers?” Cf. the discussion by Martin Heinzelmann, “Histoire Rois et Prophètes. Le role de elements autobiographiques dans les Histoires de Grégoires de Tours: Un guide épiscopal à l’usage du roi chrétien,” in De Tertullien aux Mozarabes: antiquité tardive et christianisme ancient (IIIe-Vie Siècles. Mélanges offerts à Jacques Fontaine), vol. 1 (Paris, 1992), 537–50, here 547; and now Bernard S. Bachrach, “Vouillé in the context of the Decisive Battle Phenomenon,” in The Battle of Vouillé 507 EC: Where France Began, ed. Denuta Schanzer and Ralph W. Mathisen (Berlin, 2013), 11–42. 17 Capitularia regum Francorum I, ed. Alfred Boretius (Hanover, 1883), Cap 18 c. 6. 18 Gregory, Hist. 6.41 recounts Chilperic’s responses but not in their proper chronological sequence.

11

N A R R AT I V E W O R K S

within the cities were in position to interdict the enemy’s line of communication and cut their access to supplies. A clear indication that a defense-in-depth strategy was put in place was the recognition, reported by Gregory, that both Chilperic and his commanders understood that their unprotected assets in the countryside would be at risk and likely would be seriously damaged. Thus, the king guaranteed his magnates that these assets would be replaced to their profit. The recognition that unprotected and unprotectable assets are to be sacrificed in the short term is a key aspect of a defense-in-depth strategy.19 In the wake of Chilperic’s countermeasures, Childebert’s army refrained from invading Neustria. Also of some importance in this context is the fact that Chilperic is seen to understand that the morale of his commanders and through them of his troops had to be kept high and that he accomplished this by his guaranty not only to restore their losses but to reward them as well. Gregory illustrates a wide variety of tactical principles that may be considered to have been of value to an aspiring military commander. For example, Eunius Mummolus, the son of Peonius, was a native of Auxerre and succeeded his father as comes of the civitas with responsibility for both civil and military operations. Mummolus’s military success as count led King Guntram to appoint him patricius in command of the standing army of the Burgundian kingdom.20 It is possible that Mummolus learned something of strategy and tactics from written sources, although obviously not from Gregory’s text.21 In his initial military operation as the new patricius, Mummolus found it necessary to lead his troops against a Lombard invasion force that had come through Mount Cenis. Due to the availability of sufficiently advanced intelligence, Mummolus was able to await the enemy on the Plan de Fazi near Embrun before 19 Concerning the characteristics of a defense-in-depth strategy see the discussion by Edward N. Luttwack, The Grand Strategy of the Roman Empire from the First Century A.D. to the Third (Baltimore, 1976), 130–45. Although Luttwak’s attribution of a grand strategy to the Roman Empire have sustained considerable criticism, there is little doubt that the model of defense-in-depth has enormous heuristic value for understanding the approach to frontier defense by numerous Roman emperors as well as their successors in the West. See, in this context, Alexander Sarantis, “Waging War in Late Antiquity,” in War and Warfare in Late Antiquity, ed. Alexander Sarantis and Neil Christi, 2 vols. (Leiden, 2013), 1–98, particularly 7–17. With specific regard to the preparations made by Chilperic, cf. Walter Goffart, “Byzantine Policy in the West under Tiberius II and Maurice: The Pretenders Hermenegild and Gundovald (579–585),” Traditio 13 (1957), 73–118, here 109, who concludes from Gregory’s account (6.4) that Chilperic acted like a coward. Whether it was Gregory’s intention to leave this impression is not clear. Indeed, one may even question whether Gregory understood Chilperic’s strategy. In any case, if the bishop of Tours had understood what was being done, it is doubtful that he would have been inclined to give credit to a military plan of any type, especially a plan that Chilperic had implemented. 20 Gregory, Hist. 4.4. K. Selle-Hosbach, Prosopographie merowingischer Amtsträger 511–613 (Bonn, 1974), 133–6, n. 151; and for Mummolus’s military career see Bachrach, Merovingian Military Organization, 38–41, 44, 47, 54–5, 59, 67. Cf. Weidemann, Kulturgeschichte, II: 245–6. 21 Bachrach, “Education,” 10.

12

G R E G O RY O F T O U R S A S A M I L I TA RY H I S T O R I A N

the Lombards could deploy in various directions to attack farther to the west and south (see later). However, Mummolus chose not to confront the enemy in open battle, but rather blockaded the roads in this very mountainous region, thereby cutting the Lombard line of communications and their access to supplies. In response, the Lombards were forced to go into the local forests to forage. Mummolus deployed units, probably raised locally, in these trackless woodlands, thereby taking advantage of their knowledge of the terrain. Mummolus’s troops destroyed the greater part of the Lombard force in detail as it was forced into smaller and smaller groups in order to forage. The few remaining survivors were taken as prisoners of war.22 Mummolus’s second campaign against the Lombards, the next year, was very different from the first. The patricius received intelligence that the enemy had penetrated the territory of the regnum Francorum only after they had passed through Embrun and had divided into three groups. These columns were already operating in various parts of southeastern Gaul when Mummolus moved into action. Duke Amo led one force in an extended raid through Embrun, into the region of Avignon and then south into the Arlate and the environs of Marseilles. From there Ammo moved his troops north and prepared to besiege the fortified city of Aix. Duke Zaban moved almost due west to the environs of the city of Valence and there established a fortified encampment (castra) preparatory to a siege of the urbs. Duke Rodan marched to the very well-fortified city of Grenoble on the Isère River, where he prepared to undertake a siege.23 Mummolus chose to attack Rodan’s forces first and thereby relieve the pressure on Grenoble. The Lombards, however, would seem to have felt secure in their siege encampment because they doubted that an enemy could cross the Isère. Mummolus then took advantage of the Lombards’ apparent complacency in this regard. The patricius negotiated a very difficult crossing of the river, which Gregory attributes to God’s help, launched a surprise attack, and soundly defeated the Lombards.24 The survivors of the badly mauled army, with their commander wounded, fled west from Grenoble and joined Zaban’s forces at Valence. The latter raised his siege, apparently also abandoning the booty he had collected in the countryside, and the two dukes undertook a rather precipitous retreat due east toward the Alpine pass at Mount Cenis. Rather than pursue Rodan, Mummolus obtained intelligence concerning the direction of the enemy retreat and used this information to take advantage of his interior lines of transportation. He moved his forces south-southeast to block the Lombards at Embrun. As Gregory, who surely was no

22 Gregory, Hist. 4.42. 23 Ibid., 4.44. 24 Ibid.

13

N A R R AT I V E W O R K S

supporter of Mummolus reports,25 the patricius met the Lombards at Embrun with an immense army (innumero exercitu) and cut the invaders to pieces.26 Even given the defective and hostile description of Mummolus’s military activities provided by Gregory, the reader of these chapters of the Histories had much to learn. When Mummolus had the opportunity to destroy the enemy without offering battle, as was the case at Plan de Fazi, he used various stratagems and prevailed. However, when Mummolus chose to confront the enemy in open battle at Grenoble, he took advantage of the element of surprise. By doing so he was able to select the field where he would encounter the enemy and under what conditions. At Embrun, Mummolus took advantage of the massive numerical superiority of his forces and won a decisive victory.27 Among the vast array of tactical principles illustrated by Gregory in the Histories are many far simpler to grasp then those regarding the military operations conducted by Mummolus. For example, in the spring of 531, King Theudebert I led what obviously was a substantial invasion force which was intended to bring the Thuringian kingdom of Hermanfird under direct Frankish control. Hermanfrid’s scouts reported that Theudebert’s army was on the march toward the Unstrut and identified the place where the Merovingian ruler most likely would try to cross the river. Gregory reports: When the Franks approached, the Thuringians prepared traps. They dug ditches in the future field of battle, and covered over the openings very neatly with sods so that the ditches seemed to be part of an unbroken plain. When they began the battle, many of the Frankish horsemen, therefore, charged and fell into these ditches. It was for them a very serious discomfiture. But when this stratagem was discovered, they [the Franks] began to reconnoiter.28 From Gregory’s description of this episode, several important tactical principles could easily be deduced by the intelligent reader. For the defenders, substantial forewarning of an enemy attack could be very useful. In addition, intelligence

25 It must be kept in mind that Gregory, the basic source for Mummolus’s activities, was fundamentally hostile to the general, and thus anything the latter is described as doing may be thought to be tainted by a highly negative parti pris. In addition, Gregory not only had no abiding interest in providing accurate details of military operations but was, in general, hostile to viri fortes. When such details are provided, Gregory usually has some purpose in mind other than the aggrandizement of military commanders. See Goffart, Narrators, 161, 180, 217, 219. 26 Gregory, Hist. 4.44. 27 From other engagements too numerous to discuss in detail it is clear that Mummolus employed trickery and surprise, avoided potentially risky encounters, chose the battlefield when he decided to fight, prepared fortifications carefully, looked after logistic needs, slaughtered foragers in surprise raids, forced the enemy to surrender its plunder, and concentrated overwhelming force before going into battle. 28 Gregory, Hist. 3.7.

14

G R E G O RY O F T O U R S A S A M I L I TA RY H I S T O R I A N

concerning the likely combat techniques of which the enemy were capable, i.e. a mounted charge in the present context, can be seen to be exceptionally desirable. With such intelligence provided either by spies or scouts, the force on the defensive had the opportunity to select a field of battle that favored its strengths. In addition, both time and intelligence permit the defensive force to implement whatever stratagems might most discomfort the enemy’s likely mode of attack. Finally, the specific stratagem of placing traps through the preselected field of battle is shown to have considerable value when an enemy force is wont to make a mounted charge. The lessons for the reader of this passage who might in the future be leading an offensive force are no less important. First, those planning the campaign should not make the line of march or particular destinations so obvious to those they intend to attack that traps or other insidiae might be possible. Consequently, an effort to disguise the line of march should be made in order to deprive the enemy of the time and the opportunity to choose and prepare the field of battle to its advantage. In addition, intelligence regarding the enemy, obtained either through spies or scouts, should be implemented to avoid particular kinds of traps, such as those set by the Thuringians. Siege warfare dominates the history of the Romano-German kingdoms that succeed the empire in the West.29 Consequently, like the authors of late antique military handbooks, Gregory makes clear that “cities and forts are either fortified by nature or by human hand or by both.”30 He provides good, if partial, descriptions of well-designed fortifications from which future military architects could learn. For example, Gregory discusses some of the strong points of the fortified urbs of Lugdunum Convenarum (modern Saint-Bertrand-de-Comminges) in the context of a siege carried out in 585. Gregory observes that “the city crowns an isolated height, with no other nearby”31 and emphasizes on two occasions that a deep valley surrounds the walls.32 Gregory also emphasizes the ways in which nature can be helped by man to improve fortifications. When Mummolus took command of the urbs at Avignon, Gregory indicates that the patricius recognized that the old Roman defenses were not sufficiently prepared. Gregory, who seems to have obtained information from an eyewitness, states: When he [Mummolus] first entered the urbs, he noted that a small part of it was undefended by the Rhône River. Thus, he took measures to have a channel led from the river to protect the entire part of the wall that had

29 See, Jim Bradbury, The Medieval Siege (Woodbridge, 1992); Bernard S. Bachrach, “Medieval Siege Warfare: A Reconnaissance,” The Journal of Military History 58 (1994), 119–33; and Peter Purton, A History of the Early Medieval Siege, c. 450–1220 (Woodbridge, 2009). 30 Flavius Renatus Vegetius, Epitoma rei militaris, ed. Alf Önnerfors (Stuttgart, 1995), 4.1. See also the English translaton by N. P. Milner, Vegetius: Epitome of Military Science, 2nd revised edition (Liverpool, 1996). 31 Gregory, Hist. 7.34 32 Ibid., 7.37–8.

15

N A R R AT I V E W O R K S

been exposed. Moreover, before this was done, he had very deep pits dug in the bed of the channel.33 Gregory then explains that the pits were dug so that anyone wading into the channel in the hope of crossing on foot would easily slip into the unseen holes and drown, either because of the weight of his armor or simply because he could not swim.34 In another instance, Gregory discusses the importance of a protected water supply and observes with regard to the urbs at Saint-Bertrand: “A great spring issuing from the foot of the hill is enclosed by a very strong tower; men go down to it by a covered way and thus draw water without being exposed to view.”35 It is worthwhile comparing this description with some remarks on the same topic by Vegetius: It is of great usefulness for an urbs that natural springs are contained within the walls. . . . Fortifications that have been constructed where a source of water is outside and below the walls . . . are well advised to build a small fortifications called a burgus . . . to defend the water supply from the enemy.36 For Gregory, the fortification at Dijon provided a sort of model concerning what was best: It is a fortified place (castrum) with very strong walls (firmissimis muris), built in the middle of a plain. . . . Four gates (portae) face the four corners of the earth and thirty-three towers (torres) guard the [circuit] walls. These towers are built of squared stones to a height of twenty feet and above these are courses of smalls stones. The total height of the walls comes to thirty feet and they have the thickness of fifteen feet. Why this place is not called a civitas I do not know.37 It is not without significance that Gregory’s report regarding these fortifications is exceptionally accurate, as archaeological studies have documented.38 The dissolution of the imperial government in the West did not bring with it either the disintegration of the late Roman city or the military obsolescence of its

33 Ibid., 6.26. 34 Ibid., 6.26. 35 Ibid., 7.34. See the discussion by Bachrach, Anatomy of a Little War, 122, from which it might seem that Gregory had a fundamental understanding of military matters and may have had access to another military handbook or excerpts from such a handbook. 36 Vegetius, De re militari, 4.10. 37 Gregory, Hist. 3.19. 38 Stephen Johnson, Late Roman Fortifications (Totowa, NJ, 1983), 74, 84–6, 113, 249.

16

G R E G O RY O F T O U R S A S A M I L I TA RY H I S T O R I A N

fortifications.39 Indeed, Gregory’s account of preparations for the defense of fortifications also provide lessons. As observed earlier in the discussion of the defensein-depth strategy, Gregory makes a point of the importance of keeping strongholds in defensible condition. He makes clear, as well, that when a siege is contemplated, the gathering of foodstuffs becomes critical. In this context, Gregory has one commander say to the assembled multitude of the city and its environs: “bring within the walls all of your provisions and all of your equipment.”40 In addition, Gregory takes careful note of the fully stocked granaries and warehouses that it was wise to keep within the city in case of a siege so that the garrison might be fed for as long as necessary. It is clear that Gregory not only understood that a lack of food could result in a need to surrender but he seems to have grasped the important psychological fact that a fear of not having sufficient supplies could cause severe morale problems.41 Gregory observes with approval the behavior of besieged men who kept caldrons filled with hot tar and pitch available on the walls in order to repel those enemy troops who would attempt to storm the city’s defenses. He also takes positive note of the use of stone-throwing devices by the defenders.42 Gregory provides information regarding scores of strategic principles from the importance of a protected water source within a fortification, mentioned earlier, to avoiding wars on two fronts simultaneously. Indeed, the latter point is very neatly made by the Merovingian king Chlodomer, who, after capturing the Burgundian king Sigismund, was preparing to kill him prior to attacking the latter’s brother, King Godomar. When Abbot Avitus of Micy attempted to dissuade Chlodomer from executing Sigismund, Gregory reports that the Merovingian king said: I think that it is stupid counsel if as a result I may be defeated because I have left some of my enemies behind me when I march against the rest of them. If some were to come upon me from the rear and others from the front, I should be defeated between the forces of the two battle lines of the enemy. I shall win my victory both more serviceably and more easily if I keep one army separated from the other. Thus, having killed off one enemy, the other can more easily be sent to his death.43 In the sphere of tactics, Gregory provides many examples of poor generalship from which the aspiring military officer could learn. Thus, the costs paid by commanders who failed to encamp their troops in a proper manner,44 failed to reconnoiter,45 39 In this context, see the comprehensive study of siege warfare in the late antique West as well as the Byzantine and Islamic worlds by Leif Inge Ree Petersen, Siege Warfare and Military Organization in the Successor States (400–800) (Leiden, 2013). 40 Gregory, Hist. 7.34. 41 Ibid., 7.37. 42 Ibid. 43 Ibid., 3.6. 44 Ibid., 6.31. 45 Ibid., 3.7; 6.26; 9.31.

17

N A R R AT I V E W O R K S

failed to protect foragers,46 failed to secure information regarding enemy troop movements,47 provided inadequate logistic support,48 lacked discipline,49 and failed to test equipment in order to detect enemy sabotage50 are among the many negative lessons to be learned by the attentive reader. The bishop of Tours was very much interested in emphasizing the consequences of failed military operations, and he was particularly eager to provide examples where imprudence, greed, and carelessness led to serious problems, if not to outright disaster. Indeed, it is a curious aspect of Western behavior, especially in military matters, that it is easier to learn the lesson taught by a negative example than by a positive one. This is the case perhaps because positive examples would seem to permit the student with initiative to speculate that the same good result probably could be obtained in another and even more efficient way. This habit of mind obviously can lead to important innovations, e.g. the proverbial better mouse trap. The negative example, however, shows the direct consequence of a mistake, and thus the connection of cause and effect is ineluctable. In the present context, the intelligent reader of Gregory’s Histories easily could see that the virtues attributed to Aetius were the contra positive of the vices displayed by commanders who led failed military operations.51 Indeed, even the rather inattentive reader could easily make a list or have a list made for him of those behaviors from which the would-be successful military commander was likely to profit by avoiding. To put it another way, Gregory of Tours, like most good military writers and even many not-so-good military writers, provides a panoply of maxims regarding proper and improper military behavior.52

46 47 48 49 50 51 52

Ibid., 7.35. Ibid., 3.7. Ibid., 4.42, 6.6 and cf. Vegetius, De re militari, 3.3. Ibid., 3.4; 4.49–50; 7.21 and 36. Ibid., 6.26. Ibid., 3.6; 7.29; 9.31; 10.9. See, for example, Ralph D. Sawyer with Mei-chün Sawyer (trans. and commentary), The Seven Military Classics of Ancient China (Boulder, 1993); and Vegetius, De re militari, 3.26.

18

2 N I T H A R D A S A M I L I TA RY H I S TO R I A N O F T H E C A R O L I N G I A N E M P I R E C. 833–843

Introduction Despite the substantially greater volume of sources that provide information about the military affairs of the ninth century as compared to the eighth, the lion’s share of scholarly attention concerning Carolingian military history has been devoted to the reign of Charlemagne, particularly before his imperial coronation in 800, rather than to his descendants.1 The relative neglect of the ninth century, particularly in Francia occidentalis, by specialists in military history has led to considerable confusion about the nature and conduct of war in this period.2 Indeed, the field has been dominated by a small handful of studies that draw upon an impressionistic selection of texts and consequently offer misleading conclusions about the nature of warfare in the Regnum Francorum. 1 The outpouring of studies on Carolingian military history during the reign of Charlemagne is vast. For an introduction to this topic, see Etienne Renard, “La politique militaire de Charlemagne et la paysannerie franque,” Francia 36 (2009), 1–33; Walter Goffart, “Frankish Military Duty and the Fate of Roman Taxation,” Early Medieval Europe 16 (2008), 166–90; Carroll M. Gillmor, “The 791 Equine Epidemic and Its Impact on Charlemagne’s Army,” Journal of Medieval Military History 3 (2005), 23–45; John France, “The Composition and Raising of the Armies of Charlemagne,” Journal of Medieval Military History 1 (2002), 61–82; Bernard S. Bachrach, “Charlemagne’s Cavalry: Myth and Reality,” Military Affairs 47.3 (1983), 181–7; Robert-Henri Bautier, “La Campagne de Charlemagne en Espagne (778): La réalité historique,” Société des sciences, lettres, et arts de Bayonne ns 135 (1979), 1–5. Also see Bernard S. Bachrach, Early Carolingian Warfare (Philadelphia, 2001) and Idem, Charlemagne’s Early Campaigns (768–777): A Diplomatic and Military Analysis (Leiden, 2013) with the literature cited there. 2 There are some notable exceptions to this otherwise considerable lacuna, particularly with regard to the lands east of the Rhine. See, for example, Charles R. Bowlus, Franks, Moravians, and Magyars: The Struggle for the Middle Danube, 788–907 (Philadelphia, 1995); Eric J. Goldberg, Struggle for Empire: Kingship and Conflict under Louis the German, 817–876 (Ithaca, NY, 2006), particularly 119–146; Simon Coupland, “The Carolingian Army and the Struggle against the Vikings,” Viator 35 (2004), 49–70; and Ekkehard Eickhoff, “Maritime Defence of the Carolingian Empire,” in Vikings on the Rhine: Recent Research on Early Medieval Relations between the Rhinelands and Scandinavia, ed. Rudolf Simek and Ulrike Engel (Vienna, 2004), 50–64.

19

N A R R AT I V E W O R K S

Typical in this regard is the influential article by Josef Fleckenstein concerning the putative decline of militia forces used as expeditionary levies following Charlemagne’s reign. Relying on just two capitularies, one issued in 808 and the other in 847, Fleckenstein asserted that the mobilization of small-scale landowners had ended completely by the latter date.3 Obviously, the use of just two capitularies in this manner is fraught with methodological problems. And, indeed, scholars have identified the continued mobilization of such levies by the successors of the Carolingians during the tenth and eleventh century, both east and west of the Rhine.4 Nevertheless, Fleckenstein’s essay has exercised considerable influence over subsequent views of ninth-century military organization, particularly in the Germanlanguage tradition.5 Timothy Reuter similarly has exerted significant influence over scholarly understanding of Carolingian military history during the ninth century through two articles that are based on a narrow selection of capitularies and passages from a limited group of narrative texts.6 Reuter asserted, like Fleckenstein, that militia-based expeditionary levies of small-scale farmers disappeared after c. 800. Reuter also claimed that there was a concomitant end to Carolingian military expansion because of the

3 Josef Fleckenstein, “Adel und Kriegertum und ihre Wandlung im Karolingereich,” Settimane di Studio del Centro Italiano di Studi sull’alto Medioevo 27 (1987), 2 vols., I, 67–94, here 89–92. 4 With regard to the west, see Marjorie Chibnall, “Military Service in Normandy Before 1066,” Anglo-Norman Studies 5 (1982), 65–77; Emily Zack Tabuteau, “Definitions of Feudal Military Obligations in Eleventh-Century Normandy,” in On the Laws and Customs of England: Essays in Honor of Samuel E. Thorne, ed. M. Arnold (Chapel Hill, NC, 1982), 18–59; David Bates, Normandy Before 1066 (London, 1982), 122–5, 156–7; and Bernard S. Bachrach, Fulk Nerra-the Neo Roman Consul: A Political Biography of the Angevin Count (987–1040) (Berkeley, 1993). With respect to the lands in the east, see David S. Bachrach, Warfare in Tenth-Century Germany (Woodbridge, 2012); and idem, “Feudalism, Romanticism, and Source Criticism: Writing the Military History of Salian Germany,” Journal of Medieval Military History 13 (2015), 1–26. 5 See, for example, Hagen Keller, “Grundlagen ottonischer Königsherrschaft,” in Reich und Kirche vor dem Investiturstreit: Vorträge beim wissenschaftlichen Kolloquium aus Anlaß des achtzigsten Geburtstag von Gerd Tellenbach, ed. Karl Schmid (Sigmaringen, 1985), 17–34, here 21, n. 14; and Werner Rösener, “Rittertum und Krieg im Stauferreich,” in Staat und Krieg vom Mittelalter bis zur Moderne, ed. Werner Rösener (Göttingen, 2000), 37–63, here 40. 6 These are Timothy Reuter, “Plunder and Tribute in the Carolingian Empire,” Transactions of the Royal Historical Society, 5th series 35 (1985), 75–94; and Idem, “The End of Carolingian Military Expansion,” in Charlemagne’s Heir: New Perspectives on the Reign of Louis the Pious (814–840), ed. Peter Godman and Roger Collins (Oxford, 1990), 391–405. Regarding the influence of Reuter’s model of the decline of the expeditionary levy and end of Carolingian military expansion c. 800 see, for example, Janet Nelson, “Ninth-Century Knighthood: The Evidence of Nithard,” in Studies in Medieval History Presented to R. Allen Brown, ed. Christopher Harper-Bill (Woodbridge, 1989), 255–66, repr. in The Frankish World 750–900 (London, 1996), 75–87, here 78; Karl Leyser, “Early Medieval Warfare,” in Battle of Maldon: Fiction and Fact, ed. Janet Cooper (London, 1993), 87–108, repr. in Communications and Power in Medieval Europe: The Carolingian and Ottonian Centuries, ed. Timothy Reuter (London, 1994), 29–50, here 34; Matthew Innes, State and Society in the Early Middle Ages: The Middle Rhine Valley, 400–1000 (Cambridge, 2000), 144; and Simon MacLean, Kingship and Politics in the Late Ninth Century: Charles the Fat and the End of the Carolingian Empire (Cambridge, 2003), 16–17.

20

N I T H A R D A S A M I L I TA RY H I S T O R I A N

putative decline in the numbers of men who were available for military service.7 However, these assertions are not supported by a broader corpus of sources that provide information about either the continued deployment of militia levies or the large scale of Carolingian military operations during the ninth century.8 The focus by both Fleckenstein and Reuter on the question of the supposed discontinuity of military levies of militia forces is consistent with the long historiographical tradition, dating back to Heinrich Brunner’s landmark study “Der Reiterdienst und die Anfänge des Lehnwesens,” in which he sought a sociomilitary explanation for the putative transition from “the nation in arms” to “feudal knights.”9 This social-historical approach to military matters was conspicuous for failing to address the much wider range of questions relating to the conduct of warfare. These include matters such as logistics, training, sieges, and military technology, which have been considered in much more detail by specialists in other fields, particularly Anglo-Saxon and Anglo-Norman history.10 It is clear from a methodological perspective that a reappraisal of the conduct and nature of war in the ninth-century Regnum Francorum requires the investigation of a wide range of topics that is based upon a thorough analysis of all pertinent source materials. These include not only texts but also information drawn from archaeological excavations, which are available in substantial numbers for the Carolingian period.11 However, the process of analyzing and, just as importantly, vetting source materials that shed light on the military history of the Carolingian Empire in the period after Charlemagne is in its infancy. In fact, so little analysis of the sources has been undertaken by military historians that it is not yet possible 7 Reuter, “Plunder and Tribute,” 75–94. 8 For critiques of Reuter’s methodology and conclusions, see Renard, “La politique militaire de Charlemagne,” 14 and 22; France, “The Composition and Raising of the Armies of Charlemagne,” 66; Eric Goldberg, “Ludwig der Deutsche und Mähren. Eine Studie zu karolingischen Grenzekriegen im Osten,” in Ludwig der Deutsche und seine Zeit, ed. Wilfried Hartmann (Darmstadt, 2004), 67–94, here 70; and Idem, Struggle for Empire: Kingship and Conflict under Louis the German, 817–876 (Ithaca, 2006), 120. 9 Heinrich Brunner, “Der Reiterdienst und die Anfänge des Lehnwesens,” Zeitschrift der SavignyStiftung für Rechtsgeschichte, Germanistische Abtheilung 8 (1887), 1–38, reprinted in Brunner, Forschungen zur Geschichte des deutschen und französischen Rechts (Stuttgart, 1894), 39–74. See in this regard the observations by Bachrach, “Feudalism, Romanticism, and Source Criticism,” here 3–8. 10 This literature in this area is vast. Some useful starting points include C. Warren Hollister, The Military Organization of Norman England (Oxford, 1965); Nicholas P. Brooks, “England in the Ninth Century: The Crucible of Defeat,” Transactions of the Royal Historical Society series 5.29 (1979), 1–20; Michael Jones, “The Logistics of the Anglo-Saxon Invasions,” in Naval History: The Sixth Symposium of the United States Naval Academy, ed. D. M. Masterson (Wilmington, 1987), 62–9; John O. Prestwich, “Military Intelligence under the Norman and Angevin Kings,” in Law and Government in Medieval England and Normandy: Essays in Honour of Sir. James Holt, ed. George Garnett and John Hudson (Cambridge, 1994), 1–30; and Richard Abels, “The Costs and Consequences of Anglo-Saxon Civil Defense, 878–1066,” in Landscapes of Defense in Early Medieval Europe, ed. John Baker, Stuart Brookes, and Andrew Reynolds (Turnhout, 2013), 195–222. 11 See the discussion of this issue by Bernard S. Bachrach and David S. Bachrach, “Landscapes of Defense: At the Nexus of Archaeology and History in the Early Middle Ages,” Francia 42 (2015), 231–52, and in this volume.

21

N A R R AT I V E W O R K S

to write the history of warfare in the ninth century. The burden of this present study, therefore, is to take an important first step toward this goal by assessing the information regarding military affairs that is provided in the exceptionally important Histories, written by Charlemagne’s grandson Nithard between 840 and 842. This task is two-fold, requiring first an assessment of the reliability of this author in providing accurate information about military affairs and second an analysis of what Nithard has to say about the conduct of war.

Evaluating Nithard and his text The assessment of Nithard’s Histories and the information that he provides regarding military matters must begin with an understanding of his background, his knowledge of military affairs, the purpose for which he wrote, and the intended audience for the text. The answers to these questions will illuminate not only whether Nithard had available information about the conduct of war but also whether he had an interest in providing this information in an accurate manner.12 In this context, it is noteworthy that Nithard enjoyed an exceptional education and had a lengthy career, including duties as a military commander. Nithard was the illegitimate son of Angilbert, the lay abbot of St. Riquier (790–814), and Charlemagne’s daughter Bertha. Born at the royal court no later than 800 and educated in the palace complex at Aachen, this illegitimate grandson of Charlemagne served as an official both under his uncle, Emperor Louis the Pious (814–840), and under his cousin, King Charles the Bald of West Francia (840–877). Nithard received benefices from Emperor Louis for his highly regarded service13 and very likely participated in military campaigns during the 820s and 830s.14 Nithard certainly served in a military capacity under Charles the Bald, fighting at the battle of Fontenoy on 25 June 841, participating in high-level discussions of strategy with the king, and ultimately dying in combat near Angoulême while fighting Viking raiders in 844.15 In addition to his military duties, Nithard undertook a number of high-level diplomatic missions on behalf of his cousin, including

12 Regarding the interrogation of narrative sources for the purposes of writing military history, see David S. Bachrach, “Memory, Epistemology, and the Writing of Early Medieval Military History: The Example of Bishop Thietmar of Merseburg (1009–1018),” Viator 38 (2007), 63–90. 13 Nithard, Historiarum libri quattuor, ed. G. H. Pertz, 3rd edition, MGH SRG 44 (Hanover, 1965), 2.2. There is a new French translation of the edition of this text Philippe Lauer from 1926. See Nithard: Histoire des fils de Louis le Pieux, ed. and trans. Philipe Lauer and revised by Sophie Glansdorff (Paris, 2012). There are no significant differences between the new MGH edition and the revised Lauer edition. 14 Janet Nelson, “Public Histories and Private History in the Work of Nithard,” Speculum 60.2 (1985), 251–93, here, 269 makes this suggestion with regard to Nithard, in part on the basis of the information that Nithard shared about Louis’s military campaigns during this period. 15 For Nithard’s participation in the fighting at Fontenoy, see Historiarum, 2.2. For Nithard’s death in battle, see F. L. Ganshof, “Note critique sur la biographie de Nithard,” Mélanges Paul Thomas (Bruges, 1930), 515–21.

22

N I T H A R D A S A M I L I TA RY H I S T O R I A N

acting as an emissary to Charles’s eldest brother and frequent adversary Lothair I (840–855), and as one of the twelve commissioners given the task of dividing the Carolingian Empire equally among Louis the Pious’s three surviving sons in 842– 843.16 Consequently, it is possible to conclude with great confidence that Nithard was an expert in military affairs and had the knowledge to provide detailed information about the conduct of war. The next question that must be asked, therefore, is whether he had an interest in providing an accurate account of what he knew.

Audience for the text and causa scribendi Nithard’s text has received considerable attention from scholars with regard to both the purpose for which he wrote and his intended audience.17 Some scholars, including Hans Patze and Hans-Werner Goetz, have argued that Nithard wrote in order to promote the cause of justice or to discuss the obligations of both kings and magnates to serve the public good.18 Perhaps the most detailed and comprehensive treatment of Nithard, however, has been by Janet Nelson, who made a compelling case that the Histories were intended to serve a propaganda purpose for an audience of secular magnates at Charles the Bald’s court. In particular, Nelson contended that Charles the Bald commissioned Nithard to write the first three books of the Histories in order to support Charles’s claims to rule in West

16 Nithard, Historiarum, 2.2, and 4.1. See the discussion of the process that led ultimately to the treaty of Verdun, including the role played by Nithard, in F. L. Ganshof, “Zur Entstehungsgeschichte und Bedeutung des Vertrages von Verdun (843),” Deutsches Archiv für Erforschung des Mittelalters 12 (1956), 313–30; and translated as “The Genesis and Significance of the Treaty of Verdun (843),” by Janet Sondheimer in Ganshof, The Carolingians and the Frankish Monarchy: Studies in Carolingian History (London, 1971), 289–302. 17 More recently, scholars have sought to mine Nithard’s work for the information that it might provide about ninth-century concepts of masculinity and family bonds. Although certainly an interesting area of inquiry in its own right, this particular question has only a tangential connection with an inquiry focused on the Carolingian way of war. For studies focused on these issues, see Dana Polanichka, “‘As a Brother Should Be’, Siblings, Kinship, and Community in Carolingian Europe,” in Kinship, Community and Self: Essays in Honor of David Warren Sabean, ed. Jason Coy, Benjamin Marshke, Jared Poley, and Claudia Verhoeven (New York, 2015), 23–36; Meg Leja, “The Making of Men, Not Masters: Right Order and Lay Masculinity According to Dhuoda and Nithard,” Comitatus 39 (2008), 1–40 with the literature cited there; and Stuart Airlie, “The World, the Text and the Carolingians: Royal, Aristocratic and Masculine Identities in Nithard’s Histories,” in Lay Intellectuals in the Carolingian World, ed. Patrick Wormald and Janet L. Nelson (Cambridge, 2007), 51–76; and reprinted in Stuart Airlie, Power and Its Problems in Carolingian Europe (Farnham, 2012) with the same pagination. 18 See Hans Patze, “Iustitia bei Nithard,” in Festschrift für Hermann Heimpel zum 70. Geburtstag am 19. September 1971, 3 vols. (Göttingen, 1972), 3, 147–65; and Hans-Werner Goetz, “Staatsvorstellung und Verfassungswirklichkeit in der Karolingerzeit, untersucht anhand des Regnum-begriffs in erzählenden Quellen,” in Zusammenhänge, Einflüsse, Wirkungen. Kongreßakten zum ersten Symposium des Mediävistenverbandes in Tübingen 1984, ed. Joerg O. Fichte, Karl Heinz Göller, and Bernhard Schimmelpfennig (Berlin, 1986), 229–40.

23

N A R R AT I V E W O R K S

Francia and to depict Lothair I’s actions in a negative light.19 On the basis of newly developed information regarding the manuscript tradition for Nithard’s work, as well as evidence that other contemporary writers had access to the full text of the Histories, Nelson subsequently revised her argument, concluding that the entire text was intended to serve the ends of both royal propaganda and education.20 In this vein, Nelson’s argument certainly is enhanced by the fact that Nithard wrote in a plain style that would have been accessible to an audience of secular aristocrats.21 Nelson’s conclusions regarding both the purpose and audience for Nithard’s Histories subsequently were affirmed by Stuart Airlie, who places Nithard’s work within a broad context of writing by laymen, who were motivated to comment upon and attempt to influence events during the political crises of the 830s and 840s.22 In discussing Nithard’s audience, Airlie describes the Histories as a “mirror for the aristocracy as well as for princes.”23 The compelling arguments made by Nelson and Airlie, specifically that Nithard had a fundamentally didactic as well as propagandistic purpose and that his intended audience was composed of fellow secular aristocrats and members of Charles the Bald’s court, have important implications for interpreting the information that he provides about military matters. First, insofar as Nithard’s discussion of events and the decisions made by the Carolingian kings and their commanders did not impinge directly on his anti-Lothair parti pris, the author had an incentive to ensure the goodwill of the audience by providing accurate information insofar as he was able.24 This incentive was strengthened even further by the fact that Nithard’s audience consisted of men who were intimately familiar with the subject matter of the Histories because many of them had served in the king’s army on campaign. Nithard’s didactic purpose also provided him with a strong incentive to provide accurate information so as to give proper lessons with respect to effective and ineffective military decision-making. The use of historical works for this purpose was very well established at the court of Charlemagne, where Nithard was raised, and in the courts of Louis the Pious and Charles the Bald by whom Nithard had been employed. Indeed, the Carolingians fully embraced the argument by Isidore of 19 Janet Nelson, “Public Histories and Private History in the Work of Nithard,” Speculum 60.2 (1985), 251–93. 20 Janet Nelson, “History-Writing at the Court of Louis the Pious and Charles the Bald,” in Historiographie im frühen Mittelalter, ed. Anton Scharer and Georg Scheibelreiter (Vienna, 1994), 435–42. 21 The unadorned nature of Nithard’s text was stressed by Franz Brunhölzl, Geschichte der lateinischen Literatur des Mittelalters (Munich, 1975), I, 399–400; but also see the observation by Nelson, “Public Histories,” 253 that unadorned does not mean simple or simplistic. 22 Stuart Airlie, “The World, the Text,” 61. But also see Polanichka, “‘As a Brother Should Be’,” 23–36; and Dana M. Polanichka and Alex Cilley, “A Very Personal History of Nithard: Family and Honour in the Carolingian World,” Early Medieval Europe 22 (2014), 171–200, for an interpretation of the text as an expression of Nithard’s family identity. 23 Airlie, “The World, the Text,” 71. 24 This point has recently been made with respect to Carolingian sources specifically dealing with military matters by Thomas Scharff, Die Kämpfe der Herrscher und der Heiligen: Krieg Und historische Erinnerung in der Karolingerzeit (Darmstadt, 2002), 42.

24

N I T H A R D A S A M I L I TA RY H I S T O R I A N

Seville in his Etymologies that men who wished to be effective rulers should read history because these texts allowed them to benefit from the experiences of the past.25 Charlemagne’s interest in historiae and res gestae antiquorum as sources of information about the past was well documented in Einhard’s popular Vita Karoli.26 Indeed, when writing in 798 to Charlemagne, who was campaigning in Saxony, Alcuin addressed the king in the persona of Horace (Flaccus), a veteran soldier (veteranus miles), and reminded him of the importance of historical works when fighting wars.27 Alcuin observed “that it very important for us to read in ancient books of history about the kind of strength that fighting men had so that the kind of wise temperament, which ought to be acted upon, shall guide and rule us in all things.”28 Similarly, Lupus of Ferrières presented to Charles the Bald a “brief summary of the deeds of the emperors,” emphasizing that reading histories was important for conducting public affairs.29 Consequently, Nithard’s text fits within a lengthy tradition of historiographical works that were intended for a didactic purpose and offered lessons regarding the effective conduct of military affairs.

Nithard on the conduct of war Despite the considerable scholarly attention to Nithard’s causae scribendi and his audience, there are no studies that treat in a systematic manner the overwhelmingly military content of his text, although several scholars have drawn on the Histories in studies that deal with specifically social aspects of Carolingian warfare. For example, Janet Nelson briefly addressed some military aspects of Nithard’s work in a study focused on the issue of “knighthood” in the Carolingian period, and John Gillingham utilized Nithard’s Histories as a witness to changing Frankish attitudes toward killing other high-ranking Franks in battle.30 However, no studies of Nithard consider his discussion of topics that now are central to the investigation of military history, such as military organization, logistics, tactics, or strategy. 25 Isidore of Seville, Etymologiarum sive originum libri xx, ed. W. M. Lindsay (Oxford, 1911, repr. 1957), here bk. I, ch. xliii. See the discussion by Jacques Fontaine, Isidore de Séville et la culture classique dans l’Espagne wisigothique (Paris, 1959), 180–5, regarding Isidore’s view of history. 26 Einhard, Vita Karoli, ch. 24. 27 Alcuin, Epistolae, 149, ed. E. Dümmler, MGH Epist. 4 (Berlin, 1895). 28 Ibid., “quod militantibus virtutis genus maxime necessarium esse in antiquis historiarum libris legimus, ut cuncta sapiens temperantia, quae agenda sint, regat atque gubernet.” 29 See the discussion by Rosamond McKitterick, “Charles the Bald (823–877) and His Library: The Patronage of Learning,” The English Historical Review 95 (1980), 28–47, and republished with the same pagination in Rosamond McKitterick, The Frankish Kings and Culture in the Early Middle Ages (Aldershot, 1995), here 35. 30 Janet Nelson, “Ninth-Century Knighthood: The Evidence of Nithard,” in Studies in Medieval History Presented to R. Allen Brown, ed. Christopher Harper-Bill, Christopher J. Holdsworth, and Janet L. Nelson (Woodbridge, 1989), 255–66; and John Gillingham, “Fontenoy and After: Pursuing Enemies to Death in France Between the Ninth and the Eleventh Centuries,” in Frankland: The Franks and the World of the Early Middle Ages, Essays in Honour of Dame Jinty Nelson, ed. Paul Fouracre and David Ganz (Manchester, 2008), 242–65.

25

N A R R AT I V E W O R K S

The following sections therefore will set out in detail Nithard’s observations about the conduct of war during 830s and 840s. We then will conclude with a discussion of the implications of Nithard’s depiction of this period for our understanding of Carolingian warfare in the decades following the death of Charlemagne in 814.

Military organization As discussed earlier, with a few notable exceptions, current scholarly understanding of the organization of Carolingian military forces in the period after Charlemagne is that levies of small property owners for campaign duty came to an end around the turn of the ninth century.31 By contrast, throughout his Histories, Nithard makes clear that military forces commanded by Louis the Pious and all three of his sons, who contended for rule of the regnum Francorum after 840, were drawn from both the military households of magnates and from the expeditionary levies. This included the use of these levies for offensive as well as defensive operations. For example, in his discussion of the campaign that Louis the Pious ordered against Lothair I in the spring of 834, Nithard observes that Count Wido and “all of the men between the Seine and Loire” were dispatched to the Breton frontier.32 Nithard does not identify any particular magnates when discussing this mobilization, but does mention that several of the comites in Wido’s army, who were responsible for mobilizing the expeditionary levies of their pagi, were killed in the battle. Nithard also observed that the losses suffered by Louis the Pious’s army included not only magnates but also plebis innumera multitudo, which is a clear reference to men of lower social and economic standing.33 The naming of counts and also the plebs who died in battle makes clear that levies participated in the campaign. Nithard again points to the mobilization of a very large force, which included both magnate-led elements and levies commanded by counts in his description of Lothair’s invasion of Charles’s territories in October 840.34 In his initial discussion of Lothair’s advance, Nithard observes that the senior Carolingian king mobilized all of the men, i.e. omnes, living east of the Charbonnière forest.35 Nithard returned to his commentary on these forces discussing Charles’s own advance toward the Seine in March 841. According to Nithard, once Charles arrived at the river, he learned that a number of magnates, as well as all of the counts (comites), abbots, and bishops from east of the Charbonnière, had been deployed along the Seine. Nithard observes that Lothair had positioned a rear guard there in order to prevent Charles from crossing the river.36 The presence of the counts in this

31 32 33 34 35 36

For a contrary view, see Goldberg, Struggle for Empire, 119–46. Nithard, Historiarum, 1.5. Ibid. Nithard, Historiarum, 2.3 and 2.6. Ibid., 2.3. Ibid., 2.6.

26

N I T H A R D A S A M I L I TA RY H I S T O R I A N

army, as in 834, indicates the presence of levies of militia troops, as well as the household troops of the various magnates who were serving in Lothair’s army at this time.37 In addition to observing the role played by forces led by counts in expeditionary campaigns, Nithard signals the deployment of small-scale landowners for service in offensive campaigns when he discusses the mobilization of regional levies by the Carolingian kings. The use of regional designations, e.g. Burgundians or Tournois, to denote offensive military forces that included militia levies, as well as the military households of various magnates, dates to the Merovingian period and was carried through into the eighth and ninth century.38 Gregory of Tours, for example, whose history was widely read by men of Nithard’s status and education, observed that the “men of the Touraine” served in Poitou, that is outside of their home region, as part of King Guntram’s army.39 In this case, Gregory made clear that many of the Tournois who went on campaign had not been mobilized by the count of the city, but rather were drawn by the hope of plunder.40 Other much more contemporary writers of history, e.g. from the late eighth to the mid-ninth century, also identified levies on the basis of the region from which they were mobilized. In his discussion of Charlemagne’s invasion of Spain in 778, for example, the author of the Annales regni Francorum observed that Charlemagne’s two armies were drawn de partibus Burgundaiae et Austriae vel Baioariae seu Provinciae et Septimaniae et pars Langobardorum.41 The author similarly observed with regard to Charlemagne’s campaign in 806 into Bohemia that the army was mustered de Baioaria et Alamannia atque Burgundia.42 37 Fleckenstein, “Adel und Kriegertum,”89–92, argued that it was the absence of counts in the mobilization from 847 that indicated only the dependents of magnates had been summoned for military service, in contrast to 808 when the summons to counts meant that levies were involved. 38 Indeed, historical works written in the tenth and eleventh centuries continue to use regional designations to distinguish between men of the levies, on the one hand, and members of magnate military households, on the other. The latter were denoted for the most part as milites. In this regard, see the discussion by Bernard S. Bachrach, “Dudo of Saint Quentin as a Military Historian,” in The Haksins Society Journal: Studies in Medieval History 12 (2002 appeared in 2003), 155–85; and D. Bachrach, “Memory, Epistemology,” 63–90. 39 Gregory of Tours, Libri Historiarum X, bk 7 ch. 28, “Secutique sunt eum de Toronicis multi lucre causa.” The basic study remains Bernard S. Bachrach, Merovingian Military Organization, 481–751 (Minneapolis, 1972). However, for various additions and modifications see Idem, “Quelques observations sur la composition et les caractéristiques des armées de Clovis,” in Clovis: Histoire et Mémoire, 2 vols., ed. Michel Rouche (Paris, 1997), I, 689–703; Idem., “The Imperial Roots of Merovingian Military Organization,” in Military Aspects of Scandinavian Society in a European Perspective, AD. 1–1300, ed. Anne Norgard Jorgensen and Birthe L. Clausen (Copenhagen, 1997), 25–31. 40 Walter Goffart, “Frankish Military Duty and the Fate of Roman Taxation,” Early Medieval Europe 16 (2008), 166–90 treats Gregory in this regard. 41 Annales regni Francorum, ed. F. Kurze SRG in usum scholarum separatim editi 6 (Hanover, 1907), an. 778. 42 Ibid., an. 806.

27

N A R R AT I V E W O R K S

Ermoldus Nigellus, who had military experience under King Pippin I of Aquitaine (817–838), also used regional designations to denote levies.43 In his presentation of Louis the Pious’s Breton campaign in 818, for example, Ermoldus observes that Louis mobilized the Franks and their subject peoples to serve in his army. Ermoldus then discusses the thousands of Swabians who were “organized by their hundreds (centenae),” and the cohors of Saxons, and the force (manus) of Thuringians, and the very large number of young men from Burgundy.44 It is noteworthy in this context that the Swabian peasant levies (rustici) continued to be depicted in narrative sources as being organized according to the hundreds in which they lived into the late eleventh century.45 Nithard’s exact contemporary, the author of the Vita Hludowici, usually denoted as the Astronomer, describes the army sent by Louis the Pious in 815 to intervene in the Danish succession struggle as being composed of Abodrites and forces commanded by Saxon comites.46 The Astronomer repeats this usage in discussing Louis the Pious’s spring campaign in 816 against the Sorbs, another Slavic people living between the Saale and Elbe rivers. The author observes that Louis mobilized both the men who are called East Franks, i.e. orientales Franci, and the Saxon counts.47 Nithard routinely deploys regional designations to denote a broad-based mobilization of forces rather than a summons issued only to specific magnates and their military households. In discussing the period immediately preceding the death of Louis the Pious in June 840, for example, Nithard claims that Louis the German invaded Swabia, having levied forces from among the Thuringians and Saxons.48 In a similar manner, when discussing Lothair I’s offensive against Charles in August and September 841, Nithard observes that the former had very substantial forces of Saxons, Austrasians, and Swabians.49 In this case, Nithard’s emphasis on the large number of men from these regions makes clear that he was not referring merely to small contingents of mounted troops in the military households of magnates. We see this same dynamic at play in Nithard’s discussion of events in the winter of 842, when Charles and Louis combined their armies at Mainz to face Lothair yet again. While camped in and around this fortress city, the two royal brothers were joined by reinforcements led by Louis’s son Carloman, whom

43 Ermoldus Nigellus, In honorem Hludowici, ed. Ernst Dümmler in MGH Poetae Latini aevi Carolini, vol. 2 (Berlin, 1884), bk 4 line 137–138. 44 Ermoldus Nigellus, In honorem Hludowici, bk. 3 lines 258–267. 45 Berthold of Reichenau made this point in the context of the German civil wars of the 1080s. See Die Chroniken Bertholds von Reichenau und Bernolds von Konstanz, ed. I. S. Robinson, MGH SRG nova series 14 (Hanover, 2003), 322–3. 46 Astronomer, Vita Hludowici imperatoris, ed. Ernst Tremp, MGH SRG in usum scholarum separatim editi 64 (Hanover, 1995), ch. 25. 47 Ibid., ch. 26. 48 Nithard, Historiarum, 1.8. 49 Ibid., 3.3.

28

N I T H A R D A S A M I L I TA RY H I S T O R I A N

Nithard describes leading “an immense army of Bavarians and Swabians.”50 Once again, the very large number of men who were mobilized indicates that this force included militia levies as well as the military households of magnates.

Sizes of armies and overwhelming force The continuing need for expeditionary levies to augment the military forces provided by magnates becomes clear in Nithard’s discussion of the political considerations involved in the mobilization of armies. In particular, Nithard emphasizes the positive political effect of employing what modern military specialists characterize as the doctrine of overwhelming force. This is a different issue from the widespread topos found in medieval narrative sources, where authors claimed that the opponents possessed a vast numerical superiority in order to explain away defeats by the “home side” and to enhance their victories against putatively overwhelming odds.51 Rather, Nithard makes clear that the Carolingian kings of his day, much like their grandfather Charlemagne, frequently chose a policy of mobilizing the largest possible army for operations in the field and that failing to do so could have significant negative political as well as military consequences.52 Nithard first illuminates the important political implications of having a large or an insufficiently large army in his discussion of the events at the Field of Lies in late June 833. This confrontation marked the culmination of the revolt by Lothair, Louis the German, and their brother Pippin I of Aquitaine (died 838) against their father, Emperor Louis the Pious. Nithard emphasizes that the three brothers confronted their father with a massive army (ingens exercitus).53 It was in this context that the magnates serving as subordinate military commanders on Louis the Pious’s side made the decision to abandon their ruler rather than face this overwhelmingly large host. They were willing, according to Nithard, to accept promises of future reward from the emperor’s sons in return for their betrayal of their ruler.54 Nithard did not need to add for his audience that Louis the Pious’s men were not willing to die on the battlefield in a hopeless cause.

50 Ibid., 3.7, “cum ingenti exercitu Baioariorum et Alamannorum.” 51 With regard to this practice among medieval authors, see the observations by Bernard S. Bachrach, “Early Medieval Military Demography: Some Observations on the Methods of Hans Delbrück,” The Circle of War, ed. Donald Kagay and L. J. Andrew Villalon (Woodbridge, UK, 1999), 3–20. 52 The very large size of the armies mobilized by Charlemagne has been emphasized by a number of scholars. See, for example, Karl Ferdinand Werner, “Heeresorganization und Kriegsführung im deutschen Königreich des 10. und 11. Jahrhunderts,” in Settimane di Studio de Centro Italiano sull’alto Medioevo, vol. 15 (Spoleto, 1968), 813–32, here 820–21; J.-E. Verbruggen, The Art of War in the Middle Ages from the Eighth Century to 1340, trans. Sumner Willard and S. C. M. Southern, 2nd edition (Woodbridge, 1997), 283–304; and Alessandro Barbaro, Charlemagne, The Father of a Continent, trans. Allan Cameron (Berkeley-Los Angeles, 2004), 169. 53 Nithard, Historiarum, 1.4. 54 Ibid., “ac variis affectionibus populum, ut a patre deficeret, filii compellunt.”

29

N A R R AT I V E W O R K S

The situation was reversed, however, one year later when Lothair faced Louis the Pious at Chouzy in August 834. Nithard makes clear that it was now Louis the Pious who possessed the larger army, having mobilized a powerful force (manus valida) in the Frankish heartlands. Louis the Pious’s military position was further enhanced by the decision of Louis the German to switch sides in the conflict and join him with “all of the forces from the further side of the Rhine.”55 As a consequence of the changed status of forces, it was now Lothair who was forced to concede defeat and accept humiliating conditions, including being confined to Italy, from where he was not permitted to leave without his father’s permission.56 In both of these military confrontations, it is noteworthy that no battle was fought, precisely because it was clear that one side possessed overwhelming force. As Nithard makes clear, following the death of Louis the Pious in June 840, Lothair applied the lessons of 833–834 once he made the decision to attempt to acquire rule of the entire regnum Francorum and deprive his brothers Louis and Charles of the territories that had been assigned to them by their father in 836.57 According to Nithard, immediately after Lothair received word of his father’s death, he began advancing north through the Alps at a very deliberate pace, sending out emissaries to make promises or offer threats to all of the magnates along his line of march.58 Nithard emphasizes that Lothair’s plan of building up his forces in this manner was quite successful, drawing supporters either through their greed or their fear.59 Lothair’s ability to mobilize a large army encouraged him to pursue an aggressive strategy, according to Nithard, and the emperor thus settled upon a military solution to the succession struggle against his brothers. Lothair decided first to seek out and defeat Louis the German, and he achieved an initial victory over a comparatively small element of his brother’s forces in the environs of the fortress city of Worms.60 Lothair then marched to the important royal fiscal center and palace complex at Frankfurt, where he was confronted by Louis’s main army. As Nithard makes clear, Lothair had hoped to overawe his brother’s supporters with his own exceptionally large army and win a bloodless victory of the type achieved at the Field of Lies. However, when the two armies faced each other, Louis’s partisans gave no sign of giving up, and Lothair realized that he would not be able to achieve victory without a bloody battle because he would have been forced to attack Louis’s troops, who were deployed in prepared positions.61 Consequently, Lothair withdrew in the hope that he would have an easier time of overcoming Charles’s armies.62 55 Ibid., 1.5, “insuper Lodhuwico filio suo cum universis, qui trans Renum morabantur, in auxilium sibi assumpto.” 56 Ibid. 57 See Annales Bertiniani, ed. Georg Waitz, MGH SRG 5 (Hannover, 1883), anno 836. 58 Nithard, Historiarum, 2.1. 59 Ibid., “Ergo cupiditate terroreque illecti undique ad illum confluunt.” 60 Louis’s main army, as Nithard makes clear, was located in the environs of the royal palace and fiscal complex at Frankfurt. 61 Nithard, Historiarum, 2.1, “Lodharius illum absque praelio sibi subigere diffideret.” 62 Ibid., “sperans Karolum facilius superari posse.”

30

N I T H A R D A S A M I L I TA RY H I S T O R I A N

Turning northwest, Lothair advanced during the autumn of 840 into the territories that had been assigned to Charles by Louis the Pious as part of the division of the regnum Francorum, As seen earlier, Lothair received additional reinforcements, whom Nithard identifies as all of the men (omnes) east of the Charbonnière forest, as well as the support of the powerful magnates, Abbot Hilduin of St. Denis and Count Gerard of Paris, whom Nithard’s audience would have understood commanded the substantial levies of militia troops from this densely populated region.63 According to Nithard, the decision by these magnates to abandon Charles had a ripple effect, leading even more magnates to switch sides, including Pippin, a great-grandson of Charlemagne and grandson of King Pippin of Italy (781–810), as well as other local nobles, whom Nithard again describes as succumbing either to Lothair’s promises or to his threats.64 It is in this context that Nithard observes that Lothair decided to advance as far south as the Loire, “placing his trust in the very large size of his army.”65 After gathering his forces, Lothair advanced toward Charles, who was encamped near Orleans.66 Lothair’s plan, as had been the case earlier with regard to Louis at Worms, was to force a confrontation with Charles, where it would be obvious to the latter’s supporters that they simply could not win against such overwhelming odds. The result, as Lothair hoped, would be a victory without a battle. This plan failed, according to Nithard, because Charles’s men, like those of Louis earlier, had sworn an oath to fight to the death rather than betray their king.67 This is in marked contrast to the magnates who had marched with Louis the Pious to the Field of Lies in 833. In fact, such oaths may be understood as a response to the failure of so many Carolingian magnates to keep faith with Louis the Pious in his hour of dire need. What becomes clear from Nithard’s text is that Lothair was unsuccessful in his effort to sow seeds of disunity among Charles’s forces and that the unified front presented by Charles’s supporters prevented Lothair from gaining additional strength as his brother’s army was weakened by defections. In the end, Lothair’s plan to achieve a political victory on the battlefield failed, and he did not wish to risk a military confrontation. As Nithard makes clear, Lothair again sought to utilize the strategy of bringing to bear overwhelming force in the spring of 841 by directing his attention once more against Louis. Lothair left a substantial blocking force along the Seine to protect his rear and lines of communications and supply and then mobilized an army, characterized by Nithard as “an infinite multitude” (infinitas multitudo), which he led across the Rhine to invade Louis the German’s lands.68 Lothair, according to Nithard, employed his usual tactic of sending ahead emissaries who threatened and cajoled the men of the lower orders (plebs) who lived in the regions 63 64 65 66 67 68

Nithard, Historiarum, 2.3. Ibid. This Pippin should not be confused with Pippin II of Aquitaine. Ibid., “spe multitudinis suae fretus, Ligerem usque ut procederet, deliberavit.” Ibid., 2.4. Nithard, Historiarum, 2.4, “elegerunt potius nobiliter mori quam regem proditum derelinquere.” Nithard, Historiarum, 2.6 and 2.7.

31

N A R R AT I V E W O R K S

along his line of march. It was these plebs, as members of the local militia, who were responsible for the defense of the regions in which they lived.69 Nithard makes clear that Lothair’s strategy was successful on this occasion because he convinced a number of Louis’s supporters, denoted as the populus, that they could not withstand such an enormous army (tantus exercitus). Some of Louis’s supporters switched sides. Others simply fled. As a consequence, Louis, who was now bereft of much of his military support, had no option other than to withdraw before Lothair’s advance and retreat to his base in Bavaria.70 In emphasizing the size of the armies employed by Carolingian rulers and the concomitant decisions by magnates, and perhaps the broader population as well, to reassess their original loyalties, Nithard was making clear to his audience the intertwined nature of politics and war. To be seen to be winning was, in some circumstances, tantamount to winning, and having a large army, in general, was a decisive element in being seen as a winner. As a corollary, being seen as losing, which as a practical matter meant having a small army, could have equally negative consequences in the political and, consequently, the military sphere.71 Nithard specifically addresses the problems caused when kings led small armies in the context of discussing the political and military maneuvers of the royal brothers in the months after the battle of Fontenoy (25 June 841). Following their victory over Lothair, Charles and Louis each led his army in different directions and arranged to meet again in two months on 1 September at the city of Langres, located approximately 280 kilometers southeast of Fontenoy.72 Louis marched east toward the Rhineland and Charles marched southwest toward Aquitaine, each to deal with problems in his own lands, concerning which they had received information prior to the battle at Fontenoy.73 Nithard criticized their decision, asserting that “because all opposition appeared to have been overcome” Louis and Charles went their separate ways. Nithard added that “the needs of the state were neglected through their lack of forethought.”74 Nithard then specified that each of the men

69 The plebs in this context are to be understood as local levies, i.e. those men who were too poor to serve as expeditionary levies alongside the royal and magnate military households. For a summary of Carolingian tri-partite military organization see Bernard S. Bachrach and Charles R. Bowlus, “Heerwesen,” in Reallexikon der germanischen Altertumskunde, ed. Heinrich Beck et al. (Berlin-New York, 2000), 14 cols. 122–136. 70 Ibid. 71 The political implications of having a large army were not, in Nithard’s view, the equivalent of commenting on the fighting ability of the army. Indeed, Nithard observes that a smaller but better-trained and -led force could defeat a larger but poorly led force. See Nithard, Historiarum, 1.5. This would appear to be a reference to the insights provided by Vegetius. 72 Nithard, Historiarum, 3.2. 73 Ibid. 74 Ibid., “Cumque adversa undique propulsa viderentur . . . Res autem publica inconsultius, quam oporteret, omissa.” This passage might be understood as a general condemnation of the Carolingian nobility and their failure to seek the common good rather than their personal interests. However, the context here indicates that Nithard is commenting specifically on the decision of the

32

N I T H A R D A S A M I L I TA RY H I S T O R I A N

who had departed was guilty of failing to do his duty.75 This latter comment would seem to have been directed not only at the kings themselves but also their magnates. In sum, Nithard was observing the failure of Charles and Louis to maintain the overwhelming force that had allowed them to achieve victory at Fontenoy. The cost of this error in strategic judgment became clear, as Charles sought to use the victory at Fontenoy to strengthen his support in the heartlands of West Francia. According to Nithard, Charles dispatched emissaries to meet with the “Franks” at Quierzy. However, the magnates refused to commit to Charles, according to Nithard, because they were not sure about what actually had happened at Fontenoy. Lothair’s supporters had circulated the disinformation that Charles had been killed and that Louis was wounded in the battle.76 In this context, Charles’s emissaries and his potential supporters requested that he advance into the region in order to demonstrate both his health and his military strength. However, as Charles marched through the Carolingian heartland, including the districts of Beauvais, Compiègne, Soissons, Rheims, and Châlons, the local magnates, according to Nithard, refused to join him. The overriding reason, as Nithard explains, was their contempt for the small size of Charles’s army.77 Nithard adds that this was also the attitude of the Aquitanians.78 Nithard’s explicit criticism of Charles and Louis, as well as their magnates, for failing to maintain a large, unified army after Fontenoy clearly was intended to teach his audience about the serious cost of pursuing the wrong military strategy. In this case, dividing their forces to deal with peripheral problems rather than focusing on the main threat that they faced, i.e. Lothair, gave the latter an opportunity to regroup and extend the war.

Battle-seeking and battle-avoiding campaign strategies In his important article on the science of war in the Middle Ages, John Gillingham observed the central role played by Vegetius’s Epitoma rei militaris in informing the thought of medieval military commanders about the risks inherent in battle.79 Vegetius had argued that combat was to be avoided unless the commander seeking

75 76 77 78 79

two kings to separate their forces while Lothair was still in the field, and also the likely demobilization of at least part of their forces, i.e. the idea that each man (quis) did as he desired. Nithard, Historiarum, 3.2, “quo quemque voluntas rapuit, perfacile omissus abscessit.” Nithard, Historiarum, 3.2. Ibid., “Franci vero eandem paucitatem . . . spernentes.” Ibid. The importance of Vegetius’s observations to the thinking of medieval military commanders has been the subject of considerable debate. For a useful introduction to this debate, see the discussion by John B. Gillingham, “Richard I and the Science of War in the Middle Ages,” in War and Government in the Middle Ages: Essays in Honour of J. O. Prestwich, ed. John B. Gillingham and J. C. Holt (Woodbridge, 1984), 78–91; Clifford J. Rogers, “The Vegetian ‘Science of War’ in the Middle Ages,” Journal of Medieval Military History 1 (2002), 1–19; and John B. Gillingham, “Up with Orthodoxy: In Defense of Vegetian Warfare,” Journal of Medieval Military History 2 (2004), 149–58.

33

N A R R AT I V E W O R K S

battle enjoyed an overwhelming superiority in the numbers of troops, as well as other advantages, including good logistic support, the choice of an advantageous field for battle, and having a secure line of communications.80 In the present context, it is important to understand that the epitoma was very well known to the members of the Carolingian ruling family and to their advisors. Charlemagne and Alcuin communicated in letters about Vegetius’s text.81 Louis the German and Charles the Bald both possessed copies of the text.82 Rabanus Maurus, one of Lothair’s leading supporters as the abbot of Fulda, wrote a revised version of the Epitoma when he served as archbishop of Mainz, which he sent to Lothair II (855–869), the son of Lothair I.83 Consequently, there is no doubt that Vegetius’s warnings about the dangers of seeking battle without overwhelming force were well known in the royal courts of the Carolingian kings during the tumultuous period between 833 and 843 and undoubtedly influenced Nithard as he composed his Histories. It seems likely, therefore, that the ideas available in Vegetius’s text also informed the thinking of Carolingian commanders and lay magnates in general when they made decisions about committing to a battle. Indeed, the influence of Vegetian thought is further suggested by the fact that in the course of numerous confrontations between Carolingian armies directly commanded by kings in the period 833–843, there were, in fact, only two battles, namely those at Worms and Fontenoy, mentioned earlier. However, it is also true, as Nithard makes clear, that Carolingian kings did not always avoid battle in the field and sometimes even sought battle when conditions were regarded as propitious. This was the case not only at Worms and Fontenoy but in several other cases as well, where one commander sought battle but was frustrated in executing his plan. This happened to Lothair when facing Louis the German at Frankfurt and again when facing Charles the Bald near Orleans, as discussed earlier. In the context of battle-seeking strategies, it is frequently observed in the scholarly treatments of the Histories that Nithard devotes much more attention to the preliminaries at Fontenoy than he does to the battle itself.84 This usually has been interpreted as evidence of Nithard’s efforts to establish the bona fides of Charles

80 Vegetius, Epitoma rei militaris, ed. Alf Önnerfors (Stuttgart, 1995), bk 3, ch. 20–21. 81 See Alcuin, Epistolae, ed. Ernst Dümmler, MGH, Karolini aevi, 2 (Hannover, 1895). no. 257; and the discussion by L. Wallach, Alcuin and Charlemagne: Studies in Carolingian History and Literature (Ithaca, NY, 1959), 50–1. 82 For Louis the German’s possession of a copy of Vegetius’s text, see Goldberg, Struggle for Empire, 40–2. Frechulf of Lisieux, royal chancellor of the West Frankish kingdom, provided King Charles the Bald with a specially revised edition of Epitoma rei militaris. See McKitterick, “Charles the Bald (823–877) and His Library,” 28–47, here 31. 83 Rabanus Maurus, who served both as abbot of Fulda and archbishop of Mainz, in both of which capacities he had extensive military responsibilities, oversaw the composition of a revised version of Vegetius’s handbook, which would only deal with those matters that were of value “tempore moderno.” Rabanus Maurus, De procinctu Romanae militiae, ed. Ernst Dümmler in Zeitschrift für deutsches Alterthum, 15 (1872), 443–51, here 450. 84 Gillingham, “Fontenoy and After,” 242–65; and Nelson, “Public Histories,” 255 and 262.

34

N I T H A R D A S A M I L I TA RY H I S T O R I A N

and Louis as doing everything in their power to avoid battle and the subsequent shedding of great quantities of Frankish blood.85 Undoubtedly, this is true. The propagandistic purpose of Nithard’s text required showing Charles as extremely reluctant to shed Frankish blood. Indeed, according to Nithard, Lothair’s younger brothers even agreed to a truce with the emperor in order to allow time to negotiate a lasting peace agreement.86 From the perspective of understanding the decision by Charles and Louis to cast the dice and order an attack on Lother’s forces, however, there is another factor that is of central importance. Nithard emphasizes that on the very day that Louis and Charles agreed to a truce with Lothair and returned to their camp to celebrate Mass, they received intelligence that Lothair had obtained a promise of support from his nephew, Pippin II of Aquitaine. Moreover, Pippin was marching with his army to join his eldest uncle at Fontenoy.87 It became clear to the two brothers that Lothair was using the truce simply to gain time while waiting for Pippin to arrive. It was only at this point, when the numerical superiority that they had achieved at this place and time was threatened, that Charles and Louis decided to commit their armies to battle in order to press home their numerical advantage while they still had it. As is well known, after their victory, Louis and Charles decided not to undertake a full pursuit of Lothair’s forces, and thus prevented the massive casualties that usually occur in the course of such a pursuit.88 As a consequence, as discussed earlier, their victory at Fontenoy did not alter in a fundamental manner the military and political situation in the regnum. Indeed, as Nithard makes clear, soon after Fontenoy Lothair again went on the offensive and pursued a battle-seeking strategy. Lothair’s post-Fontenoy campaign strategy was based, in large part, on the miscalculation made by Louis and Charles that their elder brother had suffered so serious a defeat that they could safely divide their armies and go their separate ways rather than pursue and administer the coup de grâce. This is the decision that Nithard thoroughly criticized.89 The result, according to Nithard, was that Lothair once again pursued a battleseeking strategy, similar to the one he employed in the summer and autumn of 840, by trying to defeat his brothers in detail while their forces were divided. Initially,

85 Frank Pietzcker, “Die Schlacht bei Fontenoy,” Zeitschrift der Savigny-Stiftung für Rechtsgeschichte, Germanistische Abteilung 81 (1964), 318–42, argued that Charles and Louis acted immorally by attacking Lothair’s forces before the end of the truce, and thus Nithard’s emphasis on timing of the battle, after the truce had expired, was intended to obscure the crime committed by the two victorious kings. Adelheid Krah, Die Entstehung der ‘potestas regia’ im Westfrankenreich während der sertsen Regierungsjahre Kaiser Karls II. (840–877) (Berlin, 2000), 76–86, attempts to reconstruct the battle at Fontenoy; however, her efforts are undermined by a thorough misunderstanding of the nature of warfare in the early medieval period. 86 Nithard, Historiarum, 2.10. 87 Ibid. 88 Gillingham, “Fontenoy and After,” 252–3 provides a valuable discussion of the losses that usually were suffered by the defeated army in the immediate aftermath of the battle. 89 Nithard, Historiarum, 3.2.

35

N A R R AT I V E W O R K S

Lothair decided that he would march against Louis and pursued his army into the middle Rhine region dominated by the fortress city of Worms with the aim of engaging him in battle.90 In the meantime, according to Nithard, Charles received intelligence regarding Lothair’s movements and learned from Louis that because of Lothair’s advance, Louis would not be able to attend the assembly that they had planned for 1 September at Langres.91 Taking these facts into consideration, Charles decided to lead his army, which Nithard depicts as small, toward St. Quentin and then Maastricht to provide aid to Louis and perhaps catch Lothair’s forces in a pincer movement.92 Nithard gives the impression that when Lothair received a report that Charles had moved his forces eastwards, the senior Carolingian ruler understood the threat posed by his brother’s maneuver to his rear. As a consequence, he broke off his advance against Louis and led his entire army northwest in order to engage in battle with his youngest brother, whose army was substantially smaller than the forces commanded by Lothair.93 Charles, according to Nithard, found himself in a very dangerous position as Lothair led his entire army against him. In light of Lothair’s volte face, Charles sent a magnate named Rabano to Louis to beg him to come to his aid as quickly as possible. Louis’s advance had the obvious tactical significance of lessening Lothair’s numerical advantage and again offered the potential option of catching Lothair in a pincer by attacking his rear. However, Charles’s appeal to Louis was in vain because Lothair had positioned his army to prevent his brothers from reuniting their forces.94 Consequently, facing Lothair alone, Charles had no choice other than to retreat to the fortress city of Paris and wait for aid to arrive from his brother and from his fideles, whom he had summoned from “all over.”95 It seems likely that these were the same men whom Nithard earlier accused of following their own interests rather than the common good after the battle of Fontenoy by returning home too early. Lothair pursued Charles, as Nithard emphasizes, very confident in the large number of east Frankish, Swabian, and Saxon troops whom he commanded.96 It was only after Charles was able to block Lothair’s crossing of the Seine by positioning forces at all of the bridges and fords that the latter decided to break off his pursuit and try again to create a rift between Charles and Louis by offering a diplomatic solution to the conflict.97 90 91 92 93 94

Ibid., 3.3 Ibid., 3.2, “venire non posset, eo quod Lodharius in regnum illius hostile manu irruere vellet.” Ibid., 3.3. Ibid., “iter arripuit et, qualiter super Karolum irrueret.” Ibid., “Rabanonem etiam ad Lodhuvicum dirigens mandat, qualiter pro suo adiutorio illis in partibus isset, quod Lodharius audiens, illo omisso, supra se cum omnibus copiis ire pararet . . . .” 95 Ibid., “tam fratris sui Lodhuvici adventum quam et ceteros fideles suos, quos undique convocaverat, praestolaturus.” 96 Ibid., “habebat enim tam Saxonum quam et Austrasiorum nec non et de Alamannis partem haud modicum secum, horumque auxilio praemaxime confisus . . . .” 97 Ibid.

36

N I T H A R D A S A M I L I TA RY H I S T O R I A N

Nithard presents Lothair seeking battle again in the late autumn of 841. Charles, according to Nithard, had dispatched part of his army to cross the Seine and march west toward a wooded region in what is today the French department of Orne, near the Breton frontier.98 Lothair thought that these troops would be defeated quite easily and believed that this victory would bring additional dividends. Nithard comments specifically on Lothair’s goal of bringing the Breton leader Nomenoi (831–851) over to his side and of causing the rest of Charles’s army to collapse in panic.99 Nithard appears to have been heartened to be able to report that Lothair was unsuccessful because Charles’s entire army was able to escape.100 However, Nithard’s observations again point out the danger of dividing one’s forces in the face of the enemy and the likelihood that such a decision, for all intents and purposes, had the negative consequence of inviting a battle on unequal terms.

Seeking battle as a moral problem Nithard’s repeated claims that Lothair sought to force his brothers to fight a battle might be understood as an element of his parti pris in blackening the image of the senior Carolingian ruler. Indeed, the repeated emphasis by Nithard on Lothair’s aggressiveness in seeking battle seems to have had this very purpose. However, Nithard also depicts Charles as actively seeking battle when the conditions were right in the Vegetian sense, meaning that he had achieved a substantial numerical superiority in a particular place and time. Consequently, the proper conclusion to draw here is that battle-seeking in Nithard’s view was not ipso facto morally wrong or undesirable either in a tactical or strategic sense, but rather was illegitimate if the military commander were seeking an evil end. In this context, Nithard discusses Charles’s effort to seek a battle in the spring of 841. In March of that year, Charles began mobilizing a large army with which he intended to force a crossing of the Seine and advance into lands currently held by Lothair. The west Frankish army successfully turned the flank of the forces that Lothair had left to guard the Seine by commandeering merchant ships downstream at Rouen.101 Marching south, Charles arrived at St. Denis, where he received intelligence both about the location of Lothair’s troops, who had withdrawn from the Seine, and also about the imminent arrival of his own men, who were hurrying to join him. In particular, Nithard records that Lothair’s men had planned an ambush of Charles’s reinforcements, but when faced with the overwhelming forces that had been brought to bear through river transports, Lothair’s troops retreated instead.102

98 Nithard, Historiarum, 3.4. 99 Ibid., “Qua quidem re sperabat se et hos facile delere et hoc terrore sibi residuos subiugare maximeque Nominoium Brittannorum ducem suo subdere dominatui posse.” 100 Ibid., “Nam exercitus Karoli omnis ab eo salvus evasit.” 101 Nithard, Historiarum, 2.6. 102 Ibid.

37

N A R R AT I V E W O R K S

Nithard explains that after receiving this information, Charles advanced from St. Denis 15 kilometers southwest to St. Germain-des-Prés in Paris, where he ostentatiously prayed, and then undertook a forced march of approximately 100 kilometers east to the confluence of the Seine and Loing at Saint-Mammès. This enabled Charles to join his reinforcements and prevent their ambush by Lothair’s now numerically inferior force.103 After marching an additional 40 kilometers southeast of Saint Mammès to Sens, Charles led his now substantially enlarged army in another forced night march toward the forest of the Othe, located 30 kilometers to the east of Sens. Nithard emphasizes that the purpose of Charles’s maneuver was to undertake an attack against Lothair’s forces, which he had learned were located in this forested area. In fact, Nithard claims that Charles “intended to attack them wherever and in whatever manner that he could.”104 However, Lothair’s troops, who obviously were greatly outnumbered, had learned of Charles’s advance and fled rather than face him in battle. In considering the implications of these events, it is clear that Nithard has once again shown the conditions under which a battle-seeking strategy was an acceptable risk, namely when a commander enjoyed a substantial numerical superiority over his opponent. Charles was in direct command of his entire army after linking up with reinforcements and was facing only a part of Lothair’s forces. Consequently, Charles can be understood to have enjoyed a significant numerical superiority in this place and time. In addition, Nithard discusses the effort by Charles to gain the element of surprise by undertaking forced marches at night. Unfortunately for Charles, Lothair’s men were not caught unprepared, which would seem to be a testament to the skill of their scouts, a topic to which we will return later. Charles, in effect, attempted to launch a surprise attack on men who had themselves prepared an ambush. For scholars enamored of the idea of chivalry, an assault of this type might seem to pass muster, because Lothair’s men already had violated a putative code of ethics by preparing to ambush Charles’s men.105 However, Nithard also depicts Charles himself seeking to ambush opposing forces even without the exceptional circumstances in play in the forest of Othe. Nithard observes, for example, that in January 841, Charles ordered Bernard of Septimania to meet with him at Bourges, where the latter indicated that he had failed to carry out his mission of negotiating a settlement with Pippin II of Aquitaine.106 Nithard explains that Charles was enraged by Bernard’s apparent lack of loyalty both toward Louis the Pious, earlier in his career, and also toward himself. Because of his deteriorating relationship with Bernard, Charles decided to take the drastic step of depriving Bernard of his offices and imprisoning him. However, Charles, according to Nithard, feared that carrying out this action would 103 Ibid. 104 Nithard, Historiarum, 2.6, “disposuerat enim, ubicumque et qualitercumque posset supra illos irruere.” 105 Nelson, “Knighthood,” 75–87 stresses the putatively “chivalric” aspects of Carolingian warfare. 106 Nithard, Historiarum, 2.5.

38

N I T H A R D A S A M I L I TA RY H I S T O R I A N

be very difficult and consequently decided to launch a surprise attack on Bernard because “he feared that he would not be able to capture him in any other way.”107 Consequently, Charles ordered this operation to take place. Bernard himself barely managed to escape from Charles’s soldiers, but a number of his men were killed and many others were captured along with Bernard’s baggage train.108 Nithard offers no explanation for Charles’s decision to launch this surprise attack other than utility and does not appear to question either the justness or morality of Charles’s decision. Rather, Nithard’s emphasis on Bernard’s history of betrayal suggests that he considered the ambush planned by Charles to have been justified. Nithard again comments positively on Charles’s decision to seek battle through an ambush in discussing events in September 842.109 After observing that Charles had driven Pippin II into Aquitaine and had left forces under the command of the dux Warin to maintain a watchful eye on the king’s nephew, Nithard turns his attention to actions of Charles’s fidelis, Count Egfrid of Toulouse. Nithard claims that Pippin II had dispatched forces to attempt to assassinate Egfrid. But the latter, apparently having learned of Pippin’s plan, set an ambush for these would-be assassins, killing a number of them and capturing others.110 Once again, Nithard gives no indication that Egfrid’s use of an insidia to get an advantage over his opponents was either unjust or immoral. Quite the opposite: because Pippin’s men were intent on murdering Egfrid, the lesson here is that they received their just deserts.

Military intelligence and communications Of primary concern for military commanders, whether actively seeking battle or not, was the acquisition of timely and accurate information about the location, disposition, and size of opposing forces. Some scholars have asserted that early medieval armies were notably inept with regard to obtaining military intelligence, and particularly in the use of scouts prior to military engagements.111 However, this view of early medieval warfare as primitive in conduct, conception, and organization cannot be sustained on even the most cursory reading of the Histories. In fact, Nithard makes clear throughout his work that all of the contenders for the rule

107 108 109 110

Ibid., “timens, ne aliter illum comprehendere posset, subito in illum irruere statuit.” Ibid. Nithard, Historiarum, 4.4. Ibid., “Insuper Egfridus comes Tolosae e Pepini sociis, qui ad se perdendum missi fuerant, quosdam in insidiis cepit, quosdam stravit.” 111 The best-known and most widely cited example in this context is Guy Halsall, Warfare and Society in the Barbarian West, 450–900 (London, 2003), 147–8, who argues that medieval military commanders routinely led their troops into ambushes and were ignorant of the basic necessities of military intelligence gathering. In light of the evidence that this broad-gauged assertion is simply untrue, Halsall adds that “nevertheless, it is unlikely that no attempt at all was ever made to scout ahead of the army. There are some references to scouting.” See the criticism of Halsall’s characterization of the use of scouting by Clifford Rogers, Soldiers’ Lives through History: The Middle Ages (Portsmouth, 2007), 106.

39

N A R R AT I V E W O R K S

of the regnum Francorum were well aware of the need to obtain detailed information about the military operations of their opponents and were highly successful in acquiring this intelligence. Nithard observes, for example, that Lothair intentionally advanced north from his Italian lands at a very deliberate pace, i.e. pedetemptim, in June 840 because “he wished to know the state of affairs before he crossed the Alps.”112 Nithard contrasts the slow movement of Lothair’s main army with the rapid departure of his nuncii, whom he dispatched throughout all of Francia in order to rally support, and to obtain information and determine which magnates would support him.113 Nithard again points to Lothair’s efforts to acquire military intelligence after he drove Louis and his army south from the Middle Rhineland into Bavaria in April 841. Following his victory over Louis at Worms, discussed earlier, Lothair marched to Aachen to celebrate Easter at Charlemagne’s imperial seat and thus gain an important propaganda advantage from this act. During the march, he began planning military operations against Charles. Significantly, Nithard observed that Lothair decided to alter his plans after learning that Charles had crossed the Seine and was headed east in order to provide aid to Louis.114 A central element in Lothair’s efforts to develop a new campaign strategy was the acquisition of additional information before he took further military action. Nithard observes that Lothair “quickly sent out men because he wished to know the state of affairs, that is where Charles was, and with whom.”115 Nithard provides a similar example of the impact of timely military intelligence in his discussion of Charles’s campaign to cross the Seine early in the spring of 841, i.e. in the same time frame that Lothair was preparing to celebrate Easter at Aachen. In March 841, following his successful effort to neutralize the threat posed by both Pippin II of Aquitaine and Bernard of Septimania, at least for the time being, Charles began recruiting a large army in order to take the initiative against Lothair in the territories between the Seine and the Rhine.116 After mobilizing forces from Burgundy, Aquitaine, and the regions between the Loire and the Seine, Charles marched toward the latter river. As Nithard makes clear, Charles found Lothair’s forces defending the right bank. Through scouting of the environs of the left bank, Charles learned that Lothair’s men had destroyed or sunk all of the local ships and wrecked the bridges to prevent him from crossing.117 However, Charles obtained information about a fleet of merchant ships that had been forced ashore near Rouen approximately 120 kilometers west-northwest of Paris. Taking advantage of this intelligence, Charles ordered the ships to be loaded with his

112 Nithard, Historiarum, 2.1, “ipse autem, pedetemptim, quo se res verteret, antequam Alpes excederet, scire volens.” 113 Ibid. 114 Nithard, Historiarum, 2.7. 115 Ibid., 2.7., “Velociter quidem praemittit rei veritatem, ubi et cum quibus esset, scire cupiens Aquis pascha celebraturus.” 116 Nithard, Historiarum, 2.6. 117 Ibid.

40

N I T H A R D A S A M I L I TA RY H I S T O R I A N

troops and sailed up the Seine, thus forcing Lothair’s troops to retreat from their positions along the banks of the river, as discussed earlier.118 Nithard’s depiction of Lothair and Charles using military intelligence to inform their campaign decisions is paradigmatic of his discussion of the acquisition of military intelligence during the entire period of civil wars. Nithard makes clear throughout his narrative that all of the Carolingian rulers regularly received detailed information about the actions of their adversaries. The only instance in which Nithard claims that there was confusion about the location of the armies under the direct command of Charles, Lothair, and Louis concerns the period immediately preceding the battle at Fontenoy, regarding which he states: “When the armies of the two sides unexpectedly caught sight of each other at Auxerre, Lothair immediately withdrew his troops a short distance from his encampment because he feared that his brothers might immediately launch an attack against him.”119 It seems likely, however, that the reader need not accept at face value Nithard’s assertion about the unexpected nature of the meeting between the armies at Auxerre. Rather, this claim fits all too well with Nithard’s overall effort to present Charles and Louis as doing everything in their power to avoid battle, rather than purposely seeking out and forcing a battle against Lothair. However, even in this case it is important not to discount the issue of the “fog of war” entirely. The striking consistency with which the Carolingian kings, in Nithard’s depiction of events, were able to gain information about both the location and plans of enemy forces, however, does require some further discussion. The fact that Carolingian commanders sought military intelligence does not explain why they apparently were so successful in obtaining it. The history of warfare is filled with intelligence failures. It seems likely that a major part of the explanation for the intelligence successes of the Carolingian rulers is that all of the armies were fighting more or less on their home ground. Consequently, each of the kings had men in his entourage who were intimately familiar with the geography and topography, including the military topography of roads, fortifications, and supply centers, in the lands through which the armies were marching. Large armies, of the type discussed by Nithard, required well-developed roads and other logistical infrastructure, such as the river ports. As a consequence, military commanders were familiar with the routes that were available to enemy forces, and confirmation of the specific line of march chosen by the enemy could be obtained relatively easily through the deployment of advanced scouts. An additional factor that likely enhanced the opportunities for gaining military intelligence was the large number of magnates and others who switched sides in the conflict, bringing with them valuable information about both the current deployment of their erstwhile comrades and their future plans. 118 Ibid. 119 Nithard, Historiarum, 2.10, “Cumque atque insperate propter urbem Alciodorensem utergue exercitus alter ab altero videretur, confestim Lodharius verens, ne forte fratres sui absque dilation supra se irruere vellent, armatus castra aliquantulum excessit.”

41

N A R R AT I V E W O R K S

Propaganda As suggested by Nelson, Nithard’s discussion of military operations should be understood as part of his broader agenda of establishing a case for “just war” carried out by Charles and Louis.120 However, in addition to the evidence from Nithard’s text itself, the author devotes considerable attention in his Histories to the efforts of the Carolingian rulers to sway “public opinion” during the course of their struggle. Nithard makes clear that communicating the proper message to both current and potential supporters about both the material and moral advantages of the “home side” was critically important for the successful pursuit of both military and political objectives. Considerable numbers of written works of propaganda were produced in the context of crises of the 830s and the civil war of the early 840s.121 However, Nithard’s main focus throughout his Histories is on the oral communications of the kings directly with their magnates in assemblies or the royal dispatch of emissaries to make the case for their patron.122 Nithard emphasizes, for example, that when Charles and Louis finally joined forces near Châlons-sur-Marne, located some 110 kilometers southeast of Fontenoy, in May 841, their first conversation focused on Lothair’s “utter lack of self control and his savage attacks on them and their supporters.”123 Nithard adds that the very next morning, the two kings held an assembly of their supporters where they set out in detail all of the calamities that they had suffered at the hands of Lothair.124 Nithard again shows the two brothers using a public assembly to influence the perspectives of their supporters in the famous oaths of Strasbourg, which were made on 14 February 842.125

120 Nelson, “Public Histories,” 251–93. 121 Janet Nelson, “The Search for Peace in a Time of War: The Carolingian Brüderkrieg, 840–843,” in Träger und Instrumentarien des Friedens im hohen und späten Mittelalter, ed. Johannes Fried (Sigmaringen, 1996), 87–114, here 98, emphasizes the large number of texts that must have been produced in the context of the crises and civil wars of the period 833–843. Among these, in February 835, Louis the Pious ordered a libellus to be produced following an assembly at the royal villa of Thionville. Louis ordered that this text, which treated in detail the unjust nature of his deposition from office in 833, be read publicly, copied, and disseminated. This libellus is noted in the Annales Bertiniani, ed. Georg Waitz, MGH SRG 5 (Hanover, 1883), an. 835. We thank Phil Wynn for his generosity in allowing us to see his manuscript dealing with the topic of propaganda in the period of the civil wars and alerting us to the rich vein of surviving texts that were produced to influence the “public.” 122 Nithard, Historiarum, 3.1, claims to have been impelled to have written the third book of his text out of fear that “forte quilibet quocumque modo deceptus res nostro in tempore gestas, praeterquam exactae sunt, narrare praesumat.” In his translation of the text Bernhard Walter Scholz, Carolingian Chronicles (Ann Arbor, 1970), 155 translated narrare as record, suggesting another work of written propaganda. This is certainly a possible translation, but it much more common to see this verb used for oral rather than written communications. 123 Nithard, Historiarum, 2.9, “quae Lodharius absque quolibet moderamine erga se suosque seviebat.” 124 Ibid. 125 Nithard, Historiarum, 3.5.

42

N I T H A R D A S A M I L I TA RY H I S T O R I A N

Each of the kings addressed both of the armies, Louis in Romance and Charles in German.126 According to Nithard, Charles and Louis laid out their case against Lothair and then promised everlasting friendship and loyalty between themselves. Nithard also presents Lothair using assemblies of his supporters to make the case for his position as the rightful ruler of the regnum Francorum. For example, Nithard draws attention to Lothair’s successful use of propaganda to enhance his political and military position vis-à-vis Charles following a significant tactical blunder by the latter. In May 841, Charles faced a choice about whether to march southwest from Attigny to join his mother and the reinforcements she was bringing from Aquitaine or to advance toward Lothair’s forces. According to Nithard, Charles summoned his magnates to discuss the two options, and the majority opted for advancing against Lothair, or at least announcing his intention of meeting Lothair at a specific location.127 This majority party counseled that if Charles withdrew to meet with his mother, this action would be characterized as a retreat by his adversaries, and those magnates who had remained neutral so far would now join Lothair.128 According to Nithard, as soon as Lothair received word of Charles’s withdrawal: “he announced to the people gathered all around that Charles had fled, and that he would pursue him as quickly as possible.” Nithard opines that “this announcement strengthened the resolve of his faithful supporters, and encouraged those who were doubtful of his cause to come over to his side more firmly.”129 This passage serves to demonstrate the important impact that propaganda could have in galvanizing support. In addition, it would seem to be a clear example of Nithard’s didactic purpose with regard to the Histories, and particularly the negative consequences that could arise when a king ignored the sound tactical advice that was proffered by his magnates. In fact, Nithard’s emphasis regarding this matter perhaps permits the conclusion that he had agreed with the majority of the magnates and opposed the withdrawal toward Aquitaine that had permitted Lothair to seize the propaganda initiative. In addition to the use of assemblies to disseminate propagandistic accounts of emerging events, Nithard observes that the Carolingian kings frequently dispatched emissaries to strengthen the resolve of their supporters and to persuade others to join their side. In September 841, for example, Charles dispatched his seneschal Adalhard and Abbot Hugh of St. Quentin (d. 844) to convince Count Gislebert of the Maasgau, located in the region of the lower Meuse, and other

126 See the discussion by Patrick Geary, “Oathtaking and Conflict Management in the Ninth Century,” in Rechtsverständnis und Konfliktbewältigung. Gesrichtliche und aussergerichtliche Strategien im Mittelalter, ed. Stefan Esders (Cologne, 2007), 239–53. 127 Nithard, Historiarum, 2.9. 128 Ibid. 129 Ibid., “Lodharius quoque ut haec ita se habere deprehendit, circumfusae plebi Karolum fugam inisse persequique ilum, quantocius posset, vell denuntiat; quo quidem nuntio fidos sibi alacriores reddidit, dubiis autem quibusque et affluendi audaciam iniecit et firmiores suae parti reddidit.”

43

N A R R AT I V E W O R K S

local magnates to join his side.130 Nithard also devotes considerable attention to Lothair’s use of emissaries and agents to ride ahead of the army in order to persuade potential supporters, or opponents, that he would be victorious and that they should take the appropriate action in light of this reality. As discussed earlier, Lothair used this strategy to great effect while campaigning against Louis the German in the spring of 841.131 In fact, Nithard emphasizes throughout his text that it was Lothair’s common practice to utilize agents in this manner. In discussing Lothair’s advance toward the Loire in October 940, for example, Nithard comments that the Carolingian king dispatched emissaries in his habitual manner (more solite) to convince the inhabitants between the Loire and the Seine to support him, either with promises or with threats.132 In Nithard’s account, all three of the royal brothers clearly utilized propaganda to encourage their supporters and to undermine the support of their opponents. Nithard’s parti pris on behalf of Charles and Louis is illuminated, however, when the content of this propaganda is considered in detail. Nithard consistently presents Lothair offering material gain to men who would support him and threatening with grave losses those who failed to do so. By contrast, Charles and Louis, according to Nithard’s telling, utilized arguments based upon justice and morality.133

Conclusion Carolingian pedagogues stressed the value of learning from books, and the sons of Louis the Pious were substantial consumers of knowledge preserved by the written word.134 Nithard certainly must be understood to be in this tradition, even if he engaged in an altogether uncommon activity as a layman by writing a history that focused almost exclusively on contemporary affairs.135 Nithard’s choice to do so permits the insight that he regarded this information as necessary for men of his

130 Nithard, Historarium, 3.3, “Hugoenm et Adhelhardum ad Gislebertum una cum ceteris foedere, quo valerent sibi adnecterent, direxit.” 131 Nithard, Historiarum, 2.7. 132 Nithard, Historiarum, 2.4, “praemittens more solite, qui ad defectionem inter Sequanam et Ligerem degentes partim minis, partim blanditiis subducerent.” 133 The role of justice in Nithard’s text is discussed in detail by Patze, “Iusticia,” 147–65. 134 See, for example, Rosamond McKitterick, “The palace school of Charles the Bald,” in Charles the Bald: Court and Kingdom, ed. Janet L. Nelson and Margaret T. Gibson (Aldershot, 1990), 326–39, and reprinted with the same pagination in Rosamond McKitterick, The Frankish Kings and Culture in the Early Middle Ages (Aldershot, 1995). 135 The writing of history by laymen was quite uncommon. The writing of strictly contemporary work of history, as contrasted with biography, without contextualizing modern affairs in a lengthy prologue dating back to the birth of Jesus or even to the Creation, also was exceptionally rare in the period before the First Crusade. There are only three such texts that survive between the ninth and eleventh centuries: Nithard’s work, Alpert of Metz’s De diversitate temporum, and Bruno of Merseburg’s Bellum Saxonum. For a discussion of this issue, see Bernard S. Bachrach and David S. Bachrach, “Bruno of Merseburg and His Historical Method c. 1085,” Journal of Medieval History 40 (2014), 381–98.

44

N I T H A R D A S A M I L I TA RY H I S T O R I A N

status and profession, and consequently can be understood to illuminate the actual conduct of war in the mid-ninth century. Of primary importance, given the attention that Nithard devotes to the topic, is the crucial relationship between political and military affairs. In Nithard’s account, the fighting and winning of battles comprised the final stage in a lengthy diplomatic, political, and propaganda struggle to demonstrate both the moral and practical superiority over one’s foes. This presentation of warfare, as distinguished from actual combat on the field, highlights Nithard’s understanding of the multifaceted relationship of the Carolingian rulers of this generation with the magnates in the regnum Francorum. In addition, Nithard’s detailed discussion of the extensive maneuvering by the Carolingian kings, on the one hand, to gain the support of the magnates and, on the other, to place their armies in a position to gain maximum long-term advantage makes clear his views regarding the political complexity of military operations in this period. As discussed earlier, Nithard also made numerous observations concerning the size of the Carolingian armies, their composition, and the strategic and tactical decisions of the contenders for the rule of the regnum Francorum. From Nithard’s perspective, and he was in a very good position to know, the Carolingian kings sought to mobilize very large armies and did so utilizing both the military households of magnates, as well as expeditionary levies drawn from militia forces. It is noteworthy, however, that little is said of the local levies who were too poor to be mobilized for the offensive operations that dominated Nithard’s account, although he did draw attention to the plebs who operated in defense of the Rhineland. In terms of military operations, Nithard’s observations permit the inference that the Carolingian rulers possessed and relied upon an extensive apparatus for obtaining accurate and timely information about the military forces of their opponents. In short, Nithard presents them as working diligently and often successfully to acquire military intelligence. Nithard’s discussion of the battle-seeking and battle-avoiding strategies of Charles, Louis, Lothair, and their various commanders also may suggest that Vegetian-style conceptions of the conduct of war were thoroughly ingrained within the leadership cadres of Carolingian society. These Carolingian rulers, as presented by Nithard, did not seek out battle very often. When they did seek battle, it would appear from Nithard’s account that this was because they believed that they had a significant numerical advantage. Finally, propaganda played an essential role in the conduct of military operations, according to Nithard. He presents the Carolingian kings working hard to show their magnates that they were fighting on the just side, and moreover the winning side. In this manner, Nithard connects effective propaganda quite closely to both political and military success. According to Nithard, having a large army showed wavering magnates that a Carolingian king already enjoyed broad support. By contrast, having a small army suggested that a ruler faced grave doubts about his ability to win.

45

3 S A X O N M I L I TA RY R E V O L U T I O N, 912–973? Myth and reality

It is generally agreed that the Saxon kings of East Francia (west and west-central Germany) in the tenth century enjoyed great military and political success. They descended from Liudolf, who became the dominant figure in the Saxon region under the later Carolingians (ca. 850). Henry the Fowler, a direct descendent of Liudolf in the male line, became king of the eastern third of the erstwhile Carolingian Empire in 919. Following in the tradition of Charles the Great, Henry’s son, Otto I (936–973) acquired the imperial title in 962.1 Over the past decades, many scholars have worked diligently to identify the conditions and causes that made possible the rapid rise of this dynasty to dominance in west-central and western Europe as well as ultimately in northern Italy.2 In 1968, Karl Leyser, Chichele

1 For an overview of the history of this period with a focus on the Ottonian/Liudolfing dynasty, see Timothy Reuter, Germany in the Early Middle Ages 800–1056 (London, 1991), 179–211. For an older account covering the Ottonian period exclusively, see Robert Holtzmann, Geschichte der sächsischen Kaiserzeit 900–1024, 4th edition (Munich, 1961). 2 The question of how the Liudolfing dynasty rose to power has generated considerable scholarly interest for more than a century and a half. For an introduction to the vast literature on this topic, see Georg Waitz, Deutsche Verfasssungsgeschichte V: Die deutsche Reichsverfassung von der Mitte des neunten bis zur Mitte des zwölften Jahrhunderts, 3rd edition (Berlin, 1874, repr. Darmstadt, 1954); Hagen Keller, “Reichsstruktur und Herrschaftsauffasung in ottonisch-frühsalischer Zeit,” Frühmittelalterliche Studien 16 (1982), 74–128; Idem, “Zum Charakter der ‘Staatlichkeit’ zwischen karolingischer Reichsreform und hochmittelalterlichem Herrschaftsausbau,” Frühmittelalterliche Studien 23 (1989), 248–64; Timothy Reuter, Germany in the Early Middle Ages 800–1056 (London, 1991), 136–48 and Idem, “The Medieval German Sonderweg?: The Empire and Its Rulers in the High Middle Ages,” in Kings and Kingship in Medieval Europe, ed. Anne J. Duggan (Exeter, 1993), 179–211; Gerd Althoff and Hagen Keller, Heinrich I. und Otto der Grosse: Neubeginn auf karolingischem Erbe, vol. 1, 2nd edition (Göttingen, 1994); and Gerd Althoff, Die Ottonen: Königsherrschaft ohne Staat (Stuttgart, 2000). In this context, Matthias Becher, Rex, Dux und Gens: Untersuchungen zur Entstehung des sächsischen Herzogtums im 9. und 10. Jahrhundert (Husum, 1996), has demonstrated conclusively that the Saxon region was not constituted as a duchy before the late tenth century and that neither Henry I nor his predecessors held office as dukes. As a consequence, we have removed all references to Henry as duke and to Saxony as a duchy in this revised version of the text. Also see in this context, Die Salier und das Reich, vol. 1: Salier, Adel und Reichsverfassung (1991), 503–47; Hans-Werner Goetz, “Das Herzogtum der Billunger- ein sächsischer Sonderweg?,” Niedersächsisches Jahrbuch

46

S A X O N M I L I TA RY R E V O L U T I O N , 912–973?

Professor of modern history at Oxford, published what today remains the most influential explanation for the success of the Saxon dynasty.3 Leyser’s thesis was that Henry I, the Saxon duke and later king, along with his son Otto I, were successful, in large part, because they were able to defeat militarily their rivals in the duchy of Saxony, overcome opposition in the kingdom of East Francia, and ultimately gain dominance in the neighboring kingdom of Italy, as well as over the Magyars and Slavs on their eastern frontiers. The key to their military success, according to Leyser, was the development of a large body of heavily armed mounted fighting men. Leyser argued that such troops, whom he often calls “knights,” dominated the battlefields not only of Saxony but also of western and central Europe from 933 onward.4 As Leyser understood military development in the early Middle Ages, the recruitment and training of significant contingents of expensive, heavily armed cavalry was initiated throughout the Frankish kingdom by Charles Martel (716– 741) following his victory over the Muslims at Poitiers in 732. These efforts putatively were pursued vigorously by Charles’s successors Pippin (741–768) and Charlemagne (768–814). Leyser asserted that the development of these forces was accomplished through the course of the later eighth and ninth centuries in West Francia, Lotharingia, and some of the more westerly parts of East Francia, as these regions are represented in the divisio regnorum that was ratified by the treaty of Verdun in 843. He argued, however, that the Saxons did not keep pace. Thus, prior to the accession of Henry as the dominant figure in the Saxon region in 912, the military forces of the Saxon duchy had not participated in or, at least had not participated as fully as necessary in, the process of building a major force of heavily armed and mounted troops. This problem was solved, according to Leyser, no later than 933 when Henry deployed a sufficiently large number of mounted and heavily armed fighting men (armati) to “encourage” a large force of lightly

für Landesgeschichte 66 (1994), 167–97, whose conclusions regarding the late date of the Saxon duchy largely are consistent with those of Becher. 3 The clearest statement of Karl Leyser’s thesis is “Henry I and the Beginnings of the Saxon Empire,” The English Historical Review 83 (1968), 1–32 and reprinted in Karl J. Leyser, Medieval Germany and Its Neighbours 900–1250 (London, 1982), 11–42. All citations are to the latter publication. Leyser deals with several of his major premises concerning the military organization of Saxony in “The Battle at the Lech, 955: A Study in Tenth-Century Warfare,” History 50 (1965), 1–25 and reprinted in Medieval Germany and Its Neighbours, 43–67 (latter pagination used in this study); and “Early Medieval Warfare,” in Communications and Power in Medieval Europe: The Carolingian and Ottonian Centuries, ed. Timothy Reuter (London, 1994), 29–50. 4 Leyser, “Saxon Empire,” 23; Idem, “Battle at the Lech,” 53, 56, and 57; and Idem, “Medieval Warfare,” 41. With regard to Leyser’s views concerning German knighthood, also see “Early Medieval Canon Law and the Beginnings of Knighthood,” in Institutionen, Kultur und Gesellschaft im Mittelalter. Festschrift für Josef Fleckenstein zu seinem 65. Geburtstag, ed. Ludwig Fenske, Werner Rösener, and Thomas Zotz (Sigmaringen, 1984), 549–66, and reprinted in Communications and Power, 51–71. Nb. This study, unlike that of Leyser, is concerned fundamentally with the military organization of the Saxon region. Those interested in the problematic mater of Saxon “identity” and “ethnogenesis” may see the relevant work of Becher, Rex, dux, gens, passim.

47

N A R R AT I V E W O R K S

armed Magyar horse archers to avoid battle at Riade and to retreat precipitously from southeastern Saxony.5 It should be made clear that Leyser appears not to have been much interested in matters of military strategy and technology. His study was focused on the social impact of the creation of heavy cavalry, and he used the term “knight” frequently in his work as a translation of the Latin term miles. In this way, he seems to have drawn on the then-current thesis, first published in 1887 by Heinrich Brunner, which was widely popularized and imaginatively augmented by Lynn T. White Jr. in 1962. The Brunner–White thesis postulated a “military revolution” in the Frankish kingdom and held that Charles Martel’s early Carolingian military establishment, which had been inherited from the last Merovingians, was radically altered. In this model, the armies of the Merovingian kings prior to 732 lacked a significant cavalry component, much less a dominant heavy cavalry component capable of sweeping lightly armed horsemen and foot soldiers from the field of battle by dint of shock combat realized through a mounted charge. Thus, it is Charles Martel who is presented as having reorganized the Merovingian military system by creating a force of heavily armed mounted troops that dominated warfare in the West not only in the Carolingian era but throughout the Middle Ages.6 It was this so-called “military revolution” in the Frankish kingdom, according to the Brunner–White thesis, that also brought about a tenurial revolution with the supposed advent of “feudalism” when Charles Martell allegedly initiated the systematic granting out of church lands as benefices to support this new force composed of royal vassals. These revolutions in military structure and tenurial practices also are claimed to have paved the way, as a result of continuous military success, for the Carolingian mayor of the palace, Pippin, to arrange his elevation to the royal throne in 751. Charlemagne, of course, continued the process and obtained the imperial title in 800. There are obvious similarities of this Carolingian success story based upon military dominance to that told by Karl Leyser. In this context, it should be emphasized that since 1970, the Brunner–White thesis has been shown to be untenable in terms of both military history and technology.7 In a parallel vein, the magisterial work of Susan Reynolds, Fiefs and

5 Leyser, “Saxon Empire,” 32. 6 Heinrich Brunner, “Der Reiterdienst und die Anfänge des Lehnwesens,” Zeitschrift der SavignyStiftung für Rechtsgeschichte, Germanistische Abtheilung 8 (1887), 1–38, reprinted in Brunner, Forschungen zur Geschichte des deutschen und französischen Rechts (Stuttgart, 1894), 39–74; and Lynn T. White, Medieval Technology and Social Change (Oxford, 1962), 1–38 and 135–53. For a recent defense of the validity of the Brunner-White model for understanding early medieval warfare, see J. F. Verbruggen, “The Role of Cavalry in Medieval Warfare,” Journal of Medieval Military History 3 (2005), 46–71, here 62. 7 Concerning the challenges to the Brunner-White thesis see, R. H. Hilton and P. H. Sawyer, “Technological Determinism: The Stirrup and the Plough,” Past and Present 24 (1963), 90–100; Donald A. Bullough, “Europae Pater: Charlemagne and His Achievement in the Light of Recent Scholarship,” English Historical Review 85 (1970), 59–105, here 84–90; and Bernard S. Bachrach, “Charles Martel, Mounted Shock Combat, the Stirrup and Feudalism,” Studies in Medieval and

48

S A X O N M I L I TA RY R E V O L U T I O N , 912–973?

Vassals, demonstrated that the supposed institutional revolution in land holding, i.e. “feudalism,” also did not take place in pre-Crusade Europe.8 Leyser, of course, argued for a military revolution in the Saxon duchy before the Brunner– White thesis was systematically discredited. While his interests, as already noted, were primarily with the political implications of social changes taking place within the nobility, in seeing the crucial development as the creation of a heavy cavalry force of soldiers, whom he felt able to designate as “knights,” his underlying interpretative scheme was heavily influenced by these received ideas. In fact, the term miles, often translated by Leyser as “knight,” retained its basic meaning of soldier during the Ottonian period and was not an indication of social standing.9 Since the time that Leyser wrote, moreover, it has been established that real warfare aimed at conquest and the control of enemy territory was dominated by sieges and not by battle in the field.10 Further, it has been established that mounted shock combat was a minor aspect even in the rare battles in the field that were fought throughout

Renaissance History 7 (1970), 49–75, and reprinted with the same pagination in idem, Armies and Politics in the Early Medieval West (Aldershot, 1993), esp. 49–52. Concerning the general acceptance by specialists in military history of Bachrach’s rebuttal of the Brunner-White thesis, see Kelly DeVries, Medieval Military Technology (Peterborough, 1992), 95–122, and now the second edition with Robert Douglas (Toronto, 2012). 8 Susan Reynolds, Fiefs and Vassals: The Medieval Evidence Reinterpreted (Oxford, 1994, repr. 2001). 9 In this context, it should be noted that Leyser used the term knight frequently in his work to denote milites, a usage that runs counter to the observations regarding the meaning of this term as developed by Joachim Bumke, Studien zum Ritterbegriff im 12. und 13. Jahrhundert, 2nd revised edition (Heidelberg, 1977), which maintains the same basic view regarding the use of the term milites as the first edition (Heidelberg, 1964). This work was translated into English as The Concept of Knighthood in the Middle Ages, trans. W. T. H. Jackson and E. Jackson (New York, 1982), where Bumke emphasizes (38) that the term miles is not an indication of social standing, but rather retained its basic meaning of soldier during the Ottonian period. In this context, also now see David S. Bachrach, “Memory, Epistemology, and the Writing of Early Medieval Military History: The Example of Bishop Thietmar of Merseburg (1009–1018),” Viator 38 (2007), 63–90; and Idem, “Milites and Warfare in Pre-Crusade Germany,” War in History 23.3 (2015), 298–343. For the west of France, also see Bernard S. Bachrach, “Milites and the Millennium,” in idem, Warfare and Military Organization in Pre-Crusade Europe (Aldershot, 2002), 85–95. 10 It has long been understood by specialists in early medieval warfare that military efforts intended to capture and hold enemy territory, and to defend one’s own territory against conquest, were dominated by sieges. See in this regard, Jim Bradbury, The Medieval Siege (Boydell, Woodbridge, UK, 1992); and Peter Purton, A History of the Early Medieval Siege, c. 450–1200 (Boydell, Woodbridge, 2010), and the enormous literature cited there; and now Leif Inge Ree Petersen, Siege Warfare and Military Organization in the Successor States (400–800 AD): Byzantium, the West, and Islam (Leiden, 2013). With regard to the centrality of siege warfare to the specifically Carolingian context, see B. S. Bachrach, Early Carolingian Warfare: Prelude to Empire (Philadelphia, 2001), passim. With regard to the centrality of siege warfare in the Ottonian context, see David S. Bachrach, “Early Ottonian Warfare: The Perspective from Corvey,” Journal of Military History 75 (2011), 393–410; Idem, “Restructuring the Eastern Frontier: Henry I of Germany 924–936,” Journal of Military History 78.1 (2014), 9–35; and Bernard S. Bachrach and David S. Bachrach, “Early Saxon Frontier Warfare: Henry I, Otto I, and Carolingian Military Institutions,” Journal of Medieval Military History 10 (2012), 17–60.

49

N A R R AT I V E W O R K S

pre-Crusade Europe.11 Finally, the tenurial history of pre-Crusade Europe needs rewriting in light of Reynold’s argument that the construct of feudalism was the result of a reading back of legal views developed centuries later.12 Leyser cannot be held accountable for failing to consider the subsequent major developments in scholarship regarding the exceptionally complicated relationships between military organization and social status. However, it must be emphasized that Leyser’s model of the rise of the army of mounted fighting men, and the concomitant emergence of a knightly class, is still the dominant model among scholars specializing in the history of early medieval Germany, particularly in the Germano-phone tradition.13 Consequently, the present state of the question regarding a supposed Saxon military revolution creates something of a paradox. There has been very little new research on this topic since Leyser’s influential studies. Thus, it is fair to suggest that the “Saxon Military Revolution” has been sustained largely as a result of inertia among scholars in the fields of Saxon political, institutional, and military history.14 The purpose of this study is to reexamine 11 See, for example, see two articles by Bernard S. Bachrach, “Charlemagne’s Cavalry: Myth and Reality,” Military Affairs, 47 (1983), 181–7, and reprinted in Bernard S. Bachrach, Armies and Politics in the Early Medieval West (London, 1993), 1–20; and “Caballus et Caballarius in Medieval Warfare,” in The Study of Chivalry, ed. H. Chickering and T. H. Seiler (Kalamazoo, 1988), 173–211 and reprinted with the same pagination in Bernard S. Bachrach, Warfare and Military Organization in Pre-Crusade Europe (Aldershot, UK, 2002). 12 Reynolds, Fiefs and Vassals, 1–16. 13 Leopold Auer, “Zum Kriegswesen unter den früheren Babenbergern,” Jahrbuch für Landeskunde von Niederösterreich 42 (1976), 9–25, here 16, argues, for example, that the central problem of medieval military history is the connection between Kriegswesen and social structure. In this regard, see also H.-J. Bartmuß, Deutsche Geschichte in zwölf Bänden, vol. 1 (Berlin, 1982), 370; Reinhard Wenskus, “Die soziale Entwicklung im ottonischen Sachsen im Lichte der Königsurkunden für das Erzstift Hamburg-Bremen,” in Institutionen, Kultur und Gesellschaft: Festschrift für Joseph Fleckenstein zu seinem 65. Geburtstag, ed. Lutz Fenske, Werner Rösener, and Thomas Zotz (J. Thorbecke: Sigmaringen, 1984), 501–14; Idem, “Die ständische Entwicklung im Sachsen im Gefolge der fränkischen Eroberung,” in Angle e Sassoni al di qua e al di là del mare (Settimane de studio del centro italiano di studi sull’alto medioevo), vol. 32 (Spoleto, 1986), 587–616, here 614; Helmut Beumann, Die Ottonen (Stuttgart, 1987), 44; Otto Ackermann, “Comites und Milites: Grafen und Krieger im Hochmittelalter,” Werdenberger Jahrbuch 7 (1994), 11–20, here 11; and Franz-Reiner Erkens, “Militia und Ritterschaft: Reflexionen über die Entstehung des Rittertums,” Historische Zeitschrift 258 (1994), 623–59, esp. 629. Special attention should be drawn to the seminal article by Karl Ferdinand Werner, “Heersorganization und Kriegsführung im deutschen Königreich des 10. und 11. Jahrhunderts” Settimane di Studio de Centro Italiano Sull’alto Medievo 15 (1968), 791–843, who, although accepting the general model of the rise of the armored Ritterheer, nevertheless argues (esp. 791–806) for fundamental continuity in military organization from Charlemagne through Otto I. In this context, Werner points out particularly that the military resources of the Saxon regnum were fully integrated into the overall Carolingian military organization by the end of the ninth century. 14 The view that heavy cavalry dominated warfare in the tenth-century Reich was common before Leyser wrote “Saxon Empire.” See, for example, Hans Fehr, “Das Waffenrecht der Bauern im Mittelalter,” Zeitschrift der Savigny Stiftung für Rechtsgeschichte: Germanistische Abteilung 35 (1914), 111–21, esp. 118; and Paul Schmitthenner, “Lehnskriegwesen und Söldnertum im abendländischen Imperium des Mittelalters,” Historische Zeitschrift 150 (1934), 229–67.

50

S A X O N M I L I TA RY R E V O L U T I O N , 912–973?

aspects of the military organization of the Saxon region from the late Carolingian era through the reign of Otto I in order to ascertain if a military revolution, in fact, took place during this period.

Leyser’s evidence for a military revolution Leyser based his view that there was a military revolution in the Saxon region on the Res gestae Saxonicae written after c. 973 by Widukind, a monk from the monastery of Corvey.15 Widukind was a member of a noble Saxon family, named after and likely related to the Widukind who led Saxon military resistance to Charlemagne during the late eighth century.16 He spent his youth and adult life as a student and then monk at the prestigious Saxon monastery of Corvey, where he lived with numerous other men from noble families, including Ruotger, the author of the vita of Archbishop Brun of Cologne, the younger brother of King Otto I (934–973).17 Widukind wrote the bulk of his Res gestae Saxonicae c. 968 and then revised and augmented the text sometime after 973.18 The reasons why Widukind wrote his history of the Saxon gens and the reigns of the first two kings of the

15 16

17

18

Leyser’s views regarding the dominance of mounted troops in the wars of the tenth-century German kingdom also have garnered considerable support over the past three decades from specialists in early medieval German history. In addition to the studies listed in the note earlier, Leopold Auer, “Zum Kriegswesen unter der früheren Babenbergern,” Jahrbuch für Landeskunde von Niederösterreich 42 (1976), 9–25, esp. 16–19 and Idem, “Formen des Krieges im abendländischen Mittelalter,” in Formen des Krieges, ed. Manrfried Rauchensteiner and Erwin A. Schmidl (Graz, 1991), 17–43, esp. 19–23; Keller, “Reichsstruktur,” 82; Gerd Althoff and Hagen Keller, Heinrich I. und Otto der Grosse: Neubeginn auf karolingischem Erbe, 2nd edition, vol. 1 (Göttingen, 1994), 87; Timothy Reuter in a series of articles, “Plunder and Tribute in the Carolingian Empire,” Transactions of the Royal Historical Society, 5th series 35 (1985), 75–94; Idem, “The End of Carolingian Military Expansion,” in Charlemagne’s Heir: New Perspectives on the Reign of Louis the Pious (814–840), ed. P. Godman and R. Collins (Oxford, 1990), 391–405; and Idem, “Carolingian and Ottonian Warfare,” in Medieval Warfare: A History, ed. Maurice Keen (Oxford, 1999), 13–35. Leyser, “Saxon Empire,” 14. For a useful brief overview of Widukind’s life and text, see Richard Engel, “Widukind von Corvey,” in Weltbild und Realität: Einführung in die mittelalterliche Geschichtsschreibung, ed. Ulrich Knefelkamp (Pfaffenweiler, 1992), 85–92. For Widukind’s age and education at Corvey, as well as his study alongside Ruotger see Helmut Beumann, Widukind von Korvey: Untersuchungen zur Geschichtsschreibung und Ideengeschichte des 10. Jahrhunderts (Weimar, 1950), 1–5. With respect to Ruotger’s study at Corvey, see Heinrich Schrörs, “Die Vita Brunonis des Ruotger,” Annalen des historischen Vereins für den Niederrhein 90 (1911), 61–100, here 61–2. Regarding the chronology of Widukind’s composition of the Res gestae, see Beumann, Widukind von Korvey, 178, where Beumann deals with the older scholarly traditions dealing with the compositional history of the text and concludes that the work was completed up to book III, chapter 69 which dealt with the death of the rebel magnate Wichmann in in 967 or 968. Beumann’s chronology, including the augmentation of the text by Widukind after Otto I’s death in 973 generally has been accepted by scholars. In this regard, see Gerd Althoff, “Widukind von Corvey: Kronzeuge und Herausforderung,” Frühmittelalterliche Studien 27 (1993), 253–72, here 258–9. However, now see the introduction to Bernard S. Bachrach and David S. Bachrach, Widukind of Corvey:

51

N A R R AT I V E W O R K S

Saxon dynasty, Henry I (919–934) and Otto I, and the sources that he had available have been the subject of considerable scholarly debate since the mid-nineteenth century.19 The answers scholars have posited to these questions have had a significant influence on their evaluation of the historical value of the information provided by Widukind in the Res gestae. The earlier scholarly tradition dealing with the historical value of Widukind’s work largely was negative. In 1899, Wilhelm Gundlach made the famous claim that Widukind was a “trickster in a monk’s habit.”20 In taking this view, Gundlach was maintaining the state of the question established three decades earlier by Rudolf Köpke, the first scholar to subject the Res gestae to systematic source criticism.21 In the new century, Widukind’s text still suffered significant criticism at the hands of scholars. Karl Hainer, in his Gießen dissertation of 1914, characterized Widukind as an epic poet who had no interest in writing history.22 Robert Holtzmann, in his revision of Wilhelm Wattenbach’s volume of source criticism, took an even harsher view of the historical value of the Res gestae and concluded in 1938 that Widukind was a naïve monk who uncritically wrote down what he heard without leaving the monastery of Corvey.23 Holtzmann added further that Widukind did not have access to written sources of information and was therefore entirely dependent on hearsay, which meant that he included numerous fables in his work.24 The process of rehabilitating Widukind’s reputation as a trustworthy source of information about the tenth-century Saxon duchy began with Helmut Beumann’s study of the Res gestae in 1950.25 In this work, and in numerous articles over the following five decades, Beumann makes a convincing case that Widukind is to be taken seriously as a trustworthy source of information about the tenthcentury Saxon duchy and the German kingdom.26 In his original study, Beumann

19 20 21 22 23 24 25 26

Deeds of the Saxons: Translation and Commentary (Washington, DC, 2014), where the scholarly tradition dealing with the chronology of Widukind’s text is treated in detail. The first systematic effort to subject Widukind’s text to source criticism was Rudolf Köpke, Widukind von Korvei (Berlin, 1867). Wilhelm Gundlach, Heldenlieder der deutschen Kaiserzeit, vol. 1 (1894), 112, “Spielmann in der Kutte.” See Köpke, Widukind, 70. Carl Hainer, Das epische Element bei den Geschichtsschreibern des früheren Mittelalters (Gießen dissertation, 1914). On the nature of Hainer’s views, see Beumann, Widukind von Korvey, 51–65. Wilhelm Wattenbach and Robert Holtzmann, Deutschlands Geschichtsquellen im Mittelalter: Deutsche Kaiserzeit (Berlin, 1938), 27–8. Ibid., 28, “Was er hörte, hat er im allgemeinen gutgläubig, aber doch keineswegs immer ohne Kritik aufgenommen und gebucht.” Beumann, Widukind von Korvey, passim. See, for example, Helmut Beumann, “Die Historiographie des Mittelalters als Quelle für die Ideengeschichte des Königtums,” Historische Zeitschrift 180 (1955), 449–88; Idem, “Historiographische Konzeption und politische Ziele Widukinds von Corvey,” in La storiografia altomedievale: Settimane di studio del centro italiano di studi sull’alto medioevo 17 (1970), 857–94; and Idem, “Imperator Romanorum, rex gentium. Zu Widukind III 76,” in Tradition als historische Kraft:

52

S A X O N M I L I TA RY R E V O L U T I O N , 912–973?

stressed that Widukind not only had available a large number of models of history writing from his study at Corvey, but that the monk adhered to a rigorous historical method when he wrote.27 Beumann did not obscure the fact that Widukind made errors, occasionally relating fiction as fact, but concluded, “it is unlikely, however, that he was aware of this when he did it.”28 In later studies, Beumann took pains to demonstrate that, contrary to the argument made by Holtzmann, Widukind had access to considerable information from a wide variety of sources. Beumann emphasizes, for example, Widukind’s close connection to leading figures of the Ottonian dynasty and his ability to obtain significant information from them.29 Even more crucially, Beumann also argued that rather than simply relying on rumor or anecdotes from acquaintances at his monastery to write his history, Widukind also had access to written documents, including those produced by the royal government and by the papacy.30 Although several of Beumann’s ideas concerning Widukind’s work have not been widely accepted by scholars, including the view that his work was written in the context of the controversy surrounding the establishment of the archbishopric of Magdeburg by Otto I, the Res gestae now generally is considered by scholars to provide a considerable corpus of accurate information about affairs in the Saxon duchy and the broader German kingdom, from both his own day and from earlier periods as well.31 The value of Widukind in presenting details about important aspects of royal ceremony, for example, was emphasized by Josef Fleckenstein.32 Karl Leyser, as noted earlier, trusted implicitly Widukind’s testimony regarding Saxon military organization. In a more wide-ranging study of historical value of the Res gestae, Gerd Althoff stressed that the quality of the information provided by Widukind must be assessed in the context of the intended audience for the work.33 Pursuing this theme, Althoff emphasized that Widukind was commissioned by a leading member of the Ottonian royal family, i.e. Archbishop William of Mainz, the queen-mother Mathilda, or Otto I himself, to write the Res gestae as a didactic tool for Princess Mathilda, the abbess of Quedlinburg, in order to prepare her for a major role in the governance of the kingdom.34 As a consequence, according to Althoff, the information provided by Widukind had to be a trustworthy representation of reality. Althoff

27 28 29 30 31

32 33 34

Interdisziplinäre Forschungen zur Geschichte des früheren Mittelalter, ed. Norbert Kamp and Joachim Wollasch (Berlin, 1982), 214–30. Beumann, Widukind von Korvey, 53–60. Ibid., 61. Beumann, “Historiographische Konzeption,” 885. Beumann, “Imperator Romanorum,” 217 and 221. With respect to Beumann’s view regarding the connection between the writing of the Res gestae and the establishment of Magdeburg, see Beumann, Widukind, 195; and Idem, “Imperator Romanorum,” 221. For a critique of this conclusion, see Althoff, “Widukind von Corvey,” 258. Josef Fleckenstein, Grundlagen und Beginn der deutschen Geschichte (Göttingen, 1974), 141. This is the gravamen of Althoff, “Widukind von Corvey”. Althoff, “Widukind von Corvey,” 262–72.

53

N A R R AT I V E W O R K S

concluded his study by observing, “In Widukind’s work, we have presentations of this century [the tenth century] which were written down in the context of a very serious work. It would require considerable intellectual arrogance for anachronists to belittle this work as unserious.”35 More recently, Karl Morrison independently came to the conclusion that Widukind’s Res gestae had the very important real-world purpose of preparing the princess and abbess Mathilda for her responsibilities as a member of the ruling dynasty.36 In this context, Morrison stressed that Mathilda did exercise considerable political power from 968, when she was the sole member of the Ottonian family north of the Alps, until her death in 999.37 In an effort to prepare her, and her entourage, for this role, Widukind, according to Morrison, was commissioned to write a text that would not only provide the princess and her advisors with considerable valuable information but would require them to participate in their own education by filling in the gaps in the narrative that Widukind purposely left there.38 Thus, rather than composing a bricolage, as some of his modern critics had claimed, Widukind, in Morrison’s view, had a high opinion of the knowledge of his audience about both historical matters and current affairs and expected them to consider what he had left out of the text and to discuss these matters.39 Widukind’s appreciation of the considerable knowledge of his audience is, of course, critical in evaluating the trustworthiness of the Res gestae as a source of accurate historical information. In addition to the requirement of providing useful, and thus accurate, information for didactic purposes, Widukind labored under the rhetorical obligation of keeping the trust of an audience that already was familiar with the events that filled his text.40 The only significant modern critique of the historical value of the information presented by Widukind in the Res gestae is Johannes Fried’s effort to return to the model presented by Holtzmann, discussed earlier, that of Widukind as a recorder of oral information.41 In this context, Fried argues, as part of a broad-gauged attack on 35 Ibid., 272. 36 Karl F. Morrison, “Widukind’s Mirror for a Princess: An Exercise in Self-Knowledge,” in Forschungen zur Reichs-, Papst- und Landesgeschichte. Peter Herde zum 65. Geburtstag von Freunden, Schülern und Kollegen dargebracht, 2 vols., ed. Karl Borchardt and Enno Bünz (Stuttgart, 1998), 49–71. 37 Ibid., 50. 38 Ibid., 53. 39 Ibid., 67–8, and passim. 40 With respect to the imperatives of style, organization, and content of historical works intended for oral presentation, with a focus on Widukind and his later contemporary Dudo of St. Quentin, see Lars Boje Mortensen, “Stylistic Choice in a Reborn Genre: The National Histories of Widukind of Corvey and Dudo of St. Quentin,” in Dudoni de San Quintino, ed. Paolo Gatti and Antonella Degl’Innocenti (Trent, 1995), 78–102, especially, 94–102. Also see the important observations by Justin C. Lake, “Truth, Plausibility, and the Virtues of Narrative at the Millennium,” Journal of Medieval History 35 (2009), 221–38. 41 See in this context, Johannes Fried, “Die Kunst der Aktualisierung in der oralen Gesellschaft: Die Königserhebung Heinrichs I. als Exempel,” Geschichte in Wissenschaft und Unterricht 44 (1993),

54

S A X O N M I L I TA RY R E V O L U T I O N , 912–973?

the historical value of all early medieval narrative sources, written texts based on the transmission of information through oral means simply cannot be trusted. However, Fried conspicuously did not address in his studies the specific arguments made by Beumann that Widukind had access to written sources of information. In addition, Fried did not answer the arguments of Althoff and Morrison that Widukind’s patrons not only thought that the monk could provide accurate information to Mathilda but commissioned him to do so. Indeed, for Widukind to have provided inaccurate information to Mathilda might have caused significant difficulties for the young abbess in her efforts to help govern the German kingdom. The failure to provide accurate information, available from other people, would also have undermined the reputation of Widukind, who, as noted earlier, benefited from royal patronage for his work. Finally, Fried did not distinguish in his studies between the information that is central to the author’s parti pris and the information that provides the context in which the central narrative is told. Thus, if Widukind can be shown to have provided a version of a particular event that was desired by his patrons, it is specifically here that one would expect to find that the information not directly related to the author’s storyline is accurate.42 For all of these reasons, it is not surprising that Fried’s efforts to marginalize the historical value of the information provided by Widukind in the Res gestae have not gained the acceptance of specialists in Ottonian history and historiography.43 From the perspective of evaluating Widukind as a source of information for specifically military matters, as contrasted with, for example, singularly important political events, three factors must be considered. First, Widukind was himself a member of a noble Saxon family and likely had close relatives, including his own father, who either served personally in a military capacity in ducal or royal armies or had to provide soldiers for military service.44 In this context, Widukind specifi493–503; Idem, “Die Königserhebung Heinrichs I. Erinnerung, Mündlichkeit und Traditionsbildung im 10. Jahrhundert,” in Mittelalterforschung nach der Wende 1989, ed. Michael Borgolte (Munich, 1995), 267–318; Idem, “Erinnerung und Vergessen. Die Gegenwart stiftet die Einheit der Vergangenheit,” Historische Zeitschrift 273 (2001), 561–93; and Idem, Die Schleier der Erinnerung: Grundzüge einer historischen Memorik (Munich, 2004). This same concern had been raised in somewhat less breathless form two decades earlier by Franz-Josef Schmale, Funktion und Formen mittelalterlicher Geschichtsscreibung (Darmstadt, 1983), especially 19–20. For a detailed criticism of Fried’s assertions, see Bachrach, “Memory, Epistemology, and the Writing of Early Medieval Military History,” particularly 64–70. 42 Thomas Scharff made this point very clearly in Die Kämpfe der Herrscher und der Heiligen: Krieg und historische Erinnerung in der Karolingerzeit (Darmstadt, 2002). 43 See, for example, the survey of this question by Johannes Laudage, “Widukind von Corvey und die Deutsche Geschichtswissenschaft,” in Von Fakten und Fiktionen: Mittelalterliche Geschichtsdarstellungen und ihre kritische Aufarbeitung, ed. Johannes Laudage (Cologne, 2003), 193–224, especially 209–24; Markus Brömel, “Widukind von Corvey: Kronzeuge und Herausforderung,” in Auf den Spuren der Ottonen III (Halle, 2002), 17–25, here18. See also the effective critique of Fried’s methodology by Althoff, “Geschichtsschreibung in einer oralen Gesellschaft,” 151–69; Keller, “Widukinds Bericht,” 406–10; and Bachrach, “Memory, Epistemology, and the Writing of Early Medieval Military History,” 64–70. 44 With regard to Widukind’s high-ranking relatives, likely including the royal family itself, see Brömel, “Widukind von Corvey,” 17.

