Journal of Medieval Military History. Volume XV. Strategies [15] 1783272570, 9781783272570

Citation preview

CONTENTS BERNARD S. BACHRACH

Later Roman Grand Strategy: The Fortifcation of the Urbes of Gaul

LEIF INGE REE PETERSEN

In Search of Equilibrium: Byzantium and the Northern Barbarians, 400–800

RICHARD ABELS MATTHEW BENNETT LUIS GARCÍA-GUIJARRO RAMOS MANUEL ROJAS GABRIEL

Evolving English Strategies during the Viking Wars Norman Conquests: A Strategy for World Domination? The Papacy and the Political Consolidation of the Catalan Counties, c.1060–1100: A Case Study in Political Strategy Alfonso VII of León-Castile in Face of the Reformulation of Power in al-Andalus, 1145–1157: An Essay in Strategic Logic

MARIA DOLORES GARCÍA OLIVA

The Treaties between the Kings of León and the Almohads within the Leonese Expansion Strategy, 1157–1230

JOHN GILLINGHAM

A Strategy of Total War? Henry of Livonia and the Conquest of Estonia, 1208–1227

JOHN FRANCE

The English Longbow, War, and Administration

Cover: Coll. Bibliothèques municipales de Chambéry, MS13, fo. 105r (circa 1170–1190). The Italian artist who illuminated this copy of Gratian’s Decretum beautifully captured the deadly seriousness of medieval warfare – for townsmen as well as knights. Image used by generous permission of the Bibliothèques municipales de Chambéry.

JOURNAL OF MEDIEVAL MILITARY HISTORY

The Journal’s hallmark is broad chronological, geographic and thematic coverage of the subject, bearing witness to ‘the range and richness of scholarship on medieval warfare, military institutions, and cultures of conflict that characterises the field’.  HISTORY

XV

J ournal of

MEDIEVAL MILITARY HISTORY

PETERSEN, ROJAS (editors)

·   XV · Edited by Leif Inge Ree Petersen and Manuel Rojas Gabriel

JOURNAL OF

Medieval Military History Volume XV

JOURNAL OF MEDIEVAL MILITARY HISTORY

Editors Clifford J. Rogers Kelly DeVries John France ISSN 1477–545X

The Journal, an annual publication of De re militari: The Society for Medieval Military History, covers medieval warfare in the broadest possible terms, both chronologically and thematically. It aims to encompass topics ranging from traditional studies of the strategic and tactical conduct of war, to explorations of the martial aspects of chivalric culture and mentalité, examinations of the development of military technology, and prosopographical treatments of the composition of medieval armies. Editions of previously unpublished documents of significance to the field are included. The Journal also seeks to foster debate on key disputed aspects of medieval military history. The editors welcome submissions to the Journal, which should be formatted in accordance with the style-sheet provided on De re militari’s website (www.­ deremilitari.org), and sent electronically to the editor specified there.

JOURNAL OF

Medieval Military History Volume XV Strategies

Edited by LEIF INGE REE PETERSEN MANUEL ROJAS GABRIEL

THE BOYDELL PRESS

© Contributors 2017 All Rights Reserved. Except as permitted under current legislation no part of this work may be photocopied, stored in a retrieval system, published, performed in public, adapted, broadcast, transmitted, recorded or reproduced in any form or by any means, without the prior permission of the copyright owner First published 2017 The Boydell Press, Woodbridge ISBN 978–1–78327–257–0 The Boydell Press is an imprint of Boydell & Brewer Ltd PO Box 9, Woodbridge, Suffolk IP12 3DF, UK and of Boydell & Brewer Inc. 668 Mt Hope Avenue, Rochester, NY 14620–2731, USA website: www.boydellandbrewer.com

A CIP catalogue record for this book is available from the British Library The publisher has no responsibility for the continued existence or accuracy of URLs for external or third-party internet websites referred to in this book, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate This publication is printed on acid-free paper

Typeset by

www.thewordservice.com

To Our Parents

The publishers acknowledge with gratitude the generosity of De Re Militari in providing financial support towards the production costs of this volume

Contents

List of Illustrations Preface

viii ix

1. Later Roman Grand Strategy: The Fortification of the urbes of Gaul Bernard S. Bachrach

3

2. In Search of Equilibrium: Byzantium and the Northern Barbarians, 400–800 Leif Inge Ree Petersen

37

3. Evolving English Strategies during the Viking Wars Richard Abels

69

4. Norman Conquests: A Strategy for World Domination? Matthew Bennett

91

5. The Papacy and the Political Consolidation of the Catalan Counties, 103 c. 1060–1100: A Case Study in Political Strategy Luis García-Guijarro Ramos 6. Alfonso VII of León-Castile in Face of the Reformulation of Power 119 in al-Andalus, 1145–1157: An Essay in Strategic Logic Manuel Rojas Gabriel 7. The Treaties between the Kings of León and the Almohads within the Leonese Expansion Strategy, 1157–1230 María Dolores García Oliva

151

8. A Strategy of Total War? Henry of Livonia and the Conquest of Estonia, 1208–1227 John Gillingham

187

9. The English Longbow, War, and Administration John France

215

List of Contributors

227

Maps and Illustrations

Later Roman Grand Strategy: The Fortification of the urbes of Gaul Map 1. Late Roman Cities in Gaul

2

In Search of Equilibrium: Byzantium and the Northern Barbarians, 400–800 Map 1. Byzantium and the North, 400-800 AD

36

Evolving English Strategies during the Viking Wars Map 1. Anglo-Saxon England under Alfred the Great (based on map 5 Alfred the Great: War, Kingship and Culture in Anglo-Saxon England, Richard Abels, 2013 Taylor & Francis, pp. 354-5)

68

Alfonso VII of León-Castile in Face of the Reformulation of Power in al-Andalus, 1145–1157: An Essay in Strategic Logic Map 1. The Campaigns of Alfonso VII

118

The Treaties between the Kings of León and the Almohads within the Leonese Expansion Strategy, 1157–1230 Map 1. The Iberian Peninsula in 1157

150

A Strategy of Total War? Henry of Livonia and the Conquest of Estonia, 1208–1227 Map 1. The Conquest of Estonia, 1208-1227

186

The English Longbow, War, and Administration Figure 1. Marginalia drawing of a Welsh archer from the Littere Wallie, c. 1282–92 (Kew, by permission of The National Archives, E 36/274, f. 356r). 219 Figure 2. Archer depicted in a thirteenth-century English bestiary, c. 1225–50 (London, The British Library Board, Harley MS 4751, f. 14v). 2 23

Preface

The topic of medieval strategy has aroused lively debate among military historians in recent years, especially whether medieval military history was characterized by a particular type of strategy, be it Grand, Vegetian or Battle-seeking. This volume, based on a conference organized by Manuel Rojas Gabriel at Cáceres in 2009 and rounded out by specially commissioned papers, brings together many of the pre-eminent military historians active today in order to examine a number of case studies that display the complexity and diversity of strategic realities, as well as exploring new models and methodological avenues in evaluating medieval strategies. The cases span the late Roman Empire through the High Middle Ages, from the Baltic and the British Isles to Iberia and the Crusader States. Bachrach examines the strategy behind the building of a defensive network of fortifications in Gaul, from the perspective of a late Roman government that was not hard pressed by barbarian invasions across the Rhine, as often argued, but did have to take extensive measures against sea-borne raiders whose bases were beyond Roman reach. Petersen follows the long-term survival and effectiveness of late Roman and early Byzantine methods of client management on the northern frontiers that persisted well into the middle Byzantine period. Abels demonstrates how the lack of coherent, long-term strategic planning led to failure under Aethelred II, despite immense military, organizational and economic efforts, whereas Alfred the Great had succeeded over a century earlier under similar circumstances by ensuring that each of his efforts reinforced the other. Bennett argues that Normans were masterful in their consistent application of the ideological component, alongside effective use of diplomacy, family connections, and technology such as ships and fortifications in what amounted to a strategic vision that characterized western Christian expansion in the High Middle Ages. Similarly, García-Guijarro demonstrates that the ideological support provided by the reform papacy was essential to the political coherence of the Catalan counties, while at the same time strengthening the papacy. Rojas and García examine the wildly fluctuating strategic needs of the Christian Iberian kingdoms in the 12th century, which shifted from truce with Muslim powers and conflict with Christian neighbors and back again to expansion at the expense of al-Andalus several times over, depending on power politics rather than ideology; however, behind the shifting alliances, the Iberian kingdoms always worked to ensure further opportunities of expansion. Gill-

x Preface

ingham examines the strategic economy of warfare in the Baltic, showing how well-attested strategies focusing on raiding and taking captives dominated the Baltic crusades, providing a model for understanding similar types of war in earlier periods. Finally, France demonstrates how strategy could form tactics, as the strategic needs and priorities of the English kings stimulated the recruitment of particular types of bowmen through the Middle Ages. The volume editors wish to thank the contributors for their stimulating papers, Stephen Morillo for his helpful feedback, Caroline Palmer, Rebecca Cribb and the whole production team at Boydell and Brewer for their patience and expert guidance, Francis Eaves for effective copyediting, Cath D’Alton for excellent maps and finally Journal of Medieval Military History editors, Clifford J. Rogers, Kelly DeVries and John France for their assistance and encouragement throughout all stages of the process. Leif Inge Ree Petersen Trondheim, May 2017

Map 1: Late Roman Cities in Gaul

1 Later Roman Grand Strategy: The Fortification of the urbes of Gaul Bernard S. Bachrach

There is widespread agreement that during the last quarter of the third century the Roman imperial government initiated what ultimately would be a vast program of military construction throughout large parts of the Empire in the West and most particularly in Gaul. These efforts resulted in a great many previously unwalled or poorly defended urbes being converted at great expense and with considerable social disruption into fortress cities with extensive defenses and supporting construction, with many walls reaching in excess of ten meters in height.1 Regarding Gaul, the focus of this study, there appears to be a rough consensus that “It was probably Probus (276–80) who instituted the policy of fortifying the western Continental urban centers.”2 These military construction 1

2

The best introduction remains Stephen Johnson, Late Roman Fortifications (Totowa, NJ, 1983); and regarding costs, see Bernard S. Bachrach, “The Fortification of Gaul and the Economy of the Third and Fourth Centuries,” Journal of Late Antiquity 3 (2010), 38–64. In addition much useful information is to be found in Michael Mackensen, “Late Roman Fortifications and Building Programmes in the Province of Raetia: The Evidence of Recent Excavations and Some New Reflections,” in Roman Germany: Studies in Cultural Interaction, ed. J. Creighton and R. Wilson (Portsmouth, RI, 1999), pp. 199–244; Marc Heijmans, “La mise en défense de la Gaule méridionale aux IVe–VIe s,” Gallia 63 (2006), 59–74; Simon Loseby, “Urban Failures in Late Antique Gaul,” in Towns in Decline AD 100–1600, ed. T. R. Slater (Aldershot, 2000), pp. 72–95 at 76–81; H. von Petrikovits, “Fortifications in the North-Western Roman Empire from the Third to the Fifth centuries A.D.,” Journal of Roman Studies 61 (1971), 178–218, provides special attention to Britain, Gaul, and the Rhineland, while more on Gaul is to be found in R. M. Butler, “Late Roman Town Walls in Gaul,” Archaeological Journal ll6 (1959), 25–50; Albert Grenier, Manuel d’archéologie gallo-romaine (Paris, 1934), vol. 6:1; Carlrichard Brühl, Palatium und civitas: Studien zur Profantopographie spätantiker civitates vom 3. bis zum 13. Jahrhundert, Vol. 1, Gallien (Cologne and Vienna, 1975); idem, Vol 2, Belgica I, beide Germanien und Raetia II (Cologne and Vienna, 1990); and Adrien Blanchet, Les enceintes romaines de la Gaule (Paris, 1897), pp. 13–219. I have not included here a discussion of the construction or reconstruction of a vast number of lesser fortifications undertaken during this period. See L’Architecture de la Gaule romaine. Les fortifications militaires, Documents d’archéologie française 100, directed by Michel Reddé with the aid of Raymond Brulet, Rudolf Fellmann, Jan-Kees Haalebos, and Siegmar von Schnurbein (Paris and Bordeaux, 2006). For the quotation, see John F. Drinkwater, The Gallic Empire: Separatism and Continuity in the North-Western Provinces of the Roman Empire, A.D. 260–274, p. 44; however, on pp. 222–26 Drinkwater provides a much more complicated scenario which will be discussed below.

4

Bernard S. Bachrach

efforts were carried on throughout the fourth, and in some places can be identified even into the early fifth, century.3 The building process continued until some 85 per cent, at least, of the 125 previously largely unfortified, or as contemporaries evaluated the situation, inadequately defended, cities located within the borders of the “hexagon” of Gaul had been converted into urban fortresses.4 The purpose of this paper is to investigate what would appear to have been the military thinking that undergirded the effort focused on the fortification of urban sites in Gaul, and to ascertain whether this lengthy and very expensive process of construction merits consideration as a “grand strategy.”5 Why Fortify the Cities of Gaul? For a considerable time, scholars have argued that the third century and especially its second half was rent by a massive “crisis” that threatened the very existence of the Empire, and that this was true especially for Gaul.6 Today, the idea of a 3

4 5

6

Johnson, Late Roman Fortifications, pp. 82–117, provides the basic information regarding the walls. However, still useful are Paul-Albert Février, “Permanence et héritages de l’antiquité dans la topographie des villes de l’occident durant le haut moyen âge,” Settimane di studio del Centro italiano di studi sull’alto medioevo 1 (1973), 41–138; Jean Hubert, “Evolution de la topographie de l’aspect des villes de gaule du Ve au Xe siècle,” Settimane di studio del Centro italiano di studi sull’alto medioevo 6 (1959), 529–58; Blanchet, Les enceintes romaines; and Grenier, Manuel d’archéologie, 281–361. Regarding the statistics of the Notitia Galliarum, see Loseby, “Urban Failures,” pp. 76–81. With regard to planning, see Everett L. Wheeler, “Methodological Limits and the Mirage of Roman Strategy,” Journal of Military History 57 (1993), 7–41; 215–40, esp. 216–17; and followed in much greater detail by Martijn Nicasie, Twilight of Empire: The Roman Army from the Reign of Diocletian until the Battle of Adrianople (Amsterdam, 1998), pp. 119–84, where all types of planning from long-term to small campaigns are examined. These two works give considerable support to some of the views of Edward N. Luttwak, The Grand Strategy of the Roman Empire: From the First Century A.D. to the Third, rev. ed. (Baltimore, 2016), pp. 126–90, and esp. pp. 132–34. Of considerable importance in this context is the rapidly developing field of landscape archaeology, which neither Luttwak nor his many critics would seem to have been able to apply. On this methodology, see 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. A. H. M. Jones, The Later Roman Empire, 284–602: A Social, Economic and Administrative Survey, 2 vols. (Norman, OK, 1964), 1:25–36, 1038–48: Jones was a strong believer in the “crisis,” and many historians and archaeologists are wont to accept his views. See, for example, the recent work of Simon Esmonde Cleary, The Roman West, A.D. 200–500: An Archaeological Study (Cambridge, 2013), pp. 18–32. For a maximalist interpretation of the threat posed by the barbarians and the damage done by them, see John Haywood, Dark Age Naval Power: A Reassessment of Frankish and Anglo-Saxon Seafaring Activity (New York, 1991). Wolf Liebeschuetz, “Was There a Crisis of the Third Century?,” in Crises and the Roman Empire. Proceedings of the Seventh Workshop of the International Workshop of the International Network, Impact of Empire, Nijmegen, June 20–24, 2006, ed. Olivier Hekster, Gerda de Kleijn, and Daniëlle Slootjes (Leiden, 2007), pp. 11–20, provides an excellent discussion of scholarly attacks on supporters of the crisis hypothesis, while defending its value as a scholarly construct. Much the same can be said of Lukas de Blois, “The Crisis of the Third Century A.D. in the Roman Empire: A Modern Myth,” in The Transformation of Economic Life



Later Roman Fortification of Gaul

5

third-century crisis is no longer as prevalent or as strongly advocated as once was the case.7 Nevertheless, the supposed destructiveness of trans-Rhenish barbarian invasions remains a vital part of a model of third-century collapse. Luttwak, for example, in the most recent edition of his The Grand Strategy of the Roman Empire, maintains his previous views and claims that “the catastrophic invasions of the mid-third century” by trans-Rhenish barbarians are to be regarded as a crucial aspect of imperial decline if not, in fact, its fundamental cause.8 Recently, Esmonde Cleary has claimed that, “In the 250s, 260s, and 270s the literary sources tell of a series of more and more devastating incursions, deep into Gaul.”9 In fact, even scholars who doubt that there was a devastating “crisis”, as the phenomenon has traditionally been understood, in the third century assert that the northwest provinces – northern Gaul, Germany, and Raetia – suffered “much greater effects” from the barbarian invasions than other parts of the Empire. In short the claim, as argued by Witschel, is that “we must not fail to acknowledge that the great Germanic invasions of the 260s and 270s constituted a turning point for settlement structures deeper within Roman territory.”10 It will be shown below that barbarian operations west of the Rhine, and particularly in Gaul, during this period were not “catastrophic” in military terms. The literary sources do not, in fact, tell a story of a series of “more and more devastating incursions, deep into Gaul” during the 250s, 260s, and 270s; nor, for that matter, do they provide a record of a “chronic succession of incursions.”11 Rather, a critical reading of these texts makes clear that there

7

8 9 10 11

under the Roman Empire. Proceedings of the Second Workshop of the International Network, Impact of Empire (Roman Empire c. 200 B.C.–A.D. 476), Nottingham, July 4–7, 2001, ed. Lukas de Blois and John Rich (Amsterdam, 2002), pp. 204–17. For a lengthy list of scholars who reject the value of the crisis construct see, Liebeschuetz, “Was There a Crisis of the Third Century?,” pp. 11–20. Among those scholars who have vigorously questioned the crisis model, see, for example, Christian Witschel, Krise, Rezession, Stagnation? Der Westen des römischen Reiches im 3. Jahrhunderts n. Chr. (Frankfurt am Main, 1999); for a brief, but no less vigorous discussion of the weaknesses of the crisis model, see too Witschel, “Re-evaluating the Roman West in the 3rd c. A.D.,” Journal of Roman Archaeology 17 (2004), 251–81. Among those who have questioned the crisis model from a less scholarly position, see Michael Grant, The Climax of Rome (London, 1968). The continuing diminution of the model’s value is well illustrated by the following observation by John Nichols, “Mapping the Crisis of the Third Century,” in Crises and the Roman Empire, ed. Hekster, de Kleijn, and Slootjes (Leiden, 2007), pp. 431–37, at 431: “Concerning the crisis of the third century, I have no means of knowing whether there was one or not, or what sort of a crisis it may have been. Many things prevent knowledge including the obscurity of the subject and the brevity of human life.” Luttwak, The Grand Strategy, rev. ed., p. 192, for the quotation; p. 199 for the invasions of the third century as constituting a “deluge” of barbarians. Esmonde Cleary, Roman West, p. 24. Christian Witschel, “Re-evaluating the Roman West,” pp. 269–70, for the quotations. Esmonde Cleary, Roman West, p. 19, for the quotation regarding chronic incursions. In this Esmonde Cleary follows a generalization made by Luttwak in the original edition of The Grand Strategy of the Roman Empire (Baltimore, 1976), pp. 130–54. It is important to emphasize that Esmonde Cleary, like Luttwak, not only fails to examine the literary sources critically but, by and large, treats them as plain text when they are mentioned.

6

Bernard S. Bachrach

were no “great Germanic invasions” of Gaul during the 260s and 270s. In fact, even a maximizing count of the so-called “invasions” of Gaul during the period c. 250–c. 275 indicates that there were perhaps as few as six relatively minor episodes in total in the course of a quarter century.12 Further, information concerning these third-century barbarian activities comes from sources written during a period far removed in time from the events concerning which they are assumed by “crisis” advocates to provide accurate and unbiased information.13 For the purposes of this paper, far more important than the distance in time between the supposed great barbarian invasions of Gaul during this period and the texts that purport to provide information concerning them are the strong biases of those texts. By and large, they are oriented toward exaggerating greatly the damage done to the Roman Empire in Gaul by barbarian invaders as well as other circumstances.14 In this regard, one merely has to take note of the fact that much of the basic information regarding the alleged barbarian invasions of Gaul is provided by the notoriously unreliable Historia Augusta.15 Many of the other sources available, as will be seen below, are considerably less prone to providing manifestly false information than is the case with the Historia Augusta. It is moreover clear that Christian accounts focused on providing information to sustain an “end of days” account of destruction at a level of hyperbole similar to that of the Historia Augusta.

12

13

14

15

For those interested in comparisons between the difficulties supposedly caused by barbarian activities, it is useful to compare what happened in Gaul in the mid-third century, as discussed here, and what happened in the early fifth century. Regarding the invasion of 406, see 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. From a methodological perspective it must be recognized that a lengthy temporal gap between the events under consideration and the sources providing information is not a basis for the a priori rejection of the latter’s potential to provide information that may be accurate. On this point see Arnaldo Momigliano, “The Rules of the Game in the Study of Ancient History,” trans. Kenneth W. Yu, History and Theory 55 (2016), 39–45. However, when these conditions do exist, as in the present case, a substantial effort must be made to validate the information found in such sources. This has not been done with regard to the sources relating to barbarian military operations during the period under discussion. The biases of these sources, many of which invalidate the information they provide, are however addressed below. Cf. Géza Alföldy, “The Crisis of the Third Century as Seen by Contemporaries,” Greek, Roman, and Byzantine Studies 15 (1974), 89–111, who casts his net well beyond Gaul and treats matters other than barbarian invasions. Alfoldi would have provided a much more valuable study had he examined in a systematic manner the overwhelming biases of the various sources that he discusses and, therefore, their limitations in providing accurate information. For a useful introduction to many of the sources that treat this period, see David Rohrbacher, The Historians of Late Antiquity (London, 2002). In addition, it is necessary to become familiar with the very flawed nature of the Historia Augusta. A useful introduction here is provided by Timothy Barnes, The Sources of the Historia Augusta (Brussels, 1978).



Later Roman Fortification of Gaul

7

The So-Called “Great Barbarian Invasions” of Gaul The state of the Gallic frontier with the lands of the German-speaking peoples throughout the 250s is best characterized as “quiet.” The only information from any source regarding barbarian military operations on the Roman frontier in Gaul concerns the year 257. At this time, the Emperor Gallienus (r. 253–68), who was in the process of touring the western reaches of the Empire, is reported to have received intelligence that there may have been some type of threat to Roman assets in the lower-Rhine region from a group of Franks. However, this putative raid of some sort can hardly have been of great moment in Gallienus’s strategic calculations, since at this time he commanded a formidable field army.16 In fact, he dealt with the “invasion” as though it were a mere annoyance – hiring a barbarian chief to lead his company of mercenaries to engage in a searchand-destroy operation against the raiders.17 Zosimus, although he did not hold a high opinion of Gallienus, reports nevertheless that the emperor’s barbarian mercenaries did their job well and destroyed the “invaders.”18 It must be noted, perhaps as an index of the order of magnitude of the threat, that there is no archaeological evidence that this Frankish “invasion” in 257 had any effect, much less a negative one, on the countryside or on the civilian population.19 Despite some hyperbole on the part of modern scholars regarding “great invasions,” as noted above, those barbarians who may have contemplated attacking imperial assets along the lower Rhine, or perhaps even crossing the Gallic frontier into Roman territory further up-river for aggressive purposes, would seem to have been frightened into inaction by the destruction of their Frankish brethren in 257. Indeed, it is known that Gallienus established a temporary headquarters at Cologne, where he showed the flag, as evidenced by an inscription on the arch of the north gate.20 Moreover, in following up on the above-mentioned “invasion,” Gallienus issued coins which depict him as Restitutor Galliarum.21 This, of course, was a traditional, even banal exercise in self-promotional propaganda, frequently indulged by Roman emperors. Such efforts are best understood as the emperor’s attempting to inflate his own reputation as a military commander, 16 17

18 19

20 21

Andreas Alföldi, “The Numbering of the Victories of the Emperor Gallienus and the Loyalty of his Troops,” Numismatic Chronicle 9 (1929), 218–79, provides some interesting information. This is according to Zosimus, Histoire Nouvelle 1.30, ed. and trans. François Paschoud (Paris, 1971). Drinkwater, Gallic Empire, pp. 9–16, tries to conflate a number of literary and epigraphic sources to conclude that this was a great barbarian invasion. He accepts (ibid., p. 88) Zosimus’s claim that the emperor hired a Frankish mercenary leader to destroy the raiders but does not discuss the obvious difficulties that would be involved in order to explain how a mere band of mercenaries could destroy a very large invasion force. Zosimus, Histoire 1.30. Esmonde Cleary, Roman West, passim, who focuses on archaeological information, is unable to call attention to a corpus of material evidence that might cast light of the supposed “invasion” of 257. For the inscription, see I. König, Die gallischen Usurpatoren von Postumus bis Tetricus, (Munich, 1981), no. 31; and discussion by Drinkwater, Gallic Empire, p. 110. Drinkwater, Gallic Empire, p. 167.

8

Bernard S. Bachrach

and as being intended as well to convince the tax-paying population that the government, and more particularly the army, was hard at work doing its job of protecting the people.22 Indeed, by 259 any whisper of problems or even potential problems on the Rhine frontier would seem no longer to have been heard. In this context, Gallienus returned to Italy with a substantial part of the formidable field army that had previously composed the garrison in Gaul, and is reported to have turned his attention to the Balkans, where he campaigned with great success. Any attempt either to identify those units which accompanied the emperor south or to enumerate those that remained in Gaul following Gallienus’s departure remains problematic. Nevertheless, it is clear that a substantial field force, along with barbarian mercenaries, remained to defend the Rhine frontier.23 In addition, as a symbol of continuing imperial attention to Gaul, Gallienus left his young son Saloninus at Trier, under the tutelage of the general Silvanus. The fact that the imperial headquarters at this time was established at Trier rather than further to the north at Cologne likely permits the inference that the lower-Rhine region meanwhile was quiet.24 As is clear from the above, there is no reason to believe as regards the decade of the 250s that there were any significant trans-border military operations by barbarians either into Gaul or affecting Gaul. A minor raid in the area of the lower Rhine, which was destroyed by a unit of barbarian mercenaries in the pay of the Roman government, can hardly be considered the stuff of a great invasion. However, early in 260, that is, in the wake of Gallienus’s departure from Gaul, Frankish and Alamannic raiders are reported to have crossed the imperial frontier.25 Some scholars also believe that a band of Franks managed to get all the way across Gaul to the city of Tours on the Loire river, some eight hundred kilometers west-southwest of Cologne as the crow flies. The text cited as evidence indicates that at Tours, however, the local citizens, unsupported by regular troops, thwarted the raiders and sent the remnants fleeing back to Free Germany.26 22

23 24 25

26

See the apt observations regarding honors and awards as instruments of propaganda by John F. Drinkwater, “The Threat to the Rhine Frontier: A Romano-Gallic Artefact,” in Shifting Frontiers in Late Antiquity: Papers from the First Interdisciplinary Conference on Late Antiquity, the University of Kansas, 1995, ed. Ralph W. Mathisen and Hagith S. Sivan (Aldershot, 1996), pp. 20–30. Drinkwater, Gallic Empire, pp. 88, 116. Zosimus, Histoire 1.38; Drinkwater, Gallic Empire, p. 23. Thomas S. Burns, Rome and the Barbarians, 100 B.C.–400 A.D.: Ancient Society and History (Baltimore, 2003), p. 414, n. 20, quite correctly does not see these activities as significant in regard to Gaul. The writer Eusebius (of Nantes) who provides this information, was an early fourth-century Greek chronicler whose work here survives in fragmenta. See F. Jacoby, Die Fragmemte der griechischen Historiker 2.101, ed. Felix Jacoby (Berlin, 1926–30), fr. 5. The basic study of this incident remains Théodore Reinach, “Le premier siège entrepris par les Francs,” Revue historique 43 (1890), 34–46, whose speculations are engaging but not compelling. The problematic nature of Eusebius’s account is ably illuminated by Drinkwater, Gallic Empire, pp. 84–85. Particular doubts must be raised also by the description of the siege machines which the invaders, actually Celts in the language of Eusebius’s account, are supposed to have used. Regarding this matter see E. A. Thompson, The Early Germans (Oxford, 1965), p. 133.



Later Roman Fortification of Gaul

9

The cives of Touraine apparently accomplished this by throwing up makeshift fortifications. These would seem to have been something like castra: traditional camps used by regular Roman troops on the march, erecting these being a defensive maneuver no doubt well known to the veterans who were undoubtedly settled among the population in this region, as they were through many parts of Gaul.27 If this story is true, and the details are highly problematic –­ for example, an otherwise uneventful 800-kilometer march across Gaul – it merely emphasizes the strength and military preparedness of the Gallo-Roman civilian population at the local level in the face of what would seem to have been an almost comically inept barbarian raiding party. It is worthwhile noting that here, too, no archaeological evidence has been unearthed to support this supposed invasion of Touraine.28 Some scholars have suggested that other groups of Franks may have raided closer to their homeland in Lower Germany and Belgica at about this time.29 However, notwithstanding substantial scholarly speculation on this matter, we have written information, apparently reliable, concerning only one such barbarian effort, and this group enjoyed very limited, and more importantly very short-lived, success.30 The only specific account of barbarian operations along the lower Rhine in this period concerns the successful effort of the Roman military commander Marcus Cassianius Latinius Postumus to deal with a single Frankish raid.31 At this time Postumus may well have been the contemporary governor of Lower Germany.32 However, Zosimus refers to him as “commander of the troops in Gaul,” and Victor makes clear that Postumus commanded the “barbarians” in Gaul as some sort of chief of imperial mercenaries.33 In any case, while operating with a military force along the lower Rhine, apparently in the area of Diessen, which is hardly part of Gaul, Postumus is reported to have trapped a booty-laden raiding party, annihilated it, and distributed the reclaimed loot to his troops.34 27

28 29

30

31 32 33 34

Eusebius, in Reinach, “Premier siège.” See Ramsey MacMullen, Soldier and Civilian in the Later Roman Empire (Cambridge, MA, 1963), pp. 103–13, regarding various roles, including contributions to the local defense, played by veterans settled in Roman installations. Esmonde Cleary, Roman West, like all previous archaeologists, has been able to find nothing in Touraine from the 260s to support this story. Here I follow, at least in a limited way, Drinkwater, Gallic Empire, p. 23, who deduces this pattern of attacks from the effective military operations undertaken by the general, later emperor, Postumus, discussed below, although the chronology is not a close fit and there is no archaeological evidence to support the notion of a flurry of raids at this time in the supposedly impacted region. Drinkwater, Gallic Empire, p. 23, concludes that, “The barbarians […] poured across the northern frontier,” but does not make clear what evidence he has for this statement insofar as it would seem to allude to a horde of invaders. Ibid., p. 24. Ibid., p. 25. Zosimus, Histoire 1.38; and Victor, Sexti Aurelii Victoris de caesaribus liber 33.10, ed. Franz Pichlmayer (Leipzig, 1882). For a discussion of this military operation, see Victor, De Caesaribus 33.8, who provides some very vague information, while Eutropius, Eutropius: Breviarium 9.9, ed. and trans. H. W. Bird (Liverpool, 1993), is of no help here. Drinkwater, Gallic Empire, pp. 25, 30, 143, 162–63, and 170, tries to focus the date of these events by reference to coinage, but on the whole we know

10

Bernard S. Bachrach

Further south, some Alamanni are believed by some modern scholars to have been successful against imperial assets. One group is said to have penetrated the Agri Decumates east of the Rhine: the “Ten Cantons” established in the Black Forest and bordered by the Neckar river and the Swabian Alps. It is generally agreed that the intention of these barbarians was to settle in the region as farmers and imperial subjects.35 However, two observations are in order here. First, the Agri Decumates are not in Gaul, and, thus strictly speaking are beyond the geographical focus of this study. Whatever may have happened there, and the question of whether or not the Alamanni conquered the region, is not relevant to a discussion specifically of supposed great barbarian invasions of Gaul. Second, it is important to emphasize that, in fact, the Agri Decumates were almost certainly not conquered by the Alamanni.36 It is far more likely that Gallienus, who is consistently given a bad press in the written sources that treat this period, but who commanded overwhelming military force, arranged to settle Alamanni there as farmers. This policy, which brought about an influx of population to the region, was intended to make it more productive both for the growing of food and for the paying of taxes to the Roman government.37 Later emperors who, like Gallienus, also commanded overwhelming military strength, similiarly made no effort either to destroy the Alamanni settled in the region or to drive them out of it. In short, these barbarians were settled there to grow food and provide taxes for the imperial government, becoming assimilated through this process into the imperial familia.38 Some scholars argue that Alamannic invaders, not the same men with their families who were settled in the Agri Decumates, but another (hypothetical)

35 36

37 38

very little of what happened. With regard to the booty, see I. Shatzman, “The Roman General’s Authority over Booty,” Historia 21 (1972), 177–205. The source at issue is Laterculus Veronensis, and Drinkwater, Gallic Empire, pp. 86–87, ably illustrates its shortcomings. With regard to the supposed Alamannic conquest of the “Ten Cantons,” the most reliable written souces, Victor and Eutropius, have nothing to say on the matter. Zosimus, representing the Greek historiographical tradition in this context, is also silent concerning Alamannic military successes in the Agri Decumates. Esmonde Cleary, Roman West, p. 24, speculates concerning the supposed barbarian conquests of the region on the basis of his unsupported assumptions regarding the possible military strength of the Alamanni and the possible military weakness of the Romans. Regarding the relative strength of Roman and Alamanni forces and the great superiority in military operations of the former, 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 Kagay and L. J. Andrew Villalon (Woodbridge, 1999), pp. 3–20. See Alföldi, “The Numbering of the Victories,” pp. 218–79. The speculation by Drinkwater, Gallic Empire, pp. 23–24, regarding an Alamannic conquest is unsupported by either written or archaeological sources. See, for example, Mackensen, “Late Roman Fortifications,” p. 201. In addition, as will be seen below, Postumus’s great success, i.e. being recognized as emperor in both Spain and Britain, as well as in Gaul, and having at his disposal an immense military force, left the Alamanni settled in the Agri Decumates. Military operations there by Probus make it clear that the Romans maintained overall control of the region: Vita Probi, in Historia Augusta (Scriptores Historiae Augustae), trans. David Magie, 3 vols. (Cambridge, MA, 1921–32) 3:3.8.



Later Roman Fortification of Gaul

11

group, enjoyed considerable success against imperial assets in Gaul during the 260s.39 According to Fredegar, a seventh-century source, the Alamanni destroyed the fortified city of Avenches in 263.40 However, there is no reason to believe that at this time any group of Alamanni, or for that matter any other group of barbarians living beyond the Roman frontiers in the West, was in possession of the organization, weapons, or manpower to capture, much less destroy, a fortress city such as Avenches.41 The remains of the fortifications at Avenches have left me with the impression that the city was rather destroyed by an earthquake. Such phenomena are common in this region even to this day;42 and comparing the remains at Avenches with those at Beit She’an (Scythopolis), which is known to have been struck by an earthquake, the nature of the damage in both cases looks very similar.43 It is also believed by some scholars that yet another Alamannic raiding party entered Gaul and terrorized the Auvergne.44 However, this account rests upon the work of Gregory of Tours, who wrote in the early 590s. Significantly, Gregory would have his readers believe that the king of the Alamanni who commanded these military operations was Chrocus. This, of course, is a gross anachronism: Chrocus is traditionally seen as a contemporary of Constantine the Great (d. 337), and is believed to have lent his support to that emperor in 306, early in the latter’s reign.45 Gregory tells us, nevertheless, that Chrocus not only burned the city of Clermont, but tore down and completely destroyed

39 40

41 42

43 44 45

Drinkwater, Gallic Empire, pp. 86–87; again, as Drinkwater shows, the relevant source, Laterculus Veronensis, in this context is not reliable. Drinkwater, Gallic Empire, pp. 23–24, discusses the matter, and points out the weaknesses inherent in Fredegar’s text; Drinkwater recognizes too that the information obtained from archaeological sources does not provide support for the tradition found in Fredegar (p. 217). Burns, Rome and the Barbarians, p. 416, n. 31, considers Fredegar’s account to be of no value, as does Justin Favrod, “La date de la prise d’Avenches par les Alamans,” Arculiana: Ioanni Boegli, anno sexagesimo quinto feliciter peracto (Avenches, 1995), pp. 172–73, who shows that the information provided by Fredegar is not reliable. Thompson, Early Germans, pp. 131–40, still remains basic regarding the inability of various barbarian groups in the West to mobilize, much less to lay siege to a fortress city. On the whole, in light of the amount of damage wrought at Avenches and the location of this fortress city on the Swiss plateau, which may be considered an earthquake-prone area, it seems likely that it was destroyed by an earthquake rather than by barbarian military action. Indeed, geologists have established that the Roman fortress city of Augusta Raurica (Kaiseraugst), only eighty-five kilometers from Avenches and also on the Swiss plateau, was rocked by an earthquake in 250 A.D. Augusta Raurica was located 47.5oN, 7.7oE, Aventicum (Avenches) at 46.9oN, 7.0oE. I owe this information to the kindness of Dr. Klaus-G. Hinzen, Priv. Doz. in the Department of Geology of the University of Cologne. The data concerning Augusta Raurica is available from the Swiss Seismological Survey at http://hitseddb.ethz.ch:8080/ecos09/query_ sum (accessed 13 August 2017). Regarding Scythopolis, see Yoram Tsafrir and Gideon Foerster, “Urbanism at Scythopolis–Bet Shean in the Fourth to Seventh Centuries,” Dumbarton Oaks Papers 51 (1997), 85–146. Drinkwater, Gallic Empire, p. 23. Anon., Epitome de Caesaribus 41.3, ed. Franz Pichlmayr (Leipzig, 1911).

12

Bernard S. Bachrach

the shrine located there that the Celts called “Vasso Galatae.”46 This story is of considerable interest because, according to Gregory, this shrine was “built in an impressive manner with a double wall. The inner wall supposedly was constructed with small stones, and the outer wall was built with very large squarely cut stones. The entire construction had a width of thirty feet.” Gregory goes on to claim that “the interior was covered with various kinds of marble and mosaics; the floor was paved with marble, and the roof was covered with lead.”47 No archaeological information has however been unearthed to sustain the view that Clermont was destroyed by the Alamanni in 260 or at any other time during the course of the later Roman Empire, to validate Gregory’s report regarding the destruction of Vasso Galatae, or even to attest to its existence in the region.48 Gregory’s imagination regarding Chrocus’s destructive campaign went well beyond the disaster he describes as having taken place at Clermont. This barbarian ruler, whom Gregory depicts as being motivated by his “wicked mother,” a not uncommon cliché, is credited with having “torn down every building constructed in ancient times.”49 This observation, prima facie, is inaccurate, even if only in regard to Clermont, but typical of the kinds of gross exaggeration purveyed at various times by certain Christian writers intent upon illustrating the disasters that struck the Roman Empire during its pagan incarnation.50 With the exception of the peaceful settlement of Alamannic farmers in the Agri Decumates, which are not located in Gaul, all of the literary information that survives concerning supposed Alamannic invasions is provided by unreliable early medieval Christian sources, written at a remove of three centuries or more by authors dedicated to demonstrating that during the era of pagan Rome, there was much suffering imposed by barbarian invasions. In contrast to these very late Christian sources, the traditional later-fourthcentury accounts written by Victor and Eutropius, and the late-fifth-century Greek history by Zosimus, putatively based upon the now lost “Imperial History” and the Chronike of Dexippus respectively, have nothing to say regarding the supposed destruction either of Avenches or of Clermont, much

46 47 48

49 50

Gregory, Gregorii episcopi Turonensis. Libri historiarum X 1.32, ed. Bruno Krusch and Wilhelm Levison, MGH SSrerMerov 1.1 (Hanover, 1951). Gregory, Libri hist. 1.32. The seventh-century embellishment of Gregory’s account of this invasion by Fredegar, Chronicarum quae dicuntur Fredegarii Scholastici libri IV. cum Continuationibus 2.60, ed. Bruno Krusch, MGH SSrerMerov 2 (Hanover, 1951), is no more plausible than that provided by his source. See the discussion by Drinkwater, Gallic Empire, pp. 84–86; and more specifically with regard to the lack of archaeological evidence see Peter Horne, “Romano-Celtic Temples in the Third Century,” in The Roman West in the Third Century: Contributions from Archaeology and History, ed. Anthony King and Martin Henig, 2 vols. (Oxford, 1981), 1:25. Gregory, Libri hist. 1.32. Favrod, “La date de la prise d’Avenches,” pp. 171–72, calls attention also to several gross errors by Fredegar, who borrows from Gregory. The material discussed by Michael Greenhalgh, The Survival of Roman Antiquities in the Middle Ages (London, 1989), gives the lie to Gregory’s hyperbole.



Later Roman Fortification of Gaul

13

less the destruction of all ancient buildings in Gaul as claimed by Gregory of Tours. In addition, there are no inscriptions or other archaeological evidence to support these accounts.51 Finally, even the author of the Historia Augusta, whose tendency to exaggerate disaster is well known, especially when the situation suits his interests, fails to provide any specific support for either of the episodes discussed above. Whatever Gregory and Fredegar may have had in mind, any attempt to use this information as evidence to claim that Gaul was ravaged by great barbarian invasions must be abandoned. The problematic nature of the information regarding the destructiveness of Alamannic operations in Gaul following Gallienus’s redeployment to the Balkans must not however be taken to imply that elements among this people never attempted to take hostile action against imperial assets during the period under discussion. It seems likely, for example, that a group of Alamanni was attracted from their homes beyond the upper Danube to northern Italy early in 260. It is reported that Gallienus thoroughly defeated this group in the environs of Milan, and by midsummer at the latest had moved with his army once again into Gaul. It seems clear that he was at this time following his peripatetic strategy of showing the flag throughout various parts of the empire.52 This destruction in Italy of an Alamannic force, concerning which there is no evidence that it had ever trodden on Gallic soil, much less gained anything there to the detriment of the Empire, would seem to follow a pattern. Like the successful resistance of the citizens at Tours and the equally successful efforts of Postumus in the region of Diessen, the destruction of these Alamanni provides a good indication of the relative strength of the Roman position, both military and civilian, and the comparative weakness of any particular group of barbarian invaders. Indeed, when specific information is provided by the relevant sources, those barbarians involved in hostile operations against the Romans within Gaul or even on its frontiers are uniformly the losers. Clearly, Gallienus was satisfied with the arrangements that he had made in Gaul, and he did not regard the trans-Rhenish barbarians as posing a serious military threat. So the emperor once again moved south, and left his son Saloninus, now elevated to the rank of caesar, to rule the region with the continued support of the general Silvanus. Before the end of 260, however, the above-mentioned Postumus executed a coup d’état, which succeeded after a very brief siege of Cologne, where Saloninus and Silvanus had taken up residence following the departure of Gallienus. This military operation is not reported to have resulted in damage either to the city or to the population. Having had Silvanus and Saloninus executed, Postumus next established his capital at Trier: such a move south from Cologne would seem to permit the inference, at the very least, that had any barbarian threat to the lower Rhineland 51 52

Esmonde Cleary, Roman West, finds no archaeological evidence to support the accounts of Alamannic destruction in Gaul provided by Gregory and Fredegar. Drinkwater, Gallic Empire, p. 24, believes that midsummer is the likely season for this battle. See also p. 104.

14

Bernard S. Bachrach

been perceived to have developed during the brief “civil war” with Saloninus, this had either greatly diminished or completely evaporated. In short, no barbarians sought to take advantage of unrest in Gaul.53 A glimpse of Postumus’s successful military administration at this time is provided by an inscription from Augsburg relating to events early in the 260s.54 It seems that the acting governor of the region of Raetia, a certain Marcus Simplicinius Genialis, received intelligence that a raiding party of “Juthingi” and “Semones,” often considered to be Alamanni by modern scholars, had moved from the upper Danube into Italy, where they enjoyed considerable success raiding south of the Alps. However, on their return to the upper Danube, as they were crossing Raetia laden with booty and captives taken in Italy, it is reported that they were trapped by Simplicinius as they tried to move through the Brenner Pass.55 Simplicinius, who apparently lacked a field army of regulars – comitatenses – had mobilized a force composed of limitanei (quasi-militia troops) who were based in the region, supplemented by “civilian” volunteers. He engaged the enemy, defeated them in battle, recaptured the booty, and freed their prisoners, as indicated in the Augsburg inscription.56 This episode is often included among the crisis-inducing events in Gaul. However, it should not be exploited as evidence for barbarian military operations in Gaul, any more than for successful Alamannic operations in Raetia, which, of course, is not part of Gaul. By contrast, Marcus Simplicinius’s effort very well illustrates the overwhelming superiority of Roman limitanei and civilians in ad hoc situations. The episode seems to have much in common with Gallienus’s successful deployment of barbarian mercenaries on the lower Rhine in 257, and the civilians who enjoyed success against a group of Frankish raiders in the environs of Tours in c. 260. In fact, as suggested above, the fundamental military weakness of barbarian raiding parties, or as some scholars would seem to believe, “great invasions,” against less-than-first-class imperial forces and even civilians is unambiguously evident from the Augsburg inscription mentioned above. Throughout his reign in Gaul, Postumus may have been pursuing a policy toward some groups of barbarians which called for the integration of selected elements of the Germanic population from beyond the frontiers into the Roman familia.57 For example, he recruited both Franks and Thuringians as merce53 54

55 56

57

For the move see Drinkwater, Gallic Empire, p. 27. This inscription has received substantial attention from scholars. See, for example, L. Bakker, “Raetien unter Postumus. Das Siegesdenkmal einer Juthungenschlacht im Jahre 260 n. Chr. aus Augsburg,” Germania 71 (1993), 369–86; Mackensen, “Late Roman Fortifications,” p. 200; and Burns, Rome and the Barbarians, p. 15, n. 24. Esmonde Cleary, Roman West, p. 19. Ibid., pp. 18–19; Esmonde Cleary seems to have little sense of quantifiable reality or logistics, believing that this defeated force of barbarians had initially taken “many thousands of Roman prisoners” north from Italy into the Alps. Burns, Rome and the Barbarians, p. 288, observes, “In both the Agri Decumates and in the former Dacia there are suggestions that the new barbarians took over with the tacit approval of the empire.”



Later Roman Fortification of Gaul

15

naries to serve imperial military interests.58 Moreover, as was made clear above, a policy of recruiting potential enemies was hardly new, and certainly cannot be construed as the result of supposed massive Roman military losses that otherwise are unreported in the sources. In 257, as already noted, Gallienus had hired a Frankish mercenary captain to engage in search-and-destroy operations against “his own people” in the region of the lower Rhine. The emperor initiated this policy whilst he commanded an imperial army of vast size; these mercenaries must be considered auxiliaries of some sort.59 It was well known to Roman military officers that using barbarian mercenaries under time-conditioned contracts was much less expensive than using regular imperial troops.60 In late 262 or part of 263, Postumus is thought to have found it necessary to take the field in northeast Gaul. There are no details regarding this operation to be found in the narrative sources, but Postumus’s coins for 263 indicate that he had celebrated a victory that year. Because the sources do not mention these military operations, there is no information regarding what if any damage was done by the enemy, or if any battles were fought by the Roman army to merit this otherwise uncorroborated burst of numismatic propaganda.61 This victory celebration is however seen by some scholars as being of key importance. Previously no victories as such can be shown to have been celebrated in Gaul during this period, despite the survival of some dubious detail concerning several of the successful operations, discussed above, that were mounted against the barbarians. Although we do not know what happened in 262, it is perhaps worth noting that following an official victory, Roman emperors often levied the aurum coronarium, by which the cities of the Empire made an offering in gold coins or bullion to the victorious emperor.62 This was a not-unimportant source of revenue for emperors, and perhaps explains the declaration of an otherwise unknown military success. Whatever we choose to make of these events, however, the information that we do possess, if reliable, can once again in no way be construed to mean that there had been a great barbarian invasion of Gaul, and still less as evidence for a successful one. In 265, however, real potential for serious damage in Gaul may indeed have presented itself, as the consequence of warfare between regular Roman armies. Gallienus decided that the time was ripe for him to “reconquer” Gaul from 58

59 60

61

62

Drinkwater, Gallic Empire, pp. 31, 225, cites the sources that indicate Postumus’s recruitment of various barbarian mercenaries. It is not clear why Drinkwater sees this as a sign of Postumus’s weakness, when the hiring of mercenaries was a normal activity for Roman generals. Drinkwater, Gallic Empire, p. 22, both for recruitment and for Gallienus’s forces. See, for example, the observations by C. R. Whittaker, “Explanations of Inflation in the Fourth Century A.D.,” in Whittaker, Land, City and Trade in the Roman Empire (Aldershot, 1993), p. 4, with references to the sources. Victor, De Caesaribus 33.8, provides some very vague information; and Eutropius, Breviarium 9.9, is of no help here. Drinkwater, Gallic Empire, pp. 30, 143, and 170, tries to focus and date these events by reference to coinage, but on the whole we know very little of what happened. A. H. M. Jones, “Inflation under the Roman Empire,” The Economic History Review, n.s. 5 (1953), 293–318 at 297.

16

Bernard S. Bachrach

Postumus. On two occasions, Gallienus tried and failed to bring Postumus to battle. Finally, when the former was accidentally injured, the invasion came to an end with no perceptible damage having been done either to the population of Gaul as a whole or to the armed forces of either emperor.63 Again, the prudent, perhaps greedy, Postumus celebrated a victory at Trier for the bloodless campaign notable for his avoidance of battle and the withdrawal of Gallienus from Gaul.64 The opportunity was not to be ignored for yet another propaganda “victory” and yet another pretext to receive the aurum coronarium.65 Despite what perhaps may be considered Postumus’s impressive personal military background and record of success or, at least, putative success, on his coinage for propaganda purposes he generally emphasized the civil rather than the military aspect of Hercules, his favorite deity.66 This tactic likely was intended to assure the people of Gaul, possibly concerned about the potential negative impact of “invasions” from beyond the frontiers and the difficulties that might be caused by civil war, that the emperor’s policy was one of peace, security, and prosperity. Indeed, the coinage issued in both 267 and 268 sends this very message.67 Postumus’s continued positive treatment of the Alamanni settled in the Agri Decumates, his lack of interest in expanding his imperium into Italy, which could only have been accomplished with considerable military effort, and his avoidance of battle with Gallienus also sustain the notion that he regarded peace as preferable to war whenever possible. Early in 269, Ulpius Cornelius Laelianus, Postumus’s governor of Upper Germany, with his provincial capital at Mainz, attempted a coup d’état. One useful index of Postumus’s military strength, in this context, is that Laelianus’s effort was put down in less than three months. This was about as long as it took Postumus to move his forces out of winter quarters, a difficult and time-consuming process, and march south from Cologne to Mainz. Postumus won a bloodless victory, and disposed of the would-be usurper. However, he is reported to have made a fatal error in the aftermath of this victory. As noted above, Postumus had won the support of his troops for a coup d’état against Saloninus in 260 by permitting his men to keep the booty that they had taken from Frankish raiders. By contrast, after his troops took Mainz, Postumus did not let his men sack the city, and they are reported as a result to have murdered him.68 The mutinous troops then proclaimed a certain Marcus Aurelius Marius, apparently a common soldier, emperor. Whether Marius’s troops sacked Mainz 63

64 65

66 67 68

Drinkwater, Gallic Empire, p. 30, tries to reconstruct events from a very weak source base which goes back to a twelfth-century Greek chronicle by Zonaras, who putatively worked with the text written more than half a millennium earlier by Hexippus. Drinkwater, Gallic Empire, pp. 31, 171–72. Concerning these propaganda efforts, see John F. Drinkwater, “The Germanic Threat on the Rhine Frontier: A Romano-Gallic Artefact?,” in Shifting Frontiers in Late Antiquity, ed. Ralph W. Mathisen and Hagith S. Sivan (Aldershot, 1996), pp. 20–30. Drinkwater, Gallic Empire, p. 32. Ibid., pp. 32, 155. Ibid., pp. 32–33.



Later Roman Fortification of Gaul

17

in the course of this “civil war” is uncertain. No written source indicates that the city was given over to the troops, and there is no archaeological evidence datable to 269 in either the urbs itself or the suburbs to support a conclusion that the city or the civitas was ravaged.69 It is known, however, that Marius was murdered by some soldiers after holding the imperial title for about twelve weeks.70 He was succeeded by Victorinus, who had been Postumus’s main field commander for several years, and had a reputation as a very able military leader.71 The murder of three emperors in less than six months certainly had the potential to stimulate feelings of insecurity among the people of Gaul, and perhaps more importantly, in light of the focus of the present study, to encourage transRhenish “barbarian hordes” to cross the frontier and ravage the countryside. No barbarian raids were actually reported, however, either as accompanying these murders or in their wake. Judging by the written sources, there appears to be no foundation for any such feelings of fear among the Gallo-Roman population in those areas where problems might be hypothesized to have occurred. So far as can be ascertained, there were no wars or battles, no sieges or even marching and counter-marching by Roman armies through the countryside. There are no written reports of devastation and depopulation by the actions either of Roman military forces or of barbarian invaders. Finally, there is no sign of that characteristic though dubious archaeological indicator of a period of serious disruption – substantial numbers of coin hoards – to illuminate this particular context.72 Whether or not the Franks or the Alamanni, in small or in large groups, sought to take advantage of what some modern scholars believe to have been a highly unstable political situation in Gaul – indeed, a crisis – no information survives to sustain the view that any invasions took place. We may hypothesize that the barbarians had poor intelligence regarding what would seem to have been happening in Gaul. It is likely, however, that the barbarians beyond the frontier were simply unable to organize an invasion in response to what would appear, in the said scholars’ perspective, to have been a propitious situation. It deserves to be emphasized that even after the murders of three emperors, and sufficient time having passed for reports of these to have been disseminated east of the Rhine, Victorinus, the man who succeeded as emperor in the Gallic provinces in the wake of the brief civil war, is not recorded by any of the sources as having found it necessary to engage in military operations against the barbarians, either pre-emptively east of the Rhine, or within the territory of Gaul itself in order to drive out invaders.73 By contrast with the apparent quiet on the eastern frontier of Gaul during this period, the Emperor Claudius II, who had succeeded Gallienus in Italy, sent his 69 70 71 72 73

Ibid., pp. 33–34. Ibid., p. 35. Ibid., pp. 35–36. Esmonde Cleary, Roman West, for example, provides no information regarding identification of coin hordes that can be dated such as to be interpretable as resulting from this situation. Drinkwater, Gallic Empire, p. 36.

18

Bernard S. Bachrach

general Julianus Placidianus with a substantial military force to take control of the Narbonensis and to establish a base at Grenoble. By the end of 270 at the latest, all of Gallic territory east of the lower Rhône valley recognized Claudius as its legitimate emperor.74 What is significant in this context is the almost bloodless nature of Placidianus’s operations, as much of southeastern Gaul easily fell under Claudius’s rule without opposition from Victorinus. To put it another way, Victorinus showed no inclination to oppose Claudius’s advances with military force. However, in light of Victorinus’s subsequent operations against Autun, which did not concern either barbarian raids or invasions, and of Claudius’s obvious disinterest in this city, it seems possible that the two emperors had come to an agreement with regard to a territorial redistribution of imperial power in Gaul.75 As a result of what would seem then to have been some sort of agreement between Claudius and Victorinus, the latter moved his primary base of operations to Cologne. The apparent aim here was the further strengthening Roman control in the region of the lower Rhine by supporting ongoing demographic and economic development, involving an increase in and protection of urban settlement and agricultural growth, as well as merchant activity, not only to supply the Roman army but also to meet the needs of the court and all that followed in its wake.76 As will be seen below, the continuously growing wealth of the region from at least the later second century attracted pirates, which in turn led to further development of the defenses of the so-called “Saxon Shore,” to protect the coasts of the North Sea and the Channel. In this context, it is important to emphasize that Cologne was the largest and most important city on the northeastern frontier. Indeed, it served as the capital of the region of the lower Rhine. It was also very important in military terms as a major base for the Roman fleet that operated in the Channel and the North Sea in concert with the fleet based at Boulogne.77 Yet, it is very clear that the establishment of an imperial capital, as compared to a regional capital, required an infrastructure far more developed than that enjoyed by a city such as Cologne at this time. A useful model for understanding how the placement of an imperial capital affects the city in which it was established and the surrounding area is the analogy of the immense development that took place in Trier and its environs during the early fourth century.78 Unfortunately, Victorinus was murdered in the course of a family scandal, and one Tetricus was elevated to the imperial title at Bordeaux to succeed him. While on his way north to Cologne, Tetricus is believed by modern scholars 74 75 76 77 78

Ibid., p. 36. Ibid., pp. 38–40. See the brief but useful observations by Drinkwater: ibid., pp. 34–40, 100–03, 135–38, 143–46. Regarding the naval base at Cologne, see Chester G. Starr, The Roman Imperial Navy: 31 B.C.–A.D. 324 (Ithaca, NY, 1941), pp. 147, 151–52. For useful information concerning Trier, see Carlrichard Brühl, “Trier,” in Brühl, Palatium und Civitas, Vol. 2, Belgica I, beide Germanien und Raetia II, pp. 65–75.



Later Roman Fortification of Gaul

19

to have encountered a raiding party of barbarians somewhere along his route in eastern Gaul. These raiders are said to have been dealt with so easily by Tetricus’s entourage that none of the literary sources thought it worthy of mention, and no inscriptions have been found to commemorate this victory. Rather, the episode is surmised only from Tetricus’s coins, which proclaim him a Victor over the Germans. So the murder of Victorinus and the succession of Tetricus, far from causing unrest among the troops, and perhaps more importantly creating opportunities for trans-Rhenish barbarians to launch immediate large-scale raids beyond the frontier, seem to have had no serious effect except for the raiding party itself that was so easily destroyed.79 As the preceding discussion of the supposed barbarian threat makes clear, the destruction of this raiding party, if the event even occurred, is the only surviving indication of hostile operations by barbarians against Roman rule in Gaul up to this point since 264: a period of more than seven years. Indeed, this victory can hardly be considered of more than personal significance to Tetricus, insofar as it provided the erstwhile desk-bound bureaucrat from the southwest of Gaul with an opportunity to trumpet his military prowess.80 In addition, the possibility cannot be ignored that Tetricus proclaimed this “victory” in order to acquire an aurum coronarium from the cities of his empire. Such funds could obviously be used to help defray the costs of the donative traditionally given to the troops at the beginning of a new reign. Where specific information is provided by the relevant sources, the barbarians involved with the hexagon are uniformly the losers. In 272, there do appear to have been barbarian incursions into Gaul, perhaps encouraged by Tetricus’s over-lengthy stay at Trier and/or his failure to show the flag at Cologne. In any case, the episode is mentioned in one of the more reliable literary sources for this period.81 The numismatic evidence indicating that Tetricus was once again victorious against the Germans thus has some corroboration, and may not have been merely empty propaganda. Indeed, in the context of his victory in 272, Tetricus would seem to have had his young son elevated to the status of co-emperor, as Tetricus II.82 Of course, however one chooses to understand this barbarian “invasion,” it clearly was unsuccessful. Whatever may have been the greater meaning of such barbarian raids, it is clear that the incursions were a military problem only insofar as it took some time, some effort, and some resources to destroy them. These raiders were pests rather than units of enemy military forces that threatened territorial conquest in 79 80

81 82

For Tetricus’s itinerary and the supposed battle, see Drinkwater, Gallic Empire, pp. 184–87. Drinkwater’s argument (ibid., pp. 40, 185) that Tetricus’s proclamation of victory over the “Germans” on his coins may be used as evidence for a large scale war puts far too much weight on numismatic information unsupported by other sources. Indeed, the chronology of the coinage and the date of Tetricus’s initial establishment at Trier all permit the inference that he encountered a raiding party before reaching his capital. Tetricus’s coinage seems to be yet one more propaganda ploy intended to bruit about, indeed exaggerate, his military ability. Victor, De Caesaribus 35.3. Drinkwater, Gallic Empire, p. 40.

20

Bernard S. Bachrach

Gaul and a crisis for imperial survival west of the Rhine.83 There was general prosperity throughout Gaul at this time: no plague or famine, and little that can pass for civil war or barbarian invasion, is reported. During Tetricus’s reign, the army would appear to have been very successful in keeping barbarian raiders from crossing the frontiers. Indeed, an overall peace was maintained, despite reports that various military units from time to time “revolted.” One of these units is reported to have been led by a certain Faustinus who commanded the troops stationed in and around Trier. However, no hostilities are recorded.84 In 273, the Emperor Aurelian, who ruled the rest of the Roman world, decided that it was time to crush Tetricus and his “Gallic empire,” and reintegrate the region under his imperium: continuing, that is, the policy that Claudius had initiated a few years earlier. Early in 274, Tetricus I and Tetricus II went south to meet Aurelian. Faced at Châlons-sur-Marne with a battle-hardened army of superior quality, which in addition enjoyed overwhelming force, and was commanded by a general of the highest quality, the Tetrici negotiated their surrender after several failed maneuvers.85 This was a bloodless victory for Aurelian, who subsequently honored his arrangements with the Tetrici: the elder was given an important administrative post as corrector of Lucania, and the younger was confirmed in his senatorial status.86 Aurelian treated in an astonishingly lenient manner not only these last of the Gallic emperors, but also all those who had gone before: none suffered damnatio memoriae, and as a result their legal acta remained valid. Indeed, the surviving senior officials who had served the Tetrici and their predecessors not only avoided proscription but, in general, were permitted to retain their offices.87 A maximalist reading of the sources for this period has enabled some scholars to identify perhaps as many as six barbarian invasions of Gaul. A more critical reading of these same sources, as above, might well lead to lowering the number of barbarian raids from six to three or perhaps four. We can possibly eliminate the controversial Frankish raid into the Touraine, and reports of both the Alamannic raid against the Auvergne and the attack on Avenches, coming from late sixth- and seventh-century sources respectively, are surely suspect. A successful military operation from beyond the Danube by Alamanni in Italy 83

84 85 86

87

The immense exaggeration by modern scholars of the threats supposedly posed by barbarian invaders is effectively treated by Walter Goffart, “The Theme of ‘The Barbarian Invasions’ in Late Antique and Modern Historiography,” in Das Reich und die Barbaren, ed. Evangelos K. Chrysos and Andreas Schwarcz (Vienna, 1989), pp. 87–107, repr. in Walter Goffart, Rome’s Fall and After (London, 1989), pp. 167–211. See the candid observations of Ammianus Marcellinus regarding the nature of the Frankish threat: Ammianus, Rerum gestarum libri qui supersunt 16.5.17, ed. Wolfgang Seyfarth, Liselotte Jacob-Karau, and Ilse Ulmann (Leipzig, 1978). Eutropius, Breviarium 9.10, for the revolts, and Victor, De Caesaribus 35.4, regarding Faustinus. See, for example, Victor, De Caesaribus 35.4; Eutropius, Breviarium 9.13. See Drinkwater, Gallic Empire, pp. 42–43, 91, who suggests that some of Tetricus’s troops failed to surrender and were slaughtered by Aurelius’s superior forces. This may be accurate, but certainly of no great demographic or economic importance. Drinkwater, Gallic Empire, p. 42.