55

N A R R AT I V E W O R K S

cally mentions in his text the service of a nobleman named Reginbern in the army of Henry I.45 Widukind notes that Reginbern was descended on his mother’s side from the famed Widukind, the author’s namesake who had fought against Charlemagne, and thus likely was a cousin of the Corvey monk.46 Second, in addition to the information gained from members of his own family about military matters, Widukind lived among a large number of men who were also from Saxon noble families and therefore had relatives who either provided military service or provided soldiers to the duke in Saxony and to the king. This group of noble monks provided a rich source of information about the military activities of their brothers, fathers, and other male relatives.47 Finally, the monastery of Corvey, as a substantial land-holder in the Saxon region, had very large military obligations, including providing a substantial contingent of well-armed mounted fighting men for service in the armies of the duke and king.48 Consequently, Widukind, who clearly had access to considerable information from both within and outside the monastery, was in a position to know the ways in which the military forces of Corvey were organized and deployed. Indeed, he may well have had conversations with the milites who served in Corvey’s household forces. However, it should be emphasized that Widukind operated within the context of an aristocratic parti pris and was writing for an audience more concerned about the leading members of society than with fighting men from the lower social and economic ranks of society. Consequently, he tends to emphasize the importance of the former at the expense of the latter. Leyser, as noted earlier, treats the Corvey monk as a reliable source and argues that Widukind, in writing about the middle years of Henry’s reign, describes the Saxon king as having “accomplished something like a military revolution in his stemland.”49 This military revolution Leyser understood to have had two

45 All citations of Widukind are to Quellen zur Geschichte der sächisischen Kaiserzeit: Widukinds Sachsengeschichte, ed. Albert Bauer and Reinhold Rau from the original edition by P. Hirsch, M. Büdinger and W. Wattenbach (Darmstadt, 1971). Widukind, Res gestae, 1.31. 46 Ibid. It is now well-recognized by scholars that writers from this period frequently used their texts to provide a memorial to their friends and relatives, and it seems likely that this is what Widukind was doing here. On this practice, see Lutz E. von Padberg, “Geschichtsschreibung und kulterelles Gedächtnis: Formen der Vergangenheitswahrnehmung in der hochmittelalterlichen Historiographie am Beispiel von Thietmar von Merseburg, Adam von Bremen und Helmold von Bosau,” Zeitschrift für Kirchengeschichte 105 (1994), 156–77, here 158 and the literature cited there. In this context, also see Gerd Althoff, Adels-und Königsfamilien im Spiegel ihrer Memorialüberlieferung: Studien zum Totengedanken der Billunger und Ottonen (Munich, 1984), particularly, 166–72. 47 Regarding the noble inhabitants of Corvey, see Brömel, “Widukind von Corvey,” 17; and Helmut Beumann, Die Ottonen (Stuttgart, 1987), 15. 48 The earliest surviving immunity for Corvey requiring the mobilization of its dependents for service in expeditione was issued in 826 by Louis the Pious. See Die Urkunden Ludwig des Frommen, 3 vols., ed. Theo Kölzer (Wiesbaden, 2016), nr. 255, with the observation by the editors that part of this charter has been interpolated. With respect to the continuing military obligation of the monastery of Corvey, see Leyser, “Saxon Empire,” 16–17. 49 Leyser, “Saxon Empire,” 14.

56

S A X O N M I L I TA RY R E V O L U T I O N , 912–973?

fundamental components. One part was ostensibly defensive and based upon the construction of numerous fortresses. The second part was offensive and based upon the development of mobile mounted forces.50 While Leyser recognized the great importance of Henry’s “castle-building” and the creation of agrarii milites, he makes clear that his subject is not the “farmer-soldiers” who were based at these fortresses, but rather chose to focus “only [on] the history of the Saxon exercitus in the early tenth century.”51 This force Leyser saw ultimately as the heavy cavalry weapon that undergirded the success of the Ottonian dynasty. Leyser maintained, in this context, that while defense was certainly not unimportant, “the pre-condition for all the opportunities open to the Liudolfing rulers and their closer adherents was . . . an effective Saxon exercitus.”52 In order to sustain this notion that Henry I built an effective exercitus for offensive military purposes, Leyser focused on Widukind’s well-known discussion of Henry’s views expressed at the fortified stronghold within the fortress at Werla (now Werlaburgdorf) located in modern Niedersachsen 4 kilometers from the city of Schladen.53 Concerning the troops who were at Werla, Widukind writes that the Saxon king saw this military force as untrained (rudus) and unaccustomed (insuetus) to bellum publicum against such savage people (saevis gens), i.e. the Magyars.54 Leyser took the situation at Werla to represent not simply the forces at this fortress where Henry could evaluate them, but rather as a characterization of the entire exercitus of the Saxon duchy as a whole.55 Leyer’s expansion of Widukind’s reference to a particular fighting force at Werla to mean the entire Saxon exercitus is problematic. First, there is nothing in Widukind’s text to suggest that the rudus and insuetus contingent at Werla 50 51 52 53

Ibid. Ibid. Leyser, “Saxon Empire,” 13. Leyser, “Saxon Empire,” 14. Widukind, Res gestae, 1.32. Whether these are, in fact, Henry’s views is problematic because the Saxon king was dead for more than a half-century when Widukind wrote. 54 Leyser, “Saxon Empire,” 14. In ibid., 33, Leyser interprets Widukind’s use of the term miles in this passage to refer to all of Henry’s milites, whom Leyser identifies as heavily armed mounted warriors who constituted the Saxon exercitus. But Leyser’s claim here is not warranted either by Widukind’s text or by the conditions at Werla. As Leyser, himself, (“Saxon Empire,” 23) observes, Widukind used the expression manus militum when he wanted to denote a body of milites. On its face, therefore, it seems unlikely that the singular miles can stand for the body of Henry’s milites. Indeed, on the few occasions that Widukind uses the term miles without referring to a specific person, the Saxon monk invariably is referring to a particular military force. Thus, for example, when Widukind wished to describe the contingent of Otto I’s army that engaged his son Liudolf’s men outside the gates of Regensburg when Otto put the city under siege in 953, the author used the term miles. See Widukind, Res gestae, 3.36. Widukind appears to use the term miles in this context as a collective noun meaning a group of high-quality fighting men. Bishop Thietmar of Merseburg (1009–1018), who drew heavily on Widukind as a source, consistently used the term miles to designate professional fighting men and eschewed using it for any other purpose. See Bachrach, “Memory, Epistemology, and the Writing of Early Medieval Military History,” 75–90. 55 Leyser, “Saxon Empire,” 14.

57

N A R R AT I V E W O R K S

was something more than a local force, that is militia troops of the general levy obligated to sustain the local defense. Second, and more importantly, in this very same chapter, Widukind makes clear that there was at least one effective Saxon military force operating against the Magyars at this time. Indeed, Widukind makes clear that this other force, i.e. not the men at Werla, had defeated a formidable Magyar army and captured one of its major leaders who then was turned over to King Henry.56 Widukind’s treatment of Henry’s subsequent use of this prestigious prisoner of war is illuminating in regard to the strength of the Saxon army in fighting against these invaders. The Magyars sent envoys to King Henry petitioning him to accept “an enormous ransom” composed of large quantities of gold and silver in exchange for the captive. The king, however, is portrayed by Widukind as rejecting the ransom and using the situation created by the victory won by his troops and the capture of the Magyar leader as an opportunity to arrange, successfully as the situation turned out, a treaty of peace (pax firmatetur).57 In this context, Widukind is portraying Henry according to the model of a good Christian monarch who prefers peace to war, as that tenet was developed during the later Roman Empire and affirmed by Carolingian political theorists.58 As a result of this treaty, Henry sent the Magyar envoys on their way with gifts (munera) and saw to it that the captured Magyar leader was set free.59 Contrary to the thrust of Leyser’s argument, Henry’s recognition that the local defense forces at Werla were not well trained was, itself, not of major importance with regard to the development of contingents of heavy cavalry. Rather, as will be seen later, it provided the stimulus for the development of well-trained local levies, i.e. the agrarii milites to defend the fortifications.60 The major series of events in 924, which serve to contextualize the information concerning Werla, were the victory won by elements of Henry’s field forces over the Magyars, the capture of an important Magyar leader, and the negotiation of a treaty of peace. The success 56 Widukind, Res gestae, 1.32. 57 Ibid. We accept the date of 924 for these events. We do so because we see Widukind’s comment on the nine-year pax as a statement of its actual duration, rather than an indication of its intended length. This point is discussed in greater detail in the appendix. 58 See particularly Sedulius Scotus, De rectoribus christianis, ed. S. Hellmann (Munich, 1906), ch. XVII, which deals with the obligations of a king to make peace when an enemy offers it. For very useful background that undermines the traditional Germanist view of Carolingian kingship see the penetrating study by J. M. Wallace-Hadrill, “The via regia of the Carolingian Age,” in Trends in Medieval Political Thought, ed. Beryl Smalley (Oxford, 1965), 22–41, and reprinted in J. M. Wallace-Hadrill, Early Medieval History, (Oxford, 1975) 181–200. See also Hubert Anton, Furstenspiegel and Herrscherethos in der Karolingerzeit (Bonn, 1968), 261–81, who discusses in detail Sedulius’s views regarding the proper behavior of a Christian king. 59 Widukind, Res gestae, I, 32. 60 For an overview of the extensive literature dealing with the agrarii milites, see Leyser, “Saxon Empire,” 15; and the more recent survey by E. J. Schoenfeld, “Anglo-Saxon ‘Burhs’ and Continental ‘Burgen’: Early Medieval Fortifications in Constitutional Perspective,” The Haskins Society Journal 6 (1994), 49–66.

58

S A X O N M I L I TA RY R E V O L U T I O N , 912–973?

of this treaty – it lasted for nine years according to Widukind – makes clear that the Magyar leaders, who had been defeated, greatly respected the strength of Henry’s army. The overall strength of Henry I’s field forces, as contrasted with militia forces responsible for local defense, should come as no surprise to readers of Widukind’s text. Indeed, it is clear that as early as 912, when Henry succeeded his father Otto as the leading figure in the Saxon region, he commanded a considerable array of military forces. In 912, King Conrad sought to limit the authority exercised by Henry, and in particular to deny him the regional dominance previously enjoyed by Otto the Illustrious, Henry’s father. However, Conrad, according to Widukind, realized that Henry commanded an exceptionally formidable military force, including a major contingent of professional soldiers ( fortium   militum manus), and also very large numbers of expeditionary levies such that he could act with the army of professionals and also with a considerable additional force (cum exercitu quoque innumera multitudine). As a result, the king concluded that he could not deal with Henry on the battlefield.61 Instead of facing Henry directly, King Conrad I used various surrogates to try to keep the new duke in his place. Henry, however, deployed the formidable military organization, noted earlier, to very good effect. For example, Henry found himself in conflict with Archbishop Hatto of Mainz, one of the great magnates of the Frankish kingdom and one of Conrad I’s major supporters. Henry outmaneuvered Hatto and confiscated all lands belonging to the archdiocese not only in the Saxon duchy but also in Thuringia.62 In the same context, Henry found it necessary to fight against King Conrad’s brother-in-law, Burchard, and the latter’s close associate Bardo.63 The king’s surrogates were defeated several times, i.e. in bellis frequentibus, and ultimately they found it necessary to surrender their holdings (terra cederent) that straddled the border between the Saxon region and Thuringia.64 In order to secure the newly acquired territory, Henry divided all of these newly acquired estates among his own milites.65 Shortly after these events, Henry decisively defeated an expeditionary force led by the king’s brother, Eberhard, in the environs of the Saxon fortress town of Eresburg on the northern flank of the Fulda Gap.66 Thus, by c. 915, not only did Henry demonstrate that he had a formidable army under his direct command, but he was able to recruit additional numbers of what would appear to be elite troops, i.e., milites, by using lands that he had taken from Archbishop Hatto, Burchard, and Bardo.67 61 62 63 64 65 66 67

Widukind, Res gestae, 1.21. Ibid., 1.22. On this point, see Leyser, “Saxon Empire,” 34. Widukind, Res gestae, 1.22; and Leyser, “Saxon Empire,” 34. Ibid. Widukind, Res gestae, 1.23. The utilization of military lands to support soldiers is not feudalism. There are a wide variety of ways to pay soldiers of which land is but one. Furthermore, the use of land or income from land for this purpose certainly is not a necessary and sufficient definition for what some people call

59

N A R R AT I V E W O R K S

With the number of defeats suffered by his surrogates increasing, King Conrad decided to take decisive action and summoned the expeditionary levies, i.e., congregata omni virtute Francorum, from all of the east Frankish kingdom for an invasion of the Saxon region.68 Henry ensconced himself in the fortress of Grone (located near modern Göttingen) to await support from his fideles. Before the king could establish a siege, however, Thietmar, Henry’s father-in-law, who commanded the Saxon forces on the eastern frontier, arrived at Grone, where Conrad’s envoys were trying to convince the Saxon ruler to make peace. Thietmar, who is characterized by Widukind as a man of very great experience in military matters (vir disciplinae militaris peritissimus) informed the envoys that he was bringing up a force of thirty units of professional troops, i.e. trigenti legiones. Although this would appear to have been some sort of ruse de guerre, both Henry’s and Conrad’s envoys would seem to have believed Thietmar, and the king’s army retreated from Saxony.69 As seen earlier, Henry, in his capacity as the leading man in the Saxon region, commanded a considerable force of his own troops, i.e. a major contingent of professional soldiers (fortium militum manus) and also very large numbers of levies (exercitus innumera multitudine). These were the forces with which Henry had defeated various of the military expeditions sent against him by King Conrad. In addition, Henry had access to the army led by Thietmar and to many other strong contingents that were commanded by his supporters, both lay and ecclesiastical, in the Saxon duchy. Among these, we may include the army commanded by Reginbern, Henry’s brother-in-law, who is credited by Widukind with destroying Danish settlements in northern Saxony and keeping the frontier safe from their periodic attacks.70 Henry’s strong force of milites would later be referred to by Widukind as the “equestri prelio probatum,” the elite troops whom he commanded at Riade in 933.71 In 919, when Henry accepted the kingship, he no longer had to rely solely on the forces that he traditionally commanded, e.g. his own milites and expeditionary

68 69

70 71

feudalism. See in this regard, Bernarch S. Bachrach, “Military Lands in Historical Perspective,” The Haskins Society Journal 9 (1997), 95–122, particularly 113–115. Widukind, Res gestae, 1.24, “congregata omni virtute Francorum.” Widukind, Res gestae, 1.24. Two points need to be made here. First, the fact that the ruse de guerre worked if, in fact such a technique was employed at this time, permits the inference that it was widely believed that Thietmar could mobilize thirty units of professional soldiers. Insofar as Widukind tells the story, it would seem that he believed it to be rhetorically plausible. Second, Widukind’s use of the term legio in this context is not at all a problem. This was a synonym used for scara by writers throughout the Carolingian and Ottonian eras to indicate a unit of professional soldiers, as contrasted with various types of levies. For the Carolingian system, see Bachrach, Early Carolingian Warfare, 81–2. Thietmar von Merseburg, Chronik, ed. and trans. W. Trillmich, 8th edition (Darmstadt, 2002), consistently used the term legio to denote a unit in an army commanded by a prince, either German or Slavic. In this regard, see Chronicon, 3.19; 4.11; 4.22; 5.36, 6.22; 6.27; 7.23; and 7.59. Widukind, Res gestae, 1.31. Widukind, Res gestae, 1.38.

60

S A X O N M I L I TA RY R E V O L U T I O N , 912–973?

levies and those of his Saxon fideles. Now, depending upon how effectively he was able to impose his royal authority on the great men of the kingdom, he would be able to mobilize the entire military establishment of the east Frankish kingdom, i.e. “he would be able to act having mobilized all the strength of the Franks” (congregata omni virtute Francorum). In this context, Widukind makes clear that in the early years of his reign, while Thietmar and Reginbern defended the eastern and northern frontiers of the Saxon region, respectively, Henry led his forces against the rulers of the southern duchies, Burchard II (917–926) of Swabia and Arnulf (907–937) of Bavaria. After subjugating both of these dukes to his authority, Henry turned his attention militarily to Charles III (898–922), the west Frankish ruler, with considerable short-term success.72 While recounting Henry’s efforts, Widukind provides considerable insight into the great military resources available to the new king and the manner in which these resources were regarded by the great men of the kingdom. For the invasion of Swabia, Henry is reported to have mobilized the entire expeditionary force available to him from the Saxon region, i.e. cum omni comitatu suo.73 Duke Buchard, who is depicted both as an exceptionally able military leader (bellator intolerabilis) and also as a very prudent man (valde prudens), recognized, according to Widukind, that he could not effectively oppose an attack by Henry. As a result, Burchard handed over, i.e. tradit, to royal command all of his fortifications (urbes) in Swabia and his entire military force (populus suus).74 By contrast, Duke Arnulf of Bavaria, who learned of Burchard’s capitulation, initially ensconced himself in the fortress city of Regensburg, but soon realized that he was not strong enough to oppose the new king. Thus, he is reported to have handed himself over to Henry in dependence along with all that he ruled.75 In light of Widukind’s presentation of the formidable array of Saxon military forces available to Henry, both as the dominant figure in the Saxon region and as king, it is hardly surprising that in 924, elements of his army decisively defeated the Magyar invaders, captured

72 Widukind, Res gestae, 1.30. For an overview of Henry I’s military campaigns, see E. Sander, “Die Heeresorganization Heinrichs I., Historisches Jahrbuch 59 (1939), 1–16; and Holtzmann, Geschichte der sächsischen Kaiserzeit, 84–96. Now also see David S. Bachrach, Warfare in TenthCentury Germany (Woodbridge, 2012), chapter 1 and passim. 73 Widukind, Res Gestae, 1.27. In this case, the comitatus clearly refers to the entire expeditionary force available to Henry, including his milites, the milites of his fideles, and the expeditionary levy from Saxony. On the difference between the milites and the men of the expeditionary levy, see Leyser, “Saxon Empire,” 23. 74 Widukind, Res Gestae, 1.27. It should be noted that urbs is the standard term used by Widukind and numerous other authors working in the German kingdom from the tenth and eleventh centuries to refer to fortified places, including fortified cities and other strongholds. On this point, see J. Ehlers, “Burgen’ bei Widukind von Corvey und Thietmar von Merseburg,” in Architektur, Struktur, Symbol: Streifzüge durch die Architekturgeschichte von Antike bis zur Gegenwart, Festschrift für Cord Meckseper zum 65. Geburtstag, ed. Maike Kozok (Petersberg, 1999), 27–32. 75 Widukind, Res Gestae, 1.27, ‘tradito semet ipso cum omni regno suo.’ Nb. regnum cannot be considered kingdom here but merely recognition of a placed that is ruled.

61

N A R R AT I V E W O R K S

one of their most important leaders, and imposed a peace that the Hungarians were afraid to break for nine years.

Saxons as mediocre soldiers? Obviously, Leyser did not give sufficient attention to Widukind’s information, described briefly earlier, regarding the formidable army Henry inherited when he succeeded his father in 912 and which he improved. Rather, Leyser mobilized several other accounts regarding the operation of Saxon military forces in the field during the ninth century, i.e. more than two decades prior to Henry’s accession to his father’s office, in order to argue that “the military standing of the Saxons in the ninth century was not particularly high.”76 Although he never explicitly makes clear exactly what this comparative statement means, Leyser adduces two arguments in its support: first, continental Saxon fighting forces were presented by ninth- and early tenth-century authors in a negative light, and second that “[w] ith one exception, Lothar’s war against his brothers in 841, they [the Saxons] played a subordinate role in the armed clashes of the contenders for possession of power in the Carolingian Reich.”77 In support of the first of these two arguments, Leyser deploys three pieces of evidence. Leyser insists first that Einhard (c. 820) “looked down” on the Saxons as auxiliaries. Leyser then attributes this same sense of “looking down” to the author of a poetic panegyric on Louis the Pious, Ermold (c. 826/28).78 Finally, Leyser states that Meginhard, to whom Leyser, following Friedrich Kurze, attributes the authorship of the Annales Fuldenses after 863, presents a “decidedly unflattering” image of the Saxons who participated in the Moravian expedition of 872.79 Because Leyser attributes such great weight to a total of three passages from the works of these ninth-century authors to justify his conclusion that the Saxons were depicted as having a military standing that was “not particularly high” in the ninth century, it is necessary to consider these particular texts in some detail. According to Leyser, Einhard’s comments regarding Charlemagne’s deployment of Saxon forces in 789 – an eighth- rather than a ninth-century campaign – permit the interpretation that the imperial biographer wanted to show the Saxons as “treacherous auxiliaries.”80

76 Leyser, “Saxon Empire,” 31. 77 Ibid. 78 Ermold, Poème sur Louis le Pieux et épitres au Roi Pépin, ed. and translated by Edmond Faral (Paris, 1932). 79 Leyser, “Saxon Empire,” 31. For a full discussion of the question of the authorship of the Annals of Fulda, see The Annals of Fulda, translated and annotated by Timothy Reuter (Manchester, 1992), 1–9. For the Latin text of the Annales Fuldenses, F. Kurze MGH SRG (Hanover, 1891). We follow the practice of Leyser and Kurze in attributing the authorship of the Annales Fuldenses to Meginhard. 80 Leyser, “Saxon Empire,” 26

62

S A X O N M I L I TA RY R E V O L U T I O N , 912–973?

In the passage in question, Einhard wrote: “during which time the Saxons fought just like the troops from the other nations that followed the royal banner when commanded to do so. [The Saxons], however, acted in deceit with less than full devotion.”81 A careful reading of this text makes clear that Einhard regards the Saxons as having served unwillingly in Charlemagne’s army – a view more than justified by the further Saxon revolts against Carolingian rule in the period after 789.82 Indeed, Einhard’s work as a whole indicates his deep hostility to the Saxons.83 But there is nothing here to suggest that Einhard “looked down” on the Saxons for serving as auxiliares. Rather, Einhard is simply observing that in 789, at least, the Saxons were now numbered among the “other nations” subject to Carolingian rule and were performing the duty incumbent upon a subject people, namely providing military service to Charlemagne. Leyser’s use of the English word “auxiliaries” for the Latin cognate auxiliares is somewhat misleading because Einhard clearly is using the term in a political rather than in the tactical sense of troops whose task it is to aid the main battle force. However, even if Leyser were correct to see Einhard “looking down” upon the Saxon troops serving as auxiliares, it would then be necessary to argue that Einhard “looked down” on the ceterae nationes as well. Of course, such an interpretation might well require that Einhard be construed as a Frankish “nationalist,” a factor that would call into question any negative comments he might make about all nationes other than the Franks. Leyser’s identification of Ermold’s very brief discussion of the deployment of a unit of Saxons by Louis the Pious in Brittany as an example of “looking down” on these men as “auxiliaries” also is problematic after an examination of the text.84 Ermold makes no comment on the fighting ability of the Saxons. He does not present them in combat. He does not even discuss the tactics in which they were trained. Rather, Ermold lists the Saxons, much as Einhard did regarding the campaign undertaken by Charlemagne in 789, as one of the subject peoples of the empire required to send troops to serve in expeditione.85 Ermold, like Einhard, is commenting on the political subordination of the Saxons, as well as, in this passage, the subordination of the Alemanni and Thuringians, to Frankish rule. There is no information in this text that can be used to show that Ermold presented the Saxons, Alemanni, or Thuringians as inferior to the Franks in a military sense. 81 Einhard, Vita Karoli, 4th revised ed. L. Halphen (Paris, 1967), 2.12, ‘In quo et Saxones velut auxiliares inter ceteras nationes, quae regis signa iussae sequebantur, quamquam ficta et et minus devota oboedientia, militabant.’ This edition is to be preferred to the edition by O. Holder-Egger MGH SRG (Hanover, 1911). 82 On the question of the Saxon revolts, see Eric Goldberg, ‘Popular Revolt, Dynastic Politics and Aristocratic Factionalism in the Early Middle Ages: The Saxon ‘Stellinga’ Reconsidered,’ Speculum 70 (1995), 467–501, particularly 475–8, where he discusses the background of revolts during the 780s and 790s. 83 On this point, see Einhard, Vita Karoli, 2. 6 and 2.7. 84 Ermold, Poème, lines 1510–1516. 85 Ibid., ‘Caesar Francos et gentesque subactacs/ esse jubet placito.’

63

N A R R AT I V E W O R K S

In turning to the Annales Fuldenses, and particularly the portion attributed to Meginhard, it must be emphasized, although Leyser neglects to mention it, that Meginhard had a strong anti-Saxon parti pris which is to be taken into account when evaluating his depiction of the Saxons.86 In the context of discussing military operations conducted in east Francia in 872, Meginhard writes that Louis the German sent an army composed of Thuringians and Saxons against the Moravians.87 The author stressed that Louis the German did not accompany the army, and, as a result, the Saxon and Thuringian contingents could not get along with each other, causing them to break off the campaign without ever engaging the enemy. Some of the counts leading the army – Meginhard does not specify whether they were Thuringian or Saxon – then were beaten by women from the region and knocked from their horses.88 Clearly, in this instance, Meginhard is presenting both the Saxons and the Thuringians as fractious and has some fun at the expense of their leaders, suggesting that even when on horseback they are weaker than women. Meginhard’s depiction here of the Saxons certainly is consistent with his anti-Saxon bias. It is to be noted, however, that Meginhard does not show the Saxons losing in battle against the Moravians, running from battle, or being deficient in any particular military sense. What is at issue is, in fact, their level of “civilization.” Nevertheless, if this were the only passage in which Meginhard discussed the Saxons engaged in war, one might conclude that he had a low opinion of their military abilities. However, although Leyser does not note this, Meginhard discusses Saxon military forces on several other occasions, depicting them as victorious in battle. In 869, for example, Louis the German sent an army of Saxons and Thuringians, under the command of Louis the Younger, against the Sorbs. According to Meginhard, Louis the Younger was very successful in this campaign, defeating not only the Sorbs but their Bohemian “mercenaries” as well.89 In 885, Saxon defense forces, aided by a small Frisian fleet, crushed a major contingent of Northmen and, according to Meginhard, killed them almost to the last man.90 The three pieces of evidence that Leyser cites to show that the Saxons were depicted as having a military standing in the ninth century that “was not particularly high” do not, in fact, show this. Indeed, in the case of Meginhard, it is clear that despite his anti-Saxon parti pris, the author of the Annales Fuldenses felt 86 The author of the Bavarian continuation of the Annals of Fulda (see Reuter, Annals, 5–9) similarly displays his hostile views of the Saxons in his discussion of the deposition of Charles the Fat in 887. Following Charles’s illness, according to the author, the Franks, and the Saxons, the latter “in their usual fashion,” entered into an “evil conspiracy” to depose the emperor. See Annales Fuldenses, an. 887, “male inito consilio Franci et more solito Saxones . . . cogitaverunt deficere a fidelitate imperatoris (emphasis added).” 87 Annales Fuldenses, an. 872. 88 Ibid. 89 Annales Fuldenses, an. 869. Cf. Annales Bertiniani, ann. 869. For the Latin text of the Annals of St. Bertin, see Annales de Saint-Bertin, ed. F. Grat, et al. (Paris, 1964). 90 Annales Fuldenses, an. 885.

64

S A X O N M I L I TA RY R E V O L U T I O N , 912–973?

compelled to present Saxon military forces as triumphant in battle against both the Slavs and the Northmen. But even if Leyser’s three pieces of evidence all pointed toward a low military reputation for the Saxons in the ninth century, this would not have been sufficient to make his case, because Leyser omits completely from his discussion of Saxon military ability the two works that have the most to say about the fighting forces from the Saxon region in the ninth century, namely the Annales regni Francorum and the Annales Sancti Bertiniani.

Saxon military effectiveness The author of the Annales regni Francorum discusses the deployment of Saxon military forces by Charlemagne and Louis the Pious on eleven occasions between 802 and 828.91 Indeed, the Annales regni Francorum consistently presents the military forces of the Saxons as an important component of the overall Carolingian military organization with particular responsibilities in the northeast against the Slavs and the Danes. In 809, for example, it is reported that Saxon forces successfully supported the Obodrite duke Thrasco, a Carolingian ally, in his campaign against the Wilzi, a Slavic polity hostile to the Carolingians.92 In 815, Louis the Pious dispatched an army of Saxons and Obodrites to help the Danish prince Heriold reacquire political control in his kingdom.93 The next year, Louis sent contingents from Saxony and Franconia to campaign against the Sorbs.94 In 819, Louis sent an army of Saxons and East Franks to put down a rebellion by the Obodrite duke Sclaomir.95 In addition to undertaking campaigns into enemy territory, Saxon forces were deployed by both Charlemagne and Louis the Pious to build and garrison frontier fortifications. In 809, Charlemagne ordered several Saxon counts to establish a fortification on the river Stör, on the Danish side of the Elbe river at a placed called Esesfelth, and to provide a garrison there to inhibit Danes from raiding into Carolinian territory.96 Similarly, in 822, Louis the Pious sent a Saxon army east of the Elbe to drive out the Slavic defenders in the area of Delbende on the Delvenau river at Lauenburg and to build a fortress there. According to the author of the Annales regni Francorum, Louis intended this fortification and its garrison to inhibit Slavic raids west of the Elbe in this region.97 Both Prudentius of Troyes and Hincmar of Rheims, as authors of the Annales Bertiniani, also present Saxon military forces as integral to Carolingian military operations in the east and northeast, together recording fifteen engagements in which 91 Military forces mobilized in Saxony are depicted going to war in 802, 808, 809 (2), 810, 815, 816, 817, 819, 820, 822, and 828. See Annales regni Francorum, ed. F. Kurze MGH SRG (Hanover, 1895). 92 Annales regni Francorum an. 809. 93 Annales regni Francorum an. 815. 94 Annales regni Francorum an. 816. 95 Annales regni Francorum an. 819. 96 Annales regni Francorum an. 809. 97 Annales regni Francorum an. 822.

65

N A R R AT I V E W O R K S

Saxons took part during the period 831–880.98 In 839, for example, while Louis the Pious was trying to put down a revolt by the Aquitanians, the Saxons are reported by Prudentius to have fought and won a major battle against the Sorbs at Kesigesburg. During the course of the fighting, the Saxons killed the Sorbian king Czimislav. In the aftermath of the battle, the Saxons captured eleven Sorbian strongholds.99 Hincmar, corroborating the account in the Annales Fuldenses, reports that in 869, Louis the Younger led a Saxon army to a major victory over the Slavs.100 Even those Carolingian authors not particularly interested in the Saxons, such as the anonymous biographer of Louis the Pious, generally referred to as the Astronomer, and Nithard, Charlemagne’s grandson who served as a military officer in the army of King Charles the Bald, present Saxon forces as being of crucial importance to Carolingian military operations against the Slavs and, to a lesser extent, against the Northmen. Echoing the Annales regni Francorum, the Astronomer reports that in 816, an army of Saxons and East Franks was sent to deal with the rebellious Sorbs and that the rebellious Abodrite leader Sclaomir was captured by Saxon military forces and handed over to Emperor Louis at Aachen in 819.101 He added that in 839, Louis the German relied entirely on a force of Saxons and Thuringians to invade Alemannia.102 Nithard confirms Louis the German’s reliance on the Saxons in 839 and then goes on to make clear the efforts of both Lothair I and Louis the German to gain the loyalty of Saxon troops during the civil wars of the next three years.103 In September 841, Lothair managed to recruit a substantial Saxon force to serve with him during his campaign against Charles the Bald in the Paris region.104 By early 842, however, Louis the German had expended considerable effort to bring the Saxons back to his side and had them with him in February of that year when he met with Charles at Worms.105 Nithard makes clear in his discussion of the meeting between the two brothers that their troops were well-trained in battle tactics, including mounted attack, and specifically records their demonstration of the tactic of the feigned retreat in war games conducted by the two Carolingian

98 Concerning the authorship of the Annals of St. Bertin, see The Annals of St. Bertin, translated and annotated by J. Nelson (Manchester, 1991), 6–13. The deployment of military forces is presented by Prudentius and Hincmar as taking place in 831, 832, 834, 839(4), 845, 858, 863, 867, 869, 876, and 880. 99 Annales Bertiniani, an. 839. 100 Hincmar records that the Slavs were Wends, and Meginhard says they were Sorbs. See Annales Bertiniani an. 869. 101 Astronomer, Vita Hluodowici imperatoris, ed. Ernst Tremp MGH SS rer. Germ. 64 (Hanover, 1995), 362 and 391. 102 Astronomer, Vita Hluodowici, 540. 103 Nithard, Historiae, 1.8 in Quellen zur karolingischen Reichsgeschichte, ed. R. Rau vol. 1, 2nd edit. (Berlin, 1955). Goldberg, ‘Popular Revolt,” 467–501, 104 Nithard, Historiae, 3.3. 105 Nithard, Historiae, 3.6.

66

S A X O N M I L I TA RY R E V O L U T I O N , 912–973?

kings.106 Among the soldiers trained in these tactics, Nithard specifically mentions the Saxons along with the Bretons and Gascons.107 Clearly, the claim that “the military standing of the Saxons in the ninth century was not particularly high” is not tenable on the basis of the accounts provided in ninth- and tenth-century narrative accounts. But what of Leyser’s secondary argument that “with one exception, Lothair’s wars against his brothers in 841, they [the Saxons] played a subordinate role in the armed clashes of the contenders for possession and power in the Carolingian Reich”? To support this broad-gauged claim, Leyser points to an account by Regino of Prüm dealing with Charles the Bald’s deployment of a unit of Saxons in Brittany in 851 and a report in the Annales Fuldenses of the battle of Andernach (876).108 Leyser asserts that in 851 Charles employed the Saxons as “unskilled military labour, not to say cannon-fodder” and that “[a]t Andernach, Louis the Younger did not treat them any better and counted on his Rhenish Franks to see him through.”109 In his account of Charles the Bald’s military campaign in Brittany in 851, Regino states that Charles deployed a unit of Saxon mercenaries (conducti) in front of his lines to blunt (excipere) the charge of the rapidly advancing Breton mounted troops.110 In tactical terms, it is clear that Charles deployed these Saxon mercenaries as skirmishers, a standard type of military deployment used to screen the main force from attack while it deploys for battle. Indeed, Charles’s army was in the process of deploying into a phalanx, which required that his mounted troops dismount and integrate with the foot soldiers. The latter also had to establish a safe place for the horses as their riders joined the battle line. Regino then records that at the first sight of the spears of the Bretons, the Saxons grew frightened (territi) and withdrew behind the line of Charles’s army.111 Regino’s characterization of the Saxon force as “terrified” notwithstanding, the withdrawal of skirmishers behind the line once the main force has completed its deployment was normal practice for skirmishers in 851, as it was in the ancient world, and as it has remained into modern times.112 106 107 108 109 110

Ibid. See the discussion of this text by Bachrach, Early Carolingian Warfare, 124–30. Nithard, Historiae, 3.6. Leyser, “Saxon Empire,” 26. Ibid. Regionis abbatis Prumiensis Chronica, ed. F. Kurze MGH SRG (Hanover, 1890), an. 851, ‘Saxones, qui conducti fuerunt, ad excipiendos velocium equorum anfractuosos recursus in prima fronte ponuntur.’ 111 Ibid., ‘sed primo impetus spiculis Brittonum territi in acie se recondunt.’ 112 The use of skirmishers (velites) is discussed thoroughly by Flavius Renatus Vegetius, Epitoma rei militaris, ed. Alf Önnerfors (Stuttgart, 1995), 3.16 and 3.20. See also the English translaton by N. P. Milner, Vegetius: Epitome of Military Science 2nd revised edition (Liverpool, 1996. For a general survey of the importance of Vegetius’s work in the early Middle Ages, see Bernard S. Bachrach, “The Practical Use of Vegetius’ De re militari during the Early Middle Ages,” The Historian 47 (1985), 239–255, reprinted with the same pagination in B. S. Bachrach, Warfare and Military Organization in Pre-Crusade Europe (London, 2002). With regard to the use of Vegetius

67

N A R R AT I V E W O R K S

Regino’s account makes clear that the skirmishers did carry out their task of providing the main force time to create a solid front. By the time the Bretons reached the Frankish ranks, the latter, Regino notes, were concerta in acie and the Saxons had done their job.113 That the soldiers in the Frankish phalanx subsequently broke ranks in an undisciplined manner to pursue the Bretons as the latter carried out a feigned retreat, suffering significant casualties in the process, can hardly be attributed to any failure of the Saxons.114 Regino himself makes clear that the Frankish casualties were not to be blamed on the Saxons, but rather insists that the Franks, who were accustomed to fighting with their swords while deployed in a phalanx, were attoniti because they were not trained or equipped to pursue the Bretons, nor adequately led to reform their phalanx once they broke ranks to pursue the apparently retreating Breton horsemen.115 Of more importance in the present context, skirmishers are not “cannon-fodder,” to use Leyser’s term, but rather an integral element of an army’s force with specific and essential tactical duties. If this passage has any bearing on the relative military standing of Saxons and Franks, Regino’s comments can only be interpreted to the detriment of the latter.116

113 114

115 116

by military leaders in east Francia, see E. Goldberg, Struggle for Empire: Kingship and Conflict under Louis the German, 817–876 (Ithaca, NY, 2006), 40–42. The particular importance of Vegetius’s text is highlighted in regard to the late Carolingian period by the work of Rabanus Maurus, who c. 856 wrote a military handbook based on De re militari for King Lothar II (d. 869). Rabanus emphasized that he was only including material that was value tempore moderno. See Rabanus Maurus, De procinctu Romanae militiae, Ernst Dümmler Zeitschrift für deutsches Altertum 15 ((1872), 413–451, 450 for this passage. Separately, a copy of Vegetius’s De re militari was commissioned by Empress Judith for the education of Charles the Bald. It can therefore be suggested with some confidence that Charles the Bald was familiar with the concept of skirmishers through Vegetius and that he knew, pace Leyser, that skirmishers could play an integral role in the conduct of battles. In this context, attention should also be paid to recent study by Guy Halsall, Warfare and Society in the Barbarian West, 450–900 (London, 2003), which has attracted some attention for its neo-primitivist model of early medieval warfare. As both the notes and bibliography provided by Halsall make clear, he is woefully ignorant of the state of the question regarding the use of Vegetius’s work in the Carolingian era. He claims (145) that the handbook written by Rabanus Maurus, archbishop of Mainz under King Lothair and based in large part on his editing of Vegetius’ De re militari, specifically for modern use, as noted earlier, “was made in a monastery . . . and we cannot be sure . . . that it was not mainly a matter of antiquarianism.” Halsall also fails to deal with the edition of Vegetius’ text by Freculf that was commissioned at the West Frankish court for use by Charles the Bald in his wars against the Vikings. Regino, Chronica, an. 851. Regino states (ibid.) that “the Bretons raced up and down on their horses in their customary manner attacking the tightly packed Frankish ranks, hurling their spears at them with all of their strength. They then carried out a feigned retreat, and thrust their spears into the chests of the men pursuing them.” Brittones more solito huc illucque cum equis ad huiusmodi conflictum exercitatis discurantes modo consertam Francorum aciem impetunt ac totis viribus in medio spicula torquent, nunc fugam simulantes insequentium nihilominus pectoribus spicula figunt. Ibid., “Franci . . . attoniti stabant . . nec ad insequendum idonei nec in unum conglobati tuti.” Karl Leyser misrepresents these Saxon mercenaries on two counts. First, it is not true that they fought poorly. Second, it is an open question whether these Saxon mercenaries serving in Charles

68

S A X O N M I L I TA RY R E V O L U T I O N , 912–973?

In a similar vein, Leyser’s view of Louis the Younger’s deployment of his Saxon troops at Andernach again betrays a fundamental misunderstanding of the tactical situation and a readiness to dismiss the parti pris of Meginhard in order to make the case for supposed Saxon military inferiority during the ninth century. The tactical situation at Andernach was exceptionally difficult for Louis the Younger. Charles the Bald was able to get his army within a short distance of Louis’s camp before the latter received intelligence of his approach. Most of Louis’s force was foraging and could not be summoned back in time to participate in the battle. Because Charles was too close to permit retreat, Louis decided to set a trap for his uncle. The bait in the trap consisted of Louis’s Saxon troops whom he deployed on foot in front of his fortified camp. The jaws of the trap were Louis’s mounted forces, who were still in camp, whom he deployed in hiding on the two flanks. When Charles arrived at Louis’s position, he saw the Saxon foot and immediately ordered a charge. The Saxons at this point received the charge and then engaged in a fighting withdrawal toward their fortified camp, thereby leading Charles’s force deeper into the trap by “refusing” the center and thereby allowing the eastern Franks to carry out a double envelopment. Once Louis saw that Charles’s men were fully into the trap and occupied with the Saxon foot, he ordered the mounted troops on the two flanks to carry out the previously mentioned double envelopment of Charles’s army, utterly surprising and then slaughtering the Western troops. Meginhard, true to his anti-Saxon bias, depicts the Saxon troops who participated in the battle as territi, and Leyser accepts this characterization as plain text and therefore as an accurate description of these fighting men.117 But the true nature of the situation is made clear by Meginhard’s statement that parumper terga verterunt, that is “they [the Saxons] retreated for a short time.”118 The crucial word here is parumper, because it makes clear that the Saxons were not routed. To retreat “for a short time” is not consistent with turning one’s back, throwing down one’s arms, and running from the field. Although he wishes to

the Bald’s army in Brittany were, in fact, recruited from more than 1,000 kilometers to the east, i.e. beyond the Weser. It is also possible that the Saxons were the descendents of military colonists in the Bayeux region who had served in the armies of various Merovingian kings in operations undertaken in Brittany. On this point, see Ferdinand Lot, “Migrations Saxonnes en Gaul et en Grande-Bretagne du IIIe au Ve siècle,” Revue Historique 119 (1915), 1–40 and reprinted in Ferdinand Lot, Recueil des travaux historiques de Ferdinand Lot, 3 vols. (Geneva-Paris, 1968–1973), II: 23–62, at 44–5, for the ninth-century Saxon mercenaries. Lot’s observations notwithstanding, we have already shown a significant number of occasions on which Saxons troops acquitted themselves well. 117 Annales Fuldenses, an. 876; and Leyser, “Saxon Empire,” 26. 118 Annales Fuldenses, an. 876. The Latin expression terga vertere in a military context does not mean to turn one’s back, but rather to retreat. The difference in meaning is crucial in a military context, since soldiers who have turned their backs cannot fight, while soldiers facing toward the enemy can, as in this case, carry out a fighting withdrawal.

69

N A R R AT I V E W O R K S

depict the Saxons in a negative light, Meginhard is still compelled by the constraints of rhetorical plausibility to show that the Saxons engaged in a fighting withdrawal, which is one of the most tactically difficult maneuvers in battle. That Louis trusted his Saxon troops to carry out this maneuver indicates a very high level of confidence in them. Far from depicting the Saxons as inferior troops used badly, as Leyser would have it, Meginhard, despite his hostility to them, makes clear in his account of the battle of Andernach that these Saxons were well trained and highly reliable. Rather more subtle is Leyser’s next argument that following the defeat and death of the Saxon leader Brun at the hands of the Danes in 880, the Saxons were only able to defeat the Danes with Frisian and East Frankish help.119 Leyser obscures the fact that the extant narrative sources do not record the Saxons as being defeated by Northmen in the period after 880. Indeed, as noted earlier, the Saxons are reported by the Saxon-hostile author Meginhard to have gained a major triumph over the Northmen in 885.120 Moreover, as Leyser himself points out, in 891 Arnulf (d. 899) successfully deployed Saxons as well as East Franks against an entrenched force of Northmen at the battle of the Dyle, a campaign in which the Alemanni, long part of the Frankish kingdom, disgraced themselves by abandoning the royal army before the battle.121 At a more fundamental level, Leyser fails to explain why the deployment of regionally diverse military forces by Carolingian rulers is ipso facto proof of the weakness of one of these units but not of the others. As shown earlier, in the period before 880, Carolingian rulers routinely deployed mixed regional forces against both the Slavs and the Northmen, drawing on Thuringians, Saxons, East Franks, and the Slavic Obodrites. Far from portraying a low military standing, a thorough reading of the relevant narrative sources for the ninth century makes clear that the Saxon region provided Carolingian rulers, from Charlemagne to Arnulf, with substantial and effective military forces for both offensive campaigns and for the defense of the frontiers. There is no question that Saxon military forces occasionally suffered defeats, even substantial defeats, at the hands of their enemies. The same, however, was true of the Frankish military forces deployed by the Carolingians. Frankish forces, for example, even suffered major defeats at the hands of the Saxons themselves, as at the famous battle of the Süntel mountains in 782 when a force of heavily armed mounted Franks was butchered by Saxon soldiers deployed on foot.122 Indeed, it should not be forgotten in the present context that Charlemagne spent more than three decades in his effort to subdue the Saxons. What is more important is the fact that for a period of more than a century prior to the accession of Henry I, Carolingian rulers relied on soldiers drawn from the Saxon region to fight on their own

119 120 121 122

Leyser, “Saxon Empire,” 31 Annales Fuldenses, nn. 885. Leyser, “Saxon Empire,” 29; and Annales Fuldenses, nn. 891. Annales regni Francorum an. 782.

70

S A X O N M I L I TA RY R E V O L U T I O N , 912–973?

and in conjunction with other regionally based military units against a broad range of enemies and that these Saxon forces frequently were successful.

Early medieval military organization and the Saxon question Being integrated into the Carolingian military system entailed the organization of Saxon military resources into the basic tripartite structure that dominated the early medieval West on the continent and later in Anglo-Saxon England.123 In this context, it should be emphasized that Anglo-Saxon England benefited from military instruction gained through military advisors recruited in Saxony.124 The

123 Both Saxony and Anglo-Saxon England were, in military terms, Carolingian successor states. Concerning the Carolingian nature of Anglo-Saxon government, see two important articles by James Campbell: “Observations on English Government from the Tenth to the Twelfth Century,” Transactions of the Royal Historical Society, fifth Series, 25 (1975), 39–54; and “The Significance of the Anglo-Norman State in the Administrative History of Western Europe,” in Histoire comparée de l’administration (IVe-XXVIIe siècles, ed. Werner Paravicini et Karl Ferdinand Werner (Munich, 1980), 117–134. Both of these studies have been reprinted with additional notes in James Campbell, Essays in Anglo-Saxon History (London-Ronceverte, 1986). In this context, it is of no little importance that Asser’s biography of Alfred owed much to Einhard’s life of Charlemagne. See M. Schutt, “The literary form of Asser’s Vita Alfredi,” English Historical Review (1957), 209–220, and the observations by Matthew Innis, “The classical tradition in the Carolingian Rennaisance: Ninth-century encounters with Suetonius,” International Journal of Classical Tradition, 3 (1997), 265–282, with regard to the importance of Suetonius to the writing of biography in early medieval England. It also has been noted that the Anglo-Saxon view of Christendom was borrowed from that of Charlemagne’s court. On this see Judith Herrin, The Formation of Christendom (Princeton, 1987), 8. Concerning the constitutional structure of the basic institutions of military organization in the Frankish kingdoms, see Bernard S. Bachrach and Charles R. Bowlus, “Heerwesen,” in Reallexikon der germanischen Altertumskunde, ed. H. Beck et al. (Berlin-New York, 2000), 14, and 122–136; Bachrach, Early Carolingian Warfare, 51–83, and 287–312; and Timothy Reuter, “The Recruitment of armies in the Early Middle Ages: What Can We Know?” in Military Aspects of Scandinavian Society in a European Perspective, AD 1000–1300, ed. A. Norgard Jorgensen and B. L. Clausen (Copenhagen, 1997), 25–31. Concerning Anglo-Saxon England, see C. W. Hollister, Anglo-Saxon Military Institutions on the Eve of the Norman Conquest (Oxford, 1962). The institutional reconstruction by Richard Abels, Lordship and Military Organizaiton in AngloSaxon England (Berkeley, 1988), has been rejected here because it relies on the unsound notion of the Königsfreien. For a thorough-going critique of Abels, see Schoenfeld, “Anglo-Saxon Burhs,” 49–66. Regarding the failure of the arguments in favor of the Königsfreien, see Timothy Reuter, “The End of Carolingian Military Expansion,” 395, with the literature cited there. See also the review of Abel’s book by Bernard S. Bachrach in Speculum 66 (1991), 109–111. Now also see Bernard S. Bachrach, “Charlemagne’s Expeditionary Levy: Observations Regarding liberi homines,” Studies in Medieval and Renaissance History 3rd series 12 (2015), 1–65. 124 King Alfred of Wessex, for example, gave preferment to a cleric named John, who had come to England from Saxony and was widely recognized by his contemporaries as a man who was very experienced in the “military arts”. The source for John’s wide reputation in military matters is Asser, De rebus gestis Aelfredi, ch. 27 (Asser’s Life of King Alfred), ed. William Henry Stevenson (Oxford, 1959).

71

N A R R AT I V E W O R K S

obsequium of the king, along with the military households of his fideles, were troops of the highest military proficiency. The king’s obsequium, of course, was the largest of all such military households, but these groups collectively constituted the smallest of the three segments of the early medieval military. Both lay and ecclesiastical magnates were required by the king, who initially acquired royal power in the wake of the dissolution of the Roman Empire, to support troops on the basis of their wealth, and the king mobilized these contingents as he needed them. The men who served in these households, of course, were the professional soldiers who had, in effect, replaced the Roman army in the West.125 When Widukind discusses Henry’s very powerful manus militum, he is referring to Henry’s obsequium. The second level of military organization was composed of expeditionary levies, i.e. men who had attained specific levels of landed and/or moveable wealth and who, on the basis of their possessions, were required to do military service either in person and/or find a substitute. If these men or women possessed more than minimum amounts of wealth, they were required to provide additional fighting men for expeditionary purposes consistent with their assets.126 In regard to Anglo-Saxon England, scholars have come to identify these levies as the “select fyrd,” as, in general, each man who possessed five hides of land was required to serve personally or to provide a substitute.127 In the Frankish kingdom, militia troops, i.e., nonprofessionals, raised from this segment of the population provided the bulk of the manpower for the large royal armies that undertook the sieges of fortress cities such as Bourges, Pavia, and Barcelona under the Carolingians.128 When Widukind called attention to the very large army (innumera exercitus) commanded by Henry I, he was referring to his expeditionary forces of the select levy, which likely were reinforced by both the king’s milites and the military households of his fideles. The many Saxon units that served various Carolingian kings were composed of such a combination of forces, as discussed earlier, as well.129

125 See Bachrach, Early Carolingian Warfare, 59–82. Reuter, “The Recruitment of Armies,” 33, gives substantial independent attention to mercenaries, rather than seeing them from an institutional perspective as a temporary extension of a royal or magnate obseqium. J. France, “The Composition and Raising of the Armies of Charlemagne,” Journal of Medieval Military History 1 (2002), 66–68, criticizes Reuter for failing to recognize the important offensive role played by what we have called here “select levies” in the armies of the Carolingians. 126 Bachrach, Early Carolingian Warfare, 54–57, with the literature cited on 289–294. In his section dealing with “conscripts” Reuter, “The Recruitment of Armies,” 34–35, is reluctant to make a sufficiently clear distinction between locally raised militia forces used for expeditionary purposes and local militia forces used only for local defense, which are discussed below. 127 Hollister, Anglo-Saxon Military Institutions, 38–102, where comparisons are made with the Carolingian system and in addition note is taken of those who have resources of fewer than five hides. 128 B. Bachrach, Early Carolingian Warfare, 202–42, and 365–76 for the notes; and D. Bachrach, Warfare in Tenth-Century Germany, 77–85 for expeditionary levies in Ottonian Germany. 129 See D. Bachrach, Warfare in Tenth-Century Germany, passim.

72

S A X O N M I L I TA RY R E V O L U T I O N , 912–973?

The third element of the military organization was the local levy, called the “great fyrd” by specialists in Anglo-Saxon military history. These forces were also a militia and were composed of all able-bodied men who lived in a particular locality, regardless of their legal status. These men were required only to fight in the defense of the locality in which they lived. Such forces were the least well trained and least well-armed of the three elements of early medieval military organization. They fought best from behind fortress walls.130 When drawn into open battle (bellum publicum) against well-trained and experienced fighting men, e.g. Vikings and Magyars, their effectiveness was greatly diminished due to their lack of training. With regard to the lack of capacity of such local levies to do battle in the open field, the chronicler Regino of Prüm, who died three years after Henry assumed his father’s office in 912, explains the situation with exceptional perceptiveness. Regino takes note of a Viking raiding party that was plundering in the region of Prüm and indicates that a force of local men who lacked armor (inermes) and horses, i.e. they were pedites, was mobilized from the nearby fields and villages (ex agris et villis) into a single unit (unum agmen) to engage the Vikings. When this local levy, which Regino admits was very large (innumera multitudo), advanced against the Vikings, the latter, although greatly outnumbered, were not frightened into retreat. Rather, they understood that because these levies were an ignobilis vulgus and, as a result, untrained, i.e. disciplina militari nudatum, they did not constitute a serious threat. Thus, the forces engaged, and the slaughter of the levies followed.131 The Saxon duchy of the later ninth century and early tenth century, i.e. before the accession of Henry I as duke in 912, possessed this tripartite military organization.132 Moreover, as seen earlier, both the obsequia of the magnates and the expeditionary levy were very effective against a variety of enemies, including forces from both within and from outside of the Frankish kingdom. The weak link in the military organization in the Saxon region, as well as in other parts of the east Francia as bemoaned by Regino, noted earlier, were the local levies. When the Magyars attacked Saxony in 924, Widukind makes clear, likely with some exaggeration, that they moved through the countryside plundering and burning the settled areas around urbes and oppida. He goes on to emphasize, again with some exaggeration, that many monasteries were destroyed and so many people were killed that the Magyars threatened to devastate the land.133 It is in this context that 130 Hollister, Anglo-Saxon Military Institutions, 25–37; Bachrach, Early Carolingian Warfare, 52–54, and 288–289 for the notes; and D. Bachrach, Warfare in Tenth-Century Germany, particularly 71–77. 131 Regino, Chronica, an., 882. 132 D. Bachrach, Warfare in Tenth-Century Germany, passim. 133 Widukind, Res gestae, 1.32. In this context, there is no reason to believe that these lightly armed Magyar horsemen actually laid siege to the urbes and oppida. Rather, their operations must be seen to have been against undefended homes and other assets in the countryside, e.g. the monasteries specifically mentioned by Widukind. It was not until the campaign that ended in the battle

73

N A R R AT I V E W O R K S

Widukind emphasizes that Henry’s forces that were supposed to be the first line of defense, that is the local levies, were insufficiently trained, i.e., that they were rudus and insuetus, to engage the Magyars in bellum publicum. As a result, they were not able to protect the territory around the fortifications (urbes and oppida), mentioned earlier. By contrast, Henry’s field forces in 924, as earlier, not only were up to the task of dealing with the Magyars in open battle but had defeated one enemy army and captured a prominent leader during this invasion in 924. In short, the local levies had failed, not the Saxon field army. King Henry recognized that the Saxon local levies, as illustrated by the rudus and insuetus troops from the area of the stronghold of Werla, mentioned previously, were unable to provide an adequate initial defense against the Magyars, much in the same way as the local men from the Prüm area had failed against the Vikings. This was because they were untrained. Consequently, Henry decided to reform this third element of the Saxon military organization. Indeed, following his discussion of the levies at Werla and immediately upon resuming discussion of Saxon military matters, Widukind takes up the question of Henry’s reform, i.e. the establishment of the agrarii milites, as the logical result of the inferiority of the rudus and insuetus locals at Werla.134 Henry’s reforms in this aspect of Saxon military organization were very successful and were maintained diligently by his successors. Thus, for example, Bishop Thietmar of Merseburg (1009–1018), the great chronicler of the Saxon dynasty from the early eleventh century, devoted considerable attention to the successful efforts of Henry I and his successors down to King Henry II (1002–1024) to maintain a system of frontier fortifications defended, in large part, by men of the local levy whom he consistently refers to as urbani.135 Thietmar draws attention, for example, to Henry I’s construction of the fortress of Meißen in 928 to protect an important crossing point over the Elbe river.136 In addition to his fortification of Meißen, Thietmar credits Henry I with providing the city of Merseburg with a new stone wall and with building a series of other fortresses (urbes) ad salutem regni.137 Following Henry I’s reforms, the local levies that protected these urbes proved very effective in defending them. Thietmar notes, for example, that the population of the fortress of Püchen, located on the left bank of the Mulde river northwest of the Wurzen, in modern Saxony, provided a safe refuge for a unit of Henry I’s royal troops that had been defeated by an overwhelming force of

134 135 136 137

of the Lechfeld in 955 that the Magyars recruited large numbers of foot soldiers and artillery men to undertake the siege of major fortifications. See on this point, Bernarch S. Bachrach, ‘MagyarOttonian Warfare: à propos a new Minimalist Interpretation,’ Francia 13 (2000), 211–230; and D. Bachrach, Warfare in Tenth-Century Germany, 54–57. It should be noted that Widukind inserts two chapters dealing with religious matters between his discussion of the troops at Werla and the subsequent discussion of the agrarii milites. Thietmar Chronicon, 3.21 and 3.23. In this context, see Bachrach, “Memory, Epistemology, and the Writing of Early Medieval Military History,” 75–90. Thietmar, Chronicon, 1.16. Thietmar, Chronicon, 1.18.

74

S A X O N M I L I TA RY R E V O L U T I O N , 912–973?

invading Magyars.138 Thietmar makes clear that it was a local defense force that manned the walls at Püchen by describing the defenders there as urbani rather than as members of a presidio, which is the author’s standard term for designating a garrison of professional soldiers. In return for their loyal service, Henry I rewarded the defenders who defended Püchen in such a manner that, according to Thietmar, their descendants enjoyed greater honor than the other people of the region (comprovinciales) up to the present day, i.e. up to the time when Thietmar was writing.139 Thietmar similarly emphasized the role that local levies played in defending fortifications in his own day. In 1012, following the death of Archbishop Walthard of Magdeburg, the Polish duke Boleslav Chrobry (992–1025) took advantage of the disturbed political situation on the frontier to launch an attack on the fortress of Lebusa, located 140 kilometers east of Merseburg.140 The stronghold was defended, in part, by a garrison composed of professional soldiers. However, the main body of defenders consisted of the urbani of Lebusa. Unfortunately for the defenders, it appears from Thietmar’s account that Boleslav Chrobry decided to employ overwhelming force to storm Lebusa. Although the duke paid a high butcher’s bill, with some 500 men killed, according to Thietmar, the assault was successful, leading to the capture of the fortress as well as the surviving defenders.141 The book of the dead (necrology) from Merseburg likewise includes a reference to the numerous defenders who bravely gave their lives defending Lebusa rather than running away.142

Conclusions Following the completion of Charlemagne’s conquest of the Saxon region early in the ninth century, the armed forces of the Saxons were integrated more or less gradually into the military organization of the Frankish kingdom.143 Throughout the greater part of the ninth century, Saxon military forces were mobilized by various Carolingian rulers for expeditionary purposes in various parts of the Frankish 138 Thietmar, Chronicon, 1.15. The date of this battle is not known despite Trillmich’s suggestion (21) that it took place in 924. 139 Thietmar, Chronicon, 1.15. 140 Thietmar, Chronicon, 6.80. 141 Thietmar, Chronicon, 6.80. For additional examples of the local militia (urbani) defending fortifications, see Thietmar, Chronicon, 5.21; 6.33; 6.34; and 7.48. 142 Die Totenbücher von Merseburg, Magdeburg und Lüneburg, ed. G. Althoff and J. Wollasch, MGH Libri Memoriales et Necrologia n. s. 2 (Hanover, 1983), fol. 5 r., p. 11, records that many fighting men were annihilated (multi peremientes (sic)) at Lebusa. 143 A full history of Saxon military organization through the end of the Ottonian period is a desideratum. Various works, such as Holtzmann, Geschichte der sächsischen Kaiserzeit; B. Scherff, Studien zum Heer der Ottonen und der ersten Salier (919–1056) dissertation from the Friedrich-Wilhelms Universität (Bonn, 1985); Leyser, “Early Medieval Warfare,” 29–50; and Reuter, “Carolingian and Ottonian Warfare,” 13–35, provide some useful insights but a complete picture is lacking. But now see D. Bachrach, Warfare in Tenth-Century Germany, passim.

75

N A R R AT I V E W O R K S

regna and beyond its frontiers. Indeed, military units mustered in Saxony were required to serve as far afield as Spain and Italy. As we have seen, Saxon armies, composed of the obsequia of the magnates and various elements drawn from the select levies, which were mobilized for expeditionary purposes, very frequently were successful against whatever enemies they faced. Contrary to Leyser’s view, Saxon military forces under Carolingian leadership cannot be considered second rate. When Henry I succeeded his father Otto the Illustrious in 912, he is reported to have commanded a strong (fortis) force of milites. Prior to his assumption of the kingship in 919, Henry further augmented this elite force of soldiers by providing some of his milites with estates that he confiscated from the archbishop of Mainz and other magnates who were used by King Conrad I to harass the Saxon leader. In this context, it is important to emphasize that each miles, who held twelve mansi, was required by legislation that governed the entire Frankish kingdom, of which Saxony clearly was an integral part in military terms, to appear at the muster with a war horse and full armor.144 Henry, like the previous Saxon leaders who served the reges Francorum, inherited a probatus force of heavily armed mounted troops (equites armati). As the leading office holder in Saxony, Henry was in a position to call upon his fideles, both lay and ecclesiastical, who were settled in the Saxon region and those parts of Thuringia that were under Henry’s control to provide auxilium by placing their obsequia under his command. In this context, specific mention has already been made of the military households of Thietmar and Reginbern, who used their armed forces with continuing success to carry out Henry’s military policies regarding the Slavs and Danes, respectively. Of course, when Henry became king he acquired all of the great men in east Francia as his fideles, by oath, and, at least in principle, their obsequia of heavily armed mounted troops were subject to his authority, in the same manner as were the fortifications of their duchies or lesser jurisdictions. As seen earlier, the dukes of both Swabia and Bavaria initially sought to resist Henry’s authority, but as soon as he brought his large army into their territories, they recognized his ditio and placed both their military forces and strongholds under his command. In addition to his own military household and those of his fideles, both within the Saxon duchy and throughout east Francia, Henry as king was able to mobilize a large exercitus of select levies both from the duchy and the remainder of the 144 On this point, see Helen M. Cam, Local Government in Francia and England (London, 1912), 127–37; F. L. Ganshof, “Charlemagne’s Army,” in Frankish Institutions under Charlemagne, trans. B. and M. Lyon (Providence, RI, 1968), 59–68, 151–161, originally published as “L’Armée sous les Carolingiens,” Settimane di Studio de Centro Italiano di Studi sull’alto medioevo, 2 vols. (Spoleto, 1968), I:109–130, with the citation to the English text (59–61); and Bachrach, “Military Lands,” 109–112. See, for example, Capitularia regum Francorum I, ed. Alfred Boretius (Hanover, 1883), nr. 44, c. 6, “De armatura in exercitu, sicut antea in alio capitulare commendavimus, ita servetur, et insuper omnis homo de duodecim mansis bruneam habeat; qui vero bruniam habens et eam secum non tullerit, omne beneficium cum brunia pariter perdat.”

76

S A X O N M I L I TA RY R E V O L U T I O N , 912–973?

kingdom. Elements of King Henry’s army, i.e. royal and magnate obsequia combined with select levies, operating in the field not only decisively defeated the Magyars in 924 but won a series of victories against various peoples, including the Bohemians, Abodrites, Redarii, Lotharingians, and the forces of west Francia, in the period prior to the Magyar invasion of 933.145 Indeed, by the time of the battle of Riade in 933, Henry’s milites and his exercitus of expeditionary levies had been winning major victories for more than two decades. The one weakness that has been identified regarding Henry’s military forces, at least in regard to the Saxon region, and especially on the frontiers, were the local levies. These forces, as was the case in most other parts of the Frankish kingdom, were ill trained and poorly armed. It would seem to have become clear to Henry that the local levies of the Saxon region, i.e. the first line of defense, were unable to stop the Magyars in 924 because they were rudus and insuetus with regard to the demands of combat in the field (bellum publicum). As a result, Henry undertook a significant military reform that saw the reorganization of these forces as agrarii milites and the construction or refurbishment of extensive fortifications. This reform was very successful, and Henry’s successors of the Saxon dynasty into the early eleventh century relied heavily on the local levies on the frontiers to keep the duchy safe. There are both myths and realities with regard to the Saxon military revolution. One myth, effectively propagated by the work of Karl Leyser and his scholarly successors for more than a quarter-century, is that Henry I and his predecessors lacked a fortis manus of heavily armed mounted troops prior to c. 924. The reality is that the integration of the Saxon region into the Frankish kingdom resulted in the Saxon military organization conforming to Carolingian regulations. In short, if we are to consider the conversion of the Saxon way of war to the Carolingian way of war a revolution, this change occurred during the earlier ninth century. A second myth, also propagated by Leyser, is that the Saxon military forces in the ninth and early tenth centuries were second rate. The reality is that forces mobilized from within the duchy were as good as any others in the Frankish kingdom. Henry I should not be deprived, however, of credit for having undertaken a revolutionary step that greatly improved the fighting capacity of Saxon defense forces on the frontiers. Widukind, as far as can be established, is correct in attributing to Henry the administrative reforms that resulted in the creation of the agrarii milites, who were to play a key role in defending the Saxon frontier throughout the tenth century and beyond. Henry’s equally successful efforts to build strongholds along the frontier were less revolutionary, as the notion of a fixed border defense system had been an important part of the strategic thinking of the West since the later Roman Empire and had been followed effectively by the Carolingians.146 145 See D. Bachrach, Warfare in Tenth-Century Germany, particularly chapter 1. 146 Concerning the continuity between Carolingian and Ottonian fortifications, see A. K. Hömberg, “Probleme der Reichsgutforschung in Westfalen mit Zwei Karten,” Blätter für deutsche Landesgeschichte 96 (1960), 1–21, especially 15–21 where he identifies at least twenty royal

77

N A R R AT I V E W O R K S

Appendix The peace treaty of 924 Traditionally among German historians, the munera given by Henry I to the Magyar envoys in 924 and in the nine years thereafter have been regarded as a tribute that was intended to purchase peace because Henry believed he could not defeat the Magyars in the open field.147 However, this is a fundamental misunderstanding of the historical situation and Widukind’s terminology. Widukind makes a very clear distinction between munera, which are voluntary gifts, and tributum, which is an obligatory payment. He notes, for example (1.34), that under Henry I the Saxon people went from being tribute payers (tributaria) to being free (libera), following the end of Frankish/Carolingian rule. Similarly, according to Widukind (1.35), following Henry’s capture of Prague, he made the Bohemians pay tribute (tributaria). In the same vein, Widukind (1.36) notes that a series of Slavic peoples, including the Abodrites, Wilzi, Hevelli, Daleminzi, and Redarii, were forced by Henry I to pay tribute after he had conquered them. In each case noted here, the word tributum is associated with the payments made by a conquered people. As noted earlier, Henry provided munera to the Magyar ambassadors after his forces had inflicted a serious defeat on the Magyars. Thus, the Saxons can in no way be seen in this context as a defeated people required to pay tribute. Rather, it was customary from the later Roman Empire onward for a great ruler to provide gifts to departing envoys if the former had been satisfied with the results of the embassy.148 It is also important to emphasize that Widukind, in noting that the treaty remained firm for nine years (1.32), is stating a chronological fact, not suggesting that the peace has a nine-year limit. The normal word for truce is treuga, and he does not use this word in this context.149 A second point needs to be made here with regard to the treaty of peace and the munera that Henry gave to the Magyar envoys, as discussed in Widukind, 1.38. Those scholars who believe that Henry

fortifications sustained in southern Westphalia by the kings of east Francia and Germany during the ninth and tenth centuries; and Walter Melzer, “Karolingisch-ottonische Stadtbefestigungen in der ‘Germania libera’,” in Die Befestigung der mittelalterlichen Stadt, ed. Gabriele Isenberg and Barbara Scholkmann (Cologne, 1997), 61–77, who emphasizes that there is substantial physical continuity between Carolingian and Ottonian fortifications, particularly in the region of Westphalia. Now also see D. Bachrach, “Restructuring the Eastern Frontier,” 9–35 147 See in this context, for example, Holtzmann, Geschichte der sächsischen Kaiserzeit, 93; and Althoff and Keller, Heinrich I und Otto der Große, I:89. 148 A. Gillet, Envoys and Political Communication in the Late Antique West, 411–533 (Cambridge, 2003), 256–258. Regarding the Carolingians, in this context, see, for example, Annales Bertiniani, an. 863 and an. 876. For examples of the treatment of envoys who did not succeed in pleasing the ruler, see ibid., an. 870 and an. 873. 149 With regard to the use of the technical term treuga for truce, see, for example, Widukind’s elder contemporary Flodoard, Les Annales de Flodoard, ed. Philippe Lauer (Paris, 1905), passim. Widukind, Res gestae, 1.38, makes clear that after nine years had passed, the Magyars violated the peace.

78

S A X O N M I L I TA RY R E V O L U T I O N , 912–973?

paid tribute to the Magyars in 924 traditionally misinterpret the withholding of munera by King Henry in 932 and see it as a cause for war. This view is fundamentally at odds with our understanding of early medieval diplomatic protocol, as noted earlier. The withholding of munera from envoys cannot be understood as the cause of a conflict. Rather, the withholding is to be interpreted as a result of the failure of the embassy to maintain cordial relations between the two parties. In dealing with the rupture of the pax that had been imposed by Henry I on the Hungarians in 924, Widukind presents the king as rousing the ire of his fideles against the Magyars through a speech presented at a royal assembly.150 In particular, Widukind presents Henry defending his heavy taxation (expoliare) of the Saxons by insisting that the funds that he had raised by these means had, in fact, gone to enrich the Magyars. Henry added further that in the future it would be necessary for him to tax the church and clergy in order to further enrich these pagans. In this context, Widukind, whose agenda is obvious, is trying to defend Henry’s overt effort to lead his people to war, i.e. to break the treaty of 924, rather than to maintain the peace, as was the warrant of a good Christian king. Two parts of this speech, which Widukind obviously constructed from whole cloth, are worth noting. First, as intimated previously, Henry’s desire to undertake an offensive war was a fundamental violation of the obligations of a Christian monarch to maintain the peace. Second, Widukind places in Henry’s mouth the justification that if he did not go to war, he would have to tax both the church and the clergy very heavily. Such taxation, at least since the time of Archbishop Hincmar of Rheims (died 882), was regarded as undesirable by the Frankish church. Indeed, Hincmar used Charles Martel as the “poster boy” for a bad ruler because of his supposed widespread use of church lands to support his military forces.151 Thus, by claiming that Henry I was faced with the unfortunate choice of taxing the church heavily or going to war, the latter is supposedly the more desirable alternative. In addition to using a variety of topoi, e.g. taxation of the church and just war, to justify Henry’s intention to the break the peace of 924, as well as the free-subjection motif, the entire economic framework that Widukind imposes on the situation is factually inaccurate.152 It is clear from Widukind’s own account (1.35–37), as noted earlier, that a series of Slavic peoples, including the Abodrites, Wilzi, Hevelli, Daleminzi, and Redarii, were forced by Henry I to pay tribute after he had 150 See Widukind, Res gestae, 1.38. 151 With respect to Charles Martel’s reputation as a despoiler of the church, see Ulrich Nonn, “Das Bild Karl Martells in mitteralterlichen Quellen,” in Karl Martell in seiner Zeit, ed. J. Jarnut, U. Nonn, and M. Richter (Sigmaringen, 1994), 9–21; and P. Fouracre, The Age of Charles Martel (New York, 2000), who summarizes the state of the question. Concerning Hincmar’s efforts to limit royal access to ecclesiastical resources, see J. Nelson, “The Church’s Military Service in the Ninth Century: A Contemporary View?,” Studies in Church History 20 (1983), 15–30 and reprinted in J. Nelson, Politics and Ritual in Early Medieval Europe (London, 1986), 117–32; and Bachrach, “Military Lands,” 113–15. 152 With respect to the importance of the freedom-subjection topos in Widukind’s text, see H. Beumann, “Imperator Romanorum,” 228.

79

N A R R AT I V E W O R K S

conquered them. Henry I also had access to a substantial source of wealth through the ongoing exploitation of the silver mines in the Harz mountains.153 In short, Henry was not in need of vast sums of money either to enrich the Magyars or for any other purpose. Whether Henry, in fact, imposed heavy taxation on the Saxons and/or on the church is a matter well beyond the scope of this study.

153 On this latter point, see Lothar Klappauf, “Zur Archäologie des Harzes im frühen Mittelalter: Eine Skizze zur Forschungsstand und Aussagemöglichkeiten,” in Bernward von Hildesheim und das Zeitalter der Ottonen, 2 vols., ed. M. Brandt and A. Eggebrecht (Hildesheim, 1993), 249–57; Idem, “Studies of the Development and Structure of Early Metal Production,” in Aspects of Mining and Smelting in the Upper Harz Mounatins (up to the 13th/14th Century) in the Early Times of a Developing European Culture and Economy, ed. Christiane Segers-Glocke and Harald Witthöft, trans. Claus Wibbelmann (St. Katharinen, 2000), 1–29; and Christoph Bartels, Michael Fessner, Lothar Klappauf, Friedrich Albert Linke, Kupfer, Blei und Silber aus dem Goslarer Rammelsberg von den Anfängen bis 1620 (Bochum, 2007).

80

4 E A R LY O T TO N I A N WA R FA R E The perspective from Corvey

Introduction Widukind, a monk of the royal monastery of Corvey (925 to after 973), is widely recognized by scholars to have been an exceptionally well-informed observer of contemporary affairs.1 He was almost certainly commissioned by a member of the royal family, perhaps Otto I (936–973) himself, to write a history of the Saxon dynasty, the Res gestae Saxonicae, from the period before the elevation of Henry I (919–936) as king, until the present day.2 Indeed, the text of the Res gestae Saxonicae, which was completed after 973, appears to have been written as part of the preparation for the royal princess and abbess Mathilda for her responsibilities as

1 Regarding the value of Widukind’s study as a source of accurate information regarding the Ottonian kingdom, see Wilhelm Wattenbach and Robert Holtzmann, Deutschlands Geschichtsquellen im Mittelalter: Deutsche Kaiserzeit (Berlin, 1938), 27–8; Helmut Beumann, “Die Historiographie des Mittelalters als Quelle für die Ideengeschichte des Königtums,” Historische Zeitschrift 180 (1955), 449–88; idem, “Historiographische Konzeption und politische Ziele Widukinds von Corvey,” La storiografia altomedievale: Settimane di studio del centro italiano di studi sull’alto medioevo 17 (1970), 857–94; idem, “Imperator Romanorum, rex gentium. Zu Widukind III 76,” in Tradition als historische Kraft: Interdisziplinäre Forschungen zur Geschichte des früheren Mittelalter, ed. Norbert Kamp and Joachim Wollasch (Berlin, 1982), 214–30; and Gerd Althoff, “Widukind von Corvey: Kronzeuge und Herausforderung,” Frühmittelalterliche Studien 27 (1993), 253–72. Also see the recent summary of the scholarly tradition regarding Widukind’s access to accurate information, particularly with respect to military matters, by Bernard S. Bachrach and David S. Bachrach, “Saxon Military Revolution, 912–973?: Myth and Reality,” Early Medieval Europe 15 (2007), 186–222, here 192–7. The failed effort of Johannes Fried to undermine the value of early medieval narrative sources, including Widukind’s Res gestae, for gaining accurate information about the past, and particularly about matters regarding military history is discussed in detail by David S. Bachrach, “Memory, Epistemology, and the Writing of Early Medieval Military History: The Example of Bishop Thietmar of Merseburg (1009–1018),” Viator 38 (2007), 63–90, here 63–70. 2 For the text of the work, see Widukind of Corvey, Res gestae Saxonicae, in Quellen zur Geschichte der sächsischen Kaiserzeit, vol. 8, ed. and trans. Reinhold Rau (Darmstadt, 1971), 3–183. Regarding the commissioning of the text, see Althoff, “Widukind von Corvey,” 262–72. For an English translation of Widukind’s text, see Bernard S. Bachrach and David S. Bachrach, Widukind of Corvey: Deeds of the Saxons (Washington, DC, 2014).

81

N A R R AT I V E W O R K S

regent in Germany during the absence of her father Otto I and her brother Otto II (973–983) in Italy.3 As a consequence of his detailed insights into contemporary affairs, specialists in early medieval military history have devoted considerable attention to Widukind’s Res gestae. This is particularly true with regard to Widukind’s discussion of the so-called agrarii milites, whom Widukind identifies as having been organized by Henry I as part of the king’s effort to secure the eastern frontiers of Saxony and Thuringia against Hungarian raids.4 A second major scholarly focus has been Widukind’s discussion of the series of military operations that ended with the destruction of an invading Hungarian army in 955, i.e. the Lechfeld campaign.5 At a broader level, Widukind has been deployed by generations of Germanlanguage scholars to support the model of early medieval warfare that was dominated by small numbers of heavily armed and mounted warriors (Panzerreiter) who were commanded by secular and ecclesiastical magnates.6 The state of the 3 Regarding the writing of the Res gestae as a mirror for Mathilda, see Karl F. Morrison, “Widukind’s Mirror for a Princess: An Exercise in Self-Knowledge,” in Forschungen zur Reichs-, Papst- und Landesgeschichte. Peter Herde zum 65. Geburtstag von Freunden, Schülern und Kollegen dargebracht, 2 vols., ed. Karl Borchardt and Enno Bünz (Stuttgart, 1998), 49–71. 4 With respect to the agrarii milites, see the survey of the German and English literature by Edward J. Schoenfeld, “Anglo-Saxon ‘Burhs’ and Continental ‘Burgen’: Early Medieval Fortifications in Constitutional Perspective,” The Haskins Society Journal 6 (1994), 49–66, which is now indispensable; Charles Bowlus, The Battle of Lechfeld and Its Aftermath, August 955: The End of the Age of Migrations in the Latin West (Ashgate, Aldershot, 2006), 45–60; and Bachrach and Bachrach, “Saxon Military Revolution,” 216 and 219. 5 Widukind’s account of the battle of the Lechfeld has been the subject of numerous studies. See, for example, Karl Leyser, “The Battle at the Lech, 955: A Study in Tenth-Century Warfare,” History 50 (1965), 1–25 and reprinted in Medieval Germany and Its Neighbours, 43–67; idem, “Early Medieval Warfare,” in Communications and Power in Medieval Europe: The Carolingian and Ottonian Centuries, ed. Timothy Reuter (Hambledon Press, London, 1994), 29–50; Lorenz Weinrich, “Tradition und Individualität in den Quellen zu Lechfeldschlacht,” Deutsches Archiv 27 (1971), 291–313; Bernard S. Bachrach, “Magyar-Ottonian Warfare: À propos a New Minimalist Interpretation,” Francia 27 (2000), 211–30; Charles Bowlus, “Der Weg vom Lechfeld: Überlegungen zur Vernichtung der Ungarn im August 955 anhand ihrer Kriegsführung,” in Bayern-Ungarn. Tausend Jahre, ed. Herbert Wilhelm Wurster, Manfred Treml, and Richard Loibl (Augsburg, 2001), 77–90; and most recently, idem, The Battle of Lechfeld, passim. 6 Regarding the putatively dominant role of heavily armed and mounted warriors in early medieval warfare, see Franz Beyerle, “Zur Wehrverfassung des Hochmittelalters,” in Festschrift Ernst Mayer zum 70. Geburtstag (Weimar, 1932), 31–91; Eugen von Frauenholz, Das Heerwesen der germanischen Frühzeit, des Frankenreiches und des ritterlichen Zeitalters (Munich, 1935); Paul Schmitthenner, “Lehnskriegswesen und Söldnertum im abendländischen Imperium des Mittelalters,” Historische Zeitschrift 150 (1934), 228–67; Heinrich Sproemberg, “Die Feudale Kriegskunst,” in Beiträge zur belgisch-niederländischen Geschichte III: Forschungen zur mittelalterliche Geschichte (Berlin, 1959), 30–55; Leopold Auer, “Der Kriegsdienst des Klerus unter den sächsischen Kaisern,” Mitteilungen des Instituts für Österreichische Geschichtsforschung, part 1, 79 (1971), 316–407; and part 2 in ibid., 80 (1972), 48–70. The continued dominance of this view in German-language scholarship is clear in Malte Prietzel, Kriegführung im Mittelalter: Handlungen, Erinnerungen, Bedeutungen (Paderborn, 2006), 21–3, and passim; and idem, Krieg im Mittelalter (Darmstadt, 2006), passim.

82

E A R LY O T T O N I A N WA R FA R E

question in regard to the dominance of heavily armed mounted troops in military operations was set forth for a generation of Anglophone readers by Karl Leyser, whose work on Ottonian warfare drew heavily on Widukind’s Res gestae. Leyser argued, for example, “The mounted warrior, the miles armatus, was now [in the tenth century] the backbone of effective belligerency.” He concluded that such heavily armed horsemen possessed a “monopoly” of military force.7 Leyser was followed by his student, Timothy Reuter, who argued forcefully that both Carolingian armies of the ninth century and Ottonian armies of the tenth were dominated by bands of warrior vassals and their lords, who were motivated to fight by the glory of combat and the hope for plunder and booty.8 In this context, Reuter has been criticized for misrepresenting in his translation of the Annals of Fulda a key text to support his contention that raiding by small bands of mounted troops was the dominant form of warfare during the late Carolingian period. Instead of accurately presenting the text, which recorded that the Carolingians wore down their Slavic adversaries with numerous sieges of their fortifications, Reuter presented the Latin as stating that the Carolingians wore down their adversaries with raids. This mistranslation conveniently supports Reuter’s general view on the nature of war in the Carolingian era by obscuring the fact that siege warfare was at issue.9 The focus by scholars on mounted warriors as the primary combatants in tenth-century warfare has led many historians, as contrasted with specialists in archaeology, to ignore or downplay the crucial role played by fortifications in both offensive and defensive campaigns carried out by the Ottonian kings.10 Indeed, the focus on Panzerreiter has led some historians to almost comical 7 See, for example, Karl Leyser, “Early Medieval Canon Law and the Beginning of Knighthood,” in Institutionen, Kultur und Gesellschaft im Mittelalter. Festschrift für Josef Fleckenstein zu seinem 65. Geburtstag, eds. Lutz Fenske, Werner Rösener, and Thomas Zotz (Sigmaringen, 1984), and reprinted in Communications and Power in Medieval Europe: The Carolingian and Ottonian Centuries, ed. Timothy Reuter (London/Rio Grande, 1994), 51–71, and for the quotation, 54–55. 8 Timothy Reuter, “Plunder and Tribute in the Carolingian Empire,” Transactions of the Royal Historical Society 5th series 35 (1985), 75–94; idem, “The End of Carolingian Military Expansion,” in Charlemagne’s Heir: New Perspectives on the Reign of Louis the Pious (814–840), ed. Peter Godman and Roger Collins (Oxford, 1990), 391–405; and idem, “Carolingian and Ottonian Warfare,” in Medieval Warfare: A History, ed. Maurice Keen (Oxford, 1999), 13–35. 9 For the pointed criticism of Reuter’s translation in The Annals of Fulda: Ninth Century Histories (Manchester, 1992), see Eric Goldberg, “Ludwig der Deutsche und Mähren. Eine Studie zu karolingischen Grenzekriegen im Osten,” in Ludwig der Deutsche und seine Zeit, ed. Wilfried Hartmann (Darmstadt, 2004), 67–94, at 70, n. 7. In his own study of the reign of Louis the German, Goldberg, Struggle for Empire: Kingship and Conflict under Louis the German, 817–876 (Ithaca, NY, 2006), 119–31, makes clear the dominant role that sieges played in the conduct of war during the late Carolingian period in the lands that would become the Ottonian kingdom. 10 Among the few exceptions to this general neglect of the role of fortifications in the conduct of war in Ottonian Germany, and its east Carolingian predecessor are Bowlus, Lechfeld, passim; and Goldberg, Struggle for Empire, particularly 119–31. For a thorough treatment of the archaeological scholarship and the central role of fortifications in Ottonian warfare, now see David S. Bachrach, Warfare in Tenth-Century Germany (Woodbridge, 2012), passim.

83

N A R R AT I V E W O R K S

conclusions regarding the role of fortifications and sieges in early medieval warfare. Heinrich Sproemberg, at a loss to explain the very large number of fortress cities and fortified sites throughout Germany, argued that it was impossible for attackers to capture these fortifications because there was a general absence of foot soldiers and siege artillery during the early Middle Ages.11 Heinrich Koller argued that “the grotesque situation was in force that the nobility built fortifications that could only be defended on foot, but continued to hold fast to the principle of only fighting on horseback.”12 In a focused study on the medieval German city under siege, Michael Toch similarly asserted the very futility of siege warfare, given armies that were composed almost exclusively of mounted warriors. Indeed, Toch argued explicitly that cavalry battles were the only manner in which wars could be settled because sieges were so rarely successful.13 The burden of the following study is to demonstrate that the efforts of scholars to identify heavily armed, mounted fighting men as playing the central role in early medieval warfare cannot be sustained by reference to the work of Widukind. Rather, the monk of Corvey devoted most of his attention to siege warfare, that is to the capture and defense of fortifications. Obviously, military operations of this type necessarily were dominated by men fighting on foot.

Siege warfare In reading the entire text of the Res gestae as a unified whole, it is clear that there is one point to which the Corvey monk returned repeatedly throughout his work. This is the central role of sieges in warfare and the crucial role that fortifications played in military operations throughout the reigns of Henry I and Otto I. To use a rather crude measurement, Widukind’s emphasis on siege warfare can be quantified numerically. Widukind identifies fifty-six military operations led by Henry I, Otto I, or their military commanders in the period c. 916–969, regarding which he provides sufficient information to determine the types of combat operations that took place.14 Of these, forty-five deal directly with efforts by the first two Ottonian kings or their commanders to defend or capture various types of fortifications, including great fortress cities such as Mainz, Regensburg, Augsburg, Rheims, Laon, Paris, and Pavia. In several cases, the campaigns encompassed the capture of numerous fortifications, which Widukind does not list separately

11 Sproemberg, “Die Feudale Kriegskunst,” 39–40. 12 Heinrich Koller, “Das mittelalterliche Stadtmauer als Grundlage städtischen Selbstbewußtseins,” in Stadt und Krieg, ed. Bernhard Kirchgässner and Günter Scholz (Sigmaringen, 1989), 9–25, here 17. 13 Michael Toch, “The Medieval German City Under Siege,” in The Medieval City Under Siege, ed. Ivy A. Corfis and Michael Wolf (Woodbridge, 1995), 35–48, here 35–6. 14 This includes several military operations commanded by Henry as duke of Saxony before his royal accession in 919. For the entire list, see the appendix.

84

E A R LY O T T O N I A N WA R FA R E

by name.15 These sieges took place in every region of the German kingdom, in the West Frankish kingdom, in Italy, and in the Slavic lands east of the Elbe and Saale river systems.16 By comparison, Widukind discusses just eleven battles in the field in which Henry, Otto, or their officers commanded one side.17 Even this small number is misleading, however, because a majority of these battles took place in the context of sieges. In 929, for example, Henry I’s officers Bernard and Thietmar inflicted a crushing defeat on a Slavic army that had come to relieve the siege of the fortress at Lenzen.18 The most famous military encounter of Henry I’s reign, at Riade in 933, was fought as a consequence of the king’s effort to relieve a Hungarian siege of an unnamed, and still unidentified, fortress that housed part of the royal treasury, as well as Henry’s sister.19 Similarly, Otto I’s defeat of the Hungarian army on the Lechfeld in 955 took place as a consequence of the king’s relief of the siege of Augsburg.20 Otto’s famous victory over the Obodrites at the battle on the Recknitz river on 16 October 955 also can be seen to have followed as a consequence of planned siege operations.21 When the supposedly enormous Abodrite army (ingens exercitus), under the command of the princes Nacco and Stoignew, blocked Otto’s passage over the Recknitz, the German king ordered the deployment of his machinae as well as archers to pin down the Slavs at the ford. He then dispatched a flying column down the river to construct pontoon bridges, which mounted troops subsequently crossed to launch a flank assault on the Abodrites massed across the river from Otto’s main army.22 The deployment of engines, which were assembled during the night of 15 October, is prime facie evidence for the presence of a siege train with Otto’s army.23 In sum, just three of the eleven battles in the field discussed by Widukind can be identified to have taken place without reference to a siege.24

15 In describing Otto I’s effort to crush the revolt by Duke Conrad the Red of Lotharingia in 953, for example, Widukind, Res gestae, 3.18 notes, “Rex autem circa Kalendas Iulii moto exercitu armis filium generumque quaerere temptavit; obvias urbes partis adversae aut armis cepit aut in deditionem acceptit.” The campaign involved the capture or surrender of numerous fortifications, but Widukind does not specify any of them by name. 16 For the sieges of these cities, see Widukind, Res gestae, 1.27, 3.3, 3.4, 3.10, 3.21, 3.36, 3.43, 3.44, and 3.64. 17 See the appendix for a full list of these battles. 18 Widukind, Res gestae, 1.36. 19 Ibid., 1.38. 20 Ibid., 3.44. 21 Widukind, 3.53–5. 22 Ibid., 3.54. 23 Ibid., 3.54, “Imperator vero de nocte consurgens iubet sagittis et aliis machinis ad pugnam provocare.” 24 These are Widukind, Res gestae, 1.38, 2.3, and 2.17. The battle at Andernach (2 October 939), in which the rebel dukes Gislebert of Lotharingia and Eberhard of Franconia were killed, took place as they were attempting to relieve Otto I’s siege of the fortress of Breisach. See Widukind, Res gestae, 2.25–6.

85

N A R R AT I V E W O R K S

Perhaps even more illuminating regarding Widukind’s presentation of the overwhelming predominance of sieges in the Res gestae is the author’s repeated emphasis on the possession of fortifications as a crucial factor in military, and subsequently political, power. In a well-known passage, Widukind places into the mouth of the dying King Conrad I (911–918) words of advice for his younger brother, Eberhard.25 He said, in part, “brother, we are able to raise and command large armies, we have fortresses, arms, the royal insignia, and everything that is fitting for king, but we lack luck and aptitude.”26 Here, the possession of fortifications is held up as one of the central requirements for effective royal rule. In a similar vein, Widukind uses the surrender of fortifications as a symbol for political subordination. Following his rapid invasion of Swabia in 919, Henry I overawed Duke Burchard of Swabia, convincing the latter that he could not resist the newly elected king.27 In describing Burchard’s submission, Widukind writes, “as a prudent man, realizing that he could not resist the king’s invasion, he surrendered himself, along with all of his fortresses, and all of his people to him.”28 Otto subsequently permitted Duke Burchard to retain fortifications in Swabia under his own control, keeping intact, however, the royal right of rendition. Following the death of Duke Gislebert of Lotharingia at the battle of Andernach in 939, his nephews surrendered to Otto I. In return, Widukind stressed, the king permitted them to retain their fortresses (urbes).29 In describing substantial military campaigns by Henry, Otto, and their military commanders, Widukind focuses almost exclusively on the results of their siege operations. In 946, for example, Otto I led a major invasion of Francia in support of his brother-in-law King Louis IV (929–954).30 In describing the three-month campaign, Widukind notes Otto’s assault on the fortress city of Laon, his successful siege of Paris, his capture of the fortress city of Rheims, and finally his failed siege of Rouen.31 Subsequent to suffering heavy losses at Rouen, according to Widukind, “Otto returned to Saxony after three months, having captured Rheims and Laon, and handing these over, as well as the other places he captured by force

25 Widukind, Res gestae, 1.25. 26 Ibid., “Sunt nobis, frater, copiae exercitus congregandi atque ducendi, sunt urbes et arma cum regalibus insignias et omne quod decus regium deposcit preter fortunam atque mores.” 27 Widukind, Res gestae, 1.27. 28 Ibid., “quia valde prudens erat, congressionem regis sustinere non posse, tradidit semet ipsum ei cum universes urbibus et populo suo.” 29 Widukind, Res gestae, 2.28. In this context, urbs is the standard term used by Widukind and many other early medieval writers to denote fortifications. The term for city, by contrast, is usually civitas. 30 Widukind, Res gestae, 3.3–4. For a French perspective on this invasion, which also focuses on sieges, see Les Annales de Flodoard, ed. Ph. Lauer (Paris, 1905), anno 946. For an English translation of this text, see Bernard S. Bachrach and Stephen Fanning, The ‘Anals’ of Flodoard of Reims, 919–966 (Toronto, 2004). 31 Widukind, Res gestae, 3.3–4.

86

E A R LY O T T O N I A N WA R FA R E

of arms, to King Louis.”32 The entire campaign is summed up with reference to the capture of the two fortress cities and other unnamed fortifications. In a similar manner, Widukind summarizes the invasion of northern Italy in 952 by Otto’s son and duke of Swabia, Liudolf (died 957), by focusing entirely on siege operations: “He led an armed force into Italy, captured several cities there, left garrisons, and returned to Franconia.”33 The most compelling aspect of Widukind’s focus on the centrality of fortifications to warfare is his thorough treatment of particular military campaigns that ended in lengthy sieges, including a discussion of the tactical and strategic errors made by Ottonian commanders. Widukind’s most detailed account of all is a discussion of the major revolt against Otto I in 953–955 led by his son Liudolf, his son-in-law Duke Conrad the Red of Lotharingia (died 955), and Bavarian magnates led by Arnulf the Younger, the son of Duke Arnulf the Bad of Bavaria (died 936).34 During the first stage of Otto I’s campaign in July 953, Widukind stresses that the king captured all of Conrad the Red’s fortifications in the duchy of Lotharingia before marching to the fortress city of Mainz, where he personally directed the siege for two months.35 During the course of the operation, Otto ordered the construction of numerous siege engines (multae machinae) that were deployed against the walls, but subsequently were destroyed when the defenders succeeded in setting them on fire.36 When both sides had wearied of the siege, according to Widukind, Otto opened negotiations with his son, Liudolf, who commanded the defenses of Mainz. The failure of the negotiations caused dissention in the ranks of the royal army, which permitted Liudolf to flee the city and take refuge in the Bavarian capital of Regensberg. There, he was joined by disaffected Bavarian magnates, who preferred the rule of Arnulf, noted earlier, to that of Henry, the younger brother of King Otto.37 It is at this point that Widukind points to an error committed by Otto that gave an unnecessary advantage to the rebels. Rather than bringing the siege of Mainz to a successful conclusion or massing overwhelming force to crush the rebellion, Otto led a small column (cum paucis) into Bavaria in pursuit of his son.38 Widukind makes it clear that Otto intended to maintain the siege of Mainz during his own absence in Bavaria. In this context, Widukind records that the dux Hermann, whom the king had left in command in Saxony, was ordered to send a new army (exercitus) from Saxony to Mainz to join the army that had begun the siege of this city.39 However, this army was ambushed by forces loyal to the rebels Liudolf and 32 Ibid., 3.4, “post tres menses Saxoniam regressus est, urbibus Remense atque Lugduno cum caeteris armis captis Hluthowico regi concessis.” 33 Widukind, Res gestae, 3.6. 34 Widukind, Res gestae, 3.13–43. 35 Ibid., 3.18 36 Ibid. 37 Widukind, Res gestae, 3.20. 38 Ibid., 3.26. 39 Widukind, Res gestae, 3.23.

87

N A R R AT I V E W O R K S

Conrad and was turned back.40 Mainz finally surrendered to Otto in the winter of 954, when the rebellion had already begun to collapse.41 Otto’s failure at Mainz is associated by Widukind with the king’s additional failure during the initial invasion of Bavaria in the late autumn of 953. Widukind’s emphasis on the small size of the royal column that participated in this campaign is in marked contrast to his frequent emphasis on the very large armies that the king normally led on campaign. In describing Otto’s invasion of Bohemia in July 950, for example, Widukind stressed that Duke Boleslav decided against attempting to hold the fortress of Prague against the German king because of the immense size of the German army.42 Gerhard, the author of the contemporary Vita of Bishop Ulrich of Augsburg (923–973), notes that in order to expand his forces, Otto mobilized troops along his line of march, including the military household of the Augsburg bishop, to join in the invasion of Bavaria. However, Otto did not take with him either his siege train or his artillery engineers. Gerhard specifically mentions that Otto left behind his siege train in order to move rapidly.43 Widukind points out that the result of this rapid and not well-prepared advance was three months of frustration as the Bavarian rebels simply took refuge in fortifications, which Otto was not able to besiege.44 Finally giving up the campaign, Otto withdrew in December 953 to Saxony.45 During the winter months of 954, Liudolf employed Hungarian mercenaries to attack his opponents in Bavaria. As a result of Liudolf’s recruitment of these hated enemies, according to Widukind, Otto was able to achieve several diplomatic successes, first by accepting the surrender of Duke Conrad the Red of Lotharingia and then by reaching a settlement with a substantial portion of the Bavarian magnates who had opposed his invasion of this regnum in 953.46 Liudolf and his supporters were now isolated, holding only the city of Regensburg and a small number of fortifications in its environs. Widukind then describes Otto’s carefully orchestrated campaign that ended with the siege and capture of Regensburg. In mid-June 954, the king led a large army to the Bavarian town of Langenzenn, approximately 23 kilometers west of

40 Ibid., 3.23. 41 Widukind, Res gestae, 3.41. 42 Widukind, Res gestae, 3.8, “Considerata itaque virtute Regis ac innumera multitudine exercitus, Bolizlav urbe egressus maluit tantae maiestati subici quam ultimam perniciem pati.” 43 Gerhard of Augsburg, Vita Sancti Uodalrici, ed. Walter Berschin and Angelika Häse (Heidelberg, 1993), 176, “dimissa parte in augusta civitate vasallorum suorum caeterisque rebus collocatis cum quibus potuit, omisso vehiculo carpenti, aequitandi in servitium regis in regionem noricorum sagaciter venit (my emphasis).” 44 Widukind, Res gestae, 3.26, “Boioarii repentino regis adventu nec ad pacem vertuntur nec bellum publicum presumunt, sed clausi muris grandem exercitui laborem suaeque regionis solitudinem parant.” 45 Widukind, Res gestae, 3.26. 46 Widukind, Res gestae, 3.31–3.

88

E A R LY O T T O N I A N WA R FA R E

Nuremberg, to a prearranged meeting with Bavarian magnates and Liudolf.47 After failing to agree to terms, Liudolf withdrew in the night, and Otto pursued him, stopping first at the powerful fortress of Roßtal, located 20 kilometers south of Nuremberg, which was held by Liudolf’s ally, the count of Hammerstein.48 The fort, originally constructed by Charlemagne circa 800, is built on a plateau west of the Rednitz River and encloses an area of some 5 hectares. In 953, it was protected by a blended wall, comprising a mortared stone surface approximately 2 meters thick fronting an earthen and timber wall 7 meters wide. The wall was protected by a 5.5-meter berm and two ditches. The inner ditch was 15 meters wide and 3.5 meters deep.49 Not wishing to waste any time in his pursuit of Liudolf, Otto decided against investing the fortress and beginning what might be a lengthy siege. Instead, he ordered an assault on the walls, which lasted all day long. According to Widukind, both sides suffered heavy losses, but when night fell, the defenders still held out.50 At this point, Widukind stresses that Otto decided it was not worth the time to remain at Roßtal any further, and so ordered the advance to continue to Regensburg, a march of three days.51 The implication here is that the decision to attack Roßtal was an error. Either Otto should have taken the time necessary to capture the fortification or he should have bypassed it completely, thereby avoiding the heavy casualties that had resulted from the one day of assaults. Parenthetically, the extensive excavations at Roßtal have shown that there was no damage done inside the walls at this time, making clear that Widukind’s depiction of the events during this assault is accurate.52 Once Otto arrived at Regensburg, Widukind notes that elaborate siege camps were established around the city, with a principle aim of blocking access to the major gates built into the east and west walls.53 Regensburg was constructed on the south side of the Danube as a Roman legionary camp, and the imperial walls of the city, rebuilt in the fourth century, were still in use in 954. They comprised

47 Widukind, Res gestae, 3.31–2. 48 Widukind, Res gestae, 3.35, for the assault on the fortress. Regarding the identity of the defending commander, see Peter Ettel, “Der Befestigungsbau im 10. Jahrhundert in Süddeutschaldn und die Rolle Ottos des Großen am Beispiel der Burg von Roßtal,” in Europa im 10. Jahrhundert: Archäologie einer Aufbruchszeit, ed. Joachim Henning (Mainz, 2002), 365–79, here 373. 49 Regarding the defenses at Roßtal, see Ettel, “Der Befestigungsbau,” 373–8. 50 Widukind, Res gestae, 3.35. 51 Ibid., 3.36. It is about 100 kilometers from Roßtal to Regensburg, indicating a marching speed of approximately 33 kilometers per day. This is likely an exaggeration of the speed of the entire force, which included a substantial siege train. However, it is possible that forward elements of Otto’s army reached Regensburg in three days. Regarding the marching rates of medieval armies, see J. W. Nesbitt, “Rate of March of Crusading Armies in Europe: A Study in Computation,” Traditio (1963), 167–82. 52 See Ettel, “Der Befestigungsbau,” 378. 53 Widukind, Res gestae, 3.36–7, “Castrorum loca occupata munitionibus circumsepta.” He notes that attacks were made against the king’s castrum facing the west gate, and against Gero’s castrum that faced the east gate.

89

N A R R AT I V E W O R K S

a rectangle measuring 2,000 meters in length and enclosing an area of 24.5 hectares. These walls were built of mortared stone that ranged in width from 2.2 to 3 meters.54 Between 920 and 936 the walls of Regensburg were massively expanded to the west by Duke Arnulf the Bad of Bavaria, noted earlier, to enclose both the merchant district and the monastery of St. Emmeram. The fortified addition measured approximately 3,000 meters and enclosed an additional 40 hectares.55 The entire circuit of exterior walls, therefore, was approximately 4,500 meters in length. It is likely that walls of the western extension, which Otto faced in 954, were built of earth and timber, with perhaps a stone face, of the same type that were constructed at Roßtal.56 The new walls also were protected by a substantial ditch. It is not known, however, whether this was filled with water diverted from the Danube.57 Otto I, according to Widukind, had planned to use siege engines (machinae) to overtop the walls. However, the size of the defending force made it impossible to bring the engines into position.58 Widukind does not elaborate here about the specific reasons why Otto could not bring his engines to bear. However, the presence of a substantial defensive trench beyond the western extension of the city’s walls and the very large number of defenders meant that enormous numbers of missile weapons such as bows and defensive artillery could be directed at the attacking forces as they tried to maneuver the siege towers into position.59 The ultimate result was that the king settled in for a lengthy siege, hoping to starve the defenders into submission.60

54 Regarding the Roman walls and a synthesis of the archaeological investigations of Regensburg in the Roman and early medieval periods, see See Carlrichard Brühl, Palatium und Civitas: Studien zur Profantopographie spätantiker Civitates vom 3. bis 13. Jahrhundert, vol. 2 (Cologne, 1990), 219–55, with a discussion of the Roman walls at 229. 55 Ibid., 239–41. 56 This type of blended fortification, with a mortared stone front, was typical of fortifications built throughout Germany during the tenth century. Regarding these techniques, see Reinhard Friedrich, “Ottonenzeitliche Befestigungen im Rheinland und im Rhein-Main-Gebiet,” in Europa im 10. Jahrhundert: Archäologie einer Aufbruchszeit, ed. Joachim Henning (Mainz, 2002), 351– 63, here 353. 57 Regarding the construction of the western expansion of Regensburg’s walls, see Brühl, Palatium und Civitas, 241. Brühl notes, ibid., that this earth and timber wall likely was replaced by a stone around 1000. 58 Widukind, Res gestae, 3.36, “Sed cum multitudo machinas muris adplicari non sineret, satis dure interdum ab utrisque pugnatum pro muris.” 59 Regarding the very heavy casualties that attackers would suffer in attacking a fortification while advancing over level ground, see Bernard S. Bachrach and Rutherford Aris, “Military Technology and Garrison Organization: Some Observations on Anglo-Saxon Military Thinking in Light of the Burghal Hidage,” Technology and Culture, 31 (1990): 1–17; and reprinted in Bernard S. Bachrach, Warfare and Military Organization in Pre-Crusade Europe (London, 2002), with the same pagination. These casualties would certainly have been much higher for troops attempting to advance over substantial ditches or trying to fill in these ditches so as to bring siege towers to bear. 60 Widukind, Res gestae, 3.36, notes that rather than face starvation, Liudolf ordered an attack on the king’s camp that had been constructed facing the western city gate.

90

E A R LY O T T O N I A N WA R FA R E

Liudolf, aware that he could not expect any relief effort to break the siege, decided, according to Widukind, to launch a daring two-pronged assault on the royal camp built facing the western gate of the city. One force, comprising the mounted troops based within Regensburg, launched a direct attack against Otto’s forces that were stationed in front of the west gate. A second force, dispatched from the northern waterfront of the city, traveled west on the Danube by ship and then attacked the royal camp after landing on the riverbank behind it.61 In describing the successful response to Liudolf’s two-pronged assault, Widukind stressed that rather than meeting the attack by Liudolf’s mounted troops with all of the men stationed in the fortified camp, Otto left behind a substantial garrison, which provided a secure base of operations. As a result, rather than finding an empty camp, Liudolf’s men came up from the riverbank to face a well-prepared and protected force of armed men (armati).62 The morale of Liudolf’s troops collapsed when they realized that they had lost the element of surprise, and they beat a hasty retreat back to their river boats. However, the loss of command and control at the tactical level among Liudolf’s men led to the almost complete annihilation of this water-borne force, as Otto’s troops advanced rapidly from the camp in good order and cut them down from behind.63 Following the failed assault, Liudolf and his commanders decided to sue for peace, but the negotiations broke down when the rebels refused to accept King Otto’s terms. As a consequence, according to Widukind, Liudolf ordered an assault on the siege camp located opposite the eastern gate of the city, which was commanded by one of Otto’s officers, the margrave Gero.64 After the failure of this assault to break the siege lines, Liudolf again sued for peace, this time acceding to his father’s terms.65 According to the terms of the agreement, defenders withdrew from the “new city” (nova urbs) comprising the walled area that had been built by Duke Arnulf the Bad. A substantial force of Saxons, who had served Liudolf in his rebellion against Otto, refused to surrender the old city that was protected by the Roman walls.66 Unable to remain in Bavaria any longer, Otto departed for Saxony and left his brother Henry in command at Regensburg. Otto returned to Bavaria with an army in April 955. On this occasion, the king was able to deploy his siege engines as well as a large number of archers against the Roman fortifications of the old city of Regensburg.67 The Saxon garrison (presidio Saxonum) recognized that they would not be able to maintain their defense of the walls and surrendered to Otto. 61 Widukind, Res gestae, 3.36. 62 Ibid., “Exilientesque de navibus irruunt in castra, offendentesque. armatos.” 63 Ibid., “dum trepidi fugae consulunt, circumfusi undique caeduntur. Alii naves ingredi nisi, timore perculsi deviantes, flumine absorbentur; pauci de pluribus superessent.” 64 Widukind, Res gestae, 3.37. 65 Widukind, Res gestae, 3.38–41. 66 Widukind, Res gestae, 3.43. 67 Ibid., “Proximum agens rex pascha cum fratre, ducit post haec exercitum contra Rainesburg, iterum armis machinisque urbem torquens.”

91

N A R R AT I V E W O R K S

The commanders (principes) of the defending forces received the rather lenient punishment of exile, while the mass of the troops in the garrison (reliquae multitdini) were spared and allowed to return home.68

Implications In two independent studies, Gerd Althoff and Karl Morrison concluded that Widukind was commissioned by the royal court, likely by Archbishop William of Mainz (954–968), Otto I’s illegitimate son, by Otto I himself, to provide a teaching text for the royal princess Mathilda and her circle.69 There was certainly nothing new about this type of commission. The reading of history had a long tradition in the Roman Empire, in the early medieval West, and under the Byzantines as a fundamental component in the training of the men (and a few women) who would have responsibilities as rulers. In the present context, reading history also was seen as fundamental to learning how to make sound military decisions.70 As such, works of history were understood to complement military manuals, such as Vegetius’s Epitoma rei militaris and Frontinus’s Strategemata.71 Indeed, the importance of learning from books was understood not only at the royal court but also by important magnates in the German kingdom. In his Vita of Bishop Ulrich of Augsburg, for example, Gerhard provided a typical example of this method of using written texts to educate future leaders in his discussion of the career of Adalbero, the nephew of Bishop Ulrich. Shortly after the death of Henry I and the accession of Otto I in 936, Bishop Ulrich sent his nephew Adalbero to study under the direction of a monk named Benedict, to whom Gerhard gives the 68 Ibid. 69 Morrison, “Widukind’s Mirror for a Princess,” 49–71; and Althoff, “Widukind von Corvey: Kronzeuge und Herausforderung,” 262–72. 70 With regard to the general pattern of elite members of society learning from books, see J. M. Wallace-Hadrill, The Barbarian West 400–1000, 3rd revised edition. (Oxford, 1967), 98–103; Rosamond McKitterick, “The Audience for Latin Historiography in the Early Middle Ages: Text Transmission and Manuscript Dissemination,” in Historiographie in frühen Mittelalter, ed. Anton Scharere and Georg Scheibelreiter (Vienna, 1994), 96–114; and Mayke de Jong, “The Empire as Ecclesia: Hrabanus Maurus and Biblical historia for Rulers,” in The Uses of the past in the Early Middle Ages, ed. Y. Hen and Michael Innis (Cambrdige, 2000), 191–226. Regarding the classical antecedents to this model of training officers, see Bernard S. Bachrach, “The Education of the ‘Officer Corps’ in the Fifth and Sixth Centuries,” in La noblesse romaine et les chefs barbares du IIIe au VIII siècle, in quarto, ed. Françoise Vallet and Michel Kazanski (Paris, 1995), 7–13. With regard to the crucial role of learning from books during the Middle Ages, see Alexander Murray, Reason and Society in the Middle Ages (Oxford, 1986), 127–30, for specific discussion of the use of military manuals such as Vegetius Epitoma rei militaris. For an overview of the problem of using books to train officers in medieval armies, see Bernard S. Bachrach, “‘A Lying Legacy’ Revisited: The Abels-Morillo Defense of Discontinuity,” Journal of Medieval Military History 5 (2007), 154–93, here 159–61 and 169–91. Now also see the discussion by Bachrach, Warfare in Tenth-Century Germany, particularly 102–34. 71 See the discussion of this topic, with a survey of the relevant literature, in Bachrach, “‘Lying Legacy’ Revisited,” 169–91.

92

E A R LY O T T O N I A N WA R FA R E

honorific doctissimus magister.72 Ulrich did not intend for Adalbero to have a career in the church, although the youth’s study of the scientia grammaticae and alia libra certainly could have prepared him for this path. Rather, as Gerhard makes clear, when the young Adalbero had become educated (doctus atque educatus) in the foundations of both good knowledge (bona scientia) and training (disciplina) and had grown into his manly strength (virile robur), he was taken from school (scola) and presented by his uncle Ulrich to Otto I, whom Gerhard refers to anachronistically as imperator.73 Ulrich recommended the young man for service in Otto I’s household (regalis servitium).74 According to Gerhard, who was well acquainted both with Ulrich and with Adalbero, the young man did exceptionally well at the royal court (curtis imperatoris), “putting in a solid effort every day.”75 After receiving training at the royal court, Adalbero then returned to his uncle’s household and assumed command of the Augsburg bishop’s military forces (militia episcopalis). Adalbero held this post when Otto I mobilized Ulrich’s troops for the ill-fated invasion of Bavaria in September 953, noted previously.76 In light of the important role that books, including histories, played in the training of leaders, including military leaders, to assume their responsibilities, Widukind’s emphasis on the central role of siege warfare has important implications for our understanding of the conception of war at the Ottonian court during the reign of Otto I. Throughout his text, Widukind stressed the crucial role that sieges played in both offensive campaigns and in the defense of territory. The Corvey monk pointed out the importance of deploying a siege train to undertake sieges successfully. Widukind also noted the importance of deploying large armies for siege operations and the problems that were caused by going on campaign with too few troops. In view of the decision to commission Widukind to write this history and the subsequent diffusion of the Res gestae, which served as a basic source of information for other historians in the German kingdom, including Bishop Thietmar of Merseburg (1009–1018), it is reasonable to conclude that Widukind’s treatment of military affairs was consistent with the expectations of his patrons in the royal family. In particular, Widukind’s focus on siege warfare can be understood to be consistent with current understanding of military reality at the royal court. In this context, Widukind was operating under the important dictates of the rhetoric of plausibility that required that truth claims made in a text conform with the expectations of the audience.77

72 73 74 75 76 77

Gerhard, Vita Sancti Uodalrici, 1.3. Ibid., 1.3, 112. Ibid. Ibid., “certo sedulitate eius servitii cottidiani.” Ibid. For an important discussion of the ways in which the rhetoric of plausibility was understood by medieval authors operating around the turn of the millennium, see the recent study of Just C. Lake,

93

E A R LY O T T O N I A N WA R FA R E

Appendix List of military operations involving the defense or capture of fortifications in Widukind’s Res gestae: 1.23 1.24 1.27 1.27 1.32 1.35 1.35 1.35 1.36 1.36 1.38 2.4 2.6 2.11 2.11 2.11 2.11 2.11 2.14 2.15 2.19 2.22 2.24 2.27 2.36 3.3 3.3 3.3 3.4 3.8 3.8 3.10

Henry I’s defense of Eresburg against Eberhard of Franconia (915) Henry’s defense of the fortress at Grone against King Conrad I (915) Henry’s invasion of Swabia and the surrender of all of Duke Burchard’s urbes to the king (919) Henry’s invasion of Bavaria and siege of Regensburg (921) Hungarian invasion of Saxony leading to their burning of urbes and oppida while Henry I took refuge at the fortified palace at Werla (926) Henry I’s wintertime campaign against the Hevelli and siege of Brandenburg (928–929) Henry I’s siege of Jahna (Gana) (928–929) Henry I’s siege of Prague (928–929) Redarii sack of the fortress of Walsleben (929) German siege of Lenzen (929) Hungarian siege of unnamed fortification that led to battle of Riade (933) Otto I’s invasion of Slavic lands with intent to besiege unnamed urbs hostium (936) Duke Eberhard of Franconia’s siege of Helmern under the command of Bruning (937) Siege by Thankmar and Duke Eberhard of Franconia of the fortress at Belecke (938) Thankmar’s siege of Eresburg (939) Siege of the fortress of Laer by Otto I’s commander Dedi (939) Siege of Eresburg by Otto I (939) Second siege of Laer by Otto I (939) Hungarian attack on the fortress of Steterburg (939) Siege of Dortmund by Otto I (939) Siege of Merseburg by Otto I (939) Siege of Chèvrement by Otto I (939) Otto I’s siege of Breisach (939) Otto I’s siege of Immo in unnamed fort (939) Capture of Aquileia by Duke Henry of Bavaria (c. 947) Otto I’s siege of Laon (946) Otto I’s siege of Paris (946) Otto I’s siege of Rheims (946) Siege of Rouen by Otto I’s troops (946) Otto I’s capture of Nimburg on the Elbe (950) Otto I’s siege of Prague (950) Otto I’s siege of Pavia (951)

95

N A R R AT I V E W O R K S

3.18 3.21 3.34 3.36 3.43 3.44 3.45 3.50 3.50 3.51 3.52 3.63 3.68

Otto I’s siege and capture of numerous fortifications held by Conrad the Red in Lotharingia (953) Otto I’s siege of Mainz (953) –35 Otto I’s siege of Roßtal (954) Otto I’s siege of Regensburg (954) Otto I’s second siege of Regensburg (955) Hungarian siege of Augsburg (955) Markgraf Thiadric’s siege of an unnamed Slavic fortress (955) Capture of several urbes by the rebels Wichmann and Ekbert in Saxon marches (954/955) Capture by Duke Hermann of Saxony of the urbes held by the rebels Wichmann and Ekbert in Saxony (954/955) Duke Hermann of Saxony’s attack on the Abodrite fortress of Suithleiscranne (955) Slavic capture of unnamed German fortress (955) Two-year siege by Otto I of an unnamed fortress (St. Leo) held by Berengar of Ivrea (961–963) Siege by Duke Hermann of Saxony of an unnamed fortress held by the rebel Wichmann’s troops (date uncertain)

List of military campaigns resulting in battles in the field identified by Widukind in the Res gestae: 1.36 1.38 1.38 2.3 2.17 2.20 2.26 3.17 3.23 3.36 3.44 3.53

Battle of Lenzen (929) Saxon and Hungarian defeat of invading Hungarian army (933) Battle of Riade (933) Defeat of Saxon and Thuringian army in Bohemia (date uncertain) Battle of Birten (939) Defeat of a German army under the command of Haika by the Slavs (939) Battle of Andernach (939) Battle between the rebel Duke Conrad of Lotharingia and the Lotharingian levies supporting Otto I (953) Battle between Saxon reinforcements sent to Mainz by Duke Hermann of Saxony and a force commanded by Liudolf and Conrad the Red (953) Battle at Regensburg (954) Lechfeld (955) Battle of Recknitz (955)

Campaigns regarding which Widukind does not provide sufficient information in the Res gestae to determine whether a siege or a battle in the field took place: 1.40

Campaign against the Danes (935x936?) 96

N A R R AT I V E W O R K S

Widukind’s observations regarding the nature of early medieval warfare are substantially at odds with the traditional scholarly approach to Ottonian military history. It was not heavily armed, mounted warriors who dominated Ottonian warfare, according to the monk of Corvey, but rather fortifications, siege engines, and men fighting on foot. Widukind’s repeated emphasis on the centrality of sieges to warfare is consistent, however, with the enormous number of fortifications from the Ottonian period that have been unearthed through the efforts of archaeologists since the Second World War. To take but one example, dendrochronological dating of wooden ramparts in the Lausitz region, located between the Bobr and Kwisa rivers and the Elbe in the modern German states of Saxony and Brandenburg and in southwestern Poland (Lower Silesian Voivodeship) and the northern Czech Republic, has demonstrated that the construction or repair of these works corresponded exactly with the military campaigns of King Henry I and then by those of Otto I and his commanders, including Margrave Gero, to conquer the district.78 To put this matter simply, in their efforts to expand east of the Elbe and Saale river systems in the period 929–968, Henry and Otto were faced with the regular necessity of besieging significant fortifications and then leaving substantial garrisons behind to defend the frontiers and to hold the territory that they had captured. Most scholars who have worked at length with Widukind’s text, including specialists in military history, have concluded that the monk of Corvey was a well-informed and reliable source of information about current affairs. The work of archaeologists, who have excavated many hundreds of sites over the past century in Germany and her neighbors, has similarly validated Widukind’s focus on sieges as dominating contemporary warfare. The necessary conclusion, therefore, is that future scholarly investigations of Ottonian warfare must take as their starting point that the armies of the Ottonians were organized for capturing and defending fortress cities and lesser strongholds, with the concomitant need for large armies of foot soldiers who were equipped with considerable quantities of siege equipment. The Panzerreiter, in turn, should be relegated by scholars to the auxiliary role that they actually played in contemporary warfare.

“Truth, Plausibility, and the Virtues of Narrative at the Millennium,” Journal of Medieval History 35 (2009), 221–38. Regarding the imperative that authors writing about military matters present their views in a manner consistent with the expectations of their audiences, see David S. Bachrach, “Memory, Epistemology, and the Writing of Early Medieval Military History: The Example of Bishop Thietmar of Merseburg (1009–1018),” Viator 38 (2007), 63–90; Bernard S. Bachrach, “Dudo of Saint Quentin as a Military Historian,” The Haskins Society Journal: Studies in Medieval History, 12 (2002), 155–85; and Idem, “Dudo of Saint Quentin and Norman Military Strategy,” Anglo-Norman Studies, 26 (2004), 21–36. 78 Joachim Henning, “Archäologische Forschungen an Ringwällen in Niederungslage: Die Niederlausitz als Burgenlandschaft des östlichen Mitteleuropas im frühen Mittelalter,” in Frühmittelalterliche Burgenbau in Mittel und Osteuropa, ed. Joachim Henning and Alexander T. Ruttkay (Bonn, 1998), 9–29, here 25–6.

94

E A R LY O T T O N I A N WA R F A R E

2.20 Abodrite revolt (939) 2.34 Campaign under Duke Berthold of Bavaria against the Hungarians (Wels, 943) 2.36 Duke Henry of Bavaria’s campaign against the Hungarians (947) 3.30 Hungarian invasion of Bavaria and Franconia (954x954) 3.58 Otto I’s campaign against the Redarii (957) 3.59 –60 Wichmann’s revolt (958x959) 3.66 Two campaigns against King Mieszko of the Licicaviki in which Margrave Gero was victorious (963) 3.66 Margrave Gero’s defeat of the Lusiki (963) 3.72 Otto I’s campaign in Calabria (969)

97

5 M E M O RY, E P I S T E M O L O G Y, AND THE WRITING OF E A R LY M E D I E VA L M I L I TA RY H I S TO RY The example of Bishop Thietmar of Merseburg (1009–1018)

Introduction It is a truism that every modern scholar is a hostage to the quantity as well as the quality of the information provided by his sources. This is nowhere more evident than in the works of historians investigating the military history of Europe during the early Middle Ages. In the absence of large numbers of generally reliable administrative documents of the type familiar to specialists working in the high and late Middle Ages, particularly in the period after c. 1200, it is necessary to rely, in large part, on narrative texts composed by clerics to address questions such as military organization, tactics, and the conduct of battles.1 However, many of these clerics had an agenda or parti pris that led to the misrepresentation of military information.2 As a consequence, it is crucial for the historian seeking to understand what happened in a military sense in the early Middle Ages to evaluate 1 By contrast, other aspects of military history can be addressed through the analysis of material information developed by archaeologists in the context of excavations, including matters such as the planning involved in the construction of systems of fortifications. In this regard, several essays included in this volume address the relatively new model of “landscape archaeology,” as well as the human, material, and financial resources that were required for the construction of various types of fortifications. These factors naturally are of considerable importance in understanding the impact of both military planning and military expenditures on the society and economy as a whole. 2 The early medieval historian who has received the most sustained analysis in this regard is Gregory of Tours. See, for example, Walter Goffart, The Narrators of Barbarian History: Jordanes, Gregory of Tours, Bede, and Paul the Deacon (Princeton, 1988), particularly 174–83; and Bernard S. Bachrach, “Gregory of Tours as a Military Historian,” in The World of Gregory of Tours, ed. Kathleen Anne Mitchell and Ian N. Wood (Leiden, 2002), 351–63. This point also has been made regarding a wide range of Carolingian writers who dealt with military matters by Thomas Scharff, Die Kämpfe der Herrscher und der Heiligen: Krieg Und historische Erinnerung in der Karolingerzeit (Darmstadt, 2002).

98

M E M O R Y, E P I S T E M O L O G Y, A N D T H E W R I T I N G

not only the potential access of an author of a narrative work to likely reliable or accurate information of a military nature but also the likely reliability of the author in presenting information to which he had access. Walter Goffart certainly was not the first scholar to emphasize that early medieval historical narratives must be treated as works of literature rather than as plain text in which medieval writers naively transmitted facts about contemporary affairs. His Narrators of Barbarian History nevertheless has had enormous influence on scholarly views about the proper manner in which to approach narrative texts in the decades since it first appeared in 1988. Indeed, in his summary address to a gathering of leading specialists in early medieval historiography and textual criticism at Zwettl in 1994, Patrick Geary discussed the “prison house” into which Goffart has placed every scholar who seeks to discern the history behind the early medieval text, that is to reconnect historiography to history.3 Geary noted that scholars have emphasized a variety of techniques to escape from the “prison” of the text, including identifying authorial intention through an analysis of the author’s biography, identifying the audience for the text and the concomitant expectations within which the author was forced to work, and identifying the manner in which a particular text was used during the Middle Ages through a careful examination of its manuscript tradition.4 Geary’s summary review of the more than two dozen studies presented at Zwettl reiterated the particularly important epistemological questions that Goffart examined in detail six years before in Narrators and provided a caution that each of the methods identified by scholars to address these questions posed dangers if pursued in isolation.5 Although the “Goffart challenge” of identifying the biography of medieval authors, the composition of the audiences for their histories, and the purposes for which these texts were deployed all pose challenges to the historian, numerous scholars have successfully used these techniques to negotiate the complex relationship between narrative texts and the historical reality they purport to depict. Over the past decade, however, Johannes Fried, professor of medieval history at the Wolfgang Goethe Universität of Frankfurt and president of the Association of Historians of Germany, has posed in a series of publications what he describes as a

3 Patrick Geary, “Zusammenfassung,” in Historiographie im frühen Mittelalter, ed. Anton Scharer and Georg Scheibelreiter (Vienna, 1994), 539–42. 4 Ibid. Geary also emphasized the difficulties inherent in each of these approaches if pursued in isolation. He suggests, for example, that scholars must be wary of “constructing” the audience for the text on the basis of the text, itself. 5 Patrick Geary, Phantoms of Remembrance: Memory and Oblivion at the end of the First Millennium (Princeton, 1994), takes up many of these same epistemological questions regarding not only narrative texts but charters as well. Many of the themes discussed at the Zwettl conference in 1994 benefited from further discussion in the collection of articles in Medieval Concepts of the Past: Ritual, Memory, Historiography, ed. Gerd Althoff, Johannes Fried, and Patrick J. Geary (Cambridge, 2002). See, in particular, Patrick Geary, “Oblivion Between Orality and Textuality in the Tenth Century,” 111–22.

99

N A R R AT I V E W O R K S

fundamental challenge to the very practice of history writing by modern scholars.6 In Fried’s view, it is not only the epistemology of modern historians that is problematic, it is the very ability and desire of medieval writers to obtain and transmit accurate information that is at issue. Focusing on what he calls the medieval “oral society,” Fried insists that modern scholars necessarily pass into the realm of fantasy when dealing with medieval narrative texts because scholars must rely on a written text that cannot, by its very nature, convey reality.7 Fried goes even further, arguing that there is no such thing as a simple “fact.” In respect, for example, to the statement that Emperor Frederick II died in 1250, Fried asserts, every [fact] requires a more or less complicated means of discovery. Every perception and bit of knowledge requires a constant, conscious intellectual effort, that is a series of abstract processes as well as verbalized and ideal connections between objects that are not bodily present at the same time.8

6 Johannes Fried, “Die Kunst der Aktualisierung in der oralen Gesellschaft: Die Königserhebung Heinrichs I. als Exempel,” Geschichte in Wissenschaft und Unterricht 44 (1993), 493–503; “Die Königserhebung Heinrichs I. Erinnerung, Mündlichkeit und Traditionsbildung im 10. Jahrhundert,” in Mittelalterforschung nach der Wende 1989, ed. Michael Borgolte (Munich, 1995), 267–318; “Wissenschaft und Phantasie. Das Beispiel der Geschichte,” Historische Zeitschrift 263 (1996), 291–316; “Erinnerung und Vergessen. Die Gegenwart stiftet die Einheit der Vergangenheit,” Historische Zeitschrift 273 (2001), 561–93; and Die Schleier der Erinnerung: Grundzüge einer historischen Memorik (Munich, 2004). 7 Fried, “Wissenschaft und Phantasie,” 296–7. It should be emphasized that although he makes universal claims about the “fantastical nature” of early medieval narrative texts, Fried himself has not engaged these numerous texts, at least those dealing with Ottonian history on an individual basis. Rather, he has restricted his actual investigation, or at least his published views, to a single narrative work, namely Widukind’s Res gestae Saxonicae, as in the previous note. Fried’s views regarding this work have not been well-received by other specialists in Ottonian history and historiography. See, for example, the survey of this question by Johannes Laudage, “Widukind von Corvey und die Deutsche Geschichtswissenschaft,” in Von Fakten und Fiktionen: Mittelalterliche Geschichtsdarstellungen und ihre kritische Aufarbeitung, ed. Johannes Laudage (Cologne, 2003), 193–224, especially 209–24; Markus Brömel, “Widukind von Corvey: Kronzeuge und Herausforderung,” in Auf den Spuren der Ottonen III (Halle, 2002), 17–25, here18; Gerd Althoff, “Geschichtsschreibung in einer oralen Gesellschaft. Das Beispiel des 10. Jahrhunderts,” in Ottonische Neuanfänge. Symposium zur Ausstellung ‘Otto der Große, Magdeburg und Europa’, ed. Stefan Weinfurter and Bernd Schneidmüller (Mainz, 2001), 151–69; and Hagen Keller, “Widukinds Bericht über die Aachener Wahl und Krönung Ottos I.,” Frühmittelalterliche Studien 29 (1995), 390–453, especially 406–10. For a more positive portrayal of Fried’s efforts to call into question the value of Widukind on the basis of the problem of “memory”, see David A. Warner, “Ritual and Memory in the Ottonian Reich: The Ceremony of Adventus,” Speculum 76 (2001), 255–83, here, 258–60. Warner does not, however, appear to have considered the criticisms Keller made regarding Fried’s conclusions and does not cite Keller’s study, “Widukinds Bericht,” published in 1995. 8 Fried, “Wissenschaft und Phantasie,” 300, “jede verlangt nach mehr oder weniger komplizierten Erkenntnisweisen, jedes Wahrnehmen und Wissen erfordert ein fortgesetztes gedankliches vorAugen stellen, eine Serie von Abstrakt prozessen, verspraclichungen und ideelen Verbindungslinien zwischen körperlich nicht gegenwärtigen Gegenständen.”

100

M E M O R Y, E P I S T E M O L O G Y, A N D T H E W R I T I N G

With respect to the ability of medieval authors to transmit accurate information about the past, Fried begins with the manifestly true statement, long understood by cognitive psychologists, that people in the present, and in the past as well, forget most things they experience and also remember selectively in order to function in the present.9 Fried also emphasizes that memories that are not anchored, either by the force of the experience itself or in writing, tend to change.10 It is in this context of the fragility of memory that Fried asserts there were no outside controls during the early Middle Ages on those who wrote down information in historical narratives.11 As a consequence, Fried concludes, the overwhelmingly “oral” society of the early Middle Ages was incapable of maintaining and then transmitting an accurate record of events.12 If taken to their logical conclusion, of course, Fried’s arguments concerning the inability of historians to glean “facts” from medieval narrative sources, much less to use these sources to ask questions about the societies in which the texts were composed, would pose serious complications in the conduct of affairs in the modern world. Leaving aside for the moment Fried’s broad-gauged and entirely undemonstrated claims about the essential “orality” of the early medieval world, it is quite clear that modern business, law, and government operate in large measure on the gathering of information from oral sources – information that is “organized” for presentation in written form at some later date.13 A wellknown example of this method of information gathering is the activity of police detectives who obtain reports from witnesses and only subsequently refine the

9 Fried, “Erinnerung und Vergessen,” 561–3. This is also the gravamen of Fried, Die Schleier der Erinnerung, 80–172. 10 Fried, “Erinnerung und Vergessen”; and idem, Die Schleier der Erinnerung, 270–91, and 313–32. 11 Fried, “Erinnerung und Vergessen,” 575. 12 Ibid., 584–5. 13 Fried’s contention regarding the “orality” of society and particularly the effect that such “orality” had on the writing of history has been criticized with devastating clarity by Keller, “Widukinds Bericht,” 406–10 where Keller (n.87) draws attention to the danger of treating all “schriftlose Kulturen” as if they were all the same “orale Kultur”; and Althoff, “Geschichtsschreibung in einer oralen Gesellschaft,” 151–69. The extent to which Ottonian society and its Carolingian predecessor was essentially oral in the conduct of public affairs has generated a substantial literature with which Fried fails to concern himself in his arguments regarding the composition of narrative texts. See, for example, Rosamond McKitterick, “Schriftlichkeit im Spiegel der frühen Urkunden St. Gallens,” in Das Kloster St. Gallen im Mittelatler: Die kulturelle Blüte vom 8. bis zum 12. Jahrhundert, ed. Peter Ochsenbein (Stuttgart, 1999), 69–82; and idem, Carolingians and the Written Word (Cambridge, 1995). With regard to the use of writing in the Ottonian period, see idem, “Bischöfe und die handschriftliche Überlieferung des Rechts im 10. Jahrhundert,” in Mönchtum-Kirche-Herrschaft 750–1000: Josef Semmler zum 65. Geburtstag, ed. Dieter R. Bauer (Sigmaringen, 1998), 231–42; Mark Mersiowsky, “Regierungspraxis und Schriftlichkeit im Karolingerreich: Das Fallbeispiel der Mandate und Briefe,” in Schriftkulturund Reichsverwaltung unter den Karolinger, ed. Rudolf Schieffer (Opladen, 1996), 109–66; and Eric J. Goldberg, Struggle for Empire: Kingship and Conflict under Louis the German, 817–876 (Ithaca, 2006), especially, 18–19, and 210.

101

N A R R AT I V E W O R K S

information they have obtained into a formal written report. This material then frequently is admitted into evidence in court proceedings through the testimony of police officers.14 Business consultants operating in large firms carry out their preliminary investigations in a similar manner.15 Simply put, in focusing on what humans forget, Fried fails to account for the substantial amount of information that humans remember – accurately. Indeed, even if one were to accept, against the findings of scholars including Rosamond McKitterick, Mark Mersiowsky, and Eric Goldberg, whose works are cited earlier, that Carolingian and Ottonian societies were essentially oral in the conduct of public affairs, of which history writing was an important part, the cognitive psychological models upon which Fried builds his case for epistemological pessimism regarding memory are far more nuanced than he would have his readers believe. As Milman Parry made abundantly clear in his path-breaking studies regarding the oral composition of epic poems in Yugoslavia during the 1920s and 1930s, nonliterate performers are able to memorize thousands of half-lines and to reproduce these half-lines in multiple performances, albeit with some variations in emphasis and order in the narration as a whole.16 Indeed, this also is the gravamen of Mary Carruthers’s work on medieval memory, which Fried cites in the bibliography of Die Schleier der Erinnerung but does not deal with in the text.17 Historians working primarily with oral sources of information in nonliterate societies have also found that certain kinds of information, including the information provided

14 The similarity in the investigative techniques of historians and police officers long has been understood by specialists in historical epistemology. A classic discussion of the relationship between the two disciplines is provided by R. G. Collingwood, The Idea of History (Oxford, 1956, 5th repr. 1977), 249–82, with a particular emphasis 266–8. The later editions of the Idea of History include additional material from Collingwood’s 1938 lectures that is not pertinent to the present discussion. 15 I thank Daniel G. Bachrach, professor of management at the Culverhouse College of Commerce and Business Administration at the University of Alabama, for bringing to my attention the enormous reliance placed by both scholars and business consultants on the results of oral surveys and interviews that only subsequently are redacted into written form. In this regard concerning the theoretical design and practical implementation of oral surveys and the extraordinary value of the data extracted from these surveys for both scholarly research and business consultation, see Edgar H. Schein, Process Consultation Revisited: Building the Helping Relationship (Reading, MA, 1999). For medieval scholars unfamiliar with the field of management, it should be noted that E. H. Schein is the distinguished professor of management at the Sloan School of Business at the Massachusetts Institute of Technology. With respect to the ubiquitous use of orally collected data for the making of business decisions worth hundreds of millions of dollars each year in the United States, see D. Stokes and R. Bergin, “Methodology or ‘Methodolatry’? An Evaluation of Focus Groups and In-Depth Interviews,” Qualitative Market Research 9 (2006), 26–37; and L. T. Wright, “Exploring the In-Depth Interview as a Qualitative Research Technique with American and Japanese Firms,” Marketing Intelligence and Planning 14 (1996), 59–64. 16 Milman Parry’s research was published after his death in The Making of Homeric Verse: The Collected Papers of Milman Parry, ed. Adam Parry (London, 1971). 17 Mary Carruthers, The Book of Memory: A Study of Memory in Medieval Culture (Cambridge, 1990, repr. 1999).

102

M E M O R Y, E P I S T E M O L O G Y, A N D T H E W R I T I N G

in poetry but also proverbs, songs, genealogies, lists of place names, and judicial precedents, can shed considerable light on earlier events that extend far beyond the past few years.18 Just as importantly in an epistemological context, it must be understood that Fried’s assault on the putative failure of memory of medieval writers is, in fact, a variation on the argument that the reconstruction of events according to the parti pris of the author creates an insurmountable barrier to historical research. Indeed, Fried makes this point when arguing in a study concerned with “memory” that writers in the Carolingian period were concerned with the “true” event rather than with the “real” event, that is, with the way things should have been rather than the way that they were.19 In making this claim, Fried echoes the arguments put forth some thirty years earlier by Amos Funkenstein and Roger Ray.20 By denying that there was any outside control on narrative sources because of the inability of anyone to remember, or rather to demonstrate, a correct memory of the past, Fried attempts to exclude an appeal to the audience of a text as a means of escaping the “prison” of the text as this term has been deployed by Geary. Fried’s position regarding the lack of controlling authorities on the production of Ottonian narrative texts, however, has been disputed quite effectively by other specialists in this field. Gerd Althoff, for example, has emphasized that far from operating in an unsupervised environment, Ottonian writers faced the necessity of writing in a manner that conformed with the expectations of their patrons and audience, often composed of leading members of the Ottonian family and court. As Althoff makes clear, these people had a very strong sense of how their own history had played out and insisted that their view of the past be presented accurately, i.e. they had a fixed view of the past.21 Hagen Keller drew attention to the danger inherent in Fried’s broad-gauged but fundamentally unsubstantiated claims by noting that Fried’s model of an oral society in which there were no fixed views about the past was based on a single anthropological study carried out by Jack Goody among the Gonja people in northern Ghana in 1956–1957.22 With respect to Fried’s assertion that Widukind, and all other early medieval authors,

18 The methods used by historians specializing in obtaining information in oral societies is discussed in considerable detail by Jan Vansina, Oral Tradition as History (Madison, 1985). Vansina focuses his research on sub-Saharan Africa but the methodologies that he describes are applicable to a range of non-literate societies. 19 Fried, “Erinnerung und Vergessen,” 575. 20 Amos Funkenstein, Heilsplan und natürliche Entwicklung: Formen der Gegenwartsbestimmung im Geschichtsdenken des hohen Mittelalters (Munich, 1965), particularly 70–7; and Roger D. Ray, “Medieval Historiography through the Twelfth Century: Problems and Progress of Research,” Viator 5 (1974) 33–60, here 43–57. 21 Althoff, “Geschichtsschreibung in einer oralen Gesellschaft,” 154–9. 22 Keller, “Widukinds Bericht,” 408. In this context, it should be emphasized that in the extensive notes and bibliography of Die Schleier der Erinnerung Fried does not cite any of the numerous works by Jan Vansina, the leading specialist in African oral history, whose theoretical studies dealing with the use of oral sources provide much of the intellectual framework for modern oral history.

103

N A R R AT I V E W O R K S

were not bound by any controlling authority, Keller observed, “Even if it is the case that Widukind’s depiction is not intended to provide a ‘precise enumeration of events’, it is anything but the written form of a history that was in flux among his contemporaries.”23 Nevertheless, having made his case for the impenetrability of medieval narrative sources to his own satisfaction, Fried has not deviated in his most recent publications from a monolithic view of early medieval “oral society” and, as a result, seems unable or unwilling to provide a single example in which it is possible to pass beyond historiography and use a narrative text to ask historical questions.24 Curiously, however, in pursuing his doctrinaire position regarding the inability of scholars to use the information provided in narrative works to ask historical questions, Fried himself fails to ask the most basic epistemological questions facing any trained historian. These include: Does a particular author provide accurate information about particular facts? What access did the author have to particular information? How can the answers to these two questions be known? Fried thus ignores the ability of the historian to corroborate information provided by a narrative source against other, independently composed narrative texts. He also fails to discuss the wide range of other sources of information available to the trained scholar, including other types of texts, e.g. charters and administrative documents, and information of a material nature such as that generated by archaeologists. In this context, Fried does not give any indication in his published works that he is familiar with or understands the controls on a narrative text that can be gleaned from deploying the methods of Sachkritik pioneered by Hans Delbrück in his multivolume history of Western warfare.25 Delbrück himself was interested most particularly in using the physical realities of battlefields to criticize the numbers of soldiers reported in ancient and medieval narrative texts, certainly a question that might be of interest when critiquing the historical value of information provided in narrative texts.26 Delbrück’s emphasis on investigating material reality 23 Keller, “Widukinds Bericht,” 408, “Auch wenn Widukinds Darstellung keine ‘präzisen Geschehensprotokolle” bieten will, so ist es doch alles andere als die Verschriftung von Geschichte, die bei den Zeitgenossen in Umlauf waren.” 24 Fried, Die Schleier der Erinnerung, 173–200, provides a series of examples of medieval written works, apparently taken at random, through which he attempts to demonstrate that they are unreliable in their transmission of information. It should be emphasized that Fried does not devote more than four pages, and usually considerably less, to any single one of these texts. His observations regarding these narrative works can, in no way, be understood to represent the type of exhaustive study that is required to be able to be able to draw epistemological conclusions regarding the value of these texts for answering particular historical questions. 25 Hans Delbrück, Geschichte der Kriegskunst im Rahmen der politischen Geschichte, 6 vols. (Berlin, 1900–1936). The first three volumes, dealing with ancient and medieval warfare, are now available in English translation by Walter J. Renfroe, History of the Art of War within the Framework of Political History, vols. 1–3 (Westport, 1975–1982). 26 Regarding Delbrück’s work in this area, and a critique of his methods as they regard the numbers of soldiers in medieval armies, see Bernard S. Bachrach, “Early Medieval Military Demography: Some Observations on the Methods of Hans Delbrück,” in The Circle of War in the Middle Ages:

104

M E M O R Y, E P I S T E M O L O G Y, A N D T H E W R I T I N G

has inspired considerable research among specialists in ancient history, although most medieval historians, along with Fried, have not taken advantage of his methodological insights.27 Fried also gives no indication in his published studies that he is familiar with the epistemological concept of “objective facts” and their division into “brute fact” and “social facts.”28 In his work, Fried misuses the term “brute fact, ” as this term commonly is used by philosophers to refer to fixed dates, e.g. Charlemagne was crowned on Christmas Day 800.29 “Brute fact,” in technical epistemological usage among Anglophone scholars, refers to those elements of reality that exist independently of what humans think of them, e.g. hydrogen atoms have one electron.30 Dates, by contrast, are “social facts” which depend for their existence on human thought.31 The concept of “brute fact” is exceptionally important in Sachkritik because the existence of a reality that does not depend on human thought, e.g. the number of calories required on a regular basis for a human male to undertake sustained heavy labor such as marching with 40 kilograms of equipment or digging ditches, creates the physical and largely immutable framework in which human action takes place. “Brute fact” also is important in the reading of narrative texts because authors such as Bishop Thietmar of Merseburg (1009–1018), who is the focus of the following study, frequently provide information about the material reality of their time that existed independently of human thought about this same material reality. For the purpose of coming to grips with the “brute facts” that conditioned the world in which Thietmar lived, it is not important whether or not Thietmar mentions in his text that there were numerous fortifications in the Slav-inhabited regions east of the Elbe river. The stone foundations and wooden remains of these fortifications exist and would still exist if Thietmar’s Chronicon had been lost. That Thietmar correctly states that these structures existed in his own day, however, certainly permits a modern scholar to observe that the bishop of Merseburg

27

28 29 30 31

Essays on Medieval Military and Naval History, ed. Donald J. Kagay and Andrew L. J. Villalon (Woodbridge, 1999), 3–20. Perhaps the single most influential work in ancient military history to use the methods of Sachkritik is Donald Engels, Alexander the Great and the Logistics of the Macedonian Army (Berkeley, 1978), which develops systematic models for evaluating the caloric requirements of the soldiers and animals deployed by Alexander the Great during his campaigns. Engel’s study has revolutionized the way in which ancient historians evaluate the narrative descriptions of military campaigns. The methods of Sachkritik are now basic to scholarship in ancient military history. In this context, see the very important work of Marcus Junkelmann, Die Legionen des Augustus: Der römische Soldat in archäologische Experiment (Mainz, 1986); and Jonathan P. Roth, The Logistics of the Roman Army at War (264 B.C.–A.D. 235) (Leiden, 1999). For a clear statement of the “realist” view regarding the nature of “objective facts” see John R. Searle, The Construction of Social Reality (New York, 1995). Fried, “Die Schleier der Erinnerung,” 573. Searl, Construction of Social Reality, 2 and passim. Ibid., 26.

105

N A R R AT I V E W O R K S

made accurate truth claims about his material reality.32 From an epistemological perspective, it is the interpretation of the stone and wooden structures as “fortifications” which is a matter of “social fact.” One final point that must be made regarding Fried’s categorical dismissal of narrative texts from the historian’s pool of sources is that he fails to distinguish among the types of information that the authors of narrative texts provide. As all historians should be aware, in every narrative text some information pertains directly to the parti pris or narrative arc that the author is presenting to his audience. Other information provides the material context in which the primary story takes place. That an author shapes certain pieces of information to fit his parti pris, through omission, emphasis, or outright falsification, does not entail that all of the information provided by the author is shaped in this manner. Indeed, as scholars working with a nuanced understanding of the epistemology of historical knowledge have made clear, in the pursuit of rhetorical plausibility, it is in the author’s interest to provide a context of accurate information, that is “brute facts” and “social facts” that are acceptable to an informed audience, precisely in those places in his narrative where he hopes to persuade an audience to accept information that is colored by his parti pris.33 In sum, Fried’s broad-gauged attack on narrative sources amounts to an abnegation of the scholar’s responsibility to test the reliability of every source on an individual basis regarding the specific information that the scholar seeks to use to answer a historical question. Nevertheless, the critiques made by Fried and Goffart regarding early medieval narrative texts have significant value because they have forced scholars to be even more aware of the weaknesses as well as the strengths of their sources. It is not enough to identify only the author’s biography, or the audience, or the reception of the text – all of these questions must be considered and answered to the extent possible before the information provided by an

32 Thietmar makes very frequent references to fortifications located within the German kingdom and beyond its eastern frontiers, including places such as Lebusa, Glogau, Lenzen, and Prague. The existence of hundreds of wooden and stone fortifications throughout the eastern frontier regions of the German kingdom as well as in Bohemia and Poland has been demonstrated by scores of archaeological explorations. For a thorough account of the state of the question regarding the archaeology of these central European fortifications from the early medieval period, see Frühmittelalterliche Burgenbau in Mittel- und Osteuropa: Tagung Nitra vom 7. bis 10. Oktober 1996, ed. Joachim Henning and Alexander T. Ruttkay (Bonn, 1998), which includes thirty-eight studies dealing with scores of excavations. Also see the catalog of Ottonian fortifications long the northeastern frontiers in David S. Bachrach, “Toward an Appraisal of the Wealth of the Ottonian Kings of Germany, 919–1024,” Viator 44.2 (2013), 1–27, here 25–7. 33 This point has recently been made with respect to Carolingian sources specifically dealing with military matters by Scharff, Die Kämpfe der Herrscher, 42. Also see in this regard, with respect to Thietmar’s direct contemporary Dudo of St. Quentin, Bernard S. Bachrach, “Dudo of St. Quentin as an Historian of Military Organization,” The Haskins Society Journal 12 (2002), 165–85, here 166–7, and passim, and in this volume. Also see the important discussion by Justin C. Lake, “Truth, Plausibility, and the Virtues of Narrative at the Millennium,” Journal of Medieval History 35 (2009), 221–38.

106

M E M O R Y, E P I S T E M O L O G Y, A N D T H E W R I T I N G

author can be used to answer a particular historical question. Once these questions have been answered, it is then possible to begin the process of evaluating the information that these narrative texts provide regarding particular historical questions. In light of the previous discussion, the present study is intended to lay the foundation for using the Chronicon of Bishop Thietmar of Merseburg (1009–1018) to write Ottonian military history. The first section presents the findings of modern scholars regarding the biography, audience, purpose, and reception of Thietmar’s text with a concomitant evaluation of the accuracy of the information that he provides. To be clear, accuracy in this context means the precision with which the “brute facts” and “social facts” dealt with by Thietmar conform to the material reality of his time, as this can be ascertained through comparative analysis of all surviving sources of information, both written and material, and the deployment of the methods of Sachkritik. The second part of this study considers what Thietmar has to say about milites, whom historians generally agree are crucial to our understanding of Ottonian warfare.34 34 The popularization of the importance of milites as the central figures in early medieval warfare generally, and in Ottonian warfare, in particular, can be traced to the exceptionally influential study by Heinrich Brunner, “Der Reiterdienst und die Anfänge des Lehnwesens,” Zeitschrift der Savigny-Stiftung für Rechtsgeschichte, Germanistische Abteilung 8 (1887), 1–38, reprinted in Brunner, Forschungen zur Geschichte des deutschen und französischen Rechts (Stuttgart, 1894), 39–74; and the restatement of this view in Grundzüge der deutschen Rechtsgeschichte (Leipzig, 1901). See the citation to the seventh edition published after Brunner’s death by Ernst Heymann (Munich, 1919, and repr. Munich 1930), particularly 68. For the discussion of this topic in the period before the Second World War, see Hans Fehr, “Das Waffenrecht der Bauern im Mittelalter,” Zeitschrift der Savigny Stiftung für Rechtsgeschichte: Germanistische Abteilung 35 (1914), 111–21; Hans Delbrück, Medieval Warfare: History of the Art of War III, trans. W. J. Renfroe (Lincoln, 1982), 365–75, originally published under the title Geschichte der Kriegskunst im Rahmen der politischen Geschichte, 2nd edition (Berlin, 1923); and Paul Schmitthenner, “Lehnskriegwesen und Söldnertum im abendländischen Imperium des Mittelalters,” Historische Zeitschrift 150 (1934), 229–67. For the period after the Second World War, see J. F. Verbruggen, “La tactique militaire des armées de chevaliers,” Revue du nord 29 (1947), 161–80; Otto Brunner, “‘Bürgertum’ und ‘Feudalgewalt’ in der europäischen Sozialgeschichte,” Geschichte in Wissenschaft und Unterricht 7 (1956), 599–614 and reprinted in Die Stadt des Mittelalters, 3 vols., ed. Carl Haase (Darmstadt, 1969–1973), III: 480–502, here 491–2, and 498; Heinrich Spromberg, “Die feudale Kriegskunst,” Beiträge zur belgisch-niederländischen Geschichte: Forschungen zur mittelalterichen Geschichte 3 (1959), 30–55; Karl Leyser, “The Battle at the Lech, 955: A Study in Tenth-Century Warfare,” History 50 (1965), 1–25 and reprinted in Medieval Germany and Its Neighbours, 43–67 (latter pagination used in this study); idem, “Henry I and the Beginnings of the Saxon Empire,” The English Historical Review 83 (1968), 1–32 and reprinted in Karl J. Leyser, Medieval Germany and Its Neighbours 900–1250 (London, 1982), 11–42. All citations are to the latter publication; and idem, “Early Medieval Warfare,” in Communications and Power in Medieval Europe: The Carolingian and Ottonian Centuries, ed. Timothy Reuter (London, 1994), 29–50; Leopold Auer, “Zum Kriegswesen unter der früheren Babenbergern,” Jahrbuch für Landeskunde von Niederösterreich 42 (1976), 9–25, esp. 16–19; and idem, “Formen des Krieges im abendländischen Mittelalter,” in Formen des Krieges, ed. Manrfried Rauchensteiner and Erwin A. Schmidl (Graz, 1991), 17–43, esp. 19–23; Wilhelm Rautenberg, “Ritter und Rotten: Zur begrifflichen und funktionalen Unterscheidung des geworbenen Kriegsvolks im Hochmittelalter,” Jarhbuch der Gesellschaft für niedersächische Kirchengeschichte 76 (1978), 87–121, especially 88–101; Hagen Keller,

107

N A R R AT I V E W O R K S

Thietmar of Merseburg and his history Born on 25 July 975 the third son of the count of Walbeck, Thietmar was educated for a career in the church, first at Quedlinburg and then at Magdeburg, where he studied at the monastery of St. John, and then at the cathedral school.35 His career culminated in 1009 when he was selected by King Henry II (1002–1024) to serve as bishop of Merseburg. Thietmar wrote his Chronicon at Merseburg between 1013 and his death in December 1018 with the aid of no fewer than eight episcopal clerks.36 The large number of trained writers who worked on the manuscript of the Chronicon in this brief five-year period is itself of some interest in the discussion about the essential “orality” of early medieval society. One might question why in a period in which all important matters supposedly were dealt with only or primarily by the spoken rather than the written word, it would be necessary for a relatively poor frontier bishop such as Thietmar to have so many expensively trained writers available for work on his history. One might also ask what these same writers did when not taking dictation for sections of the Chronicon. These questions, however, must await a different study. The manuscript of the Chronicon remained in the cathedral library of Merseburg until 1091, when Bishop Werner (1059–1093) gave it as a gift to the newly reestablished monastery of St. Peter located at Altenburg, just outside the city walls.37 The manuscript was kept there until c. 1160 when it was lent to the Benedictine monastery at Corvey, where it was copied and augmented with additional material from the work of the famous historian of that house, Widukind (died c. 975).38 The autograph manuscript of Thietmar’s Chronicon then was returned to St. Peter at Altenburg, where it remained until the early sixteenth century, suffering significant damage to fifteen folios.39 The text of the Chronicon was first printed in 1580 and thereafter attracted significant attention, as it was published in at least four further editions and translated into German several times by the beginning of the nineteenth century.40 Thietmar’s focus on the kings of the Ottonian dynasty earned him an honored place in nineteenth-century German historiography. In 1858, Wilhelm Wattenbach

35

36 37 38 39 40

“Reichsstruktur und Herrschaftsauffasung in ottonisch-frühsalischer Zeit,” Frühmittelalterliche Studien 16 (1982), 74–128, esp. 82; and Gerd Althoff and Hagen Keller, Heinrich I. und Otto der Grosse: Neubeginn auf karolingischem Erbe, 2nd edition, vol. 1 (Göttingen, 1994), 87. Concerning Thietmar’s family and education, see Nicole Lederer-Herkenhoff, “Thietmar von Merseburg,” in Weltbild und Realität: Einführung in die mittelalterliche Geschichtsschreibung, ed. Ulrich Knefelkamp (Bamberg, 1992), 111–20, specifically, 111–12; and Helmut Lippelt, Thietmar von Merseburg: Reichsbischof und Chronist (Cologne, 1973), 46–58, and 71–87. Robert Holtzmann, “Über die Chronik Thietmars von Merseburg,” Neues Archiv der Gesellschaft für ältere deutsche Geschichtskunde 50 (1935), 159–209, here 168. Concerning the provenance of the manuscript of the Chronicon, see Holtzmann, “Chronik,” 161–5. Ibid., 163; and Lippelt, Thietmar, 185. Holtzmann, “Chronik,” 163. Lippelt, Thietmar, 1–2.

108

M E M O R Y, E P I S T E M O L O G Y, A N D T H E W R I T I N G

wrote: “Thietmar’s conception of history is very healthy and truly German. He is a proud Saxon who loves his ruling family and experiences pain when he sees kings going along the wrong path.”41 This view of Thietmar as an honest and German historian was fully shared by Robert Holtzmann, who accepted Wattenbach’s views regarding Thietmar in a revised edition of Wattenbach’s study in source criticism published in 1938.42 Holtzmann, whose edition of Thietmar’s Chronicon remained the basic text until it was revised and substantially annotated by Werner Trillmich, also emphasized that Thietmar relied to a great extent on now-lost charters (perdita) to write about many of the bishopric’s losses and gains of property.43 Although the lessons of the Second World War have tended to diminish the emphasis on Thietmar’s Germanness, more recent historiography has affirmed Thietmar’s status as a capable historian who also was exceptionally well informed about contemporary affairs.44 Helmut Lippelt, whose monograph on Thietmar remains the standard work in this area, defends the basic reliability of the Chronicon.45 Nevertheless, Lippelt does not treat Thietmar as a transparent transmitter of information. Rather, he recognizes that the bishop focused on those matters that were of interest to him in his episcopal office or because they involved his friends, relatives, or enemies.46 Lippelt points out, for example, that Thietmar discusses King Henry I’s decision to fortify the frontier settlement at Meißen because a century later Thietmar himself was required by Henry II to provide soldiers to garrison the fortress there.47 On the other hand, Thietmar did not stress the major revolts

41 Wilhelm Wattenbach, Geschichtsquellen in Mittelalter bis zur Mitte des dreizehnten Jahrhunderts (Berlin, 1858) revised by Robert Holtzmann, Deutschlands Geschichtsquellen im Mittelalter: Deutsche Kaiserzeit, 2nd edition, 4 vols. (Tübingen, 1948) I: 57. The first edition of Holtzmann’s revision of Wattenbach’s study was published in 1938. Regarding the positive views of Thietmar in the nineteenth century, see Lippelt, Thietmar, 138–9. 42 Holtzmann’s acceptance of Wattenbach’s favorable views of Thietmar is discussed by Lippelt, Thietmar, 138. Thietmar von Merseburg, Chronik, ed. and trans. Werner Trillmich (Darmstadt, 1974). The newest edition is now Thietmar von Merseburg, Chronik, ed. and trans. Werner Trillmich, 8th edition (Darmstadt, 2002). All citations to the Latin text of Thietmar are to this Trillmich edition. The English translation is Ottonian Germany: The Chronicon of Thietmar of Merseburg, trans. and annotated by David Warner (Manchester, 2001). 43 Wattenbach, I: 57. Regarding Thietmar’s use of charters to obtain information that he subsequently incorporated into his history, see ibid., 59 and 68. 44 For a critique of the “echt Deutsch” label attached to Thietmar, see Lippel, Thietmar, 117. 45 In this context, reliability means that Thietmar provides information about brute facts and social facts that accurately represents the objective material reality of his own time, as accuracy and objective material reality are defined by specialists in epistemology. On this point, see Searl, Social Construction of Reality, passim. 46 It is now recognized that it was common for medieval German writers to use their historical works as a means of remembering their friends and relatives, particularly those who had died. In this regard, see Lutz E. von Padberg, “Geschichtsschreibung und kulterelles Gedächtnis: Formen der Vergangenheitswahrnehmung in der hochmittelalterlichen Historiographie am Beispiel von Thietmar von Merseburg, Adam von Bremen und Helmold von Bosau,” Zeitschrift für Kirchengeschichte 105 (1994), 156–77, here 158 and the literature cited there. 47 Lippelt, Thietmar, 145.

109

N A R R AT I V E W O R K S

against King Otto I, although they were of considerable importance to his source, Widukind, presumably because Thietmar was not as interested in these events as was the Saxon monk.48 Lippelt also recognizes that Thietmar, on occasion, shaped his account to his own advantage or to the advantage of his diocese. This was particularly true when Thietmar dealt with specific people or with situations that were related to the reestablishment of the diocese of Merseburg.49 Lippelt notes, for example, that in discussing Henry II’s dealings with Archbishop Giselher of Magdeburg (981–1004), who as the bishop of Merseburg had acquiesced in the elimination of his own diocese, Thietmar emphasized Giselher’s negative qualities.50 It is important to stress in this context, however, that Lippelt, as well as later scholars who have noted this particular episode, nevertheless emphasize that once Thietmar’s parti pris is understood, it is possible to use the information he provides to gain an understanding of contemporary affairs.51 Indeed, it is Thietmar’s rather obvious parti pris that has enabled scholars to identify those areas where the bishop is likely to have shaded the reality of events purposefully to suit his own ends. The comment of Philip Buc regarding Liudprand of Cremona, who wrote historical works at the German court in the generation before Thietmar, are particularly apropos on this point. Noting the earlier negative reception of the works of Liudprand by modern scholars, Buc concludes: Generations of positivitistic historians found this quest for vengeance one of the grounds to dismiss the “Antapodosis” (like Liutprand’s two other surviving historical works, the “Historia Ottonis” and the “Legatio Constantinopolitana”) as an unreliable historical source. Their very work, however, in identifying key points where he twisted reality provides the solid basis of an assessment of Liutprand’s techniques and purposes.52

48 Lippelt, Thietmar, 151. 49 Lippelt, Thietmar, 94–5. 50 Ibid. This point had earlier been raised by Robert Holtzmann, “Die Aufhebung und Wiederherstellung des Bistums Merseburg,” Sachsen und Anhalt 2 (1926), 35–75, reprinted in Aufsätze zur deutschen Geschichte im Mittelelberaum, ed. Albrecht Timm (Darmstadt, 1962), who nevertheless argued that Thietmar was fundamentally trustworthy. Most recently, John W. Bernhardt, “King Henry II of Germany: Royal Self-Representation and Historical Memory,” in Medieval Concepts of the Past, 39–69, here 52, notes that “even Thietmar, arguably our best single source for Henry II’s reign, was not always an unbiased observer of events when he had pragmatic interests at stake.” In particular, Thietmar failed to record the inducements offered by Henry II to the archbishop and chapter at Magdeburg to lessen the sting of losing the properties now held by the reestablished diocese of Merseburg. This was a point also made several years before by David A. Warner, “Henry II at Magdeburg: Kingship, Ritual and the Cult of Saints,” Early Medieval Europe 3 (1994), 133–66, esp. 162–5. It should be emphasized, that like Holtzmann, Bernhardt suggests that it was only in cases where Thietmar had a pragmatic reason affecting his own or his bishopric’s personal interest that he shaped his history to suit these needs. 51 See previous note. 52 Philippe Buc, “Italian Hussies and German Matrons: Liutprand of Cremona on Dynastic Legitimacy,” Frühmittelalterliche Studien 29 (1995), 207–25, here 210.

110

M E M O R Y, E P I S T E M O L O G Y, A N D T H E W R I T I N G

In addition to emphasizing the scholar’s ability to read “against the grain” once Thietmar’s parti pris has been identified, Lippelt stresses that Thietmar wrote the most about those periods for which he had the most information. Thietmar’s treatment of contemporary events and the reign of Henry II is substantially longer, for example, than his book on the reign of Henry I (919–936).53 Thietmar, as Lippelt emphasizes, had considerable access to King Henry II who visited Merseburg twenty-six times during his reign.54 Because Merseburg was the major staging point for Henry II’s numerous campaigns in Poland and Bohemia, Thietmar was an eyewitness to many of the major military events of Henry II’s reign as well as serving as one of the king’s leading advisors and therefore, was someone who was conversant with his plans.55 The most recent studies dealing with Thietmar qua author affirm the traditional conclusions that he was knowledgeable about current events and that the information that he provides, particularly when this information is not directly related to his parti pris, is accurate. Nicole Lederer-Herkenhoff argues that Thietmar is indispensable for understanding the Ottonian period.56 She stresses with regard to Thietmar’s knowledge of current events and the recent past that the political requirements of restoring the diocese of Merseburg, as well as the constant military threat to the frontier, kept the bishop in constant contact with the king.57 David Warner, Thietmar’s English translator, notes that “[a]side from its very existence in an age relatively poor in narrative sources, the work has been appreciated for its attention to detail, breadth of interests, and, above all, its author’s well-informed perspective.”58 Marcus Brömel points to Thietmar’s importance for understanding Ottonian policies regarding the frontier marches, the early history of the West Slavs, Slavic-German relations, and German military campaigns in the East.59 Brömel, like Lederer-Herkenhoff, focuses on the extensive information available to Thietmar from his personal contacts with King Henry II, as well as from Thietmar’s own family, which included his two elder brothers, who held royal offices as the count of Walbeck and the Burggraf of Magdeburg.60 In addition to works focusing specifically on Thietmar as a historian, numerous scholars have implicitly 53 Lippelt, Thietmar, 145. 54 Lippelt, Thietmar, 115. Henry II also spent more time at Merseburg than any other German king. 55 Lippelt, Thietmar, 166–71. 56 Lederer-Herkenhoff, “Thietmar,” passim. 57 Ibid., 112–14. 58 David A. Warner, “Thietmar of Merseburg on Rituals of Kingship,” Viator 26 (1995), 53–76, here 64. Warner cites specifically, Karl Leyser, The Ascent of Latin Europe (Oxford, 1986), 21; Werner Goetz, “Thietmar von Merseburg, Geschichtsschreiber,” in Gestalten des Hochmittelalters: Personengeschichtliche Essays im allgemeinhistorischen Kontext (Darmstadt, 1983), 70–83, especially 72; and Trillmich, Chronik, xx. 59 Markus Brömel, “Thietmar von Merseburg: Ein Geschichtsschreiber der Ottonenzeit,” in Auf den Spuren der Ottonen II, ed. Ronald Mecke, Robert Werner, Axel Voigt, and Jan Brademann (Halle, 2000), 131–39. 60 Ibid., 133.

111

N A R R AT I V E W O R K S

recognized Thietmar’s exceptional value as a source of accurate information by invoking his testimony as dispositive in understanding the Ottonian century.61 In considering the critical evaluation by scholars over the past century and a half regarding the Chronicon of Thietmar of Merseburg, it is difficult to escape the conclusion that he is among the best surviving witnesses to the Ottonian period and the single best witness to the reign of Henry II. Within this broad area of inquiry, Thietmar’s office as bishop of Merseburg, a territory of crucial strategic significance for King Henry II’s eastern policies, and Thietmar’s close relationship with the ruler, as well as to many other leading office holders, including his own two elder brothers, make Thietmar’s testimony regarding military affairs particularly important for understanding contemporary military organization. Indeed, not only was Thietmar able to gain substantial information from his father, who was a leading military commander in eastern Saxony, and his brothers, who commanded significant military forces under the king’s command as the count of Walbeck and the Burggraf of Magdeburg, respectively, Thietmar personally commanded military forces in independent commands and under the king’s orders as part of the royal army.62 In addition, and perhaps even more importantly, the city of Merseburg, as noted earlier, served as the staging area for most of Henry II’s military undertakings in Poland and Bohemia.63 Consequently, even when not personally serving in the field, Thietmar was an eyewitness to the campaign preparations undertaken by Henry II. Moreover, as a close advisor to Henry II regarding the Slavs and the eastern frontier, Thietmar played an important role in planning these same campaigns and had the opportunity to witness discussions among the king’s councilors that 61 Among the most important of these works is Stefan Weinfurter, Heinrich II: Herrscher am Ende der Zeiten (Regensburg, 1999), 12 and passim, who argues for the centrality of the information provided by Thietmar for the writing of the history of Henry II’s reign. For the exceptional importance of the information provided in Thietmar’s Chronicon for writing the history of Ottonian-Slavic relations, imperial politics, the contemporary conception of royal power and government, and missionary activity, see Lorenz Weinrich, “Der Slawenaufstand von 983 in der Darstellung des Bischofs Thietmar von Merseburg,” in Historiographia Mediaevalis: Studien zur Geschichtsschreibung und Quellenkunde des Mittelalters. Festschrift für Franz-Josef Schmale zum 65. Geburtstag, ed. Dieter Berg and Hans-Werner Goetz (Darmstadt, 1988), 76–87; Alexandra Pütz, “Zum Verhältnis von Reichsgewalt und Partikulargewalt. Herzog Heinrich II von Bayern in der Chronik des Bischofs Thietmar von Merseburg,” in Aus Überrest und Tradition: Festschrift für Anna-Dorothee von den Brincken, ed. Peter Engels (Erlangen, 1999), 10–53; Gerd Althoff, “Saxony and the Elbe Slavs in the Tenth Century,” in New Cambridge Medieval History III: 900–1024, ed. Timothy Reuter (Cambridge, 1999), 267–92; Stanislaw Rosik, “Der Christianisierungsprozeß von Schlesien am Anfang des 11. Jahrhunderts nach der Chronik des Thietmar von Merseburg,” in Religions-und Kulturgeschichte in Ostmittel-und Südosteuropa, ed. Joachim Köhler and Rainer Bendel, vol. 1 (Münster, 2002), 191–8; and Sverre Bagge, Kings, Politics, and the Right Order of the World in German Historiography c. 950–1150 (Leiden, 2002), 95–188. 62 Thietmar discusses his personal service at the fortresses of Meißen and Lebusa with his milites, and his military planning sessions with leading figures in the realm, including Queen Cunigunda and Archbishop Tagino of Magdeburg (1004–1012). See, for example, Thietmar, Chronicon, 6.33, 278; 6.59, 306; and 6.80, 326. 63 On this point, see Lippelt, Thietmar, 115, 145, and 166–71.

112

M E M O R Y, E P I S T E M O L O G Y, A N D T H E W R I T I N G

ultimately turned into reality and others that did not.64 Given these circumstances, Thietmar’s information regarding military affairs is of crucial importance for a thorough understanding of the military history of Ottonian Germany.

Milites as professional soldiers in Thietmar’s Chronicon As will be shown in this study, Thietmar diligently applied the term miles to men whose primary occupation was military service, that is to say professional soldiers.65 Understood in historical context, the milites of the German kingdom, who are discussed by Thietmar, play the same role as the royal obsequium and magnate obsequia deployed by Carolingian kings.66 These professional soldiers in Germany stand in contrast to the militia forces of the select and general levies.67 It

64 See, in this context, David S. Bachrach, “The Eastern Campaigns of Henry II of Germany, 1003–1017,” Journal of Medieval Military History 17 (2020), 1–36. 65 In the few instances in which Thietmar uses the term miles to refer to individuals who were not soldiers, the bishop generally intends the term to be understood in a topological sense. Thus, for example, when discussing Saint Maurice, Thietmar refers to the holy man as a miles Christi. See Thietmar, Chronicon, 7.16. Warner, Ottonian Germany (as n. 41 earlier) 160, n. 38, suggests that Thietmar used the term milites loosely, referring specifically to his use of the term at 4.13 to describe the Liutizi employed by Duke Boleslav II of Bohemia (972–999). It is clear from this passage, however, that these Liutizi were serving as mercenaries, that is as professional soldiers. Thietmar’s use of the term miles to discuss mercenaries will be discussed in greater detail later. 66 With respect to the obligations of both secular and ecclesiastical magnates to provide expeditionary military forces to the king on the basis of their landed holdings and income, see Bernard S. Bachrach, Early Carolingian Warfare: Prelude to Empire (Philadelphia, 2001), 59–65; and David S. Bachrach, Warfare in Tenth-Century Germany (Woodbridge, 2012), particularly, 87–92. 67 Concerning the obligation of wealthy individuals to provide expeditionary military service on the basis of their property, see Helen M. Cam, Local Government in Francia and England (London, 1912), 127–37; F. L. Ganshof, “Charlemagne’s Army,” in Frankish Institutions under Charlemagne, trans. Bryce and Mary Lyon (Providence, RI, 1968), 59–68, 151–61, originally published as “L’Armée sous les Carolingiens,” in Settimane di Studio de Centro Italiano di Studi sull’alto medioevo, 2 vols. (Spoleto, 1968), I:109–30, with the citation to the English text (59–61); and Bernard S. Bachrach, “Military Lands in Historical Perspective,” The Haskins Society Journal 9 (1997), 95–122, here 109–112; and idem, “Charlemagne’s Expeditionary Levy: Observations Regarding liberi homines,” Studies in Medieval and Renaissance History 3rd series 12 (2015), 1–65. With respect to the military obligations of the local levy, see Bachrach, Early Carolingian Warfare, 52–7; and D. Bachrach, Warfare in Tenth-Century Germany, 77–85. As will be seen later, when Thietmar wishes to differentiate between the forces of the select levies and the professional soldiers of the magnate and royal households, he uses the standard technique of referring to the former with a generic regional identification, e.g. the “Bavarians,” while the latter are denoted by the term milites. The practice of describing levies, as contrasted with professional soldiers, with a generic regional designation was very common among medieval writers. For example, Widukind, one of Thietmar’s sources, distinguished between the milites of the ducal household and the exercitus of the Saxons whom Henry I, before his election as king, could mobilize from his duchy. On this point, see Leyser, “Henry I,” 23. Similarly, Dudo of St. Quentin, Thietmar’s contemporary, distinguished between milites and the local men of a particular region, e.g. Frisones, Walgrenses, who served in the levy. On this point, see Bachrach, “Dudo of St. Quentin as an Historian of Military Organization,” 171.

113

N A R R AT I V E W O R K S

should be emphasized that Thietmar was not interested, or at least not sufficiently interested, in the private lives of the men he denotes as milites to discuss what they did while not engaged in military service. This is not to say, however, that the milites always appear in Thiemtar’s text engaging in combat. Indeed, they are rather more frequently shown serving in garrisons, as will be made clear later.68 In this context, the bishop of Merseburg deployed the term miles/milites frequently to depict members of the military household of one or another secular or ecclesiastical magnate in the German kingdom.69 Numerous counts and dukes holding office during the reigns of Otto I, Otto II, Otto III, and Henry II are presented by Thietmar as employing and deploying contingents of milites. Otto I’s younger brother Henry, the duke of Bavaria, for example, is presented as deploying his milites during the king’s siege of the rebellious royal son Liudolf at Mainz in 953.70 Thietmar’s own father Count Siegfried commanded a group of milites whom he deployed in support of the milites of Margrave Hodo in 972 against Duke Miesco I of Poland.71 In 995, according to Thietmar, a margrave named Ekkehard, the count palatine Frederick, as well as Thietmar’s three paternal uncles and his maternal uncle, each led contingents in a combined force to relieve the siege of the fortress of Brandenburg that was being threatened by a force of Slavic pagans, known to the Germans as Liutizi.72 The men in these military companies are described by Thietmar as milites.73 In discussing a military conflict between 68 In this context, it is important to observe that Thietmar’s practice of denoting professional soldiers as milites and reserving the term milites exclusively for professional soldiers was maintained by authors later in the eleventh century as well. See David S. Bachrach, “Milites and Warfare in Pre-Crusade Germany,” War in History 23.3 (2015), 298–343. 69 The maintenance of military households by magnates had a very long tradition in the West, dating back to the late Roman Empire. During the Carolingian period, down to the end of the ninth century, both secular and ecclesiastical magnates were required to maintain military households that were subject to mobilization by the Crown as part of the tripartite military organization of the regnum Francorum, the other two elements of which were the general and select levies. Regarding the nature of military households and their deployment by the central government in the late Roman, Merovingian, and Carolingian periods, see Bachrach, Early Carolingian Warfare, 51–82. With respect to the nature of the magnate military household on the continent and in England in the period after the collapse of the Carolingian empire, see Bachrach, “Dudo of St. Quentin as an Historian of Military Organization,” 165–85; J. O. Prestwich, “The Military Household of the Norman Kings,” in Anglo-Norman Warfare, ed. Matthew Strickland (Woodbridge, 1992), 93–128, originally published in English Historical Review 96 (1981), 1–35; and Marjorie Chibnall, “Mercenaries and the familia regis under Henry I,” Proceedings of the Society for Antiquaries of Scotland 62 (1977), 15–23. 70 Thietmar, Chronicon, 2.6. Cf. Thietmar’s source Widukind, Quellen zur Geschichte der sächisischen Kaiserzeit: Widukinds Sachsengeschichte, ed. Albert Bauer and Reinhold Rau from the original edition by Paul Hirsch, Max Büdinger and Wilhelm Wattenbach (Darmstadt, 1971), 3.18, who does not use the term miles in this context. Now also see the English translation of Widukind by David S. Bachrach and Bernard S. Bachrach, Widukind of Corvey: Deeds of the Saxons (Washington, DC, 2014). 71 Thietmar, Chronicon, 2.29. 72 Ibid., 4.22. 73 Ibid.

114

M E M O R Y, E P I S T E M O L O G Y, A N D T H E W R I T I N G

Hermann, the margrave of Meißen and his uncle Gunzelin in 1009, Thietmar notes that the former had garrisoned his fortresses (civitates) with his milites in order to hold them against assault.74 Similarly, Gunzelin’s brother, Count Brun, following the command of King Henry II, deployed his own milites to garrison the fortress at Meißen.75 Thietmar also routinely describes the men in the military households of ecclesiastical magnates as milites. Bishop Folkmar of Brandenburg (980–983), for example, employed a company of milites commanded by a man named Dietrich, who singularly failed to protect episcopal properties in the region during the massive Slavic revolt of 983.76 Archbishop Giselher of Madgdeburg, noted earlier, had a force of his milites serving as the garrison of Arneburg in 997.77 In 1002, Bishop Werner of Strassburg’s milites provided a last-ditch defense of the city’s cathedral, a defense characterized by Thietmar as ineffective, against a very large force of Swabian militiamen recruited by their duke Hermann for an assault on Strassburg.78 In that same year, according to Thietmar, the bishops of Strassburg and Basel provided contingents of milites to serve in the garrison of the fortress of Breisach.79 In his discussion of the death in 1010 of Bishop Ansfrid of Utrecht, Thietmar makes clear that the prelate’s daughter, who was an abbess in her own right, employed a company of milites whom she ordered to be deployed to seize her father’s remains from the urban militia of Utrecht who had arrived armatis manibus in order keep possession of them.80 In the context of describing a property agreement he had made with Archbishop Gero of Magdeburg (1012–1023) in 1015, Thietmar mentions that the former deployed a contingent of milites to garrison the fortress of Zörbig, which belonged to Gero.81 In addition to the men serving in the military households of magnates from the German kingdom itself, Thietmar applied the term milites to the members of the military households of important figures living outside the Reich. Duke Boleslav Chrobry of Poland (92–1025), for example, is depicted by Thietmar as providing a substantial force of 300 of his own milites, who were moreover equipped with hauberks (loricae), for service with Emperor Otto III.82 When discussing the efforts of this same duke to defend the fortress of Glogau in 1010, Thietmar uses the term milites to describe the members of the military company that he kept with 74 75 76 77 78

79 80 81 82

Ibid., 6.53. Ibid., 6.55. Ibid., 3.17. Ibid., 4.38. Ibid., 5.12. Here it should be noted that Hermann’s household troops are characterized by Thietmar as a miles armatus, using the singular as a collective noun, while the great bulk of Hermann’s force is depicted by the bishop of Merseburg as a turba Alemannorum. Thietmar, Chronicon, 5.21. Ibid., 4.37. Ibid., 7.24. Ibid., 4.46.

115

N A R R AT I V E W O R K S

him in the garrison there.83 Indeed, it is worth noting that Thietmar used the term milites twice when discussing these troops rather than deploying some other term that denoted fighting men.84 Thietmar’s consistency here may well indicate that he thought the term milites and milites alone could adequately describe the nature of Boleslav’s troops. Thietmar also used the term milites to describe the military forces of Aethelred, a claimant to the English throne, as the latter mobilized his fighting men to invade England following the death of Sven Forkbeard in 1014.85 The members of the military household of Aethelred’s widow, who held London against the Dane Cnut in 1016, likewise were described by Thietmar as milites.86 Although Thietmar usually deploys the term milites to refer to members of the household military forces of magnates, he does on occasion also describe as milites professional fighting men who served for pay as mercenaries. In particular, Thietmar describes the pagan Liutizi employed by King Henry II and the Moravians employed by Boleslav Chrobry as milites. In the first instance, Thietmar records that the Liutizi, whom Henry II had employed for operations against Boleslav Chrobry in 1017 lost fifty milites when they attempted to cross the river Mulde while it was in flood.87 Regarding events in this same year, Thietmar recorded that Boleslav’s Moravian milites overwhelmed a much larger force of Bavarians whom King Henry had ordered into the field against the Polish duke.88 In addition to identifying companies of fighting men as milites, Thietmar frequently uses the term miles in the Chronicon to denote individual men who had careers in military service. The single highest ranking of these was John Tzimiskes, the Byzantine general, whom Thietmar identifies as the miles who assassinated and then replaced Emperor Nikephoros Phokas (963–969) in 969.89 It was John Tzimiskes who sent his niece Theophanu to Germany as a bride for Otto II. Other high-ranking men identified by Thietmar as serving as milites include Count Brun of Arneburg, who died during Otto II’s invasion of Francia occidentalis in 978; Margrave Ricdag, who served as the royal garrison commander (civitatis custos) of Meißen; the garrison commander of Brandenburg named Kizo; and Thietmar’s own brothers Henry, the count of Walbeck, and Frederick. The bishop describes his brothers as optimi milites.90 Thietmar also used the term miles when discussing the reinstatement to royal favor of Duke Hermann of Swabia, the son of Thietmar’s maternal great uncle.91 In this case, the bishop of Merseburg notes Hermann’s two separate relationships with King Henry II, first as a royal miles and second as the king’s faithful friend (fidus amicus), which may be understood 83 84 85 86 87 88 89 90 91

Ibid., 6.58. Ibid. Ibid., 7.37. Ibid., 7.40. Ibid., 7.64. Ibid., 7.57. Ibid., 2.15. Ibid., 3.8; 4.5; 4.22; and 4.41, respectively. Ibid., 5.22.

116

M E M O R Y, E P I S T E M O L O G Y, A N D T H E W R I T I N G

here as synonymous with fidelis.92 One of the highest-ranking Germans identified as a miles by Thietmar is Conrad the Red, Otto I’s son-in-law, who was duke of Lotharingia until he was removed from office during his revolt against the king in 953. In this instance, Conrad, who during the battle on the Lechfeld in 955 commanded a noteworthy body of troops from the Rhineland, is described by Thietmar as miles egregious, or outstanding soldier, who gave his life fighting against the Hungarians.93 In addition to identifying as milites those high-ranking men whose responsibilities included the regular command of soldiers in the field or in garrisons, Thietmar identified by name numerous individual soldiers of lower rank, whom he also designated as milites. Moreover, as in the case of companies of milites, noted previously, many of these men served magnates not subject to the German Crown. Thietmar draws attention, for example, to a miles named Wagio, who served Duke Boleslav I of Bohemia (935–972).94 This particular soldier was of interest to Thietmar because he was responsible for the murder of the garrison commander Ricdag, noted earlier.95 Thietmar similarly identified a miles in the service of Duke Boleslav II of Bohemia (972–999) named Slopan, who served as a scout in the ducal army.96 Of course, Thietmar also uses the term miles to identify individuals who served in the military households of German magnates. In discussing the failure of Margrave Ekkehard I of Meißen (985–1002) to seize the German throne in 1002, Thietmar commented on the loyalty of the margrave’s personal troops, including the miles Herman and the miles Athulf, who both gave their lives to protect Ekkehard from assassination.97 In 1003, according to Thietmar, the miles Magnus, who served the rebellious Margrave Henry of Schweinfurt, led a body of troops under his command to capture a royal baggage train near the settlement of Hersbruck.98 In one final example, Thietmar identifies as milites four men, named Thiedbern, Bernhard, Isi, and Benno, in the military household of Bishop Arnulf of Halberstadt (996–1023).99 These soldiers distinguished themselves by their stupidity, in Thietmar’s view, by assaulting in an uncoordinated manner a force of Slavs that not only was ensconced in a fixed position but was equipped with archers as well.100 It is a noteworthy aspect of Thietmar’s style that in addition to using the term miles to denote individual professional soldiers, whether German, Slav, Anglo-Saxon, or 92 Ibid. 93 Ibid., 2.10. Cf. Widukind, Res gestae, 3.47, who does not use the term miles to describe Conrad in this context. 94 Thietmar, Chronicon, 4.5. 95 Ibid. 96 Ibid., 4.12. 97 Ibid., 5.6. 98 Ibid., 5.34. 99 Ibid., 6.22. 100 Ibid.

117

N A R R AT I V E W O R K S

Byzantine, he occasionally used the same term miles, although a singular noun, to denote a body of men whom he might otherwise have described as milites. In discussing, for example, Otto I’s conquest of Rome in 961, Thietmar notes that the German king went into battle at the head of his miles armatus.101 Given the large size of the military force involved in this operation, Thietmar very likely intended the miles armatus to be understood not only to include the milites of Otto’s own military household, but the milites of the numerous German and Italian magnates serving in the royal army as well. Similarly, in his discussion of the campaign by Otto III to crush a Slavic uprising in 997, Thietmar notes that the emperor proceeded against the enemy with an armatus miles.102 Given that this revolt was a major problem facing the kingdom as a whole, it again seems likely that the miles armatus here was intended by Thietmar to denote all of the milites taking part in the operation, rather than only to Otto III’s personal military household. In other instances, however, Thietmar clearly uses the term miles to denote either the entire military household of a single magnate, or at least a substantial number of the milites whom he employed. Thietmar uses the term miles, for example, to denote the military household of Margrave Ekkehard I of Meißen, noted previously. In this particular case, men living in the village of Görschen, subject at that time to Archbishop Giselher of Magdeburg, caught several men who were committing crimes and then hanged them. These men were subject to the margrave.103 At the time of this incident, according to Thietmar, Archbishop Giselher and Margrave Ekkehard were engaged in a competition to gain royal favor, and Ekkehard held a grudge against the archbishop. In order to demonstrate his displeasure with Giselher, therefore, Ekkehard claimed that the men of Görschen had violated his judicial jurisdiction and dispatched his miles under the command of a man named Rambold to vindicate his rights by arms (ad vindicanda armari).104 Rambold, according to Thietmar, led a large force (multitudo grandis) that was sufficient to encircle the entire village and capture all of the men living there.105 In concluding this section, it must be stressed that all of the men whom Thietmar identifies either individually or collectively as milites made their living through

101 Ibid., 2.13. 102 Ibid., 4.29. Nb., Thietmar did not use the term exercitus here, which would have indicated the inclusion of expeditionary levies, as well as professional soldiers in this campaign. 103 Thietmar, Chronicon, 4.73. Thietmar likely was interested in this story because Görschen was among the properties that were awarded to Magdeburg during the suppression of the diocese of Merseburg under Otto II, and Thietmar was able to gain it back for his diocese once he became bishop in 1009. 104 Thietmar, Chronicon, 4.73. 105 Ibid. For other instances in which Thietmar uses the phrase armatus miles or simply miles to refer to a company of professional soldiers, see Thietmar, Chronicon, 5.12 (to be discussed below); 6.80; and 8.34. In one instance, discussing the soldiers of Margrave Gero II of the east Saxon March (993–1013) who were serving under the command of Gero’s officer named Hugal, Thietmar first denoted the men as milites, and then as an armatus miles, using miles as a collective noun. See Thietmar, Chronicon, 6.96 and 6.97, respectively.

118

M E M O R Y, E P I S T E M O L O G Y, A N D T H E W R I T I N G

military service or, in the case of the high-ranking men, noted earlier, devoted a considerable portion of their careers to commanding soldiers in campaigns and in garrisons, that is, serving as military officers. It is particularly important to recognize that Thietmar viewed the term miles as appropriate for discussing professional soldiers from the German kingdom as well as from far beyond its borders. Indeed, the only characteristic that the pagan Slavic Liutizi; the Christian Bohemians; the Bavarians, Saxons, and Swabians serving in the households of secular and ecclesiastical magnates; and the Byzantine general and emperor John Tzimiskes – all of whom Thietmar identifies as milites – had in common was that they were professional soldiers. Their military service is to be contrasted with that of men who provided occasional military service in militia forces of the type with which Thietmar was familiar, namely the select and general levies of the Ottonian kingdom. It should also be emphasized that Thietmar does not use the term miles as an honorific, e.g. knights as some Anglophone scholars prefer anachronistically, nor does he use it to denote so-called “feudal” relationships, e.g. as a synonym for vassal.106 Thietmar’s diligent reservation of the term miles for men who served as professional soldiers is made particularly clear, ironically, in his single use of the term to describe a bishop. In 972, Bishop Michael of Regensburg (941–972), in conjunction with numerous other Bavarian magnates, took the field against the Hungarians.107 Once the two sides made contact, the Bavarian forces quickly were overwhelmed, and Michael, who was wounded in the fighting, killed a Hungarian soldier. Finally reaching safety, Bishop Michael, according to Thietmar, was regarded by everyone not only as an optimus pastor but as a bonus miles as well.108 From the context, it is quite clear that Thietmar was referring to Michael’s skill as a soldier, demonstrated by his survival in hand-to-hand combat.

Deployment of milites Given the exceptionally important role, as identified by both historians and archaeologists, of fortifications in the military organization of the Ottonian period, it is unsurprising that Thietmar very frequently discusses the service of milites as garrison troops.109 Thietmar emphasizes, for example, the policy of Otto III to 106 The most important statement of the nature of the milites as members of a putatively feudal warrior nobility is Leyser, “Henry I,” passim, who argues that the milites in Thietmar’s text as the lowest rung of the ranks of the nobility. 107 Thietmar, Chronicon, 2.27. Concerning the practice of Ottonian bishops leading troops into the field, see Leopold Auer, “Der Kriegsdienst des Klerus unter den sächsischen Käiser,” Mitteilungen des Instituts für österreichischen Geschichtsforschung 79 (1971), 316–407; and part 2 in MIÖG 80 (1972), 48–70. 108 Thietmar, Chronicon, 2.27. 109 Regarding the extensive construction and maintenance programs for fortifications undertaken by the kings of the Ottonian dynasty, as well as many of their magnates, see the exceptionally valuable survey of the literature in Edward J. Schoenfeld, “Anglo-Saxon ‘Burhs’ and Continental

119

N A R R AT I V E W O R K S

maintain strong fortifications along the Slavic frontier that were intended to inhibit large-scale incursions into German territory.110 One of the principal fortresses was Arneburg, located in modern Sachsen-Anhalt on the western bank of the Elbe river, where Otto III devoted considerable resources in 997 to strengthening the structure in order to withstand a siege.111 In order to garrison this position, Otto mobilized the milites from the military households of a number of his magnates. Thietmar pays special attention to the milites deployed by Archbishop Giselher of Magdeburg because Arneburg was lost to the Slavs in the interval between the time Giselher’s milites ended their term of service and departed the fortress and the arrival of the milites commanded by Thietmar’s paternal uncle, Margrave Liuthar.112 Thietmar places the blame here on Giselher, against whom Thietmar already held a grudge because of his role in the suppression of the diocese of Merseburg, noted previously, rather than on his uncle during whose “watch” the fortress actually was destroyed.113 Giselbert’s successor but one as archbishop of Magdeburg, Gero, with whom Thietmar was on good terms, similarly provided milites from his household

110 111 112 113

‘Burgen’: Early Medieval Fortifications in Constitutional Perspective,” The Haskins Society Journal 6 (1994), 49–66. Also see Konrad Schünemann, “Deutsche Kriegsführung im Osten während des Mittelalters,” Deutsches Archiv 2 (1938), 54–84, particularly 56–61, where he emphasizes that Charlemagne and his successors in the Ottonian period not only grasped the importance of controlling rivers to transport supplies, but also devoted considerable energy to protecting these rivers with fortifications and bridges.; Heinrich Büttner, “Zur Burgenbauordnung Heinrichs I.,” Blätter für deutsche Landesgeschichte 92 (1956), 1–17; Albert K. Hömberg, “Probleme der Reichsgutforschung in Westfalen mit Zwei Karten,” Blätter für deutsche Landesgeschichte 96 (1960), 1–21, especially 15–21 where he identifies at least twenty royal fortifications sustained in southern Westphalia by the kings of Francia orientalis and Germany during the ninth and tenth centuries; Kurt-Ulrich Jäschke, Burgenbau und Landesverteidigung um 900: Überlegungen zu Beispielen aus Deutschland, Frankreich und England (Sigmaringen, 1975), especially 116 and 118–20; Auer, “Zum Kriegswesen unter der früheren Babenbergern,” 11, who emphasizes continuity in usage and maintenance of fortifications between the Carolingian and Ottonian periods; Leyser, “The Battle at the Lech, 955,” who stresses the use of fortifications by the Ottonians for counter-raids against the Hungarians; Walter Melzer, “Karolingisch-ottonische Stadtbefestigungen in der ‘Germania libera’,” in Die Befestigung der mittelalterlichen Stadt, ed. Gabriele Isenberg and Barbara Scholkmann (Cologne, 1997), 61–77, who emphasizes that there is substantial physical continuity between Carolingian and Ottonian fortificationns, particularly in the region of Westphalia; and Charles Bowlus, The Battle of Lechfeld and Its Aftermath, August 955: The End of the Age of Migrations in the Latin West (Aldershot, 2006), who provides a detailed account of the establishment of a system of fortifications in eastern Bavaria. In his dissertation, Studien zum Heer der Ottonen und der ersten Salier (919–1056) (Bonn, 1985), 10–11. Bruno Scherff makes clear that military campaigns culminating in sieges far outnumbered those that culminated in combat in the field during the entirety of the Ottonian period and half of the Salian period as well. Thietmar, Chronicon, 4.38. Ibid., “Harnaburg civitatem opere muniens necessario.” Ibid. Ibid., Thietmar suggests that the fortress was already burning when Liuthar arrived even though the archbishop had remained in garrison at Arneburg until the end of his tour of duty (custodivit tamen . . . urbem ad dictum diem Gisilherus tristisque reversus obviam habuit . . Liutharium, cuius curam civitas predicta tunc respiciebat).

120

M E M O R Y, E P I S T E M O L O G Y, A N D T H E W R I T I N G

forces to serve as garrison troops in the king’s service. Thietmar records that following the surrender, noted previously, of Margrave Henry of Schweinfurt in 1004, Archbishop Gero deployed a unit of his milites to guard this prisoner and hold the fortress (castellum) of Giebichenstein, located near Magdeburg.114 Milites did not, however, only serve as garrison troops in royal service. German magnates also deployed their milites in internecine struggles. As noted earlier, Count Hermann, the margrave of Meißen, deployed his milites in garrisons to hold fortresses against his uncle Gunzelin in 1009.115 This strategy was successful at Hermann’s stronghold of Strehla, located in the modern German state of Saxony about 110 kilometers east of Merseburg. However, at Rochlitz, located about 55 kilometers southwest of Strehla, Hermann deployed an insufficiently large force of his milites, and Gunzelin captured the fortification. Subsequently, according to Thietmar, Gunzelin burned it down.116 Events in this same year also shed light on the ability of King Henry II to intervene in the deployment of milites by his magnates by requiring them to provide these men for service as garrisons in royal strongholds. Thietmar notes that at the same time Gunzelin was locked in a conflict with his nephew Hermann and struggling to keep control of his own strongholds, Henry II mobilized the milites of Gunzelin’s brother Brun, from whom Gunzelin might otherwise have drawn support, to serve in a garrison in the royal service.117 Henry II’s control over the deployment of Brun’s milites at this point is especially significant because Gunzelin lost his fortress (castellum) on the Saale river, in which he kept his treasury, to an assault by Hermann, indicating that Gunzelin had an insufficiently large garrison there to hold the stronghold.118 Thietmar makes a point of associating the conflict between Gunzelin and Hermann to Henry II’s continuation of the long-standing royal policy of the Ottonian kings to require German magnates, including both secular and ecclesiastical officials, to provide garrisons of their own milites to hold frontier fortifications in the general interest of the kingdom.119 Thietmar notes explicitly that at the same time Gunzelin was losing his struggle against Hermann, it was Brun’s turn to provide a garrison of milites to hold the frontier fortress of Meißen. Thietmar takes the opportunity here to emphasize King Henry II’s power by pointing out not only that Brun’s company of milites completed its tour of duty as required by the king rather than sending these men to aid his brother Gunzelin but that Brun then turned over command of the fortress to his nephew Hermann, against whom Brun’s brother Gunzelin was fighting.120 In order that his audience not miss his

114 115 116 117 118 119 120

Thietmar, Chronicon, 6.2. Ibid., 6.53. Ibid. Ibid., 6.55. Ibid., 6.53. Ibid., 6.55. Ibid.

121

N A R R AT I V E W O R K S

point about the extent of royal power, Thietmar takes special care here to emphasize that the rendition of the fortress was carried out under the direction and in the presence of a royal official whom he identifies as a nuncius.121 Not surprisingly, given his very good sources of information regarding the Slavic magnates to the east of his diocese of Merseburg, Thietmar also comments on the practice of the Polish duke Boleslav Chrobry and the Abodrite ruler Mistislav of installing garrisons of milites in their fortresses. Thietmar discusses Boleslav’s deployment of his milites at the strategically located fortress of Glogau, as noted earlier, in the context of King Henry II’s campaign against the Polish duke in 1010.122 Thietmar comments on the deployment by Mistislav of his milites as garrison troops while reviewing events that took place in 1018, the year in which Thietmar died.123 In this latter instance, Thietmar focuses on Mistislav’s decision to mobilize a picked (electi) force of his milites to hold the citadel (municio) inside the fortified settlement (civitas) of Schwerin, located north of the Elbe and west of the lower course of the Elde, in an effort to withstand a major incursion by the Liutizi. Ultimately, Mistislav’s plan failed because the population of Schwerin revolted and forced the Abodrite ruler and his picked force of milites to abandon this position to the Liutizi.124 Corollary to the deployment by both German and Slavic magnates of their milites to serve as garrison forces in fortifications is the fact that these same strongholds frequently were under threat of assault, either by rivals from within the Reich or from incursions by external forces. In this context, Thietmar discusses the deployment of milites not only to defend fortresses but to attack them as well. In 1002, for example, Duke Hermann of Swabia launched an assault on the city of Strasbourg, in the course of which his army successfully stormed the walls and engaged with the defenders in a pitched battle near the cathedral.125 Thietmar emphasizes that Hermann’s attacking force was formed from two separate elements. The first consisted of the duke’s military household, denoted by Thietmar as his miles armatus, and the second of levies from Swabia, whom the bishop of Merseburg describes, with some scorn, as an execrata Alemmanorum turba.126 This passage illuminates Thietmar’s disdain of the men who served in militia forces. Moreover, the distinction that Thietmar draws here is significant with respect to the author’s view of the relative discipline of the two elements of Hermann’s army. Rather than serving the purpose for which they were mobilized, the militia levies actually assaulted the cathedral at Strasbourg and looted it, despite the defense of the building by the milites of the bishop of the city. Thietmar is quick to point out regarding this event, however, that these Swabian levies, presumably unlike Hermann’s milites, acted without the duke’s knowledge, i.e. inscio

121 122 123 124 125 126

Ibid. Ibid., 6.58. Ibid., 8.5. Ibid. Ibid., 5.12. Ibid.

122

M E M O R Y, E P I S T E M O L O G Y, A N D T H E W R I T I N G

duce, indicating their lack of discipline.127 But given his own biases, Thietmar does not mention that the duke had lost command and control over these troops. In discussing the military actions of the next year, 1003, Thietmar reports the deployment by King Henry of a substantial force of milites in the course of a series of siege operations directed against the rebel magnate, Margrave Henry of Schweinfurt, noted previously.128 Following the assault made by the military forces of Henry of Schweinfurt, commanded by his miles, Magnus, against a royal baggage train, the royal army cornered Magnus’s troops at the fortress (civitas) of Ammerthal, located near the Bavarian city of Amberg. The royal army then prepared for a siege, setting up engines (instrumenta belli) for this task.129 Magnus realized that he could not hold out against this substantial royal army, particularly one equipped with siege engines, and surrendered. King Henry then moved on to the stronghold (castellum) of Creußen commanded by Margrave Henry’s brother Bukko, located some 55 kilometers to the northeast of Ammerthal, and established a siege there.130 In a strategy typical of siege warfare, Margrave Henry attempted to aid the defenders of Creußen by harassing the royal forces besieging this citadel. Thietmar notes, in particular, the margrave’s attacks on those of the king’s foragers, who were incautious in seeking fodder for the army’s horses.131 In order to prevent further attacks of this type, King Henry deployed a unit of 400 milites to protect his foragers and keep the margrave’s forces at bay, a task that, according to Thietmar, they successfully completed, thereby helping to ensure the success of the siege.132 The following year, in 1004, King Henry dispatched a contingent of royal milites to help Jaromir, the exiled duke of Bohemia, capture the fortified city of Prague from Boleslav Chrobry.133 Thietmar mentions in this context that these milites were an elite force (optimi), as contrasted with the Bohemian militia levies (incoli) loyal to Jaromir.134 Boleslav Chrobry, for his part, also committed his milites to assaults on fortresses. According to Thietmar, in 1017 Boleslav dispatched a very substantial force of Moravian milites, here to be understood as mercenaries rather than the Polish duke’s household troops, to attack an as yet unidentified fortress (urbs) in Bohemia.135 Although the Moravians were initially successful, capturing the urbs and a substantial quantity of booty, they were not so fortunate during their retreat. Margrave Henry, of the Bavarian eastern march, caught them as they were burdened with their loot and killed more than 1,000 of the Moravian milites.136 Thietmar notes that in this same time frame, i.e. mid-August 1017, Boleslav Chrobry 127 128 129 130 131 132 133 134 135 136

Ibid. Ibid., 5.24. Ibid. Ibid. Ibid., “incaute frumentum equis pregregantes.” Ibid. Ibid., 6.12. Ibid. Ibid., 7.61. Ibid.

123

N A R R AT I V E W O R K S

dispatched another force of milites, who may have been drawn from his household forces, to assault the stronghold (urbs) of Belgern, located on the right bank of the Elbe in the modern district of Torgau-Oschatz.137 On this occasion, however, the Polish duke’s milites failed to obtain their objective even after devoting themselves to a long siege.138 Although Thietmar focuses most of his attention regarding military affairs on siege warfare, that is on the capture or defense of strongholds, the bishop also discusses the deployment of milites in other capacities.139 Thietmar records in several instances, for example, the deployment by magnates of their milites in the field in an effort to secure their jurisdictional rights or to pursue their private interests. As noted earlier, Margrave Ekkehard of Meißen dispatched a group of his milites to the village of Görschen in an effort to enforce his rights of judicial jurisdiction over the men living on his own lands, against the usurpation of these rights by the men of Görschen.140 Similarly, the daughter of Bishop Ansfrid of Utrecht, in her role as abbess, deployed her milites to gain control of her deceased father’s remains from the citizens of Utrecht.141 Thietmar also commented on the efforts of his own cousin Werner, the margrave of the Saxon northern march, to deal with the threat posed to his position by Count Dedi of Wettin.142 In this case, Dedi had made accusations at the royal court against Werner that were sufficiently serious to threaten the latter’s possession of his office and the beneficia that he held through it.143 After the case at the royal court was postponed through the intercession of Werner’s friends, the margrave decided to take direct action against Dedi, leading a troop of twenty milites, including Thietmar’s brother Frederick, for this purpose. When Werner located Dedi, he launched an attack, killing the count along with one of the count’s soldiers (miles), a man named Egilard.144 As punishment for taking matters into his own hands, Werner was stripped of his comital office and all of his beneficia by King Henry II. The latter were transferred to a count named

137 Ibid. 138 Ibid., “bello . . . inpugnantes diutino.” 139 Thietmar’s focus on siege warfare is consistent with the model provided by his source Widukind who similarly focuses the great majority of his attention on sieges as contrasted with battles in the field. The basic work on medieval siege warfare is now Jim Bradbury, The Medieval Siege (Woodbridge, 1992). See also Bernard. S. Bachrach, “Medieval Siege Warfare: A Reconnaissance,” Journal of Military History 58 (1994), 119–33, repr. with the same pagination in Bernard. S. Bachrach, Warfare and Military Organization in Pre-Crusade Europe (Bury St. Edmunds, 2002); Peter Purton, A History of the Early Medieval Siege, c. 450–1200 (Boydell, Woodbridge, 2010), and the enormous literature cited there. Cf. Guy Halsall, Warfare and Society in the Barbarian West 450–900 (Routledge, London, 2003), 215–27, whose claims are thoroughly outside the mainstream of scholarship dealing with medieval warfare because of his efforts to downplay the importance of sieges. 140 Thietmar, Chronicon, 4.73. 141 Ibid., 5.21. 142 Ibid., 6.48–98. 143 Ibid., 6.48. 144 Ibid., 6.49.

124

M E M O R Y, E P I S T E M O L O G Y, A N D T H E W R I T I N G

Bernard, whose father Dietrich had served as the margrave of the northern Saxon march from 965 to 985.145 These incidents all were relatively small-scale affairs. Thietmar does, however, discuss on occasion the deployment of larger forces of milites by magnates to pursue their own limited interests, as contrasted with the interests of the state (res publica). In 972, for example, Margrave Hodo of the eastern Saxon march attacked Duke Miesco I of Poland (d. 992) despite the fact, as Thietmar stresses, that Miesco had faithfully paid tribute to the German king.146 Because of the questionable nature of this operation, Hodo was unable to find many other magnates to join him in his campaign and had the support only of Count Siegfried, Thietmar’s father. Thietmar tried to excuse Siegfried for his lack of judgment on the basis of his youth, marked most notably by the fact that he was not yet married.147 As the campaign turned out, the optimi milites of Hodo and Siegfried were largely wiped out by the Polish forces, and the two commanders were lucky to escape with their lives.148 In his discussion of milites serving in the field, Thietmar also does not ignore the role they played in royal campaigns. In 1004, for example, Henry II marshaled a substantial army (exercitus) at Merseburg, having issued a general mobilization order to magnates throughout the Reich in mid-August of that year.149 Although publicized as an expedition to Poland, this was a ruse designed to lull Boleslav Chrobry into a false sense of security with regard to his control of Prague and much of Bohemia. Henry II’s real target, according to Thietmar, was Bohemia.150 Unfortunately for the German king, Boleslav’s spies received word of Henry’s intention to invade Bohemia and installed a force to block the German line of march. It is in this context that Thietmar notes Henry dispatched a force of picked (electi) milites to break through the ambush and establish a clear route for the rest of the army.151 Regarding this same year, Thietmar also discusses the deployment of royal milites, as well as the milites of numerous magnates, to quell an uprising in the city of Pavia, when Henry II was vindicating his claims to rule in northern Italy.152 145 146 147 148 149 150 151

Ibid., 6.50. Ibid., 2.29. Ibid., “pater meus Sigifridus, tunc iuvenis necdumque coniugali sociatus amori.” Ibid. Ibid., 6.10. Ibid., 254. Ibid. Henry II used a similar deployment to force a way through the Alpine passes being held by King Arduin of Italy’s forces in 1004. See Thietmar, Chronicon, 6.4. In this case, however, rather than utilizing milites regis, Henry I deployed Carinthian levies, who were familiar with the district, to clear the pass. Thietmar, following his normal usage, did not describe these Carinthians as milites since they were militiamen rather than professional soldiers. Nevertheless, since they served in the context of a royal army, Thietmar does describe their units as legiones, following his normal practice of reserving this term for units of armies led or authorized by German kings or Slavic princes. 152 Thietmar, Chronicon, 6.7–8.

125

N A R R AT I V E W O R K S

According to Thietmar, the king’s household troops (domestici) at first were hard put to defend the royal palace against the rampaging Pavian mob. However, as soon as word of the revolt reached the royal camp outside the city, Swabian, Franconian, and Lotharingian troops forced their way into the city and attacked the mob from the rear.153 The fighting ended in vicious house-to-house combat with the milites regis as well as the other German troops burning down a substantial number of buildings to root out the Pavians, i.e. local militia forces, who were using the rooftops to shoot their arrows down on the king’s soldiers.154 In one final example, on this occasion regarding events that took place in 1015, Thietmar discusses the deployment and loss in battle of a noteworthy group of milites serving in a royal exercitus. He records that at the end of a rather unsuccessful campaign against Boleslav Chrobry, Henry II left a rearguard under the command of Archbishop Gero of Magdeburg, Margrave Gero of the east Saxon march, and Burchard, the count palatine of Saxony (1002–1017), to protect the army’s withdrawal from Boleslav’s lands.155 Thietmar laments that soon after Henry departed, the rearguard sustained a massive assault, resulting in very heavy losses, including Margrave Gero, a count named Folkmar, and 200 optimi milites who died in the fighting.156

Equipment of milites Although the milites discussed by Thietmar in the Chronicon are to be found most frequently deployed either in the defense of fortifications or in operations intended to capture strongholds – types of warfare that generally necessitated combat on foot – Thietmar nevertheless makes clear that some milites were provided with horses. Indeed, Thietmar even indicates that there was a soldierly manner of sitting a horse. In discussing Beleknegini, a Hungarian noblewoman, the bishop of Merseburg remarked that she rode more militis.157 The fact that Thietmar did not feel compelled to discuss what this peculiarly military type of horsemanship entailed indicates that he expected his audience to know how soldiers sat their horses and thus how Beleknegini did as well. In Thietmar’s discussion of the use of horses by some milites, particularly by high-ranking men, it seems clear that such soldiers owned their own mounts. In discussing the fall from a horse by his father Siegfried during his last campaign, for example, it is clear that Thietmar was discussing an accident involving 153 These fighting men are designated by Thietmar by their ethnic/regional origin rather than as milites. 154 Thietmar, Chronicon, 6.8. 155 Ibid., 7.21. 156 Ibid. Die Totenbücher von Merseburg, Magdeburg und Lüneburg, ed. G. Althoff and J. Wollasch, MGH Libri Memoriales et Necrologia n. s. 2 (Hanover, 1983), fol. 5 r., 11, records the deaths of “Gero et Vuolcmar comites cum sociis.” In this case, socii is a synonym for milites, i.e. household troops. 157 Thietmar, Chronicon, 8.4.

126

M E M O R Y, E P I S T E M O L O G Y, A N D T H E W R I T I N G

Siegfried’s own animal.158 In other instances, however, the context of Thietmar’s discussion of the deployment of milites with horses suggests that the soldiers in question did not own their animals. For example, in discussing the service by milites of the bishops of Strasbourg and Basel at the fortress of Breisach, noted previously, Thietmar notes that the soldiers went out each day “ob acquirenda equorum pabula,” that is they went out to gather fodder for their horses.159 Obviously, the collection of fodder is part of the workload required of low-ranking soldiers, though not of wealthy men who themselves employed servants to undertake time-consuming and mundane labor.160 Whether low-ranking soldiers could afford expensive horses trained for military service is certainly an open question, but it does seem unlikely.161 It should also be noted that although Thietmar described some milites as being deployed with horses, the author by no means treats the term miles as synonymous with mounted soldiers. In his discussion of King Henry II’s campaign against Boleslav Chrobry in 1017, for example, Thietmar notes that following the failure of the German king’s army to capture the fortress of Glogau, the Polish duke rejoiced with his milites.162 In the very next sentence, Thietmar records that Boleslav undertook a counterstroke against Henry II by dispatching a column of 600 pedites from his army to ravage Bohemia.163 A careful reading of this passage makes clear that far from distinguishing these 600 pedites from milites understood as mounted troops, Thietmar was making clear that Boleslav dispatched these 600 foot-soldiers from among the milites with whom he had celebrated the victory won at Glogau. Indeed, when Thietmar wishes to distinguish foot soldiers from men with mounts, he contrasts these pedites with equites, as was the case in his discussion of the “thirty legiones” of Slavs who rose in revolt against Otto II in 983, rather than contrasting pedites with milites.164 In this context, it cannot be ignored that in contrast to his frequent discussion of milites in deployments in which it was necessary for them to fight on foot, Thietmar never explicitly discusses the deployment of mounted soldiers in combat operations, although some cases do permit the inference that some

158 Ibid., 4.16. 159 Ibid., 5.21. For another example of milites, themselves, leaving the fortification that they were garrisoning in order to gather fodder, see Thietmar, Chronicon, 5.9. 160 Regarding the low-status nature of such work on campaign, see Leyser, “Henry I,” 21. 161 Commenting on this very question, Leyser, “Henry I,” 41, emphasized that magnates in Ottonian Germany frequently were provided mounts by the king. The issue here is one of “payment.” The high value of horses made them desirable rewards even for magnates. Concerning the very high costs of war horses, see Bernard S. Bachrach, “Animals and Warfare in Early Medieval Europe,” in Settimane di studio del centro italiano di studi sull’alto medioevo, vol. 31 (Spoleto, 1985), 707–64, and reprinted with the same pagination in Armies and Politics in the Early Medieval West (London, 1993). 162 Thietmar, Chronicon, 7.64. 163 Ibid. 164 Ibid., 3.19.

127

N A R R AT I V E W O R K S

milites did operate militarily on horseback, rather than simply riding horses from place to place, as was the case among the Anglo-Saxons in contemporary England.165 Finally, it must be emphasized that Thietmar never discusses the deployment of soldiers carrying out the tactics of what has been termed by some scholars as “mounted shock combat.” The other element of equipment in addition to mounts that Thietmar occasionally discusses in the context of milites are hauberks (lorica).166 As noted earlier, Duke Boleslav Chrobry indicated his acceptance of Emperor Otto III’s superior status by providing him with a contingent of 300 milites for service with the royal army. Thietmar emphasizes that these men were loricati and that this fact was especially pleasing (maxime sibi placuit) to the emperor.167 Thietmar also points out that the picked milites whom Henry II dispatched in 1010 in an effort to clear a path through Boleslav Chrobry’s archers in the forest of Miriquidui were loricati.168 In the one final case in which Thietmar explicitly connects the term lorica with miles, he is discussing Cnut’s siege of London in 1016. In this context, the Danish ruler demanded that the defenders hand over 24,000 loricae, a number that Thietmar describes as unbelievable (numerus incredibilis), as part of the terms of surrender.169 Later in the

165 For example, the deployment of 400 milites to protect King Henry II’s foragers at the siege of Creußen, noted earlier, may well have included a mounted element. Thietmar’s hesitancy on this point stands in marked contrast to his source Widukind who does discuss the deployment of Saxon and Thuringian mounted forces in combat operations using the tactic of the feigned retreat, most notably at the battles of Lenzen and Riade. See in this regard, Widukind, Res gestae, 1.36, and 1.38. 166 In this context, it should be emphasized that Thietmar uses the classical word lorica and not the medieval term brunia. 167 Thietmar, Chronicon, 4.46. 168 Ibid., 6.10. 169 Ibid., 7.40. In his translation of this passage, Warner, Ottonian Germany, 335, construes the number 24,000 as referring to the pounds of silver required by Cnut to redeem the lives of the soldiers serving in the defense of the city. The pertinent section of the sentence reads as follows: “si regina voluisset dare filios suos in mortem seque cum XV milibus argenti ponderibus et episcopos cum XII milibus et omnibus loricis, quarum milia XXIIII numerus incredibilis erat redimere.” The grammatical structure of this sentence clearly mitigates against Warner’s translation of this sentence as “if the queen were willing . . . to redeem herself for fifteen thousand silver pounds, redeem the bishops for twelve thousand, and all of her armed men for the unbelievable sum of twenty-four thousand pounds.” Both the queen (se) and the bishops (episcopos) are placed in the accusative as the direct objects of “redimere.” The word “loricis,” by contrast, is in the ablative following from “cum” meaning with. Moreover, the term “loricis” is not an adjective meaning “armed,” but rather a noun, that is haubkerks. When Thietmar wished to describe a man as armed or a force of armed men, he used the term armatus or armati. However, even if lorica could be used adjectivally in place of the proper term loricatus, Thietmar does not use it this way. Instead, when he wants to describe men as being equipped with hauberks, he describes them as loricati. Finally, the antecedent of the possessive relative pronoun “quarum” must be the feminine noun loricae rather than the third declension neuter noun pondera as indicated by the ending “arum” rather than “orum.” The correct translation of this sentence, as recognized in the translation by Trillmich (p. 396), is as follows: “If the queen wished . . . to redeem herself for fifteen thousand silver pounds, redeem the bishops for twelve thousand pounds, along with all twenty-four

128

M E M O R Y, E P I S T E M O L O G Y, A N D T H E W R I T I N G

same passage, Thietmar uses the loricae as a synecdoche for the queen’s milites, i.e. the armored ones.170 Given the frequency with which Thietmar discusses the deployment of milites, the rarity of his references to loricae requires some consideration. It does not necessarily follow that Thietmar’s infrequent references to hauberks means that this item of equipment was relatively rare among the professional fighting men of Ottonian Germany. However, it must be asked why, if all milites are to be thought of as having been equipped with loricae, Thietmar felt that it was necessary on occasion to emphasize the nature of their equipment. Perhaps in regard to Duke Boleslav’s troops, noted previously, Thietmar’s comment can be taken to suggest that Polish milites were not as well equipped, in general, as Germans, and a German audience would not normally expect Polish troops to have hauberks. In this context, it is worth recalling the comments that Thietmar put into the mouth of the miles Slopan, a scout serving Duke Boleslav Chrobry, when the former reported his observations regarding four legiones of German troops, one of whose commanders was Thietmar’s father Siegfried.171 According to the bishop of Merseburg, Slopan told Boleslav that although the German force (exercitus) was small, it consisted of elite troops (qualitate sua optimus) and, just as importantly, the men were all equipped with armor (ferreus).172 In this context, Slopan suggested to his duke that although the Poles might be victorious, their losses to such a well-equipped foe would be so great that they would be unable to face another enemy. One possible explanation for Thietmar’s decision to emphasize that the milites deployed by Henry II at the forest of Miriquidui also were equipped with loricae is that not all German milites normally had hauberks. Thietmar may have wished his audience to understand that Henry dispatched the soldiers with the correct equipment to undertake the dangerous task of clearing a road guarded by archers. Unfortunately, without the type of administrative documents available for studies of military equipment in later periods, e.g. thirteenth-century England, it cannot be known with any degree of certainty whether the lorica was a standard piece of equipment for milites. Thietmar’s discussion of loricae does suggest,

thousand hauberks, an incredible number.” Warner’s confusion on this point is understandable given that Thietmar does refer synecdochally to the term loricae later in the same passage mentioning the “prenominatis militibus.” 170 Thietmar, Chronicon, 7.40. Thietmar also, occasionally, notes that fighting men, whom he does not refer to as milites, were equipped with loricae. In his discussion, for example, of the forces commanded by several German magnates as well as Duke Jaromir of Bohemia as they passed near the fortress of Glogau in 1010, Thietmar notes that these men were loricati. See Thietmar, Chronicon, 6.58. Although Thietmar does not describe these troops as milites, he does place a speech in the mouth of Boleslav Chrobry, who saw them march by, in which the Polish duke, according to Thietmar, described the contingents as a picked (electus) force, indicating that they probably were professional soldiers. 171 Thietmar, Chronicon, 4.12. 172 Ibid.

129

N A R R AT I V E W O R K S

however, that even if a significant portion of milites in any particular context were equipped with hauberks, this was by no means a universal phenomenon.

Conclusion In his well-informed and detailed observations regarding the military organization of Ottonian Germany, Bishop Thietmar of Merseburg devotes considerable attention to milites. A thorough examination of all of the contexts in which Thietmar uses the term makes clear that the bishop reserved the term to refer to professional soldiers. These milites served in the military households of both ecclesiastical and secular magnates, including the kings of Germany and Slavic princes. Indeed, Thietmar consistently employs the term miles to refer to professional soldiers from a variety of ethnic backgrounds and ranks, including pagan Slavs, AngloSaxons, and even the Byzantine general John Tzimiskes. In addition, the bishop uses the term miles to refer to those quintessentially professional soldiers, that is mercenaries. It should be noted that Thietmar does not use the term miles in a socalled “feudal” sense to denote the holders of beneficia or vassals. Thietmar also does not use the term miles to denote a man’s social status or as an honorific. In Thietmar’s usage, miles has a strictly military meaning without any of the social connotations so famously claimed by Georges Duby for the contemporary milites of the Mâconnaise, which was highly anachronistic.173 As might be expected given the exceptional importance accorded to fortifications by political and military leaders in the German kingdom and surrounding lands, the majority of the milites depicted by Thietmar in military operations served in the context of siege warfare. The Ottonian kings from Henry I to Henry II maintained a policy of building frontier fortifications that were to be garrisoned by milites provided by both secular and ecclesiastical magnates. Slavic princes, including Duke Boleslav Chrobry of Poland, similarly garrisoned strategic strongholds with their own milites. Within the German kingdom, bishops, counts, margraves, and dukes regularly garrisoned their strongholds with household milites. The corollary to these defensive preparations was the regular deployment of milites by kings, princes, and magnates to besiege fortifications, fortress cities, and other strongholds. Although milites occasionally are seen in Thietmar’s Chronicon to serve in the field, the bishop of Merseburg clearly presents the capture and defense of fortifications as the primary military occupation of milites in the Ottonian period. It is in this context that Thietmar’s infrequent references to horsemen and very rare discussion of equipment, such as hauberks, utilized by heavily armed

173 With respect to the enoblement of the milites in the mâconnaise during the late tenth as well as the eleventh century, see Georges Duby, La Société aux XIe et XIIe siècles dans la région mâconnaise (Paris, 1953, and reprinted with different pagination in 1971), 191 ff. For a critique of Duby’s methods and analysis regarding the noble nature of the milites in the French kingdom during this period, see Bernard S. Bachrach, “The Milites and the Millenium,” The Haskins Society Journal 6 (1994), 85–95.

130

M E M O R Y, E P I S T E M O L O G Y, A N D T H E W R I T I N G

mounted troops are to be understood. On the basis of Thietmar’s depiction of military affairs, it is hard to escape the conclusion that mounted combat of all types, and particularly the type of warfare characterized by scholars as “mounted shock combat” in which heavily armed and armored warriors charged the enemy in serried ranks, was rare in Ottonian Germany. In short, Thietmar’s milites, particularly those in the German kingdom and Slavic lands, were professional soldiers employed primarily, if not exclusively, to serve the needs of siege warfare.

131

6 D U D O O F S T. Q U E N T I N A S A N H I S TO R I A N O F M I L I TA RY O R G A N I Z AT I O N

Modern scholars have long considered the early eleventh-century author Dudo of St. Quentin to have been a failure as a historian.1 This judgment has been rendered because Dudo’s De Moribus et Actis primorum Normanniae ducum (sometimes referred to by scholars as Gesta Normannorum) is filled with inaccuracies.2 Dudo’s failure to get the facts right is believed by many scholars to have been intentional for the most part. It is argued that apparently Dudo cared little or not at all about setting out in an accurate manner the details of the political narrative that he constructed.3 It is clear, nevertheless, that the contemporary ruling 1 Francis Palgrave, History of Normandy and England, 4 vols. (Cambridge, 1852–1864), II: 908, questioned Dudo’s accuracy in a fundamental manner as early as 1857. A quarter century later Henry Howard, “A Criticism of the Life of Rollo, as told by Dudo de St. Quentin,” Archaeologia 45 (1880), 223–50, make clear many of his delicts. Finally, Henry Prentou, Étude critique sur Dudon de S. Quentin et son Histoire (Paris, 1916), mustered the massive detail which demolished De Moribus as a source for early Norman political history. Prentout also reviews much of the earlier scholarly literature (3–13, 114–32). This critical tradition is dealt with, as well, in considerable detail by David S. Douglas, “Rollo of Normandy,” English Historical Review 57 (1942), 417–36, and more generally in idem, “The Rise of Normandy,” Proceedings of the British Academy 33 (1947), 101–30. Both are reprinted in David C. Douglas, Time and the Hour (London, 1977), 95–119 and 121–40, respectively. 2 The basic edition remains Dudo: De Moribus et Actis primorum Normaniae ducum, ed. J. Lair (Caen, 1865). Gerda C. Huisman, “Notes on the Manuscript Tradition of Dudo of St. Quentin’s Gesta Normannorum,” Proceedings of the Battle Conference on Anglo-Norman Studies 6 (1984), 122–35, has promised a new edition. In addition, Eric Christiansen, Dudo of St. Quentin, History of the Normans (Woodbridge, 1998), has provided a complete translation with a useful introduction and helpful notes. Unfortunately, as noted below, in dealing with technical military matters Christiansen’s translation is often misleading and occasionally wrong. All translations herein are my own unless otherwise noted. 3 Emily Albu (Hanawalt), “Dudo of Saint-Quentin: The Heroic Past Imagined,” Haskins Society Journal 6 (1994), 111–18, here 113 correctly observes that this negative view by modern scholars is encouraged by the fact that Dudo “apparently had access to archival materials and to the duke’s own memories and family records” but chose rather to write a “Vergilian” epic of no historical value in tracing the political narrative of the Norman rulers. For additional observations, see Emily Albu, The Normans and their Histories: Propaganda, Myth, and Subversion (Woodridge, 2001). Some scholars, however, have tried to vindicate Dudo as an historian who reported the facts in an essentially accurate manner. The strongest proponent of this view has been Johannes Steenstrup,

132

D U D O O F S T. Q U E N T I N A S A N H I S T O R I A N

family in Normandy, Dudo’s patrons, commissioned him to write a plausible story in praise of their ancestors. In short, the story, which pretended to be history, had to be credible as fact rather than fantasy, at least to the patrons and very probably to those in Dudo’s likely audience, whose opinions these patrons may be thought to have valued.4 Dudo was not unaware of the need for plausibility in this context.5 He is known to have provided accurate material details to contemporaries about matters with

“Normandiets Historie under de syv forste hertuger,” in Det Kongelige Dansk Videnskaberess Selskabs Skrifter, Historisk og Filosophisk Afdeling, 7th series, vol. 1 (Copenhagen, 1925), 1–292 with a French résumé 293–310. This effort, it is widely agreed, has failed. As David Bates, Normandy Before 1066 (London, 1982), 10–11 observes, “A series of studies by Professors Lucien Musset and Jean Yver, published over the last forty years and culminating in the latter’s magisterial survey of the province’s institutions, have consigned Dudo’s opinions to an oblivion from which they will surely never return.” However, a recent popular work by François Neveux, La Normandie des ducs aus rois Xe-XIIe Siècle (Rennes, 1998), 22–7, 40, relies strongly on Dudo and disparages previous critiques. By contrast with the critical approach to Dudo’s work illuminated in the commentary by Bates, a growing number of scholars have argued that it is essentially wrong to judge Dudo by the standards normally applied to historians, even eleventh-century historians, because he was trying to write in the Norse saga tradition. See, for example, August Nitschke, “Beobachtungen zur normannische Erziehung,” Archiv für Kulturgeschichte 43 (1961), 273–82; Frederick Amory, “The Viking Hasting in Franco-Scandinavian Legend,” in Saints, Scholars and Heroes: Studies in Medieval Culture in Honor of Charles W. Jones, ed. Margot H. King and Wesley M. Stevens, 2 vols. (St. Johns, MN, 1969), 265–86; Elisabeth van Houts, “Scandinavian Influence in Norman Literature of the Eleventh Century,” Anglo-Norman Studies 5 (1983), 107–21; Eleanor Searle, “Fact and Pattern in Heroic History: Dudo of St. Quentin,” Viator 15 (1984), 75–85; Jan Ziolkowski, Jezebel: A Norman Latin Poem of the Early Eleventh Century (New York, 1989); and Marcello Meli, “Dudone di S. Quintino e la preistora vichinga,” in Dudone di san Quintino, ed. Paolo Gatti and Antonella Degl’Innocenti (Trent, 1995), 29–48. Christiansen, Dudo, xvi–xvii, effectively points out the weaknesses inherent in the saga model. He also effectively argues (xviii–xix) against efforts to demonstrate that early French vernacular written and/or oral accounts provided Dudo with his sources and models. This tradition of “romance” origins goes back to the nineteenth century and was still being argued in the late twentieth century. See, for example, R. H. C. Davis, The Normans and their Myth (London, 1976), 58–9. Karsten Friis-Jensen, “Dudo of St. Quentin and Saxo Grammaticus,” in Dudone di San Quintino, 11–28 casts strong doubt on the idea that Dudo uses a Scandinavian model. However, it should be noted that van Houts, “Scandinavian Influence,” makes some useful points with regard to connections between Normandy and Scandinavia even though the importance of a few of her examples is exaggerated. 4 Scholars have provided a broad spectrum of arguments regarding the aims of Dudo’s patrons. For example, Prentout, Étude critique, 19–23, saw the patrons of De Moribus as wanting a work of “propaganda” that was intended to make the Normans respectable. Searle, “Fact and Pattern,” 122–3, sees the patrons as desiring a work of family praise in the “Germanic” tradition of the “literature of house and kindred.” Felice Lifschitz, “Dudo’s Historical Narrative and the Norman Succession of 996,” Journal of Medieval History 20 (1994), 101–20 argues that Richard I wanted a work that was intended to help secure the succession of his son and heri Richard II. Idem, “Translating ‘Feudal’ Vocabulary: Dudo of Saint-Quentin,” Haskins Society Journal 9 (2001 for 1997), 39–56 begins with this premise. With regard to Dudo’s patrons, themselves, see Christiansen, Dudo, xxiii–xxvii, where previous scholarship is well summarized. 5 With respect to the question of rhetorical plausibility see Nancy Partner, “The New Cornificius: Medieval History and the Artifice of Words,” in Classical Rhetoric and Medieval Historiography, ed.

133

N A R R AT I V E W O R K S

which they were well acquainted. These include, for example, the physical realities of the ducal palace at Fécamp and various types of legal documentation.6 It will be argued here that as a result of the need to be plausible, Dudo worked diligently to treat the matter of military organization in a largely accurate manner. Indeed, this was certainly a subject about which he knew that his patrons in the ducal family and many in his potential audiences at the Norman court likely were well informed.

Military organization Throughout the early Middle Ages prior to the First Crusade and, indeed, for a long time thereafter, Western Europe’s military strength was organized on the basis of three interrelated institutional structures. At the most basic level, all able-bodied men, regardless of legal or social status, were required to participate in the local defense when needed. The majority of these militia men, however, were mobilized only for local purposes and were not required to participate in military operations outside their home districts because, in general, they were regarded as being too poor to sustain such efforts on a regular basis. A relatively small percentage of militia men, however, were required to participate in military operations outside their home district, that is to undertake what the sources call expeditiones. This military service was based on an assessment of the wealth of the person who was required to perform such service. Finally, the important people in society, both lay and ecclesiastical, abbesses as well as abbots, maintained household military forces. Among these important people, the king generally was the most important and his obsequium, based both as presentales in and around the court and in various other venues throughout the regnum, especially when joined with the military Ernst Breisach (Kalamazoo, 1985), 5–59, here 12; Roger Ray, “Rhetorical Scepticism and Verisimilar Narrative in John of Salisbury’s Historia Pontificalis,” in ibid., 61–102, here 83–4; and Justin C. Lake, “Truth, Plausibility, and the Virtues of Narrative at the Millennium,” Journal of Medieval History 35 (2009), 221–38. Also see the series of articles by John Bliese, “Rhetoric and Morale: A Study of Battle Orations from the Central Middles,” Journal of Medieval History 15 (1989), 201–26; idem, “The Courage of the Normans: A Comparative Study of Battle Rhetoric,” Nottingham Medieval Studies 35, 91991), 1–16; and “Rhetoric Goes to War: The Doctrine of Ancient and Medieval Military Manuals,” Rhetoric and Society Quarterly 24 (1994), 105–30. However, cf. Christiansen, Dudo, xxi, who asserts that Dudo’s “first consideration was aesthetic”. By contrast, Christiansen, at xxix–xxxiv, provides a very satisfactory treatment of Dudo’s rhetorical techniques and observes (xxxii): “The plain passages are no less artificial than the coloured, since they observe the rhetorical plan of inspiring confidence by lucidity, verisimilitude, and brevity.” 6 For example, as shown by Annie Renaux, “Fouilles sur le site du château ducal de Fécamp (XeXIIe siècle),” Proceedings of the Battle Conference on Anglo-Norman Studies 4 (1982), 133–52, here 138–9 and 144, Dudo’s descriptions of the ducal château are demonstrated to be accurate. In a similar vein, Emily Tabuteau, Transfers of Property in Eleventh-Century Norman Law (Chapel Hill, NC, 1988), after noting the traditional wisdom (12 and 383 n. 39) that Dudo is untrustworthy, goes on to demonstrate in considerable detail (51–5, 61–2, 103, 198, 302–3, and 330–1) that with regard to the legal or institutional aspects of property transfers he is, by and large, quite trustworthy.

134

D U D O O F S T. Q U E N T I N A S A N H I S T O R I A N

households of the magnates, constituted a standing army of professional soldiers that could be mobilized rapidly at governmental command and which might be of considerable size. On occasion early medieval governments also reinforced their armies with mercenaries.7

Local levies During the later Roman Empire, the imperial authorities at the highest level made a successful decision to militarize the civilian population. This process was well in train by the fifth century in the countryside, as well as in the walled cities and other fortified population centers of the West.8 In Gaul, these developments saw all able-bodied men, regardless of social or legal status, mobilized for the defense of the local area (civitas, pagus, gau, county) in which they dwelled.9 7 With respect to military organization in this period, see Bernard S. Bachrach, “Medieval Military Historiography,” in Companion to History, ed. Michael Bentley (London, 1997), 203–20; and now Bernard S. Bachrach and David S. Bachrach, Warfare in Medieval Europe c. 400–c.1453 (London, 2016), particularly chapter 3, with the literature cited there. Cf. the works of Timothy Reuter, “Plunder and Tribute in the Carolingian Empire,” Transactions of the Royal Historical Society, 5th series 35 (1985), 75–94; “The End of Carolingian Military Expansion,” in Charlemagne’s Heir: New Perspectives on the Reign of Louis the Pious (814–840), eds. Peter Godman and Roger Collins (Oxford, 1990), 391–405; “Carolingian and Ottonian Warfare,” in Medieval Warfare: A History, ed. Maurice Keen (Oxford, 1999), 13–35; and Guy Halsall, Warfare and Society in the Barbarian West, 450–900 (London, 2003), whose efforts to insist that early medieval armies must have been small because of the putatively primitive nature of contemporary institutions founder on their failure to assimilate the vast and growing volume of source material that points to the centrality of sieges to early medieval warfare and the concomitant requirement for large, effectively organized armies. In this regard, see the work by Leif Inge Ree Petersen, Siege Warfare and Military Organization in the Successor States (400–800 AD): Byzantiume, the West and Islam (Leiden, 2013); as well as Peter Purton, A History of the Early Medieval Siege, c. 450–1200 (Boydell, Woodbridge, 2010), and the enormous literature cited there for siege warfare throughout the early and high Middle Ages. 8 In this context, see the important studies by Stefan Esders, “‘Öftenliche’ Abgaben und Leistungen im Übergang von der Spätantike zum Frühmittelalter: Konzeption und Befund,” in Von der Spätantike zum frühen Mittelalter: Kontinuitäten und Brüche, Konzeptionen und Befunde, ed. Theo Kölzer and Rudolf Schieffer (Ostfildern, 2009), 189–244; idem, “Zur Entwicklung der politischen Raumgliederung im Übergang von der Antike zum Mittelalter – Das Beispiel des pagus,” in Politische Räume in vormodernen Gesellschaften: Gestaltung, Wahrnehmung, ed. Ortwin Dally et al. (Rahden, 2013), 185–201; idem, “Nordwestgallien um 500: Von der militarisierten spätrömischen Provinzgesellschaft zur erweiterten Militäradministration des merovingischen Königtums,” in Chlodwigs Welt: Organisation von Herrschaft um 500, ed. Mischa Meier and Steffen Patzold (Stuttgart, 2014), 339–61; as well as Walter Goffart, “Frankish Military Duty and the Fate of Roman Taxation,” Early Medieval Europe 16 (2008), 166–90. 9 See Bernard S. Bachrach, “Imperial Roots of Merovingian Military Organization,” in Military Aspects of Scandinavian Society in a European Perspective, AD 1–1300, ed. Anne Norgard Jorgensen and Birthe L. Clausen (Copenhagen, 1997), 25–31; idem, “Quelques observations sur la composition et les caractéristiques des armées de Clovis,” in Clovis: Histoire et Mémoire, ed. Michel Rouche 2 vols. (Paris, 1998), I: 689–703; and Esders, “Nordwestgallien um 500,” 339–61.

135

N A R R AT I V E W O R K S

The use of both urban and rural populace for local defense continued throughout the Merovingian period.10 In the Carolingian world as well, the deployment for defensive purposes of these militia forces at the local level was continued.11 It is notable, however, that as the Carolingian Empire and its successor states were more frequently on the defensive over the course of the ninth and tenth century due to attacks by Vikings, Magyars, and Saracens, these local levies received more attention from contemporary authors. While royal government in Francia Occidentalis declined more thoroughly than in Francia Orientalis during the period following the death of Charlemagne’s grandsons, the local levies were still widely used for defensive purposes in both realms.12 This obligation to participate in the local defense was the norm throughout the West. Indeed, even in England, where direct continuity from the institutions of the later Roman Empire to the AngloSaxon kingdoms of the early Middle Ages is certainly problematic and, indeed, unlikely, the mobilization of all able-bodied men for the local defense was the norm.13 Dudo’s account of military matters in De Moribus treats the local levies in a manner that fundamentally is consistent with the history of this practice up to

10 Bernard S. Bachrach, Merovingian Military Organization, 451–751 (Minneapolis, MN, 1972), although in need of updating makes this basic point clear. 11 Bernard S. Bachrach, “Military Organization in Aquitaine under the Early Carolingians,” Speculum 49 (1974), 1–33; and reprinted with the same pagination in idem, Armies and Politics in the Early Medieval West (London, 1993). For a broader view of local militia forces, see Bernard S. Bachrach, Early Carolingian Warfare: Prelude to Empire (Philadelphia, 2001), 52–4; David S. Bachrach, Warfare in Tenth-Century Germany (Woodbridge, 2003), 71–7; and idem, “Civilians and Militia in Ottonian Germany: Warfare in an Era of Small Professional Armies,” in Civilians and Warfare in World History, ed. Nicola Foote and Nadya Williams (London, 2017), 110–31. 12 For the strength of the military institutions in Ottonian Germany, see D. Bachrach, Warfare in Tenth-Century Germany, passim; and Bernard S. Bachrach, “Magyar-Ottonian Warfare: À-propos a New Minimalist Interpretation,” Francia 27 (2000), 211–30. With respect to the control over local levies by regional rulers in West Francia during the tenth and early eleventh century, see Bernard S. Bachrach, “The Angevin Strategy of Castle-Building in the Reign of Fulk Nerra, 987– 1040,” American Historical Review 88 (1983), 533–60; and idem, “Angevin Campaign Forces in the Reign of Fulk Nerra, Count of the Angevins (987–1040),” Francia 16 (1989), 67–84. 13 The basic treatment of the great fyrd remains C. Warren Hollister, Anglo-Saxon Military Institutions on the Eve of the Norman Conquest (Oxford, 1962), 25–37. The effort by Richard Abels, Lordship and Military Obligation in Anglo-Saxon England (Berkeley, 1988), to undermine Hollister’s formulation was not successful as he relied on the no-longer accepted notion of the “king’s free” to argue against the public nature of military obligation and to assert instead a private relationship between lord and man. This model was demolished during the 1960s and 1970s by Hermann Krause, “Die Liberi der Lex Baiuvariorum,” in Festschrift für Max Spindler zum 75. Geburtstag, ed. Dieter Albrecht, Andreas Kraus, and Kurt Reindel (Munich, 1969), 41–73; and Johannes Schmitt, Untersuchungen zu den liberi homines der Karolingerzeit (Frankfurt, 1977). For a thorough discussion of this problem, now see Bernard S. Bachrach, “Charlemagne’s Expeditionary Levy: Observations Regarding Liberi Homines,” Studies in Medieval and Renaissance History 3rd series 12 (2015), 1–65, and the literature cited there; and David S. Bachrach, “Royal Justice, Freedom, and Comital Courts in Ottonian Germany,” forthcoming in Zeitschrift der Savigny-Stiftung für Rechtsgeschichte, germanistische Abteilung.

136

D U D O O F S T. Q U E N T I N A S A N H I S T O R I A N

the early eleventh century and, indeed, beyond. For example, in the course of discussing the manner in which Rollo putatively acquired the city of Rouen and its civitas or pagus, that is the Rouenais, Dudo tells his audience that the only forces defending the entire region were inerme vulgus. These, of course, were people who lacked defensive armament or armor. They were not, as some scholars unschooled in military history believe, men who engaged in combat without weapons, that is unarmed.14 Indeed, the city of Rouen itself, Dudo reports, was defended by a rather small force (parva manus) of such similarly lightly armed men (inermes).15 In a battle harangue to the troops that is treated as direct discourse, Dudo has Rollo, a possible ancestor of his patrons, develop the theme of the general military unpreparedness of the Rouenais in the context of Viking preparations to conquer the region. In the process of “quoting” Rollo’s harangue or battle oration, as such speeches are traditionally characterized from a rhetorical perspective, Dudo has the Scandinavian commander aver that the region is lacking in regular professional soldiers, that is milites. Rollo even goes so far as to allege that the region is also lacking in men who regularly were provided with defensive military equipment, that is armigeri.16 An audience of senior military men at the Norman court, including the duke, who could be made aware of the essence of Dudo’s account, would likely understand that battlefield orations were supposed to contain a modicum of exaggeration because they were crafted for the purpose of raising the morale of the troops.17 Parenthetically, it should be emphasized that Dudo provides a noteworthy example of his understanding of contemporary prebattle practice while at the same time evidencing a considerable rhetorical subtlety by having his protagonist exaggerate the enemy’s weakness in a context in which such exaggeration generally was regarded as the norm in order to sustain morale. Dudo, however, does make clear that both he and Rollo are exaggerating the lack of readiness of the men of the Rouenais to defend the region. In this context, Rollo depicts the Viking commander telling his troops that they must win control

14 This matter is discussed in some detail by Bernard S. Bachrach, “The Northern Origins of the Peace Movement at Le Puy in 975,” Historical Reflections/Reflexions historiques 14 (1987), 405–21; and reprinted with the same pagination in idem, State Building in Medieval France: Studies in Early Angevin History (London, 1995), 417–18. 15 Dudo, De Moribus, 2.2. Cf. Christiansen, Dudo, 35, who translates “inerme vulgus” as “weaponless commoners” and “parvam manum et inermem” as “a small and weaponless garrison,” respectively, while at 191 n. 160 he claims that “inerme vulgus” is a cliché meaning unarmed. Inermes in the present context, as in most military contexts in pre-Crusade Europe, does not mean weaponless, but rather lacking in armor. See Bachrach, “Northern Origins,” 405–21. 16 Dudo, De Moribus, 2.12. Christiansen, Dudo, 36 translates milites as knights, which is an impossible meaning for the north of Francia Occidentalis in the early eleventh century, much less a century or so earlier. See the discussion later. 17 See Bliese, “Rhetoric and Morale,” 201–26; idem, “The Courage of the Normans,” 1–16; and idem “Rhetoric Goes to War,” 105–30.

137

N A R R AT I V E W O R K S

of the oppida, vici, castra, and urbes in battle (praelia) against the inhabitants (gentium) of the pagus. Clearly, the putative lack of first-class and perhaps even second-class troops, for example, milites and armigeri, in the region during this period would not result in a walkover by the Scandinavian invaders. Rather, the militarized general population, or as modern scholars refer to them, the local levies, had to be reckoned with, and this was especially the case when the capture of fortifications was at issue.18 For obvious reasons, local levies were likely to be more effective when protected by defenses rather than in the open field. Dudo makes this point regarding the local levies on several occasions in the context of sieges where, despite their lack of heavy armament, the militia of cives living within the walls of a city were able to operate effectively. Dudo’s accounts of the defense of the erstwhile Roman fortress cities of Paris and Chartres makes this point clearly.19 Dudo obviously was correct when he decided to emphasize that local levies of the defense militia largely were lacking in armor, that is that they were inermes. However, these forces, because they were locals living near the scene of the conflict, compensated for this limitation in several ways. In particular, they were available for rapid mobilization in large numbers near the point of enemy attack. Dudo’s account rings true when he indicates that the territorii pagenses, having been informed of a Viking landing, rapidly mobilized an aciem maximam to oppose the invasion force.20 In addition, these militia troops could be deployed behind city walls in order to defend an urban fortress or some lesser stronghold such as a castrum, which were obvious targets for attack. These fortifications, especially the fortress cities, were the places where the greatest quantities of moveable wealth were to be found and were thus attractive as targets for plundering expeditions. Moreover, armies bent on territorial conquest had to capture the fortifications of the region, as the quotation earlier from Dudo’s account of Rolo’s harangue to his troops makes clear. Although Dudo varies his style in describing the mobilization of local levies, the institutional substance remains constant. We learn that the Frisiones mobilized a multitudo of plebs rapidly (concite) to deal with a Viking invasion force that made landfall in their area.21 Dudo provides somewhat greater detail when describing the actions of the Walgrenses (inhabitants of Walcheren) who, having learned that a barbarian people had arrived at their shores, mobilized a large force of the local population of the pagus.22 With regard to another episode, Dudo 18 Dudo, De Moribus, 2.12. 19 Ibid., 2.17. 20 Ibid., 2.5. It is of considerable interest here that Dudo is describing the local defenders of one of England’s eastern shires. It is not clear whether he was familiar with the institution of the great fyrd in the Anglo-Saxon kingdom or is merely describing the latter in terms common to the northern parts of the erstwhile Regnum Francorum. 21 Dudo, De Moribus, 2.9. 22 Ibid., “audientes . . . quod gens barbara . . . suis littoribus esset advect, congregata multitudine pagensium.”

138

D U D O O F S T. Q U E N T I N A S A N H I S T O R I A N

observed that the rustici were mobilized (congregantes) in an innumerable multitude (incomprehensibilis multitudo).23 In the latter stages of the fighting with the Vikings in this raid, the rustici of the local levy are described as taking up the weapons and armor of those regular fighting men (rotustimissimi bellatores et . . . pugnatores) who had fallen in battle. However, Dudo reports that these rustici were unaccustomed to using this sophisticated equipment and as a result were unsuccessful in their efforts.24 As Dudo and his audience undoubtedly were aware, considerable training was required to use military equipment such as swords and shields effectively.25 When considering Dudo’s treatment of local levies, it is exceptionally important to recognize his emphasis on the order of magnitude of these forces. Military historians often have observed the tendency of medieval chroniclers to exaggerate, often greatly, the number of men engaged in combat and the number of casualties that were suffered in battle.26 Writers sometimes used clichés such as “the soldiers were more numerous than the fish in the sea” or borrowed from the biblical promise to Abraham that his descendants would be more numerous than the grains of sand along the sea shore and more numerous than the stars in the sky.27 Still other medieval chroniclers wrote in terms of immense but logistically impossible actual figures, such as armies of 300,000 or 700,000 men.28 23 Ibid., 2.22. 24 Dudo, De Moribus, 2.22. The phrase desueta arma should not be translated as “disused weapons” (Christiansen, Dudo, 42, 193 n. 185) but rather from the context clearly means that they were using more sophisticated arms (arma), and perhaps armor of the regular soldiers (robustissimi bellatores et pugnatores) who had fallen in battle and which the rustici were not accustomed to using. See the comments of Prentout, Étude critique, 189–90, who believes that his represents an episode that took palce in the north in 859, that was treated in the Annals of St. Bertin. See Annales de Saint-Bertain, ed. F. Grat, J. Vielliard, and S. Clémencet (Paris, 1864), page 80. However, this controversial episode says nothing about sophisticated weapons and the term used by the author of the Annales is vulgus not rustici. 25 Bernard S. Bachrach, Early Carolingian Warfare: Prelude to Empire (Philadelphia, 2001), 84–131, regarding military training and equipment. 26 Hans Delbrück, History of the Art of War within the Framework of Political History, trans. Walter J. Renfroe Jr. (Westport, CT, 1982), made a break-through in volume 3 in criticizing the misuse of numbers by medieval chroniclers. However, he took his arguments too far and often used flawed methods. For a critique of Delbrück’s methods, see Bernard S. Bachrach, “Early Medieval Military Demography: Some Observations on the Methods of Hans Delbrück,” in The Circle of War, ed. Donald Kagay and L. J. Andrew Villalon (Woodbridge, 1999), 3–20. 27 Delbrück, History of the Art of War, III: 151–2 comments on the excessive numbers “reported” by chroniclers for the Norman Conquest. . For a recent review of the scholarly debate concerning the size of the conqueror’s army, see Bernard S. Bachrach, “Baudri of Bourgeuil, Adela of Blois, and the Norman Conquest,” Anglo-Norman Studies 35 (2013), 65–79. For the no longer supportable view of a small army, see John Gillingham, “1066 and Warfare: The Context and Place of (Senlac) of the Battle of Hastings,” in 1066 in Perspective, ed. David Bates (Leeds, 2018), 109–22. 28 Delbrück, History of the Art of War, vol. 3, who generates a great deal of amusement with regard to the excesses of medieval chroniclers at the expense of scholars who have failed to treat numbers critically.

139

N A R R AT I V E W O R K S

In the case of Dudo, however, who was constrained by the knowledge possessed by his audience at the court of the Norman duke regarding contemporary military affairs, the emphasis on the large size of the militia forces operating against the Vikings is credible. Dudo ensures this credibility in several ways. First, he avoids citing actual figures that were so large as to undermine the plausibility of his account among an audience of real soldiers, who had expectations of accuracy in an historical work rather than epic exaggeration.29 He also avoids using metaphors such as the army had more men than there are fish in the sea. In a positive sense, Dudo’s claims for the mobilization of large forces at the local level are credible because all of the local men, or as Dudo has it, all of the pagenses or rustici, were summoned to defend against the Vikings, without observing any differentiation on the basis of wealth, legal status, or social position, or their places of habitation whether rural or urban. In this context, urban sites throughout northeastern Francia were in a demographic growth phase during this period.30 In the context of understanding the large numbers of men available for the local defense, it is important to observe that archaeological studies point to substantial population growth throughout the tenth and early eleventh centuries throughout the northern and western reaches of the erstwhile western realm of the Carolingian Empire, as well as in England.31 The growth of the population was so widely understood and appreciated that Pope Urban II is reported to have proclaimed

29 A good example of epic exaggeration is to be found in the Chanson de Roland, which finally took written form about a century after Dudo wrote. See Gerlad J. Brault, The Song of Roland: An Analytical Edition, 2 vols. (University Park, PA, 1978), which includes the French text, and English translation, and a detailed commentary. For the argument that the Song of Roland really was a Norman parody at the expense of the French, see Bernard S. Bachrach, “Is the Song of Roland’s Rconcevalles a Military Satire?,” in Prowess, Piety, and Public Order in Medieval Society: Studies in Honor of Richard W. Kaeuper, ed. Craig M. Nakashian and Daniel P. Franke (Leiden, 2017), 13–35. 30 See, for example, David Herlihy, “Demography,” in Dictionary of the Middle Ages, ed. J. R. Strayer, 12 vols. (New York, 1984), IV: 136–48, who recognizes the growth pattern but nevertheless emphasizes a minimalist position. By contrast, see the important studies on urban growth by Paul Bairoch, Jean Batou, and Pierre Chèvre, La Population de ville européens: Banque de données et analyse sommaire des résultat, 800–1850 (Geneva, 1988), 4–69; and Adriaan Verhulst, “On the Preconditions for the Transition from Rural to Urban Industrial Activities, 9th–11th Centuries,” in Labour and Labour Markets Between Town and Countryside, Middle Ages -Nineteenth Century, ed. Bruno Blondé, Eric Vanhaute, and Michèle Galand (Turnhout, 2001), 33–41. 31 Extensive archaeological research over the past thirty years has made clear that the population growth of the tenth and eleventh centuries was based upon an already rapidly expanding population in the Carolingian period. See, for example, Pierre Toubert, “Le moment carolingien (VIIIe – Xe siècle) in Histoire de la famillle, ed. André Burguière, Christiane Klapisch-Zuber et al., 2 vols. (Paris, 1986), I: 333–59; and Verhulst, “On the Preconditions,” 34. With respect to demographic growth in Normandy and Anjou see Bates, Normandy Before 1066, 95–8; and Bernard S. Bachrach, “The Angevin Economy, 960–1060: Ancient or Feudal?,” Studies in Medieval and Renaissance History new series 10 (1988), 3–55; and reprinted with the same pagination in idem, State Building.

140

D U D O O F S T. Q U E N T I N A S A N H I S T O R I A N

during his famed sermon at Clermont-Ferrand that “this land which you inhabit, shut in on all sides by the seas and surrounded by mountain peaks, is too small of our large population.”32 Finally, Dudo compares the scale of the local levies to the size of the Viking forces in a manner that is consistent with our current understanding of the order of magnitude of the latter. He depicts the local levies operating somewhat effectively against the relatively small bands of hundreds of Vikings arriving in dozens of ships, while faring very poorly against the exceptionally large armies of Vikings arriving in hundreds of ships.33

Select levies The processes by which the select levy was developed in the Romano-German kingdoms of early medieval Europe are rather more obscure than those by which levies for the local defense were brought into being through late imperial efforts to militarize the general population.34 It is clear, however, that by the very early sixth century, the institution of the select levy operating beyond the borders of its own pagus already was in development. There is a reliable report, for example, that indicates in 507 Sidonius Appolinaris the Younger commanded “the men of Auvergne” in combat on the side of the Arian Visigoths against the imperially orchestrated military operations of the Frankish king Clovis at the battle of Vouillé in the Poitou.35 No later than the later sixth century, the institutional structure of

32 Robert the Monk, Historia Ihierosolymitana in Recueil des Historiens des Croisades: Historiens Occidentaux, 5 vols. (Paris, 1844–1895), III: 728. For the value of Robert’s account, see Dana C. Munro, “The Speech of Pope Urban II at Clermont,” The American Historical Review 11 (1906), 231–42, which remains the basic study of the speech. 33 The size of Viking armies and their military effectiveness is the subject of considerable discussion. Peter Sawyer, The Vikings (London, 1962), is the leading figure among those who argue for very small Viking armies. Sawyer’s arguments were challenged using the methods of Sachkritik by N. P. Brooks, “England in the Ninth Century: The Crucible of Defeat,” Transactions of the Royal Historical Society 5th series 29 (1979), 1–20, who combined an analysis of the narrative accounts regarding the sizes of Viking fleets with a treatment of the size of the fortifications that they occupied in England. Brooks concluded that the larger Viking forces must have numbered in the thousands of men rather than only hundreds. These methods subsequently were adopted by Richard Abels, Alfred the Great: War, Kingship and Culture in Anglo-Saxon England (London, 1988); and John Peddie, Alfred the Warrior King (Phoenix Mill, UK, 1999). Bernard S. Bachrach reviewed these works positively in Albion 32 (2000), 468–9, and 620–1, respectively. Another effective method for analyzing the scale of Viking armies is an assessment of the payments made to them. See, in this context, Einar Joransen, The Danegeld in France (Chicago, 1923). 34 Bernard S. Bachrach, “Early Medieval Europe,” in War and Society in Ancient and Medieval Worlds: Asia, the Mediterranean, Europe, and Mesoamerica, ed. Kurt Raaflaub and Nathan Rosenstein (Cambridge, MA, 1999), 271–397. 35 See Gregory, Hist. 2.37 (Liberi Historiarum, ed. B. Krusch and W. Levison, MGH Scriptores Rerum Merovingicarum, Hanover, 1951), and the compelling discussion by Michel Rouche, L’Aquitaine des Wisigoths aux Arabes, 418–781: Naissance d’une Région (Paris, 1979), 351, 662 n. 120. Also now see Bernard S. Bachrach, “Vouillé and the Decisive Battle Phenomenon in Late Antique Gaul,” in The Battle of Vouillé, 507 CE: Where France Began (Boston, 2012), 11–42.