Later Roman Fortification of Gaul

21

which was crushed in Raetia may also be crossed off the list. However many barbarian “invasions” of Gaul one might settle upon. moreover, all were of minor importance and were crushed brutally and effectively, largely by forces composed of irregulars such as mercenaries and cives. Finally, none of these barbarian efforts has left archaeological evidence, in the form of ostensiblyrevealing layers of black earth at significant habitation sites and numerous coin hoards, that might lend support to the view of some modern scholars that they amounted to great invasions.88 The Invasion of 275–276 Not long after settling matters with the Tetrici in Gaul, Aurelian moved on to Illyricum where he was murdered. His two immediate successors, Tacitus and Florianus, the latter perhaps being the former’s brother or half-brother, had the responsibility of ruling in Gaul for a period of fewer than nine months between them. On the basis of the Historia Augusta, it is believed by some modern scholars that during the years 275–76 – so before Probus, Florianus’s successor, arrived in Gaul with yet another impressive imperial field army – barbarian invaders engulfed the Gallic provinces with large armies and placed perhaps as many as seventy “cities” under their rule.89 As Arthur E. R. Boak put it, “274 [sic] saw Gaul devastated by land and sea, and two years later the whole of that country was overrun again.” He continues, “in the year 275 alone, sixty towns of Gaul were said to have been captured and plundered by various barbarian tribes.”90 It is implausible, prima facie, that a substantial portion of the Gallo-Roman population of Gaul, perhaps overall somewhere in the ten to twelve million range, could have been “conquered” without evidence of a single battle having been fought, much less of some sort of barbarian victory having been won or perhaps even a number of barbarian military successes.91 In addition, as a corollary to a great but unrecorded barbarian military victory – an accomplishment comparable at least to Arminius’s victory in the battle of the Teutoburg Forest 88

89

90

91

Esmonde Cleary, Roman West, p. 32, points out that archaeologists are wont to identify all evidence for urban fires to have been the result of barbarian military operations. Thus he makes clear that it is important to take note of Witschel, “Re-evaluating the Roman West,” pp. 259–60, who points out that most fires at inhabited sites likely were due to accidents and not barbarian attacks. The state of the question recorded more than two generations ago by Inza Jan Manley, Effects of Germanic Invasions on Gaul (Berkeley, 1934), pp. 42, 91–103, has not been much altered since. Cf. Esmonde Cleary, Roman West. Arthur E. R. Boak, Manpower Shortage and the Fall of the Roman Empire in the West (Ann Arbor, 1955), pp. 24 and 56, for the quotations. One of the problems with Boak’s chronology is that he cites various scholarly works which give different dates for the “same” events and then treats these as different events. Despite the obvious conflation, which not incidentally increases the numbers of disasters, which is Boak’s aim, he does not look critically at the detail, still less the biases of the sources. However, as will be noted below, not only are there serious problems regarding the reliability of some sources, but the chronology is far from firm. Concerning estimates of the size of the population, see J. F. Drinkwater, Roman Gaul, The Three Provinces, 58 B.C.–A.D. 260 (Ithaca, NY, 1983), pp. 168–70.

22

Bernard S. Bachrach

– a vast number of problems would surely have been inherent in the imposition of some sort of barbarian rule over a very large Gallo-Roman population, which, as we have seen above, had shown itself to be militarily active and successful against various barbarian incursions during the previous two decades. Further, at this time, the imperial government in Gaul was not limited to relying on civilians and militia forces, but was supported by both a field army of considerable size and numerous units of limitanei deployed along the frontiers. When the written sources are properly evaluated – the sources which have been pasted together uncritically by modern scholars in search of great barbarian invasions and victories to sustain a narrative of imperial defeats and disasters meriting characterization as a crisis – the situation looks very similar to that of the tale of gloom and doom provided by Christian writers smitten with apocalyptic fever.92 First, it must be emphasized that the traditional Latin texts based upon the now lost “Imperial History,” which are regarded by modern scholars as the most reliable sources, do not provide any support for this catastrophic view of the period 275–76 in Gaul. Indeed, it is upon the fundamentally unreliable Historia Augusta that scholars depend for an account of the alleged success of the great barbarian invasion at that time.93 The vita of the Emperor Tacitus, and particularly his so-called “proclamation” regarding this supposed invasion, recounts the successes of the invaders and is the focus of the account provided in the Historia Augusta.94 The Life of Probus, who is depicted in the Historia Augusta as having come to Gaul in order to clean up the mess left by Tacitus and Florianus, provides additional information regarding the alleged disaster of 275–76. This is done while cataloging Probus’s effective reversal of the situation as “Victor” and “Restorer of the Peace” in Gaul.95 The possibility that such “honors” were mere propaganda proclamations cannot be ignored.96 The account in the Historia Augusta claims that Probus killed some 400,000 of the barbarian invaders.97 This type of exaggeration, of course, is but an extreme illustration of what may rightly be characterized as “the panegyric effect,” a rhetorical technique employed by a most untrustworthy author to praise his hero, at the expense even of plausible demographic detail. It must be emphasized that the traditional and most reliable sources for this period, Eutropius and Victor, do not provide this catastrophic view of the situation in Gaul during 275–76. Both authors, each in his own way, introduce information regarding this supposed great barbarian 92

93 94 95 96 97

Manley, Effects, p. 92, readily admits that the literary sources, which provide information about the invasion, are far from exact or accurate, and that modern scholars must in the end estimate damage from “some vague references by ancient authors.” Manley, Effects, pp. 42, 91–103; and H. Schönberger, “The Roman Frontier in Germany: An Archaeological Survey,” Journal of Roman Studies 59 (1969), 178. Vita Taciti 3.4, in Historia Augusta. Vita Probi 8.7–8, 14.1. The basic scholarly study of Probus is Gerald Kreucher, Der Kaiser Marcus Aurelius Probus und seine Zeit (Wiesbaden, 2003). Drinkwater, “The Germanic Threat,” pp. 20–30, provides a useful introduction to various propaganda techniques. Vita Probi 8.8.



Later Roman Fortification of Gaul

23

invasion largely for the purpose of proclaiming the skill demonstrated by the new emperor, Probus, and the success he enjoyed in dealing with the problems left behind by his incompetent predecessors.98 Eutropius says that “Probus, a man of remarkable military reputation, […] restored the Gauls, who had been occupied by barbarians.” However, Eutropius does not at all make clear what he means by “occupied,” and more importantly fails to claim that a large number of cities had been captured and taken under barbarian control by huge numbers of invaders. Nor, like the author of the Historia Augusta, does he mention by name any city that had supposedly been captured, taken under barbarian control, and subsequently liberated.99 Victor meanwhile, although also a propagandist of sorts for Probus, is even less forthcoming than Eutropius in regard to alleged barbarian successes. From a military perspective, Probus is credited with having “crushed the barbarian peoples who had invaded when our emperors had been killed.” Indeed, Victor makes no allusion to any “great invasion” of Gaul. Rather, Probus’s victory is left undifferentiated from any other military effort of the type discussed earlier.100 When we turn from the traditionally-cited Latin historians of the fourth century to later Greek sources and Christian accounts, there is no independent information to be found. Zosimus, apparently basing his work on that of Dexippus at this point, is intent upon discussing the effort by Probus in crushing the barbarians, and thus may also exaggerate the supposed “disasters” of 275–76. However, he makes no mention of the capture and rule of cities in Gaul by barbarian invaders.101 The gloom-and-doom scenario for 275–76 is also to be found, as might be expected, in the works of the “usual Christian suspects” writing during the fifth century. Not surprisingly they too call attention to the success of Probus, perhaps influenced by the inflated accounts of his accomplishments found in the Historia Augusta.102 Thus Orosius, for example, observes that, “at last [Probus] completely freed the provinces of Gaul, which had been occupied for such a very long time by the barbarians.”103 By contrast with the exaggerated claims found in the Historia Augusta and some Christian sources from the fifth century, there is a modicum of archaeological information that is positive in nature from an imperial perspective regarding this period. This evidence suggests that the imperial government carried on business as usual during the supposed invasions and occupations of 275–76. For example, a substantial corpus of inscriptions, especially from milestones, makes it clear that government road-building efforts in Gaul continued at a normal pace.104 In Eutropius, Breviarium 9.17; Victor, De Caesaribus 37. Eutropius, Breviarium 9.17. 100 Victor, De Caesaribus 37. 101 Zosimus, Histoire 1.67–68. 102 See the discussion by Manley, Effects, p. 100, n. 2, for a conflation of a selection of the sources. 103 Zosimus, Histoire 7.24 (my italics). 104 See, for example, CIL, XII, nos. 8923, 8928, 9001, 9002, 9076. Manley, Effects, p. 41, exaggerates the nature of the problems in Gaul during the reign of Tacitus and Florianus. Thus, he relegates these inscriptions to a note (p. 58, n. 147), which would seem to be a way of indicating their lack of importance. 98 99

24

Bernard S. Bachrach

addition, the mints continued in operation in order to pay the troops and facilitate commerce. This vital governmental service was still functioning despite a putative barbarian conquest of perhaps as many as seventy cities. Finally, there is sound evidence that a substantial Roman field army, at least a part of which was commanded by Bonosus and Proculus, was stationed in the northeast. It survived the alleged barbarian conquest and occupation of 275–76, and remained battleready but unchallenged during this period.105 Together, these indicators permit the suggestion that the Historia Augusta purveys a massive literary exaggeration concerning a supposed great invasion and its more or less extended occupation of Gaul. Before dismissing the the Historia Augusta’s account of a huge and successful barbarian invasion of Gaul in 275–76 as greatly overstated, if not a complete literary hoax, however, it must be recognized that from time to time even such seriously flawed texts have been known to provide traces of accurate information. It is thus particularly significant this context, with regard to rejecting both the information provided by the Historia Augusta and some of the wildly overheated modern interpretations of military events in 275–76 based upon it, that a meta-study of previous archaeological works has led to the conclusion that there is “a complete absence so far of any archaeological evidence to support the thrust of these literary references.”106 In consonance furthermore with the present state of the question as seen from this archaeological perspective, Ammianus Marcellinus makes clear to the careful reader that whatever information may have circulated in the later fourth century regarding significant disasters inflicted as a result of barbarian military efforts in Gaul following the death of Aurelian, such reports are simply not to be believed.107 An “Extension” of the Saxon Shore The very few and ineffective overland “invasions” of Gaul by trans-Rhenish barbarians, discussed above, cannot be considered to have been militarily significant, much less to have been important in either economic or demographic terms.108 Therefore, it is useful to call into question the widely held view that Probus undertook a massive long-term revision of imperial strategy in regard to the defense of Gaul in order to thwart barbarian invasions from east of the

De Caesaribus 37.3; Eutropius, Breviarium 9.17, for this force which was later defeated by Probus in the region of Cologne. 106 Schönberger, “The Roman Frontier in Germany,” p. 178, for the quotation recognizing the absence of archaeological evidence. However, Schönberger does take note of a statement attributed to the Emperor Julian (Convivium, 314b) in the 350s, obviously based on the claim of the Historia Augusta that some seventy cities had fallen in Gaul. 107 Historiae, in Ammianus Marcellinus, ed. and trans. John C. Rolfe, 3 vols. (Cambridge, MA, 1935–40). 108 As noted above (n. 89), Manley, Effects, pp. 42, 91–103, remains a good summary of the state of the question. 105 Victor,



Later Roman Fortification of Gaul

25

Rhine.109 Rather, it will be suggested here that the imperial effort to fortify a very large selection of cities, and indeed sites of lesser demographic and economic importance throughout the greater part of Gaul, was, in fact, the extension of a strategy initially developed in regard to the North Sea and Channel coast. This strategic matrix of forts and fleets, which came to be called the “Saxon Shore” during the later fourth century if not perhaps earlier, likely was initiated in either the late second or early third century to defend against an increasing number of pirate attacks on merchant ships and population centers lacking walls along the economically flourishing coasts of the North Sea and the English Channel.110 In light of the general dates for the construction of urban defenses in Gaul and Britain, the decision made by the imperial government to fortify these coasts was not based upon a static plan with a definitive end point. Rather, it was a process conceived as a continuous expansion, as needed, intended to deal with a situation in which barbarian blue and brown water operations in Roman territory were becoming more numerous and effective.111 Ultimately, fortifications were built west of Dover along the Channel coast and extended to the fort at Bitterne on the British side. On the mainland coast, the process of building fortifications moved west from Boulogne to Lillebonne, Rouen, Bayeux, Avranches, Aleth, St. Brieuc, Coz-Yaudet, Morlaix, and Brest on the northwestern tip of the Breton peninsula.112 This North Sea–Channel defensive system also saw fortification on 109 There

is general agreement that Probus built defenses throughout various parts of Gaul beyond the work already done on the Saxon Shore. See Johnson, Late Roman Fortifications, pp. 136–42; and John F. White, Restorer of the World: The Roman Emperor Aurelian (Staplehurst, 2005), pp. 152–53. 110 Stephen Johnson, The Roman Forts of the Saxon Shore (London, 1976), pp. 96–115, makes clear that trying to date the early history of the Saxon Shore from stylistic attributes of the fortifications alone is not fruitful. J. C. Mann, “The Historical Development of the Saxon Shore,” in The Saxon Shore: A Handbook, ed. Valerie A. Maxfield (Exeter, 1989), pp. 1–11, has a similar view. It seems likely that the origins go back to the period of the early empire as the Roman economy grew and the government saw the need to provide increased measures to defend against pirates. Initially, it is likely that various locations along the coasts of the North Sea and the Channel were provided with defenses, supported by the Roman fleet which had major bases at Boulogne and Dover. See Raymond Brulet, “The Continental Litus Saxonicum,” in The Saxon Shore, ed. Maxfield, pp. 45–77. 111 Piracy was thought to be such an important topic in the ancient world that it became an important theme in literature, as noted by Philip de Sousa, Piracy in the Greco-Roman World (Cambridge, 1999), pp. 214–18. Haywood, Dark Age Naval Power, pp. 30–32, reads as plain text, as is his custom, unreliable stories in unreliable texts regarding Frankish seamen operating as pirates in the Mediterranean and ultimately returning to their homes in the region of the lower Rhine. These exaggerated and highly unlikely stories do have value for us, however, as they indicate that some members of the authors’ later Roman audiences might well believe them, because of the real threat known to be posed by pirates, which required the government to invest extensive resources in defensive assets. 112 Johnson, Late Roman Fortifications, provides useful information regarding the fortifications at Boulogne, pp. 96–98; Lillebonne, pp. 86–88, 117; Rouen, pp. 87, 209; Bayeux, p. 87; Avranches, p. 94; Aleth, pp. 87, 93–94; Saint-Brieuc, p. 94; le Coz-Yaudet, p. 94; Morlaix, p. 94; Brest, p. 93–94; Vannes, pp. 89, 93–94; Nantes, pp. 92–94; Saintes, pp.106–07; Bordeaux, pp. 106–07; and Bayonne, pp. 110–12. See also p. 83, fig. 25, for a very good map.

26

Bernard S. Bachrach

the mainland extend along the Atlantic coast of Gaul from Vannes to Nantes, Saintes, Bordeaux, and Bayonne.113 The defensive strategy inherent in the development of the “Saxon Shore” complex of fortifications and fleets, including its Atlantic extension, provides prima facie evidence that the imperial government regarded an offensive approach to the problem of curtailing piracy to have been impractical. Of course, it was possible for the Romans to deploy naval assets along with troops against the barbarians settled northeast and east of the Rhine even as far east as the Elbe, where various elements of peoples called in the source Franks and Saxons dwelled. However, the Roman government had long since abandoned the idea of trying to conquer and occupy this region.114 As a result of the curtailment of a strategy of conquest and settlement, expeditions against these peoples would have to be repeated on a regular basis. This was because the barbarian infrastructure could easily be destroyed in the course of such operations, but since it was primitive it could easily be rebuilt. In addition, Roman offensive operations would be expensive, and the return in booty from such efforts, in terms both of material wealth and slaves, would be comparatively slight.115 Roman Strategy From a strategic perspective, the Roman fortifications discussed above played a dual role. First, the barbarian pirates, who raided along the coasts and up river, lacked the logistical support and the technological sophistication to undertake siege warfare.116 Thus the Roman population ensconced within the defenses of the Saxon Shore and its Atlantic extension was safe, to all intents and purposes, from direct enemy attacks. Second, troops stationed in these fortifications were positioned to seek out and harass pirate raiding parties which had landed in

“The Continental Litus Saxonicum,” pp. 45–77. the very useful study by Colin M. Wells, The German Policy of Augustus: An Examination of the Archaeological Evidence (Oxford, 1972). 115 A variety of sources from the fourth century provide a basis for understanding Roman policy in regard to the Gallic frontier in the east. Probus, for example, is reputed to have defeated the “barbarians,” some of whom are credited with being able to operate on the sea, and to have taken extensive booty, such as cattle, and levied a tribute in horses as well as in grain. In order to secure these gains, Probus is reported to have established fortified encampments (castra) on the east bank of the Rhine and beyond the previously defined border. He is said also to have stationed troops (milites) in these trans-Rhenish strongholds. These soldiers were provided with homes, fields, grain rations (annona), and storehouses for their grain from the resources of the newly occupied territory, i.e. the defeated barbarians paid for the support of the soldiers who were deployed to assure that they paid the required tribute. See e.g. Eutropius, Breviarium 9.17, and Victor, De Caesaribus 37.5, with discussion by C. R. Whittaker, “Agri Deserti” in Studies in Roman Property, ed. Moses I. Finley (Cambridge, 1976), pp. 148, 196. These accounts likely embody some considerable exaggeration, but they make clear that movement deep into “Free Germany” was not an imperial goal. 116 Thompson, Early Germans, pp. 131–40. 113 Brulet, 114 See



Later Roman Fortification of Gaul

27

their neighborhood and were trying to loot the countryside.117 In coordination with the above-mentioned fortifications, the Roman government stationed major fleets at Dover and Boulogne, which could operate in both the North Sea and the Channel.118 A third fleet was established on the Rhine just north of Cologne in order to support the Boulogne fleet and deal with pirates who attempted to raid imperial assets up-river. Finally, a fourth fleet was established at Étaples, just at the mouth of the river Canche. This site was to become the base for the later Merovingian emporium at Quentevic.119 The Romans, despite building extensive coastal defenses and deploying substantial naval assets, found it difficult to protect up-river assets, as earlymedievalists who study the Vikings are in a position to understand very well. For example, the very broad estuaries of the Seine or of the Gironde dictated that, in an era that lacked even telescopes, a substantial spectrum of difficulties faced the Romans in regard to the location, identification, and interdiction of seaborne hostiles. The problems inherent in locating and identifying seaborne raiders and pirates were exacerbated for the Romans when the invader had the benefit of the tides and currents, which were predictable, and of the wind, which often was not. The Roman riverine assets which were under threat because of these types of technical reality at times could not make sufficient speed to carry out an interdiction even if the enemy were sighted. It also should be noted that the pirates had the tactical advantage of being able to concentrate their assets at a particular time and in a specific place. By contrast, the defending Roman naval forces and the land-based forces with which they coordinated their efforts were likely deployed, more often than not, in comparatively small numbers because of patrolling obligations. On the whole, it was much easier for invaders to obtain intelligence regarding Roman military strength, routine patterns of operation, and deployment than it was for the imperial forces to learn about the plans of one or another band of pirates prior to a raid.120 Therefore, it was generally only after a raid had been undertaken up-river, during which Roman assets had been attacked, and the barbarians were trying to exit the mouth of river, that imperial naval forces were able to interdict a pirate fleet returning home with its booty. This defense in depth strategy on the part of the Romans was executed successfully, for example, by Carausius (d. 293).121 Early Germans, pp. 131–40. Louis Bonnard, Le navigation intérieure de la Gaule à l’époque gallo-romaine (Paris, 1913), pp. 217–29. 119 De Sousa, Piracy, pp. 225–27. 120 Regarding various debates concerning intelligence gathering see A. D. Lee, Information and Frontiers: Roman Foreign Relations in Late Antiquity (Cambridge, 1993); Fergus Millar, “Emperors, Frontiers, and Foreign Relations, 31 BC to AD 378,” Britannia 13 (1982), 1–25; C. R. Whittaker, Frontiers of the Roman Empire: A Social and Economic Study (Baltimore, 1994). 121 De Sousa, Piracy, pp. 226–29, provides a useful summary of these tactics, which remain controversial. See, for example, the discussions by P. J. Casey, Carausius and Allectus: The British Usurpers (London, 1994); and Norman Shiel, The Episode of Carausius and Allectus: The Literary and Numismatic Evidence (Oxford, 1977), pp. 155–56. 117 Thompson, 118 See

28

Bernard S. Bachrach

Fortifying the Interior of Gaul The fortifications and fleets of the Saxon Shore and its Atlantic extension were developed to protect imperial assets along the northern and western coasts of the Gallic region and their environs from raids executed by pirates. However, in the course of dealing with piratical raids for several decades, it became clear that defending only the coastal areas was insufficient to ensure security. Once a fleet of pirate ships had bypassed the coastal defenses and sailed up rivers where only the mouth or estuary was defended, the unwalled riverine cities and their suburbs in the interior were vulnerable to attack. As a result, it is argued here, many of the urbes of the Gallic provinces, which in large part were located in the major river valleys of Gaul, were fortified as extensions of the defenses of the Saxon Shore. As noted above, scholars familiar with the Viking invasions of Gaul during the ninth century will have little difficulty in seeing the pirates of the third century as forerunners of their northern counterparts more than half a millennium later. It is likely that the more general process of providing defenses for the up-river cities of Gaul was begun under Probus and ultimately resulted in the investment of huge quantities of surplus material wealth and labor.122 For example, the valley of the Seine and its tributaries were defended by major fortifications at Lillebonne and Rouen at the mouth of the river. Subsequent fortifications were provided through Evreux, Paris, Melun, Sens, Troyes, and as far south as Dijon, which could be reached in an efficient trek from the headwaters of the Seine.123 Cities located on the tributaries of the Seine were also provided with walls, as exemplified by the conversion of Senlis, Soissons, and Meaux into fortress cities.124 A pattern of fortifications very similar to that established along the Seine is evident throughout the major river valleys of the hexagon.125 For example, the lower Loire east of the fortress port city of Nantes saw the fortification of the urbes of Angers, Tours, and Orleans, along with numerous lesser strongholds scattered throughout the river valley.126 Amiens on the Somme, where the river narrows south of the estuary, also was fortified.127 Most of these newly fortified centers of population and wealth, on the Seine and along its tributaries as well as along the Loire, were, in general, not vulnerable to trans-Rhenish overland attacks. For example, when St. Jerome produced his much-exaggerated account of the military operations of the Alan, Vandal, and Sueve invaders who crossed the Rhine in 406, none of the cities he mentioned 122 Regarding

costs and the economy see Bachrach, “The Fortification of Gaul,” pp. 38–64, and the extensive literature cited there. 123 Johnson, Late Roman Fortifications, pp. 87–88 for Evreux; Paris, 101–02, 116; Melun, pp. 100–01, 116; Sens, pp. 99, 100, 104, 113; Troyes, pp. 100–01; and Dijon, pp. 84–86. See also Johnson’s map, p. 83, fig. 25. 124 Ibid., Senlis, pp. 95–97, 114; Soissons, pp. 95–98, 114; and Meaux, p. 100. 125 Ibid., p. 83, fig. 25 126 Ibid., Angers, p. 85; Tours, pp. 76–77; and Orleans, p. 101. 127 Ibid., Amiens, pp. 97, 113–14.



Later Roman Fortification of Gaul

29

as having been attacked and devastated by the barbarians was located in the northwest of Gaul. Even this barbarian force, which initially likely boasted a combined total of some 50,000 fighting men between the ages of fifteen and fifty-five, and is often characterized as having executed the most formidable invasion of later Roman Gaul, was unable to undertake effective overland operations in this region.128 Coastal and up-river defenses of urban centers which were vulnerable to attacks by pirates based in the Mediterranean are also a part of the larger picture. It is of course the case that those scholars who focus their attention on the frontier of Gaul and the territory of Free Germany tend to give scant attention to Mare Nostrum, and of course Mediterraneanists tend to repay the favor;129 however, in trying to understand the extensive construction of fortress cities during the later Roman Empire we should not ignore the fact that there was a long history of attacks on imperial shipping along the Mediterranean coasts and up various river valleys. Thus, it is hardly surprising that along the Rhône, north of the fortified ports of Arles and Marseilles, the urbes of Avignon, Orange, Valance, and Lyons at the confluence with the Saône were transformed into fortress cities.130 In fact, even Grenoble and Die, located on tributaries of the Rhône, were provided with high quality fortifications.131 The Romans were aware that the densely inhabited and wealthy population centers of the river valleys were at risk from pirates and sea raiders who successfully bypassed the coastal defenses. This is made clear by the fact that the cities and towns of Gaul which the imperial government did not license for wall construction during this period were those not easily accessible to attack on the coast or along one or another riverine system. For example, in the province of Aquitania Prima, Albi, Cahors, Rodez, Javols, and St. Paulien, as well as the nearby city of Lodève in First Narbonensis, were not transformed into fortresses during the period under discussion here.132 Similarly, in Novempopulana several urban centers in the interior – Lescar, Aire, Tarbes, Boas, Oloron, and Eauze – were not fortified.133 In the same province, by contrast, Auch, and to the north Lectoure, both on the banks of the river Gers, then Agen at its confluence with the

128 See

the strategic analysis by Bachrach, “Some Observations Regarding Barbarian Military Demography.” 129 Regarding the neglect of the Mediterranean, see Bernard S. Bachrach, “Charlemagne’s Mediterranean Empire,” in Mediterranean Identities in the Premodern Era: Entrepôts, Islands, Empires, ed. John Watkins and Kathryn L. Reyerson (Farnham, Surrey, 2014), pp. 154–72; and De Sousa, Piracy. 130 Johnson, Late Roman Fortifications, Avignon, Orange, p. 14; Valance, p. 15; and Lyons, p. 82. Concerning the fortification at Avignon, see Bernard S. Bachrach, The Anatomy of a Little War: A Diplomatic and Military History of the Gundovald Affair: 568–586 (Boulder, CO, 1994), pp. 75–76. 131 Johnson, Late Roman Fortifications, Grenoble, p. 104; and Die, p. 104. 132 Ibid., p. 83, fig. 25. It is possible that Rodez (ibid., p. 104) had some fortifications, but the surviving information is so slight these cannot be dated. 133 Ibid., pp. 108–09.

30

Bernard S. Bachrach

Garonne, flowing on to the port city of Bordeaux, were all fortified;134 as was Dax, one hundred kilometers up the Adour river which runs to the coast at Bayonne.135 Did the Romans Have a Grand Strategy? To ascertain whether the imperial government in fact pursued a long-term or perhaps even a grand strategy for the development of massive urban fortifications throughout Gaul depends, in part, upon understanding the institutional decisionmaking processes that were required for such an effort to be undertaken. Clearly, before deciding to fortify even one particular urbs, it was necessary for the requisite authorities of the central government to have gathered a substantial corpus of information regarding such a city’s relation to whatever strategic aims were being considered at the time. Under Roman law, conversion of an open city into a fortress city, or for that matter any significant military construction of a permanent nature, such as improving the defensive situation of an already fortified city, was carried out in the name of the emperor, and presumably with his written approval in some form of document that is no longer extant.136 Initially, the emperor was unlikely to have immediately at hand the kind of detailed information regarding many of the undefended urbes of Gaul that at one time or another were to be considered for the construction of extensive and expensive fortifications. As a result, the emperor’s local commanders in Gaul would be required to provide relevant military-related data, while tax assessors provided economic information.137 It should be noted, in addition, that various types of “maps” were available which served as the basis for the creation of the “Peutinger Table”; the emperor who made the decision and his advisers had accurate information concerning relevant geographical matters.138 With sound information regarding the location of various urbes in relation to other cities, the road systems, rivers, mountains, mountain passes, and other relevant topo134 Ibid.,

and 106, for Auch, Lectoure, and Agen, respectively. pp. 108–11, for Dax, and p. 83, fig. 25 for the map. 136 The Digest of Justinian, 1.8.9.4; 50.10.6, ed. Theodor Mommsen, Paul Krueger, and Alan Watson, 4 vols. (Philadelphia, 1985). Regarding the uncontroversial primacy of the central government in these matters, see S. T. Loseby, “Decline and Change in the Cities of Late Antique Gaul,” in Die Stadt in der Spätantike – Niedergang oder Wandel?, ed. Jens-Uwe Krause and Christian Witschel (Stuttgart, 2006), pp. 67–104 at 76; Johnson, Late Roman Fortifications, pp. 62–63; and Bachrach, “The Fortification of Gaul,” pp. 41–44. 137 See Bachrach, “The Fortification of Gaul,” pp. 41–44, and Bachrach’s synthesis, “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), pp. 192–218. 138 Concerning maps, see the useful material provided by O. A. W. Dilke, Greek and Roman Maps (Ithaca, NY, 1985), pp. 166–82; and idem, The Roman Land Surveyors: An Introduction to the Agrimensores (Newton Abbot, 1971); R. J. Tarrant, “Agrimensores,” in Texts and Transmission: A Survey of the Latin Classics, ed. L. D. Reynolds (Oxford, 1993, corrected repr. 1986), pp. 1–13; and the basic work by J. B. Campbell, The Writings of the Roman Land Surveyors: Introduction, Text, Translation and Commentary (London, 2000). Also of value is P. D. A. Harvey, Medieval Maps (Toronto, 1991), pp. 7–37. 135 Ibid.,



Later Roman Fortification of Gaul

31

graphical features of military value, the emperor and his advisers were able to identify which urban centers it would be appropriate to fortify in regard to one or another strategic imperative.139 Since we know that the fortification of any particular city required an imperial warrant, several observations are in order. First, it is safe to conclude that unwalled urbes were not converted into fortress cities willy-nilly because one or another local curia made a decision to fortify. An undefended city that was not subsequently fortified likely had not received permission from the emperor to have walls constructed, and without such permission construction was illegal. In addition, whatever funds and other support, such as military engineers and architects, were forthcoming from the central government for approved defense projects obviously would not be available for work that had not been approved. Unless the contrary can be demonstrated, we should presume that the urban fortifications that were constructed or repaired were approved from a military perspective because the emperor and his advisers were convinced that important enough strategic imperatives were at issue to merit the vast expenditure of surplus resources and the social dislocations inherent in such construction projects.140 Definitions Since imperial permission was required to construct fortress cities and the central government continued for more than a century to issue such permissions, it would seem reasonable to conclude that a long-term imperial strategy was being pursued in Gaul. Whether we can consider this lengthy policy of fortifying cites to be a grand strategy, is, however, another matter. Among modern scholars it has become common to see strategy as having two components: “grand strategy” and “campaign strategy;”141 but despite lengthy discussions by military officers and academics no definition of “grand strategy” has been developed that satisfies the epistemological criteria of both necessity and sufficiency. There is no consensus regarding what may be considered formally even a flawed definition of the term.142 In addition, it is clear that most historians

139 Regarding

the “reading” of the landscape, see Bachrach and Bachrach, “Landscapes of Defense.” 140 As noted by Bachrach, “The Fortification of Gaul,” pp. 38–64, these facts are uncontroversial. 141 Paul Kennedy, “Grand Strategy in War and Peace: Toward a Broader Definition,” in Grand Strategies in War and Peace, ed. Paul Kennedy (New Haven, CT, 1991), pp. 1–10 at 1. 142 It is useful to compare here the observations by Edward Mead Earle, “Introduction,” in Makers of Modern Strategy: Military Thought from Machiavelli to Hitler, ed. Edward Mead Earle (Princeton, NJ, 1943), p. viii; and those of Basil Henry Liddell Hart, Strategy, 2nd ed. (New York, 1974). In this context, Brian Bond, Liddell Hart: A Study of His Military Thought (London, 1973), is also of use. The dearth of philosophical rigor in this tradition is made evident in the treatment by Kennedy, “Grand Strategy in War and Peace,” pp. 1–7, and 185–86 of the same volume, Grand Strategies, ed. Kennedy, for additional literature. Kennedy and the tradition that he represents in this regard seem to have little interest in either the methodological or the epistemological problems involved in the development of definitions.

32

Bernard S. Bachrach

who, in fact, use the term “grand strategy” tend to avoid any attempt to develop a formal definition.143 Rather, scholars seem to focus on situational descriptions. Then following these descriptions, the intellectual baggage carried by each reader with regard to his own view of grand strategy, however flawed, results in either his rejection or acceptance of whether a grand strategy was at issue.144 The hope, it would seem in such contexts, is to avoid having one’s critics focus on what appears at this time to be the insoluble problem of labels as, for example, was the case in the futile discussion of whether one or another society was “feudal.”145 Thus, historians are wont to manipulate their readers through traditional rhetorical techniques and this, not surprisingly, may on occasion generate more heat than light.146

143 It

is perhaps ironic that the most important historical discussion of “grand strategy” in recent times for ancient and medieval historians, Luttwak, The Grand Strategy, 1976 ed., avoids definitions. Perhaps even more ironic is that his critics, while arguing vigorously that the Roman Empire had no grand strategy, also fail to offer definitions. This point is made by Everett Wheeler, “Methodological Limits and the Mirage of Roman Strategy,” Journal of Military History 57 (1993), 7–41, 215–40. 144 For example, Benjamin Isaac, The Limits of Empire: The Roman Army in the East (Oxford, 1990), vigorously attacks Luttwak, without taking him to task for failing to providing a methodologically valid definition of “grand strategy.” Isaac himself does not venture a definition and thus perhaps also saw the rhetorical folly of such an approach when proper definitions are so obviously lacking. However, this lacuna in Isaac’s work is particularly interesting from a rhetorical perspective in regard to his chapter 9, “Frontier Policy – Grand Strategy?” (pp. 372–418), and “Epilogue” (pp. 419–26). For example, Isaac writes (p. 418), “there was no Grand Strategy: Roman decision-making proceeded at a different conceptual level” – different, one supposes, from that of those who do conceive of a grand strategy. However, since there is no proper definition of “grand strategy,” the basis for asserting the “necessity” of a particular “conceptual level” for purposes of decision-making as a sine qua non for grand strategy cannot be entertained. It is not my purpose here to take sides in the debate concerning whether there was some reality that might usefully be considered Roman grand strategy; but rather merely to suggest that without proper definitions, the process of discussion as to whether or not one or another polity developed a grand strategy resembles blind men in a boxing ring who are likely to make contact more by accident than intention. At the very least, everyone must admit from the outset that there are no proper definitions of “grand strategy.” Thus, there is absolutely no point in trying to give the impression that X or Y has failed to meet the putative taxonomic specifics of a formally-valid definition. Isaac may well be able to demonstrate in compelling detail that Trajan and Hadrian functioned at a “conceptual level” very different from NATO planners in 1950 or 1960. However, until a particular “conceptual level” is established as a sine qua non for the existence of grand strategy, any argument regarding conceptual levels is not going to be convincing from an epistemological perspective. 145 See the magisterial deconstruction of the feudal model by Susan Reynolds, Fiefs and Vassals: The Medieval Evidence Reinterpreted (New York, 1994). 146 Wheeler, “Methodological Limits,” pp. 7–41, 215–40, on the whole does an excellent job of exposing the rhetorical devices used by the critics of Luttwak. He is less interested in Luttwak’s own exceptional mastery of techniques of persuasion, and, on occasion, presents arguments that are less than transparent. See also the examples provided by Arther Ferrill, “The Grand Strategy of the Roman Empire,” in Grand Strategies, ed. Kennedy, pp. 71–85 and 196–99; and Isaac, The Limits of Empire, both of whom are skilled at persuasive presentation of their case, and know the sources well.



Later Roman Fortification of Gaul

33

There seems to be no doubt, however, among both professional soldiers and academics, that there is some sort of “reality” underlying the term “grand strategy,” even if no valid definition has been developed. It would appear that the concept, however vague and putatively abused, is widely viewed as having heuristic value even if it is only for the purpose of stimulating research and discussion.147 Some, rather broad, areas of agreement have been adumbrated regarding various of the items that specialists tend to agree are useful to consider when discussing what has justly been characterized as “an elusive concept.”148 For example, it is now taken for granted that a grand strategy concerns the long term. However, there is no agreement on thresholds. Thus, while some commentators would seem to be satisfied that only a few years are sufficient, others prefer to think in terms of a few decades, and some would even have us think in terms of centuries.149 A strategy that is limited to the short term, say a year or so, is not to be considered “grand,” but rather, according to modern scholarly consensus, “campaign strategy.” However, if a campaign lasts perhaps three or four years then the question of labels can well become more controversial.150 In addition, there now seems to be a tendency to discard, or at least nuance, the notion of campaign strategy, and replace the label, either wholly or in part, with “operations.” Again, no definitions in a proper epistemological sense have been developed. Scholars instead rely on catalogs of attributes: bundles of characteristics. For example, one popular catalog reads, “Operational military activity involves the analysis, planning, preparation, and conduct of the various facets of a specific campaign.”151 Other apparent areas of agreement concerning what modern scholars are willing to regard as “grand strategy” are also identifiable in the literature. For example, grand strategy is generally regarded as transcending solely military matters. To put this aspect of the discussion in more positive terms, in order for a strategy to be considered grand it must have not only a long-term but also an important political and diplomatic component. Grand strategy is not primarily about winning battles or even about winning wars, although both certainly are important. Rather, grand strategy is about how the state develops the complex of its resources, including military assets, in order to shape itself within a process of defining its place among its neighboring polities.152 147 The

immense response to the work of Luttwak, cited above, is a good example of the heuristic value of discussions regarding grand strategy. For an excellent examination of many of the arguments, see Wheeler, “Methodological Limits,” pp. 7–41, 215–40. 148 Ferrill, “The Grand Strategy of the Roman Empire,” p. 78. 149 Kennedy, “Grand Strategy in War and Peace,” p. 4. 150 Ibid., pp. 1–2. 151 Allan R. Millett, Williamson Murray, and Kenneth H. Watman, “The Effectiveness of Military Organizations,” in Military Effectiveness, ed. Allan R. Millett and Williamson Murray, 3 vols. (London, 1988), 1:12. This pseudo-definition, in an epistemological sense, is cited with apparent approval by Kennedy, “Grand Strategy in War and Peace,” p. 185, n. 3. 152 Kennedy, “Grand Strategy in War and Peace,” pp. 2–3, with the cited literature.

34

Bernard S. Bachrach

This shaping process may perhaps be called “long-term policy,” and in such a process diverse military and non-military elements are brought together. These include, for example, how the military is to be organized, and the social basis for decisions regarding who shall serve and under what conditions. Sufficient economic resources, vital to the pursuit of a successful military policy, must be at the disposal of the government to sustain its grand strategy. In addition, the support of those people in a position either to make the policy work, or to thwart it, is of the highest importance. As a result, high morale is necessary. In short, much of what is needed to win a war in the short term is also required in the long term, as an integral part of a society, in order for it to be able to implement a grand strategy.153 The Roman government’s successful efforts over a period of more than a century to fortify numerous cities in Gaul, primarily to thwart barbarian piratical behavior, would seem to permit the conclusion that a long-term strategy was developed and sustained. When these imperial efforts in the long term are examined in regard to various bundles of characteristics often adduced by scholars, as discussed above, in one or another configuration to identify a grand strategy, it is seems likely that the Romans were engaged in what may perhaps be considered a grand strategy. The effort to fortify coastal and up-river cities in Gaul on a long-term basis transcended, or at least included more than, purely military goals. The strategy was aimed at protecting assets that were vital to a prosperous economy and to the well-being of the Gallo-Roman population as a whole. This long-term fortification strategy altered various fundamentals regarding the nature of urban life throughout Gaul. Nevertheless, people at the local level likely saw their morale raised as a result of the security provided by the walls erected to protect their cities, despite the economic and social costs incurred in the process.154

153 Cf.

the summary by Kennedy, ibid., pp. 4–6. the importance of piracy and responses to it see, for example, De Sousa, “Rome’s Contribution to the Development of Piracy,” in The Maritime World of Ancient Rome: Proceedings of the “Maritime World of Ancient Rome” Conference held at the American Academy in Rome, 27–29 March 2003, ed. Robert L. Hohfelder (Ann Arbor, 2008), 71–96; and De Sousa, Piracy.

154 Regarding

Map 1: Byzantium and the North, 400-800 AD

2 In Search of Equilibrium: Byzantium and the Northern Barbarians, 400–800 Leif Inge Ree Petersen

Understanding late Roman and early Byzantine long-term strategy towards the northern barbarians is a problematic endeavor. Recent historiography on late Roman frontier dynamics proposes a systematic Roman practice of building up a belt of client rulers, in the process profoundly changing societies brought within the Roman orbit. In return, these clients provided the Romans with extra sources of manpower and layers of security that stretched far beyond the frontiers.1 Unsurprisingly, the Classical frontier system of small client polities under Roman hegemony has attracted relatively little attention after it is assumed to have broken down in the north with the Gothic victory at Adrianople, the invasion of 406, and the subsequent establishment of federate groups within Roman territory. Arguably this was a problem the Romans brought upon themselves by integrating barbarians into their political structures and then involving them in Roman civil conflicts. The death knell for the Classical system of recognized client rulers in the north was the Hunnic hegemony that subverted what remained of the Roman client system.2 In contrast to its Roman predecessor, early Byzantine client management strategy has until recently been treated as rather reactive and ad hoc, with relations with barbarians regarded as an aspect of inter-state diplomacy.3 Obolensky 1

2 3

See C. R. Whittaker, Frontiers of the Roman Empire. A Social and Economic Study (Baltimore, 1994); Peter Heather, “The Late Roman Art of Client Management: Imperial Defence in the Fourth Century West,” in The Transformation of Frontiers: From Late Antiquity to the Carolingians, ed. Walter Pohl, Ian Wood, and Helmut Reimitz (Leiden, 2001), pp. 15–68; Michael Kulikowski, “Constantine and the Northern Barbarians,” in The Cambridge Companion to the Age of Constantine, ed. Noel Lenski (Cambridge, 2006), pp. 347–76. See in general Guy Halsall, Barbarian Migrations and the Roman West, 376–568 (Cambridge, 2007). Byzantinist contributions have until recently tended toward the more minimalist side in the “grand strategy” debate. Benjamin Isaac, The Limits of Empire: The Roman Army in the East (Oxford, 1990) is rightly critical of many of Luttwak’s conclusions, but underestimates Roman control beyond the frontier (see works cited in n. 1 above). Cf. John Haldon, Warfare, State and Society in the Byzantine World 565–1204 (London, 1999), pp. 34–44 and passim; Byzantine Diplomacy, ed. Jonathan Shepard and Simon Franklin (Aldershot, 1994), especially the contributions by Alexander Kazhdan, “The Notion of Byzantine Diplomacy,” pp. 3–24; and Evangelos Chrysos, “Byzantine Diplomacy, A.D. 300–800: Means and Ends,” pp. 25–40. For a comprehensive overview of late

38

Leif Inge Ree Petersen

identified a Byzantine system of satellite or allied states north of the Caucasus, Black Sea, and lower Danube that were deliberately cultivated to secure the Empire’s northern border, a system that persisted for centuries and gave rise to the Byzantine Commonwealth.4 However, his overall framework has not been further developed despite new archaeological and textual evidence, methodological advances, and a flourishing historiographical environment that have far advanced our knowledge. The focus of research has rather been Byzantine relations with particular geographical regions or barbarian groups, such as the Heruls, Gepids, Lombards, Avars, or Slavs, with fairly strict chronological delineations appropriate for each.5 Long-term developments are rarely followed much after the early seventh century, which becomes notoriously difficult due to the complex transition from the late Roman Empire to her medieval Byzantine successor, and to the extremely fragmentary nature of the sources, as is often pointed out with regard, for example, to Byzantine–Khazar relations.6 While Byzantine relations with the Khazars can hardly be described in terms of client politics, they undeniably constituted a successful strategic alliance against the Umayyads. For our purposes, they also had an important role in displacing Old Great Bulgaria, a very large Byzantine client state north of the Black Sea in the 660s, a process which led to the rise of multiple lesser Bulgarian groups, one of which disrupted Byzantine efforts to regain control of the Balkans in the late seventh and eighth centuries, and famously went on to become a long-term challenge.7 Even though the sophistication of early Byzantine political and diplomatic efforts against any number of polities or ethnic groups, such as Bulgars or Slavs, is well recognized, the conceptual framework within which to treat these efforts over time and space varies according to the interest and focus of the individual researcher. Thus scholars often interpret the sources’ description of events on

4

5

6

7

Roman and early Byzantine research, see Conor Whately, “Strategy, Diplomacy and Frontiers: A Bibliographic Essay,” in War and Warfare in Late Antiquity: Current Perspectives, ed. Alexander Sarantis and Neil Christie (Leiden, 2013), pp. 239–54. Dimitri Obolensky, Byzantium and the Slavs (Crestwood, NY, 1994), especially the essays “The Principles and Methods of Byzantine Diplomacy,” pp. 1–22, and “The Empire and its Northern Neighbors 565–1018,” pp. 23–74. Alexander Sarantis, “The Justinianic Herules: From Allied Barbarians to Roman Provincials,” in Neglected Barbarians, ed. Florin Curta, Studies in the Early Middle Ages 32 (Turnhout, 2011), pp. 361–402; idem, “War and Diplomacy in Pannonia and the North-West Balkans during the Reign of Justinian: the Gepid Threat and Imperial Responses,” Dumbarton Oaks Papers 63 (2009), 15–40; Walter Pohl, “The Empire and the Lombards: Treaties and Negotiations in the Sixth Century,” in Kingdoms of the Empire: The Integration of Barbarians in Late Antiquity, ed. Walter Pohl (Leiden, 1997), pp. 75–134; idem, Die Awaren: ein Steppenvolk im Mitteleuropa, 567–822 n. Chr. (Munich, 1988); Florin Curta, The Making of the Slavs (Cambridge, 2001). Thomas Noonan, “Byzantium and the Khazars: A Special Relationship?” in Byzantine Diplomacy, ed. Shepard and Franklin, pp. 109–32; James Howard-Johnston, “Byzantine Sources for Khazar History,” in The World of the Khazars: New Perspectives, ed. Peter Golden, Haggai Ben-Shammai, and András Róna-Tas (Leiden, 2007), pp. 163–94. See most recently Panos Sophoulis, Byzantium and Bulgaria, 775–831 (Leiden, 2011).



Byzantium and the Northern Barbarians

39

a particular frontier within a reactive ebb-and-flow model, in relation to the strength of the Empire and the intensity of challenges elsewhere. In such a view, Hunnic, Slav, and Avar raids on the Balkans during the sixth century indicated imperial overstretch punctuated only by a few short-term successes during the early years of Justinian, whose conquest policies exacerbated structural weaknesses. His appeasement of the Huns towards the end of his reign only further paved the way for the Lombard invasion of Italy and Avar and Slav encroachment in the Balkans in the years which followed it.8 Even if the later years of Maurice (582–602) saw significant success in driving out the raiders, he was famously deposed because of his policy of taking the war beyond the frontiers while the Empire’s finances could not sustain pay and equipment for the troops.9 The ensuing civil war in the early seventh century, along with Avar and Slav pressure, led to a total collapse of the Balkan frontier, giving way to a much weaker, but still occasionally effective, diplomatic approach towards the northern peoples, with military means only employed when Arab pressure was relieved, although some groups, such as the Balkan Slavs, were barely even worthy of notice until c. 800.10 In all cases, however, recent research has produced a much more nuanced picture. The Byzantines’ influence on their neighbors was profound, the strategic benefits of their presence significant, and very often their activities and movements can be directly tied to internal Byzantine conflict, which lends a whole new dimension to Byzantine relations with the northern barbarians. In fact, when the northern ethnic groups and polities from the Caspian Sea to the Balkans are examined within a consistent framework, a pattern of continuous and largely successful client management emerges. Despite the inevitable ebb and flow of pressure and relief, recognizable steps were consistently taken by Byzantine authorities, very often leading to the same long-term results: potential threats were contained, favored groups were allowed orderly settlement on Byzantine terms, and those safely within the Byzantine orbit were groomed for service and eventual integration into the Empire’s political structures. Of course there were serious challenges to this system. Some of these arose from external factors, such as political and military events beyond Byzantine control, especially in Central Asia, while others were of the Byzantines’ own making, through civil wars or usurpations. However, regardless of the potential severity of crises or their origins, the Byzantines were quick to respond and find a new equilibrium that ensured long-term Byzantine political and military goals were achieved. Furthermore, it is precisely during crises and ruptures that our scant sources, often infused with invective relevant to political conflict lines within Byzantine society which must be taken into account, reveal the mechanisms of Byzantine–barbarian relations. When read in the wider historical 8 9 10

E.g., John Moorhead, Justinian (London, 1994), passim. See in general Michael Whitby, The Emperor Maurice and his Historian: Theophylact Simocatta on Persian and Balkan Warfare (Oxford, 1988). Warren Treadgold, The Byzantine Revival 780–842 (Stanford, 1988), pp. 70–72.

40

Leif Inge Ree Petersen

context, such evidence illuminates not just the failures, but also the normal functioning of an effective system of client management and alliances that had its roots in the late Roman Empire, and was maintained with more success than is normally assumed through the eighth century and beyond. In many respects the Classical “toolkit” remained in use throughout the early Byzantine period, although the changed strategic situation of the fifth century created a number of new challenges as well as potential instruments. The basic diplomatic, ideological, and military repertoire consisted of imperial permission granted to client groups to settle on a frontier or (after the fifth century) within former Roman provinces, along with recognition of their rulers through inclusion in the imperial hierarchy of honors and offices. Some of the latter were symbolic, conferring prestige in varying degrees; others were of practical import, often in recognition of past or future service in imperial armies, and opening the road to settlement within the Empire and full integration into Byzantine administration and society. The line between the two types was often blurred, depending on the success of Byzantine long-term policies, and both included prestige gifts, subsidies in money or kind, trade rights, and a share in the glory of Byzantine power, still a coveted honor in the eighth century.11 However, there were significant modifications to Classical client management in the fifth and sixth centuries. The Huns and subsequent new nomadic groups represented an empire-building tradition that had arisen in Central Asia from Chinese, Iranian, and autochthonous steppe traditions. While most of the groups encountered are regarded as ethnically Turkic, they often represent hybridizations with various ethnicities, including nomadic Iranians and Mongolians. This capability to build hybrid political structures was routinely applied in later empire-building efforts. Possessing both a sophisticated ideological foundation for and a well-developed political tradition of forming large confederations from disparate ethnic groups, political traditions, and economic systems, such groups were uniquely capable of challenging Roman hegemony. This certainly left a profound imprint on successors, such as the Gepids, who controlled former Roman territory and were far less amenable (though not impervious, as we shall see) to Byzantine client management efforts.12 The loss of erstwhile Roman provinces to such groups would in part be offset by the presence of large sub-Roman populations stranded by military setbacks or deported from Roman provinces, especially in the fifth and seventh centuries. These populations played a significant role in shaping many of the ethnic

11 12

For the Classical means of client management, see references in n. 1 above; for Byzantine examples, see below generally. Hyun Jin Kim, The Huns, Rome, and the Birth of Europe (Cambridge, 2013); Peter Golden, An Introduction to the History of the Turkic Peoples: Ethnogenesis and State-Formation in Medieval and Early Modern Eurasia and the Middle East (Wiesbaden, 1992); Iver B. Neumann and Einar Wigen, The Steppe Tradition and Eurasian State Building, 4000 BCE–2017 AD (forthcoming). I thank Einar Wigen for a copy of the manuscript in preparation and our fruitful discussion on Turkic empire-building.



Byzantium and the Northern Barbarians

41

groups and polities that arose in the Balkans, allowing for an accelerated pace of integration or even assimilation into Byzantine cultural, economic, and political structures; thus many seemingly independent barbarian groups operating on the fringes of Byzantine territory were just as likely to (re)join the Empire in the right circumstances. Conversion to Christianity was not yet fully developed as a tool, but was increasingly applied to try to formalize Byzantine hegemony north of the Black Sea and the Caucasus, in the latter case often aided by local Christian client rulers. In return, the Byzantines acquired buffers against further nomadic threats, strategic allies in warfare, pools of manpower that could be raised for defense or internal power struggles, and a deep fringe territory that could be and indeed was (re)integrated into the structure of the Empire as resources and circumstances allowed. In sum, these policies formed a coherent strategic vision with consistent outcomes in the seventh and eighth centuries, ensuring that no largescale threat emerged in the north despite existential threats from the east and south. These measures could backfire and had to be recalibrated during periods of acute internal or external crisis; it was rather the very success of consistent policies, however, that led to the rise of Bulgaria as a real threat around 800, due to her close integration into the Byzantine political system. But even this threat was contained, and the subsequent recalibration of Byzantine policies in the ninth century, based on tried and tested means developed since the late Roman Empire, was so effective as to allow for the creation of the Byzantine Commonwealth. The Long Afterlife of the Classical Client System The cultural and political formation of barbaricum north of the Rhine, Danube, and Black Sea was largely the result of deliberate Roman client management over the course of centuries. As opposed to an inherently antagonistic view of barbarians as a threat to the Roman Empire, the weight of scholarship now recognizes a deep frontier zone beyond the borders that was largely under careful Roman supervision and management. While Roman control varied with distance, the minor, favored polities on the frontier, the “inner band” in Heather’s terms, stretching approximately 100  km out from the frontier, were within striking distance of Roman frontier garrisons and were thus held to high standards of loyalty.13 The most important means and ends were, as noted, subsidies, prestige gifts, access to markets, and Roman diplomatic recognition. These established the various minor chieftains and petty kings against other, competing barbarian groups with less favorable positions, further away from the frontier and with less access to political and economic advantages available with support from Roman authorities. The groups on the frontier were responsible for security, introducing 13

Heather, “Late Roman Art.”

42

Leif Inge Ree Petersen

or funneling in newcomers, and raising mercenary or auxiliary contingents that might eventually be integrated into the regular army. This also meant that they frequently played a role in civil wars, as the Franks and Goths did in the early fourth century during the conflict between Constantine and Licinius.14 There were of course severe consequences if things went wrong, as Roman retribution could be swift and merciless. The small client polities’ precarious position on the frontiers was well within Roman grasp (as noted by Bernard Bachrach in this volume regarding most of the invading groups of the third century), so they ran great risks in backing the wrong side in a civil conflict, as the Goths did during the revolt of Procopius in the 360s. Another aspect of the precarious position of petty rulers and statelets was that they were completely dependent upon Roman subsidies and recognition to maintain their status. If the client system temporarily broke down or was controlled by different parties, this would have unpredictable effects in barbaricum, as Romans could use barbarians against each other, while less-favored groups tried to exploit the situation by staging plundering attacks, either to acquire captives with skills that increased their economic basis, to blackmail Roman authorities into paying ransoms or subsidies, or even to win diplomatic recognition and a place in the client system as a result of demonstrated military prowess. This is probably one explanation for many of the raids recorded with particular frequency in the third quarter of the third century. While ostensibly an extremely risky venture, some low-level brigandage across the borders seems actually to have been tolerated by the Roman authorities, as long as it affected only lowly provincials; it was only taken seriously when it became too persistent, coincided with internal Roman strife, or afforded an opportunity for an emperor or governor to establish his military credentials.15 The Roman client system was famously credited by Whittaker with creating Germanic society, which developed from an economically primitive and socially egalitarian society in the first centuries A.D. to become, through Roman cultivation of favored elites and groups, highly stratified and economically complex by the fourth century. This system ostensibly suffered a crisis due to the formation of larger “confederations” – Franks, Alemanni, Goths – before major collapse at Adrianople in 378. However, serious doubts have been cast on the importance of that event, as the Goths were largely subdued by 382, and later Gothic movements cannot be genetically linked to the participants in the battle.16 There was already a long tradition of Gothic client or federate troops, of diverse origins,

14 15

16

Noel Lenski, “The Reign of Constantine,” in Age of Constantine, ed. Lenski, pp. 59–90. Noel Lenski, “Captivity, Slavery, and Cultural Exchange between Rome and the Germans from the First to the Seventh Century CE,” in Invisible Citizens: Captives and their Consequences, ed. Catherine Cameron (Salt Lake City, 2008), pp. 80–109. Thus Heather, “Late Roman Art,” and many other works on the Goths; but see especially Halsall, Barbarian Migrations; see also Walter Goffart, Barbarian Tides: The Migration Age and the Later Roman Empire (Philadelphia, 2006); Michael Kulikowski, Rome’s Gothic Wars from the Third Century to Alaric (Cambridge, 2007).



Byzantium and the Northern Barbarians

43

in the Balkans, and they continued to play a major role at the center of the East Roman military organization until the early fifth century, and even more so on the fringes of Roman control until the end of that century or beyond. Indeed, both major Gothic groups that crystallized in the west took shape on Roman territory and more often than not as recognized parties in internal Roman conflicts; Gothic troops were effectively integrated into the Roman army’s command and logistical system, with Gothic kings bearing titles such as magister militum and fighting in Roman civil wars. The degree of Gothic integration into (post-) Roman society is eloquent testimony to the transformative power of Roman client management.17 The Classical system arguably survived in some fashion until the consolidation of Hunnic power around 440, although its efficacy was compromised in the west by the settlement of semi-autonomous barbarian groups on Roman soil during endemic civil wars. At this crucial juncture, the Huns provided an alternative vision to the lone gravitational pull of the vast Roman Empire; at the same time Visigoths and Vandals, in the wake of West Roman civil conflicts, established their independence in 439.18 What the Hunnic hegemony demonstrated was a profound new reality – not so much their military power, but their political and ideological commitment to creating a vast confederation under their rule from elements primed to serve Roman imperial power for centuries. The origins of the Huns are disputed, but recent research accepts their connection with the Xiung-nu of Chinese sources, with a long intermediate period in close symbiosis with Iranian political traditions, both settled and nomadic, in Central Asia. Thus the Huns, when they established their power in the west under the sole rule of Attila after a period of fragmented client service in Roman armies in the 420s and 430s, were heavily informed by the cultural and political environment of the steppes, drawing on ideological and political traditions of Chinese, Iranian, and autochthonous steppe elements. This included a particular ability to form a genealogical superstructure that integrated the elites of a host of diverse ethnic groups, and even utilize a combination of nomadic and settled economic and political structures. Their empire was largely composed of erstwhile Roman clients, and many of their most important administrators were Roman officials and generals.19 The fallout of the Hunnic collapse after the death of Attila in 454, along with the wreckage of civil war in the west, set the stage for a new situation. Large swathes of previously Roman territory now fell outside the grasp of either Constantinople or any of the major successor states, particularly along the upper and middle Danube, where sub-Roman populations continued to live 17

18 19

H. G. W. Liebeschuetz, Barbarians and Bishops: Army, Church and State in the Age of Arcadius and Chrysostom (Oxford, 1990); Peter Heather, Goths and Romans 332–489 (Oxford, 1991); Halsall, Barbarian Migrations. Halsall, Barbarian Migrations, argues that this was indeed a symptom of failure to integrate into Roman politics, in large part caused by the Romans themselves. Kim, Huns; Golden, History of the Turkic Peoples; Neumann and Wigen, Steppe Tradition.

44

Leif Inge Ree Petersen

under different minor barbarian groups. In the Balkans and immediately north of the Danube, Germanic and Hunnic groups jostled for previously Roman territory or mere favors from Constantinople. In addition to elements from Attila’s empire, new Hunnic groups announced their arrival north of the Black Sea in 463, seeking diplomatic recognition from Constantinople, with significant recruitment to the army or even settlement inside the Empire. Gothic groups on the fringes of Roman control drifted in and out of Roman service, mostly conditioned by commitment to various factions in court politics in 470s, before Theoderic’s departure with the majority of them to Italy in 489. Others remained in the Balkans to be gradually integrated under Anastasios (491–518) and Justinian (527–65). To take just two examples: the usurper Vitalian’s army was a melting pot of different client groups that achieved formal status within the East Roman army; while Mundos went from Gepid warlord to Byzantine army commander in the 530s.20 By 500 or so, a situation recognizable throughout most of the reign of Justinian had begun to stabilize, with the Ostrogoths firmly established in Dalmatia and Italy, Lombards, Heruls and Gepids in the northwestern Balkans, any remaining Goths firmly integrated into the regular Roman army and being assimilated into mainstream Roman society, and Hunnic groups dominating the steppes from the lower Danube to the Kuban and north of the Caucasus and beyond. Most of these groups should have been firmly within the Roman client orbit. However, since the Classical client system, there had been a sea-change in ideology and political geography. The Huns had opened the possibility of a power not directly dependent upon Rome or Constantinople. This tendency was strengthened by the successful establishment of successor states on formerly Roman territory in the west, with infrastructure, administration and military capabilities that could, at least locally, challenge Byzantine hegemony. This was less evident on the lower Danube and the steppes, as no potential hegemonic power was yet on the horizon, but that situation soon changed after the 550s, with the Hunnic group requiring continuous supervision to prevent it from creating a larger political structure. In the meantime, the sheer number, and in the case of the Slavs amorphous nature, of groups that could threaten the Empire and needed political and military management was a challenge that Justinian handled far better than he has been given credit for, in face of internal discontent, plague, and large-scale wars on two fronts. The Gepids were the foremost example of changes brought on by Hunnic rule. As one of the main supporters of Attila’s regime, they inherited the core lands of Attila’s empire with control over Sirmium, which represented a major element of infrastructure and population and provided a strategic key to the Balkans. Their political ambitions were significant, as they aimed for regional hegemony and used their position along the Danube and her tributaries to blackmail the

20

Patrick Amory, People and Identity in Ostrogothic Italy (Cambridge, 2003), pp. 127–31 (Vitalian), 397–98 (Mundo), and passim.