141

N A R R AT I V E W O R K S

the select levy was fully developed in the Regnum Francorum.36 It is now clear that the development of the select levy was tied closely to the gradual elimination of the tax on land in the Merovingian regna. All people, both men and women, secular and religious, with sufficient real or moveable wealth to sustain military service in expeditione were required to serve beyond the limits of their home pagus in lieu of paying land taxes.37 Those individuals, including women, children, and clerics, who met the wealth qualifications but who were not capable of doing military service in person were required to send substitutes.38 Because men of means in the regnum Francorum already were required to perform expeditionary service, the need for substitutes led to the formation of a large pool of well-trained part-time soldiers, who offered their service on one or another campaign. These men were drawn in general from among the able-bodied men of the lower strata of society, such as servi and coloni, who lacked the property that would have obliged them to serve in the select levy.39 In the post-Merovingian Regnum Francorum, the Carolingians relied upon the select levy for the large majority of troops who served in the very large armies of conquest that were mobilized by King Pippin I and Charlemagne.40 These forces, which on occasion reached several tens of thousands of men under the early Carolingians, were deployed to establish lengthy sieges of great fortress cities such as Avignon, Pavia, and Barcelona.41 These massive fortifications had been con36 See the discussion in Bernard S. Bachrach, The Anatomy of a Little War: A Diplomatic and Military History of the Gundovald Affair: 568–586 (Boulder, 1994), 161–7. 37 See Walter Goffart, “Old and New Merovingian Taxation,” Past and Present 96 (1982), 3–21, reprinted in idem, Rome’s Fall and After (London, 1989), 213–31; idem, “Frankish Military Duty and the Fate of Roman Taxation,” 166–90; Esders, “‘Öftenliche’ Abgaben,” 189–244; idem, “Zur Entwicklung der politischen Raumgliederung,” 185–201; and idem, Nordwestgallien um 500,” 339–61. 38 Bachrach, Early Carolingian Warfare, 57–9. 39 Ibid., 59–60. 40 Ibid., 54–9, 207–42. 41 Contemporary research on the order of magnitude of Charlemagne’s armies is based on the magisterial work of Karl Ferdinand Werner, “Heeresorganization und Kriegsführung im deutschen Königreich des 10. und 11. Jahrhunderts,” in Settimane di Studio de Centro Italiano sull’alto Medioevo, vol. 15 (Spoleto, 1968), 791–843, which is supported by a detailed examination of the administrative infrastructure of Carolingian society in idem, “Missus-Marchio-Comes. Entre l’administration central et l’administration locale de l’Empire carolingien,” in Histoire comparée de l’administration IVe-XVIIe siècles, ed. Werner Paravicini and Karl Ferdinand Werner (Munich, 1980), 191–239. Werner’s views regarding large armies have been embraced by most specialists in medieval military history. See, for example, Philippe Contamine, La Guerre au Moyen Age (Paris, 1980, 4th edition 1994), 101–3. The first edition of this book has been translated into English by Michael Jones, War in the Middle Ages (Oxford, 1984). J. F. Verbruggen, Art of Warfare in Western Europe, ed. and trans. S. Willard and R. W. Southern, 2nd revised edition (Woodbridge, 1997), 283–4, not only accepts the notion of large armies but comes very close to arguing that Charlemagne understood the doctrine that we call “overwhelming force”. See also Charles R. Bowlus, Franks, Moravians, and Magyars: The Struggle for the Middle Danube 788–907 (Philadelphia, 1995), 19; and Eric J. Goldberg, Struggle for Empire: Kingship and Conflict under Louis the German,

142

D U D O O F S T. Q U E N T I N A S A N H I S T O R I A N

structed by the Romans in the later third and early fourth centuries and were kept in good repair for many hundreds of years.42 The Carolingians also deployed large armies to sustain the conquest of very populous and well-fortified regions such as Aquitaine, Lombard Italy, and much of northeastern Spain, where the enemy could mobilize substantial armies of both local and select levies.43 Following the dissolution of the Carolingian Empire and the further dissolution of its Western successor states, local governments at the comital and ducal level, as well as the rulers of the rump of Francia Occidentalis, maintained the effective use of select levies.44 The Ottonian rulers of the German successor of the Carolingian Empire maintained substantially greater direct control over the select levies at the local level than their Western counterparts.45 The obligation owed by individuals with sufficient wealth to participate in offensive military campaigns, either personally or through a substitute, was the

42

43 44

45

817–876 (Ithaca, 2006), who adopts Werner’s model regarding the scale of Carolingian armies for estimating the size of those deployed by Louis the German. With respect to large armies operating in the early medieval period, also see Bernard S. Bachrach, “The Hun Army at the Battle of Chalons (451): An Essay in Military Demography,” in Ethnogenese und Überlieferung: Angewandte Methoden der Frühmittelalterforschung, ed. Karl Brunner and Brigitte Merta (Vienna, 1994), 59–67; and John France, Victory in the East: A Military History of the First Crusade (Cambridge, 1994), 122–41, who estimates the order of magnitude of the crusader army at 50,000 to 60,000 including noncombatants. However, some scholars trained in the German minimalist/primitivist tradition are still wedded to the notion of small, poorly organized, and poorly supplied military forces. See, for example, Maximilian Georg Kellner, Die Ungarneinfälle im Bild der Quellen bis 1150: Von der “Gens destanda” zur “Gens ad fidem Christi conversa” (Munich, 1997). However, as noted earlier, the lead with regard to asserting the small size of early medieval armies has been taken by Timothy Reuter. Reuter’s myopic insistence on the small size of armies leads him in “Carolingiand Ottonian Warfare,” 28–30 to assert that if Charlemagne’s armies were more than 3,000 to 4,000 men in size (a tenth of what most scholars sustain) they would have starved due to lack of a satisfactory logistic infrastructure. However, in trying to support these untenable claims, Reuter omits information from the sources that runs counter to his position. For example, Reuter claims with regard to the battle of Firenzuola that only 50 men were killed, when the source for this information, Flodoard, Annales, an. 923, ed. Philiippe Lauer (Paris, 1905), indicates that 1,500 were killed. See Bernard S. Bachrach, “Imperial Walled Cities in the West: An Examination of their Early Medieval Nachleben,” in City Walls: The Urban Enceinte in Global Perspective, ed. James T. Tracy (Cambridge, 2000), 192–218, with the literature cited there. Also now see Bernard S. Bachrach, “Later Roman Grand Strategy: The Fortification of the Urbes of Gaul,” Journal of Medieval Military History 15 (2017), 3–34. Bachrach, Early Carolingian Warfare, 94–5, 221–6, 229–30; and in more detail in idem, “Military Organization in Aquitaine,” 1–33. There is, to my knowledge, no basic treatment of this subject in the scholarly literature for the period between c. 900 and c. 1100. However, a careful reading of Flodoard, for example, Annales an. 923, makes clear that the rustici operating under King Charles were select levies deployed outside their home pagus. See Bachrach, “Magyar-Ottonian Warfare,” 211–30; and Bachrach, Warfare in Tenth-Century Germany, passim.

143

N A R R AT I V E W O R K S

norm throughout the West.46 Even in England, where military institutions developed on the basis of imitatio of the Frankish exemplars rather than directly on Roman foundations, men who possessed the minimum wealth qualifications were mobilized for expeditionary service. Under Norman leadership in the period after 1066, these English expeditionary levies were even deployed on the continent.47 The minimum property qualification for an individual to serve in expeditio during the early ninth century would appear to have been four dependent mansi, although both smaller and larger assessments were utilized at various times, as were schemes for having men club together to send one of their number on campaign.48 Dudo’s accounts of military operations in De Moribus make it clear that in addition to local levies, he was aware of the select levies and their importance for military operations. As seen earlier, Dudo commented on the efforts of the Walgrenses to oppose a Viking raiding party. However, the local leaders of the Walgrenses quickly realized that they were no match for the attackers, who were reinforced by twelve shiploads of soldiers (milites) from England and another twelve shiploads of supplies.49 As a consequence, according to Dudo, the Walgrenses sent word to the dux, who commanded the region in which they lived, and also the princeps, who ruled the entire Frisian duchy, to ask for military support. These two officials subsequently mobilized a very large army of pagenses from outside the region of Walcheren (conglobato exercitu aliorum pagorum) to aid the Walgrenses.50 As is self-evident, the pagenses summoned by the dux and the princeps were mobilized to serve beyond the boundaries of their own pagi in order to combat the Viking forces commanded by Rollo. Dudo is not explicit in identifying the pagi from which this exercitus was mobilized except for noting that the troops represented a number of different peoples 46 With regard to Christian Spain see, for example, James F. Powers, A Society Organized for War: The Iberian Municipal Militias in the Central Middle Ages, 1000–1284 (Berkeley, 1988), 13–39, where there is much information that is similar to that found in the French and German kingdoms but treated rather differently. Also see idem, “The Origins and Development of Municipal Military Service in the Leonese and Castillian Reconquest, 800–1250,” Traditio 26 (1970), 91–111. 47 As discussed previously in note 13. 48 See Bernard S. Bachrach, “Military Lands in Historical Perspective,” Haskins Society Journal 9 (2001 for 1997), 95–122; Simon Coupland, “The Carolingian Army and the Struggle against the Vikings,” Viator 35 (2004), 49–70, here 54; and Walter Goffart, “Defensio Patriae as a Carolingian Military Obligation,” Francia 43 (2016), 21–39. 49 Dudo, De Moribus, 2.9. It is of some interest that Christiansen (Dudo, 33), adheres to the feudal construct that calls for the introduction of “knights” into England by William the Conqueror, translates Dudo’s English milites as “armed warriors”. For the most articulate proponent of this now discredited model, see R. Allen Brown, Origins of English Feudalism (New York, 1975). Also see the discussion by Prentout, Étude critique, 163–7, regarding this passage. 50 Dudo, De Moribus, 2.9. In recounting this episode, it is clear that Dudo is using much “recent history” regarding both people and places so that the personae and geography sound authentic. Christiansen, Dudo, 187–8, clarifies these matters consistent with an overall treatment of this region by Léon Vanderkindere, La formation territorial des principalites belges, 2nd edition, 2 vols. (Brussels, 1902), which still remains basic.

144

D U D O O F S T. Q U E N T I N A S A N H I S T O R I A N

(populi).51 On other occasions, however, Dudo is very specific. When describing the mobilization of an army by the Norman duke for operations in the field, Dudo explains that he summoned the whole army (omnem . . . exercitum ascivit) of Normandy and Brittany for an invasion of Flanders. Dudo specifically mentions the men levied from the pagus of the Côtentin, one of the regions that comprised the bishopric of Rouen and later came to be known as Normandy. The Constantinenses, as Dudo refers to these levies, were chosen by the Norman leader to spearhead the attack on the stronghold of Montreuil, which was located within Flanders, and thus well beyond the borders of the Côntentin.52 On another occasion, Dudo explicitly calls attention to levies from the Bayeux region and the Côtentin operating together under ducal command against levies from the Rouenais. This encounter took place on the banks of the river Dives, which marked the border between the civitates of Bayeux and Lisieux.53 The men of the Côtentin were deployed some 150 kilometers from their homes, while some of their adversaries, that is the levies of the region of Rouen, were 125 kilometers from their home pagus.54 By referring to the levies mobilized in the Côtentin and Bessin as Constantinenses and Bajocacenses, respectively, Dudo uses the same terminology for the select levies, that is militia forces fighting outside their home pagus, that had been used regularly by writers such as Gregory of Tours from the later sixth century onward.55

Standing armies Throughout the medieval West, the rulers of the empire’s Romano-German successors and their posterity radically reduced the direct economic cost to the central authority that previously had been incurred for the maintenance of standing armies. This vast saving was accomplished through a process of decentralization.56 As already discussed, the numerically preponderant element in campaign forces was composed of the expeditionary levies of individuals with sufficient wealth to support this obligation, or the substitutes whom they provided. Many ecclesiastical institutions, and some wealthy individuals, opted for a more permanent solution to meet their obligation to provide military forces for the ruler’s army. Rather than

51 De Moribus, 2.9. 52 Ibid., 3.60. 53 Ibid., 4.86. It is noteworthy that the Constantinenses and Bojacacenses are referred to as a multitudo who overwhelm the enemy, in large part, because of their numbers. In addition, Dudo is careful to distinguish these levies from their Viking allies who are neither select levies nor Christians. 54 Dudo, De Moribus, 4.86. 55 See Bachrach, Merovingian Military Organization, passim, but especially 42–5, 51–3, 66–8, with the clarifications provided by Bachrach, Anatomy of a Little War, 105–7, 161–7; and now David S. Bachrach and Bernard S. Bachrach, “Nithard as a Military Historian of the Carolingian Empire, c. 833–843,” Francia 44 (2017), 29–55, and in this volume. 56 Bachrach, “Early Medieval Europe,” 271–307.

145

N A R R AT I V E W O R K S

employing fighting men on an ad hoc basis, they developed permanent military followings of trained soldiers, sometimes denoted in the sources as an obsequium or familia. This practice dated back to the late Roman Empire and was actively encouraged by the imperial government.57 The highly trained and well-equipped units of professional soldiers based in the households of kings, high government officials, and magnates (both lay and ecclesiastical) constituted only a small part of the very large armies mustered by early medieval governments for offensive military operations. The select levies remained the numerically superior element throughout the pre-Crusade period and were required in large numbers for the sieges of large-scale fortifications and fortress cities. These were the foci of both secular and ecclesiastical administration and government, the location of regional markets, and remained the centers of the greatest population and wealth through the Middle Ages.58 Despite their relatively small numbers, however, the men serving in the military households of the kings and magnates generally have received disproportionate attention from contemporary observers. The focus by contemporary writers on the men of the obsequia was due, in large part, to the fact that they were closely associated with royal and aristocratic patrons of the authors of the surviving histories and chronicles.59 This bias in contemporary sources often has resulted in an overemphasis by modern scholars on the military importance of the obsequia. In some cases, medieval historians, clinging to nineteenth-century romantic-nationalist traditions, describe the military households of magnates as comitatus in an effort to squeeze them into an otherwise irrelevant Tacitean or Beowulfian model of primitive Germanic societies.60 However, the comparatively greater detail provided by contemporary writers regarding the military households of magnates does have the distinct advantage of permitting a closer examination of the organization of these units of professional soldiers in pre-Crusade Europe.61

57 Bachrach, Early Carolingian Warfare, 59–65; and Alexander Callander Murray, “From Roman to Frankish Gaul: ‘Centenarii’ and ‘Centenae’ in the Administration of the Merovingian Kingdom,” Traditio 44 (1988), 59–100. 58 Bachrach, “Imperial Walled Cities in the West,” 192–218. 59 This is true of Dudo’s account as well. See the literature cited earlier in note 4. 60 See Steven C. Fanning, “Tacitus, Beowulf, and the Comitatus,” Haskins Society Journal 9 (2001 for 1997), 17–38, who demonstrates chapter and verse the gross misuse of Tacitus’s account by modern scholars. Of particular interest in the present context is Reuter, “Plunder and Tribute,” 84 who argues in defense of his primitivist/minimalist conception of the early Middle Ages that vassals in the regnum Francorum during the ninth century were “essentially the comitatus of Tacitus.” Now also see John D. Niles, “The Myth of the Feud in Anglo-Saxon England,” Journal of English and Germanic Philology 114.2 (2015), 163–200, who demonstrates the ongoing pernicious influence of nineteenth-century romantic-nationalist historiography on modern scholarly treatments of the early Middle Ages, and the effort to present society in Beowulfian terms. 61 The military household emphatically was not, as shown by Fanning a replication of Tacitus’s comitatus, but rather a force of professional soldiers. The complex organization of the military household of the Norman dukes at the beginning of the eleventh century foreshadows that of the Norman kings two generations later. See regarding the latter J. O. Prestwich, “The Military

146

D U D O O F S T. Q U E N T I N A S A N H I S T O R I A N

Dudo, who wrote under the patronage of the Norman ruling family, presents considerable information about the military organization of the “ducal” household.62 Dudo refers to this military unit as the militia domus and has the duke speak in direct discourse of militia nostra. The commander of the duke’s militia is described as the princeps militiae and princeps domus.63 Dudo also has the duke speak of the commander of his household troops as major and as praecipuus comes (the special count) of the domus suae militiae.64 Dudo also denotes the commander of the military household simply as comes.65 A certain Both, who appears to have been an older contemporary of Dudo and therefore a real person of some considerable importance rather than a figment of the author’s literary strategy, is depicted as the commander of the duke’s household troops in De Moribus and is given considerable attention. In addition to assigning him what seem to have been the official or semi-official titles of comes and princeps, Dudo refers to Both as the most excellent of all soldiers (cunctorum militum praecellentissimum) who served the duke. Dudo also describes him as a “rough” or perhaps more accurately as a “very tough” soldier (asperrimus miles).66 The soldiers in the ducal military household obviously were professionals who had made their careers in the military service of the ruling house. As a result, Dudo, as will be seen later, consistently refers to them as milites. To translate this term as “knight,” “chevalier,” or “Ritter” is fundamentally misleading and represents a basic misunderstanding of military organization during the tenth and early eleventh centuries.67 Knighthood, as a legal construct with specific rights and obligations, is a phenomenon of the later twelfth century. The great majority of the men serving as milites in the military household of the dukes of Normandy and their contemporaries possessed neither an elevated economic status nor a specific

62 63 64 65 66 67

Household of the Norman Kings,” English Historical Review 96 (1981), 1–37, and reprinted in Anglo-Norman Warfare: Studies in Late Anglo-Saxon and Anglo-Norman Military Organization and Warfare, ed. Matthew Strickland (Woodbridge, 1991), 93–127. Much of the detail noted by Dudo is to be found later in the Norman royal military household. See, Prestwich, “Military Household,” 93–127. Dudo, De Moribus, 3.37, 38, 40. Ibid., 4.66. Ibid., 2.16 praecipuus comes; 3.36, ditissimus comes. Ibid., 3.46; 2.16, respectively. Regarding Both, see Prentou, Étude critique, 174–6, 280–1, 311 n.1, 332–3, 347, 355, 420. Christiansen, Dudo, passim and 72, 81, 101, 122, 143 insists on using the term “knight,” which is anachronistic. He does not seem to recognize this as a problem (xxxvi). First, it must be emphasized that a knight is a man at the lowest rung of the aristocratic ladder of society and later will become the base of the nobility. Most mounted soldiers were neither knights nor members of an elevated social class, much less an elevated legal class. See Bernard S. Bachrach, “The Milites and the Millennium,” Haskins Society Journal 6 (1994), 85–95; and William Delehanty, Milites in the Narrative Sources of England, 1135–1154 (Unpublished Doctoral Dissertation, University of Minnesota, 1975). With respect to the contemporary German kingdom, see David S. Bachrach, “Memory, Epistemology, and the Writing of Early Medieval Military History: The Example of Bishop Thietmar of Merseburg (1009–1018),” Viator 38 (2007), 63–90 (and in this volume); and idem, “Milites and Warfare in Pre-Crusade Germany,” War in History 23.3 (2015), 298–343.

147

N A R R AT I V E W O R K S

legal status, other than that gained through direct service to the ruler. In short, they were professional fighting men and not members of a military aristocracy. By contrast, the dukes of Normandy, like both their Carolingian predecessors and their own contemporaries, had places in their military households for young men from elite families who were preparing for a career of military leadership. Dudo consistently referred to such boys and young men as tirones, drawing upon a terminology that was already current in the late eighth century for the description of such “trainees.” In some cases, Dudo refers to these individuals as the tirones of the duke’s domus.68 Dudo does not use the term schola in the context of a school for the training of professional officers at the court, as the Carolingians did in a frank effort at imitatio imperii.69 However, the notion that these youths were being educated in military matters is clear. Dudo observes, for example, that a military instructor taught (educavit) the youth moribus et studiis belli.70 The man who was in charge of the military education of the tirones living at the ducal court is referred to by Dudo as a princeps.71 While billeted at the court, the tirones were provided with a cash stipend, the diarium, which Dudo indicates was to be used to pay for their own food on a daily basis.72 The organization of the tirones at the Norman ducal court thus differs from the situation at the Carolingian royal court, where the tirones ate together in a designated mess as part of the process of developing a sense of group cohesion.73 According to Dudo’s narrative, the milites who served in the ducal household were organized and deployed in several different ways. A man named Bernard the Dane, for example, is presented as commanding men whom he personally had recruited from Denmark and whom he personally had organized into a military unit. Dudo characterizes these men as Bernard’s fideles, suggesting that they had taken a personal oath to him. Bernard himself is presented as a fidelis of the duke.74 68 Dudo, De Moribus, 2.33. Christiansen, Dudo, 53 demonstrates that he is enthralled by the model of anthropological determinism when he translates tirones as “young braves.” The very notion that even heuristic value can be found in alluding to nonliterate North American Indians, who knew neither iron weapons nor the wheel, for comparison with the technologically advanced and literate (for practical purposes) post-Carolingian ducal society and government of Normandy is ludicrous. See Dudo, De Moribus, 3.59 for reference to tirones who were being trained in the household of Hugh the Great. Dudo makes the point that there is a significant difference in the various kinds of servitium that can be performed by a miles as compared with a tiro. Therefore, the former is due greater respect. With regard to training of young men for command positions in the contemporary German kingdom, see D. Bachrach, Warfare in Tenth-Century Germany, 102–34. 69 Bachrach, Early Carolingian Warfare, 71–5. 70 Dudo, De Moribus, 3.37; and cf. 3.50, in which King Louis’s sojourn as a youth in the ducal household is satirized: “domo ducis ut donigena et vernula.” 71 Dudo, De Moribus, 3.51. The comments by Christiansen, Dudo, 202 n. 265 on grammar are well taken here. 72 Dudo, De Moribus, 4.92, emphasizes this point by noting how a certain Radulfus Torta, who had taken charge of the royal household following the death of the king, was niggardly with the payment of the appropriate stipendium to the tirones. The figure being derided was 18 denarii. This stingy affect is regarded as a mark of Radulfus’s lack of true nobility. 73 Bachrach, Early Carolingian Warfare, 71–5. 74 Dudo, De Moribus, 3.45; 3.60. Cf. Christiansen, Dudo, 80, who characterizes the men who have sworn faithfulness as “vassals,” although neither the word vassus nor vassallus is used by Dudo in

148

D U D O O F S T. Q U E N T I N A S A N H I S T O R I A N

On occasion, one or another unit of milites that normally served at the court was deployed under their princeps outside of the court.75 Dudo records, for example, units of milites were deployed in this manner to serve as garrison troops in a newly obtained stronghold or to reinforce the garrison of a stronghold that was in danger of being attacked.76 Dudo routinely refers to the commanders of these units of milites as principes. The term princeps, therefore, would seem to have had two possible meanings as the leading men of the ducal court and as leaders of individual military units. Dudo devotes particular attention to the principes, who commanded the various units of milites stationed at Rouen, the de facto Norman capital. These principes appear to have been based in a domus, which was located within the walls of the city and which may have been part of a complex of buildings that comprised the ducal residence in Rouen.77 The milites who served under the command of these princes were billeted in special quarters in a specific section of the city.78 This aspect of the organization of the ducal military household recalls the disposition of household troops at the Carolingian court, where officers were housed in quarters separate from those of their men.79 This pattern also calls to mind the vicus of milites located in the monastic town of Centulla, which was governed and apparently owned by the monks of St. Riquier from the later eighth century onward.80 In the context of a rather complicated story, Dudo also provides some insight into the processes by which units of the duke’s standing army might be recruited. Dudo relates the request by the magnate leader of an armed force to the duke to provide him with a substantial tract of land (terra) as a gift (donum). This land was to be used, according to Dudo, to support a large force of milites who would serve under the magnate’s immediate command, but in the service of the duke. These soldiers, as Dudo makes clear, are represented to the duke by their leaders as would-be members of the ducal militia domus. However, they were to be based regularly on the terra obtained by their leader rather than within the walls of Rouen or on other ducal property. In this way, they would constitute a type of military colony. This, of course, was not an uncommon way to organize troops in the West from the later Roman Empire through the early Middle Ages.81 Despite the apparent attractiveness of the offer made in the petition by the magnate to increase the size of the ducal standing army at a rather small economic cost, Dudo presents the duke himself as reluctant to accept the deal. Instead, the duke made a counteroffer. He suggested to the magnate that these soldiers

75 76 77 78 79 80 81

this context. This kind of misleading translation lies at the heart of the failed “feudal” construct so effectively demolished by Susan Reynolds, Fiefs and Vassals: The Medieval Evidence Revisited (New York, 1994). See the review of Reynolds by Bernard S. Bachrach in Albion 27 (1995), 466–7. Dudo, De Moribus, 3.60; 4.117. Ibid., 3.60; 4.117. Ibid., 4.70. Ibid. Bachrach, Early Carolingian Warfare, 65–8. Ibid., 62–4. Dudo, De Moribus, 3.43, 44. See Bachrach, “Military Lands,” 95–122, for the variety of ways in which these matters were handled.

149

N A R R AT I V E W O R K S

(milites) whom he commanded be established at the ducal court and serve, in effect, as praesentales, that is men immediately attendant upon the ruler. The duke promised that in return for this service, the men would receive both preferential treatment and rewards. The latter was to include, Dudo observes in an effort to demonstrate the generosity of the duke, military equipment for use in combat such as axes (secures), mail armor (loricae), helmets (galea), and very high-quality swords that were wonderfully decorated with gold (enses . . . praecipuos auro mirabiliter ornatos). In addition, the men were to be provided with horses. The use of the masculine equus in this context indicates that Dudo intended his audience to understand that these were battle-trained stallions.82 This equipment, of course, would make it possible for the milites to serve as mounted troops if this should be required of them in one or another type of tactical deployment. However, such mounted deployments were quite rare, and wellequipped milites generally dismounted to fight on foot, a fact that was much better known and appreciated by Dudo’s contemporaries than would appear to be the case among modern historians.83 Dudo mentions additional gifts that would be provided by the duke, including armbands (armillae) and sword belts (baltei). Finally, after they joined the ducal military household, these soldiers would be given access to money changers (cambitores) located at Rouen, who would be in a position, likely with a ducal license, to purchase booty that the milites acquired in ducal service in various military operations.84 As might be expected from the terms set out in this offer, the soldiers who served in the ducal military household very likely were prepared to operate as heavily armed mounted troops when necessary. They would have benefitted from the helmets, mail, and high-quality swords. The gold decoration, by contrast, must 82 Dudo, De Moribus, 3.43, 44. 83 Hollister, Anglo-Saxon Military Institutions, 131, observes “in every important battle of the Anglo-Norman age, the bulk of the feudal cavalry dismounted to fight.” C. W. Hollister, The Military Organization of Norman England (Oxford, 1965), strengthens this argument, which is adopted by Jim Bradbury, “Battles in England and Normandy,” Proceedings of the Battle Conference on Anglo-Norman Studies 6 (1984), 1–12; and reprinted in Anglo-Norman Warfare, ed. Strickland, 182–93, provides additional evidence regarding the role of dismounted troops in Anglo-Norman warfare. Compare with Matthew Strickland, War and Chivalry: The Conduct and Perception of War in England and Normandy, 1066–1217 (Cambridge, 1996), 132–76, whose emphasis is on what can only be considered “romantic knighthood” or proto-chivalry. He tries to avoid discussion of fighting on foot and relegates this major theme to a footnote (167 n. 57). See the review of Strickland by Bernard S. Bachrach in The Journal of Interdisciplinary History 28 (1998), 568–70. 84 Dudo, De Moribus, 3.43. Cambitores presents some problems as Ducange, Glossarium, II:45, has this as a variant of cambarius or beer maker. By contrast, J. F. Niemeyer, Mediae Latinitatis Lexicon Minus (Leiden, 1997), 117–18 has money changer. Because Rouen was an economically vibrant emporium where much war booty was traditionally bought and sold, the need for cambitores, that is money changers, by successful soldiers is self-evident. Cf. Christiansen, Dudo, 65, who seems to believe that the cambitores were horses and translates the word as “coursers”. Regarding Rouen’s economic position, see Bates, Normandy Before 1066, 23, 28–30, 96. The observation by van Houts, “Scandinavian Influence,” 108 that “in about 1000 Rouen was both a commercial and literary centre” is right on the mark.

150

D U D O O F S T. Q U E N T I N A S A N H I S T O R I A N

be regarded as a status symbol rather than an integrated aspect of military technology.85 The detail that Dudo provides in the context of this story makes it clear that milites in the duke’s service were heavily armed when deployed in combat. Dudo emphasizes this point somewhat poetically when he described one unit of household soldiers as dressed in iron (ferro indutis) while launching an attack on an enemy force.86 Readers of the account by Notker may recognize echoes of his description of the army of Charlemagne as it besieged Pavia in 773.87 Dudo’s treatment of the tripartite military organization that was in operation not only during the Carolingian era, the period that he ostensibly is described in his account of the invasion by Rollo but also in the early eleventh century, is set out clearly in his discussion of events at Rouen, mentioned previously. Dudo speaks of cives, suburbani, plebs, and rustici, four very different groups, all of whom were in arms. He also makes a point of calling attention to the milites of the ducal household, whom he describes as armati. Dudo contrasts the heavy armament of these professional soldiers with the much less sophisticated equipment of the cives, suburbani, plebs, and rustici.88 This distinction is made quite explicit by Dudo on several occasions, but perhaps nowhere more clearly than in his discussion of the defense of the fortified urbs of Chartres. Dudo describes the bishop, who governed the city, sallying forth “pressed in with the citizens and with the iron clad ranks” (cum civibus ferratisque aciebus constipatus) and driving off the enemy. As Dudo’s use of the enclitic “que” in this sentence makes clear, the bishop commanded both heavily armed soldiers (ferrates) and a citizen militia (cives), the latter of which was not heavily armed.89 Parenthetically, a bishop as a military commander was hardly an anomaly to Dudo’s audience.90

85 Ian Pierce, “Arms, Armour, and Warfare in the Eleventh Century,” Proceedings of the Battle Conference on Anglo-Norman Studies 10 (1987), 237–57, provides a useful introduction to this subject. 86 Dudo, De Moribus, 3.46. 87 Notker, Gesta Karoli Magni, ed. H. F. Haefele, MGH SRG new series (Berlin, 1959), 2.17. 88 Ibid., 4.70–1. 89 Ibid., 2.23. The observation by Christiansen, Dudo, 194 n. 188, that ferratis aciebus is “a well worn phrase” surely may be taken to indicate that Dudo was intent upon having his audience recognize the difference between heavily armed professional troops in the garrison and lightly armed cives who were serving in the capacity of local levies. 90 For the general background regarding the role of clerics in warfare in pre-crusade Europe, see Friedrich Prinz, Klerus und Krieg im früheren Mittelalter (Stuttgart, 1971, and the review by Bernard S. Bachrach in Catholic Historical Review 60 (1974), 478–9. Particular attention in this context may be paid to Guy, the count-bishop of Le Puy and the brother of the Angevin Count Geoffrey Greymantle (d. 987), who was a well-known contemporary of Dudo. Regarding Guy, see Bachrach, “Northern Origins,” 405–21. With regard to the participation of clerics in warfare, now also see Craig M. Nakashian, Warrior Churchmen of Medieval England 1000–1250: Theory and Reality (Woodbridge, 2016), which considers the question of militant clerics in the broader AngloNorman world.

151

N A R R AT I V E W O R K S

Conclusions Dudo’s depiction of military organization in the northeastern region of Francia Occidentalis c. 1000 is consistent with the tripartite system of local levies, select levies, and professional soldiers seen in dozens of other contemporary and nearcontemporary narrative and legal texts. This system functioned from the fifth century through the end of the eleventh century and beyond not only in Gaul but throughout much of the Latin West. In this regard, modern scholars generally argue that Dudo, despite his failure to get the political narrative right in objective terms, was required by the needs imposed upon him by his Norman patrons to tell a story that was plausible. I agree with that position at least insofar as Dudo found it necessary in accounts that focused primarily on military affairs to avoid alienating his audience at the ducal court that was well-versed in the actual practices of warfare. Dudo was not commissioned to write an epic literary fantasy such as those crafted by the authors of Beowulf and the Chanson de Roland, where authors had much more freedom to deviate from reality. While it is highly unlikely that many details of Dudo’s political narrative, especially with regard to the earlier periods of his account, will be found accurate, his treatment of contemporary military matters beyond the subject of organization, for example, tactics and strategy, may well repay further study. I intend to address these in the future.91

91 Now see Bernard S. Bachrach, “Dudo of St. Quentin and Norman Military Strategy c. 1000,” Proceedings of the Battle Conference: Anglo-Norman Studies 26 (2003), 21–36; and idem, “Writing Latin History for a Lay Audience c. 1000: Dudo of Saint Quentin at the Norman Court,” Haskins Society Journal 20 (2009 for 2008), 58–77.

152

7 ADEMAR OF CHABANNES AS A M I L I TA RY H I S TO R I A N

Ademar of Chabannes (d. 1034), the monk of Cybard, stands out as one of the most prolific literary figures of the early eleventh century.1 Although he produced a considerable corpus of material on religious matters, Ademar is best known by modern scholars for his secularly oriented Chronicon, which is referred to on occasion as a Historia.2 This text is regarded by medievalists as the basic narrative of events in Aquitaine covering the later tenth and the first third of the eleventh century.3 It is of great importance that our honoree, Daniel Callahan, has been in the forefront among medievalists who have worked to illuminate Ademar’s career, his intellectual values, and the mentalité of the society in which he flourished.4 Indeed, my own work owes much to Professor Callahan’s scholarship, as I have given special attention to Ademar’s interests in politics and military matters.5 As will be seen later, Ademar himself devoted considerable attention in his Chronicon to the details of armed conflict. Ademar was very much interested not only in secular history, as evidenced by the Chronicon, but also in religious history. As a result, he copied large numbers of

1 Richard Landes, Relics, Apocalypse and the Deceits of History: Ademar of Chabannes, 989–1034 (Cambridge, 1995), 11–14 with appendices on 342–3 and 346–68. 2 The broad context for Ademar’s work is provided by Robert-Henri Bautier, “L’Historiographie en France au Xe et XIe Siècles,” in Settimane di Studio del Centro Italiano di Studi sull’alto Medioevo, 2 vols. (Spoleto, 1970), I: 793–850. For the best edition of the text, see Ademari Cabannensis Chronicon, ed. P. Bougain, Georges Pon, and Richard Landes (Turnhout, 1999). 3 John Gillingham, “Ademar of Chabannes and the History of Aquitaine in the Reign of Charles and Bald,” in Charles the Bald: Court and Kingdom, ed. Margaret Gibson, Janet Nelson, and David Ganz (Oxford, 1981), 3–14, here 3. 4 See the bibliography below. 5 Bernard Bachrach, “Early Medieval Fortifications in the ‘West’ of France: A Technical Vocabulary,” Technology and Culture 16 (1975), 531–69, and reprinted with the same pagination in Idem, Warfare and Military Organization in Pre-Crusade Europe (London, 2002); Idem, “A Study of Feudal Politics: Relations between Fulk Nerra and William the Great, 999–1030,” Viator 7 (1976), 111–22, and reprinted with the same pagination in Idem, State Building in Medieval France: Studies in Early Angevin History (London, 1995); Idem, “Toward a Reappraisal of William the Great, Duke of Aquitaine (995–1030),” Journal of Medieval History 5 (1979), 11–20; and Idem, “‘Potius Rex quam esse Dux putabatur’: Some Observations Concerning Ademar of Chabannes’ Panegyric on Duke William the Great,” The Haskins Society Journal 1 (1989), 11–21.

153

N A R R AT I V E W O R K S

texts in both of these genres and produced a substantial corpus of original works on these subjects as well.6 For Ademar, as for his contemporaries, e.g. Dudo of St. Quentin, a major problem, if not the major problem, was to write history that would be believed by his audience. Authors were concerned with the need to have their histories perceived as factually accurate both by their patrons and their audiences in a broader sense.7 In this context, the teaching of Isidore of Seville that “historiae were things that really happened” dominated the mentality of both the authors of the later tenth and eleventh centuries and the people for whom they wrote.8 For some authors, the matter of credibility, i.e. providing information concerning what really was done, became an issue because the further back into the past the historian ventured, the more difficult it was for him to document what, in fact, had happened. In this context, it is noteworthy that Ademar’s contemporary Dudo of St. Quentin, writing for the Norman ducal court, worked diligently to maintain rhetorical credibility regarding events that supposedly had occurred in the distant past. His method was to provide whenever possible accurate information concerning matters about which his audience had sound knowledge. Thus, having established the factual nature of his account in this manner, he was somewhat insulated from criticism when he introduced fabricated or dubious information, concerning which the members of his audience were not informed or not well informed. As a result, his audience was led to accept this dubious information as true. In short, Dudo, like his older contemporary Richer of Rheims, engaged in the rhetoric of plausibility.9 Ademar encountered a problem similar to that dealt with by Dudo (and Richer), who had been faced with writing the history of the “Normans” for a period concerning which there was little surviving documentation. However, where Dudo had 6 Landes, Relics, 346–68, for a detailed list of the relevant manuscripts. 7 Gillingham, “Ademar of Chabannes,” 3–14, demonstrates brilliantly how Ademar deduced “facts” in order to fill lacunae in the information available to him. In fact, as Gillingham makes clear, Ademar was so effective that he fooled many highly qualified medievalists who accepted his work as accurate. 8 Isidore, Etymologiae, ed. W. M. Lindsay (Oxford, 1911), I:44.5. For Background, see Bautier, “L’Historiographie,” 793–850. For more specific studies, see Bernard S. Bachrach, “Dudo of Saint Quentin as a Military Historian,” The Haskins Society Journal 12 (2002), 155–85; Karl Ferdinand Werner, “Die literarischen Vorbilder des Aimoion von Fleury und die Entstehung seiner ‘Gesta Francorum’,” in Medium Aevum Vivum, Festschrift für Walther Bulst (Heidelberg, 1960), 69–103; Jason Glenn, Politics and History in the Tenth Century: The Work and World of Richer of Rheims (Cambridge, 2004); Bernard S. Bachrach and David S. Bachrach, “Saxon Military Revolution, 912–973?: Myth and Reality,” Early Medieval Europe 15 (2007), 186–222; and also Lars Boje Mortensen, “Stylistic Choices in a Reborn Genre: The National Histories of Widukind of Corvey and Dudo of St. Quentin,” in Dudone di san Quintino, ed. Paolo Gatti and Antonella Degl’Innocenti (Trent, 1995), 79–102. Cf. Suzanne Fleischmann, “On the Representation of History and Fiction in the Middle Ages,” History and Theory 22 (1983), 278–310. Perhaps the most relevant is the recent study by Justin C. Lake, “Truth, Plausibility, and the Virtues of Narrative at the Millennium,” Journal of Medieval History 35 (2009), 221–38. 9 Bachrach, “Dudo of Saint Quentin,” 155–85. It is noteworthy that Dudo’s fabricated “facts” were so well integrated that for a very long time scholars accepted them as accurate, much the same way that some of Ademar’s work fooled some very prominent medievalists.

154

A D E M A R O F C H A B A N N E S A S A M I L I TA RY H I S T O R I A N

to go back perhaps some two centuries, Ademar was separated from his primary subject of interest, St. Martial of Limoges, by a millennium. It was Ademar’s ultimate aim to write a history of the mission of St. Martial of Limoges so that the holy man would be accepted as an apostle. What was known of St. Martial in Aquitaine, either from oral tradition or surviving texts, during the period when Ademar wrote provided no reason to believe in the apostolicity of the holy man. Thus, in the late 1020s, several years after writing at least one draft of his Chronicon, Ademar set out to fabricate an exceptionally sophisticated and detailed religious “history,” which was intended to prove that St. Martial’s place in the pantheon of Christendom’s heroes was much more elevated than traditionally believed.10

Ademar’s literary strategy Ademar began his Chronicon in c. 1025, i.e. several years before he began fabricating the text of his history of St. Martial. The Chronicon was conceived by Ademar to have two interrelated purposes. It was to be the vehicle by which he established his credibility as a writer of history, who was to be trusted to report the facts accurately.11 In this context, it is argued here that he wanted to dedicate the Chronicon to Duke William of Aquitaine (d. 1030) and have him accept the work as its patron. Ademar undertook this strategy because, as he saw the religiopolitico situation in the later 1020s, the duke would be the single most important person in any process that might be held in Aquitaine, which could result in the acceptance of St. Martial’s apostolicity. Consequently, it is argued here that Ademar worked to establish his rhetorical credibility, particularly to the duke and his courtiers, by providing detailed and accurate information in his Chronicon regarding events that took place over the course of the previous half-century. These matters treated in the Chronicon were facts about which the people whom Ademar needed to approve his history of St. Martial, i.e. Duke William and his court, either were well informed or easily could be informed. As a result of establishing his credibility concerning what was known to be true by contemporaries, Ademar positioned himself to tell a story about St. Martial concerning which his audience largely was ignorant but which they would be given no reason to reject. The confidence of the ducal court in Ademar would be established, in large part, because William and his entourage would have come to trust their historian to tell the truth about events that really had occurred.

Establishing rhetorical plausibility Ademar treats a great many topics in the Chronicon that deal with events which took place within living memory of his audience. For example, Ademar would appear to have been very much interested in what was happening in the Middle 10 Landes, Relics, 204–7, and passim. 11 Cf. Landes, Relics, 122–9.

155

N A R R AT I V E W O R K S

East. He gave particular attention to the manner in which the Muslim rulers of the Holy Land mistreated Christians and gave special notice to the destruction of the church of the Holy Sepulcher by the Caliph al-Hakim II in 1009.12 Pilgrims, likely in considerable numbers, returning from Jerusalem are known to have brought back to the West relevant information, which was widely disseminated in Aquitaine. Ademar, in treating events such as the destruction wrought by al-Hakim and the persecution of Christians, was presenting information that already was known to Duke William and his court.13 In telling this story, therefore, Ademar’s reputation would be enhanced as a writer who set out the facts accurately. In addition to subjects such as the anti-Christian activities pursued by Muslims in the Middle East, Aquitanian politics, i.e. relations between the duke and his various fideles, as well as the interactions of the latter with each other, obviously were of great importance. This was the case especially for the lay members of the ducal court. On the whole, Ademar’s account of these various political interactions were accurate.14 It is to be noted, however, that consistent with his desire that Duke William patronize his history, Ademar tended to treat his would-be benefactor in a flattering manner, which sometimes led him to present information in a manner that some modern scholars find to be inaccurate.15 In the course of dealing with political matters, Ademar also found it useful, both historically and rhetorically, to convey considerable quantities of genealogical information concerning the families of various Aquitanian magnates. This is especially the case with regard to the men whom William established in episcopal and abbatial offices as well as of various counts, viscounts, and lesser officials who were dependent upon the duke. Because the families of the great men in Aquitaine were at issue in these discussions, this type of information likely was known to those magnates who frequented the ducal court. Ademar understood that genealogical relationships were of importance in providing accurate information with regard, for example, to claims by various magnates regarding preferment to office and lands. In rhetorical terms, the accurate presentation of genealogical information regarding his contemporaries assured Ademar’s audience that he was in command of the relevant factual information, which he conveyed accurately. In addition, the inclusion of this kind of genealogical information served to flatter the magnates and make them more inclined to support Ademar and his goals.16 12 See the examination of this by Daniel F. Callahan, “The Cross, the Jews, and the Destruction of the Church of the Holy Sepulcher in the Writings of Ademar of Chabannes,” in Christian Attitudes Toward the Jews in the Middle Ages, ed. Michael Frasseto (New York, 2007), 15–23. 13 Daniel F. Callahan, “Al-Hakim, Charlemagne, and the Destruction of the Church of the Holy Sepulcher in Jerusalem in the Writings of Ademar of Chabannes,” in The Legend of Charlemagne in the Middle Ages: Power, Faith, and Crusade, ed. Matthew Gabriele and Jace Stucky (New York, 2008), 41–57. 14 Bachrach, “Feudal Politics,” 111–22. 15 Bachrach, “Reappraisal of William the Great,” 11–20. 16 Léon Levillain, “Adémar de Chabannes, généalogiste,” Bulletin de la Société des Antiquaires de l’Ouest 3rd series 10 (1934–1935), 237–63; and Lutz E. von Padberg, “Geschichtsschreibung

156

A D E M A R O F C H A B A N N E S A S A M I L I TA RY H I S T O R I A N

Military matters were a subject of no less importance to the ducal court or, at least, to the lay members of William’s entourage, than was genealogy. However, it might seem far more difficult for a monk to have been able to demonstrate a sound understanding of the details of military strategy and tactics than regarding other subjects. In this context, it is important that Ademar had two maternal uncles, Abbo and Raymond, who were men of considerable military experience. In fact, Ademar characterizes these men as vigorous military commanders (strenuissimos duces) whom he describes as physically strong (robustos) and warlike in spirit (in animo bellicosos).17 Ademar’s close relation with these men, who served as loyal advisors, i.e. fidelissimi consiliarii, to Abbot Peter of the monastery of Dorat, likely was one of the ways in which he came to understand strategy and tactics.18 In addition to learning about warfare from his uncles, viva voce, Ademar undoubtedly acquired much information regarding the basics of strategy, military operations, and battle tactics from books such as Vegetius’s famous handbook De re militari.19 This text and its author were regarded as the auctoritas in military matters from the later Roman era throughout the course of the Middle Ages.20 Ademar not only knew the text of De re militari very well but made a copy of this handbook for his own use.21 In addition, as part of his effort to obtain the support of William to be his patron, Ademar either presented or perhaps, as is more likely, planned to present a more elaborate exemplar of Vegetius’s text to the duke for his personal use.22 Ademar knew very well that the duke, in the classical and medieval tradition of those who were thought of as great military commanders, took books with him while on campaign.23

17 18 19

20

21 22 23

und kulterelles Gedächtnis: Formen der Vergangenheitswahrnehmung in der hochmittelalterlichen Historiographie am Beispiel von Thietmar von Merseburg, Adam von Bremen und Helmold von Bosau,” Zeitschrift für Kirchengeschichte 105 (1994), 156–77, who discusses the same phenomenon with regard to historical writers in the lands east of the Rhine. Ademar, Chronicon, 3.45. Ibid. Flavius Renatus Vegetius, Epitoma rei militaris, ed. Alf Önnerfors (Stuttgart, 1995), 4.1. See also the English translaton by N. P. Milner, Vegetius: Epitome of Military Science 2nd revised edition (Liverpool, 1996). Philippe Contamine, La Guerre au Moyen Age, 4th edition (Paris, 1994), 353–6; and in greater detail for the period under consideration here, Bernard S. Bachrach, “The Practical Use of Vegetius’ De re Militari during the Middle Ages,” The Historian 47 (1985), 239–55, and reprinted with the same pagination in Idem, Warfare and Military Organization in Pre-Crusade Europe (London, 2002). Now also see Christopher Allmand, The De Re Militari of Vegetius: The Reception, Transmission and Legacy of a Roman Text in the Middle Ages (Cambridge, 2011). Philippe Richardot, Végèce et la culture militaire au moyen âge (Ve-Xe siècle) (Paris, 1998), 20, 30, 127; and Landes, Relics, 112. Cf. Landes, Relics, 112. Ademar, Chronicon, 3.54. Regarding the theme of educated military commanders, see Bernard S. Bachrach, “L’art de la guerre angevin,” in Plantagenêts et Capétiens: Confrontations et héritages, ed. Martin Aurell and Noël-Yves Tonnerre (Turnhout, 2006), 267–84; and for an eastern perspective on this same question, David S. Bachrach, Warfare in Tenth-Century Germany (Woodbridge, 2012), 102–34.

157

N A R R AT I V E W O R K S

In a previous study, I examined in detail a recondite aspect of Ademar’s treatment of military matters, which focused on the terminology that he used to label different types of strongholds. The purpose of that study was to ascertain whether Ademar considered it to be important to be precise in providing accurate labels of various fortifications that he described. However, no narrative account can be treated as plain text, i.e. taken at face value, even when the author in general is making truth claims regarding the accuracy of this reportage. Therefore, I compared and contrasted the terminology used by Ademar to label fortifications with the material construction of those same fortifications as developed through excavations. As a result, it became clear that Ademar cared greatly about attaining precision insofar as he consistently used specific terms that were relevant to the description of different types of military strongholds. In fact, when he found that he had made a mistake in an earlier version of his work, he corrected the error so that his terminology was consistent with the physical realities of the type of fortification that was under discussion.24 Obviously, this type of detail, if mangled, might lead some of the military men at the court, including perhaps the duke himself, to question Ademar’s ability to convey information, concerning which they had specific knowledge, in an accurate manner.

Ademar’s audience It is argued here that it was Ademar’s intention when he was satisfied with the Chronicon to dedicate this history to Duke William in the hope that the latter would accept the work as patron and give it the secular counterpart of an ecclesiastical imprimatur.25 Ademar had two purposes in mind. First, for the apostolicity of St. Martial to be accepted in Aquitaine, Duke William had to lead the way and have his courtiers follow his lead in embracing the holy man’s newly established status.26 Second, Ademar composed a lengthy panegyric dedicated to William as a major theme in the Chronicon. In this panegyric, Ademar says of William that “he was thought of more as a king than as a duke (Potius rex quam dux putabatur).” This implicitly elevated the duke’s status and thereby would make his support even more important than were he a mere duke. Ademar also placed information in other chapters of the Chronicon that praised the duke very highly.27 24 Bachrach, “Early Medieval Fortifications,” 531–69. 25 Of course, it is important to point out that Ademar never finished the Chronicon, although he penned several drafts and corrected his work numerous times. It is my view that William died before Ademar was satisfied that he had completed a satisfactory draft of the Chronicon that was ready for presentation. 26 It is clear that without ducal support and the wide circulation of an approved version of the Chronicon, Ademar’s efforts to have his treatment of St. Martial accepted not only failed but led to serious consequences. Regarding Ademar’s failure, see Landes, Relics, 236, 253, 257, 264, who makes clear that Ademar became a “laughingstock” as a result of his forgery of a “history” of St. Martial which claimed that he was an apostle. 27 Bachrach, “Potius Rex quam esse Dux Putabatur,” 11–21. On occasion Ademar manipulated certain facts so that William would be seen in a more favorable light. Cf. Landes, Relics, 112, who mentions the panegyric but does not develop a sense of its importance.

158

A D E M A R O F C H A B A N N E S A S A M I L I TA RY H I S T O R I A N

If William accepted the Chronicon, Ademar could expect that the duke would follow normal practice and have the text, or at least parts of it, read aloud to his court. In the early Middle Ages, this tradition had been developed under Charlemagne and was recorded by Einhard, whose Vita Karoli was very widely copied and read throughout the early medieval West. Thus, for example, the contemporary court of the Norman duke at Rouen, c. 1016, adhered to this custom with respect to the work produced by Dudo of Saint Quentin.28 Such a reading would establish Ademar before the ducal court as a historian who got his facts right. Ademar probably hoped as a result that he would find the duke and his fideles more likely at a later date also to accept as true what he had to write about the apostolicity of St. Martial. In order for Ademar to be successful when reading the Chronicon or parts of it aloud at William’s court, it would be necessary for the duke and his lay courtiers to be able to understand Latin. The duke himself was highly educated and even carried books in Latin with him when on military campaign. Ademar emphasizes this point in the course of his panegyric dedicated to the duke.29 However, any claim regarding William’s education and learning in the context of a panegyric rightly should be treated as suspect by modern scholars. In this context, therefore, it is also noteworthy that Duke William exchanged letters in Latin, some of considerable complexity, on numerous occasions with highly educated men, including Bishop Fulbert of Chartres, Bishop Leo of Vercelli, Abbot Aribertus of Charroux, and Hildegar, the head of the cathedral school at Chartres.30 Even the most skeptical modern scholar will have to recognize that William would have been able to understand Ademar’s simple and straightforward Latin prose if Ademar were fortunate enough to have the text of the Chronicon read aloud at the ducal court. In addition to Duke William, as well as the high-ranking clerics at his court, there is substantial information that other important lay magnates also were proficient in Latin. Ademar observes, for example, that Count Alduin of Angoulême was well educated in both military and literary matters, and it is reasonable, therefore, to conclude that he, too, could understand the straightforward and unadorned Latin of the Chronicon, particularly if the text were read in an Aquitanian accent.31 In this context, it is very likely the case that various schools in Aquitaine, like those in contemporary Normandy, and, indeed, throughout much of the Latin West, including in the German kingdom, accepted students from among the magnate 28 Bachrach, “Writing History for a Lay Audience,” 58–77, where the tradition is discussed in detail. In the present context, it is of special importance that Ademar, Chronicon, 3.27 observes that the Normans “gave up their pagan language and learned to speak Latin.” 29 Bachrach, “Potius Rex quam esse Dux Putabatur,” 11–21. 30 The Letters and Poems of Fulbert of Chartres, ed. and trans. Frederick Behrends (Oxford, 1976), numbers 107, 116, 119, 120, and 122 for Bishop Fulbert; numbers 103, 112, 113 for Bishop Leo; number 196 for Aribertus, and numbers 69 ad 79 for Hildegar. It should be noted that each of these letters, which has survived, provide substantial information for additional correspondence that is now lost. 31 Landes, Relics, 140 n. 44, regarding Alduin’s learning.

159

N A R R AT I V E W O R K S

families who were slated to have a lay career.32 On an even broader level, the ability of Aquitanian lay magnates to understand simple Latin when read aloud is secularly evidenced by the conventum that Duke William, in his capacity as count of Poitou, negotiated with Hugh of Lusignan.33 Ademar expected, as shown by his efforts with regard to his treatment of fortress terminology, that military officers at the ducal court, who might hear his Chronicon read aloud, would understand those matters that concerned their professional duties as soldiers. In this context, Ademar, like his contemporary Dudo, chose to write Chronicon in a style characterized as sermo humilis. This unadorned style had been advocated by Cicero, whose work on rhetoric written more than a millennium earlier was a basic text studied in the schools of the erstwhile Carolingian Empire during the later tenth and early eleventh centuries.34 No less importantly, St. Augustine also advocated the use of the sermo humilis as the proper style to be used for educational purposes, as he believed that it was well known to facilitate understanding.35 As a consequence of these factors, the capacity of the Aquitanian secular magnates to understand the Latin of Ademar’s text when read aloud in the proper cadence and accent seems highly likely. As indicated earlier, such performances were the norm at both royal and ducal courts throughout the polities that emerged in the wake of the dissolution of the Carolingian Empire. Even the descendants of the Scandinavian adventurers who established Normandy and had a very different cultural background than that of the Latin-Romance Aquitanian milieu patronized and listened to Latin histories of this type.36 Ademar himself was aware of this practice and drew attention in his Chronicon to the fact that the Norman leaders had given up their native language and learned Latin.37

32 Regarding the Norman schools, see Bachrach, “Writing Latin History for a Lay Audience,” 58–77; and in more general terms, see Pierre Riché, Écoles et enseignement dans le Haut Moyen Age (Paris, 1979), 119–86. For the education of laymen in monastic and cathedral schools in the east, see Bachrach, Warfare in Tenth-Century Germany, 102–34 and the literature cited there. 33 The Conventum inter Guillelmum Aquitanorum comitem et Hugonem Chiliarchum is an interesting example of what may be considered “vernacular” Latin conversation. For the text, see Jane Martindale, “Conventum inter Guillelmum Aquitanorum comitem et Hugonem Chiliarchum,” The English Historical Review 84 (1969), 528–53. This text has received considerable attention in recent years and much of the bibliography is covered by Jane Martindale, “Dispute, Settlement, and Orality in the Conventum inter Guillelmum Aquitanorum comitem et Hugonem Chiliarchum: A Postscript to the Edition of 1969,” in Status, Authority and Regional Power: Aquitaine and France, 9th to 12th Centuries, ed. Jane Martindale (Aldershot, 1997), 1–36. 34 Lake, “Truth, Plausibility,” 221–38. 35 Augustine, De doctrina christiana, ed. and trans. M. Thérèse Sullivan (Washington, DC, 1930), 4.21; and the discussion by Giles Constable, “The Language of Preaching in the Twelfth Century,” Viator (1994), 131–52, here 137, who observes that medievalists have shown that the divergence between Latin and the vernacular, “were less sharp than scholars whose training was in classical Latin once assumed.” 36 Bachrach, “Writing Latin History for a Lay Audience,” 58–77. 37 Ademar, Chronicon, 3.27, “gentilium linguam obmittens, latino sermone assuefacta est.”

160

A D E M A R O F C H A B A N N E S A S A M I L I TA RY H I S T O R I A N

Listening to Latin prose works was part of a broader pattern of participation in Latinity by lay magnates. They also often witnessed charters and other acta that were written in Latin and read aloud at courts. Magnates were required to swear that they understood the acta and subsequently appended their signa in their capacity as witnesses, indicating that they would help enforce the terms of the various documents. This participation in the legal life of the court was made possible because, as Giles Constable observed: “The pronunciation of Latin in the Middle Ages resembled that of the vernacular . . . and the evidence of charters shows that Latin was read aloud like the vernacular.”38 Indeed, H.A. Kelly regards “legal Latin” and “Romance vernacular” as part of the same language.39 In addition to the practical understanding of legal Latin as well as straightforward Latin prose, it is important that even the least educated of William’s courtiers had been integrated into the Latin-Christian liturgical life of the church in Aquitaine, which fundamentally was based upon an understanding of Latin. As Jan Ziolkowski has observed: “What should be appreciated is not just the language or the literature of Medieval Latin but the very culture of Medieval Latin permeated wherever the Western Church exerted its influence.”40 Christian liturgical practice, in fact, constituted a system that led to the widespread exposure of its adherents to the Latin language.41 Over time, such a frequent and consistent process amounted to a form of linguistic conditioning for many on a daily basis.42 Of particular importance in the present context is the fact that it was normal for soldiers to make battlefield confessions and to hear both battlefield sermons and secular orations before going into combat. This undoubtedly contributed to the soldiers’ ability to attain an additional acquaintance with Latin in a context where understanding the language was deemed to be of great importance.43 These speeches,

38 Constable, “The Language of Preaching,” 137 for the quotation. 39 H. A. Kelly, “Lawyers’ Latin: Loquenda ut vulgus,” Journal of Legal Education 38 (1988), 195–207, here 201. 40 Jan Ziolkowski, “Towards a History of Medieval Latin Literature,” in Medieval Latin: An Introduction and Bibliographical Guide, ed. F. A. C. Mantello and A. G. Rigg (Washington, DC, 1996), 505–36, here 512. 41 With particular regard to the Latinity of sermons, see Beverly Mayne Kienzle and David L. D’Avray, “Sermons,” in Medieval Latin, 659–69. 42 A valuable introduction to this topic is provided by Daniel Sheerin, “The Liturgy,” in Medieval Latin, 157–82; and Constable, “The Language of Preaching,” 131–52, which is not limited to the twelfth century. 43 See Kienzle and D’Avray, “Sermons,” 660, who observe “A sermon designed for catechetical public preaching employs a much simpler syntax and style than a sermon aimed at a select monastic audience . . . .” Concerning battlefield religion, see David S. Bachrach, Religion and the Conduct of War, c. 300-c. 1215 (Woodbridge, 2003). The basic work in the field is by John Bliese. See, for example, “The Courage of the Normans: A Comparative Study of Battle Rhetoric,” Nottingham Medieval Studies 35 (1991), 1–16; idem, “Deliberative Oratory in the Middle Ages: The Missing Millennium in the Study of Public Address,” Southern Communication Journal 59 (1994), 273–82; and Idem, “Rhetoric Goes to War: The Doctrine of Ancient and Medieval Military Manuals,” Rhetoric Society Quarterly 24 (1994), 105–30.

161

N A R R AT I V E W O R K S

both religious and secular, not only were presented in simple and repetitive Latin but also in a straightforward manner. They resembled in this manner the developing Romance syntax and lacked the periodic sentence structure of classical Latin.44 The purpose of these speeches was to build the soldiers’ morale prior to entering combat.45 The ongoing tradition from the ancient world through the Middle Ages, and beyond, of providing various types of orationes prior to battle indicates that these efforts were regarded as useful.46 Necessarily, the quality of understanding attained by those men who had no formal Latin education, i.e. the vast majority, varied greatly. Understanding depended, at least in part, upon the local accent with which the Latin was articulated.47 As Michael Herren observed: “Latin continued as a spoken language throughout the Middle Ages. There is evidence to show that it was pronounced according to the phonetic systems of the various regions.”48 This very brief account of the acquisition by laymen of the ability to understand Latin, even those who were not formally trained in the language, lends support to the view that many soldiers could understand the sermo humilis of Ademar’s Chronicon. As a result, an audience of important laymen, or at least a large part of such a group, based at the ducal court and serving in William’s military household were able to understand simple Latin when read aloud with an Aquitanian accent. As a consequence, in pursuit of his agenda of establishing himself as a credible purveyor of true information, Ademar found it necessary to present information regarding military matters in an accurate manner in order to maintain his rhetorical plausibility.49

44 Bernard S. Bachrach, Early Carolingian Warfare: Prelude to Empire (Philadelphia, 2001), 114–46, and 153–9, regarding the style used for battlefield orations and sermons, respectively. With regard to vernacular syntax for Latin sermons, see Constable, “The Language of Preaching,” 138. 45 See, for example, John Bliese, “Rhetoric and Morale: A Study of Battle Orations from the Central Middle Ages,” Journal of Medieval History 15 (1989), 201–26; and Idem, “When Knightly Courage May Fail: Battle Orations in Medieval Europe,” The Historian 53 (1991), 489–504. Also of use here is Bachrach, Religion and the Conduct of War, 11–13, 42–3, 62–4, 70, 74, 80, 83, 88, and 107. 46 John Bliese, “Fighting Spirit and Literary Genre,” Neuphilologische Mitteilungen 96 (1995), 417–36. 47 In this area of research two works of particular importance are Michel Banniard, Viva voce: communication écrite et communication orale du IVe au IXe siècle en occident Latin (Paris, 1992); and Roger Wright, Late Latin and Early Romance in Spain and Carolingian France (Liverpool, 1982). See also the effective treatment of these matters in two studies by Rosamond McKitterick, “Latin and Romance: An Historian’s Perspective,” in Latin and the Romance Languages in the Early Middle Ages, ed. Roger Wright (University Park, PA, 1991), 130–45; and idem, “The Written Word and Oral Communication: Rome’s Legacy to the Franks,” in Germania Latina, ed. R. North and T. Hofstra (Groningen, 1992), 89–112. The efforts by József Herman, “The End of the History of Latin,” Romance Philology 49 (1990), 364–84, to undermine this widely held view have not been accepted by most scholars. 48 Michael Herren, “Latin and Vernacular Languages,” in Medieval Latin, 122–8, here 123 for the quotation. Constable, “The Language of Preaching,” 132–52, takes a similar position. 49 See Lake, “Truth, Plausibility,” 221–38.

162

A D E M A R O F C H A B A N N E S A S A M I L I TA RY H I S T O R I A N

Ademar as a military historian In the remainder of this chapter, I will continue my earlier work by examining Ademar’s treatment of military history in broader terms, i.e. beyond the description and labeling of fortifications. As has been seen thus far, Ademar both had a desire to establish himself as a purveyor of accurate information and was constrained in his presentation of this information by an audience of laymen, who were well-informed about military matters. Moreover, this audience was capable of understanding the simple Latin of Ademar’s Chronicon and therefore would be able to detect errors of commission and omission were this text to be read aloud at the duke’s court. On this basis, there is a prima facie case that Ademar worked diligently to present information about military matters as accurately as he could.50 Nevertheless, it would be naïve in the extreme to conclude that Ademar, or any other contemporary author of a narrative history, was operating in a Rankian paradigm of providing the truth, the whole truth, and nothing but the truth. Rather, it is well understood that in the context of the main narrative arc of the historical account, medieval authors routinely selected particular facts to highlight, or even fabricated “facts” from whole cloth. Rather, it is the contextualizing information, i.e. the background to the main narrative arc, in which medieval authors sought most diligently to present information in an accurate manner so as to maintain their rhetorical plausibility. Distinguishing between the narrative arc, with the author’s concomitant parti pris, and the background information involves, as discussed previously, ascertaining the objectives of the author, his values, and interests, as well as those of his audience.51

Ademar’s discussion of military organization The military institutions that had been well-established throughout the Regnum Francorum under both the Merovingians and Carolingians continued to function in many parts of the erstwhile Carolingian Empire for centuries after its dissolution. This was true both in the western parts and eastern parts of the former Regnum Francorum.52 These institutions even were maintained in the Norman 50 For a discussion of the methods employed to ascertain whether a medieval narrative source is providing accurate information, see two studies by Bernard S. Bachrach, “Gregory of Tours as a Military Historian,” in The World of Gregory of Tours, ed. Kathleen Mitchell and Ian Wood (Leiden, 2002), 351–63; and Idem, “Dudo of Saint Quentin as a Military Historian,” 155–85. Also see David S. Bachrach, “Memory, Epistemology, and the Writing of Early Medieval Military History: The Example of Bishop Thietmar of Merseburg (1009–1018),” Viator 38 (2007), 63–90. 51 The most important study opening up the field of historical criticism in this manner is Walter Goffart, The Narrators of Barbarian History (550–800). Jordanes, Gregory of Tours, Bede, and Paul the Deacon (Princeton, 1988). 52 See, for example, Simon Coupland, “The Carolingian Army and the Struggle against the Vikings,” Viator 35 (2004), 49–61; David S. Bachrach, “The Military Organization of Ottonian Germany, c. 900–1018: The Views of Bishop Thietmar of Merseburg,” Journal of Military History 72 (2008), 1061–88; David S. Bachrach and Bernard S. Bachrach, “Early Saxon Frontier Warfare: Henry I,

163

N A R R AT I V E W O R K S

duchy, founded early in the tenth century. This is of considerable importance because the Scandinavians, who took control in 910 of the region that came to be called Normandy, had entered and settled in the West Frankish kingdom with very different military customs and traditions from those that had flourished in the Carolingian world.53 It is to be noted in addition that the soundness of these institutions led to their adoption in Anglo-Saxon England, which developed into a Carolingian-style polity.54 Perhaps even more indicative of the usefulness and more particularly the adaptability of these institutions was their adoption in Slavic lands. These areas not only had never been under direct Carolingian rule but also had never been a part of the Roman Empire.55 Carolingian military organization was based upon a tripartite division of service. At the most fundamental level, all able-bodied men, whether free or unfree, were required to participate in the local defense. Scholars tend to refer to these men as members of “general levies.” In addition, those able-bodied men whose real or moveable wealth reached a legally stipulated minimum were required, when called upon by the government, not only to serve in the local defense but also to serve as members of an expeditionary levy for military operations beyond the locality in which they resided. Those who personally were unable to serve were required to find substitutes. All those who possessed multiples of the minimum wealth requirement were obliged, when called upon by the government, to provide numbers of troops consistent with the value of their holdings. These expeditionary levies often are described by scholars as “select levies.” Finally, the king, or in the case of Aquitaine, the duke and his magnates, both lay and ecclesiastical, maintained military households composed of men who earned their living by military service, i.e. professional soldiers. These latter forces were an important element in the contingents which the great landholders were required to mobilize, in accordance with their great wealth, at ducal command for military action of an expeditionary nature.56

53

54

55 56

Otto I, and Carolingian Military Institutions,” The Journal of Medieval Military History 9 (2011), 17–60. More generally, see Bernard S. Bachrach, “Early Medieval Europe,” in War and Society in the Ancient and Medieval Worlds: Asia, The Mediterranean, and Mesoamerica, ed. Kurt Raaflaub and Nathan Rosenstein (Cambridge, 1999), 271–307; and Idem, “Medieval Military Historiography,” Companion to Historiography, ed. Michael Bentley (London, 1997), 203–20. J. Yver, “Les premières institutions du duché de Normandie,” Settimane di Studio del Centro Italiano di Studi sull’alto Medioevo 16 (1969), 299–366; Bachrach, “Dudo of Saint Quentin as a Military Historian,” 155–85; and David Bates, Normandy Before 1066 (London, 1982). See two important articles by James Campbell, “Observations on English Government from the Tenth to the Twelfth Century,” Transactions of the Royal Historical Society 5th series 25 (1975), 39–54; and Idem, “The Significance of the Anglo-Norman State in the Administrative History of Western Europe,” in Histoire comparée de l’administration (IVe – XVIIe siècles), ed. Werner Paravicini and Karl Ferdinand Werner (Munich, 1980), 117–34. Both of these studies have been reprinted with additional notes in James Campbell, Essays in Anglo-Saxon History (London, 1986). See Alexander Ruttkay, “The Organization of Troops, Warfare and Arms in the Period of the Great Moravian State,” Slovenska archeological 30 (1982), 165–98. For the outline of the system with reference to the relevant studies see Bernard S. Bachrach and Charles R. Bowlus, “Heerwesen,” in Reallexikon der germanischen Altertumskunde, ed. Heinrich Beck et al. (Berlin, 2000), cols. 122–136.

164

A D E M A R O F C H A B A N N E S A S A M I L I TA RY H I S T O R I A N

Ademar of Chabannes, while discussing military matters, makes clear that all three types of military organization, i.e. general levies, expeditionary levies, and professional soldiers, were part of the system that was operative in Aquitaine during the later tenth and early eleventh centuries. For example, Ademar treats an effort to besiege the fortress city of Poitiers by the count of La Marche and calls attention to the cives, i.e. the able-bodied inhabitants of the city and its suburbs, who were mobilized for the local defense.57 In this context, the cives included general levies who were from the lower economic levels of society and therefore were required to participate only in the local defense within the borders of the Poitou. In addition, the cives included men who were sufficiently well off to have obligations for both the local defense and expeditionary military service beyond the borders of the Poitou. Finally, as will be discussed later, the duke as well as his higher-status fideles maintained military households and in the situation under consideration here, made these men available for the defense of Poitiers as well. However, the men of these military households likely were not considered cives either by Ademar or his potential audience.58 The fortress city of Poitiers was defended by a massive stone circuit wall dating back to the later Roman Empire. These defenses were more than 3,000 meters in circumference and rose to a height of some 10 meters. The walls were reinforced with numerous mural towers that overtopped the enceinte and heavily defended gates, each of which was defended by a complex arrangement of towers.59 In order to defend these walls, a minimum force of approximately 2,500 able-bodied men drawn from a mix of general and expeditionary levies as well as military households was required. For the count of La Marche to have posed a serious threat to take the city by storm, it was necessary for him to mobilize an army of some 10,000 to 12,000 effectives, composed largely of expeditionary levies and strengthened by the professional soldiers who served in his own military household and those of his fideles.60 Scattered throughout the Chronicon are numerous examples of general levies, expeditionary levies, and household troops. Because the overwhelming majority of the population, amounting to at least 90 percent, lived in the countryside and were directly involved in agricultural production, most of the men who were obligated to serve in the general and expeditionary levies also lived in the 57 Ademar, Chronicon, 3.34. 58 For a detailed discussion of these matters, see Bernard S. Bachrach, “Imperial Walled Cities in the West: An Examination of their Early Medieval Nachleben,” in City Walls: The Urban Enceinte in Global Perspective, ed. James T. Tracy (Cambridge, 2000), 192–218. 59 Carl-Richard Brühl, Palatium und Civitas: Studien zur Profantopographie spätantiker Civitates vom 3. bis 13. Jahrhundert, 2 vols. (Cologne, 1975, 1990), here I: 168–76. 60 Bernard S. Bachrach and Rutherford Aris, “Military Technology and Garrison Organization: Some Observations on Anglo-Saxon Military Thinking in Light of the Burghal Hidage,” Technology and Culture 31 (1990), 1–17; and reprinted with the same pagination in Warfare and Military Organization.

165

N A R R AT I V E W O R K S

countryside.61 Most of these men, including those lesser landholders who by virtue of their wealth were obligated to participate in the expeditionary levy, were of lower social status, or at least had less well-developed ties to the mighty, than the professional soldiers, often styled milites, based in the military households of the magnates. This was the case even though many of the men who served as milites were themselves recruited from among the lower social and economic strata of society.62 Consequently, when Ademar observes that a force of rustici was mobilized to attack the stronghold of Melle, his use of this term indicates that these men were members of the local levy, or at least that the force was dominated by men of this type, rather than the professional soldiers employed by various magnates.63 By contrast, when Ademar wants to draw attention to the participation of members of these military households in various operations, he uses a different set of terms. For example, Ademar calls attention to an element or unit in Duke William’s military household by observing that its commander held the title of praepositus.64 In this case, Ademar observes that the unit of William’s military forces, led by the praepositus, came into conflict with the homines of the abbot of the monastery of Saint-Jean-D’Angély, the result of which was considerable bloodshed and the death of the praepositus.65 The division of Duke William’s military household into subunits, each under its own commander, can be seen to parallel the organization of the military household of the contemporary dukes of Normandy.66

Ademar’s discussion of strategy It is now established beyond any doubt that military operations that were directed toward territorial conquest during the medieval millennium were dominated by sieges.67 This was certainly true in the world in which Ademar lived, where, for example, the Angevin count Fulk Nerra (987–1040), whom Ademar discusses with some frequency, has been given the sobriquet “le grand bâtisseur” by posterity for his military 61 See, for example, Fredric Cheyette, “The Disappearance of the Ancient Landscape and the Climatic Anomaly of the Early Middle Ages: A Question to be Pursued,” Early Medieval Europe 16 (2008), 127–63, here 128; and Chris Wickham, Framing the Early Middle Ages: Europe and the Mediterranean, 400–800 (Oxford, 2005), 12. 62 For details regarding the social status of milites with additional literature see Bernard S. Bachrach, “The Milites and the Millennium,” The Haskins Society Journal 6 (1994), 85–95, and reprinted with the same pagination in Warfare and Military Organization. Also see David S. Bachrach, “Milites and Warfare in Pre-Crusade Germany,” War in History 23.3 (2015), 298–343, which stresses the lower social and economic status of many men who were recruited to serve as milites. 63 Ademar, Chronicon, 3.48. 64 Ibid., 3.56. 65 Ibid. 66 Bachrach, “Dudo of Saint Quentin as a Military Historian,” 155–85. 67 In this context, see Jim Bradbury, The Medieval Siege (Boydell, Woodbridge, UK, 1992); and Peter Purton, A History of the Early Medieval Siege, c. 450–1200 (Boydell, Woodbridge, 2010), and the enormous literature cited there.

166

A D E M A R O F C H A B A N N E S A S A M I L I TA RY H I S T O R I A N

strategy based upon the construction, defense, and capture of fortifications.68 The Norman dukes also based their military strategy on the defense and attack of various types of strongholds, including fortress cities.69 Consequently, it should hardly be surprising that Ademar, who was trying to assure William and his entourage at the ducal court of his sound grasp of military matters, devoted considerable attention to fortifications. This was not only true in his use of precise terminology, as discussed earlier, but also the role that strongholds played in the conduct of military campaigns. In the course of discussing military history over a period of approximately a half a century, corresponding to his own memory and that of his uncles, treated previously, Ademar calls attention to dozens of fortifications. These span the range from great fortress cities such as Poitiers, which were constructed initially in the late Roman period, to castra with stone donjons and curtain walls, to simple towers some of which were so primitive that they were constructed of wood and even lacked a curtain wall.70 In most cases, Ademar deals with these fortifications either at the time when they were being constructed or when they came under attack.71 These descriptions make clear that Duke William and the magnates of the region were thoroughly conversant with the strategic criteria that were required for the optimal siting of fortifications for both offensive and defensive purposes. This led, in part, to a pattern of warfare that resulted in frequent attacks on strongholds and on the defense of these fortifications. In two noteworthy cases, Ademar calls attention to the fortifications of Gençay and Bouteville, which were attacked and destroyed, then rebuilt, and then once again attacked and destroyed, and then rebuilt.72

Ademar’s discussion of siege tactics Throughout premodern history prior to the development of effective gunpowder weapons capable of destroying stone fortifications, there were basically two ways to capture a fortress city through military action.73 The least time consuming, but 68 Bernard S. Bachrach, “The Angevin Strategy of Castle-Building in the Reign of Fulk Nerra, 987–1040,” American Historical Review 88 (1983), 533–60, and reprinted with the same pagination in Warfare and Military Organization. In a broader context see Idem, Fulk Nerra-the Neo Roman Consul: A Political Biography of the Angevin Count (987–1040) (Berkeley, 1993). 69 Bernard S. Bachrach, “Dudo of Saint Quentin and Norman Military Strategy,” Anglo-Norman Studies 26 (2004), 21–36. 70 Bachrach, “Early Medieval Fortifications,” 531–69. It is worthwhile here to call particular attention to efforts made to keep the walls of the fortress cities in repair. See Ademar, Chronicon, 3.23, where he notes the efforts of the count of Angoulême to keep the Roman walls of this fortress city in repair. Ademar, Chronicon, 3.25, also calls attention to the construction of a rather primitive wall that was placed around the convent of Saint-Maixent. 71 Ademar, Chronicon, 3.29, 34–5, 41–3, 45, 48; 3.50–2, 60, 67. In many chapters Ademar discusses several sieges. 72 Ibid., 3.34 and 3.60. 73 For useful background concerning the complexity of capturing fortress cities, see Paul Bentley Kern, Ancient Siege Warfare (Bloomington, 1998), 12, 49–50; and Constantin Nossov, Ancient and Medieval Siege Weapons (Guildford, CT, 2005), 78–80, 248–9.

167

N A R R AT I V E W O R K S

generally the most costly in dead and wounded, was to mobilize large numbers of men to storm the walls with scaling ladders. Such an operation required that the attacking force outnumber the defenders by a ratio of at least 4:1 or 5:1.74 Attacking forces in these efforts could be supported by siege towers, battering rams, catapults, and handheld missile weapons.75 The second option was to starve the defenders into submission by cutting off their food supplies. Obviously, in order to bring about such conditions, the besieging force had to have available the means to establish a blockade in order to deprive the besieged of relief and a logistic system that provided the men maintaining the investment with sufficient food and other supplies for themselves.76 As Ademar describes the tactics of siege warfare, it becomes clear that he understood that large numbers of men were required by the force on the offensive. For example, when King Robert besieged the Aquitanian stronghold of Bellac, the monarch is described as having mobilized forces from all parts of Francia.77 In another example, Ademar tells his readers that the siege of the stronghold of Rochemaux was carried out by a large force of brave men (multitudine fortium).78 In yet another case, Ademar indicates that the stronghold of Brosse was placed under siege by Duke William and four other counts with a very large force (cum valida manu). Although Ademar was uncertain regarding how much detail to present in this case, he knew that the successful relief force also had to be large, and he notes that Viscount Guy relieved the siege with the Limousin levies (cum Lemovicinis), who drove off the besieging army.79 In addition to recognizing and conveying to his audience that he understood that there was a need to mobilize large forces in order to undertake the investment of a stronghold, Ademar provides some information regarding the establishment of a siege. He recounts in some detail an attack by Muslims from the Caliphate of Cordova on the old Roman fortress city of Narbonne. The Muslims are described as coming by sea and making a surprise landing (subito . . . appulerunt) at night with many ships (cum multa classi). Upon landing, the Muslims deployed around the entire city (in circuitu civitatis), and with the coming of dawn, it could be seen that they had established a defensive encampment, i.e. that they had dug themselves in with their siege weapons (cum armis).80 In response to the establishment of the Muslim siege, the men charged with the defense of Narbonne decided that the armed forces under their command, whom Ademar characterizes as Christiani, would be deployed most effectively 74 75 76 77 78 79

Bachrach and Aris, “Military Technology,” 1–17 regarding the ratio of attackers to defenders. Bradbury, The Medieval Siege, 241–80, and passim. Ibid., 78–88. Ademar, Chronicon, 3.34. Ibid., 3.41. Ibid., 3.34, where it is clear that Ademar added and excised various bits and pieces of information regarding the details of the siege and its relief. 80 Ademar, Chronicon, 3.52. It is notable that Ademar uses the term Mauri, Sarasceni, and Hagares to describe the attacking force.

168

A D E M A R O F C H A B A N N E S A S A M I L I TA RY H I S T O R I A N

by executing a sortie against the enemy force. In describing this episode, Ademar accurately calls attention to the common Christian prebattle ritual, i.e. after deciding to attack the enemy, the men heard Mass and took Communion in order to be prepared for death (de preparentes se ad mortem).81 Following this ritual, the men of Narbonne attacked the Muslims and carried the day. According to Ademar, a great many of the enemy were killed, while many others were taken prisoner and sold as slaves. As a result of this victory, according to Ademar, the Christians captured the Muslims’ ships (naves) and acquired large quantities of booty from these vessels.82 This episode highlights the long-recognized fact that while battles in the field were unusual, encounters such as that won by the men of Narbonne usually were fought in conjunction with a siege.83 Ademar provides numerous other examples of battles that took place in the context of sieges. The sieges of Poitiers and Brosse, mentioned earlier, resulted in such battles. Ademar also notes the siege of Rochemaux by Duke William and the effort by Count Boso to raise the siege with a large and powerful relief force. According to Ademar, Boso deployed his men in battle formation (in eum aciem struxit) and then attacked the besieging force. Duke William’s men, after being discomforted severely by the onslaught of the attacking battle line, recovered and scattered Boso’s forces, which subsequently fled. After achieving this victory in the field, William reestablished the siege and took the stronghold by storm (vi castrum cepit).84

Conclusion Ademar never finished a presentation copy of his Chronicon for Duke William. This may have been because the duke died in 1030, before Ademar had completed what he considered to have been a satisfactory final draft. In addition, from Ademar’s surviving autograph manuscripts, we cannot ascertain to whom this history was to be dedicated. Nevertheless, on the basis of the internal evidence, especially the panegyric, it seems clear that Ademar intended ultimately to dedicate the work to Duke William. Consequently, it seems reasonable to conclude that it was Ademar’s aim to have the contemporary part of the Chronicon read aloud at the ducal court. He hoped that the great men of Aquitaine, both lay and ecclesiastical, would conclude that he wrote accurate history and was to be trusted to provide accurate information and to tell the truth when writing other types of history. In light of

81 Ademar, Chronicon, 3.52. See Bachrach, Religion and the Conduct of War, 64–107, regarding pre-battle religious rituals during this period. 82 Ademar, Chronicon, 3.52 83 See Jim Bradbury, “Battles in England and Normany, 1066–1154,” Anglo-Norman Studies 6 (1983), 1–12, who notes (1) “One of the most obvious features of Norman warfare is the close, almost inseparable relationship between sieges and battles.” See also Bachrach, “L’art de la guerre,” 267–84. 84 Ademar, Chronicon, 3.42.

169

N A R R AT I V E W O R K S

his broader agenda, Ademar understood that it was imperative for the duke and his fideles to place their trust in him as a historian so that when he made his case for the apostolicity of St. Martial, these men and especially Duke William, would provide the necessary support for having the holy man recognized as an apostle. With this plan in mind, Ademar understood that he had to work very carefully in writing those parts of the Chronicon that provided information with which various members of the ducal court and their entourages likely were familiar. Ademar could not afford to make mistakes that could easily be recognized by members of his audience. Such errors would have the potential to undermine his reputation as an accurate historian or teller of the truth. If this were to take place, then the “facts” that he hoped to present later regarding St. Martial’s apostolicity might be called into question and his entire long-term project thereby undermined. Among the many subjects with which Ademar dealt in his Chronicon, with which he hoped to gain the confidence of his audience, military matters received a considerable amount of attention. Ademar had several sound reasons for his treatment of this subject. First, military operations were an important part of the political life in Aquitaine, and these matters were well-known and of interest to an important part of Ademar’s potential audience at the ducal court. Second, Ademar understood at least the basics of military matters as he garnered this information from Vegetius’s De re militari and likely from his uncles as well. By providing an accurate treatment of matters such as fortification terminology, the military organization of contemporary Aquitaine and her neighbors, and siege warfare, Ademar could be confident that both Duke William and his fideles would conclude that he was a truth teller and also be pleased that he mentioned them and their relatives in his text. Modern scholars engaged in the writing of military history can benefit from this verdict.

170

8 B R U N O O F M E R S E B U R G’S SAXON WAR A study in eleventh-century German military history

Background to the Saxon Wars During the 1070s and 1080s, the German kingdom was wracked by a civil war that pitted King Henry IV (1056–1106) against an evolving coalition of opponents, which ultimately included most of the population of the Saxon duchy as well as numerous great magnates in southern Germany.1 The war began in 1073 with a great rebellion of the Saxon secular and ecclesiastical office holders and common people against the royal government. However, the origins of the war date back to the reign of Henry III (1039–1056), the father of Henry IV. During the course of the 1040s and 1050s, Henry III undertook a rigorous policy of recovering royal estates that had been alienated either as benefices or which had been usurped from royal control.2 This policy, known in German as Henry III’s Revindikationspolitik, faced considerable resistance from magnates, particularly those in the Saxon duchy, who had benefitted from the largess of previous kings or who had availed themselves of royal property by less-than-legal means.3 The death of Henry III in 1056, leaving his six-year old son Henry IV under the care of regency government, meant that the efforts by the Crown to regain royal properties were put into abeyance. This hiatus would only last for a decade. When Henry IV came of age in 1065, he renewed his father’s policy of regaining direct control over royal lands.4 The 1 Thomas Zotz, “Der südwestdeutsche Adel und seine Opposition gegen Heinrich IV.,” in Welf IV. Schlüsselfigur einer Wendezeit: Regionale und europäische Perspektiven, ed. Dieter Bauer and Matthias Becher (Beck, Munich, 2004), 339–59. 2 For an overview of these issues, see Wolfgang Giese, “Reichsstrukturproblem unter den Saliern- der Adel in Ostsachsen,” in Die Salier und das Reich, 3 vols., ed. Odiolo Engels, Franz-Josef Heyen, Franz Staab, and Stefan Weinfurter (Thorbecke, Sigmaringen, 1991), I: 273–308. 3 See, for example, the discussion by Manfred Stimming, Das deutsche Königsgut im 11. und 12. Jahrhundert (Husum, Berlin, 1922), particularly 96–100 for the Revindikationspolitik of the Salian king; and Karl Leyser, “The Crisis of Medieval Germany,” Proceedings of the British Academy 69 (1983), 409–43, particularly 423–6. 4 See the brief survey of these events by I. S. Robinson, Henry IV of Germany 1056–1106 (Cambridge University Press, Cambridge, 1999), 63–77, as well as the more detailed study by Mathias Becher, “Die Auseinandersetzung Heinrichs IV. mit den Sachsen: Freiheitskampf oder Adelsrevolte?,” in

171

N A R R AT I V E W O R K S

young Henry put increasing pressure on recalcitrant magnates by constructing royal strongholds from which the king’s garrisons could compel compliance with his commands through the threat of violence if necessary. In addition, Henry IV spent an enormous amount of time in the Saxon duchy, requiring local magnates, both secular and ecclesiastical, to undertake the burden of supporting the large royal court at their own expense. Ultimately, Henry IV’s increasing pressure on the Saxon land-holding magnates and his efforts to assert ever greater royal control over the economic resources of the kingdom led to a violent response. In 1069, Margrave Dedi of Lower Lausatia raised a rebellion against King Henry’s rule in an effort to assert his family’s claim to properties in the eastern regions of Saxony.5 This rebellion failed, but the very next year Otto of Northeim raised the standard of rebellion against the king in response to having been removed from his office as duke of Bavaria (1061–1070). Otto was joined in his rebellion by Magnus Billung, the heir to the ducal seat in Saxony, whose family had suffered considerable losses as a result of Henry IV’s policy of recovering royal estates from secular magnates.6 The king was able to defeat this rebellion as well. However, in 1073, the entire Saxon duchy was convulsed by a broad-based, and apparently popular, rebellion against the king.7 In addition to the efforts by magnates to reverse the policy of revindicatio of royal estates, there appears to have been a groundswell of opposition among the lower orders of society, including small-scale landowners, to Henry IV’s efforts to impose taxation on private property and to increase the service obligations of the general population to the royal government, including work on royal fortresses.8 This violent uprising was to shake the foundations of the German monarchy and led to a civil war that lasted for a dozen years (1073–1085). The initial phase of the war went well for Henry IV, who decisively defeated a large Saxon army at the battle of the Unstrut (9 June 1075).9 In the aftermath of the battle, the king is depicted by hostile contemporary sources as pursuing a vindictive policy of

5 6

7

8

9

Vom Umbruch zur Erneuerung? Das 11. und beginnende 12. Jahrhundert, Positionen der Forschung, ed. Nicola Karthuas, Jörg Jarnut, and Matthias Wemhoff (Fink, Munich, 2006), 357–78. See Robinson, Henry IV, 64–5. Magnus Billung held office as duke of Saxony 1072–1106. For an overview of the relationship between the Billung dukes of Saxony and the Salian kings, see Gerd Althoff, “Die Billunger in der Salierzeit,” in Die Salier und das Reich, I: 309–30. See the discussion by Karl Leyser, “From Saxon Freedoms to the Freedom of Saxony: The Crisis of the Eleventh Century,” in Idem, Communications and Power in the Middle Ages 2: The Gregorian Revolution and Beyond (London, 1994), 51–67. These taxes are mentioned by Bruno of Merseburg in his Saxonicum Bellum, chapters 17 and 84. The best edition of Bruno’s Saxonicum Bellum is now to be found in Quellen zur Geschichte Kaiser Heinrichs IV.: Die Briefe Heinrichs IV., Das Lied vom Sachsenkrieg, Brunos Sachsenkrieg, Das Leben Kaiser Heinrichs IV., ed. and trans. by Franz-Josef Schmale and Irene Schmale-Ott (Wissenschaftliche Buchgesellschaft, Darmstadt, 1974), 192–405 with facing page German translation. This battle is discussed in all of the contemporary narrative sources, which are discussed later.

172

B R U N O O F M E R S E B U R G ’ S S A X O N WA R

persecution against the Saxon nobility and destruction of the Saxon people.10 By contrast, sources favorable to the German ruler depict him as imposing condign punishment on recalcitrant rebels.11 While both of these interpretations are driven by contemporary biases, it is undeniable that Henry IV was unable to use his victory at the Unstrut to bring about peace. The king’s conflict with Pope Gregory VII (1073–1085), long denoted in the scholarly literature as the Investiture Contest, provided the impetus for the estranged aristocratic leadership of southern Germany to join with the Saxons in rebellion against Henry IV.12 The proximate cause for this multifront rebellion was the excommunication of King Henry by the pope in early 1076. Dukes Welf of Bavaria (1070–1077, 1096–1101), Berthold of Carinthia (1061–1077), and Rudolf of Swabia (1057–1077) joined in a renewed rebellion, and in March 1077 elected Rudolf as king in opposition to Henry IV. The royal and rebel forces fought three major battles over the next several years at Mellrichstadt (7 August 1078), Flarchheim (1 January 1080), and the Elster river (15 October 1080). The first of these battles had an inconclusive result, but the latter two were significant defeats for Henry IV’s armies. Nevertheless, the German king prevailed over his opponents. Rudolf died of the wounds that he suffered at the Elster river, and his successor as anti-king, Hermann of Salm (1081–1088), was unable to drive Henry IV from the throne. By 1084, Henry had reestablished his authority throughout the German kingdom and was able to depose all of the secular and ecclesiastical magnates who had opposed him.13

Scholarly treatment of the Saxon Wars Several of the battles between Henry IV and his opponents have received attention from scholars.14 However, these battlefield studies have not been integrated into broader questions that now animate much of the scholarly discussion regarding

10 See the discussion of these and other charges made by contemporaries against Henry IV in Hermann Kamp, “Die Vorwürfe gegen Heinrich IV. – eine Zusammenfassung,” in Heinrich IV., ed. Gerd Althoff (Thorbecke, Ostfilder, 2009), 355–67. 11 These sources are discussed in detail by Tilman Struve, “Der ‘gute’ Kaiser Heinrich IV: Heinrich IV. im Lichte der Verteidiger des salischen Herrschaftssystems,” in Heinrich IV., 161–88. 12 The investiture contest has been the subject of enormous scholarly attention. For a useful introduction to basic issues and sources, see Uta-Renate Blumenthal, The Investiture Controversy: Church and Monarchy from the Ninth to the Twelfth Century (University of Pennsylvania Press, Philadelphia, 1991). 13 See the discussion by Robinson, Henry IV, 239–74, who describes the successes of the king against his opponents both in Germany and against Pope Gregory VII. 14 See Hans Delbrück, Geschichte der Kriegskunst im Rahmen der politischen Geschichte, 6 vols. (De Gruyter, Berlin, 1900–1936), the first three volumes of which translated under the title History of the Art of War: Within the Framework of Political by Walter J. Renfroe (Greenwood Press, Westport, CT, 1975–1982), here III: 131–45; Leopold Auer, “Die Schlacht bei Mailberg am 12. Mai 1082,” Militärhistorische Schriftenreihe 31 (1975), 1–31; and John Gillingham, “An Age of Expansion, c. 1020–1204,” in Medieval Warfare: A History, ed. Maurice Keen (Oxford University Press, Oxford, 1999) 59–88, here 73–6. The recent monograph by Hartmut Lauenroth, Die

173

N A R R AT I V E W O R K S

medieval warfare, including the organization of armies, the role played by fortifications and sieges in campaigns, and logistics. In fact, the conduct of war more generally in the German kingdom during the eleventh century has not benefitted from significant investigation.15 Fortunately, this lacuna in the scholarship regarding the conduct of war in the German kingdom during the later eleventh century is not based on a lack of source materials. Charters and letters survive in great numbers from this period,16 and there is an increasingly large volume of information available from archaeological excavations.17 In addition, there is a rich collection of narrative histories written during and immediately following the Saxon Wars of the 1070s and 1080s, which provide considerable information about many aspects of the conduct of war.18 However, none of these narrative texts has benefitted from

15

16

17

18

Sachsenkriege unter Heinrich IV. und Heinrich V (Schäfer, Langenbogen, 2010), despite its title, is a popular treatment of political events rather than a military history. The one lengthy study of this period from a military perspective by Bruno Scherff in his 1985 dissertation at the University of Bonn, Studien zum Heer der Ottonen und der ersten Salier (919–1056) ends with the death of King Henry III (1039–1056) and is notable for the topics that it explicitly avoids, including military equipment, logistics, the sizes of armies, military levies, fortifications, and, above all, sieges. Concerning the lack of scholarship on German military history more generally in the period after the Second World War, see Hans-Hennig Kortüm, “Der Krieg im Mittelalter als Gegenstand der historischen Kultur-Wissenschaften: Einer Annäherung,” in Krieg im Mittelalter, ed. idem (Akademie Verlag, Berlin, 2001), 13–43; and by David S. Bachrach, “Feudalism, Romanticism, and Source Criticism: Writing the Military History of Salian Germany,” Journal of Medieval Military History 15 (2015), 1–25. particularly 1–3. There are almost 500 surviving royal charters from Henry IV’s reign. Die Urkunden der deutschen Könige und Kaiser: Die Urkunden Heinrichs IV, 3 vols., ed. D. von Gladiss, and A. Gawlik (Monumenta Germaniae Historica, Hanover, 1941–1978). In addition, there are 42 surviving letters issued by this king. These were published in a facing-page Latin and German edition by Franz-Josef Schmale in Quellen zur Geschichte Kaiser Heinrichs IV (Darmstadt, 1968), and are available in English translation in Imperial Lives and Letters of the Eleventh Century, trans. and annotated by Theodor E. Mommsen and Karl F. Morrison, 2nd edition (New York, 2000). See, for example, Helmut Peinhardt, Wehrhaftes Erfurt: Die mittelalterliche Stadtbefestigungen (Jena Hain Verlag, Rudolstadt, 1996); Karl Bernhard Kruse, “Die Bernwardsmauer in Hildesheim. Befestigung vom Domhügel und Stadt im Mittelalter,” in Stadtarchäologie in Norddeutschland westlich der Elbe, ed. Heiko Steuer and Gerd Biegel (Habelt, Bonn, 2002), 199–210; Franz Seberich, Die Stadtbefestigung Würzburgs Teil 1. Die mitteralterliche Befestigung mit Mauern und Türmen (Freunde Mainfränkischer Kunst und Geschichte, Würzburg, 1962); Marcus Trier, “Die Kölner Stadtbefestigung im Mittelalter und in der frühen Neuzeit,” Lübecker Kolloquium zur Stadtarchäologie im Hanseraum 7 (2010), 535–52; and Wolf-Dieter Steinmetz, Geschichte und Archäologie der Harzburg unter Saliern, Staufern un Welfen 1065–1254 (Braunschweigisches Landesmuseum, Bad Harzburg, 2001). These include Lampert of Hersfeld, Annales, ed. Oswald Holder-Egger, MGH Scriptores rerum Germanicarum in usum scholarum separatism editi 38 (Monumenta Germaniae Historica, Hanover, 1894), now available in translation as The Annals of Lampert of Hersfeld, translated and annotated with an introduction by I. S. Robinson (Manchester University Press, Manchester, 2015); Hermann of Reichenau, Chronicon, ed. G. H. Pertz, MGH SS 5 (Monumenta Germaniae Historica, Hanover, 1844), Die Chroniken Bertholds von Reichenau und Bernolds von Konstanz 1054–1100, ed. I. S. Robinson, MGH Nova Series 15 (Monumenta Germaniae Historica, Hanover, 2003), which are available in translation by I. S. Robinson, Eleventh-Century Germany: The Swabian Chronicles

174

B R U N O O F M E R S E B U R G ’ S S A X O N WA R

a systematic analysis with respect to the information that they can shed on military matters. The burden of this study, therefore, is to begin the process of illuminating these questions through an investigation of Bruno of Merseburg’s Saxon War, which provides one of the most detailed accounts of the conduct of military operations by both King Henry IV and his opponents.19

Bruno and his Bellum Saxonicum It is generally agreed by scholars that Bruno was commissioned to write his history by Bishop Werner of Merseburg (1059–1093), to whom Bruno dedicated the work, although the purpose for the writing of the text has been the subject of some debate.20 The Bellum Saxonicum is characterized by Bruno’s obvious parti pris in favor of the Saxons and his sexually charged attacks against the character of Henry IV.21 However, despite his highly polemical attacks against the king, Bruno presented his text as a work of history, and he worked diligently to convince his audience that he was relaying matters in a truthful manner.22 Nevertheless, Bruno’s various biases do color his work to a significant degree. Bruno’s negative depiction of the participation of men of the lower orders in the conduct of military

19 20

21

22

(Manchester University Press, Manchester, 2008); Carmen de bello saxonico, ed. and trans. Franz-Josef Schmale in Quellen zur Geschichte Kaiser Heinrichs IV,142–189 along with and Vita Heinrici IV. Imperatoris, in ibid, which is available in English translation in Imperial Lives and Letters of the Eleventh Century, trans. and annotated by Theodor E. Mommsen and Karl F. Morrison, 2nd edition (Columbia University Press, New York, 2000); Frutolf of Michelsberg, Chronicon, ed. G. Waitz, MGH SS 6 (Monumenta Germaniae Historica, Hanover, 1844), now available in English translation in Chronicles of the Investiture Context; Frutolf of Michelsberg and His Continuators, trans. and annotated T. J. H. McCarthy (Manchester University Press, Manchester, 2014). For an introduction to the scholarly traditions regarding these texts see Simon M. Karzel, Nihil crudelius a barbaris perpeti potuissent: Die Darstellung von Krieg und Gewalt in den historiographischen Quellen zur Zeit Heinrichs IV. (Tectum-Verl, Marburg, 2008). We hope to write a similar analysis of the Annales by Lampert of Hersfeld, whose treatment of military matters provides a similar level of detail as Bruno’s work. For a summary of the early scholarly treatment of Bruno, see Schmale and Schmale-Ott, Quellen, 28–31; and the more recent survey by Karzel, Nihil crudelius, particularly 44–7. Concerning Bruno’s causa scribendi see Gerd Althoff, “Pragmatische Geschichtsschreibung und Krisen. Zur Funcktion von Brunos Buch von Sachsenkrieg,” in Pragmatische Schriftlichkeit im Mittelalter: Erscheinungsformen und Entwicklungsformen, ed. Hagen Keller (Fink, Münster, 1992), 95–107; and Wolfgang Eggert, “Das Wir-Gefühl bei fränkischen und deutschen Geschichtsschreibern bis zum Investiturstreit,” in Wir-Gefühl und Regnum Saxonum bei frühmittelalterlichen Geschichtsschreibern, ed. Wolfgang Eggert and Barbara Pätzold (H. Böhlaus, Weimar, 1984), 13–179, here 158–61; and Idem, “Wie ‘pragmatisch’ ist Brunos Buch vom Sachsenkrieg?,” Deutsches Archiv 51 (1995), 543–53. The nature and purpose of these attacks have been discussed in detail by Megan McLaughlin, “’Disgusting Acts of Shamelessness’: Sexual Misconduct and the Deconstruction of Royal Authority in the Eleventh Century,” Early Medieval Europe 19.3 (2011), 312–31. The issue of Bruno’s self-conception as a historian is treated in detail by Bernard S. Bachrach and David S. Bachrach, “Bruno of Merseburg and His Historical Method c. 1085,” Journal of Medieval History 40 (2014), 381–98.

175

N A R R AT I V E W O R K S

operations can be understood to have been shaped by his disdain for such men.23 Conversely, Bruno’s focus on the successes of aristocrats on campaign went hand in hand with his overall effort to depict favorably men of this social and economic status. In a similar manner, Bruno’s effort to defame Henry IV in general certainly had an effect on the author’s depiction of the king in specifically military contexts. However, when the account of military matters simply provided the background for matters that were central to Bruno’s agenda, it is likely that he presented this information in as accurate a manner as he could so as to increase the plausibility of his account to an audience familiar with the conduct of war.24

Military organization Much like his contemporaries Lampert of Hersfeld and Berthold of Reichenau, Bruno devotes attention to two broadly different types of fighting men in his account of the wars between the Saxons and Henry IV. The first of these groups can be characterized as part-time fighting men or militia forces, who were mobilized for a particular military operation.25 These included military actions both in the defense of their home districts and offensive campaigns that took men much further afield. Consistent with his general parti pris in favor of the aristocracy and against the lower social orders, Bruno frequently is quite hostile toward the peasants (rustici) and townsmen (cives) who served as militia troops. He regularly impugns their intentions and negatively depicts their military effectiveness. It is notable that Bruno is even hostile toward the men of the lower social orders who served in the armies fighting against Henry IV. The second group of fighting men discussed by Bruno were those who served as professional soldiers in the military households of secular and ecclesiastical magnates. Consistent with other eleventh-century authors in Germany, Bruno reserved the term miles, plural milites, to denote these men.26

Militia forces It is one of the ironies of Bruno’s account that although he is very interested in highlighting the inherent superiority of aristocrats over the lower orders of society, he devotes considerable attention to the military activities and organization of 23 See the discussion of this issue by Bachrach, “Feudalism, Romanticism, and Source Criticism,” 9–13. 24 This point has been made with respect to Carolingian sources specifically dealing with military matters by Thomas Scharff, Die Kämpfe der Herrscher und der Heiligen: Krieg Und historische Erinnerung in der Karolingerzeit (Wissenschaftliche Buchgesellschaft, Darmstadt, 2002), 42. With respect to the question of rhetorical plausibility, also see the discussion by Justin C. Lake, “Truth, Plausibility, and the Virtues of Narrative at the Millennium,” Journal of Medieval History 35 (2009), 221–38. 25 See Bachrach, “Feudalism, Romanticism, and Source Criticism,” 13–22. 26 This issue is treated in detail with respect to a range of late eleventh-century German historians by David S. Bachrach, “Milites and Warfare in Pre-Crusade Germany,” War in History 23.3 (2015), 298–343.

176

B R U N O O F M E R S E B U R G ’ S S A X O N WA R

those whom he deems his social inferiors. In part, this is because Bruno wished to damage the reputation of King Henry by tying the latter to disreputable elements in society, such as townsmen, or even worse, merchants! But even more importantly, Bruno’s emphasis on the participation in war by the middling levels of society, and even peasants, is born from a desire to ridicule them, including those on his own side, and to blame them for setbacks on campaign, as well as for the inherent horrors of military conflict. In discussing the very beginning of the great military struggle between the Saxons and Henry IV in 1073, Bruno records that an assembly of all of the Saxons, from the highest rank to the lowest, gathered together at Hoetensleben to discuss their common defense against the aggression of the king.27 After listening to the speeches of the magnates who had suffered at Henry’s hands and hearing the catalog of injustices that the common people (parvi) had suffered, everyone there, whom Bruno describes as a very large army (maximus exercitus), swore to defend their homeland and their liberty by force of arms.28 Bruno returns to the point that armies mobilized by the Saxons to defend their homeland were drawn from every rank of free men, down to the lowest peasants, in his discussion of Henry IV’s invasion of the Saxon duchy in February 1074. Bruno indicates that the king was able to make the plausible claim to his own troops that they had nothing to fear because the Saxon army that they faced consisted only of peasants (rustici).29 Bruno tied this low social status to the idea that the Saxon forces lacked mounted troops and that the individual men lacked military training, i.e. they were bellicarum rerum imperiti.30 Bruno once more drew attention to the broad cross-section of Saxon society that was mobilized to fight in his discussion of the events leading up to the battle of the Elster river. On this occasion, according to Bruno, the Saxons met King Henry’s invading army at a place called Künkel near the frontier between Thuringia and Saxony. Rather than face the Saxons, however, Henry devised a ruse de guerre, in which a rapid strike force of mounted troops was deployed toward Goslar in Saxony, while the main army headed toward the Thuringian fortress city of Erfurt.31 In describing their response to Henry’s maneuver, Bruno distinguishes between the bulk of the Saxon army, which consisted of a great crowd of foot soldiers (magna turba peditum), and the smaller element composed of Saxon mounted troops.32 As the campaign progressed, elements of the Saxon army, including both mounted troops and eventually foot soldiers as well, caught up with Henry’s forces at the Elster river. In describing the course of the ensuing battle, Bruno allowed his bias against the lower social orders to come to the fore and claimed 27 Bruno, Saxonicum Bellum, ch. 23–24. The location of Hoetensleben has not been established by scholars. 28 Bruno, Saxonicum Bellum, ch. 26–27. 29 Ibid., ch. 31. 30 Ibid. 31 Bruno, Saxonicum Bellum, ch. 121. 32 Ibid.

177

N A R R AT I V E W O R K S

that after their victory in taking the royal camp, the Saxon foot soldiers wished to have the entire benefit of the victory for themselves and to take all of the booty.33 These low-class Saxons only were restrained, according to Bruno, by the good sense of Duke Otto of Northeim, who warned the foot soldiers to maintain their formation and keep out of the tents in the royal camp until it was clear that all of the enemy troops had been defeated.34

Militia troops in expeditionary armies In addition to his focus on differences in social status among the troops who comprised the Saxons’ defensive armies, Bruno devotes considerable attention to the issue of social and economic status while discussing the expeditionary armies of both King Henry and his opponents. For example, in his treatment of the campaigning season of 1077, after Duke Rudolf Swabia raised the standard of rebellion against Henry IV, Bruno observed that the Salian king had increasing difficulty in mobilizing a large army with which to invade Saxony.35 As a consequence, aside from forces that Henry summoned from Bavaria and Bohemia, Bruno asserted that the king’s army was small and weak, commenting specifically that the majority of his troops were merchants (mercatores).36 In support of his social rather than his political agenda, Bruno was not content to point out the nonprofessional element in the armies of Henry IV alone. He also observed the presence of men of low social and economic status in the armies of Henry’s opponents. In discussing the Saxon invasion of the Rhineland in October 1076, for example, Bruno commented that the goal of the Saxon magnates was to unite their forces with an army of Swabians, who recently had joined their duke’s revolt against Henry IV.37 However, although they were now allies, the leaders of the Saxons and Swabians feared to bring their troops into too close proximity. They were worried about the potential for violence between their men resulting from lingering hostilities because Saxons and Swabians had fought on opposite sides in the battle on the Unstrut in June of the previous year. Bruno specifically comments that it was the baser people (viles personae) who posed the greatest source of concern for the commanders.38 In a similar manner, Bruno also drew attention to the low-status “vulgar crowd” (vulgus) who served in the army of Rudolf of Swabia after the latter had been consecrated as king.39 Bruno claims, for example, that during the siege of Würzburg in

33 34 35 36 37 38 39

Ibid., ch. 122. Ibid. Bruno, Saxonicum Bellum, ch. 95. Ibid. Bruno, Saxonicum Bellum, ch. 88. Ibid. Bruno, Saxonicum Bellum, ch. 94.

178

B R U N O O F M E R S E B U R G ’ S S A X O N WA R

August 1077, Rudolf was reluctant to capture the city by storm because he feared that he would not be able to keep the “vulgar crowd” in check and stop them from looting churches and stealing clerical property once they had overwhelmed the defenders.40 Consequently, according to Bruno, Rudolf preferred to engage in a lengthy siege rather than risk the dishonor that would result from allowing the destruction of sacred sites.41 Obviously, however, Bruno’s comments here do not bear close scrutiny from a military point of view.

Professional fighting men Although Bruno clearly believed that his audience would find it plausible that there were large numbers of farmers and townsmen in the armies of Henry IV and his Saxon opponents, he also indicates that a central element of the military forces on both sides in the conflict consisted of the military households of the magnates, that is milites. In his discussion of these milites, Bruno indicates that the individual military households of the various magnates had a hierarchical structure and that some milites were far more important than others. For example, in the course of discussing the union of the Saxons and Swabian armies at Oppenheim near Worms, mentioned previously, Bruno observed that there was still hostility not only between the viles personae but also among the milites on both sides.42 As a consequence, according to Bruno, after the leaders of the Saxons and Swabians swore oaths of loyalty to each other, the soldiers of the second and even the third rank (ordinis secondi sive tertii partis utriusque milites) “gave each other the kiss of peace and forgave each other, not without a considerable shedding of tears, for the injuries that they had given each other.”43 Similarly, in discussing the assembly at Kaufungen, located near Cassel, in February 1081, at which representatives of the rebels and Henry IV sought to work out a peace settlement, Bruno claimed that the lower-ranking soldiers (milites plebei partis) in the royalist army favored the Saxon proposals, but their leaders refused to accept them.44

Royal military household Throughout the Bellum Saxonicum Bruno devotes considerable attention to the military forces that were employed directly by the king. The royal military household, as Bruno presents it, was divided into two parts. The first of these consisted of men who were attendant directly on the king, and the second, numerically larger group, was composed of those royal troops whom the king delegated to serve as garrisons in various strongholds, including both free-standing fortresses, 40 41 42 43 44

Ibid. Ibid. Bruno, Saxonicum Bellum, ch. 88. Ibid. Bruno, Saxonicum Bellum, ch. 128.

179

N A R R AT I V E W O R K S

and fortified cities. These elements of the royal military household should not be understood as entirely separate, however, as there would, in Bruno’s telling, appear to have been considerable overlap between the two. For example, in discussing the return of Henry IV from his meeting at Bardowick, located near Lünaburg, with King Sweyn II of Denmark (1047–1074), Bruno notes that the German king was accompanied by a small force (pauci) of fighting men.45 After his meeting with Sweyn, King Henry inspected the fortress of Lüneburg and then detached a subunit of seventy men, whom our author denotes as the king’s most faithful (fidelissimi) supporters, to garrison the stronghold, which was located some 5 kilometers south of Bardowick.46 In discussing these events, therefore, Bruno makes clear that the forces normally attendant on the king also were utilized, at least on an ad hoc basis, to perform garrison duties. In addition to the events at Lüneburg, Bruno emphasizes on several other occasions that fortresses throughout Saxony were garrisoned by royal troops of this type and provides details that indicate these men were drawn from the royal military household.47 Among these fortresses, for example, was the citadel in the episcopal city of Meißen, where the garrison commander, Burchard, was represented by Bruno as a close confidant of the king.48 The large scale of the royal forces that were placed into garrisons in Saxony is suggested by Bruno in his discussion of the temporary peace settlement that was arranged between Henry IV and the Saxons in the early spring of 1076.49 According to Bruno, following the decision of the Saxons not to fight at Oberspier in October 1075 and their surrender of hostages, King Henry gained control over the entirety of the Saxon duchy.50 The king then secured his hold over the Saxon territory by establishing garrisons in all of the cities and fortresses in the duchy.51 The men who were established are described here by Bruno as the king’s fideles, the same term our author used in denoting the seventy men, discussed earlier, who were delegated by the king to garrison Lüneburg in 1071.52

Military households of magnates Bruno routinely identifies by name numerous magnates, both secular and ecclesiastical, who served in the armies of both Henry IV and of his opponents. Bruno also makes clear that these magnates, like the king, led their military households on campaign. At the battle of Mellrichstadt, for example, both Archbishop Werner 45 46 47 48 49 50 51 52

Bruno, Saxonicum Bellum, ch. 19–20. Bruno, Saxonicum Bellum, ch. 21. Bruno, Saxonicum Bellum, ch. 11, 16, 29, 60, and 84. Bruno, Saxonicum Bellum, ch. 11. Bruno, Saxonicum Bellum, ch. 60. Bruno, Saxonicum Bellum, ch. 54–55. Bruno, Saxonicum Bellum, ch. 60. Ibid.

180

B R U N O O F M E R S E B U R G ’ S S A X O N WA R

of Magdeburg and Bishop Werner of Merseburg are depicted by Bruno as present with their men on the battlefield.53 In this case, Bruno was critical of both bishops, despite the fact that they both had been his patrons, because they had presumed to lead their “legions” (legiones) into battle rather than acting properly as churchmen and singing psalms.54 Bruno offered another example of the negative consequences that could arise from bishops maintaining military households, on this occasion with regard to supporters of King Henry, when discussing events that took place in late June 1076 at Mainz.55 Earlier that year, the Saxons regained control over a number of royal fortresses in their duchy and expelled the king’s garrisons.56 In response, Henry mobilized his army and marched to Mainz, where he hoped to use several high-ranking hostages to restore his earlier position in Saxony. However, according to Bruno, once the royal army arrived, a fight broke out between the soldiers (milites) of Mainz and Bamberg, that is between the military households of the bishops of these two cities, which led to a major fire at Mainz. Bruno gloats that during the confusion, the hostages made their escape and thwarted Henry’s plans.57 In addition to observing the regular maintenance of military households by bishops, Bruno points to the same practice by secular magnates. For example, in recounting events during the battle at Mellrichstadt, mentioned previously, Bruno observed that Count William of Kamburg rashly advanced ahead of the main rebel army with his own small military household. Because of this poor decision, William was captured and only escaped once a larger unit of Saxons overwhelmed the soldiers of Eberhard the Bearded, King Henry’s close advisor, who had captured him.58

Battlefield operations Although infrequent by the standard of Clausewitzian-inspired battle-seeking warfare of the modern age, by a medieval standard the four battles discussed by Bruno in the four and half years between June 1075 and October 1080 represent a very high density of large-scale encounters in the field.59 Bruno sought in his nar53 54 55 56 57 58 59

Bruno, Saxonicum Bellum, ch. 96. Ibid. Bruno, Saxonicum Bellum, ch. 85. Bruno, Saxonicum Bellum, ch. 84. Bruno, Saxonicum Bellum, ch. 85. Bruno, Saxonicum Bellum, ch. 99. Regarding the battle-averse nature of medieval military commanders, and the relative increase in battles in the field in the modern era, see the discussion by Martin van Creveld, Supplying War: Logistics from Wallenstein to Patton (Cambridge University Press, Cambridge, 1977); and Christopher Duffy, The Fortress in the Age of Vauban and Frederick the Great 1660–1789 (Routledge, London, 1985). However, also see the critique of this model by Stephen Morillo, “Battle Seeking: The Contexts and Limits of Vegetian Strategy,” Journal of Medieval Military History 1 (2002), 21–41. For the orthodox view regarding the relative rarity of battles in the medieval period, and the general desire of medieval commanders to avoid battle, see John Gillingham, “Richard I and the

181

N A R R AT I V E W O R K S

rative to provide exceptionally detailed accounts regarding the battles themselves and also the wide range of preparations made by both sides before they committed to a set-piece confrontation. This high level of detail permits a close examination of a number of factors that Bruno presents as having an important impact on the decision by military commanders to join the battle and also as having an impact on the final result of these confrontations.

Military intelligence and the control of information Among the most important potential assets that a military commander has in planning a campaign is maintaining secrecy, thereby denying the opponent an opportunity to react in a timely and effective manner.60 Bruno makes it clear that the military commanders of the Saxons, like their contemporaries in the AngloNorman world, were fully aware of the dangers of being surprised by enemy forces and therefore sought to obtain intelligence about Henry IV’s intentions. In recounting the events of February 1074, for example, Bruno noted that as Henry IV prepared to invade Saxony, the Saxons were warned in advance, i.e. they received reliable military intelligence, that the king had mobilized an army and was marching through Thuringia.61 In fact, according to Bruno, the Saxons received this military intelligence with sufficient time to mobilize a large army of their own and to take up a strong defensive position at the fortification of Vacha, which was located just 30 kilometers east of the fortress city of Erfurt in Thuringia, well south of the Saxon frontier.62 The ability of the Saxons to obtain information about the king’s plans and to prepare accordingly for an invasion of their homeland in the winter of 1074 was, in Bruno’s presentation, the first in a series of successes in gaining military intelligence that would go on until 1081, when he ended his account. Bruno recounts six

Science of War in the Middle Ages,” in War and Government in the Middle Ages: Essays in Honour of J. O. Prestwich, ed. John Gillingham and J. C. Holt (Boydell, Woodbridge, 1984), 78–91; and idem, “‘Up with Orthodoxy!’ In Defense of Vegetian Warfare,” Journal of Medieval Military History 2 (2002), 149–58. For the general rarity of battles in the German kingdom of the Ottonians and Salians as compared with the overall number of military campaigns, and the comparatively greater number of sieges conducted by German armies, see the discussion by Scherff, Studien zum Heer der Ottonen und der ersten Salier, 11. 60 The topic of military intelligence in the medieval period is one that has received comparatively little attention from scholars. One important exception is the study by J. O. Prestwich, “Military Intelligence under the Norman and Angevin Kings,” in Law and Government in Medieval England and Normandy: Essays in Honour of Sir James Holt, ed. George Garnett and John Hudson (Cambridge University Press, Cambridge, 1994), 1–30, which covers a period contemporaneous with the civil war in Germany, but from a more western perspective. Also see Scherff, Studien zum Heer, 121 and 138, who observes that the Ottonian and Salian kings of Germany must have utilized sources of both military and political intelligence, but without providing a detailed discussion of the topic. 61 Bruno, Saxonicum Bellum, ch. 31. 62 Ibid.

182

B R U N O O F M E R S E B U R G ’ S S A X O N WA R

further efforts by King Henry to invade Saxony in June 1075, October 1075, the autumn of 1078, January 1080, October 1080, and December 1081.63 In each case, the Saxons received sufficient prior warning, according to Bruno, to mobilize a substantial army to defend the frontiers of their homeland, or even to advance into Thuringia to blunt the king’s advance. Despite his strong parti pris against Henry IV, Bruno acknowledged that the king also successfully acquired intelligence about the operations of the Saxons and used this to his advantage. In the early summer of 1078, the Saxons, under the leadership of Rudolf of Swabia, sought to undertake an offensive against Henry IV in eastern Franconia. To this end, the Saxons mobilized an army that would invade from the north and successfully gained the commitment of the Swabians to launch an invasion of Franconia from the southwest.64 However, as Bruno records, King Henry obtained intelligence of this pincer operation and deployed his army so as to keep the Saxons and Swabians from joining forces. As a consequence, Henry led his army to Mellrichstadt, which is located some 80 kilometers northeast of the fortress city of Würzburg, and thereby blocked the route of the Saxons as the latter marched south.65

Scouting Bruno makes clear throughout his narrative that military leaders on both sides of the Saxon Wars regularly employed scouts (exploratores). These men were to obtain specific information about the numbers of troops they might face on the battlefield, how these soldiers were equipped, and how they were deployed. Bruno indicates that in many cases, the information provided by scouts played a decisive role in the decision of commanders whether to commit to a battle or to withdraw their armies and avoid an engagement. In discussing Henry IV’s initial invasion of Saxony in January 1074, for example, Bruno observes that the king’s army was quite large, but that the Saxon forces that mobilized to defend the frontier at the fortress of Vacha, located on the Werra river some 20 kilometers east of Hersfeld, were nearly double the size of the royal army.66 As part of his ongoing effort to depict Henry IV in negative terms, Bruno claims that the king misled his own men about the type of force that they faced, insisting that the Saxon army consisted of untrained peasant levies, who lacked both individual skill in combat and had only limited mounted forces, who might prove difficult to defeat on the battlefield.67 However, as Bruno recounts, both sides dispatched scouts “to investigate carefully the strength of each of the armies,” and the scouts on both sides “gave very clear reports of what they had seen.”68 As a result of the new information brought by

63 64 65 66 67 68

Bruno, Saxonicum Bellum, ch. 46, 54, 103, 117, 121, and 125. Bruno, Saxonicum Bellum, ch. 96. Ibid. Bruno, Saxonicum Bellum, ch. 31. Ibid. Ibid.

183

N A R R AT I V E W O R K S

the scouts, namely that the Saxon force was larger than the king’s army and that Saxons had both highly trained soldiers and mounted forces, the men who commanded various contingents in the royal army made clear to Henry that they would not enter battle. As a consequence, Henry was forced to enter negotiations with the Saxons and establish a temporary peace with them.69 Bruno recounts a similar anecdote regarding the role of scouts in the course of Henry IV’s planned invasion of Saxony in the late autumn of 1078. On this occasion, the royal army advanced to the Thuringian-Franconian frontier, but soon discovered that the rebel forces had received advanced warning and were already mobilized to face the king’s troops well to the south of Saxony.70 Bruno suggests that when reports filtered into the royal camp about the size of the Saxon army, the king’s men did not at first believe them. However, after scouts had been dispatched and returned with word that the Saxon force was exceptionally large, Henry ordered a retreat and marched instead toward Swabia.71 Consequently, Bruno again makes clear that information provided by scouts played a decisive role in determining whether to enter battle. In yet another case, Bruno claims that information provided by scouts enabled King Henry to deploy his forces in a manner that severely discomfited the Saxons.72 In October 1080, Henry launched yet another invasion of Saxony. When the royal army arrived near Eisenach, located some 50 kilometers west of Erfurt in Thuringia, Henry learned that the Saxons were encamped in and around the fortress of Künkel, blocking his passage northward into Saxony.73 Bruno once again points out that Henry dispatched scouts to determine the strength of the Saxon forces and again realized that their numbers made a pitched battle in the field too dangerous.74 On this occasion, however, Henry, as Bruno admits, devised a cunning ruse de guerre whereby he forced the Saxons to hesitate and divide their forces, while he led his army to the fortress city of Erfurt, which he captured and sacked.75 The king knew, as Bruno observed, that the Saxons under normal circumstances would be warned well ahead of time by their own scouts of any effort to march against Erfurt and would be in a position to block Henry’s line of march eastwards. Therefore, as discussed previously, the king sent a flying column of his fastest mounted troops in the direction of the royal palace complex at Goslar, located some 140 kilometers to the north in Saxony. Their orders were to burn everything in their path and give the impression that Henry was heading into the heart of Saxony.76 While the Saxons were distracted, Henry then undertook a rapid march to Erfurt, 69 70 71 72 73 74 75 76

Ibid. Bruno, Saxonicum Bellum, ch. 103. Ibid. Bruno, Saxonicum Bellum, ch. 121. Ibid. Ibid. Ibid. Bruno, Saxonicum Bellum, ch. 121.

184

B R U N O O F M E R S E B U R G ’ S S A X O N WA R

which he sacked before the Saxons could respond, because their scouts led them to focus on the actions of the flying column that had been dispatched toward Goslar.

Battlefield tactics Bruno does not provide hyperbolic tributes to the military genius of the rebel commanders, nor does he hide the fact that the tactics employed by Henry IV’s opponents on the battlefield led, on occasion, to significant reverses. Indeed, in his account of the battle of the Unstrut in June 1075, Bruno makes clear that the Saxons were undisciplined, unprepared for combat, and disorganized as they met Henry’s troops.77 Bruno does his best to ameliorate this negative depiction of the Saxons by focusing on the magnates on the king’s side who were killed in battle and by lauding the bravery of those of his countrymen who stood their ground. Bruno says of the Saxons: So, as happens when people are taken by surprise, the few, who were brave and had their arms to hand, went into battle. The many, however, who lacked bravery and their arms, turned in flight. But in truth, the few who remained steadfast in the battle not only did their duty, but also did the duty, insofar as they were able, of those who had fled as well. If God had not chosen to humble our arrogance, our few men would have driven off their entire army.78 However, when the opportunity presented itself to depict the military performance of the entire rebel army in more positive terms, Bruno certainly took advantage. In discussing the battles of Flarchheim and Elster river in January and October 1080, respectively, Bruno makes clear that the Saxons were deployed effectively and that their commanders utilized sound tactics to inflict very heavy losses on royalist troops. Moreover, at the battle of Elster river, Bruno emphasizes that these effective tactics allowed the Saxons to achieve a major victory over King Henry. The military operations that led to the battle of Flarchheim began with the decision by King Henry to undertake a winter campaign against the Saxons in order to press the political advantage he had achieved by negotiating a separate peace with a number of prominent Saxon magnates, including Margrave Eckbert of Meißen.79 Henry IV, following the pattern established over the previous two years, again marched through Thuringia and made his camp 50 kilometers eastnortheast of Erfurt along the banks of the Eichbach, a right tributary of the Unstrut river. The Saxon army was deployed on the opposite bank of the Eichbach. Bruno emphasizes that the rebels had a very strong defensive position. Not only would Henry’s troops have to cross the deep river to fight them, but the rebels also held 77 Bruno, Saxonicum Bellum, ch. 46 and 96–99. 78 Bruno, Saxonicum Bellum, ch. 46. 79 Bruno, Saxonicum Bellum, ch. 117.

185

N A R R AT I V E W O R K S

the higher ground so that royalist troops who did manage to cross the water would have to advance uphill in order to engage the enemy.80 In addition to establishing a strong defensive front along the Eichenbach, the anti-king Rudolf held a substantial force in reserve. As Bruno goes on to explain, this demonstration of tactical prudence by Rudolf was to prove decisive during the course of the battle. King Henry recognized that the Saxons held a virtually impregnable position. He therefore did not attempt to assail their forces at their strongest point and force a crossing of the Eichenbach against a well-positioned opponent, who also held the higher ground. Instead, showing tactical prudence of his own, Henry secretly dispatched a substantial part of his army to cross an unguarded section of the river and assault the rebel forces from the rear. In describing Henry’s maneuver, Bruno observed, “as was their custom, they craftily came around to attack us unexpectedly.”81 However, the king’s use of stealth did not, at least in Bruno’s account, achieve its desire aim. Instead, the troops whom Rudolf had kept in reserve were able to meet Henry IV’s assault and drove off the royalists after a bitter fight.82 In discussing the conduct of the battle, Bruno explains that as soon as Henry’s troops were sighted to the rear of the Saxon army, Rudolf sent a messenger to Otto of Northeim asking him to withdraw his men from the riverbank of the Eichenbach and come to his aid. However, in a clear illustration of both the difficulty involved in changing battle plans in the course of combat and of the sound decision that Rudolf made in keeping a reserve force, Bruno records that Otto sent back a messenger stating that this would not be possible. Instead, Otto urged Rudolf to fight as best he could, promising only that his men would be redeployed from the river when this was possible.83 Although Bruno does not say so explicitly, the implication of this exchange of messages between Rudolf and Otto was that elements of Henry’s army were still on the opposite bank of the Eichenbach and that a precipitous withdrawal of the Saxons from this front would allow the royal troops the opportunity to cross and join the battle. Bruno’s discussion here also illustrates excellent battlefield communications enjoyed by the rebel commanders in this battle.

Command and control The sound tactical decisions made by Rudolf at the battle of Flarchheim illustrate the value of a defensive posture on the battlefield, particularly when holding a strong position.84 This sound decision-making by the military commanders 80 81 82 83 84

Ibid. Ibid. Ibid. Ibid. The value of assuming the tactical defensive on the battlefield, even when taking the strategic offensive in a campaign, is discussed in considerable detail by Clifford J. Rogers, “The Offensive/Defensive in Medieval Strategy,” Von Crécy bis Mohács. Kriegswesen im späten Mittelalter (1346–1526) XII Congreß der internationalen Kommission für Militärgeschichte (International Commission of Military History, Vienna, 1997), 158–71; and idem, “Edward III and the Dialectics

186

B R U N O O F M E R S E B U R G ’ S S A X O N WA R

and their ability to maintain control over their troops on the battlefield can be contrasted with Bruno’s negative comments about poor decisions that were made by a number of military commanders, chiefly on the rebel side. A concomitant matter treated by Bruno in some detail is the connection between loss of command and control over the troops and subsequent battlefield defeat. In recounting the battle of the Mellrichstadt, for example, Bruno makes clear the disaster that can result from the failure to maintain command and control over the troops. In the prelude to the battle, the rebel army came upon Henry IV’s troops already established on the field. However, rather than waiting to deploy in an effective battle line, Bruno indicates that some of the Saxon troops charged into battle willy nilly and without a plan of action. The result was disastrous, as the individual units in this uncoordinated assault were destroyed piecemeal or fled from the field, exposing the remaining rebel troops to an even greater danger of being enveloped and destroyed by the royal army.85 Bruno places the blame for the failure in this battle on several magnates, including both of his patrons Archbishop Werner of Magdeburg and Bishop Werner of Meißen, of whom Bruno says: “From our side, the first to flee were those who never should have come to battle. These were the two bishops, both called Werner, who shared the same name but not, so to speak, the same fate.”86 Bruno also draws attention to a magnate named William, the count of Kamburg, discussed earlier, whom he accuses of advancing “rashly.”87 In the course of the rout of the rebel army, numerous other magnates shared Count William’s fate of being captured, including the archbishop of Mainz and the bishop of Worms.88 The most detailed discussion by Bruno concerning the issue of maintaining command and control on the battlefield comes in the context of his description of the battle of the Elster river. By contrast with the failures of command and control among the rebel forces in the battle of Mellrichstadt, Bruno observes that Otto of Northeim’s ability to maintain discipline and control at the battle the Elster river, discussed previously, secured the victory for Henry IV’s opponents.89 In drawing attention to Otto’s prudence in military matters, Bruno also illustrated the poor decision-making by the commanders of Henry IV’s army, who helped transform a draw on the battlefield into an overwhelming enemy victory. Following the failure of his invasion of Saxony in the winter of 1080, Henry IV withdrew southeastwards toward Erfurt and then toward the White Elster river, with the anti-king Rudolf’s army in pursuit.90 Bruno records that once Rudolf

85 86 87 88 89 90

of Strategy, 1327–1360,” Transactions of the Royal Historical Society 6.4 (1994), 265–83, both of which are reprinted with the same pagination in Idem, Essays on Medieval History: Strategy, Military Revolutions and the Hundred Years War (Ashgate, Burlington, VT, 2010). Bruno, Saxonicum Bellum, ch. 96. Ibid. Ibid. Bruno, Saxonicum Bellum, ch. 96. Bruno, Saxonicum Bellum, ch. 122. Ibid.

187

N A R R AT I V E W O R K S

arrived and discovered that Henry’s forces were deployed with their backs to the Elster on the east and the Grone swamp to their north, the erstwhile Swabian duke divided his army into two elements.91 Rudolf retained the majority of the mounted forces, who comprised his best troops, under his direct command, and sought a western route around the swamp in order to engage with the royal army from the west.92 The foot soldiers were placed under the command of Otto of Northeim.93 The latter then led his foot soldiers on an arduous march through the swamp in order to attack the royal army from the north.94 In discussing the initial stages of the battle, Bruno admits that Henry’s forces had the upper hand. As the royalist troops met Rudolf’s men at the western end of the swamp, there was a brutal battle in which Henry’s men emerged victorious. Bruno again attempts to pillory the king by claiming that Henry fled at the first sign of combat. But he then added: But his army resisted ours with such great bravery that some of our men turned in flight. So the false report came to the enemy’s camp, and it was announced out of the mouth of a liar that the Saxons had been defeated.95 But in this moment of triumph for Henry IV’s mounted troops, Otto of Northeim finally cleared the swamp and attacked the unprepared royal camp from the north. In their premature celebration of victory, Henry’s men did not take care to guard their camp or post sentries to provide warning. Consequently, the charge of the Saxon foot soldiers caught them completely unprepared and led to a rout of the royalist troops. Duke Otto’s men kept a close pursuit of the king’s men through their camp and then to the banks of the Elster river, driving the royalists into the deep waters where, according to Bruno, many of them drowned. Commenting on this phase of the battle, Bruno observed that more men died in the river than died in combat.96 The failure of command and control among the royal forces, and particularly the failure to take precautions against the chance of attack through the swamp, led to disaster. Bruno illuminates this failure of battlefield communications among the royalist commanders even more clearly in his discussion of the decisions made by Otto of Northeim after the initial rout of the royal troops from their camp. Rather than allowing the Saxons to loot the king’s camp, Otto ordered them to maintain their formation. As Bruno emphasized:

91 This was a strong defensive position with the major drawback that there was no place for the royal army to retreat if they suffered defeat in battle. In fact, Bruno goes to great lengths to try to explain the decision by Henry to allow himself to be cornered by the Saxon army in this location. 92 Bruno, Saxonicum Bellum, ch. 122. 93 Ibid. 94 Ibid. 95 Ibid. 96 Ibid.

188

B R U N O O F M E R S E B U R G ’ S S A X O N WA R

But Duke Otto, who was prudent in military matters, feared there might still be enemies remaining to their rear. He warned them keep their hands from the booty for now until they were certain there was no enemy hiding behind them. They could then safely seize the enemy camp.97 Otto’s prudence in this matter was revealed in the next phase of the battle. The royal forces under the command of Henry of Laach, the count palatine, had defeated Rudolf of Rheinfelden’s men at the western edge of the Grona swamp. Confident in his victory, Henry of Laach did not post any scouts or otherwise safeguard his troops while they celebrated. As a consequence, Otto of Northeim was able to advance with his foot soldiers undetected and launch a surprise assault from the rear. Bruno implies that Otto’s men were outnumbered in this battle, but claims that with God’s aid, they were able to drive the royal troops from the field.98 Only then, after having defeated both elements of King Henry’s army, did Otto of Northeim permit his men to return to the royal camp and loot it. Bruno records Otto as telling his men: “Seek out their camp in safety. Now that you are safe, take whatever you find there. Whatever belonged to the enemy, you can call your own because it was gained by your strength!”99

Logistics Bruno regularly draws the attention of his audience to the problem of logistics. For example, Bruno observes that even in the aftermath of a substantial victory over the Saxons in the battle of the Unstrut on 9 June 1075, King Henry could not effectively exploit the opportunity to impose a complete military settlement on his enemies because of a lack of supplies. According to Bruno, the king did pursue the Saxons after his victory in the field, looting and burning along their line of march in a manner consistent with, or indeed, even worse than the heathen enemies of the Saxons to the east, that is the pagan Slavs.100 The ultimate goal of Henry’s march was the recapture of the royal palace and fortress at Goslar, which was located approximately 120 kilometers due north of the battlefield at contemporary Bad Langensalza.101 Henry arrived at Goslar with only a small portion of his army, but was received by a number of Saxon bishops with what Bruno describes as a “glorious triumph.”102 Henry was able to divide his forces in this manner because the Saxon magnates, whom Henry had defeated at the Unstrut, had scattered and could not offer any effective resistance to the royal army.

97 98 99 100 101 102

Ibid. Ibid. Ibid. Bruno, Saxonicum Bellum, ch. 47. Bruno, Saxonicum Bellum, ch. 53. Ibid.

189

N A R R AT I V E W O R K S

Nevertheless, despite the fact that the Saxons had been thoroughly defeated, Henry was not able, as Bruno put it, to do “the only thing that he wanted to do, namely immediately to subject them all to servitude.”103 Bruno goes on to explain: He was not able to capture all of the leading men because they were dispersed in several places, nor could he keep his army in this land for very long because hunger, for once, was useful to us this year, and in the month of July the crops were not yet ripe.104 It is evident from Bruno’s account of the month following the battle on the Unstrut that Henry had not refrained from looting, burning, and ravishing the countryside, as well as towns and even monasteries. Nevertheless, the royal army could not simply live off the land, and the king had not prepared adequate supplies or transport for a lengthy stay in the Saxon region. Having withdrawn from the Saxon duchy in July 1075, Henry IV worked diligently to raise yet another army to invade the region again in October of that year. Bruno again commented that Henry’s decision regarding the timing of this campaign was due to concerns about supply. The king, according to Bruno, had observed the crops growing in vast abundance in the fields while marching northwards toward Goslar.105 Now that the crops had been harvested, Henry IV planned to use them to feed his own troops, or burn them to deny them to the Saxons. If the king had been successful in this plan, Bruno contends that “he would either devour the entire people with his sword, or subject them as humble petitioners to perpetual servitude.”106 To avoid this fate, the Saxons mobilized their forces and advanced beyond the frontiers of Saxony in order to meet and deter the royal army to their south in Thuringia. However, recognizing the weakness of their position, the Saxons surrendered without a battle and agreed to have their leaders imprisoned by the king.107 In addition to the vital importance of logistics for maintaining an army in the field, Bruno draws attention to the difficulties that arose when garrisons that were defending fortifications lacked adequate supplies. In his discussion of Henry IV’s decision to seize the fortress of Lüneburg in 1073, Bruno observed that in addition to doing something that was illegal, i.e. seizing the private property of a free man without justification, the king acted stupidly.108 Bruno claims that “because the garrison had been installed without any preparation” Hermann, the brother of the Saxon duke, “simply waited until the king departed from the region, and quickly

103 104 105 106 107 108

Ibid. Ibid. Bruno, Saxonicum Bellum, ch. 54. Ibid. Ibid. Bruno, Saxonicum Bellum, ch. 21.

190

B R U N O O F M E R S E B U R G ’ S S A X O N WA R

thereafter surrounded the fortress with all of his own forces.”109 Bruno goes on to ask rhetorically, “What were the men inside to do?” He adds: The fortress was very strong, and could not be conquered by anything other than hunger. But aside from a few loaves of bread that the monks had left behind when they departed, there was nothing to eat. Hunger demanded that they exit the fortress, but the strength of the opposing swords did not permit them to leave. Ultimately, Henry’s garrison was forced to surrender, and to gain their release, the king had to hand over Magnus Billung, Hermann’s nephew and the heir to the Saxon ducal seat, who was being held as a royal prisoner.110

Morale Bruno returns repeatedly throughout his narrative to the idea that the Saxons were engaged in a defensive war and consequently enjoyed high morale because they were confident in the justice of their cause. In discussing Henry IV’s campaign of October 1075, noted earlier, Bruno contrasts the war aims of the king and of the Saxons. The former, Bruno claims, intended to slaughter the Saxons or reduce them to slavery. By contrast, the Saxons came to fight for their liberty. Indeed, Bruno claims that “they would either retain their liberty with God’s help, or they would lose it along with their lives.”111 Similarly, in his treatment of Henry IV’s autumn campaign in Saxony in 1080, Bruno again emphasized that the Saxons saw their defense of their homeland as making them worthy of divine aid, claiming: “The Saxons took up their positions there after establishing their encampment so that they could defend their frontiers, with God’s aid, against this invasion by the enemy.”112 The seeking of proper justification for war was an element in a broad array of religious rites and ceremonies in the Western tradition that were intended to gain divine aid for one’s cause.113 This was crucial for morale because soldiers who believed that God was on their side could be confident not only about the battle in which they were fighting but also that their souls would be safeguarded if they fell on the battlefield. It is in this context that Bruno emphasizes the religious ceremonies of the Saxon armies both before and after they went into battle. In his discussion of the encounter on the Elster river, for example, Bruno notes that before Rudolf of Rheinfelden led his men around the Grona swamp to engage Henry IV’s troops, the bishops and priests in the Saxon army sang Psalm 82 with

109 110 111 112 113

Ibid. Ibid. Bruno, Saxonicum Bellum, ch. 54. Bruno, Saxonicum Bellum, ch. 121. For the treatment of this issue with regard warfare in the medieval period, see David S. Bachrach, Religion and the Conduct of War c. 300–c. 1215 (Boydell, Woodbridge, 2003).

191

N A R R AT I V E W O R K S

great devotion.114 This biblical passage was particularly apropos because it calls upon God to aid Israel against her enemies in battle. Bruno also observes that the Saxons, and their royal opponents as well, were diligent in their efforts to praise God for victories on the battlefield. Following the battle at Mellrichstadt, for example, Count Palatine Frederick of Sommerschenburg held the field with his troops but did not allow them to plunder the dead, at least not immediately. Instead, Frederick “collected all of his men from the various parts of the battlefield and spent that night in great joy and especially in praise of God.”115 The next day, after taking considerable quantities of booty, the Saxons marched home “joyfully singing hymns to God.”116 In a similar manner, following the initial defeat of Rudolf’s men at the battle of the Elster, the royal troops, who remained in their encampment, sang the Te Deum laudamus, a hymn in praise of the Lord, led by their bishops and priests.117 Their companions, under the command of Count Henry of Laach, who had defeated Rudolf’s men, loudly sang the kyrie eleyson.118 In both cases, however, the songs of praise to God were premature, as discussed previously. Belief in God’s power to determine who would win a battle certainly plays an important role in Bruno’s narrative. However, he also drew attention to the size of the armies that were involved in a conflict, making clear the important positive effects on morale that came from serving in a large army and the concomitant negative effects on morale among the soldiers who realized that they were outnumbered. In describing Henry’s first attempt to invade the Saxon duchy in February 1074, Bruno emphasizes that although the king had a large army, the Saxon army was twice its size.119 The royal troops, according to Bruno, were already hesitant about fighting because they were not sure of the justice of their cause.120 This sentiment obviously is the opposite of how the Saxons, as presented by Bruno, viewed their defense of their homeland. Nevertheless, believing that their army was sufficiently larger, better equipped, and better trained than the Saxons to win a victory, the royalists were prepared to go into battle. But once the scouts returned and reported that the Saxons not only included well-trained troops in their ranks but also that they thoroughly outnumbered Henry’s army, the royalists “had no desire at all to fight because along with the lack of a cause, they also lacked the numbers to go safely into battle against such a multitude.”121 Bruno explains the connection between serving in a large army and having high morale even more explicitly in his depiction of the Saxon response to King Henry’s invasion in October 1075. In this case, Bruno states: “Saxons, who had become prudent because of 114 115 116 117 118 119 120 121

Bruno, Saxonicum Bellum, ch. 122. Bruno, Saxonicum Bellum, ch. 101. Ibid. Bruno, Saxonicum Bellum, ch. 122. Ibid. Bruno, Saxonicum Bellum, ch. 31. Ibid. Ibid.

192

B R U N O O F M E R S E B U R G ’ S S A X O N WA R

the great danger they had experienced, came with a large army so that they would not turn their backs in flight, but rather would fight bravely for their liberty.”122 Moreover, Bruno also indicates that his audience would understand the concept that modern theorists denote as the doctrine of overwhelming force. This theory holds that one method of avoiding combat is by demonstrating to an opponent that victory is impossible, or at least very unlikely, because of the scale of forces that have been brought to bear against them. In discussing the Saxon response to Henry IV’s invasion in October 1078, which followed upon the battle at Mellrichstadt that summer, Bruno declared that his people gathered in numbers never seen before, claiming that they raised a force of 60,000 troops to defend their homeland.123 While this figure almost certainly is a significant exaggeration, the crucial point is that Henry’s troops retreated before ever testing the Saxons in battle because of their fear of the exceptionally large army that they faced.124 Bruno’s consistent depiction of the Saxons mobilizing very large armies for the defense of their homeland likely is due, at least in part, to his effort to emphasize the unity of his people in the face of Henry IV’s putative tyranny. However, it is noteworthy that by depicting the forces on the “home side” as particularly numerous, and even as outnumbering the enemy, Bruno was at odds with the rhetorical tradition throughout the ancient and medieval period that sought to exaggerate the numerical strength of the enemy and to minimize the strength of the side which the author sought to depict positively.125 The purpose of this rhetorical device was to inflate the meaning of victories of the “few” over the “many” and to explain away defeats. In addition, pious writers could use the topos of the few defeating the many to invoke the image of divine intervention on the side of the “just.” That Bruno, who was writing for an audience that was intimately familiar with the details of the campaigns of the Saxon Wars, chose not to indulge in this rhetorical trope suggests that Saxons really did mobilize very large armies and that the size of these armies did, in fact, serve to raise the morale of the Saxon troops. Given his intention of justifying the Saxon cause and impugning Henry IV, it is hardly surprising to see Bruno emphasize factors such as the Saxons’defense of their homeland, their unity in the face of oppression, and their efforts to secure divine support in their military endeavors as contributing to their high morale. By contrast, Bruno seeks to downplay the importance of material concerns for the Saxon leadership and to highlight the desire for booty among the royalists as a testament to their inherent moral corruption. Nevertheless, consistent with his parti pris against men of lower social and economic status, Bruno does not attempt to hide interest in booty among the great mass of rebel troops and the role that the acquisition of plunder

122 123 124 125

Bruno, Saxonicum Bellum, ch. 54. Bruno, Saxonicum Bellum, ch. 103. Ibid. See the discussion of this topic by Bernard S. Bachrach, “Early Medieval Military Demography: Some Observations on the Methods of Hans Delbrück,” in The Circle of War, ed. Donald Kagay and L. J. Andrew Villalon (Boydell, Woodbridge, UK, 1999), 3–20.

193

N A R R AT I V E W O R K S

played in raising their morale. Instead, Bruno uses this base desire as a further means of distinguishing both the moral superiority of the noble opponents of the king and the moral degradation of the magnates on Henry IV’s side, who are depicted as having more in common with Saxon peasants than with men of their own rank. For example, according to Bruno, following the battle of Mellrichstadt, Henry IV held an assembly of his magnates in Regensburg in which the king sought to persuade these men to launch an invasion of the Saxon duchy that finally would crush their rebellion.126 Henry is presented as telling his magnates that the rebels were all but defeated following Mellrichstadt and that those who followed the king to war would be enriched with the rich lands of all those Saxons who had been killed. Bruno depicts this promise as having the desired effect, insisting: “Deceived by this vain hope, and as if they already did possess Saxony, they grew puffed up in spirit.”127 Bruno then goes on to explain that going to war in the hope of material gain caused substantial problems. Because Henry had insisted that the rebels were all but defeated, his magnates had no fear of going to war and so did not mobilize an army that was commensurate with the task that they were about to undertake. Rather, they sought to limit their own numbers so as to avoid both the cost of mobilizing a new large force and the necessity of sharing out the riches of the Saxons among too many hands. As Bruno put it: But they did not wish their army to be very large, in fear that each one of them might receive less of the region the more of them there were to divide it. Because it is natural that the size of the individual portions would be smaller, the greater the number of portions there were.128 As a consequence, when Henry’s troops arrived in Thuringia and found a massive rebel army waiting there, the royalist troops withdrew in fear.129 The lesson of this episode, in Bruno’s account, was that high morale built on the base desire for material gain was ephemeral, particularly when contrasted with determination that derives from defending one’s homeland. Bruno implicitly contrasts the greed of the royalist magnates with the pure motives of the Saxon leaders in his discussion of the aftermath of the defeat of Henry IV’s army at the battle of the Elster river. As they withdrew in confusion from the battlefield, many of Henry’s men were killed by members of the local Saxon levies. Bruno recounts the deaths of soldiers at the hands of peasants armed with simple axes and even clubs.130 He asserts that when the peasants captured rather than killed Henry’s men, the latter were starved and held for ransom. By contrast, according to Bruno, “if any of the enemy were 126 127 128 129 130

Bruno, Saxonicum Bellum, ch. 103. Ibid. Ibid. Ibid. Bruno, Saxonicum Bellum, ch. 123.

194

B R U N O O F M E R S E B U R G ’ S S A X O N WA R

brought to one of our upright men, they were healed if they were wounded, and were sent back to their fatherland free of charge, well outfitted with clothing and arms.”131 In refusing to demand ransoms, or even to take the equipment of their prisoners, the Saxon magnates, in Bruno’s presentation, demonstrated both their personal honor and the justness of their cause. By contrast, the Saxon peasants demonstrated their moral inferiority, a characteristic that they shared not only with the baser men serving in the king’s army but also Henry IV’s aristocratic supporters.

The role of fortresses in warfare As the previous discussion indicates, Bruno worked diligently to provide considerable information to his audience about battles in the field. Nevertheless, Bruno makes clear throughout his text that sieges played a central role in warfare between the king and the Saxons, particularly for the purpose of territorial conquest and control. The account in Bruno’s Saxon War consequently is consistent with the treatment of warfare in both contemporary narrative sources and those written throughout the medieval millennium.132 In discussing, for example, Henry IV’s decision to seize control over the fortress at Lüneburg, noted earlier, Bruno claims that the king’s “unquenchable desire for the fortress” was based on his view that no one in the region would be able to resist his rule once he held this stronghold.133 Bruno returns to this idea by presenting what purports to be a speech by Otto of Northeim to the assembled Saxon host at Hoetensleben on the eve of the great rebellion in 1073. Otto is presented as appealing to the fears of the Saxons that the royal fortresses that were built throughout their lands, if left in the king’s hands, would serve as bases to terrorize and control the population.134 The catalog of infamies to which the Saxons had been subjected included the imposition of labor duties on free men, the theft of their property, and even the misuse of their daughters and wives.135 However, Otto is presented as saying that these abuses were only a taste of what would come once Henry completed construction of his strongholds and garrisoned them with troops. At this point, according to the speech, the king would be able to seize the Saxon people wholesale and reduce them all to slavery.136 131 Ibid. 132 In this context, see Jim Bradbury, The Medieval Siege (Boydell, Woodbridge, UK, 1992); and Peter Purton, A History of the Early Medieval Siege, c. 450–1200 (Boydell, Woodbridge, 2010), and the enormous literature cited there. Cf. Guy Halsall, Warfare and Society in the Barbarian West 450–900 (Routledge, London, 2003), 215–27, whose claims are thoroughly outside the mainstream of scholarship dealing with medieval warfare because of his efforts to downplay the importance of sieges. 133 Bruno, Saxonicum Bellum, ch. 21. 134 Bruno, Saxonicum Bellum, ch. 25. 135 Ibid. 136 Ibid.

195

N A R R AT I V E W O R K S

In addition to purporting to quote speeches in which fortresses and garrisons are presented as a crucial tool in the king’s arsenal to impose direct control over the territory of Saxony, Bruno quotes a letter written by his patron, Archbishop Werner of Magdeburg, to Archbishop Siegfried of Mainz, which makes precisely the same point.137 In the course of discussing the king’s sins against the Saxon people, Archbishop Werner reminds his fellow prelate that once Henry IV came of age, he began the process of constructing large numbers of fortresses in the Saxon region. When these centers were completed and provided with large garrisons, according to Werner, the king then began to oppress the people living in those districts. The archbishop of Magdeburg then provided a list of the atrocities committed by royal garrison troops.138 Because maintaining well-garrisoned strongholds was so important for territorial control, these fortifications necessarily were a primary focus of military action by the Saxons, who sought to free themselves from royal domination. In the case of Lüneburg, noted previously, Bruno explains that Hermann, the brother of the Saxon duke, waited until Henry IV had departed from the region with his army and then immediately mobilized all of his own military forces to undertake a siege of the stronghold.139 Hermann realized, according to Bruno, that although the royal garrison was strong enough to defend the walls of Lüneburg, the troops did not have sufficient supplies to withstand even a brief siege. As a consequence, the king’s men were forced to capitulate very rapidly.140 By contrast with the relatively easy and bloodless military operation that culminated in the capture of Lüneburg in 1071, Bruno makes clear that the efforts of the Saxons to besiege the royal fortress of Harzburg in 1073 were as daunting as the danger posed by this stronghold to their rebellion.141 After recognizing that an effort to storm the walls of Harzburg would lead to unacceptably high numbers of casualties, even if assaulted by an enormous army, the Saxons decided to construct their own fortress as a base of operations.142 According to Bruno, the Saxons chose a site higher up on the side of the mountain, so that they would be able to look down inside the walls of Harzburg and also deploy stone-throwing engines against this stronghold. Because of their relative positions, Henry IV’s men had great difficulty in reaching the newly constructed Saxon fortress with their own artillery.143 The Saxons then established a schedule by which the men serving as a garrison in their siege fortress would be relieved at regular intervals by fresh troops. In addition to providing a focus for military operations against Henry IV within the Saxon duchy, the rebels against Henry’s rule identified major fortified strongholds as campaign objectives when they took the war into lands that were loyal to

137 138 139 140 141 142 143

Bruno, Saxonicum Bellum, ch. 42. Ibid. Bruno, Saxonicum Bellum, ch. 21. Ibid. Bruno, Saxonicum Bellum, ch. 27–29. Bruno, Saxonicum Bellum, ch. 29. Ibid.

196

B R U N O O F M E R S E B U R G ’ S S A X O N WA R

the king. For example, after discussing the election and consecration of Rudolf of Swabia as king in the spring of 1077, Bruno emphasizes that Rudolf first went on a rapid tour of the lands that were loyal to him in Swabia and Thuringia and then returned to the Saxon region in June to plan an offensive campaign against Henry IV.144 Rudolf then mobilized what Bruno describes as a large army and marched to the royalist city of Würzburg in August 1077. His intention was to capture this major fortress and thereby severely weaken Henry IV’s power in the region of Franconia that stretched along the Main river.145

Conclusion Throughout his narrative, Bruno demonstrates a keen interest and eye for detail regarding an array of topics concerning the conduct of military operations and the overall organization of war in the German kingdom during the 1070s and 1080s. Insofar as Bruno can be understood as a reliable witness for the Saxon Wars, his account illuminates the complex military organization of both the royal forces of Henry IV and those of his opponents. Both sides relied extensively on militia troops not only for ancillary purposes such as making camp and collecting supplies but for military operations in both the field and in sieges. Bruno also comments extensively on the preparations for campaigns and battles made by both sides in the war, including the efforts to obtain military intelligence, to organize their logistical capacity, to develop plans of battle, and to devise ruses des guerre. Bruno’s commentary on the actual campaigns undertaken by Henry IV as well as his Saxon and southern German enemies indicates the centrality of fortifications and consequently sieges to gaining and maintaining long-term control over territory. Indeed, despite the high concentration, by medieval standards, of battles during the Saxon Wars, defeat or victory in these contests did not prove to be decisive for the overall conflict. This investigation of Bruno’s history is a first step in bringing greater light to the heretofore obscure nature of warfare in the German kingdom during the period of the Salian dynasty. However, some tentative conclusions are possible based upon what Bruno has written. First, it appears that the conduct of war in Germany was quite similar to that in the West, where the armies of Henry IV’s contemporaries in Anjou, Normandy, and England, as well as the Iberian kingdoms of Castile and Aragon, employed large numbers of militia troops alongside a core of professional fighting men. Similarly, the centrality of sieges to warfare in Germany has close parallels to the military campaigns undertaken, for example, by William the Conqueror. As more research is done with respect to Salian Germany, undoubtedly other similarities to warfare in the West will emerge.

144 Bruno, Saxonicum Bellum, ch. 93. 145 Bruno, Saxonicum Bellum, ch. 94.

197

9 F E U D A L I S M, R O M A N T I C I S M, AND SOURCE CRITICISM Writing the military history of Salian Germany1

The military history of the German kingdom in the eleventh century remains to be written. The most recent monographic treatment that considers the Salian kings at war, Bruno Scherff’s 1985 dissertation Studien zum Heer der Ottonen und der ersten Salier (919–1056), ends with the death of King Henry III (1039–1056) and is notable for the topics that it explicitly avoids, including military equipment, logistics, numbers, military levies, fortifications, and, above all, sieges.2 There are no monographic studies of the conquest of the Burgundian kingdom by King Conrad II of Germany (1024–1039) in his two-year campaign of 1033–1034.3 Since the end of the nineteenth century, scholars have assiduously avoided investigating the wars of Conrad II, Henry III, and Henry IV (1056–1106) in Poland and Hungary.4 Even the civil wars within Germany during the final quarter of the eleventh century and the first decade of the twelfth, which have received considerable attention from a political perspective, have inspired only limited commentary regarding military matters.5 At an even more basic level, the enormous volume of 1 This chapter was originally delivered as the De re militari Plenary Lecture at the medieval studies conference at Western Michigan University in May 2014. I thank all of the participants for their helpful questions and insights. 2 Bruno Scherff, Studien zum Heer der Ottonen und der ersten Salier (919–1056) (Doctoral Dissertation, University of Bonn, 1985), 1–5. 3 The very dated dissertation by Otto Blümcke, Burgund unter Rudolf III. und der Heimfall der burgundischen Krone an Kaiser Konrad III (Doctoral Dissertation, University of Greifswald, 1869) focuses on the transition in power between the last Burgundian ruler and the Salian Conrad II, but does not do so from a military perspective. 4 Julius Reinhard Dieterich, Die Polenkriege Konrads II. und der Friede von Merseburg (Giessen, 1895); and Emil Friedrich Kuemmel, Die zwei letzten Heereszüge Kaiser Heinrichs III. nach Ungarn (1051 und 1052) mit Rüchsichtsnahme auf die bairisch-kärtnische Empörung (Strassnitz, 1879). A general treatment of German-Hungarian relations can be found in the studies by Friedrich Schuster, Ungarns Beziehungen zu Deutschland von 1056–1108 (Hermannstadt, 1899); and Otto Rademacher, Ungarn und das deutsche Reich unter Heinrich IV (Merseburg, 1885). 5 Two notable exceptions to the otherwise spare historiographical tradition are the articles by Leopold Auer, “Die Schlacht bei Mailberg am 12. Mai 1082,” Militärhistorische Schriftenreihe 31

198

FEUDALISM, ROMANTICISM, AND SOURCE CRITICISM

source materials pertaining to warfare in the Salian century, including narrative texts, letters, charters, and archaeological evidence, has not received the kind of detailed critical analysis that has become the norm in Anglo-Saxon, Carolingian, and increasingly Ottonian military history.6 The lack of scholarly attention to the military history of the German kingdom generally, and to the Salian period in particular, is due to several factors. First, until very recently non-German-speaking scholars, including military historians, generally avoided the history of Germany altogether.7 Second, German-speaking historians largely have avoided topics in military history, especially since the end of the Second World War.8 Finally, the one consistent exception to this aversion to military matters, namely the focus on social and economic questions regarding the relationship among military service, nobility, and “feudalism,” actually has served to hinder a thorough investigation of the nature and conduct of war and the concomitant examination of its institutional, material, and human elements. Even before the 1940s, many topics that currently are understood as the province of military history largely were ignored in the German scholarly tradition. These include but are not limited to logistics, siege warfare, military technology, military organization, recruitment, military institutions such as local and expeditionary (1975), 1–31; and John Gillingham, “An Age of Expansion, c. 1020–1204,” in Medieval Warfare: A History, ed. Maurice Keen (Oxford, 1999), 59–88. Several of the major battles of the Saxon wars also were outlined by Hans Delbrück, Geschichte der Kriegskunst in Rahmen der politischen Geschichte, vol. 3, 3rd edition (Berlin, 1923, originally published 1907), trans. by Walter J. Renfroe as History of the Art of War, Volume III: Medieval Warfare (University of Nebraska Press, Lincoln, 1982, repr. 1990), 131–45. 6 David S. Bachrach, “Milites and Warfare in Pre-Crusade Germany,” War in History 23.3 (2015), 298–343, is an attempt to redress this imbalance, particularly in regard to narrative texts. 7 Fortunately, there has been a recent blossoming of English-language scholarship on the German kingdom. An early precursor of this tradition was John W. Bernhardt, Itinerant Kingship and Royal Monasteries in Early Medieval Germany c. 936–1075 (Cambridge, 1993). Now also see, for example, Jonathan Lyon, Princely Brothers and Sisters: The Sibling Bond in German Politics, 1100–1250 (Cornell Ithaca, NY, 2013); and John Eldevik, Episcopal Power and Ecclesiastical Reform in the German Empire: Tithes, Lordship, and Community, 950–1150 (Cambridge, 2012). From a military perspective, see Charles R. Bowlus, The Battle of Lechfeld and Its Aftermath, August 955: The End of the Age of Migrations in the West (Aldershot, 2006). 8 Concerning the lack of scholarship on German military history in the period after the Second World War, see Hans-Hennig Kortüm, “Der Krieg im Mittelalter als Gegenstand der historischen Kultur-Wissenschaften: Einer Annäherung,” in Krieg im Mittelalter, ed. idem (Berlin, 2001), 13–43; and Scherff, Studien zum Heer der Ottonen und der ersten Salier, 2–4. One notable exception has been the Austrian scholar Leopold Auer whose numerous publications on military matters include “Der Kriegsdienst des Klerus unter den sächsischen Kaisern,” Mitteilungen des Instituts für Österreichische Geschichtsforschung part 1, 79 (1971), 316–407; and part 2 in ibid., 80 (1972), 48–70; “Zum Kriegswesen unter den früheren Babenbergern,” Jahrbuch für Landeskunde von Niederösterreich 42 (1976), 9–25; “Formen des Krieges im abendländischen Mittelalter,” in Formen des Krieges: Vom Mittelalter zum “Low-Intensity-Conflict”, ed. Manfried Rauchensteiner and Erwin A. Schmid (Graz, 1991), 17–43; and “Mittelalterliches Kriegswesen im Zeichen des Rittertums,” in Krieg im mittelalterlichen Abendland, ed. Andreas Obenau and Christoph Kaindel (Vienna, 2010), 65–79.

199

N A R R AT I V E W O R K S

levies, pay, morale, education, and military training. Instead, scholars devoted their energies to explaining the supposed transition from a putative “nation in arms” among the early Germanic peoples to small retinues of “feudal knights” who ostensibly dominated both warfare and politics through their monopolistic mastery of equestrian combat. A brief overview of this historiographical tradition will help to illuminate the current (limited) state of understanding of warfare in Germany in the pre-Crusade period.

From levies to feudal knights In his classic study first published in 1887, Heinrich Brunner argued for Charles Martel’s rapid transformation of Frankish military organization from large armies of men on foot to much smaller armies of mounted warriors.9 Brunner asserted that Charles Martel, having defeated an army of mounted Muslims with his infantry phalanx, now sought to develop a cavalry army that could face the Muslims in battle. The implicit paradox in this argument was never fully explained by Brunner. In order to achieve this new army, Charles is depicted as having “borrowed” large numbers of estates from the church and granting these as beneficia to his loyal supporters so that they could purchase and maintain horses that were suitable for mounted combat. Brunner averred that because these mounted troops were so much more militarily effective than the foot soldiers whom they replaced, Charles Martel simply stopped mobilizing the latter and relied on his new cavalry army. The ultimate consequence of this change in military organization of the Frankish kingdom was that only those men who could equip themselves and their dependents as mounted warriors remained militarily viable. The bulk of the population, which now was excluded from military service, declined in social, economic, and political status. The warrior aristocracy and their beneficed retainers thus took their place alongside the king as the major figures in the new “feudal” order. Military historians in the late nineteenth and early twentieth century generally did not accept Brunner’s model of a rapid transition of Frankish military organization towards an army that was composed largely of mounted warriors – not least because Carolingian military legislation from the ninth century clearly indicated that foot soldiers continued to be mobilized for combat operations.10 However, 9 The information in the following paragraph is drawn from Heinrich Brunner, “Der Reiterdienst und die Anfänge des Lehnwesens,” Zeitschrift der Savigny-Stiftung für Rechtsgeschichte, Germanistische Abtheilung 8 (1887), 1–38, reprinted in Brunner, Forschungen zur Geschichte des deutschen und französischen Rechts (Stuttgart, 1894), 39–74. 10 In this context, Delbrück, History of the Art of War, 3: 16–18, took the view that the transition was very rapid, while numerous other scholars argued for a much more gradual approach. In his review essay of the third volume of Delbrück’s massive study on the history of war, in which Delbrück argued in favor of Brunner’s rapid transition to an army of mounted fighting men, W. Erben, “Zur Geschichte des karolingischen Kriegswesens,” Historische Zeitschrift 101 (1908), 321–36, drew attention to a series of errors that Delbrück had made, and specifically noted Delbrück’s failure to account for the capitulary evidence that discussed the continued mobilization of foot soldiers

200

FEUDALISM, ROMANTICISM, AND SOURCE CRITICISM

Brunner’s argument that mounted military service provides the key to understanding both the establishment of a warrior nobility and the development of feudalism did win widespread approval from scholars in the early twentieth century. Thus, for example, in his study of the constitution of the state and the army in early medieval Europe, published in 1906, Otto Hintze started with the premise that there was a transition in Frankish military organization from large armies of foot soldiers to smaller armies of mounted troops. Hintze argued that this transformation was accompanied by a professionalization of the army, with the concomitant requirement that mounted fighting men have sufficient wealth not only to equip themselves with horses and arms but also the leisure time that was necessary to train to fight on horseback. The difference in wealth combined with the difference in status as warriors resulted, according to Hintze, in the development of a knightly class (Ritterstand). These knights were socially as well as economically superior to the bulk of the population that did not have sufficient wealth to participate in mounted combat.11 The distinction between those who fight and those who farm was given an ostensible legal foundation by Hans Fehr in his 1914 study on the rights of farmers to bear arms.12 Fehr compared the discussion of fighting men in Carolingian capitularies of the ninth century with the texts of royal commands for territorial peace (Landfrieden) issued by Salian and Staufen kings from the eleventh through the thirteenth century. Fehr observed that although the latter documents clearly permit all men to engage in self-defense, the Landfrieden do not make provisions for Bauern to participate in expeditionary campaigns. He concluded on this basis that farmers no longer had the right to bear arms (Waffenrecht) and consequently suffered a social decline as a class.13 It is noteworthy that Fehr did not utilize any narrative sources in his study. Nor did he find an adequate solution to the problem that as late as 1179 the Landfrieden issued by Frederick Barbarossa banned farmers from carrying swords while in their home villages but did not ban them from carrying swords while they went on journeys.14

11

12 13 14

for campaign duty. Many other scholars called into question the rapid transition to a mounted army on the basis of references to foot soldiers in narrative sources. See, for example, Erich Sander, “Die Heeresorganisation Heinrichs I.,” Historisches Jahrbuch 59 (1939), 1–26; Hans von Mangoldt-Gaudlitz, Die Reiterei in den germanischen und fränkischen Heeren bis zum Ausgang der deutschen Karolinger (Berlin, 1922); Franz Beyerle, “Zur Wehrverfassung des Hochmittelalters,” in Festschrift Ernst Mayer zum 70. Geburtstag (Weimar, 1932); and Paul Schmitthenner, “Lehnskriegswesen und Söldnertum im abendländischen Imperium des Mittelalters,” Historische Zeitschrift 150 (1934), 228–67. Otto Hintze, Staatsverfassung und Heeresverfassung (Dresden, 1906); and reprinted in idem, Staat und Verfassung. Gesammelte Abhandlungen zur allgemeinen Verfassungsgeschichte, ed. Gerhard Oestreich, 3rd edition (Göttingen, 1970), 52–83. Hans Fehr, “Das Waffenrecht der Bauern im Mittelalter,” Zeitschrift der Savigny-Stiftung für Rechtsgeschichte. Germanistische Abteilung 35 (1914), 111–211. Ibid., 158–61. Ibid., 139.

201

N A R R AT I V E W O R K S

In the interwar period, many scholars adopted the theme that men of the lower social orders, variously identified as farmers or simply as free men (Freie), were demilitarized and that warfare came to be dominated by a knightly class of mounted warriors. Franz Beyerle, for example, argued: “From the later Carolingian period, the army increasingly was dominated by vassalic contingents.”15 In support of this argument, Beyerle drew attention to Regino of Prüm’s discussion of the defeat of the tenants of his monastery at the hands of the Vikings in the late ninth century, which Beyerle used as evidence of the poor fighting skills of local levies that resulted in their increasing exclusion from military affairs.16 In this context, Beyerle did not draw attention to Regino’s specific observation that the men in question were untrained, i.e. disciplina militari nudatum.17 Eugen von Frauenholz also depicted the ninth century as a period of transition between armies of foot soldiers and smaller armies of mounted warriors. Like Hintze, von Frauenholz initially depicted this development as bringing about a concomitant professionalization of the army. However, this professional force quickly developed, in the view of von Frauenholz, into a new hereditary social class (Geburtsstand) with rules governing not only military life but all aspects of newly developing knighthood (Ritterschaft).18 In the decades after the Second World War, scholars continued to adhere to the intertwined argument that there was a transition from armies of men on foot to armies composed predominantly of mounted warriors and that this transition played a crucial role in development of both a warrior aristocracy and feudalism. In his study of the participation of German ecclesiastical magnates in the wars of the Ottonian kings, published in 1971–1972, the Austrian scholar Leopold Auer asserted: “Undoubtedly, vassalic contingents gained considerably in importance because of the efforts of the Carolingians, and the number of free men in the army, who had been quite numerous up to this point, now shrank.”19 A few years later, Auer returned to this idea in a study focused on the military organization of the early Babenberger in Austria and argued with regard to vassalic contingents that “we are dealing here with a fundamental issue of medieval military history, namely the central question of the relationship between war and social structure,

15 Franz Beyerle, “Zur Wehrverfassung des Hochmittelalters,” in Festschrift Ernst Mayer zum 70. Geburtstag (Weimar, 1932), 31–91, here 31. 16 Ibid., 40. 17 Regino of Prüm, Regionis abbatis Prumensis chronicon cum continuatione Treverensi, ed. F. Kurze, MGH SRG (Hanover, 1890), 118. 18 Eugen von Frauenholz, Das Heerwesen der germanischen Frühzeit, des Frankenreiches und des ritterlichen Zeitalters (Munich, 1935), 60. 19 Leopold Auer, “Der Kriegsdienst des Klerus unter den sächsischen Kaisern,” Mitteilungen des Instituts für Österreichische Geschichtsforschung part 1, 79 (1971), 316–407; and part 2 in ibid., 80 (1972), 48–70. Auer, “Kriegsdienst,” part 1, 320; and idem, part 2, 50, “Zweifellos hat aber das vasallitische Aufgebot durch die Maßnahmen der Karolinger . . . stark an Bedeutung gewonnen und die Zahl der Freien im Heer, die in dieser Zeit noch recht zahlreich vertreten waren, eingeschränkt.”

202

FEUDALISM, ROMANTICISM, AND SOURCE CRITICISM

or put another way, the relationship between status and the obligation to undertake military service.”20 He added that heavily armored mounted men increasingly comprised the core of the army from the time of Pippin III [died 768] onwards, while foot soldiers retreated in importance. The material demands that resulted from this situation (above all the possession of a horse and armor) were too high for the great mass of the free men, particularly in light of the increasing number of wars. The result was that participation in offensive campaigns, that is expeditionary duty, increasingly was limited to the wealthier strata of the population.21 However, Auer did accept that members of the lower social orders continued to be mobilized for territorial defense (Landfolge) into the eleventh century in periods of emergency.22 By contrast, Auer claimed, those members of the lower social orders found on campaign were there merely as servants or attendants in the baggage train.23 In sum, Auer believed that lower social and economic orders were of no direct military value. Auer’s views regarding the changes in military organization under the Carolingians were echoed by Josef Fleckenstein in his study on the intertwined development of the nobility and the nature of military organization in the Frankish empire. Fleckenstein focused his attention on Carolingian capitularies and identified a three-stage process in which the position of the free fighting man was weakened relative to the vassals of the great magnates. The first phase, Fleckenstein argued, came with Charlemagne’s 808 capitulary that called for a mobilization of the entire army with men coming to the host either with their senior or with the count. Fleckenstein interpreted this to mean that the army was now divided into two elements. The first consisted of vassals led by their feudal lords, and the second consisted of nonvassals who were led to war by the count in whose pagus they resided. Fleckenstein argued that the next stage of development was marked by the 847 capitulary of Meersen that summoned all men to come to the host with their seniores, without making any mention of men coming to the host with their counts. Fleckenstein interpreted this to mean that nonvassalic forces no longer participated on campaign. The third stage, Fleckenstein argued, came with the increasing number of nobles who appear as vassals and the concomitant numerical domination of vassalic contingents by men of high social status who held benefices from their superior lords.24 In making this argument, however, Fleckenstein

20 Leopold Auer, “Zum Kriegswesen unter den früheren Babenbergern,” Jahrbuch für Landeskunde von Niederösterreich 42 (1976), 9–25, here 16. 21 Ibid. 22 Ibid., 17. 23 Ibid., 21. 24 Josef Fleckenstein, “Adel und Kriegertum und ihre Wandlung im Karolingereich,” Settimane di Studio del Centro Italiano di Studi sull’alto Medioevo 27 (1987), 2 vols., I, 67–94, here 89–92.

203

N A R R AT I V E W O R K S

studiously ignored the fact that for many free landowners within the comitatus, the royally appointed count was their senior, but they were not his vassals. The arguments regarding the transformation of Frankish military organization and the concomitant domination of warfare by vassalic contingents of noble mounted warriors were further developed in the context of Ottonian history by Hagen Keller and Karl Leyser. The former observed in his study on the foundations of Ottonian royal rule that the transition from the Carolingian period, namely the accentuation of the personal character of the structure of lordship, had a particularly clear effect in the change in military organization. The Ottonians abandoned the territorially-based mobilization of the free – aside from territorial defense – and relied entirely on the vassals of the bishops and secular magnates.25 In a series of essays published from the 1960s to the 1980s, Karl Leyser also developed the theme that warfare in the Ottonian period was dominated by nobles, and their noble vassals, who were distinguished by their mastery of equestrian combat.26 Unlike Keller, however, Leyser looked for this change in the tenth rather than the ninth century. He argued that the Saxons lagged behind their Frankish neighbors in military development, and so the first Saxon king, Henry I (919–936), was compelled to bring his people into modernity by developing a force of heavily armed mounted warriors in order to deal with the Magyar threat. Although he did not cite Heinrich Brunner’s study on mounted military service and the origins of feudalism, Leyser’s argument paralleled Brunner’s model concerning Charles Martel, namely that the creation of an army of mounted vassals led to the creation of a knightly class.27 The concomitant effect on military organization was to eliminate the need for foot soldiers, other than in lowly support roles, such as maintaining the wagon train.28 Leyser was willing to concede that “the ignobile vulgus was by no means yet without weapons.” However, he added that “the honour of arms had become the hallmark and was already the hereditary possession primarily of nobles.”29

25 Hagen Keller, “Grundlagen ottonischer Königsherrschaft,” in Reich und Kirche vor dem Investiturstreit: Vorträge beim wissenschaftlichen Kolloquium aus Anlaß des achtzigsten Geburtstag von Gerd Tellenbach, ed. Karl Schmid (Sigmaringen, 1985), 17–34, here 21, n. 14. 26 See, for example, Karl Leyser, “The Battle at the Lech, 955: A Study in Tenth-Century Warfare,” History 50 (1965), 1–25 and reprinted in Medieval Germany and Its Neighbours, 43–67; idem, “Henry I and the Beginnings of the Saxon Empire,” The English Historical Review 83 (1968), 1–32 and reprinted in K. J. Leyser, Medieval Germany and Its Neighbours 900–1250 (London, 1982), 11–42; and idem, “Early Medieval Canon Law and the Beginnings of Knighthood,” in Instutionen, Kultur und Gesellschaft im Mittelalter: Festschrift für Josef Fleckenstein zum 65. Geburtstag (Sigmaringen, 1984), 549–66, repr. in Communications and Power in Medieval Europe: The Carolingian and Ottonian Centuries, ed. Timothy Reuter (London, 1994), 51–71. 27 Leyser, “Henry I and the Beginnings of the Saxon Empire,” 40–1. 28 Ibid., 21. 29 Leyser, “Early Medieval Canon Law and the Beginnings of Knighthood,” 70–1.

204

FEUDALISM, ROMANTICISM, AND SOURCE CRITICISM

The views of scholars such as Fleckenstein and Leyser have had an enormous influence on subsequent scholarship. With just a few notable exceptions, over the past forty years, the transition to armies consisting of vassalic contingents of mounted knights with the concomitant disappearance of Carolingian-style expeditionary levies has been taken for granted by specialists in the history of Ottonian and Salian Germany.30 I have argued elsewhere that this model is illsuited to describe warfare under the Ottonians.31 My focus in the remainder of this chapter will be on ascertaining whether this consensus among scholars is justified with regard to the eleventh century. Was warfare in the Salian period truly dominated by small groups of knightly vassals who fought cavalry battles against other knightly vassals?32 Was it really the case, as one scholar averred, that “the grotesque situation was in force that the nobility built fortifications that could only be defended on foot, but continued to hold fast to the principle of only fighting on horseback”?33 In sum, is it true that military forces, which might be characterized as non-noble levies, were not deployed for military operations in eleventh-century Germany? In order to answer these questions, I will address two central issues. First, how does the parti pris of the narrative sources from the eleventh century affect their depiction of warfare? Second, what do these same authors have to say about the role played by non-milites in the conduct of war? Before turning to these

30 Numerous studies can be cited in this context. See, for example, Heinrich Koller, “Das mittelalterliche Stadtmauer als Grundlage städtischen Selbstbewußtseins,” in Stadt und Krieg, ed. Bernhard Kirchgässner and Günter Scholz (Sigmaringen, 1989), 9–25, here 17; Charles R. Bowlus, “The Early Kaiserreich in Recent German Historiography,” Central European History 23.4 (1990), 349– 67, here 357; Franz-Reiner Erkens, “Militia und Ritterschaft: Reflexionen über die Entstehung des Rittertums,” Historische Zeitschrift 258 (1994), 623–59, here 629; Otto Ackermann, “Comites und Milites: Grafen und Krieger im Hochmittelalter,” Werdenberger Jahrbuch 7 (1994), 11–20, here 11; Michael Toch, “The Medieval German City Under Siege,” The Medieval City Under Siege, ed. Ivy A. Corfis and Michael Wolf (Woodbridge, 1995), 35–48, here 35–6; and Malte Prietzel, Kriegführung im Mittelalter: Handlungen, Erinnerungen, Bedeutungen (Paderborn, 2006), 21–3, and passim; and idem, Krieg im Mittelalter (Darmstadt, 2006), passim. 31 See, for example, David S. Bachrach, Warfare in Tenth-Century Germany (Woodbridge, 2012). 32 A number of scholars have called into question whether there was a knightly class in Germany during the eleventh century. See the discussion by Johann Johrendt, Milites und Militia im 11. Jahrhundert. Untersuchung zur Frühgeschichte des Rittertums in Frankreich und Deutschland (Nüremberg, 1971); idem, “‘Milites’ und ‘Militia’ im 11. Jahrhundert in Deutschland,” Das Rittertum im Mittelalter, ed. Arno Borst (Darmstadt, 1976), 419–36; and Joachim Bumke, Studien zum Ritterbegriff im 12. und 13. Jahrhundert mit einem Anhang zum Stand der Ritterforschung 1976 (Heidelberg, 1977), and translated into English as The Concept of Knighthood in the Middle Ages (New York, 1982), who made the case that there is no evidence of a knightly class in Germany during the eleventh century, and the term miles cannot properly be translated as knight in this period. For a broader commentary on the problem of scholars using the term inappropriately before 1100, see the discussion by Jean Flori, “Knightly Society,” in The New Cambridge Medieval History: c. 1024–c. 1198, vol. 4, ed. Rosamond McKitterick (Cambridge, 2004), 148–84, particularly 149. 33 Koller, “Das mittelalterliche Stadtmauer,” 17.

205

N A R R AT I V E W O R K S

questions, however, it is useful to make a few observations about the milites who populate the pages of eleventh-century narrative works. In general, it is these milites who have been identified by many scholars either as feudal knights or beneficed vassals of lords who dominated warfare through their monopoly on mounted combat.34 However, the identification of milites as either knights or vassals cannot be sustained for eleventh-century Germany. First, the term miles/milites retained its traditional meaning as a fighting man, and more particularly as a professional soldier, throughout the eleventh century. Second, the sources for this period provide no basis for concluding that men denoted as milites possessed a legal or social status that differentiated them from other laymen on the basis of their identity as milites. The term designates a professional rather than a legal or social status. Nor is there any basis in the sources for concluding that most or even many milites were men with high social status. Rather, the opposite is true. Most milites were men of low social status and could be killed with impunity in sieges and on the battlefield, as illuminated by the fact that contemporary observers felt no need to identify them by name. In addition, it is clear that although some milites did possess beneficia, many did not and instead received wages, i.e. stipendia, for their military service. Finally, although some milites can be shown to have possessed horses, and even to have ridden them in combat, most milites appear in the narrative sources of the eleventh century as garrison troops in fortifications or in armies besieging fortifications where horses were of limited utility.

Authorial bias and non-milites at war in Salian Germany In turning back to problem of the non-milites who served in warfare, it must be emphasized that much of the argument for the transformation of the military organization of the Frankish empire and its successor states, whether under the Carolingian or the Ottonians, depends upon the reading of legal texts, such as capitularies and Landfrieden, in isolation from narrative sources.35 Those scholars who adduce evidence from narrative sources to support the contention that military levies gradually disappeared (whether in the ninth or the tenth century) tend to use examples that are not treated in context, particularly with regard to the role that these passages were intended to play in the author’s overall account.36 From a methodological perspective, selecting isolated or unrepresentative passages from narrative works in order to suit a scholar’s chosen conclusion is particularly problematic. It is well understood that medieval authors, in general, and

34 For a discussion of the historiography treating the meaning of the term miles in pre-Crusade Germany and a fuller discussion of the points made in the remainder of this paragraph, see Bachrach, “Milites and Warfare,” passim. 35 See above for the use of legal sources by scholars, including Hans Fehr and Josef Fleckenstein. 36 With regard to the selective use of narrative texts that leads to a misrepresentation of the text as a whole, see the discussion by Bernard S. Bachrach and David S. Bachrach, “Saxon Military Revolution, 912–973?: Myth and Reality,” Early Medieval Europe 15 (2007), 186–222.

206

FEUDALISM, ROMANTICISM, AND SOURCE CRITICISM

eleventh-century authors in particular, were not engaged in a transparent effort to reveal the past “as it really happened.”37 Rather, authors in the pre-Crusade period, most of whom were aristocratic clerics, were heavily biased not only in favor their particular patrons but also toward members of their own families, and more broadly toward individuals characterized by an elevated social and economic status.38 In the context of writing military history, these biases are manifested in the considerable attention given to relatively small numbers of aristocrats engaged in combat operations, but also, on occasion, in the denigration of members of the lower social orders for rhetorical purposes. As a consequence, when medieval writers note only the deaths of aristocrats in combat, this cannot be understood ipso facto as evidence that nonaristocrats were absent from the battlefield. Similarly, when these same authors write disparagingly about men with low social status on military campaign, it is incumbent upon the historian to determine whether this passage is reflective of reality or whether the author is indulging his biases for rhetorical purposes. A few examples will help to illuminate this problem. Bruno of Merseburg, a Saxon aristocrat and cleric who was employed by both the archbishop of Magdeburg and the bishop of Merseburg during the course of the 1070s and 1080s, devoted considerable attention throughout his history of the civil war in Germany, the Saxonicum Bellum, to lauding and highlighting the positive contribution of nobles and the crude, violent, and uncivilized behavior of peasants and others in the lower strata of society.39 Bruno’s hostility toward the lower social orders manifests itself on several occasions in military contexts. Bruno claims, for example, that during the siege of Würzburg in August 1077, the anti-king Rudolf 37 The crucial observations by Walter Goffart, The Narrators of Barbarian History: Jordanes, Gregory of Tours, Bede, and Paul the Deacon (Princeton, 1988) with regard to the highly politicized approach to the past by early medieval writers certainly are apropos to the discussion of the authors of eleventh-century texts. For an overview of the main narrative sources for the Salian period and a discussion of their biases with a focus on what they have to say about warfare and violence, see Simon M. Karzel, Nihil crudelius a barbaris perpeti potuissent: Die Darstellung von Krieg und Gewalt in den historiographischen Quellen zur Zeit Heinrichs IV (Marburg, 2008). 38 See, for example, David S. Bachrach, “The Military Organization of Ottonian Germany, c. 900–1018: The Views of Bishop Thietmar of Merseburg,” Journal of Military History 72 (2008), 1061–88, here 1067–68. With regard to authorial focus on aristocrats in narrative texts so as to memorialize the author’s own family and friends, see Lutz E. von Padberg, “Geschichtsschreibung und kulterelles Gedächtnis: Formen der Vergangenheitswahrnehmung in der hochmittelalterlichen Historiographie am Beispiel von Thietmar von Merseburg, Adam von Bremen und Helmold von Bosau,” Zeitschrift für Kirchengeschichte 105 (1994), 156–77. 39 The best edition of Bruno’s Saxonicum Bellum is now to be found in Quellen zur Geschichte Kaiser Heinrichs IV.: Die Briefe Heinrichs IV., Das Lied vom Sachsenkrieg, Brunos Sachsenkrieg, Das Leben Kaiser Heinrichs IV., ed. and trans. by Franz-Josef Schmale and Irene Schmale-Ott (Darmstadt, 1974), 192–405 with facing page German translation, hereafter Saxonicum Bellum. See the discussion of Bruno’s background and employment by David S. Bachrach and Bernard S. Bachrach, “Bruno of Merseburg and His Historical Method c. 1085,” Journal of Medieval History 40 (2014), 381–98. With regard to Bruno’s tendency to highlight the deaths of aristocrats and to downplay the deaths of members of the lower social orders in combat, see Saxonicum Bellum, ch. 46, 98, 102, and 117.

207

N A R R AT I V E W O R K S

of Rheinfelden was reluctant to capture the city by storm because he feared that he would not be able to keep the lower-class elements in the army in check and stop them from looting churches and stealing clerical property once they overwhelmed the defenders.40 Consequently, according to Bruno, Rudolf preferred to deploy his troops in a lengthy and expensive siege rather than risk the dishonor that would result from allowing a successful assault on the city by the “vulgar crowd,” an assault that he believed would lead to the destruction of sacred sites.41 In evaluating this passage, it seems unlikely that Rudolf refrained from launching an assault on Würzburg simply because he feared what the undisciplined vulgus would do once inside the walls. It is rather more probable that he feared to sustain the heavy casualties that would result from attempting to storm this welldefended fortress city. What is clear, however, is that Bruno wanted his audience to believe that the vulgus was responsible for Rudolf’s failure to capture the city and that he expected his readers to find it plausible that the vulgus not only was present as part of the besieging army but that these lower-class men would play an important role in any assault against the city. Another author who sometimes indulges in gratuitous attacks on those whom he viewed as social inferiors is Lampert of Hersfeld. In his account of the rebellion by the citizens of Cologne against their archbishop in 1074, Lampert castigates the former for their soft manner of life and their inability to understand the true rigors of military life.42 Lampert claims: It is not difficult to move men of this kind in any way that you might wish, just like a leaf is seized by the wind, because they are raised from their youth with the delights of the city, and never have any experience in war. After selling their wares, they are accustomed to discuss military affairs over wine and food, and think that anything that might occur to them is no sooner said than done. They have no idea how this will end.43 Lampert was a strong partisan of the Thuringian and Saxon nobility, who were engaged in a lengthy rebellion against King Henry IV. In addition, he demonstrates throughout his Annales an acute sense of the rights of ecclesiastical magnates.44 So it is hardly surprising that he chooses to denigrate townsmen who not only opposed their bishop but also were strong supporters of Henry IV. Even 40 Bruno, Saxonicum Bellum, ch. 94. 41 Ibid. 42 Lamperti Annales, in Lamperti Monachi Herfeldensis Opera, ed. Oswald Holder-Egger (Hanover, 1894, repr. 1956), 187. 43 Ibid., “Nec difficile fuit id hominum genus in omne quod velles tamquam folium quod vento rapitur transformare, quippe qui ab ineunte aetate inter urbanas delicias educati nullam in bellicis rebus experientiam habebant, quique post venditas merces inter vina et epulas de re militari disputari soliti omnia quae animo occurrissent tam facilia factu quam dictu putabant, exitus rerum metiri nesciebant.” 44 See the discussion by Karzel, Nihil crudelius, 41–4.

208

FEUDALISM, ROMANTICISM, AND SOURCE CRITICISM

when keeping these biases in mind, however, when one considers this passage in isolation, the reader is left with the image of fat burghers boasting around the dinner table about the great feats of military prowess they planned to achieve. Certainly, such men could be of little value on the battlefield. However, reading further, one finds that Lampert paints a very different picture of the actual military effectiveness of the population of Cologne. He presents the populace marching to the archiepiscopal palace, equipped with helmets and armor as the light of flaming torches gleamed off the points of their swords.45 After a brief fight, they drove the archbishop and his men from the city.46 As these examples suggest, the biases of the narrative sources are toward the depiction of the kind of aristocratic warrior nobility identified by scholars such as Auer, Fleckenstein, and Keller. In addition, the tendency of aristocratic authors to use images of their social inferiors to make rhetorical points often has been interpreted, as in the case of Beyerle and Leyser, as indicative of the lack of military effectiveness of non-noble combatants. However, these scholars have not considered the biases of the aristocratic authors of narrative sources, authors who were focused on the exploits of their aristocratic audiences and frequently used men of lower social and economic status as foils. In light of these pervasive biases, it is clear that arguments from silence regarding the absence of men of low social status on campaign have a rather high hurdle to meet. It is methodologically unsound to conclude that such men were not on campaign simply because authors who were not interested in them did not mention them. Concomitantly, when authors do discuss the presence of fighting men from the lower orders on campaign, this must be understood as going against their natural tendencies, and therefore having rather more significance as a result. Similarly, when aristocratic authors mention the effective military performance of men with low social and economic status, this, too, must be taken seriously as going against their general tendencies and biases. When evaluating the information presented in eleventh-century narrative sources, several other factors also require attention. First, authors of the main narrative texts for eleventh-century Germany such as Berthold of Reichenau, Lampert of Hersfeld, Bruno of Merseburg, Bernold of Constance, and the anonymous author of the Vita Heinrici were very careful in their discussion of military matters to distinguish between milites, who were the members of the military households of secular and ecclesiastical magnates, and other types of fighting men.47 Second, these authors also were inclined to see themselves as operating in the Isidorean tradition of history, in which the narrator was responsible for discussing events that actually happened in the past, in so far as this reality served the rhetorical objectives of the text.48 Finally, these authors also appear to have accepted the 45 46 47 48

Lampert of Hersfeld, Annales, 187, “galeatum, loricatum, igneo mucrone terribiliter fulgurantem.” Ibid. This point is treated in depth by Bachrach, “Milites and Warfare,” passim. Isidore of Seville, Etymologiarum sive originum libri xx, ed. W. M. Lindsay (Oxford, 1911, repr. 1957), 1.40. ‘Historiae est narratio rei gestae, per quam ea, quae I praeterito fact sunt,

209

N A R R AT I V E W O R K S

imperatives of the Ciceronian tradition of historical writing that required the presentation of a plausible framework in which to tell a story to an audience that was familiar with the broader topic, but not necessarily with the details of particular events.49 As a consequence, when the authors indicate that the men in a particular military force were not milites, the proper conclusions to draw are first that such troops were not members of the military households of magnates, and second that the audiences for these narrative texts found it plausible that such nonprofessional fighting men served in military operations. In this context, the authors under discussion here tend to treat the participation of non-milites in offensive military operations in two ways: (1) the use of the presence of such men in order to make a rhetorical point and (2) the formal distinction between milites and other troops who are participating in a battle or siege.

Rhetorical depiction of non-milites on campaign As mentioned earlier, Bruno of Merseburg chose to blame the failure of Rudolf of Rheinfelden to capture Würzburg in 1077 on his ostensible fear that the vulgus in his army would get out of hand when the city was taken. This use of men from the lower social orders as a rhetorical foil in support of the author’s parti pris is a common topos in the narrative texts of this period.50 Bruno himself utilized this ploy again when discussing the outbreak of a widespread rebellion against Henry IV and the joining of rebel Saxon and Swabian armies in the Rhineland in the autumn of 1076. Although they were now allies, the Saxon and Swabian princes feared to bring their armies into a single camp, according to Bruno, because they dinoscuntur.’ Also see the translation of the work, The Etymologies of Isidore of Seville, trans. Stephen Barney, J. W. Lewis, J. A. Beach, and Oliver Berghof (Cambridge, 2006). Regading the wide-spread acceptance of Isidore’s dictum see Benoît Lacroix, L’Historien au moyen âge (Paris, 1971), particularly chapter one, and the review article of this work by Robert W. Hanning, History and Theory 12 (1973), 419–34; A. J. Minnis, Medieval Theory of Authorship: Scholastic Literary Attitudes in the Later Middle Ages (London, 1984), particularly 1–20; D. H. Green, Medieval Listening and Reading: The Primary Reception of German Literature 800–1300 (Cambridge, 1994), 246–8; Gabrielle M. Spiegel, The Past as Text: The Theory and Practice of Medieval Historiography (Baltimore, 1997), 88–90; Almut Sauerbaum, “Accessus ad auctores: Autorkonzeption in mittelalterlichen Kommentartexten,” in Autor und Autorschaft im Mittelalter, ed. Elizabeth Andersen, Jens Haustein, Anne Simon, and Peter Strohschneider (Tübingen, 1998), 29–37; and Sebastian Coxon, “Zur Form und Funktion einiger Modelle der Autorenselbstdarstellung, ‘Wolfdietrich’ und ‘Dietrichs Flucht’,” in ibid., 148–62, particularly 151 and 162. A full study of the approach toward history by German writers in the Salian period is a desideratum. For an introduction to this issue and a discussion of the approach taken by Bruno of Merseburg, see Bachrach and Bachrach, “Bruno of Merseburg and His Historical Method c. 1085,” 381–98. 49 With regard to the impact of Cicero’s ideas about the role of the historian on the writing of historical works in the period after c. 1000, see Justin C. Lake, “Truth, Plausibility, and the Virtues of Narrative at the Millennium,” Journal of Medieval History 35 (2009), 221–38; and Matthew Kempshall, Rhetoric and the Writing of History, 400–1500 (Manchester, 2011), here 255. 50 Concerning the depiction of violence by members of the lower social orders as well as city dwellers, see Karzel, Nihil crudelius, 124–34.

210

FEUDALISM, ROMANTICISM, AND SOURCE CRITICISM

were worried about the potential for violence between their men. The two armies had fought on opposite sides the previous year at the battle of the Unstrut (9 June 1075), and many men in both the Swabian and Saxon armies were nursing grievances. Bruno specifically comments that it was the men of the lower social orders (viles personae) who posed the greatest source of concern for the commanders.51 As it turned out, there was no violence between the Saxon and Swabian armies, and as a result of their unity, Henry IV was forced to sue for peace and promise to seek absolution from Pope Gregory VII (1073–1085) for his sins. However, Bruno was able to use this scene to pursue his agenda of demonstrating the moral superiority of aristocrats as contrasted with the vulgus. Bruno again points to the difficulties of employing men of the lower social orders on campaign in his discussion of the battle between Rudolf of Rheinfelden and King Henry IV on the Elster river (14 October 1080). Before the battle began, the Saxon army was divided into two parts with the intention of catching the royalist forces in a pincer. Mounted troops under the direct command of Rudolf approached the royal camp from the west but were met and defeated by Henry IV’s own mounted forces. The Saxon foot soldiers, under the command of Otto of Northeim, approached the royal camp from the north, cutting through a densely wooded area that Henry and his advisors had believed was impassable. When the Saxon foot soldiers launched their attack on Henry IV’s camp, the surprise was complete, and the Saxons drove the king’s men into the waters of the White Elster river where many of them are reported to have drowned.52 Clearly, the foot soldiers of the Saxon levy acquitted themselves quite well in this battle. However, Bruno again permitted his bias against men from the lower social orders to come to the fore when describing the aftermath of this victory. According to Bruno, once the royalists had been driven off, the entire focus of the Saxon foot soldiers was on looting King Henry’s camp, so that they could satisfy their greed and keep all the booty for themselves. Bruno asserts that they were only restrained from indulging their foolish desires by the good sense of the nobleman, Otto of Northeim, who warned the foot soldiers to stay away from the royal camp until it was clear that all of the enemy troops had been defeated.53 The Saxon foot soldiers did obey Otto’s commands, which indicates that they could follow orders even when these were contrary to their perceived interests. Lampert of Hersfeld, despite his hostility toward the urban levy of Cologne, discussed previously, was less inclined than his contemporary Bruno to engage in 51 Bruno of Merseburg, Saxonicum Bellum, ch. 88. 52 Ibid., ch. 122. 53 Ibid. As a postscript to the battle on the Elster river, Bruno drew attention to the many calamities that were suffered by King Henry’s troops. Among these were the subsequent death of many powerful men at the hands of peasants defending their own homes, who were armed with axes and clubs. Yet other magnates were captured and held for ransom by local inhabitants, whom Bruno denotes as low-born people (personae viles). By contrast, Bruno claimed: “if any of the enemy were brought to one of our upright men, they were healed if they were wounded, and were sent back to their fatherland free of charge, well outfitted with clothing and arms.” See ch. 123.

211

N A R R AT I V E W O R K S

the wholesale denigration of men from the lower social orders in military matters. This was particularly the case in regard to those who opposed King Henry IV, the antihero of Lampert’s Annales. However, when discussing military matters Lampert did play on what would seem to have been the anti-peasant biases of many aristocrats. For example, in his account of the days leading up to the battle of the Unstrut, Lampert took pains to explain why King Henry and his advisors were so confident about engaging their Saxon opponents. Henry had dispatched scouts (exploratores) to determine the size and disposition of the Saxon army. The scouts reported back that although the Saxons equaled the royal army in numbers and were equipped with roughly the same armament, the king’s troops were otherwise superior in every other way for battle.54 In discussing whether to advance against the Saxon camp, the king’s advisors, according to Lampert, stressed that they had a picked force (miles lectissimus). By contrast, the majority of the Saxons were peasants (rustici) and comprised, therefore, an inept rabble (ineptus vulgus) who were more accustomed to agricultural labor than to military service.55 The king’s advisors added that the Saxons lacked the military spirit that was necessary for war and had been mobilized because of their fear of their princes. Consequently, these peasants could not be expected to stand and fight once the battle began and sword was raised against sword. In fact, they would certainly flee before the battle even began, terrified by the noise as the two armies approached one another.56 In discussing the battle itself, Lampert takes considerable pains to show that this assumption by the king’s advisors was ill-founded. The Saxon troops, including the vulgus, initially gave a very good account of themselves in the fighting before ultimately being driven from the field.57 While not so subtly mocking the assumptions by Henry and his advisors about the lack of military skill of the Saxon peasantry, Lampert himself indulges in this very same rhetorical theme in his description of the aftermath of the battle with regard to the men of the lower social orders who were serving in the royal army. As Lampert makes clear, following a lengthy battle, the Saxon line was shattered by simultaneous attacks on both flanks by King Henry’s troops.58 In discussing the subsequent rout of the Saxons, Lampert, drawing upon Sallust’s description of the retreat of an army in the Jugurthine war, observed “as is always the case when the enemy flees, the most cowardly and the most courageous are equally brave [in pursuit].”59 Lam54 Lamperti Annales, in Lamperti Monachi Herfeldensis Opera, ed. Oswald Holder-Egger (Hanover, 1894, repr. 1956), 216. 55 Lampert, Annales, 216, “Illinc vulgus esse ineptum, agriculturae pocius quam miliciae assuetum.” 56 Ibid., “et ideo non expectaturum, ut commisso certamine comminus stricto ense feriret et feriretur, sed ante commissum prelium solo concurrentis exercitus strepitu et clamore terrendum, fugandum et fundendum esse.” 57 Ibid., 217–21. 58 Ibid., 220. 59 Ibid., “ut semper in fuga hostium ignavissimis et fortissimis par solet esse audacia.” Here, Lampert draws on Sallust, Jugurtha, ch. 53.

212

FEUDALISM, ROMANTICISM, AND SOURCE CRITICISM

pert then made the severe observation that all of the royal troops went off in pursuit of the fleeing Saxons, “even the plebeians and the peasants who usually hung around the camp doing servile work.”60 Lampert’s contention here regarding the plebei ac rustici could, if taken out of context by scholars seeking “evidence” for the military marginalization of the rustici, be utilized to show that men of this status had only ancillary roles in military campaigns as servants, cooks, attendants for horses, and the like. However, as we will see later, this is not accurate.61 In addition to using the rustici and plebei as a rhetorical foil in his discussion of the pursuit of the Saxons at the battle of the Unstrut, Lampert addresses Henry IV’s mobilization of non-milites in an effort to show how little support the king had from the magnates for his war against the Saxons. In the context of the initial Saxon and Thuringian rebellion in July 1073, Henry tried to divert an army that he had mobilized for operations in Poland to suppress the widespread revolt against his rule in Saxony and Thuringia. However, Henry’s secular and ecclesiastical magnates refused to consider this plan.62 Consequently, Henry dismissed the army but sent messengers to all of the principes in the kingdom announcing that he would be mobilizing a new army in early October that was to join him at the villa of Bredingen near Hersfeld. The specific purpose of this army was to undertake operations in Saxony.63 Lampert takes pains to stress that on this occasion, the royal messengers were dispatched not only to the princes but to the people as well (non solum principibus sed et popularibus). As Lampert makes clear, Henry’s decision to mobilize the populus was inspired, in no small measure, by his fear that the magnates either would not come to the muster at Bredingen or would not fight if they did go on campaign.64 It turns out that King Henry was right to be worried. The German princes refused to fight, and instead compelled him to make a peace agreement with the Saxons and Thuringians at the villa of Gerstungen in the autumn of 1073.65 With his options limited in regard to Saxony, Henry turned his attention to the potentially advantageous request by his brother-in-law Salomo to help him regain the Hungarian throne. In return for his aid, Henry would receive a series of frontier fortifications in Hungary that would help ensure that the Hungarians would no longer be able to raid into Bavaria.66 Henry therefore issued summonses in July 1074 to his principes ordering them to mobilize for a campaign 60 Lampert, Annales, 220, “omnes in exercitu regis legiones confusis ordinibus, omnes etiam plebei ac rustici, qui castrorum usibus servilem operam dependebant.” 61 As noted previously, both Leopold Auer and Karl Leyser presented men with lower social and economic status serving only menial tasks on campaign. 62 Lampert, Annales, 157. 63 Lampert, Annales, 158. 64 Ultimately, the German princes refused to fight the Saxons, and Henry IV was forced to make concessions to the Saxons in an agreement that is commonly known as the peace of Gerstungen. See Lampert, Annales, 163. 65 Ibid. 66 See the discussion by Lampert, Annales, 197–8.

213

N A R R AT I V E W O R K S

in Hungary. But according to Lampert, the princes refused, citing the costs that they had just undertaken for the expedition to Saxony the previous year that had led to the peace of Gerstungen. As a consequence, Lampert avers, Henry was forced to begin his Hungarian campaign with a very limited force comprising just the royal military household (miles privatus) and a force of nonprofessional fighting men, whom Lampert denotes as the common herd (gregarius).67

Milites and other troops In addition to the mention of rustici, plebei, viles personae, gregarii, and the vulgus in military campaigns for the purpose of making a rhetorical point, the authors of narrative sources routinely distinguish between milites and other types of troops when discussing the mobilization and deployment of fighting men for military operations. The Swabian author Berthold of Reichenau, for example, emphasized that when Henry IV returned to Germany from Canossa in June 1077, he immediately mobilized a large army in order to crush the rebels who were opposed to his rule.68 Berthold distinguished among these various forces, noting the presence of Henry’s own milites as well as the troops provided by the Swabian bishops of Basel and Strassburg and the secular magnate Count Palatine Hermann. These household troops are contrasted with a substantial additional force (non modica auxiliariorum militia) of Bavarians and Carinthians, almost the entire strength (virtus) of Burgundy, and substantial force (non modica pars) of Franconians.69 Berthold, who was very precise in his use of military terminology, reserved alternative terms to denote the Bavarians, Carinthians, Burgundians, and Franconians. This was his way of signaling to his audience that these forces included nonmilites in their ranks. As we will see later, many of these men were, in fact, rustici. Lampert of Hersfeld also frequently used the technique of distinguishing between milites and other types of troops and used a variety of terms to denote the latter, including exercitus and copiae.70 In his discussion of the initial phases of the battle on the Unstrut, noted earlier, Lampert conspicuously does not draw the reader’s attention to the milites in the royal army, but rather observes that the exercitus of the Swabians and the exercitus of the Bavarians retreated after suffering considerable losses at the hands of Otto of Northeim’s Saxon troops.71 As noted previously, these Saxons included men of all social statuses, including rustici. Lampert then makes clear that the royal position on the battlefield was salvaged

67 Ibid., 198. 68 Die Chroniken Bertholds von Reichenau und Bernolds von Konstanz, ed. I. S. Robinson, MGH SS rerum germanicarum nova series 14 (Hanover, 2003), hereafter Chronicon, here 276. 69 Berthold, Chronicon, 276. 70 See, for example, Lampert, Annales, 105, 107, 119, 123, 142, 148, 161, 188, and 214. 71 Lampert, Annales, 220.

214

FEUDALISM, ROMANTICISM, AND SOURCE CRITICISM

when the milites from the military households of the count of Gliesburg and the bishop of Bamberg struck the Saxon line on the two flanks.72 As noted earlier, Bruno of Merseburg, writing independently of Lampert, observed that the Swabian forces at the battle of the Unstrut included viles personae, who were still angry about the losses that they had suffered in battle at the hands of the Saxons. The presence of such men in the Swabian battle line certainly helps to explain Lampert’s decision to denote the entire Swabian force as the exercitus of the duchy rather than as a congeries of military households of Swabian magnates, i.e. milites, because the Swabians included a mixed group of professional and nonprofessional fighting men. Lampert similarly distinguished between the mobilization of a large force consisting of both non-milites and professional forces in his discussion of a campaign undertaken by Margrave Ekbert of Meissen against Henry IV’s Bohemian allies in August 1076.73 In the aftermath of his resounding victory at the battle on the Unstrut, Henry had handed over a series of fortifications in the march of Meissen to Duke Vratislav II of Bohemia (1061–1092) as a reward for his aid in this campaign. Ekbert, joined by “the Saxons,” as Lampert has it, recaptured all of these strongholds. After the Saxons withdrew and returned to their home, Ekbert installed garrisons of his own milites to keep the region secure.74 In a similar manner, Lampert distinguishes between the Thuringians who besieged royal fortifications in the summer of 1073 and the royal milites who defended these strongholds.75 Lampert observes that a multitude (multitudo) of Thuringians from the nearby area (ex vicinis locis) undertook a siege of the fortress of Heimenburg. Clearly, this was not a small force of heavy cavalry. Rather than attempting to starve the garrison into submission, they assaulted the castellum and captured it.76 The Thuringians, whom Lampert now denotes as an exercitus, then began operations at the fortress of Asenburg. But the strong natural defenses of the fortress, as well as the toughness of the garrison, whom Lampert denotes as milites regis, were sufficient to thwart the efforts of the Thuringians, who maintained their siege from August through December 1073.77 The anonymous author of the Carmen de bello Saxonico, a panegyric written on behalf of King Henry just after his victory on the Unstrut in June 1075, similarly

72 Ibid. The successful assault by these milites was then followed up by the copiae of the Bohemians and the Lotharingians, including a substantial number of mounted troops (equites). 73 Lampert, Annales, 273. 74 Ibid. 75 Lampert, Annales, 161. 76 Ibid. The author of the Carmen de bello Saxonico, claims that some 3,000 men participated in the attack on Heimenburg. See Quellen zur Geschichte Kaiser Heinrichs IV.: Die Briefe Heinrichs IV., Das Lied vom Sachsenkrieg, Brunos Sachsenkrieg, Das Leben Kaiser Heinrichs IV., ed. and trans. by Franz-Josef Schmale and Irene Schmale-Ott (Darmstadt, 1974), here after Carmen, bk 1: 87–124. 77 Lampert, Annales, 161 and 172.

215

N A R R AT I V E W O R K S

distinguishes between milites and other types of troops in his discussion of the mobilization of the Saxons to defend their region in early 1074.78 He claims that as word went throughout Saxony to prepare for war, the rustici cast aside their farming implements, the shepherds left their flocks, and the merchants abandoned their wares. These men from every station and every profession girded themselves for war. The poet then declared that this varied force of farmers, shepherds, and merchants, equipped with whatever arms they had available, joined with the milites and rushed to war.79 Many of the non-milites considered up to this point are identified in the sources as rustici, that is men from the countryside. However, urban levies also played an important role in military operations. As discussed earlier, Lampert of Hersfeld drew attention to the urban levy of Cologne and their successful effort to drive the archbishop’s household troops from the city. The author of the Carmen de bello Saxonico, in the passage just discussed, mentions that mercatores were to be found in the Saxon host in early 1074. Earlier in the poem, he had observed that cobblers, smiths, bakers, and butchers had joined with milites from Goslar to participate in the Saxon siege of the royal fortress of Harzburg.80 Bruno of Merseburg, for his part, scornfully observed that Henry IV’s army in the summer of 1077 was exceptionally weak and that most of his men were mercatores.81 The anonymous author of the Vita Heinrici, which was written shortly after the emperor’s death in 1106, presents a much more positive image of urban levies in his discussion of the conflict between Henry IV and his son Henry V.82 After the elder Henry’s escape from captivity in early 1106, he made his way to Liège, where he was warmly received by both the bishop Otbert (1091–1119) and Duke Henry of Lower Lotharingia (1101–1106). According to the Vita, when Henry V began preparing to march against his father, the Lotharingian duke, along with the citizens of Cologne and Liège, prepared their arms, gathered their forces, and improved the defenses of their cities. In addition, they promised Henry IV their full support in the coming struggle.83 The Vita goes on to make clear that the citizens of Cologne, Liège, and other cities did not intend simply to defend their walls against Henry V – they also planned to take action against his forces directly in the field. When Henry V began his siege of Cologne, his initial assaults were

78 79 80 81

Carmen, Bk II: 130–44. Ibid. Carmen, bk I: 198–9. Bruno of Merseburg, Saxonicum Bellum, ch. 95, “nam maxima pars eius ex mercatoribus erat.” 82 For the text of the Vita Heinrici see Quellen zur Geschichte Kaiser Heinrichs IV.: Die Briefe Heinrichs IV., Das Lied vom Sachsenkrieg, Brunos Sachsenkrieg, Das Leben Kaiser Heinrichs IV., ed. and trans. by Franz-Josef Schmale and Irene Schmale-Ott (Darmstadt, 1974), hereafter Vita Heinrici, c. 13. 83 Ibid., “Cum igitur audisset H dux et Colonienses cum Leodicensibus, quod super se rex exercitum ducere vellet, arma parabant, copias colligebant, urbes firmabant et ad resistendum pari voto studioque se accingebant.”

216

FEUDALISM, ROMANTICISM, AND SOURCE CRITICISM

driven back by the citizens. At the same time, the citizens of other cities prepared a fleet and blockaded the Rhine river so that the younger Henry’s army could not receive supplies by water. Then when parties of foragers were sent out from Henry V’s siege camp, the citizens of Cologne and others, who knew the local terrain, attacked the foraging parties.84

Military organization of non-milites The examples treated thus far indicate that non-milites, that is men outside of the military households of magnates, played a substantial role in combat operations. One important question, however, is the extent to which the mobilization of such men can be understood as routine rather than as ad hoc. Two authors working independently from each other and in different regions of the German kingdom, Berthold of Reichenau and Lampert of Hersfeld, give indications that the mobilization of such men was not an ad hoc response to the exigencies of a particular campaign. Rather, non-milites expected to serve on campaign when called upon to do so by leaders with the legal jurisdiction to summon them. Moreover, both authors indicate that this jurisdiction was territorially based. For example, Berthold of Reichenau observes that in the summer of 1078, Henry IV dispatched Bishops Burchard of Basel and Thiepald of Strassburg, noted earlier, to deploy their forces to ravage the lands of Duke Berthold of Carinthia.85 The chronicler comments that the military forces of the two bishops were composed of two separate elements. The first of these consisted of their household troops, that is their milites. The second component consisted of rustici. These rustici, Berthold of Reichenau explains, were mobilized on the basis of the counties (comitatus) in which they lived. Moreover, these men were required to serve in this campaign because of the oaths they had sworn to the bishops, i.e. they were adiurati.86 The transfer of comital authority to German bishops with the concomitant responsibility to mobilize the levies of the comitatus for military operations dates back to the reign of Otto III (983–1002).87 The bishops of Strassburg had enjoyed comital rights and undertaken comital obligations since the late tenth century.88 The bishops of Basel had received specific comital authority in Augstgau and

84 85 86 87

Ibid., 462–4. Berthold of Reichenau, Chronicon, 332. Ibid. See the discussion of the utilization of bishops to serve comital functions by Hartmut Hoffmann, “Grafschaften in Bischofshand,” Deutsches Archiv für Erforschung des Mittelalters 46 (1990), 375–480. 88 Die Urkunden Otto des II und Otto des III., ed. Theodor Sickel (Hanover, 1888), DO II, nos. 72 and 73; Die Urkunden Heinrichs II. und Arduins, ed. Harry Bresslau, Robert Holtzmann, and Hermann Reincke-Bloch (Hanover, 1900–1903), no. 367; and Die Urkunden der deutschen Könige und Kaiser: Die Urkunden Heinrichs III., ed. Harry Bresslau and Paul Kehr (Berlin, 1931), no. 220.

217

N A R R AT I V E W O R K S

Sisgau in 1041 from King Henry III.89 Consequently, the mobilization of these rustici is consistent with the ongoing practice of both Ottonian and Salian kings to utilize their bishops to undertake comital duties on their behalf. Berthold again draws a clear connection between territorially based mobilization of non-milites and the taking of oaths in his discussion of the mobilization of Franconian rustici, who fought just a short time later against Swabian, Carinthian, and Bavarian forces under the command of Duke Welf of Bavaria and Duke Berthold of Carinthia.90 The latter were marching to join the Saxon army under the command of Rudolf of Rheinfelden, who was advancing into eastern Franconia. Henry IV was attempting to keep his Saxon and southern opponents from joining their forces, with the plan of facing and defeating them seriatim. Henry therefore marched into eastern Franconia to face the Saxons, whom he faced at the battle of Mellrichstadt (7 August 1078). Concomitant with his own military operations, the king ordered western Franconian rustici to mobilize in an effort to delay Duke Welf and Duke Berthold. According to Berthold of Reichenau, these Franconians, whom the author denotes as both comprovinciales and rustici, were mobilized from their centenariae, i.e. their local subcomital districts. Moreover, he observes that they were bound together for military service by their oaths (coniurati). Not incidentally, Berthold also notes that these rustici were equipped with weapons of war (arma militaria), likely in contrast to farm implements such as scythes and pitchforks.91 This final observation permits the inference that these were wealthier rustici who could afford specifically military equipment. The purchase and possession of arma militaria, moreover, would seem to be prima facie evidence of the expectation by these men that they could be called upon to do military service. Lampert of Hersfeld, for his part, illuminates the legal and jurisdictional basis for the mobilization of men in Lotharingia and in Saxony for military service. In discussing the army that King Henry mobilized for operations in Saxony in the autumn of 1075, Lampert notes that Duke Theoderic II of Upper Lotharingia (1075–1115) and Duke Godfrey IV of Lower Lotharingia (1069–1076) brought enormous forces to join the royal army. In fact, Lampert mocks the dukes for attempting to show off by standing out among the other princes on the basis of the size of their armies.92 In commenting on these troops, Lampert observed that the Lotharingians were very well equipped. The parallel with the Franconian rustici is quite clear. Lampert then explains that the two dukes had assured the high quality of their military forces by imposing a very strict selection criteria on the men whom they mobilized, that is they only picked the best men for campaign

89 90 91 92

Die Urkunden Heinrichs III, no. 77. Berthold of Reichenau, Chronicon, 332–3. Ibid. Lampert, Annales, 234.

218

FEUDALISM, ROMANTICISM, AND SOURCE CRITICISM

duty. Lampert explicitly states, moreover, that these men were mobilized from the regions over which the two dukes had jurisdiction.93 Lampert’s comments in this regard illustrate that Theoderic and Godfrey had the authority to mobilize troops because of their ducal offices. In addition to being territorially based, this authority to mobilize fighting men had an intrinsic selection mechanism that allowed the two dukes to mobilize the men who were best equipped and best suited for service on military campaign. In light of Lampert’s consistent effort to mention when particular magnates and milites were mobilized for service, his failure to do so in this case indicates that the copiae in question included large numbers of non-milites. In one final example, Lampert describes the mobilization of a vast Saxon army in January 1074 to repel what the Saxons feared would be an invasion of their region by King Henry.94 Almost certainly exaggerating, Lampert claims that as many as 40,000 Saxons of all ranks came to Werra river.95 This is the same force regarding which the author of the Carmen de bello Saxonico, noted previously, had claimed the mobilization of farmers, shepherds, and merchants alongside the milites. As it became clear that there would be no immediate invasion, the Saxons sent a substantial part of their army home. In discussing the demobilization of part of the Saxon force, Lampert specifies that 12,000 of the men of lower social and economic status (plebei) were released. Lampert explains that these men, in particular, had been summoned with so little notice that they had not brought food with them and had to go home.96 The clear implication of Lampert’s comment is that even men of “plebian” status had the expectation not only of going on campaign outside their home districts to fight but that under normal circumstances they would bring sufficient supplies with them to be able to remain in the field for a substantial period.

Conclusions As the many examples of plebei, rustici, vulgus, gregarii, viles personae, mercatores, and just plain cives treated here make clear, warfare in the eleventh century, including combat operations, was not monopolized by milites, whether or not one wishes to construe such men as mounted knights. In fact, non-milites played a central role both in battles in the field and in sieges. In fact, they likely were the numerically preponderant element in any large army. It is notable that 93 Ibid., “ita militaribus armis instructas, ita de tota cui preerat regione severissimo delectu habito exquisitas.” 94 Lampert, Annnales, 176. 95 The vast size of the Saxon army is confirmed by both the anti-Henry Bruno of Merseburg and the pro-Henry author of the Carmen de bello Saxonico. See Bruno of Merseburg, Saxonicum bellum, c. 32; and Carmen, bk II, 130–60. 96 Lampert, Annales, 176–7, “quoniam subito clamore in expeditionem evocata cibos secum non sumpsissent, in domos suas, tamquam minus sibi necessaria remitterent.”

219

N A R R AT I V E W O R K S

all of the combat operations that I have discussed took place during Henry IV’s reign, and most of these in the context of the civil wars of the 1070s and 1080s. One might conjecture on this basis that the mobilization of the vulgus was the consequence of the enormous demands of the lengthy military conflict, and therefore, a return to the type of military organization that typified the Frankish empire before the alleged disappearance of military levies. However, such a conjecture would be incorrect. First, none of the authors discussed here give any indication that the mobilization of non-milites for combat operations marked a new or surprising turn of events. Rather, the opposite is the case, in that the mobilization of rustici is depicted as a normal and well-developed practice. Second, the vulgus and rustici, as well as mercatores, appear in both Saxon and royal armies right from the beginning of the civil wars in 1073, long before there could have been any question of the exhaustion and depletion of magnate military forces. The weight of the evidence for the deployment of non-milites during Henry IV’s reign is largely an artifact of the production of a large number of lengthy historical works, which devoted enormous attention to military matters because they were inspired by the dramatic events during the final three decades of the eleventh century. Finally, it should be observed that the main basis for the explanation of the elimination of levies as an important element in early medieval military organization is the claim that battles in the field, and moreover battles between mounted forces, came to dominate warfare. As virtually all specialists in military history are well aware, this was never true. Sieges dominated warfare in both the Carolingian Empire and in its various successor states.97 Henry IV’s reign was actually quite conspicuous for the relatively large number of battles that were fought, with at least five and perhaps six between 1075 and 1086.98 Nevertheless, these battles were outnumbered by a factor of at least six to one by sieges, including operations against fortress cities such as Würzburg and powerful fortresses such as the Harzburg. In order to undertake these numerous siege operations, military commanders required far more troops that could be mobilized from among the milites employed by magnates. Thus, the rustici and other non-milites of the eleventh century continued to play the same role as their forbears had under the Ottonian kings, and before them the Carolingians and Merovingians as well.

97 The scholarship in this area is too vast to summarize here. However, for valuable entry points, see Jim Bradbury, The Medieval Siege (Woodbridge, Boydell, 1992); and Peter Purton, A History of the Early Medieval Siege, c. 450–1200 (Woodbridge, Boydell, 2010), and the enormous literature cited there. Cf. Guy Halsall, Warfare and Society in the Barbarian West 450–900 (London, Routledge, 2003), 215–27. 98 For the relatively large number of battles fought during Henry IV’s in the context of the Saxon wars, see Gillingham, “An Age of Expansion, c. 1020–1204,” 73–6. By comparison, there were at least twenty-seven significant sieges during this same period. See the appendix below.

220

FEUDALISM, ROMANTICISM, AND SOURCE CRITICISM

Appendix Sieges during Henry IV’s Saxon Wars, 1072–1088 1072: Siege of Lüneburg by Saxon forces commanded by Hermann Billung99 1073: Siege of Heimenburg by Thuringian forces opposing Henry IV100 1073: Siege of Asenburg by Thuringian forces opposing Henry IV101 1073: Siege of Harzburg by Saxon forces102 1073: Siege of Wiganstein by Saxon forces103 1073: Siege of Moseburg by Saxon forces104 1073: Siege of Sassenstein by Saxon forces105 1074: Siege of Spatenberg by Thuringian forces opposing Henry IV106 1074: Siege of Vokenrot by Thuringian forces opposing Henry IV107 1077: Siege of Sigmaringen by Rudolf of Rheinfelden108 1077: Siege of fortification in Burgundy by bishops of Basel and Lausanne on behalf of Henry IV109 1077: Swabians loyal to Henry IV besiege fortifications of Swabian magnates opposed to Henry IV110 1077: Siege of Würzburg by Rudolf of Rheinfelden111 1077: Siege of the fortifications of Count Eckbert I of Formbach in Bavaria by Henry IV112 1078: Siege of a fortification in Bavaria by Henry IV113 1078: Siege of Tübingen by Henry IV114

99 100 101 102 103 104 105 106 107 108

109 110 111 112 113 114

Lampert of Hersfeld, Annales, 160–1. Lampert of Hersfeld, Annales, 159 and 161; Carmen de bello Saxononico, I.87–130. Lampert of Hersfeld, Annales, 159 and 161. Bruno of Merseburg, Bellum Saxonicum, ch. 28–29; Lampert of Hersefeld, Annales, 159; Carmen de bello Saxonico, 1. 139–237 and 2.74–114. Lampert of Hersefeld, Annales, 159. Ibid. Ibid. Lampert of Hersfeld, Annales, 159 and 174. Ibid. Berthold of Reichenau, Chronicon, 277; Bernold of Constance, Chronicon, 413; Casuum Sancti Galli continuation II, ed. G. H. Pertz MGH SS 2 (Hanover, 1829), 157; and Casus monasterii Petrishusensis, ed. Otto Abel and Ludwig Weiland, MGH SS 20 (Hanover, 1869), 646. Berthold of Reichenau, Chronicon, 289; and Gesta episcoporum Lausannensium, ed. I. Heller, MGH SS 24 (Hanover, 1879), 799–800. Berthold, Chronicon, 290. Bruno of Merseburg, Bellum Saxonicum, ch. 94; Berthold of Reichenau, Chronicon, 292–3; Bernold of Constance, Chronicon, 458–9; Vita Heinrici imperatoris, 422. Berthold of Reichenau, Chronicon, 301. Berthold of Reichenau, Chronicon, 313. Annales Zwifaltenses, ed. D. Otto Abel, MGH SS 10 (Hanover, 1852), 54.

221

N A R R AT I V E W O R K S

1078: Siege of a fortification in eastern Franconia by Duke Welf of Bavaria and Duke Berthold of Carinthia115 1079: Storming of a fortification held by Count Otto II of Buchhorn in Raetia by Duke Welf116 1079: Siege of a fortification near Ulm by Duke Welf of Bavaria117 1080: Siege of Saxon fortifications by Margrave Eckbert on behalf of Henry IV118 1084: Siege of Augsburg by Duke Welf of Bavaria119 1084: Siege of Burgdorf in Burgundy by forces loyal to Henry IV120 1086: Siege of Regensburg by Duke Welf and Swabian magnates121 1086: Siege of Würzburg by Swabian, Saxon, and Bavarian magnates122 1087: Siege of a fortification in Bavaria by Henry IV123 1088: Siege by Henry IV of the fortification of Gleichen in Thuringia, southwest of Erfurt124 1088: Siege of Quedlinburg by Margrave Eckbert II, Margrave of Meißen125

115 116 117 118 119 120 121 122 123 124 125

Berthold of Reichenau, Chronicon, 335; Bernold of Constance, Chronicon, 420. Berthold of Reichen, Chronicon, 348–9. Berthold of Reichenau, Chronicon, 360. Berthold of Reichenau, Chronicon, 379. Bernold of Constance, Chronicon, 438. Bernold of Constance, Chronicon, 444. Bernold of Constance, Chronicon, 457–8. Bernold of Constance, Chronicon, 458–60; Vita Heinrici imperatoris, 422. Bernold of Constance, Chronicon, 462. Bernold of Constance, Chronicon, 473. Vita Heinrici, 426.

222

10 RALPH OF CAEN AS A M I L I TA RY H I S TO R I A N

For those seeking glory, it is a perilous undertaking to contribute a paper on the First Crusade to a collection of essays honoring John France.1 His Victory in the East remains the dominant account of the military history of the First Crusade almost 20 years after it was written.2 There is, however, one important area in regard to the study of the First Crusade, which John has recognized deserves fuller attention, i.e. the evaluation of the narrative sources that we use for the study of military history.3 Therefore, we are offering this examination of Ralph of Caen as a military historian not only to honor John’s exceptional contributions to the study of the First Crusade but also to support his effort to encourage a more thorough examination of the value of the various narrative sources.4

Writing medieval military history Modern scholars who undertake to write medieval military history are greatly disadvantaged by the tendency exhibited by many of our narrative sources, most of which were written by clerics, to harbor a highly negative attitude toward war.5 These same sources also tend to find the men who, as the biblical characterization 1 Obviously, we allude here to John’s Perilous Glory: The Rise of Western Military Power (New Haven, 2011). 2 John France, Victory in the East: A Military History of the First Crusade (Cambridge, 1994). 3 See The Crusades and their Sources: Essay Presented to Bernard Hamilton, ed. John France and William G. Zajac (Aldershot, 1998). 4 The Gesta Tancredi of Ralph of Caen: A History of the Normans on the First Crusade, trans. Bernard S. Bachrach and David S. Bachrach (Aldershot, 2005), and for the Latin text we relied on Ralph of Caen, Gesta Tancredi, RHC Occ. vol. 3. Our translation was not able to take into consideration Radulphi Cadomensis Tancredus, ed. Edoardo D’Angelo (Turnhout, 2011). However, we found no reading of the new Latin edition that required us to alter our translation or understanding in those parts of the text used for this chapter. 5 Regarding the traditional negative parti pris of clerical writers concerning war, see Walter Goffart, The Narrators of Barbarian History: Jordanes, Gregory of Tours, Bede, and Paul the Deacon (Princeton, 1988), especially 174–83; and Thomas Scharff, Die Kämpfe der Herrscher und der Heiligen: Krieg und historische Erinnerung in der Karolingerzeit (Darmstadt, 2002).

223

N A R R AT I V E W O R K S

puts it “lived by the sword” (Matthew, 26.52), to be less than admirable.6 Such negative treatment also is evident in accounts of the First Crusade despite the fact that these men were fighting what was regarded as a just war that had been preached by Pope Urban II for the ostensible purpose of reconquering Jerusalem from the Muslims.7 The latter, not incidentally, were widely understood to be persecuting Christians in horrible ways throughout the Middle East and consequently as deserving of punishment.8 In order for scholars to use these biased Latin sources for the purpose of writing history in general, and military history in particular, it is necessary to understand the ways in which their authors selected particular facts to include in their works or made up “facts” from whole cloth in order to vindicate their various agendas.9 It has been known for a very long time that the Latin histories, which provide narratives concerning the First Crusade, cannot be read naively as plain text.10 Nevertheless, even recent studies, despite this long tradition of criticism, have engaged in problematic methods.11 For example, one scholar in a recent study has tended to focus on the sensational highlights of implausible reports of supposed large-scale cannibalism by the Crusaders and exaggerated account of mass killing following the capture of Jerusalem.12 In the latter case, he does so despite an extended critique of these exaggerations and contradictions in the Latin texts and a sober treatment of the often-neglected Hebrew sources.13 For those who would follow John France’s own work, it is necessary to evaluate critically these narratives that provide information in regard to military history. One of several methodological approaches that has been found effective is the deployment of Sachkritik, which makes clear how narrative sources must be 6 See collection of essays in The Peace of God: Social Violence and Religious Response in France around the Year 1000, ed. Thomas Head and Richard Landes (Ithaca, 1992), but cf. Georges Duby, The Three Orders: Feudal Society Imagined, trans. Arthur Goldhammer (Chicago, 1980). 7 Elizabeth Siberry, Criticism of Crusading, 1095–1274 (Oxford, 1985), 72–5. 8 See the extended discussion of the historiography by Benjamin Z. Kedar, “The Jerusalem Massacre of July 1099 in the Western Historiography of the Crusades,” Crusades 3 (2004), 15–75; and John France, “Siege Conventions in Western Europe and the Latin East,” in War and Peace in the Ancient and Medieval World, ed. John France and Philip de Souza (Cambridge, 2008), 158–72. 9 See, for example, Bernard S. Bachrach, “The Crusader March from Dorylaion to Herakleia,” in Shipping, Trade and Crusade in the Medieval Mediterranean: Studies in Honour of John Pryor, ed. Ruthy Gertwagen and Elizabeth Jeffreys (Farnham, 2012), 231–54. 10 The tradition is traced to Heinrich von Sybel, Geschichte des ersten Kreutzzug (Düsseldorf, 1841). See Kedar, “The Jerusalem Massacre,” 15–75. 11 See Leopold Auer, “Krieg und Fehde als Mittel der Konfliktlösung im Mittelalter,” in Bericht über den achtzehnten österreichischen Historikertag in Linz (Salzburg, 1991), 231–8; and Peter Frankopan, The First Crusade: The Call from the East (London, 2010), 26–100, who provides an excellent critique of the Greek sources, particularly the work of Anna Comnena, but tends to treat the Latin source with a less critical eye. 12 Jay Rubenstein, Armies of Heaven: The First Crusade and the Quest for the Apocalypse (New York, 2011), 151–2, 240–1 regarding cannibalism, and 290–6 regarding mass killing at Jerusalem. 13 Kedar, “The Jerusalem Massacre,” 15–75.

224

R A L P H O F C A E N A S A M I L I TA RY H I S T O R I A N

examined in light of the physical realities that they purport to treat.14 Not only must each text be evaluated individually in light of the author’s bias but any particular information that is to be accepted as evidence must be carefully scrutinized. Recent claims that “equal weight” is given to all of the texts that provide information regarding the First Crusade are to be rejected.15

Ralph of Caen Ralph of Caen’s Gesta Tancredi has a very strong claim to be given serious consideration in regard to providing accurate information concerning military operations undertaken during the course of the First Crusade and subsequently in the early years of the principality of Antioch. Traditionally, however, the Gesta, perhaps because of its difficult Latin and offputting prosimetric style, has tended to be given far less attention by modern scholars than it deserves.16 As Susan Edgington observed, the Gesta is the least studied and least well known of all of the major contemporary sources for the First Crusade.17 Because of this lack of study, many of the details uniquely treated by Ralph have remained obscure, and this, too, may have hampered its use for the writing of military history.18 However, even very recent monographs dealing with the military history of the First Crusade continue to give little attention to Ralph’s work despite the fact that it is now available in English translation with substantial notes regarding many once-obscure details.19 Ralph was not an eyewitness to military operations during the First Crusade. However, like many other contemporary writers who also were not eyewitnesses but whose accounts have received far more attention than the Gesta Tancredi, Ralph’s text has far greater claim to our attention than traditionally has been 14 For introductory observations, see Bernard S. Bachrach, “Early Medieval Military Demography: Some Observations on the Methods of Hans Delbrück,” in The Circle of War in the Middle Ages, ed. Donald J. Kagay and L. J. Andrew Villalon (Woodbridge, 1999), 3–20; and for the application of Sachkritik to the crusade narrative sources, see Idem, “Some Observations on the Administration and Logistics of the Siege of Nicaea,” War in History 12 (2005), 249–77. 15 For the quotation, see Rubenstein, Armies of Heaven, 341; but cf. Kedar, “The Jerusalem Massacre,” 15–75, who makes clear that “equal weight” is not merited. 16 See, for example, Yuval Noah Harari, “Eyewitnessing in Accounts of the First Crusade: The Gesta Francorum and Other Contemporary Narratives,” Crusades 3 (2004), 77–99, who also treats non-eyewitnesses but ignores Ralph. Cf. Luigi Russo, “Tancredi e i Bizantini. Sui Gesta Tancredi in expeditione hierosolymitana di Rodolfo di Caen,” Medioevo Greco 2 (2002), 193–230. 17 Susan Edgington, “The First Crusade: Reviewing the Evidence,” in The First Crusade: Origins and Impact, ed. Jonathan Phillips (Manchester, 1997), 57–77. 18 See the notes in Gesta Tancredi, trans. Bachrach and Bachrach, passim, and 13–14 regarding independent information. 19 Rubenstein, Armies of God, 343–85, relies on Ralph fewer than 20 times in more than 350 notes. For Frankopan’s use of Ralph, see The First Crusade, 127–8, 158, 160. Also see the very recent discussion of violence on the First Crusade by David Crispin, “Ihr got kämpft jeden Tag für Sie”: Krieg, Gewalt und religiöse Vorstellungen in der Frühzeit der Kreuzzüge (19095–1187) (Leiden, 2019), who ignores Ralph altogether.

225

N A R R AT I V E W O R K S

recognized. Ralph was a close associate of three very important men who played major roles in the First Crusade. Arnulf of Chocques, who was Ralph’s teacher, initially served as chaplain of the Norman duke Robert Curthose. He became episcopal leader of the Crusade following the death of the papal legate Bishop Adhémar of Le Puy, and subsequently was chosen as the first Latin patriarch of Jerusalem. In addition, Arnulf likely was the editor of the Gesta and the man to whom Ralph dedicated the text.20 The two other leaders of the First Crusade with whom Ralph had very close relations and from whom he claims to have obtained eyewitness information were Tancred, the protagonist of the Gesta itself, and Bohemond de Hauteville, the latter’s uncle and first prince of Antioch. Beginning in 1106, Ralph likely served as the personal chaplain in Bohemond’s entourage during the prince’s travels in the West and during his military operations against the Byzantine Empire.21 After Tancred succeeded Bohemond as ruler of Antioch, Ralph became the former’s personal chaplain.22 In addition, various men who served in the military households of Bohemond and Tancred are said by Ralph to have made themselves available to him while was researching the Gesta. These men also were in a position to provide eyewitness information.23 Ralph came from the region of Caen in Normandy, where his family had wellestablished contacts at a high social level. During the late 1080s and early 1090s, Ralph studied at Caen’s cathedral school with the previously mentioned Arnulf of Chocques before the latter joined Robert Curthose for the iter to the East in 1096. Arnulf’s prominence is evidenced by his relations with the leaders of Norman society. In addition to serving as Duke Robert’s chaplain, Arnulf was earlier attached to the convent of Trinité of Caen, where he had been tutor to William the Conqueror’s daughter Cecilia, probably also to her younger sister, Adela of Blois, and perhaps to some of the English king’s grandchildren.24 As has been ascertained quite easily from Ralph’s writing, he had a broadly based education in the liberal arts that progressed well beyond mastery of the trivium and quadrivium. In regard to Ralph’s intellectual accomplishments, it is evident that he was very well read in the classics of Latin literature, with command of Ovid, Vergil, and Horace. In addition, he was well served in regard to preparation for writing the Gesta by his mastery of the works of Roman historians, 20 Regarding Ralph’s relations with Arnulf, see Russo, “Tancredi,” 214; and concerning Arnulf, see William Aird, Robert Curthose: Duke of Normandy, c. 1050–1134 (Woodbridge, 2008), 165, 168–9. 21 Nicholas L. Paul, “A Warlord’s Wisdom: Literacy and Propaganda at the Time of the First Crusade,” Speculum 85 (2010), 534–66, here 559 for Ralph as Bohemond’s chaplain. 22 See Jean Flori, Bohemond d’Antioche: Chevalier d’Aventure (Paris, 2007); Luigi Russo, Bohemondo: figlio de Guiscardo e principe di Antiocha (Avellino, 2009); and Robert L. Nicholson, Tancred: A Study of His Career and Work in their Relation to the First Crusade and the Establishment of the Latin States in Syria and Palestine (Chicago, 1940). 23 Gesta Tancredi, trans. Bachrach and Bachrach, 3–4. 24 Kimberly A. LoPrete, Adela of Blois, Countess and Lord (c. 1067–1137) (Dublin, 2007), 31.

226

R A L P H O F C A E N A S A M I L I TA RY H I S T O R I A N

such as Livy, Caesar, Lucan, and Sallust.25 Further research likely will be able to show that he had command also of the Latin translations of Josephus’s History and Jewish Antiquities. These were widely disseminated throughout the cathedral and monastic libraries in the West. Ralph’s contemporary, Baudri of Bourgueil, used Josephus’s work in his own history of the First Crusade.26

Ralph’s philosophy of history Ralph possessed a sophisticated view of the historian’s craft as it had developed in the West over more than fifteen centuries from the ancient Greek writers, e.g. Herodotus and Thucydides, through Roman authors, such as Livy and Caesar, to his own time.27 Ralph was influenced greatly by Isidore of Seville (d. 635), who conveyed to medieval writers the key imperative, highly valued by the ancients, that it was the historian’s duty to tell readers what “really happened” and to make sure that he got the “facts right.”28 Isidore’s teaching on history was particularly important among contemporary Norman writers throughout the eleventh century, when Ralph was in school.29 Therefore, it is hardly surprising that Ralph maintained a philosophy of history that was in consonance with what he had learned from his study of Roman writers and the teaching of Isidore. Of particular importance to Ralph, in addition to getting the facts right, was his emphasis on the primacy of information that was provided by those who were eyewitnesses to events. This information was to be especially prized and given precedence over secondhand accounts. By contrast, Ralph’s contemporaries Baudri of Bourgueil, Robert the Monk, Guibert of Nogent, and to a much lesser extent Albert of Aachen relied primarily on “published” accounts, which they greatly embellished, with the result of providing their readers with inaccurate information concerning what really happened but which served to advance the various agendas of the authors.30 It was Ralph’s aim to tell the story of Tancred’s military career. Nevertheless, he claims that he made clear to Tancred that he would not make public the Gesta

25 See Laetitia Boehm, “Die ‘Gesta Tancredi’ des Radulf von Caen: Ein Beitrag zur Geschichtsschreibung der Normannen um 1100,” Historisches Jahrbuch 75 (1956), 47–72; and Henri Glaesener, “Raoul de Caen: historien et écrivain,” Revue d’histoire ecclésiastique 46 (1951), 5–21. 26 Gesta Tancredi, trans. Bachrach and Bachrach, 5. 27 Ibid., but cf. Ernst Breisach, Historiography, Ancient, Medieval and Modern, 3rd edition (Chicago, 2007). 28 Isidore of Seville, Etymologiarum sive originum libri xx, ed. W. M. Lindsay 2 vols. (Oxford, 1911), I: 40. 29 See Leah Shopkow, History and Community: Norman Historical Writing in the Eleventh and Twelfth Centuries (Washington, DC, 1997), 20–1. 30 Edgington, “First Crusade,” 57–77; and idem, “Albert of Aachen and the Chansons de Geste,” in The Crusades and their Sources, 23–37. In regard to Guibert, see Harari, “Eyewitnessing,” 91–5. See Bachrach, “The Crusader March,” 231–54, for a detailed examination of the distortion of information by various crusade sources.

227

N A R R AT I V E W O R K S

until the prince of Antioch himself had died (preface). It was Ralph’s view that if he were to make public the text while Tancred were living, contemporaries would complain that he had manipulated the facts to please the prince and thereby obtained unmerited benefits (preface). Ralph, however, knew that when he wrote or at least made public the Gesta, c. 1112, i.e. after Tancred’s death, many contemporaries were not only aware of many of the facts that he would be reporting but also had participated as eyewitnesses in the events that he planned to treat. This list not only included very important secular leaders such as King Baldwin I of Jerusalem but also Arnulf of Chocques, then Jerusalem’s patriarch, who had had his own conflicts with the de Hauteville leaders (chs. 135–36). Despite Ralph’s vigorous effort to emphasize that he was writing fact and not fiction, the Gesta is a prosimetric work, i.e. part prose and part poetry. Ralph certainly was aware that men of his background and intellectual training roundly condemned the use of poetry to write history. Ralph’s older contemporary, William of Poitiers, while criticizing works written about William the Conqueror made this point very sharply when he observed that poets “are permitted to expand their knowledge in any way they liked by roaming through the fields of fiction.” As a result, William the Conqueror’s biographer rejected such sources as untrustworthy.31 A century prior to Ralph’s efforts, Dudo of Saint Quentin had used a prosimetric style to write De moribus et actis primorum Normanniae ducum.32 Dudo made clear that the prose sections of De moribus were intended to provide an accurate depiction of military organization and operations as these were understood c. 1000 by the duke and the leaders of his military household, while the poetry was intended ostensibly for various types of commentary, and perhaps even to show off his literary versatility.33 The great popularity of De moribus in Normandy during the eleventh century leaves no doubt that the text itself, as well as the works of those who used it, e.g. William of Jumièges, not only were read but even diligently studied.34 As will be seen later, it seems likely that Dudo’s prosimetric approached influenced Ralph. In this context, Ralph, who was thoroughly familiar with the use of poetry in historical texts, indicates in his own work that when he wrote poetry the truth value of his account was not guaranteed. Rather, he treated poetically events concerning which he lacked sufficient information to provide a factual account, e.g. chapters 27–32. Therefore, in the sections that were written in verse, he was following the 31 Gesta Guillelmi of William of Poitiers, ed. and trans. R. H. C. Davis and Marjorie Chibnall (Oxford, 1998), book 1, chapter 20 (28–9), for the translation as slightly altered by the authors. 32 Dudo of Saint-Quentin, De moribus et actis primorum Normanniae ducum, ed. Jules Lair (Caen, 1865). 33 See Bernard S. Bachrach, “Writing Latin History for a Lay Audience c. 1000: Dudo of Saint Quentin at the Norman Court,” Haskins Society Journal 20 (2008), 58–77. 34 Regarding Dudo’s great popularity, see Gerda H. Huisman, “Notes on the Manuscript Tradition of Dudo of St. Quentin’s Gesta Normannorum,” Anglo-Norman Studies 6 (1984), 122–35, here 124.

228

R A L P H O F C A E N A S A M I L I TA RY H I S T O R I A N

tradition of literature that was intended for entertainment purposes, i.e. the genre of the chanson de geste. In fact, at one point, after providing a poetic account of a battle in which he elucidated what he regarded to be epic values, Ralph concluded, “You would say that Roland and Oliver had been reborn if you saw the raging of the count, this one with a spear and that one with a sword” (ch. 30). By contrast, when Ralph wrote in prose, he was confident that he knew the facts.35

The battle of Dorylaeum: phase one John France’s reconstruction of the battle of Dorylaeum is the most nuanced and complete account of all modern scholarly efforts to describe this key encounter between the Crusader army and that of Kilij Arslan early in July 1097.36 As John makes clear, a matrix of contemporary and near-contemporary narrative sources provide information, often contradictory or confused information, concerning the battle. Therefore, scholars have found it necessary to make critical choices concerning which information may, in fact, be used as evidence for what actually happened and which information is to be rejected.37 It is in this context that, with limited space available, we review Ralph’s account of the first phase of the battle. It is our aim to ascertain whether Ralph’s contribution regarding military operations at Dorylaeum has potential for illuminating some key points that previously have not been stressed. The now-accepted treatment of the first phase of the battle suggests that Kilij Arslan, the Seljuk Turkish leader, commanded a large army composed overwhelmingly of mounted archers. This force enjoyed “numeric superiority . . . over the knights of the vanguard,” led by Bohemond and Duke Robert of Normandy.38 It was the Turkish strategy to take the Crusaders of the vanguard by surprise and win a victory in a “mobile battle” over this force of heavily armed mounted troops, which was outnumbered. However, according to this reconstruction, “the crusaders were alert and their foot soldiers prepared to pitch camp while an element of the knights confronted the enemy and were put to flight.” Following this retreat, these heavily armed mounted troops deployed in a solid formation and held the attacking force at bay until relief arrived from the main body of the army.39

35 Cf. Natasha Hodgson, “Reinventing Normans as Crusaders? Ralph of Caen’s Gesta Tancredi,” Anglo-Norman Studies 30 (2007), 117–32, who tries to rehabilitate for historical purposes the poetic elements of the Gesta (124–6) but who does not give contemporary Norman hostility to this type of “history” writing its due. It should be emphasized with regard to Hodgson’s aims that poetry, where the facts are in doubt, can also be used for purposes of studying the ideas in which she has an interest. Also note the curious case of the Chanson d’Antioche, discussed by Edgington, “Albert of Aachen,” 23–37. 36 France, Victory in the East, 169–85; Cf. Bernard S. Bachrach, “Crusader Logistics from Victory at Nicaea to resupply at Dorylaion,” in Logistics and Warfare in the Ages of the Crusades, ed. John H. Pryor (Aldershot, 2006), 43–62. 37 France, Victory in the East, 169–81. 38 Ibid., 182. 39 Ibid.

229

N A R R AT I V E W O R K S

Ralph’s potential contribution In light of Ralph’s sources, it is important that he emphasizes that the Crusader army marched at night, not during the day (ch. 20). Generally, this point is not given sufficient attention in discussions either of the course of the battle and/or the chronology of the march to Dorylaeum.40 In addition, this Crusader force was out in the open, i.e. nowhere near the marching camp that had been used the previous day. Therefore, this entire force, and not only the heavily armed horsemen, was highly vulnerable to Kilij Arslan’s mounted archers, who could shower the line of march with arrows from a considerable distance. Upon discovering that they were in danger of being attacked in the early morning on what would be the first day of the battle, the “foot soldiers prepared to pitch camp,” which was “made quickly” and played a key role in blunting the Turkish attack.41 According to Ralph, Bohemond’s force had defended the baggage train of the “lesser folk” and this may account, in large part, for the defenses that the foot soldiers initially constructed (ch. 22). Ralph claims that Tancred was in command of the scouting party that first discovered the enemy force early that morning (ch. 21). This information, like much else we are told regarding Tancred’s activities and which is not to be found in other sources, requires some discussion. Despite Ralph’s claim that he was guided by the facts and that, in fact, he published after Tancred’s death to avoid criticism of favoritism, the possibility cannot be ignored that the protagonist was given credit that he did not merit. On the whole, Ralph depicts Tancred as being deployed in the most dangerous tasks, whether with the rearguard or in the vanguard. This, of course, was fully consistent with his role as Bohemond’s second in command, but also placed Tancred in the positions of greatest danger.42 As a consequence, Ralph’s parti pris with regard to demonstrating Tancred’s remarkable deeds would appear to have been consistent with contemporary military practice. It is notable in this context that other contemporary, and independent, sources present Tancred undertaking similar tasks and seeking the most dangerous positions, as seen in the prose portions of Ralph’s Gesta.43 Therefore, if Ralph were, in fact, fabricating information regarding Tancred – and there is good reason to believe that he was not acting in such a manner – such fabrications would have been consistent with the way that Tancred was presented by others both on the field of battle, and more generally, as the ruler of Antioch while Bohemond was held captive and also after the latter died. Ralph’s purpose, in his own words, was to provide details of his protagonist’s activities that had not been covered adequately by others, not to create a story from whole cloth.

40 41 42 43

Cf. France, Victory in the East, 175. Ibid., 182. Gesta Tancredi, trans. Bachrach and Bachrach, passim. See the dated but still useful monograph by Nicholson, Tancred; and Thomas S. Asbridge, The Creation of the Principality of Antioch, 1098–1130 (Woodbridge, 2000).

230

R A L P H O F C A E N A S A M I L I TA RY H I S T O R I A N

Following Tancred’s discovery of the enemy, Ralph’s account differs from the scholarly synthesis, as he indicates that there was no noteworthy engagement at this time between the crusading force and the Turks. Rather, the former made its camp on one side of a river, left unnamed in the Gesta, and the latter was encamped on the other side (ch. 21). The emphasis by France that the Crusaders made their camp with its right bank protected by a marsh likely supports Ralph’s observation that a river of sorts was between the two forces, as significant inland marshland rarely exists without a fresh water source.44 In addition, the ever-present need for fresh water for men, female camp followers, and particularly animals strongly supports Ralph’s assertion that the Crusaders encamped near such a source. Indeed, the presence of Byzantine guides with the Crusader army likely facilitated the choice of a route that was the most likely to provide fresh sources of water. According to Ralph, it was as a result of Tancred’s early warning that Kilij Arslan’s plan for a surprise attack was thwarted and that the opening of hostilities was delayed until much later in the day (chs. 21, 23). Once the Crusader force was ensconced in its encampment, and contrary to the traditional scholarly view, Ralph claims the Normans spent the remainder of the day resting and very likely sleeping, which obviously was a necessity in light of the fact that they had been on the march throughout the previous night.45 Their war horses undoubtedly needed rest as well as food and water after the nightlong march. At this time, Ralph reports that groups of foot soldiers worked on improving the fortified camp that initially had been developed from the vehicles of the baggage train. Among the refinements made to the encampment were a screen of sharpened wooden stakes (ch. 23). These were probably fashioned from wood obtained by breaking up some of the baggage wagons or carts, as a force of some 20,000 effectives and assorted camp followers required numerous vehicles to transport its food and equipment.46 Toward dusk, according to Ralph, when the sun was setting and the air was cooling, a now-rested unit of heavily armed mounted troops was sent forward to attack a large force of Muslim horse archers. From Ralph’s description, it is clear that these Muslims were deployed in a mass forward of the Turkish encampment and on the Crusaders’ side of the river. The Norman attack was intended to dislodge the Turks and drive them back across the river and thereby put the Crusader camp outside the effective range of the Turkish bows. Initially, the Crusaders inflicted heavy casualties as they drove through the lines of the enemy archers, 44 France, Victory in the East, 182 regarding the marsh. 45 The claim by Fulcher of Chartres that the battle began during the first hour of the day or shortly thereafter and lasted until the sixth hour should be taken to refer to the morning of the second day as seen from the perspective of the Norman vanguard. See Fulcher of Chatres, Historia Hierosolymitana, ed. H. Hagenmayer (Heidelberg, 1913), book 1, chapter 11. The claim by the Anonymous that the battle lasted from the third to the ninth hour provides the perspective of the relief force on the second day. See Gesta Francorum et aliorum Hierosolimianorum, ed. and trans. Rosalind Hill (Oxford, 1967), 20. 46 For the numbers, see France, Victory in the East, 169.

231

N A R R AT I V E W O R K S

who likely had been taken by surprise. However, the Turks retreated rapidly and scattered, as it was their tactical imperative to avoid hand-to-hand combat with the heavily armed mounted troops. Once outside the reach of the slower-moving Norman cavalry, the archers reformed and launched intensive barrages of arrows, which forced the Crusaders to return to their own lines (ch. 21). It may be suggested that if Ralph had been interested in raising Tancred’s profile by giving him credit where none was due, he easily could have projected him as part of this initially successful attacking force. It is notable, therefore, that Ralph does not mention Tancred’s participation in this phase of the battle. The initial attack by the Norman horsemen against the Turkish archers was followed up before sundown, according to Ralph, by a two-pronged attack executed by heavily armed mounted troops, who were led by Bohemond and Duke Robert of Normandy, respectively. The advance was supported at the rear by a tightly packed phalanx of foot soldiers armed with spears, which moved up slowly behind the cavalry. The foot soldiers were deployed in order to provide the horsemen with protection should a retreat become necessary (ch. 22). Such a deployment was a commonly used tactic that had been developed for combined cavalry–infantry forces, which lacked the high levels of “fire power” that were available to mounted archers.47 It may be suggested that the Crusaders were influenced in their decision to use this tactic by Tatikios, the Byzantine military advisor to the Crusaders, who is thought to have accompanied the vanguard.48 The tactics also possibly could have been suggested by one of the Byzantine officers among the “foreign troops” whom, Ralph notes, were accompanying Bohemond’s forces at this time (ch. 20). The Norman attack, like that of the initial assault, discussed earlier, apparently enjoyed some initial success. However, the great numerical superiority of the Turkish mounted archers made the trading off of casualties an unsustainable strategy. As a result, Duke Robert initiated a retreat toward the Crusader camp, and he was followed by Bohemond. The retreat was halted when the mounted troops reached the phalanx of their own foot soldiers. The latter had stood firm, presenting a front of spears. This tactic, however, required that the infantry open its ranks so that the retreating horsemen could pass through. It seems likely that the foot soldiers, or at least many of them, lacked sufficient training to execute this maneuver in a proper manner. As a result, additional men, both mounted troops and foot soldiers, were lost to the Turks (ch. 22). Again, Ralph does not give Tancred a role in this operation. Once Robert’s and Bohemond’s forces had returned to the Crusader encampment, Ralph credits the Norman duke with suggesting to his colleague that they should not think any further in terms of retreat but should stand their ground and protect the camp. The account of this exchange, provided as direct discourse by Ralph, surely is not a verbatim version of the conversation, if only because 47 See Bernard S. Bachrach, A History of the Alans in the West (Minneapolis, 1973), 126–32. 48 France, Victory in the East, 175, hypothesizes that Tatikios was with the vanguard.