Byzantium and the Northern Barbarians

45

Byzantines for concessions. Most of Justinian’s efforts in the Balkans were in fact aimed at containing them and preventing the rise of a larger state under their hegemony. Two other major Germanic groups were effectively used to ensure balance: the Lombards and the Heruls. The Heruls had been a mainstay of the West Roman army in the fifth century, but forced out after the Ostrogothic takeover of Italy; they performed well and loyally in East Roman armies in the sixth century. The Lombards too were objects of Byzantine diplomacy and client relations, serving in Byzantine armies, often alongside Herul contingents. Sarantis demonstrates a successful policy of containment, whereby the most ambitious group, the Gepids, was kept in check. The Heruls were gradually integrated into Roman society and disappear from view as they become fully assimilated. The Lombards were geographically more distant, but shared in a common military culture and seem to have fused with the local sub-Roman population and other groups in Pannonia, with imperial support and blessing, as a result of seeking and receiving settlement from Justinian in 547. This allowed the use of Lombard mercenary or allied contingents on many fronts from Italy to the Caucasus in a number of different capacities. By the time the Avars arrived, the northwestern Balkans were under strong Byzantine influence. In light of the fate of the Heruls, the profound Byzantine influence on the Lombards, and their fusion with former Roman provincials in Pannonia, we cannot exclude the possibility that Justinian had similar designs on them as on the Heruls.21 In the meantime, a new ethnic group had appeared on the horizon. The origin of the Slavs is a controversial and difficult topic and will largely be passed over here; briefly, they may have spread from a region between the northeastern slopes of the Carpathians, the upper reaches of the Dniepr, Dniestr, and Vistula rivers, and the Pripet Marshes. The earliest Slavs had a peculiar social organization as semi-nomadic pastoralists practicing slash-and-burn agriculture, but they seem to have been highly militarized and astonishingly effective in expanding over large territories in a short period of time. Seemingly egalitarian, their obscure origins before 500 were presumably on the fringes of, or subsumed under, more powerful Germanic and Hunnic elites operating in the same region. They had no major state structures of their own, and little in the way of conspicuous consumption preserved in archaeological finds. However, a defining trait in their social organization seems to have been that all free males were warriors, giving them an edge over more populous but less militarized groups, and that captives after a time were assimilated into their society as free members, allowing for rapid expansion, demographic growth and mutual assimilation with autochthonous populations. While they may have participated in Hunnic ventures under other groups, with the movement of most (Germanic) elites onto Roman territory or clustering along the old frontier on the middle Danube, the Slavs seem to have found room for rapid expansion on the lower Danube. They are recognized as difficult for the Byzantines to deal with, due to their simple agricultural economy, 21

With the literature cited in nn. 1 and 5, see also Neil Christie, The Lombards (Oxford, 1995).

46

Leif Inge Ree Petersen

amorphous political organization in loosely associated villages, talent at guerrilla-style warfare in forests and broken landscapes, and effective use of dugout canoes on tributaries of the Danube.22 The Byzantines soon adapted to this potential new threat. For one, there may in fact have been somewhat different types of social and political organization among different groups of Slavs (although the nature of this is controversial), reinforced by their ability to establish themselves in and fuse with different types of socio-economic and political environment. There was also some influence from Turkic nomadic groups (on which see further below), which may have affected the Antae, a Slavic-speaking group living on the steppes northeast of the Danube delta; some of the first Slavs drawn into the Byzantine gravitational zone provided cavalry alongside Hunnic clients for expeditions as early as 537.23 There are also signs of a consistent Byzantine policy of trying to encourage more permanent political structures among Slavs to allow for effective client management, but this was difficult as the Slavs were lacking in permanent leaders of significance – Procopius goes as far as to describe Slav political organization as “democratic,” which was hardly meant as a compliment. Curta is very positive as to the effect of Byzantine policies, but identifies a rather abrupt Byzantine withdrawal from engagement, based upon the evidence of a lack of coin hoards and the simultaneous building of fortifications in the 540s.24 However, the sources describe consistent use of Slavs and Antae in Byzantine armies into the 550s, while annual subsidies continued to be paid out to Hunnic groups, as related by Procopius. Jankowiak has pointed out a significant feature of later Byzantine diplomacy towards Scandinavians and Rus which has puzzled scholars, as the vast majority of silver hoard finds in Scandinavia and Norse settlements in Rus are from Islamic sources: while coins certainly attest to close and voluminous economic relations, they were also the ultimate objective of interaction between the Islamic world and Scandinavia – the interchange ended with the monetary transaction. In contrast, Norse relations with Constantinople relied far more on cultural capital, recognition, non-monetary gifts and subsidies, and opportunities for service in Byzantine armies. Similarly in the sixth century, once the imminent threat of endemic raiding across the frontier had been contained, relations stabilized to the extent that the Byzantines could dispense with cash payments and rely on other means of dominance which are less easily observable in the archaeological record, but still witness to a distinct cultural zone north of the lower Danube and around the eastern bend of the Carpathians.25 22 23 24 25

Curta, Making of the Slavs; Paul Barford, The Early Slavs: Culture and Society in Early Medieval Eastern Europe (Ithaca, 2001). Procopius, History of the Wars, trans. H. B. Dewing, Loeb Classical Library (London, 1914– 28), 5.27.1–5 Curta, Making of the Slavs. Note the observations of Marek Jankowiak, “Byzantine Coins in Viking-Age Northern Lands,” in Byzantium and the Viking World, ed. Fedir Androshchuk, Jonathan Shepard, and Monica White (Uppsala, 2016), pp. 117–39, pointing out that even though the volume of coins found in the north from the Samanid emirate was greater than from Byzantium, they reflected a



Byzantium and the Northern Barbarians

47

Client management techniques began to pay off by 560s, when we find the signs of a political elite in the preserved fragments of Menander. The Antae seem to have had a fully-fledged ruling class and diplomatic relations with the Huns further east, while the famous Daurentius or Davritas, ruling a group of Balkan Slavs, expressed a conquest ideology challenging Avar hegemony.26 When the Avars took over, they found Slav groups effective tools for their expansion, and used them for military engagements alongside Gepid and Hunnic groups. We begin to see a clear differentiation into political zones based on degrees of Avar or Roman influence, with some Slavs under direct Avar rule, others as tributaries or allies, as attested in the Miracles of St. Demetrius, while others again (such as Davritas) did their best to remain free of Avar hegemony. While their influence may well have played its part in the formation of the Slavs, the Avars’ hegemony was relatively short-lived, with control over significant Slavic populations outside the Carpathian basin only briefly in the 580s and again c. 610–26. This must be measured against a century or more of Roman political, cultural, and military influence, during which, in the 570s, the Avars themselves were used by the Byzantines, admittedly with unfortunate long-term results, by the to punish the Slavs. Indeed, Curta’s overarching argument is that the Slavs as an ethnic group were created by interaction with the Byzantines.27 However, as we shall see, the benefits in the form of a manageable client system, and how this worked through Byzantine integration of a Slav ruling elite, only becomes clear in our evidence in the wake of the Avar collapse, when the Byzantines could establish a fully-fledged client system using the recentlydeveloped elite ruling the embryonic Slav polities. In the meantime, the lowest reaches of Danube and the Pontic steppes were dominated by Hunnic (as far as we can tell, Turkic-speaking) groups, some of which constituted fragments of the former Hunnic empire while others were more recent arrivals. Despite their ostensibly devastating raids, these groups actively and consistently sought Byzantine recognition, titles, and subsidies, and actually seemed to take offense and sulk at Byzantine diplomats when such recognition was not forthcoming. Several of these groups were soon to be co-opted into the Byzantine military system. The most famous during the later years of Justinian’s reign were the Kotrigur, Utigur, and Sabir Huns, settled respectively west of the Don (Kotrigurs), east of the Sea of Azov (Utigurs), and north of the Caucasus (Sabirs), close to the Alans, an Iranian-speaking nomadic group that had the closest and most longstanding ties with the Byzantines.28 The Kotrigurs

26

27 28

far more straightforward economic relationship, while the complex cultural and political relations between the Rus and Byzantium were regulated by a host of other means in addition to monetary transactions. Menander, The History of Menander the Guardsman, ed. and trans. Roger C. Blockley (Liverpool, 1985) [henceforth Menander], fragment 5.3. For further discussion and later examples of Slav petty rulers, see Curta, Making of the Slavs, pp. 324–31. Curta, Making of the Slavs. For an archaeological perspective, see Irina Arzhantseva, “The Alans: Neighbours of the Khazars in the Caucasus,” in World of the Khazars, ed. Golden, Ben-Shammai, and Róna-Tas, pp. 59–73.

48

Leif Inge Ree Petersen

provided significant contingents for Byzantine armies in Africa and Italy, while the Sabirs entered Byzantine service in the east but also found service in the Sasanian Empire.29 The 540s and 550s were ostensibly a time of grave crisis for the Byzantines. After the plague, recommencement of the interminable Ostrogothic war, and a horrific war against the Persians beginning in 540 and lasting until 562, Justinian was able to use Hunnic groups against each other while the field armies were engaged in Italy and the east. Inter-Hunnic rivalry was clearly based on competition for favors from Byzantium, where raids were deliberate demonstrations of military prowess, intended to pressure the Byzantines into providing privileges and opportunities for service. Utigur–Kotrigur rivalry played out in several rounds. The Utigurs were deliberately provided with rich annual subsidies, both to keep them settled where they were and, as subsequent events demonstrate, to use them as a counterweight against other groups on the steppes. The Kotrigurs to their west, jealous of their recognition and subsidies, were in reach of Roman territory and engaged in significant raids in the Balkans, at one point in collusion with the Gepids, who feared Justinian’s deployment of the Lombards and so provided ferry services to the Kotrigurs. Justinian however mobilized the Utigurs to attack Kotrigur lands, with the quadruple benefit that, first, the main force of Kotrigurs had to withdraw to protect their homelands (Utigur raids having been announced to Kotrigur raiding parties for shock effect by Byzantine ambassadors); second, thousands of Byzantine captives were able to flee back to Roman territories; third, a significant Kotrigur contingent which had previously served the Empire, and had not participated in the Gepid-allied expedition, now sought and received settlement inside it; and fourth, the two major Hunnic powers were then locked in conflict with each other. A fifth benefit could be adduced, as the Gepids’ attempt at using Huns to build up regional hegemony failed.30 A similar situation was described by Agathias a few years later in 559, although this at first sight seemed even more serious, as Kotrigur armies reached as far as the fort of Thermopylai in Greece, the Chersonese wall in Thrace, and even plundered near Constantinople. Although this is often taken as a sign of the Empire’s weakness, the effective program of fortification in the previous decades had forced the Kotrigurs to penetrate extremely deep into the Balkans to find worthwhile plunder or make a lasting political impression.31 The Kotrigurs again engaged in attacks with the explicit objective of forcing the emperor to give them recognition, gifts and subsidies – in other words, they were angling for “recognized client” status, which had not been forthcoming. However, their attempts at Thermopylai proved miserable and costly failures. The great Belisarius was brought out of retirement to defeat the Kotrigur raiding party near Constantin29 30 31

Procopius, Wars, trans. Dewing, 8.11.22–31; 8.13.6–7; 8.14.4–11. For the background to and development of the Hunnic rivalry, see Procopius, Wars, trans. Dewing, 8.18–19; for Huns in the Byzantine army, see, e.g., ibid., 8.26.13. See most recently Alexander Sarantis, Justinian’s Balkan Wars: Campaigning, Diplomacy and Development in Illyricum, Thrace and the Nothern World A.D. 527–65 (Prenton, 2016).



Byzantium and the Northern Barbarians

49

ople with a small force of his old retainers, circus factions, and local farmers. Obviously most garrison troops were busy guarding their forts and cities dotting the Balkans. Agathias and contemporaries, who did not at first understand why the emperor let the Kotrigurs get away after inflicting significant suffering on the rural population of Thrace, soon learned of the emperor’s master plan: the Utigurs were again mobilized to attack Kotrigur homelands, and the two groups were locked in a struggle which ensured that neither could gain the upper hand and control the whole of the Pontic steppes.32 The success of this policy has to be measured in the context of Central Asian traditions of large-scale tribal confederacies that had proved so effective under Attila and still existed under the Eurasian Avar khaghanate, about to be displaced by the even greater Kök Türk khaghanate, with powerful later groupings on the Pontic steppes under the Old Bulgarian and Khazar khaghanates, and even in the Balkans under the Avars and later the Bulgars. As the Utigurs, Kotrigurs, and Sabirs played a significant role in several of these confederacies, and shared in their political, material, military, and ideological culture, it is quite astonishing that no long-term threat appeared near Byzantine territories during the reign of Justinian. Instead, however, through the creation of ties of dependency unmistakably similar to earlier Roman client politics, all the Hunnic groups’ energy was absorbed by competition for attention and relatively small favors from Constantinople, making Justinian the arbiter of the steppes, even if no Byzantine armies ventured onto them, from the Danube delta to the Caucasus. Given the economic and manpower shortages caused by plague, the Balkans held very well. However, the rise of the great Kök Türk khaghanate in the early 550s posed an enormous challenge, as the fallout threatened and eventually disrupted Justinian’s client system shortly after his death. The origins of the European Avars and their connection with the powerful Central Asian Avar khaghanate are obscure. Briefly, the Central Asian Avars had ruled the steppes from the fifth century until c. 552, when they were displaced by the Türk Ashina clan who organized the Kök Türk (Celestial Turk) khaghanate. It is likely that the Avars who fled westwards did not represent the ruling elite that gave the original khaghanate its name, but were largely refugees from the main tribal groups that had formed the basis of Avar power and thus used the Avar name as an indication of ideological legitimacy. However, their arrival within the Byzantine orbit was remarkably humble and is in fact revealing of the Byzantine system of detecting and absorbing newcomers from Central Asia. The Avars made their arrival known through the Alans, whom we have encountered. A long-established client group and in effect the Empire’s sounding posts on the northern side of the Caucasus, the Alans introduced the Avars to the commander of the army in Lazica (western Georgia). The 32

Agathias, Agathiae Myrinaei Historiarum libri quinque, ed. Rudolf Keydell, Corpus fontium historiae Byzantinae 2 (Berlin, 1967); trans. as Agathias: The Histories, trans. Joseph D. Frendo (Berlin, 1975), 5.11.1–25.6.

50

Leif Inge Ree Petersen

first encounter is recorded in a preserved fragment of Menander and worth quoting in full: Concerning the Avars: after many wanderings they came to the Alans and begged Sarosius, the leader of the Alans, that he bring them to the attention of the Romans. Sarosius informed Germanus’ son Justin, who at that time was general of the forces in Lazica, about the Avars, Justin told Justinian, and the Emperor ordered the general to send the embassy of the tribe to Byzantium. One Kandikh by name was chosen to be the first envoy from the Avars, and when he came to the palace he told the Emperor of the arrival of the greatest and most powerful of the tribes. The Avars were invincible and could easily crush and destroy all who stood in their path. The Emperor should make an alliance with them and enjoy their efficient protection. But they would only be well-disposed to the Roman state in exchange for the most valuable gifts, yearly payments and very fertile land to inhabit. Thus spoke Kandikh to the Emperor.33

Menander goes on to blame Justinian’s age and feeble state for not crushing the Avars, but this was of course written with the benefit of hindsight, when the Avars had established themselves in the Carpathian basin and become a considerable nuisance in the Balkans. Interesting for our purposes, however, is the nature of the gifts, which says something of the true relative strengths of the parties in the late 550s: The Emperor put the matter up for discussion, and when the holy senate had praised his plan and its shrewdness, he immediately sent the gifts: cords worked with gold, couches, silken garments and a great many other objects that would mollify the arrogant spirits of the Avars.34

Interestingly, few of these objects are likely to be preserved archaeologically, but they were clearly among the most significant gifts in the Roman diplomatic repertoire. Their value lay primarily in the cultural capital of the Romans as producers and dispensers of exclusive luxurious materials, the distribution of such items being carefully controlled and their conferred prestige dependent upon their use as symbols of power at the court in Constantinople and hence their embodiment of the grace of the Emperor to a supplicant.35 Through them the Emperor signaled that he deigned to include the recipient within the structures of imperial power, thereby establishing a symbolic hierarchy that would immediately be obvious to contemporaries, Byzantine and barbarian alike. This is also clear from Justinian’s expectations in return for such gifts, as he sent one of his personal bodyguards to encourage (perhaps better put, to instruct, as subsequent favors depended on how well these instructions were carried out) the Avars to subdue specific tribes north of the Caucasus and on the Kuban: Unigurs (probably Onugurs), Zali, and, significantly, Sabirs. The rationale behind this 33 34 35

Menander, fr. 5.1. Menander, fr. 5.2. Robert Lopez, “Silk Industry in the Byzantine Empire,” Speculum 20 (1945), 1–42.



Byzantium and the Northern Barbarians

51

was surely that the Sabirs had gone from Byzantine to Persian service in the 550s and had thus proven themselves unreliable as clients. This was presumably also the case for the other groups that were similarly punished.36 After their performance of these and possibly other services on the way across the Pontic steppes, the Byzantine plan was to grant the Avars land in Pannonia II, just south of Gepid territories and previously settled by the Heruls, who by now appear to have been fully absorbed into mainstream Roman society and dispersed with army postings from Italy to the Caucasus. Such settlement was completely normal (compare the Kotrigur settlement noted above) and an efficient way of integrating clients into the formal defenses and eventually social structures of the Empire. As opposed to most other client groups, however, the Avars, according to Menander, deliberately exploited these client management procedures for their own purposes. Not satisfied with the territory offered, they wanted to stay on the steppes. An Avar envoy revealed to the general Justin, whom he had befriended, a more sinister plan, to use the grant of access across the Danube to attack (and presumably conquer, rather than raid) Roman territory in the Balkans. However, the Avars’ hesitance in accepting settlement so far away from the steppes proper should be explained by the fact that they needed sufficient pasturage for their large cavalry armies and control over, or at least access to, other nomadic tribes as a necessary precondition for any independent power, if that was indeed their plan at this early date.37 Justin’s warning and the Avars’ refusal to accept regular client settlement occurred after they seem to have gone beyond their mandate by attacking the Antae and allying with or establishing hegemony over the Kotrigurs, who later became important elements in Avar armies. When lengthy negotiations over Avar settlement failed, Justinian made sure to provide their ambassadors with the now accustomed gifts. He also provided another client boon by allowing them to make purchases of other prestige clothing and arms, which must have been made possible through subsidies in cash, or perhaps plunder or tribute from other client groups, which ensured that subsidies returned to the Byzantine economy. However, aware of the danger, the Byzantines mobilized along the Danube and confiscated the purchased arms as the Avar embassy left the Empire, which Menander describes as the beginning of hostilities.38 Yet in the end, it took well over two decades for the Avars to establish more than a regional hegemony in the Carpathian basin and seriously threaten the Balkans. The operational and diplomatic details are well described elsewhere; our focus remains upon how the Byzantines tried to contain, and when possible manage, their rogue client. The emperor Justin (r. 565–74), clearly suspicious after his namesake’s experiences of their treacherous intentions, flatly rejected the customary prestige gifts when Avar ambassadors arrived upon his acces36 37 38

Leif Inge Ree Petersen, Siege Warfare and Military Organization in the Successor States (400– 800 AD): Byzantium, the West and Islam (Leiden, 2013), pp. 543–46; Menander fr. 5.2. Menander, fr. 5.4. Menander, fr. 5.3.

52

Leif Inge Ree Petersen

sion. With hindsight this may be regarded as a foolish move, but for contemporaries the loss of imperial favor meant an appalling loss of prestige, and the Avars were reduced to trying their luck against the Austrasian Franks, with rather mixed results. While still militarily formidable and victorious in battle at a second attempt, the Avars had to rely on supplies from the Franks to survive after they had concluded a treaty. Their famous alliance with the Lombards against the Gepids in 567 (leading to the emigration to Italy of the former in 568 and submission of the latter to the Avars) clearly upset the carefully balanced system set up by Justinian, but the Romans were able to regain Sirmium, thus securing the Balkans and frustrating the Avars, who claimed the city by right as the conquerors of Gepid lands. The Avars attempted to take the city, but failed miserably, and were unable to extort even token subsidies in return for ceasing hostilities, which they explicitly stated would cause them serious loss of face. Although they partly made up for this by sending subject Kotrigurs on a raid against Dalmatia, real Avar power was still well over a decade away.39 The bone of contention was Sirmium, which the Avars still claimed as conquerors of the Gepids; but meanwhile they also sought a formalized relationship with the Byzantine state based on the precedent of recent client arrangements. This was in line with the steppe political conception that subduing a group legitimized a claim to their previous territories, sources of income, and diplomatic status. Having defeated or incorporated both Utigurs (a rather dubious assertion) and Kotrigurs, they likewise claimed back-pay of the annual subsidies awarded to them by Justinian, plus, of course, the resumption of the subsidies granted to them in 558 but halted by Justin. Military action was inconclusive, with both sides suffering defeats in battle in the early 570s, and Byzantium was embroiled in yet another Persian war from 573 that required most available military resources. Tiberius, who had faced the Avars both in battle and as a diplomat, finally conceded when he came to power in 574, negotiating a vast annual payment of 80,000 solidi, which allowed him to make a major recruitment drive in the Balkans in order (successfully) to stabilize the frontier with Persia. Depleting his defenses came at a cost, as the Slavs used the opportunity to organize vast raids, but Tiberius’s treaty with the Avars paid dividends when they were ferried down to Wallachia to attack the homelands of the Slavs while most of their men were raiding in the Balkans in 577 or 578. While they ravaged Slav villages and freed many Roman prisoners, the performance of their cavalry against Slav guerrilla tactics in the woodlands was not satisfactory, and the Slavs flatly rejected Avar claims to hegemony. However, perceived Byzantine weakness led to a far more aggressive Avar policy as the Avars mobilized their Pannonian clients, subject Slavs, allied Lombard shipwrights, and captured Roman engineers to methodically besiege Sirmium from 579. The city fell around 581, and despite a brief peace treaty, the Avars then besieged and captured Singidunum in 583.40 39 40

Menander, fr. 12.1–7. Menander, fr. 21; see further Whitby, Emperor Maurice, pp. 87–89; Pohl, Die Awaren, pp. 61–76 with references.



Byzantium and the Northern Barbarians

53

The problem for the Byzantine authorities, of course, was the existence of only one hegemonic power in the Carpathian basin (as opposed to the two or more groups that could be set off against each other under Justinian). They now had to face a simpler, but potentially far more threatening scenario, with an emerging Avar khaghanate in Pannonia, Slavs raiding the Balkans from Wallachia, and the establishment of the Kök Türk khaghanate north of the Caucasus that threatened to swallow up Roman clients. The Avar hegemony resembled the threat posed by the Huns in the fifth century, as they subverted long-established Byzantine clients and used them against the Empire. With a shared enmity against both Persians and Avars, the Turks took the initiative to forge an alliance with Byzantium in 568, eagerly seized upon by Justin II. The Byzantines carefully cultivated their alliance with the Turks through several difficult diplomatic missions deep into Central Asia, and let themselves be drawn into war with Persia at the Turks’ behest from 572.41 Not only did the war go poorly at first, but it would allow the Avars and Slavs a comparatively free hand in the Balkans. Thus the appearance of the Turks had doubly unbalanced the Byzantine client system. Roman client management tools normally included military means to prevent this when diplomacy failed, but with most military resources focused on the east, the Byzantines had to bide their time and contain the danger for the time being. It needs to be emphasized that the existence of two expansive khaghanates, one along the Balkan frontier, another north of the Caucasus and Black Sea, in place of a dozen or more minor client groups, was an unprecedented situation in Roman and Byzantine history. The fact that such a situation had been held off for so long, and that it did not do more damage once it arose, is testament to Byzantine client management strategy. On a larger scale, the Byzantines managed to balance the great threats by using the Avars against the Slavs, as noted above, while the Kök Türks, Byzantium’s strategic ally against Persia, constituted a potential threat against the Avars, whom the Turks regarded as subjects and escaped slaves. Furthermore, as Pohl points out, the seemingly vast payments extorted by the Avars in the decades around 600 were not necessarily regarded as tribute in contemporary eyes: gifts that could not be requited through counter-gifts or military favors actually established the superiority of the giver.42 The relationship with the Kök Türks, in contrast, was on a more equal footing; but even in this case we have indications that the Byzantines used their listening posts on the Crimea and in the Caucasus in preparation for subverting the Kök Türk confederation. The complexity of this diplomatic balancing act becomes clear from the great displeasure expressed by the Kök Türk khaghan when a Byzantine embassy arrived at his court in 576. The Kök Türks argued that the treaty established with the Avars was in breach of the first treaty they had made with the Byzantines, and 41 42

Menander, fr. 10. 1–5; for beginning of the Persian War, see Whitby, Emperor Maurice, pp. 250–58. Pohl, Die Awaren, p. 212 and the revealing exchange in Menander, fr. 8.

54

Leif Inge Ree Petersen

an affront to their claims to hegemony over a defeated group (just as the Avars in relation to the Gepids). They threatened to take matters into their own hands and subdue the Avars. While presenting this as a favor to the Byzantines, they were themselves none too eager to see the whole of their northern marches under the control of one gargantuan steppe empire. In a show of force, the Turks sent an expedition to besiege the Crimean Bosporus, just west of the Kuban steppes. The Avars, conversely, claimed control over the Utigurs, now under Turkish suzerainty, and clearly had ambitions towards the Pontic steppes should the occasion arise. The Utigur area of settlement lay on the Kuban steppes just across the straits from the Crimea, which was the main Byzantine point of contact with nomads on the steppes north and immediately east of the Black Sea, while Byzantine armies in the Caucasus had secured hostages from among the Sabirs and Alans, presumably refugees from the expanding Turks. These could obviously be used to disrupt the Kök Türk confederation.43 The Turks’ target is a clear indication that they had a healthy respect for Byzantine the ability to subvert the tribal confederations, having had long experience managing the steppes and close relations with many of the constituent groups in the Kök Türk khaghanate. In face of these threats, especially those in the Balkans, the fortress-building program under Justinian proved its worth. Although the Byzantines were still unable decisively to defeat or to conduct punitive expeditions against Avars and Slavs, the great raiding armies of the 580s were far less successful than is often imagined, as many reported sieges were actually failures, and no further territory was permanently lost.44 The situation changed steadily after the favorable peace with the Sasanians in 591, when the Byzantines could devote the necessary manpower to defeating all raiders and begin to go on the offensive across the Danube against the Slavs, a devastating strategy well described in the Strategikon.45 The emperor Maurice (r. 582–602) was brought down by troops dissatisfied with their pay and service conditions as they were chasing Slavs along the frosty banks of the tributaries of the Danube (loss of foliage and tracking in the snow meant that they could no longer use the effective guerrilla tactics which had foiled the Avars). Nevertheless, Byzantine dominance in arms was such that even a renewed Persian war in 603 could not bring down the edifice constructed by Tiberius and Maurice. The northern frontier was, as far as we can tell, stable for another decade, and it is in this period we must postulate the formation of a series of Slav tribal groups that remained relatively stable throughout the seventh century and formed the backbone of the Byzantine client system on the Balkans. A renewed Persian war, then, was not enough to break the northern frontier, but conditions deteriorated rapidly with the Byzantine civil war, when Heraclius revolted against Phocas in 608–10, and the Persian breakthrough on the Euphrates frontier in 610 or 611. This finally stretched military resources to 43 44 45

Menander, fr. 18.5. Petersen, Siege Warfare, pp. 381–82 and passim. Maurice’s Strategikon: Handbook of Byzantine Military Strategy, trans. George T. Dennis (Philadelphia, 1984), 11.4.



Byzantium and the Northern Barbarians

55

breaking point, as the strengthened Balkan field armies had to withdraw to fight for the survival of the Empire. This is the background for renewed Avar expansion, with the actual capture of important cities in the interior of the Balkans attested around 614, followed by two great, but failed, expeditions against Thessalonica in the late 610s, and finally, the great siege of (and even greater failure at) Constantinople in 626.46 The reach, capacity, and organization of the Avars appeared formidable. Again a major steppe dynasty had managed to subvert decisively the Roman client system and turn it against the Empire. However, this was only possible because the convergence of civil war and Persian invasion. The Avars’ primary advantage lay in their military organization and ideology that allowed them to establish hegemony over erstwhile or potential Roman clients when the Empire was otherwise engaged in a fight for its survival. Yet the success of the Avars was ephemeral and largely hinged upon their ability to overawe the component groups of their confederation. An important element of this was the capacity to capture cities and show success in the form of plunder and extortion of subsidies from the Byzantines. Their first successful raids and sieges in the 610s certainly yielded valuables, not least numerous captives, which would feature prominently in future client relations. However, the Byzantines had already laid the groundwork for subverting Avar power. Much of the Avars’ coalition was rather shaky; many Slav groups, especially, seem not to have been content with Avar rule, even if they accepted their direction and exploited Avar invasions for their own purposes. Byzantium’s intermittent and somewhat difficult alliance with the Turks paid off in 628 when they joined Heraclius’s expedition in the Caucasus, a major stepping stone in out-maneuvering and then defeating the Persians and forcing them to conclude peace in 629. On the Pontic steppes another opportunity arose when the Kök Türks, with impeccable timing from a Byzatine point of view, became embroiled in a civil war. The first mover in those events had been the Tang dynasty, which had recently gained control over the Tarim basin and could directly interfere in Turkish rule and establish its own client system. While some the constituent tribes in the Kök Türk confederation had briefly been under Avar hegemony, the disappearance of the Kök Türks as a ruling group opened up new possibilities but also the threat of Avar encroachment. The Byzantines provided support to the Bulgars (associated with the Onogundurs), who had thrown off Avar dominance around the Sea of Azov, by granting their ruler Koubrat the title patrikios, gifts, and subsidies which have been amply attested in the archaeological record. Importantly, the Pontic steppes remained stable and peaceful throughout the first onslaught of the Islamic conquests and concurrent collapse of the great Turkic khaghanate.47 46 47

For a basic narrative, see Pohl, Die Awaren, pp. 237–54. Nikephoros, Patriarch of Constantinople: Short History, ed., trans., and comm. Cyril Mango (Washington, DC, 1990) [henceforth Nikephoros], ch. 22; discussion in Sophoulis, Byzantium and Bulgaria, pp. 105–08. The seals are usefully listed in translation, with references, by Kiril Petkov, The Voices of Medieval Bulgaria, Seventh–Fifteenth Century: The Records of a Bygone Culture (Leiden, 2008), pp. 1ff.

56

Leif Inge Ree Petersen

Punctuated Equilibrium The collapse of Avar power was sudden and complete, but it was also prepared by Byzantine diplomatic efforts. The Bulgars were clearly and deliberately built up to dislodge the Avars on their eastern flank (and compensate for the collapse of the Kök Türk khaghanate), while their western Slav marches were under the control of Samo as early as 623. There is no direct indication of Byzantine hands in this, but Byzantine diplomacy was a concern to Merovingian rulers at least as late as the 650s. It would hardly be surprising if the Byzantines helped engineer a situation that would lead to the immediate implosion of Avar power as soon as its expansion was halted by defeat before Constantinople.48 Heraclius clearly had diplomatic relations with two other significant Slav groups that emerged from the rubble of Avar hegemony, the Croats and the Serbs, who were allowed to settle in Dalmatia.49 While the intractable question of the latters’ ethnic roots falls outside the range of this study, Turkic or Iranian elite origins, as have been suggested, are certainly possible, as most of the “Turkic” confederations often had complex interwoven ethno-political arrangements which we can follow among both Avars and Bulgars quite closely.50 Heraclius’s settlement of the north in 630, or shortly thereafter, thus stretched from the Kuban steppe east of the Sea of Azov to Croatia, with all major players owing allegiance and legitimacy to the Byzantines, which is indeed a stunning reversal of fortune. The interior of the Balkans, on the fringes of or beyond erstwhile Avar influence, was not recovered, however. This area was under the control of a number of Slav tribes in an arch following the Aegean from Thessaly through Macedonia to Thrace. The distinct nature of these small polities, each with its own king and names that remained stable for generations, was a world away from the amorphous political organization on the lower Danube in the early sixth century, and had a very different political arrangement with the Empire. As we have seen, it is only for the years covered by Menander’s account that we know of named political leaders arising with greater regularity, and these leaders’ rivalries intensified during the late sixth century. While Slav political organization appears obscure in the Miracles of St. Demetrius’s description of the first siege of Thessalonica in 586, a distinct set of tribal entities had crystallized by the 610s. We know next to nothing about these groups again until the Miracles of St. Demetrius describes in some detail political and cultural relations in the second half of the seventh century. Otherwise, only brief notices in Theophanes’s Chronicle of expeditions against Slavs under Constans II (641–68/69) and Justinian II (685–95, 705–11), which seem to have resulted in the forcible recruitment of large Slav contingents that later defected to the 48 49 50

This point is also made by Obolensky, “Principles and Methods”; see further Pohl, Die Awaren, pp. 256–61. Constantine Porphyrogenitus, De administrando imperio, ed. Gyula Moravcsik, trans. R. J. H. Jenkins, Corpus fontium historiae Byzantinae 1 (rev. ed., Washington, DC, 1967), cc. 31–32. For the debate and further evidence, see Pohl, Die Awaren, pp. 261–68.



Byzantium and the Northern Barbarians

57

Arabs, provide snapshots that make better sense in terms of a client-management framework thrown out of kilter by internal Byzantine conflict.51 The event that at first sight seems best to confirm the existence of a hostile or at least ambiguous relationship between Byzantium and the Slavs is described in the fourth miracle of the anonymous collection of the Miracles of St. Demetrius, most often dated to around 677, but actually occurring around 662. A number of Slav tribes settled in Macedonia arose against Thessalonica when one of their kings, Perboundos, was arrested and brought to Constantinople after being denounced for plotting treason while visiting the former city. The course of the ensuing political debacle is highly revealing. Slav raiding around Thessalonica was, according to the hagiographer, exceptional, and the treason Perboundos was accused of involved high-ranking members of the court in Constantinople. The assistance rendered by Slavs in Thessaly and Thrace to Byzantine authorities reflects a political situation in which the Slavonic tribes in the southern Balkans, at least, were largely within the Byzantine political and cultural orbit as clients.52 Before we investigate how such a situation arose, however, we need to devote attention to some basic source criticism, as the dating of this and later events recounted in the Miracles plays a major role in our understanding of developments in the Balkans in the seventh century. Lemerle dated the events of the fourth miracle on the basis of an indiction reference to 677, settling on that date rather than the possible 662 because it fitted with the then-accepted date for the first siege of Constantinople, in the mid- to late 670s. This has pushed the fifth miracle, concerning the failed attempt of the Bulgars Kouber and Mavros to subvert and take the city, to shortly after 680, since internal evidence shows that it happened a few years after the fourth miracle.53 However, the text states that the appearance of the ex-Byzantines under Bulgar leadership was sixty years after the deportation of a large group of Romans from the Danube provinces, which itself happened at the time of the War of Chatzon, dated to the late 610s. The events of the fifth miracle clearly happened after those referred to in the fourth, but must have taken place after some time had passed, which led Lemerle, having dated the fourth miracle to 677, to place the fifth in c. 682. Apart from the critique by Howard-Johnston, no attempt has been made to re-date these events so far.54 The question requires a major study in itself; we need only survey the points most relevant to our present purposes. Concerning the fourth miracle, the recent 51

52 53 54

Theophanes, The Chronicle of Theophanes the Confessor: Byzantine and Near Eastern History AD 284–813, ed., trans., and comm. Cyril Mango and Roger Scott (Oxford, 1997) [henceforth Theophanes], pp. 484, 487 (Greek text Theophanis Chronographia, ed. Carl de Boor [Leipzig, 1883], pp. 347, 348) and pp. 508, 511 (de Boor, 364, 366). Les plus anciens receuils des miracles de saint Démétrius. I Texte, ed. and trans. Paul Lemerle (Paris, 1979), Le quatrième miracle, cc. 230–82. Paul Lemerle, Les plus anciens receuils des miracles de saint Démétrius. II Commentaire (Paris, 1981), pp. 128–33. See the preliminary discussion in Petersen, Siege Warfare, pp. 662–68, with references. James Howard-Johnston, Witnesses to a World Crisis: Historians and Histories of the Middle East in the Seventh Century (Oxford, 2010), pp. 152–54

58

Leif Inge Ree Petersen

reinterpretation and re-dating of the siege of Constantinople to naval raids around 670 seriously weakens Lemerle’s argument, as he had to distort the narrative found within the miracle to make it fit the then-accepted chronological framework. A close reading of the internal evidence further excludes his version. Instead, the surrounding events can easily be related to the political upheaval surrounding the trial of Maximus the Confessor and Constans’s suppression of the long-simmering opposition of a party, including generals, provincial governors, and family members associated with hardline Chalcedonians who refused to accept the compromise with Miaphysite Christians championed by Heraclius and later Constans himself.55 He also conducted a major military expedition to the east, establishing control of the Transcaucasus deep into present-day Azerbaijan and northwestern Iran. The trial and exile of Maximus can be dated precisely to May–June 662, while Thessalonica was relieved by a fleet and a land army in late summer. Thus the whole affair lasted from late 659 to the summer of 662, when Constans II set out on a major expedition that reached Athens, necessarily marching via Thessalonica, and eventually Italy in 663.56 The Slavs in the miracle are presented as having been peaceful and loyal to the Roman state in recent memory; the majority of them largely remained so during the military conflict. The Belezegetes, settled in Thessaly and by now agriculturalists, provided supplies for the beleaguered city when the fields around Thessalonica were too dangerous to tend. When Perboundos tried to flee Constantinople with assistance from a courtier, they had to avoid the Slavs living on the Thracian border, who remained completely loyal. In fact, the assistance of a high-ranking interpreter, courtiers, and military personnel indicate a complex plot far beyond the ambitions of the small Slav statelets: the objective was to depose and replace the reigning emperor, which was a real threat to Constans for most of his reign. The situation around Thessalonica was restored when Constans’s grand western expedition soundly defeated the Slav client kinglets in 662. The nature of Byzantine–Slav relations is not difficult to gauge: the loss of the Balkans to the Avars and Slavs in the 610s was compounded by the withdrawal of military units to fight the Persians, then the Arabs, and the deportation of much of the remaining Roman population. Much as the Byzantines recovered after the first bout of Avar expansion in the 580s, the collapse of the frontier around 615–26 might have been reversible had it not been for the immediate Arab conquests that began in 633 and soon absorbed all available manpower in the east. The loss of fortifications in the interior, the deportation or flight of communities who could anchor a frontier, and continuous war in the east, meant that any attempt to re-establish control of the interior Balkans could not succeed. Instead Constans, continuing Heraclius’s northern settlement, limited himself to strengthening Thrace, Thessalonica, Athens, Corinth (Acrocorinth), and 55 56

For the political context, see Phil Booth, Crisis of Empire: Doctrine and Dissent at the End of Late Antiquity (Berkeley, 2013) Petersen, Siege Warfare, pp. 669–70.



Byzantium and the Northern Barbarians

59

Monemvasia, and establishing hegemony over the lesser Slav communities from such secure points. The political map as described in the Miracles seems to have been largely stable until the Byzantines re-established direct rule in the late eighth century. The origins of this system can fairly obviously be traced to Heraclius and his efforts to establish Old Great Bulgaria on the steppes and the Croats and Serbs in Dalmatia. Any treaties with the Slavs, about which we are not informed, must have been concluded around 630. The cultural make-up of these statelets, however, is a world away from that of the lower-Danube villages. Numerous changes are attested in material culture. The typical Grübenhäuser were no longer to be found in the Balkans, while slash-and-burn agriculture and pastoralism had given way to more permanent agricultural settlements. This accelerated change was of course not just clever client management: despite the breakdown of the frontier, the Balkans remained dotted with numerous, though isolated, settlements of erstwhile Romans who had neither fled nor been deported, and were absorbed into the Slav states politically and economically but not necessarily ethnically, as continuously demonstrated in the following centuries. This allowed the Byzantines to deal with the Slavs, hitherto so difficult to manage, by using more permanent agricultural settlement to encourage and cultivate kinglets in a handful of minor states. The connection was made through the supply of titles, prestige gifts, and trade. While there is some evidence in the archaeological record, most important is a series of lead seals that denote the place and role of minor Slav rulers within the Byzantine political system.57 In return for granting such favors, Byzantine influence stretched deep into the Balkans, especially the agriculturally productive river valleys of Macedonia, the plains of Thessaly, and the northern littoral of the Aegean. It may even have stretched as far as the Danube, especially along the Black Sea, as Constantine IV clearly regarded this as sovereign Byzantine territory in 680 when he set out against the Bulgars. Indeed, excepting the Perboundos affair, which may be attributed to internal Byzantine political dynamics, there were few indications of aggression or threats in the Balkans between the collapse of the Avar hegemony and the arrival of the Bulgars. This state of affairs might explain the behavior of Slav contingents in Byzantine armies in Anatolia, who defected to the Arabs. We know of two significant cases, noted above, one under Constans and another under Justinian II. Usually this is taken as evidence of Byzantine failure to integrate unwilling Slav conscripts, but we only learn about these because the sources tend to be critical, finding fault with various emperors’ policies. Five thousand Slavs, presumably pressed into service by Constans II, went over to the Arabs around 664, but this may in fact have been in the context of a major revolt under the general Saporios, who temporarily detached the Armeniac and parts of the 57

Florin Curta, The Edinburgh History of the Greeks c. 500–1050 (Edinburgh, 2011), pp. 116ff and passim; idem, “L’administration byzantine dans les Balkans pendant la ‘grande brèche’: le témoignage des sceaux,” Bizantinistica. Rivista di studi bizantini e slavi 6 (2004), 155– 90.

60

Leif Inge Ree Petersen

Anatolic command (later themes – themata – of the Empire) from the Arabs during a large-scale attempted usurpation around 664–69, while Constans II was in Sicily. The Slavs in question would thus not have been acting on their own initiative, but following orders from their commanders who were in revolt.58 The Saporios revolt also had consequences for the eastern, Caucasian reaches of the Byzantine client system as it existed in the early 660s. The great Turkic khaghanate had collapsed due to internal power struggles exacerbated by pressure from the Tang dynasty in China. The khagahanate having provided useful service to the Byzantines under Heraclius, the latter were probably none too eager for a power vacuum to develop and cause chaos on the steppes, which had been quiet during the Arab onslaught. The rebels under Saporios tried to force the defection of the easternmost clients of Constans, led by the indomitable Juansher of Caucasian Albania (in present-day Azerbaijan), who had long experience as a general in the Sasanian army and commanded considerable respect among both Armenian and Persian nobles. One of Maximus the Confessor’s disciples was still in the Caucasus in the 660s, and his letters to supporters elsewhere in the Empire written in 666 are proof of a sudden regime change in part of Anatolia and the Caucasus shortly beforehand, which must be associated with the revolt of Saporios.59 This had repercussions in the Caucasus forts where Maximus had died and his disciples were still held, but now suddenly received much more lenient treatment from newly-appointed commanding officers. This, as we have seen, was also the entry-point for negotiations between Byzantium and nomads on the steppes north of the Caucasus. Some scholars have recently argued that the Khazars, an erstwhile constituent group of the Kök Türk khaghanate, began to fill the vacuum they had left behind in the early 660s.60 The biography of Juansher attests to how the Khazars, rather surprisingly, rather than concentrating on the Arabs, instead made a fierce attack on the Caucasian Albanians in the winter of 664/65. Juansher, already cut off by an Armenian revolt in 663, and now separated from the regime of Constans too by a large segment of Byzantine Anatolia in revolt, surrounded by Khazars to the north and Arabs to the south, was forced to submit to the Huns, as he called them, and seek an alliance with the caliph Mu‘awiya.61 Thus the rise of the Khazars, who are traditionally regarded as reliable, if distant, allies of Byzantium from the seventh to the tenth century, was in fact closely intertwined with internal Byzantine politics, although discerning cause and effect is difficult with the level of evidence we 58 59

60

61

See n. 51 above and discussion in Petersen, Siege Warfare, Appendix III, pp. 439–53. Anastasius Apocrisiarius, Epistula Anastasii ad Theodosium, 4–7, in Maximus the Confessor and his Companions: Documents from Exile, ed. and trans. Pauline Allen and Bronwen Neil, Oxford Early Christian Texts (Oxford, 2002. repr. 2004), pp. 132–47. Constantin Zuckerman, “The Khazars and Byzantium – The First Encounter,” in World of the Khazars, ed. Golden, Ben-Shammai, and Róna-Tas, pp. 399–432; but cf. the alternative interpretation of the role of the Khazars in the invasion of 665 and its effects in Petersen, Siege Warfare, Appendix III. wHoward-Johnston, Witnesses to a World Crisis (Oxford, 2010), pp. 108–28.



Byzantium and the Northern Barbarians

61

have. If this conjecture is correct, however, it was internal Byzantine political developments that set off the chain of events that led to the arrival of the Bulgars on the Danube fifteen years later. Constans, in the meantime, having foiled numerous attempted plots, coups, and revolts, ultimately fell to traitors’ hands in Sicily (where the usurper was another general from the Caucasus, the Armenian Mzez Gnuni) in 668 or 669. His son Constantine IV was able to stabilize the situation, but the revolt and its possible repercussions upon reorganization of the steppes had important knockon effects. The Khazars, either exploiting the vacuum created by internal dissension or actively encouraged by the party in revolt, soon expanded at Old Great Bulgaria’s expense. The latter may have been weakened by an (archaeologically discernible) decrease in subsidies from the 650s. We cannot tell whether this was policy – hardly likely, as the Bulgars had apparently kept the peace so effectively for a generation – or marked a shift to a new supply system relying more on valuables in kind that cannot be detected archaeologically, such as silks and foodstuffs, in contrast to the rich gifts provided by Heraclius and Constans early in his reign. It may simply have been another consequence of Byzantium’s life-and-death struggle with the Caliphate. Vast Arab fleets had raided Cyprus twice, in 649 and 650, occupied Rhodes, Kos, and eastern Crete in 653, and tried (unsuccessfully) to storm Constantinople by sea in 654, while most military resources were tied up defending Anatolia and the Caucasus.62 The result is nevertheless clear. Old Great Bulgaria collapsed under Khazar pressure, and the five famous sons of Koubrat migrated elsewhere. While one segment of the Onogurs ended up as a constituent part of the Khazar khaghanate (which according to later Arab authors comprised some twenty-five distinct groups), at least three other groups ended up affecting Byzantium. The first to do so, however, was not that of the renowned Asparukh, whose followers settled north of the Danube and became the forebears of medieval Bulgaria, but another Bulgar contingent that made it to the Avar khaghanate.63 Reeling from their great failure against Constantinople in 626, and with revolts already underway on their flanks – Slavs under Samo to the west and Old Great Bulgaria in the east – the Avar center collapsed into civil war between Avar and (other) Bulgar (perhaps Kotrigur) elements in the 630s. The latter lost, and nine thousand of them sought refuge with the great Merovingian king Dagobert, only to be massacred with only seven hundred survivors integrated into Merovingian society. However, the Avars still ruled over a multi-ethnic society which included many of the constituent groups listed above, but now in addition a large number of Romans who had been deported northwards shortly before 620. We know of this from the Miracles of St. Demetrius: the fifth miracle describes how the Avars decided to place a Bulgar, Kouber, and his second-in-

62 63

See further Petersen, Siege Warfare, Appendix III and the relevant sieges s.aa. Lemerle, Plus anciens receuils, I, Le cinquième miracle.

62

Leif Inge Ree Petersen

command Mavros, in charge of the Roman population, referred to collectively as “Sermesians.” The context is somewhat obscure, but this group under Kouber’s leadership decided to break away from Avar rule. The ex-Romans were yearning for their homelands, but the Bulgar leadership had other, more nefarious intentions that involved taking control of Thessalonica. Had they succeeded, we might have seen a Bulgaria on the Aegean shortly before Asparukh founded his Bulgaria on the Danube. As events turned out, the plot was discovered and averted. However, the Miracles of St. Demetrius reveals that the “Sermesians” were cautiously received by Byzantine authorities, who settled them some distance inland from the city and ordered the local Slavs (who had been in revolt a few years earlier according to the fourth miracle) to provide supplies. According to lead seal finds, Kouber’s associate, named Mavros, found a career in the Byzantine army, and the “Bulgarians” near Thessalonica seem to have been a distinct group at least into the reign of Justinian II.64 Thus the potential threat was not only contained, but became a net gain, as it provided new manpower and assimilated what had potentially been an enemy into Byzantine political structures. This is far from being the only such example. Several lead seals attest to the presence of Slav “archons,” petty rulers with a recognized place within the Byzantine hierarchy, gradually becoming fully integrated into the formal administration of the Empire.65 Byzantium’s long-term strategy in this area is thus beginning to crystallize: from the days of Heraclius, the Empire had constructed a buffer of Slav statelets that projected influence deep into the Balkans, providing extra security and forming a reserve of logistical support and manpower that was mobilized for either external warfare or internal conflicts. The costs of this system were modest, consisting of small subsidies, trade rights, titles, extremely rare military interventions that were largely shows of strength, and, in only the most serious cases, the arrest or killing of rulers who misbehaved. But as we have seen, such misbehavior seems only to have occurred at the request of Byzantines plotting against the sitting regime. Petty kings and chieftains did dual service as archons, informal lower-level Byzantine officials, who were charged with looking after imperial interests in their area of influence. This was also a prelude to (re) integration of land areas or people as the opportunity arose. For instance, the land corridor from Thrace to Thessalonica was maintained or re-established by Constans, while Slavs, Bulgars, or ex-Romans all seem to have been welcome in the Byzantine state. It is against this background we should see the first encounter between the Byzantines and Asparukh’s Bulgars around 680. Byzantine client policies and demonstrations of force were largely successful, and the Bulgars must have been settling for a while to the north of the Danube delta before Constantine

64 65

For the seals, see Petkov, Voices of Medieval Bulgaria, pp. 2–3. Curta, “Le témoignage des sceaux.”



Byzantium and the Northern Barbarians

63

IV decided to intervene. The early 670s were inauspicious for a drive against the Bulgars or an attempt at regaining more of the Balkans. Constantine’s early reign was disturbed by the fallout from Mzez Gnuni’s revolt, which was possibly repeated by a son, John, in 672.66 Around the same time, an Arab fleet temporarily anchored at, and perhaps also attempted a naval blockade of, Constantinople. Several large-scale Arab incursions raided Lycia in 674–75, but were roundly defeated. These crises were all resolved, and so impressed neighboring states that they sent embassies to Constantinople. These defensive wars were immediately followed by the considerable effort of sending the Mardaites to Lebanon, wreaking havoc in the heartlands of the early Umayyad caliphate. The Sermesian affair described above would have happened at about the same time. There are in fact indications that the Byzantines, as an alternative to confrontation, were using their client networks to contain the Bulgars, sending subsidies and gifts to elites located in present-day Wallachia. However, as had happened so many times before on the steppes, Asparukh’s Bulgars were quite successful in building up their influence, which is probably the reason for Constantine IV’s ill-fated expedition to the Danube delta around 680. According to Theophanes and Nikephoros – both accounts deriving from Trajan the Patrician’s lost history – the Byzantines botched the expedition and the Bulgars were able to establish themselves south of the Delta in the Dobrudja, as well as expand up-river along the lower Danube.67 In light of the long and often bitter history of conflict between Byzantium and Bulgaria and the potential danger posed by a nomadic warrior elite ruling over large tracts of the Balkans (as the Avars and the Huns had ruled from the region of Pannonia), subverting Roman clients, raiding Roman territory, and extorting subsidies and ransoms, the potentially disastrous fallout was actually very well contained. There is in fact little evidence of Bulgar encroachment on the Byzantine client system along the Aegean. If there was, Justinian II effectively re-established Byzantine hegemony on his expedition to Thessalonica in 688–89, noted above, which again netted a substantial transfer of not only Slav soldiers but also their families to Anatolia – a policy we learn of only on account of its ostensible failure. Little noted is that Justinian also instituted the policy of transferring peoples from the eastern frontier to Thrace. The brief report by Theophanes goes on to describe a defeat by the Bulgars but, given that this occurred in the region of Thessalonica, the group settled by Constantine IV in the Vardar valley may be meant: these were still distinct from the Danube Bulgars in the early eighth century, as is evident from the Bulgars’ own inscriptions. As we might expect, the Bulgars on the Danube did take control over substantial Slav groups further inland, resettling them along their frontiers for defensive

66 67

I owe this observation to discussions with Vivien Prigent in Rome, December 2015. Nikephoros, cc. 35–36 and Theophanes, 497–501 (de Boor, 356–60), with commentaries.

64

Leif Inge Ree Petersen

purposes. However, Byzantine connections with the Avars seem to have intensified in response. There was also a significant rapprochement with the Khazars (further described below), which ensured stability on the steppes and prevented any further accretions to a hegemonic Bulgar power. Similarly, a potential threat from that direction probably had a dampening effect upon Bulgar activities. Meanwhile the Bulgars’ internal propaganda is itself highly revealing. While their great rock-relief at Madara shows the hero-horseman typical of nomadic lore, they also produced a handful of inscriptions dating to our period describing political relations. The Madara inscriptions dating to the reign of Khan Tervel (c. 700–20) demonstrate that there was indeed communication between the Bulgars near Thessalonica and those on the Danube: the inscriptions make reference to the Tervel’s “uncles” at Thessalonica, whom, despite the misgivings of their leaders, the Byzantines seem to have effectively co-opted, as reflected in a number of seals. Furthermore, the treaty between Justinian and Tervel, the second khan, is well known. This established trade, military assistance, and demarcation of borders.68 Arguably much more threatening to Byzantium’s long-term survival was the Umayyad caliphate, which persisted until its defeat at battle at Akroinon in Anatolia in 740, and never recovered after a massive Berber revolt in the same year. Leo III (r. 717–41) pursued a highly successful, but defensive, policy which was considerably facilitated by the Byzantine alliance with the Khazar khaghanate, maintained through marriages by both Justinian and Leo. The Arabs expended enormous resources on the eastern Caucasus and into present-day Dagestan during most of the Umayyads’ rule, which immensely relieved the pressure upon Byzantium. The connections with both Khazars and Bulgars also proved useful for Justinian when he was deposed and exiled to the Crimea in 695. After fleeing and spending a stint with the Khazars, where he acquired a Khazar princess as his bride, he moved on to Bulgaria where (after being rebuffed by the Bulgars at Thessalonica, who presumably remained loyal to the sitting regime) he received assistance from Tervel to regain his throne: five thousand men, according to Tervel’s inscription. His gratitude, expressed in a joint ceremony in Constantinople at which Tervel received the title of caesar, did not last long, however. He soon instigated a war against Tervel, probably to re-establish Byzantine dominance (and perhaps also exploiting his familiarity with internal Bulgar affairs), but Tervel managed to force Justinian to honor the treaty terms agreed upon. While not to Justinian’s liking, this was clearly advantageous to the Empire in the long run. Shortly afterwards, the Arabs launched their largest and most dangerous assault yet upon Constantinople, wintering in Thrace in 717–18 while a large fleet anchored at Kyzikos in the Sea of Marmara. During this conflict the Bulgars proved their worth as allies, harassing the Arabs in Thrace and contributing to their utterly miserable camp conditions with months of snow and no supplies to be pillaged from the local countryside.69

68 69

Petkov, Voices of Medieval Bulgaria, pp. 3–4 and 5–6; Sophoulis, Byzantium and Bulgaria. Theophanes, 546 (de Boor, 397).



Byzantium and the Northern Barbarians

65

It is clear, then, that the potentially disastrous appearance of the Bulgars was not only contained, but managed so as on the whole to benefit the Byzantines. The Bulgars of Thessalonica largely remained loyal and a source of prominent military officers (and probably manpower), while Tervel, strengthened by his intervention in Byzantine politics and rewarded with the title of caesar, nevertheless recognized his place as the junior partner and never seems to have followed a policy that threatened Byzantine control of the Balkans or domination of the minor Slav client statelets along the Aegean. This remained the case through the early eighth century. The danger from the Bulgars, however, lay in their ability to provide assistance to a Byzantine pretender, a practice first exploited by Justinian II but ended after the accession of Leo III. The deposed emperor Anastasios II, in exile at Thessalonica, had in 718 called for Bulgar assistance, which at first was given, but after a stern rebuke from Leo (who had grown up and was well-connected in Thrace) the Bulgars abandoned Anastasios II to his fate.70 Subsequent Byzantine civil wars (such as the revolt of Artavasdos in 741–43) and Umayyad pressure were never exploited by the Bulgars in a manner significant enough to be noted in the chronicles. Ironically the situation was to change because the Byzantines were relieved of pressure in the east, and the Bulgars, thanks to their close symbiotic relationship with Byzantium and the resulting presence among them of a large number of Byzantine exiles, again became a threat around 800. Constantine V (r. 741–75) came to power shortly before the Umayyad dynasty was paralyzed by a series of civil wars that culminated in the Abbasid revolution. This had major repercussions upon Byzantine–Bulgar relations. Constantine captured Germanikeia and led other successful expeditions in the east, whence he deported large populations of Armenians and Syrian Christians to Thrace. The subsequent events are well known and need only be summarized briefly. The Bulgars, insisting on observation of treaty commitments, opposed new settlements and fortifications. However, Constantine launched a series of successful campaigns that threw the Bulgar leadership into complete disarray, to the extent that the Byzantines successfully dominated Bulgar politics, and at times even decided who would be khan. In the following years under Irene and Constantine, the Slav archontates along the Aegean were formally annexed and integrated into Byzantine territory, speaking to the degree of cultural, economic, and political symbiosis established in the course of the seventh and eighth centuries. By 790, Byzantine hegemony in the Balkans was overwhelming; but it had been painstakingly built up ever since the collapse of the 620s and the setback caused by the arrival of the Bulgars in 680. However, increased external pressure could also lead to internal consolidation, while at the same time there was widespread Bulgar familiarity with Byzantine culture, economy, strategy, and tactics through long-term symbiosis and the presence of a large number of Greek-speakers in Bulgaria, both popu70

Theophanes, 552 (de Boor, 401), Nikephoros, c. 57.

66

Leif Inge Ree Petersen

lations stranded from the seventh century and recent exiles from Byzantine internal conflicts. The rise of Khan Krum resulted from a combination of Bulgar resurgence, Byzantine infighting, Constantine VI’s incompetence, and Abbasid pressure in Anatolia. Nikephoros I decided the time had come to stamp out the Bulgar state completely. His initially highly successful campaign caused the Bulgars to sue for peace, but the intended coup de grâce turned into utter defeat in a pass the Haemus mountains.71 The ostensible success of the resurgent Bulgars has arguably thrown historians off the scent of successful Byzantine client management. Most important was Bulgar integration into the Byzantine political scene; thus Bulgaria became a refuge for Byzantines who fell out of favor in the bitter internal conflicts from the late 780s until the establishment of the Amorian dynasty and beyond. In the prelude to Nikephoros’s ill-fated campaign, there are numerous indications of high-ranking Byzantine defectors in the harsh political climate caused by regime change, loss of patronage, or outright persecution of the opposition party around 800. The Bulgars were even actively used as a foil by Leo V to gain power. While the crisis was made possible by Byzantine infighting or errors of judgment (such as those displayed by Constantine VI during his sole reign and Nikephoros’s disastrous expedition in 811), and allowed the Bulgars breathing space, Krum’s threat to Constantinople turned out to be purely contingent and opportunistic. Bulgar aggression was soon contained by the Byzantines, and during the ninth century Bulgar expansion was directed elsewhere, towards former Avar lands, now that Charlemagne had crushed the Avars. This set the stage for continued Byzantine influence on developments in the Balkans, where the coastal regions of Greece were fully integrated into the Byzantine administrative structure. While Bulgaria remained unconquered, its subsequent conversion to Orthodox Christianity involved adopting and adapting Byzantine cultural and political models. In the meantime, Byzantine lead seals attest to contacts with Scandinavia that have been interpreted as a deliberate drive to recruit mercenaries and adventurers who could help shape the vast tracts north of the steppes.72 While Bulgaria was gradually being brought under Byzantine cultural hegemony, its expanding power had to be contained. At the same time, the Khazar hegemony on the steppes that had served Byzantium so well was growing problematic. A new equilibrium had to be found. For Byzantium, the greatest coup of all awaited, as they brought the Varangians, and through them the whole territory north of the steppes, under Byzantine cultural and political influence in the course of the ninth and tenth centuries, with tools perfected during the long transition from the late Roman Empire. 71 72

For the preceding account, see in general Sophoulis, Byzantium and Bulgaria, and Treadgold, Byzantine Revival. I owe the information on Byzantine lead seals to John Ljunkvist (Uppsala), presenting on “Vikingtidens skandinaviske krigerelite” (“The Scandinavian Warrior Elite in the Viking Age”) at the Migration and Viking Age archaeology conference “Scandinavia – One, Three, or Many?” (“Skandinavia – ett, tre, eller mange?”) at the University of Oslo, 3–5 December 2014.

Map 1: Anglo-Saxon Burhs and Travel Routes

3 Evolving English Strategies during the Viking Wars Richard Abels

The history of middle and late Saxon England was shaped by the Viking invasions and the military responses of English kings, in particular, those of the two kings who have become respectively bywords for success and failure: Alfred the Great and Æthelred II “Unræd.” Thirty years have passed since Simon Keynes in an influential article ruled direct comparisons of the two “out of court,” and warned against the temptation “to ask […] why [Æthelred] failed where Alfred had succeeded.” Keynes makes a cogent argument against judging “one king in the light of the other,” aptly pointing out the disparate quality of the sources that have survived for the two men’s reigns.1 Alfred is unique among Anglo-Saxon rulers in the quantity and quality of the sources we have for his reign, much of which originated in his court, whereas Æthelred was less fortunate in the chronicler of his reign, who wrote in the wake of defeat and under the shadow of his declining posthumous reputation.2 But I disagree with Keynes that the military challenge each faced, and their responses to those threats, are not comparable. The thesis of this paper will not surprise those familiar with my previous work, namely that Alfred’s success was based on his ability to plan strategically on a grand scale, and that Æthelred II’s failure was due, at least in part, to his and his advisors’ inability to develop a coherent strategy against a similar threat. Alfred developed and expanded the military institutions he inherited into a new civil defense system that represented a defense-in-depth strategy tailored to meet the particular threat offered by the Great Heathen Army. The threat that Æthelred II faced between 980 and 1016 was in many respects similar, as were some of the main elements in his strategic responses. He too refurbished and constructed burhs and forts, strove to improve the effectiveness of royal armies, and ordered the construction of a royal fleet. As did Alfred, Æthelred II and his advisors understood the Viking threat as a divine punishment for sins, and, like 1 2

Simon Keynes, “A Tale of Two Kings: Alfred the Great and Æthelred the Unready,” Transactions of the Royal Historical Society 36 (1986), 195–217 at 217. See Simon Keynes, “The Declining Reputation of King Æthelred the Unready,” in Ethelred the Unready: Papers from the Millenary Conference, ed. David Hill, British Archaeological Reports, British Series 59 (1978), pp. 227–53; Richard Abels, “Alfred and his Biographers: Images and Imagination,” and Simon Keynes, “Rereading King Æthelred the Unready,” in Writing Medieval Biography, 750–1250. Essays in Honour of Frank Barlow, ed. David Bates, Julia Crick, and Sarah Hamilton (Woodbridge, 2006), pp. 61–75 and 77–98.