232

R A L P H O F C A E N A S A M I L I TA RY H I S T O R I A N

it is reported in Latin rather than in Norman French. However, because Ralph had access to both Bohemond and Arnulf, the duke of Normandy’s chaplain, the essence of the discussion between the two Norman leaders should be considered by modern scholars as essentially accurate. In any case, Robert’s advice was taken, and a screen of Norman mounted troops was deployed in front of at least part of the Crusader camp to await a Muslim attack (ch. 22). At this time, that is in the early evening of the first day of the battle, Ralph indicates that Robert and Bohemond sent a messenger to the main part of the army, which as yet had not been informed that the vanguard was under attack. They urged that a relief force be sent. This information is very different from that provided in the other sources, which have been taken to indicate that immediately upon seeing the enemy, the leaders of the vanguard sent messengers for help.49 Because the request for aid was made, according to Ralph, at dusk or shortly thereafter, it is very likely that Bohemond and Robert understood that a relief force would not be sent before sun-up. This makes clear why the author of the Gesta Francorum indicates that hostilities began during the third hour, at about nine the following morning.50 After nightfall, while Bohemond’s and Robert’s mounted troops stood guard, a force of Muslim spearmen on foot slipped past the cavalry screen and penetrated that part of the Crusader encampment where the Crusader infantry forces were deployed and bivouacked (ch. 23). Ralph notes the “great number” of foot soldiers but makes clear that he had little regard for their quality as fighting men. He claimed that they “lacked military ability.” In any case, Ralph highlights the success of this Muslim attack, as he reports that the Crusaders suffered large numbers of casualties, that this part of the encampment was looted, and that numerous prisoners were carried off to be sold as slaves (ch. 23). Ralph concludes, undoubtedly with some exaggeration, that “not ever in all of the events of war that preceded or followed this one was so much Latin blood shed by gentile swords” (ch. 23). During this slaughter, the Norman leaders remained firm in their decision to hold their position and neither to attack the Muslims as they were in the process of penetrating part of the Crusader encampment, nor to pursue them when they withdrew. Ralph’s account permits the inference that he believed, very likely reflecting the opinion of the far more offensive-minded Tancred, that a counterattack by the main body of the Norman mounted troops against the Muslims heavily laden with booty and prisoners – and thus necessarily returning slowly to their own lines – could have deprived them of their gains before they reached the safety of their own encampment (ch. 23). While the Muslim troops penetrated the Norman defenses and the screen of mounted men did nothing either to stop them or to recover the booty and prisoners, Ralph makes clear that Tancred, who was deployed on the other side of the camp, was able to take effective action. At this time, Muslim forces took control 49 Cf. France, Victory in the East, 175. 50 Gesta Francorum, 20, as discussed earlier.

233

N A R R AT I V E W O R K S

of a small hill facing the Crusader camp and were positioned to shower with arrows the area where Tancred and his men, including his brother William, were deployed. Upon obtaining this information, Tancred requested from Bohemond permission to attack the enemy position. This point indicates that at this time the chain of command was effective. As Ralph makes clear, it was Bohemond’s plan to remain on the defensive in accord with his agreement with Duke Robert and to maintain his mounted forces intact in order to fight off the Turks. Tancred, however, convinced Bohemond that he should be allowed to attack. In a rapid advance Tancred’s troops took the hillock and drove off the enemy (ch. 24). When Tancred’s force took the high ground, a breach was established in the Muslim effort to encircle the Crusader camp. As a result, Kilij Arslan ordered a counterattack in order to retake this key position as part of his plan to cut off the vanguard from reinforcements should aid arrive and thereby restrict their movement. In accord with Kilij Arslan’s orders, Turkish forces rushed up the side of the hillock in large numbers, but the superiority of the arms and armor of the Norman troops caused immense casualties in hand-to-hand combat, with the result that the Muslims were driven back. After several such failed sorties, which were intended to retake the little hill rapidly, the Muslim commander decided to change tactics. He had his forces stand off at a distance and shower the Normans with barrages of arrows (chs. 24–26). Finally, near daybreak, having held the hillock for the better part of the night, the barrages of arrows drove Tancred and his men from their position. As a result, Kilij Arslan’s forces were able to close the breach that Tancred’s attack had opened in the Muslim lines and to resume their effort to encircle the Crusader camp. This battle, which Ralph sees as sufficiently important to describe over the course of three chapters, likely because it both featured the efforts of Tancred and an all-night battle, has not been given sufficient attention in traditional accounts of the battle of Dorylaeum. The death of Tancred’s younger brother William in this part of the battle also is noteworthy, as it illuminates the potential that significant numbers of casualties were suffered by the Normans in this phase of the battle (chs. 24–26). At dawn, that is during the first hour of the day, with Tancred’s troops having been driven from the hillock beyond the boundaries of the Crusader encampment, Kilij Arslan’s forces began to deploy for an all-out attack.51 At this point, however, Ralph claims that the Crusader relief was seen because of the dust cloud that it raised in the distance (ch. 26). Nevertheless, the actual fighting between the Crusader relief force and the Muslims likely began only during the third hour, as Anonymous claims.52 Between the time that the vanguard and the Muslims saw the dust cloud, it likely took until about the third hour for the relief force to reach the field, scout the situation, and deploy fully. The ensuing combat with Kilij

51 Fulcher of Chartres, Historia, book 1, chapter 11. 52 Gesta Francorum, 20.

234

R A L P H O F C A E N A S A M I L I TA RY H I S T O R I A N

Arslan’s forces, in our reconstruction based on Ralph’s text, begins the second phase of the battle, which ended with a Crusader victory at about three in the afternoon when the Turks retreated and left the field, as well as their own camp, to the Christians.

Some conclusions In light of Ralph’s highly placed sources, it seems reasonable to conclude that the account of the military operations on the first day of the battle as presented in the Gesta is accurate. It is important that the Crusaders, at least those of the vanguard, marched during the night and not during the day. Thus, the vanguard’s first contact with the enemy by Tancred’s scouting party was around dawn or shortly thereafter, i.e. after a night on the march. After this contact, the vanguard established camp near a water source and rested throughout most of the remainder of the day. Hostilities were initiated early in the evening with two attacks by the Norman mounted troops, which failed to drive the Turks across the river and thus away from the Crusader encampment. A decision then was made by Duke Robert and Bohemond to remain on the defensive and send for help. Through the night, the cavalry remained in position but failed to deal with a successful penetration of the camp by Muslim foot soldiers. Subsequently, Tancred captured a key Muslim position and held it through the night. This action delayed an all-out night attack on the Crusader camp. When Tancred was forced to retreat, the Turks began to deploy for an all-out attack. At dawn, however, the approach of a Crusader relief force was spotted and what we have characterized as the first phase of the battle of Dorylaeum came to an end.

235

Part 2 MATERIAL SOURCES AND IMAGES

11 LANDSCAPES OF DEFENSE At the nexus of archaeology and history in the early Middle Ages

It is axiomatic that for some two generations, the study of the military history of pre-Crusade Europe has been one of the major academic casualties of the Second World War.1 This conscious neglect is obvious despite the uncontroversial fact that preparation for war, war itself, and its aftermath consumed an enormous part of the surplus human and material resources in most regions of the erstwhile Roman imperial West throughout the early Middle Ages, particularly in the construction and maintenance of fortifications and related infrastructural elements, including roads and bridges. However, following Western Europe’s recovery from World War II, diligent efforts by archaeologists, especially in Britain and Germany, with important contributions also from France and Italy, have resulted in a massive increase in knowledge about medieval fortifications, as demonstrated quite clearly by the volume of essays under consideration here. As a result of this vast and increasing body of research, the construction of fortifications has come to play a role in some historians’ efforts to grasp the magnitude of the costs of war, broadly understood, and the sophisticated administrative organization required to undertake such projects on a large scale successfully.2

1 The lack of proper attention to military history is made clear by Hans-Werner Goetz, “Social and Military Institutions,” in The New Cambridge Medieval History: c. 700–c. 900, ed. Rosamond McKitterick (Cambridge, 1995), 451–80, esp. 479–80, where the subject of military organization receives very little attention. Nb. Ead, Charlemagne: The Formation of a European Identity (Cambridge, 2008), in her otherwise excellent book neglects both military organization and warfare in her effort to craft a picture of Carolingian identity despite the fact that Charlemagne and his government were overwhelming focused on military matters. This neglect of military affairs is particularly evident in German scholarship. See, for example, the discussion by Hans-Hennig Körtum, Der Krieg im Mittelalter als Gegenstand der Historischen Kulturwissenschaften: eine Annäherung, in Krieg im Mittelalter, ed. Idem (Berlin, 2001), 13–43. 2 See, for example, Bernard S. Bachrach, “The Cost of Castle-Building: The Case of the Tower at Langeais, 992–994,” in The Medieval Castle. Romance and Reality, ed. Kathrin L. Reyerson and Faye Powe (Dubuque, IA, 1984), 46–62; and reprinted in Bernard S. Bachrach, Warfare and Military Organization in Pre-Crusade Europe (London, 2002) (with the same pagination; and Idem, “The Fortification of Gaul and the Economy of the Third and Fourth Centuries,” Journal of Late Antiquity 3.1 (2010), 38–64. For a study centered on the chronological period under consideration

239

M AT E R I A L S O U R C E S A N D I M A G E S

Of growing importance to our understanding of the role of fortifications in early medieval history have been the efforts of scholars working in the new tradition of landscape archaeology, largely dominated at present by British specialists, to synthesize the findings from a broad array of fields in order to decipher patterns of human agency that are embedded in the landscape over relatively wide geographical spaces. Archaeologists working within this new investigative paradigm are in the process of helping us understand the undertaking of what would appear to have been long-term defensive military strategy in Anglo-Saxon England.3 In large part, this work in landscape archaeology has been developed on the basis of advances that have been made in the study of individual fortifications along with roads and road systems, as well as ancillary sites such as watch towers and fire beacons.4 In addition, the intensive study of place names that are connected with military installations of all types as well as with roads, which would seem to have had an important, if not a primarily military, purpose, has been used to explain how defensive systems were developed.5 Specialists within the new investigative paradigm argue on the basis of their findings that the landscape can be read in a manner similar to the ways in which historians read written texts.6 One central and inescapable result of this research is to highlight the role of government officials and military strategists in constructing and carrying out large-scale plans, such as the Burghal Hidage system, which was expanded substantially by King Alfred of Wessex († 899).7 As a result of the work of archaeologists, including those who focus on the “landscape,” specialists in early medieval history now are required, whether or not they are yet aware of this imperative, to integrate many of these aspects of military construction into their view of early medieval government, demography, and economy. It is necessary for historians to recognize that neither natural terrain nor manmade military topography can be ignored in regard to our understanding of the role of government in sophisticated military planning. In large part due to the work of archaeologists and especially landscape archaeologists, the death knell has been sounded for those historians who see the governments of Rome’s successor states in the West and their administrative institutions as primitive or prestate manifestations of a long-supposed Dark Age. We are now long past the point where any scholar familiar with the exceptionally large and growing corpus

3

4 5 6 7

here, see Bernard S. Bachrach and David S. Bachrach, “The Costs of Fortress Construction in TenthCentury Germany: The Case of Hildagsburg,” Viator 45.3 (2014), 25–58. For example, John Baker and Stuart Brookes, “Anglo-Saxon Civil Defence: Theory and Historical Context,” in Beyond the Burghal Hidage. Anglo-Saxon Civil Defence in the Viking Age (Leiden, 2013), 1–41, provide something of an introduction to the value of discussing “civil defense” and the place of civil defense in state-formation theory. Ibid., 48. Ibid., 49. This is the gravamen of the entire collection of the essays discussed in this review. See the very useful collection of essays in The Defence of Wessex: The Burghal Hidage and AngloSaxon Fortifications, ed. David Hill and Alexander R. Rumble (Manchester, 1996).

240

LANDSCAPES OF DEFENSE

of archaeological studies dealing with military installations can still opine about the destruction of ancient civilization and its replacement with a barbarian warrior culture modeled on epic fantasies such as Beowulf or putatively more accurate historical works such as Tacitus’s Germania.8 This process of integrating archaeology and history, however, will not be easy. In some parts of Europe, particularly in the British Isles and in Germany, archaeologists have been hard at work dealing with these artifacts that demonstrated highly organized and very expensive governmental control of fortification strategy.9 It is unfortunate, therefore, that, in general, disciplinary boundaries among academics in both countries are vigorously guarded and overall there is far too little cooperation between archaeologists and historians. In 1963, Walter Schlesinger, chair of medieval history at the University of Marburg, sought to help to bridge the divide between the fields with his investigation of the fortress at Merseburg as a model 8 In this regard, see the discussion by the leading specialist in Ottonian history, Hagen Keller, “Grundlagen ottonischer Königsherrschaft,” in Reich und Kirche vor dem Investiturstreit. Beiträge beim wissenschaftlichen Kolloquium aus Anlaß des 80. Geburtstags von Gerd Tellenbach, ed. Karl Schmid (Sigmaringen, 1985), 17–37, who argues that the material resources of the Ottonian kings were peripheral to their real basis of power, and that their lack of institutional continuity with the Carolingian empire is most obvious with regard to military matters. Keller made these claims despite the well-known efforts by the Ottonians, based upon Carolingian precedent and institutions, to maintain systems of fortifications along their eastern frontiers. In this regard, see Walter Schlesinger, “Burgen und Burgbezirke. Beobachtungen im mitteldeutschen Osten,” in Von Land und Kultur. Beiträge zur Geschichte des mittledeutschen Ostens. In gemeinsamer Arbeit mit Wolfgang Ebert et al., zum 70. Geburtstag Rudolf Kötzschkes, ed. Werner Emmerich (Leipzig, 1937), 158–87; Dietrich Claude,“Der Königshof Frohse,” Blätter für deutsche Landesgeschichte 110 (1974), 29–42; and Berthold Schmidt, “Das Westsaalegebiet im Verband des fränkischen Staates und die Ostexpansion des 9./10. Jahrhunderts,” Zeitschrift für Archäologie 18 (1984), 23–32. Also see David S. Bachrach, “Restructuring the Eastern Frontier: Henry I of Germany 924–936,” Journal of Military History 78.1 (2014), 9–35. With respect to efforts to build Beowulfian models into discussion of early medieval governmental operations, see Steven C. Fanning, “Tacitus, Beowulf and the Comitatus,” The Haskins Society Journal 9 (1997), 17–38, which deserves special attention for showing how generations of scholars wedded to a primitivist approach to the early Middle Ages have misused both Tacitus’ Germania and the notion of the comitatus as a part of an effort to sustain the “warrior culture” myth. 9 The archaeological literature on early medieval fortifications is too vast to summarize here. However, readers can gain a valuable introduction to the foci and methodologies of archaeologists specializing in this topic in Frühmittelalterlicher Burgenbau in Mittel und Osteuropa, ed. Joachim Henning and Alexander Ruttkay (Bonn, 1998); and Europa im 10. Jahrhundert. Archäologie einer Aufbruchszeit, ed. Joachim Henning (Mainz, 2002). Also see the important work of Matthias Hardt, “Linien und Säume. Zonen und Räume an der Ostgrenze des Reiches im frühen und hohen Mittelalter,” in Grenze und Differenz im frühen Mittelalter, ed. Walter Pohl and Helmut Reimitz (Vienna, 2000), 39–56; idem, “Hesse, Elbe, Saale and the Frontiers of the Carolingian Empire,” in The Transformation of Frontiers from Late Antiquity to the Carolingians, ed. Walter Pohl, Ian Wood, and Helmut Reimitz (Leiden, 2001), 219–32; and Idem, “The Limes Saxoniae as part of the Eastern Borderlands of the Frankish and Ottonian-Salian Empire,” in Borders, Barriers, and Ethnogenesis: Frontiers in Late Antiquity and the Middle Ages, ed. Florin Curta (Turnhout, 2005), 35–50, who has demonstrated with great skill the tactical and strategic efforts of both the Saxons and the Franks to utilize systems of fortifications for territorial defense.

241

M AT E R I A L S O U R C E S A N D I M A G E S

and inaugural study for the long-term project on Deutsche Königspfalzen. Here, Schlesinger established an important paradigm for synthesizing both historical and archaeological sources to develop a comprehensive model of the physical structure, purpose, and use of royal palaces, many of which were also fortresses.10 Despite these pioneering efforts in both Germany and elsewhere it remains commonplace that archaeologists often complain that historians are too firmly wedded to their texts, which, in any event, they correctly note are more or less biased and therefore cannot be relied upon to provide accurate information. Historians, in turn, complain that archaeologists are not trained in the methods needed to explicate the written evidence, which, in fact, requires taking into account the textual problems inherent in these sources, including the parti pris of authors for which a sound control both of Latin and philology is required.11 By contrast, many archaeologists would appear to believe that the artifacts speak for themselves and often seem willfully ignorant that their own interpretations of the artifacts likely are biased. Several decades ago, Phillip Grierson highlighted this point when he wrote: “it is said that the spade cannot lie, but it owes this merit in part to the fact that it cannot speak.”12 Landscapes of Defence in Early Medieval Europe, edited by John Baker, a research fellow at the Institute for Name-Studies in the University of Nottingham; Stuart Brookes, research associate in the Institute of Archaeology at University College London; and Andrew Reynolds, professor of medieval archaeology at the Institute of Archaeology, University College London, have brought together thirteen essays by archaeologists and historians that illustrate many aspects of the status questionis that exists at the nexus of early medieval archaeology and history, particularly military history. This volume is one of two studies produced with the help of a grant to Baker and Brookes from the Leverhulme Trust in 2005, which was intended to deepen understanding of Anglo-Saxon warfare and military organization by treating in exceptional detail the military landscape of some parts of Britain. These scholars published an initial volume, Beyond the Burghal Hidage, in 2013, which has much to recommend it.13 Of the thirteen articles published in Landscapes of Defence, eight treat AngloSaxon England. There is one that treats the Low Countries, one that deals with the

10 Walter Schlesinger, “Merseburg–Versuch eines Modells künftiger Pfalzbearbeitungen,” in Deutsche Königspfalzen. Beiträge zu ihrer historischen und archäologischen Erforschung, vol. 1 (Göttingen, 1963), 158–206. 11 See, for example, the discussion of this issue by Bernard S. Bachrach, “Writing Latin History for a Lay Audience c. 1000: Dudo of Saint Quentin at the Norman Court,” The Haskins Society Journal 20 (2008), 58–77; and Idem, “Early Medieval Fortifications in the ‘West’ of France: A Revised Technical Vocabulary,” Technology and Culture 16.4 (1975), 531–69; and reprinted with the same pagination in Bachrach, Warfare and Military Organization. 12 Phillip Grierson, “Commerce in the Dark Ages: A Critique of the Evidence,” Transactions of the Royal Historical Society 5th Series 9 (1959), 123–40, here 129. 13 For an overall positive review of Baker and Brookes, Beyond the Burghal Hidage, see Bernard S. Bachrach, Journal of Military History 78 (2014), 1103–6.

242

LANDSCAPES OF DEFENSE

German-Slavic frontier, one on Scandinavia, and two on Christian Spain. For all intents and purposes, the heartland of Western Europe, or to put it another way, the Frankish regna and empire, are ignored despite the vast numbers of studies that have been made of fortifications and defensive systems constructed both within the borders of the hexagon and to the east between the Rhine and the Weser. In addition, there is no study that looks south of the Alps toward Italy where the investigation of fortifications and systems of defense is a very well-developed tradition, as evidenced, for example, by the work of Ricardo Frankowich.14 On the whole, the volume is Anglo-Saxon centered and recalls the bromide “Channel fogged in, continent isolated.” Collectively, the thrust of this work can be divided into four broad categories. First, it is argued that patterns of fortification can now be more easily detected in the early medieval landscape with the help of a wide variety of modern devices such as ground-penetrating radar. When these patterns are understood in light of the natural topography of a particular area, they serve as evidence for policy or strategic decisions by governments that possessed knowledge of current political situations and initiated plans to shape future behavior. Second, because the military landscape can be read in this way, it provides prima facie evidence for governmental sophistication that merits these polities to be classified as “states” according to traditional anthropological definitions of a state. Third, periods of military crisis provided governments with an opportunity to expand the scope of their authority and to impose both administrative oversight and economic burdens, e.g. taxes and corvée obligations of various types, in a more intensive manner than was possible when external threats are less exigent. These situations also enabled governments to inculcate ideological support for their actions by successfully orchestrating an effective defensive posture in times of peril. Finally, large fortifications served important governmental roles in addition to their military functions. In particular, governments utilized these strongholds as centers of administrative activities, including courts, mints, and tax collection, as well as refuges when enemies invaded the region. We agree that all four of these factors are exceptionally important, although they are treated with different levels of skill in the individual essays. Most of the essays in this useful and provocative volume focus on military landscapes, various ways of identifying in detail the complexity of such landscapes, and the fortifications that comprised such landscapes. However, four essays do not fall into this category and tend to blur the overall focus on military landscape. The lead essay by Reynolds, “Archaeological Correlates for Anglo-Saxon Military Activity in Comparative,” which would appear to have been intended as an introduction to the volume as a whole, focuses on the notion that Anglo-Saxon England was dominated by a warrior culture. His additional goal to “concentrate 14 For a very useful introduction, see Ricardo Francovich, “The Beginnings of Hilltop Villages in Early Medieval Tuscany,” in The Long Morning of Medieval Europe: New Directions in Early Medieval Studies, ed. Michael McCormich and Jennifer R. Davis (Burlington, 2008), 55–82.

243

M AT E R I A L S O U R C E S A N D I M A G E S

on social explanations for military activity (2),” takes this essay even further from the military aspects of landscape archaeology. Reynold’s bias in regard to the study of military matters from a military perspective is illuminated by his personal view that these are “one of the most undesirable features of human behavior (2).” Reynold’s primary focus is not on landscape archaeology but on “burial archaeology,” which he believes provides evidence for the notion that so-called “weapons graves” are evidence that a warrior elite dominated Anglo-Saxon England. It is certainly clear that the decision to bury some men with swords entailed the permanent disposal of items that had substantial intrinsic value both because of the raw metal content of high-quality iron and the high-cost labor that was required to transform this iron into a weapon. Contrary to the model set out by Reynolds, however, such an observation does not lead ineluctably to the conclusion either that the political elite of early Anglo-Saxon England was also a “warrior elite” or that the specific men buried with swords were, in fact, members of the political elite. First, it must be recognized that in many cases only families with a certain minimum level of wealth likely were in a position to dispose permanently of a very expensive tool in a grave. The model of a warrior elite requires that only a small, indeed elite, element of society was equipped for combat with high-quality weapons. But to assert that such a small elite existed on the basis of the limited number of graves that included high-quality weapons is a tautology that ignores the economic reality involved in disposing of such expensive tools. Put another way, just because we find swords in only a limited number of graves does not mean that there were only a limited number of men who possessed high-quality swords of this type. One profitable avenue of comparison in this regard would be a comparison of the types of weapons found in graves with those found in other environments such as battlefields where there is a possibility for a more random selection of former weapon owners. A second problem with the “warrior elite” model is that it presupposes that the decision about whether to bury a particular weapon lay exclusively with the heirs of the man in the grave. However, the possibility cannot be excluded on the basis of the surviving evidence that the men in the graves actually were paid “soldiers” who served in the military households of the magnates, e.g. kings and princes, who in fact, ruled society. Under these conditions, the man in the grave might perhaps have been well off but could not be considered a member of the ruling elite. A comparative example might be adduced that some of the men who served in the mercenary forces of captains such as John Hawkwood became very wealthy; however, they can hardly be considered to have dominated Florentine society.15 The second of the essays that does not truly fit the paradigm of dealing with a landscape of defense is the study by Peter Ettel, professor of pre- and early history

15 On this topic, see William Caferro, John Hawkwood: An English Mercenary in Fourteenth-Century Italy (Baltimore, MD, 2006).

244

LANDSCAPES OF DEFENSE

at the Friedrich-Schiller-Universität, Jena, who has undertaken a survey of the vast quantity of “Frankish and Slavic Fortifications in Germany from the 7th to the 11th century.” As Ettel points out, more than 1,600 strongholds are known from archaeological sources for this period. There are more than 1,000 fortifications identified largely on Germany’s eastern frontier and more than 600 from the western borders of the Slavic region. Most of these data, however, were collected more than twenty years ago, and undoubtedly more sites have been identified in recent years. Most of these fortifications, both those located on the frontiers of the German kingdom and those identified in Slavic territory, are not mentioned in any surviving written text. Many of the fortifications that eventually are mentioned in one or another medieval text are to be found in records that were compiled a century and more after the stronghold was constructed, as dated by archaeological evidence, e.g. dendrochronology. Ettel points out that traditionally archaeological research was focused on dating the initial and subsequent development of fortifications and on creating typologies. These data undergird the view that following the so-called “Migration Period” (Völkerwanderungszeit) of the fourth and fifth centuries, there was a hiatus in the construction of fortifications during the first century and a half of the Merovingian era, i.e. before the fragmentation of royal power and the emergence of the rois fanéants. From the mid-seventh century onward, there was a period of continuous fortress construction up through the end of the tenth century. Ettel identifies three types of Frankish fortifications built during the period, which are thought to have developed progressively from largely wood and earth constructions to the increasing use of stone. This pattern would seem to be evidence, on the one hand, for the increasing wealth of those who were responsible for the construction of these strongholds, while on the other hand, evidence also for an increasing capacity, e.g. the development of siege technology, by active or perceived enemies to threaten less well-constructed fortifications. In addition to the traditional aspects of German archaeology, Ettel calls attention to an increased interest in life within the fortifications and their environments. As an excellent example of the high levels of information archaeological research can obtain, Ettel draws close attention to the fortification complex at Karlburg on the river Main where he himself made significant contributions. The fortifications at Karlburg likely were begun under the auspices of King Pippin I († 768) and saw continuous development for centuries. Ettel sees Karlburg as a “large early medieval village,” which can be compared with early urban sites. Ettel’s terminology is not transparent and, therefore, it is important that he describes major aspects of the site. The heart of the complex was the stronghold whose walls were expanded with an increased use of stone to a width of 9 to 10 meters. The height of the walls now cannot be recovered, but the entire complex was surrounded by a ditch. In addition, there was a second stronghold on the other side of the river called the Grainberg. There was also a monastery, a harbor, and a ship landing indicative both of trade and for the transport of troops along the river and of logistical support. There is evidence for dwellings, stables, granaries, some fifty “pit houses” 245

M AT E R I A L S O U R C E S A N D I M A G E S

illustrating iron working and weaving, and finds indicating trade in high-quality goods. The defenses ultimately enclosed an area along the river a kilometer in length and 200 meters in width. Whether modern scholars should consider this complex a large village, a small town, or perhaps even an emporium for the buying, selling, and transport of agricultural products westward and manufactured goods eastward certainly raises questions regarding classification. Demographic estimates are desideratum for both the population of the enclosed area and the regions close by that supplied the population with food, as only occasional gardens are to be found within the walls as there was no space for the grazing of herds for food. Some contemporary texts of the later eighth century refer to Karlburg as an urbs, and this provides food for additional thought.16 Indeed, the classification of other fortifications along the frontiers of the German kingdom with regard to economic and demographic criteria remains to be ascertained. Another desideratum in this context is an assessment of the labor resources that were required to construct defenses on this scale, which would help to illuminate the ability of governmental authorities to mobilize human, material, and financial resources on a large scale. Ettel devotes a total of four pages of discussion to Slavic fortifications. The main point to be taken away from this part of the essay is Ettel’s support for the contemporary Germanist view that traditionally Slavic fortifications have been dated two or more centuries too early. The basis for this late dating is not made clear by Ettel, and he does not explain the weakness in the methods of Slavic archaeologists. In fact, not a single work in a Slavic language is cited in Ettel’s bibliography.17 Without a detailed examination of the Germanist and Slavicist positions, one wonders whether traditional German views of Slavs as comparatively primitive are at work or whether traditional Slav views that they were at least the equals of their German neighbors are operative. One is reminded here of long-held views by the English of the inferiority of their Celtic neighbors and the efforts of the latter to counter those views. When, according Ettel, the Slavs began large-scale construction of fortifications throughout the western parts of the region that they dominated, their efforts seem to be very similar to those undertaken by their German neighbors. In the way Ettel presents this information, the reader seemingly is encouraged to believe that the Slavs learned a great deal from the Germans, which, in fact, may be an accurate appreciation of Slavic-German interactions in regard to construction technology and techniques. In this context, it is to be noted that the literate Germans learned much from Roman books, such as the works of Vitruvius.18 In fact, 16 See the discussion by Bernard S. Bachrach, Charlemagne’s Early Campaigns (768–777): A Diplomatic and Military Analysis (Leiden, 2013), 563–4. 17 P. M. Barfort, The Early Slavs. Culture and Society in Early Medieval Eastern Europe (Ithaca, NY, 2001), particularly 144–6. 18 Bachrach and Bachrach, “Hildagsburg,” 25–58.

246

LANDSCAPES OF DEFENSE

parts of Euclid’s geometry were translated into German, while the Slavs in general lacked access to such texts. As with his treatment of German fortifications, Ettel describes in considerable detail several large and well-excavated Slavic strongholds, which are very similar to large German fortifications with administrative functions, churches, manufacturing infrastructure, and trade. They obviously played host to populations of a considerable size. However, no demographic estimates are ventured. Ettel draws a series of conclusions from his tour of the horizon. First, and most importantly, he recognizes that these fortifications were built for military purposes, although he does not venture to review the scholarship that explains how this was done. He emphasizes, particularly in regard to large strongholds, that they played a role as places of refuge during enemy raids, but importantly makes clear that there also were fortifications that were inhabited with permanent populations and did not remain empty awaiting a military crisis. Ettel also emphasizes that fortifications had a “sacral” function by which he means that churches often were built within the walls. However, “sacral function” does not seem to be the correct language. Fortifications also had economic functions, e.g. manufacturing and trade, administrative functions, and as “centres of [. . .] lordship” which remains unclear. He also notes that some strongholds “were seats of sovereigns, aristocracy, and clergy.” A more reasonable phrasing would seem to be that kings, bishops, and lay magnates sought the safety of fortifications for their homes. While Ettel has provided a very useful introduction to German and Slavic fortifications in the course of fewer than thirteen pages of text, there is nothing to be found in regard to landscapes of defense, despite the fact that much has been done by German scholars in this vein. One need only review, for example, the extensive literature dealing with the limes Saxonicum, the Germar Mark, and the Sorbian march, the latter two of which certainly were comparable to the Burghal Hidage, in order to gain a proper appreciation of Carolingian use of fortifications for strategic and tactical purposes.19 In addition, the strategy undergirding the frontier fortifications constructed by Henry I († 936) and his son Otto the Great († 973) make clear that the rulers of the German kingdom were at least as attentive to landscapes of defense and, indeed, of offense as their Anglo-Saxon contemporaries.20 In short, otherwise uninformed readers of Ettel’s essay will come away believing that rulers such as Alfred the Great were more attuned to the role of fortifications in military strategy than, for example, were Charlemagne († 814) and Otto the Great. This would be a highly misleading conclusion. The majority of the studies in this volume, including those by Ettel and Reynolds, treat numerous sites and, as will be seen later, these often include multiple fortifications. By contrast, Neil Christie with Oliver Creighton and Matt Edgeworth, who are connected to the University of Leister, the University of Exeter, 19 See, for example, the thought-provoking works of Hardt, “Hesse, Elbe, Saale,” 219–32; and Idem, The Limes Saxoniae as part of the Eastern Borderlands,” 35–50. 20 See, for example, Bachrach, “Restructuring the Eastern Frontier,” 9–35.

247

M AT E R I A L S O U R C E S A N D I M A G E S

and the University of Leister, respectively, devote their attention to the single, albeit very important, fortress at Wallingford. Their goal is to examine whether Wallingford “was primarily a frontier foundation or a burh with planned specific urban and economic functions from the outset.” A second goal of this study is to ascertain the ways in which Wallingford was able to “endure the transition to Norman rule.” The highly nuanced and exceptionally well-documented study by Christie et al. suggests that the questions posed earlier are rather tendentious and should be answered in a manner that takes into account the chronology of particular situations and the availability of information that is uncontroversial. With regard to the second question, for example, there is some reason to suggest that in the wake of Duke William’s conquest, the position of Wallingford may have declined precisely because Norman strategy was not the same as that of some Anglo-Saxon rulers. However, Wallingford again assumed a very important role during the civil war between Stephen and Matilda, as their strategic imperatives once again were different from those of the earlier Norman rulers of England. In the context of understanding the changing role of the fortification at Wallingford, it is important to consider that it was a major element in Alfred’s Burghal Hidage, which controlled a key crossing of the Thames west-northwest of London and required garrison defense forces of the same order of magnitude as the late Roman walls of the fortress city of Winchester. Nevertheless, neither the AngloSaxon Chronicle nor Asser’s Life of Alfred mentions the fortress. The failure of either text to discuss Wallingford is even more noteworthy in light of the fact that Orosius’s Universal History, which claims that Julius Caesar fought a battle at Wallingford, was translated at Alfred’s court. Thus, the failure to mention Wallingford may be taken as illuminating the lacunose nature of both texts and suggests that relying on these works alone for information about the extent of the burghal system in Wessex is likely to lead to errors. Although not treating a “landscape,” the cooperative effort by Christie, Creighton, and Edgeworth has produced an important study with regard to the state of the question concerning Wallingford, and just as importantly, a model for additional studies of other elements in the Burghal Hidage, as well as longitudinal views of the areas involved both before and after Alfred, which are much needed. In all, these data will help military historians develop a deeper understanding of the strategy of fortress building over time, as differing conditions led to creativity among military planners. One additional point touched upon by Christie et al. is the Roman background of the Anglo-Saxon landscape and what decision makers such as Alfred and those educated men at the court knew both from written and topographical sources about various aspects of the past. It is to be hoped that scholars following the model set out in the essay on Wallingford will delve more deeply into this particular issue. The fourth of the studies to eschew a specific focus on landscapes is the essay by Juan Antonio Quiró Castillo, professor of archaeology at Euskal Herriko Unibertsitatea, that offers “a brief summary of recent archaeological studies of defensive sites of the early Middle Ages (5th to 10th centuries) in the north-western 248

LANDSCAPES OF DEFENSE

part of the Iberian Peninsula.” This survey, however, is not focused on a discussion of “landscapes of defense,” but its aim is “to discuss in social terms the role that these structures played in the formation of medieval societies” (303). Quiró Castillo does not approach this topic by asking how the people who decided that these fortifications were to be built viewed their efforts. Instead, he suggests that the expenditure of extensive human and material resources to build, maintain, and garrison strongholds was intended for largely nonmilitary purposes. Quiró Castillo begins with an all-too-brief and, therefore unsatisfactory, survey of various “theories” put forth, largely by French and Italian scholars, regarding what is thought to be the social history of fortifications. A more detailed examination of these ideas might well have been very useful to the attentive reader. Absent from this limited discussion of the scholarship are the works of German archaeologists and even of Anglo-Saxonists, who have put forth volumes of ideas in this area. What has most impressed us regarding Quiró Castillo’s introductory effort is his attachment to the “feudalism” construct. This suggests to us that Quiró Castillo either has not absorbed the implications of Susan Reynolds’s magisterial and very widely accepted deconstruction of “feudalism” in her 1994 monograph Fiefs and Vassals or that he simply has chosen to ignore this paradigm-breaking work so as to support an argument that relies on an approach that lacks even heuristic value. If one is to ignore Reynolds, it is necessary to defend one’s position in detail and at length. Quiró Castillo’s intention is to “explore the active social role played by castle and fortress in the early Middle Ages by means of an integral analysis of territory with a comparative analysis at the regional level” (308). To this end Quiró Castillo divides early medieval fortification construction into three periods. Initially, he focuses on structures built or reused during the period of the dissolution of Roman imperial power in the northwestern part of the Iberian Peninsula. These he calls “first generation castles,” which are dated to the period from the fifth to the seventh century. The “second generation castles” are dated to the eighth and ninth centuries, and the “third generation castles” are dated from the tenth to the eleventh century. On the whole, there is rather little archaeological work available to draw meaningful conclusions regarding the social role of these fortifications. In fact, the location of many strongholds mentioned in written sources has yet to be identified, much less excavated. Thus, Quiró Castillo finds solace in various theories generated for which there is a dearth of evidence in Northwest Spain. In general, with regard to the first generation of strongholds, some types are marked by the poor quality of the available archaeological record (309), and the immense diversity of examples within this chronological period seem to defy characterization. As a result, Quiró Castillo’s generalizations seem to be less than compelling. One point, however, worth noting regarding Quiró Castillo’s discussion of “urban planning” is that these castles frequently have a cemetery beyond their walls (315). It is regrettable, in this context, that Quiró Castillo does not provide a valid definition of what constitutes an urban site. In addition, given the early date of these fortifications, it is surprising that Quiró Castillo does not 249

M AT E R I A L S O U R C E S A N D I M A G E S

raise the question of the survival of the Roman tradition of the pomerium. Overall, Quiró Castillo grasps at straws to try to explain various phenomena that are very poorly documented in the archaeological record in a period that is largely bereft of written sources. For example, he presents an argument regarding what he believes to have been a pattern of behavior indicating that there are greater numbers of “castles” in those areas where the collapse of urban networks was most pronounced. To support the complex hypothesis for which there is no evidence, he calls attention to what seems to be a well-documented fact that there were “urban abandonments” in the Duero Basin a millennium later. Trying to explain events, much less policies, by conditions separated by 1,000 years cannot be tolerated, at least not by specialists in medieval European military history. As noted previously, Quiró Castillo finds there to be a dearth of archaeological evidence in key areas during the first generation. Similarly, concerning the second generation, he observes that there are also very few archaeological studies of fortifications from the eighth and ninth centuries. This problem, for which the author provides no compelling cause, raises questions regarding the validity of his three-generation model. This problem is further exacerbated by the fact that those fortifications that are dated to the eighth and ninth centuries by archaeological evidence often are to be found treated in the written sources only from the late ninth and tenth centuries. The major point to be emphasized regarding this “second generation” is its orientation toward stopping or slowing down Muslim advances both in terms of territorial conquest and raiding. This defensive mode oriented toward Muslim aggression is very well exemplified by clusae of considerable size and sophistication that were constructed to control mountain passes, as well as the reutilization and reorientation of older fortifications. While Quiró Castillo likely is correct in seeing these efforts in terms of what he identifies as state development, he simply ignores the fact that the main purpose of these efforts was to defend against the Muslims. Whether it is politically incorrect to see the Muslims as enemies even at a remove of a millennium and more is a question best left to Spanish academics. However, in a more serious vein, this “second generation” in Quiró Castillo’s model provides exceptionally fruitful opportunities to compare the defensive methods undertaken in the northwest of Christian Spain with the efforts of the Mercian kingdom and then of Wessex to defend against the Vikings. Indeed, even more fruitful comparisons might be drawn with other regions of Iberia, as will be seen later. As throughout, Quiró Castillo consistently confuses the English usage “early Middle Ages” and refers to the period as the “late Middle Ages.” However, it is to be emphasized that the third generation of fortifications does not concern the late Middle Ages. Here Quiró Castillo focuses his findings in the “Basque Country,” where some forty “castles” in the “late Middle Ages (sic),” i.e. the tenth and eleventh centuries, have been identified (328). Most of the fortifications that have been excavated or studied in recent years “are located high up on steep hilltops from which they dominate large territories” (329). Exactly how this domination 250

LANDSCAPES OF DEFENSE

took place is not explained, as both handheld missile weapons and catapults of various types had limited range. Such strongholds, if properly sited, could control a road that passed by the fortifications. However, after characterizing the strongholds as dominating large territories, Quiró Castillo concludes that they were “symbol” castles and lacked any offensive function. Rather, he believes that they “serve to mark territory” (329). Quiró Castillo does not discuss the likelihood that these strongholds were constructed to hold territory in order to halt Muslim territorial conquests and as magazines to provide government forces with logistic support when moving against enemy assets. On the whole, Quiró Castillo exposes his readers to large quantities of information, which if properly digested and related to major research questions and/or to the matter of landscapes of defense could be of exceptional value to historians. Two major areas of research clearly could be meaningfully affected. In the discussion of the first generation of fortifications, the examination of later Roman survivals into and through the Visigothic period is of considerable importance. In regard to fortifications of the second and third generations, an important focus should be on the question of developing landscapes of defense in Christian territory against Muslim aggression. Special attention here to place names, roads, and signal stations are of primary importance. Finally, comparisons between AngloSaxon defensive activities against the pagan Vikings and those taken by Christians against Muslim aggression could be very useful. Of the nine essays in this volume that focus specifically on landscapes, three are written by historians. Barbara Yorke, professor emerita of medieval history at the University of Winchester, provides observations on “West Saxon Fortifications in the Ninth Century: The Perspective from the Written Sources.” Richard Abels, professor of history in the United States Naval Academy, focuses on “The Cost and Consequences of Anglo-Saxon Civil Defence, 878–1066.” The essay by Julio Escalona, senior researcher at the Instituto de Historia, Centro de Ciencias Humanas y Sociales – CSIC Madrid, is intended to examine the role of fortifications on the political frontier marked by the river Duero as a means of treating Christian advances against the Muslims. All three studies would seem to break new ground insofar as they work at a multidisciplinary level to integrate history and archaeology. It is Yorke’s aim to demonstrate that references to fortifications in narrative sources for the ninth century enable us to expand the number of known defensive sites in use in Middle Saxon Wessex (104). She initiates an effort to compare the terminology used in Anglo-Saxon sources for fortifications with those used regarding the same strongholds in Latin sources and also calls attention to aspects of imitatio imperii, i.e. copying the way in which the Romans behaved, as an important element in the construction of new fortifications. However, when discussing marching camps used by both Anglo-Saxon and Viking armies, she neglects to call attention to the working of imperial influence. Here, attention is due to the broad reading of Roman history at the court of Alfred the Great († 899), as well as the fact that many Norse adventurers in English service had previous military experience 251

M AT E R I A L S O U R C E S A N D I M A G E S

under Byzantine emperors. While Yorke’s innovative approach in regard to AngloSaxon studies is to be applauded, it must be remarked that interdisciplinary efforts regarding the continent, which deploy terminology, archaeological remains, and accounts of military operations in the course of sieges to expand our understanding of fortifications, have been a commonplace for a long time.21 Abels, one of the leading specialists in Anglo-Saxon military history and especially Alfred the Great, argues that the development of a complex and costly civil defense system in late ninth-century Wessex played a critical role in the state formation of late Anglo-Saxon England (217). Whether Alfred’s Wessex can be considered a “state” depends, at least in part, on the definition of state one chooses to employ. Abels, unlike some of his colleagues in this volume, does not choose to enter the lists on this question. In addition, it is controversial, at the least, whether the term civil defense is so anachronistic that even its heuristic use may be challenged. In much less controversial terms, it is clear that the government of Alfred the Great saw increasing sophistication and centralization during the process of developing the military resources of Wessex for the purpose of combating the Vikings. Moreover, the concomitant development of military and institutional sophistication cannot be isolated to periods of impending danger. While Alfred developed the Burghal Hidage for defensive purposes, his heirs used these relevant fortifications as bases to press offensive operations. Abels’s thesis, i.e. the development of military assets played an important role in the growing sophistication of governmental institutions, in itself is hardly novel, although throughout his essay he adds evidence to an already solid case for King Alfred and his successors. Abels’s recognition that Alfred’s major military innovation was the development of a system of defense that was constituted strategically as a defense in depth also is not new but no less accurate for that. What is refreshing is Abels’s willingness to focus on military decision-making that looked toward a long-term strategy on a large scale and his further observation that these projects were exceptionally expensive in terms of the investment of human and material resources. The examination of aspects of the economic underpinnings of governmentally directed military building is somewhat of a departure for Abels. Thus, he begins by accepting the “revisionist” view of Offa’s Dyke and recognizing that it was, in fact, a military project. However, when he enters the discussion regarding its costs, which is a subject of great importance for understanding the availability of surplus resources for military projects, Abels’s estimates regarding these matters represent only about 30% of the actual costs (198–199, esp. n. 2).22 Abels’s too low levels for the cost of building Offa’s Dyke diminishes our understanding of the great expense of building such monumental military projects. 21 See, for example, Bachrach, “Early Medieval Fortifications,” 531–69; and idem, “Fortifications and Military Tactics,” Technology and Culture 20 (1979), 531–49. Both are reprinted with the same pagination in Warfare and Military Organization. Also see Bachrach, “Restructuring the Eastern Frontier,” 9–35. 22 See the studies cited in note 2.

252

LANDSCAPES OF DEFENSE

However, despite the relatively low level of resources that he associates with the construction of this landscape of defense Abels recognizes that the sheer scale of construction implied by the Burghal Hidage is enormous (201). This is an important step in the right direction, as scale obviously evokes matters of costs. However, as with Offa’s Dyke, Abels’s citation of cost estimates, based largely on claims adduced by English scholars such as Halsam, are too low, as they are intended to prove that the construction of the entire system could have been completed in about fifteen months. Abels also takes note of the costs involved in maintaining the fortifications and garrisoning them. Once some sort of estimate regarding the costs of systems such as the Burghal Hidage has been established, it will be necessary to use these data for the purposes of trying to make sense of the Wessex economy, i.e. the goods and services produced in Alfred’s kingdom during the later ninth century. The final essay by a historian, which is also the final essay in the volume, is Julio Escalona’s “Military Stress, Central Power, and Local Response in the County of Castile in the Tenth Century.” Escalona sees a multifaceted process of combined central government and local efforts undertaken in a period of “intense social change, largely triggered by military stress and incorporation in a largescale, and more complicated system” (342). However, as will be seen later, the documentation presented by Escalona does not permit a clear demarcation of the orchestration by governmental authorities of what may seem to be a fortressbuilding strategy. One way to treat this problem is to ascertain how and under what circumstances resources were laid under contribution to affect a particular strategy. However, before Escalona focuses on his main goals, he provides an overview of the history of fortifications in this area, which shows that very little is known of the region often referred to as Old Castile from the dissolution of imperial authority until sometime in the later eighth century or perhaps somewhat earlier. By ca. 850, the original name of the region, which was Bardulias, came to be called Castella in Latin and al-Qila in Arabic (344). This phenomenon of castle building, recognized by both Christians and Muslims, already was well developed, likely for a generation or more, when the name change became so firmly established as to be recorded in written documents, which by their nature are exceptionally conservative in their usage. Since the kingdom of Asturias of which Bardulias was a distant frontier ca. 750 cannot be shown to have much administrative reach southward toward the Duero even in the later eighth century, it should be concluded that the castle-building phenomenon was spearheaded by local aristocrats intent upon establishing the defense of Christian territory against Muslim aggression. From an archaeological perspective, Escalona does not explain whether the making of old Castile was done on the basis of already existing strongholds that were reused or through largely new construction. It is also unclear whether this newly baptized “land of castles” was thickly covered with stone strongholds or fortifications constructed of timber and/or earth. Whatever the case, it is important that future studies give some sense both of the costs of castle construction 253

M AT E R I A L S O U R C E S A N D I M A G E S

or refurbishment and also of the demography of the region so as to ascertain the economic and social conditions that made available the surplus human and material resources that were required to build these strongholds. Without a better sense of these material realities, Escalona’s hypotheses regarding competition between local elites and elites connected to the royal government must remain little more than interesting possibilities. Or to put it another way, Escalona’s claim that local elites likely were farmers themselves “even if they had access to more lands and extra workforce by means of their control of royal relations and community leadership,” remains only theoretical (348). The question of the strategy undergirding the castle-driven defense of the Duero frontier in the late ninth and early tenth centuries, like most scholarly matters, is divided between those who see the direction of the central government as of primary importance and those who focus on local initiative to the detriment of royal initiative. However, at this time the formulation of hypotheses seems fruitless, because as Escalona bitterly recognizes, scholars, and here one might read government support, are focused on prehistoric sites and on the archaeology of Muslim fortifications. Escalona’s efforts to bring further attention to watch towers and signaling systems also are undermined by a lack of archaeological evidence, as are his efforts to look into the matter of fortifications constructed as refuges in the face of Muslim raids. By contrast, studies treating place names that are of value to understanding military matters seemed to have fared very well by comparison with more costly archaeological efforts. Unfortunately, Escalona’s treatment of military organization still suffers from the voluminous efforts by Claudio Sánchez-Albornoz to place Spanish military institutions within the now-discredited Germanic freeman model, which supposedly developed over time into the equally discredited European-wide feudal model. In addition, Escalona’s reference to Guy Halsall’s primitivist nondiscussion of the Duero frontier is rather more misleading than useful (356). The reader is encouraged to examine James F. Powers’s magisterial A Society Organized for War: The Iberian Municipal Militias in the Central Middle Ages, 1000–1284 that is exceptionally useful for the background material that it provides. Of the remaining six essays in the volume, the most important from the perspective of the study of military history is that by Stuart Brookes, “Mapping AngloSaxon Civil Defence.” Brookes begins with the long-held observation by military planners that field craft, as this term is used by specialists in military science, makes clear that “elements of the landscape interrelate” and as a result landscapes can be read as “palimpsests of past action” (39). Brookes’s aim is to discuss militarism, which he defines as “geographies of military preparedness,” in terms of “continued preparations which states make for war and the geographical impacts such measures have” (41). Brookes’s focus throughout the essay is on the mapping of various aspects of militarism, so defined, in southern England during the Viking Age. Brookes’s initial concern is with borders and their defense by “states,” a term that he believes can be used without reservation to denote Anglo-Saxon polities. 254

LANDSCAPES OF DEFENSE

Brookes further enters into controversy by adopting or at least adapting some of the ideas of Edward Luttwak, whose effort to apply the idea of grand strategy to the defense of the Roman Empire was widely condemned by British specialists in ancient history, although often accepted by scholars elsewhere.23 Brookes, like Luttwak, does not provide an epistemologically valid definition for “Grand Strategy.” Brookes consequently is in the ironic position of setting out a conceptual framework for examining the Viking Age that many, if not most, British specialists in Roman history have vigorously rejected as beyond the capabilities of the empire’s intellectual, planning, and material resources. Of course, Rome commanded a wide variety of assets that were far superior to those available to Alfred the Great and his fellow Anglo-Saxons. This being the case, it is to be noted that Brookes has sufficient information available from a variety of sources without adducing vague and undefined concepts such as grand strategy or the work of Luttwak.24 Brookes has available substantial bodies of information regarding important aspects of militarism in regard to assets such as fortifications, signal systems, place names, and roads in relation to the likely movements of men and information, and these can be shown to articulate effectively in regard to the topography. Brookes alludes to matters of logistics, which are important both in constructing infrastructural components (see Abels earlier) and for the supply of both troops on the move and garrison troops who were deployed to defend various fortifications. In order to add logistics to the range of materials that support Brookes’s compelling arguments it will be necessary to examine the written sources, particularly Domesday Book, from which it will be possible to identify estates that were positioned to provide logistic support. John Baker, one of the co-editors of this volume, has contributed “The Language of Anglo-Saxon Defence,” which is focused on place names that have military and military-related importance. This work supplements part of Brookes’s efforts to integrate fortifications, signal systems, and roads in relation to the likely movements of men and information, as demonstrated in regard to both natural and manmade topography. The large number of illuminating place names for military roads, e.g. here-paed and fyrd road names such as fyrd-straet, are of great importance in understanding connections between fortifications and troop movements, which also have implications for logistics. However, Baker makes clear that terms such burh do not always have the same meaning of fortification over time, or at least not that primary meaning, as the function of places can change. Also, it is

23 For a survey of the scholarship, see Alexander Sarantis, “Waging War in Late Antiquity,” in War and Warfare in Late Antiquity, ed. Alexander Sarantis and Neil Christi, 2 vols. (Leiden, 2013), 1–98, particularly 7–17. 24 This model has been defended successfully by Everett Wheller, “The Methodological Limits and the Mirage of Roman Strategy,” Journal of Military History 57 (1993), 7–41, 215–40; and Kimberly Kagan, “Redefining Roman Grand Strategy,” Journal of Military History 70 (2006), 333–62, although these authors have significant disagreements. Also see the broad survey by Sarantis, “Waging War,” 7–17.

255

M AT E R I A L S O U R C E S A N D I M A G E S

important that fortifications are not always called burhs. These nuances are not unimportant, as particular places that at one time likely were constructed for military purposes over time can be seen to cease to play that role at a later date. Like Brookes, Baker sees the interrelated aspects of forts, roads, and signal stations as evidence for governmental planning. Gareth Williams, curator of early medieval coinage in the British Museum, aims to explain the multivalent meaning of the term burh in a useful but sometimes misleading essay entitled “Military and Non-Military Functions of the Anglo-Saxon Burh, c. 878–978.” Williams begins on the wrong foot by following the notion, espoused by Abels and others, that Anglo-Saxon military organization was based upon “personal lordship” and that this view has been “expanded to early medieval warfare in general” (133). Here Tacitus’ early second-century Germania, largely cleansed of his views regarding the centrality of kingship, and literary fantasies such as Beowulf traditionally have served as the framework for this argument. Thus, Hollister’s accurate assessment of Anglo-Saxon military organization, which owes much to Frankish influences, e.g. Charlemagne’s military organization, is ignored. Williams and other defenders of the lordship argument raise the red herring of a dichotomy between lordship and an archaic military organization, often based on notions of Gemeinfreiheit, that it was the obligation of the so-called Volk as a whole to bear arms.25 The late antique military system in Gaul that influenced the Anglo-Saxons was tripartite in structure. For purposes of local defense, all able-bodied men, even including slaves, were mustered to defend their home territory regardless of military training. This was done on the basis of later Roman legal enactments dating at least to the early fifth century, if not even earlier. Also following imperial legislation, locals whom the government determined possessed sufficient means were mobilized personally or required to provide one or more substitutes on the basis of their wealth in order to undertake offensive military operations outside their home territory. Magnates who owned large expanses of land were obligated to attend the muster and potentially bring with them very large numbers of men. Finally, the king and his magnates, both lay and ecclesiastical, supported military households, obsequia, which were composed of professional soldiers. These men who served in a magnate household were mobilized for war as part of their employers’ obligation to produce high-quality fighting men when called upon to do so by the king or relevant prince. In short, neither ill-defined notions of lordship nor so-called Volk obligations determined the basis for military service. Rather, wealth was the basis for assessing military obligations.26

25 See C. Warren Hollister, Anglo-Saxon Military Institutions on the Eve of the Norman Conquest (Oxford, 1962); and Idem, The Military Organization of Norman England (Oxford, 1965). For later Anglo-Saxon England as a Carolingian-type state, see James Campbell, Essays in AngloSaxon History (London, 1986). 26 This system is described in detail in Bernard S. Bachrach, Early Carolinglian Warfare. Prelude to Empire (Philadelphia, 2001), 51–83.

256

LANDSCAPES OF DEFENSE

In addition to the tripartite Frankish system that the English adopted, the trinoda necessitates, which the Anglo-Saxons also adopted, are a well-known example of borrowing from later Roman legal enactments, likely mediated through Carolingian usage. Very similar obligations were geared to the support of military and military-related operations in the regnum Francorum. Indeed, Williams argues that burhs were modeled on Frankish civitates as the focus of both royal military authority and administration (145) and also accepts the view that the Danish fortress system was constructed under Frankish influence. In this context, it should be noted that the Saxons similarly constructed fortifications in the eighth century along their frontiers with the Regnum Francorum and that large fortifications served this same function in the Ottonian kingdom during the tenth century.27 In addition to the copying of Frankish military institutions, it bears emphasis that imitatio imperii was well developed in England with regard to the design of some new strongholds and the refurbishment and use of others such as the fortress city of Winchester, which played an important role in the Burghal Hidage. In fact, Williams sees the “Germanic” north as much influenced by Rome. Williams himself recognizes Roman influence, as he claims that Alfred the Great was influenced directly by his personal experience of the fortifications at Rome (149). Once Williams gets beyond matters of early medieval military organization, he goes on to examine various functions that would appear to have occurred in more than a few royal burhs. Given his position at the British Museum, it is hardly surprising that he gives most attention to mints and coinage. He also notes that some burhs likely had the capacity to serve as refuges for inhabitants in the neighborhood. However, in his discussion of the Burghal Hidage, Williams observes that “the ability of the civilian population to outdistance pursuit by Viking forces” required that “civilians of all ages and in all states of health could travel twenty miles [. . .] in the course of a day” (131). This model assumes, contrary to fact, that some pairs of Burghal Hidage fortifications were 40 miles apart, when on the whole they were a maximum of 20 miles apart, and, thus, fugitives were required at a maximum to travel only approximately 10 miles to reach the safety of their walls. It is also assumed that there were no refuges along the roads between burhs. On the whole, Williams is on the right track in seeing the importance of burhs and this especially in the context of mainland influence. His approach would be far better served if he also saw military institutions in terms of Frankish influence. Andrew Agate, a PhD candidate at the Institute of Archaeology, University of London, currently is undertaking research regarding “suburbs” in Anglo-Saxon England, and in this volume he treats “Aspects of Suburban Settlement at Early Urban Centres in England.” Agate devotes several interesting pages to how suburbs are discussed, especially by sociologists and urban planners, and concludes, that for the most part, this work, while perhaps stimulating in various ways,

27 Werner Emmerich, “Landesburgen in ottonischer Zeit,” Archiv für Geschichte von Oberfranken 37 (1957), 50–97; and Hardt, “Hesse, Elbe, Saale,” 219–32.

257

M AT E R I A L S O U R C E S A N D I M A G E S

especially in terms of post-medieval views of suburbs, is not of much value to a specialist working primarily with archaeological evidence available for the AngloSaxon period. What emerges from this tour of the horizon is that there are no valid epistemological definitions for cities or for suburbs, although Agate seems to find Susan Reynolds’s bundle of characteristics to denote a city to be of value. Here it is to be noted that American English calls for the use of the word city, while the British tend to prefer town. This may confuse some readers who are not familiar with this difference. There are two important problems with Agate’s approach to the matter of Anglo-Saxon suburbs. Although Agate focuses on an examination of post-medieval observations regarding suburbs, markedly absent from his tour of the horizon is the uncontroversial fact that suburbs were connected to cities and lesser walled population centers throughout the ancient world and especially in the Roman Empire, not excluding Britain. It seems to us that developing various ideas regarding suburbs, however defined, in the Anglo-Saxon period would benefit as much from an understanding of the subject in the pre–Anglo-Saxon period as from the treatment of discussions of this topic by modern city planners. A second problem is the lack of focus on suburbs within the framework of landscapes of defense. For example, if a suburb is seen to develop or to be developed in the neighborhood of a fortification, it is certainly important to know whether this was due to government action or resulted from ostensibly private initiative. A second important question concerns the fate of suburbs outside the walls of a fortress when the city was under attack. Here one might expect that archaeological evidence could cast light, for example, on suburbs that were destroyed while the settlement within a walled fortification survived. Contemporary efforts in the East Frankish/German kingdom to expand the walls of fortress cities such as Worms and Mainz to enclose suburbs make clear that government planners, here bishops, were aware of this problem.28 The efforts of the Anglo-Saxon specialists in landscape archaeology in this volume are complemented by two additional works that focus in turn on the Low Countries and Scandinavia. The first of these is by Letty ten Harkel who is a researcher at the Institute of Archaeology in the University of Oxford and holds a PhD from the University of Sheffield. Her work focuses on “A Viking Age Landscape of Defence in the Low Countries: The Ringwalburgen in the Dutch Province of Zeeland.” It is ten Harkel’s goal to reexamine the status questionis regarding the characterization of five Zeeland ring wall fortifications with the aim of ascertaining whether they belong to a single defensive system of some kind, which has been suggested by some scholars on the basis of their style and construction methods, as well information provided by some contemporary written sources. She also 28 See, for example, the discussion by Heinrich Büttner, “Zur Stadtentwicklung von Worms im Frühund Hochmittelalter,” in Aus Geschichte und Landeskunde. Forschungen und Darstellungen. Franz Steinbach zum 65. Geburtstag gewidmet von seinen Freunden und Schülern (Bonn, 1960), 389–407; and Ludwig Falck, Mainz im frühen und hohen Mittelalter (Mitte 5. Jahrhundert bis 1244) (Düsseldorf, 1972), 75–83.

258

LANDSCAPES OF DEFENSE

seeks to reexamine the argument that all of these strongholds were vluchtburgen, i.e. refuges constructed so that the local population could find safety during Viking raids. On the whole, ten Harkel rejects both of these arguments. A study of various contemporary written sources has led some historians to see these fortifications as the work, for example, of the Frankish king and emperor Louis the Pious († 840), who is recorded to have had strongholds constructed to deal with Viking invasions. However, as ten Harkel emphasizes, the dating of these strongholds according to archaeological methods fits neither the chronology of Louis’s reign nor any other period for fortress construction that is indicated in the written sources. If one accepts these archaeological dates, then the obvious question must be asked, but not in this study, is whether archaeologists have been able to identify any of the fortifications that fit the information provided by these contemporary written sources? In addition, ten Harkel points out that the archaeological dates for the construction of these fortifications do not fit with times when all of the strongholds were under the control of a single government. Therefore, she argues that they cannot have been a system or part of a single system. She suggests instead (249) that they may have been elements of competing systems, as the region of Zeeland formed the frontier between emerging polities. These fortifications, especially Domburg with a diameter of likely 265 meters, undoubtedly required the mobilization of considerable quantities of human and material resources, likely amounting to many tens of thousands, if not hundreds of thousands, of man hours of labor. The presence of what clearly were permanent populations in these strongholds, engaged in some cases both in manufacturing, e.g. spinning of wool, and iron work, also would seem to support ten Harkel’s suggestion that these were elements of well-developed defensive systems.29 In conclusion, we must agree with ten Harkel’s observation that the study of Ringwalburgen in Zeeland will benefit from further comparative research. However, it seems to us that more archaeological work on the Ringwalburgen themselves is required. This is the case both concerning their role in the local economy, with respect to manufacturing and trade, as well as their use as fortified points. Also, demographic estimates must be ventured both in terms of the labor required to construct these fortifications and to defend them. Finally, more archaeological effort must be expended in regard to chronology, and it would seem necessary for archaeologists working in the region to examine the landscape more thoroughly in relation to information provided in the written sources regarding the construction of fortifications during the reign of Louis the Pious and others. Charlotte Hedenstierna-Jonson, researcher at Statens Historiska Museum, and Lena Holmquist, assistant professor at the Arkeologiska forskningslaboratoriet, 29 Concerning the order of magnitude of fighting men required for wall defense as laid out in the Burghal Hidage, see Bernard S. Bachrach and Rutherford Aris, “Military Technology and Garrison Organization: Some Observations on Anglo-Saxon Military Thinking in Light of the Burghal Hidage,” Technology and Culture 31 (1990), 1–17; and reprinted with the same pagination in Warfare and Military Organization.

259

M AT E R I A L S O U R C E S A N D I M A G E S

both in the University of Stockholm, along with Michaeol Olausson, curator and researcher at Stockholm’s Länsstyrelsen, have collaborated to author “The Viking Age Paradox: Continuity and Discontinuity of Fortifications and Defence Works in Eastern Scandinavia.” It is the aim of this essay to focus on the knotty question of chronological patterns in efforts to build fortifications and the abandonments of such efforts, or as the authors put it, discontinuity and continuity of fortifications in eastern Scandinavia. Their effort is aimed at coming to an understanding of what would seem to be the paradox of the landscape of defense in the region during the Viking Age, when there was little construction of fortifications despite the warlike nature of society. In their discussion of the pre-Viking era, Hedenstierna-Jonson et al. begin with an obvious paradox by claiming that a rampart does not “necessarily indicate military or defensive connotations” (286). While the English is not as clear as one might prefer, the authors would seem to want us to believe that the effort and expense entailed in the construction fortifications, or so we understand “ramparts,” were invested when no military purpose was at issue. They go on to assert: “Only a minor part of the enclosures represent actual fortifications” (286). Either something very important has been lost in translation or these observations constitute meaningless posturing. In short, the authors seem to be using the English word ramparts in ways that English speakers do not, and one wonders why the editors permitted this type of double-talk to remain uncorrected. However, after denying a military purpose for most ramparts, the authors go on to describe defense functions for these enclosures but seem to insist that they were constructed for purposes of social status. The basis for such conclusions is unclear. The paradox that most concerns the authors, however, is not the matter of constructing expensive fortifications for nonmilitary purposes in the pre-Viking era. Rather, their focus is on the lack of fortifications constructed or apparently reused during the Viking period in Scandinavia. Here Hedenstierna-Jonson et al. are struck by the fact that while operating within the region of the regnum Francorum and neighboring lands, the Vikings constructed fortifications, apparently for military purposes, but at home, by and large, this was not done for any purpose. They are adamant, however, in asserting that the lack of fortification cannot be taken as an indication of general peacefulness. However, an atmosphere of general peacefulness might be concluded from a study of the planned town of Sigtuna in the mid-tenth century, which the authors do not mention.30 In their discussion of Scandinavian operations in the regnum Francorum, the authors also do not consider the possibility that the Vikings learned to construct fortifications of the contemporary type from the Franks, who themselves were influenced by Roman imperial traditions. In this context, it is noteworthy that among the few Viking Age fortifications constructed in Scandinavia during the Viking era, the strongholds or 30 See Makt och människor i kungens Sigtuna. Sigtunautgrävningen 1988–90. 28 artiklar om de preliminära resultaten från den arkeologiska undersökningen i kv. Trädgårdsmästaren 9 och 10, ed. Sten Tesch (Sigtuna, 1990).

260

LANDSCAPES OF DEFENSE

“camps” built in Denmark, such as at Trelleburg, were measured out according to the Roman foot. In any case, the authors seem to have been led to claim on the basis of the rampart and pile barricades at Stegeborg that during the Viking Age there were efforts to develop a defense-in-depth strategy. Unfortunately, additional strongholds, roads, and signal stations simply have not yet been discovered to sustain such a model of a landscape of defense as evidence for supposed growing central power. The project, Strongholds and Fortification in Central Sweden, AD, 400–1100, initiated in 1998, may well provide evidence, especially if the necessary work is done on roads, signal stations, and place names, so that in the future more compelling evidence for a defense in depth may be developed. Any generalizations at this point with regard to landscapes of defense are premature as evidence to support arguments for centralized power. By contrast, the detailed studies that have exposed much of the emporium at Birka, the examination of which is the strongest part of the article, does provide support for the development of central authority capable of controlling trade, among other key aspects of government.

Some conclusions Landscapes of Defense is an especially important collection of essays for specialists in pre-Crusade military history and for archaeologists from outside England, who have much to learn from the works of specialists on the Anglo-Saxon period. Unfortunately, the editors of Landscapes chose not to provide a badly needed conclusion that relates the essays to each other and to outline areas that are important for future research, and so we have undertaken some aspects of that task here. The most important lesson of the volume taken as a whole, and a point that is made with particular cogency in the article by Stuart Brookes, is that the study of early medieval history, in general, and military history in particular, requires an interdisciplinary approach in which a wide range of sources of information are brought to bear to answer salient questions. Such an approach, however, should not be taken to mean that specialists in one field pick and choose individual pieces of information out of context that, as is often the case, suit the ends of a particular argument. Nor may historians simply provide a photograph or illustration of a site without explaining its meaning as if the example were self-explanatory. Rather, it is incumbent upon historians, archaeologists, numismatists, specialists in onomastics, and others to develop an understanding of the methodological imperatives and practices of other fields so as to be able to draw intelligently from studies beyond their own areas of specialization. In addition, it is crucial to understand that specialists in fields outside of history also have biases, and these must be understood. For example, efforts to diminish the importance of military motivations and to highlight what often euphemistically are labeled as social concerns distort our understanding of both. A second major lesson to be drawn from this collection is the essential role of economic resources, including most importantly demography, for the 261

M AT E R I A L S O U R C E S A N D I M A G E S

conceptualization and actualization of landscapes of defense. Here, it is necessary to understand demography both in relation to labor resources and men of fighting age and capacity.31 Several of the studies in this volume, most notably that by Richard Abels, draw attention to the massive expenditures in manpower, money, and materiel that were required for the military infrastructure of early medieval polities. In particular, Abels’s work makes clear that it is necessary to ascertain the costs of these efforts and to develop models for the construction of landscapes of defense. This approach has much in common with the practices of Sachkritik pioneered by the great German military historian Hans Delbrück. As we suggested earlier, however, much more research remains to be done. This volume also demonstrates that archaeologists have a great deal to teach historians about the ways in which material reality can illuminate both human action and human thought. Of particular importance is the understanding by archaeologists that material remains of a landscape of defense can be understood as illuminating planning undertaken by human agents. This insight has the possibility of revolutionizing historical inquiry regarding the intellectual activities of early medieval rulers and their advisors in the military sphere, who often were very well educated not only in regard to history but also were familiar with technical texts such as Vegetius’s Epitoma rei militari, Frontinus’s Stratagemata, and Vitruvius’s De architectura. This process of reading systematic planning into the material record may have the most impact for those periods in which the written word is particularly scarce. Indeed, one very compelling lesson for historians is that our surviving written sources provide information about only a small percentage of the fortifications that have been unearthed by archaeologists, and this percentage likely will become smaller as more excavations are undertaken.

31 See, for example, Bernard S. Bachrach, “Some Observations Regarding Barbarian Military Demography: Geiseric’s Census of 429 and Its Implications,” Journal of Medieval Military History 12 (2014), 1–37; and specifically with regard to military demography see Bachrach and Aris, “Military Technology and Garrison Organization,” 1–17.

262

12 T H E C O S T S O F F O RT R E S S C O N S T R U C T I O N I N T E N T HC E N T U RY G E R M A N Y The case of Hildagsburg

Introduction The ninth and tenth centuries witnessed a rapid increase in the number of fortifications that were constructed along the eastern frontiers of the Carolingian Empire and subsequent German kingdom.1 The kings of the Ottonian dynasty, Henry I (919–936), Otto I (936–973), Otto II (973–983), Otto III (983–1002), and Henry II (1002–1024), constructed several hundred fortifications both within the borders of the German kingdom itself and in lands that Henry I and his heirs conquered, especially in the east.2 1 An initial synthesis of the archaeological studies dealing with these fortress systems was published by Rafael von Uslar, Studien zu frühgeschichtlichen Befestigungen zwischen Nordsee und Alpen (Cologne, 1964). Subsequent regional syntheses include Rolf Gensen, “Frühmittelalterliche Burgen und Siedlungen in Nordhessen,” in Ausgrabungen in Deutschland, Monographien des RömischeGermanischen Zentralmuseums (Mainz, 1975), 313–37; Klaus Schwarz, “Der frühmittelalterliche Landesausbau in Nordost Bayern-archäologisch gesehen,” in Ausgrabungen in Deutschland, 338– 409; and Konrad Weidemann, “Archäologische Zeugnisse zur Eingliederung Hessens und Mainfranken in das Frankenreich vom 7. bis zum 9. Jahrhundert,” in Althessen im Frankenreich, ed. Walter Schlesinger (Sigmaringen, 1975), 95–119. The ongoing excavation of fortified sites, both in the German kingdom and in the lands to the east, including Bohemia and Poland, has led to new syntheses, including the collections of articles in Frühmittelalterliche Burgenbau im Mittel und Osteuropa, ed. Alexander Ruttkay and Joachim Henning (Bonn, 1998); Europa im 10. Jahrhundert. Archäologie einer Aufbruchzeit, ed. Joachim Henning (Mainz, 2002); and Joachim Henning and Matej Ruttkay, “Frühmittelalterliche Burgwälle und der mittleren Donau im ostmitteleuropäische Kontext: Ein deutsch-slowakisches Forschungsproject,” in Frühgeschichtliche Zentralorte in Mitteleuropa, ed. Jiri Machácek and Simon Ungerman (Bonn, 2011), 259–88. 2 Large numbers of fortifications appear in royal possession during the course of the tenth and early eleventh century in a wide range of written documents, including royal charters and narrative sources. In addition, excavations of numerous sites have unearthed fortifications that are not mentioned in any surviving sources, but which nevertheless can be identified as having been constructed by the royal government on the basis of their location, size, and the techniques used in their construction. A catalogue of more than 200 fortifications in royal possession over the period 919–1024 is available in David S. Bachrach, “Toward an Appraisal of the Wealth of the Ottonian Kings of Germany, 919–1024,” Viator 44 (2013), 1–27, here 24–7. In addition to the massive expenditure of resources

263

M AT E R I A L S O U R C E S A N D I M A G E S

The political and military importance of these fortresses have long been understood.3 By and large, however, scholars have neglected the economic and social implications of fortress construction on such a massive scale.4 In light of this glaring lacuna, the burden of the present study is to cast light, using archaeological evidence, on the broad range of costs that were involved in the construction of royal fortifications during the course of the tenth and early eleventh centuries, both within the German kingdom and in territory that the Ottonian kings conquered from the Slavs.5 The primary focus is on providing estimates regarding the expenditure of human, animal, and material resources that were of necessity involved in the effort to build the fortification at Hildagsburg, located 16 kilometers north of Magdeburg, which will be used here as a model. After establishing estimates of the costs involved in the construction of Hildagsburg, we will offer some tentative suggestions regarding the overall costs of the investment in fortifications by the kings of the Ottonian dynasty over the long century 919–1024.

General patterns of construction The efforts of the Ottonian kings to secure their control over the German kingdom and to protect their frontiers were accompanied by the construction of very large numbers of fortifications. In doing so, they followed the model of their

on fortifications by the Carolingian and Ottonian Empires, archaeological research also has demonstrated similar expenditure of resources by their Slavic opponents. To date, more than 600 Slavic fortifications have been identified in the area between the Elbe/Saale rivers and the Oder/Neiße rivers from this period, of which at least a third can be dated to the ninth century. See the observations by Sebastian Brather, “Karolingerzeitliche Befestigungbau im wilzisch-abodritischen Raum: Die sogenannten Feldberger Höhenburgen,” Frühmittelalterliche Burgenbau, 115–26, here 116. 3 See, for example, the important study by Werner Coblenz, “Boleslav Chrobry in Sachsen und die archäologische Quellen,” Slavic Antiqua 10 (1963), 249–85, which demonstrates the crucial role played by frontier fortifications in the fifteen-year military conflict between Duke Boleslav Chrobry of Poland (992–1025) and King Henry II of Germany (1002–1024). Also see the recent studies by Eric J. Goldberg, “Ludwig der Deutsche und Mähren. Eine Studie zu karolingischen Grenzekriegen im Osten,” in Ludwig der Deutsche und seine Zeit, ed. Wilfried Hartmann (Darmstadt, 2004), 67–94; David S. Bachrach, Warfare in Tenth-Century Germany (Woodbridge, 2012); and idem, “The Eastern Campaigns of King Henry II of Germany, 1003–1017,” Journal of Medieval Military History 14 (2020), 1–26, which trace out the important role played by fortifications in the conduct of war by the eastern Carolingian kings and their Ottonian successors. 4 One recent effort to address this imbalance in the scholarship is by Harriet Bönisch, and a team of archaeologists from the Brandenburg office for archaeological research, to calculate the quantity of wood that was required to construct the Slavic fortress at Raddusch, which was located some 250 kilometers east-southeast of Magdeburg in the region of Lower Lausatia. See Harriet Bönisch, “‘Ne Menge Holz’: Überlegungen zum Holzbedarf in altslawischer Zeit am Beispiel des Burgwalls Raddusch,” in Ausgrabungen im niederlausitzer Braunkohlenrevier (Wünsdorf, 2011), 191–207. 5 With regard to the Ottonian conquest of the lands east of the Elbe, see Herbert Ludat, An Elbe und Oder um das Jahr 1000. Skizzen zur Politik des Ottonenreichs und der slavischen Mächte in Mitteleuropa (Cologne, 1971, reprinted 1995).

264

T H E C O S T S O F F O RT R E S S C O N S T R U C T I O N

Carolingian predecessors, who also devoted enormous resources to frontier defenses.6 The Ottonian investment in fortifications was particularly intensive during the decades-long struggle to conquer and maintain their authority in the region between the Elbe-Saale frontier and the Oder river.7 These fortifications were built to a wide variety of specifications, usually dictated by the terrain, which dominated the area where they were constructed, and also in regard to the estimated enemy threat with which the particular fortresses were intended to deal. Nevertheless, we can state with considerable confidence that the walls, in general, averaged about 10 meters in height.8 Of course, there were exceptions. For example, the walls of Zehren, located on the left bank of the Elbe 6 kilometers downstream from the major royal fortress of Meißen, are calculated to have risen to a height of some 11 and a half meters, while the walls of Mecklenburg would seem to have risen only to some 8 meters.9 Ottonian fortifications that were constructed of earth and timber generally lacked significant numbers of mural towers or towers that were intended to defend the corners of a stronghold.10 By contrast, Ottonian fortifications that were protected by largely mortared stone walls did have mural towers. For example, the fortified royal palace at Magdeburg, the so-called Altenburg at Kinzgheim, the fortified episcopal seat at Hildesheim, and the Rhineland fortress at Altenburg am Neckar, all of which were constructed of mortared stone, also had mortared stone towers.11 6 See, in this context, Andrea Stieldorf, Marken und Markgrafen: Studien zur Grenzsicherung durch die fränkisch-deutschen Herrscher (Munich, 2012), who although emphasizing the royal duty to defend the frontiers, does not address the actual defense of the frontiers through the construction of fortifications, which was the primary activity of the Ottonian rulers, as well as their Carolingian predecessors. 7 Ibid. 8 The height of earth and timber walls, not including the palisade at their summit, whether of wood or of stone, ranged anywhere from 8 to 11 meters. See, for example, Werner Coblenz, “Wallgrabung auf dem Burgberg Zehren,” Ausgrabung und Funde 2 (1957), 41–5; Corpus archäologischer Quellen zur Frühgeschichte auf dem Gebiet der Deutschen Demokratischen Republik (7. bis 12. Jahrhundert), ed. Joachim Hermann and Peter Donat (Berlin, 1973), 11–12 with regard to the fortress at Mecklenburg; Volker Schmidt, Drense: Eine Hauptburg der Ukrane (Berlin, 1989), 16; and Heinz-Joachim Vogt, “Die Ausgrabungen auf der Wiprechtsburg in Groitzsch,” Ethnographischarchäologisch Zeitschrift 21 (1980), 695–710, here 697. The average height of the earth and timber walls of fortifications in the region of Lausatia, where the building techniques were very similar to those used at Hildagsburg, was 10 meters. See the observations on this point by Bönisch, “‘Ne Menge Holz’,” 194–5. 9 Coblenz, “Wallgrabung auf dem Burgberg Zehren,” 41–5; and Corpus archäologischer Quellen, 11–12. 10 There are, however, exceptions. For example, the earth and timber “Wolfgangwall” at Frauenberg near the monastery of Weltenburg had a mural tower. For a detailed discussion of this fortification, see M. Hensch, “Neue archäologische Aspekte zur mittelalterlichen Geschichte des Frauenbergs,” in Höhenbefestigung der Bronz-und Urnenfeldzeit. Der Frauenburg oberhalb Kloster Weltenburg, ed. M. M. Rind (Regensburg, 2006), 341–422. 11 Ernst Nickel, “Magdeburg in karolingisch-ottonischer Zeit,” Zeitschrift für Archäologie 7 (1973), 102–42, here 136; Reinhard Friedrich, “Ottonenzeitliche Befestigungen im Rheinland und im Rhein-Main-Gebiet,” Europa im 10. Jahrhundert, 351–63, here 358; Hans Wilhelm Heine,

265

M AT E R I A L S O U R C E S A N D I M A G E S

In addition to massive walls and, in some cases, mural towers, Ottonian strongholds, like those of their predecessors dating back to the Romans, have been shown to have employed complex defenses for the protection of gates, which were long-regarded as the most vulnerable part of any fortification.12 The example of Hildagsburg’s gate defenses, discussed later, provides a good example of the means by which entry into a stronghold could be protected. As a rule, Ottonian fortifications also were equipped with fosses, or defensive ditches, in much the same manner as Carolingian, Merovingian, and indeed, Roman fortifications.13 Some of these fosses can be considered true moats, as these were filled with water, at least in part, by the diversion of nearby rivers and streams through the application of high-quality hydrological engineering techniques.14 After taking account of the basic elements of these tenth and early eleventhcentury fortifications, the reader should be alerted to the fact that merely to sketch, as is done later, a preliminary estimate of some of the costs in both labor and resources, which were required to fortify any one of the installations constructed by the Ottonians, will, at best, convey a minimalist impression. This is the case, in part, because a great many thousands of relevant details, much less estimates of

“Frühmittelalterliche Burgen in Niedersachsen,” in Frühmittelalterliche Burgenbau, 137–49, here 139; Gösta Ditmar-Trauth, “Die Ausgrabung an der Großen Klosterstraße in Magdeburg (Gerberei, Uferbefestigung, Stadtmauer),” Jahresschrift für mitteldeutsche Vorgeschichte 86 (2003), 213–71; and Brigitte Kunz, “Die Befestigungen des Steilufers der Pfalzen Magdeburg und Duisburg im 10. Jahrhundert,” in Finden und Verstehen: Festschrift für thomas Weber zum sechzigsten Geburtstag, ed. Hans-Jürgen Beier et al. (Langen Weißbach, 2012), 299–302. 12 Regarding Roman gate defenses, see Stephen Johnson, Late Roman Fortifications (Totowa, NJ, 1983), 44–50; and in regard to those of the Ottonians see, for example, Hermann Schroller, “Die Ausgrabung der Pfalz Werla und ihre Probleme,” in Deutsche Königspfalzen: Beiträge zu ihrer historischen und archäologischen Erforschung, vol. 2 (Göttingen, 1965), 14–149, particularly 143–4; and Herbert Küas, “Die Leipziger Burg des 10. Jahrhunderts,” Arbeits und Forschungsberichte zur sächsischen Bodendenkmalspflege 20/21 (1976), 299–332, here 325. 13 See, for example, the discussion of the ditches that protected the Carolingian fortress at Bad Münstereifel, which was further strengthened and improved by the Ottonians in Friedrich, “Ottonenzeitliche Befestigungen im Rheinland,” 352; and also the system of ditches that protected the Carolingian fortification at Roßtal, which was similarly strengthened and improved under the Ottonians. See Peter Ettel, “Der Befestigungsbau im 10. Jahrhundert in Süddeutschland und die Rolle Ottos des Großen am Beispiel der Burg von Roßtal,” in Europa im 10. Jahrhundert, 365–79, particularly 366–7. 14 Regarding Roman examples, see Albert Grenier, Manuel d’archéologie gallo-romaine, VI.2 (Paris, 1935), 5, 330, 334, 407, 412–13, 416, 426; Jean-Pierre Adam, La construction romaine: matériaux et techniques 4th edition (Paris, 2005), 235–9; and L’Architecture de la Gaule romaine. Les fortifications militaires, ed. Michel Reddé et al. (Paris, 2006), 88–9. For German examples, see Paul Grimm, “Zur Lage der Pfalz Wallhausen, Kr. Sangerhausen,” Ausgrabungen und Funde 5 (1960), 33–5; Harald W. Mechelk, “Vorberichte zur Grabung in Magdeborn, Kr. Leipzig,” Ausgrabungen und Funde 10 (1965), 84–8; and Wolfgang Bünnig, “Ausgrabungen auf einem slawischen Fundplatz in Hohennauen, Kr. Rathenow,” Ausgrabungen und Funde 32 (1987), 82–9. Also see the important observations by Robert F. Heizer, “Ancient Heavy Transport, Methods and Achievements,” Science 153 (1966), 821–30, who demonstrates that complex building projects required sophisticated governmental organization and that this was the case even in prehistory.

266

T H E C O S T S O F F O RT R E S S C O N S T R U C T I O N

their costs, simply cannot be provided in a brief compass. In addition, all estimates of the labor and materials that were utilized in the construction of Hildagsburg here are made in the best case, i.e. the smallest quantities of materials and labor. We have done so in order to avoid the criticism that the expenses have been exaggerated as supposed proof of the great wealth and administrative skill of the Ottonians. With these warnings in mind, it is hoped that the reader will be impressed positively by the order of magnitude of the costly construction efforts undertaken by the Ottonian kings.

Building Ottonian fortresses Our understanding of governmental decision-making in regard to whether a particular place was to be fortified in a specific chronological context provides a brief glimpse of a rather complicated but rather poorly documented process. First, like the Carolingians, Merovingians, and Romans, the decision to build a new fortification, which obviously had serious military as well as political and economic implications, lay with the Ottonian king.15 Indeed, in many cases, the German monarchs, who traditionally led military campaigns, were on the spot when the decision was to be made to build a permanent fortification at a particular place.16 It is obvious that the complexity inherent in constructing a fortification, not to mention an effectively functioning moat, required, as will be seen later in detail, various engineering and architectural specialists, who were part of the royal entourage on campaign and attached to the central government.17 Such men certainly were very well educated. In this context, specialists in the history of early medieval education have pointed to the preservation of Roman as well as early medieval manuals and handbooks that were utilized by both architects and engineers in the pre-Crusade era. For example, copies of Vitruvius’s De architectura not only are well known to have been preserved and frequently recopied following the dissolution of Roman imperial authority in the West during the later fifth

15 The need to obtain a royal license to construct fortresses is stressed in numerous contemporary sources. See the discussion of this point by Bachrach, Warfare in Tenth-Century Germany, 97 and the sources cited there. 16 For example, King Henry I ordered the construction of the fortification at Meißen to replace the Slavic stronghold of Jana, which he sacked during the winter campaign of 929. Dendrochronological dating of the oldest wooden elements at Meißen indicate that the trees used in its construction were felled in 929, which strongly suggests that Henry ordered that the construction begin here immediately, even as he continued on his march northwards toward Prague. For the dating of the fortress, see Arne Schmid-Hecklau, Die archäologische Ausgrabungen auf dem Burgberg in Meißen: Die Grabungen 1959–1963 (Dresden, 2004), 205. 17 See the highly suggestive grant by Emperor Arnulf of Carinthia to his head of building operations (operarius) in 897 in Die Urkunden Arnolfs, ed. Paul Kehr (Berlin, 1940), nr. 152. The subject of military engineers recently has been addressed by Peter Purton, The Medieval Military Engineer: From the Roman Empire to the Sixteenth Century (Woodbridge, 2018), although he seriously underestimates the technological sophistication of military architecture in the early Middle Ages.

267

M AT E R I A L S O U R C E S A N D I M A G E S

century but also to have been utilized by both Carolingian and Ottonian builders.18 In addition, Roman surveying books, the agrimensores, were of exceptional value to engineers, architects, and surveyors throughout the early medieval period, including in the German kingdom under Ottonian rule.19 In the context of constructing a particular fortification, the initial step was for the king and his advisors to make a tactical and strategic assessment as to whether a stronghold was to be built and where this construction was to take place. After the military concerns were evaluated, the relevant government experts in surveying, architecture, and engineering were given the task to lay out the plan for the fortification and design its particular characteristics within the area that the king and his advisors wanted the structure to be built. In addition to being well-versed in the technical aspects of military construction, at least some of these specialists had to understand not only how the defenses of a particular stronghold were to be articulated in order to obtain maximum defensive efficiency but also how to evaluate and secure non-military resources within a unique environment: for example, where supplies of stone were to be found and how they could be transported to the building site.20 All of these considerations were ineluctable if, as was the case, fortifications were to be built and, once constructed, to survive.21 This process is not at all controversial in the contemporary construction of stone cathedrals, but historians and archaeologists up to this point largely have neglected the implications of this process for military architecture. As part of the process of mobilizing resources for the construction of fortifications, the number of skilled workers, e.g. masons, lime makers, oven and furnace builders, wood cutters, smiths, and nonskilled laborers of various types, had to be estimated. This was the case as well in regard to the number of draft animals and vehicles of various types, including barges and ships, that were needed to carry out the project. In addition, the quantities of food that would be needed to sustain both the men and the animals had to be estimated. Once all of these logistical estimates 18 With regard to the survival and use of Vitruvius’s text in both the Carolingian and Ottonian Empires, see the studies by P. L. Butzer, “Mathematics in the West and East from the Fifth to Tenth Centuries: An Overview,” in Science in Western and Eastern Civilizations in Carolingian Times, ed. P. L. Butzer and D. Lohrmann (Basel, 1993), 443–81, here 453; and Bernhard Bischoff, “Die Überlieferung der technischen Literature,” in Artigianato e tecnica nella società dell’alto medioevo occidentale, Centro italiano di studi sull’alto medioevo (Spoleto, 1971), 267–97, here 274. 19 See J. B. Campbell, The Writings of the Roman Land Surveyors: Introduction, Text, Translation and Commentary (London, 2000). 20 In this context, Markus C. Blaich and Henning Zellmer, “Die ottonische Pfalz Werla-Überlegungen zu Baugrund und Baugestein,” in Landschaft lesen lernen: 12. Internationel Jahrestag der Fachsektion GeoTop der deutschen Gessellschaft für Geowissenschaften (Hannover, 2008), 27–39, provides an important model for hauling costs for bringing stone to the building site at the Ottonian royal palace at Werla. 21 The histories of Ottonian fortifications, where they can be traced, make clear that they were built to last. This is made evident by the fact that for most of these strongholds, the first written evidence for their existence long past-dates their actual construction. This pattern is particularly clear in the case of Hildagsburg. For a fuller discussion of this issue, see the appendix later.

268

T H E C O S T S O F F O RT R E S S C O N S T R U C T I O N

had been established, the royal government then had to mobilize the resources of the royal fisc, church holdings, as well as labor and material obligations on the broader population, which were then to be deployed to the construction site.22 Finally, if the new stronghold were to be constructed in a dangerous area, such as a newly conquered region on the eastern frontier, troops had to be deployed in order to defend the work site if necessary.

Hildagsburg: the archaeological record As noted previously, we have chosen for our analysis the Ottonian-era fortifications at Hildagsburg, which was located at Elbeu some 16 kilometers north of the Ottonian “capital” at Magdeburg, along a branch of the Elbe river called the Renze (Ronne).23 This site, including the fortifications, benefitted from an initial evaluation at the hands of a talented amateur archaeologist named Heinrich Wilhelm Schultheiß when a train line was to be constructed through the site in 1849–1850.24 A second detailed inventory of the topography of the site was undertaken in the period 1926–1929 by professional archaeologists employed by the state archaeological service when a canal was to be built through the area.25 The initial excavations were undertaken by Hans Dunker, with subsequent excavations carried out by Otto Hahne, Walter Schultz, Alt Albrecht, and Paul Grimm, who later would become one of the leading archaeologists in the German Democratic Republic.26

22 The Ottonian kings exercised very close control over their fiscal resources for logistical purposes and worked diligently to assure that they had accurate and up-to-date accounts of the stocks of food and other supplies that were available for use by the royal court and by the army. See the treatment of this topic by David S. Bachrach, “Exercise of Royal Power in Early Medieval Europe: The Case of Otto the Great 936–973,” Early Medieval Europe 17 (2009), 204–20; and idem, “Feeding the Host: The Ottonian Royal Fisc in Military Perspective,” Studies in Medieval and Renaissance History 3rd series 9 (2012), 1–43. Now also see Idem, “The Benefices of Counts and the Fate of the Comital Office in Carolingian East Francia and Ottonian Germany,” Zeitschrift der SavignyStiftung für Rechtsgeschichte, germanistische Abteilung 126 (2019), 1–50. 23 The basic study is Hans Dunker, Die Hildagsburg: Der Burgwall von Elbeu, Kreis Wolmirstedt, with an introduction by Paul Grimm (Berlin, 1953). For the date of the construction of the fortress at Hildagsburg, see the appendix. New excavations have begun at Hildagsburg, but to date the published findings have not altered the basic conclusions drawn by Dunker. See Ottilie Blum, “Hildagsburg und Schlossberg Wolmirstedt-zwei frühgeschichtliche-mittelalterliche Burgen im Elbraum nördlich Magdeburgs,” in Soziale Gruppen und Gesellschaftsstrukturen im westslavischen Raum: Beiträge der Sektion zur slawischen Frühgeschichte des 20. Jahrestagung des Mittel-und Ostdeutschen Verbandes für Altertumsforschung in Brandenburg (Havel), ed. Felix Paul Biermann (Langenweißbach, 2013), 247–54; and idem, “Eine kruezformige Fibel von der Hildagsburg bei Wolmirstedt-Elbeu,” in Die frühen Slawen – von der Expansion zu gentes und nationes, ed. Paul Felix Biermann, Thomas Kersting, and Anne Klammt (Langenweißbach, 2016), 165–72. 24 Dunker, Die Hildagsburg, 192. 25 Ibid. 26 Ibid., 193.

269

M AT E R I A L S O U R C E S A N D I M A G E S

The main fortress at Hildagsburg As with many mid-to-large Ottonian fortifications, Hildagsburg was constructed as a “double fortress.” Its main defenses consisted of a walled central structure, called a Hauptburg in German, which was protected further by a large earthen berm and two water-filled moats. The Hauptburg enclosed an area of approximately 2 hectares. There was also an outer fortification or bailey called a Vorburg in German, which enclosed an area of 7 hectares. The earth and timber wall that enclosed the Hauptburg was approximately 600 meters in length and was some 15 meters wide, both at its base and at its summit, i.e. it was not tapered, as was the tradition with stone walls.27 This wall is estimated to have reached a height of 10 meters.28 Along three-fourths of its length, this earth and timber structure was surmounted along the front side of its 15-meter width by a mortared-stone upper wall, which was some 450 meters in length. On the south and west, the earth and timber subwall also was surmounted on its rear side by a second mortared stone wall some 150 meters in length.29 The space between the two stone walls along the top of the earthen subwall was approximately 10 meters in width. Along the eastern side of the earth and timber wall, which faced the Renze river, there is no evidence of mortared stone, and this portion of the wall likely was surmounted only by a wooden palisade.30 The mortared stone components of the Hauptburg, i.e. those that surmounted the front and back side of the earth and timber substructure, measured a total of some 600 meters in length.31 Archaeological excavation has shown that the foundations of the stone wall were embedded in the earth and wooden base upon which they were constructed to a depth of approximately 70 centimeters.32 In light of Roman usage as discussed 27 Ibid., 219. 28 Ibid., 192 and 201. In this context, while measurement of the width of this wall at its base presented archaeologists with no difficulties, estimating the height was problematic. These estimates rest both on comparative evidence (see later) and the pattern of debris available to Dunker and his team when they carried out their studies. In addition, Dunker had available unpublished observations of men in the late nineteenth century who reported that after a 1,000 years, the walls of the Hauptburg were still more than 5 meters in height (pp. 192, 201). Recent investigations of fortifications in the region of Lausatia, approximately 100 kilometers southeast of Hildagsburg, indicate that earth and timber fortifications constructed there during the ninth and tenth centuries, including those utilized by the Ottonians, often reached 10 meters in height, excluding their parapets, as discussed earlier. Wood comprised approximately half of the height of these fortifications. As this wood disintegrated over time, the walls were reduced in height to between 5 and 6 meters. At Hildagsburg, the vestigial remains of both horizontal and vertically oriented timbers discovered by the excavators indicate that this earthen wall similarly was reinforced by a wooden internal structure that also “shrank” as a result of the rotting of the timbers used in its construction. On this point, see Dunker, Die Hildagsburg, 219. 29 Ibid., Tafel 41. 30 No evidence was found of a mortared stone addition to the earth and timber subwall along the eastern side of the Hauptburg, which faced the Renze river. See Dunker, Die Hildagsburg, 219. 31 This is a minimum estimate for the stone wall based on Dunker’s incomplete excavation. 32 Dunker, Die Hildagsburg, 220.

270

T H E C O S T S O F F O RT R E S S C O N S T R U C T I O N

by Vitruvius, whose work, as noted previously, was well known during the early Middle Ages, and also taking into account the modern “rule of thumb” regarding this matter, we find it likely that the height of the stone wall at Hildagsburg did not exceed a ratio of three times the depth of its foundation.33 As a consequence, we see the stone walls rising about 2 meters or so above the earth and wood foundation upon which they were constructed. The thickness of these stone walls is to be estimated at approximately 2 meters in width on the basis of the well-preserved stone foundation.34

The berm There was a distance of approximately 6 meters between Hildagburg’s Hauptburg and the first of three water-carrying moats that were constructed to protect the fortress as a whole.35 This space between the wall and the moat was filled by an earthen berm. However, it is unlikely that this manmade topographical feature had been created primarily for defensive purposes, as the walls rose 6 or 7 meters above it. The earth from which the berm was constructed likely was excavated initially from the moat. The berm probably grew in height largely as a result of the need from time to time to dredge the moat. However, the berm did have value to the defenders insofar as it raised the thither bank of the moat from the perspective of an attacking force. Once potential attackers negotiated the moat, the height they had to climb in order to get out of the moat was considerably higher than what would have been the case were there no berm.

The Hauptburg moats As mentioned earlier, there were three water-filled moats protecting the complex of fortifications at Hildagsburg. Two of these moats were an integral part of the defenses of the Hauptburg.36 Just beyond the berm, as discussed, was the innermost moat, i.e. the one closest to the walls of the Hauptburg itself. This moat averaged 8 meters in width, and its trench or ditch was “U” shaped in form, i.e. it was smaller at the bottom than at the surface. The inner moat was approximately 450 meters in length, and its average depth was approximately 2 meters.37 The 33 For the modern treatment of this issue, see “Masonry Fence Walls and Retaining Walls,” a pamphlet issued by the city of Mesa Arizona (p. 16) that the ratio of the foundation to wall in brick structures is to be 3:1. By contrast, Vitruvius, De architectura I.v.i, advised that the depth of the foundation should be approximately one-fifth to one-sixth of the height of the wall. However, in order to maintain a minimalist impression of the overall labor and costs involved in the construction of the mortared stone wall at Hildagsburg, we have decided to use the lower estimate of the height of the wall as implied by the modern formula. 34 Dunker, Die Hildagsburg, 200. 35 Ibid., 219. 36 Ibid., 220 and the sketch of the moat system in Plan 7, Tafel 41. 37 Ibid., 194.

271

M AT E R I A L S O U R C E S A N D I M A G E S

outer moat was 4 meters wide, and thus only about half the width of the inner moat, but was somewhat longer, at 470 meters. Like the inner moat, the outer moat ran through a “U”-shaped trench, which reached on average a depth of 2 meters.38 On the strip of land between the two moats, the Ottonians constructed a second berm. However, the excavation at Hildagsburg was not able to determine the details of its construction, and it would not be prudent to suggest that it had a wooden infrastructure of a palisade. Both the inner and the outer moats drew their water from the Renze river that flowed northward along the east side of the Hauptburg. The water passed directly into the ditches of the two moats to fill a double defensive system along the southern, western, and northern sides of the main fortification. In fact, the Ottonian engineers created a continuous hydrological system to feed the moats, and the water, after going through the northern part of the two moats, flowed back into the Renze.39 In order to avoid erosion at key points where the water flow was vigorous, the engineers oversaw the construction of a wooden embankment with pilings at the southeastern “corner” of the Hauptburg, where the river entered the inner moat. A somewhat smaller system of pilings was constructed at the point where the river entered the outer moat.40 One can only marvel at the work accomplished by the men who designed this effective system.41

The entry system Gate complexes were the most vulnerable element of a walled defensive system because they were essentially an already existing “hole” in the wall. The gate system at Hildagsburg was particularly well protected, which might be expected given the high level of engineering skill seen in the construction of the moat system. A wooden bridge extending from beyond the outer moat of the Hauptburg rose up to an opening in the forward stone wall, which was constructed atop the earth and timber subwall, discussed earlier. This opening was at least 4 meters in width, i.e. large enough to permit a large wagon to pass through without difficulty.42 The path into the fortress then progressed approximately 30 meters eastwards, between the forward and rearward stone walls that surmounted the earth and timber subwall of the Hauptburg. The path into the fortress then passed through a gated opening in the southeastern “corner” of the rearward stone wall that also was at least 38 39 40 41

Ibid. Ibid., 220. Ibid. For similar kinds of expertise in the management of large flows of water, see Robert Koch, “Fossa Carolina. Neue Erkentnisse zum Schiffahrtskanal Karls des Großen,” in Häfen, Schiffe, Wasserwege: Zur Schiffahrt des Mittelalters, ed. Konrad Elmshäuser (Hamburg, 2002), 54–70, who makes clear, on the basis of newly undertaken excavations, that despite its ultimate failure, Charlemagne’s canal project demonstrated enormous expertise in hydrological matters. 42 The postulated width of the gate is based upon the estimate by Dunker, Die Hildagsburg, 220, that the bridge that approached the gate was a minimum of 4 meters in width.

272

T H E C O S T S O F F O RT R E S S C O N S T R U C T I O N

4 meters in width. The wooden ramp, comparable in length to the wooden bridge that afforded initial access to the upper wall, then extended into the courtyard of the Hauptburg. The opening in the rearward stone wall for the gate was protected by two wooden towers. These were constructed on both sides of the gate on the inner side of the Hauptburg wall. It is not entirely clear how, or whether, these wooden towers were connected to the wall of the Hauptburg.43 We estimate that these wooden towers reached a minimum of 14 meters in height in light of the fact that the wall next to which they stood was 12 meters in height, including the earth and timber subwall and the mortared stone upper wall. The wooden bridge over the water defenses of the Hauptburg was anchored at least 54 meters beyond the outer moat on the southeastern side of the Hauptburg. The ramp that extended from the wall of the Hauptburg into the courtyard similarly was at least 54 meters in length. The length of the ramps is calculated on the basis of the need to have a maximum of a 12.5% grade for the movement of wagons or carts to reach the bridge passageway atop the Hauptburg wall.44

The Vorburg At Hildagsburg, the second and larger element of the fortification in terms of surface area was the Vorburg, or bailey. The defenses of the bailey were laid out in a semicircular pattern, which varied in distance from between 120 meters and 140 meters from the walls of the main fortress. The bailey’s walls stretched a total of 750 meters from the bank of the Renze river on the south side of the Hildagsburg fortress complex to the bank of the river north of the fortress. As a whole, the bailey had a surface area of 7 hectares, which was approximately three and a half times larger than the surface area of the Hauptburg.45 In light of the area enclosed by the walls of the bailey, some 2,500 men, women, and children could find comfortable refuge there for a long period should the area as a whole come under enemy attack. For shorter periods, far more refugees could be accommodated.46

43 Ibid., 197. 44 The bridge at Hildagsburg reached a height of 10 meters as it surmounted the earth and timber subwall. In order to accommodate a slope that would permit ingress by wheeled vehicles, the length of the bridge had to have a minimum ratio of 8:1 for distance over height. Thus, inclusive of the 26-meter width of the two moats and berms that protected the walls of the Hauptburg, the bridge had to extend a distance of 54 meters into the territory of the Vorburg. In identifying the minimum length of the bridge at 13 meters (220), Dunker did not account for the ramps that would be necessary to permit an accessible gradient for traffic. 45 Dunker, Die Hildagsburg, 192. 46 For comparative purposes, the fortress of Eresburg, with a protected area of 28 hectares, could house approximately 10,000 people. See Carl Schuchhardt, Die frügeschichtlichen Befestigungen in Niedersachsen (Berlin, 1925), 39. By analogy, the 7 hectares of the Vorburg at Hildagsburg could accommodate approximately 2,500 people.

273

M AT E R I A L S O U R C E S A N D I M A G E S

The bailey defenses The entire semicircular area of the bailey was protected by an earthen wall that was 10 meters wide and reached a minimum of 2 meters in height, but probably was higher than this.47 No surviving wooden elements were discovered in the excavations of the Vorburg wall.48 Beyond the earthen wall of the Vorburg was another earthen berm that was 3 meters in width at its base.49 It likely was intended to serve the same purpose as the berm that had been constructed between the main wall of Hildagburg’s Hauptburg and the first moat. This berm, as will be recalled, likely had been constructed from earth that had been excavated from the moat and was intended to raise the thither bank of the later, as seen from the perspective of the attacking force. As alluded to previously, the third and last of Hildagburg’s moats was dug to increase the defensibility of the Vorburg wall that encloses the bailey. The ditch for the moat followed the contours of the wall and was more than 800 meters in length and 11 meters in width. Like the other two moats discussed previously, it was constructed in a “U” shape.50 The excavators of this moat found that it was approximately 2 meters deep, with the result that the enemy forces could not wade across it.51 The water for this third moat, as was true for the other two, was drawn from the Renze river in a hydrological scheme that was no less technologically complex than that which had been used in regard to the other two moats.52 The sophistication of the hydrological learning demonstrated by these Ottonian engineers from the artifacts that we can study at Hildagsburg is both remarkable and prima facie evidence of the continuity of knowledge from the later Roman era into the early Middle Ages.

Methodology The basic method of research in regard to the estimation of construction costs is, in fact, rather simple, although arriving at plausible results is considerably more difficult. First, one selects a structure, such as Hildagsburg, concerning which a considerable quantity of information is known due to effective archaeological and/or architectural studies.53 From these data a “blueprint” can be developed that

47 The stone wall survived into modern times to a height of 1.75 meters. See Dunker, Die Hildagsburg, 192 and 197. 48 Ibid. 49 Ibid., 197. 50 Ibid. 51 Ibid. 52 Ibid. 53 See the discussion of this method by Bernard S. Bachrach, “The Cost of Castle-Building: The Case of the Tower of Langeais, 992–994,” in The Medieval Castle: Romance and Reality, ed. K. Reyerson and F. Powe (Dubuque, 1984), 46–62 and reprinted with the same pagination in Warfare and Military Organization in Pre-Crusade Europe (London, 2002); and idem, “The Fortification

274

T H E C O S T S O F F O RT R E S S C O N S T R U C T I O N

is comparable to, but hardly as detailed as, a document executed prior to construction by a modern architect or building contractor. Then, like a modern construction engineer with blueprints and reference works at hand, the historical researcher estimates the costs of the construction, taking into consideration the details that he has been trained to treat in order to make an accurate estimate of what it would cost to build the structure. Accuracy in arriving at cost estimates for the construction engineer is crucial so that a profit can be made. For the historian, accuracy is essential so that the cost to society can be assessed.54 Of course, estimating late antique or medieval building costs must be done in consonance with the technology that was in use at the time, when possible, or at least prior to the invention of various mechanical techniques developed in more modern times.55 Nevertheless, it is also important to recognize that there are at least some important constants. For example, neither the weight of a cubic meter of granite nor the weight of a cubic meter of oak has changed in any noteworthy manner over the centuries, although there are, of course, different grades of many types of stone and different types of oak, and its relative desiccation is of importance as well.56 Moreover, as in past centuries, human beings, depending on size, weight, and age, who are engaged in heavy construction work require, in addition to comparable quantities of water, approximately the same quantities of carbohydrates and proteins to maintain energy levels and remain reasonably healthy.57 Even draft animals, e.g. horses, mules, and oxen, when hauling heavy loads required quantities of food and water in the late antique and early Middle Ages similar to what they need in more modern times adjusting, of course, for size, climate, and the nature of the work being undertaken.58 Perhaps more surprising

54 55

56 57

58

of Gaul and the Economy of the Third and Fourth Centuries,” Journal of Late Antiquity 3 (2010), 38–64. Also of great value in developing models for building estimates is Janet DeLaine, The Baths of Caracalla: A Study in the Design, Construction, and Economics of Large-Scale Building Projects in Imperial Rome (Portsmouth, 1997). Bachrach, “The Cost of Castle-Building,” 52–3. Of considerable importance is the pioneering work of Hans Hubert Hofmann, “Fossa Carolina Versuch einer Zusammenschau,” in Karl der Grosse: Lebenswerk und Nachleben, ed. W. Braunfels and Helmut Beumann (Düsseldorf, 1965), 437–53. Since the appearance of his work, there has been much fruitful research in estimating costs by scholars of the ancient world and late antiquity. See D. J. Breeze and B. Dobson, Hadrian’s Wall (London, 1987); Andrew Pearson, The Construction of the Saxon Shore Forts (Oxford, 2003), 84–111; Roger Kendal, “Transport Logistics Associated with the Building of Hadrian’s Wall,” Britannia 27 (1996), 129–52; Delaine, The Baths of Caracalla, 85–224; A. Simon Esmonde Cleary and Jason Wood, Saint-Bertrand-de Comminges III, Le rampart de l’Antiquité tardive de la ville haute (Bordeaux, 2006), 43–146, 215–16; and Bachrach, “The Fortification of Gaul,” 38–64. See Adam, La construction romaine, 24 for stone, and 91–2 with regard to various types of wood. For a vast array of data concerning work, rest, and food under a substantial variety of conditions, see Colin Clark and Margaret Haswell, The Economics of Subsistence Agriculture, 4th edition (London, 1970). Jonathan P. Roth, The Logistics of the Roman Army at War (264 B.C.-A.D. 235) (Leiden, 1999), 66–7 for oxen and 62–5 for horses; and Donald W. Engels, Alexander the Great and the Logistics of the Macedonian Army (Berkeley, 1978), 12–16 for oxen, and 126–9 for horses.

275

M AT E R I A L S O U R C E S A N D I M A G E S

is the fact that some purely physical activities such as mixing mortar by hand and laying brick or stone by the hand of a mason has not changed over time, although the modern worker often can be more rapidly supplied with building materials on the scaffold than his earlier counterpart.59 As will become clear, in total, the building of the highly complex wood, earth, and stone fortifications at Hildagsburg required, at a minimum, between 2 and 3 million man-hours of labor from literally scores of craft specialties. It would require a monograph-length work merely to describe in detail all of the interrelated administrative, logistical, and technical challenges that obviously were met and overcome in the course of the construction of Hildagburg’s defenses. Such an investigation, for example, would have to deal with the matter of whether the area in which the fortress was constructed was forested or even heavily forested, or already cleared of large numbers of trees when the decision was made to begin building there.60 Unfortunately, the necessary pollen studies have not been undertaken. Consequently, given the limitation of time, space, and the reader’s patience, the following pages will provide a brief introduction to a rather limited number of the “costs” that were involved in order to bring the construction of the Hildagsburg fortifications to a successful conclusion. Before beginning the discussion of the efforts that were involved in the Ottonian construction of the fortress at Hildagsburg, it is appropriate to note that various teams of excavators have found that this structure was independent of the Slavic fortifications that were built on a part of this site during the later eighth century. This latter site was abandoned, likely under pressure from Charlemagne, before the turn of the eighth century.61 As is clear from the excavation reports published by Dunker, the earthen elements of the Slavic fortress, insofar as they had not eroded, undoubtedly were cleared to make way for the new Ottonian fortress. The latter, as will be discussed later, was constructed in a lattice-type framework of logs, which were filled in with earth. Some of the earth that had been used to construct the Slavic fortress could have been used for this purpose, although it still had to be dug up and moved. Therefore, this did not diminish the labor inputs that were necessary to accomplish the task of filling in the lattice.

Digging ditches From the few figures provided earlier, and with additional archaeological work perhaps more could be generated, it is obvious that more than 30,000 cubic meters

59 Bachrach, “The Cost of Castle-Building,” 31–2. 60 Bönisch, “‘Ne Menge Holz’,” 191–207 makes clear that although large numbers of trees were cut down to build the fortifications at Raddusch, this did not amount to deforestation. Although the number of trees utilized at Hildagsburg was far larger than at Raddusch, we believe that deforestation was not a concern in this instance either. 61 Dunker, Die Hildagsburg, 209.

276

T H E C O S T S O F F O RT R E S S C O N S T R U C T I O N

of earth, stones, and vegetation were removed in the process of excavating the ditches for all of the three moats combined that were constructed to protect the Hauptburg and the Vorburg of Hildagsburg. The inner moat was 450 meters in length, 8 meters wide, and 2 meters deep. The outer moat was 470 meters in length, 4 meters wide, and 2 meters deep. Finally, the third moat was approximately 800 meters in length, 11 meters in width, and 2 meters in depth. In addition, it seems unlikely that sufficient earth was excavated from the first and third ditches in order to construct the respective berms that were built up behind each of these moats. Therefore, the excavation of 30,000 cubic meters of earth and other material should be understood as a minimalist figure. Under what might be considered normal conditions, a cubic meter of excavated earth, rocks, and vegetation weighs approximately one metric ton.62 In ideal conditions, one experienced construction worker, using tools of the period and with appropriate support provided by what we will refer to as an earth removal team, is estimated to have been able to excavate approximately 3 metric tons in the course of a ten-hour work day.63 Therefore, in the circumstances at Hildagsburg, digging the “ditches” for the three moats, i.e. excavating 30,000 cubic meters of earth, rocks, and vegetation, required approximately 3,000 ten-hour work days for the diggers and an additional 3,000 ten-hour work days for their helpers. To put this figure into more manageable terms, 100 men digging under optimal conditions could do the necessary excavation for the moats in 30 days, if they were properly supported by 100 additional men. However, optimal conditions rarely obtain for lengthy periods of time in the course of carrying out complicated construction projects, even under modern conditions.64

The wooden walls of the Hauptburg The timber elements identified by the excavators in the subwall of the Hauptburg provide compelling evidence that the builders used an internal wooden framework.65 A common type of framework utilized by both Ottonian and contemporary Slavic fortress builders was constructed of overlapping layers of wooden logs with empty spaces filled in with earth.66 However, as is normal in regard to the haphazard survival of wooden remains, some Ottonian and Slavic fortifications

62 These estimates are based upon modern figures and certainly must be considered averages. See Bachrach, “The Cost of Castle-Building,” 51–2. 63 Cf. Hofmann, “Fossa Carolina,” 345 for a slower rate; and Esmonde Cleary and Wood, SaintBertrand-de-Comminges, 143–6, 215–16 for faster rates. 64 Bachrach, “The Fortification of Gaul,” 58–9. 65 At Hildagsburg, the vestigial remains of both horizontal and vertically oriented timbers discovered by the excavators indicate that this earthen wall was similarly reinforced by a wooden infrastructure of “cells”. Dunker, Die Hildagsburg, 219. 66 See Bönisch, “‘Ne Menge Holz’,” 191–207.

277

M AT E R I A L S O U R C E S A N D I M A G E S

provide more complete information than others regarding the engineering details of such walls.67 In regard to the remains of the wooden structures used for the walls and fortifications during this period, one of the best-preserved examples is the stronghold at Raddusch. In the context of her detailed examination of this site, Harriet Bönisch and her team estimated that somewhere in the neighborhood of 1,200 to 1,600 oaks were used in the construction of a wood and earthen wall structure that was approximately 144 meters in length and approximately 11 meters in width. These oaks, the archaeologists have concluded, ranged in length between 11 and 14 meters, once appropriately trimmed of their too narrow tops, and were between 15 and 60 centimeters in diameter. These figures used for the timbers at Raddusch, when averaged, come to 12.5 meters in length and 37.5 centimeters in diameter.68 Based on the surviving wooden remains, it is likely that the 600-meter subwall of the Hauptburg at Hildagsburg was constructed following engineering principles very similar to those used at Raddusch and many other contemporary and near-contemporary fortifications such as Wiprechtsburg at Groitzch, the stronghold at Gars/Thunau, the “Landeskrone” near Görlitz, and the fortress at Lenzen.69 Simple arithmetic makes clear that the relevant walls at Hildagsburg, which not only were much longer than the sample from Raddusch but also considerably higher and wider, were some 800% greater in volume than those at the fortification excavated by Bönisch and her team. Therefore, it seems reasonable that the walls under consideration here at Hildagsburg required some eight times the number of trees, i.e. an average of about 11,000 rather mature oaks, with an average diameter of 37.5 centimeters and a length of 12.5 meters.

Costs of wood cutting To fell approximately 11,000 rather mature oak trees was not an easy task, nor could it be accomplished without large numbers of trained wood cutters, who had the proper tools available to them.70 The importance of experienced woodcutters cannot be overestimated, as both great skill is required of the men and

67 Vogt, “Die Ausgrabungen auf der Wiprechtsburg,” 696; Erik Szameit, “Zum frühmittelalterlichen Burgwall von Gars/Thunau. Bemerkungen zu den Fortifikationsresten und der Innenbebauung. Ein Vorbericht,” in Frühmittelalterliche Burgenbau, 71–8, here 74; Jaspar von Richthofen, “Die Landeskrone bei Görlitz- eine bedeutende slawische Befestigung der östlichen Oberlausitz,” Arbeits-und Forschungsberichte zur sächsischen Bodendenkmalpflege 45 (2003), 263–300, here 293; and Heike Kennecke, “Forschung Burg Lenzen, Wo Slawen gegen Sachsen kämpften,” Archäologie in Deutschland 22 (2006), 8–13, here 12. 68 Bönisch, “‘Ne Menge Holz’,” 193–8. 69 See note 65 earlier. 70 Bachrach, “Cost of Castle-Building,” passim; and Carroll Gillmore, “Naval Logistics of the CrossChannel Operation, 1066,” Anglo-Norman Studies 7 (1984), 105–31, especially 117–18, which provide much of the information used in the following pages regarding the use of timber.

278

T H E C O S T S O F F O RT R E S S C O N S T R U C T I O N

the work itself is highly dangerous.71 As with estimates for excavations of earth, calculated previously, initially we must use modern information for comparative purposes. Then we can make allowances for various inferiorities inherent in medieval technology, e.g. lower-quality iron or steel for axes. For example, an experienced woodcutter with a modern high-quality steel axe, can fell in the course of a normal work day, three oak trees, each with a 30-centimeter diameter.72 Thus, to fell 11,000 oaks would require approximately 3,700 man-days of labor. With axes of softer metal than those available to modern woodcutters but with assistants to keep the axes sharp, it is suggested here that the men who cut the wood for the fortification of Hildagsburg also likely were able to fell three trees per day, although with double the number of personnel, i.e. approximately 7,400 man-days, each of ten hours each. If the project manager at Hildagsburg employed 100 experienced wood cutters, these 11,000 oaks could be felled in somewhat less than 40 days, under ideal conditions, i.e. with no rest for the men during the work day and no weather conditions that undermined the pace of the work. It should be noted that the 100 wood cutters required 100 assistants, who were charged primarily with keeping the axes sharp, but these men also could have been employed to strip the branches from the felled trees and to cut off their tops at the point where they were no longer useful for inclusion in the construction of the wall. It is not possible to identify the sites from which the trees utilized in the construction of Hildagsburg were brought. As a consequence, developing a model for the labor involved in their transport cannot be undertaken here. However, for heuristic purposes, it is worth noting that different scenarios for the movement of these trees required drastically different labor inputs. The simplest method, if available, would have been to cut down oak trees in the immediate vicinity of the Elbe river, to roll them down to the riverbank, and then to float them downstream to the construction site.73 By contrast, if the forests from which the oaks were taken were located further afield, it would have been necessary to haul or drag these trees over various distances to the river, perhaps with the concomitant

71 Regarding these issues, see Charles F. Carroll, The Timber Economy of Puritan New England (Providence, 1973), 45, 49–50, 62, who discusses the considerable difficulties faced by colonists in New England who lacked timber cutting skills. 72 E. Legrand-Girard and H. Plessix, Manuel complet de fortification (Paris, 1909), 198. This figure is accepted by J. F. Finó, Fortresses de la France medieval, 3rd edition (Paris, 1977), 83; and Gillmore, “Naval Logistics,” 117. For additional information, see Adam, La construction romaine, 93–4. 73 Michael Rosholt, The Wisconsin Logging Book: 1839–1939 (Rosholt, 1980), 120–42, discusses the large numbers of men who were required to drive logs on a river. Particular attention is paid to log jams, which can cause a blockage on the river and completely stop movement down river. In more modern times, it was found necessary on occasion to blow up the jams with dynamite. Agnes M. Larson, The White Pine Industry in Minnesota (Minneapolis, 1949), 188, observes that it could take 200 men as long as six weeks to break up a big log jam. When everything was working well on the river, a big collection of logs could be moved at a pace of 1 mile per day (190).

279

M AT E R I A L S O U R C E S A N D I M A G E S

necessity of constructing lumber roads to facilitate the movement of tree trunks over rough or overgrown territory. In the latter case, the labor involved easily could have surpassed many thousands of man-days, as well as the deployment of hundreds of draft animals.74

Constructing the subwall of the Hauptburg Approximately 11,000 oak trees with an average length of 12.5 meters and an average diameter of 37.5 centimeters likely were used in the construction of the subwall of the Hauptburg defenses. In light of the surviving examples of trees that were utilized in the construction of the fortress at Raddusch, discussed earlier, it would appear that it was common practice to utilize preexisting knots and vestigial protrusions where tree limbs were cut off rather than notches to keep the tree trunks in place once they were set down to form “cribs” that subsequently were filled with earth.75 As is true of several other aspects of this building project, it is necessary to draw some inferences from modern experience to develop a model for the total number of man-hours that likely were necessary to build the wood framework in which earth was layered to form the subwall of the Hauptburg. This framework can be conceptualized as a series of interconnected but roofless “log cabins.” In this context, it is still quite common to construct log cabins today utilizing technology such as ropes and ramps, as well as animal power, including oxen, that are quite comparable to those devices and animals that were available for use during the tenth century, although the animals from this earlier period were smaller than those available today. The following model draws upon the experience of modern master builders, who continue to use what might be termed “primitive” methods when constructing log cabins.76 In total, a group of eight men and four oxen, with the latter organized into two teams, can maneuver one log of the type under consideration here into position in approximately one hour. As the body of the wall rose higher, it would be necessary to roll or drag the logs up ramps, the length of which would necessarily increase as the wall grew toward its summit of 10 meters in height. Maneuvering the logs into position requires the tying of rope loops around the logs and attaching the ropes over simple pulleys and then back to the team of oxen, which would have the task of slowly pulling the ropes in order to raise the logs into the air.77 74 Joseph A. DeLaittre, A Story of Early Lumbering in Minnesota (Minneapolis, 1959), 18, notes the great effort that was required to build logging roads because it was necessary to avoid slopes with the concomitant need to level the ground over which the logs traveled. 75 Bönisch, “‘Ne Menge Holz’,” 194–5. 76 We thank Robert W. Chambers, a consultant and leading expert on the construction of log cabins, for his insights regarding the construction of these buildings utilizing human and animal labor, and tools that are comparable with those available in the tenth century. The following reconstruction is based on information Mr. Chambers provided to us in a private communication. 77 The use of pulleys of varying degrees of complexity certainly was well understood in the West, even by nonspecialists. See the discussion in Richer von Saint-Remi, Historiae, ed. Hartmut

280

T H E C O S T S O F F O RT R E S S C O N S T R U C T I O N

As each of the 12.5-meter logs would weigh approximately 800 kilograms, the task of maneuvering the logs into position was both taxing and dangerous for the men stationed atop the wall. Their task would have been made somewhat easier, however, if the earthen fill was layered and tamped down within each of the many thousands of “cribs” or “cells” that comprised the subwall as the entire structure was being built, rather than waiting until the wall was complete to haul earth up 10 meters in height and try to fill in the gaps from the top of the wall. In this scenario, the teams of men who were constructing the wooden framework would have been able to operate on a relatively level and stable surface. We have not attempted to calculate how many of the 11,000 oak trees were cut into multiple sections to provide shorter lengths for the construction of the cribs. As it required eight men and four oxen an hour, on average, to maneuver one of these logs into position, the construction of the framework required a minimum of 88,000 man-hours, as well as 44,000 oxen-hours to complete this part of the fortress. This does not count the time necessary to bring the logs into position before being hauled up for placement on the wall structure, as this aspect of the construction process cannot be calculated. If thirty teams of men and oxen were engaged in this task, working ten-hour days with no time to rest on Sundays, they could have completed their work, barring accidents, bad weather, or other negative circumstances, in just over five weeks. But because neither men nor oxen can work in this manner for that length of time, it likely would have required thirty teams closer to fifty days rather than thirty-six to carry out this project.

Basket carriers For both logistical reasons and to increase safety, it is very likely that the thousands of “cribs” that comprised the structure of the subwall were filled in with earth and tamped down as the wooden framework was being constructed. The total volume of the earth and timber structure of this wall was 90,000 cubic meters, of which 15,000 cubic meters comprised wood. This left 75,000 cubic meters of earth to be filled into the structure. If the earth for the Hauptburg wall was drawn from open areas within the Hildagsburg fortress, each load of earth would have to travel approximately 100 meters on average to the dump site. Here, it would then be poured into one of the thousands of “cribs” that comprised the earth and timber subwall. The earth would then have to be tamped down so as to compact the fill and thereby strengthen the integrity of the wall as a whole. In light of the relatively short distances involved, it is likely that men equipped with baskets carried the earth from the dig sties to the wall, rather than attempting to fill carts or wagons with earth. Studies on average human carrying capacity indicate that a man could carry a

Hoffmann, MGH Scriptores 28 (Hanover, 2000), book 3.105. Also see the translation of this text in Histories: Richer of Saint-Rémi, ed. and trans. Justin Lake (Cambridge, MA, 2011).

281

M AT E R I A L S O U R C E S A N D I M A G E S

maximum of 40 kilograms over substantial distances, particularly when the labor involved both the lifting and subsequent dumping of this weight.78 We have calculated that in an absolute best-case scenario, a man could not make more than forty trips in the course of a day, carrying a 40-kilogram basket of earth, walking 100 meters to the dump site and then carrying this load up a ramp, dumping the earth, and tamping it down, then walking back to pick up another load of earth. In the course of a ten-hour workday, forty trips would entail just fifteen minutes per trip, with no breaks at all for meals, broken equipment, falls, or more serious accidents. As noted earlier, 1 cubic meter of earth weighs approximately 1 metric ton. Thus, a basket carrier making forty trips in the course of a workday, carrying 40 kilograms on each trip, could move approximately 1,600 kilograms, i.e. 1.6 metric tons. Thus, filling in the 75,000 metric tons of earth in the Hauptburg wall would require 470,000 man-hours or 47,000 man-days of labor. In addition, as seen earlier in the context of digging the moats at Hildagsburg, it required approximately one full man-day to excavate 3 cubic meters of earth. Consequently, an additional 25,000 man-days of labor were required to fill the baskets of the carriers as they made their way to the Hauptburg wall. In addition to this 72,000 or so man-days of labor that were required to fill in the Hauptburg wall with earth, it is necessary to add the labor required to construct the earthen wall of the Vorburg. This wall, as seen earlier, measures some 750 meters in length, 10 meters in width, and a minimum of 2 meters in height, for a total of at least 15,000 cubic meters of earth. Utilizing the model set out for the Hauptburg wall, the movement and piling up of the Vorburg wall required an additional 14,000 man-days of labor, bringing the total for both major earth-moving components of the Hildagsburg wall systems to 86,000 man-days of labor.

The stone walls of the Hauptburg As mentioned previously, for a distance of some 450 meters, the workers at Hildagsburg constructed a mortared stone wall approximately 2.5 meters in height atop most of the front side of the earth and timber subwall of the Hauptburg. This 2.5-meter height includes a 70-centimeter foundation embedded in the earth and timber wall on which it stood.79 A similar wall, this one only some 150 meters in length, was constructed on the rear side of the earth and timber subwall, and thus opposite part of the front stone wall. These stone walls, each of which had a width 78 Regarding the importance of baskets on Roman construction sites, see DeLaine, The Baths of Caracalla, 93–4. Calculations regarding the capacity of men to carry basket loads in the 40-kilogram range are based upon experiments undertaken with regard to the equipment of Roman soldiers by Marcus Junkelmann, Die Legionen des Augustus: Der römische Soldat in archäologische Experiment (Mainz, 1986), 200–2; and accepted by Roth, The Logistics of the Roman Army at War, 75–7. See also Maurice Daumas, A History of Technology and Invention: Progress Through the Ages, trans. Eileen B. Hennessy, 3 vols. (New York, 1970), I: 433, figures 3 and 4 for shoulder baskets to carry stones, and figure 1.2 for hod carrying to bring mortar to the top of a wall. 79 Dunker, Die Hildagsburg, 200.