70

Richard Abels

Alfred, attempted to appease God by a program of spiritual reform. The strategic responses of both kings, moreover, evolved along with the changing threat posed by the Viking invaders. But in the end, Æthelred II failed. His responses were piecemeal and ad hoc and, ultimately, ineffective. He lacked the strategic vision of his great, great grandfather, and that cost him his kingdom. Over the last two decades, John Gillingham’s “Vegetian Strategy” thesis has become accepted as the reigning orthodoxy about military strategy in the Middle Ages. In a series of important articles, Gillingham took on the received opinion, best represented by Sir Charles Oman, that strategic thought and planning had all but disappeared from Western Europe between the eleventh and the mid-fourteenth century. A key piece of evidence cited by Oman was the paucity of battles during these centuries, a paradox given the warrior ethos of the knights who dominated the warfare of the time – explained by Oman as a consequence of the lack of strategic planning and the inability of armies to locate each other in the absence of reconnaissance. Gillingham, using largely the same historical sources as Oman, demonstrated that military commanders from the mid-eleventh through early fourteenth centuries did in fact follow a well-defined strategic doctrine, which he characterized as “Vegetian,” since it reproduced some of the general precepts set forth by the late Roman military manualist Vegetius whose Epitoma rei militaris (or De re militari) was among the most widely copied works in the Middle Ages. The key point they took from Vegetius, according to Gillingham, was that good generals should starve the enemy into submission rather than risk battle. The reason this strategy was followed was that it made military sense given the highly militarized landscape of Western Europe in the High Middle Ages and the inherent risk of battle that could turn suddenly on the death of a king or count leading from the front. Consequently, the warfare practiced by Duke William the Bastard (later King William the Conqueror), Richard the Lionheart, William Marshal, and other experienced military commanders of the High Middle Ages featured ravaging and sieges; battles were only risked when a commander believed he enjoyed a decisive advantage and the disadvantaged opponent was unable to escape.3 This was not the type of warfare fought in pre-Viking England. The primary military activity featured in the endemic small wars between neighboring tribal kingdoms in the seventh and eighth centuries was battle. For the period A.D. 600–835, there are a total of fifty-eight references to individual wars and major 3

John Gillingham, “Richard I and the Science of War in the Middle Ages,” in War and Government in the Middle Ages, ed. John Gillingham and J. C. Holt (Woodbridge, 1984), pp. 78–91; idem, “William the Bastard at War,” in Studies in Medieval History Presented to R. Allen Brown, ed. Christopher Harper-Bill, Christopher Holdsworth, and Janet Nelson (Woodbridge, 1989), pp. 41–58; and idem, “War and Chivalry in the History of William the Marshal,” Thirteenth Century England 2 (1991), 1–13. Cf. Clifford J. Rogers, “The Vegetian Science of Warfare in the Middle Ages,” Journal of Medieval Military History [henceforth JMMH] 1 (2002), 1–19; Stephen Morillo, “Battle Seeking: The Contexts and Limits of Vegetian Strategy,” ibid., 21–41; and Gillingham’s response, “‘Up with Orthodoxy’”: In Defense of Vegetian Warfare,” JMMH 2 (2003), 148–59.



English Strategies during the Viking Wars

71

military conflicts in Bede, the Anglo-Saxon Chronicle [ASC], the Annales Cambriae, and twelfth-century Latin chronicle sources. In thirty-seven, the text refers explicitly to battles. These battles, moreover, were often decisive, due to the deaths of commanders. In more than half of these battles – twenty of thirtyseven – one or more king, ætheling, or commanding ealdorman is recorded as having been killed. Ravaging is mentioned in only eleven of the fifty-eight general references to warfare, mostly as a prelude to battle. Even less common, however, are sieges, which are mentioned only twice, both occurring in Bede’s Historia ecclesiastica.4 The sources, other than Bede, are at the earliest ninthcentury, and it is possible that siege warfare became less common as Roman town defenses decayed. Military strategy in pre-Viking England thus defied the maxims of Vegetius. Invaders might ravage enemy territory to weaken the enemy and enrich themselves, but both sides actively sought a decisive engagement. Strategy in pre-Viking England was shaped by the relative lack of fortified towns and strongpoints, and by the goals kings sought to achieve through warfare: territorial expansion, the imposition of overlordship/tribute, and the acquisition of booty. The three were related. Warfare was an opportunity for a king or chieftain to exhibit his leadership and to gain the material wealth necessary to reward his followers. Success led to more success, since the military power of kings grew alongside their wealth and reputation, as they attracted more and more warriors eager to serve them in hope of reward. To the battles’ victors literally went the spoils. As the poet of the Old English Judith sang, “The dwellers in the land [the Hebrews] now had a chance to spoil the most hateful ones, their ancient foes now lifeless, of bloody booty, beautiful ornaments, shields and broad swords, brown helmets, precious treasures.”5 The Staffordshire Hoard discovered near Lichfield in 2009 proves that this is no mere literary trope. The date of the Hoard is still uncertain, but it seems likely that it was deposited in the mid-seventh to early eighth century. It contains over 3,500 gold and silver objects, most of which have been identified as fragments of ornamental fittings stripped from swords, helmets, and shields. Unlike contemporary Anglo-Saxon burial deposits, we find here no brooches, hairpins, buckles, or domestic items. And unlike later hoards, there are, unfortunately, no coins. What we do have are “as many as eighty-four sword and dagger pommel caps, seventy-one hilt collars, two or three gold crosses, a number of twisted4

5

Bede, Historia ecclesiastica gentis Anglorum [hereafter HE], 3.16: Bede writes that Penda, after “cruelly devastating the kingdom of the Northumbrians far and wide,” beseiged the royal Bernician stronghold of Bamburgh. When the Mercians “could not capture it by assault or siege,” Bede reports that they attempted to set the city ablaze, but that it was saved by a sacred wind supposedly sent in response to a plea from the saintly Aidan; Bede’s Ecclesiastical History of the English People, ed. and trans. Bertram Colgrave and R. A. B. Mynors (Oxford, 1969), p. 263. Cf. HE, 3.1, where Bede explains how Cædwalla king of the Britons killed King Osric of Northumbria. Osric was “rashly” besieging him in a fortified town, perhaps referring to York, when Cædwalla suddenly rushed out of the town with his forces and destroyed him and his army (“in oppido municipio temerarie obsedisset”). R. K. Gordon, trans., Anglo-Saxon Poetry (London, 1970), p. 295.

72

Richard Abels

metal rings, what is probably a shield decoration, and at least one cheek-piece from a helmet. Tellingly, several of the items have bent pins still sticking out of them, which means they were ripped from their original mounts.”6 The logical inference is that this is battle spoil collected by a Mercian king, ealdorman, or other magnate.7 Given the laconic entries of the Anglo-Saxon Chronicle, it is impossible to discuss in more than general terms the military strategic planning of AngloSaxon kings before Alfred. From Bede’s Historia ecclesiastica it would seem that the element of surprise was a critical factor, and that the king who could raise the largest force most expeditiously would emerge as the victor.8 That pre-Viking English kings were capable of strategic planning on a grand scale is attested by archaeological evidence, in particular the great dykes that ran along the Mercian–Welsh border. Offa’s Dyke and Wat’s Dyke remain monuments in the landscape not only to the power and authority of late eighth- and early ninth-century Mercian rulers but to their strategic planning. The construction of Offa’s Dyke alone required a minimum of 9 million and perhaps as many as 18 million man hours, representing the labor of tens of thousands of conscripted peasants.9 Once interpreted as a frontier marker, Offa’s Dyke is now thought to have been a defensive military barrier, which is more likely given the labor and expense that went into its construction. Excavations, field work, and surveys conducted by the Department of Extra-Mural Studies at the University of Manchester under the supervision of David Hill and Margaret Worthington reveal that Offa’s Dyke proper only ran from Rushock Hill north of the Herefordshire plain to near Mold in Flintshire, some 103 km. This was approximately the border between Mercia and the Welsh kingdom of Powys in the mid-eighth century. Hill and Worthington posit a connection between Offa’s 6 7

8

9

Alex Burghardt, The Times Literary Supplement [TLS], 14 Oct. 2009. Illustrations and maps for this paper are posted on my website: http://www.usna.edu/Users/history/abels/index.htm Among the few non-military objects in the hoard are two or three bent gold crosses. The defeated enemy was Christian. This is attested not only by the crosses, which appear to have been processional ones, but by an inscription quoting a few lines from Numbers 10:35 found on a strip of gold that may have run across the crest of a helmet (like the inscription on the Coppergate helmet): “surge d[omi]ne [et] dispentur inimici tui et fugent qui oderunt te a facie tua” (“rise up, o Lord, and may thy enemies be scattered and those who hate thee be driven from thy face”). The casual reduction of Christian religious objects into scraps of gold suggests that the victor may have been pagan. It is tempting to think that this hoard belonged to Bede’s notorious slayer of Christian kings, King Penda of Mercia, or to one of his ealdormen. See, e.g., HE, 2.12. When the East Anglian king Rædwald decided to help Edwin gain the Northumbrian throne, he quickly raised a large army: “Not giving King Æthelfrith time to summon and assemble his whole army, Rædwald met him with a much greater force and slew him.” Bede’s Ecclesiastical History, trans. Colgrave and Mynors, p. 181. David Hill, “Offa’s and Wat’s Dyke: Some Aspects of Recent Work 1972–1976,” Transactions of the Lancashire and Cheshire Antiquarian Society 79 (1977), 21–33; idem, “The Construction of Offa’s Dyke,” The Antiquaries Journal 65 (1985), 140–42. My calculations assume that a worker could excavate one cubic yard of earth per hour from a five foot deep trench in ordinary soil. This is based upon Capt. William Beach’s Manual of Military Field Engineering for the Use of Officers of the Line, 3rd ed. (Fort Leavenworth, KS, 1897), p. 73.



English Strategies during the Viking Wars

73

Dyke and “Eliseg’s Pillar,” erected in the mid-ninth century by the Powys king Cyngan ap Cadell: “Concenn son of Catell […] built this stone for his great grandfather Eliseg […] who annexed the inheritance of Powys […] from the power of the English both by his sword and by fire.” The Annales Cambriae reports that Offa fought a battle against the Welsh at Hereford in 760, and this might be where Eliseg freed his kingdom from Mercian domination. At the time of the dyke’s construction, however, the Mercians apparently enjoyed control over this frontier area, since Offa’s engineers were able to choose the location of the dyke so as to provide the best view possible into Powys. As the University of Manchester project discovered no traces of associated forts and no planned gateways, Worthington and Hill reasonably concluded that it was unlikely that the dyke was continuously manned. Strategically, the dyke would not have been intended as a preclusive barrier but as a defense against raiding, in particular cattle raiding.10 At least until the autumn of 865, the objectives of Vikings who raided England differed little from the goals of warring English and Welsh kings. Like those kings, they sought wealth though the use or threat of force. Viking chieftains, however, did so not as territorial kings with the intention of establishing lasting control over a region, but as pirates. This meant that the “strategy” they followed was designed to obtain as much portable wealth as possible with a minimum of risk. Using shallow-draft boats, they rowed up river until they found a convenient bank located near their designated targets on which to encamp, built field fortifications or improved existing defenses to a seized estate to protect their boats and loot, either seized or bought horses from the locals, raided the locality, and then moved on. They relied on speed and discipline. Their main targets were monasteries and churches, crammed with wealth and poorly if at all defended. If necessary, Viking bands would engage in battle if intercepted by an English army, but, unlike English kings in the wars they fought with their neighbors, Viking leaders did not seek battle, since battle was not necessary, and perhaps even detrimental, to the achievement of their goals. Those goals are materially represented by the contents of the hoards they buried for safe keeping. The Cuerdale Hoard discovered in 1840 on the banks of the river Ribble, near Preston, Lancashire, is one example. It contains about 7,500 silver coins and 1,000 other silver objects buried in lead lined chest, in A.D. 903–05. The hoard’s coins and objects are from widely distant regions (various parts of England, Francia, Italy, Ireland, Pictland), reflecting perhaps the areas plundered by the treasure’s owner. The differences between the Cuerdale and Staffordshire hoards are illustrative: the former is mainly a collection of coins and hack silver; the latter has a wealth of gold objects and, with few exceptions, all the pieces are weapons and armor fittings. The contents of the two hoards

10

David Hill and Margaret Worthington, Offa’s Dyke: History and Guide (Stroud, 2003); Ryan Lavelle, Fortifications in Wessex c. 800–1066 (Oxford, 2003), p. 11.

74

Richard Abels

are both plunder, but the Cuerdale Hoard is the wealth obtained from sacking monasteries and towns whereas the Staffordshire Hoard is battlefield booty. Though Vikings’ objectives were familiar to equally predatory Anglo-Saxon kings, the manner in which they waged war was new and disconcerting. While Anglo-Saxon commanders sought battle, Vikings avoided it. Their modus operandi involved seizing a defensible site, often a royal estate, and fortifying it further with ditches, ramparts, and palisades. From that base they would ride through the countryside, plundering as they went. If confronted by a superior military force, they would retreat to their camp. As slight as were its makeshift defenses, they nonetheless proved effective against an enemy unfamiliar with siege warfare and saddled with a logistical system designed only for short, decisive campaigns. A besieged Viking army would try to out-wait the enemy, knowing that once the besieging force exhausted its supplies, it would either have to leave or offer a profitable peace. Or, if the besiegers grew careless, the Vikings might burst out suddenly from behind their defenses in a furious counter-attack or sneak away under cover of night. Anglo-Saxon commanders often found themselves outmaneuvered or stalemated. Between 835, when sustained Viking raiding began in England, and 865, when “a great heathen raiding-army” (ASC: micel hæðen here] overwintered in East Anglia, Viking raiding fleets became larger and operated opportunistically on both sides of the Channel, though usually not at the same time. At first royal reeves and ealdormen dealt with the raiders. As the fleets grew in size, however, local forces became inadequate to meet the threat. From 836 on, kings responded to larger Viking incursions by raising armies from several shires with the intention of intercepting the raiding bands before they could do too much damage and force them to fight a battle. Then, in the autumn and winter of 865, the very nature of the threat changed. The ambitions of the leaders of the Great Heathen Army took on a territorial dimension.11 Between 865 and 878 every English kingdom with the exception of Wessex came under the political control of the Danes, either directly as in East Anglia, or intermediately through the establishment of native client kings, as in Mercia. If that failed, they contented themselves with extorting payment for what proved to be temporary peace.12 11

12

The word in the Anglo-Saxon Chronicle [hereafter ASC] translated as “army” is here. Michael Swanton translates it more precisely as “raiding-army” to distinguish it from fyrd. The AngloSaxon Chronicle, ed. and trans. Michael Swanton (New York, 1998), pp. xxxiii-xxxiv. On the composition and threat of this force, see Richard Abels, “Alfred the Great, the micel hæðen here and the Viking Threat,” in Alfred the Great: Papers from the Eleventh-Centenary Conferences, ed. Timothy Reuter, Studies in Early Medieval Britain 3 (Aldershot, 2003), pp. 265–79. On the difficulties of making peace with Vikings, see Niels Lund, “Peace and Non-Peace in the Viking Age,” in Proceedings of the Tenth Viking Conference: Larkollen, Norway, 1985, ed. J. E. Knirk (Oslo, 1987), pp. 255–69; Richard Abels, “King Alfred’s Peace-Making Strategies with the Vikings,” The Haskins Society Journal 3 (1992), 23–34; idem, “Paying the Danegeld: Anglo-Saxon Peacemaking with Vikings,” in War and Peace in Ancient and Medieval History, ed. Philip de Souza and John France (Cambridge, 2008), pp. 173–92; Ryan Lavelle, Alfred’s Wars: Sources and Interpretations of Warfare in the Viking Age (Woodbridge, 2010), pp. 315–34.



English Strategies during the Viking Wars

75

The goal of conquest and consolidation of territory required changes in strategy. The leaders of the Great Heathen Army continued to use rivers for transport, secure winter quarters, and obtain horses and supplies. But now they would seize, fortify, and make base in royal tuns with the intention of using them as long-term operating bases. A greater emphasis was placed on logistical planning, since the Danes needed to secure supplies for an entire winter rather than a few weeks. They relied on speed and surprise for “surgical strikes” intended to decapitate the targeted kingdom. From Danish traders and locals they obtained intelligence about the whereabouts of enemy kings/leaders, in order to launch surprise attacks, often in winter, before English kings or ealdormen could mobilize forces. The aim was to kill or expel these kings. Rather than engage in long-drawn-out wars with local magnates, they were, at least initially, content to set up disaffected Anglo-Saxon æthelings as client kings. In the first seven years of his reign, Alfred alternated between fighting the Danes, as he did in several battles in 871, and paying them off. Alfred’s great victory at Edington after Easter in 878 was due to the king’s ability to retain the loyalty of his nobility even after his near-capture by Guthrum the previous Christmas, and to mobilize forces to surprise an enemy who, by then, probably thought that he had been rendered helpless. But if Edington highlighted Alfred’s ability to inspire and lead troops, the events preceding the engagement, in particular Guthrum’s seizure of Chippenham, and with it the kingdom, illustrated just as dramatically the limitations of the military system Alfred had inherited. The West Saxon military establishment had been shaped by the kind of warfare that prevailed among the kingdoms of early England. The logistical inadequacies of the existing West Saxon military system were further exacerbated by the manner in which armies were raised. Assembling levies of local landowners and their followers was timeconsuming; Viking raiders could ravage an entire region before the king’s army appeared in the field. The towns of Wessex, as yet undefended, lay open to attack, a point dramatically underscored by the Viking sack of the kingdom’s greatest trading depot, Hamwic, in 840. The closest things to strongpoints in the kingdom were the royal villas or tuns, the defenses of which probably amounted to little more than ditches and palisades. The king’s army consisted of his household retainers, numbering perhaps a hundred or so warriors, and the shire levies led by his ealdormen. The former, a standing force, may have been the professional core of the king’s army, but its numbers were too few for it to conduct full-scale campaigns on its own. For that shire levies were needed. These territorial forces, consisting mainly of landowners and their followers, were raised on an ad hoc basis, a method of recruitment that severely limited their effectiveness against the Vikings. By the time the warriors could be gathered from the various localities, a highly mobile raiding party could have devastated a region and moved on. We know little about the king’s naval forces, or whether the royal fleet mentioned in the Anglo-Saxon Chronicle entries for 882 and 885 was an invention of Alfred’s or an inheritance from his predecessors. Alfred’s near disaster in 878 impressed upon him the need to reorganize the military resources of his kingdom in order to counter the specific threat

76

Richard Abels

posed by the Great Heathen Army. And this he did. Thirteen years later, when the Vikings returned in force, they found the kingdom defended by a standing, mobile field army, and a network of garrisoned fortresses that commanded its navigable rivers and Roman roads. Alfred had analyzed the problem and found a solution. If under the existing system he could not assemble forces quickly enough to intercept mobile Viking raiders, the obvious answer was to have a standing field force. If this necessitated transforming the West Saxon royal army from a sporadic levy of king’s men and their retinues into a mounted standing army, so be it. If his kingdom lacked strong points to impede the progress of an enemy army, he would build them. If Wessex lacked ships to confront the Vikings at sea, Alfred would build a fleet of “longships” larger, swifter, and more stable than the Viking warships of the period.13 Characteristically, Alfred’s innovations were firmly rooted in traditional West Saxon practice, drawing as they did upon the so-called “common burdens” of bridge work, fortress repair, and service on the king’s campaigns that all holders of bookland and royal loanland owed the Crown. To counter the mobility of the Danish invaders, Alfred reinterpreted fyrd service to mean two months of service in the field, with each soldier required to come not only armed but with a horse and, possibly, with supplies sufficient to last him for his tour of duty. The concept of fortresses was not original to Alfred. Decades earlier Pope Leo IV had ordered the construction of the fortified Leonine City to protect Rome from Saracen raids, and in the 860s Emperor Charles the Bald had spanned the Loire and the Seine with a number of fortified bridges.14 The Danes themselves routinely fortified their camps. Many if not most royal residences in Wessex would have had palisades and defensible gates, as evidenced by the Anglo-Saxon Chronicle’s story of King Cynewulf’s death at the hands of a rebel ætheling Cyneheard who took the king by surprise while he was visiting his mistress at the royal vill of Merantune (ASC, s.a. 757, referring to an event of 786). Whether or not pre-Alfredian Wessex had burhs, Mercia certainly had; at least one historian believes that these Mercian burhs formed a system of 13

14

ASC, s.aa. 893, 896. For a discussion of Alfred’s navy, Alfred the Great. Asser’s Life of King Alfred and Other Contemporary Sources, trans. Simon Keynes and Michael Lapidge (Harmondsworth, 1983), pp. 211, 289–90, 291. Alfred P. Smyth, King Alfred the Great (Oxford 1995), 109–13, and Richard Abels, Alfred the Great: War, Kingship and Culture in Anglo-Saxon England (London, 1988), pp. 194–207, 305–07, point out the practical problems of Alfred’s design. But cf. Edwin and Joyce Gifford, “Alfred’s New Longships,” in Alfred the Great, ed. Reuter (Aldershot, 2003), pp. 281–89. Ole Crumlin-Pedersen has suggested that Alfred’s design served as the model for the Viking longship: “Viking and Anglo-Saxon Longships,” FortySecond International Conference on Medieval Studies, Kalamazoo, MI, 10 May 2007. In terms of physical characteristics and methods of construction, the closest Continental analogues are the Carolingian and Ottonian Burgen of the German marches, though these tended to be quite a bit smaller than Alfred’s burhs. The similarities, however, may be due less to cultural transmission and conscious imitation than to a shared environment and similar threats. See Edward J. Schoenfeld, “Anglo-Saxon Burhs and Continental Burgen: Early Medieval Fortifications in Constitutional Perspective,” The Haskins Society Journal 6 (1995), 49–66 at 59–60, 65–66.



English Strategies during the Viking Wars

77

fortifications in the eighth century, although the archaeological evidence for this is slight.15 What was unique about Alfred’s scheme was its sheer scale, the strategic disposition and purpose of the burhs, and the administration through which he manned and maintained them. Alfred’s intention was not merely to fortify a few towns. He planned the construction of a network of burhs, and for this he could find no model in Britain, except perhaps the ancient Roman forts of the “Saxon Shore.”16 The strategic vision that underlay Alfred’s building program was as foreign to previous Anglo-Saxon strategic thinking as was the threat of the Vikings. Even if one dates the Burghal Hidage to Edward the Elder’s reign rather than that of his father, there can be little doubt that this document provides details of Alfred’s civil defense system. Alfred did not merely build fortresses; he created a unified system of West Saxon civil defense based upon an integrated network of permanently garrisoned forts and fortified towns that were designed to operate in tandem with one another and with Alfred’s mobile field force, the fyrd. Whether or not Alfred planned it, his reforms created an integrated defense-in-depth system.17 Defense-in-depth strategy assumes that an invader will enter the defender’s territory but establishes conditions that make it difficult for the invading force either to take land or profit from raiding. This is exactly how Alfred’s system operated in practice. Well-garrisoned burhs along the primary travel routes presented a major obstacle for Viking invaders. (See Map 1.) Even if a Viking force avoided the English field army and successfully raided the interior, the booty-laden marauders would face burghal garrisons as they attempted to return to their ships or strongholds. The most expensive element in Alfred’s strategic system was the thirty fortified centers of varying sizes he ordered either built or refurbished. These func15

16 17

Stephen Bassett, “Divide and Rule? The Military Infrastructure of Eighth- and Ninth-Century Mercia,” Early Medieval Europe 15 (2007), 53–85. In this article Bassett surveys the archaeological evidence for pre-Alfredian Middle Saxon defenses at the Mercian burhs of Tamworth, Winchcombe, and Hereford. Based on the ASC entry for 868, Nottingham probably also had defenses. The same is true of Derby, Leicester, Lincoln, and Stamford, which together with Nottingham were called the “Five Boroughs” of the Danelaw, by the early tenth century, although their defenses may have been Danish in origin. Bassett’s argument for an eighth-century Mercian burghal system is highly speculative, however. Certainly there is no evidence for a Mercian defensive network like that outlined in the Burghal Hidage in the reign of King Burgred. The best general discussion of the use and development of Anglo-Saxon fortifications is now Lavelle, Alfred’s Wars, pp. 209–63. Edward Luttwak was the first to characterize a premodern grand strategy as “defense-in-depth.” He contended that early fourth-century Roman emperors adopted a defense-in-depth strategy based on the construction of multiple layers of fortifications and a mobile field force: E. N. Luttwak, Grand Strategy of the Roman Empire: From the First Century AD to the Third (Baltimore, 1976), pp. 125–90. Cf. the critiques by J. C. Mann, “Power, Force and the Frontiers of the Empire,” Journal of Roman Studies 69 (1979), 175–83, and Benjamin Isaac, The Limits of Empire: The Roman Army in the East (Oxford, 1990), pp. 372–418. As a general caveat, given the limitations of the extant sources, assuming that one can ascertain strategic and tactical planning from the outcomes of medieval campaigns and battles risks succumbing to the “post hoc, propter hoc” fallacy.

78

Richard Abels

tioned both as refuges for the local population – no place in Wessex was more than twenty miles from a burh – and, more importantly, as impediments to Viking movement. Despite their apparent use of stone-throwing siege machines during the Great Heathen Army’s unsuccessful eleventh-month-long siege of Paris in 885–86, Danish raiding-armies in England showed no familiarity with siegecraft. (Nor, for that matter, did the English.) Burhs defended by ditch-andearthen rampart defenses, even without stone walls, seemed nearly impregnable to ninth-century raiders. Alfred and his planners situated the burhs so as to command all the major navigable rivers, estuaries, Roman roads, and trackways crossing or leading into his kingdom.18 A Viking fleet rowing up the Thames would encounter no fewer than five burhs in succession.19 Raiders sailing along the southern coastline of Wessex or the northern shores of Devonshire or Somerset would find few places where they could beach their ships without a fight. An extensive network of roads and trackways connected the burhs, making it possible for the garrisons to support one another and to work in tandem with the field force, while an integrated beacon system permitted the defenders in these burhs to be apprised of and respond to enemy movements well in advance. If, as seems likely, Alfred used the burhs to store his food rent, his field forces would never be more than a day’s march from food and supplies.20 New light has been shed upon the development of Alfred’s burghal system by a team of archaeologists at University College London headed by John Baker and Stuart Brookes, who over the last few years have been exploring civil defense in Viking-era England. Integrating archaeological, toponymic, geographical/topographical, and documentary evidence for civil defense systems in early England, Baker and Brookes add to our understanding about how Alfred’s burhs and the West Saxon beacon system were interrelated and how both fitted in with the 18

19

20

David Hill, “Gazetteer of Burghal Hidage Sites,” in The Defence of Wessex: The Burghal Hidage and Anglo-Saxon Fortifications, ed. David Hill and Alexander R. Rumble, (Manchester, 1996) pp. 189–228, provides an excellent overview. For the military perspective, see Richard P. Abels, Lordship and Military Obligation in Anglo-Saxon England (Berkeley, 1988), pp. 68–71, 236 nn. 67, 68; idem, “English Logistics and Military Administration, 871–1066: The Impact of the Viking Wars,” in Military Aspects of Scandinavian Society in a European Perspective, AD 1–1300, ed. A. N. Jørgensen and B. L. Clausen (Copenhagen, 1997), pp. 257–65 at 260–61. On the siting of the burhs and their archaeology, see David Hinton, Alfred’s Kingdom: Wessex and the South, 800–1500 (London, 1977), pp. 29–58. Assuming, of course, that Oxford was one of Alfred’s burhs. Oxford lay within Mercia and belonged to Ealdorman Æthelred until his death in 911, when Edward the Elder assumed control over both it and London (ASC s.a.). See Jeremy Haslam, “The Origin of the Two Burhs of Oxford,” Oxoniensia 75 (2010), 25–34. Coins were struck in Alfred’s name at Oxford, perhaps as early as the 880s. See M. A. S. Blackburn, “The London Mint during the Reign of King Alfred,” and Simon Keynes, “King Alfred and the Mercians,” in Kings, Currency, and Alliances: The History and Coinage of Southern England, AD 840–900, ed. M. A. S. Blackburn and David Dumville (Woodbridge, 1998), pp. 1–46. Cf. C. S. S. Lyon, “Historical Problems of Anglo-Saxon Coinage (4), the Viking Age,” British Numismatic Journal 39 (1970), 193–204 at 196–97. Barbara Yorke, Wessex in the Early Middle Ages, Studies in the Early History of Britain (London, 1995), p. 121.



English Strategies during the Viking Wars

79

military topography of Wessex.21 Pointing out, for example, that the Thames was probably not navigable beyond Oxford – if that far – they conclude that the burhs in the Thames River valley, with the exception of Sashes, were sited not to impede attacks up the Thames but to deter invasion from the north. The Thames in western Wessex formed a natural barrier to invasion, but it was a permeable frontier that could be crossed at fords or bridges. The need to block these crossing-points and to control network hubs for Roman roads explains the siting of the burhs of Cricklade, Southwark, Wallingford, and Oxford.22 The reason that a burh was built at Cricklade was not to protect against Vikings coming up the Thames, but to control the main Roman road connecting Cirencester to Silchester where that road crossed the river, which was one of the major routes into Wessex from Mercia. Wallingford, possibly a double burh, and Oxford, which lay on the river’s north bank, were similarly well placed to police the Icknield Way and the several river crossings that lay between them. The vulnerability of this area to attack is underscored by the seven herepaths – military roads – that lie in that area. Wallingford’s strategic importance is attested by 2,400 hides allocated to its defense. Sashes is the one clear exception: an island fort that lay up-river of London, Sashes impeded Viking fleets coming from the mouth of the Thames. That the burghal system on the Thames looks toward the north is an argument for the implementation of that system of fortifications at a time when invasion from Mercia was still a threat. One might date this to the period between Alfred’s victory at Edington in the spring of 878 and the “renovation” of London in 886 and establishment of Alfred’s son-in-law Ealdorman Æthelred as ruler of Mercia.23 21

22 23

John Baker and Stuart Brookes, Beyond the Burghal Hidage: Anglo-Saxon Civil Defence in the Viking Ages (Leiden 2013), pp. 269–333. Baker and Brookes reject the characterization of the Burghal Hidage as a record of an Alfredian defense system. They argue that the burhs listed in it were built at different times, some earlier than Alfred’s reign and others later, to meet specific strategic needs (at pp. 327–33), and that the document is both post-Alfredian in origin and just a snapshot in time of civil defense. The last I accept; the rest I believe to be mistaken. For a persuasive critique of their arguments, see Jeremy Haslam, “The Burghal Hidage and the West Saxon Burhs: A Reappraisal”, Anglo-Saxon England 45, 2017 (forthcoming). Baker and Brookes, Beyond the Burghal Hidage, ch. 5. Jeremy Haslam in a series of articles has argued strenuously for the dating of the entire Alfredian burghal system outlined in the Burghal Hidage (and the Burghal Hidage itself) to 878–79: Jeremy Haslam, “King Alfred and the Vikings – Strategies and Tactics, 878–886,” Anglo-Saxon Studies in Archaeology and History 13 (2005), 121–53; “Origin of the Two Burhs”; “King Alfred, Mercia and London, 874–886: A Reassessment,” Studies in History and Archaeology 17 (2011), 120–46. (These articles are posted on his website at http://jeremyhaslam.wordpress. com/, accessed 11 August 2017.) Haslam summarizes his thesis as follows: “Following his defeat of Guthrum’s army at Edington in 878, King Alfred put in place the system of forts and fortresses in Wessex and eastern Mercia which is listed in the contemporary Burghal Hidage document, which system reflected a policy both for the defence of the West Saxons as well as a strategic offensive against the Viking presence in Mercia and in London. The construction of this burghal system was arguably one of the principal factors which forced Guthrum to retreat from Mercia and London to East Anglia in late 879, their respective spheres of influence being redefined by a new boundary to the east of London which was set out in the contemporary

80

Richard Abels

Alfred intended his burhs to work in tandem with the fyrd and with each other. To do this effectively required a signaling system to warn the garrisons of the burhs of the approach of an enemy along any of the routeways, and a road system that would allow for the quick mobilization and movement of troops. Roman roads and ancient trackways were mainly used for the latter, but there is also evidence that they were supplemented by military pathways (herepaths), as seen above. A West Saxon beacon system interconnected with the burhs can be reconstructed by plotting out on a map OE placenames that combine words for “hill” or “look-out” with those relating to observation or fire.24 The UCL team and mathematician Keith Briggs with his student Jake Shemming independently examined these sites for intervisibility.25 The results indicate that the beacons in Wessex (but, interestingly, not in Kent) formed a single coherent system in which Cricklade played a critical role as a nodal point connecting Malmesbury to Chisbury or Wallingford. The location of the burhs of the upper and middle Thames at fords and Roman road river crossings and a beacon network designed to warn against incursions from Mercia suggest that, at least initially, Alfred thought in terms of a preclusive strategy rather than defense-in-depth. This stretch of the burghal system was meant to keep Guthrum or any other Danish chieftain from stealthily crossing the Thames and surprising Alfred as he had been surprised at Chippenham. The defensive system that Alfred created in the 880s and 890s was designed to defend against the simultaneous attacks of a number of different heres, precisely as happened in 892–95, and limit their ability to raid. Heres could enter his kingdom, but if they did, they were unlikely to make it back to their ships with their booty. As a result, Alfred was able to fight Vikings simultaneously in the eastern, northern, and western frontiers of his kingdom. These new Viking raiders discovered that English towns were no longer easy prey. It was dangerous to leave a garrisoned burh intact, but it was equally dangerous to attempt to take one. Possessing neither siege engines nor doctrine, they could not storm burhs protected by ditches, earthworks strengthened by wooden revet-

24

25

Treaty between Alfred and Guthrum.” “King Alfred, Mercia and London,” p. 120. Although Haslam demonstrates that it is physically possible for the thirty burhs of the Burghal Hidage to have been built in a single year, his thesis has not achieved wide acceptance. Many question whether Alfred would have had the power and wherewithal to launch so ambitious and labor-intensive effort so soon after Edington. Based on their research, Baker and Brookes (Beyond the Burghal Hidage) believe that the burghal system as depicted in the Burghal Hidage only gradually took shape. Guthrum’s sojourn at Cirencester in 878–79, however, provides an attractive historical context for Baker and Brookes’s findings about the northern orientation of Alfred’s civil defense system in the Middle Thames. Word elements indicating that the site was a vantage point include dūn, hyll, beorg, hlæw, and hnoð, all meaning “hill” or “mound”, and hōh, meaning “promontory.” Word elements meaning “watch” or “look-out” include weard and tōt. Those for fire include ād. See Baker and Brookes, Beyond the Burghal Hidage, pp. 184–89. Baker and Brookes, Beyond the Burghal Hidage, pp. 192–99, 312–27. The findings of Keith Briggs and Jake Shemming for an Anglo-Saxon communication network are posted online at http://keithbriggs.info/AS_networks.html.



English Strategies during the Viking Wars

81

ments, and palisades. If they attempted to starve a town into submission, the hunter was likely to become the hunted, as the fyrd and garrisons from neighboring burhs would come to the relief of the besieged. Alfred’s system, in short, had worked precisely as he conceived it would. The very geography of his last war attested to its effectiveness. In 871, 876, and 878, the Great Heathen Army had attacked and ravaged the very heartland of Wessex. In 892–94, an even larger army, with allies in Northumbria and East Anglia, had to content itself with raiding along the frontiers of Wessex and Mercia. Only once had Viking raiders penetrated the countrysides of Surrey or Hampshire, and those marauders had paid for their daring at Farnham. When the men of Somerset and Wiltshire fought, it was well beyond the borders of their shires. Alfred had proved to his enemies and his friends alike the wisdom of his demands “with regard to the building of fortresses and the other things for the common profit of the whole kingdom.”26 In the 880s, at the same time as he was “cajoling and threatening” his nobles into building and manning the burhs, Alfred undertook an equally ambitious effort to revive learning. It entailed the recruitment of clerical scholars from Mercia, Wales, and abroad to enhance the tenor of the court and of the episcopacy; the establishment of a court school to educate his own children and those of his nobles; an attempt to require literacy in those who held offices of authority; a series of translations into the vernacular of Latin works the king deemed “most necessary for all men to know”; the compilation of a chronicle detailing the rise of Alfred’s kingdom and house; and the issuance of a law code that presented the West Saxons as a new people of Israel and their king as a just and divinely-inspired lawgiver. This enterprise was to Alfred’s mind as essential for the defense of his realm as the building of the burhs. “The temptation we must resist,” Simon Keynes wisely admonishes, “is to stand back and admire a multiplicity of “different” Alfreds: the soldier, the law-maker, the statesman, the educator, and the scholar […] The genuine Alfred of the late ninth century was […] the integrated Alfred, for whom all these things were inseparable aspects of his determination to discharge the responsibilities of his high office for the good of his subjects and in the service of God.”27 Burhs, fyrds, and ships were the material expressions of civil defense; wisdom and piety were its spiritual dimensions. As Alfred observed in the preface to his translation of Gregory the Great’s Pastoral Care, kings who fail to obey their divine duty to promote learning can expect earthly punishment to befall their people.28 Conversely, the pursuit of wisdom, he assured the readers of his Boethius, is the surest path to power. The portrayal of the West Saxon resistance to the Vikings by Asser and the authors of the Anglo-Saxon Chronicle 26 27 28

Alfred the Great, trans. Keynes and Lapidge, pp. 101–02 (Asser, ch. 91). Simon Keynes, “The Power of the Written Word: Alfredian England 871–899,” in Alfred the Great, ed. Reuter, pp. 175–97 at 197. Alfred, King Alfred’s West Saxon Version of Gregory’s Pastoral Care, ed. and trans. Henry Sweet, 2 vols., EETS Original Series 45 and 50 (London, 1871–72), 1: Preface, pp. 2–3.

82

Richard Abels

was more than mere rhetoric or “propaganda.” It reflected Alfred’s own belief in a doctrine of divine rewards and punishments rooted in a vision of a hierarchical Christian world-order in which God is the Lord to whom kings owe obedience and through whom they derive their authority over their followers. If measured by effectiveness, Alfred’s military establishment was worth the money and manpower expended upon it. Not only did it prove the salvation of Wessex in the 890s, but in the hands of Alfred’s children and grandchildren it became a finely-honed instrument of aggression. Whether or not Alfred conceived of the burhs he ordered built as “islands of royal authority,” there can be little doubt that Edward and Æthelflæd did. There burhs were less intended as elements of a civil defense system than as anchors for the consolidation of conquest. The result was the creation through conquest of a unified kingdom of England. The true fruit of Alfred’s success was the halcyon reign of his great-grandson Edgar the Peaceable (959–75). As I have argued elsewhere, the consequence of peace was the abandonment of the more costly elements of Alfredian civil defense.29 When the Vikings returned in 980 to begin a new age of raiding and invasion, they found a peaceful and wealthy England ripe for pillaging. It was certainly a well-administered, or at least highly administered, kingdom, in which the central government had in place effective mechanisms for the maintenance of internal order and the raising of revenues. But one should not mistake bureaucratic efficiency and ideological sophistication for military strength. The civil defense systems that Æthelred II and Alfred inherited had much in common. Both relied on ad hoc levies summoned to meet crises, and while Æthelred’s England may have had many more defended towns, these burhs lacked the permanent military garrisons that had made them into something more than passive refuges. At least in this sense, Æthelred was unready to meet the new Viking threat. One popular explanation advanced for Alfred’s victory and Æthelred’s defeat is that the two kings faced quite different threats.30 On this view, the looselyknit bands of Danish thugs who ravaged England and Francia in the ninth century had little in common with the organized state armies led by Swein Forkbeard and his son Cnut the Great a century later. But is this “Whig” view of the Viking raids correct? Was there a radical change in the scale, organization, and objectives of Viking ventures over the course of the tenth century? The evidence for such a change is not compelling.31 If we go simply by the sources, Alfred in 892 and Æthelred in 1015 each faced fleets of 200–250 29

30 31

Richard Abels, “From Alfred to Harold II: The Military Failure of the Late Anglo-Saxon State” in The Normans and their Adversaries at War, ed. R. Abels and B. Bachrach (Woodbridge, 2001), pp. 15–30. Keynes, “Tale of Two Kings,” pp. 205–07. Jens Ulff-Moller, “The Vikings in England: A Reappraisal of Peter Sawyer’s Minimization Theory,” unpublished paper given at the Tenth International Conference of the Haskins Society, October 1991.



English Strategies during the Viking Wars

83

ships,32 carrying in the range of 3,000–12,000 combatants.33 And if we jettison the sources as untrustworthy, we are left with nothing but speculation. In short, there is no reason to believe that Æthelred faced significantly larger Viking armies than had Alfred.34 Nor may we assume that the forces of Swein and Cnut were organized in a manner radically different from previous Viking armies. Niels Lund has argued cogently that Danish fleets of the early eleventh century were organized along traditional lines, as loosely-knit gangs of warriors known as liths, rather than as royal, national levies (the leding).35 The leaders and “fellows” of these liths were motivated alike by the desire for plunder and tribute that would enhance their standing back home.36 That it is a mistake to see the invading Danish armies of 1013 and 1015 as “state” armies is underscored by the dubious role played in this period by Thorkell the Tall, whose transfers of loyalty made him the Danish Eadric Streona. It is clear from the sources that Thorkell acted as a free agent, and it is certain that he was not unique in this among the Danes.37 Alfred’s victory and Æthelred’s defeat cannot be explained simply by changes in Danish military organization and the consequent differences in the forces 32

33



34

35 36

37

As Nicholas Brooks has pointed out, there is basic agreement among ninth-century Irish, English, and Frankish sources that the “great” Viking armies of the period consisted of 100–250 ships. Occasionally these sources even concur about the specific number of ships in particular armies. Large round figures obviously represent estimates, but the agreement of independent contemporary observers suggests that we ought to take seriously the possibility that the Vikings ravaging England and Francia had fleets of 200 or so ships: Nicholas Brooks, “England in the Ninth Century: The Crucible of Defeat,” Transactions of the Royal Historical Society, 5th ser., 29 (1979), 1–20 at 2–11. See also C. Patrick Wormald, “Viking Studies: Whence and Whither?” in The Vikings, ed. R. T. Farrell (London, 1982), pp. 134–7. Cf. Carroll Gillmor, “War on the Rivers: Viking Numbers and Mobility on the Seine and Loire, 841–886,” Viator 19 (1988), 79–109, who argues for smaller numbers on the basis of logistical needs. For the Vikings of 892, see ASC, s.a. 892 A: 250 ships; B, C, D: 200 ships. (These figures do not count Hasteinn’s 80 ships.) Cf. Annales Fuldenses, ed. F. Kurze, MGH SSrerGer, Hanover, 1891, s.a. 882, which allows 200 ships to the fleet that a decade later was to set up shop at Appledore; ASC, s.a. 892. For Cnut’s fleet see Encomium Emmae Reginae, Camden Third Series 72, ed. Alistair Campbell (London, 1949), 2.1, 4, pp. 17, 19: 239 ships. Though it would be unwise to rely on any detail provided by this notoriously ill-informed writer, one should note that, for him, at any rate, a fleet of this size constituted an enormous and magnificent armada. The estimates for the complement of Viking warships vary considerably, from a low of twenty to more than fifty: Peter Sawyer, Age of the Vikings, 2nd ed. (London, 1971), pp. 126–27; Wormald, “Viking Studies,” p. 135; Gillmor, “War on the Rivers,” pp. 81–85. In 994, for instance, Swein Forkbeard, king of Denmark, and Olaf Tryggvason, soon to be king of Norway, raided the coast of southern England with only ninety-four ships: ASC, s.a. 994 C,D,E. Cf. ASC, s.a. 991 A, which gives Olaf a naval force of ninety-three ships, but on which see Janet M. Bately, “The Anglo-Saxon Chronicle,” in The Battle of Maldon, 991, ed. Donald Scragg (Oxford, 1991), pp. 37–50 at 43–49. Niels Lund, “The Armies of Swein Forkbeard and Cnut: ledung or lith?” Anglo-Saxon England 15 (1986), 105–18. See also idem, “The Danish Perspective,” in Maldon, ed. Scragg, pp. 114–42. Wormald, “Viking Studies,” pp. 145–48; Lund, “The Armies of Swein Forkbeard and Cnut,” pp. 105–06, 111–18. See also Timothy Reuter, “Plunder and Tribute in the Carolingian Empire,” Transactions of the Royal Historical Society, 5th ser., 35 (1985), 75–94. Lund, “Armies,” pp. 112–18.

84

Richard Abels

arrayed against them. Rather, we must return to the kings themselves, to their strategic planning and vision, and to the manner in which they implemented them. The Viking threat changed dramatically in magnitude and intent between 980 and 1016, and Æthelred and his advisors tried as best they could to match their strategic response to that changing threat. When the battle of Maldon was fought in August of 991 King Æthelred had no clearly defined strategy for dealing with the Vikings. This is not at all surprising, for at first the threat must have seemed modest. The raids of the 980s certainly caused local devastation, but they were sporadic and seemed a problem for local authorities and shire fyrds rather than the king. The king did act, however, to deprive the raiders of their safe haven in Normandy. By 990, relations between England and Normandy had become sufficiently hostile to be noticed by Pope John XV (985–96). As a result of Pope John’s intervention, Æthelred and Duke Richard agreed to terms, and a treaty was signed, by their representatives, in Rouen, on 1 March 991. Between 991 and 1005, the raiding fleets grew larger, as did the devastation, beginning with the ninety-four ship fleet led by Olaf Trygvasson from Norway and King Swein Forkbeard from Denmark that ravaged along the southeast coast in the summer of 991 and defeated Ealdorman Byrhtnoth and the Essex fyrd at Maldon. England enjoyed two years of peace after Olaf contracted a treaty with Æthelred in 994 and departed for Norway, but a new fleet ravaged England in 997–99, wintered in Normandy in 1000/1 and resumed raids in 1002. King Swein returned (possibly in response to the St. Brice’s day massacre) with a fleet in 1003 and pillaged freely until, compelled by a great famine throughout England, he left in 1005. As Alfred had done a century before, King Æthelred and his advisors took steps to deal with the growing threat. They began by addressing the vulnerability of the boroughs. Perhaps as early as the 990s the king began an ambitious program of military construction. New burhs were raised on the sites of iron-age hill-forts at South Cadbury in Somerset, Old Sarum in Wiltshire, and Cissbury in Sussex, and the defenses of existing burhs were refurbished, as stone walls replaced timber revetments and palisades.38 Æthelred and his advisors, however, lacked Alfred’s strategic vision. The construction was conducted in a piecemeal fashion, exemplified by the new hill-forts, each apparently built as a consequence of a threat to or the sacking of an existing mint. The haphazard character of Æthelred’s civil defense program is also suggested by the siting of Kent’s beacons. Unlike the later Armada beacon system, the views-

38

See C. A. R. Radford, “The Later Pre-Conquest Boroughs and Their Defenses,” Medieval Archaeology 14 (1970), 81–103; M. Biddle, “Towns,” in The Archaeology of Anglo-Saxon England, ed. D. M. Wilson (London, 1976), pp. 99–150 at 140–41. On the excavations at Cadbury, see Leslie Alcock, “By South Cadbury is that Camelot…”: The Excavations of Cadbury Castle 1966– 1970 (London, 1972), pp.194–201. There is evidence for the refurbishing of defenses during the late Saxon period at Wallingford, Cricklade, Amesbury, Malmesbury, Wareham, Christchurch, Lydford, and Daws Castle near Watchet. See Anglo-Saxon Towns in Southern England, ed. Jeremy Haslam (Chichester, 1984), pp. 76, 109, 128, 137, 153, 188, 193, 258.



English Strategies during the Viking Wars

85

heds of Kent’s Anglo-Saxon beacons look landward and over navigable rivers rather than toward the coast. Baker and Brookes found that Kent’s beacons, in contrast to the Alfredian beacon system of the Thames River valley, “operated in three discrete groups centring on the Medway, Wantsum and eastern Weald, with no apparent point of intervisibility between these networks.”39 These findings are certainly curious given the seaborne character of the Viking threat to Kent in the late tenth and early eleventh centuries. They may indicate no more than that Kent’s beacon system, which looked north and west, was set up at an earlier time when coastal attacks represented less of a threat. On the other hand, Æthelred understood the importance of coastal defense and was responsible for refurbishing the defenses of Dover, including building there the look-out/beacon of St. Mary-in-Castro. That he did not set up an intervisible network of beacons in Kent is indicative of his deficiency in strategic vision, as is his failure to develop an overall strategy of response in which the newly refurbished burhs could play a role beyond that of local refuges. There was no plan to establish permanent garrisons in the burhs or to use them as bases to mobilize troops for defense. In short, Æthelred’s civil defense projects failed to result in an Æthelredrian burghal system to match Alfred’s, and, as a consequence, this ambitious and expensive military construction program as a whole fell considerably short of the sum of its parts. Confronting a disciplined and mobile predatory enemy, Alfred had responded first by attempting to defeat them in battle, and, when victory proved elusive, had adopted Charles the Bald’s strategy of purchasing peace. Only when that failed did he undertake his ambitious program of fortress construction and reconstituting the fyrd into a standing army. Without his victory at Edington, Alfred probably would not have had the prestige or leverage with the West Saxon nobility to impose upon them his draconian new interpretation of the common burdens. Æthelred did not have an Edington. He did not even enjoy the reputation of a warrior king. Unlike his predecessors (and successors), King Æthelred preferred to appoint generals than lead his troops on campaign, and for the first two decades of his reign had placed the defense of his realm solely in the hands of his ealdormen and reeves. Even if he had thought of returning the fyrd to the professional standing army it had been a century earlier, the king probably did not possess the political capital to persuade or coerce his nobles into compliance. So Æthelred resorted instead to tribute-paying, diplomacy, and, finally, the employment of Vikings as mercenaries. Ealdorman Byrhtnoth’s disaster at Maldon in 991 convinced Æthelred and his counselors of the gravity of the situation and of the wisdom of purchasing peace. The ten thousand pounds offered to the raiders in 991, however, proved to be only the first of many such payments that became increasingly expensive as the invading armies grew larger and hungrier.40 From the early 990s, 39 40

Baker and Brookes, Beyond the Burghal Hidage, ch. 6. On the reliability of the Chronicle’s escalating figures for the payment of Danegeld, see the exchange between John Gillingham and M. K. Lawson in The English Historical Review: Gillingham, “‘The Most Precious Jewel in the English Crown’; Levels of Danegeld and Here-

86

Richard Abels

Æthelred used diplomacy and cash to divide his enemies and deprive them of foreign support. Unlike Alfred, whose wartime diplomacy focused on neighboring Mercia and Wales, Æthelred’s foreign policy was conducted against the backdrop of continental politics, reflecting how much more England – and Scandinavia – was now integrated into the medieval European state system. The dukes of Normandy, for example, were alternately threatened and courted, as the king tried to close their ports to his enemies, which he finally accomplished by marrying Emma, sister of Duke Richard, in 1002.41 Though Alfred was well informed about Viking activities on the continent and knew that the same bands were ravaging both kingdoms, he apparently did nothing to coordinate defenses with his West Frankish contemporaries, probably because his overall policy was to make Wessex a less inviting target than Francia. This was not an option for Æthelred, which should serve as a reminder that the two kings lived in quite different political worlds. King Æthelred attempted to divide his enemies by hiring Vikings to fight other Vikings. As with the payment of tribute, this was a well-established method for dealing with raiders, one that the West Frankish King Charles the Bald had used in the 860s. In 994 Æthelred managed to separate the Norse chieftain Olaf Tryggvason from his erstwhile ally the Danish King Swein, even standing sponsor at the savage young chieftain’s confirmation.42 In return for twenty-two thousand pounds, gifts of friendship, and provisions for his men, Olaf agreed to aid Æthelred against his enemies.43 That year Olaf, with Æthelred’s blessing, departed England, never to return. Although Olaf never served Æthelred as a mercenary captain, his activities in Norway drew Swein’s attention and kept the Danish king occupied until the battle of Svold in A.D. 1000.44 But from the treaty’s provisions regulating feuds and trading between Danes and

41

42

43 44

geld in the Early Eleventh Century,” EHR 104 (1989), 373–84; Lawson, “‘Those Stories Look True’; Levels of Taxation in the Reigns of Æthelred II and Cnut,” EHR 104 (1989), 385–406; Gillingham, “Chronicles and Coins as Evidence for Levels of Tribute and Taxation in Late Tenth and Early Eleventh Century England,” EHR 105 (1990), 939–50; Lawson, “Danegeld and Heregeld Once More,” EHR 105 (1990), 951–61. James Campbell, “England, France, Flanders and Germany in the Reign of Ethelred II: Some Comparisons and Connections,” in Ethelred the Unready, ed. Hill, pp. 255–70, repr. in James Campbell, Essays in Anglo-Saxon History (London, 1986), pp. 199–201; Elisabeth M. C. van Houts, “The Political Relations between Normand and England before 1066 according to the Gesta Normannorum Ducum,” in Les Mutations socio-culterelles au tournant des XIe–XIIe siècles, ed. R. Foreville and C. Viola (Paris, 1984), pp. 85–97. Theodore M. Andersson, “The Viking Policy of Ethelred the Unready,” Scandinavian Studies 59 (1987), 284–95. Cf. Phyllis R. Brown’s and Peter Sawyer’s responses in the same issue of the journal. For the text of Æthelred’s treaty with Olaf, see Simon Keynes, “The Historical Context,” in Maldon, ed. Scragg, pp. 103–07; English Historical Documents, I, no. 42. Lund, “The Danish Perspective,” pp. 138–40. Fourteen years after Olaf Tryggvason’s defeat, Æthelred helped another Norwegian Olaf, St. Olaf, obtain the throne, undoubtedly with an eye toward creating mischief for his enemies at home: Gwyn Jones, The Vikings (New York, 1968), p. 375.



English Strategies during the Viking Wars

87

Englishmen, it would seem that at least part of the fleet remained in England, serving Æthelred as a mercenary army to deter future raiders. Æthelred endowed some of the fleet’s leaders – notably the Danish chieftain Pallig – with estates in return for pledges of loyalty, in an attempt to embed them into the existing political and social structures. This may not have proved a good bargain as matters turned out. In 997 a Viking fleet, perhaps including some of those who were supposed to be in Æthelred’s service, ravaged the West Country. Four years later, when a new Viking fleet appeared off the coast of Devonshire, Pallig joined the raiders with as many ships as he could assemble, “in spite of all the pledges he had given” and the gifts of land and gold and silver he had received from the king.45 Æthelred’s response was to purchase another peace with the Vikings, for twenty-four thousand pounds. On St. Brice’s day in 1002, Æthelred made a bold attempt to eliminate the problem of untrustworthy Danish mercenaries in one fell swoop by ordering (in the words of a royal charter of 1004) a “most just extermination” of “all the Danes who had sprung up in this island, sprouting like cockles amongst the wheat.”46 The St. Brice’s day massacre was the act of a king confident that he could handle the existing Viking threat without mercenaries, especially since some of them had proved unreliable. If that is what King Æthelred believed, he was wrong. The level of Viking activity intensified once more with the arrival of a “great fleet” under the command of Tostig in the autumn of 1006. Æthelred and his advisors responded as Alfred had a century before, acting to improve England’s military forces. The basic strategy that the king adopted was preclusive, centering on the construction of a massive royal fleet to intercept raiders at sea. In 1008 he ordered the kingdom to be divided into naval districts of 300 or 310 hides, to facilitate the construction and maintaining of a great armada. At the same time, he ordered that a helmet and corselet be supplied from every eight hides “unremittingly over all England” (ASC, s.a. 1008). If we go by the hidage total of Domesday Book, this would have meant a fleet of about two hundred ships and armor for nine thousand warriors. This is in itself testimony to the effectiveness of English institutions of governance in the early eleventh century. (It also suggests that many soldiers in the English fyrds had been less well armored and armed than their Viking enemies.) One can only speculate where the king’s armory or armories were, and how and to whom his officers distributed the weapons stored there. What is certain, though, is that Æthelred 45

46

For Pallig, see ASC, s.a. 1001. Simon Keynes suggests that Pallig received his “great gifts, in estates and gold and silver” in connection with this treaty: S. D. Keynes, “The Vikings in England: c. 790–1016,” in The Oxford Illustrated History of the Vikings, ed. Peter Sawyer (Oxford, 1997), pp. 42–82 at 77. See Ann Williams, Æthelred the Unready, The Ill-Counselled King (London, 2003), pp. 52–55. The charter is Sawyer no. 127, “Electronic Sawyer: Online catalogue of Anglo-Saxon charters,” http://www.esawyer.org.uk/charter/127/html (accessed 11 August 2017), translated in EHD, 1, ed. Whitelock, no. 127, renewing title-deeds to St. Frideswide’s, Oxford. St. Frideswide’s lost its charters in a fire set by the burgesses, intent on killing the Danes who had taken refuge in the church.

88

Richard Abels

used the powerful institutions of governance of late Saxon England to remedy the deficiencies in his military forces. In 1009 King Æethelred ordered his new fleet to be stationed off Sandwich to guard against the return of the Vikings. But the naval preparations came to nothing. Internal disputes led to a noble accused of treason “enticing” the crews of twenty ships to raid the southern coast of England, while the thegn charged with capturing him lost an additional eighty ships to a sudden storm. King Æthelred and his ealdormen decided to leave the fleet; the surviving ships were brought back to London. In the words of the chronicler, “we had neither the luck nor the honour that the ship-army was useful to this country, any more than it often was before.”47 The disheartening dispersal of the national fleet was followed by the unimpeded arrival of Thorkell’s great fleet in August 1009. Æthelred, probably after consulting Archbishop Wulfstan II and his other clerical advisors, concluded that the new micel here was God’s punishment for the sins of the English people. Echoing Carolingian practice, Æthelred responded with an impressive program for national repentance, prayer, and alms-giving.48 At Elmham and Bath in 1009 the king issued legislation drafted by Archbishop Wulfstan (V–VI Æthelred and VII Æthelred) that began with the admonition that “one God shall be loved and honoured above all, and all men shall show obedience to their king in accordance with best traditions of their ancestors and cooperate with him in defending the realm.” The legislation decrees that every adult Christian give alms, fast, and pray, and prescribes the litanies that the clergy were to chant to obtain the divine favor necessary for victory.49 Associated with this legislation is the litany preserved in the eleventh-century Winchester Troper: “that you may see fit to preserve King Æthelred and the army of the English.” These litanies “contra paganos” generally concluded with the three-fold invocation of the Lamb of God calling upon Christ to “grant us peace.” Simon Keynes reasonably associates the issuance of Æthelred’s agnus dei coin type in 1009 with the legislation and the enforced litanies.50 Æthelred’s program of repentance, prayer, and alms-giving proved no more effective than the ship-sokes. Even with helmets and coats of mail, English fyrdmen proved no match for Thorkell’s tested Vikings. After defeating Ulfcytel of East Anglia in battle at Ringmere in May 1010, Thorkell’s Danes spent the next six months ravaging East Anglia, the East Midlands, and Wessex without significant resistance, before returning to their base in Kent. According to the writer of the Anglo-Saxon Chronicle, King Æthelred summoned his witan for advice about how the country should be defended, but whatever policy the king adopted would be discarded within a few weeks for another. The chronicler 47 48 49 50

ASC, s.a. 1009. What follows is based on Simon Keynes, “An Abbot, an Archbishop and the Viking Raids of 1006–7 and 1009–12,” Anglo-Saxon England 36 (2007), 151–220. Ibid., pp. 177–201). Ibid.



English Strategies during the Viking Wars

89

attributed the inability of the English to deal with Thorkell to the king’s and his counselors’ poor timing and inconstancy of purpose: “All these misfortunes befell us through lack of decision, in that they were not offered tax [C: nor fought against] in time; but when they had done great evil, then a truce and peace was made with them. And nonetheless for all this truce and peace and tax, they travelled about everywhere in bands and raided and roped up and killed our wretched people.”51 Incapable of mounting any effective military resistance to Thorkell, Æthelred decided to hire him instead, and somehow managed to persuade the Viking captain that it would be more profitable to eat from the king’s table than to steal food from it. Thorkell agreed that he and his forty-five ships would defend Æthelred’s realm in return for being fed and clothed. To fulfill his end of the bargain Æthelred instituted a regular tax, the much-hated impost known as the heregeld.52 The new policy also failed, as Thorkell was either unable or unwilling to oppose King Swein when he invaded in the following year. King Æthelred II’s third and final strategy, one adopted and followed with greater constancy by his son Edmund Ironside, was a return to battle-seeking. After Swein’s death in February 1014, the magnates of England, lay and clerical, invited Æthelred to return from Normandy to be their king once again. At the same time, the Danish fleet and the chief men of Lindsey chose Swein’s son Cnut to be their king. Emboldened by the sworn support of the nobility and needing to demonstrate his worthiness to be king, Æthelred at the head of a large army marched north to confront Cnut in Lindsey. Cnut set sail for Denmark, putting ashore at Sandwich long enough to mutilate the hostages that the people of London had given to his father, while Æthelred contented himself with ravaging Lindsey as brutally as any Viking here. In the autumn of 1015, Cnut returned with a fleet of 160 ships to raid the south of England. The ætheling Edmund Ironside, who had established an independent power base in the Danelaw, without the blessing of his father, responded by raising an army in the north, intending to join up with the forces of the Mercian ealdorman Eadric Streona and confront Cnut with their full power. Edmund, however, grew suspicious of Eadric’s intentions, and returned north without engaging Cnut. In midwinter Cnut, now with the open support of Eadric Streona, began a systematic campaign of conquest. His strategy was simple: he would ravage a territory until the demoralized local leaders submitted to his authority, and then move on to the next shire. The obvious strategic response was to raise an army, force Cnut into battle, and destroy his forces. King Æthelred, never comfortable with military command and in failing health, chose, however, not to take the field. His son attempted to fill the vacuum. The first army that Edmund raised in the name of his father dispersed when it became clear that the king had no intention of joining it. A second “national army” disintegrated when

51 52

ASC, s.a. 1011 E, trans. Swanton. ASC, s.a. 1012.

90

Richard Abels

King Æthelred abandoned it because of rumors that he was to be betrayed. It was not until King Æthelred’s death on 23 April and Edmund’s election as his successor that the English were finally able to mount a forceful resistance. As Cnut continued to ravage across Mercia, East Anglia, and Northumbria, King Edmund raised a large army. He pursued Cnut and brought him to battle, inconclusively, at Penselwood in Somerset and then at Sherston in Wiltshire. A third battle at “Assandun” (probably modern Ashingdon) in Essex proved decisive. Edmund lost the battle when Ealdorman Eadric Streona of Mercia, who had ostensibly returned to Edmund’s side, treacherously fled the battlefield with all his forces. Edmund was forced to sign a treaty with Cnut, conceding to him all of England except Wessex. When Edmund died about a month later, Cnut became king of all England, bringing to an end the Viking Wars. The parallels between the threats of 865–96 and 980–1016 were greater than often recognized. Both represented military threats that began with raiding parties numbering perhaps a few hundred men and ended with “great armies” of several thousand or more bent on conquest and consolidation. In both cases, existing English strategic doctrine and military institutions were incapable of countering the new threat represented by Vikings, and in both cases the leaders of those states attempted both to reconstruct and reorganize their militaries in order to counter the evolving threats, and to appease God’s apparent wrath by correcting spiritual deficiencies. Both Alfred and Æthelred recognized the inadequacies of the systems they inherited, and both understood the importance of strategic planning. Nonetheless, the former succeeded and the latter failed. There are many reasons for this, but one critical factor was certainly the strategic vision of Alfred that led to the creation of an effective and coherent civil defense system and the lack of such a vision on the part of Æthelred and his advisors. If Alfred’s strategic responses to the military threat posed by the Great Heathen Army were synergistic, Æthelred’s were the opposite. The individual actions he took were impressive; taken together, however, they were ineffective. Probably the best judgment on Æthelred’s strategic planning remains that of the writer of the Anglo-Saxon Chronicle: When the enemy was in the east, then the army was in the west; and when they were in the south, then our army was in the north. Then all the councillors were ordered to the king, and it then had to be decided how this country should be defended. But whatever was then decided, it did not stand for even a month. In the end there was no head man who wanted to gather an army, but each fled as best as he could; nor in the end would one shire help another.53

53

ASC, s.a. 1009 E, trans. Swanton.

4 Norman Conquests: A Strategy for World Domination? Matthew Bennett

Strategy is the highest level of a tripartite view of warfare (in which the operational and tactical levels are the other two). It is concerned with the creation of plans and decision-making in order to lend success to military operations. The word is Greek in origin, the strategos being the commander, the brains behind what the soldiers (and sailors) are told to do. In the modern day the word is used to describe any kind of idea or process for achieving an end: for example, a “sales strategy,” or the way for one sports team to defeat another. Key to its meaning, in my interpretation, is that it is in the realm of the mind. Strategy involves knowing what you want to do, or at least devising an approach in order to achieve a goal. This means understanding not only your own mind, but also that of your opponent. A leader’s strategy requires a vision of what needs to be achieved, for only then can this be broken down into means and approaches to that end. I have linked this to ideology because every cause needs some kind of justification (even that of naked greed) if it is to appeal to followers. Most leaders, if they are canny, will seek to demonstrate the highest possible ethical grounds for their intentions, for their aggression might bring down upon them a weight of criticism that impairs the effectiveness of their plans.1 The reason I have chosen the Normans to provide examples for strategies of warfare is not that I believe them to be some sort of super-race (although they themselves, and some historians subsequently, might represent them as such), but because I know their history well. This is the result of having been taught by that great “Normanist” R. Allen Brown, some thirty years ago, and also of my own inclination to follow their adventures from the borders of Scotland through France and Italy and across the seas to Greece and Syria. When we view this broad span of Norman conquests, it is impossible not to wish to investigate the strategic considerations of these ambitious campaigns, far from the rolling hills and orchards and meadows of fertile Normandy, with its long sea coast, but poorly provided with harbors, and the great diversity of landscapes and seas of the regions which they eventually reached.

1

Beatrice Heuser, The Evolution of Strategy: Thinking War from Antiquity to the Present (Cambridge, 2010).

92

Matthew Bennett

Vision and Ideology Of course, it took a vision to make men venture so far in search of fame and fortune. If we look at the Duke William’s invasion of England, we can see a man inspired by his belief that he possessed a right to the country’s throne. His claim, not as strong as many, came in essence from his great-aunt’s marriage to King Æthelred II in 1002. This dynastic match was made for the very sound strategic reason that the kingdom of England was under attack from Scandinavian adventurers, who shared a common ancestry and culture with the original creators of Normandy, Hrolf the Ganger’s Danish followers.2 Æthelred’s plan was to end the succor these Viking raiders received from across the Channel, where they were able to winter and so resume attacks on England as soon as the weather allowed it. In the end the plan failed, and the Danes conquered England in 1016, with Cnut replacing Æthelred both as king and as husband to Emma. When Cnut died in 1035, to be replaced first by a bastard and then his legitimate heir until 1041, when Edward son of Æthelred returned to take the throne, William was only a teenager, and besides, his father and his regents had given Edward refuge in the Norman court. Edward’s return upset a political structure established for twenty-five years, and during a dispute with the powerful family of the AngloDanish Earl Godwin in 1051, which led to the earl and his entire family being exiled, Duke William may have been nominated as Edward’s heir in the absence of any issue of his own. (This was because Edward had been obliged to marry Godwin’s daughter Edith, but she had borne no children, possibly because she was unable to, but also perhaps because Edward was unwilling to have children by the daughter of the man responsible for his brother’s death.) As it turned out, Godwin and his clan returned in triumph the following year and the Godwinsons dominated the rest of Edward’s reign. So it was that in January 1066 Harold was on hand at Edward’s deathbed to have himself denominated the dying king’s successor, and to be promptly crowned the next day in Westminster Abbey.3 William believed, however, that he still had a claim to the throne based on the old king’s wishes of fifteen years earlier, and also upon the oaths sworn to him by Harold when he was William’s prisoner in 1064. According to the Norman version, these were that on Edward’s death Harold, his emissary in France, would support William’s accession and indeed hold and maintain important fortresses such as Dover for the duke’s arrival.4 So William’s vision was based upon his sense of righteousness, and of the perjury of King Harold; in his mind God was surely on his side. In its way, this amounted to an ideology: a belief that the Divine intervened in the affairs of Man. This attitude was 2 3 4

See David Crouch, The Normans: The History of a Dynasty (London, 2002), with dynastic tables on pp. 5 and 31. For an analysis of the strategic context and the crisis of 1051, see R. Allen Brown, The Normans and the Conquest (London, 1969) pp. 109–27. Ibid. pp. 127–31, drawing on The Gesta Guillelmi of William of Poitiers, ed. and trans. R. H. C. Davis and Marjorie Chibnall (Oxford, 1996), 1.41–42, pp. 68–71.



Norman Conquests: World Domination Strategy?

93

confirmed when William sought aid from Pope Alexander II in order to justify his invasion. Supposedly Alexander entrusted the duke, upon his embassy to the Vatican, with a banner bearing the papal arms. This may or may not have been true, and it is possibly a post facto justification of William’s success; but it is entirely in keeping with the Norman approach.5 After all, Count Robert Guiscard, a Norman adventurer in Italy, had asked precisely the same thing of Pope Nicholas II in 1059, and had been granted it as a symbol of his dukedom and in recognition of his protection of the pope against hostile elements in Rome.6 Furthermore, Duke William was able to cast his invasion of England as a necessity to reform a corrupt Church there, headed by a man, Archbishop Stigand, who was a pluralist and who had received his pallium from a pope who had later been declared illegitimate. This was in the context of the “reform papacy,” which Alexander represented and which was the dominant ideological program in Western Christendom at the time. Certainly William seems to have had no trouble raising troops to fill his invasion force, and the legitimacy conferred by papal support may have been as important as the desire for gain amongst his mercenary followers.7 That this desire was central to the spirit of the age is confirmed by the operations of the Normans in Italy, where they first arrived half a century before 1066, serving as mercenary troops to the Byzantines. The Greeks were always keen to recruit warriors with a fierce reputation and specific skills into their armies. Indeed, from the mid-eleventh century onward, Normans commanded one of the foreign regiments, of Franks, and played a crucial role in Byzantine campaigns as far afield as Sicily and eastern Asia Minor.8 During the troubles of Duke William’s minority, and later when the duke wanted to get rid of troublemakers, more Normans followed in the footsteps of the first, and as their numbers grew they proved able to set up their own regime in defiance of the Greeks and Lombards who had first employed them. This striving for independence can also be seen as a kind of ideology: the belief that they possessed superior military virtues to the natives of southern Italy and that this gave them the right to rule. The Hauteville clan proved the most successful, and the greatest of these proved to be Robert Guiscard “The Wily,” who eventually established a dynasty that ruled in both southern Italy and Sicily9 5

6 7 8

9

R. A. Brown always admitted that he put forward the Norman version, considering it to be true, while John Gillingham has expressed doubts about whether papal blessing came before the Conquest (personal communications). G. A. Loud, The Age of Robert Guiscard: Southern Italy and the Norman Conquest (Harlow, 2000) pp. 127–30. The structural and territorial reorganization of the English Church in the 1070s was as much a part of a control strategy as was imposing castles and knight-service on the country. Alicia J. Simpson, “Three Sources of Military Unrest in Eleventh Century Asia Minor: The Norman Chieftains Herve Frankopoulos, Robert Crispin and Roussel de Bailleul,” Mesogeios, 9–10 (2000), 181–207, provides the most recent survey of the careers of Normans in Byzantine military service. Loud, Age of Robert Guiscard, passim but especially chapter 6, pp. 234–90.

94

Matthew Bennett

The reformed papacy was meanwhile developing another ideology that justified conquest. The Church had been striving for decades to create a situation in which the troublesome activities of the knightly classes, forever fighting for fun, sport, and profit, and caring not how much damage they inflicted upon Christian society as a result, could be brought to serve a more beneficial cause. Gregory VII, the greatest reformer of all (although also the most intransigent, which was to lead to his downfall), conceived the idea of a militia Christi in which the energies of the knights would be put to good use rather than expended in struggle between one another.10 Their deaths, instead of resulting needlessly from godless fighting, thus putting their immortal souls in peril, could rather serve to aid the cause of (Latin) Christianity, wherever was needed. This was in the context of the irruption of the Seljuk Turks into the Middle East in the mid-eleventh century, and their defeat of the both the Byzantines, at Manzikert in 1071 (where the Norman mercenaries contrived to be absent from the field), and the forces of the Muslim world almost everywhere; as a result, the city of Jerusalem, which had become a pivotal center for pilgrimage since the millennium, and especially the thousandth anniversary of the Passion in 1033, fell into the hands of a band of warriors who seriously interrupted the pilgrim trade.11 Thus, when in November 1095 Pope Urban II preached that the churches of the East must be rescued from oppression, and his audience took that to mean the physical liberation of Jerusalem, the ideological movement to be known as the Crusades got under way.12 This justification for adventuring in God’s cause was immediately popular with the military classes (and with a number of others, followers of Peter the Hermit), and was a critical element in the creation of armies and navies sent eastward in the papal cause. It is difficult to judge the extent to which individual popes and their advisors and administrators had a “strategic” conception in relation to these eastern expeditions; but if war is to be considered the “continuation of policy by other means,” and diplomacy has always been part of the designs of political leaders, then it is not unreasonable to assign the papal curia an eye for strategy.13 Certainly, although popes often found it difficult to influence crusading expeditions once they had left for the 10

11

12

13

H. E. J. Cowdrey, “The Peace and Truce of God in the Eleventh Century,” Past and Present 46 (1970), 42–67; I. S. Robinson, “Gregory VII and the Soldiers of Christ,” History 58 (1973), 169–92. When the present paper was delivered at the 2009 Caceres conference there was no easily accessible monograph available on the impact of the Seljuks; but now see A. C. S. Peacock, The Great Seljuk Empire (Edinburgh, 2015). The bibliography of crusade histories is now vast. The most influential recent historical writing in English has been by Jonathan Riley-Smith, The Crusades: A History, 2nd ed. (1987, repr. London, 2005), and his pupils. Christopher Tyerman, God’s War: A New History of the Crusades (London, 2006), proposes an alternative critique, while John France, The Crusades and the Expansion of Catholic Christendom 1000–1714 (London, 2005) provides both a comprehensive and innovative study, together with a thorough exploration of the military aspects of the topic. The quotation is from Carl von Clausewitz, On War, ed. and trans. Michael Howard and Peter Paret (Princeton, 1989), p. 87; although Maurice Keen argued the reverse for medieval warfare: M. H. Keen, The Laws of War in the Late Middle Ages (London, 1965).



Norman Conquests: World Domination Strategy?

95

East – the Fourth Crusade of 1202–04 being possibly the most outstanding example – this does not mean that they were incapable of thinking in military terms. Indeed, the most recent research into the Second Crusade of 1146–48 suggests that Pope Eugenius III had a very clear concept that non-Christians should be confronted wherever they might be found, and that he endorsed a threefold strategy, in Iberia, in the Baltic lands, and in Syria, in order to achieve the aim of both protecting and expanding Christendom.14 There were of course Normans on the First Crusade, although not the king himself (on this occasion no kings went), but rather his brother Robert, duke of Normandy. Robert was able to attract followers owing to his reputation as a warrior. Italo-Normans too were present: the progeny of Guiscard and his clan, Mark Bohemond (named after a legendary giant) and his nephew Tancred d’Hauteville.15 All these Normans played a great role on the Crusade, although its prominence may have been heightened thanks to the attentions of supportive chroniclers, notably the anonymous author of the Gesta Francorum and Ralph of Caen, who wrote the Gesta Tancredi.16 Strangely, perhaps, Duke Robert did not receive his own biographer, although he was allegedly offered the throne of Jerusalem and turned it down in favor of a return to Normandy and an attempt upon the throne of England.17 Bohemond established a dynasty ruling Antioch and later invaded Epirus (eastern Greece), in a bid for the Byzantine throne which failed. Clearly, all these men were thinking strategically, and on a continental scale, of where their best opportunities might lie, and how to exploit them. Family and Politics The foregoing has served to indicate the role played by family and politics in the machinations of the ruling elites in Western Europe and beyond. Duke William’s successful conquest of England was dependent in large part upon the problems of the succession caused by the inability of King Edward and his family by marriage to live in harmony. A major figure not previously mentioned in regard 14

15

16

17

Jonathan Phillips, The Second Crusade: Extending the Borders of Christendom (New Haven, 2007): for his analysis of the purpose of the papal bull Quantum praedecessores see pp. 50ff.; for motivations of the identifiable crusading dynasties in the various theaters of operation, chapter 6. Ralph Bailey Yewdale, Bohemond I, Prince of Antioch 1058–1111 (Princeton, 1917), and Robert Lawrence 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) remain the only biographies, although these high-profile figures feature in all primary chronicles and secondary studies of the First Crusade. Gesta Francorum et aliorum Hierosolimitanorum: The Deeds of the Franks and Other Pilgrims to Jerusalem, ed. and trans. Rosalind Hill (London, 1962); 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). William M. Aird, Robert Curthose, Duke of Normandy (c. 1050–1134) (Woodbridge, 2011) is the most recent modern biography.

96

Matthew Bennett

to the events of 1066 is King Harold Godwinson’s brother Tostig, whom Harold had arranged to have exiled in the autumn of 1065 as a result of an uprising in Tostig’s earldom of Northumbria.18 Tostig proceeded actually to launch an invasion from Flanders in the spring of 1066, which proved unsuccessful, such that he had to take succor at the Scottish court before later joining King Harald of Norway in his invasion of Yorkshire in September. After an initial victory over English levies at Fulford Gate, just outside York, and while still in the full flush of their victory (in other words, drunk), their forces were surprised by Harold, who had marched up quickly from the south, attacking and massacring the Viking host, including his own brother. It was precisely because Harold was up north that William was able to land unopposed on the south coast; so it could be argued that it was a dispute between the two brothers, a family rift, that left England with a Norman king. Generally, however, family was supposed to help in creating a stronger strategic position; indeed, that was the whole point of dynastic politics. It was possible to see women as, in the marvelous Old English phrase, “peaceweavers” – people who brought an end to fighting between rival families – with inter-dynastic matches clearly being intended to create the possibility of dynastic alliances, or at the very least to keep a potential enemy at bay for a while.19 Marriages were also a statement of status and a declaration of intent. Bohemond, duke of Antioch, came from rather undistinguished stock, even if successful in carving out territories for itself; but when he sought a bride, he did not choose locally in Syria, or from families in Italy or Normandy, but rather a daughter of the king of France. Bohemond was making a statement to the effect that he had arrived in the top rank of international rulers. When his marriage was combined with an appeal to crusade, not against the Muslims who threatened the Christian enclaves in the East, but rather against the Byzantine Empire (whose people might at best be considered, as a justification for fighting them, the “wrong kind” of Christian), he was using both family connections and ideology to boost his attempt on the imperial throne (or, if nothing else, expansion of Norman lands across the Aegean). Of course, the very reason Bohemond was in a position to seize Antioch in 1098 was that he had already fought the Greeks, in support of his father, at Durazzo (on the Albanian coast) in 1081 and 1084.20 Meanwhile, his cousin, Roger de Hauteville had conquered Sicily, creating a rich kingdom at the heart of the Mediterranean and thus, strategically, crucial to almost all crusading efforts thereafter. The Sicilian fleet, which grew 18

19

20

The Life of King Edward who Rests at Westminster attributed to a Monk of St Bertin : Vitae Ædwardi Regis qui apud Westmonasterium requiescit S. Bertini monacho ascripta), ed. and trans. Frank Barlow (London, 1962) tells this story (whilst feigning disbelief), at pp. 52–53. Prof. Ann Williams gives credence to the claim (personal communication). See Ryan Lavelle, “Towards a Political Contextualization of Peacemaking and Peace Agreements in Anglo-Saxon England,” in Peace and Negotiation: Strategies for Coexistence in the Middle Ages and Renaissance ed. Diane Wolfthal (Turnhout, 2000) pp. 39–55. Bohemond’s pre-crusade history is best seen in Anna Komnene, The Alexiad, trans. E. R. A. Sewter, 2nd ed. (1969, repr. 2009), s.v.



Norman Conquests: World Domination Strategy?

97

to be of great significance in the twelfth century, provided a strategic insurance against Muslim counter-attack, and even sponsored its own attack on Alexandria in 1174 (and the Byzantine Empire in 1185).21 Horse and Arms The Normans’ initial attack on Sicily as far back as 1061 (although they did not subdue the island and its outliers such as Malta and Gozo for another thirty years) was launched in just a few ships. In fact there were so few that they had to return to pick up a second wave of warriors in order to bring them across the Straits of Messina; for while it was important for the Norman knights to cross over, even more crucial was that they should be accompanied by their warhorses. This required specially constructed vessels to load the animals on and off, which were scarce – hence the double journey.22 Perhaps the supreme image of the Norman knight is armored almost head-to-toe in a long mail coat, carrying his lance with its pennant blowing in the breeze, with his sword, the symbol of knighthood, strapped to his side, and both horse and man protected by a long, kite-shaped shield. He presents an impression of invulnerability and fearsome determination which must have struck terror into the hearts of opponents all over the known world. According to the daughter of a Byzantine emperor, a Frankish knight could bore his way through the walls of Babylon, so impetuous and unstoppable was his charge.23 The style of fighting developed in Western Europe from the Carolingian era onwards had proved enormously successful. At the battle of the Lech in Bavaria in 955, the armored horsemen of the German emperor Otto I had swept away the Magyar cavalry who raided Germany, France, and Italy virtually unopposed for two generations.24 Whether fighting for or against the Byzantines, such knights had proved their particular utility in battle. In Iberia, where for centuries the Muslims had dominated, from the mid-eleventh century onwards Christian armies were beginning to assert themselves and to sweep their opponents from the field. On crusade in the east, the Frankish method of war caught the Egyptians by such surprise at Ascalon in 1099 that their superior forces broke and ran.25 There was something special 21 22

23 24 25

Donald Matthew, The Norman Kingdom of Sicily, Cambridge Medieval Texts (Cambridge, 1992), pp. 277–79. Matthew Bennett, “Norman Naval Activity in the Mediterranean c. 1060–1108,” Anglo-Norman Studies 15 (1993), 41–58; idem, “Amphibious Operations from the Norman Conquest to the Crusades c. 1050-c. 1250,” in Amphibious Warfare 1000–1700: Commerce, State Formation and European Expansion, ed. Mark Fissel and David Trim (Leiden, 2005), pp. 51–68, drawing on the work of J. H. Pryor. Komnene, Alexiad, p. 416. Charles Bowlus, The Battle of Lechfeld and its Aftermath, August 955 (Aldershot, 2006), is the most recent and detailed study of the campaign and battle. Gesta Francorum, ed. Hill, pp. 92–97. John France, Victory in the East. A Military History of the First Crusade (Cambridge, 1994), provides the best modern account, pp. 361–65, with two maps, p. 362.

98

Matthew Bennett

about the Western knight that made the effort to transport them to the scene of the action worth a commander’s while. Of course, while an ordinary knight might own only one warhorse (together with some baggage mounts), richer lords had several, all of which required feeding and watering on campaign and made demands upon the logistic support for expeditions. It was noticeable that the crossings of Asia Minor in 1097–98, 1101, and 1147 all resulted in very heavy loss of animals, and hence of the tactical effectiveness of armies involved.26 The situation only improved when transportation by sea became preferred for the movement of the valuable warhorses, which meant that although, inevitably, smaller numbers were sent east, they generally arrived in better condition for use on campaign and in battle.27 (In claiming that the Western knight could play a dominant role in combat I am conscious that I am moving from the strategic to the tactical sphere; but believe this to be justified in so far as the delivery of “effect” to the operational area was a key strategic consideration.) Wood and Stone The final topic which I wish to examine concerns especially the use of fleets and fortifications. As we shall see, these had a “symbiotic” relationship, particularly in a coastal scenario such as that of the Latin states in the east, but also in a wider perspective. First to consider, of course, is the invasion fleet of 1066. It was only by this means that Duke William was able to deploy his forces rapidly across the English Channel in September of that year. There has been a great deal written about the size of the fleet and its construction during the spring, but most of this is speculation. There is also a Ship List of dubious value, which suggests that William was able to raise some seven hundred ships from his main tenants.28 It is certainly very unlikely that the whole fleet was built from scratch, as there would have been insufficient time, and although some ships may have been built as a result of Duke William’s urgings it is more likely that, as at other times during the era, vessels were acquired through rights of lordship or bought in for the campaign. William’s father-in-law, as count of Flanders, could have helped in this latter respect. The story of the movements of the fleet in 1066 is instructive because it demonstrates that naval activity was fraught with danger. When William collected his fleet in the estuary of the river Dives, a rare large anchorage on the Norman coast, his most likely target would have been the excellent natural harbors on the southern coast of England lying behind the Isle of Wight. A landing here would also allow him to advance on Winchester, the ancient capital of the English kings, and from 26 27 28

Gesta Francorum, ed. Hill, p. 27. Odo of Deuil, De profectione Ludovici VII in Orientem, ed. and trans. V. G. Berry (New York, 1948). John H. Pryor, “Transportation of Horses during the Era of the Crusades, Eighth Century to 1285, Part I: to c. 1285,” The Mariner’s Mirror 70 (1984), 9–27. Bennett, “Amphibious Operations,” pp. 51–68; Elisabeth M. C. van Houts, “The Ship List of William the Conqueror,” Anglo-Norman Studies 10 (1988), 159–83.



Norman Conquests: World Domination Strategy?

99

there by a good road to London; but while the English forces on land and sea remained in being even this was an optimistic option. By waiting until 8 September, when King Harold was forced to stand down his forces (due to the expiry of service obligations and the requirements of the harvest), William had the opportunity to attack an undefended shore. When he set out a few days later, however, a seasonal storm scattered his fleet and drove it northwest into another large estuary, the bay of the Somme. Here it took several days to reorganize the fleet, and doubtless some ships had been lost; although it had been worse in his father’s time, for an entire fleet intended to support Edward the Confessor’s claim to the English throne had then been entirely scattered. One of the myths of 1066 is that William was forced to wait in Normandy throughout the summer because of contrary winds. This is most unlikely to be true, as the prevailing winds in the English Channel are southwesterlies, ideal for sailing northwards from Normandy. It is possible nevertheless that northerly winds may have kept the Normans penned-up in Saint-Valery for ten days or so.29 This may explain the romantic tale that the town’s eponymous saint changed the wind direction after the Normans had paraded his relics (and covered them in gold oblations). I assume that when William set out again on the night of 21 September he had changed his target: for, unless the winds were very favorable, it would then have proved difficult to sail westwards to the Isle of Wight, and a different landing place was required. Dover, with its powerful hill-top fortification (which had been one place Harold had promised to hold for William), lacked a harbor. This is probably why William chose to land at Pevensey, which boasted a wide lagoon into which his shallow draft ships could be beached, and also a peninsular fortress, based on an old Roman fort, to protect the anchorage. Indeed, there were no harbors on the south coast of England that could hold a large fleet. The move to Hastings, which did have a harbor and also led to routes inland, was accompanied by the construction of a castle there. Hastings subsequently became the base for the initial phase of the Conquest; but not for long, because once the English had been defeated at Hastings, William needed to move on other strategic objectives: first to Dover, then Canterbury, seat of the archbishop, and finally, via a circuitous route, to London. The same storm which disrupted William’s crossing also scattered the English fleet(s) which had been making their way back eastward to London via the east Kent ports and anchorages.30 These vessels do not seem to have been reorganized in time either to prevent William’s crossing or attack Hastings, or to blockade the Norman supply lines, all of which might have been considered possibilities, had there still been a fleet in existence. 29

30

Christine Grainge and Gerald Grainge, “The Pevensey Expedition: Brilliantly Executed Plan or Near Disaster?,” The Mariner’s Mirror 79 (1993), 261–73, repr. in The Battle of Hastings, ed. Stephen Morillo (Woodbridge, 1996), pp. 129–42; personal communication from the late Marjorie Chibnall, and alluded to in Gesta Guillelmi, ed. Davis and Chibnall, Introduction, xxiv–xxvi. The Anglo-Saxon Chronicle: A Revised Translation, C, ed. Dorothy Whitelock with David C. Douglas and Susie I. Tucker (London, 1961), p. 142.

100

Matthew Bennett

In Italy, the Normans made similar use of fleets to achieve their strategic objectives.31 This was first because they were seeking to occupy a peninsula which contained several important ports; second, because they sought to control the rich island of Sicily; and third, because Robert Guiscard’s ambition drew him further east to attack Byzantine possessions in Albania, eastern Greece, and the Ionian. Once the Normans had gained control of the Italian ports they were also able to commandeer a rich source of naval vessels and experienced crews, with which they could take on and defeat both the Byzantines and the Arabs of Sicily and North Africa, so assuring their rule over a wealthy region along with the trade routes which passed through it. In the early twelfth century Norman colonies were even established on the coast of Africa, which shows how powerful the Sicilian fleet had made the dynasty that ruled the island.32 As regards the use of fleets on crusade, these came initially from the Italian maritime cities. The Genoans played a crucial role in 1099, during the siege of Jerusalem. It was their sailors, and indeed the timber from their ships, that helped to construct the siege engines used in the storming of the city that July.33 As early as the 1101 crusade there was an Anglo-Norman fleet engaged in supplying Antioch; but it was at the siege of Lisbon in 1147 that fleets from northern waters really proved their worth in crusading expeditions.34 What made the fleet particularly interesting on that occasion was that it was manned not by knights and nobles, but by ordinary merchants from the shores of the North Sea. This was in fact a natural development from the activities of traders who had been seeking Atlantic routes into the Mediterranean. The combination of a Portuguese king anxious to win a vital strategic fortress and the merchants’ greed proved to be a siege-winning one. In the event, owing to the strong currents of the Tagus (Tejo), the ships themselves played only a small role in the capture of the city, but the crews fought like demons on land to acquire the prize. King Afonso Henriques had been trying to advance his cause for years, but it was the strategic reach of the Anglo-Norman and German fleets that provided a crucial force-multiplier. While fleets proved important in transporting forces and indeed in storming coastal cities, it is also important to understand that they depended upon fortified bases in order to be most effective. In 1066, for example, the castle at Hastings with its garrison helped to secure the cross-Channel lines of supply which the ships provided. The fortified ports of the much richer Italian and Sicilian shores did the same. In the Holy Land, the coastline was secured by walled

31 32 33 34

Bennett, “Norman Naval Activity.” David Abulafia, “The Norman Kingdom of Africa,” Anglo-Norman Studies 7 (1985), 26–49; H. E. J. Cowdrey, “The Mahdia Campaign of 1087,” English Historical Review 92 (1977), 1–29. Randall Rogers, Latin Siege Warfare in the Twelfth Century (Oxford, 1992). The Conquest of Lisbon : De expugnatione Lyxbonensi, trans. Charles Wendell David, 2nd ed. with foreword and bibliography by Jonathan Phillips (1936, repr. New York, 2001); Matthew Bennett, “Military Aspects of the Conquest of Lisbon, 1147,” in The Second Crusade: Scope and Consequences, ed. Jonathan Phillips and Michael Hoch (Manchester, 2001), pp. 71–89.



Norman Conquests: World Domination Strategy?

101

coastal cities, affording fleets protected harbors which were in turn supplied by those same fleets from the sea, and so could not be starved into submission,.35 The dominance which crusader fleets established over those of Muslim Egypt from the 1120s onwards helped to gain and secure the coastal fortresses even after Saladin’s conquest of the crusader states in 1187. Here I am moving away from purely Norman concerns; but as I mentioned at the outset, my purpose in invoking the Normans was as a template for a strategic model employed by the Franks in general. In this I am not original, for the model was established in a seminal article by Robert Bartlett in Annales in 1974.36 Why, however, do I go on to propose that the Normans, or other westerners for that matter, had a strategy for world domination? I make this claim because it was the combination of vision, ideology, family, and diplomacy with the physical elements of men and their ships and fortresses which enabled relatively small groups of warriors to conquer territories and set up ruling dynasties in a truly remarkable way from around 1050 onwards. This early “Rise of the West” had profound consequences for later history, even when, as in the case the Holy Land, it proved impossible for the westerners to retain their lands in the face of a reorganized Muslim world. It was, of course, a very similar impetus that fueled the movement in the Iberian Peninsula which some call the Reconquista. It was the essence of medieval strategy, and it worked.

35 36

John H. Pryor, Geography, Technology, and War: Studies in the Maritime History of the Mediterranean, 649–1571 (Cambridge, 1988). Robert Bartlett, “Technique militaire et pouvoir politique, 900–1300,” Annales 41 (1986), 113–59.

5 The Papacy and the Political Consolidation of the Catalan Counties, c.1060–1100: A Case Study in Political Strategy Luis García-Guijarro Ramos

The relationship between the Catalan counties and the papacy was intense from the time when Sunyer, monk of the abbey of Cuixà in the Pyrenean region of Conflent, traveled to Rome in 950 to seek papal protection for his monastery.1 At the beginning of the preceding century, the Carolingian empire had organized into several dependent administrative units the Roman-Visigothic population of the northeastern part of Iberia which had remained free from Muslim power or, in many cases, had regained freedom from it as a result of Frankish campaigns which brought Christian control over the cities of Girona in 785, and Barcelona in 801.2 The restructuring of the region affected also some less thoroughly romanized groups which had originally dwelt deep in the Pyrenees and whose members were called hispani by contemporary Carolingian sources.3 Once the western empire entered an irreversible decline as an effective continental power in the second half of the ninth century, these borderline districts facing al-Andalus became political units autonomous to all intents and purposes from increasingly powerless emperors – a development which was shared by many of their counterparts in other parts of the Carolingian world.4 The Catalan 1

2

3

4

Petrus de Marca, Marca Hispanica sive limes Hispanicus […] (Paris, 1688, repr. Barcelona, 1972), nos. 164 and 165, cols. 978–87; Ramon d’Abadal i de Vinyals, “Com neix i com creix un gran monestir pirinenc abans de l’any mil: Eixalada-Cuixà,” Analecta Montserratensia 8 (1954–55), 125–337, no. 73, pp. 292–93. Conflent was a dependency of the county of Cerdanya. Josep M. Salrach, El procés de formació nacional de Catalunya (segles VIII–IX). 1: El domini carolingi (Barcelona, 1978), pp. 10–11 and 14–26; idem, Història de Catalunya dirigida per Pierre Vilar. 2: El procés de feudatlizació (segles III–XII) (Barcelona, 1987), pp. 129 and 135–38. Abilio Barbero, “La integración social de los ‘hispani’ del Pirineo oriental al reino carolingio,” in Melanges offerts à René Crozet, 1, ed. Pierre Gallais and Yves-Jean Rivu (Poitiers, 1966), pp. 67–75; Abilio Barbero and Marcelo Vigil, La formación del feudalismo en la Península Ibérica (Barcelona, 1978), pp. 354–57. For criticisms of the historiographical idea of a weakening of Frankish royal power in the second half of the ninth century, see the recent overview on the Carolingians down to 888: Marios Costambeys, Matthew Innes, and Simon MacLean, The Carolingian World (Cambridge, 2011), pp. 314 and 316. Classic studies of the Carolingian world going beyond that time are Rosamond McKitterick, The Frankish Kingdoms under the Carolingians, 751–987 (London, 1983), and Pierre Riché, Les Carolingiens: une famille qui fit l’Europe (Paris, 1983).

104

Luis García-Guijarro Ramos

counties followed the course from dependent administrative centers into fullyfledged self-ruling political structures, as did other territories in the empire.5 This complex and multifaceted process underlay their history from the late ninth to the mid-eleventh centuries. It basically meant the gradual strengthening of so far incipient feudal relations at the bottom and at the top of society, at the rural base and among high political circles. Modern historiography on these eastern Iberian counties at that time has concentrated mainly on the transformation, or to be more precise on the “mutation,” of allegedly “free” peasants into serfs which had taken place in the latter stages of the period; this thesis implies the absence of any previous feudal features in social relations and a tacit agreement on the ascription of Carolingian and early post-Carolingian southern Pyrenean counties to “Late Antiquity.”6 The line of thought shared by this type of study has clearly followed the path opened by the seminal but rather schematic, and somewhat misleading, book by Pierre Bonnassie.7 Much research has been devoted to the mutation féodale, and comparatively little to political construction of nascent feudal powers in the area, on the assumption that the thorough works of Paul Kehr, Ramon d’Abadal, and Archibald Lewis had settled the question of the lack of an undisputed postCarolingian auctoritas which could provide a firm basis for comital ambitions to build strong polities in their respective zones.8 These historians selected the basic evidence and identified the main legitimacy problem which the counts had to face when the Carolingian political reference point ultimately disappeared. However, constrained by the intellectual framework they worked within, of clearly positivist or empiricist overtones, none of these three scholars went beyond a sound description of the landmarks which announced the search for and development of fluid contacts with the papacy. Nor did they sufficiently dwell on the basic fact that the counts could act in no other way, because popes represented the only universalist authority left in the Latin Christian world, alone able to offer solid grounds on which to base legitimacy for their secular powers in the making. Certain of the monasteries located in comital territories were of paramount importance in connecting the Catalan counts with the popes, mainly in the period c. 950–c. 1050. Monastic exemption, which tied abbeys directly to the 5

6 7 8

On the origins and development of principalities in Francia occidentalis the reference is still Jan Dhont, Études sur la naissance des principautés territoriales en France, IXe–Xe siècle (Bruges, 1948). This is the generally accepted view since the 1980s and it is best expressed in Salrach’s contribution to Pierre Vilar’s history of Catalonia (see note 2 above). Pierre Bonnassie, La Catalogne du milieu du Xe à la fin du XIe siècle. Croissance et mutation d’une société, 2 vols. (Toulouse, 1975–6). Paul Kehr, Das Papsttum und der Katalanische Prinzipat bis zur Vereinigung mit Aragon (Berlin, 1926, Catalan translation by Ramon d’Abadal i Vinyals, Barcelona, 1931); Ramon d’Abadal i Vinyals, Els Primers Comtes Catalans (Barcelona, 1958); idem, Dels Visigots als Catalans, 2 vols. (Barcelona, 1969); Archibald Lewis, The Development of Southern French and Catalan Society, 718–1050 (Austin, 1965).



The Papacy and the Catalan Counties

105

papacy while restricting the intervention in their affairs of bishops whose sees were occasionally located in foreign lands and who might thus be inclined to foster alien secular interests, became a useful method for reinforcing the political power of counts over their districts in northeastern Iberia. Abbeys fulfilled there a double role, at the two extremes of society: they brought comital territories together by actively expanding bonds of dependence at the bottom of the social hierarchy, and at the same time, by means of the monastic exemption which afforded them an especially privileged connection to the pope, they indirectly provided local rulers with the auctoritas they most eagerly sought, namely the universal authority of the Latin Church. The counties of Cerdanya and Besalú offer the clearest evidence of this state of affairs in the first two decades of the eleventh century. Oliba, abbot of Ripoll and of Cuixà from 1008, and bishop of Vic ten years later down to his death in 1046, along with his brothers the count of Cerdanya Guifred II (988/994–1035) and the count of Besalú Bernard I (988/994–1020), developed a joint policy which used the prestige of Roman ecclesiastical authority to strengthen the power of their lineage in both districts, by way of monastic exemptions or of the “peace councils” which met in the region at that time and curbed violence potentially disruptive to the counts’ power.9 This strategy also shows that the profits drawn from this type of relationship were meanwhile one-sided. The papacy had the auctoritas that the Catalan counts were in need of, but not enough potestas to benefit from a relationship with secular powers. The popes seemed to have the upper hand in those contacts, but in practice they did not.10 Institutional historians have touched in different ways upon this fundamental contribution to the rise of comital polities in northeastern Iberia. Paul Kehr reckoned that the counts and not the papacy held the initiative, but did not press this point to its logical conclusion in terms of the crucial relevance of that connection to the political formation of those counties.11 Ramon d’Abadal declared that the travels to Rome of counts and ecclesiastics had no political purpose down to 970, but that after that date they were “fatally” stimulated by such motives.12 His traditional perception, of an ideal dichotomy between religion and politics which gave way to a pernicious mingling of the two from the late tenth century, offers little help in understanding the rise of counties as distinct political entities. This picture changed from the middle of the eleventh century. The so-called Gregorian Reform aimed precisely at balancing the undisputedly potent ecclesiastical auctoritas with a similarly strong potestas, tending to over-

9

10 11 12

Luis García-Guijarro, “Los monasterios catalanes y la exención en época pregregoriana: los casos de Ripoll y de Cuixà,” in Medievalis Historia Pyrenaica, II Congrés Internacional Història dels Pirineus, ed. Julià Maroto, Sònia Ramió, and Eduard Ripoll (Girona, 2005), pp. 255–69. Thomas Deswarte thinks otherwise: “Rome et la specificité catalane. La papauté et ses relations avec la Catalogne et Narbonne (850–1030),” Revue historique 294 (1995) 3–43. Kehr, Das Papsttum, pp. 7 and 17. R. d’Abadal, Els Primers Comtes, p. 306.

106

Luis García-Guijarro Ramos

flow from the spiritual realm of the Church to invade the secular field; this new ecclesiastical approach is most clearly perceived in Gregory VII’s set of propositions known as Dictatus Papae and dated to 1075.13 When these new principles began to operate, the relationship of the reformist popes with the Catalan counties kept its previous forms, but its content was altered considerably. While some historians have seen continuity between pre-Gregorian and Gregorian times, there were in fact deep changes.14 From the 1060s, the secular powers were not the only ones engaged in a struggle to achieve political control and stability: the Roman see was also much involved and working vigorously along similar lines to enhance its status as a super-power. That meant collaboration, and at times competition and confrontation, with kingdoms and counties. It remains true that the balance was usually tilted in favor of the latter, which had real localized foundations to work on and continued to use papal authority to strengthen the power of nascent polities, while the Roman see had the will but not the ability to widen its potestas to fulfill the over-ambitious reformist aims it had set itself; nonetheless, the entirely new historical setting from the middle of the eleventh century implies that any event in or written evidence from eastern Iberia at that time has to be placed against a complex background of secular and spiritual powers, each of them striving to achieve specific objectives which often collided and led to silent or to open confrontation. The prominent new role of the Roman see is attested by the frequent contacts of the reformist popes with the Catalan counts in the second half of the eleventh century and beyond. An “entente cordiale” operated between the two powers as long as both parties considered that their collaboration was mutually beneficial in terms of their power strategies; this was also the case as regards apostolic relations with the Aragonese kings, although in the new kingdom a brief conflict arose in the late 1070s.15 Three specific instances, ranging from the late 1060s to the 1090s, show the complementary policies of the popes and many of the Catalan counts at that time. The last example bears upon the first attempt to conquer Tarragona and to restore its metropolitan see, which had passed to Narbonne after the Islamic conquest of the former city in the early eighth century and the recapture of the latter by the Franks in 759. In this particular case, the dual and complex meaning of Urban II’s letter of 1 July 1089 to Count Berenguer Ramon II, which might be considered (wrongly) a crusade bull years before Clermont, but was in fact a papal answer to the count’s appeal focused only on apostolic interests, should warn us against literal, deterministic, and exclusively Roman-directed interpretations of this type of pre-conquest bull in 13

14 15

Das Register Gregors VII, ed. Erich Caspar, 1 (Berlin, 1955), 2.55a, pp. 201–8; The Register of Pope Gregory VII, 1073–1085: An English Translation, ed. H. E. John Cowdrey (Oxford, 2002), 2.55a, pp. 149–50. H. E. John Cowdrey, The Cluniacs and the Gregorian Reform (Oxford, 1970), p. 220. Luis García-Guijarro Ramos, “El papado y el reino de Aragón en la segunda mitad del siglo XI,” Aragón en la Edad Media 18 (2004), 245–64 at 257–58.



The Papacy and the Catalan Counties

107

eastern Iberia in the following centuries, be it the case of Tortosa in 1148, of Mallorca in 1228, or of Valencia in 1236.16 The two previous instances in which first the count of Barcelona and then those of Besalú and Cerdanya were involved are even more revealing of the nature of the relationships between counts and popes at the time. By the mideleventh century, the county of Barcelona, which had included Girona and Vic since late Carolingian times, was already the most powerful political unit in what would later be called Catalonia. The 1040s and 1050s were difficult times for its count, Ramon Berenguer I the Old (1035–76).17 He had to face dangerous rebellions stemming from disputes with distant or closely-related members of his own comital dynasty. A remote relative, Mir Gelabert, grandson of count Borrell II (947–92), questioned Ramon Berenguer’s ultimate control of certain southern territories from 1041 to 1059, and even awarded himself the title of princeps of Olérdola. Serious conflicts with the bishop of Barcelona, also related to Borrell II, and with his nephew, Viscount Udalardo, arose as well. Ramon Berenguer’s grandmother, the countess Ermessenda, sternly opposed her grandson and managed in 1056 to get him briefly excommunicated by Pope Victor II, who had already sanctioned him in that way, along with his third wife Almodis, for the repudiation of his previous spouse, Blanca.18 These difficulties were on the brink of seriously weakening comital power. Once they were finally dealt with successfully in the years 1057–59, Ramon Berenguer aimed at tightening his grip over the feudal hierarchy finally emerging from the conflicts which had torn the county during two decades.19 The fact that other Catalan counts became vassals of Ramon Berenguer, as in the cases of Ermengol III of Urgell in 1063 and Pons I of Empùries in 1067, shows clearly that strife with his own nobility had not hampered the count’s capacity for extending his direct 16

17 18 19

For the bull attributed to the Tortosa campaign, see Luis García-Guijarro, “Reconquest and the Second Crusade in Eastern Iberia: The Christian Expansion in the Lower Ebro Valley,” in The Second Crusade. Holy War on the Periphery of Latin Christendom, ed. Jason T. Roche and Janus Møller Jensen (Turnhout, 2015), pp. 219–55 at 234–39. For the bulls related to the conquests of Mallorca and Valencia, see Luis García-Guijarro Ramos, “Guerra y religión en el contexto ibérico del siglo XIII: una mirada sobre el reinado de Jaime I,” in Jaume I: commemoració del VIII centenari del naixement de Jaume I, ed. María Teresa Ferrer i Mallol, 2 (Barcelona, 2013), pp. 317–38 at 326–32; also idem, “Reconquista and Crusade in the Central Middle Ages: A Conceptual and Historiographical Survey,” in Crusading on the Edge: Ideas and Practice of Crusading in Iberia and the Baltic Region, 1100–1500, ed. Torben Kjersgaard Nielsen and Iben Fonnesberg-Schmidt (Turnhout, 2016), pp. 55–88 at 86–87. A short overview is in Santiago Sobrequés, Els Grans Comtes de Barcelona (Barcelona, 1980), pp. 39–56. Sobrequés, Els Grans Comtes, p. 46. In 1057, Ermessenda gave way to her grandson, assuring him that she would urge the pope to suppress the excommunications of the count and Almodis: Liber Feudorum Maior, ed. Francisco Miquel Rosell, 2 vols. (Barcelona, 1945), 1, no. 215, pp. 223–26. In 1059, a sentence of a court of justice, specifically set up to deal with the case, put an end to Mir Gerebert’s long and bitter dispute with the count of Barcelona; the viscount had disregarded the previous sentence of 1052, but this time he abided by the decision of the court: Sobrequés, Els Grans Comtes, pp. 54–55.

108

Luis García-Guijarro Ramos

political dominance over nearby territories.20 From the time of Alexander II, the papacy played a key role in reinforcing this internal and external pre-eminence and also in protecting comital power in the difficult years after Ramon Berenguer I’s death in 1076. One of the first instances of mutually beneficial collaboration between this secular power and the apostolic see was the council of Girona of November 1068. Ramon Berenguer I and his wife Almodis attended that synod, convoked by the pope at their request and presided over by the papal legate Cardinal Hugh Candidus.21 Their presence was thus not just a matter of protocol; it indicated that the ecclesiastical assembly addressed points which affected both the Church’s and the count’s interests. Canon 11 of that council is perhaps the best example of how measures taken in favor of the former also benefited the secular power: it stipulated that ecclesiastical property enjoyed by laymen in usufruct should return to the Church when the donor or the recipient died, because “worldly goods devoted to God should not be in the hands of secular men through inheritance.”22 By this ruling the pope, through his legate Hugh Candidus, tried to avoid lands and premises which had been bestowed by high clerics upon laymen in the past passing into hereditary possession and so slipping out of clerical control. This decision of the synod ran counter to the tendency of tenentes to consolidate their holdings; it was favored by the count of Barcelona precisely on that account. He did not wish his own dependents to consolidate their grip on beneficia granted to them by himself or by his predecessors; his bitter struggle with a section of the nobility, which had only ended less than a decade earlier, was a potent reminder of the importance of indirectly reaffirming his own supreme power, albeit at a regional council apparently dealing only with the reformist agenda of the Roman Church. Ecclesiastical and secular potestates over vassals relied on the control they could exert over grants, which were the physical expression of the social links of personal dependence which bound societies together whether at a local or at a universal level. The pope and the count thus strove hard toward the same goals. The “Peace and Truce of God” decreed in 1068 by the council of Girona, and somewhat later by an assembly which took place at Vic, provide supplementary evidence of the convergent policies of count and pope.23 Pacifying decrees had been common in southern France and Catalonia since the end of the tenth century, and had allowed the Church to extend its influence into the secular world thereafter, as well as permitting some regional secular powers to strengthen grip

20 21

22 23

Liber Feudorum Maior, ed. Rosell, 1, no. 150, pp. 150–51; 2, no. 519, pp. 32–33; no. 837, p. 320. “sub praesentia domni Remundi, Barchinonensis Comitis et domne Almodis Comitissae quorum cura et instancia haec Sinodus congregata est,” Jaime Villanueva, Viage literario a las iglesias de España, 13 (Madrid, 1850), no. 25, p. 262. “quia res Deo dicatae non debent a laicis ereditario jure possideri,” Villanueva, Viage literario, 13, no. 25, p. 263. Villanueva, Viage literario, 13, no. 25, p. 264. For the peace decrees of Vic, see P. de Marca, Marca Hispanica, no. 269, col. 1141.



The Papacy and the Catalan Counties

109

on their lands – Bishop Oliba’s assemblies have already been mentioned as exemplifying the double function of such meetings in pre-Gregorian times.24 This trait was even more perceptible after the 1050s, at a time when the reform movement aimed at expanding the pope’s potestas. In the case of Ramon Berenguer I, nothing could have been more profitable to his purposes than to check unrest amongst his higher noble vassals through propositions sanctioned by local synods. Having just managed to control rebellious nobles, he saw in these ecclesiastical measures a way to increase his power, which had been on the point of crumbling only two decades previously. The decrees approved at Vic, shortly after the council of Girona, included very precise details of the responsibilities of fathers or superiors with regard to violence caused by their sons or dependents: the former would be severely punished with excommunication should they fail to apply the treatment specified for those under their command who had broken the peace.25 This obviously represented an ecclesiastical intrusion into secular life, but the interference was welcomed by Ramon Berenguer I as a means to curb the aggression of sectors of the nobility. A new set of peace decrees, also agreed upon at Vic and dated to the mid1070s,26 offers the most astonishing evidence of the extension of spiritual potestas to the pax christiana, in the form of remissio peccatorum: “To whoever has kept and observed the above-mentioned peace and the above-mentioned pact with the Lord, which we call truce, the God of peace and of all consolation will be with him, and will offer him the grace of His benediction, and for this observance he will be able to receive remission for all his sins.”27 Well before Urban II’s momentous address of November 1095 at Clermont, the apostolic see was stretching to the utmost his spiritual potestas by extending it to Christian peace as well as to Christian warfare, which together formed an indivisible whole in reformist ecclesiastical thinking and subsequently in crusading ideology and doctrine. Thus it is not surprising that, about fifteen years after these decrees had been issued at Vic, Urban II offered remissio peccatorum to those participating in the Tarragona campaign. These and other episodes make us wonder whether the importance attached to Clermont as a turning point in the papal view with respect to a certain type of warfare and to the spiritual rewards attached to it is not yet another historiographical construct. This type of policy, common to the apostolic see and the Catalan secular powers, was not limited to the county of Barcelona. One of the most interesting instances of how Roman reformist endeavors mingled with the political aims

24 25 26 27

A classic short study on the Peace and Truce of God is H. E. J. Cowdrey, “The Peace and the Truce of God in the Eleventh Century,” Past and Present 46 (1970), 42–67. “Si filius faciat malum, an pater teneatur,” P. de Marca, Marca Hispanica, no. 269, col. 1140. Villanueva, Viage literario, 6 (Valencia, 1821), pp. 211–12. “Quicumque pacem praedictam et pactum Domini praescriptum, quam treguam vocamus, bene custodierit et servaverit, Deus pacis et totius consolationis cum illo erit, et gratiam suae benedictionis illi praebebit, et per hanc observationem remissionem omnium peccatorum suorum habere poterit,” Villanueva, Viage literario, 6, no. 36, pp. 322–23.

110

Luis García-Guijarro Ramos

of the Catalan counts is presented by the events that took place at Girona and Besalú in between December 1077 and March 1078. The problems that the counts of Cerdanya and Besalú had to face to reaffirm their political power were different from those prevailing in the more powerful county of Barcelona. The latter had three bishoprics (Barcelona, Girona, and Vic), Cerdanya and Besalú none. The absence of a diocesan who could offer ecclesiastical and political coherence to territories in need of it marked the history of these two counties in the eleventh century. Several schemes were attempted to tackle the problem. The most lasting was the simoniacal control of the metropolitan see of Narbonne by a member of the comital dynasty ruling Cerdanya and Besalú: he himself or close relations forced the appointment of other relatives to bishoprics – Elna, Girona, or Urgell – whose jurisdiction extended over lands included in either of the family’s two counties, but whose sees were based in foreign polities. This dynastic policy, which continued for decades, was so noteworthy and consistent that it has come to symbolize the condition of the pre-Gregorian Church in some general views of the Middle Ages.28 Simony has always had a very bad press among historians, mainly because many of them have adhered to Roman reformists’ negative view of the selling of Church dignities and against the supposed interference of secular rulers in spiritual matters.29 What was at stake was the control of high ecclesiastical posts, which determined in its turn political grip over certain territories, rather than strictly moral issues. For this reason, historiographical positions – I have explicitly mentioned Ramon d’Abadal – which regard the blend of religion and politics as an evil to be uprooted are unhistorical if projected to a time when Christianity was the ideology which bound tight all polities in Latin Christendom. For most of the eleventh century, the archbishopric of Narbonne was in the hands of Guifred, son of the count of Cerdanya Guifred II, who had bought the dignity for his still infant son in 1016 for the considerable sum of 100,000 solidi; three years later he was consecrated.30 When in power at Narbonne, the metropolitan reproduced the system by purchasing the bishopric of Urgell for his brother William in 1041.31 The family also attained control over the sees of Elna in 1035, and Girona in 1051;32 in the 28

29

30 31 32

“If we wish to catch a glimpse of the old church life of Europe before it was transformed by the zeal of the late eleventh century, we cannot do better than turn to the counts of Catalonia.” Richard W. Southern, The Making of the Middle Ages (London, 1953, repr. 1975), p. 115; for his detailed treatment of Cerdanya and Besalú at that time, pp. 117–19. The volume (7) devoted to the pre-Gregorian period in Histoire de l’Église depuis les origines jusqu’à nos jours, ed. Augustin Fliche and Victor Martin, has a most significant title: Émile Amann and Auguste Dumas, L’Église au pouvoir des laïques (888–1057) (Paris, 1942). In relation to Catalonia, see, for instance, Lewis, Development, pp. 317–18. Claude Devic and Joseph Vaissète, Histoire générale de Languedoc, 5 (Toulouse, 1875), no. 251-CCXXI, col. 497. Devic and Vaissète, Histoire générale, 5, no. 251-CCXXI, col. 498. Villanueva, Viage literario, 13, pp. 102–4; Antonio S. Ruiz, “Bérenguer (Wilfrid), évêque de Gironne,” Dictionnaire d’histoire et géographie ecclésiastiques, 8 (Paris, 1935), cols. 369–70.



The Papacy and the Catalan Counties

111

years 993–1003 the former dignity had been held by Berenguer of CerdanyaBesalú, with the influence of his brothers Guifred (the future archbishop of Narbonne) and Bernard and Oliba (who ruled Cerdanya and Besalú).33 A revealing account of these policies may be found in the exposé made by Berenguer, viscount of Narbonne, at a council in Toulouse in 1056:34 he was a relative of the archbishop Guifred, but probably resented the latter’s wide powers in the region, which clearly extended well beyond spiritual affairs.35 In the course of Guifred’s long government over the see he was excommunicated several times. Neither he nor his close relatives the counts of Cerdanya and Besalú paid much attention to that severe papal punishment. The metropolitan was fulfilling a political role for the sake of the lineage, behavior that was not in opposition to his religious duties in his own eyes or in the eyes of the dynasty to which he belonged. In the late 1060s the situation began to change, however. The new counts Bernard II of Besalú (1066–85) and William I of Cerdanya (1068–95) began to consider other alternatives which could be more effective than simony to achieve the goal of tighter political control in the face of an already overpowering county of Barcelona. The events of 1077–78 have to be looked at with this change of policy in mind. In December 1077 the papal legate Amatus, bishop of Oloron, arranged for a synod to be held at Girona, the main object of the meeting being to condemn once again the metropolitan Guifred and all the diocesans who followed him. The archbishop’s party caused such uproar that the assembly had to be cancelled, and Bernard II offered Besalú as the new meeting place.36 But at that precise moment he and his cousin William I of Cerdanya overtly changed policies and abandoned their uncle, the metropolitan, to his fate. They both realized that the old methods were no longer feasible at a time when the Roman see was more than ready to press its influence and power. They saw a bigger political advantage in complying with the pope’s wishes than in following an autonomous ecclesiastical policy. By becoming a peculiaris miles of Saint Peter, Bernat II of Besalú showed the extent and depth of the move from simony to strict allegiance to Rome and papal protection.37 The canons hastily approved at Besalú, which were hard on all simoniacal dignitaries, were later confirmed at a new council which took place at Girona in March 1078.38 33 34 35 36 37

38

Lewis, Development, p. 349. Bishop Berenguer died while fighting against the Muslims at Albesa in the summer of 1003: R. d’Abadal, Dels Visigots, 2:174–75. G. D. Mansi, Sacrorum Conciliorum Nova et Amplissima Collectio, 31 vols. (Florence and Venice, 1759-98), 19:col. 850. Devic and Vaissète, Histoire générale, 5, no. 251-CCXXI, col. 496–502; they dated the querimonia to around the year 1159. Liber Feudorum Maior, ed. Rosell, 2, no. 501, pp. 16–17. “Preter hoc autem, ut sanctus Petrus me habeat peculiarem militem, censum sibi mee milicie constituo singulis annis C mancusos aureos, et ut meus filius idem faciat post mortem meam vel quicumque meum honorem habuerit,” Liber Feudorum Maior, ed. Rosell, 2, no. 501, p. 17. Villanueva, Viage literario, 13, p. 110, and no. 36, pp. 264–66.

112

Luis García-Guijarro Ramos

All these assemblies were in theory strictly ecclesiastical and dealt with questions which at first sight concerned only the Roman Church. However, at Girona and Vic in the late 1060s, and again at Girona and Besalú in the late 1070s, serious political issues were also at stake. At a time when the Church was striving for power, as kings and princes too were doing, councils were not simply religious assemblies (nor had they ever been, in fact, since Nicaea in 325). Christianity being the ideology which firmly integrated the Latin West, the institution which symbolized that ideology, the Church, could not remain confined to its religious function. It was in itself a political protagonist which interacted with the secular powers. The policies common to Rome and the Catalan counties were also at the very core of the last topic that I am going to examine briefly: the conquest of Tarragona and the restoration of its late Roman and Visigothic metropolitan see. This was planned and achieved in the last decade of the eleventh century, although Christian control and consequently the recovery of the city’s ancient ecclesiastical status was short-lived due to its prompt recapture by the Almoravids, and did not become permanent until the 1120s. The campaign in the early 1090s led to a temporary territorial expansion through war encouraged by the papacy, and might thus be considered either part of the Reconquista or, were we to see it in terms of a kind of retrospective “pluralism,” as a crusade which had thus shown its traits some years before that meritorious type of warfare was officially proclaimed at Clermont. Some scholars who have dealt with this subject in the past have tended to regard the restoration of the archbishopric of Tarragona as a purely ecclesiastical affair. It was certainly not so: it had deep political undertones, as Lawrence McCrank has pointed out.39 I have already mentioned that Narbonne had become the supradiocesan center for the north- and southeastern zone of the Pyrenees in Carolingian times, after the old metropolis of Tarragona had been taken over by the Muslims in the early eighth century. The territory of Narbonne was not directly controlled by any of the leading Catalan comital dynasties, however, and the contradiction between secular and ecclesiastical boundaries, which was a serious handicap at diocesan level for political stability in Cerdanya and Besalú, became a problem at the superior archiepiscopal level common to all Catalan counts when, in the second half of the tenth century, the autonomy of the old Carolingian districts south of the Pyrenees began to turn them into solid centers of power. The counts of Barcelona were among the strongest supporters of the establishment of a metropolitan see south of the Pyrenees. The first attempt dated back to the late tenth century and was conceived by Count Borrell II (947–92). It aimed at transferring Tarragona’s ecclesiastical status to the diocese of Vic and at promoting its bishop, Aton, to that higher dignity. There were no prospects at that moment of conquering the old Roman-Visigothic city; the alternative was 39

Lawrence J. McCrank, “Restauración canónica e intento de reconquista de la sede tarraconense,” Cuadernos de Historia de España 61–62 (1977), 145–245. This article in Spanish is based on part of the author’s doctoral thesis in English, “Restoration and Reconquest in Medieval Catalonia: The Church and Principality of Tarragona, 971–1177” (University of Virginia, 1974).



The Papacy and the Catalan Counties

113

thus to upgrade Vic, which was then one of the most significant urban centers in the count’s territories. Borrell II, Aton, and a young Gerbert of Aurillac journeyed to Rome in 970 and received papal approval for their project.40 The life of the new archbishopric was very short, but the idea of establishing a supradiocesan center within the county of Barcelona did not die away. A hundred years later, the conquest of Tarragona and the restoration of its see were much more feasible. The Visigothic past underpinned the right of conquest and was a potent ideological motive, but the factor of greatest relevance was that the old city was located in the most obvious direction for the territorial expansion of the strongest county south of the Pyrenees. The social conditions which made expansion possible, as well as the set of ideas which fueled it, clearly associate this campaign with the Reconquista rather than identifying it as a “crusade before Clermont.” However, the occupation of Tarragona did not just mean territorial accretion. It also brought with it the possibility of reinstating the metropolitan see over the whole of northeastern Iberia. The restoration was in itself a potent tool to increase the political influence of the comital House of Barcelona over the area, as well as to put a brake on Leonese-Castilian designs upon the region at a time of internal weakness in the county due to the violent death of Count Ramon Berenguer II in 1082 and strong suspicions of instigating the murder which were cast upon his brother and co-ruler Berenguer Ramon II.41 The evidence that two bishops of Vic, Aton in the late tenth century and Berenguer Sunifred of Lluça a hundred years later, actively participated in the project of secession of the Catalan sees from Narbonne, and that they travelled to Rome to get papal approval for their schemes, cannot obscure the fact that both plans were carefully studied political moves. Both bishops were undoubtedly ambitious ecclesiastics, but their promotion to metropolitan status concerned the secular powers at least as much as clerical circles. The upgrading of Vic and the restoration of the Visigothic metropolitan see of Tarragona were so actively supported respectively by Count Borrell II in the tenth century and by the feeble count-regent Berenguer Ramon II in the late 1080s that both might be considered instigators of initiatives which, if realized, would elevate the standing of the county not only in the region but in the whole Christian world. In these and in the other cases that we have dealt with, ecclesiastical policies were tightly intertwined with the aspirations of the Catalan counts. It could not be otherwise at a time when the Church was playing such a decisive social and ideological role in the emergence and consolidation of polities in northeastern Iberia and elsewhere. The chronological sequence of papal and comital letters about the Tarragona campaign and its motives might suggest an entirely ecclesiastical lead on the question of the conquest of the city and the restoration of its metropolitan see. At the instance of the bishop of Vic, Berenguer Sunifred of Lluça, who had trav40

41

Kehr, Das Papsttum, pp. 13–15; R. d’Abadal, Els Primers Comtes, pp. 308–11. For Pope John XIII’s bull offering metropolitan status to Vic in January 971, see Demetrio Mansilla, La documentación pontificia hasta Inocencio III (965–1216) (Rome, 1955), no. 1, pp. 1–2. Sobrequés, Els Grans Comtes, pp. 107–12.