282

T H E C O S T S O F F O RT R E S S C O N S T R U C T I O N

of 2 meters, were constructed with a building technique modern scholars call ashlar. This technique employs a system of two parallel walls strengthened with fill.80 The two parallel walls here should not be confused with the two stone walls proper, each of which was built using the ashlar method. The 600 meters in aggregate of the stone walls of the Hauptburg were constructed of greywacke (German Grauwacke), which is very hard sandstone found extensively in the Harz mountains in the southern reaches of the Saxon region within 100 kilometers of Hildagsburg.81 However, it was not necessary for the builders at Hildagsburg to have the many tons of greywacke that were used in the construction of the Hauptburg wall transported from such a great distance. This type of stone was available much closer to the construction site. For example, there was a quarry at Alt Olvenstedt in the environs of Magdeburg, which provided greywacke during the Middle Ages both for the city walls and for the construction of the cathedral there.82 The maximum distance for bringing the greywacke from the quarry in the environs of Magdeburg by boat or barge to Hildagsburg, therefore, was likely approximately 16 kilometers along the Elbe and Renze rivers. For the entire 450 meters length of the front stone wall, the men who constructed Hildagsburg built two parallel stone frames. These frames were each approximately 50 centimeters wide. The same technique, with the same dimensions, was utilized in the construction of the 150-meter rearward stone wall. These stone frames were built of finished or moderately faced stone spaced approximately 1 meter apart. After the stone framework, walls were completed to a height of 2 meters, and the space between the walls was filled with stone debris, earth, and clay.83 The ashlar framing walls for the total length of 600 meters of the front and rear stone walls, which were constructed atop the earth and timber subwall, had a total volume of approximately 1,600 cubic meters of greywacke. The total volume of rock, earth, and clay used as fill for the ashlar walls also comes to approximately 1,600 cubic meters.

Quarrying greywacke for the Hauptburg Estimating the average size of the stone blocks from which any wall has been constructed is not an easy task and requires numerous sample areas to be assessed.84 However, because greywacke is a particular type of sandstone, there seem to have been particular ways in which it could be won from a quarry and then shaped and/ 80 Ibid., 201 and 219. 81 There are, for example, important greywacke quarries at Ballenstedt and Rieder in and near Quedlinburg. 82 For the use of greywacke in the construction of Magdeburg Cathedral, see Rainer Kuhn, “Die ottonische Kirch amMagdeburg Domplatz. Baubefunde und stratigraphische Verhältnisse der Grabungsergebnisse 2001–2003,” in Aufgedeckt: Forschungsgrabungen am Magdeburger Dom 2006–2009, ed. Harald Meller (Halle, 2009), 9–50. 83 Dunker, Die Hildagsburg, 201 and 219. 84 Regarding this method, see Bachrach, “Costs of Castle-Building,” passim.

283

M AT E R I A L S O U R C E S A N D I M A G E S

or faced. Because the samples of greywacke from Hildagsburg were not averaged by the excavation team, it is fortunate that the Ottonian fortifications at nearby Leipzig, also a part of the German defensive frontier against the Magyars and Slavs, provides us with useful and reliable information in this regard.85 As a result of the excavations at Leipzig by Herbert Küas, it seems that on average, the greywacke stone used for the walls was cut into blocks about 30 centimeters in length, 10 centimeters in height, and 10 centimeters in width.86 Thus, each stone is to be understood to have the approximate dimensions of many types of modern bricks but about a third or so the size of modern cinder blocks. These similarities are of importance in estimating the amount of labor required to process the stones that were utilized at Hildagsburg.

Stones quarried Using these figures for the size of the blocks that were employed in the construction of the ashlar stone frames, it may be observed that the 50-centimeter width of each of the ashlar framing walls required five stones placed side by side, and the 2.5-meter height of the framing walls required twenty-five stacked stones. In its entirety, the 1,200-meter length of all of the ashlar frames required 500,000 stones. The stone cap for these two sets of ashlar walls along the entire 600-meter length of the front and rearward stone walls had a surface area of 900 square meters and a total of 90 cubic meters, if we assume a depth of just 10 centimeters. This amounts to an additional 30,000 blocks of stone. In order to construct the ashlar frame and cap for the stone walls of the Hauptburg of more than 500,000 blocks of greywacke, each individual stone measuring on average 30 × 10 × 10 centimeters had to be won from the quarry in the environs of Magdeburg. To provide an estimate of the work required to win this stone, it is necessary to use some comparative data from the later Middle Ages, for which better records survive. It also must be emphasized that masons had available higher-quality steel tools in the later Middle Ages than was the case during the tenth century.87 In the later period, a mason working with an assistant, who kept the tools sharpened, could win between six and seven blocks of stone of the size used at Hildagsburg in the course of each ten-hour working day. Using the “best case” figures, the quarrying of the stone for the Hauptburg would have required in the neighborhood of 150,000 work-days. Thus, 100 masons working with 100 assistants could win the stone under consideration here in approximately

85 Herbert Küas, Das alte Leipzig in archäologischer Sicht (Berlin, 1976), 43. 86 Ibid., 43, 56–9. 87 For example, during the later Middle Ages, when high-quality steel tools, greatly superior to those used by the Ottonians, were available in large numbers, it took as much time for smiths and masons to repair and sharpen their tools as it did for the latter to win the stone, i.e. each hour devoted to cutting stone required one hour of tool preparation. See Bachrach, “The Cost of Castle-Building,” 50.

284

T H E C O S T S O F F O RT R E S S C O N S T R U C T I O N

750 days.88 Of course, if the number of masons and assistants were increased, the number of days required to win the stone would decrease. In addition to these figures for masons and assistants, it is necessary to take into account the vast quantity of unskilled labor necessary for gathering up the debris that was to be used to provide the rubble fill that comprised much of the fill between the ashlar walls. However, it is not possible to calculate this labor with any precision.

Hauling costs for stone The volume of stone blocks cut for the ashlar walls come to approximately 1,600 cubic meters with a weight of some 1,600 metric tons. If the fill in the ashlar wall comprised only an additional 1000 cubic meters of stone rubble (and therefore some 600 cubic meters of earth and clay), a total of approximately 2,600 cubic meters of cut stone and rubble had to be transported from the quarry to Hildagsburg. Easy access to the Elbe river would have made the entire process of transportation much less complicated, as the transport of stone by water was many orders of magnitude cheaper and easier than land transport.89 In addition, the Renze river, a length of which ultimately contributed to the defenses at Hildagsburg, was a tributary of the Elbe and was navigable all the way to the site at modern Elbeu.

More basket carriers It was very likely that unskilled workers using baskets were delegated to carry both the quarried stone and the rubble that was to be used at the Hauptburg to the banks of the Elbe river so that it could be loaded onto barges that traveled downstream to the work site.90 As discussed earlier with regard to the movement of earth, stone carriers each could carry a load on average of about 40 kilograms from the quarry to the river along what undoubtedly was a well-trod path of approximately 2 kilometers in each direction.91 In order to move some 2,600 metric tons 88 Ibid. 89 In the premodern period, land transportation always was substantially more expensive than transportation over the same distance by water. However, the hauling power of horses and mules was increased during the Middle Ages and thus costs for overland transportation decreased over time. See the observations by A. Burford, “Heavy Transport in Classical Antiquity,” The Economic History Review 2nd series 13 (1960), 1–18, which are supported by Lynn T. White, Medieval Technology and Social Change (Oxford, 1962), 156; and Paul Gans, “The Medieval Horse Harness: Revolution or Evolution? A Case Study in Medieval Technological Change,” in Villard’s Legacy: Studies in Medieval Technology, Science and Art in Memory of Jean Gimpel, ed. Marie-Thérèse Zenner (Aldershot, 2004), 175–88, here 178 for the late Roman and early medieval periods. For the late Middle Ages, see James Massachaele, “Transport Costs in Medieval England,” The Economic History Review new series 46 (1993), 266–79, here 273. 90 The Elbe and its tributaries were navigable all the way to the site at Hildagsburg up through the fifteenth century. See Dunker, Die Hildagsburg, 193. 91 Junkelmann, Die Legionen des Augustus, 200–2.

285

M AT E R I A L S O U R C E S A N D I M A G E S

in 40-kilogram allotments, a total of 67,500 such treks were necessary. If 100 carriers were employed and each carrier made five trips per day from the quarry to the barges, then the task could be accomplished in 135 average work-days, for a total of 13,500 man-days of labor. If the number of workers were doubled, the work could be done in just 67 days. When evaluating the labor input for these basket men, we are including the process of loading the basket at the quarry and unloading it at dockside. Under these conditions, each of the workers would carry loads for 10 kilometers each day and walk free of such burdens for 10 kilometers each day.

Barges The types of river boats and barges that were available to the Ottonians for the hauling of heavy loads averaged in capacity approximately 10 tons. This average is based upon the ranges in carrying capacity from 6.5 tons for the Utrecht-type river boat, which had been in service since at least the early ninth century, to the largest ninth- and tenth-century barges, which had a carrying capacity of some 15 tons.92 Given the need to transport a total of some 2,600 metric tons of stone a distance of approximately 15 kilometers down to Hildagsburg, it would be possible to accomplish this transportation task with about 270 barge loads of greywacke and rubble.

Even more basket carriers Once the greywacke reached the landing place at Hildagsburg, which had been designated for the stone to be unloaded, several labor-intensive tasks had to be undertaken. First, workers had to unload the cargo of some 10 metric tons, or 10,000 kilograms from each barge. Then both the stone blocks and the rubble had 92 The Utrecht ship, uncovered in 1930, which gives its name to the basic river boat type, has received extensive scholarly attention. In this context, see Johannes P. W. Philipsen, “The Utrecht Ship,” Mariner’s Mirror 51 (1965), 35–46; Aleydis Van de Moortel, “A New Look at the Utrecht Ship,” in Boats, Ships and Shipyards. Ninth International Symposium on Boat and Ship Archaeology, ed. C. Betrame (Venice, 2003), 183–9; idem, “The Utrecht Type: Adaptation of an Inland Boatbuilding Tradition to Urbanization and Growing Maritime Contact in Medieval Northern Europe,” in Between the Seas: Transfer and Exchange in Nautical Technology, ed. Ronald Bockius (Mainz, 2009), 321–7; and idem, “The Utrecht Ship Type: An Expanded Logboat Tradition in Its Historical Context,” in ibid., 329–36. Regarding barges, see the discussion by D. Ellmers, “Post-Roman Waterfront Installations on the Rhine,” in Waterfront Archaeology in Britain and Northern Europe: A Review of Current Research in Waterfront Archaeology in Six European Countries Based on Papers Presented to the First International Conference on Waterfront Archaeology in Northern European Towns, ed. Gustav Milne and Brian Hobley (London, 1981), 88–95, here 88–91. For barges with heavier carrying capacities of up to 15 tons, see Aleydis Van de Moortel, “Medieval Boats and Ships of Germany, the Low Countries and Northeast France: Archaeological Evidence for Shipbuilding Traditions, Timber Resources, Trade and Communications,” in Settlement and Coastal Research in the Lower North Sea Region (Köthen, 2011), 67–105.

286

T H E C O S T S O F F O RT R E S S C O N S T R U C T I O N

to be carried to the work sites where the walls were being constructed. Using the same calculations as discussed earlier for the loading of the barges, 67,000 basket loads were required. However, because the distances involved were much shorter, probably forty to fifty trips could be made each day rather than just five, meaning that the work could be accomplished in about 10% of the time, equivalent to about 1,350 man-days of labor.

Building the stone walls In order to start the process of building the stone walls, a ditch 70 centimeters deep, 2 meters in width, and 600 meters in length (including both the forward and rearward mortared stone walls) had to be dug into the top of the already constructed main earth and timber subwall of the Hauptburg in order to create the foundation. The earth excavated from these ditches, a bit less than 1,000 cubic meters in total, likely was thrown down along the front side of the earth and timber part of the wall to help build the berm, discussed earlier. This ditch atop the wall likely could have been excavated by 300 men in about a day. The laying of the ashlar frame of more than 500,000 large brick-sized blocks of greywacke, measuring 30 × 10 × 10 centimeters was a highly labor-intensive task. It is generally estimated that a modern mason or bricklayer can process – the technical term for laying brick or stone – approximately forty-five units per hour.93 If we assume that the Ottonian masons worked a ten-hour day without rest periods enjoyed by modern specialists and that the work site was supplied with mortar rapidly, as is the case on modern construction jobs, then each day one highly skilled tenth-century mason could process 450 pieces of building stone. The processing of 500,000 pieces of already shaped greywacke, therefore, would require in excess of 1,000 average work-days. In a best-case evaluation, 100 skilled masons supported by a sufficient number of assistants could process the greywacke for the stone wall at Hildagsburg in just over ten days.

Making mortar Here an attempt will be made to provide some estimates regarding a few aspects of the cost of making the mortar used to fix in place the greywacke stone that had been won from the Magdeburg quarry and shaped into blocks for use at the top part of the Hauptburg wall.94 Although these calculations make no pretense of being exhaustive, it was necessary for the Ottonian workers to produce sufficient

93 The authors would like to thank Richard Paetow, who as assistant director of Minnesota Masonry Society, provided this information to Bernard Bachrach in 1984 for this study, “The Cost of Castle-Building.” 94 Regarding recent observations concerning the making of mortar, see Adam, La construction romaine, 69–79; Bachrach, “The Cost of Castle-Building,” 48; DeLaine, The Baths of Caracalla, 88–9; and Pearson, The Construction of the Saxon Shore Forts, 87.

287

M AT E R I A L S O U R C E S A N D I M A G E S

mortar to hold together more than 500,000 stone blocks. Examination of medieval stone walls indicates that the mortared joints between blocks on average were about 2 centimeters in width.95 Therefore, each stone block of greywacke used at Hildagsburg, with an average dimension of 30 × 10 × 10 centimeters, required that approximately a centimeter of mortar be spread over its six sides, a surface area of 1,400 square centimeters, or 0.14 square meters. With a thickness of 1 centimeter, a total of 1,400 cubic centimeters or 0.0014 cubic meters of mortar were required for holding each stone in place. For 500,000 blocks, therefore, 700 cubic meters of mortar were required. The initial expenditure undertaken by the builders for making the mortar was a capital investment, i.e. the construction of a kiln in which to “cook” the limestone in order to produce lime.96 However, kilns throughout the West in premodern times and, indeed, even into more modern times, not only varied greatly in size but also in design and efficiency. They varied, in addition, in terms of the cost to construct them, both in man-hours and materials and in regard to their potential output.97 It must be emphasized that the construction of an efficient kiln in which to cook lime was not a haphazard venture. Rather, it required skilled workers who understood not only how to build the kiln but also the design principles under new conditions each time an apparatus was to be built.98 The manufacture of mortar under the technologically primitive conditions of the premodern and early modern world, however, largely was a constant. It was understood that mortar was a mixture composed of approximately one-quarter lime and three-quarters clean sand by volume.99 This ratio means that for making the mortar that was to be used to set the greywacke stone blocks at Hildagsburg, 525 cubic meters of clean sand were required. In addition, 175 cubic meters of lime were required. From a scientific perspective, it is understood historically that 2 cubic meters of limestone were required to produce each cubic meter of lime.100

95 Bachrach, “The Cost of Castle-Building,” 48. 96 An excellent introduction to the wide variety of types of kilns used from the Roman Empire well into the modern period when “primitive” technology is still being used is provided by Adam, La construction romaine, 69–76. 97 Ibid. The size of the kiln ultimately reflects its maximum output, but also the work effort necessary to load it and feed it with fuel as well as the “cooking” time. Smaller kilns provide smaller amounts of lime than bigger kilns but do so at a proportionally more rapid pace. 98 Adam, La construction romaine, 69–76, reviews some of the knowledge required for those who construct kilns under primitive conditions and without a theoretical knowledge of the physical laws at issue. 99 These figures are provided by Vitruvius, and they are discussed in detail by Adam, La construction romaine, 76–9. In pre-Crusade Europe and undoubtedly throughout the Middle Ages, the same system was used, as noted in the Mappae Clavicula, chapters 102, 103, and 254. For the text of this work, see Cyril Stanley Smith and John G. Hawthorne, “Mappae clavicula: A Little Key to the World of Medieval Techniques,” Transactions of the American Philosophical Society 64 (1974), 3–128. 100 A. B. Searle, Limestone and its Products (London, 1935), 354–5; and Bachrach, “Costs of CastleBuilding,” 48.

288

T H E C O S T S O F F O RT R E S S C O N S T R U C T I O N

Therefore, approximately 350 cubic meters of limestone were required to produce the 175 cubic meters of lime for making mortar that was used at Hildagsburg. Modern calculations suggest that a cubic meter of limestone could be quarried and broken up into pieces that could be easily fed into a kiln in the course of a day’s work by an average mason so long as his tools were kept sharpened.101 Each cubic meter of limestone weighs approximately 500 kilograms.102 In order to transport the approximately 350 cubic meters of limestone to produce 175 meters of lime, 350 carts each with a 500-kilogram capacity load were required. Each cart required two draft animals, e.g. horses, mules, or oxen, and was serviced by a two-man team.103 Large quantities of wood also were required to cook the lime in the kilns, as the heat necessary to carry out this process must reach 1,000 degrees centigrade, which is about 200 degrees centigrade greater than needed for the manufacture of pottery.104 Thus, for each kilogram of limestone that was put into the kiln to be made into lime, 2 kilograms of wood were required.105 For the Hildagsburg construction project, the 350,000 kilograms of limestone required 700,000 kilograms of wood. Oak weighs approximately 700 kilograms per cubic meter, which means that approximately 1,000 cubic meters were required for burning of the limestone. The assumption here is that the tops of the oak trees and their branches, which were not used for the wall framework and palisade as well as for the ramps, were used to provide the wood to heat the kilns. In this context, 1,000 cubic meters of oak that were burned to heat the kilns represent approximately 6% by volume of

101 Adam, La construction romaine, 23–32; and Bachrach, “The Costs of Castle-Building,” 50. 102 Regarding different types of limestone, see DeLaine, The Baths of Caracalla, 88. For limestone weights, see C. Ramsey and H. Sleeper, Architectural Graphic Standards, 4th edition (New York, 1951), 546. 103 The basic work on this topic remains Richard Lefebvre des Noëttes, L’attelage et le cheval de selle à travers les âges, 2 vols. (Paris, 1931), which has a strong experimental underpinning. For the state of the question, see White, Medieval Technology, 59–61; Albert Leighton, Transport and Communication in Early Medieval Europe (New York, 1972), 77 and 161; Bernard S. Bachrach, “Animals and Warfare in Early Medieval Europe,” Settimane die Studio del Centro Italiano di Studi sull’alto Medioevo 31 (Spoleto, 1985), 707–64 and reprinted with the same pagination in Armies and Politics in the Early Medieval West (London, 1993), 716–17; John Langdon, Horses, Oxen and Technological Innovation: The use of Draught Animals in English Farming from 1066– 1500 (Cambridge, 1986), 1–10; and for a recent review of the state of the question, Gans, “The Medieval Horse Harness,” 175–87. 104 Adam, La construction romaine, 65 and 63, respectively. 105 Bachrach, “The Cost of Castle-Building,” 48 regarding the biothermal units (BTUs) required to cook the lime. For various types of wood, their BTU values at various levels of moisture content, and burning efficiency, see L. T. Hendricks, “Economics of Wood Burning,” in Heating the Home with Wood, University of Minnesota Extension Bulletin 436 revised edition (Minneapolis, 1980), 14–15. With regard to wood cutting in different contexts, see Adam, La construction romaine, 91–104 and 91–2 for the density of wood. Also see Janet DeLaine, “Bricks and Mortar: Exploring the Economics of Building Techniques at Rome and Ostia,” in Economics Beyond Agriculture in the Classical World, ed. David J. Mattingly and John Salmon (London, 2001), 230–68, here 260–1.

289

M AT E R I A L S O U R C E S A N D I M A G E S

the wood that was utilized in the construction of the Hauptburg wall. The use of this “scrap wood” is, of course, a best-case solution for the acquisition of the fuel that was required for use in the kilns. The transportation of the 700 metric tons of wood from the forest, where it was procured, to the site of the kiln where it was to be burned required at least 1,400 cart loads each drawn by two draft animals and serviced by two men. Two additional facts need to be noted here. First, as mentioned previously, we do not know where the logs were cut that were used for the fortification, of which the treetops and branches were a by-product. Therefore, we do not know the distances that the fuel for the kilns had to travel from the forest to Hildagsburg. Second, it is possible, and perhaps even likely, as noted earlier, that after the trees for the fortification were felled and prepared for transport by having their tops and branches lopped off, they could be sent down the Renze river to Hildagsburg. However, the treetops and branches could not be effectively transported by water and thus had to be either carried in barges or hauled in carts, or having been cut into small pieces, carried in baskets. With an average load of 40 kilograms per basket, it would have required 17,500 basket loads to move 700 metric tons. In the next stage of the preparation for the processing of the stone at Hildagsburg, the 175 metric tons of lime had to be arranged into approximately 500-kilogram segments to be hauled from the kiln(s) by cart, where the product was manufactured, to the work site where it had to be offloaded. In this process, from the time of manufacture to the time of use, it was absolutely essential to keep the lime dry and, thus, it likely was kept in waterproof barrels and not simply left in piles like sand so that it could be loaded onto carts with shovels.106 It is highly unlikely that lime was transported in baskets, which could not assure that the lime remained dry. It is noteworthy that the Carolingian government, from which the Ottonians inherited a wide range of institutions, worked with standard size barrels, each of which had a capacity of about 35 kilograms, i.e. 110 Roman pounds. Thus, a standardized cart, a basterna, which was also waterproof, could carry about fifteen such barrels.107 These barrels, of course, had to be manufactured and transported to the site of the kilns in order to be loaded with the newly produced lime.108 It is also likely that once filled, these barrels were kept under cover in sheds or huts that also had to be built at the work site for temporary use.

106 Esmonde Cleary and Wood, Saint-Bertrand-de-Comminges, 146 comment, with typical British understatement, on the need for dry weather when using lime. The Carolingians commanded the technology to keep containers watertight. See Bernard S. Bachrach, “Carolingian Military Operations: An Introduction to Technological Perspectives,” in The Art, Science, and Technology of Medieval Travel, ed. Robert Bork and Andrea Kann (Aldershot, 2008), 17–29. 107 Bachrach, “Carolingian Military Operations,” 28–9, indicates that standardized Carolingian barrels carried 110 Roman pounds, which comes to approximately 35 kilograms. 108 While baskets could be used for carrying lime in quantities of up to about 40 kilograms, these would not be appropriate as storage containers for lime.

290

T H E C O S T S O F F O RT R E S S C O N S T R U C T I O N

At the work site, the lime would be mixed with clean sand and water by the mason’s assistants, as the mortar was needed for the skilled men who were processing the stone. In order to make the mortar, some 525 cubic meters of clean sand were required at approximately 500 kilograms per cubic meter. Consequently, a total of some 260 metric tons of sand had to be dug up, likely from a riverbank where it had been washed clean by running water, in order to make the mortar.109 One man could dig up approximately 3 cubic meters of sand during a ten-hour work day and so all of the sand required could be made available for the mixing of the mortar in the course of 175 man-days of labor. To carry this sand to the work site at Hildagsburg, where it was to be mixed with lime, required more than 6,500 basket loads. In addition to bringing the lime and sand to the stone wall being constructed at Hildagsburg, 17,500 decaliters of clean fresh water had to be obtained from the Renze river for the purpose of mixing the lime and the sand to make mortar.110 Lacking hoses and pumps by which to move the water, 4,800 barrels, of the type discussed earlier, had to be filled and moved approximately 100 meters, on average, from the river to the wall and thereafter hauled up 10 meters to the building site of the stone walls. Using carts with a normal 500-kilogram capacity, each of which could transport fifteen barrels of water, 320 cart loads would be required.

The costs of fortress construction In estimating the labor costs involved in the construction of the fortress complex at Hildagsburg, we have focused on several of the most labor-intensive elements of the building process, i.e. digging moats and constructing the walls of the Vorburg and Hauptburg. In doing so, we have noted in passing but have not calculated the labor involved in several other elements of the project that also required significant technical skill and manpower resources, including the construction of two substantial wooden towers; the building of two wooden ramps, each measuring 54 meters; the construction of a wooden palisade along the eastern wall of the Hauptburg; and the emplacement of wooden pilings to direct the flow of the Renze into the tripartite moat system. To these large-scale activities must be added numerous other, relatively less labor-intensive projects, including the construction of scaffolding, the building of ramps and pulley systems to bring logs into position on the under-wall and to haul water up to the top of the wall to make mortar, the production of thousands of baskets and barrels, the making of huge quantities of rope, the sharpening and repair of tools, the care of animals, the construction of housing for laborers, the 109 Adam, La construction romaine, 78, argues on the basis of Vitruvius that river sand was to be preferred. Sand that is constantly washed by salt water has a negative effect on the mortar. 110 Regarding the quantities of water in proportion to lime and sand, see Adam, La construction romaine, 78. These proportions remain constant during the early Middle Ages, as shown in Mappae clavicula, chapters 102–3, and 254.

291

M AT E R I A L S O U R C E S A N D I M A G E S

building of bake ovens for bread and other cooked victuals, the making of kilns for producing lime, the mending of worn clothing, and a host of other details that are involved in supplying the needs of large numbers of men over a lengthy period. As a consequence, the estimates for the labor involved in the construction of Hildagsburg must be understood not only as a best-case scenario in terms of the minimum labor involved but also as excluding a significant quantity of the man-days that necessarily were part of the project. With these figures in mind, it is clear that the construction of the main elements of the fortress at Hildagsburg required a bare minimum of 2.7 million man-hours or 270,000 man-days of labor, assuming a standard ten-hour day that excluded times for meals and breaks. Under these conditions, 2,700 men could have constructed the moats and walls at Hildagsburg, if all of the other ancillary work also were completed, in approximately 100 days, or more likely four months when due consideration is given to the need for days of rest, particularly on Sundays, as well as interruptions of the work due to the weather. To put this figure into perspective, 1,000 men require approximately 1 metric ton of carbohydrates and protein each day to maintain energy levels that are necessary to carry out hard physical labor.111 Over the course of four months, 2,700 laborers would require approximately 320 metric tons of food supplies. This does not include the grain and other fodder necessary to maintain oxen, horses, and other draught animals that were used in the construction project.

Some statistics While statistics of any kind are very hard to develop for early medieval history, it is possible to provide some context for the material investment made by the royal government to feed the workers at Hildagsburg. In 874, King Louis the German (840–876) issued a charter in which he dealt with the surplus production at the royal villa of Hornau, located in the region of Frankfurt. The charter makers clear that the eight mansi attached to this villa produced an annual surplus of 130 modii of grain, twenty pigs, and four cartloads of wine.112 The surplus revenues derived from Hornau can be compared with the surplus produced at the erstwhile Ottonian royal estate of Bruoch, located in Lotharingia, the dues from which are recorded in a charter issued by the monastery of Gorze in 984.113 111 For the caloric intake see the discussion by Roth, The Logistics of the Roman Army at War, 7–55; and Bernard S. Bachrach, “Crusader Logistics from Victory at Nicaea to Resupply at Dorylaion,” in Logistics of Warfare in the Age of Crusades, ed. John H. Pryor (Aldershot, 2006), 43–62. 112 Die Urkunden Ludwig des Deutschen, Karlomanns und Ludwigs des Jüngeren, ed. Paul Kehr (Berlin, 1932–1934), here Louis the German nr. 155. 113 Mettensia II: Cartulaire de L’Abbaye de Gorze, MS 826 de la Bibliothèque de Metz, ed. A. D’Herbomez (Paris, 1898), nr. 116, 211–13. This document was discussed briefly by Liudolf Kuchenbuch, “Abschied von der ‘Grundherrschaft’: Ein Prüfgang durch das ostfränkish-deutsche Reich 950–1050,” Zeitschrit der Savigny-Stiftung für Rechtsgeschichte, germanistiche Abteilung

292

T H E C O S T S O F F O RT R E S S C O N S T R U C T I O N

According to the information provided in the charter issued by Gorze, the estate as a whole consisted of 21¾ mansi, whose holders owed a combination of labor and renders to the monastery. Each mansus was required to pay 8 denarii each year at Christmas, and each inhabitant of the estate owed 6 denarii on the feast of St. Remigius (1 October). If we take an average of five inhabitants per mansus, then the yearly cash dues amounted to 38 denarii per mansus, or a total of 826.5 denarii a year from the entire estate. Each mansus also owed 8 modia of grain each year, and every 1.5 mansi owed either a cart of wine or the sale price of a cart of wine, which is an important indication of the penetration of the use of money into the rural economy. The estate as a whole owed 56 chickens and 350 eggs at Easter to Gorze. In addition to these food renders and cash payments, each mansus had to provide two cartloads of wood and 200 bundles of reeds a year to Gorze.114 While these texts should not be understood as having demonstrated either the entire production of private or royal estates, nor as an example of either Carolingian or Ottonian fiscal administration, they do provide a useful heuristic tool for understanding the scale of the economic resources that were required to feed the workers who constructed the fortification at Hildagsburg. There is no reason to believe that the villa of Hornau or the estate of Bruouch were particularly productive or unproductive. Therefore, the surplus production here may be understood as providing a rough guide to average production on royal estates (as well as private estates) in the lands of the German kings. Second, the provision of information about the surplus production of the eight mansi at Hornau and the 21.75 mansi at Bruoch, that is the quantity of agricultural goods that could be extracted from these lands after the caloric requirements of the agricultural workers were met and the seed corn for the next year was set aside, can be used with caution to illustrate more broadly how much surplus production might be achieved per mansus from lands held by the Ottonian government or from which they could demand renders. It is important to note that many magnates, particularly bishops and abbots, possessed vast quantities of landed resources, that is many thousands of mansi, the production from which the Ottonian kings had no difficulty in accessing for governmental use.115

121 (2004), 1–99, 89–90, who commented on some of the unusual terms used in the text to denote the population, but did not discuss the fact that this was a formerly fiscal property, and that the obligations of the inhabitants, according to the record, were those that dated back to the time when it was a fiscal estate. Kuchenbuch, himself, relied heavily on the discussion by Charles Edmond Perrin, Recherches sur la Seigneurie Rurale en Lorraine, d’après les plus anciens censiers (IxeXIIe siècles) (Paris, 1935), 141–69 and 693–703, who also was not focused on the role played by this estate as part of the royal fisc, but only in examining the putative development of rural lordships in tenth-century Lotharingia. 114 Cartulaire de L’Abbaye de Gorze, 212–13. 115 The use of ecclesiastical resources to support the royal court has received extensive attention. See, for example, John W. Bernhardt, Itinerant Kingship and Royal Monasteries in Early Medieval Germany c. 936–1075 (Cambridge, 1993).

293

M AT E R I A L S O U R C E S A N D I M A G E S

Clearly, agricultural production varies from year to year based on the vagaries of weather and other contingencies. However, because the grant of property to the chapel of St. Mary at Frankfurt was intended for possession over the long term, the surplus production enunciated in the charter issued by Louis the German can be understood as the “best guess” for an average year. Certainly, underestimating the production would undermine the positive impression that was to be gained by donating to the church. Conversely, overestimating the production might incur the hostility of the recipients when the promise of exaggerated yields did not materialize. The charter issued by Gorze deals with property that the monastery had acquired from the royal fisc in the late 960s or early 970s that previously had been held as a beneficium by a count named Raimbald. In response to requests from the tenants of Bruoch, the abbot ratified the dues that they owed at the same level that they had paid when the estate was held by the count. As a result, the payments must be understood as a best-case scenario for the dependents on the estate, that is a state of affairs that they wished to commemorate and preserve in writing. Consequently, for heuristic purposes, taking as our starting point the surplus production per mansus at Hornau we arrive at a surplus of 16 modii of grain, i.e. half a metric ton, as well as 2.5 pigs and half a cart load of wine per mansus. When we consider Bruoch, each mansus was required to pay 8 modia of grain, two-thirds of a cart of wine, or the sale price of two-thirds of cart of wine, as well as 36 denarii in cash dues. Using these two data points, feeding the workers involved in the construction of Hildagsburg would have required the entire surplus production from approximately 650 mansi of the Hornau type, or 490 mansi of the Bruoch type, if we can convert the cash renders into grain.116 This is equivalent of about ten to thirteen royal estates, if we use the conservative estimate of just 50 mansi per estate.117 In this context, it is useful to recall that fortifications were built for specific military purposes and that the construction process, under normal circumstances, was completed as rapidly as

116 Georg Waitz, Die Deutsche Reichsverfassung von der Mitte des neunten bis zur Mitte des zwölften Jahrhunderts, vol. 4 (Kiel, 1878), 224, using Carolingian figures, pointed to a price of two denarii for a modius of grain. The revenues from the fiscal estate at Bruoch show a cash render of 38 denarii a year per mansus, which would be sufficient, using Waitz’s model to purchase 19 modii of grain, which is slightly more than half a metric ton. To this can be added the 8 modii of grain that also was owed, on average, by each mansus at Bruoch. 117 Many royal villae had far more than 50 mansi. For example, the Carolingian royal estate at Friemersheim, which was granted to the monastery of Werden, possessed 129.5 mansi. See Die Urbare der Abtei Werden A. D. Ruhr: Die Urbare vom 9.-13. Jahrhundert, ed. Rudolf Kötzschke (Bonn, 1906), 15–18. The Ottonian royal estate of Kitzingen, which Henry II granted to Bamberg, possessed 260 mansi. See the discussion of this estate and the grant by Henry II in Wilhelm Störmer, “Heinrichs II. Schenkungen an Bamberg: Zur Topographie und Typologie des Königsund bayerischen Herzogsguts um de Jahrtausendwende in Franken und Bayern,” in Deutsche Königspfalzen. Beiträge zu ihrer historischen udn archäologischen Erforschung, vol. 4, ed. Lutz Fenske (Göttingen, 1996), 377–408, particularly 392–3; and Dieter Hägermann, “Wandel der klösterlichen Grundherrschaft im 11. Jahrhundert? Beobachtungen an dem Urbar des Benediktinerklosters Kitzingen in Unterfranken,” Grundherrscaft und bäuerliche Gesellschaft im Hochmittelalter (1995), 162–83.

294

T H E C O S T S O F F O RT R E S S C O N S T R U C T I O N

possible. Uncompleted fortifications could not serve their intended military function for either territorial defense or as a base of operations for offensive campaigns. This is in marked contrast, for example, to the construction of churches, and particularly stone cathedrals, which could take years or even decades to complete. Even when considered in isolation, the investment of resources on this scale to supply the laborers constructing Hildagsburg is quite remarkable. However, the Ottonian kings did not construct fortifications in isolation, but rather as elements in well-defined defensive systems. For example, in the period c. 929–936, Henry I constructed approximately three dozen fortifications in the valleys of the White Elster, Mulde, and Elbe rivers, for an average of about five each year.118 The construction of these fortifications likely was concentrated toward the first half of this period, immediately following Henry’s conquest of the region, with as many as ten, or perhaps more, fortresses being built in a single year. These fortresses were, on average, about half of the size of Hildagsburg, at least in their initial phase of construction. Thus, for example, the fortification at Magdeborn, located on a western tributary of the Mulde, had a circuit of walls measuring some 250 meters.119 Consequently, the labor involved in the construction of these strongholds, on average, likely was about half that required for Hildagsburg, i.e. 135,000 man-days per fortress. For heuristic purposes, if we use the figure of ten Magdeborn-sized strongholds being constructed in 930, we come to a total of 1.35 million man-days of labor being deployed by the royal government. Using the same models as discussed earlier with regard to Hildagsburg, feeding these men for four months of labor would require the use of the entire annual surplus production from approximately 3,250 mansi of the Hornau type, equal to 65 royal villae, or 2,400 mansi of the Bruoch type, equal to about 48 royal villae, which in either case would represent a truly massive undertaking. Similarly large expenditures of labor can be identified during the reign of Otto I in the construction of a series of least sixteen strongholds along the course of the Havel river following the German conquest of this region in 955, as well as in the construction of a dense network of strongholds in eastern Bavaria.120

Conclusion This investigation of the material reality and technology involved in the construction of the stronghold at Hildagsburg provides the possibility of gaining insights 118 With regard to the fortresses in this region, see Bachrach, Warfare in Tenth-Century Germany, 26–32 and map 1. 119 See Harald W. Mechelk, “Vorbericht zur Grabung 1965 in Magdeborn, Kr. Leipzig,” Ausgrabungen und Funde 11 (1966), 96–100, here 97. 120 See the discussion of the fortress system along the Havel by Joachim Hermann and Richard Hoffmann, “Neue Forschungen zum slawischen und frühdeutschen Burgwall ‘Räuberberg bei Phöben, Kr. Potsdam Land,” Ausgrabungen und Funde 4 (1959), 294–306, here 302; and for the creation of a defensive system in eastern Bavaria under Otto I and Otto II, see Hubertus Seibert, “Adlige Herrschaft und königliche Gefolgsschaft. Die Grafen von Schweinfurt im ottonischen Reich,” Zeitschrift für bayerische Landesgeschichte 65.3 (2002), 839–82, here 853–7.

295

M AT E R I A L S O U R C E S A N D I M A G E S

of a social, economic, and political nature into Ottonian Germany. The extrapolation of these data provide a tentative framework and possible model for the discussion of royal expenditures on military building projects during the Ottonian century.121 In addition, however, the great scale of the resources involved in the building of relatively simple fortresses (and the picture becomes even more complex when palaces and cathedrals constructed of mortared stone are taken into account) also impinges upon the ways in which the German royal government should be conceptualized. First, the engineering work at Hildagsburg, including the high-quality hydrology that was involved in the construction of the triple-moat system, provides compelling evidence for the continuity of late antique learning and construction techniques into the post-Carolingian German kingdom and the availability of this expertise to the Ottonian kings. Moreover, the use of this technology on the frontier, rather than simply in the domiciles of the secular and ecclesiastical elites in Rhineland cities of Roman origin, clearly is evidence of the broad diffusion of high-level engineering skills and concomitant systems for the transfer of this knowledge throughout the German kingdom. Second, the very existence of Hildagsburg and hundreds of additional royal strongholds from this period necessarily demonstrates the capability of the royal government to mobilize the human, financial, and material resources that were necessary to carry out these building projects. To make this point somewhat differently, it is often claimed that medieval writers exaggerated the successes of their patrons in order to glorify them beyond their just desserts. By contrast, the simple existence of fortresses such as Hildagsburg, when we come to understand the technology and costs involved in its construction, is a demonstration of the fact that the royal government of Germany not only carried out the major building programs claimed by writers such as Widukind of Corvey but that these writers, if anything, undersold the actions of their patrons. In understanding the administrative capacity to undertake building efforts on the scale illuminated here, it is helpful to understand that the Ottonian royal government had under its direct management a vast congeries of estates and dependents. At a bare minimum, Otto I controlled 900 estates of various sizes and more than a quarter of a million dependents, although this number could easily have exceeded half a million.122 Thus, Otto I, and the situation was very similar for his father Henry I, and his successors could call upon the labor services of a population of direct royal dependents that approximated the entire population of the contemporary kingdom of Wessex.123 In this context, the kings of Wessex were

121 Karl Leyser, “Ottonian Government,” The English Historical Review 96.381 (1981), 721–53 drew attention to a number of aspects of Ottonian administration in Saxony, with a particular focus on the organization of fiscal resources. 122 Bachrach, “Toward and Appraisal,” 10–16. 123 See Richard Abels, Alfred the Great: War, Kingship and Culture in Anglo-Saxon England (Harlow, 1998), 207 for the population estimate of half a million people in Wessex during Alfred’s reign.

296

T H E C O S T S O F F O RT R E S S C O N S T R U C T I O N

able to construct, refit, and refurbish thirty-three fortifications, many of stone and larger than Hildagsburg in the span of just a few decades.124 In addition to the considerable population of royal dependents on whom the Ottonian kings could mobilize for labor duties, Henry I, Otto I, and their successors had the authority and power to mobilize the entire population of the realm to participate in the construction of fortifications. They inherited this authority from their Carolingian predecessors, who had enunciated the obligation of all able-bodied people to participate in the common defense. The most thorough enunciation of this principle, based upon earlier Carolingian legislation by Charlemagne and Louis the Pious, was Charles the Bald’s edict of Pîtres.125 Charles the Bald’s younger contemporary and relative, Arnulf of Carinthia, also possessed this authority. Indeed, this east Carolingian ruler’s charter on behalf of his fidelis Heimo in 888 takes for granted the existence of the general obligation on the population to participate in the construction and defense of royal strongholds in the eastern regions of the Carolingian Empire.126 This obligation in the Carolingian Empire had a parallel in the trimoda necessitas in Anglo-Saxon England. A number of immunities granted by Otto I to ecclesiastical institutions such as Corvey, St. Maurice at Magdeburg, and Sts. Peter and Paul at Weißenburg (French Wissembourg) make clear that the Ottonians retained the authority to mobilize the entire population, including both the free and unfree dependents of other magnates, to provide labor for the construction of fortifications.127 In these cases, rather than issuing orders to the counts of numerous pagi to mobilize the population to construct fortresses, Otto I issued commands to the abbots of the monasteries in question to mobilize not only their own dependents but also the populations living in the nearby pagi, who were not otherwise under the abbot’s jurisdiction, to take part in the construction and repair of royal strongholds. Finally, in considering whether the Ottonian kings utilized the resources of the royal fisc, or required great magnates to expend their own resources, or called upon the labor duties of the entire population, or some combination of all three to build fortifications, the essential factor remains that Henry I, Otto I, and their successors set out policy goals and oversaw their implementation. Whether in the case of a specific fortification Henry or Otto ordered the mobilization of “private” resources rather than those of the fisc is trivial in comparison to the fact that the 124 For a useful introduction to many of the questions regarding the origin, size, and organization of the defensive scheme described in the Burghal Hidage, see the collection of essays in The Defence of Wessex: The Burghal Hidage and Anglo-Saxon Fortifications, ed. David Hill and Alexander Rumble (Manchester, 1996). 125 Capitularia regnum Francorum, 2 vols., ed. A. Boretius and V. Krause (Hanover, 1890–1897), II: 273 c. 27. 126 Die Urkunden Arnolfs, nr. 32. 127 For a discussion of these grants of immunities and the creation of special defensive districts, see David S. Bachrach, “Immunities as Tools of Royal Military Policy under the Carolingian and Ottonian Kings,” Zeitschrift der Savigny-Stiftung für Rechtsgeschichte, germanistische Abteilung 130 (2013), 1–36, here 31–4.

297

M AT E R I A L S O U R C E S A N D I M A G E S

Ottonian kings commanded that this work be done and that it was done. Moreover, it must be observed that for “private” resources to have been utilized, the Ottonian kings required descriptiones of these resources in a manner consistent with the ubiquitous demands of the Carolingian mayors of the palace, kings, and emperors. As observed earlier, the mobilization of these “private” resources was not an ad hoc tool. Rather, it was well established that the ruler had the unique legal authority to mobilize ostensible private resources for the public good, and particularly for the common defense, which included the construction of royal fortifications.

298

APPENDIX The dating of the construction of the Ottonian fortress at Hildagsburg

Hildagsburg is typical of Ottonian fortifications insofar as it is not mentioned in any surviving documents or written sources produced during the course of the tenth century.128 The vast majority of early medieval fortifications throughout the German kingdom, which have been identified by archaeologists, are mentioned for the first time, if at all, in texts of a much later period.129 As a consequence, the dating of the initial period of the construction of Hildagsburg, like all other such early fortifications, must be executed on the basis of four types of evidence traditionally used by archaeologists and historians. In the first instance, archaeologists undertake a comparative analysis of archaeological finds, e.g. pottery, that has been identified as coming from particular strata located within the fortification itself. But this does not include coins, which are really a form of written document.130 Second, archaeologists tend to rest dating heavily on dendrochronology.131 Third, archaeologists use comparative studies with fortifications that already would seem to be well-dated. In this context, not only is the structure as a whole subject to comparison but also the construction techniques and technology. Finally, historians, often with the help of archaeologists, work with geopolitical criteria in order to develop the political, military, and economic position of the fortification under consideration to provide dates that make sense for the government to have invested substantial resources at a particular place and time. 128 The fortification at Hildagsburg first attracted attention in the surviving written sources in the context of its destruction by Margrave Albrecht of Brandenburg (1100–1170), in 1129. See Annalista Saxo, ed. Georg Waitz, MGH SS 6 (Hanover, 1844), 766. 129 This point has been emphasized by Peter Ettel, “Ergebnisse der Ausgrabungen auf der Burg Horsadal, Roßtal bei Nürnberg,” Frühmittelalterliche Burgenbau, 127–36, here 127, who points out that of the more than 250 fortifications identified through archaeological research in northern Bavaria dating from the period c. 700–c.1000, only 30 are named in the surviving written sources. 130 Dunker, Die Hildagsburg, 221, discusses this issue in some detail. 131 Unfortunately, the excavators of Hildagsburg did not have available the detailed archaeological framework to undertake dendrochronological dating of the surviving wooden elements of the fortification. It is certainly a desideratum for any future investigation of the site to include a dendrochronological analysis.

299

M AT E R I A L S O U R C E S A N D I M A G E S

Based upon the variety of material finds within the fortification, as well as the highly advanced technologies utilized in its construction, including mortared stone walls, Dunker concluded that Hildagsburg was built by the Ottonian royal government when it was at the height of its power. He visualized the Ottonian apogee as the later stages of the reign of Otto I, who died in 973, eleven years after he took the title of emperor in 962. Using a geopolitical approach, Dunker argued that Hildagsburg was constructed to protect Otto’s capital at Magdeburg, located just a half-day’s ride to the south.132 In light of these geopolitical insights, Dunker surmised that the elevation of Magdeburg to the status of an archiepiscopal see in 968 served as the stimulus that moved Otto I to build the fortifications at Magdeburg.133 It is not impossible that the initial construction of the stronghold at Hildagsburg was begun late in the reign of Otto I. The archaeological evidence is not conclusive.134 For example, the ceramic remains admit a rather broad possibility for establishing the chronology in the tenth century, and the dendrochronological evidence has yet to be developed.135 However, in our view, also using geopolitical criteria, it seems rather more likely that the fortification was begun almost half a century earlier than the date suggested by Dunker. We argue that this stronghold was constructed in the context of Henry I’s effort to reestablish the middle and lower Elbe as the frontier of the East Carolingian/German kingdom in the face of Slavic opposition. In this temporal context, it is to be noted that Magdeburg already was established as an important fortified trading center and administrative headquarters for the exercise of governmental functions under Charlemagne in the early ninth century.136 In addition, there were numerous Ottonian fortifications established farther to the east of Magdeburg long before it was made an archiepiscopal see. These included the episcopal seats at Brandenburg and Havelberg.137 To put it simply, from the perspective of military strategy, there was no pressing danger to Magdeburg in the late 960s. As a result of numerous conquests farther to the east that had been carried out by Henry I and Otto I during the course of the previous half-century, Magdeburg no longer was situated on the frontier.138 By the late 960s, the efforts of Henry I and Otto I, in fact, had extended the frontier of the German kingdom all the way east to the Oder river, some 250 kilometers

132 133 134 135 136

Dunker, Die Hildagsburg, 221. Ibid. Ibid. Ibid., 210 and 225. See Matthias Hardt, “Hesse, Elbe, Saale and the Frontiers of the Carolingian Empire,” in The Transformation of the Frontiers from Late Antiquity to the Carolingians, ed. Walter Pohl, Ian Wood, and Helmut Reimitz (Leiden, 2001), 219–32. 137 See Klaus Grebe, “Die Ergebnisse der Grabung Brandenburg,” in Berichte über den II. Internationalen Kongreß für slavische Archäologie, vol. 3 (Berlin, 1973), 269–73; and Christian Popp, “Gründung und Frühzeit des Bistums Havelberg,” Mitteilungen des Vereins für Geschichte der Prignitz 3 (2003), 6–82. 138 Bachrach, Warfare in Tenth-Century Germany, 14–69.

300

T H E C O S T S O F F O RT R E S S C O N S T R U C T I O N

beyond Magdeburg. As a result, the German emperor was unlikely to entertain the realistic possibility from a strategic perspective that Magdeburg was threatened by an invasion of Slavic forces. In short, Otto I had no reason to fortify the site at Hildagsburg simply because Magdeburg had been elevated to archiepiscopal status. It is clear, however, that Otto did wish to aggrandize the status of Magdeburg and that he did so by supporting its development with various gifts and his personal attention through visits.139 The construction of an obviously redundant stronghold 16 kilometers to the north of Magdeburg served to increase the security of neither the new archbishop nor his episcopal city. By contrast, there is considerable reason from both a geopolitical and archaeological perspective to sustain the argument that Hildagsburg was constructed toward the end of the second decade of the tenth century or early in the third decade. Henry, as both primary leader of the Saxon region from 912 and as German king from 919, undertook a series of expeditions against the Slavic polities along his eastern frontier. These campaigns culminated in the well-known expedition of 929 that led to the sack of the fortified base at Brandenburg and the assertion by Henry of claims to the entire region between the Saale and Elbe rivers.140 In the geopolitical context, these successes, at least in the short term, required that Henry build strategically located fortifications east of the Elbe in order to maintain control of the newly conquered region and to provide a defense, ultimately a defense in depth, against the likelihood of Slavic counterattacks or even plundering raids.141 From an archaeological perspective, recent dendrochronological studies make clear that it was precisely during the first two decades of the tenth century that local Slavic leaders were expanding and improving their own fortifications east of the Elbe.142 The obvious conclusion would seem to be that the Slavs were acting

139 In September 937, Otto I issued a charter on behalf of the newly established monastery of St. Maurice at Magdeburg that lists 31 royal villae, which formerly had been attached to the royal curtis at Magdeburg, and which were not being transferred to the monastery. See Die Urkunden Konrad I., Heinrich I., und Otto I., ed. Theodor Sickel (Hanover, 1879), here Otto I, nr. 14. Also see the discussion of this grant by Gerhard Billig and Gerhard Böttcher, “Burgen und Burgbezirke im Erzstift Magdeburg vom 10. bis zum 12. Jh.,” Magdeburger Blätter 3 (1984), 24–38, here 27. 140 Bachrach, Warfare in Tenth-Century Germany, 27–30; and idem, “Henry I of Germany’s 929 Military Campaign,” 307–37. 141 The summer of 929 saw a major Slavic raid that culminated in the sack of the frontier fortress of Walsleben. In this context, the Ottonian system of defense in depth was based upon earlier Carolingian military schemes, including the so-called Germar Mark, the Limes Saxonicum, and the Sorbian March. In this context, see Bachrach, Warfare in Tenth-Century Germany, 32–3 and 73–7. 142 This process is described in some detail by Joachim Henning, “Archäologische Forschungen an Ringwällen in Niederungslage: Die Niederlausitz als Burgenlandschaft der östlichen Mitteleuropas im frühen Mittelalter,” in Frühmittelalterliche Burgenbau, 9–29, here 13 and 24–5; and idem, “Der slawische Siedlungsraum und die ottonische Expansion östlich der Elbe: Ereignisgeschichte-Archäologie-Dendrochronologie,” in Europa im 10. Jahrhundert, 131–46, here 132.

301

M AT E R I A L S O U R C E S A N D I M A G E S

in response to military pressure from the West.143 It is in this context that we see the need for Henry I to undertake the construction of the fortification at Hildagsburg, as well as other trans-Elben strongholds. This matrix of fortifications had a dual strategic purpose. From a defensive perspective, it was necessary to protect the already important center at Magdeburg. Offensively, Henry was developing salient strategically placed strongholds, e.g. Biederitz, Schartau, Burg, Grabow, and Dretzel, along the frontier to sustain offensive operations against the Polabian Slavs and ultimately to complete the conquest and pacification of their territory.144 In carrying out this dual defensive and offensive strategy, Henry was following a model that had been developed by Charlemagne, who had made the lower Elbe valley the northeastern frontier of the Frankish Empire.145 Among the most important fortifications established by Charlemagne in this area was Wolmirstedt, located just 2 kilometers north of Hildagsburg at what had been in the eighth century the confluence of the Ohre and Elbe rivers.146 It was well understood in the early medieval period that control of river crossings was of primary strategic importance.147

143 Bernard S. Bachrach and David S. Bachrach, “Early Saxon Frontier Warfare: Henry I, Otto I, and Carolingian Military Institutions,” Journal of Medieval Military History 10 (2012), 17–60, here 48–9 and the literature cited there. 144 For these fortifications, see Corpus archäologischer Quellen, 363, 371, 364, 366, 357, and 362, respectively. 145 Hardt, “Hesse, Elbe, Saale,” 219–32. 146 Regarding the history of Wolmirstedt, see Corpus archäologischer Quellen, 224; and Nickel, “Magdeburg in karolingisch-ottonischer Zeit,” 104. 147 In this regard, see Bernard S. Bachrach, Charlemagne’s Early Campaigns: A Diplomatic and Military Analysis (Leiden, 2013).

302

13 S O M E O B S E RVAT I O N S O N T H E B AY E U X TA P E S T RY*

During the past decade, a consensus has developed which maintains that the creation of what now is commonly called the Bayeux Tapestry was sponsored or patronized by Bishop Odo of Bayeux, William the Conqueror’s half-brother, sometime between 1066 and 1082, and that it was probably executed in the workshop of Saint Augustine’s Canterbury by an Anglo-Saxon artist. In simplest terms, these conclusions rest upon several well-established facts. Aside from Duke William, Bishop Odo is the most prominently depicted Norman on the tapestry. In addition, “apart from the major historical figures . . . only three persons are singled out by name in the Tapestry – Turold, Wadard, and Vital.” These men appear in no other account of the Conquest, but all three have been shown to have been important members of Odo’s entourage. The evidence for the workshop is equally straightforward. The artistic style exhibited on the tapestry most closely fits drawings and paintings done in Canterbury and particularly at Saint Augustine’s. In addition, Odo was earl of Kent and this placed him in close proximity to Canterbury. He and his men also had a strong presence in the city itself.1 The development of this consensus has provided an important advance over previous formulations and hypotheses by considerably narrowing our field of study and speculation. However, the new consensus does raise significant questions. Many of these can be considered under the very broad heading: How did Odo, or more likely a representative of his household, interact with the designer * The Bayeux Tapestry is an extremely large embroidery. It measures 230 feet in length and 20 inches in height. There are eight different colors of wool on linen strips sewn together. The final fragment is missing. 1 N. P. Brooks and H. E. Walker, “The Authority and Interpretation of the Bayeux Tapestry,” in Proceedings of the Battle Conference 1978, ed. R. Allen Brown (Ipswich, 1979), 1–34, and p. 8 for the quotation. This work also provides a good introduction to the huge bibliography on the Bayeux Tapestry. Additional bibliography is to be found in the Anglo-Saxons: Synthesis and Achievement, ed. J. D. Woods and D. A. E. Pelteret (Waterloo, Canada, 1985), 153–4. The best edition of the Tapestry remains: The Bayeux Tapestry, ed. F. M. Stenton et al., 2nd edition (London, 1965), and all references to the pictorial representations are to this edition. All references to the written text, unless otherwise indicated, are to my edition and translation in the appendix. David M. Wilson, The Bayeux Tapestry (New York, 1985), provides the best available color reproduction of the tapestry, but it is not a critical edition, and the commentary is very weak at many points.

303

M AT E R I A L S O U R C E S A N D I M A G E S

of the tapestry? Or to put it another way, what was the relation between the patron and the artist? Traditionally, scholars have tended not to confront this problem. Rather, they have evaded it. The approach of Shirley Brown in a recent article is representative. Brown asks: “Was the Tapestry’s goal merely to repeat what had already been established by the history-makers, or was its narrative meant to be more complex than that?”2 Brown is using a simple stylistic device to avoid the question of the relationship between patron and artist. The tapestry had no goal at all. Only those who contributed to it had a goal, or perhaps several goals, and it cannot be assumed that all of the contributors shared either a single goal or that they shared common goals. It should be emphasized here that Brown uses the word “narrative” to indicate the combination of words and pictures in their present sequence in the tapestry, as if these were the work of a single “originator.” Brown continues and asks: “Was the originator consciously referring to several different sources for his information, and were deliberate parallels drawn with other art or literary forms?”3 It is only with the publication of David Bernstein’s soundly researched and highly imaginative The Mystery of the Bayeux Tapestry that more than a passing effort has been made to deal with the question of the relation between patron and artist. Bernstein focuses upon the apparent anomaly of a Norman patron contracting with an Anglo-Saxon artist to design what most scholars perceive to be a piece of pro-Conquest propaganda. Bernstein presents a provocative and at many points compelling argument that the Anglo-Saxon artist, in fact, provided an attack on the Conquest through the subtle manipulation of the written text, traditional iconographical motifs, and deft distortions. Bishop Odo, in Bernstein’s view, got “a lot more than he bargained for.”4 According to Bernstein, the Anglo-Saxon artist was able to present his own view of things because he was “a consummate master of double meaning” while Bishop Odo was “rather naïve and literal minded.” The patron, Bernstein believes, 2 “The Bayeux Tapestry: History or Propaganda,” in The Anglo-Saxons, 11–25, and p. 14. I have been unable to consult Brown’s Cornell University doctoral dissertation: “The Bayeux Tapestry: Its Purpose and Dating,” (1977). Since the original publication of this article a substantial corpus of studies have had the Bayeux Tapestry as their focus. See, for example, the important bibliography gathered by Shirley Ann Brown, The Bayeux Tapestry: Médiathèque Municipale Ms. I: A Source Book (Turnhout, 2013); and the collection of essays in The Bayeux Tapestry and its Contexts: A Reassessment, ed. Elizabeth Carson Pastan and Stephen D. White (Woodbridge, 2014). However, none of these studies, to our knowledge, have addressed the thesis of this chapter. Nevertheless, note should be made of Michael J. Lewis, “Questioning the archaeological authority of the Bayeux Tapestry,” Cultural and Social History 7 (2010), 1–16, who presses the point that the depiction of material civilization in the Bayeux Tapestry should not be read as plain text. However, Lewis apparently was unaware of the study being republished here nor of Bernard S. Bachrach, “On the Origins of William the Conqueror’s Horse Transports,” Technology and Culture 26 (1985), 505–31, and reprinted in Idem, Warfare and Military Organization in Pre-Crusade Europe (London, 2002). Both of these studies preempt much of what Lewis has to say. 3 Brown, “The Bayeux Tapestry,” 14. 4 David J. Bernstein, The Mystery of The Bayeux Tapestry (Chicago, 1987), 112 for the quotation.

304

S O M E O B S E R VAT I O N S O N T H E B A Y E U X T A P E S T R Y

was unlikely to have been involved in the details of the composition, but upon seeing the tapestry, he, “like most observers, would have been swept along by the flow of the narrative.” Bernstein also emphasizes the practical and points out that Odo likely “never studied the work closely enough to work out its multiple levels of meaning.”5 Bernstein is much more concerned with the artist than with the patron and would appear to believe that Odo merely gave the commission for the tapestry and did nothing else. Nicholas Brooks, by contrast, appears closer to the way things were done when he suggests, “The artist who was responsible for the design of the Tapestry must have been provided with a narrative of the Conquest, and from this produced his designs and probably also the explanatory inscriptions.”6 Brooks’s view of the process explains how the artist would know where to highlight Odo’s performance and also the names of the bishop’s men: Turold, Wadard, and Vital, as well as where to place them in the narrative. If Odo’s party did provide the artist with a written account of what it wanted highlighted, then there would have been little reason for them to oversee his work. The artist, in turn, would have little except the written text to guide him on matters of importance to the Normans. Thus, the nature of the written text, i.e. Odo’s instructions, provided one can plausibly demonstrate its likely existence, is crucial to understanding the role of the patron. I will suggest here that Brooks is partially correct and that Odo’s party did provide a written text. This text, it will be suggested, is probably the written narrative, edited by the artist, which is now embroidered along the top of the pictorial narrative on the tapestry. The language of the text, i.e. words such as parabolare, suggest a Norman linguistic base.7 I will also argue that Bernstein is correct in maintaining that Odo’s party played no role in the design of the tapestry. Thus, in fact, there are two documents to be considered. There is the written text, i.e. Odo’s Gesta Wilhelmi, which the Anglo-Saxon artist occasionally edited but more frequently misrepresented because he misunderstood it. There is also the pictorial narrative which may well have had an anti-Norman double meaning; this likely accounts for a small part of the editing of the text and perhaps for some of the apparent misunderstandings. If Odo’s party provided the Anglo-Saxon artist with a written text, it is highly likely that it did not include the ubiquitous appearance of hic and the less frequent use of ubi. I am led to this conclusion because normal practice with regard to manuscript illustrations (the production of these is likely the closest analogue to the tapestry) was to provide instructions in this and similar ways. However, it was also customary to cover over these instructions.8 The inclusion of these 5 Ibid., 163–4 for the quotations. 6 Brooks and Walker, “Authority,” 3. 7 Cf. R. Lepelley, “Contribution à l’étude des inscriptions de la Tapisserie de Bayeux: Bagias et Wilgelm,” Annales de Normandie 14 (1964), 313–21. 8 C. Davis-Weyer, Early Medieval Art, 300–1150 (New York, 1971), 23–4.

305

M AT E R I A L S O U R C E S A N D I M A G E S

“instruction words” in the finished version of the tapestry may perhaps suggest that the designer or artist was not supervising his workers in a very careful manner. It is also problematic whether the written narrative itself was intended by the Norman patron to be embroidered on the tapestry. The vast majority of the Normans who were likely to see the tapestry could not read.9 Those Normans who possessed rudimentary reading skills would likely be frustrated in their efforts to decipher the written text because of the Anglo-Saxon script and spelling. If it were intended that the tapestry be displayed at a considerable height, as many scholars believe,10 even a literate person who was familiar with both Anglo-Saxon script and spelling would have had trouble reading the inscriptions. Good eyesight was needed. The lighting in any eleventh-century building was not likely to be good. There were no optical aids. The letters of the written text measure less than an inch in height and about half that in width. However, to have made the letters much taller and wider and thus easier to see would have seriously altered the composition of the tapestry as a whole. Further technical questions are raised by the often-discussed confusion in the scenes depicting the death and burial of King Edward. As most medievalists are aware, the Bayeux Tapestry has King Edward the Confessor in his coffin being carried to his burial at the church of Saint Peter the Apostle and then in the next scene as a corpse in bed. Not only are the pictures confused in chronological order but so is the written text that accompanies them.11 Scholars have devised elaborate interpretations to account for this chronological confusion.12 However, the simplest explanation is carelessness. The two bedroom scenes with King Edward both alive and dead begin the third of seven strips of linen, which when stitched together comprise the tapestry. The burial scene, however, appears at the end of the second linen strip.13 Clearly, the latter should have been allocated to the group of embroiders working on strip two, while the former should have gone to those doing strip three if the accurate chronology were to be maintained. In the context of its own strip, each scene makes sense and creates no chronological confusion. It is only when the strips are joined that the chronological disorder becomes obvious.14 If Bernstein, who ignores the potential technical impact of the strip-method of production, is correct and the chronological confusion of these scenes is crucial 9 Pierre Riché, Ecoles et enseignement dans le Haut Moyen Age de la fin du Ve siècle au milieu du XIe siècle (Paris, 1979); and Colin Morris, The Discovery of the Individual: 1050–1200 (New York, 1972), provide the background for this generalization. 10 Bernstein, Mystery, 104–7 covers the literature well regarding display but does not deal with the problem of reading the written narrative where the tapestry might be displayed. 11 Wilson, Tapestry, plates 32–33. 12 Bernstein, The Mystery, 1171–21, reviews the problem and the literature well. 13 S. Bertrand, La Tapisserie de Bayeux (La Pierre-qui-Vire, 1966), 24–5. Both Brooks and Walker, “Authority,” 190 n. 5; and Berstein, Mystery, 199 n. 7, recognize that the tapestry was made in strips but fail to associate this with the “error” regarding King Edward’s death and burial. 14 See the experiments depicted by Bernstein, Mystery, 120–1.

306

S O M E O B S E R VAT I O N S O N T H E B A Y E U X T A P E S T R Y

to the designer’s Anglo-Saxon interpretation of events, then several observations may be made. It is arguable that the artist understood that this confusion would upset literal-minded Normans if they saw it isolated at some stage of production and thus had the opportunity to concentrate their attention on it without being carried along by the narrative of the completed whole. Thus, it may be suggested that the designer, perhaps fearing a surprise visit early in the production process, when the patron was still enthusiastic about the project, carefully arranged for the scenes and the text that were to be chronologically confused to appear at the end of one working strip and at the beginning of another. With this precaution taken, the naïve Norman observer would see the chronological cogency of each separate part and not be exposed to the chronological confusion of the united whole until some later day when the impression created by the completed tapestry would perhaps obscure this “error.” The items examined thus far suggest that the artist added hic and ubi throughout the written text as instructions to his workers. Also, the Norman patrons may well not have intended that the written text appear on the tapestry. Finally, a very important scene was presented in a chronologically confused manner, both in the written text and in the pictorial narrative, in a way that cannot have been considered acceptable to the Normans. These three points tend to suggest that the AngloSaxon artist treated the written text much as he saw fit and that the Normans were not overseeing the design. Lack of Norman supervision, however, would seem to have resulted in the Anglo-Saxon artist making many errors, which not even the grand sweep of the completed tapestry could keep from literal-minded members of Odo’s entourage. These errors can be seen in the rendering of what broadly can be defined as military technology and combat techniques. Concerning the former, it may be noted that the Breton fortifications at both Dol and Dinan are depicted in the tapestry as motte and bailey strongholds, i.e. a hill or motte upon which stood a tower that was surrounded by a wall.15 Both of these fortifications, however, were old castra and probably built of stone. They were likely much more similar to the strongholds built by Fulk Nerra and his son Geoffrey Martel between 992 and 1060 in neighboring Anjou, Maine, and western Brittany than to the temporary earth and timber fortifications built by Duke William in the first few years after the conquest of England. This interpretation is supported by the terminology found in the written sources.16 Even R. Allen Brown, who throughout his career has tended

15 Wilson, Tapestry, plates 23, 25, 26. 16 All references to William of Poitiers should follow The Gesta of Guillelmi of William of Poitiers, ed. and trans. R. H. C. Davis and M. Chibnall (Oxford, 1998). book 1, ch. 45. With regard to William of Poitiers’s work, see David Bates, “The Conqueror’s Earliest Historians and the Writing of his Biography,” in Writing Medieval Biography, 750–1250. Essays in Honour of Professor Frank Barlow, ed. David Bates, Julia Crick, and Sarah Hamilton (Woodbridge, 2006), 129–44; and E. A. Winkler, “The Norman Conquest of the Classical Past,” Journal of Medieval History 42 (2016), 456–78.

307

M AT E R I A L S O U R C E S A N D I M A G E S

to defend the accuracy of the tapestry in military matters, was constrained to recognize apologetically, “the designer of the Tapestry, forced here and elsewhere to use an artistic short-hand, represents castles by their donjons, whether mottes or freestanding towers.”17 In short, Brown admits that the designer got it wrong. Perhaps even more revealing of the missed connection between the tapestry designer and Norman military reality is the depiction of the motte and bailey fortification at Hastings. The written text of the tapestry indicates that Robert of Mortain, William the Conqueror’s half-brother and Odo of Bayeux’s full brother, “ordered that a fortified encampment be dug at Hastings -castrum.”18 There was a major castrum at Hasting with a perimeter wall of almost 700 yards when the Normans arrived there.19 From the Latin of the written text it is clear that the castellum that Robert ordered to be dug was a ditched encampment that was added to the already existing castrum. This was likely a wall-and-ditch affair of the Roman type, which was well known throughout the west of France from the often-used handbook of Vegetius who wrote c. 400 A.D.20 The castellum dug at Hastings along with the already existing walls of the castrum were required to provide a defensible perimeter enclosing at least 25 acres in order to give protection to some 12,000 to 13,000 men and from 2,000 to 3,000 horses.21 This is likely the same defensive complex that William of Poitiers, who was very knowledgeable about military matters and served in Duke William’s

17

18 19

20

21

Concerning terminology, see Bernard S. Bachrach, “Early Medieval Fortifications in the ‘West’ of France: A Revised Technical Vocabulary,” Technology and Culture 16 (1975), 531–69. R. Allen Brown, English Medieval Castles (London, 1954), 35. Also now see Barbara English, “Towns, Mottes, and Ring-Works of the Conquest,” in The Medieval Military Revolution: State, Society, and Military Change in Medieval and Early Modern Europe, ed. Andrew Ayton and J. L. Price (London, 1995), 45–61. The Latin text is “Iste iussit ut foderetur castellum” at “Haestingacaestra”. Wilson, Tapestry, plate 50, 51. However, cf. Wilson, Tapestery, 229 n. 40, who seems to struggle with this part of the text. Regarding the question of logistical matters considered here, see Bernard S. Bachrach, “Some Observations on the Military Administration of the Norman Conquest,” in Anglo-Norman Studies VIII, ed. R. Allen Brown (Woodbridge, 1986), 1–25, here 22–3. Concerning the use of Vegetius’ handbook see Bernard S. Bachrach, “Practical Use of Vegetius’ De Re Militari During the Early Middle Ages,” The Historian 47 (1985), 239–55. The importance of Vegetius and particularly to William the Conqueror is emphasized by John Gillingham, “William the Bastard at War,” in Studies in Medieval History Presented to R. Allen Brown, ed. C. HarperBill, C. Holdsworth, and J. Nelson (Woodbridge, 1989), 78–91; and reprinted in Anglo-Norman Warfare, ed. Matthew Strickland (Woodbridge, 1992), 143–60. Also see A. J. Taylor, “Evidence for a Pre-Conquest Origin for the Chapels at Hastings and Pevensey Castle,” Château Gaillard 3 (1969), 144–51. Bachrach, “Military Administration,” 1–7. For a more recent review of the scholarly debate concerning the size of the conqueror’s army, see Bernard S. Bachrach, “Baudri of Bourgeuil, Adela of Blois, and the Norman Conquest,” Anglo-Norman Studies 35 (2013), 65–79. However, for the no longer supportable view of a small army, see John Gillingham, “1066 and Warfare: The Context and Place of (Senlac) of the Battle of Hastings,” in 1066 in Perspective, ed. David Bates (Leeds, 2018), 109–22.

308

S O M E O B S E R VAT I O N S O N T H E B A Y E U X T A P E S T R Y

household, described as a receptaculum for the Conqueror’s men.22 The written text on the tapestry, which is being argued here was provided by Bishop Odo’s party, agrees with William of Poitier’s account. Both written accounts tend to conform to the known topographical and military realities at Hastings.23 By contrast, the tapestry artist depicted a motte and bailey fortification. This is contrary to the meaning of the written text on the tapestry itself and strongly suggests that the artist lacked a Norman advisor to interpret the written text for him in a correct manner. The designer’s use of the motte and bailey “shorthand,” if Brown’s terminology is to be accepted, also led him astray in the depiction of the civitates of Rennes and Bayeux. Both of these walled cities were constructed during the later Roman Empire and had nothing in common with the motte and bailey fortifications depicted in the tapestry. These old Roman fortifications, which were kept in good repair in the west of France throughout the early Middle Ages, usually had a perimeter wall more than 30 feet in height and at least 10 feet thick. The perimeter walls were strengthened at regular intervals of from 20 to 30 yards by round towers. In addition, the civitas was traditionally strengthened by an internal citadel or arx of formidable proportions.24 At Rennes, for example, the perimeter wall was more than 1,300 yards and the Roman arx was supplemented by a second great tower built of stone during the later tenth century.25 For an Anglo-Saxon artist, who likely was a cleric but who lacked Norman supervision of his work, not to have been well informed about or perhaps even interested in the military architecture of Brittany and Normandy is hardly remarkable. However, for the Norman milites of Odo’s entourage, men such as Turold, Wadard, and Vital, not to mention the bishop himself, to have been ignorant of or not to have taken a keen interest in these kinds of details which determined their military planning and siege operations, hardly seems likely. In addition, it should not go unnoticed that the glory to be attained from the capture of a massive civitas or urbs was much greater than from taking a mere motte and bailey stronghold. Perhaps of greater importance in this context is that traditions of manuscript illustration were well developed in pre-Crusade Europe for the depiction of great cities that either were fortified or refortified during the later Roman Empire, as had been the situation with Bayeux and Rouen. Models were available by which such cities could be distinguished from lesser fortifications, and labeled exemplars of 22 Gesta Guillelmi, book 2, ch. 9; and the discussion by Bachrach, “Military Administration,” 23. 23 The hill which presently is part of the topography at Hastings was in the past thought to be the motte built by William and depicted in the Bayeux Tapestry. However, recent excavations have shown that this hill was not a motte at all but merely a pile of sand incapable of supporting a tower. See P. A. Barker and K. J. Barton, “Excavations at Hastings Castle, 1968,” Archaeological Journal 134 (1977), 80–100. 24 Bachrach, “Early Medieval Fortifications,” 539–49. For the latest literature dealing with these fortifications, see Bernard S. Bachrach, “Late Roman Grand Strategy: Fortification of the Urbes of Gaul,” Journal of Medieval Military History 15 (2017), 3–34. 25 Bachrach, “Early Medieval Fortifications,” 544.

309

M AT E R I A L S O U R C E S A N D I M A G E S

the latter are also known.26 At least some of the manuscripts from which models for fortified cities could have been copied were used for other purposes by the Anglo-Saxon artist designing scenes from the tapestry.27 One may perhaps ask rhetorically, in the spirit of Dr. Bernstein’s thesis, whether the tapestry artist intentionally downgraded the military architecture further to degrade the Normans? In concert with the failure to get the fortifications right, and here I mean the failure to use the correct models consistent with the type of stronghold being depicted, the designer of the tapestry erred in depicting the Norman attack techniques. At Dol, the Norman horsemen are shown riding about and throwing spears at the defenders on the walls.28 At Dinan the Norman horsemen are also seen to ride up to the walls and throw spears at the defenders. Several of the attackers – they appear from their armor to be dismounted horsemen – are depicted attempting to set fire to the bailey wall with torches.29 No foot soldiers are clearly identified in either engagement, no archers are evident, no siege works are depicted, and no siege engines are shown. From a Norman perspective, these are ridiculous depictions of attacks on fortifications.30 Duke William was a master of siege warfare, and this is made clear in the Norman sources.31 For example, William of Poitiers provides a description of part of the duke’s operations against Domfront in the following manner: William . . . advanced with an army against the Angevin territories . . . on the approach to Domfront he turned aside with fifty mounted troops. . . . As soon as possible 300 mounted and 700 foot soldiers were sent out and attacked him from the rear . . . they fled back to the stronghold. . . . The duke then was more determined than ever and decided to besiege Domfront. . . . He set up four siege works (castelli). The site of the stronghold denied its rapid capture either by storm or by sapping, for the rugged rock of the escarpment kept even foot soldiers at bay unless they approached by two narrow and difficult paths. . . . The Normans, however, harassed them with continuous and ferocious attacks. The duke, himself, was the first among those who inspired terror in the defenders. Sometimes he was for a long period of time on the look out, riding through the night or hidden under cover to see if attacks could be launched against those 26 The illustrations of Prudentius’s Psychomachia provide very good examples of major fortified cities as model types. See R. Stettiner, Die illustrierten Prudentius-Handschriften (Berlin, 1895– 1905), for the illustrations. For lesser fortifications see the reconstructed tradition of the Notitia Dignitatum. Useful examples are found in Stephen Johnson, Late Roman Fortifications (Totowa, NJ, 1983), 42–3, along with other sources. This is a subject that requires monographic research. 27 For the tapestry artist’s use of the Psychomachia see Bernstein, Mystery, 39, 153–4. 28 Wilson, Tapestery, plate 23. 29 Ibid., plates 25, 26. 30 See Gillingham, “William the Bastard at War,” 143–60. 31 For a good treatment of William as a general see David C. Douglas, William the Conqueror (Berkeley, 1964), 53–80; and Gillingham, “William the Bastard at War,” 143–60.

310

S O M E O B S E R VAT I O N S O N T H E B A Y E U X T A P E S T R Y

attempting to bring supplies, or against those carrying messages, or those seeking to ambush his foragers.32 William of Poitiers’s account, much abbreviated here, permits the observation that there was a Norman audience for accurate and credible military history of the recent past.33 The account also makes clear that mounted troops were not deployed in direct attacks on the walls of fortifications that sat atop hills or in other difficultto-reach locations. Perhaps more important than the tapestry artist’s misrepresentation of Norman combat techniques is the fact that his pictures of events at Dol and Dinan are inconsistent with the written text of the tapestry and other Norman written material. The written text of the tapestry reads: “They came to Dol and Conan turned in flight toward Rennes. Duke William’s milites fight against the men of Dinan, and Conan hands over the keys.”34 As mentioned earlier, the artist depicts the Normans attacking Dol with spears from horseback. Then Conan, the count of Rennes, is shown escaping from the tower at Dol down a rope. However, the written text of the tapestry says nothing about the Normans attacking the stronghold of Dol. It merely says that they came to Dol. The situation at Dol is described by William of Poitiers in a manner thoroughly consistent with the written account on the tapestry and fundamentally different from the pictorial representations. According to William of Poitiers, Conan was laying siege to Dol; the Normans came to relieve the siege. Conan was outside the stronghold of Dol and not in the tower. Duke William and his men did not attack Dol. When Conan learned of the duke’s approach, he fled.35 The written text of the tapestry has Conan flee toward Rennes. The tapestry artist places the Normans in pursuit so that he can carry them on to the next scene where they besiege Conan at Dinan. William of Poitiers, however, has Conan flee across the borders of Brittany, where he meets with the Angevin count, Geoffrey the Bearded, and the two men come to an agreement.36 The artist, however, ties Conan’s flight toward Rennes to a Norman pursuit and an attack on Dinan so that a continuous action is created. This is not supported by William of Poitiers’s account, it has no basis in the written text of the tapestry and makes no logistic sense. Indeed, there is no reason to believe that Duke William was even at Dinan. The text of the tapestry reads: “The soldiers of duke William fight against the men 32 Gesta Guilellmi, book 1, chapters 16–17. 33 Readers may be somewhat surprised by the large number of men who engaged William at Domfront. In light of the context, this is not a large number. On this general problem, see Bernard S. Bachrach, “Angevin Campaign Forces in the Reign of Fulk Nerra, Count of the Angevins (987–1040),” Francia 16.1 (1989, appeared in 1990), 67–84. 34 Wilson, Tapestery, plates 23–26. 35 Gesta Guilellmi, book 1, chapter 45. Brooks and Walker, “Authority,” 3, base the case “That this narrative was written and not derived orally from eyewitnesses is suggested by one sentence, where the artist appears to have misunderstood his narrative” upon the scene of Conan at Dol. 36 Gesta Guilellmi, book 1, chapter 45.

311

M AT E R I A L S O U R C E S A N D I M A G E S

of Dinan” (milites Willelmi ducis pugnant contra Dinantes). There is no mention of the duke himself. When we turn to another area of military technology, William the Conqueror’s horse transports, it is even more clear that the Anglo-Saxon designer erred significantly in depicting the kind of straightforward reality likely to have been important to the literal-minded Normans.37 A thorough examination of the literary and technological evidence concerning the duke’s horse transports indicates that he had special ships designed and built along the lines of those used by the Byzantines and Normans in southern Italy and Sicily. These ships had a high stern or poop deck, sophisticated anchoring devices, and internal structures such as stalls and ramps so that the horses could be safely transported and landed in battle-ready condition.38 By contrast, the tapestry artist depicted William’s ships as Norse vessels, both in design and in construction, with low freeboard, no structures to protect the fragile horses from the rigors of a late September crossing of the English Channel, no raised stern, no disembarkment apparatus, and no anchors. The horses are shown being transported in the very same types of vessels that are shown to be used for the transport of the men. There is no difference in the technology of the troop transports and of the horse transports.39 The total failure of the tapestry designer to recognize the importance of the special needs and problems concerning the transportation of horses is inadvertently demonstrated by Bernstein. He ingenuously remarks: “horses too were carefully observed in different settings and depicted with evident joy. We see horses obviously enjoying the sea trip.”40 Anyone remotely acquainted with horses is aware that not only do horses vigorously dislike being transported but that transporting them in any way can be dangerous to their health. This is most especially the case when transporting them by sea.41 No Norman miles, or for that matter squire or groom who had taken part in the invasion of England, could possibly have accepted the picture of war horses “obviously enjoying the sea trip.” In the same vein, the depiction of the disembarkation of the Norman war horses at Pevensey is totally without reference to what was either technically possible or militarily feasible. With the aid of a groom or sailor, two horses are seen to jump or step into shallow water from a ship with very low gunwales, i.e. about 24 inches in height.42 Such a ship could not cross the English Channel under sail loaded with 37 Wilson, Tapestery, plates 42–45. 38 Bachrach, “On the origins of William the Conqueror’s Horse Transports,” 505–31. This argument recently has been strengthed by the research of Krijnie Ciggaar, “Byzantine Marginalia to the Norman Conquest,” in Anglo-Norman Studies IX, ed. R. Allen Brown (Woodbridge, 1987), 43–63. Now also see Bachrach, “Baudri of Bourgeuil,” 65–79 for a reexamination of matters involving William’s ships. 39 Bachrach, “William the Conqueror’s Horse Transports,” 505–31. 40 Bernstein, Mystery, 71. 41 Bachrach, “William the Conqueror’s Horse Transports,” 521 n. 18. 42 Ibid., 506–8.

312

S O M E O B S E R VAT I O N S O N T H E B A Y E U X T A P E S T R Y

ten or twenty horses. Efforts by Danish Sea Scouts to create an experiment in which this scene from the tapestry was duplicated failed by a wide margin. Duke William, in fact, landed from 2,000 to 3,000 war horses in battle-ready condition at Pevensey, and the tapestry is not only incorrect in its depiction of the nature of the landing but provides a rendering of it that is technically not possible.43 This list of errors in matters of military concern to the Normans can be increased by mention of the often-mentioned mail trousers, which the tapestry artist insisted upon depicting the Norman horsemen wearing. It is well established that such equipment was not used on horseback, and to use it would cause serious injury to both the horse and the rider.44 Another error that is occasionally mentioned concerns the failure of the tapestry artist to depict the Norman foot soldiers using crossbows at the battle of Hastings when these weapons are well-attested in written accounts of the engagements.45 The Anglo-Saxon artist’s inaccurate depiction of Norman military technology and combat techniques strongly suggests that members of Bishop Odo’s entourage were not supervising the design. The specific instances where the artist’s pictorial narrative contradicts the written text or where he interpreted the written text’s possible ambiguity in a manner contradicted by other evidence (e.g. the Normans did not attack the stronghold of Dol, Count Conan of Rennes was not inside the fortifications at Dol, the defenses at Hastings were not of the motte and bailey variety), make it likely that the artist was not the author of the written text. Indeed, at points he clearly did not understand what it meant. Further inconsistency between the written text and the pictorial narrative is demonstrated by the artist’s treatment of Turold, Wadard, and Vital, the three important men from Bishop Odo’s entourage whom the prelate chose to be honored by having their names included in the written text so that they might be represented on the tapestry. Turold, the first of Odo’s men to be mentioned in the text of the tapestry, was the bishop’s tenant in Kent for the estates of Kartley, Addington, Witenamers, Eccles, Milton, Luddesdown, Stockbury, Waterbury, Little Wrotham, and Ockley. Turold is also identified as holding about the same number of estates in Essex, and he also had estates in Normandy. His son, Radulf, who was also in Odo’s entourage, appears to have inherited these holdings.46 With these facts in mind, it comes as a complete shock that Turold is depicted in the tapestry as a dwarf-like character playing the role of a groom or stable boy holding

43 Ibid., 505–31. 44 This is discussed effectively by Brooks and Walker, “Authority,” 19–20. 45 The Carmen de Hastingae Proelio of Guy Bishop of Amiens, ed. C. Morton and H. Muntz (Oxford, 1972), 112–15. But now see the new edition of this text in The Carmen de Hastingae Proelio of Guy, Bishop of Amiens, ed. and trans. Frank Barlow (Oxford, 1999). Recent efforts to see this text as not being contemporary with the Conquest have failed. On this point, see Bachrach, “William the Conqueror’s Horse Transports,” 529 n. 62. 46 The evidence is summarized by Brooks and Walker, “Authority,” 192–3, n.22.

313

M AT E R I A L S O U R C E S A N D I M A G E S

the horses of messengers sent by Duke William to Count Guy of Ponthieu.47 In short, this great landholder and honored miles is ridiculed by the tapestry artist.48 The depiction of Turold is not only wholly inconsistent with his importance as a landholder in Bishop Odo’s entourage, but it is generally agreed by scholars that both the written text and the pictures in the scenes concerning William’s messengers where Turold appears are greatly confused. The text of the tapestry reads: “Ubi nuntii Willelm ducis venerunt ad Widonem. Turold. Nuntii Willelmi. Hic venit nuntius ad Wigelum ducem.” If we ignore the word ubi as an instruction to the embroiderers, the first sentence reads: “Duke William’s messengers come to Guy.” This scene is clearly represented in the tapestry by two well-armed men dismounted and conferring with Count Guy. After this, however, the trouble begins both with the written text and the sequence of pictures. The name “Turold” undeclined in the Latin text does not follow after “Widonem” but is placed by the designer in the middle of the pictorial sequence above the head of the dwarf and just below “ducis” of the first sentence. The words “nuntii Willelmi” which follow “Widonem” in the normal line of the inscription are depicted as two mounted figures riding toward an unidentified building. Then the final nuntius who comes to Duke William is shown as an Anglo-Saxon apparently from Harold’s entourage. It may be, as scholars have often suggested, that these messenger scenes and the written texts are totally confused and out of order. It is arguable that the AngloSaxon messenger from Harold to William should lead the sequence, followed by the messengers sent by the duke to Count Guy and then the negotiations with him for Harold’s release. How or why such a serious confusion took place is not clear, nor can it be ascertained from such a reconstruction why Turold is ridiculed. Bernstein has not found that the distortions in these scenes lend support to his theory of a hidden Anglo-Saxon attack on the Normans. The view that the tapestry artist was a great master of detail makes it difficult to argue that we are looking at a mere workplace error.49 In light of the unlikelihood of a workshop error, let us see what happens if we put “Turold” back into the main line of the text after “Widonem” and thus relieve this important man of the ignominy of being depicted as a stable boy. In addition, since the general scholarly consensus is that the scene depicting William’s nuntii is out of place, might not this provide a warrant for emending the written text as we have it? A reading something like “Turold [us dux] nuntii (sic) Willelmi” as

47 Wilson, Tapestry, plate 12. Much nonsense has been written concerning Turold and often repeated. See ibid., 176, where it is noted that efforts have been made to dissociate the inscription from the picture. 48 Despite his finely tuned sense for the Anglo-Saxon attack, Bernstein, Mystery, 19, 72 does not see the ridicule of Turold. This is likely because the ridicule is obvious, and Bernstein insists that the Anglo-Saxon artist was very subtle and did not want the Normans to guess that they were being criticized. 49 Bernstein, Mystery, 19, 199 n. 16.

314

S O M E O B S E R VAT I O N S O N T H E B A Y E U X T A P E S T R Y

appositional to the first sentence may perhaps be proposed. Thus, Odo’s man gets his due as the leader of the duke’s embassy and the putatively out-of-place messengers are not out of place at all. Finally, the third element of the sequence, “A messenger comes to William,” should be seen as conveying the result of Turold’s discussion with Count Guy to the Norman duke. We learn from William of Poitiers that it was arranged for Guy to bring Harold west to the neighboring county of Eu where Duke William, who was coming east, would meet him. From Eu, William took Harold to Rouen.50 The depiction of this messenger as an Anglo-Saxon makes no sense at all. Indeed, it is hardly likely than an Anglo-Saxon in Harold’s entourage would have sufficient geographical knowledge to travel around Normandy to locate Duke William or the linguistic skills to obtain proper directions. If these emendations of the written text are accepted, it should read something like: “Before the messengers of Duke William came to Guy. Turold is the leader of William’s messengers. A messenger comes to Duke William.” This reading is consistent with Bishop Odo’s intention of honoring Turold by having him mentioned by name and with the latter justly deserving such an honor. However, some explanation is required if we are to account for the confusion in these messenger scenes and the ridicule of Turold with the suggestion that the Anglo-Saxon artist mutilated the written text provided by Bishop Odo to honor Turold. A possible reason for the Anglo-Saxon artist’s attack on Turold may be found in the latter’s role as a despoiler of the church in and around Canterbury. He was also well known to have been very much involved in the famous trial of Peneden Heath in 1072, which concerned church property.51 Finally, some reasons may be suggested as to why one of Odo’s men headed Duke William’s embassy to Guy. There are some more or less plausible speculations. For example, Turold may have had particular skills in this area, and it is for this reason that Bishop Odo valued him so highly and rewarded him so well. Turold may also have been a fidelis of Duke William as well as of Bishop Odo. It is also possible that William was in the company of Odo and Turold, somewhere west of Eu, when the duke learned of Harold’s capture and simply sent Turold as the best available man to do the job. Finally, it was a commonplace of Carolingian government, much of which either continued to be practiced in Normandy or was resuscitated by the dukes, for men such as Turold to do embassy service regardless of who his immediate dominus might be.52 Unlike Turold, who is ridiculed in the tapestry, Wadard is shown honorably in armor on a war horse. However, the written text says nothing of his role. It merely reports: “Here is Wadard.” The scene immediately prior to Wadard shows a troop 50 Gesta Guilelmi, book 1, chapter 41. 51 See note 45 above. Bernstein, Mystery, 33, was well aware of Turold’s activity in this regard. 52 Concerning Carolingian continuity and the potential family relations of Odo’s man Turold which would tie him closer to the duke see David Bates, Normandy Before 1066 (New York, 1982), 112, 155, 173.

315

M AT E R I A L S O U R C E S A N D I M A G E S

of armed and mounted Normans riding as a group and then apparently threatening someone of another group (undepicted) with their weapons. The written text reads: “and here the milites have hurried to Hastings in order to seize food.” Then in the next scene, Wadard is shown fully armed on horseback among various unarmed men who are seizing all kind of rations.53 As it has already been shown to be highly unlikely that the Normans supervised the designer of the tapestry, it may be asked how he knew what to do with Wadard who from the pictures seems to be in command of the foraging operation. One way to solve this problem is to suggest that the artist, after reading and understanding the written text provided by the patron, altered it. However, the information that he obtained from the text enabled him to depict Wadard in a command role. It seems reasonable to suggest that the original text may have read in some such manner as: “Mounted troops hurried off to Hastings to seize food, and after Wadard is in command.” This emendation thus results in a reading that is similar stylistically to the emended text recounting Turold’s role in the embassy. The treatment of Vital also indicates that the tapestry artist altered the information received from Odo’s people. Vital is treated in the following manner: “Hic Wellelm dux interrogat Vital si videsset exercitus Haroldi. Iste nuntiat Haroldum regem de exercitu Wilelmi ducis.”54 The Anglo-Saxon artist depicts the initial action by showing Duke William in conversation with a fully armed Norman horseman. This is followed by two fully armed and mounted Normans riding off in the direction of Harold’s army. After this scene comes a depiction of one of Harold’s foot soldiers looking in the direction of the Normans. This is the “iste” of the written text, which is followed by either the same observer or another of Harold’s foot soldiers, who “gives news to King Harold concerning the army of Duke Williams.”55 The sequence constructed by the Anglo-Saxon artist makes very good pictorial sense. It even makes good military sense, but it is wrong. First, it is not consistent with what is actually said in the written text. Second, it is at odds with what is known from other Norman sources. Finally, it robs Vital of his reason for being one of the three men mentioned by name in Odo’s account. From a grammatical perspective, the pronoun iste in the text quoted earlier can only refer to Vital. The two others previously mentioned in the sentence, Duke William and Kind Harold, cannot be intended for reasons too obvious to belabor. What is no less important here, however, is that the tapestry artist frequently abuses pronoun references in similar situations. For example, following the banquet at Hastings, the written text refers to a conference between “Odo episcopus, Willelm [et] Robert.” This is followed by the sentence “Iste iussit ut foderetur castellum et Hestinga-caestra.” The grammar is clear that Robert is “that very 53 Wilson, Tapestry, plates 46–48. 54 Ibid., plates 56–58. 55 Brooks and Walker, “Authority,” 30–2. However, They see a consistency between the language of the written text and the pictures.

316

S O M E O B S E R VAT I O N S O N T H E B A Y E U X T A P E S T R Y

man” who gave the order. As Duke William’s half-brother and a major figure in the Norman government, it was certainly appropriate for him to give this important responsibility after a conference with his brothers. However, the Anglo-Saxon artist not only fails to make clear in his illustrations that Robert was in command of this project but he depicts the commander in a way that might well lead a casual observer to believe that someone else had that important responsibility.56 In yet a third situation, this time with the plural isti, the artist manipulates the pictorial narrative so that it misrepresents the written text. The text reads: “Hic dederunt Haroldo coronam regis. Hic resident Harold rex Anglorum. Sitgant Archiepiscopus. Isti mirant stellam.”57 The sense is clear that just after Harold was given the crown and with Archbishop Stigand at his side, the two men were awed by Halley’s comet. This was intended by the Normans to evoke the ill omen that the comet was to have for Harold and his bishop because of the former’s perjury and the latter’s misbehavior.58 However, the Anglo-Saxon artist designed his pictorial narrative to undermine this effect by depicting the “isti” as a nondescript group of Anglo-Saxons. In each of these cases, the tapestry artist has misrepresented the pronoun iste in his pictorial narrative to the detriment of Norman interests. Thus, Vital is deprived of the honor of being identified in the tapestry as the man sent under very adverse conditions to King Harold in a last-ditch effort to avoid war. Much is made of this episode in the Norman written sources and of Harold’s answer that God would decide through battle who was right.59 By providing a distorted picture of Vital’s role, the artist also avoids the matter of Harold’s alleged boast and of Duke William’s putative efforts to make peace. Unlike Turold, who was apparently attacked for his depredation of the church, and Wadard, who was edited out of his command role, Vital’s glory would seem to have been sacrificed to a much larger cause. It is of some interest perhaps that Bernstein, despite his vigorous effort to uncover the duplicity of the Anglo-Saxon artist, missed the attack on Turold, the diminution of Wadard’s role, and the minimizing of Vital’s adventure. It is also important to note that Bernstein did not see how the tapestry artist manipulated pronoun references in order to further his anti-Norman goals. In part, I would suggest that Bernstein missed these important points because he assumed as a working hypothesis that the artist also composed the written text. The evidence discussed thus far would seem to suggest that not only did Bishop Odo patronize the tapestry but that he provided the written text to the artist. I would also suggest 56 57 58 59

Wilson, Tapestry, plates 50–51. Ibid., plates 34–35. Douglas, William the Conqueror, 181 with the literature and sources cited there. The two most important sources here are William of Poitiers, Gesta Guillelmi, book 2 chapters 11–13; and Guy, Carmen de Hastingae Proelio, lines 269–334, who both identify William’s messenger in this case as a monk. It seems more likely that he was dressed as a monk in order to carry the duke’s message so as to avoid complications. These sources discuss this man’s bravery, and Guy recounts how the messenger made a complete report to William. The entire question of negotiations and messengers between William and Harold needs more study.

317

M AT E R I A L S O U R C E S A N D I M A G E S

that the artist felt some constraint to work within the written text that had been provided for him. Just how restrained he felt, however, is problematical.

The purpose of Odo’s Gesta Guillelmi From what does appear in the written text of the tapestry, despite the Anglo-Saxon designer’s editing and distortion, it is possible to ascertain at least some of what Bishop Odo had in mind when the Gesta Guillelmi was composed and the artist received his commission. The oft-mentioned efforts to honor Turold, Wadard, and Vital, though perhaps not of primary importance, are nevertheless among the most obvious. Yet it would seem that it was not simply the bishop’s intention to have three of his fideles given notice of the kind that no other account of the Conquest thought them worthy. If the reconstruction presented earlier of their roles is correct, then these men were far from minor figures in the story. Turold headed the embassy from Duke William to Count Guy, which won the release of Harold. Wadard headed the foraging mission that scoured the countryside from Pevensey to Hastings, and the success of his mission is symbolized by the banquet that followed. Vital was perhaps the duke’s chief scout and one of his messengers in the last-minute effort to convince Harold that war should be avoided. It is important here to emphasize that the failure of other accounts of the Conquest to give credit to these men not only diminished their personal reputations but also undermined Bishop Odo’s overall contribution. A view of the written text as a whole, e.g. in a manner less obvious than the highlighting of Turold, Wadard, and Vital, indicates that Odo and his interests dominate the central section from the ordering of the invasion fleet (line 28) until the beginning of the battle of Hastings (line 45). In the first phase of this middle third of the written text, the artist edited Bishop Odo’s name out of the written text (Hic Willelm [sic] dux iussit naves edificare), but in the illustration shows the prelate playing a major role.60 Because it should appear well established by now that the Normans did not oversee the tapestry design, the most likely way for the Anglo-Saxon artist to have known of Odo’s role in the decision to build and provision the invasion fleet was to have had that information in the written text. Something to the effect that “Duke William with the advice of Bishop Odo ordered the ships to be built” may well have been in the original written text provided to the artist, which he then edited. The unambiguous representation in the illustration of a person’s role combined with a corresponding failure in the written text to inform that depiction has been noticed previously in the treatment of Wadard. Odo’s role in the building of the invasion fleet was not a fiction that the prelate sought to propagate. According to the “ship list,” a later and edited copy of a document that noted the major contributions made by Duke William’s primary 60 Wilson, Tapestry, plate 37. Bernstein, Mystery, 137–8, treats the illustrative material very well but does not seem to realize that the information depicted was not easily available except from Odo or his party.

318

S O M E O B S E R VAT I O N S O N T H E B A Y E U X T A P E S T R Y

supporters to the first phase of the Conquest, Bishop Odo is credited with providing 100 ships. Only Odo’s brother, Robert of Mortain, is indicated to have provided a larger number of vessels.61 Thus, the bishop of Bayeux’s desire to have this part of his role in the invasion highlighted was certainly reasonable. The opening phase of the Conquest in which the fleet is built and loaded (lines 28–33), i.e. the description of Odo’s contribution, is shared between the duke and his half-brother. The entire next phase of the invasion, i.e. the crossing of the Channel, is covered in the written text with “Duke William in a great ship crossed the sea and came to Pevensey (lines 33–34).” From the landing, however, Odo and his men are once again put into the forefront in the written text. Wadard, as mentioned earlier, commanded the foraging mission that left from Pevensey and went to Hastings. This leads directly to the use of the victuals that he and his men acquired in the banquet where “the bishop blesses the food and drink.” This account is then followed by a conference which according to the written text included “Bishop Odo, William [and] Robert.” Following this major role given to the duke’s two half-brothers, Robert of Mortain is shown ordering the defenses at Hastings to be constructed (lines 38–39). Less than three lines then follow in which William prepares for battle. Then Vital, Odo’s man and perhaps the duke’s chief scout, becomes the focus of attention. He is queried by the duke on the location of Harold’s army and, if the reconstruction made earlier in this chapter is correct, Vital also is given the task of trying to arrange peace between William and Harold (lines 42–45). In the battle of Hastings itself, Bishop Odo is given a major role in rallying the mounted troops in the attack (line 50). Here again, moreover, we can see how the explicit intent of the written text is subverted by the Anglo-Saxon artist. The written text reads: “Odo episcopus baculum tenens confortat pueros,” i.e. “Bishop Odo holding a staff encourages his boys.” (line 50). This is important because according to the written text Odo is not carrying a true weapon, i.e. one that was intended to draw blood. Thus, he is described verbally as acting in harmony with the canons. However, the Anglo-Saxon artist depicts Odo carrying a mace, i.e. an implement forbidden to him by church law, rather than a staff. In fact, the weapon shown for Duke William at Hastings is exactly the same as the baculum put into Odo’s hand by the tapestry designer.62 61 Brevis relation, 21–2, ed. J. A. Giles in Scriptores Rerum Gestarum Willelmi Conquestoris (London, 1845). See also Bachrach, “Military Administration,” 17 n. 33. All work on the ship list is based upon E. C. M. Van Houts, “The Ship List of William the Conqueror,” Anglo-Norman Studies 10 (1987), 159–83 and reprinted with the same pagination in idem, History and Family Traditions in England and the Continent, 1000–1200 (Aldershot, 1999). 62 See Wilson, Tapestry, plates 55, 56, 59, 68. For the background, see Friedrich Prinz, Klerus und Krieg im früheren Mittelalter (Stuttgart, 1971). Now see the recent studies on the participation of Norman and Anglo-Norman prelates in combat by Daniel M. G. Gerrard, The Church at War: The Military Activities of Bishops, Abbots and Other Clergy in England c. 900–1200 (Abingdon, 2017); and Craig C. Nakashian, Warrior Churchmen of Medieval England 1000–1250: Theory and Reality (Woodbridge, 2016).

319

M AT E R I A L S O U R C E S A N D I M A G E S

Aside from projecting himself and his followers to the forefront, Odo’s written text appears to have been aimed at seeing that others who were of importance in the Conquest were not given attention. For example, among the most prominent people in the Conquest was Bishop Geoffrey of Coutances, who actually took part in the invasion.63 Geoffrey is not mentioned in the written text of the tapestry. By omitting Geoffrey, Odo appears as the only prelate involved in the invasion and by default, he is its spiritual leader. Moreover, Odo does not leave to chance the impression that he is the spiritual head of the Norman conquest and highlights this role by giving the blessing at the banquet prior to the battle (lines 36–37). The Anglo-Saxon artist, however, would appear once again to have edited the written text which reads: “Episcopus cibum et potum benedicit.” Thus we see that Odo’s name does not appear, but the context of the artist’s depictions both in this scene and in the following one where the bishop is identified by name make it clear that the Bayeux prelate was indicated as the man who said the blessing.64 As with the case of Odo’s role in the construction of the invasion fleet, discussed previously, the Anglo-Saxon artist would appear to have found the bishop’s name mentioned in the narrative text that was provided to him, went on to depict the Bayeux prelate in the illustration, but then edited his name out of the written text that was put on the tapestry.65 The relegation of important men to obscurity by not mentioning their roles in the Conquest would appear to have been one important aim of the text provided by Odo’s party to the tapestry designer. During the years prior to the Conquest, while the campaign itself was in progress, and following the victory, seven men are conspicuous by their frequent presence at Duke William’s side. These men are Odo, Robert of Mortain, Robert of Eu, William Fitz Osbern, Richard of Evreux, Roger of Beaumont, Roger II of Montgomery, and Viscount Hugh of Avranches.66 Of these men, Roger of Beaumont was left in Normandy to head the council along with the duke’s wife. Roger II of Montgomery appears to have had the responsibility for organizing the invasion fleet and the encampment at Dives-sur-Mer. Robert of Eu likely had the same duties at Saint Valéry that Roger II had at Dives.67 Of the others, William Fitz Osbern is given more attention in the narrative sources of the invasion itself than any other figure of the duke’s entourage; this includes both Odo and Robert of Mortain.68 However, of these seven men, Odo included only his brother Robert in the written text of the tapestry. All the rest are, insofar as the tapestry is concerned, nonpersons.

63 Concerning Geoffrey, see Douglas, William the Conqueror, 119–32, 306–10, and passim; and Bernstein, Mystery, 141. 64 Wilson, Tapestry, 49–50. 65 Cf. Bernstein, Mystery, 138–41, who grasps the importance of this scene but does not address the question of where the artist obtained his information. 66 Douglas, William the Conqueror, 184–92. 67 Bachrach, “Military Administration,” 16–17. 68 Douglas, William the Conqueror, passim.

320

S O M E O B S E R VAT I O N S O N T H E B A Y E U X T A P E S T R Y

The only member of the Norman side depicted in the tapestry who is neither in Bishop Odo’s entourage nor a family member is Eustace of Boulogne.69 It was generally recognized by contemporaries of the Conquest that he played an important role at the battle of Hastings. However, Eustace fell out of favor with Duke William very shortly after the victory and remained so perhaps as late as 1080.70 Eustace’s inclusion in Odo’s account strongly suggests that it was written after the count of Boulogne was restored to the king’s good graces.71 However, from a political perspective, it may be suggested that Odo saw his brother Robert and Eustace of Boulogne as part of a factio in the constant struggle for power and influence against men such as William Fitz Osbern and others in the conqueror’s entourage. Aside from pressing his personal interests discussed earlier, Odo also would seem to have intended to detail the events leading up to the Conquest in a manner which helped to legitimize his brother William’s success. However, as Bernstein has ably argued, the Anglo-Saxon artist worked diligently to undermine this aspect of Odo’s project. Thus, in the first five lines of the written text, the purpose of Harold’s journey to the continent is obscured. The text as it appears on the tapestry reads: “EDUUARD REX: UBI; HAROLD DUX: ANGLORUM: ET SUI MILITES: EQUITANT: AD BOSHAM: ECCLESIA: HIC HAROLD: MARE NAVIGAVIT: ET VELIS: VENTO: PLENIS VENIT: IN TERRA VVIDONIS COMITIS.” In order to make some sense from the first line, it is worth noting that “Harold” is not declined. If, following the hypothesis established earlier in the chapter, we also ignore “ubi” as an artist’s instruction, it is possible to suggest that the first line read something like “King Edward [ordered] Harold.” This is what is intimated by the illustration through the king pressing his finger aggressively into Harold’s chest.72 As noted in the discussion earlier concerning Bishop Odo’s involvement with the ships and with the banquet, the artist provides information in the illustration which is not available in the written text as found on the tapestry. The most likely way the artist obtained the information for this picture, however, is from the written text provided by Odo’s party, which was then edited for the tapestry. The next line of the text provided by the Normans would then have read “The duke of the English and his milites went to the church at Bosham.” It is of importance here to recognize that the use of the word milites is an indicator of a Norman written text behind the Anglo-Saxon edition. A description such as Harold “and his knights” would not be meaningful to an Anglo-Saxon.73 Also, the reference

69 70 71 72

Wilson, Tapestry, plate 68. Carmen, xxii–xxv, xxvii–xxix. Exactly when this took place is not clear. Bernstein, Mystery, 38. Wilson, Tapestry, plate 1. Carmen, lines 289–300, for example, set out this idea. For a review of the entire problem, see Carmen, 56–72. 73 R. Allen Brown, The Normans and the Norman Conquest (New York, 1968), 94–9; and idem, “The Status of the Norman Knight,” in War and Government in the Middle Ages: Essays in Honour of

321

M AT E R I A L S O U R C E S A N D I M A G E S

to Bosham here is found only in the text of the tapestry, and such information was likely available to Odo’s party, which had direct contact with Harold through Turold’s embassy just after the Anglo-Saxon duke’s ship had landed. Indeed, there would be no more simple way for the tapestry artist to learn of this aspect of the landing. Surely, word of Harold’s brief stop at Bosham in Sussex prior to his departure for the continent is unlikely to have carried all the way to Canterbury by word of Anglo-Saxon mouth. Lines 3–5: “Harold [sic] nagavit et velis vento plenis venit in terra Widonis comitis” constitute the only part of the written narrative on the tapestry that could have literary value. The unique prose description of Harold’s sailing may perhaps be an example of the style of the original Gesta Guillelmi before the Anglo-Saxon artist edited it, either as instructions for his embroiders or for inclusion on the tapestry. Wherever the truth may lie here, the vast difference stylistically between these lines and the rest of the written text merits attention. Thus it is of considerable importance that the italicized part of the lines quoted earlier would appear to be an allusion to or an echo of a line from Virgil’s Aeneid (book 5, line 281): “Vela fecit tamen et plenis subit ostia velis” with various manuscript readings such as “velis . . . plenis” and “plenis . . . velis.” Indeed, even as the line from the tapestry now stands, it almost scans.74 The significance of this Virgilian allusion or echo goes well beyond the putative literary talent of the man who used it; it has great importance for Duke William’s case against Harold. The Aeneid (book 5 line 281) deals with Sergestus who is a military dependent or armed follower of Aeneas. Sergestus almost wrecks his ship while bringing it into port. The ship is described as an “inglorious vessel” and Sergestus’s performance is “sine honore” (line 272). This Virgilian echo would seem to evoke the Norman accounts of Harold having been shipwrecked and of him earlier having recognized Duke William as his dominus.75 Most important, however, in establishing the Virgilian context of this echo is the phrase “sine honore,” which was of great importance to men like William the Conqueror and his contemporaries. During this period, the Latin word honor had two significant meanings which are relevant to the present discussion. Honor could mean a feudum or beneficium held by a fidelis or vassus (vassalus) from his dominus or senior and it could also mean dignitas.76 The locus classicus from a juridical perspective for this dual meaning is provided by Fulbert of Chartres in a letter that he composed defending Count Odo II of Blois against charges brought by King Robert c. 1023. Robert asserted that

J. O. Prestwich, ed. J. Gillingham and J. C. Holt (Woodbridge, 1984), 18–32. Winkler, “The Norman Conquest of the Past,” 456–78 for the broad picture. 74 I am grateful to my former student Dr. Henrietta Warwick, author of A Virgil Concordance (Minneapolis, 1975), for giving precision to my intuition on this point. 75 Douglas, William the Conqueror, 175–8. 76 F. L. Ganshof, Feudalism, trans. Philip Grierson, 2nd English edition (New York, 1961), 106–13.

322

S O M E O B S E R VAT I O N S O N T H E B A Y E U X T A P E S T R Y

Odo was unworthy of holding any beneficium from him, and he makes known his intention to strip the count of all of his honores. Odo responds that he has hereditary rights to these holdings over and above the fact that he received them as beneficia from Robert. This claim is rejected, as is Odo’s argument that because he served Robert loyally in military operations he deserves his honores. Finally, the case is made that if Robert deprives Odo of his honores in the sene of beneficia, he also deprives the count of his honor in the sense of dignitas.77 The Norman claim that Harold was William’s military dependent is presented in the written text of the tapestry in the following manner (lines 13–21): Duke William and his army came to Mont Saint-Michel and they crossed the river Couesnon. Here Duke Harold dragged them from the sand and they came to Dol and Conan turned to flight [toward] Rennes. Duke William’s milites fight against the men of Dinan and Conan handed over the keys. William gave arms to Harold. William came to Bayeux where Harold made an oath to Duke William. Harold bravely fights in William’s service. Then the Anglo-Saxon is granted arms by the Norman duke. Finally, Harold takes the oath to William. To the Normans this was the classic manner of establishing honorable dependence relative to the situation in Count Odo’s letter to King Robert. The point to be emphasized here, however, is that the Normans saw this from Robert’s side, and Duke William’s uncle, Richard, had been the king’s representative in the conflict with Count Odo in 1023. From the Norman-Capetian perspective in 1023, neither hereditary claims nor military service would keep a lord from “dishonoring” his dependent if he violated his oath.78 As mentioned earlier and largely in accord with Bernstein’s arguments, it seems clear that the Anglo-Saxon artist sought to undermine the Normans’ position. The death and burial scenes were reversed, Harold’s and Stigand’s association with the comet as an ill omen is obscured, and in lines 29–31, Harold’s rejection of William’s claims seems to have been edited out. These distortions of the Norman claims thus give even greater weight to lines 3–5 where the Virgilian echo, apparently undetected by an Anglo-Saxon monk who was likely not well informed concerning continental juridical arguments concerning honores, let slip through the allusion or echo which highlights Harold “sine honore” as a dependent of Duke William.

77 The Letters and Poems of Fulbert of Chartres, ed. and trans. Frederick Behrends (Oxford, 1976), number 86. 78 Ibid. See also number 51, and the study by Bernard S. Bachrach, “Enforcement of the Forma Fidelitatis: The Techniques used by Fulk Nerra, Count of the Angevins (987–1040),” Speculum 59 (1984), 796–819.

323

M AT E R I A L S O U R C E S A N D I M A G E S

Norman reaction to the tapestry After the tapestry was completed, probably sometime before 1082, it dropped from historical sight for almost four centuries. It is first mentioned in an inventory of the church of Notre-Dame of Bayeux that was drawn up in 1476. In this inventory it is stated that the tapestry “is hung round the nave of the church on the feast of the relics and throughout the Octave.”79 It is thus of great importance to emphasize that between c. 1082 and 1476 no one was sufficiently impressed by this unique piece of historical artistry to find it noteworthy in a written work that has survived. When the practice of displaying the tapestry began at Notre Dame or any other place is unknown. It can be stated with confidence, however, that no one writing about the Conquest within a century of Duke William’s victory can be shown to have used certain important information provided by the tapestry. Here, matters such as Harold’s stop at Bosham and detention at Beaurain are noteworthy. No less significant is the omission of Turold, Wadard, and Vital from all accounts. The omission of the capture of Dinan is also important. Finally, Bishop Odo’s supposed central role in the building of the fleet, the planning of the battle, and the battle itself went unnoticed outside the tapestry. Of great importance in emphasizing the absence of tapestry influence, the works of Wace and Baudry of Bourgeuil deserve special attention. Wace was a canon of Bayeux cathedral with the mentalité of an antiquary and a vivid interest in the Norman conquest. However, scholars not only have been unable to prove that Wace used the tapestry, it cannot even be shown that he ever saw it. The one noteworthy effort to demonstrate a connection between Wace and the tapestry is methodologically unsound and proceeds from the assumption that the tapestry artist copied from Wace’s work.80 Baudry of Bourgeuil, who apparently took a lively interest in pictorial representations of the Norman Conquest, never mentions the Bayeux Tapestry. This despite writing a poem describing a “Tapestry” belonging to Adele of Blois, Duke William’s daughter. She is alleged to have possessed a wall hanging on which the Conquest is depicted.81 Indeed, in those places where one might compare Baudry’s poem with the Bayeux Tapestry, such as with regard to Duke William’s horse transports, the accounts are diametrically opposed. Baudry’s treatment is consonant with that of the Carmen and Wace where different words are used to

79 Wilson, Tapestry, 12. 80 Matthew Bennett, “Poetry as History? The Roman de Rou of Wace as a Source for the Norman Conquest,” in Anglo-Norman Studies V, ed. R. Allen Brown (Woodbridge, 1983), 21–39, here 22–3. Also see Elisabeth van Houts, “Wace as a Historian,” in Family Trees and the Roots of Politics: The Prosopography of Britain and France from the Tenth to the Twelfth Century, ed. Katharine S. B. Keats-Rohan (Woodbridge, 1997), 103–32. 81 Cf. Bernstein, Mystery, 209 n. 4–6.

324

S O M E O B S E R VAT I O N S O N T H E B A Y E U X T A P E S T R Y

describe the boats used to transport foot soldiers and those for horses.82 The tapestry artist depicts these ships as not differing at all.83 The apparent obscurity of the Bayeux Tapestry during the first three or four centuries following its completion permits the hypothesis that it was rejected by its Norman patrons. Prima facie, the remarkably good condition of the tapestry, indeed even its very survival – a unique case for such a wall hanging of the eleventh century – strongly suggests that it was not hung very often and certainly that it was not displayed in some cold, damp Norman church or hall for several years, much less for several generations or centuries. There is no physical evidence of the kind of damage that would have been the result of gravity’s impact on such a wall hanging.84 There were many good reasons why Odo and his party would not have been pleased with the tapestry, why they would likely have stuffed it in a sea chest and shipped it off to Normandy where it was relegated to the obscurity that they felt it deserved. Even if we assume that all or at least most of the duplicity of the AngloSaxon artist that Bernstein has so imaginatively identified went unnoticed by the patrons, there was much about which the “literal-minded” Normans could object. As we have seen earlier, numerous depictions represented the Norman military reality in a corrupt manner. Among the most obvious are: 1 2 3 4 5 6 7 8 9

Important secondary fortifications were depicted as minor motte and bailey defenses Major urban fortifications were depicted as minor motte and bailey defenses The castellum constructed under the direction of Robert, Bishop Odo’s brother, was misrepresented as a motte and bailey fortification The horse transports were depicted in an absurd and grossly inaccurate manner The disembarkation of the Norman troops was depicted in a grossly inaccurate manner The mail worn by the Norman horsemen was depicted in a grossly inaccurate manner Crossbowmen were not depicted in the battle at Hastings Norman siege tactics were depicted in a grossly inaccurate manner The campaign at Dol was depicted in a grossly inaccurate manner

In a no less distorted manner, Turold, one of the three men chosen by Bishop Odo to be honored by having him included in the tapestry, was ridiculed by the Anglo-Saxon artist and depicted as a dwarf-like stable boy. In a somewhat more sophisticated way, the artist diminished Wadard’s role and also that of Vital. The

82 Bachrach, “William the Conqueror’s Horse Transports,” 510. 83 Bachrach, “Baudri of Bourgeuil,” 65–79. 84 Cf. Bernstein, Mystery, 104–7, who imagines a very arduous and bruising routine for the tapestry, which includes transportation to different sites.

325

M AT E R I A L S O U R C E S A N D I M A G E S

messenger scenes were confused, as were the death and burial of King Edward and the purpose of Halley’s comet. In conclusion, it may be emphasized that the Anglo-Saxon artist or designer was, as Bernstein has argued, an exceptionally talented and able man. He surely subverted the written text provided to him by his Norman patrons in very many subtle ways. However, in order to carry out his program, he likely avoided any effort by the patron to oversee his work. This paradoxically resulted in the commission of numerous “mistakes,” as seen from the perspective of a “literal-minded” man such as Bishop Odo.85 The Normans thus rejected the tapestry and did not display it. Ironically, rejection of the tapestry was responsible for its survival, and this in turn makes it possible for us to recognize the Anglo-Saxon who designed it as a great artist.

85 Ibid., 164 for Odo not only as literal minded but also as “rather naïve.” Cf. David Bates, “The Character and Career of Odo, Bishop of Bayeux (1049/1050–1097),” Speculum 50 (1975), 1–20.

326

APPENDIX Odo’s Gesta Guillelmi86

King Edward [instructs] Harold . . . the duke of the English and his milites ride to Boscham and visit the church. Harold sailed the sea and with the wind full in his sails he came to the country of Count Guy. Harold [lands]. Guy seizes Harold and led him to Beaurain and held him there. Harold and Guy confer. Duke William’s messengers came to Guy. Turold [is the leader of] Williams messengers. A messenger comes to Duke William. Guy led Harold to William, duke of the Normans. Duke William comes with Harold to his palace where a clericus and Aelfygyva [argue].87 Duke William and his army come to Mont Saint-Michel and they crossed the river Couesnon. Here Duke Harold dragged them from the sand and they came to Dol and Conan turned in flight [toward] Rennes. Duke William’s milites fight against the men of Dinan and Conan handed over the keys. William gave arms to Harold. William came to Bayeux where Harold made an oath to Duke William. Duke Harold returned to English soil and came to King Edward. King Edward in bed addresses his fideles and he is dead. The body of King Edward is carried to the church of Saint Peter the Apostle. They have given the king’s crown to Harold Harold remains sitting as king of the English. Arch-bishop Stigand [is at his side]. These men wonder at the star. Harold [rejects William’s claim]. An English ship came to the land of Duke William [with Harold’s rejection]. Duke William [with the advice of Bishop Odo] ordered ships to be constructed. The ships are dragged to the sea. These same men carry arms to the ships and drag cart[s] loaded with wine and arms [to the ships]. Duke William in a great ship crossed the sea and came to Pevensey. The horses leave the ships, and the milites have hurried off to Hastings to seize food. Wadard is [in command]. The meat is being cooked, and servants serve. They made a banquet and Bishop [Odo] blesses the food and drink. Bishop Odo, William, [and] Robert [confer]. The latter ordered that a fortified 86 What follows is a partial reconstruction and translation of the putative written text provided by Bishop Odo to the Anglo-Saxon artist. Justification for the reconstruction is to be found throughout the text of the chapter. Those arguments not presented in the chapter directly will be found in the notes. 87 By and large, I have followed the identification of Aelfgyva made by J. Bard McNulty, “The Lady Aelfgyva in the Bayeux Tapestry,” Speculum 55 (1980), 659–68.

327

M AT E R I A L S O U R C E S A N D I M A G E S

encampment be dug at the stronghold of Hastings. News is brought to William concerning Harold. A house is burned. The militess went out of Hastings and came to do battle against King Harold. Duke William asks Vital if he has seen Harold’s army. The latter carries a message to King Harold concerning [the superiority] of Duke William’s army. Duke William encourages his milites so that they might prepare themselves manfully and wisely for battle against the army of the English Leofwine and Gyrth, the brothers of King Harold, fell. The English and the French in battle at the same time. Bishop Odo holding a staff encourages his boys. Duke William [raises his visor to show he is alive] and Eustace [says]: “Here is William.” The French attack and those who are around Harold are cut down. King Harold is killed, and the English turned in flight.88

88 As seen passim, the Bayeux Tapestry has much in common with William of Poitiers’s Gesta Guillelmi. However, it is obvious that the “style” of the tapestry text has little in common with the writing of William, whose work was dominated by his classical education. However, the allusion to Virgil (book 5.281) seems to lead to the suspicion that the original text submitted by Odo to the designer may have contained a greater number of classical allusions that were bowdlerized. The suspicion that William of Poitiers played some role in the text of the tapestry may be strengthened by the fact that King William’s chaplain had a close relationship with Bishop Odo.

328

14 H E N RY I O F G E R M A N Y’S 929 M I L I TA RY C A M PA I G N IN ARCHAEOLOGICAL PERSPECTIVE

Introduction In early 929, King Henry I of Germany (919–36) began a three-phased military campaign against the Slav polities along the entire length of his eastern frontier. The first stage culminated in the capture and sack of the fortress of Brandenburg. This was the princely seat of the Hevelli, one of the powerful peoples of the large Weleti confederation, who lived in the lands between the lower Elbe and Oder rivers.1 Henry then turned south, where he captured and destroyed the major fortress of Gana, the main seat of the Daleminzi, a Sorbian people living between the middle Saale and the middle Elbe.2 Henry then marched farther south into Bohemia, where he undertook a siege of Prague and forced Duke Wenceslaus (921–935) to surrender and accept his authority.3 The campaign of 929 was a military triumph and demonstration of Henry I’s power, particularly the skill of his army in the dangerous undertaking of siege warfare. The campaign also marked a fundamental change in east Frankish/German policy toward the east.4 The entire Daleminzi region was brought under German 1 See the basic work on the Hevelli by Lothar Dralle, Slaven an Havel und Spree. Studien zur Geschichte des hevellisch-wilzischen Fürstentums (6. bis 10. Jahrhundert) (Berlin, 1981). 2 Regarding the broad division of Slavic peoples of trans-Elben and trans-Saale regions into three major confederations, the Weleti, Obodrites and Sorbs, see the discussion by Gerhard Lebuda, “Zur Gliederung der slawischen Stämme in der Mark Brandenburg (10. – 12. Jahrhundert),” Jahrbuch für die Geschichte Mittel und Ostdeutschland 42 (1994), 103–39, particularly 103–11. 3 With respect to the establishment of Bohemia as a stable political entity, see the discussion by Herbert Ludat, “Böhmen und die Anfänge Ottos I.,” in Politik, Gesellschaft, Geschichtsschreibung: Giessener Festgabe für Frantisek Graus zum 60. Geburtstag (Cologne, 1982), 131–64. 4 Scholars generally have seen Henry’s policy with regard to the trans-Saale region as a basic continuation of Carolingian policy, which entailed the imposition of tribute upon the Slavic polities but not their conquest and integration into the eastern kingdom. In this regard see, for example, Walter Mohr, König Heinrich I. (919–936): Eine kritische Studie zur Geschichtschreibung der letzten hundert Jahre (Saarlouis, 1950), 54–5; Robert Holtzmann, Geschichte der sächsischen Kaiserzeit, 3rd edition (Munich, 1955), 93; Josef Fleckenstein, Early Medieval Germany, trans. Bernard S. Smith (New York, 1978), 133–5 (in original, Grundlagen und Beginn der deutschen Geschichte);

329

M AT E R I A L S O U R C E S A N D I M A G E S

control, and the frontier was pushed 130 kilometers to the east, from the river Saale to the Elbe. From 929 onward, Henry I actively campaigned in the lands east of the Elbe river and established a policy of territorial conquest that ultimately would lead, under his son Otto I (936–73), to the acquisition by the German kingdom of all of the lands between the Elbe and the Oder. The importance of the 929 campaign for the expansion and development of the German kingdom is widely recognized in the scholarly literature and need not be belabored here. What is striking, however, is that there is no book-length, or even article-length, study dedicated to this campaign.5 This is due, in large part, to the relative dearth of information about the events of 929 in the written sources. Widukind of Corvey, who began writing his history of the reigns of Henry I and Otto I some three decades after the campaign of 929, is our most important informant and provides an outline of the three phases of Henry I’s campaign.6 Two brief accounts in the Annales Ratisponenses and Auctarium Garstense, which shed additional light on the siege of Prague, round out the narrative accounts.7 One charter, issued by Henry I on 30 June 929 at Nabburg, some 55 kilometers northnortheast of Regensburg, provides a terminal date for the campaign.8 By contrast with the relative dearth of written sources, vast quantities of information have been developed over the past sixty years by archaeologists regarding the military topography of both the east Frankish/German kingdom and her Slav neighbors. Many scores of fortifications have been excavated, and there is now detailed information available regarding the number, size, and strength of the defenses of the Hevelli, Daleminzi, and Bohemians, circa 929.9 These data, when analyzed in the context of Sachkritik, that is the material reality within which military operations were carried out, provide insights regarding King Henry’s likely

5

6

7

8 9

Wolfgang Giese, Heinrich I.: Begründer der ottonischen Herrschaft (Darmstadt, 2008), 114–15. However, several scholars have developed the idea that Henry intended to use the trans-Saale region as a buffer against Hungarian attacks. See, for example, Franz Lüdke, König Heinrich I. (Berlin, 1936), 104–5; and more recently, Dralle, Slaven an Havel und Spree, 109, raised the possibility that Henry saw this region as a possible “glacis” that would protect the “Festung” of Saxony against Hungarian attack. The only scholarly investigation of the military nature of the events of 929 of which I am aware is Werner Hülle, Westausbreitung und Wehranlagen der Slawen in Mitteldeutschland (Leipzig, 1940), 41, which is based, in part, on the limited archaeological material available at that time. Widukind of Corvey, Res gestae Saxonicae, in Quellen zur Geschichte der sächsischen Kaiserzeit, ed. and trans. Reinhold Rau, vol. 8 (Darmstadt, 1971), 1.35. This text is now available in English translation in Bernard S. Bachrach and David S. Bachrach, Widukind of Corvey: Deeds of the Saxons (Washington, DC, 2014). Annales Ratisponenses, ed. W. Wattenbach, MGH SS 17 (Hanover, 1861), 583, “Heinricus Boemiam vicit cum Arnolfo.” Also see Auctarium Garstense, MGH SS 9 (Hanover, 1851), 565, “Hainricus rex cum Arnolfo duce Boemanos vicit.” Die Urkunden Konrad I., Heinrich I. und Otto I., ed. Theodor Sickel (Hanover, 1879–1884), here Henry I., nr. 19. See, for example, the important collection of essays in Frühmittelalterliche Burgenbau in Mittel und Osteuropa, ed. Joachim Henning and Alexander T. Ruttkay (Bonn, 1998).

330

H E N RY I O F G E R M A N Y ’ S 9 2 9 M I L I TA RY C A M PA I G N

campaign route and strategy, as well as the scale of his undertaking. The focus of this study, therefore, will be to provide a detailed military analysis of King Henry’s 929 campaign, including the material, logistical, and manpower requirements that his operations entailed.