114

Luis García-Guijarro Ramos

elled to Rome to favor the cause, Urban II issued an introductory bull on 1 July 1089.42 A year later, and following the defeat of Berenguer Ramon II at Tevar by El Cid, the count submitted to the pope all his honores and the city of Tarragona when it should be conquered.43 The apostolic see gave its final approval to the project on 1 July 1091.44 Tarragona was probably semi-deserted at the time and of little interest to the Muslims, the center of whose petty kingdom – taifa – was further south at Tortosa. By mid-1092, the old Roman-Visigothic city had been taken over by the Christians, who kept control of it until the Almoravids raided the region in the summer of 1102, having in the spring of that year occupied Valencia. The latter city had been abandoned shortly before by El Cid’s widow Jimena and by Alfonso VI of León and Castile, who had arrived to secure the evacuation of its Christian inhabitants.45 My concern here is not to stress what might be regarded as a distinct crusading flavor to Urban II’s texts, so much as to focus upon papal interests, in particular the internal affairs of the county of Barcelona following the violent death in 1082 of Ramon Berenguer II, and upon wider Iberian politics at a time when the king of León-Castile Alfonso VI, “lord of the two religions” after the conquest of Toledo in 1085, had his sights set on other Muslim and Christian Iberian lands. Such a “macro-view” offers a wider perspective on the campaign and allows historians to perceive that the basic impulses underlying it and its timing were conditioned by a severe crisis in the county of Barcelona and in Iberian politics more generally. The papacy played an important but subsidiary role, comprising support for the count’s ambitions whilst not alienating Alfonso VI of León-Castile. In this and in other cases, popes seldom had the initiative. The Gregorian reform aimed at securing a dominant position for the Church of Rome in both spiritual and secular affairs. It never completely achieved that privileged position. On most occasions, the supreme authority of popes simply provided backing for kings or counts; the latter looked to the papacy for reinforcement, not for inspiration or leadership. This was precisely the case as regards the Tarragona campaign. On 15 October 1088, eight and a half months before the first text devoted to Tarragona, Urban II granted the archbishop of Toledo Bernard of Sedirac primacy over all Iberian sees.46 The bull was addressed to “[episcopis] Terraconensi et ceteris yspaniarum archiepiscopis”, according to the manuscript kept at Toledo,47 it thus indicated the apostolic interest in detaching the Catalan bish42 43

44 45 46 47

Mansilla, La documentación pontificia, no. 29, pp. 46–47. Le Liber censuum de l’Église romaine, ed. Paul Fabre and Louis Duchesne, 3 vols. (Paris, 1889–1952), 1, no. 216, p. 469; also in Demetrio Mansilla, La documentación pontificia de Honorio III (1216–1227) (Rome, 1965), no. 433, pp. 314–15. Mansilla, La Documentación pontificia no. 32, pp. 49–52. Bernard F. Reilly, The Contest of Christian and Muslim Spain, 1031–1157 (Cambridge, MA, 1992), p. 125. Mansilla, Documentación hasta Inocencio III, no. 27, pp. 43–45. McCrank, “Restauración canónica e intento de reconquista de la sede tarraconense, “ pp. 162–69



The Papacy and the Catalan Counties

115

oprics from the jurisdiction of Narbonne, probably because this metropolitan see was still obstructing the reformist program. The pope’s rescript reinstated the old Visigothic hierarchy (Toledo had received primatial jurisdiction in 681, as part of the Visigothic kings’ efforts to reinforce their powers and prevent disruptive tendencies). This move was probably prompted by King Alfonso VI of León-Castile. The recovery of the old Toledan ecclesiastical privileges, made possible by the conquest of that city in 1085, went beyond the limited field of supra-diocesan rights. Primacy allowed the king to make his presence felt in the Catalan counties by attempting to press the final jurisdiction of Toledo in an area where the metropolitan authority of Narbonne was openly questioned. Alfonso VI was specifically interested in the county of Barcelona, which was going through critical times; his policy was directed toward having a say in the regency which followed Ramon Berenguer II’s violent death in 1082. The motives of the Catalan count-regent and his entourage of nobles, as well as those of other Catalan counts, were manifold. Berenguer Ramon II’s position was very weak after his brother’s death in 1082, for which he was held responsible. The power of the nobility, which had been curbed by his father decades earlier, was growing. The bishop of Vic was a member of the party which worked out a compromise with Berenguer Ramon whereby the regency would remain in his hands only until his brother’s son had reached the age to govern the county himself. Berenguer Sunifred of Lluça had thus a personal, an ecclesiastical, and also a political interest in promoting the cause of the restoration of Tarragona. Although Vic had lost its previous importance among Catalan cities, or perhaps due to that very fact, he pressed the issue and led the negotiations. By engaging the papacy in the project, he both ensured his own promotion and guaranteed an extension of papal potestas to pilot a sacred war for the restoration of a see that would firmly comply with Roman reformist policy. Nonetheless his efforts would have probably been reduced to the level of a futile personal enterprise if he had not been expressing a united stance of all Catalan counts against Leonese-Castilian interference, and also a social need for a territorial expansion of the county of Barcelona that a weak count-regent could not fulfil. Berenguer Ramon II was not successful against the Muslims; El Cid had much to do with this failure. It is telling that after defeat at the battle of Tevar in 1090 “the Fratricide,” as the count-regent was later known, had to submit to the pope to make realizable the aspirations of the House of Barcelona for an expansion southwards. Berenguer Ramon II’s political survival had depended greatly on the prestige which a campaign in al-Andalus would bring back to him, and on the lands and rents that he could offer the nobility after the conquest. When he failed on that account, he had to fall back on the pope to reinforce his waning power and make possible what he had not been able to achieve by himself. The Tarragona project which was conceived and carried out in the years 1088–92 shows the many implications of an apparently straightforward papal action. If the pluralist approach to the concept of “crusade” is extended backwards in time, the Tarragona campaign might be seen to reduce the impact of

116

Luis García-Guijarro Ramos

Clermont as the starting point for the history of the crusading movement. But that would be misleading. What is most relevant in the conquest of Tarragona and the restoration of its metropolitan see is that, from the second half of the eleventh century, campaigns which in respect of the forms they took might nowadays be regarded as holy wars or crusades were launched by the concerted action of secular and ecclesiastical powers, each struggling to achieve a growth in its influence and control. In the case of the county of Barcelona, the military action well beyond the traditional frontier on the Mediterranean coast, which was completed by mid-1092, had much to do with internal politics, with Barcelona’s relationship with other Iberian states, and with the ideology of fighting the Muslims that was shared by Christian Iberian rulers: in other words, with the Reconquista. When papal crusade bulls appeared on this and other occasions, an occurrence which was by no means always characteristic of campaigns against the Saracens or other infidels, they were merely one of various institutional elements which expressed the common policy of popes and secular rulers. These bulls in themselves explain nothing, and consequently modern historians should not magnify their importance: in doing so, they concentrate on outward forms, rather than upon the roots from which the documents in fact derive their significance. From the assemblies of Girona and Vic in the late 1060s and early 1070s to the restoration of Tarragona in the 1090s, there were many instances of profitable joint work by Catalan counts and the Roman Church leading to common ends and to the increase of their powers – predominantly the comital powers. Popes tried to give some real substance to their aspirations to universal jurisdiction; counts relied on a distant authority to strengthen their hold over territories in northeastern Iberia against baronial centrifugal tendencies, and also to keep a check on the ambitions of the king of León-Castile who, styling himself “emperor,” dreamed of supreme authority over the whole peninsula. Church reformist principles and their implementation came to be useful in many ways as reinforcement for the political units of what was soon to be known collectively as Catalonia.

Map 1: Conquest and Loss of Almería, 1147-1157

6 Alfonso VII of León-Castile in Face of the Reformulation of Power in al-Andalus, 1145–1157: An Essay in Strategic Logic1 Manuel Rojas Gabriel

The following pages will try to prove two basic points. The first of these is that the military operations which Alfonso VII of León-Castile (1126–57) undertook against the Muslims during the long decade between 1145 and 1157 were not random events. These acts of war rather followed the logical strategic guidelines that arose from the readjustment to a new political situation in al-Andalus, of the plan the monarch had been developing since c. 1130 for a gradual attrition of the basis of Almoravid power within the Iberian peninsula. The second point is that Alfonso’s conquest of Almería in 1147, and his concern to keep the city under his control down to its loss in 1157, did not spring from crusading ideals, still less from what an influential historiographical trend of recent decades has designated the “international” movement of the Second Crusade. On the contrary, this entire episode was just a product of dynamic strategies in the light of regional concerns. Background and Context Alfonso VII is not normally remembered for his military or political achievements against the Muslims. With regard to León and Castile Alfonso VI (1065/72–1109), who conquered the city of Toledo (1085) and dominated the taifa kingdoms, Alfonso VIII (1158–1214), who defeated the Almohads at the famous battle of Las Navas de Tolosa (1212), and Fernando III “the Saint” (1217/30–52), who subdued the territorial heart of al-Andalus, spring more readily to mind. With regard to Aragón, Alfonso I “the Battler” (1104–34), who conquered Zaragoza (1118) and frequently defeated the Almoravids at the height of their military power, and Jaime I (1213–76), accurately named “the Conqueror,” who took Mallorca and Valencia, enjoy similar fame. Another king who deserves to be regarded as a great military leader was Afonso Henriques 1

I wish to express my gratitude to José David Cortes Gallardo and Leif Inge Ree Petersen for their help in the translation into English of this paper, and to José Juan Merino del Viejo for his aid in drawing the map. Of course, any mistakes are my own responsibility.

120

Manuel Rojas Gabriel

of Portugal (1139–85) who, within four decades, doubled the territories he had inherited by fighting against the Muslims. Nevertheless, any contemporary of Alfonso VII would without any doubt have described him as a monarch of formidable stature.2 Indeed, in 1135 Alfonso was proclaimed emperor on account of the homage he received from every other Christian leaders within the Iberian peninsula and from some beyond the Pyrenees.3 Until his death in 1157 nobody questioned this status.4 Behind him lay the hard times during which he had had to fight to put an end to the problems inherited from the reign of his mother Queen Urraca (1109–26) and the interference of Aragón in Leonese-Castilian affairs caused by her troublesome marriage to Alfonso I “the Battler.”5 These were circumstances that, for years, diverted him from the conflict against an al-Andalus which, since the previous century, had been absorbed into the fearsome Almoravid empire: “When the war with the king of Aragón had finished, there arose another war in Castile with King García of Pamplona and with King Alfonso of Portugal […] On account of these wars the Emperor did not campaign in the land of the Saracens, and for this reason the Saracens prevailed over the land of the Christians.”6 Admittedly Alfonso VII’s kingship was not embellished by any one outstanding military success. He was involved in some skirmishes, but he never fought a pitched battle. He conquered vital strongholds in order to maintain his frontiers and support his lines of advance, but he never managed to add any of his enemy’s primary strongpoints permanently to his domains. Nor could he guarantee the survival or maintain the loyalty of the various Muslim leaders who acted on his behalf within the Islamic community or, later, became trusted allies against the Almohad hurricane. Long-term conflicts and, particularly, the warfare of positions – so characteristic of medieval warfare – were not to be won through brilliant and outstanding military actions alone.7 The enemy could also be defeated by a tenacious undermining which could lead to the depletion of his resources or to levels of internal disorder and demoralization too high for him to keep on fighting. In this sense, Alfonso VII was a remarkable strategist. 2

3

4

5

6 7

In general, for Alfonso VII’s kingship, see Bernard F. Reilly, The Kingdom of León-Castilla under King Alfonso VII, 1126–1157 (Philadelphia, 1998). See also Manuel Recuero Astray, Alfonso VII, Emperador: El Imperio Hispánico en el siglo XII (León, 1979), and idem, Alfonso VII (1126–1157) (Burgos, 2003), although both works are too dependent on the narrative sources. Chronica Adefonsi imperatoris, ed. Luis Sánchez Belda (Madrid, 1950) [hereafter CAI], i. 71, p. 194. For practical reasons, I have used here the English translation by Simon Barton and Richard Fletcher in The World of El Cid: Chronicles of the Spanish Reconquest (Manchester, 2000). For a brief account of relations between Alfonso VII and his peninsular Christian neighbors, see Esther Pascua Echegaray, Guerra y pacto en el siglo XII: La consolidación de un sistema de reinos en Europa occidental (Madrid, 1996), pp. 212–17. On Urraca’s reign and on the serious internal problems in León-Castile, see Bernard F. Reilly, The Kingdom of León-Castilla under Queen Urraca, 1109–1126 (Princeton, 1982); María del Carmen Pallares and Ermelindo Portela, La reina Urraca (San Sebastián, 2006). CAI, ii. 20, p. 213. John France, Victory in the East: A Military History of the First Crusade (Cambridge, 1994), p. 41.



Alfonso VII and Power in al-Andalus

121

When Alfonso VII became king of León-Castile in 1126, the Almoravid empire was still at the height of its power. The golden years, during which most of the petty kingdoms were dominated by his grandfather, were long gone, liquidated by the power of this African Empire. The Tagus frontier, in particular its core around Toledo, was threatened with collapse by their intermittent but firm pressure. Alfonso’s priority was to avoid any worsening of this situation, and to reverse it. This became possible after 1135 when he established a hegemony over the other Christian rulers which permitted him to take the offensive. The chosen method was a strategic program based on certain guidelines that could be classified as an “indirect approach.”8 By 1145, only a decade later, the achievements of this strategic project were evident. The Leonese-Castilians were just one step away from expanding the Tagus frontier into the area of the upper and middle Guadalquivir valley; Almoravid unity had practically evaporated in al-Andalus; a myriad of new petty kingdoms had blossomed, exploiting the opportunity afforded by the disintegration of North African authority over Iberia. This was the fruit of both political and military strategems. In the first place, a systematic attrition of the enemy’s economic, political, and psychological base was achieved by periodic and damaging raids throughout Islamic territory. Second, certain valuable positions were recovered around the river Tagus. Third, the promotion and support of dissidents was sustained, and anti-Almoravid revolts by the Andalusian population were fostered. Finally, there was the construction of a local political alternative to Maghrebian authority offered by Sayf al-Dawla b. Hud, known as King Zafadola in the Christian sources, who was the emperor’s vassal from the 1130s.9 This created a situation comparable to that which had prevailed under Fernando I (1035/37–65) and in the first half of Alfonso VI’s reign: the military initiative was in Christian hands and al-Andalus was fragmented into disunited petty kingdoms fighting amongst 8

9

Despite his poor opinion of the conduct of medieval warfare, it is possible to adopt the concept developed by Basil Liddell Hart, The Strategy of Indirect Approach, 2nd ed. (London, 1946), pp. 184–216. Zafadola’s father had worked closely with Alfonso I “the Battler” against the Almoravids from his eagles’ nest of Rueda, near Zaragoza; Ibn ‘Idari al-Marrakusi, Al-Bayān al-mugrib. Nuevos fragmentos almorávides y almohades, ed. and trans. Ambrosio Huici Miranda (Valencia, 1963), pp. 131–32, 146. Ambrosio Huici Miranda, “Los Banū Hūd de Zaragoza, Alfonso I el Batallador y los almorávides (nuevas aportaciones),” Estudios de Edad Media de la Corona de Aragón, 7 (1962), 7–38, at 32, discusses information about the alliance between the lord of Rueda and the Aragonese king after the conquest of Zaragoza (1118) that does not appear in his edition of Al-Bayān al-mugrib. On the homage performed by Zafadola to Alfonso VII, CAI, i. 27–29, pp. 162–63, and Ibn al-Kardabūs, Historia de al-Andalus, ed. and trans. Felipe Maíllo Salgado, 2nd ed. (Madrid, 1993), pp. 145–46. Alfonso VII’s ideas about using Zafadola are reported in official records as well as narrative sources: the fuero of Guadalajara, granted on May 3 1133, when Zafadola still had almost no recognition in al-Andalus, lists him as a witness “over the moors antelucinos” i.e., Andalusis: Colección de fueros municipales y cartas pueblas de los reinos de Castilla, León, Corona de Aragón y Navarra, ed. Tomás Muñoz y Romero (Madrid, 1847), p. 511. He is recorded as having been at Alfonso VII’s imperial coronation before he had any real authority, under the title “rex […] Sarracenorum”; CAI, i. 70, p. 182.

122

Manuel Rojas Gabriel

themselves and unable to resist aggression. This enabled the emperor to demand parias (tribute payments).10 In short Alfonso VII recreated what Angus MacKay, writing of the eleventh century, memorably called a “protection racket.”11 The great problem that emerged, with success so close, was that events did not evolve as Alfonso VII and his counselors had expected. It is perfectly reasonable to suppose that, had the sum of the political and military pressure exerted by the Hispanic kingdoms, together with the clearly evident sentiment of rejection toward the Berber regime on the part of the Andalusian population followed their natural course, Almoravid power would had been inexorably defeated. However, the Almoravid empire’s dissolution was in large part an indirect consequence of the appearance, progress, and rapid triumph of the Almohad movement in North Africa.12 The unstoppable advance of those who proclaimed themselves believers “in the unity and oneness of God” demonstrated that the “Andalusian issue” always constituted a secondary front of interest for the Almoravids, who showed no hesitation in deploying most of their military power to try to stop the Almohads in the Maghreb.13 This was really why the Lamtunas (that is, Almoravids) lost the military initiative to the Christian powers between approximately 1139 and 1142, as the Chronica Adefonsi imperatoris relates with a certain optimism: “The strength 10

11 12

13

See further Manuel Rojas Gabriel, “A Decisive Century in the Struggle against Islam in Iberia, ca. 1031–1157: Grand Strategic Perspectives,” in The Medieval Way of War: Studies in Medieval Military History in Honor of Bernard S. Bachrach, ed. Gregory I. Halfond (Farnham, 2015), 185–204; and idem, “The Iberian Christian Kingdoms in the Face of the Almoravids (ca. 1086–ca. 1146): Strategic Responses and Defeat,” forthcoming, and the bibliography in both papers. For a narrative and occasionally analytical account focused on political-military dynamics, see Francisco García Fitz, Relaciones políticas y guerra: La experiencia castellano-leonesa frente al Islam, siglos XI–XIII (Seville, 2002), pp. 77–98. Angus MacKay, Spain in the Middle Ages: From the Frontier to Empire, 1000–1500 (London, 1977), p. 15. The best account of the Almohad empire is the magisterial work of Ambrosio Huici Miranda, Historia política del Imperio Almohade, 2nd ed., ed. Emilio Molina López and Vicente Carlo Navarro Oltra, 2 vols. (1956–57, repr. Granada, 2000). See also Roger Le Tournau, The Almohad Movement in North Africa in the Twelfth and Thirteenth Century (Princeton, 1969); Allen James Fromherz, The Almohads: The Rise of an Islamic Empire (London, 2010); Hugh Kennedy, Muslim Spain and Portugal: A Political History of al-Andalus (London, 1996), pp. 196 ff.; and María Jesús Viguera Molins, “Los Almohades,” in El retroceso territorial de al-Andalus: Almorávides y Almohades, siglos XI al XIII, ed. María Jesús Viguera Molins, Historia de España Menéndez Pidal 8/2 (Madrid, 1997), pp. 73–111. See now too Amira K. Bennison, The Almoravid and Almohad Empires (Edinburgh, 2016). A bibliographical update for the period of the African empires, highlighting some of the most representative tendencies of recent research, can be found in María Isabel Fierro Bello, “Almorávides y almohades,” in 711–1616: de árabes a moriscos. Una parte de la historia de España, ed. Maribel Fierro, Juan Martos, Juan Pedro Monferrer and María Jesús Viguera (Córdoba, 2012), pp. 35–52. Witnessed by the fact that the successful governor of al-Andalus, Tashfin b. ‘Ali was sent to North Africa in 1138 on a mission to try to stop the Almohads: Jacinto Bosch Vilá, Los Almorávides, 2nd ed., ed. Emilio Molina López (1956, repr. Granada, 1998), pp. 241–42. On the military role played by this emir on the Tetouan side of the peninsula, see Ambrosio Huici Miranda, “Contribución al estudio de la dinastía Almorávide: el gobierno de Tasufin ben Ali ben Yusuf en al-Andalus,” Études d’orientalisme dédiées à la memoire de Lévi-Provençal, 2 vols. (Paris, 1962), 2:605–21.



Alfonso VII and Power in al-Andalus

123

of the Saracens and their great power continued until the Emperor Alfonso went to Jerez [de la Frontera, 1133] and until he captured Oreja [1139] and Coria [1142].”14 The combination of factors derived from the gradual withdrawal of the best Almoravid troops to the other side of the Strait of Gibraltar is fundamental if we wish to understand in its entire complexity and magnitude what happened then and immediately after, during the years between 1145 and 1157. The achievements of the Leonese-Castilians, greater than those of any other kingdom, such as Portugal or Aragón-Catalonia, were made possible in the second half of the 1140s not so much by strategic or tactical master-strokes, or occasional foreign aid during the conquest of Lisbon and Almería in 1147 and Tortosa in 1148, as by the abrupt collapse of Almoravid power. As early as 1142, when Alfonso VII conquered Coria, symptoms of weakness could be observed on the Islamic side, because the Almoravids were unable to muster an efficient relieving army to lift the siege.15 Between 1147 and 1148, the petty kingdoms allied with the dying remnants of the Almoravid government and tried to attack the besiegers of Lisbon, Almería, and Tortosa; but by then their military capabilities were too limited to be considered a serious threat. It was at this point that, for a short period – a few years only – and simultaneously, three key political circumstances converged in al-Andalus. First was the disappearance of the monolithic Islamic power represented by the Almoravids.16 Second came the birth of the so-called “second petty kingdoms”: these generally had a rather short existence, except in the eastern regions of al-Andalus17 where Ibn Hamushk of Murcia lasted until 1169, and, until his 14

15

16

17

CAI, ii. 20, 213. For the great punitive expedition of 1133, CAI, i. 33–42, pp. 178–82, and Esperanza Díaz y Providencia Molina, “Las campañas de Alfonso VII en Andalucía: Un precedente de la conquista de Córdoba,” in Andalucía entre Oriente y Occidente (1236–1492), ed. Emilio Cabrera Muñoz (Córdoba, 1988), pp. 63–70 at 64. That was not just an incursion with a military end, but also a consequence of a clear political strategy. It was recommended by Zafadola, who accompanied the Leonese monarch. The fact that the Almoravids did not react led some Andalusian leaders to say to Sayf al-Dawla b. Hud, “Speak to the King of the Christians and, with his aid, free us from the hands of the Moabites [Almoravids]. We will give to the King of León royal tributes larger than those our forefathers gave to his, and with you we will serve him free from fear, and you and your son will reign over us.” After consulting with Alfonso VII, Zafadola’s answer was, “Go, and tell my brothers the princes of the Hagarenes [Andalusians]: Take some strong castles and some strong towers within the cities, wage war in all parts, and the King of León and I will swiftly come to your aid” (CAI, i. 41, p. 181). This was the first hint that the option Zafadola represented was feasible – that at least part of the Andalusian population was willing to face the Almoravids, and that a return to the tutelary and coercive system of the period of the first petty kingdoms was possible. María Dolores García Oliva, “Un espacio sin poder: la Transierra extremeña durante la época musulmana,” Studia Histórica. Historia Medieval 25 (2007), 89–120 at 106–07. Also, CAI, ii. 64–66, pp. 230–31. Significantly, this very situation could be observed during the Siege of Oreja (1139); CAI, ii. 57–58, pp. 227–28. Francisco Codera y Zaidín, Decadencia y desaparición de los almorávides en España, 2nd ed., ed. María Jesús Viguera Molins (1899, Pamplona, 2004), remains the best narrative of the Almoravids’ last years in the peninsula. Besides Codera, Decadencia, for the “second petty kingdoms” period see María Jesús Viguera Molins, “Taifas post-almorávides,” in El retroceso territorial, ed. Viguera Molins, pp. 63–72.

124

Manuel Rojas Gabriel

death in 1172, Ibn Mardadish of Valencia (“the Wolf King” to the Christian chroniclers) maintained a virulent resistance against the Almohads.18 Finally, there was the arrival and progressive implantation of a new Muslim regime from North Africa, which would inherit, albeit partially, the strategic role the Almoravids had previously played. As Jean-Pierre Molénat explains, what both the Almoravids and the Almohads always intended was to consolidate a “front line” situated on the left bank of the river Tagus. As long as Toledo remained under Leonese-Castilian control, it would be impossible to maintain permanent strongpoints north of the river. To a great extent the sources concur that the ultimate intention of both the Almoravids and the Almohads was not so much to conquer the realms of their Christian adversaries as to restore the territorial distribution that had existed before 1085.19 There was, however, for León-Castile and the Christian kingdoms in general, another important and troublesome matter to consider in addition to the need to reformulate their strategic guidelines, and the tactical implementation of these, when the fall of Almoravid power was becoming perceptible. Against whom, where, and how intensely should they proceed? To a degree, these questions arose as a consequence of the Christians’ lack of information about what was happening in every corner of al-Andalus and, particularly, in the Maghreb.20 True, it is evident from the narrative of the Chronica Adefonsi 18

19

20

In general on the Mardanisi regime, see Ambrosio Huici Miranda, Historia musulmana de Valencia y su región, 5 vols. (Valencia, 1970), 3:127–82; Pierre Guichard, Les musulmans de Valence et la Reconquête (XIe–XIIIe siècles), 2 vols. (Damascus, 1990), 1:116 ff.; idem, Al-Andalus frente a la conquista cristiana: Los musulmanes de Valencia (siglos XI–XIII), trans. Josep Torró (Madrid, 2001), p. 113 ff.; Julio Navarro Palazón, Sharq al-Andalus: resistencia frente a los almohades (Murcia, 1993–94). Jean-Pierre Molénat, “Les diverses notions de ‘frontière’ dans la région de Castilla-La Mancha au temps des Almoravides et des Almohades,” in Alarcos 1995: Actas del Congreso Internacional Conmemorativo del VIII Centenario de la Batalla de Alarcos, ed. Ricardo Izquierdo Benito and Francisco Ruiz Gómez (Ciudad Real, 1996), pp. 103–23 at 112–15. On the frontier realities of the period studied here, see Pascal Buresi, La Frontière entre chrétienté et Islam dans la péninsule Ibérique: Du Tage à la Sierra Morena (fin XIe–milieu XIIIe siècle) (Paris, 2004): although Buresi’s methodology is perhaps open to debate, his work has the rare merit of considering both sides of the frontier. See also the papers in Fronteras en discusión: La Península Ibérica en el siglo XII, ed. Juan Martos Quesada and Marisa Bueno Sánchez (Madrid, 2012). There is clear evidence that in general terms, and from the comprehensive application of their own social and cultural parameters, the Christians had a certain level of information about the Islamic sphere: Simon Barton, “Islam and the West: A View from Twelfth-Century León,” in Cross, Crescent and Conversion: Studies on Medieval Spain and Christendom in Memory of Richard Fletcher, ed. Simon Barton and Peter Linehan (Leiden, 2008), pp. 153–74. For example, it is symptomatic that the Chronica makes a clear distinction between “Agari,” “Moabitas,” and “Muzmutos”: that is, between Andalusians, Almoravids, and Almohads. See further Ron Barkai, Cristianos y musulmanes en la España medieval (el enemigo en el espejo), trans. M. Bar-Kochba and A. Komay (Madrid, 1984), pp. 140–44. The Latins understood that not all Muslims were the same and that they should therefore treat them differently. While the first group could be captured, and treaties with them were always feasible, the only possible fate for the Africans was their total elimination, at least until their political fragmentation



Alfonso VII and Power in al-Andalus

125

imperatoris that, grosso modo, the circles closest to the emperor were well informed regarding events throughout the Islamic world.21 Nevertheless, with the means available at the time it was not easy for even the highest levels of goverment genuinely to appreciate the magnitude of the political and doctrinal movement then transforming the geography of power. Nor could they be fully aware of the expansionist ambitions harbored by the Almohads with respect to al-Andalus. Since plentiful and trustworthy information constitutes one of the foundational pillars upon which any efficient strategic planning rests, this circumstance helps to explain why the Leonese-Castilians could not calculate to a sufficiently accurate time-scale what their definitive lines of action should be; not, at least, until the threat from the Maghreb’s new rulers became unequivocally clear in 1146. Three root factors therefore come together to explain the strategic conduct Alfonso VII displayed between 1145 and 1146: first, assessment and comprehensive understanding of the continuous change the Muslim side was undergoing; second, reformulation and re-adaptation of his strategic conduct to a different political context, itself plunged into permanent flux; and third, the management of his military resources and their adjustment to meet a series of new targets which only a few years earlier had seemed chimerical. To an extent, it would seem that the course of events in the Islamic sphere had rather caught León-Castile by surprise. But the Leonese king’s duty was to take advantage of the favorable conjunction of circumstances presented by the loss of unity afflicting the Muslim bloc. This meant that the strategic pattern he had developed until then was no longer the only or the best option. It had simply become obsolete. What was now needed was a directly offensive operational strategy whose ultimate aim was to advance the frontier from the river Tagus to the northeastern sectors of what today is Andalucía. That implied, in turn, the occupation, either by force or submission, of the main strongpoints located within that territory. The strategic guidelines of both the emperor and the other Hispanic princes had therefore to be modified, and evolve gradually from a notion of agile fighting based on rapid movement to one of siege warfare. However, as was by then generally appreciated, defensive procedures were superior to offensive, and experience had proven that even the humblest strongpoint could be hard to overcome, even when circumstances favored the aggressors. Offensive dynamics thus became slower, as they had to be prepared thoughtfully, choosing appropriate targets with care. Moreover the Leonese-Castilians had to plan to take account of the changing forces in an Islamic political scene which was neither stable nor

21

had become such as to make their domination by the Christians conceivable. Meanwhile, and again in general terms, the Muslims either did not know the Christian world, or had only very vague ideas about it. See, for example, Aziz a-Azmeh, “Mortal Enemies, Invisible Neighbours: Northerners in Andalusi Eyes,” in The Legacy of Muslim Spain, ed. Salma K. Jayyusi (Leiden, 1992), pp. 259–73. CAI, ii. 98, 101–05, pp. 245–47.

126

Manuel Rojas Gabriel

uniform. Undeniably that in itself could make Christian expansionist ambitions easier to fulfil. However, the Andalusian political ambience had become so volatile that it was exceedingly difficult to judge which were the Muslim leaders who could be relied upon to establish lasting agreements and alliances that would favor the Christians, or, failing that, who were willing to be used to harass their Muslim neighbors. Furthermore, and as if the situation were not already sufficiently complex, other Latin powers – especially Aragón-Catalonia, Genoa, and, to a lesser extent, the Republic of Pisa – also decided they had the right to take advantage of the condition al-Andalus presented. These too began to calculate their options and proceed to act upon them, with the result that they started to interfere in areas of al-Andalus which León-Castile considered as belonging to her own zone of prospective territorial expansion. Stage One: Alfonso VII versus the Post-Almoravid taifas, and the Beginnings of the Almohad Irruption (mid-1145–mid-1147) This complex context with its many variables explains why Alfonso VII’s strategy between 1145 and 1157 can be divided into three great stages. Until mid-1147 there was no serious Almohad threat to Leonese-Castilian military actions. This enabled Alfonso to prepare as effectively as possible for the forthcoming conflict against the new Berber empire. It would not be entirely unreasonable to suggest that Alfonso VII knew, or at least suspected, that the Almohads were going to land in al-Andalus at the first opportunity, and that he was therefore acting pre-emptively, trying to cushion the impact of their arrival. After all, the situation resembled that at the end of the eleventh century, when the taifas called on the Almoravids for help. Moreover, these second petty kingdoms of the 1140s were much weaker than their predecessors had been in 1086. An early indication of the likelihood of appeals to the Almohads was that made by the sufi mystic, Abu’l-Qasim Ahmad b. al-Husayn al-Qasi, when he became the leader of a furious anti-Almoravid rebellion in the Algarve (Portugal) in 1144: at that point they had answered that they were not able to heed the request – yet. 22 More generally, it is hardly conceivable that the Almohads, who were replacing the Almoravids 22

CAI, ii. 94, p. 242, proves that the Emperor had news of the revolt. For this rebel movement see Josef Dreher, “L’imamat d’Ibn Qasi à Mértola (automne 1144–été 1445): Légitimité d’une domination soufie?” Mélanges de l’Institut Dominicain d’Études Orientales du Caire 18 (1988), 195–210; Christophe Picard, Le Portugal musulman (VIIIe–XIIIe siècle): L’Occident d’al-Andalus sous domination islamique (Paris, 2000), pp. 92–100; Rachid El Hour, “El Algarve en época almorávide: Aspectos políticos y jurídicos,” Codex Aqvilarensis 12 (1998), 35–49 at 41–47; Manuela Marín, “A l’extrémité de l’Islam medieval: Élites urbaines et islamisation en Algarve,” Annales 53 (1998), 361–81 at 373–79. Unfortunately, I have not had access to D. R. Goodrich’s “A Sufi Revolt in Portugal: Ibn Qasi and his Kitab khal ‘al-na’ layn” (Unpublished Ph.D. thesis, Columbia University, 1978).



Alfonso VII and Power in al-Andalus

127

everywhere, would ignore Iberia: after all, the physical gap between Africa and Iberia represented by the Strait of Gibraltar was considered neither a political nor a psychological frontier between the two different Islamic orbits in medieval Muslim mentalities.23 Crossing the Strait and intervening in the territory on the other side was a natural pattern of behavior for the Maghrebins, as it was inversely for the Andalusians on several occasions: the result of a certain degree of “solidarité messianique.”24 Five main events marked this first phase of Alfonso’s strategic development: the partial conquest and subjugation of Córdoba; the unexpected death of Zafadola; the Almohad landing in Cádiz; the arrangement of a pact of vassalage between Yahya b. ‘Ali b. Ghaniya, last Almoravid governor in Spain and the emperor; and the taking of the strongpoint of Calatrava by Alfonso VII. Excepting the last episode, this sequence took place between February and May 1146. Bearing in mind the slow pace more characterisitic of the political and military rhythms of the Middle Ages, all this is indicative of at least two circumstances: first, that the decline of Almoravid power in al-Andalus had accelerated to the point of collapse in a remarkably short period of time; and second, that Alfonso VII was obliged to react to the new political set-up so rapidly that he was very probably unable to furnish the means, and did not have the time, to confont the situation in the most effective manner. He had to improvise some of his maneuvers as the events were occurring. The Leonese king’s efforts were still aimed primarily at trying to control the regions around the upper and middle Guadalquivir (as Fernando III was later to succeed in doing, in the second quarter of the thirteenth century). This was the most logical target for León-Castile. After submitting Córdoba to his authority, Alfonso’s obvious next goals were Jaén and perhaps even Seville, if the strategic conditions were conspicuously favorable – and he could well have considered them to be so, given the astonishing Almoravid collapse and the politico-doctrinal totum revolutum into which the Islamic territory had lapsed. Thus, the most reasonable way to proceed was to attack as soon as possible and attempt to conquer the territorial heart of al-Andalus. There is little point in recounting once more here a generally familiar series of events. What must be considered are those facts that are most relevant to an understanding the strategic guidelines adopted by Alfonso VII. Two occurrences determined the path to be followed by the Leonese king or, at least, made much clearer the choices required of him in response to a sudden volatility of the 23 24

Mohamed Kably, Société, pouvoir et religion au Maroc à la fin du “Moyen Âge” (XIVe–XVe siècle) (Paris, 1986), pp. 80, 93. Ibid., p. 80. Regarding the strategic importance of the Strait during the Middle Ages see Joaquín Vallvé, “Las relaciones entre al-Andalus y el norte de Africa a través del Estrecho de Gibraltar (siglos VIII–XV),” in Actas del Congreso Internacional “El Estrecho de Gibraltar”, ed. Eduardo Ripoll Perelló and Manuel Fernando Ladero Quesada, 4 vols. (Madrid, 1988), 2:9–36. For the events of this period, see B. Rosenberger, “Le contrôle du Détroit de Gibraltar (XIIe–XIIIe siècles),” in L’Occident musulman et l’Occident chrétien au Moyen Âge, ed. Mohammed Hamman (Rabat, 1995), 15–42.

128

Manuel Rojas Gabriel

situation. The first was the murder of Sayf al-Dawla b. Hud (Zafadola or Çahedola in Christian chronicles) by Christians in February 1146 in very obscure circumstances.25 With his death the plan of imposing an indirect sovereignty over a fragmented and vulnerable Muslim Spain evaporated. It was time now to negotiate pacts and alliances with various Islamic regional powers. The second event was the irruption of the Almohads into southern Spain in April–May 1146.26 During the remainder of the year, they took over Jerez de la Frontera, Arcos, Ronda, Niebla, Mértola, Silves, and the Algarve. In January 1147 they were already besieging Seville. As already mentioned, these factors added to the instability of Hispanic-Muslim political geography: some petty kingdoms resisted the Almohads, and active remnants of Almoravid power persisted. Their effect was to make the planning of a new strategy adequate to the changed political scene an urgent task. That meant undertaking direct offensive campaigns which hastened territorial annexations, and finding support among the Andalusian leaders, fearful of Almohad dominion. The result was a fierce duel between the Christians from the north and the Almohads from the south, the trophy being the annexation of as many pieces as possible of the political jigsaw-puzzle created by the breakdown of Almoravid Andalusia. Movements and target selection could not be disorderly or random: every step needed to be considered with care in order to avoid irremediable errors. The general circumstances indicated that Alfonso VII should act like Alfonso VI when he conquered Toledo in 1085 and immediately made the city his main headquarters, seeking to control its surrounding territory in order to secure the frontier and to develop his operational strategy against the Muslims. In Alfonso VII’s case, it was clearly Córdoba which would enable him to fulfill equivalent aims, especially if his ultimate purpose was to advance from the Tagus to the upper and middle Guadalquivir valley. Conditions in 1146 were not the same as in 1085, however: a formidable and expansionist African power constituted an immediate and dire threat against which the Christians would have to fight sooner or later. This created uncertainty, especially as very little was known about Almohad capabilities. If the Almohads were destroying the Lamtunas everywhere, would their power be even greater? Whether or not that turned out to be the case, the outlook could scarcely be less pleasing or of more concern. The upshot was that Alfonso VII simply did not have the time to devise an occupation plan for Córdoba and then compel its population to accept Leonese-Castilian control; quite the contrary. At the beginning of 1146, he was forced to proceed in the most expeditious manner, before the situation changed and he lost the only chance 25

26

We only have the Christian version of the events: CAI, ii. 96–98, pp. 243–45. Possibly Zafadola was murdered because he was trying to release himself from dependence on Alfonso VII: breaking from his lord would have been considered treason, and would have further complicated the situation for the Leonese monarch. The chronology of the first Almohad intervention in al-Andalus is confusing; but see, Viguera Molins, “Los Almohades,” pp. 83–84.



Alfonso VII and Power in al-Andalus

129

he had. The emperor had to improvise, and quickly; hardly the conditions for designing and executing a flawless operational strategy. Furthermore, inside Córdoba, as in most of al-Andalus, the balance of power was very changeable, almost chaotic. In its concision the Anales Toledanos 1 eloquently summarizes the fluctuations between the beginning of 1145 and the spring of 1146, that is, between the moment when the effective dissolution of the Almoravid regime became obvious and the entry of Alfonso VII into the city: Çahedola [Sayf al-Dawla b. Hud] went in the month of January [1145] to Córdoba and killed Farach Adali [farax, adalid, from the Arabic ad-dalil, ‘guide’, here ‘military commander’, a warrior of great fame who had as a base the fortress of Calatrava], and he fled to Granada, and after Çahedola fled they raised Aben Hamdin [Ibn Hamdin, a qadi who filled the power vacuum and gained control over the city] as king in Córdoba [March 1145]. Çahedola fought against the Christians and they killed him [February 1146]. King Abengama [Yahya b. ‘Ali b. Ghaniya, the last Almoravid governor in al-Andalus] expelled king Aben Hamdin from Córdoba [February 1146]. Later on, in the month of May [1146] the Emperor captured Córdoba, and after that he gave it to Abengama.27

This was the point at which the Almohads burst on to the scene, an event that would transform enemies into allies. Their appearance threatened the annexationist ambitions of the northern kingdoms, particularly those of León-Castile and Portugal, as well as Almoravid power in al-Andalus, and the HispanicMuslim petty kings. Thus, although at first sight it might seem a violation of the normal logic, representing a paradoxical union between opposite sides, Yahya b. ‘Ali b. Ghaniya agreed to perform homage to Alfonso VII in exchange for being allowed to continue ruling Córdoba on his behalf.28 The old adversaries, for the time being, needed each other: having taken so much care, both political and military, Alfonso VII achieved, albeit provisionally, the submission of Córdoba, 27

28

Anales Toledanos 1, ed. Enrique Flórez, España Sagrada 23 (Madrid, 1767), 381–400 at 389. For Ibn Hamdin’s political maneuvers during these months, CAI, ii. 99–100, p. 245; for his personal trajectory, Rachid El Hour, “La transición entre las épocas almorávide y almohade vista a través de las familias de ulemas,” in Biografías Almohades I, ed. Maribel Fierro y María Luisa Ávila (Madrid, 1999), pp. 261–305 at 263–64. For the Almoravid leader and his lineage, see Alfred Bel, Les Benou Ghaniya, derniers représentants de l’empire almoravide et leur lutte contre l’empire almohade (Paris, 1903). It is surprising that CAI does not register the partial conquest of Córdoba, as Yahya b. ‘Ali Ghabiya kept the medina, where he had gained strength. Besides the sketchy Anales Toledanos 1, we have the Muslim sources: Histoire de l’Espagne musulmane, extraite du Kitab A’mal al-A’lam, ed. and trans. Évariste Lévi-Provençal (Rabat, 1934), pp. 291–92. The royal chancellery could not resist proclaiming this situation: in August 1146, a document was dated in Toledo, “xiv Kalendas” of September, “In the era of 1183 [1146], after the return of the campaign with which the Emperor made his vassal the Moor Prince Abigania [Yahya b.’Ali b. Ghaniya] and plundered part of Córdoba along with its major mosque”: Peter Rassow, “Die Urkunden Kaiser Alfons’ VII. von Spanien. Eine palaeographisch-diplomatische Untersuchung,” Archiv für Urkundenforschung 10 (1928), 328–467, and 11 (1930), 66–137 at 97–98.

130

Manuel Rojas Gabriel

one of the most important urban centers of al-Andalus. Symbolically, it was a joyous triumph, producing great prestige, but above all it represented a remarkable strategic victory, as the city was indispensable to control of the upper and middle sectors of the Guadalquivir valley. Official propaganda, predisposed to praise royal achievements, portrayed the event as a true de facto conquest.29 If, however, Alfonso VII wanted to keep developing his new strategy based on direct action and the exploitation of the divisions within the Muslim orbit, he also intended to maintain his authority over Córdoba, so it was vital not to neglect in his wake any important stronghold. That meant taking any Muslim position that could undermine the advancing Leonese-Castilian front and from which it was possible to harass the long Christian communication lines. At that moment, Córdoba was the spearhead of León-Castile; nonetheless, it was not a consolidated conquest and, what is more, was virtually isolated within a disputed area, with no potential aid from local allies in case of need. Indeed the true Leonese-Castilian base of operations was still Toledo, around 300 km away from the old Ummayad capital city.30 This set of factors made the conquest of Calatrava (the Old) an ineluctable obligation, for this was a stronghold located on the river Guadiana’s middle–upper basin, a mandatory halt between Córdoba and Toledo which also lay across the routes which connected eastern and western Spain. The fact that Calatrava, and the nearby fortress of Almodóvar de Campo, were taken during the winter, in January 1147, is proof of their importance.31 The short campaign was possible because of temporary Almohad setbacks which prevented the sending of a relief force. Alfonso tried immediately to consolidate his hold by repopulating the area in order to transform it into a key position for the maintenance of the front opened in the Guadiana’s upper and middle regions.32

29

30

31

32

In November 1146 royal documents declared that Alfonso VII was “imperatore in Toleto, in Legione, in Saragoza, in Nauara, Castella, et Galicia, et Cordube”: Juan del Álamo, Colección diplomática de San Salvador de Oña (822–1284), 2 vols. (Madrid, 1950), 1: no. 199. If, as Martin van Creveld explains in Supplying War: Logistics from Wallenstein to Patton, 2nd ed. (Cambridge, 2004), “the usual method, indeed the very aim of warfare, was to live at the enemy’s expense” (p. 23), then it is worth remembering that during previous years the Christian expeditions attacking or returning from what today is Andalucía must have had serious logistic problems between the Tagus border and Sierra Morena. An example is the great incursion of 1133: see CAI, ii. 34, p. 179. Until that moment, the Muslim Calatrava had acted as a nest of almogávares (violent men from the frontier) and like a base from which to attack the Christian territories; CAI, ii, 15, pp. 210–11; 83–88, pp. 237–39. The document in which Alfonso VII delivered the mosque of Calatrava to the cathedral of Toledo, remembered and emphasised precisely what Toledo had suffered from this stronghold during the past years; José Antonio Garcíá Luján, Privilegios reales de la cathedral de Toledo (1086–1462): Formación del patrimonio de la S. I. C. P. a través de las donaciones reales, 2 vols. (Toledo, 1982), 2, no 18, pp. 58–60. On the stronghold, Amador Ruibal Rodríguez, Calatrava la Vieja: Estudio de una fortaleza medieval (Cuidad Real, 1984); Enrique Rodríguez Picavea, “Calatrava: una villa en la frontera castellano-andalusí del siglo XII”, Anuario de Estudios Medievales 30/2 (2000) 807–49



Alfonso VII and Power in al-Andalus

131

Stage Two: The Conquest of Almería within the Leonese-Castilian Strategic Context (mid-1147–1148) A few months after the conquest of Calatrava, a great military operation took place, considered by some historians as integrating Iberian warfare into the Second Crusade. In this view the conquest of Almería was part of a general offensive launched by Latin Christendom against the Muslim and pagan worlds. This is the substance of the so-called “Constable thesis,”33 which, despite some dissenting voices,34 has migrated from revisionism to orthodoxy, at least as far as Anglophone crusader studies are concerned.35 However, it is a judgment which may prove not to be well founded when all contemporary circumstances are considered. Was it possible, in the context of the Middle Ages, to conceive of major military guidelines to “extend the frontier of [Latin] Christendom” (to employ the subtitle of Jonathan Phillips’s monograph)? Could such an initiative, involving expeditions on three fronts separated by thousands of kilometers, be comprehended within the geopolitical parameters of the time?36 Finally, given the difficulties of organizing, undertaking and maintaining hostilities, was it feasible to execute such a plan as that envisaged for the Second Crusade? Even with the communication networks, the flow of information and the logistic, 33

34

35

36

Giles Constable, “The Second Crusade as Seen by Contemporaries,” Traditio 9 (1953), 213–79. Years later, Constable reaffirmed his position in “A Further Note on the Conquest of Lisbon in 1147,” in The Experience of Crusading, edited by Marcus Bull and Norman Housley, 2 vols., 1: Western Approaches (Cambridge, 2003), pp. 39–44 at 44. Both essays are reprinted with some revisions in Giles Constable, Crusaders and Crusading in the Twelfth Century (Farnham, 2008), chap. 10, 229–300, and chap. 11, 301–09. For instance, Alan J. Forey, “The Second Crusade: Scope and Objectives,” Durham University Journal 86 (1994), 165–75, and idem, “The Siege of Lisbon and the Second Crusade,” Portuguese Studies 20 (2004), 1–13; John France, The Crusades and the Expansion of Catholic Christendom, 1000–1714 (London, 2005), pp. 130–38, and idem, “Logistics and the Second Crusade,” in Logistics of Warfare in the Age of the Crusades ed. John H. Pryor (Aldershot, 2006), pp. 77–94. For example, Jonathan Riley-Smith, The Crusades: A History, 2nd ed. (New Haven, 2005), p. 124; Nikolas Jaspert, The Crusades, trans. P. G. Jestice (London, 2006), pp. 47–48 and p. 121. Jonathan Phillips, The Second Crusade: Extending the Frontier of Christendom (New Haven, 2007), while accepting Constable’s ideas, nonetheless suggests that there was no predetermined plan to expand the Latin Christendom on three fronts designed at the highest ecclesiastical level, but rather a papal reaction to local impulses, which gave unity and sense to the diverse Christian offensives against Muslims and Slavs. This idea was already suggested in Phillips and Hoch’s introduction to The Second Crusade: Scope and Consequences, ed. Jonathan Phillips and Martin Hoch (Manchester, 2001), pp. 1–14 at 3 and 8, which accepts the existence of “twelfth-century Spanish crusading narratives” (p. 2); see also Phillips’s “Second Crusade (1147–1149),” in The Crusades: An Encyclopedia, ed. Alan V. Murray, 4 vols. (Oxford, 2006), 4:1084–90. It is symptomatic that Otto of Freising, The Deeds of Frederick Barbarossa (Gesta Friderici I imperatoris), ed. and trans. Charles Christopher Mierow and Richard Emery, 2nd ed. (1953, repr. New York, 2004), 2, p. 130, makes the Genoese participants in the conquest of Lisbon in 1147, while Eugene III’s bull Divina dispensatione, 11 April 1147, PL, 180, cols. 1, 203–04, does not even mention the northern European fleet which helped to besiege and take the city.

132

Manuel Rojas Gabriel

economic, material and military resources available nowadays, developing an undertaking on such a scale would be extremely complicated. In the mid-twelfth century such a thing would simply have been impossible to accomplish, if not indeed unimaginable. We are, then, facing an explanation derived from an overly idealized a posteriori analysis: “a neat version of the past,” in the words of Christopher Tyerman.37 In the context of twelfth–century Iberia, the “Constable thesis” is thus highly controversial.38 This is especially true as regards the siege and conquest of Almería between mid-August and October 1147 by Leonese-Castilian, CatalanAragonese, Genoese, Navarrese and Provençal contingents.39 It is not however the aim of this paper to try to amend the hypothesis in its entirety; nor will it analyze the possible doctrinal motives or the epic-historic aspects associated with the episode in question,40 or the factors which might have inspired the Republic of Genoa to be the main architect of the campaign.41 Still less is it 37 38

39

40

41

Christopher Tyerman, God’s War: A New History of the Crusades (London, 2006), p. 304. This has not prevented some scholars of medieval Spain from characterizing the conquest of Almería as a “Crusade.” Amongst these are Joseph P. O’Callaghan, Reconquest and Crusades in Medieval Spain (Philadelphia, 2003), pp. 41–49, Ricardo da Costa, A guerra na Idade Média: Estudo da mentalidade de cruzada da Peninsula Ibérica (Rio de Janeiro, 1998), p. 69 ff., and the pluralist avant la lettre José Goñi Gaztambide, Historia de la bula de cruzada en España (Vitoria, 1958), p. 76 ff. By contrast, although he focuses mainly on northern Spain, Luis García-Guijarro Ramos, “Reconquest and the Second Crusade in Eastern Iberia: The Christian Expansion in the Lower Ebro Valley,” in The Second Crusade: Holy War on the Periphery of Latin Christendom, ed. Jason T. Roche and Janus Møller Jensen (Turnhout, 2015), pp. 219–55, refutes this approach. For an overview of Muslim Almería, see Leopoldo Torres Balbás, “Almería islámica,” Al-Andalus 22 (1957), 411–57; on the fortifications of the town, Lorenzo Cara Barrionuevo, La Almería islámica y su alcazaba (Almería, 1990). The exact date by which the petty kingdom of Almería was established is not known, but 1145 is a likely date according to Abu Muhammad Abd al-Wahid al-Marrakusi, Kitab al-Mu’yib fi Taljis Ajbar al-Magrib, ed. and trans. Ambrosio Huici Miranda (Tetouan, 1955), pp. 170–71. See, for example, H. Salvador Martínez, El “Poema de Almería” y la épica románica (Madrid, 1975); Raymond McCluskey, “Malleable Accounts: Views of the Past in Twelfth-Century Iberia,” in The Perception of the Past in Twelfth-Century Europe, ed. Paul Magdalino (London, 1992), pp. 212–27 at 217–19. On the trading importance of Almería, see Olivia Remie Constable, Trade and Traders in Muslim Spain: The Commercial Realignment of the Iberian Peninsula, 900–1500 (Cambridge, 1994), pp. 18–19 and passim; Emilio Molina López, “Almería Islámica: Puerta de Oriente, objetivo militar. Nuevos datos para su estudio en el Kitab Iqtibas al-anwar de Abu Muhammad al-Rusati,” Actas del XII Congreso de la Union Europea de Arabistas e Islamistas, Málaga 1985 (Madrid, 1986), pp. 559–608, and “Algunas consideraciones sobre la vida socio-económica de Almería en el siglo XI y primera mitad del XII,” Actas IV Coloquio Hispano-Tunecino, Palma de Mallorca 1979, vol. 1 (Madrid, 1983), pp. 181–86; al-Sayyid ‘Abd al-‘Aziz Salim, “Algunos aspectos del florecimiento económico de Almería islámica durante el período de las taifas y de los almorávides,” Revista del Instituto de Estudios Islámicos 20 (1979–80), 7–22. For Genoese commercial relations with the city in the first half of the twelfth century see Miracula b. Egidii auctore Petro Guillelmo, ed. Phillipp Jaffé, in MGH SS 12 [Historiae aevi Salici, ed. Georg Heinrich Pertz] (Hanover, 1852), pp. 316–23 at. p. 321; Luigi Tommaso Belgrano, “Il registro della curia arcivescovile di Genova,” Atti



Alfonso VII and Power in al-Andalus

133

concerned to relate how the siege of the city evolved,42 and how its inhabitants finally capitulated because of the pressure exerted by the besiegers and the absence of aid from allied Muslim forces.43 Although each one of these points is in itself of considerable interest and worthy of the greatest attention, our main concern here is to analyze the strategic guidelines developed by Alfonso VII from 1145 onwards. To view the taking of Almería as part of the Second Crusade underestimates, if it does not ignore, the specific ideological features and the particular politicalmilitary dynamisms which were being developed by the three main HispanicChristian powers against Islam: León-Castile, Aragón-Catalonia, and Portugal. These Christian powers were competing against each other, and it was unusual for them to act together against the Muslims, seeking a common tactical goal which could imply the defeat and systematic occupation of Andalusia. It is clear that, for a number of different reasons – normally volatile or unstable – such as ties of clientage, family bonds at a high level, shared interests on certain occasions and almost always for short periods of time, the Christian entities were able to forge transitory agreements to combine efforts against the Muslims. But such situations were more the exception than the norm. Fighting against the Islamic enemy was usually an enterprise undertaken individually by Christian kingdom; thus the political strategies developed, objectives set, and operational plans were

42

43

della Società Ligure di Storia Patria 2/2 (1860), 9–11. Some general indication of the city’s importance can also be found in Blanca Garí, “Why Almería? An Islamic Port in the Compass of Genoa,” Journal of Medieval History 18 (1992), 211–31. For the campaign of 1147, see John Bryan Williams, “The Making of a Crusade: The Genoese Anti-Muslim Attacks in Spain, 1146–1148,” Journal of Medieval History 23 (1997), 29–53. Because of its extent, although with minor errors, the most complete account of the campaign is José Ángel Tapia Garrido, Historia general de Almería y su provincia, 3: Almería musulmana (711–1172) (Almería, 1986), pp. 347 ff.; also, for example, Phillips, Second Crusade, pp. 250 ff., although his interpretative orientation must be kept in mind. The best contemporary description is provided by Caffaro, Cafari ystoria captionis Almarie et Turtuose ann. 1147 et 1148, in Annali Genovesi di Caffaro e de’ suoi continuatori, vol. 1, ed. Luigi Tommaso Belgrano, Fonti per la storia d’Italia 11 (Genoa, 1890), pp. 77–89, translated into English in Williams, “Making of a Crusade,” Appendix 1, pp. 48–51, and also in Caffaro, Genoa and the Twelfth-Century Crusades, ed. and trans. Martin Hall and Jonathan Phillips, Crusade Texts in Translation 24 (Farnham, 2013); although the Prefatio de Almaria, in Chronica Hispana saeculi XII, Corpus Christianorum, Continuatio medievalis 71/1, ed. Juan Gil (Turnhout, 1990), pp. 253–67, translated as “The Poem of Almeria” in World of El Cid, ed. Barton and Fletcher, pp. 250–63, adds further valuable information. The Muslim sources are rather meager. For example, the eastern historian Ibn al-Athir, The Chronicle of Ibn al-Athir for the Crusading Period from al-Kamil fi’l-Ta’rikh, 2: The Years 541–589/1146–1193: The Age of Nur al-Din and Saladin, ed. and trans. D. S. Richard (Farnham, 2007), only indicates that “In Jumada I of this year [October 1147] the Franks besieged Almeria in Andalusia and pressed it hard by land and by sea. They took it by force of arms and carried out much killing and plundering.” On the siege operations, see Randall Rogers, Latin Siege Warfare in the Twelfth Century (Oxford, 1992), pp. 177–79, which, however, includes little about tactics and techniques, merely summarizing the siege briefly.

134

Manuel Rojas Gabriel

individual to each Christian power.44 As a matter of fact, and as I hope to make very clear, in spite of the ties of vassalage which linked Ramón Berenguer IV of Catalonia-Aragón to Alfonso VII, one of the reasons that forced the emperor to join the Almería campaign was precisely to prevent the Catalan-Aragonese from opening potential lines of advance within the region in the future. As José María Mínguez argues, The joint Catalan-Aragonese and Leonese intervention in Almería would introduce a potentially conflict-causing element in the relationship between the emperor and his vassal Ramón Berenguer IV, inasmuch as it indicated a convergence of interests over the same territory. It is not necessary to recall here the repeated Leonese actions, both in the Ebro valley and in Valencia, in an attempt to control the entire east. The Aragonese and Catalans, for different reasons, were interested in controlling this territory.45

Alfonso VII did not decide to join and lead the campaign of Almería on a whim or in response to a request for aid from Genoa, let alone from simple ambition or “achievement-motivation.”46 The Leonese king had demonstrated too much strategic talent and wisdom in conflicts, whether against Muslims or Christians, to let himself be dragged into a situation of that kind. Temptations which might well enhance propaganda, fame, and epic were bad counselors when it came to the conduct of warfare, and have lured many political and military leaders to failure throughout the course of history. That the conquest of Almería must rather be placed within the strategic context of political fragmentation in al-Andalus at the time is patently clear. Nonetheless, besides this, other important variables which triggered Alfonso VII’s participation in the campaign and, therefore, his promotion to its leadership, must also be taken into consideration. Firstly, in spite of the fact that the Muslims were the natural enemies of the Christian kingdoms, they were not the only ones. Each Christian power was suspicious of the ambitions of the others. Alfonso VII had to deal with Genoese ambition, which although Christian might well disturb Leonese-Castilian plans, and meanwhile restrain too the expansionist claims the Catalan-Aragonese had been advancing for decades with respect to the eastern regions of Iberia. Indeed, although Genoa was an ally, and Ramón Berenguer IV of Catalonia-Aragón a client, of Alfonso VII, these circumstances could obviously change to the detriment of León-Castile. Moreover, if Almería ended up in Alfonso VII’s hands, he could properly re-shape his operational strategy with regard to the southwestern regions of the peninsula. He could conquer Jaén and Granada, or at least dominate them through vassals or allies, containing the Almohads. It is important to recognize that Ibn Mardanish of Valencia and Ibn Hamushk of Murcia were the 44 45 46

Rojas Gabriel, “Decisive Century,” passim. José María Mínguez, La España de los siglos VI al XIII: Guerra, expansión y transformaciones. En busca de una frágil unidad, 2nd ed. (San Sebastián, 2004), p. 281. Norman F. Dixon, On the Psychology of Military Incompetence (London, 1976), pp. 238 ff.



Alfonso VII and Power in al-Andalus

135

strongest of the petty kings who dared to challenge the Almohads. The siege of Almería thus made perfect sense and was not a departure from Alfonso VII’s strategic logic; it should not be contrasted with the neat lines of action of Afonso Henriques towards Lisbon, or of Ramón Berenguer IV and the conquests of Tortosa, Fraga and Lérida. The Iberian Christian monarchs never allowed foreign forces to act by themselves without clear tutelage. The Christian contingents from beyond the Pyrenees that fought in Iberia during the Middle Ages were always under the command of local political leaders, becoming part of their hosts and acting according to local strategic plans and tactics. Alfonso VII was no exception, and would certainly not permit an alien Christian power over which he had no control to occupy an important Muslim city. This was especially true when the city was located in the heart of a territory over which León-Castile considered itself to have indisputable rights of annexation and conquest. The ambitions of Genoa were already patent: during the summer of 1146 the Genoese had attacked Menorca and then besieged Almería, although the end of the most propitious season for undertaking military activities had forced them to lift the siege.47 Alfonso VII was conscious that Genoa was prepared to make major human, material, and logistic efforts to take the city. This would eliminate its harbor as a competitor in the western Mediterranean, and provide entry to the maritime trading routes which connected al-Andalus with “Berbería.”48 Such a situation posed grave problems for León-Castile and its allies, who could hardly welcome a new and aggressive political-military power bursting on to the Iberian scene. They would have no hesitation in treating with the Muslims to prevent this if there was money involved or their own interests were at stake.49 However, were Alfonso VII to join the campaign, and furthermore to enlist the auxilia of those main vassals who could actively collaborate in the undertaking – Ramón Berenguer IV, García Ramírez IV of Navarre and Guillaume IV of Montpellier – things would be very different; not least because this way of proceeding would raise him immediately to the position of highest authority over the hosts, thanks to his superior status in the hierarchy, and, even more importantly, because Almería would remain under his control by right of conquest. It is true that most of the booty would go to the Genoese, who would acquire a third of the city itself, besides many other privileges like keeping their own factories, markets, churches, baths, and ovens; but this was a minor issue compared to the grand prize of possessing this valuable urban center. 47 48 49

Caffaro, Annales Ianuenses, in Annales Genovesi, ed. Belgrano, pp. 3–75 at 33–35; translated in Caffaro, Genoa, ed. Hall and Phillips, pp. 49–101. For a brief summary, see Steven A. Epstein, Genoa and the Genoese, 958–1528 (Chapel Hill, 1996), pp. 40 ff. It is hard to forget that two centuries later the Chronicle of Alfonso XI of Castile would remark, “the Genoese were always willing to help those who gave them money, but beyond this they did not favor Christianity or any other good.” Crónica del rey don Alfonso el Onceno, in Crónicas de los reyes de Castilla, vol 1, ed. Cayetano Rosell, 2nd ed., Biblioteca de autores españoles 66 (1875, repr. Madrid, 1953), pp. 171–392 at 309.

136

Manuel Rojas Gabriel

It is unlikely that, had Genoa not asked for aid in taking Almería,50 Alfonso VII would have decided to undertake its conquest by himself. His attention was focused at the time on other troublesome sectors, of higher priority. He immediately recognized, however, that the Genoese proposal could be turned to his own benefit. All the southwestern areas of the peninsula – from Seville to the Algarve and Badajoz – were falling under Almohad rule, so that a line of positions formed by Córdoba, Jaén and Almería (plus a series of secondary strongpoints) was the only practicable option for containing the North Africans. Such a front line had to be supported logistically by fortified sites located between that advanced front and the Tagus basin and, to its eastern flank, by Andalusian petty kings willing to keep their autonomy by fighting the Maghrebins. In other words, it was necessary to deepen the front line. Nevertheless, there were two serious impediments preventing any such strategic plan from being put into practice. First, the vast Castilian-Manchegas prairies were in a state of semi-depopulation, which meant that, except for Calatrava, in the forthcoming years there would be no strong bases to support the Leonese-Castilian forces or provide security for conquests achieved in what today is Andalucía. It was perfectly possible that the situation experienced in the fortress of Aledo in 1092 and Valencia in 1109 might happen again: both were untenable against the Almoravid thrust because they were isolated and surrounded by enemy territory.51 Second, it was a high-risk project which could end up with nothing. In the summer of 1147, Alfonso VII could only count on his weak control over Córdoba, although with tenacity he would soon obtain Almería and would also try to occupy Jaén. With these three cities under his dominion, perhaps Granada – one of the last important bastions the Almoravids still had – would become a realistic target. Anxiety about Aragón-Catalonia was a further factor propelling Alfonso into the alliance against Almería. Its rulers looked beyond the Ebro border and towards the Levant, Murcia, and the northern areas of the present-day Andalucía. Once we understand the competitive relationship between these two Christian states, we can see this was a decisive strategic reason for the Leonese monarch to intervene in the campaign. If Alfonso VII seized Almería he would be in a very advantageous position to check such Catalan-Aragonese ambitions. Aragón’s objectives had become clear since the great expedition commanded by Alfonso I “the Battler” in 1125–26 through the Levant and the northern areas of Andalucía,52 and the attack on Valencia in 1129.53 The succession crisis after Alfonso I’s death in 1134 and the union with Catalonia by marriage in 1137 had 50 51

52 53

CAI, ii. 107, p. 248; Codice diplomatico della Repubblica di Genova, ed. Cesare Imperiale di Sant’Angelo (Rome, 1936), nos. 166–68, pp. 204–14. See Ambrosio Huici Miranda, Las grandes batallas de la Reconquista durante las invasiones africanas (almorávides, almohades y benimerines) (Madrid, 1956), pp. 85–99, or Guichard, Al–Andalus, pp. 63–83. José Ángel Lema Pueyo, Alfonso I el Batallador, rey de Aragón y Pamplona (1104–1134) (Gijón, 2008), pp. 196–213. Documentos para el estudio de la Reconquista y la repoblación del valle del Ebro, ed. José María Lacarra (Zaragoza, 1982–1985), no. 181; Ambrosio Huici Miranda, “Los Banu Hud de



Alfonso VII and Power in al-Andalus

137

slowed the confederation’s expansionist movements down for a while; however, it would not be long before these were reactivated, reinforced by Catalan maritime ambitions. More recently, in 1144, Aragón-Catalonia had directed a great predatory raid against the lands of Almería and Granada, which meant that “all the kingdom of the Hagarenes [i.e. Andalusis] from Almería to Calatrava was destroyed, and there remained nothing but a few really strong cities and castles.”54 Such interventions led to a vassalic pact between Ibn Mardanish and Ibn Hamushk and Ramón Berenguer IV. In 1147–50 Alfonso VII was in a position to control this situation, as he was at the head of the feudal hierarchy. Nonetheless, the attainment of this status had to be clearly demonstrated. In January 1151, the Treaty of Tudején, or Tudilén, was signed, motivated by the need to end tensions with Navarre and to delimit expansion zones in the Muslim south. Of course, the terms of the treaty responded to the political logic which linked a lord (Alfonso VII) and a vassal (Ramón Berenguer IV); the latter was assigned Valencia, Denia, and Murcia as potential areas of conquest, although for the fortresses of Vera and Lorca – directly to the south of Murcia, in lands adjacent to Almería’s area of influence – he had to pay homage to the emperor, who kept the rest of al-Andalus for himself.55 Mutatis mutandis, and except for small alterations introduced in later treaties,56 this division of the expansionist spaces of both kingdoms against Islam remained virtually unaltered for the rest of the Middle Ages.57

54 55

56

57

Zaragoza, Alfonso I el Batallador y los almorávides,” Estudios de Edad Media de la Corona de Aragón 7 (1962), 7–38 at 29–33. CAI, ii. 92, p. 241. The text of the treaty is in Colección de documentos inéditos del Archivo General de la Corona de Aragón, ed. Próspero de Bofarull y Mascaró (Barcelona, 1849) 4, no. 62, pp. 168–74; and Liber feudorum maior, ed. Francisco Miquel Rosell, 2 vols. (Barcelona, 1945), 1:39–42. See Juan Torres Fontes, La delimitación del Sudeste peninsular (tratados de partición de la Reconquista) (Murcia, 1950), pp. 5–14. Besides Torres Fontes, Delimitación see, for example, Julio Valdeón Baruque, “Las particiones medievales en los tratados de los reinos hispánicos. Un precedente de Tordesillas,” in El Tratado de Tordesillas y su proyección (Valladolid, 1973), pp. 21–32; Miguel Ángel Ladero Quesada, “Sobre la evolución de las fronteras medievales hispanas (siglos XI a XIV),” in Identidad y representación de la frontera en la España medieval (siglos XI–XIV), ed. Carlos de Ayala Martínez, Pascal Buresi, and Philippe Josserand (Madrid, 2001), pp. 3–49. Even so, Almería remained a potential Catalan-Aragonese target. In 1309, in the midst of the War of the Strait of Gibraltar (1275–1350), Aragón tried to conquer the city. See Andrés Giménez Soler, El sitio de Almería de 1309 (Barcelona, 1904); “Expedición de Jaime II a la ciudad de Almería,” Boletín de la Real Academia de Buenas Letras de Barcelona 2 (1903–04), 377–99; I. S. Allouche, “La relation du siège d’Alméria en 709 (1309–1310) d’après de nouveaux manuscrits de la Durra al-Higāl,” Hespéris 16 (1933), 122–38; Évariste Lévi-Provençal, “Un ‘zayyal’ hispanique sur l’expédition aragonaise de 1309 contre Alméria,” Al-Andalus 6 (1941), 377–99; Carmen María Marugán Vallvé, “El sitio de Amería de 1309: El desarrollo de la campaña militar,” in Almería entre culturas, siglos XIII–XIV: Actas del coloquio (Almería, 1990), pp. 171–86; María Desamparados Martínez de San Pedro, “Jaime II y la cruzada de Almería,” in Jaime II setecientos años después, ed. Juan Antonio Barrio Barrio, José Vicente Cabezuelo Pliego, and Juan Francisco Jiménez Alcázar (Alicante, 1996–97), pp. 579–86.

138

Manuel Rojas Gabriel

It is evident that Alfonso VII was acting according to the pragmatic guidelines dictated by the following obvious facts: he could not be everywhere at one time; he could not do everything at the same time; and he could not risk all his military operations on a single bet. Almería was a high-level target, but it was not the only one. Even though the emperor had to head the alliance against Almería, he also had to use a large part of his resources against other enemy strongpoints. These were fortresses which, once the Mediterranean harbor was under his control, would afford a degree of coherence to this new frontier sector and to the Christian lines of military action. In fact, and in spite of the reasonable doubts we might have in this respect, the operational dynamic developed by the Leonese king while he was heading for the siege of Almería explains some of the features of the Leonese-Castilian participation in this military undertaking. The emperor arrived very late at the siege, where the rest of the combined army was already waiting for him, and with limited forces, because after campaigning around Jaén he had been forced to let a great part of the host go. Was this deliberate and premeditated? Here, in the absence of reliable data, we move on to the shifting sands of mere speculation, as happens quite often when it comes to analyzing strategic conduct.58 However, given that Alfonso VII very possibly did not want to deploy the bulk of his forces in an operation which, although it could provide him with an important urban center and its dependent territories, nevertheless implied during those months a notable diversion of troops from areas of conflict where he had to fight with no external aid, the possibility must not be rejected. It is entirely probable that Alfonso VII intended to avoid human and material losses which would be difficult to replace should he have to deploy his forces in other actions. Such tactical decisions might perhaps garner less renown but, pragmatically speaking, they were essential for reaching his goals. In such a manner, with minimal forces, the emperor might reap the maximal benefits that could be obtained from joining and leading the conquest of Almería. In brief, León-Castile would gain the trophy, while the rest of the participants would provide most of the necessary resources. What had Alfonso VII been doing during the weeks preceding the campaign of Almería? His movements seem to make it fairly clear that, after the occupation of Córdoba, the king had been fighting in the upper Guadalquivir valley, that is, the sector indispensable for articulating a front-line border as solid as possible before the Almohads’ inexorable arrival in the area. As the sequence of events understood as a whole seems to prove, Alfonso VII’s offensive acts had a manifest end. There was little hope of establishing a strong defensive line in the western regions of the present-day Andalucía, Extremadura, and southern Portugal, where the Christians barely had significant strongpoints to conquer – except, perhaps, Badajoz – and then re-using these bases against the Maghrebins. By that time, and if on any occasion this project was pondered, 58

Michael Prestwich, Armies and Warfare in the Middle Ages: The English Experience (New Haven, 1996), p. 187: “Evidence for the discussion of strategy is rare, partly for the very good reason that it was often kept secret. It is necessary to rely largely upon examination of the course of events to deduce what plans had been made – a procedure that obviously has its dangers.”



Alfonso VII and Power in al-Andalus

139

taking over Seville and other nearby fortresses was an impossible task, so the monarch was focusing on the submission of Jaén and Almería. If he took both cities, perhaps later on he could thrust towards Granada, the only meaningful stronghold that was still in Almoravid hands, and which therefore was isolated from any possible assistance that might prevent the Christians from taking it. Alfonso VII still had to deal with several difficult problems. Providing logistical and human support was difficult because behind his front there was probably a huge Christian demographic vacuum. Moreover, Muslim strongpoints were scattered across the disputed regions, and taking them required time, even in the case of minor fortresses lacking external aid. Time was precisely what the king did not have. The main targets of this strategic program might be clear, but if one of those key locations could not be taken, or was recovered by the enemy, the compact front being formed to stop the Almohads would collapse like a house of cards, following from the resulting breach of territorial integrity. Indeed, for the triangular territory whose corners were Jaén–Almería–Granada to become a serious obstacle to the Almohads, it was absolutely necessary that those three places be in Christian hands. The reason was simple: if Córdoba or even Calatrava could not work as advanced operational bases for the Leonese-Castilians, those three towns and their nearest fortresses had to cooperate – for, if any of these urban nuclei was besieged by the Almohads, the garrisons of the other two were close enough to lift the siege. This was especially important because the Almohads, even more than the Almoravids, were not very skilled in poliorcetics and were suffering severe hardships when they had to take walled towns and castles which offered a determined resistance.59 Thus, in the summer of 1147, during the weeks before the beginning of the siege of Almería, it seems that Alfonso VII was trying to kill two birds with one stone. Firstly, he was distributing his military efforts in the most balanced way, employing a substantial part of his resources against the still semi-autonomous remains of Almoravid power around the headwaters of the river Guadalquivir. If, besides Almería, the emperor wanted to conquer Jaén immediately, then he needed to seize a series of nearby Muslim strongpoints, which would serve as attacking bases to isolate the city. So for most of the summer he devoted himself to occupying these secondary locations, with the aim of facilitating his actions against Jaén. It is obvious that this was not simply a campaign of aggression. Alfonso was taking the first steps that would allow him to establish a territorial wedge which would then be difficult for the Almohads to attack when they arrived in the northern regions of Andalucía. The idea was that the North Africans would be forced to get involved in a fierce war of positions, this being the worst predicament to find themselves in when attempting rapidly to expand their 59

This constitutes a historiographical cliché. On Almohad military organization see, for instance, Victoria Aguilar Sebastián, “Instituciones militares: el ejército,” in El retroceso territorial, ed. Viguera Molins, pp. 187–208, and Jean–Pierre Molénat, “L’organisation militaire des Almohades,” in Los Almohades: problemas y perspectivas, ed. Patrice Cressier, Maribel Fierro, and Luis Molina, 2 vols. (Madrid, 2005), 2:547–65.

140

Manuel Rojas Gabriel

domain over what was left of the Andalusian-Almoravid world. Furthermore, if Alfonso VII managed to subdue the Muslim sectors he was attacking, the natural road for the Almohads to follow to reach the Levantine lands would be blocked. These were zones where the Wolf King of Valencia and his father-inlaw Ibn Hamushk of Murcia were able – and willing – to join forces with the Christians at any time to fight the new Maghrebian onslaught. So, although in the eyes of his allies and vassals Alfonso VII was late in arriving before Almería in the summer of 1147, he was actually carrying out a shrewd campaign en route. His march was slow because he was exploiting the difficulties afflicting the remaining possessions of his Almoravid vassal Yahya b. Ali b. Ghaniya. As a matter of fact, if we follow the king’s steps during those summer weeks, his line of action could hardly be clearer. By mid-July, Alfonso VII was “in ripa de Godalqueuir,” besieging Andújar.60 This was an operation which must be placed in relation to the conquests of “Itilurgi” (or “Vrgi),” Baños de la Encina, Baeza and other fortified centers – some of which are difficult to identify – whose taking by force of arms is mentioned in the Prefatio de Almaria.61 At the same time, a source that usually proves to be well informed points out that Yahya b. Ali b. Ghaniya turned Baeza over to his Christian lord,62 which has been interpreted by some scholars as an attempt in extremis by the Almoravid leader to try to consolidate his alliance with the Leonese in order to counteract the Almohad advance jointly.63 However this is not a correct reading. What was really happening was that Alfonso VII was taking advantage of the Muslim leader’s weakness to act fiercely against him. Besides the fact that the Poema de Almería leaves no room for doubt about the aggressive nature of the actions undertaken by the emperor against those localities, Ibn Khaldun also confirms that the former Almoravid governor had no choice but to surrender Úbeda and Baeza to Alfonso VII, against his will.64 This circumstance is corroborated by a contemporary document.65 But Alfonso VII’s pressure produced an unexpected reaction from Yahya b. Ali b. Ghaniya. The emperor probably thought that, whatever the circumstances, the Almoravid leader would hand over his few but strategically important posses-

60

61

62 63 64 65

Luis Sánchez Belda, Documentos reales de la Edad Media referentes a Galicia. Catálogo de los conservados en la Sección de Clero del Archivo Histórico Nacional (Madrid, 1953), no. 245, pp. 118–19. The Prefatio (also known as El Poema de Almería) mentions the conquests of Andújar, “Vrgi,” Baños, Bariona, and Baeza during the progress of the Leonese-Castilian host to Almería. Juan Gil, its most recent editor, identifies “Vrgi” with Iliturgi, a locality situated between Linares and Andújar, and offers varioius identifications for “Bariona,” such as Bailén or El Ballón, near Montoro: Prefatio de Almaria, pp. 297–316. Anales Toledanos 1, ed. Flórez, p. 389. For example, Recuero Astray, Alfonso VII, Emperador, p. 259, and Derek Lomax, The Reconquest of Spain (London, 1978), p. 91. Ibn Khaldoun, Histoire des Berbères et des dynasties musulmanes de l’Afrique Septentrionale, trans. William MacGuckin (baron de Slane), ed.Paul Casanova, 4 vols. (Paris, 1969), 2:187. Peter Rassow, “Die Urkunden,” (1930), no. 28, pp. 99–100.



Alfonso VII and Power in al-Andalus

141

sions. His reckoning was wrong. The demand for the surrender of Córdoba, the Christian monarch’s aggression toward his territories, and an increase in parias provoked the Almoravid to ask for the protection of the only power with enough military force to provide it: the Almohads. By mid-1148, Yahya b. Ali b. Ghaniya proposed to the new Maghrebian power the secession of Córdoba and Carmona in exchange for keeping Jaén. The Almohads might have been religious fanatics, but they were not lunatics, and they accepted the arrangement immediately.66 For the time being, this did not divert Alfonso VII from applying his strategic planning: in the summer of 1148 he attacked Jaén, but the operation failed.67 This was an evident turning point in the race between the Christian political entities – and particularly León-Castile – and the Almohads to annex as much territory as possible in al-Andalus. The obviously favorable situation the Christian kingdoms had enjoyed with the fall of the Almoravid empire and the growth of various petty kingdoms was quickly fading. The Almohads were imposing their rule over a great part of Andalusia, which meant that many of the Muslims were gathering again under a single authority. From a political-military perspective, this implied that al-Andalus was again becoming a solid bloc. Stage Three: Collapse (1149–1157) Save for Granada and the Balearic Islands – a petty kingdom which was obviously beyond Alfonso VII’s reach – it seemed that by the end of 1148 the Almoravids had already passed into Andalusian history, and that the Leonese king was only a step away from confronting the Almohads. The emperor was not entirely alone: the Muslim rulers of Valencia and Murcia refused to bow to the authority of the new North African power. So, for the third time in a short period – although with fewer options each time – Alfonso VII remodeled his strategic program to try to adapt to the new circumstances. One of the expressions of this plan is familiar to us: it involved forging arrangements with Ibn Mardanish and Ibn Hamushk, as it was not necessary to have a great talent in political-military matters to conclude that the best possible way of facing the Almohads was to add together as many military forces as possible and to try to redirect the Africans’ attacks to the eastern areas of the peninsula – that is, as far away as possible from the sectors which were vital to the Leonese-Castilians in the core regions of Iberia. Indeed, while independent Andalusian leaders survived, compromising Almohad claims of unity, the Christians could conclude pacts with them and 66

67

Ibn Khaldoun, Histoire des Berbères, 2:187. Cordova surrendered to the Almohads in 1148; Ibn Abi Zar’, Rawd al-qirtas, ed. and trans. Ambrocio Huici Miranda, 2 vols. (Valencia, 1964), 1:382–83. In late August, the Almohad emir ‘Abd al-Mu’min sent a letter to Yahya b. Ali b. Ghaniya inviting him to surrender, since he had been informed that Yahya was prepared to accept the Almohad regime: Évariste Lévi-Provençal, “Un recueil des lettres officielles almohades. Étude diplomatique et historique,” Hespéris 28 (1941), 1–80 at no. 4, pp. 23–24. It is analyzed in detail by Simon Barton, “A Forgotten Crusade: Alfonso VII of León-Castile and the Campaign for Jaén (1148),” Historical Research 73 (2000), 312–20.

142

Manuel Rojas Gabriel

so use them to their own benefit. This circumstance, furthermore, meant that the Berbers had to devote part of their energies to trying to subdue them, thus employing resources they could not use to attack Christian domains. At the beginning of 1149, barely a month after Yahya b. Ali b. Ghaniya’s death,68 when the interesting alternative of exploiting the last vestiges of the Almoravid regime was a thing of the past, Alfonso VII met Ibn Mardanish and Ibn Hamushk.69 Both petty kings were eager to reach any kind of arrangement with a Christian power that could help them sustain their fight against the Almohads. Indeed, before this encounter, the Wolf King had sought the aid of Pisa, and a few months later of Genoa: in January 1149 the Muslim king had signed a ten-year treaty with the Pisans, to whom he granted factories in Valencia and Denia and, in June, he signed an almost identical pact with the Genoese and agreed additionally to respect the latter’s trading stations in Almería and Tortosa.70 Conversely, both taifas were perfectly aware of the fact that they were also valuable allies for the Christians, and that their collaboration was vital as they had become the only noteworthy hubs of Andalusian anti-Almohad resistance, and they were thus unwilling to embrace anyone for nothing in exchange. Of course, the best of their options was Ramón Berenguer IV, who by then had already conquered Tortosa, was about to undertake an offensive to take Lérida, Fraga, and Mequinenza, and consequently wished to avoid any Muslim interference on his southwestern flank.71 Alfonso VII was therefore unable to establish any kind of agreement with Ibn Mardanish, who commited himself to pay parias to the Catalan-Aragonese.72 This was hardly good news from the Leonese-Castilian point of view, but was still perhaps the lesser evil as, after all, Ramón Berenguer IV was a direct vassal of the Leonese monarch. Yet the fight against the Muslims was not the only concern Alfonso VII had to face. Medieval monarchs, besides leading their armies on campaign, had to 68

69

70

71 72

Yahya b. Ali b. Ghaniya withdrew to Granada and tried to convince the last Almoravid redoubt that opposed the new African invaders that the struggle should cease and they should allow their fellows from the other side of the Strait to absorb them. In January 1149 he died in that city; Ibn Abi Zar’, Rawd al–qirtas, ed. Huici Miranda, 1:382–83. On Grenade as the last strongpoint of resistance against the Almohads see, for example, Rafael Gerardo Peinado Santaella, “De la conquista musulmana al reino nazarí (711–1232),” in Historia de Granada, 4 vols. (Granada, 1987), 2:230–33. In addition to the titles cited in n. 18, on the relationships between the Wolf King and Castile, see Ignacio González Cavero, “Una revisión de la figura de Ibn Mardanish: Su alianza con el reino de Castilla y la oposición frente a los Almohades,” Miscelánea Medieval Murciana 31 (2007), 95–110. In January 1149, the Muslim king had signed a ten-year treaty with the Republic of Pisa, granting that city factories in Valencia and Denia, and in June he signed an almost identical pact with the Genoese, and agreed in addition to respect the latter’s trading stations in Almería and Tortosa: Michele Amari, I diplomi arabi del R. archivio fiorentino (Florence, 1863), pp. 239–40; Codice diplomatico, ed. Imperiale di Sant’Angelo, pp. 247–49. See García-Guijarro Ramos, “Reconquest and the Second Crusade,” pp. 249–54, and the literature quoted there. Julio González, El reino de Castilla en la época de Alfonso VIII, 3 vols. (Madrid, 1960), 3:886.



Alfonso VII and Power in al-Andalus

143

rule and manage their realms, and those tasks not only diverted their attention from maintaining hostilities, but also prevented them from spending as much time as was necessary near the fronts. Although León-Castile’s bureaucratic and financial structures had reached a certain level of sophistication in the first half of the twelfth century, their machinery was not yet sufficiently developed to allow “administrative kingship,” that is, routine management that could function without constant royal supervision.73 Nor was Alfonso VII an “Andaluz” king, as Fernando III and, above all, Alfonso X or Sancho IV were, who spent most of their time in the south, next to the frontier with the Islam. It is true that, as Bernard F. Reilly remarks, “Of the thirty-one years of the reign of Alfonso VII, only seven were marked by the absence of a major campaign in which the monarch personally participated.”74 But this also meant that the king was waging war throughout the year without respite, the result being that when the situation on the frontiers became truly dangerous between 1149 and the first months of 1150, due to the fact that the Almohads had got rid – at last – of most of the local entities that stood between them and the Christians, the emperor was obliged to devote himself urgently to domestic affairs instead of fighting the North Africans, or even preparing to do so. He was in fact unable to return to the struggle against the Maghrebins until the spring/summer of 1150.75 By then, time was already inexorably running out for the Leonese king, who had to accept that the happy days of 1145–47, intense and promising as they had been, were over. His strategic targets may still have still been the same, and focused on the same areas, but from 1150 his military achievement in the south went into a clearly descending curve. The Almohads were providing those parts of al-Andalus which were already under their control with the kind of unity not seen since the high days of Almoravid rule. In that year the Leonese-Castilians tried to take Córdoba in a surprise attack. The defenders could not be subdued, and the siege had to be lifted despite the defeat of a relieving Muslim host.76 Loyal to his strategic program, in August 1151 Alfonso besieged Jaén again while, theoretically, some “Frankish” ships were supposed to attack Seville simultaneously.77 But the campaign failed.78 The emperor kept trying to maintain the initiative wherever possible before the Almohads could definitely go on the offensive. His best option was to 73 74 75 76 77

78

On this concept, see C. Warren Hollister and John W. Baldwin, “The Rise of Administrative Kingship: Henry I and Philip August,” American Historical Review 83/4 (1978), 867–905. Reilly, León-Castilla under Alfonso, p. 213. Ibid, pp. 109–11. Anales Toledanos 1, ed. Flórez, p. 390; Ibn Khaldoun, Histoire des Berbères, 2:188. Anales Toledanos 1, ed. Flórez, p. 390. The date of a document of August 4 1151 is glossed “quando imperator iacebat super Gaen, expectando naues francorum que deberant uenire ad Sibillam”: Colección documental del Archivo de la Catedral de León, vol. 5 (1109–1187), ed. José María Fernández Catón (León, 1990), no. 1470, pp. 261–63. Alfonso VII perhaps expected some extra-peninsular help which never materialized. Even in March 1152, the royal chancellery remembered the unsuccessful siege of Jaén: Manuel Villar y Macías, Historia de Salamanca, 3 vols. (Salamanca, 1887), 1: Appendix 8, p. 232.

144

Manuel Rojas Gabriel

approach Ibn Mardanish, giving him in effect a blank check for harassing the North Africans. During the fruitless siege of Jaén, the emperor and the Wolf King met in Lorca, and as a result of the commitment they made there worked together in the conquest of Guadix.79 From this act of war we can deduce that the notion of threatening Granada – which was, incredibly, still in Almoravid hands – had not been totally abandoned. Failing that, it might be possible to obstruct the way to Almería should Granada finally fall quickly under Almohad control. Alfonso VII’s strategic plan was essentially a logical military project, and practically that which Ibn Mardanish would keep attempting to realize for almost the next two decades, until his demise in 1172. Nevertheless, 1155 marked the beginning of the definitive collapse of Alfonso VII’s strategic program of exploiting the political-doctrinal divisions within Islamic Spain in order to counter the arrival of the Almohads. In March of that year he managed to wage a limited campaign in the mid-Guadalquivir valley, where he took Andújar and other minor centers in the Sierra de Córdoba, such as Pedroche, and Santa Eufemia.80 But events quickly proved these victories to be ephemeral: later that same year the new Almohad governor of Córdoba reconquered Pedroche, Montoro, and Almodóvar del Río. At the same time the governor of Seville undertook a fruitful punitive expedition against Portuguese lands which penetrated so deep into the interior of the Portuguese kingdom that he even razed the fortress of Trancoso, in the present-day region of Beira Interior. Of greater significance than anything else, however, was that Granada, the last Almoravid redoubt in the peninsula, finally surrendered to the Almohads, materially and psychologically unable to maintain firm resistance without future prospects.81 Official propaganda, manifested in the intitulations of documents from the royal chancellery, claimed that at the end of 1156 Alfonso VII was at the height of his power. He was king of Galicia, León, Castile, Nájera, Zaragoza, Toledo, Almería, Baeza, and Andújar.82 In some of these places the titles reflected a genuinely imperial dignity, but in the last three cases the reality was very soon to change. With Granada in their hands the Almohads embarked upon the conquest of Almería. The city could not endure the attack and fell in the summer of 1157. Neither the rescue organized by the Leonese monarch nor 79

80 81

82

Mariano Gaspar y Remiro, Historia de Murciana musulmana (Zaragoza, 1905), p. 207. However, the activities of the Wolf King began to cause some unease in the Almohad court. As was his custom, in September 1153 the caliph ‘Abd al-Mu’min sent him a letter in which he urged him to submit: Lévi-Provençal, “Recueil des lettres,” no. 10, pp. 31–32. Anales Toledanos 1, ed. Flórez, p. 390; Peter Rassow, “Die Urkunden,” (1930), no. 56, pp. 134–35. Ibn ‘Idari al-Marrakusi, Al-Bayan al-mugrib fi ijtisar ajbar muluk al-Andalus wa al-Magrib, ed. and trans. Ambrosio Huici Miranda, vol. 1 (Tetouan, 1953), pp. 302–10; Ambrosio Huici Miranda, “Un nuevo manuscrito de al-Bayan al-Mugrib: Datos inéditos y aclaraciones sobre los últimos años del reinado de Alfonso VII, el Emperador,” Al-Andalus 24 (1959), 63–84. José Antonio García Luján, Privilegios reales de la catedral de Toledo (1086–1462), 2 vols. (Toledo, 1982), 2: no. 21, p. 64; Colección documental, ed. Fernández Catón, no. 1492, p. 291.



Alfonso VII and Power in al-Andalus

145

the profit-seeking support provided by Ibn Mardanish to retake the city proved sufficient to the task, and during his withdrawal the emperor lost Baeza and Úbeda as well. He died before reaching Toledo.83 It was a sad end for a king who had once been just a step away from reaching the strategic goals he had set to himself in his struggle against Islam. Most painful of all, however, must have been to witness how, in just a couple of years of offensive dynamism, the Almohads had stripped him of all the triumphs he had achieved in the northeastern areas of Andalucía. That, for a warrior of Alfonso VII’s caliber and his rank, was surely almost unbearable. Some Conclusions At first sight events in 1157 seemed to be repeating those at the end of the eleventh century, when the Almoravids burst on to the Iberian peninsula to impose their regime within al-Andalus. It seemed once more that a great African power would decide the development of the struggle between Christendom and Islam in Iberia. However, stricto sensu, the circumstances were now perceptibly different. The kingdoms of León and Castile separated after nearly a century and a quarter of political unity, the former going to Fernando II (1157–88) and the latter to Sancho III (1157–58). Union had provided enough resources and energy to make León-Castile the greatest Christian power of Hispania and had therefore facilitated, whenever possible, the initiation of aggressive action against the Muslims. Furthermore, union had given Castile rather than León a hegemonic role, largely due to the conquest of the taifa of Toledo and the widely expanded front which was a Castilian responsibility and encompassed territories from Portugal to the Levant. This reduced the possible sectors for Leonese advance to a narrow strip in the present-day Extremadura, toward Coria–Cáceres–Mérida/Badajoz.84 This division between Alfonso VII’s sons and heirs was perfectly natural. However, from the viewpoint of expansionist or defensive strategies vis-à-vis the Muslims, the appearance of two polities where previously there was one complicated matters. Until the definitive union of both kingdoms took place in 1230, each pursued its own political and military interests, often competing against each other either openly or in a veiled manner. Already in 1158 they became embroiled in a brief war, and although this was ended by the Treaty of Sahagún (23 August 1158)85 dynastic problems diverted 83

84 85

Lucas de Tuy, Chronicon mundi ab origine mundi usque ad eram MCCLXXIV, ed. Andreas Schottus, Hispaniae Illustratae 4 (Frankfurt, 1608), p. 105; Rodrigo Jiménez de Rada, Historia de rebus Hispaniae sive Historia gothica, ed. Juan Fernández Valverde, Corpus Christianorum, Continuatio medievalis 72 (Turnhout, 1987), pp. 232–33; Ibn ‘Idari al-Marrakusi, pp. 311–12; Ibn Abi Zar‘, Rawd al-qirtas, ed. Huici Miranda, 1:386–87, though the wrong the year is given; Évariste Lévi-Provençal, “Recueil des lettres,” no.16, pp. 41–43. See the Maria Dolores García Olica article in the present volume, and the literature quoted there. Julio González González, “Fijación de la frontera castellano-leonesa en el siglo XII,” in Estudios en memoria del Profesor D. Salvador de Moxó, ed. Miguel Angel Ladero Quesada, 2 vols. (Madrid, 1982), 1:411–24 at 419–20.

146

Manuel Rojas Gabriel

their energies from fighting against the Berbers. In the following years, discord between Castile and León was common. Nonetheless, 1157 also saw a set of positive cirumstances which made things very different from the time when the Almoravids had appeared in 1090. In the first place, colonizing activities and territorial organization had made the Tagus and Ebro frontiers virtually invulnerable to any offensives the Almohads might undertake. The North Africans mounted great predatory expeditions and inflicted much suffering, but they could not take important strongholds by siege. It has become a commonplace for scholars to note that the unitary empire was as incapable of keeping great armies operational over long periods as it was capable of recruiting them in the first place. Secondly, the Almohads found it difficult to impose Muslim unity, enabling Ibn Mardanish, for example, in connivance with the Christians, to resist until 1172. Nor did they ever fully exploit such military successes as they did enjoy against their northern rivals. By 1157 the Christian leaders had moreover learnt a lesson with respect to the end of the eleventh century: this time, instead of wearing down the Wolf King, in the manner adopted at Zaragoza and Valencia when the Almoravids occupied al-Andalus, they decided to support and use him against the Almohads through an alliance. Nobody doubted, of course, that the Africans constituted a formidable enemy, certainly much more so than an al-Andalus fragmented into petty kingdoms. But it is also true that they were not the Ummayad caliphate; nor did they have the vigor that for decades characterized the Almoravids. Indeed, although it is tangential to the aim of this paper, we can posit the rule that each North African danger the Hispanic kingdoms had to confront – from the Almoravids through the Almohads to the Marinids – was successively less powerful than the previous one. Meanwhile the Christian societies had reached a level of internal development, maturity, and dynamism which allowed them to cushion more efficiently each blow from Africa. Thirdly, since the fitna which ended in 1031 with the disintegration of the caliphate of Córdoba, the Christian entities in Iberia had created and perfected a series of direct and indirect political-military methods, designed to defeat the successive Muslim adversaries they had to face, which proved to be extremely efficient. In 1157, when Alfonso VII died, these strategic formulae were already well established thanks to their regular implementation. So, whether each individually or through temporary pacts, normally arranged in order to achieve a particular goal or to combat a specific threat, the Hispanic kingdoms already knew by the mid-twelfth century how to defeat the Muslims, even though at first sight the enemy showed notable internal cohesion and great military might. And one of the reasons for their success is to be sought, beyond the intrinsic expansionistic aggressiveness that was typical of the Christian societies – what José María Mínguez calls “the activating components of conquest”86 – in the inherent nature of the Muslim societies. 86

Mínguez, La España, pp. 239–42.



Alfonso VII and Power in al-Andalus

147

Military functions were almost exclusively the responsibility of state structures and therefore barely impinged on the social fabric in comparison with their northern rivals. On the other hand they had a persistent tendency toward political fragmentation for ethnic, doctrinal, or sociological reasons.87 Because of these and other factors it was only a matter of time before the African empires or the Andalusian petty kingdoms began to show grievous symptoms of internal weakness. Such signs were inmediately exploited by the Christians. Of course, in the Middle Ages the wherewithal to maintain an uninterrupted military effort, even in the most favorable circumstances, was limited. Thus Christian advances into disputed areas were never either linear or sustained, save for short periods during which the Islamic front simply collapsed due to the combination of pressure exerted from the north and internal problems, as in 1145–47. As Derek Lomax explains (and in this way I pay a particular tribute to this British historian), the Christian conquest (the traditional idea of “Reconquista”) “was not, as is often implied, a slow and gradual one. The Christians did not advance steadily, step-by-step; they took great leaps forward.”88 Taking advantage of the rapid Almoravid breakdown and the division of Andalucía into petty kingdoms between 1145 and 1157, Alfonso VII tried to make one of those leaps, rushing from the Tagus frontier to the northern regions of today’s Andalucía. The fact that this plan was premature, because of the arrival of the new Maghrebian lords and the lack of depth on his front, does not mean that its strategic logic was wrong. It was virtually the same strategy as successfully executed by Fernando III in the second quarter of the thirteeth century. Too often historians – influenced by “pluralistic” notions of crusading and overrelying on a few lines composed by a single chronicler writing far from the location of the events he described – have interpreted Alfonso VII’s political-military actions between c. 1145 and 1157 – particularly the conquest of Almería in 1147 – as military operations forming part of the Second Crusade.89 If doctrinal matters are left aside (which, by the way, would make for another interesting debate), this interpretation suffers from at least two weaknesses. In the first place, the impression is given that the Christian acts of war in Iberia must have been stimulated at a specific moment in line with the guidelines established by the Holy See and its agents, and moreover that the goals reached must have been attained thanks to the collaboration of foreign contingents. This approach disregards the prolonged tradition of offensive and defensive military operations undertaken by the many native political entities which had long been developing their own strategic guidelines to try to defeat their Muslim enemies. In the second place, 87 88 89

On this topic, a good interpretative synthesis is Felipe Maíllo Salgado, De la desaparición de al-Andalus, 2nd ed. (Madrid, 2011). Lomax, Reconquest, pp. 175–76. On the concept of “Anglolexia” – as the late Timothy Reuter called it – and its problems, see Manuel Rojas Gabriel, “Introducción: De historiografía y otras calamidades,” in La conducción de la guerra en la Edad Media: Historiografía y otros estudios, forthcoming.

148

Manuel Rojas Gabriel

the “crusader” interpretation pays insufficient attention to the events of the long decade examined here; the conquest of Almería must be viewed in relation to the aggressive strategy undertaken by Alfonso VII against al-Andalus because of a Muslim political conjuncture which had started a few years before. In other words (to make a counterfactual case), if the Second Crusade had not been proclaimed, the emperor would still have acted in the same way against the Muslims and, equally, Genoa would have forced León-Castile to attack that Mediterranean harbor. It is in terms of the strategic logic of the various Christian powers, including the Leonese-Castilians, that this set of events must be analyzed, as they set about exploiting the acute if temporary political fragmentation into which al-Andalus had sunk, and subsequently attempting to forge a coherent territorial front line against the Almohad avalanche. This makes much more sense, and accords more with the facts of the case, than a preoccupation with ultra-Pyrenean crusades. In actual fact the Iberian war against Islam was waged by the Christian kingdoms, not as pan-European crusades, though Iberian kings were quite ready to make use of foreign warriors – provided these were under their control. The goals of such wars were not of a crusading nature, but a direct consequence of the political and military situation peculiar to this theater of operations. It is true that the campagins undertaken by Alfonso VII from 1145 onwards and, especially, the conquest of Almería in 1147, coincided chronologically with the Second Crusade, and that in the taking of that city forces which did not belong to León-Castile intervened. However, if we had the opportunity to ask the Leonese monarch if his acts of war and political maneuvers against Islam were part of a major “international” undertaking, driven by different and distant forces, he would surely have been perplexed. The emperor would surely respond that his strategies were executed in the interests of his realms. His efforts might or might not have been beneficial to the whole of Latin Christendom, but, above all, he was the heir to a long secular native tradition of war against Islam. He might have added that he did not even know who the Wends were, or where Edessa and Damascus lay.

Map 1: The Iberian Peninsula in 1157

7 The Treaties between the Kings of León and the Almohads within the Leonese Expansion Strategy, 1157–1230 María Dolores García Oliva

The main object of this paper is to analyze how the kings of León used diplomacy in order to make possible the territorial expansion of their kingdom at the expense of Muslim lands, a subject hitherto practically ignored by historians. It is well known that León had less potential for expansion than Castile and Portugal. The ambitions of these three adjacent kingdoms converged on Muslim territory, a fact that would cause countless armed conflicts between these Christian entities throughout the period in question. An appraisal of the diplomatic relations established by the Leonese attests that the truces they concluded with the Almohads allowed the kings of León to keep their area of expansion open despite the actions of their direct rivals. In the same manner, during the last stage of León’s existence as an independent kingdom, its leaders finally focused on territorial conquest. As we shall see, for those campaigns, the kings of León relied quite often on the aid or cooperation of their neighboring kingdoms, thanks to a series of agreements reached with the respective rulers of those states. I Censorious judgments regarding the treaties, pacts, and alliances concluded between the Christian kingdoms and the Muslims are not common in the historiography of recent decades. Nonetheless, such remarks have not completely disappeared from the copious writing on the topic. For instance, for Manuel González Jiménez, the agreements between the kings of León and the Almohads were “alianzas contra natura” – unnatural alliances.1 Maria Alegria Marques and João Soalheiro describe as an “aliança espúria” – an illegitimate alliance – the pact arranged between Sancho VII of Navarre (1194–1234) and the caliph ‘Abd Allah al-Nasir (1199–1213).2 Pascal Buresi’s view is slightly more nuanced, recognizing a difference between the Christian–Muslim treaties signed in the early years of the thirteenth century and those concluded previously, the last 1 2

M. González Jiménez, “Reconquista y repoblación del occidente peninsular,” in Actas das II Jornadas Luso-Espanholas de História Medieval, 4 vols. (Porto, 1987), 2:455–89 at p. 464. M. A. Marques, and J. Soalheiro, Corte dos primeiros reis de Portugal: Afonso Henriques, Sancho I, Afonso II (Somonte-Cenero, 2009), p. 23.

152

María Dolores García Oliva

of which was between Sancho VII of Navarre and the Almohads, from 1198 to 1201. In positing such a distinction, he notes that in the earlier arrangements made among the Christian kings it was established that they should respect the truces which some of them might already have made, and that consequently they would not be able to aid other monarchs in their fights against the Almohads. However, this proviso was substantially modified in the treaties signed during the first decades of the thirteenth century, by conceding that the fighting men of the signatories, even when their lords had truces with the Muslims, would be allowed to help the Christian king with whom the arrangement was established. Buresi quotes as an example of this situation the armistices concluded by Alfonso IX of León (1188–1230) with Fernando III of Castile (1217–52) in 1218, and with Afonso II of Portugal (1211–23) in 1219, which allowed Castilian and Portuguese warriors, whose kings were maintaining truces with the Muslims at the moment of the signature, to join the expeditions undertaken by the Leonese king against al-Andalus. Buresi believes that the absence of this option in the previous treaties – those of the last decades of the twelfth century – demonstrates that the Christian and Muslim kingdoms were then considered equals, and this impels him to conclude that the pacts of that time were “unnatural alliances,” concurring in this with González quoted above.3 This view seems to be inspired by the notion of Reconquista, which views fighting against the Muslims as a task to be shared by all the Christian kings, and also by the ideal of the crusade, whose doctrinal corpus encouraged Christians to merge in the struggle against the infidel, tolerating no other connection with them but one of confrontation. It is a perspective that would appear to be shared by Stephen Lay, as might be deduced from the value judgments that at times attach to his account of the performance of the Christian kings. Thus, regarding the truce signed by Sancho I of Portugal (1185–1211) with the Almohads in 1191, he remarks that the Portuguese kings, like the rest of the Hispanic monarchs, had no qualms about negotiating with the infidels.4 The comparison permits us, perhaps, to glimpse a certain condescension toward the conduct of the Portuguese monarch, seen as merely behaving in the manner of the other Christian leaders. It would appear to be in this spirit that Lay justifies the alliance between Afonso Henriques (1139–85) and the Almohads in terms of the actions of Fernando II of León (1157–88). He relates that in 1169, when the Portuguese king and Giraldo Sempavor – “the Fearless” – had Badajoz under siege, Fernando II led his troops to the city, not in fact to aid the besiegers, but to make sure that Badajoz would remain under Muslim control, by allying himself with the Almohads and thus acting in concert with such “unlikely allies.” Far from feeling uneasy about dealing with the Almohads, Fernando II presented his maneuver as a triumph and, two years later, he reaffirmed their alliance. In 3 4

P. Buresi, La Frontière entre chrétienté et Islam dans la péninsule Ibérique: Du Tage à Sierra Morena (fin XIe–milieu XIIIe siècle) (Paris, 2004), pp. 244–45. S. Lay, The Reconquest Kings of Portugal: Political and Cultural Reorientation on the Medieval Frontier (London, 2009), p. 160.



Treaties between León and the Almohads

153

these circumstances, Afonso Henriques had no choice but to conclude truces with the Almohads, notwithstanding the calamitous consequences that such a course would have for the Christian defense, as the Castilians would find themselves isolated against the Almohad menace.5 Lay goes on to explain that, at the beginning of his reign, Alfonso IX, due to the enmity between himself and the Castilian king, accepted the truces offered by the Almohads despite the hazard of such a settlement for his uncle Sancho I of Portugal – given that the Saracens, who had also arranged truces with Castile, once rid of those adversaries could deploy their troops against the Portuguese king.6 The contemporary Christian chronicles tend to not report the arrangements made between the Christian kings and the Muslims or, when they do, generally adduce some kind of justification, or refrain from making value judgments. The exception is the Crónica latina de los Reyes de Castilla, in which this attitude is not adopted with regard to Alfonso IX: its presumed author Juan of Osma condemns the alliance between the Leonese king and Muslims. However, such a judgment can be explained by Juan’s animosity toward the monarch – an open animosity made manifest throughout his works. Thus, with regard to the alliance between Alfonso IX and the Almohads in 1196, as a result of which the former obtained soldiers and economic assistance to attack Castile, Juan states that the king had only sought cooperation with the caliph Yusuf al-Mansur (1184–99), induced by bad counselors, because of his resentment toward Alfonso VIII (1158–1214), who had knighted him in Carrión and refused to grant him certain castles he had claimed. According to the chronicler, Alfonso IX had considered the ceremony a humiliation, as it demonstrated the superior status of the Castilian king. The Leonese king’s association with the Almohads, in which the Navarrese were also involved, although in the chronicle this is alluded to only indirectly, is labeled by Juan a “coalición de impiedad” (coalition of impiety) and, furthermore, as designed to destroy Castile. Juan de Osma’s bias in favor of Alfonso VIII, on the other hand, becomes apparent when he goes on to note that truces were eventually signed between the Almohad caliph and the Castilian king, yet abstains from any value judgment in this regard.7 Similarly, the Castilian–Almohad truce signed in 1214 is presented as closely associated with the famine prevailing at that time, especially in the Transierra8 and the Extremadura,9 thus justifying the concord with 5 6 7 8

9

Ibid., pp. 137–38. Ibid., p. 154. Crónica latina de los Reyes de Castilla, ed. Luis Charlo Brea (Madrid, 1999) [hereafter Crónica latina], 14, p. 40 and 15, p. 42. The term “Transierra” refers to the southern slope of the mountain range known as the Spanish Central System and its adjoining territories. Thus, in the following pages references to the “Leonese Transierra” and the “Castilian Transierra” allude to those territories within the area in question that belonged to each kingdom. In its origins, the toponym “Extremadura” referred to the territories situated immediately beyond the south bank of the river Duero. As in the case of “Transierra” explained above, the expressions “Leonese Extremadura” and “Castilian Extremadura” allude to those territories

154

María Dolores García Oliva

the infidels. For this reason too Alfonso VIII had been forced to lift the siege of Baeza, begun with great effort as he had not entirely recovered from a severe illness. It is revealing therefore that Juan de Osma did not recognize the logistical hardships inherent in both siege and famine as applying to the Leonese kingdom, whose own Extremadura and Transierra must have been affected by the latter. On the contrary, he asserts that Alfonso IX, whom the king of Castile had to urge not to hinder his attacks against the Muslims, refused to maintain the offensive against them after the conquest of Alcántara in 1214 (an achievement that Juan attributed moreover to don Diego López de Haro, a vassal sent by Alfonso VIII with six hundred men-at-arms to help the Leonese king).10 Similarly, the chronicler shows no uneasiness about the alliance of Fernando III of Castile with ‘Abd Allah “el Baezano,” with his brother Abu Zayd, king of Valencia, or with Abu l-‘Ala, governor of Seville, who proclaimed himself caliph in 1227. Such an attitude may be explained by el Baezano and Abu Zayd having become vassals of Alfonso VIII and, more importantly still, the Castilian king’s support for el Baezano having brought him Martos, Andújar, Salvatierra and other places in al-Andalus, while in exchange for the truce with the caliph he obtained a significant amount of money, as the chronicler informs us.11 Nevertheless, the fact remains that it was the Christian ruler Fernando III who participated in these arrangements with the Muslims. Jiménez de Rada is less partisan in his account, and he avoids making any value judgment. Thus he relates that Alfonso IX had used Muslim troops when he attacked Castile in 1196, having sought the alliance with the Almohads, but he does not condemn this coalition at any point. He also says that Alfonso VIII, after the third campaign led by Yusuf al-Mansur against Castile and after the battle of Alarcos, “established for some time a truce with the king of the Arabs, in order to be able to focus on his neighbors.”12 Jiménez tried to suggest that Alfonso VIII’s lack of response to the Muslim offensive was due neither to cowardice nor to a possible military inferiority. He also alludes to the truce signed by the Alfonso with the Almohads in 1214, justified by the famine being suffered in Castile, but he does not mention the pacts between Alfonso’s successor Fernando III and the Andalusi leaders, perhaps because they did not fit with the reason for which he claims the king decided not to prolong the peace with the Muslims. According to Jiménez, the king decided to take up arms against the infidels and to offer to God the first profits of his fight: it would be difficult to regard such an attitude as compatible with an alliance

10 11 12

within the zone in question which were controlled by those kingdoms. However, the toponym soon acquired an indeterminate sense when the Christian kingdoms expanded their domains to the south. In the context upon which this paper is focused, the term “Leonese Extremadura” refers both to the southern territories of León and to the lands located further south within the Muslim area. Crónica latina 26, pp. 56–57. Ibid., 45, 46, 47, 53 at pp. 76–79, and 84. Jiménez de Rada, Historia de los hechos de España, ed. and trans. Juan Fernández Valverde (Madrid, 1989) 7.20, p. 301.



Treaties between León and the Almohads

155

or cease-fire with those same enemies. Only with respect to the acquisition of Baeza, Andújar, and Martos might we conceive – although misleadingly – of the existence of some kind of covenant with el Baezano, as Jiménez says that Alfonso VIII “gathered again his armies and, being as they were granted to him by Avomahomat, who was the noble Prince of the Arabs, […] took ownership of Baeza, Andújar, and Martos.”13 Within the first paragraph quoted above, Jiménez de Rada also notes Alfonso VIII’s interest in concluding truces with the Muslims in order to pacify one of his main fronts, a goal of a special importance when he had to face open conflicts with more than one adversary. Around 1196, Castile was fighting against Navarre and León, and to make peace with al-Andalus would allow its king to focus his resources on those opponents. The chronicler is conscious of the fact that the Christian kingdoms – independent political entities – disregarded religious differences and concluded pacts with the Muslims according to their circumstances or particular interests. It is clear that the Almohad empire was just another political entity to keep in mind within the game of peninsular inter-monarchical relations, whether pacific or hostile. However, the conclusion of these agreements did not imply renouncing, even temporarily, expansion into Andalusi territories. On the contrary, such agreements could be part of a strategy designed to achieve that very goal. Nowadays it is certainly recognized that military pressure was not the only means used for extending boundaries because, as Francisco García Fitz has pointed out, in line with recent Anglophone historiography on medieval warfare, “in the same way that the application of military might generate direct territorial gains or, depending on the case, weaken the opponent, political relations with the Muslims – alliances, treaties, or any other kind of arrangement, more or less temporary – sought the same end.”14 Manuel Rojas meanwhile has stated that “it is worth considering the relations of a political nature between opposing parties in conflict as an integrated part of the strategy the adversaries could develop.” In Rojas’s view, the notion of “total war” was unknown to medieval strategy due to the fact that institutional and socioeconomic structures did not allow the gathering of sufficient human, financial, and logistic resources to prolong a war sine die. For this reason, the deployment of a long-term or, to use his own expression, “grand” strategy had necessarily to combine both military and diplomatic resources in order to achieve the annexation of new territories.15 From this perspective the arrangements made between the Leonese kings and the Almohads make absolute sense, as they form part of a Leonese strategy designed for the expansion of their territories at the expense of the Muslims. Such a strategy was 13 14

15

Ibid., 9.12, p. 344. Francisco García Fitz, “¿Hubo estrategia en la Edad Media?: A propósito de las relaciones castellano musulmanas durante la segunda mitad del siglo XIII,” Revista da Faculdade de Letras-História, 2nd s., 15 (1998), 837–54 at 844. Manuel Rojas Gabriel, “De la estrategia en la Batalla del Estrecho durante la primera mitad del siglo XIV (c. 1292–1350),” in El Siglo XIV: El Alba de una Nueva Era (Soria, 2001), pp. 223–69 at 229.

156

María Dolores García Oliva

conditioned fundamentally by two factors: the first related to the demographic situation of the meridional area of the Leonese kingdom and the adjoining Andalusi territories, and the second to the politico-military relations of León with the other Christian kingdoms, especially with Castile and Portugal, its direct rivals in the expansion to the south. Regarding the conquest of Muslim lands, León’s starting point was less favorable than those of the other Christian kingdoms. Both Portugal and Castile held significant positions in advance of the Leonese ones. Afonso Henriques had taken Lisbon and Santarem in 1147, before the death of the emperor Alfonso VII (1126–57), and was concerned to maintain the momentum of the conquest of Muslim enclaves in the interests of securing Portugal’s recently-acquired independence, an objective which boosted its expansionist dynamism. Castile, despite the Muslim offensive, had managed to hold on to the northern area of the kingdom of Toledo, including its capital city. This was the most heavily populated zone, and its annexation had increased Castilian demographic, economic, and therefore military power. Thus Castile had displaced León from its preeminent position, as is proven by the fact that Sancho III, the emperor’s firstborn, inherited Castile and not León, despite the latter kingdom being the more ancient. Fernando II’s domains were also close to the Tagus, as Coria, conquered by Alfonso VII in 1142, had remained under his jurisdiction. However, the territory in which his demographic and economic bases were located lay far from the Guadiana river basin, where the concentration of population enabled rulers to enforce effective political authority. A wide sector of León’s southern area was poorly populated by the middle of the twelfth century. Thus, while repopulation had progressed to the north of the river Tormes, the western area and the sierras of the Leonese Extremadura remained barely developed, still with a low population density nearly a century later. In fact, the most important base of operations was Salamanca, a city fairly distant from the Muslim domains. And as regards the Transierra, the region presented an impression of emptiness little changed from that of previous centuries. Coria certainly does not seem to have produced a significant city council with the authority to control its territory after its definitive conquest by Christian forces, the changes of jurisdiction the city underwent during the 1160s perhaps being indicative of its slow development, as these would seem to reveal the difficulties experienced in establishing solid power-bases.16 This weakness of the Leonese positions in the Transierra and in 16

The church of Santiago de Compostela, to which Alfonso VII had already granted a third of the city of Coria, received the remaining two thirds in 1162; in 1168 Fernando II recovered the city in order to give it to the Templars, but shortly afterwards it returned to the royal domain: G. Velo Nieto, Coria: Reconquista de la Alta Extremadura (Cáceres, 1956), pp. 186–87, 190–93. For the repopulation of the area of Salamanca, see L. M. Villar García, La Extremadura castellano-leonesa: Guerreros, clérigos y campesinos (711–1252) (Valladolid, 1986), pp. 260–61; A. Barrios, “El poblamiento medieval salmantino,” in Historia de Salamanca: Edad Media, 5 vols. (Salamanca, 1997), 2:264 ff.; and for the repopulation of the Transierra, José Luís Martín Martín, “La repoblación de la Transierra (siglos XII y XIII),” in Estudios dedicados a Carlos Callejo Serrano (Cáceres, 1979), pp. 477–97 at 483; M. D. García Oliva, “En torno al pasado



Treaties between León and the Almohads

157

the middle area of the Extremadura contributes to an explanation of the failure of the attempt to impose Leonese dominion over Alcántara and Cáceres. Fernando II was possibly dragged into this in order to preserve his expansion route to the south, threatened as it was by Portugal from the west; but such a venture was premature, and therefore failed, as both cities were recovered by the Almohads in the campaign of 1174 against the kingdom of León.17 To the inauspicious situation, as regards undertaking a solid offensive against the Muslim positions, of the Leonese southern territories must be added that pertaining beyond the borders, characterized by weak demographic growth between the Tagus valley and the Guadiana river basin. This – a factor contributing to the loss of the cities just mentioned – deprived any potential conqueror of the infrastructure to facilitate the establishment of a new power. Two sets of events may be taken as illustration. First, the conquest of Coria in 1142 by Alfonso VII’s troops: for while it is true that this was eased by the Almoravid empire’s internal decay, it can be specifically explained by the absence of any great centers in the vicinity from which the Muslims could aid the city by lifting the siege. The people of Coria capitulated because of the lack of a response to their request for relief, and upon hearing the news of Coria’s fall the people of Albalat did not even try to summon reinforcements, opting to leave the fortress in the sure knowledge that they had been left to their fate.18 The second sign of the feebleness of the Muslim positions in the area may be observed in the deeds of Giraldo Sempavor (mentioned above), who in less than two years took Trujillo, Cáceres, and Montánchez, among other husûn (plural of hisn, Arabic for castle or fortification).19 According to Eva

17

18 19

medieval de la Transierra Occidental,” in Actas del I Congreso sobre A fala ed. A. Salvador Plans, González J. Carrasco, and Oliva M. D. García (Mérida, 2000), pp. 31–50 at 37–39. The basic works of reference for tracking the political events are Julio González, Regesta de Fernando II (Madrid, 1943); Alfonso IX (Madrid, 1944); idem, El reino de Castilla en la época de Alfonso VIII. 1: Estudio, 3 vols. (Madrid, 1960); idem, Reinado y Diplomas de Fernando III. 1: Estudio, 2 vols. (Córdoba, 1980); Ambrosio Huici Miranda, Historia política del imperio almohade, 2 vols. (Tetuoan, 1956); L. Suárez Fernández and F. Suárez Bilbao, “Historia política del reino de León (1157–1230),” in El reino de León en la Alta Edad Media IV: La monarquía (1109–1230) (León, 1993), pp. 214–350, a work which closely follows Julio González’s monographs about the Leonese kings; Manuel Terrón Albarrán, “Historia política de la Baja Extremadura en el período islámico,” in Historia de la Baja Extremadura. 1: De los orígenes al final de la Edad Media (Badajoz, 1986), pp. 283–556. More recently, Francisco García Fitz has dealt with this aspect in Relaciones políticas y guerra: La experiencia castellano-leonesa frente al Islam. Siglos XI–XII (Sevilla, 2002). That is how the facts are narrated, at least, in the Crónica del emperador Alfonso VII, ed. Maurilio Pérez González (León, 1997), pp. 114–15. The narrative sources do not entirely agree when it comes to enumerating Giraldo Sempavor’s conquests. Thus, Ibn Sahib al-Sala informs us that he took over Trujillo, Cáceres, Montánchez, Serpa, and Juromenha, conquests to which must be added Lobón, given that, according to the same writer, this fortress was taken from his men by the Almohads: Al-Mann bil-imama, ed. and trans. Ambrosio Huici Miranda (Valencia, 1969), pp. 137–39, 187. However, Ibn ‘Idari, though following the latter narrator, includes Évora but omits Serpa: Al-Bayān al-mugrib: Nuevos fragmentos almorávides y almohades, ed. Ambrosio Huici Miranda (Valencia, 1963), pp. 402–04. In the Christian chronicles we can find a greater diversity: The Chronicon Lusitanum attributes

158

María Dolores García Oliva

Lapiedra, before selling his services to the Portuguese king, Giraldo would have been a slave-soldier of Galician-Portuguese origin, under Muslim orders. As such he would have gathered valuable information about the places to be attacked, the surveillance routines of their guards, and would also perhaps have learned their language: knowledge that would help explain the ease of his conquests.20 It might also be worth taking into consideration the idea that the fortresses Giraldo captured no longer any great importance, and hence were scarcely guarded – a circumstance that would surely contribute to the success of the frontero.21 With regard to politico-military relations with the rest of the Christian kingdoms, development of royal power in the period under scrutiny must be taken into account, as the inter-monarchical struggles inherent to it absorbed most of these states’ military resources. Nonetheless the very progress made in strengthening the positions of the Christian kings in itself doubtless nurtured the great Christian offensive against Muslim territory from the second decade of the thirteenth century, which in the case of the Leonese meant reaching the Guadiana river valley. The process was shared with the other kingdoms and, as Esther Pascua Echegaray has demonstrated, one of the mechanisms that contributed decisively to the strengthening of royal as against noble jurisdictional power was military conflict with other political entities and the subsequent agreements, truces, and treaties. The conflicts among Christian kingdoms began to subside in the second decade of the thirteenth century, as royal power became increasingly established. Such fights often arose from border rivalries, fed by the fact that the territorial delimitation of the different kingdoms was in an “assembly stage,” as was royal power, and was hindered by the hierarchical organization and overlapping of jurisdictions of the time, and the lack of any absolute coincidence between ascription of a territory and the vassalage ties of those who held its jurisdiction. There was nevertheless some progress in the fixing of boundaries between the different kingdoms as royal power solidified, and by the same means: war, and above all settlements signed after armed confrontations, or treaties approved in the face of the threat of war. These were arrangements of great fragility, on account of the low level of institutionalization and highly personal nature of political relations.22

20 21 22

the conquest of Évora to Giraldo, and to Afonso Henriques the conquests of Mora, Serpa, and Alconchel: Chronicon Lusitanum, ed. Enrique Flórez, in España Sagrada 14 (Madrid, 1905), p. 415. The Crónica latina relates that, after Fernando II”s troops and the Almohads had relieved Badajoz from the Portuguese siege, Giraldo surrendered to Fernando Rodríguez de Castro, in exchange for his freedom, the strongpoints of Montánchez, Trujillo, Santa Cruz de la Sierra, and Mofra (Monfra – Monfragüe? – in the Latin text), suggesting thus that these last two strongholds were also under his control (Crónica latina, p. 36). Eva Lapiedra Gutiérrez, “Giraldo Sem Pavor, Alfonso Enríquez y los almohades,” in Bataliús: El reino taifa de Badajoz, ed. Fernando Díaz Esteban (Madrid, 1996), pp. 147–58 at 152–53. The Spanish expression frontero alludes to the figure of an independent or semi-independent military leader who operated along the frontier of one or several politico-territorial entities. Esther Pascua Echegaray, Guerra y pacto en el siglo XII: La consolidación de un sistema de reinos en Europa Occidental (Madrid, 1996) is a work of special interest for the period



Treaties between León and the Almohads

159

Thus, successive armed conflicts occurred intermittently between the various parties, just as alliances between them were continuously shifting too, given that any modification of the internal or external situation of a political entity could trigger a “cascade effect” and redraw the map of inter-monarchic coalitions, with a view either to achieving a new balance or reaping as much benefit as possible from the current state of affairs. Likewise, although the constant mutability of relations between the different agents – between friendship and hostility – gives an impression of oscillating policy, reacting to specific situations or guided by improvisation, in fact it amounted to nothing but changes in tactics, while always aimed at the same end. This is the context in which the strategy of the Leonese kings is inscribed, oriented to the enlargement of their realms at the expense of Muslim territories. In addition, it is essential to bear in mind the extent to which their interests were directed against the same targets as those of the neighboring Christian kingdoms, also engaged in the southward extension of their borders. This rivalry made the task of the Leonese leaders more laborious. Indeed, the overlapping nature of their aims meant that the strategies designed by each of the contenders with a view to certain Muslim territories was apt to suffer interference from the actions of opponents aiming to neutralize or counteract such measures of their adversaries as could hinder their own expansionist interests, or driven by the possibility of participating in the advantages derived from a rival’s actions, or even, where possible, of manipulating those advantages to their own benefit rather than that of the actions’ instigators. This same political game includes, of course, the Almohads, who resorted selectively to the arrangement of truces with one or several Christian kingdoms in order to be able to focus their offensive capacity against those who had stayed out of the deal, and who might aid an ally against a third party so as to contribute to its weakening, in an attempt both to obstruct the expansionist tendency of all of them and at the same time, again where possible, to recover lost territories. II The factors outlined above explain the delay of the Leonese territorial advance until the early decades of the thirteenth century, and it is therefore essential to take them into account when it comes to analyzing the strategy of León’s leaders for assuring their participation in the annexation of Muslim lands. Their “grand strategy” was aimed at obtaining the Guadiana river valley, the occupation of which would later make it possible to sustain territorial expansion toward the Mediterranean coast. However, it seems clear that for some time the kings of León had insufficient resources to annex and subsequently maintain a territory so far from their economic and social bases. Thus, initially, their operational

under review, both for its general approach and for its particular references to the peninsular Christian kingdoms.