Historical background During the first decade of the tenth century, the political and military situation along the eastern frontiers of the east Frankish kingdom was altered dramatically by the rise of the Hungarians as a power in the Carpathian basin. In the southeast, the traditional Carolingian policy of establishing hegemony over the Moravian and Bohemian polities was shattered by the crushing defeat of a Bavarian army at Pressburg in 907 at the hands of the Hungarians.10 The effort by the Franks to restore the status quo ante failed when a large Carolingian army under the nominal command of Louis the Child (899–911) was defeated by the Hungarians in a battle near Augsburg in 910.11 Following their victory at Pressburg, the Hungarians began to develop a diplomatic strategy to construct an anti-Frankish coalition among the Slav polities along the eastern frontier of the kingdom.12 In doing so, the Hungarians were able to build upon the existing alliances between the Bohemians, with their political center at Prague, and the Hevelli. The latter dominated the region along the Havel river, as well as the hilly country to the south now called Fläming.13 In 906/907, the Hevelli princess Drahomir was married to Duke Vratislav of Bohemia (died 921). Moreover, the familial relationship between the Hevelli ruling family and the Premyslids in Bohemia likely dated back to the ninth century.14 A fourth member

10 Regarding the battle at Pressburg, see Adalberts Fortsetzung der Chronik Reginos (Adalberti continuatio Regionis) in Quellen zur Geschichte der sächsischen Kaiserzeit, trans. and ed. Albert Bauer and Reinhold Rau (Darmstadt, 2002), based on the translation of Paul Hirsch, Max Büdinger and Wilhelm Wattenbach, hereafter Adalbert, Continuatio, an. 907. Also see the English translation of this text in Simon MacLean, History and Politics in Carolingian and Ottonian Europe: The Chronicle of Regino of Prüm and Adalbert of Magdeburg (Manchester, 2009). For a detailed discussion of the political implications of this battle for Carolingian policy in the south-east, see R. Hiestand, “Preßburg 907: Eine Wende in der Geschichte des ostfränkischen Reiches?,” Zeitschrift für bayerische Landesgeschichte 57 (1994), 1–20; and Charles R. Bowlus, Franks, Moravians, and Magyars: The Struggle for the Middle Danube, 788–907 (Philadelphia, 1995), 258–67. 11 Regarding the battle at Augsburg, see Adalbert, Continuatio, an. 910; Liudprandi Cremonensis Opera Omnia: Antapodosis, Homelia Paschalis, Historia Ottonis, Relatio de Legatione Constantinopolitana, ed. P. Chiesa (Turnholt, 1998), here Antapodosis, 2.3–4. 12 See the discussion by Bernhard Friedmann, Untersuchungen zur Geschichte des abodritischen Fürstentums bis zum Ende des 10. Jahrhunderts (Berlin, 1986), 180–1. 13 For an overview, see Dralle, Slaven an Havel und Spree. 14 See the discussion of these political ties by Lothar Dralle, “Zur Vorgeschichte und Hintergründen der Ostpolitik Heinrichs I.,” in Europa Slavica, Europa orientalis: Festschrift für Herbert Ludat zum 70. Geburtstag, ed. Klaus-Detlev Grothusen and Klaus Zernack (Berlin, 1980), 99–126, here 115; and Friedmann, Untersuchungen zur Geschichte des abodritischen Fürstentums, 180–1.

331

M AT E R I A L S O U R C E S A N D I M A G E S

of the anti-Frankish community of interests was comprised by the Daleminzi, one of the Sorbian peoples, whose main area of settlement was between the Saale and the Elbe river valleys. They had a long and often hostile relationship with the east Frankish kingdom. Regular raiding by the Sorbs into Thuringia and southern Saxony had led to the establishment and maintenance by the Carolingian government of a very substantial system of fortifications along the Saale and Unstrut rivers, which ultimately came to be known as the Sorbian March (Limes Sorabicus).15 The Hungarians very quickly made good use of their relationships with the Bohemians and Daleminzi to launch substantial raids into the long-settled and wealthy regions of Thuringia and Saxony. As early as 908, the year after Pressburg, the Bohemians permitted Hungarian horsemen to pass unmolested through the Bohemian plain to descend down from the Erzgebirge, which marks the northern edge of the high Bohemian plateau, into the valleys of the Zwickauer and Freiburger Mulde, two rivers which were located in the lands of the Daleminzi.16 The Daleminzi, for their part, are reported by Widukind of Corvey to have provided guides to the Hungarians to lead them into the lands of the east Franks.17 The Hungarians launched additional raids into Saxony and Thuringia in 912, 915, and 924.18 The reported savagery of the Hungarian raids led Henry I, first as leading man in Saxony (from 912) and then as king, to broaden and strengthen his defenses in the east by pushing his frontier beyond the Saale to the Mulde river valley, which provided a major route of entry for the Hungarians as they came north through Bohemia.19 The written testimony for this policy comes from a rather late source, Thietmar of Merseburg (d. 1018), who, although writing considerably after the events in question, was very well informed about military matters in Saxony and Thuringia. He recorded that Henry was surprised by a vastly superior Hungarian force while campaigning along the Mulde River.20 Rather than risk battle against a larger force, Henry sought refuge in the fortification at Püchen (modern Püchau),

15 See the discussion by Walter Schlesinger, Die Entstehung der Landesherrschaft, Teil I (Dresden, 1941), 79; Paul Grimm, Handbuch vor-und frühgeschichtlicher Wall-und Wehranlagen Teil I: Die vor- und frühgeschichtlichen Burgwälle der Bezirke Halle und Magdeburg (Berlin, 1958), 38–56; Hansjürgen Brachmann, “Der Limes Sorabicus – Geschichte und Wirkung,” Zeitschrift für Archäologie 25 (1991), 177–207; and Matthias Hardt, “Hesse, Elbe, Saale and the Frontiers of the Carolingian Empire,” in The Transformation of Frontiers from Late Antiquity to the Carolingians, ed. Walter Pohl, Ian Wood, and Helmut Reimitz (Leiden, Boston and Cologne, 2001), 219–32. 16 Regarding the, at least, passive assistance provided by the Bohemians to the Hungarians, see Friedmann, Untersuchungen zur Geschichte des abodritischen Fürstentums, 180–1. For the raid of 908, see Adalbert, Continuatio, an. 908. 17 Widukind, Res gestae, 1.20. 18 For the earlier raids, see Adalbert, Continuatio, an. 912 and 915. For the raid of 924, see Widukind, Res gestae, 1.32. 19 See note 4. 20 The newest edition is now Thietmar von Merseburg, Chronik, ed. and trans. Werner Trillmich, 8th edition (Darmstadt, 2002), hereafter, Thietmar, Chronicon, 1.15. For an English translation of this text, see David Warner, Ottonian Germany: The Chronicon of Thietmar of Merseburg (Translation, 2001).

332

H E N RY I O F G E R M A N Y ’ S 9 2 9 M I L I TA RY C A M PA I G N

located on the west bank of the Mulde River, some 60 kilometers east of the Saale frontier at Merseburg.21 The inhabitants, according to Thietmar, were still praised in his own day for the glory that they had earned in helping to defend the king.22 The dating of this encounter with the Hungarians is problematic because Thietmar sets the encounter within the general framework of Henry’s efforts to keep the Hungarians at bay.23 In the absence of an excavation of the fortification at Püchen with dendrochronological dating of its wooden elements, Henry’s establishment of a substantial outpost at the Mulde cannot be dated more precisely than 908–924, marking, respectively, the first Hungarian raid into this region and Henry’s peace agreement with the Hungarians.24 The latter came about following the capture of the commander of a Hungarian raiding party in 924.25 Following his peace agreement with the Hungarians in 924, Henry turned his attention to the west, where he successfully reintegrated Lotharingia into the east Frankish kingdom.26 Henry’s final military operation in the west took place in 928, when he personally led a large army to lay siege successfully to the fortification (castrum) of Durofostum on the Meuse, which belonged to Count Boso of Pérthois, the brother of King Raoul.27 With affairs settled in the west, Henry was able to turn his attention again to the eastern frontier and undertake his long-cherished goal of establishing a major defensive bulwark against future Hungarian raids.

The campaign of 929 The very extensive campaign during the winter and spring of 929 required considerable preliminary planning. The three main objectives of the campaign, at Brandenburg, Gana, and Prague, spanned a significant length of the eastern frontier of Henry’s kingdom, separated from north to south by a distance of 300 kilometers as the crow flies. These three political, economic, and military centers were protected, in aggregate, by several dozen other fortified strongholds, many of which were quite large and strongly built as well. As a consequence, in addition to detailed intelligence, a major campaign of this size and duration required the mobilization of large military forces and substantial quantities of supplies. 21 22 23 24

Thietmar, Chronicon, 1.15. Ibid. Ibid. Widukind, Res gestae, 1.32. See the discussion of these events by Bernard S. Bachrach and David S. Bachrach, “Saxon Military Revolution, 912–973?: Myth and Reality,” Early Medieval Europe 15 (2007), 186–222, here 220–2 for the nature of this peace agreement. 25 For the capture of the Hungarian commander, and the subsequent peace agreement, see Widukind, Res gestae, 1.32. 26 See the discussion of these events in David S. Bachrach, Warfare in Tenth-Century Germany (Woodbridge, 2012), 36–7, and the literature cited there. 27 Les Annales de Flodoard, ed. Ph. Lauer (Paris, 1905), an. 927.

333

M AT E R I A L S O U R C E S A N D I M A G E S

Phase one: the capture of Brandenburg When Henry I and his advisors began to plan their campaign to capture Brandenburg, they benefitted from a wealth of intelligence that was available, at least in part, from large numbers of merchants who regularly traveled along the wellestablished trading route between Magdeburg and Brandenburg that dated back to the eighth century.28 The interrogation of merchants was fundamental to basic intelligence gathering throughout the early medieval period and was well known to the Ottonian court from long-standing Carolingian practice.29 Henry and his commanders almost certainly had gained additional intelligence in the course of military operations to the east and northeast of Magdeburg, a principal Ottonian fortification and trading center on the middle Elbe, that resulted in the capture of fortifications at Biederitz, Schartau, Burg, Grabow, and Dretzel, during the first or second decade of the tenth century.30 This group of erstwhile Slavic fortifications, all of which were located within a day’s march of Magdeburg, provided an important defensive screen against attacks from the east and northeast. The complex of defensive works in the immediate vicinity of Magdeburg were matched farther north along the course of the middle Elbe by fortifications at Tangermünde, Osterburg, and Walsleben.31 This line was anchored at the confluence 28 Regarding the well-established trade route between the two centres, see J. Hermann, MagdeburgLebus. Veröfftenlichungen des Museums für Ur-undFrühgeschichte Potsdam 2 (Potsdam, 1963), 89–106, here 90. 29 See the detailed discussion of intelligence-gathering by the Carolingian government in Bernard S. Bachrach, “Charlemagne and the Carolingian General Staff,” The Journal of Military History 66 (2002), 313–57, particularly 319–22. For intelligence gathering by the Ottonian kings, see David S. Bachrach, “Military Intelligence and Strategic Planning under the Ottonian Kings of Germany, c. 919–1024,” in Military Cultures and Martial Enterprises in the Middle Ages: Essays in Honor of Richard P. Abels, ed. John Hosler (Woodbridge, 2020), 61–87. 30 With respect to the military and economic importance of Magdeburg in the early tenth century, see the discussion by Walter Schlesinger, “Zur Geschichte der Magdeburger Königspfalz,” Blätter für deutsche Landesgeschichte 104 (1968), 1–31; Ernst Nickel, “Magdeburg in karolingischottonischer Zeit,” Zeitschrift für Archäologie 7 (1973), 102–42; Eike Gringmuth-Dallmer, “Die Siedlungsgeschichtlichen Grundlagen für die Enstehung Magdeburgs als Zentrum der sächsischen Macht im 10. Jahrhundert,” in Archäologie des Mittelalters und Bauforschung im Hanseraum: Eine Festschrift für Günter P. Fehring (Rostock, 1993), 113–18; Gösta Ditmar-Trauth, “Die Ausgrabung an der Großen Klosterstraße in Magdeburg (Gerberei, Uferbefestigung, Stadtmauer),” Jahresschrift für mitteldeutsche Vorgeschichte 86 (2003), 213–71; and Brigitte Kunz, “Die Befestigungen des Steilufers der Pfalzen Magdeburg und Duisburg im 10. Jahrhundert,” in Finden und Verstehen: Festschrift für thomas Weber zum sechzigsten Geburtstag, ed. Hans-Jürgen Beier et al. (LangenWeißbach, 2012), 299–302. The Slavic fortifications at all of these sites were replaced by German fortifications early in the tenth century. See Corpus archäologischer Quellen zur Frühgeschichte auf dem Gebiet der Deutschen Demokratischen Republik (7. bis 12. Jahrhundert), ed. Joachim Hermann and Peter Donat (Berlin, 1973), 363, 371, 364, 366, 357, and 362, respectively. 31 Regarding the fortifications at Tangermünde and Osterburg, see Corpus archäologischer Quellen, 175 and 189. For German control of Walsleben see Widukind, Res gestae, 1.36.

334

H E N RY I O F G E R M A N Y ’ S 9 2 9 M I L I TA RY C A M PA I G N

of the Havel river with the Elbe by the two fortifications of Havelberg and Klietz.32 Ultimately, a further fortification was constructed just 3 kilometers to the west of Klietz at Arneburg, which would become a major bulwark of German rule along this portion of the lower Elbe through the remainder of the tenth and early eleventh centuries.33 These Ottonian fortifications marked the re-establishment under Henry I of the defensive line along the course of the lower Elbe that Charlemagne had organized during the final decade of the eighth and the first decade of the ninth century.34 In addition to debriefing merchants and other travelers, as well as obtaining information from earlier military operations, it is very likely that Henry followed the well-established precedent of sending out scouts (exploratores) to obtain fresh intelligence about the state of the roads, defensible camping sites, sources of water, and locations and strengths of enemy fortifications.35 In this context, the advice of Vegetius, whose handbook on military affairs circulated widely in both the east Frankish and German kingdoms, is crucial.36 In discussing the proper preparations 32 Concerning the large fortification at Klietz that was controlled by Henry I, see Hubert Reimer, “Der slawische Burgwall von Klietz, Kreis Havelberg – ein Vorbericht,” Jahresschrift für mitteldeutsche Vorgeschichte 75 (1992), 325–45, particularly 343 for the establishment of a German garrison. With respect to the Slavic fortification at Havelberg, see Christian Popp, “Gründung und Frühzeit des Bistums Havelberg,” Mitteilungen des Vereins für Geschichte der Prignitz 3 (2003), 6–82, here 18. 33 For the important role played by the fortress at Arneburg in securing this portion of the frontier, see Thietmar, Chronicon, 4.38 and 6.28. Also see the discussion of Arneburg by David S. Bachrach, “The Military Organization of Ottonian Germany, c. 900–1018: The Views of Bishop Thietmar of Merseburg,” Journal of Military History 72 (2008), 1061–88, here 1072–3. 34 See the detailed discussion of both the Carolingian and Ottonian frontiers along the lower Elbe in Matthias Hardt, “Linien und Säume. Zonen und Räume an der Ostgrenze des Reiches im frühen und hohen Mittelalter,” in Grenze und Differenz im frühen Mittelalter (Vienna, 2000), 39–56; and idem, “Hesse, Elbe, Saale and the Frontiers of the Carolingian Empire,” in The Transformation of Frontiers from Late Antiquity to the Carolingians, ed. Walter Pohl, Ian Wood, and Helmut Reimitz (Leiden, Boston and Cologne, 2001), 219–32, particularly 223–6. 35 Regarding the deployment of scouts by Ottonian military commanders, see Ekkehard IV, Casus Sancti Galli, ed. Hans F. Haefele (Darmstadt, 1980), 120–2; Richer, Histoire de France (888–995), vol. 2 954–995, ed. Robert Latouche (Paris, 1937), 16; Vita Sancti Bernwardi episcopi Hildesheimensis auctore Thangmaro, ed. Walter Berschin and Angelika Häse (Heidelberg, 1993), 328; Adalbold, Vita Heinrici II, MG SS 4 (Hanover, 1841), ch. 35; and Alpert of Metz, De Diversitate Temporum, MGH SS, 700–723, bk.2 ch. 10. Now see the English translation of this text by David S. Bachrach, Warfare and Politics in Medieval Germany, c. 1000: On the Variety of Our Times by Alpert of Metz (Toronto, 2012). 36 Regarding the large number of Vegetius manuscripts in circulation during the early Middle Ages, see Michael D. Reeve, “The Transmission of Vegetius’ Epitoma Rei Militaris,” Aevum 74 (2000), 243–354; and now Christopher Allmand, De Re Militari of Vegetius: The Reception, Transmission and Legacy of a Roman Text in the Middle Ages (Cambridge, 2011). In the mid-ninth century, Rabanus Maurus, who served both as abbot of Fulda and archbishop of Mainz, in both of which capacities he had extensive military responsibilities, oversaw the composition of a revised version of Vegetius’s handbook, which would only deal with those matters that were of value ‘tempore moderno’. See Rabanus Maurus, De procinctu Romanae militiae, ed. Ernst Dümmler in Zeitschrift für deutsches Alterthum 15 (1872), 443–51, here 450. Louis the German also had a copy of

335

M AT E R I A L S O U R C E S A N D I M A G E S

to make before embarking on campaign, Vegetius emphasized regarding the commander that: First, he should have itineraries of all regions in which war is being waged written out in the fullest detail, so that he may learn the distances between places in terms of the miles and quality of roads, and examine short-cuts, by-ways, mountains and rivers, accurately described.37 Consequently, before setting out, it is very likely that Henry was well informed that the main seat of the Hevelli princes at Brandenburg was a double fortress, built on an island in the Havel river with walls measuring approximately 870 meters.38 These walls were protected by a ring of ditches. Not least, the fortification as a whole also was defended by the waters of the Havel.39 This princely seat derived substantial further protection from a belt of fortifications that were located within a day’s march to the west and northwest. Ten kilometers due west of Brandenburg was the stronghold at Plau, built along the shore of the Plauersee.40 Twenty kilometers west of Plau was the Slavic fort at Altenplathow, which was constructed on an island located in a branch of the Elbe.41 Twenty-six kilometers northwest of Brandenburg was the fortification at Milow.42 Fourteen kilometers

37

38 39 40 41 42

Vegetius’s handbook. See Eric J. Goldberg, Struggle for Empire: Kingship and Conflict under Louis the German, 817–876 (Ithaca, 2006), 40–2. Of the eleven surviving manuscripts of the Epitoma rei militaris from the tenth century, six are from the Ottonian kingdom, including Lotharingia. This includes manuscripts composed at Freising, St. Gall, Einsiedeln and Ghent. See Philippe Richardot, Végèce et la culture militaire au moyen âge (Ve – Xv siècle) (Paris, 1998), appendix 1. Cf. Guy Halsall, Warfare and Society in the Barbarian West, 450–900 (London, 2003), 146; and Timothy Reuter, “Carolingian and Ottonian Warfare,” in Medieval Warfare: A History, ed. Maurice Keen (Oxford, 1999), 13–35, here 19–21, who suggest that important manuals for training future officers, such as Vegetius’s Epitoma rei militaris and Frontinus’s Strategemata were of merely antiquarian interest to monks, and had no practical value in either preparation for, or the conduct of military affairs. However, it should be emphasized that neither Reuter nor Halsall attempted to explain either the large number of extant manuscripts of Vegetius from the early medieval period, nor the frequent reference to Vegetius’s text by members of both the Carolingian and Ottonian royal courts. See the discussion of this topic by Bachrach, Warfare in Tenth–Century Germany, particularly 102–34. Vegetius, Epitoma rei militaris, ed. Alf Önnerfors (Stuttgart, 1995), 3.6. I am quoting the translation of this passage by N. P. Milner, Vegetius: Epitome of Military Science, 2nd revised edition (Liverpool, 1996), 73. Klause Grebe, “Die Ergebnisse der Grabung Brandenburg,” in Berichte über den II. Internationalen Kongreß für slavische Archäologie, vol. 3 (Berlin, 1973), 269–78, here 270–4. Ibid., 270–4. Corpus archäologischer Quellen, 116. Johannes Schneider, “Ausgrabungen auf dem Burwall Genthin-Altenplathow 1976–1977,” Ausgrabungen und Funde 25 (1980), 209–12, here 210. Wolfgang Bünnig, “Ein spätslawischer Abschnittswall von Milow, Kr. Rathenow,” Ausgrabungen unde Funde 22 (1977), 93–6. The fortification at Milow, which dates back to the eighth century, was built on a stretch of land that extended into the Havel River. It was a typical Abschnittswall fortified on the east side and protected on the other three sides by the waters of the Havel. The

336

H E N RY I O F G E R M A N Y ’ S 9 2 9 M I L I TA RY C A M PA I G N

farther north and west of Milow was the fortification at Hohennauen, which was constructed on the northwest end of the Hohennauenersee on a spur of land that projected into the lake.43 As a group, Brandenburg and the strongholds to its north and west created a salient between Magdeburg to the southwest and HavelbergKlietz to the northwest, thereby posing a potential source of danger to both regions of the Frankish frontier. Along the southwestern and southern approaches to Brandenburg, the Hevelli princely seat derived further protection from a series of fortifications located in the Fläming hills. Tucheim was constructed on a northern extension of the Fläming hills directly on the edge of the marshlands formed by the Fiener Bruch 35 kilometers southwest of Brandenburg.44 Thirty kilometers due south of Brandenburg was Belzig, which was the most important fortification in the western Fläming hills. This fort, which was built on a high plateau over the Belziger Bach, had a circuit of walls measuring 600 meters, making it almost as large as Brandenburg itself.45 Two other strongholds in the Fläming hills at Niemegk and Mörz commanded the upper reaches of the Plane river, which flowed into the Havel just west of Brandenburg.46 To the east, the Hevelli prince ruled both banks of the Havel up to the major fortress at Spandau located 80 kilometers from Brandenburg on an island in the river. Spandau was protected by a massive earth and timber wall that was further reinforced by a stone front.47 Its construction probably dates to the decade just before Henry I’s campaign against Brandenburg.48 These numerous Hevelli strongholds provided a system of defense in depth for Brandenburg that Henry I’s army would have to neutralize before a close siege of the island fortress could be undertaken. The garrisons at Plau, Milow, and Altenplathow posed a threat to Henry’s line of communications between

43 44 45

46

47 48

west-facing wall protected an area with a diameter of 80 meters. The wall was 17 meters wide, protected by a 15-meter wide ditch. Wolfgang Bünnig, “Ausgrabungen auf einem slawischen Fundplatz in Hohennauen, Kr. Rathenow,” Ausgrabungen und Funde 32 (1987), 82–9. Corpus archäologischer Quellen, 362. Joachim Herrmann, “Belizi 997-Beltz-Belzig: Von der Slawenburg zur kursächsischen Festung zwischen Havelland und Fläming. Eine archäologisch-historische Topographie,” Veröffentlichungen des brandenburgischen Landesmuseums für Ur-und Frühgeschichte 28 (1994), 191–221, here 191–7. Hermann, “Belizi,” 196. The fortification at Mörz has been partially excavated. It had an outer defensive perimeter of 400 meters that enclosed a settled area of 2,800 square meters. The Slavic fortification was originally built in the ninth century and gave way to a German garrison in the context of Henry’s campaign against Brandenburg. See the discussion by Joachim Herrmann, “Der slawische und frühdeutsche Burgwall bei Mörz, Kr. Belzig, und seine Probleme,” Ausgrabungen und Funde 4 (1959), 286–93. Adriaan von Müller, “Der Burgwall von Berlin Spandau,” in Europas Mitte um 1000, ed. Alfried Wieczorek and Hans Hinz (Stuttgart, 2000), 278–81, here 279. Joachim Henning, “Der slawische Siedlungsraum und die ottonische Expansion östlich der Elbe: Ereignisgeschichte–Archäologie-Dendrochronlogie,” in Europa im 10. Jahrhundert: Archäologie einer Aufbruchszeit, ed. Joachim Henning (Mainz, 2002), 131–46, here 132. Henning (132) argues that the fortifications at Brandenburg also date to approximately 919.

337

M AT E R I A L S O U R C E S A N D I M A G E S

Magdeburg and Brandenburg from the north. From the south, the garrisons at Belzig, Niemegk, and Mörz were poised to interrupt Henry’s siege operations at Brandenburg by moving down the Plane river valley. In addition, the Havel river provided an important route for aid from other Hevelli forces to approach Brandenburg from the east. In his account of the events of 929, which almost certainly drew upon information provided by members of the royal court with direct experience of the campaign, Widukind makes clear that Henry’s campaign against the Hevelli involved several subsidiary military operations before the urbs at Brandenburg was captured. He stressed that the German forces wore out the Slavs in numerous battles (multis preliis fatigans) over a lengthy period. Whether these praelia involved any battles in the field must remain an open question. It is rather more likely that they comprised raiding operations that were bent on the destruction of infrastructure and weakening of Hevelli defenses. In the course of these operations, it is likely that Henry worked diligently to capture many, if not all, of the fortifications in the immediate vicinity of Brandenburg, including Altenplathow, Milow, Plau, and Tucheim. With garrisons along the line Burg, Grabow, Dretzel, Parchem, Tucheim, Altenplathow, Milow, and Plau, Henry could ensure that his operations at Brandenburg would not be disrupted by an intervention from allies of the Hevelli to the northeast of the Havel.49 Among these allies were the Redarii, another people in the Weleti confederation. The danger posed by the Redarii was demonstrated dramatically by their capture in the summer of 929 of the German fortification at Walsleben, which was located 20 kilometers west-southwest of Havelberg and served an important role in protecting the frontier.50 In undertaking this substantial operation, which resulted in the massacre of the entire garrison as well as the local population that had taken refuge at Walsleben, the Redarii were acting in vengeance for Henry’s successful campaign the previous winter against the Hevelli. In addition to protecting his forces against intervention from the northeast, the German king required a blocking force along the lower course of the Plane river south of Brandenburg to inhibit efforts by Hevelli garrisons located in the Fläming hills to come to the aid of their prince. A third blocking force was required east of Brandenburg on the Havel to interdict any relief force that might attempt to come down the river. Excavations of the fortress at Spandau, noted earlier, show that it was completely destroyed by a catastrophic fire in the first half of the tenth century, before being rebuilt, likely during Henry I’s reign, i.e. before 936. It is

49 Excavations at Parchem show that a German fort was established on the site of an open Slavic settlement dating from the tenth century. See Coropus archäologischer Quellen, 360. For the line of march from Magdeburg to Brandenburg, see Werner Radig, “König Heinrich I. und die ostdeutsche Archäologie,” Mannus 8 (1931), 60–70; and subsequently accepted by Werner Hülle, Westausbreitung und Wehranlagen der Slawen in Mitteldeutschland (Leipzig, 1940), 41. 50 The main seat of the Redarii was located at Riedegost near Strelitz, approximately 130 kilometers northeast of Brandenburg. See the discussion in Res gestae Saxonicae, in Quellen zur Geschichte der sächsischen Kaiserzeit, ed. and trans. Reinhold Rau, vol. 8 (Darmstadt, 1971), 70, n. 90.

338

H E N RY I O F G E R M A N Y ’ S 9 2 9 M I L I TA RY C A M PA I G N

possible, therefore, that Henry’s blocking force on the Havel actually consisted of a substantial contingent that moved 80 kilometers upstream, requiring a three-day march, which then captured and destroyed the Hevelli fortress at Spandau.51 After isolating Brandenburg completely, Henry’s army undertook a close siege of the fortress. Widukind draws attention to the fact that by the time Henry had prepared the way for his final attack on Brandenburg through a series of military operations, the coldest part of the winter had arrived. As a consequence, the army was forced to camp on the ice (castris super glaciem positis).52 The final phase of the operation against Brandenburg, which culminated in Henry’s army storming the fortress, surely resulted in heavy losses to the German forces. The contemporary Burghal Hidage, which was developed by the royal court of Wessex to provide a system of defense in depth that utilized a wide range of fortifications, established a ratio of one defender for approximately 1.3 meters of wall.53 In a world in which technology was pretty much constant across the board, i.e. the same weapons were used in Anglo-Saxon England, by the Vikings, by the Ottonians, and in the Slavic lands, it is likely that the rulers of the Hevelli came to the same general conclusions regarding the minimum number of men required to defend a given length of wall. In this context, 700 men, at a minimum, were needed to defend the 870-meter circuit of the walls at Brandenburg.54 In light of the ongoing campaign against Hevelli assets in the weeks prior to the final attack on Brandenburg, however, it is almost certainly the case that far more than the minimum number of defenders manned the walls of the island fortress. In order to capture Brandenburg by storm, Henry required an advantage of at least four to one over the defenders, who were armed with missile weapons, including both self-bows and spears. This estimate is based on a detailed statistical analysis utilizing probability theory regarding the likely number of arrows

51 For the suggestion that Spandau was burned down in the context of Henry I’s capture of Brandenburg, see Müller, “Der Burgwall von Berlin Spandau,” 279. 52 Widukind, Res gestae, 1.35. In describing the siege and capture of Brandenburg, Widukind alludes to Cicero’s speech against Lucius Calpurnius Piso. Cicero, In Pisonem, 17 reads, “exercitus nostri interitus ferro, fame, frigore, pestilentia.” Here, Cicero was listing the causes for the casualties suffered by Piso’s troops when he served as governor of Macedonia. Widukind paraphrases, listing only hunger, arms (literally iron), and cold as the causes for the casualties suffered by the defenders of Brandenburg. Widukind, Res gestae, 1.35, “cepit urbem quae dicitur Brennaburg fame ferro frigore.” 53 For a useful introduction to many of the questions regarding the origin, size and organization of the defensive scheme described in the Burghal Hidage, see the collection of essays in The Defence of Wessex: The Burghal Hidage and Anglo-Saxon Fortifications, ed. David Hill and Alexander R. Rumble (Manchester, 1996). 54 For a detailed discussion of the requirements set out in the Burghal Hidage for the defence of all types of fortifications, and the general applicability of these requirements to other regions of Europe, see Bernard S. Bachrach and Rutherford Aris, “Military Technology and Garrison Organization: Some Observations on Anglo-Saxon Military Thinking in Light of the Burghal Hidage,” Technology and Culture 31 (1990), 1–17; and reprinted in Bernard S. Bachrach, Warfare and Military Organization in Pre-Crusade Europe (London, 2002), with the same pagination, here 1–4.

339

M AT E R I A L S O U R C E S A N D I M A G E S

shot from tenth-century bows that would land in a given area over a given period. Crucially, this analysis assumes a best-case scenario for attackers, i.e. that they utilized the best possible strategy for minimizing casualties and thus required the least possible numerical advantage. If, in fact, attackers did not utilize the best possible methods for minimizing casualties, then a higher than four-to-one ratio would be needed to capture a well-defended stronghold, such as Brandenburg.55 Consequently, a conservative estimate utilizing the four-to-one figure puts Henry’s assaulting force in the range of 3,000 to 3,500 men. This was not, however, the entirety of Henry’s army, merely the minimum number of troops who had to participate in the actual attack on Brandenburg’s walls with scaling ladders and perhaps other siege equipment. Troops attacking over level ground and covering approximately 5 feet per second while carrying scaling ladders could expect to suffer 25% to 30% casualties before ever reaching the base of the wall.56 If Henry dispatched the minimum number of 3,000 men to assault the walls, it is likely that his forces suffered as many as 1,000 casualties, including dead and wounded, before their scaling ladders could be put into place.57 This number of casualties is consistent with the observation by Thietmar of Merseburg that the Polish duke, Boleslav Chrobry (992–1025), lost 500 men in his assault on the fortress of Lebusa (1012), which was defended by 1,000 men.58 Henry’s forces may have been somewhat able to 55 Regarding the need for attacking forces that were much larger than those of the defenders, see Bachrach and Aris, “Military Technology and Garrison Organization,” 5–10. Numerous excavations of Slavic sites have identified bows and arrows. See, for example, Corpus archäologischer Quellen, 138; Michael Müller-Wille, “Krieger und Reiter im Spiegel früh und hochmittelalterliche Funde Schleswig-Holsteins,” Offa: Berichte und Mitteilungen zur Urgeschichte, Frühgeschichte und Mittelalterarchäologie 34 (1977), 40–74, here 44; and Volker Schmidt, Drense: Eine Hauptburg der Ukrane (Berlin, 1989), 49–50. With regard to the diffusion of Western spears, particularly the so-called winged lance, into the Slavic world, and the continued use of the weapon into the eleventh century, see Janusz Górecki, “Waffen und Reiterausrüstungen von Ostrów Lednicki: Zur Geschichte des frühen polnischen Staates und seines Heeres,” Zeitschrift für Archäologie des Mittelalters 29 (2001), 41–86, here 46–53; and Tomasz Kurasinski, “Waffen im Zeichenkreis. Über die in den Gräbern auf den Gebieten des frühmittelalterlichen Polen vorgefundenen Flügellanzenspitzen,” Sprawozdania Archeologiczne 57 (2005), 165–96, here 171–4. 56 Regarding the speed of attack and casualty rates, see Bachrach and Aris, “Military Technology and Garrison Organization,” 7. 57 For the use of both arrows and spears by Slavic troops, see Corpus archäologischer Quellen, 138 for the excavations at Schwaan near Burg Werle on the Warnow River. 58 Thietmar, Chronicon, 6.80. Thietmar was careful in using specific numbers so that his observations here likely represent the correct order of magnitude for the number of men killed in the attack. Regarding Thietmar’s trustworthiness with regard to using numbers, see Bachrach, “The Military Organization of Ottonian Germany,” 1083–4. The book of the dead (necrology) from Merseburg includes a reference to the numerous defenders who were killed at Lebusa. Die Totenbücher von Merseburg, Magdeburg und Lüneburg, ed. G. Althoff and J. Wollasch, MGH Libri Memoriales et Necrologia, n.s. 2 (Hanover, 1983), fol. 5r, p. 11, records that many fighting men were annihilated (multi peremientes [sic]) at Lebusa. The fortress at Lebusa was probably the Löbsal Burgberg, which has been excavated, in part, by Ralf Gebuhr, “Liubusua-Löbsal. Ein bemerkenswerter Ort früher Geschichte der Mark Meißen

340

H E N RY I O F G E R M A N Y ’ S 9 2 9 M I L I TA RY C A M PA I G N

reduce the number of casualties they suffered through sustained suppressing fire on the part of their own archers against the defenders. However, this archery support would have to cease once the assaulting force neared the walls, in order to avoid losses from “friendly fire.” In describing the successful siege of Brandenburg, Widukind does not belabor the casualties suffered by Henry’s forces, noting only that once the fortress fell, the German king acquired the entire territory (regio) that had been subject to the Hevelli prince.59 In order to secure his hard-won gains at Brandenburg and the entire Hevelli region, however, Henry had to leave a substantial garrison at Brandenburg itself and also at several other fortifications.60 These men served the double function of beginning the integration of Haveland into the German kingdom and providing a secure base of operations for the next stage of the campaign against the Daleminzi. Given the circuit of the walls of the fortress at Brandenburg, it is likely that the German force left there numbered at least 700 men, that is enough men to protect the walls against an effort to recapture the fortress. Henry probably also left smaller garrisons at fortifications along the route from Magdeburg to Brandenburg, including Burg, Grabow, Dretzel, Parchen, Tucheim, Altenplathow, Milow, and Plau. In light of the size of these fortifications, these garrisons, in aggregate, including the force at Brandenburg, cannot have numbered many fewer than 1,500 men.61

Phase two: the capture of Gana Once he had secured his base of operations at Brandenburg, Henry marshaled his forces to begin the second phase of the campaign, which was directed toward the important Daleminzi center at the “fortress called Gana.”62 Scholars long have

59

60

61

62

lag im Großenhainer Land?,” Heimatkalender für die Grossenhainer Pflege: Grossenhainer Stadtund Landkalendar 11 (2007), 50–60; idem, Jarina und Liubusua: Kulturhistorische Studien zur Archäologie frühgeschichtlicher Burgen im Elbe-Elster-Raum (Bonn, 2007), 145–72. Widukind, Res gestae, 1.35: “Cumque illa urbe potitus omnem regionem.” It is not entirely clear how long this occupation of the Hevelli lands lasted. Widukind, Res gestae, 2.21, himself, indicates that the Hevelli lands were not under Ottonian control when Otto I came to power in 936. Regarding the need for substantial garrisons in each of the numerous fortifications held by the Ottonian rulers east of the Elbe and Saale river systems see Herrmann, “Belizi,” 195–6 and 203. Thietmar, Chronicon, 4.38, 6.2, 7.23 and 7.25, provides details regarding the use of milites provided by secular and ecclesiastical magnates to serve as garrison troops in the frontier zones. See the discussion of these rotating garrison forces by Bachrach, “The Military Organization of Ottonian Germany,” 1070 and 1073. The fortification at Milow had a circuit of 250 meters. See Wolfgang Bünnig, “Ein spätslawischer Abschnittswall von Milow, Kr. Rathenow,” Ausgrabungen unde Funde 22 (1977), 93–6, here 94. The circuits of the other seven fortifications have not yet been identified. However, if they were roughly the same size as Milow, the total aggregate length of the walls of these eight strongholds was 2,000 meters, which would require approximately 1,600 men to serve as garrison troops. If the seven other fortifications were only half the size of Milow, then 800 men would be required. Widukind, Res gestae, 1.35: “signa vertit contra Dalamantiam, adversus quam iam olim reliquit ei pater militiam; et obsidens urbem quae dicitur Gana.”

341

M AT E R I A L S O U R C E S A N D I M A G E S

disputed the exact site of this fortress, located west of the Elbe, in the valley of the Jahna river after which the site took its name. Werner Radig suggested two locations for Gana, namely Burgberg near Zschaitz, and the Zöthener Schanze south of Lommatzsch.63 Werner Coblenz rejected these and suggested instead a location in the valley of the Jahna river between Hof and Stauchitz.64 The Hof/Stauchitz fortification is located 11 kilometers southwest of the modern Elbe river town of Riesa. The fortification sits in the valley of the small Jahna river, which rises at Döbeln and enters the Elbe at Riesa. Recent excavations at the Hof/Stauchitz location have identified a massive complex that may be the site of the fortress captured by Henry.65 Even if this fort was not Gana, however, the Hof/Stauchitz complex was contemporary to Gana and provides a useful comparative model for the effort that was required to capture the Daleminzi central fortress. The Hof/Stauchitz fortress was protected on three sides by swampy lowlands created by the Jahna River, although these likely would freeze in a hard winter such as the one experienced by this region in 928–929. The one dry approach to the fortification was from the south.66 During the final stage of its construction, the walls of the main fortress were approximately 15 meters wide at the base and 6 meters high at its crown. The wall itself was protected by a 15-meter-wide and 5-meter-deep ditch.67 The earth and timber wall also was provided with a dry stone wall front.68 The circuit of walls of this fortification measured 700 meters and enclosed an area of 4 hectares, which does not include a postulated outer fortification (Vorburg).69 The heart of the fortress was therefore approximately the same size as the entire complex at Brandenburg. To its north, Gana was protected by a chain of Slavic fortifications that were constructed along the course of the Elbe. These included Zahna,70 Auberg,71 Torgau,72 63 Werner Radig, Der Burgberg Meißen und der Slawengau Daleminzien (Augsburg, 1929), 48–9. 64 Werner Coblenz, “Archäologische Betrachtungen zur Gana-Frage im Rahmen der älterslawischen Besiedlung des Gaues Daleminzien,” in Beiträge zur Archivenwissenschaft und Geschichtsforschung, ed. R. Gross and M. Kobuch (Weimar, 1977), 354–70, here 358. For an overview of the historiographical discussion regarding the location of Gana, see Judith Oexle and Michael Strobel, “Auf den Spuren der ‘urbs, quae dicitur Gana’, der Hauptburg der Daleminizier. Erste archäologischer Untersuchungen in der slawischen Befestigung von Hof/Stauchitz,” Arbeits und Forschungsberichte zur sächsischen Bodendenkmalpflege 46 (2004), 253–63, here 253. 65 Oexle and Strobel, “Auf den Spuren,” 253–63. 66 Ibid., 255. 67 Ibid., 260. 68 Ibid., 261. 69 Ibid., 263. I thank Dr. Michael Strobel for the information concerning the length of the walls of Hof/Stauchitz, which he provided in a private communication. 70 Berthold Schmidt and Waldemar Nitzschke, “Untersuchungen in slawischen Burgen zwischen Saale und Fläming,” Ausgrabungen und Funde 20 (1975), 43–51, here 45. 71 Gerhard Billig, “Mittelalterliche Burgen in Dommitzsch nördlich Torgau,” in Landesgeschichte und Archivwesen: Festschrift für Reiner Gross zum 65. Geburtstag (Dresden, 2002), 21–34, here 26. 72 Heinz-Joachim Vogt, “Archäologische Untersuchungen im Altstadtbereich von Torgau,” Ausgrabungen und Funde 37 (1992), 46–53, here 47.

342

H E N RY I O F G E R M A N Y ’ S 9 2 9 M I L I TA RY C A M PA I G N

Fichtenberg near Mühlberg,73 and Görzig.74 To its west, Gana was guarded by fortifications at Zschaitzer Burgberg or Hof/Stauchitz or Zothain, depending on the true identify of Gana, as well as Döbeln, Mügeln, Hwoznie/Ziegra, Mochau, and Schreibitz.75 Both the size and the location of the fortress at Stauchitz were well known to Henry from his own campaigns against the Daleminzi, which he had led over the previous two decades.76 So, too, were the large number of Slavic fortifications that protected the Daleminzi central fortress from attacks originating to its north and west. From his stronghold at Püchen, noted earlier, and Leipzig, which will be discussed in greater detail later, Henry’s scouts had ample opportunity to undertake reconnaissance operations to the south and east.77 In conceptualizing Henry’s route to Gana from Brandenburg, Werner Radig, noted previously, and Werner Hülle envisioned a single column marching west around the Fläming hills and crossing the Elbe in the neighborhood of Dessau. From here, they postulated a march down the west bank of the Mulde until a point south of Püchen.78 After leaving the vicinity of Püchen, Henry’s army moved, according to Radig and Hülle, up the right bank of the Mulde until reaching the Jahna river valley and Gana.79 This route is implausible for two reasons. First, if Henry did not move up the Plane river into the Fläming hills after his capture of Brandenburg, the garrisons at Belzig, Niemegk, and Mörz, as well as the dense surrounding population, would 73 Günther Wetzel, “Der erste slawische Burgwall des Kreises Bad Liebenwerda in Fichtenberg bei Mühlberg,” Ausgrabungen und Funde 22 (1977), 76–85, here 85–6. 74 Werner Coblenz, “Boleslav Chrobry in Sachsen und die archäologische Quellen,” Slavia Antiqua 10 (1963), 249–85, here 274. 75 See the discussion of these fortifications, all of which were converted to German use, in Gerhard Billig, Die Burgwardorganisation im obersächsisch-meissnischen Raum: Archäologisch-archivalisch vergleichende Untersuchungen (Berlin, 1989), 64. 76 Widukind, Res gestae, 1.17; and Thietmar, Chronicon, 1.15. Regarding the location of the fort at Püchau, held by Henry against the Hungarians, on the Mulde south of Eilenburg, see D. Vöttger and Franz Winter, “Die Diöcese Magdeburg,” Geschichtsblätter für Stadt und Land Magdeburg 3 (1868), 162–82, here 179; and Hülle, Westausbreitung und Wehranlagen, 41. 77 The fortification at Leipzig has received considerable attention since the Second World War. Herbert Küas, who undertook the initial excavations of Leipzig, saw this fortification as a bridgehead (Germ. Brückenkopf) from Merseburg into the valley of the White Elster. See Herbert Küas, “Wehrtürme und Wohntürme auf ausgegrabenen deutschen Burgen zu Leipzig, Meißen und Groitzch,” Sächsische Heimatblätter 19 (1973), 145–55, here 146. Küas’s dating of the first phase of fortification to the early tenth century has been confirmed in subsequent excavations. See Friedemann Winkler, Leipzigs Anfänge: Bekanntes, Neues, offene Fragen (Beucha, 1998), 33. Recent investigations have revised downward the size of the fortification at Leipzig in the earlier period, so that it is now seen to have been a typical Abschnittsbefestigung in the early tenth century that was constructed of earth and timber, rather than the massive stone fortress imagined by Herbert Küas. For the most recent view, see Thomas Westphalen, “Die frühen Burgen Leipizgs,” in Archäeologie und Architektur: Das frühe Leipzig, ed. Wolfgang Hocquél (Leipzig, 2003), 43–50. 78 Hülle, Westausbreitung und Wehranlagen, 41, and the discussion of Henry’s defense of this fortification against the Hungarians by Thietmar, Chronicon, 1.15. 79 Hülle, Westausbreitung und Wehranlagen, 41.

343

M AT E R I A L S O U R C E S A N D I M A G E S

pose a serious threat to his line of communications back to Brandenburg and also west to Magdeburg.80 Second, as will be discussed, Henry’s decision to establish a major fortification at Meissen, which was located on the Elbe 25 kilometers southeast of Gana, required that he control the course of the Elbe river to Magdeburg. Otherwise, the garrison at Meissen would be completely isolated by the series of Slavic fortifications, noted earlier, constructed along the middle course of this river. Excavations at Auberg,81 Torgau,82 Fichtenberg,83 and Görzig84 make it clear that all four of these Slavic fortifications were rapidly replaced by German fortifications that were constructed in the immediate vicinity of the original Slavic sites. Görzig, for example, was replaced in 929 by the German fortification at Strehla, which protected a ford over the Elbe.85 Similarly, Auberg was replaced by Osterburg.86 Another German fortification built at this time was Dahlen, located 44 kilometers east of Leipzig and 22 kilometers south of Strehla.87 All three of these fortifications – Leipzig, Dahlen, and Strehla – sat astride the major east–west highway, later called the Hohenstrasse, which ultimately led eastward from the major economic and military center of Merseburg on the Saale to Silesia.88 A German column under Henry’s direct command, therefore, likely marched up the valley of the Plane into the Fläming hills, where the fortifications at Belzig, Niemegk, Mörz, and Zahna were neutralized, and likely received German garrisons. The last named of these fortifications has benefitted from excavations, and it is clear that Zahna was protected by a substantial earth and timber circuit of approximately 300 meters, with two defensive ditches located 25 and 50 meters, respectively, from the face of the wall.89 Widukind does not record whether Zahna was defended, but Henry’s army likely was able to capture, or at least neutralize, the fortification and continue its march, with a probable crossing of the Elbe in the vicinity of Pratau. An Ottonian fortification was quickly established at this latter

80 Regarding the dense Slavic population in the Fläming hills, see Thomas Kersting, “Die Burg von Luckenwalde am Niederen Fläming, Brandenburg: Bodendenkmalpflege und Landesgeschichte,” in Cum grano salis: Beiträge zur europäischen Vor und Frühgeschichte: Festschrift für Volker Bierbrauer zum 65. Geburtstag (Friedberg, 2005), 331–8, here 332. 81 Billig, “Mittelalterliche Burgen in Dommitzsch nördlich Torgau,” 26. 82 Grimm, Handbuch, 113; and Heinz-Joachim Vogt, “Archäologische Untersuchungen im Altstadtbereich von Torgau,” Ausgrabungen und Funde 37 (1992), 46–53. 83 Wetzel, “Der erste slawische Burgwall des Kreises Bad Liebenwerda,” 75. 84 Werner Coblenz, “Boleslav Chrobry in Sachsen und die archäologische Quellen,” Slavia Antiqua 10 (1963), 249–85, here 274. 85 Billig, Die Burgwardorganisation, 68; and Johannes Herrmann, “Lorenzkirche, Markt des Burgwards Strehla im Daleminzergau der Mark Meißen,” Herbergen der Christenheit: Jahrbuch für deutsche Kirchengeschichte 19 (1993), 17–27, here 18. 86 Billig, “Mittelalterliche Burgen in Dommitzsch nördlich Torgau,” 23. 87 Herrmann, “Lorenzkirche, Markt des Burgwards Strehla,” 18. 88 Ibid., 19; and the earlier discussion of this route by Billig, Die Burgwardorganisation, 68. 89 Berthold Schmidt and Waldemar Nitzschke, “Untersuchungen in slawischen Burgen zwischen Saale und Fläming,” Ausgrabungen und Funde 20 (1975), 43–51, here 45.

344

H E N RY I O F G E R M A N Y ’ S 9 2 9 M I L I TA RY C A M PA I G N

site to protect an important ford over the Elbe.90 From Pratau, Henry’s army probably marched along the course of the Elbe, capturing or receiving the surrender of Slavic fortifications until reaching the vicinity of the stronghold at Hof/Stauchitz. As Henry’s column was marching south toward Gana, it seems likely that a second column was marching east to join with the king in a pincer movement. The deployment of multiple columns that were to join up in the vicinity of the enemy had been a hallmark of Carolingian warfare throughout the ninth century, and it is hardly likely that this was forgotten.91 The deployment of multiple columns in a pincer movement also was characteristic of Henry I’s successors over the course of the tenth and early eleventh centuries.92 There are two important reasons why Henry would mobilize a second column for the phase of his campaign directed at Gana. First, the troops moving east from the major Ottonian strongholds along the Saale, including both Halle and Merseburg, could provide tactical support to Henry’s army as it made its potentially hazardous advance through hostile territory. Any Slavic force that attempted to mobilize in the valleys of the Mulde or Elbe in order to interfere with Henry I’s advance along the latter river would face attack from both the west and the east. Second, when the two columns converged at Gana, the German garrisons deployed at Leipzig, Püchen, Grimma, and Oschatz ensured a secure route for the transport of supplies from Merseburg, Halle, Magdeburg, and even farther west from the major royal estates and monastic complexes in the Harz.93 These supplies, which moved along the Hohenstrasse, would complement supplies transported up the Elbe directly from Magdeburg and then up the Jahna to Gana. The need for a secure supply route to Gana likely recommended that the second column be mobilized to begin its march at Merseburg, which so frequently served as a base of operations in the next century for King Henry II.94 From Merseburg, it was 30 kilometers east to Leipzig. It was a further 25 kilometers southeast from Leipzig to Grimma on the Mulde. From Grimma, it was 32 kilometers to Oschatz, 90 Paul Grimm, Tilleda: Eine Königspfalz am Kyffhäuser, Teil I: Die Hauptburg (Berlin, 1968), 112. 91 Regarding the regular deployment of multiple columns that would converge on a single target by Carolingian rulers from Charlemagne to Louis the Child, see Bowlus, Franks, 58, 88, 177–8, 226–7, 264; Bernard S. Bachrach, Early Carolingian Warfare: Prelude to Empire (Philadelphia, 2001), 42 and 193; and Goldberg, Struggle for Empire, 138–9. 92 For the repeated use multiple columns that were intended to converge on the enemy during the reign of Henry II (1002–1024), see Bruno Scherff, Studien zum Heer der Ottonen und der ersten Salier (919–1056) (Bonn, 1985), 99. The deployment of multiple columns by Henry I’s son Otto I (936–73) is also attested in the written sources. See, for example, Flodoard, Annales, an. 948; and Adalbert, Continuatio, an. 953. 93 Regarding the extensive royal estates in the Harz region under the Ottonians, see Karlheinz Mascher, Reichsgut und Komitat am Südharz im Hochmittelalter (Cologne, 1957), particularly 1–12. 94 With respect to the role played by Merseburg as a base of operations for Henry II’s campaigns against Boleslav Chrobry in the east, see David S. Bachrach, “Memory, Epistemology, and the Writing of Early Medieval Military History: The Example of Bishop Thietmar of Merseburg (1009–1018),” Viator 38 (2007), 63–90, here 73–5.

345

M AT E R I A L S O U R C E S A N D I M A G E S

which was located on the route that later would be named the Hohenstrasse.95 A column could cover the distance between each fortification in a day. The route Leipzig–Grimma–Oschatz and then on to Meissen became the standard line of march for German armies entering Lausatia after 929.96 Once troops reached Oschatz, they would be within close striking distance of Gana 11 kilometers to the south. Upon arriving at Gana, King Henry’s forces faced a very difficult task. Widukind emphasized that the siege lasted twenty days before the German army successfully stormed the walls. The massive 15-meter-wide and 5-meter-deep ditches had to be filled in, almost certainly under enemy fire, before an advance against the walls of the main fortress could be undertaken. To fill in just a 200-meter section of the defensive ditch would require 15,000 cubic meters of earth weighing approximately 15,000 tons.97 Under ideal conditions, an experienced construction worker could excavate about 3 tons of earth in the course of ten hours.98 Thus, simply to excavate the earth needed to fill in a 200-meter section of ditch at Gana would require 50,000 man-hours under ideal conditions. Of course, conditions in wintertime are far from ideal because the ground is frozen. Consequently, far more than 50,000 man-hours of labor are at issue here. Once the earth was removed from a surface area measuring 1.5 hectares, assuming a uniform depth of excavation of 1 meter, it had to be carried to the fortification at Gana and dumped into the massive defensive ditch that was to be traversed by the army.99 Under ideal circumstances, Henry’s men would have used carts, perhaps brought from Merseburg over the Hohenstrasse to Gana, to load, carry, and unload the earth. The carts in use during the early tenth century had a carrying

95 The fortification at Oschatz was located on an oval plateau measuring 200 × 150 meters. The drywall stone fortification at the summit has been dated to the tenth century and probably was the German fortification. See Hans Kaufmann, “Die frühgeschichtliche Wall-und Wehranlage auf dem oschatzer Collm,” in Archäologische Feldforschungen in Sachsen: Fünfzig Jahre Landesmuseum für Vorgeschichte Dresden (Berlin, 1988), 331–3. 96 Arne Schmid-Hecklau, Die archäologischen Ausgrabungen auf dem Burgberg in Meißen: Die Grabungen 1959–1963 (Dresden, 2004), 12. 97 For this figure see the detailed analysis by Bernard S. Bachrach, “The Cost of Castle-Building: The Case of the Tower at Langeais, 992–994,” in The Medieval Castle: Romance and Reality, ed. K. Reyerson and F. Powe (Dubuque, IA, 1984), 46–62 (four plates); and reprinted with the same pagination in Bernard S. Bachrach, Warfare and Military Organization in Pre-Crusade Europe (London, 2002), here 51–2. 98 Hans Hubert Hofmann, “Fossa Carolina, Versuch einer Zussammenschau,” in Karl der Grosse: Lebenswerk und Nachleben, 5 vols, ed. W. Braunfels and Helmut Beumann (Düsseldorf, l965), I: 437–53, here 345. For a somewhat faster estimate, see A. Simon Esmonde-Cleary and Jason Wood, Saint-Bertrand-De-Comminges III: le rempart de l’antiquité tardive de la ville haute. Etudes d’Archéologie Urbaine (Bordeaux, 2006), 143–6, and 215–16. 99 It is easiest to dig earth close to the surface. As the hole from which the earth is being removed becomes deeper, it becomes increasingly difficult to lift the soil to the surface. As a consequence, it is easier to remove earth from a large surface area than it is to dig deep trenches.

346

H E N RY I O F G E R M A N Y ’ S 9 2 9 M I L I TA RY C A M PA I G N

capacity of approximately half a metric ton of earth, if drawn by two horses or oxen.100 Loading earth onto carts and then in turn unloading them required, at a minimum, ninety minutes under ideal conditions.101 This does not include, however, any time for breaks during the loading or unloading process. Nor does this figure account for the difficulties of working in freezing conditions. The figure for loading and unloading should likely be increased by at least a third for a total of about two hours to load and offload each cart. As approximately 30,000 cart-loads of earth had to be transported to the ditch for dumping, the loading and unloading process would require, at a minimum, 60,000 man-hours. Presumably, the site from which the earth was excavated was a sufficient distance from the walls to protect the workers from enemy fire but close enough to minimize transportation time to the ditch. Horse-drawn carts operating under difficult conditions could move at approximately 3.2 kilometers per hour.102 If the earth were excavated at an average distance of 1 kilometer from the edge of the defensive ditch, each cart deployed could make three trips, under best-case conditions, in the course of the eight or so hours of sunlight that shines in January and February in northern Germany. This includes two hours for loading and unloading the carts, as well as forty minutes for the carts simply to travel 2 kilometers. In the course of a twenty-day siege, a minimum of 750 carts would be required to make 30,000 trips to load and offload earth into the defensive ditch. Excavating, loading, and unloading these carts required a minimum of 110,000 man-hours. With an average of just eight hours of daylight, it would require 1,000 men about two weeks, including work on Sundays, to carry out this excavation and unloading operation. The difficulties inherent in merely filling in the ditch in front of Gana’s walls certainly provide insight into the twenty-day length of the siege before Henry’s army finally stormed the walls. Widukind stresses that once the fortification at Gana was captured by storm, Henry permitted his troops to slaughter the male defenders. He then had the young boys and girls divided up among his professional soldiers (milites) as slaves.103 In light of the viciousness with which Henry handled the defeated Daleminzi, it is likely that his forces had suffered significant losses in the course of the siege and in storming the main fortification, described earlier. Given the need to fill in the defensive ditch, German casualties almost certainly were as high as they were at Brandenburg, and perhaps even higher, i.e. more than 1,000 dead and wounded.

100 Albert C. Leighton, Transport and Communication in Early Medieval Europe AD 500–1100 (New York, 1972), 72, gives the figure of 1000 pounds for the carrying capacity of a two-wheeled Roman vehicle. Marcel Girault, Attelages et Charrois au Mayen-Age (Paris, 1992), 138, sees pulling capacity of medieval carts limited to 500 kilograms, or 1,100 pounds. 101 Esmonde-Cleary and Wood, Saint-Bertrand-De-Comminges III, 145.` 102 Ibid. 103 Widukind, Res gestae, 1.35, ‘Preda urbis militibus tradita, puberes omnes interfecti, pueri ac puallae captivitatai servatae.’

347

M AT E R I A L S O U R C E S A N D I M A G E S

Phase three: the surrender of Prague After his capture of Gana, Henry began the march up the Elbe toward Prague. While underway, the German king ordered the construction of a new fortification at Meissen on the Elbe, some 25 kilometers southeast of the fortification at Hof/ Stauchitz. Dendrochronological dating of the wooden structures at Meissen indicates that construction began in 929, which suggests that Henry detached forces to build the fortress even as he continued his campaign.104 Meissen immediately became a key fortress in maintaining German control of the middle Elbe and sat astride the main route from the west to the upper Lausatian region.105 From Meissen, it was a march of 180 kilometers up the Elbe valley to its confluence with the Ultava river. From here, it was a further 50 kilometers upstream along the Ultava to Prague. In the course of his march, Henry’s army had to traverse the region now known as Saxon Switzerland (Sächsische Schweitz), which begins near the modern town of Pirna, some 45 kilometers south of Meissen. As it descends through the Erzgebirge, the Elbe has carved deep and winding valleys through limestone cliffs that provide numerous opportunities for defenders to lay ambushes. In order to move securely through this area, the German army required constant vigilance and effective scouting of the terrain. During Henry’s reign, the Elbe likely was navigable all the way to its confluence with the Ultava. Indeed, in 1744, 1756, and 1757, well before the construction of the modern river controls during the later nineteenth century, Frederick the Great of Prussia (1740–1786) led tens of thousands of troops along this same route. In order to supply these troops, Frederick deployed 500 transport ships all the way up the Elbe to the Bohemian plain. Although navigation was possible up to the city of Melnik, which was located at the junction of the Ultava with the Elbe, Frederick decided to halt his fleet at the major trading port of Leitmaritz, which had been developed at the confluence of the Elbe with the Eger (Czech Ohre), some 33 kilometers downstream from Melnik.106 There is no indication in the surviving written sources that Henry faced any significant difficulty on the march into Bohemia. Rather, Widukind rather laconically observed that the German king “came to Prague, the fortress of the Bohemians,

104 Regarding the dating of the earliest construction, see Schmid-Hecklau, Meißen, 205. 105 For the strategic significance of the fortress at Meissen, see Werner Coblenz, “Zur Frühgeschichte der meißner Burg: Die Ausgrabungen im meißner Burghof 1959/1960,” Meißner Heimat Sonderheft (1961), 3–32. Thietmar, Chronicon, 1.16 noted that Meissen was carved out of virgin forest lands. 106 See the discussion by Christopher Duffy, Frederick the Great: A Military Life (London, 1985), 94–5. Frederick chose to disembark his supplies, and particularly his heavy artillery, at Leitmaritz and then march overland to Prague, a distance of 68 kilometers (307), because the facilities at Leitmaritz were superior for the off-loading of heavy guns. This consideration obviously was not operative for Henry I.

348

H E N RY I O F G E R M A N Y ’ S 9 2 9 M I L I TA RY C A M PA I G N

with his entire army, and accepted the surrender (ditio) of its king.”107 It is perhaps not surprising that the Saxon monk does not report the presence at Prague of Duke Arnulf of Bavaria (907–937) with his army. Widukind was well aware that Arnulf had opposed Henry’s accession as king for two years (919–921) and that the sons of the Bavarian duke revolted against Henry’s son Otto I in both 938 and 953–954.108 However, Arnulf’s participation in this campaign is recorded in two Bavarian chronicles.109 The participation of Bavarians in this final phase of the campaign marks an important continuity with Carolingian military practice. The use of a pincer against Bohemia, with armies attacking from the north, south, and west, was standard operating procedure for the Carolingians dating back to the early ninth century. In 805, for example, an army based in Saxony joined both a Bavarian and an east Franconian army for an invasion of Bohemia.110 Indeed, Charlemagne recognized that Saxon armies regularly would be deployed against Bohemia in support of troops based in Bavaria and Franconia. In 807, Charlemagne formalized the military requirement that in wars against the Bohemians, one-third of the eligible Saxon troops were to be mobilized for service.111 Louis the German (840–876), Charlemagne’s grandson, deployed a Saxon army in 846, which was to march along the Elbe and then through Bohemia to meet a Bavarian army in Moravia.112 In 872, Louis the German’s campaign against Moravia included three armies, one of Saxons and Thuringians, a second of Franconians, and a third of Bavarians.113

107 Widukind, Res gestae, 1.35, ‘Post haec Pragam adiit cum omni exercitu, Boemiorum urbenm, regemque eius in deditionem accepit.’ 108 Regarding Duke Arnulf’s resistance to Henry I’s rule in the period 919–21, see Liudprandi Cremonensis Opera Omnia: Antapodosis, Homelia Paschalis, Historia Ottonis, Relatio de Legatione Constantinopolitana, ed. P. Chiesa (Turnholt, 1998), Antapodosis, 2.21; Annales Magdeburgenses a. 1–1188. 1453–1460, ed. G. H. Pertz, MGH SS 16 (Hanover, 1859), 142; and Widukind, Res gestae, 1.27. With respect to the revolt by Arnulf’s sons in 938 see Adalberti continuatio Regionis, 198; Annales of Einsiedeln, MGH SS 3 (Hanover, 1839), an. 938; Herimanni Augiensis Chronicon, ed. G. H. Pertz, MGH SS 5 (Hanover, 1844), 113; Annales Magdeburgenses a. 1–1188. 1453–1460, ed. G. H. Pertz, MGH SS 16, 143. Regarding the Bavarian revolt in 953–4, see Adalbert, Continuatio, an. 953; Gerhard of Augsburg, Vita Sancti Uodalrici, 1.10, ed. Walter Berschin and Angelika Häse (Heidelberg, 1993), 1.10; Herimanni Augiensis Chronicon, ed. G. H. Pertz, MGH SS 5, 114; Die Annales Quedlinburgenses, ed. Martina Giese (Hanover, 2004), 466–7; and Widukind, Res gestae, 3.34–35. 109 Annales Ratisponenses, 583: “Heinricus Boemiam vicit cum Arnolfo.” Also see Auctarium Garstense, 565: “Hainricus rex cum Arnolfo duce Boemanos vicit.” 110 The campaign was recorded in the Annales regni Francorum, an. 805; and Annales Mettenses priores, an. 805. See the discussion by Charles R. Bowlus, Franks, Moravians, and Magyars: The Struggle for the Middle Danube, 788–907 (Philadelphia, 1995), 58. 111 Capitularia regum Francorum, ed. Alfred Boretius and Viktor Krause, MGH, 2 vols. (Hanover, 1890–1897), I: Cap. 49, c. 2. 112 Annales Fuldenses, an. 846, and the discussion by Goldberg, Struggle for Empire, 138–9. 113 Annales Fuldenses, an. 872, and the discussion by Bowlus, Franks, Moravians and Magyars, 177–8.

349

M AT E R I A L S O U R C E S A N D I M A G E S

Even in Henry I’s own lifetime, in 892, Saxon troops served in Moravia under King Arnulf (887–899) alongside forces from Franconia, Bavaria, and Swabia.114 The most likely line of march for Duke Arnulf’s column from Bavaria into Bohemia followed the route Regensburg–Cham–Furth im Walde–Domazlice– Pilsen–Prague. From Regensburg to Furth im Walde the army traveled along the Regen river valley. The Bavarian column then entered the valley of the Radbuza until its confluence with the Berounka at Pilsen. From here, the Bavarians followed the Berounka to Prague and a rendezvous with the king’s troops.115 Did the final phase of the campaign to Prague include troops from Franconia? Certainly Duke Eberhard of Franconia (919–939) was allied with Henry I from the time of the latter’s accession. In addition, several important Franconian magnates, including Bishop Theodo of Würzburg (908–931), had very good relations with the German king.116 Given the frequent deployment of east Franconian armies into both Bohemia and Moravia by Carolingian rulers, it is certainly plausible that Henry I did mobilize these forces as well. Furthermore, an argument from silence that Franconian troops did not participate in the campaign is difficult to sustain. Widukind would not be expected to list Franconian contingents, especially if Eberhard commanded them, because of the latter’s revolt against Otto I in 939, which ended in the duke’s death at the battle of Andernach. Direct royal control over the Franconian duchy until the end of the Ottonian dynasty in 1024 also meant that there was no ducal patronage of Franconian chronicles that might record the presence of Eberhard or other Franconian magnates at Prague in 929. Nevertheless, in the absence of positive evidence, it must remain an open question whether the omnis exercitus commanded by Henry at Prague included Franconian contingents. Widukind’s emphasis that Henry had his “entire army” with him at Prague is appropriate, given the enormous challenge posed by the Bohemian fortress. Prague itself was protected by a substantial double fortification of approximately 1,200 meters, which enclosed some 6 hectares.117 The walls of the fortress were built in a blended fashion. An earth and timber core was protected by a stone front.118 The walls varied in breadth from 5 to 9 meters.119 Prague was protected further by two substantial ditches.120 Finally, the fortress was built on a steep hill.

114 See the discussion by Bowlus, Franks, Moravians, and Magyars, 226–7. 115 Ibid., 59. 116 Henry confirmed several important grants of revenues and property for the bishopric of Würzburg in 923. See Henry I, nr. 5, 6, and 7. 117 Jan Frolik, “Prag und die prager Burg im 10. Jahrhundert,” in Europa im 10. Jahrhundert: Archäologie einer Aufbruchszeit, ed. Joachim Henning (Mainz, 2002), 161–9, here 162. 118 Ivana Bohácová, “Zum Befestigungssystem der Premyslidenburgen (am Beispiel der archäologischen Untersuchungen in der Prager Burg und in Stará Boleslav),” in Frühmittelalterliche Burgenbau in Mittel und Osteuropa, ed. Joachim Henning and Alexander T. Ruttkay (Bonn, 1998), 37–47, here 39. 119 Ibid., 40. 120 Ibid.

350

H E N RY I O F G E R M A N Y ’ S 9 2 9 M I L I TA RY C A M PA I G N

In addition to the main fortress at Prague, the de facto capital of Bohemia was protected by two further substantial fortifications. The first of these was Levy Hradec, which was constructed on the Ultava 9 kilometers north of Prague.121 It was approximately the same size as Prague in the early tenth century with a fortified area of 6.4 hectares.122 The 4-meter-wide walls also were constructed in a blended fashion so that a 1.15-meter dry stone front protected an earth and timber core of approximately 3 meters. This wall was further protected by a ditch approximately 3 meters wide and 1 to 1.5 meters deep.123 The second protecting fortress, called Vyserad, was built just to the south of the Prague walls.124 Like Prague, Vyserad was a double fortress constructed on a steep hill whose perimeter walls measured approximately 1,300 meters.125 According to the formulae set out in the contemporary Burghal Hidage, noted earlier, 1,900 men, at a minimum, were required to defend the 1,200-meter circuit of the walls at Prague and the similar sized fortification at Levy Hradec.126 The 1,300 meters at Vyserad required a further 1,000 men, suggesting that Duke Wenceslaus’s defensive forces within the immediate environs of Prague were in the range of 3,000 men, at a minimum. Given the considerable and growing population density in the Prague region during the ninth and tenth centuries, it is very likely that Wenceslaus had ample men to defend the walls of Prague as well as Levy Hradec and Vyserad.127 Moreover, this grouping of fortifications posed a serious danger for Henry’s army if the king planned an extensive siege. Raiding parties based in any one of the three fortresses with intimate knowledge of the local topography could inflict considerable damage on foraging parties, and perhaps even interdict supplies that were being shipped up the Ultava or down the Berounka to the German army. However, despite their impressive fortifications, the Bohemians declined to fight. Instead, according to Widukind, when Henry arrived cum omni exercitu, the Bohemian “king” (rex) surrendered, i.e. accepted Henry’s ditio, rather than attempt to defend the city.128 It seems likely that Wenceslaus made this decision with the knowledge of events at Brandenburg and Gana, namely the successful capture by storm of these two impressive fortifications.

121 Katerina Tomkova, “Die Stellung von Levy Hradec im Rahmen der mittelböhmischen Burgwälle,” Frühmittelalterliche Burgenbau, 329–39. 122 Ibid., 329. 123 Ibid., 330. 124 Frolik, “Prag und die prager Burg,” 165. 125 Ibid., 67. 126 Bachrach and Aris, “Military Technology,” 1–4. 127 Regarding the dense population in and around Prague in the ninth and tenth centuries, see Frolik, “Prag und die prager Burg,” 164. Alexander Ruttkay, “The Organization of Troops, Warfare and Arms in the Period of the Great Moravian State,” Slovenska archeologia 30 (1982), 165–98, observes that Bohemia, as well as Moravia, utilized a tripartite military organization that drew heavily upon local levies for the defense of fortifications in their districts. 128 Widukind, Res gestae, 1.35.

351

M AT E R I A L S O U R C E S A N D I M A G E S

In order to make clear to Wenceslaus the futility of armed resistance, Henry required an army that was large enough to pose a credible threat of carrying the walls of Prague by direct assault. Settling down for a lengthy siege in the late winter or early spring in the middle of hostile territory at the end of a lengthy supply route could not have been an attractive option for the German king. Wenceslaus also certainly was aware of the logistical challenges posed by a winter campaign when the last dregs of supplies were being consumed in hopeful expectation of a bountiful harvest of winter grains. If Henry were to capture Prague by force, he would have to assault the fortress before supplying the army became too difficult. At a ratio of just four to one, Henry would have required an army of 12,000 men to pose a credible threat to the fortresses at Prague, Levy Hradec, and Vyserad, if he intended simultaneous assaults on all three. However, it is likely that Wenceslaus brought most, if not all, of his men into the primary fortress of Prague, to assure that his “capital” at least was well defended so that Henry’s army would be faced with the challenge of storming this very well-manned stronghold. In this context, a ratio of four attackers to every defender likely is too low, given the difficult approaches to the hilltop fortifications, which would slow down the German assault and provide more time for the Bohemian archers to inflict casualties on the attackers. A ratio of five to one would place Henry’s army in the 15,000 range. This is on the high side for expeditionary forces dispatched by Carolingian kings during their invasions of Bohemia, but certainly within the realm of possibility, especially if troops were drawn from Franconia as well as Saxony, Thuringia, and Bavaria.129 129 Goldberg, Struggle for Empire, 126, estimates that each of the five regions within the east Frankish kingdom (Saxony, Thuringia, Bavaria, Franconia and Swabia) could each provide 6,000 men for expeditionary service, for a total of 30,000 men, but that the king usually mobilized far fewer than the maximum. In this context, he contends that Louis the German’s (840–876) armies, when they were mobilized from just one or two of the five regions, ranged in size from 2,000 to 12,000 men. Goldberg then suggests (126), given the repeated emphasis by the contemporary chroniclers on the great size of Louis the German’s armies, that the military forces mobilized by this Frankish king “often approached 10,000 men.” When due consideration is given to all of the military tasks that had to be completed on campaign in addition to providing a sufficiently large force of foot soldiers to provide a credible threat when storming large-scale fortresses, it seems likely that the upper range, estimated by Goldberg, is too low, and that consequently, the large armies that Louis mobilized were significantly larger than 10,000 men. In this context, Goldberg (125) accepts the basic model postulated by Karl Ferdinand Werner, “Heeresorganization und Kriegsführung im deutschen Königreich des 10. und 11. Jahrhunderts,” in Settimane di Studio de Centro Italiano sull’alto Medioevo, vol. 15 (Spoleto, 1968), 813–32, here 820–1, regarding the pool of manpower that was available for mobilization by Charlemagne north of the Alps for offensive campaigns. However, Goldberg misread Werner’s argument that 30,000 to 35,000 heavily armed mounted fighting men and a minimum of 100,000 lightly armed mounted fighting men and foot soldiers were available for expeditionary military service. Instead, Goldberg indicates that Werner had argued for a total of 100,000 men of whom 30,000 were heavy cavalry, thereby providing an underestimate of approximately one-third, i.e. 30,000 rather than 45,000 for the east Frankish kingdom. If the range of forces identified by Goldberg as being mobilized by Louis the German, i.e. 2,000 to 12,000, is increased in a manner

352

H E N RY I O F G E R M A N Y ’ S 9 2 9 M I L I TA RY C A M PA I G N

Logistics for the campaign The capture by storm of two heavily fortified and strongly defended princely seats at Brandenburg and Gana, along with their attendant belts of lesser strongholds, followed by the surrender of Duke Wenceslaus at Prague points to the substantial military strength brought to bear by Henry I in the course of the campaign of 929. Indeed, the very large army that Henry almost certainly was able to deploy at Prague is even more impressive in light of the significant losses that Henry’s forces suffered in bloody assaults against the Hevelli and Daleminzi, the need to detach forces to garrison recently captured strongholds, and the beginning of the construction of a major new fortification at Meissen. The question at hand is how did King Henry keep his forces supplied? The army that overawed the Bohemian duke at Prague will serve as a model for the logistical operation of the entire campaign. Twelve thousand to fifteen thousand men require on a daily basis a minimum of 12 to 15 metric tons of food supplies.130 If the armies marching to Prague had, in aggregate, 3,000 pack animals and war horses, an additional 15 metric tons of food and fodder would be required per day.131 In total, some 27 to 30 metric tons of food and fodder had to be delivered to Henry I’s men and beasts every day over the entire third phase of the campaign. Over land, the movement of supplies could be accomplished by carts that, in the tenth century, were much the same as those employed by the late Roman Empire, as well as the Roman successor states in the West. It was not until the development of the rigid shoulder collar for horses in the ninth century was combined with the

consistent with Werner’s model, we come to a maximum of 18,000 men. In line with the estimates by Werner, I suggest that armies of more than 10,000 men commonly were deployed by Louis the German, a half century before Henry I’s campaign of 929. Since Henry ruled a larger, more populous, and wealthier kingdom than did Louis the German, it does not seem unreasonable that he could raise armies at least as large as those of his Carolingian predecessor. 130 Fighting men require approximately 3,000 calories a day to maintain their effectiveness on campaign. In this context, see Jonathan P. Roth, The Logistics of the Roman Army at War (264–335) (Leiden, 1999), 7–55; and Bernard S. Bachrach, “Crusader Logistics; from Victory at Nicaea to Resupply at Dorylaion,” in Logistics of Warfare in the Age of the Crusades, ed. John H. Pryor (Aldershot, 2006), 43–62. For relatively brief campaigns of just a few weeks, fighting men could be supplied with a rough and ready diet of hard-baked biscuit or bread and some dried or smoked meat or fish, with approximately 1 kilogram per man of these food stuffs providing 3,000 calories. See the observations by Franz Stolle, Der Römische Legionär und sein Gepäck (‘Mulus Marianus’): eine Abhandlung über den Mundvorrat, de gepälast und torister des Römischen Legionärs und im anjhand Erkläring der Apokolpse 6.6 (Strassburg, 1914), 28; and Roth, The Logistics of the Roman Army, 43, 51–2, 70–1, with a synthesis by Bachrach, “Crusader Logistics,” 54–6. Kathy L. Pearson, “Nutrition and the Early-Medieval Diet,” Speculum 72 (1997), 1–32, here 8, identifies some of the ways in which meat and fish were processed so that they would not spoil. 131 Regarding the feeding of horses, see Bernard S. Bachrach, “Caballus et Caballarius in Medieval Warfare,” in The Story of Chivalry, ed. H. Chickering and T. Seiler (Kalamazoo, 1988), 173–211; and reprinted with the same pagination in Bernard S. Bachrach, Warfare and Military Organization in Pre-Crusade Europe (London, 2002), 179–80.

353

M AT E R I A L S O U R C E S A N D I M A G E S

development of the pivoted axle and wipple tree for wagons in the twelfth century that there was a significant increase in the carrying capacity of vehicles.132 As noted previously, carts, which were drawn by two horses or mules, could carry a load of approximately 500 kilograms.133 Thus, the equivalent of some fifty-four to sixty carts had to reach Henry’s forces every day during the third phase of his campaign. Pack animals could also be used to transport supplies, although this was far less efficient than using vehicles. Mules could bear loads of approximately 75 kilograms and horses 100 kilograms.134 Part of the load for these animals, however, would consist of their own food. Horses required 5 kilograms of grain as well as 5 kilograms of grass each day.135 The latter generally was not available in the winter, so this also had to be carried. Mules required approximately 3.5 kilograms of grain a day, as well as an equal quantity of grass or hay.136 In addition to land-based transportation, the route of Henry’s campaign afforded him the option of carrying supplies by water. As noted earlier, Henry’s army likely marched along the valleys of the Elbe and then the Ultava. The Bavarian column that marched to Prague, as mentioned previously, likely followed the Regen to the Radbuza and then the Berounka, which meant that it could be supplied by water all the way to Prague as well. If a Franconian column participated, it is very likely that it followed the Eger (Czech Ohra) river valley to its confluence with the Elbe south of Melnik, joining Henry I’s column the rest of the way to Prague. So the Franconians, too, could be supplied by water.137 One barge of the type typically used for transportation on these rivers during the mid-tenth century could carry up to 15 tons of supplies, or the equivalent of thirty cart loads.138 At this rate, two barges would have to reach the army every day to meet its food and fodder needs. 132 With regard to the slow diffusion of the rigid shoulder harness for horses after its initial development, see Paul Gans, “The Medieval Horse Harness: Revolution or Evolution? A Case Study in Technological Change,” in Villard’s Legacy: Studies in Medieval Technology, Science, and Art in Memory of Jean Gempel, ed. Marie-Thérèse Zenner (Aldershot, 2004), 175–88. Concerning the development of the pivoted axle during the twelfth century, see Marjorie Nice Boyer, “Medieval Pivoted Axles,” Technology and Culture 1 (1960), 128–38; which is placed within a broader context by John Langdon, Horses, Oxen and Technological Innovation: The Use of Draught Animals in English Farming from 1066–1500 (Cambridge, 1986). 133 Regarding the carrying capacity of carts, see the basic studies by R. Lefebvre des Noëttes, L’attelage et le cheval de selle à travers les âges, 2 vols. (Paris, 1931); and Albert Leighton, Transportation and Communication in Early Medieval Europe (Devon, 1972). 134 Roth, The Logistics of the Roman Army, 202–8. 135 Regarding the feeding of horses, see Bachrach, “Caballus et Caballarius in Medieval Warfare,” 179–80. 136 Roth, The Logistics of the Roman Army, 62, 65. 137 See the discussion of the route from Franconia along the Eger by Bowlus, Franks, Moravians, and Magyars, 176. 138 See, in this context, Aleydis van Moortel, “Medieval Boats and Ships of Germany, the Low Countries, and Northeast France: Archaeological Evidence for Shipbuilding Traditions, Timber Resources, Trade and Communications,” in Settlement and Coastal Research in the Lower North Sea Region (Kögthen, 2011), 67–105.

354

H E N RY I O F G E R M A N Y ’ S 9 2 9 M I L I TA RY C A M PA I G N

What was the source of this considerable quantity of food? From the north, Henry could easily mobilize vast quantities of supplies from the thirty-one villae that were attached to the royal complex at Magdeburg.139 Ottonian villae in this region appear to have been quite large. For example, the royal villa of Wolmirsleben, some 25 kilometers southwest of Magdeburg, had sixty-four familiae of dependents divided among servi and liti.140 Similarly, the villa of Fermersleben had fifty-six familiae of dependents working its lands.141 It was generally the case that a familia farmed a dependent mansus, so that these two estates had a minimum of sixty-four and fifty-six dependent mansi, respectively, with perhaps the same number held in demesne.142 To put these figures into perspective, a charter issued by Louis the German in 874 indicates that just eight mansi attached to the villa of Hornau, in the region of Frankfurt, produced a surplus of 130 modii of grain, as well as twenty pigs, and four cartloads of wine.143 This amounts to a surplus beyond that needed to feed the cultivators consisting of sixteen modii (or half a metric ton) of grain, as well as two and half pigs and half a cart-load of wine per mansus.144 There is nothing in Louis the German’s charter to indicate that Hornau was a particularly productive villa, and it can be seen as providing a rough model for the surplus production of royal estates in the east. In this context, if the thirty-one villae attached to the royal curtis of Magdeburg averaged just 50 dependent mansi each, i.e. fewer than are known to have been attached to Wolmirsleben and Fermersleben, these 1,550 mansi can be postulated to have produced a surplus of 775 metric tons of grain and 3,750 pigs. These supplies would be enough food to keep 15,000 men and 3,000 horses in the field for an entire month. If one takes the dependent mansi as representing just half of the total mansi attached to an estate, then these sums can 139 In September 937, Otto I issued a charter on behalf of the newly established monastery of St Maurice at Magdeburg that lists thirty-one villae that were attached to the royal curtis of Magdeburg, whose income was now being assigned to the monastery. Otto I, nr. 14. Also see the discussion by Gerhard Billig and Gerd Böttcher, “Burgen und Burgbezirke im Erzstift Magdeburg vom 10. bis zum 12. Jh.,” Magdeburger Blätter (1984), 24–38, here 27. 140 Otto I, nr. 16. 141 Otto I, nr. 14 and 21. 142 For the ratio of one familia to one mansus, see, for example, Otto I, nr. 395. On the relationship of demesne to dependent mansi see the valuable discussion by Walter Goffart, “Frankish Military Duty and the Fate of Roman Taxation,” Early Medieval Europe 16.2 (2008), 166–90, with the literature cited there, particularly 171–3. 143 Die Urkunden Ludwig des Deutschen, Karlomanns und Ludwigs des Jüngeren, ed. Theodor Schieffer (Berlin, 1966), here Louis the German, nr. 155. 144 Regarding the weight of a modius of grain, see the discussion by Konrad Elmshäuser, “Facit Navigium: Schaffahrt auf Seine, Marne, Mosel und Rhein in Quellen zur frühmittelalterlichen Grundherrschaft” in Häfen, Schiffe, Wasserwege: Zur Schiffahrt des Mittelalters, ed. idem (Bremerhaven, 2002), 22–53, here 31; and the broader survey of the literature by Jean-Pierre Devroey, “Units of Measurement in the Early Medieval Economy: The Example of Carolingian Food Rations,” French History 1 (1987), 68–92, particularly 80–6. Devroey (86) concludes that the modius of grain ranged in weight from 30 to 40 kilograms, so that sixteen modii per mansus would amount to 480 to 640 kilograms.

355

M AT E R I A L S O U R C E S A N D I M A G E S

be doubled to account for the support of a large army of 15,000 men and 3,000 horses for two months. The considerable resources that were available at Magdeburg were mirrored at numerous other fiscal complexes throughout Thuringia and eastern Franconia, such as Halle, Merseburg, Erfurt, and Würzburg.145 Moreover, Duke Arnulf of Bavaria had access to both royal and ducal fiscal resources in his duchy, including the vast panoply of properties that were attached to Regensburg and Forchheim.146 In short, neither the quantity of supply nor an inability to transport food stuffs can be understood to have limited the ability of Henry I to put a very large army into the field for the purpose of besieging large, well-defended fortifications such as those at Prague.

Conclusion The campaign of 929 was a great military as well as political success. Henry I gained, at least temporary, control of the Havel river valley and permanently added the entire region between the Saale and Elbe rivers to the German kingdom. The first Ottonian king also acquired the fidelity of Duke Wenceslaus of Bohemia in a manner consistent with his Carolingian predecessors. Such conclusions are hardly novel, buttressed as they are by the surviving narrative sources.

145 In 923, Henry I confirmed the receipt by the bishop of Würzburg of 10% of the revenues of twentythree fiscal complexes, including the exceptionally wealthy center at Ingelheim. See Henry I, nr. 6 for the grant. With regard to the fiscal complex centered on Ingelheim, see H. Schmitz, Pfalz und Fiskus Ingelheim (Marburg, 1974). Concerning the importance of the royal fiscal center at Erfurt, see Walter Schlesinger, Die Entstehung der Landesherrschaft: Untersuchungen vorwiegend nach mitteldeutschen Quellen, revised edition (Darmstadt, 1964), 45; and Michael Gockel, “Erfurt,” in Die deutschen Königspfalzen: Repertorium der Pfalzen, Königshöfe udn übrigen Aufenthaltsorte der Könige im deutschen Reich des Mittelalters, vol. II Thüringien (Göttingen, 2000), 103–48. With regard to the wealth and military importance of the royal complexes at Merseburg and Halle, see A. Koch, Veste über en Wassern. Die Hochseeburg und die Königs- und Bischofsbauten in Merseburg (1932); Paul Grimm, “Archäologische Beobachtungen an Pfalzen und Reichsburgen östlich und südlich des Harzes mit besondere Berücksichtigung der Pfalz Tilleda,” in Deutsche Königspfalzen: Beiträge zu ihrer historischen und archäologischen Erforschung, vol. 2 (Göttingen, 1965), 273–99, here 281; and Volker Hermann, “Der ‘Limes Sorabicus’ und Halle im frühen Mittelalter,” in Siedlung, Kommunikation und Wirtschaft im westslawischen Raum: Beiträge der Sektion zur slawischen Frühgeschichte des 5. deutschen Archäologenkongresses in Frankfurt an der Oder (Langenweissbach, 2007), 133–43. 146 Regarding the fiscal resources attached to Regensburg, see Peter Schmid, Regensburg: Stadt der Könige und Herzöge im Mittelalter (Kallmünz, 1977). With regard to Forchheim’s role as the centre of a massive fiscal complex, see Die Urkunden Otto des II und Otto des III., ed. Gesellschaft für ältere deutsche Geschichtskunde (Hanover, 1888), here Otto II, nr. 132; Die Urkunden Heinrichs II. und Arduins, ed. Gesellschaft für ältere deutsche Geschichtskunde (Hanover, 1900–1903), Henry II, nr. 170; Die Urkunden der deutschen Könige und Kaiser: Die Urkunden Heinrichs IV, 3 vols., ed. Dietrich von Gladiss and Alfred Gawlik (1941–1978), Henry IV, nr. 72 and 88.

356

H E N RY I O F G E R M A N Y ’ S 9 2 9 M I L I TA RY C A M PA I G N

However, the enormous, and ever-growing, corpus of information developed through the efforts of archaeologists regarding the military topography of the Slavic lands permits a much more detailed analysis of the campaign of 929 than has been possible relying on the laconic observations of contemporary and nearcontemporary authors alone. In particular, the scale of Henry’s military operations, as detailed here, is far larger than has been appreciated by scholars who specialize in the history of the Ottonian kingdom.147 Siege warfare, in general, required the deployment of significant numbers of men who were trained to fight on foot. Putative warriors on horses, however brave the man and beast, have little role to play in the storming of high walls. When due consideration is given both to the size and number of fortifications that were captured by Henry I during the course of the 929 campaign, it is clear that he fielded a very large army indeed. This army, necessarily, was deployed to fight on foot, particularly in the storming of Brandenburg, Gana, and whichever other Hevelli and Daleminzi fortifications resisted Henry’s troops. Moreover, as the success of Henry’s campaign makes clear, the German king had the capacity to feed these troops and their horses over an extended period. One of the likely sources of supply was the royal fisc, including the well-organized curtis of Magdeburg, with its thirty-one dependent villae. The campaign of 929 was just one of many successful military operations undertaken by Henry I that include the siege and capture of important fortifications. Henry’s son Otto was an even more prolific campaigner than his father, waging war and besieging strongholds, including great fortress cities, across the length and breadth of the Latin West. Future research regarding the scale and complexity of warfare under the Ottonian kings, their Carolingian predecessors, and their Salian successors, must consider the enormous trove of information that has been developed by our archaeologist colleagues.

147 See, for example, the suggestion by Timothy Reuter, “Carolingian and Ottonian Warfare,” in Maurice Keen (ed.), Medieval Warfare: A History (Oxford, 1999), 13–35, that Ottonian armies rarely numbered even a few thousand men.

357

INDEX

Abels, R. 71n123, 251–3, 262 Abraham (biblical figure) 139 acie maxima 138 acta 161 Adela of Blois 226 Ademar of Chabannes 153–70, 153n2, 157n23, 158n25–6; audience 158–62; establishing rhetorical plausibility 155–8; Historia 153; literary strategy 155; as a military historian 163; military organization 163–6; siege tactics 167–9; strategy 166–7 Adhémar, Bishop 226 ad vindicanda armari 118 Aetius (general) 8–9 Agate, A. 257–8 agrarii milites 57–8, 58n60, 74, 77, 82, 82n4; see also milites agricultural production 294 agrimensores 268 Airlie, S. 24, 24n22 Albert of Aachen 227 Albrecht, A. 269 Albu, E. 132–3n3 Alcuin (scholar) 25, 34 Alemannia 66 Alexander the Great 10 Alfred the Great 247, 251–2, 257 Altenburg 265 Althoff, G. 53–5, 75n142, 92, 103, 175n20 Amo (duke) 13 Anabasis of Alexander (Arrian) 7 Angilbert (saint) 22 Anglo-Norman 21, 182 Anglo-Saxon 21, 117, 128, 136, 199, 255, 256–8, 314n48; artist 304–5, 307, 309–10, 313, 315–21, 325; attack

on the Normans 314; designer 312; interpretation of events 307; military history 252; military organization 256; monk 323; script and spelling 306 Anglo-Saxon Chronicle 248 Anglo-Saxon England 71–2, 71n123, 164, 240, 243–4, 252, 257, 297, 339 Angoulême 22 Annales 208, 212 Annales Bertiniani 65 Annales Fuldenses 62, 64, 66–7 Annales Ratisponenses 330 Annales regni Francorum (Kurze) 27, 65–6 Annales Sancti Bertiniani 65 Annals of Fulda (Reuter) 83 Ansfrid, Bishop 115, 124 Antapodosis (Bishop Liudprand of Cremona) 110 anti-Saxon bias 64, 69 apostle 155, 170 apostolicity 155, 158–9, 170 architectura, De (Vitruvius) 262, 267 Arian Visigoths 141 Aribertus of Charroux 159 Aridius (Gallo-Roman advisor) 9–11 arma militaria 218 armati 47, 91, 151 armatis manibus 115 armatus miles 118 armed conflict 153 armigeri 137–8 armillae 150 Arnold, M. 20n4 Arnulf, Bishop 117 Arnulf, King 350 Arnulf of Carinthia 267n17, 297

358

INDEX

Arnulf of Chocques 226, 228 Arnulf the Bad of Bavaria (duke) 61, 87, 90–1, 348 Arnulf the Younger 87 Arrian (Greek historian) 7 Arslan, Kilij (sultan) 229–31, 234–5 asperrimus miles 147 Astronomer 28, 28n46, 66 attacking forces 340n55 Auctarium Garstense 330 auctoritas 157 Auer, L. 50n13, 198–9n5, 202–3, 202n19, 209, 224n11 Augsburg 331n11 Augustine, Saint 160, 160n35 Austrasians 28 baculum 319 Bajocacenses 145 Baker, J. 242, 240n3, 255 Baldwin I, King 228 Banniard, M. 162n47 Bardulias 253 barges 286 basket carriers 281–2, 285–7 baskets 282n78, 290n108 basterna 290 Bates, D. 20n4 battle-avoiding campaign strategies 33–7 battlefield operations 181–2 battlefield tactics, Saxon Wars 185–6 battle-seeking campaign strategies 33–7 Baudri of Bourgueil 227, 324 Bavaria 32, 40, 172, 213, 295, 350, 352 Bavarians 214, 349 Bayeux Tapestry 303–26, 304n2, 309n23; Norman reaction to 324–6; Odo’s Gesta Guillelmi 318–26 Beleknegini (Hungarian noblewoman) 126 bellum publicum 57, 73–4, 77 Bellum Saxonicum 44n135, 175–6, 179 beneficia 124, 130, 200, 206, 294, 322–3 Bennett, M. 324n80 Beowulf 152 Beowulfian model 146 Bernard of Septimania 38–40, 85 Bernhardt, J. W. 199n7 Bernold of Constance 209 Bernstein, D. 304–6, 310, 323, 326 Berthold of Carinthia (duke) 173, 217–18 Berthold of Reichenau (monk) 28n45, 176, 209, 214, 217–18

Bertrand, S. 306n13 Beumann, H. 52–3, 52n26, 53n1, 55 Beyerle, F. 82n6, 202, 209 Beyond the Burghal Hidage (Baker and Brookes) 242 biothermal units (BTUs) 289n105 Bliese, J. 162n45 Blümcke, O. 198n3 Bohemia 111–12, 123, 125, 331–2, 350–2 Bojacacenses 145n53 Boleslav Chrobry (duke) 75, 115–16, 122–3, 125, 128–30, 340 Boleslav II of Bohemia (duke) 117 Boleslav I of Bohemia (duke) 117 Bönisch, H. 264n4, 276n60, 278 Boso (king) 169 Bowlus, C. R. 19n2, 205n30 Bradbury, J. 15n29, 49n10, 195n32, 220n97–8 Brandenburg 334–42, 337, 344, 353 Brandenburg, Henry I, capture of 334–41 Brault, G. J. 140n29 Brömel, M. 111 Brookes, S. 240n3, 242, 254–5 Brooks, N. 303n1, 305 Brown, R. A. 307, 308n17 Brown, S. 304 Brühl, C. 90n54 Brunner, H. 21, 21n9, 48, 48n6, 107n34, 200–1, 200n9, 204 Brunner-White thesis 48, 48n7 Brun of Arneburg 116, 121 Bruno of Merseburg 44n135, 175–6, 207, 209–11, 215–16; see also Saxon Wars Bruno sheriff 174n15, 198n2 “brute fact” 105, 107 Buc, P. 110 Bukko 123 Bünnig, W. 336n42, 337n43 Burchard, Bishop 217 Burchard II of Swabia 61 Burchard of Swabia 86 Burggraf of Magdeburg 111–12 Burghal Hidage 240, 248, 252–3, 257, 259n29, 339, 339n54, 351 Burgundians 214 Burgundy 214 burgus 16 burhs 255–7 Butzer, P. L. 268n18 Byzantine 118

359

INDEX

Caesar, J. 7, 9, 227, 248 Caliphate of Cordova 168 Callahan, D. 153, 156n12–13 Cam, H. M. 76n144, 113n67 cambitores 150 campaign of 929, Henry I 333 Campbell, J. B. 7n1, 71n123, 164n54 capture of Brandenburg, Henry I 334–41 capture of Gana, Henry I 341–7 Carinthians 214 Carloman (king) 28–9 Carmen de bello Saxonico 215–16, 219 Carolingian 46, 78n148, 142–3, 163, 176n24, 199, 202–3, 220, 266–7, 290n106; armies 83; builders 268; capitularies 201, 203, 206; court 149; era 151; fortress 266n13; government 334n29; kings 29, 34, 41, 43, 66–7, 72, 113, 352; military 19–21, 19n1, 26, 48, 65, 71, 164, 200, 349; policy 331; political theorists 58; predecessors 148, 265, 297; regulations 77; Reich 62; rulers 32, 41, 70, 75, 345n91, 350; ruling family 34; style expeditionary levies 205; warfare 25–6 Carolingian Empire 21, 23, 46, 136, 140, 143, 160, 163, 220, 263, 297 Carroll, C. F. 279n71 Carruthers, M. 102 Castillo, J. A. Q. 248–51 causa scribendi 23–5 Cecilia, daughter of William the Conqueror 226 centenariae 218 Centulla 149 Chambers, R. W. 280n76 chanson de geste 229 Chanson de Roland 152 Charbonnière 26 Charlemagne 11, 19n1, 29, 29n52, 34, 47, 56, 142, 142n41, 159, 247, 349; 808 capitulary 203; army 63, 151; Carolingian military forces 26; conquest of the Saxon 75; court of 24; deployment of Saxon forces 62, 65; grandsons 136; historiae 25; infancy 21; invasion of Spain 27; military forces 70; reign 19; res gestae antiquorum 25; Saxon military forces 65; Saxon military resistance 51 Charles III 61 Charles the Bald of West Francia (king) 22–6, 28, 30–42, 34n82, 44, 66–7, 68n112, 69, 297

Charles the Great 46 Charles Martel 47–8, 79, 200, 204 Cheyette, F. 166n61 Chibnall, M. 20n4 Childebert II (king of Austrasia) 11 Chilperic (king) 11–12 Chlodomer (king) 17 Christi, N. 12n19 Christiani 168 Christians 156, 224, 251, 253 Christiansen, E. 147n67 Christie, N. 247–8 Christmas Day 800 105 Chronicon 105, 107–9, 112n61, 113–30, 114n70, 118n103, 153, 155, 158–60, 158n25–26, 163, 170 Cicero (Roman statesman) 9, 160, 210n49 cives 151, 165, 219 civil war 171–2, 207, 220 civitas 16, 123, 135, 137, 309 civitates 11, 115, 145, 257, 309 civitatis custos 116 Clausewitzian-inspired battle-seeking warfare 181 Clovis (king) 10–11, 141 Coblenz, W. 264n3, 265n8, 342, 342n64 Collingwood, R. G. 1, 102n14 Collins, R. 20n6 coloni 142 comitatus 146, 204, 217 comites 11, 26, 28, 147 command and control, Saxon Wars 186–9 Commentaries (Caesar) 7 comprovinciales 75, 218 Conan of Rennes (Count) 313 conduct of war, Nithard (military historian) 25–6 Conrad, King 60 Conrad I, King 59, 76, 86 Conrad II, King 198 Conrad the Red (duke) 87–8, 117 Constable, G. 161 Constantinenses 145, 145n53 Contamine, P. 157n20 control of information, Saxon Wars 182–3 conventum 160 coup de grâce 35 Coupland, S. 163n52 Creighton, O. 247–8 Creveld, M. van 181n59 Curthose, R. 226 Czech Republic 94 Czimislav (Sorbian king) 66

360

INDEX

Danish Sea Scouts 313 Dark Age 240 decision-making 186, 267 Dedi of Lower Lausatia (margrave) 172 Dedi of Wettin (count) 124 DeLaine, J. 282n78 Delbrück, H. 104, 104n25–6, 139n26–8, 173n14, 200–1n10, 262 denarii 293–4 De re militari (Vegetius) 157, 170 D’Herbomez, A. 292–3n113 Dieterich, J. R. 198n4 Dietrich 115 Domesday Book 255 Dorylaeum 229 Drahomir 331 Dralle, L. 331n14 Duby, G. 130 Dudo of St. Quentin 54n40, 132–52, 133n4, 137n15–16, 139n24, 144n49–50, 148n68, 148n70–2, 148n74, 150n84, 154, 159, 228; De moribus et actis primorum Normanniae ducum 228; local levies 135–41; military organization 134–5; select levies 141–5; standing armies 145–51 Duffy, C. 348n106 Dunker, H. 269, 269n23, 270n28, 272n42, 300 early medieval military history 98–131; Chronicon 113–19; milites 113–30; Thietmar of Merseburg 108–19 East Francia 46–7 East Franks 28 Eberhard of Franconia (duke) 86, 350 Eberhard the Bearded (duke) 181 Edgeworth, M. 247–8 Edward, King 306, 326 Egilard 124 Eichenbach 186 Einhard 25, 62–3, 63n81 Ekbert of Meissen (margrave) 215 Ekkehard I of Meißen (margrave) 117–18, 124 Engel, U. 19n2 Engels, D. 105n27 England 116, 136, 140, 144 English Channel 312 Epitoma rei militaris (Vegetius) 15n30, 33, 92, 262 Ermold (author) 62–3, 62n78 Escalona, J. 251, 253–4

Esders, S. 135n8 Ettel, P. 244–7 Etymologies (Isidore of Seville) 25, 154n8 Euclid’s geometry 247 Eunapius (historian) 10 Europe 46–7, 98, 201, 241 Eustace of Boulogne 321 exploratores (scouts) 183, 335 Fanning, S. C. 146n60–1 feeding of horses 353n131 Fehr, H. 201 feudalism 48–50, 59–60n67, 119, 199, 201, 204, 249; see also Salian Germany feudum 322 fideles/fidelis 36, 39, 60–1, 72, 76, 79, 117, 148, 156, 159, 165, 170, 180, 297, 315, 322, 327 Fiefs and Vassals: The Medieval Evidence Reinterpreted (Reynolds) 48–9, 249 Field of Lies 31 First Crusade 223–7 Fläming 331 Fleckenstein, J. 20–1, 20n3, 27n37, 53, 203–5, 209 Folkmar of Brandenburg 115 Fontenoy, battle of 32 Forkbeard, S. 116 fortifications 83n10, 90n56, 90n59, 106, 119–120n109, 241n9, 263–4n2, 335n32, 341n60–61, 343n77, 346n95; castellum 121, 123, 215, 308, 325; castra 13, 138, 167; castrum 16, 138, 308, 333; of Gençay and Bouteville 167; Hildagsburg 276; logs 290; oppida 73–4, 73n133, 138; Ottonian and Slavic 277; presidio 75; Roman 309; stone 167 fortress construction: bailey defenses 274; berm 271; costs of 291–2; digging ditches 276–7; entry system 272–3; general patterns of 264–7; hauling costs for stone 285; making mortar 287–91; Ottonian 267–9; stone walls 287; wood cutting for 278–80 fortress construction in tenth-Century Germany 263–299; bailey defenses 274; barges 286; basket carriers 281–2, 285–7; berm 271; building Ottonian fortresses 267–9; building the stone walls 287; constructing the subwall of the Hauptburg 280–1; costs 278–80, 291–2; digging ditches 276–7; entry system 272–3; general

361

INDEX

patterns of 264–7; hauling costs for stone 285; Hauptburg moats 271–2; Hildagsburg 269–71; making mortar 287–91; methodology 274–6; quarrying greywacke for the Hauptburg 283–4; statistics 292–5; stones quarried 284–5; stone walls of the Hauptburg 282–3; Vorburg 273; wooden walls of the Hauptburg 277–8 France, J. 72n125, 223, 229, 229n36 Francia 33, 40, 168 Francia Occidentalis 19, 116, 143, 152 Francia Orientalis 136 Franconia 183, 350, 352 Franconians 214, 218, 349 Frankish military organization 200–1, 204 Frankowich, R. 243 Frauenholz, E. von 202 Frederick II of Germany 100 Frederick Barbarossa 201 Frederick of Sommerschenburg, Count Palatine 192 Frederick the Great of Prussia 348 Fried, J. 54–5, 54–5n41, 81n1, 99–106, 100n6–8, 101n13, 104n24 Frigeridus, R. 9 Frisiones 138 Frontinus 92, 262 Fulbert, Bishop 159, 322 Fulcher of Chartres 231n45 Fulda Gap 59 Fulk Nerra (Angevin count) 166, 307 Funkenstein, A. 103, 103n20 fyrd-straet 255 Gana capture, of Henry I 341–7 Geary, P. 43n126, 99, 99n3–5, 99n5, 102 Gemeinfreiheit 256 generalship 9 Geoffrey, Bishop 320 Gerard of Paris 31 Gerhard of Augsburg 88n43 German Democratic Republic 269 Germania (Tacitus) 241, 256 Germany 50, 82, 84, 113, 171, 173, 206, 214; archaeology 245; civil war 207; fortifications 247 Gero (margrave) 94, 126 Gero, Archbishop 115, 121, 126 Gerstungen 214 Gesta Normannorum 132 Gesta Tancredi 225–7 Gesta Wilhelmi 305

Ghana, Gonja people 103 Giese, W. 171n2 Giles, J. A. 319n61 Gillet, A. 78n148 Gillingham, J. 25, 33, 33n79, 35n88, 153n3, 154n7, 308n20 Giselher, Archbishop 110, 115, 118, 120 Gislebert of Lotharingia (duke) 86 Godfrey IV of Lower Lotharingia (duke) 218–19 Godman, P. 20n6 Goetz, H.-W. 23 Goffart, W. 8n6, 27n40, 99, 142n37, 163n51, 207n7, 223n5 “Goffart challenge” 99, 106 Goldberg, E. 19n2, 63n82, 83n10, 102, 352n129 Gonja people 103 Goody, J. 103 Görschen 124 Gregory VII, Pope 173, 211 Grenier, A. 266n14 greywacke 283 Grierson, P. 242 Grimm, P. 269 Guibert of Nogent 227 Gundlach, W. 52 Guntram (king) 11–12, 27 Gunzelin (king) 115, 121 Hahne, O. 269 Hainer, K. 52 Hakim II, Caliph al- 156 Halsall, G. 39n111, 254 Harari, Y. N. 225n16 Hardt, M. 335n34 Harkel, Letty ten (archaeologist) 258 Harold, King 316, 318, 321, 324 Harper-Bill, C. 20n6 Hatto, Archbishop 59 Hauptburg (citadel) 270–4, 277 Hauteville, Bohemond de (prince) 226, 229–30, 232–5 Hawkwood, J. 244 Head, T. 224n6 Heath, P. 315 Hedenstierna-Jonson, C. 259–60 Henning, J. 94n78, 337n48 Henry I, King 56–7, 73–6, 78–82, 84–6, 109, 247, 263, 267n16, 296, 300, 302, 329–57, 335n32, 349n108, 356n145; accession 70; bring people into modernity 204; campaign of

362

INDEX

929 333; campaigns in Poland and Bohemia 111; capture of Brandenburg 334–41; capture of Gana 341–7; death 92; field forces 59; fortify the frontier settlement at Meißen 109; history 331–3; logistics for the campaign 353–6; military campaigns 94; scholarly debate 52; success 47; surrender of Prague 348–52 Henry II, King 108–11, 115–16, 124, 125n151, 127, 128n165, 129, 263, 345, 345n92, 345n94; campaign against Polish duke 122; eastern policies 112; fortifications 74; military undertakings in Poland and Bohemia 112; milites 114, 121; substantial army 125 Henry III, King 171, 174n15, 198, 218 Henry IV, King 114, 116, 171–3, 174n16, 179–80, 182–8, 190–2, 198, 210–15, 211n53, 212–13, 217, 218–20, 330–1; aristocratic supporters 195; armies fighting against 176; army 189, 216, 346; fortifications 215; inspection of the fortress of Lüneburg 180; military operations 175; milites 57n54, 123; moral degradation of the magnates 194; putative tyranny 193; reputation damage 177; seize control over fortress at Lüneburg 195; sieges during 221–2 Henry V, King 216–17 Henry of Laach (Count) 189, 192 Henry of Lower Lotharingia (duke) 216 Henry of Schweinfurt (margrave) 117, 121, 123 Henry the Fowler 46 here-paed 255 Heriold (Danish prince) 65 Herlihy, D. 140n30 Hermann of Salm (duke) 173 Hermann of Swabia (duke) 116, 122 Herodotus 227 Herren, M. 162 Herrmann, J. 337n45–46 Hildagsburg 273n44, 274, 277n65; construction costs 291; engineering work 296; fortifications 276; fortress construction 269–71; Hauptburg 278; processing of stone 290; trees 279; Vorburg 277 Hilduin, Abbot 31 Hincmar of Rheims 65–6 Hintze, O. 201

Historia 153 Histories (Gregory of Tours) 8, 9, 11, 11n1614 Histories (Nithard) 22–3, 25–6, 34, 39, 42–3 History (Frigeridus) 9 Hodo (margrave) 114, 125 Hofmann, H. H. 275n55, 346n98 Hollister, C. W. 21n10, 71n123, 72n127, 73n130, 136n13, 150n83, 256n25 Holmquist, L. 259 Holtzmann, R. 52–4, 81n1, 109, 109n42, 110n50 Holy Land 156 Holy Sepulcher 156 Hömberg, A. K. 77n146 honores 323 Horace (Flaccus) 25 Hugh of Lusignan 160 Hülle, W. 330n5, 343 Hungarians 331, 333 Hungary 198, 214 Iberia 250 Idea of History, The (Collingwood) 1, 102n14 imitatio imperii 148, 251, 257 infantry phalanx 200 Isidore of Seville (scholar) 7, 7n1, 25, 25n25, 153n8, 154, 209–10n48, 227 Italy 46–7, 76, 82, 87, 143, 312 Jerusalem 156, 224, 226, 228 Jewish Antiquities 227 Johnson, S. 266n122 John Tzimiskes 116, 119, 130 Jugurthine war 212 Kedar, B. Z. 224n8 Keller, H. 20n5, 103–4, 103n22, 104n23, 161, 204, 204n25, 209, 241n8 Klappauf, L. 80n153 knighthood/knights (Ritterschaft) 47–9, 147, 202 knightly class (Ritterstand) 201, 205n32 knightly vassals 205 Koch, R. 271n41 Koller, H. 84 Köpke, R. 52 Kortüm, H.-H. 199n8 Küas, H. 284 Künkel 184 Kurze, F. 27, 62 kyrie eleyson 192

363

INDEX

Lair, J. 132n2 Lampert of Hersfeld (chronicler) 174n18, 176, 208–9, 211–19 Landes, R. 224n6 “Landeskrone” 278 Landfrieden 201, 206 Landscapes of Defence in Early Medieval Europe (Baker) 242 Larson, A. M. 279n73 Late Roman Fortifications (Johnson) 266n12 Latin-Romance Aquitanian 160 Laudage, J. 55n43, 100n7 Lebuda, G. 329n2 Lechfeld campaign 82 Lederer-Herkenhoff, N. 111 Lefebvre, R. 289n103 Legrand-Girard, E. 279n72 Leighton, A. C. 347n100 Leipzig–Grimma–Oschatz 346 Leo, Bishop 159 Leverhulme Trust 242 Leyser, K. 46–8, 47n3–4, 49n9, 50–1n14, 50–62, 57n54, 62n54, 62n79, 68, 68–9n116, 77, 82n5, 83, 83n7, 127n161, 204, 204n26, 209; identification of Ermold 63; Saxons military standing 64–5; secondary argument 67; view of Louis the Younger 69 Life of Alfred 248 limes Saxonicum 247 Limes Sorabicus 332 limestone 289n102 Limoges, cum Lemovicinis 168 Lippelt, H. 109–11 liti 355 Liudolf (duke) 46, 46n2, 87, 89, 91, 114 Liudprand of Cremona 110 Liuthar (margrave) 120 Liutizi 116 Livy 227 logistics: Saxon Wars 189–91; logistics for campaign, Henry I 353–6 loricae 115, 128–9, 150 loricati 128 Lothair I 23–4, 26–8, 30–3, 36–8, 40–2, 66 Lothair II 34 Lotharingia 47, 292, 333 Louis IV, King 86–7 Louis the Child 331 Louis the German 28–30, 34, 34n82, 44, 64, 66, 294, 349

Louis the Pious (emperor) 22, 24, 28–33, 35, 38, 41–2, 42n121, 56n48, 62–3, 65, 259, 297; consumers of knowledge 44; losses suffered 26; put down revolt by the Aquitanians 66; Saxon military forces 65 Louis the Younger 64, 66–7 Low Countries 258 Ludat, H. 264n5 Lugdunum Convenarum (modern SaintBertrand-de-Comminges) 15 Lupus of Ferrières 25 Luttwak, E. 12n19, 255 Magdeburg Cathedral 283n82 Magnus (duke) 191 Magyars 47, 57–8, 73–5, 77, 136, 284 mansi/mansus 144, 292–4, 294n117, 295, 355, 355n142 Marcellinus, A. 7 Martel, G. 307 Martial, St., in Aquitaine 155, 159 Martial of Limoges, St. 155 Mary, St., at Frankfurt 294 Mathilda (princess) 53 McKitterick, R. 25n29, 44n134, 102, 239n1 McLaughlin, M. 175n21 Medieval Latin 161 Meginhard (Count) 62, 64, 69–70 Meißen 74, 115–16, 180, 185 mentalité 153, 324 mercatores 216, 219–20 Merovingians 48, 163, 220, 266–7; king 17; military organization 8; period 136 Merseburg 110–11 Mersiowsky, M. 102 Michael of Regensburg, Bishop 119 Middle East 155–6, 224 Middle Saxon Wessex 251 Miesco I of Poland (duke) 114, 125 “Migration Period” 245 military: aristocracy 148; campaigns 67, 167, 213–14; colony 149; commanders 151, 157, 181n59, 182, 186–7, 220; communications 39–41; conflict 220; customs and traditions 164; decision-making 24; development 47; education 148; effectiveness 65–71, 209; equipment 150, 218; establishment 61; galea (helmet) 150; genius 185; historian 139, 163; history 98n1, 199n8, 207, 239n1; households 114n69,

364

INDEX

146n61, 148–50, 164, 166, 179–81, 228, 244, 256; installations 241; institutions 136n12, 199; instructor 148; intelligence 39–41, 182–3, 182n60; levies 220; marginalization 213; obligations 256; obsolescence 16; operations 18, 144, 146, 150, 157, 164, 175–6, 196, 214, 217, 225–6; organization 135n7; planning 240, 309; purposes 294; reality 308; revolution 49, 51–62; Leyser’s evidence for 51–62; strategy 48, 157, 166–7; strength in Western Europe 134; technology 199, 313; violence 8 military organization 26–9, 50–1, 72–3, 75, 147, 151–2, 197, 199–200, 203, 220, 254; Ademar of Chabannes 163–6; Anglo-Saxon 256; concomitant effect 204; domus suae militiae 147; Dudo of St. Quentin 134–5, 137; early medieval 71–5; elite force (optimi) 123; elite troops (qualitate sua optimus) 129; execrata Alemmanorum turba 122; exercitus 11, 57, 76–7, 118n102, 125–6, 129, 144, 214–15; familia/familiae 146, 355, 355n42; fortis manus 77; fortium militum manus 59–60; Frankish 200–1, 204; ignobilis vulgus 73; inermes 73, 137–8; inerme vulgus 137; insuetus 57, 74, 77; levies, to feudal knights 200–6; Limousin levies 168; local levies, Dudo of St. Quentin 135–41; medieval 257; militia domus 147, 149; militia episcopalis 93; militia forces 21, 136, 176–8; militia troops, in expeditionary armies 178–9; of non-milites 217–19; obsequia 72, 73, 76–7, 113, 146, 256; ordinis secondi sive tertii partis utriusque milites 179; plebei 213–14, 219; plebei ac rustici 213; plebis innumera multitudo 26; plebs 31–2, 32n69, 151; praesentales 134, 150; princeps 144, 147–9; princeps domus 147; princeps militiae 147; principes 92, 213; professional soldiers 59–60; royal military household, Saxon Wars 179–80; rustici 139–40, 151, 166, 177, 212–14, 216–20; Saxon 74; Saxon Wars 176; select levies 141–5, 164; standing armies, Dudo of St. Quentin 145–51; territorii pagenses 138; tirones 148; tripartite military organization 151; vulgus 208, 212, 214, 219–20

military training 200; bellator intolerabilis 61; bellicarum rerum imperiti 177; disciplina 93; disciplina militari nudatum 73, 202; ineptus vulgus 212; moribus et studiis belli 148 milites 27n38, 56, 57n54, 59–60, 72, 76–7, 107, 107n34, 113n67, 119n106, 128n165, 129n170, 130n173, 138, 144, 147–51, 166, 166n62176, 179, 206, 209–10, 215n72, 217, 219, 316, 321, 347; bonus miles 119; deployment of 119–26; early medieval military history 113–30; equipment of 126–30; miles armatus 83, 118, 122; miles egregious 117; miles lectissimus 212; miles privatus 214; milites plebei partis 179; milites regis 126, 215; Norman miles/milites 309, 312; optimi milites 125–6; ordinis secondi sive tertii partis utriusque milites 179; Salian Germany 214–17 Mistislav (Abodrite ruler) 122 modii 292, 294, 355 morale, Saxon Wars 191–5 Moribus et Actis primorum Normanniae ducum, De 132, 136, 144, 147, 228 Morrison, K. 54–5, 54n36, 92 Mount Cenis 12 Mummolus, E. 12–15, 14n25, 14n27 munera 58 municio 122 Muslim rulers 156 Muslims 47, 168–9, 200, 224, 231, 233–4, 250–1, 253–4 Mystery of the Bayeux Tapestry, The (Bernstein) 304 Nacco (prince) 85 naves 169 Nelson, J. 20n6, 22n14, 23–5, 24n19–22, 25n30, 42n121, 66n98 Neustria 12 Nickel, E. 265–6n11 Nigellus, E. 28, 28n43–4 Nithard (military historian) 19–45, 22n13, 22n15, 23n16–18, 32n71, 41n119, 42n122, 66; armies/force 29–33; battleavoiding campaign strategies 33–7; battle-seeking campaign strategies 33–7; causa scribendi 23–5; conduct of war 25–6; evaluation 22–3; Histories 22–3, 25–6, 34, 39, 42–3; military

365

INDEX

communications 39–41; military intelligence 39–41; military organization 26–9; propaganda 42–4; seeking battle as moral problem 37–9; text 22–5 Nomenoi (Briton leader) 37 Norman ducal court 154 Norman military technology 313 Notre-Dame of Bayeux 324 nuncius 40, 122, 314 Odo, Bishop 303–5, 308, 313–17, 318–26 Odo II of Blois (count) 322 Olausson, M. 260 Old Castile 253 Onasander (philosopher) 9 optimus pastor 119 orationes 162 Osbern, W. F. 320–1 Otbert 216 Otto I, King 46–7, 51–3, 81–2, 84–5, 85n15, 87, 89, 91–2, 94, 110, 114, 117–18, 263, 296–7, 330, 349–50, 35n139 Otto II, King 82, 114, 263 Otto III, King 114–15, 118–20, 217 Ottonian 101n13, 206, 335n35, 339; barges 286; builders 268; dynasty 53, 57, 108; engineers 272; fortifications 265, 268n21, 344; fortress 276; kingdom 119; kings 83, 218, 220, 267, 269n22, 293, 295–6, 297; military history 199; royal rule 204; rulers 143 Ottonian Germany 128n169, 296 Ottonian Warfare 81–94; implications 92–4; siege warfare 84–92 Otto of Northeim (duke) 172, 178, 186–8, 195, 211, 214 Otto the Great 247 pagenses 140, 144 pagus 26, 135, 137–8, 141–2, 144, 145, 203, 297–8 Palgrave, F. 132n1 Panzerreiter 82, 83, 94 Parry, M. 102 parti pris 14n25, 24, 37, 55–6, 64, 69, 98, 103, 106, 110–11, 163, 175–6, 183, 193, 205, 210, 230, 242 Partner, N. 133–4n5 patricius 12–15 Patze, H. 23, 23n18 Pavia 151

peace treaty of 924 78–80 Peinhardt, H. 174n17 Pippin I, King 28–9, 31, 35, 39, 142, 245 Pippin II, King 35, 38–40 Pippin III, King 203 Peter, A. 157 Phokas, Emperor Nikephoros 116 Pietzcker, F. 35n85 pilgrims 156 Plessix, H. 279n72 Poitou 160, 165 Poland 111–12, 198, 213 Powers, J. F. 144n46, 254 praepositus 166 Prague 125, 348–52 Pressburg 331n10 Prinz,F. 151n90 Procopius 7 propaganda 42–4 Prudentius of Troyes 65, 310n26 Psalm 82 191 Psychomachia (Prudentius) 310n26 Purton, P. 15n29, 49n10, 195n32 quadrivium 226 Rabanus Maurus 34, 34n83, 68n112 Radig, W. 342–3 Ralph of Caen (author) 223–35; battle of Dorylaeum 229; Gesta Tancredi 225–7; philosophy of history 227–9; potential contribution 230–5; writing medieval military history 223–5 Raoul, King 333 Ray, R. 103 reges Francorum 76 Regino of Prüm (monk) 67–8, 68n114, 73–4, 202 regna 11, 76, 88, 134, 142, 243 regnum Francorum 13, 19, 21, 26, 30–1, 40, 43, 114n69, 142, 163, 257, 260 Renaux, A. 135n6 Renze (Ronne) 269 Rerum Gestarum Libri (Marcellinus) 7 Res gestae 52–4, 61n72–5, 73n133, 78n149, 82–4, 82n3, 86 Res gestae Saxonicae 51, 81 res publica 125 Reuter, T. 20–1, 20n6, 21n8, 46n1, 71n123, 83, 83n8–9 Revindicatio 172 Revindikationspolitik 171

366

INDEX

Reynolds, A. 242–4, 247 Reynolds, S. 48–50, 49n8, 249, 258 rhetorical plausibility 93n77, 155–8, 162 Rhineland 210 Ricdag, M. 116 Richard of Evreux 320 Richer of Rheims 154 Ringwalburgen 259 Robert, King 168, 322 Robert of Eu 320 Robert of Mortain 308, 320 Robert of Normandy (duke) 229, 232–4 Robert the Monk 141n32, 227 Robinson, I. S. 171n4 Rochemaux 169 Rodan (duke) 13 Roger II of Montgomery 320 Roger of Beaumont 320 Rogers, C. J. 186n84 rois fanéants 245 Rollo 137, 144 Roman Empire 77–8, 92, 135–6, 146, 149, 164–5, 255, 258, 309 Roman fortifications 266, 309; see also fortifications Roman gate defences 266n12 Romania 8 Roßtal 89, 89n51 Rouenais 145 Rubenstein, J. 224n12, 225n19 Rudolf of Rheinfelden (duke) 173, 183, 189, 191, 197, 207–8, 210–11, 218 Ruotger (archbishop) 51 ruse de guerre 60, 177, 184, 197 Ruttkay, A. 351n127 Sachkritik 104–5, 105n27, 107, 224, 262, 330 Sachsen-Anhalt 120 Saint-Jean-D’Angély 166 Saint Peter the Apostle 306 Salian Franks 10 Salian kings 218 Sánchez-Albornoz, C. 254 Saracens 136 Sarantis, A. 12n19 Saxon 28, 177, 180, 192, 349; as mediocre soldiers 62–5; military effectiveness 65–71; military organization 74, 75n143; opponents 179; proposals 179; question 71–5 Saxonicum Bellum 207, 207n39

Saxon Military Revolution 46–77; early medieval military organization 71–5; Leyser’s evidence for military revolution 51–62; Saxon military effectiveness 65–71; Saxon question 71–5; Saxons as mediocre soldiers 62–5 Saxon Switzerland 348 Saxon Wars 171–97; battlefield operations 181–2; battlefield tactics 185–6; Bellum Saxonicum 175–6; command and control 186–9; control of information 182–3; logistics 189–91; military households of magnates 180–1; military intelligence 182–3; military organization 176; militia forces 176–8; militia troops in expeditionary armies 178–9; morale 191–5; professional fighting men 179; role of fortresses in warfare 195–7; royal military household 179–80; scholarly treatment 173–5; scouting 183–5; sieges during Henry IV 221–2 Scandinavia 243, 258, 260 Scharff, T. 24n24, 55n42, 106n33, 176n24 Scherff, B. 198 Schlesinger, W. 241, 332n15, 334n30 Schmid, K. 241n8 Schmid, P. 356n146 Schoenfeld, E. J. 81n4 schola 93, 148 Schultheiß, H. W. 269 Schultz, W. 269 Sclaomir (Obodrite duke) 65–6 scouting, Saxon Wars 183–5 Second World War 94, 107n34, 199, 202, 239 Sedulius Scotus 58n58 seeking battle as moral problem, Nithard (military historian) 37–9 Selle-Hosbach, K. 12n20 Sergestus 322 sermo humilis 160, 162 servi 142, 355 Sicily 312 Sidonius Appolinaris the Younger 141 siege engine: instrumenta belli 123; machinae 85, 90; multae machinae 87 siege warfare 15, 124n139, 168 Siegfried (count) 125–7 Sigismund (king) 17 signa 161 Simek, R. 19n2

367

INDEX

Slavs 264 Slopan 117 Society Organized for War: The Iberian Municipal Militias in the Central Middle Ages, 1000–1284, A (Powers) 144n46, 254 Sorbian March 332 Spain 11, 76, 143, 249 Sproemberg, H. 84 Stieldorf, A. 265n6 Stigand, Archbishop 317 Stimming, M. 171n3 Stoignew (prince) 85 Stratagemata (Frontinus) 92, 262 Struve, T. 173n11 Swabia 184, 350 Swabian armies 179 Swabians 28–9, 183 Sweyn II, King 180 Ten Books of History (Gregory of Tours) 8 Theoderic II of Upper Lotharingia (duke) 218–19 Theodo, Bishop 350 Theudebert I, King 14 Thiepald, Bishop 217 Thietmar of Merseburg 60–1, 74–5, 85, 93, 105, 106n32, 107–8, 108n35, 109n45, 110, 112n62, 113n65, 115, 118n103, 118n105, 120n113, 124n139, 332, 340; Chronicon 112–30, 112n61, 114n70, 119n107, 125n151, 129n170, 340n58; Germanness 109; parti pris 111 Thrasco (Obodrite duke) 65 Thucydides 227 Thuringia 182, 332, 352 Toch, M. 84 Toubert, P. 140n31 Trillmich, W. 109 Ulrich of Augsburg 88 Universal History 248 Urban II, Pope 140, 224 urbs/urbes 10, 11, 13, 15–16, 61, 73–4, 73n133, 123–4, 138, 151, 246, 309, 338 Uslar, R. von 263n1 Utrecht 115 Utrecht ship 286n92

Vegetius 15n30, 33, 33n79, 67–8n112, 92, 157, 157n19, 170, 262, 308n20, 335n36, 336, 336n36; velites 67n112 Victory in the East: A Military History of the First Crusade (France) 223 Viking Age 255, 261 Vikings 73, 136–41, 141n33, 144, 202, 250, 259, 339 Virgil 322 Viscount Hugh of Avranches 320 Vita Heinrici 209, 216 Vita Hludowici (the Astronomer) 28, 28n46 Vita Karoli (Einhard) 25, 63n81, 159 Vitruvius 246, 262, 267, 271, 271n33, 288n99 vluchtburgen 259 Völkerwanderungszeit 245 Vorburg 270, 274, 282, 342 Vratislav (duke) 331 Vratislav II of Bohemia (duke) 215 Vyserad 351–2 Waffenrecht 201 Waitz, G. 294n116 Walgrenses 138, 144 Walker, H. E. 303n1 Wallace-Hadrill, J. M. 92n70 Walthard, Archbishop 75 Warner, D. 111 Wattenbach, W. 81n1, 109, 109n41 Weidemann, M. 8n4 Weinfurter, S. 112n61 Welf of Bavaria (duke) 173, 218 Wenceslaus (duke) 329, 351, 353 Werner, Archbishop 180–1, 187, 196 Werner, Bishop 108, 115, 187 Werner, K. F. 15n13, 142–3n41 Western Europe 134 West Francia 47 White, L. T., Jr. 48 Wido 26 Widukind (monk) 51–62, 51n17–18, 54n40, 56n45–6, 57n57, 58n57, 60n69, 81n1–2, 82n5, 91–2, 100n7, 103–4, 110, 296, 330, 330n6, 338, 341, 344, 346–7, 350; centrality of sieges 94; contemporary affairs 81; Henry’s reform 74; manus militum 72; military operations 82; Otto’s failure at Mainz 88; Res gestae 61n72–5, 73n133, 78n149, 83, 85n15, 85n24, 86, 86n29–30,

368

INDEX

88n42, 88n44, 89n48, 89n53, 93, 339n52, 341n59 William of Aquitaine (duke) 53, 155–62, 166–7, 169–70, 187, 248, 303, 307, 313–18, 321 William of Jumièges 228 William of Kamburg (Count) 181 William of Mainz 92 William of Poitiers 228, 307n16, 308, 309, 310–11, 315, 317n59 Williams, G. 256–7

William the Conqueror 197, 226, 228, 303, 308, 312, 322 Wollasch, J. 75n142 Würzburg 183, 207–8, 210, 220 Yorke, B. 251–2 Yugoslavia 102 Yver, J. 164n53 Zaban (duke) 13 Ziolkowski, J. 161

369