160

María Dolores García Oliva

strategy was to keep the front for expansion open, as it were: preventing Portugal and Castile, with their more advanced positions and greater power to expand, from absorbing certain areas east and west of their southern domains, and from connecting their borders in a pincer movement which would confine León by separating its frontiers from Muslim territory, bringing to an end its expansionist ambitions. It is this goal which explains the tendency to establish cordial relations with the Almohads at this early stage. Nevertheless, in the last decades of Alfonso IX’s reign, when the population of the frontier sectors had begun to grow once more, the Almohad empire had reached a stage of irreversible decline, and discord among the Christian kings had decreased, we can identify the beginnings of a new and clearly offensive phase, during which the assistance of military reinforcements from other kingdoms was sought. Two different substages of Leonese operational strategy may be distinguished, each characterized by features of the politico-military performance enacted in its course. In the first stage, which lasted until the second decade of the thirteenth century, the actions of the kings of León were focused, as indicated above, on keeping open their potential area of expansion. With that purpose in mind, to begin with Fernando II envisaged the elimination of the kingdom of Portugal, his younger, and therefore less consolidated, rival, through an alliance with the hegemonic kingdom of Castile, with which he also arranged the distribution of the future conquered Muslim territories. The agreement materialized in the Treaty of Sahagún, signed with Sancho III (1157–58) in 1158, a pact that allowed the Leonese king to settle on his throne and to recover the lands taken by the Castilian monarch after an irruption into his domains which had occurred days before. Among the points of the treaty was, as mentioned above, the delimitation of their respective expansion areas within Muslim space, with the western sector located between Montánchez, Mérida, and Seville – a city that would be divided and shared by the two kingdoms – and the Atlantic Ocean, including Lisbon, belonging to León. Fernando II acknowledged his brother’s right to seize the rest of al-Andalus but at the same time secured the acceptance of his own rights by Sancho, the only adversary he would have once Portugal had been annihilated, as the arrangement also contemplated the conquest of that kingdom and its distribution between the two kings once annexation was completed.23 Fernando II was perhaps even more ambitious than the Treaty of Sahagún suggests. Although Fernando committed himself to sharing the Portuguese kingdom with his brother (who would choose which of two shares to take after the Leonese monarch had divided the conquered lands), such a promise was the necessary compensation for involving Castile in the elimination of the western rival. However, the fulfillment of that promise, if necessary, could be held in abeyance. If we consider that, in the event of the conquest being accomplished, the Leonese domains would be interposed between Castile and its new

23

The document has been published in Julio González, El reino de Castilla en la época de Alfonso VIII. 2: Documentos (1145–1190) (Madrid, 1960), pp. 79–82.



Treaties between León and the Almohads

161

acquisitions, regardless of the way in which the distribution was made, and that León would emerge strengthened by the annexation, it is not unreasonable to suspect Fernando II of harboring the idea of adding the Portuguese lands in their entirety to his realm. The project of eliminating Portugal – perhaps excessively ambitious and, like any other plan, with no absolute guarantee of success – constitutes the clearest evidence of the Leonese king’s political outlook. Sancho, conscious of the expansionary potential of his rivals, strove to ensure that his own aspirations would not be choked off before they could be pursued. Though ambitious, the treaty was indisputably rational: given the hegemony of Castile, it would be a far greater challenge to compete with her in the advance southward than it would be, having joined forces with her, to undertake the obliteration of the other opponent, Portugal, before it had properly consolidated. Moreover, successful implementation of the plan would open the possibility of Leonese enlargement in an unsuspected way and, better still, at the expense of a rival. However, on 31 August 1158, Sancho III passed away, and his death profoundly changed the situation in which the agreement in question had been forged. The political crisis that erupted on account of Alfonso VIII’s minority not only deprived Fernando II of the expected aid for annihilating Portugal, but also diverted his attention to Castilian issues. The new situation was thus a great opportunity for Afonso Henriques, whose independence was soon acknowledged by the Leonese king, de facto at least, as may be deduced from the meetings the two held in Cabrera in 1158 and Celanova at the end of 1159.24 Fernando II’s intervention in Castile was used by Afonso Henriques to harass the adjoining Leonese territories. The Castilian attacks against the Leonese frontier line seemed to require the concentration of the military resources on that front, so Fernando II tried secure peace on his western front by signing the so-called “Paz del Lérez” with Afonso Henriques, sealing their new friendship with the marriage of the king of León to Urraca, the only daughter of the Portuguese monarch. Given that this treaty was signed on 30 April 1165, it is possible that the actions of Giraldo Sempavor also influenced its arrangement, as these forced Fernando II to attend to his southern boundary. Hesitation arises in this regard because, according to Ibn Sahib al-Sala’s account, Giraldo’s first conquest (of Trujillo) took place between 15 April and 13 May of that year; so, if this occurred during the early days of the corresponding lunar month, and the news arrived punctually at the Leonese court, it could have accelerated the peace negotiations with Portugal.25 Fernando II did not lead his forces against Muslim territory until 1166, but we cannot be certain as to whether or not the expedition was delayed for a year. 24

25

J. González, Regesta, pp. 37–8, 41. In the record of the meeting in Celanova, Fernando II proclaims himself “hispaniarum rex,” a fact that, as Luis Suárez thinks, may indicate that he considered himself first among the peninsular Christian kings: Suárez Fernández and Suárez Bilbao, “Historia política,” p. 238. Al-Mann, ed. Huici Miranda, p. 137.

162

María Dolores García Oliva

The suspicion that it may have been arises from the fact that during the summer of 1165 troops were gathered around Ciudad Rodrigo, a recently-created base of operations in the western sector of the kingdom from which attacks against the Muslims and the Portuguese could be launched. It seems, however, that news arrived of a Castilian offensive from the east: Nuño de Lara, who had custody of Alfonso VIII, was heading to Medina de Rioseco in order to recover from there the fortresses Fernando II had occupied along the Castilian border. This news forced the king to change his plans and the army departed from Ciudad Rodrigo in order to lay siege to Medina.26 The objective of the already-mentioned expedition of 1166 against al-Andalus was Alcántara, a choice which permits us to appreciate the lucidity of Fernando II’s geopolitical vision if we consider the context in which the Leonese offensive was launched: that is, the occupation of Trujillo, Évora, Cáceres, Montánchez, and Juromenha, among other strongholds, by Giraldo Sempavor, and of Moura and Alconchel, at the very least, by Afonso Henriques.27 All these areas were located within the natural expansion zone of the Leonese kingdom, so Fernando II’s initial reaction might seem oddly modest, as he attacked only Alcántara. It should be borne in mind, however, that this last was a strongpoint of a vital importance for keeping his front for expansion open, in that its possession guaranteed control over the only local passage across the river Tagus. This had become the main route linking Fernando’s kingdom with the south because the so-called “Calzada de la Guinea” – Guinea Road – had become devalued as an expansion route owing to its establishment as the dividing line between Castile and León. At the same time, the occupation of Alcántara would obstruct Portuguese eastward expansion through this sector, a principal factor behind Fernando’s interest in conquering a strongpoint his conduct shows he considered to be of special importance. While he was besieging Alcántara, Fernando Rodríguez de Castro asked for his help in defending Toledo, which was under siege by the Castilian king and his supporters, but Fernando did not respond to his vassal’s request: a clear sign of abandonment of the attempt to hold the city, due no doubt to the fragility of his power within that region, but also telling evidence of his overriding interest in assuring control of the expeditious route to the south. It is probable that financial difficulties, added to the pressure exerted on the eastern front by the Castilians, contribute to an explanation of the Leonese king’s moderate reaction to his rivals’ recent gains. Nonetheless, we must also suspect that Fernando II considered it a waste of his depleted forces to attempt to take back some of the strongholds Giraldo had conquered, being aware of the fragility of those conquests due to the lack of a solid demographic base that would enable effective control of the territory. To this must be added, moreover, the fact that these actions had been undertaken by a gang

26 27

Suárez Fernández and Suárez Bilbao, “Historia política,” p. 253. See note 16.



Treaties between León and the Almohads

163

of fronteros, acting very much independently and in the hope of acquiring a large degree of autonomy in imposing their authority over the conquered lands, and whose bonds with any established power were nothing if not lax, susceptible to modification at the fronteros’ convenience.28 This helps explain why Giraldo’s achievements, detailed in the Muslim historiography, go nearly unnoticed in the Christian narrative sources: neither Lucas de Tuy nor Jiménez de Rada provide information about his deeds. The Crónica latina de los Reyes de Castilla only mentions his capture after his unsuccessful attempt to conquer Badajoz, while the Chronicon Lusitanum simply records his conquest of Evora and its granting to the Portuguese king, apprising us thereafter of the places occupied by the latter.29 However, we do know that in 1169 Giraldo and his men, with the assistance of Afonso Henriques and his forces, besieged the city of Badajoz and managed to break through the first defences, compelling the population to take shelter within the citadel, where they held out awaiting the arrival of relieving troops. This finally provoked a reaction from Fernando II, quite rapidly as it happened, as he was apparently in Galicia when he received the unfortunate news, and rushed immediately to frustrate the besiegers’ conquest of the city. According to Lucas de Tuy, the vigorous intervention of the Leonese king reflected the fact that the occupation of the city by the Portuguese king would have infringed Fernando’s claimed rights of conquest.30 But that alone can hardly explain the speed of his reaction since, as noted above, the capture by Giraldo or by the Portuguese king of other sites located within the Leonese area of expansion had not provoked such a response. In this case, the importance of the city itself must be considered, as it was indeed the main center of that sector of the Guadiana river basin, and a zone that had besides a denser population, which would facilitate the implantation of power structures. If the Portuguese king captured Badajoz, he could easily gain access to the east through the Guadiana river valley, seriously threatening if not destroying completely the Leonese expansion project. The presence of Afonso Henriques himself in support of Giraldo constituted a clear sign of the city’s prime importance, as his cooperation during the siege not only contributed to accomplishing the mission, but also reflected his interest in preventing the warlord from taking the city independently. 28

29 30

Pedro Gomes Barbosa, “Organização defensiva na fronteira Beirã oriental: ‘Extrematura’ e Riba Cõa até ao século XIII,” in IV Jornadas luso-espanholas de História Medieval: As Relacões de fronteira no seculo de Alcañices, 2 vols. (Porto, 1998) 1:199–212 at p. 200. Similarly, Huici Miranda propounds a hypothesis based on the idea that Fernando II considered Giraldo as acting independently of the Portuguese king and, for that reason, did not think that the latter’s conquests would prejudice his own rights (Historia política, p. 233). Crónica latina, p. 36; Chronicon Lusitanum, ed. Flórez, p. 415. The expression used by the chronicler is “civitatem maurorum ad regem Fernandum de iure spectantem.” Lucas de Tuy, Crónica de España, ed. J. Puyol, 2nd ed. (Madrid, 1926, repr. Valladolid, 2007), p. 404; Jiménez de Rada equally connects Fernando II’s intervention with the fact that Badajoz was included in his expansion zone (Historia, p. 292).

164

María Dolores García Oliva

Another familiar circumstance of the area under contention raised the stakes for both monarchs still further: the city of Mérida had been the head of the archdiocese of Lusitania, a status later transferred to Compostela in 1120. Its conquest, therefore, interested both Portugal, as the bishoprics created to the south of the river Duero depended – theoretically at least – on the Lusitanian metropolitan, and León, as its occupation by any other political entity would jeopardize Santiago’s archiepiscopal status, which could in turn mean the subordination of the bishoprics located within Leonese territories to sees situated in other kingdoms. Nevertheless, Fernando II was well aware of being in no position to seize the territory in question with any confidence in being able to hold on to it. In the event of an attack, whether Portuguese or Muslim, even were this not to coincide with another struggle along his borders, he would find it very difficult to aid the territories under threat, due to the vast, virtually uninhabited space lying between these and the domains from which he could send help of any kind. The depopulated state of the intervening area was unlikely to be amenable to change in the short term, as it seems that there was not enough of a demographic surplus in León to repopulate such an extensive territory. For this reason, Fernando’s intervention sought not so much to conquer Badajoz, as to avoid the city’s falling into Portuguese hands. Unquestionably, the best way of safeguarding his future chances of expansion was to ensure that the city remained under Muslim authority. The assistance Fernando II provided seems to be the result of the alliance concluded with the caliph, which was bowdlerized from the Christian chronicles that were closer to the events. Thus, Lucas de Tuy reports that, when the king of León lifted the siege of the city laid by Giraldo and Afonso Henriques, its inhabitants bestowed it upon him and they committed themselves to be his vassals – a promise that, as was only to be expected of infidels, they soon broke, transferring their allegiance to the Almohads. Jiménez de Rada reports these events in a similar fashion, although he places the commitment to vassalage after a second intervention of the Leonese king.31 However, Ibn Sahib al-Sala, a contemporary historian of the events, who was very well informed, tells us that Fernando II concluded a pact with the Almohads in 1168, thanks to which he obtained military aid against Nuño de Lara and his allies, who had custody of Alfonso VIII of Castile, and who were striving for control in Castile. The Muslim warriors spent five months cooperating with Fernando II in his offensive against Castile and, once they considered the campaign was finished, the Leonese king ratified the peace he had signed with their government; and they arranged with him that, when he heard of a Christian expedition, he would head for the country of the Amir al-Mu’minin […] and that he would go against that enemy with them to repel [the expedition] and defend them with his protection […]32 31 32

Lucas de Tuy, Crónica, pp. 404–05 and Jiménez de Rada, Historia, pp. 292–93 Al-Mann, ed. Huici Miranda, p. 136.



Treaties between León and the Almohads

165

The chronicler then immediately reports that the king honored his commitment and succored the citizens, as he would detail later on, and indeed, he goes on to explain that Fernando II, after defeating the besiegers, remained faithful to his oath and left the city to its inhabitants, refusing even to enter the citadel unless it should be by order of the caliph, to whom he acknowledged the fortress belonged.33 This story, even if it hardly gives a true account of events, derives from an impeccable logic. Through the alliance with the Almohads, the Leonese king received military aid against one of his adversaries, Castile, and by cooperating with the Muslims he hindered Portuguese access to the east via the Guadiana river valley. When Giraldo and his men persisted in harassing Badajoz, Fernando II returned in 1170 to prevent them from taking the city, and there he came together with a host led by Abu Hafs, a son of the caliph, with whom he ratified once more his alliance with the Almohads.34 The overall character of the pact is not known, but it does not seem that Fernando II obtained any territory from the Muslims by diplomatic means. Ibn Sahib tells us that, after the king’s first intervention in Badajoz, he received “in favors and rewards, a prize equivalent to the graciousness of his pact that he himself had not even imagined,” this being a generic allusion that Ibn Sahib promises to detail later on and whose particular content is known thanks to Ibn ‘Idari al-Marrakusi, an author who closely follows him and who informs us that the “prize” Fernando II received was a sort of pallium covered with gemstones.35 The capture of Afonso Henriques and Giraldo Sempavor during the fighting at Badajoz, however, did bring territorial benefits to Fernando II and some of his supporters. Lucas de Tuy and Jiménez de Rada assert that in exchange for his freedom the Portuguese king returned Limia, Toroño, and other places he had – unfairly – withheld, without specifying which those places were; the Crónica latina echoes this report, adding that Giraldo, in exchange for his deliverance, surrendered Montánchez, Santa Cruz, Trujillo, and Monfragüe to Fernando Rodríguez de Castro.36 The inclusion of this last information in the Castilian chronicle may be influenced by the fact that the nobleman was a native of Castile, or to the fact that those places were located within the area of expansion claimed by that kingdom, most of them coming under Alfonso VIII’s dominion some time later. However, the sources omit to say how Fernando II attained Cáceres, Alburquerque, and the rest of the fortresses situated along the Guadiana river valley, which he subsequently granted to the Church and the Order of Santiago between 1170 and 1171. Three different hypotheses have

33 34

35

36

Ibid., p. 145. Ibid., p. 156. With regard to Fernando II”s second expedition to Badajoz, Julio González had already stated that the king did not mean to add the city to his domains due to the huge distance between regions (Regesta, p. 88). Ibid., p. 144; Ibn ‘Idari al-Marrakusi, Al-Bayan al-mugrib fi ijtisar ajbar muluk al-Andalus wa al-Magrib 2.1, ed. Ambrosio Huici Miranda, in Colección de crónicas árabes de la reconquista, 4 vols. (Tetuoan, 1952–55), pp. 14–15. Lucas de Tuy, Crónica, p. 405; Jiménez de Rada, Historia, p. 292; Crónica latina, p. 36.

166

María Dolores García Oliva

been advanced regarding the acquisition of Cáceres. The first is that the city was conquered by the Leonese forces on their way back from Badajoz.37 The second is that it was gained through a pact with the Almohads or in appreciation for the assistance given, which would imply that the city had been recovered by the Almohads either by force of arms or after lifting the siege of Badajoz.38 Both speculations are distinctly implausible: in the first case, due to the weakness of the Leonese position in that area at that time; and in the second, due to the fact that there is no proof the Muslims recovered the fortresses under control either of the Portuguese ruler or his paladin after their defeat. There is a third possibility: that the cession of Cáceres to the Leonese king was a consequence of the agreement made with Afonso Henriques for his liberation, and therefore that Fernando II received the city from the Portuguese.39 Regarding the acquisition of Alburquerque, M. A. Ladero Quesada argues that Fernando II took it over, along with Cáceres, during the expedition of 1169, when he procured the submission of Badajoz; García Fitz believes, however, that Fernando II acquired the city during the campaign undertaken a year later, since he granted the city to the church of Santiago in December 1170.40 This donation indeed indicates that Fernando II had taken over the area by then – but that does not necessarily imply that he had come into its possession immediately before the grant was made. It is plausible that he made the grant during the second expedition to Badajoz, judging by the distrust that is said by Ibn Sahib al-Sala to have been provoked by this new march of the Leonese king into Almohad terri37

38

39

40

Antonio C. Floriano Cumbreño, Estudios de historia de Cáceres (desde los orígenes a la Reconquista) (Oviedo, 1957), p. 122. Although correcting the erroneous date, Floriano Cumbreño probably had in mind the chronicles of Alonso de Torres y Tapia, who asserts that Fernando II conquered the city in 1169 on his way back from Santarem, where he had gone to aid Afonso Henriques, under attack from the Almohads, an event that did not occur until 1184: Crónica de la Orden de Alcántara, new ed., 2 vols. (1763, repr. Salamanca, 1999) 1:74. Francisco de Rades y Andrada also indicates that the city was conquered by the king in 1171: Crónica de Santiago, in Chronica de las tres ordenes y cauallerias de Sanctiago, Calatraua y Alcantara, new ed. (1572, repr. Valencia, 1997). Leopoldo Torres Balbás, “Cáceres y su cerca almohade,” Al-Andalus 13 (1948), 446–72 at 451–52; his opinion is shared by Gervasio Velo y Nieto, Castillos de Extremadura (tierra de conquistadores): Cáceres (Madrid, 1968), p. 170. Julián Clemente Ramos also links the occupation of Cáceres with the assistance given to the Almohads, although he does not declare explicitly that the city was granted by them to the king. (“La Extremadura musulmana (1142–1248): Organización defensiva y sociedad,” Anuario de Estudios Medievales 24 (1994), 647–701 at 650–51). José Mattoso and Armindo de Sousa do not refer to the way in which Cáceres was obtained by Fernando II, but they do aver that the fortress fell back into Almohad hands after the siege of Badajoz: A Monarquia Feudal (1096–1480), in História de Portugal, 8 vols. (Lisbon, 1993–95), 2:78. José-Luís Martín, Orígenes de la Orden Militar de Santiago (Barcelona, 1974), p. 6; Derek W. Lomax, “La fecha de la reconquista de Cáceres,” Archivos Leoneses 23 (1979), 309–19 at 309; Lay, Reconquest, p. 183. M. Á. Ladero Quesada, “1135–1217. 1: Castilla y León,” in La Reconquista y el proceso de diferenciación política (1035–1217), vol. 9 of the Historia de España Menéndez Pidal (Madrid, 1998), p. 468; García Fitz, Relaciones políticas, pp. 124–25.



Treaties between León and the Almohads

167

tory: the chronicler relates that in 1170 Sayyid (“Lord”) Uthman, with Almohad and Andalusi troops, departed for the defense of Badajoz, which was once again on the verge of being taken by the Portuguese, who were taking advantage of the weakened situation of the city caused by Giraldo’s harassment. At this point Uthman was informed that Fernando II was also marching with his forces toward the city “to recover its possession from Muslim hands,” in light of which the Sayyid sent emissaries “to settle with him and learn about his departure and about whether he would respect the arrangement they had established with him or not.” The Leonese king ratified his commitment to the arrangement.41 This report is evidence, however, of a profound mistrust toward Fernando II’s actions, which might well have been the result of the Leonese king’s taking over certain strongholds – including Alburquerque and some of the fortresses he had donated to the Order of Santiago by that time – either by force of arms or, if they were nearly deserted, through their occupation. Or it might have been caused by the Almohads noticing some indication that the Leonese king was contemplating an assault on Badajoz. The Almohads, whose primary interests were focused on Morocco and, as far as al-Andalus was concerned, on the eastern region where they were confronting the resistance of Ibn Mardanis, did not seem to be capable of re-establishing control over this sector, as can be deduced from Giraldo’s conduct after his release. That same year of 1169, he defeated the force led against him by the governor of Badajoz and, in May 1170, he captured a convoy with supplies and weapons sent from Seville to supply the city. A while later the second siege of the city (already referred to above) took place. This time the Almohads managed to take Juromenha – Giraldo’s main headquarters – from him. Giraldo continued exerting pressure from Lobón, leading the Almohads to send provisions to the city in the autumn of 1171, an action that had to be repeated the following year.42 Against this background, Fernando II, fearing that the threat of the Portuguese spreading throughout the Guadiana river area would eventually be fulfilled, seems indeed to have thought about attacking Badajoz, considering the attention he paid to the territories located to the south of the river Tagus between 1170 and 1171. Between August and December 1170 he granted Cáceres to the Order of Santiago;43 in December of the same year, he endowed the church of Compostela with the city of Aramenia and the castle of Alburquerque. The church then handed the tenure of that fortress over to the Order of Santiago in February 1171, along with several assets and rights so that the Order could afford the expense that holding and defending the place, as well as the taking of other locations, would involve.44 In that year, although again on an unknown 41

42 43 44

Al-Mann, ed. Huici Miranda, pp. 155–56. The date of renewal of the pact was the month of “rabi-l-awawal, 566” which corresponds with the period between 12 November and 11 December 1170. Ibid., pp. 149–50, 153–56, 187, 233. J.-L. Martín, Orígenes, p. 47. Tumbo A de la Catedral de Santiago, ed. Manuel Lucas Álvarez (Santiago, 1998), doc. 118, pp. 244–45; J.-L. Martín, Orígenes, doc. 42, p. 214.

168

María Dolores García Oliva

date, Fernando II granted to the same military order the Albuera valley with Luchena, Cantiñana, and Monmaior castle (the fortress of Alconchel, situated to the southwest of Badajoz, “apte situm ad expugnandos Christi inimicos” – “well sited for the launching of attacks against the enemies of Christ” – according to its donation document), and Monfragüe: territory that had been given by Giraldo to Fernando Rodríguez in exchange for his liberation, as already noted.45 Frankly, I have reservations about the idea that the Order, despite its recent foundation – 1st August 1170, according to J.-L. Martín’s meticulous study of its origins – could already count on enough men and economic resources to take care of and defend all the places it had received. Regardless of this, however, the donation in itself suggests that Fernando II was willing to take Badajoz. Quite clearly, holding the positions to the south of the city, as in the cases of Alconchel, Monsmayor, or la Albuera, would be inconceivable were the city itself not under his control. If we bear in mind the date of these donations, when the Almohads, immersed in the struggle against Ibn Mardanis, had not put an end to the activities of Giraldo and the Portuguese around the Guadiana river basin, we must surely conclude that the Leonese king aspired to acquire certain positions within the area in question, since he was willing to make an attempt at conquering the city regardless how premature such an action might be. At the same time, Fernando II’s conduct in this area reveals his interest in protecting his expansion zone from Portugal, since Alconchel, Monsmayor, and Alburquerque were located in this western strip. To these positions, and slightly more to the north, by the river Tagus, Alcántara – the fortress that Fernando II rushed to conquer in 1166 at the beginning of Giraldo’s campaign – would be added shortly afterwards.46 However, in the next year, 1172, Ibn Mardanis was dead and the Almohads finally reunified al-Andalus, which obviously strengthened their position within the Peninsula. In the meantime, and after the loss of Lobón, Giraldo’s situation was gradually becoming more difficult, especially as regarded maintaining his independence. In this new political context, after responding to the Castilian offensive in 1173 with an incursion into the surroundings of Talavera and Toledo, the Almohads signed truces with Castile and with Portugal. According to Ibn Sahib, the caliph accepted the armistice “in order to repopulate the deserted country in this peninsula and because of the surprises caused by the god-damned foreigner Giraldo along his fragmented borders.”47 The author focuses on Giraldo but, by then, the warlord’s activities had run out of steam and he was no longer 45 46

47

J.-L. Martín, Orígenes, doc. 49, pp. 222–23, doc. 50, pp. 223–24, doc. 51, pp. 224–25. José-Luis Martín had already suggested that Fernando II could have used the Order of Santiago as a shield against the Almohads and the Portuguese (J.-L. Martín, Orígenes, p. 7). I have myself previously developed this approach regarding the policy followed by Fernando II this matter in María Dolores García Oliva, “Consideraciones sobre la estructura defensiva almohade y la expansión leonesa,” in La Península en la Edad Media treinta años después: Estudios dedicados a José-Luís Martín ed. José María Mínguez Fernández and Gregorio del Ser Quijano (Salamanca, 2006), pp. 159–74 at 171–72. Al-Mann, ed. Huici Miranda, pp. 135, 233.



Treaties between León and the Almohads

169

a serious threat to al-Andalus. By contrast, the advanced Leonese positions, acquired thanks to the Muslim political instability of the previous years, did indeed represent a potential hazard. For that reason, Almohad forces were led against them just a year later, in 1174. It was at that point that Ya’qub (1163–84) organized a campaign in which even Ciudad Rodrigo was besieged, aiming to destroy what was about to become a dangerous Leonese base of operations. Fernando II managed to frustrate this Muslim attempt, but the Almohads did recover Alcántara and the rest of the king’s fortresses situated to the south of the Tagus – which, incidentally, was undoubtedly the main goal of the expedition.48 Ibn ‘Idari confirms that the Almohad offensive was a response to Fernando II’s rupture of the pact with the Muslims, and this has led some scholars to conclude that the Muslim expedition was a reaction to an incursion allegedly undertaken by the king of León against al-Andalus, embarked upon by the monarch as a demonstration of his disapproval of the arrangements made between the caliph and his most direct Christian rivals.49 Nevertheless, this version given by the Moroccan chronicler seems to reflect his interest in trying to justify the caliph’s actions without damaging the latter’s image of integrity. It makes sense that Ibn ‘Idari offers no explanation of what triggered the rupture mentioned above. He also refers to the breaking of truces, rather than of their expiration, when Portuguese razed Beja and dashed toward Seville in 1178, or when a year before that the Castilians laid siege to Cuenca.50 The fact that Ibn Sahib al-Sala, in the paragraph quoted above, focused his attention on Giraldo rather than Fernando may likewise be due to the fact that at those times the truces agreed between the Almohads and the Leonese king were still in force. A point in favor of this hypothesis is that the Almohad caliph, having pacified al-Andalus and having concluded truces with Portugal and Castile by that time, was in an extremely propitious situation for launching an offensive against León and, obviously, would stand to benefit the most from breaking the armistice with the Leonese king. His ultimate end was to conquer the lost fortresses before the Leonese presence within the area could become more resistant; delay would not only make the region’s recovery harder, but would also, in the intermediate term, result in a serious menace to the northern Muslim positions. As we have already mentioned, the caliph achieved his goal, pushing the border back as far as the river Tagus in this sector. Fernando II did not try to retake the strongpoints he had lost, which is a clear sign of the hardships he would encounter not just in capturing them, but also 48

49

50

According to Ibn ‘Idari, Al-Bayan 2:15, the Muslims took back Alcántara and Nadus – an unidentified place – but, in the aftermath of that expedition, the Leonese domains were pushed back north of the river Tagus and, consequently, they lost the rest of their remaining possessions within that region. Ibn ‘Idari, Al-Bayan, p. 15; Huici Miranda, Historia política, p. 272; Terrón Albarrán, “Historia política de la Baja Extremadura,” p. 432. This last scholar, however, despite accepting Ibn ‘Idari’s version, warns us that the disreputable chronicler’s attempts to justify the Almohad offensive should not be taken at face value. Ibn ‘Idari, Al-Bayan, pp. 21, 28.

170

María Dolores García Oliva

in keeping them under his control, due to the extreme emptiness the southern area of his domains and the neighboring Muslim territories. His first offensive reaction was delayed until 1177, when he forayed into the enemy’s territory, arriving at Arcos and Jerez through the Guadalquivir river valley, and taking advantage of the siege of Cuenca laid by Alfonso VIII. This report is offered by Ibn ‘Idari, who goes on to tell us that, on his way back, the army of Seville pursued him and that these forces reached a group of Christians from Talavera from whom they not only took their booty but also their lives.51 According to Julio González, whose opinion is shared by L. Suárez, the real victims of the Muslim attack were the Leonese troops, as may be inferred from the manner in which the Muslim chronicler presents the news. Ambrosio Huici Miranda, by contrast, considers that – as the source itself indicates – it was the militia of Talavera who suffered the defeat: a force which could either have been acting in conjunction with the Leonese, or merely at the same time as them, in response to incursions they had suffered into their territory.52 Either way, what seems clear is that the Leonese king opted for a punitive expedition. Take into account the precise circumstances – which undoubtedly favored the Leonese initiative, as their adversaries had had to divert their attention to other conflict zones – Fernando II’s attack was evidently a forceful reminder to the Almohads of the threat that León represented; and such a menace could well lead to a quest for understanding. The difficulties for both sides were increasing, and each could hope to make an ally or, at least, to neutralize a rival. Thus, the Leonese action may have contributed to the signing of the truce agreed toward the end of 1178, when hostilities between Castile, Portugal and al-Andalus had been resumed. With regard to León, a rise in its belligerence along its eastern front can clearly be observed – immediately leading to the Portuguese exerting pressure on the kingdom’s western frontier. In the meantime, and as regards the contiguous Christian states, this campaign against the lands of what nowadays is Cádiz could only mean that the king had not given up his claims to the lower Guadalquivir as his expansion zone. The acknowledgement of those claims by Castile had not been confirmed after Sancho III’s death, and the Portuguese – who had been left out of the agreement – had not of course sanctioned them, and had moreover taken up arms in order to recover what the Treaty of Sahagún had taken away from them. It is no coincidence, in this regard, that the year after the Leonese expedition, in 1178, Afonso Henriques’s forces were deployed against Seville in particular, an incursion that was followed by others in 1181 and 1182 against the Aljarafe. Until then, the Portuguese troops had focused their activities on the Guadiana river valley, although it is also true that those activities had been directed toward a certain end – taking over as many strongholds as possible – and were not intended merely to weaken the Muslim positions by attrition. 51 52

Ibid., pp. 29–30. J. González, Regesta, p. 117; Suárez Fernández and Suárez Bilbao, “Historia política,” p. 270; Huici Miranda, Historia política, p. 273.



Treaties between León and the Almohads

171

For all these reasons we can conclude that the Leonese and the Almohads might be circumstantial but could not be unconditional allies, as their ultimate ends were completely opposed. Furthermore, they were not the only actors on the stage; and nor was the Christian–Muslim struggle was the only source of concern for the various peninsular political entities. These all had many interests simultaneously and, employing similar strategic mechanisms, embraced whatever pacts were most convenient in relation to the moment, the circumstances, and the priorities at stake. Thus, the Leonese tendency to come to understandings with the Muslims, at this early stage, did not prohibit the conclusion of pacts between Fernando II and other Christian kingdoms in which military actions against al-Andalus were considered. For example, in June 1183 the kings of León and Castile signed the Treaty of Fresno-Lavandera. Fernando II was at the time keeping a peace agreement with the Almohads, either because the one arranged in 1178 was still in force or because it had been renewed. However, by the treaty with Castile, the king committed himself to abandon the armistice with the Muslims, a fact that would remain secret until Christmas – perhaps because that was when the truce expired. From the date that had been set, both kings would be obliged to wage war against the Saracens and not to conclude any truce or peace agreement with them at the risk of excommunication, interdict over their lands, and loss of the castles their vassals were holding on their behalf.53 The outcome of this commitment was the siege of Cáceres, laid by Fernando II between January and May 1184, which the king was finally forced to lift after discovering that there was an Almohad army planning an expedition against the Christians and that one of the main aims of that army might be to relieve the besieged; to which must be added the weariness of his forces due to the duration of the siege. On that occasion, the Leonese king did not opt for a punitive expedition, but instead tried to capture a stronghold located in the eastern sector of his theoretical expansion zone, next to the “Guinea Road,” rather than in the western strip as in 1166 when he took over Alcántara. This choice reveals unequivocally Fernando II’s fear that Castilian expansionary dynamism would eventually cut the Leonese off from the Muslim lands, ending his chances of participation in the annexation of al-Andalus. That fear might seem unfounded, as the recently formalized deal with Alfonso VIII envisaged the simultaneous fight of both kings against the Muslims and hence recognized, implicitly at least, the right of the Leonese king to share in the conquest of Muslim territories. Additionally, barely two years previously, in March 1181, both kings had signed the Treaty of Medina de Rioseco, whereby they agreed to respect the division of the kingdoms as it had been established by the emperor, both north and south of the river Tagus,54 which theoretically meant acknowledging a Leonese expansion front. Fernando’s 53 54

J. González, Regesta, doc. 46, pp. 315–21, also published in El reino de Castilla, 2, doc. 407, pp. 701–08. J. González, Regesta, doc. 46, pp. 299–304, also in El reino de Castilla, 2, doc. 362, pp. 614–23.

172

María Dolores García Oliva

apprehensiveness is nevertheless understandable, given the Castilians’ expansionary impetus, discernible in their southward advance on both flanks or the realm: to the east, with the conquest of Cuenca, and to the west – with greater impact on León – with the retention of Trujillo, Santa Cruz, and other fortresses located south of the river Tagus which had fallen into the Castilian king’s hands. It was evident that neither Alfonso VIII nor his vassals neglected the southwestern sector. A mere month after the signing of the Treaty of Medina de Rioseco, the Council of Ávila would ensure that Alfonso VIII ratified the terms that, according to its preamble, the Council had had from the time of Alfonso VII. In this document it is made explicit that the boundary of the Council’s lands to the west extended down from Puerto Muñoz, along the Gata stream bed, to the river Alagón, which was then the dividing line until it flowed into the Tagus.55 Despite certain mistakes and inaccuracies, such as the omission of the section of the Arrago that connects the Gata stream with the Alagón, this boundary line took in Coria, no less, and reached Alcántara, where the Alagón joins the Tagus. There can be no doubt that such a border could not at the time be effectively enforced, but what is significant about the document is that it evinces the Council of Avila’s intentions toward the territory adjacent to its lands. These intentions were shared by the Castilian king, as the foundation of Plasencia by Alfonso VIII would indicate just a few years later, in 1186. I do not mean to assert that the Leonese king knew of the document under discussion, nor that he knew in 1184 that his nephew was going to found Plasencia – something that perhaps not even the founder himself could have imagined at that point; but rather to point out that Fernando II was aware of Castilian expansionary potential, and also of the general fragility of agreements. In line with the argument developed above, the siege of Cáceres seems to me a logical response to the geopolitical situation along existing the border at the time, the Leonese king trying as he did to take over a stronghold located by the Guinea Road in order to contain the pressure exerted by Castile, and to secure the Leonese presence within that sector. The stronghold must have been well defended and victualed, however, as it withstood the siege, which, as mentioned above, Fernando decided to lift when he learned that the caliph was about to send an expedition against the Christian kingdoms. The Muslim expedition may originally have been aimed at Cáceres, but in the event was directed at Santarem; and there Fernando II went to aid the Portuguese. This intervention must have been in fulfillment of his commitment to fight the Muslims contracted in Fresno-Lavandera: no other similar agreement with the Portuguese king is known of and, according to Lucas de Tuy, when Afonso Henriques knew that the Leonese king was heading to Santarem, he feared that his intention was to help the Muslims rather than to aid him against them, and he hurried to send emissaries offering peace between their kingdoms.56 However, Fernando

55 56

J. González, El reino de Castilla, 2, doc. 365, p. 629. Lucas de Tuy, Crónica, p. 405.



Treaties between León and the Almohads

173

II stood to benefit from the failure of the Muslim expedition, because if the Muslims managed to capture Santarem, their presence would be stronger within that zone of the Lower March. On the other hand, the union of the Leonese and the Portuguese military forces could force the Muslims to retreat, preventing them from prosecuting their offensive against other Christian territories – a risk from which the kingdom of León was not exempt, given its recent assault on Cáceres. As the thorough analysis of the Christian and Muslim sources carried out by Huici Miranda shows,57 the Almohads lifted their siege of the city before Fernando II’s arrival, and in their retreat made no incursion into Leonese territory, which thus did not have to endure the consequences of the transit of an enemy force. As a result of the injuries he suffered in Santarem, shortly after this episode, toward the end of June 1184, Abu Ya’qub died and A’bu Yusuf al-Mansur was proclaimed the new caliph. Fortunately for the Christians, it was necessary for him to overcome serious resistance in North Africa, a process that diverted his attention until the end of the decade. This situation was exploited especially by Alfonso VIII of Castile, who to the conquest of Alarcón in 1184 added that of Iniesta in 1186 and, in 1189, besieged Magacela for a month, plundered Reina, and reached Alcalá de Guadaira while raiding Andalusi territory. In the same way, Sancho I of Portugal, who ascended the throne in 1185, directed his troops against the Muslims and, with the assistance of some crusaders, managed to conquer Silves in 1189. Fernando II, however, was not able to exhibit a similar offensive capability; he could not take advantage of the crisis among the Almohads to return to Cáceres or to undertake any other intervention against al-Andalus, even though by then he did not face any other significant conflict with neighboring Christian kingdoms. When Alfonso IX ascended the throne, the new king had to face armed pressure from Castile in Tierra de Campos, and from Sancho I of Portugal in the eastern strip. Besides the politico-military relations of Alfonso IX with the adjacent Christian kingdoms, which absorbed nearly all his military resources, the demographic situation of the southern part of his realm, which had barely changed notwithstanding Fernando II’s efforts at fostering repopulation of the sector, must necessarily be taken into account. It was precisely that demographic weakness which was one of the main factors hampering Leonese territorial expansion, as it had been for Alfonso’s predecessors. Under such circumstances, it is understandable that Alfonso IX soon sought an understanding with the Almohads; an understanding that was also of interest to the latter, insofar as it would allow them to focus their offensive on the more militarily active of their enemies. According to Huici Miranda, in 1190, Ya’qub al-Mansur reaffirmed the truces he had signed with León and also concluded a peace agreement with Castile, enabling him to fight against

57

Huici Miranda, Historia política, pp. 290–303.

174

María Dolores García Oliva

Portugal under very favorable conditions.58 We do not know the duration of this armistice or whether the Leonese king attempted to extend it, as Ibn ‘Idari offers only general information, reporting that after the caliph’s successful campaign against Silves in 1191, “the Christian kings yielded to him and truces were signed for the glory of the Word of the Islam.” A few pages below, he narrates that in the year 588 of the Hijra – that is, from 18 January 1192 to 6 January 1193 – “the ambassadors of the Christian kings arrived to renew the pact with the Muslims,” but their demands were unacceptable to the caliph, so he dismissed them without reaching any agreement. The chronicler also relates that when the Andalusi governors came to him in 1194, “the truces with the king of Castile had already expired.”59 In the first quotation, he clearly refers to the Christian kings as a group. The usage may be precise, but it is also possible that it has a figurative sense, in light of the independent behavior of each Christian kingdom and the explicit reference to the kingdom of Castile in the latter quotation – although it is also plausible that the explicit reference to Castile reflects the fact that it was that kingdom against which the Almohads marched once the armistice was over. We lack the information to elucidate whether the king of León was unable to renew the truce, or simply did not try to do so, perhaps induced to end the peace with the Almohads by the intervention of Cardinal Gregorio de Sant’Angelo, who was sent by Pope Celestine III to the Iberian peninsula on a mission to ensure that the Christian kings joined their forces against the infidels. This papal legate dictated peace between the kings of León and Castile, the agreements being ratified in the Treaty of Tordehumos, dated 20 April 1194. Combined operations against the Muslims are not envisaged in this pact, but the resulting climate of peace nevertheless allowed for the military aid with which Alfonso IV was about to provide the Castilian king in 1195, when the latter was defeated at Alarcos. It was not long before the concord between the kings of León and Castile was broken, however, as Alfonso VIII, in spite of what had been agreed in Tordehumos, did not consent to give his cousin the strongholds he was demanding. In this context, Alfonso IX tried again to parley with the Almohads in order to form a new alliance that would benefit both parties, as it would force their common adversary – Castile – to split her military resources. Alfonso IX obtained soldiers and economic aid from the Muslims and, in the summer of 1196, in conjunction with Ya’qub al-Mansur’s campaign into the Castilian Transierra, he pushed into enemy territory through Tierra de Campos, ravaging that area.60 During this offensive the Almohads laid waste to Plasencia and recovered Montánchez and Trujillo, thus halting the Castilian advance in that sector. There can be no doubt that this result was especially profitable for Alfonso IX, as it pushed back a potential threat to his kingdom’s territorial expansion to the south. 58 59 60

Ibid., p. 347. Ibn ‘Idari, Al-Bayan, pp. 173–80. Crónica latina 14, p. 40, and Jiménez de Rada, Historia 7.30, p. 300.



Treaties between León and the Almohads

175

When the Almohads ended to their campaign against Castile, toward the end of July 1196, Alfonso VIII went on the offensive against León. This time he also turned to what we might call “spiritual weapons” in an attempt to undermine the Leonese king. Through the Castilian prelates, on 31 October 1196 Pope Celestine III promulgated the papal bull Cum renatis, in which he excommunicated Alfonso IX, authorized the archbishop of Toledo to summon a crusade against him if he maintained his alliance with the Muslims, and furthermore exempted his vassals from their ties of fidelity and obedience if he attacked any other Christian kingdom with Muslim forces. His adversary attacking him by such means, Alfonso IX too resorted to the spiritual font to counteract the effects of the initiative: in January 1197, he was girded with the belt of knighthood in front of the relics of St. James the Greater during a solemn ceremony that took place in the cathedral of Santiago de Compostela. Nevertheless, it seems that he did not terminate his alliance with the Almohads, as in the same year the pope issued another bull, Cum auctores, in which he promised to those who fought against the Leonese king the same indulgence he had promised to those who had gone on crusade to the Holy Land and at the same time authorized the king of Portugal, or any other king, to seize whatever lands they could take away from the Leonese monarch.61 Despite not breaking his alliance with the Almohads, it seems that Alfonso IX received no more military aid from them against his Christian adversaries. Taking advantage of Ya’qub al-Mansur’s second campaign against the Castilian Transierra, he took back Castro de los Judíos, which had been occupied by Alfonso VIII; but as soon as the Muslims retreated to al-Andalus, Castilian and Aragonese troops ravaged the Leonese Extremadura. The Portuguese, in the meantime, were harassing southern of Galicia. In this delicate situation, Alfonso IX went to Seville to meet with the caliph, probably intending to ask for aid against his enemies. Unfortunately for him, the caliph rejected his requests because he had already arranged a truce with the king of Castile.62 The shifting game of alliances had thus changed: Alfonso VIII, thanks to his pact with the Almohads, not only eliminated a constant threat on one of his borders, but also deprived his rival of human and economic resources that could be used against him. In this new situation, with potential Almohad aid neutralized, and enduring Castilian and Portuguese military pressure as well as the spiritual pressure exerted by the Holy See, Alfonso IX opted for peace with Castile, a peace that would be formalized through his marriage to the infanta Berenguela, Alfonso VIII’s daughter, at the end of 1197. We have no more reports of possible pacts between León and the Almohads in either the Christian or the Muslim sources. It is unclear whether other allusions to truces or friendly pacts are general, or refer specifically to Castile, the hegemonic kingdom. Thus Ibn ‘Idari relates that in the year 594 of the Hijra – 13 November 61 62

J. González, Alfonso IX, 1:83–84, and 2, doc. 102, p. 149; José Goñi Gaztambide, Historia de la bula de cruzada en España (Vitoria, 1958), pp. 101–02. Huici Miranda, Historia política, p. 378; Crónica latina 15, p. 42.

176

María Dolores García Oliva

1197 to 2 November 1198 – “The Christian kings […] sent their ambassadors to ask for peace as it had been laid out in the agreements, and it was granted to them according to the precepts of the Islamic Law”; and those truces were reaffirmed after Abu Yusuf al-Mansur’s death, in 1199, by the new caliph, ‘Abd Allah al-Nasir.63 The following references are specifically to Castile: in the year 600 of the Hijra – 10 September 1203 to 28 August 1204 – the king of Castile sent an emissary to Morocco in order to conclude certain truces with ‘Abd Allah al-Nasir; and in the letter in which he informed his governors of the conquest of Salvatierra, in September 1211, the emir says that by then “the existing peace between the Almohads and the lord of Castile was about to expire.”64 The allusion to Alfonso VIII in particular in the last quotation seems to reflect the fact that his territory had been attacked on this occasion, but it does not necessarily imply that the Almohads did not maintain peaceful relations with other Christian kingdoms, including, obviously, León. It must be remembered that in the first decade of his rule, the emir had to face several revolts in North Africa, and therefore relegated peninsular issues to second place. Under such circumstances, we can presume an interest in securing the cessation of the hostilities with Castile – the hegemonic kingdom – but also with other Christian states. Alfonso IX faced no other armed conflict with Castile from the signing of the peace between both kings in 1197 until his divorce from Berenguela in 1204, when hostilities arose once more, albeit sporadically. There is no testimony to armed confrontations between León and Portugal from 1199 until Sancho I’s death in 1211. Given that there is no evidence of Leonese offensive actions against al-Andalus during this period, especially during the years in which the kingdom was not at war with any adjoining Christian entity, we might assume that Alfonso IX remained faithful to his armistice with the Almohads during that time. I do not believe it can be asserted, nonetheless, that such inactivity was due only to the existence of peace agreements between the king of León and the Almohad emir, considering the lassitude shown by León in relation to al-Andalus until the last years of its independence. Whatever the case in this regard, according to the surviving reports it does not seem that Alfonso IX was maintaining the truce with the Almohads on the eve of the 1212 campaign of Las Navas de Tolosa. Alfonso VIII’s deep mistrust of the Leonese king in this period is evident. (This may have arisen, as Julio González has suggested, from Alfonso IX’s belief that the Castilian king was aware that the peace agreements contracted in 1206 and 1209 contained certain injustices with regard to León).65 In February 1211 the pope ordered the bishops of Toledo, Coimbra, Zamora, and Tarazona to sanction by ecclesiastical means those Hispanic kings who had truces with Castile if, during the offensive that 63

64 65

Ibn ‘Idari, Al-Bayan, p. 203. The reports about the renewal of the truces are compiled by Al-Nuwayri, Historia de los musulmanes de España y Africa, trans. Mariano Gaspar Remiro (Granada, 1917–1918), p. 243. Ibn ‘Idari, Al-Bayan, pp. 228, 265. J. González, Alfonso IX, p. 140.



Treaties between León and the Almohads

177

its king was about to undertake against the Almohads, they violated those truces and attacked his realms. Inasmuch as Pedro II of Aragon (1196–1213) had manifested his willingness to cooperate with Alfonso VIII, the other neighboring kings who could have prompted this pontifical threat were those of Navarre and León, and a subsequent document shows that it was indeed the latter that raised the pope’s suspicions: barely two months had passed when Pope Innocent III (1198–1216) sent another missive, this time addressed to the archbishops of Toledo and Santiago, in which he enjoined them to make the Christian kings maintain the peace between them and help each other in the struggle against the Saracens. The pope also ordered that if any Christian king joined the Muslims, or with them attacked the Christians, the archbishops must excommunicate him immediately and his kingdom in interdict, mentioning explicitly the king of León as the presumed potential offender.66 The document itself alludes to the possibility that Alfonso IX might ally with the Almohads to wage war against Castile, but it does not confirm that he had already agreed anything with them. Just as Alfonso VIII had feared, when the Castilian forces were busy campaigning against the Almohads, the king of León attacked his domains and recovered certain fortresses the Leonese monarch considered to belong to him. Conditions for the offensive might in theory have been more advantageous if Alfonso IX had had some sort of pact with the Almohads, as by such means he would have secured his southern border. Nonetheless, if we consider the politico-military context at that point, we might easily conclude that such a pact was not indispensable in order to launch an offensive, as no Muslim attack against Leonese territory was expected. The tranquility in the Leonese–Almohad border had not been disrupted by the Muslims since 1174, when they regained the strongholds situated to the south of the Tagus, and from 1184 onwards the Leonese had undertaken no offensive against al-Andalus. There was thus no evident reason to fear a Muslim invasion, the more so since the Almohads had to reckon with the increasing likelihood of a Castilian incursion. Furthermore, neither the narrative nor the documentary sources indicate that the Leonese king employed Muslim forces in this attack, or that he had any kind of agreement with the Almohads. All in all, it seems most likely that Alfonso IX acted alone, albeit taking advantage of the circumstances. According to Goñi Gaztambide, Alfonso VIII may have considered responding to the Leonese attack by sending his army to fight them instead of deploying it to intercept the advance of the Almohad emir; but if so, he had to abandon the idea in the face of the opposition of the kings of Aragon and Navarre.67 After 66 67

Goñi Gaztambide, Historia, doc. nº 446, p. 475 and doc. nº 471, pp. 501–02. Ibid., p. 124. The news is extracted from a letter sent by Blanche, wife of the future Louis VIII of France and Alfonso VIII’s daughter, to the countess of Champagne, Sancho VI of Navarre’s daughter. As the author, who reproduces the corresponding fragment (footnote 51, pp. 124–25) indicates, the news seems to be trustworthy given the identity of its writer who, it can be presumed, was revealing first–hand information and, would not divulge anything that might hurt her father’s honor if it were not absolutely true.

178

María Dolores García Oliva

the battle of Las Navas de Tolosa, the king focused on holding the strongpoints he had captured and, taking advantage of the instability prevailing along the frontier, on seizing other Muslim strongholds. He therefore tried to appease the other Christian kings in order to arrange with them combined and simultaneous attacks against al-Andalus – a coordinated offensive that could enable them to push the borders of Christendom southward. So reports Lucas de Tuy, who relates that after the battle of Las Navas, Alfonso VIII, instead of turning against the king of León in anger, as was expected, invited him to make peace. In order for this to happen he renounced the fortresses Alfonso IX had occupied, and even granted others to him. The peacemaking orchestrated by Alfonso VIII involved Portugal too, including Alfonso IX’s commitment to return to Afonso II of Portugal such parts of his kingdom as remained under the former’s control. The chronicler assures us that all of this was being done by the “sapientissimus” king so that, once all the kings of Hispania had been appeased, he could unite them against the Saracens.68 The Crónica latina relates something similar but it does not forgo the opportunity to make accusations, once again, against the Leonese king. First, it states that the king of Castile was prepared to ignore the Leonese invasion because he preferred to fight against the Moors; consequently, he decided to make peace with his cousin so that the two kings could aid each other in the war against the Muslims. Later, however, it is asserted that Alfonso VIII did not trust the king of León, and for that reason summoned a large army to intimidate him and thereby force him to sign the peace, hoping that, if he did not want to aid him against the Moors, he would at least not hinder the Castilian offensive. Although Alfonso IX accepted the deal and took up arms against the infidel, the king of Castile still distrusted him. This was why, according to the chronicler, Alfonso VIII sent Diego López de Haro with a force of six hundred men to cooperate with Alfonso IX in his expedition against Muslim territory. During this campaign, undertaken in 1213, the Leonese troops along with their Castilian reinforcements conquered Alcántara and, after a raid through the area of Mérida, Alfonso IX ordered the troops to retreat against the will of don Diego, who wanted to carry on with the offensive.69 The partiality of the author very clearly influences his interpretation of events; however, I do not believe that his report completely lacks veracity. It seems that Alfonso IX was not determined to undertake an offensive action against Muslim territory, and the fact that Alfonso VIII provided him with soldiers invites us to ponder whether he would otherwise have had the military resources to launch a campaign, even in conditions that offered some guarantee of success. On the other hand, the sudden withdrawal of the Leonese forces after the taking of Alcántara may have resulted from the hardships that

68 69

Lucas de Tuy, Crónica 73, p. 416. Crónica latina 25, p. 56; 26, pp. 56–57.



Treaties between León and the Almohads

179

arose that year due to the famine suffered in the peninsula, as happened in the case of the siege of Baeza, lifted by Alfonso VIII.70 The limited enthusiasm manifested by Alfonso IX no doubt also reflected the fact that, despite his endeavors to secure the effective occupation of the southern sector, there had been as yet no substantial achievement; this would have to await the penultimate decade of his reign, when the progress in repopulation became more noticeable, continuing with even greater vigor in his last ten years. With this in mind we can appreciate that the real priority for the kingdom of León during most of its independent existence was to maintain an open border with al-Andalus. From this perspective, the role of the agreements with the Almohads within the strategy the Leonese king had followed until then is comprehensible. We still have no reports of any pact between León and the Muslims in the aftermath of the conquest of Alcántara. From such “radio silence” it may be deduced that diplomatic relations between Alfonso IX and the Almohads maintained a low profile, with no significant arrangement emerging. Castile nonetheless signed truces in 1214 with Abu Ya’qub al-Mustansir (1213–24) that were periodically reaffirmed until the emir’s death. The Almohads were immersed in North African affairs and were no imminent threat to León, and Alfonso IX, in return, did not seem to be a source of concern for the Andalusis – which would explain the absence of arrangements between the two parties. It must moreover be taken into account that, even if conflicts between the Christian kings had not completely disappeared, they were, in this second decade of the thirteenth century, abating. This too helps to explain the lack of agreements between Alfonso IX and the Almohads, as it was at moments of greater conflict with his Christian neighbors that he tended to seek their support, as we have seen above. The conquest of Alcántara might be considered as having set a new orientation for the Leonese expansion strategy, not only because Alfonso IX captured a Muslim fortress but also because, in doing so, he had relied on Castilian military reinforcements, assistance he would again seek in his subsequent expeditions against al-Andalus. I have already highlighted this aspect of the story elsewhere, but I believe it is worth briefly recapitulating events.71 The next expedition of the Leonese king against Muslim territory took place in November 1218. During this campaign Cáceres was besieged, but

70

71

Juan de Osma is silent regarding the difficulties of provisioning when he relates the Leonese campaign, but then he reports that Alfonso VIII had to lift the siege of Baeza due to the lack of victuals, which was especially acute in the Transierra and Extremadura (Crónica latina 26, p. 57). Jiménez de Rada expands the food scarcity to the whole kingdom of Castile (Historia 8.14, p. 326). His words have a hyperbolical nature but we can guess that this situation affected León as well and, in particular, its Transierra and its Extremadura, as much as it did the corresponding Castilian zones. García Oliva, “Consideraciones,” pp. 163 ff.

180

María Dolores García Oliva

the king was forced to lift the siege near Christmas due to weather conditions.72 Warrior-monks of the military orders, knights from Castile and other peninsular kingdoms, and crusaders from beyond the Pyrenees all cooperated in the siege. This participation of different forces was favored by the Pope’s exhortation for peace among the Christian princes, intending that they use their military power against the Muslims instead of wasting it in internecine fights, and the spirit of crusade had been encouraged by Rome since the Fourth Lateran Council. Thus, in the peace agreement concluded between León and Castile in 1218, both monarchs commit themselves to aid each other against all their adversaries – Fernando III introducing the exception that if he had truces signed with the Almohads, he would respect them and would assist Alfonso IX only after these expired. However, he admitted the possibility of allowing his vassals to join forces, if they wanted to, with the king of León in his clashes with the Muslims.73 This contrived interpretation of the Castilian king’s obligations to the Almohads was clearly of benefit to his father (that is, Alfonso IX), as Fernando III’s military inactivity on the Muslim border facilitated the cooperation of Castilian knights in the campaigns organized by the latter – in the siege already mentioned, for example. If the Leonese king had previously used Almohad assistance to help him face Castile, now he was using the Christians to attack al-Andalus. A similar commitment was agreed in the Peace of Boronal, concluded with Afonso II of Portugal in 1219. As in the previous case, both kings swore to help each other against every one of their adversaries except for the Muslims, as the Portuguese king had truces in force with them; but permission for Portuguese knights to participate in the king of León’s wars against the Moors was also granted. The arrangement of truces between Alfonso IX and the Almohads was also permitted, but only until those the Portuguese had contracted with the Muslims expired. From that moment, if any pact was to be made with the “infidels,” it would have to be in concert by both kings.74 License to make truces with the Almohads until those between the Muslims and the Portuguese had expired must have been requested by Alfonso IX, as he was the only king among the direct rivals – the kings of Portugal, Castile and León – who at that point had no peace agreement with them. It is true that an attack by Abu Ya’qub al-Mustansir was unlikely, as the pressure being exerted by the Banu Marin was a more urgent concern for the emir than peninsular affairs.75 However, if Alfonso IX decided to go on the offensive it would be he who was exposed to a Muslim counterattack; hence he wished to be able to stop the fighting if 72

73 74 75

In accordance with the Anales toledanos 1, ed. Enrique Flórez, in España Sagrada 23 (Madrid, 1767), the siege was lifted because “facia tan grandes aguas que non pudieron y durar” (“it was pouring so much with rain that the men could not resist it”), p. 400. J. González, Alfonso IX, 2, doc. 352, pp. 460–61. Ibid., doc. 373, p. 488. See, for instance, Miguel Ángel Manzano Rodríguez, La intervención de los benimerines en la Península Ibérica (Madrid, 1992).



Treaties between León and the Almohads

181

necessary. Alternatively, the commitment to signing truces in the future only in concert with the other king can be taken to mean that some mistrust between the two remained and, more significantly as regards the matter at issue, shows the role that diplomatic contacts may have had in the context of politico-military relations between the peninsular kingdoms. In 1222, Alfonso IX organized one more expedition, in which he was joined by monks from every military order. The king struck at Cáceres and this time, according to the Anales toledanos 2, the aggressors “attacked the city with trebuchets and delibra [sic] battering down towers and walls, and they were about to seize it, but the king of Morocco parleyed with the king of León, and this latter promised him that he would lift the siege of the city and that he would not enter the lands of the Moors; and he did so, and many Christians were captured and many others died, but later on the Moorish king passed away due to the injuries he had suffered.”76 This story is probably not devoid of truth, but I do not believe that it faithfully reports what happened, on the basis of its parallelism with another episode in which the same formula is used but the final result is clearly different: the chronicler describes the siege laid to Requena by Jiménez de Rada, archbishop of Toledo, in 1219, and concludes by saying that the Christians could not take the stronghold and suffered numerous losses.77 Thus, although the report of the damage caused to the defences of Cáceres was possibly accurate, it does not imply that the place was about to yield; given which, it is understandable that Alfonso IX abandoned the siege; not out of negligence, as the chronicler insinuates, but on account of presumed casualties and the impossibility of forcing the besieged to surrender. Viewing the episode in this perspective, there is a lack of logic in the story of an agreement between the king of León and the Almohad emir to lift the siege that may well give rise to hesitation regarding its occurrence. The chronicler affirms that after Alfonso accepted the arrangement proposed by the Almohads, the latter caused many deaths among the attackers and took many prisoners. If the delegates of Abu Ya’qub al-Mustansir – the emir was not in al-Andalus, but in Morocco – opted for an agreement with Alfonso IX, we would be forced to conclude that they had no means of repelling the attack, and if that were the case, the version offered by the annalist would be open to suspicion. Thus it seems more likely that an Andalusi counterattack was what drove Alfonso IX to abandon the siege. If so, we must doubt the existence of any pact, and therefore doubt too that the Almohads committed themselves to pay any presumed sum of money to the Leonese

76 77

Anales toledanos 2, ed. Enrique Flórez, in España Sagrada 23 (Madrid, 1767), p. 406; “delibra” must refer to an unknown siege machine or tactic. The fragment of the document we have referred to is the following: “Despues cercó à Requena dia de S. Miguél, è lidiaronla con almajanequis, è con algarradas, è con delibra, è derrivaron torres, è azitaras, è non la pudieron prender, à murieron y mas de dos mil Christianos” (“Later on he besieged Requena on the Day of St. Michael and attacked it with trebuchets and with catapults […], and they brought down towers and walls but they could not capture the city, and more than two thousand Christians died”). Anales toledanos 2, ed. Flórez, p. 400.

182

María Dolores García Oliva

king. In my opinion, the bias of the author against Alfonso IX, and his providential interpretation of events, have markedly influenced his account. What the chronicler wants to convey is that Alfonso IX, moved by his indolence and greed, decided to leave the fortress in the hands of the infidel in spite of being about to conquer it, and such an abominable action was deservedly punished, as his army suffered great losses and, even worse, he did not receive the sum he was expecting. Unfortunately there are no other accounts of this episode available to confirm what may in fact have happened; I believe, however, that there are certainly sufficient grounds, as outlined above, to doubt that Alfonso IX established any sort of agreement with the Almohads. The reports we have of subsequent offensive actions do not inform us of the composition of the forces deployed or, quite often, even the kind of military operation that was undertaken. Thus Julio González quotes a document dated 1223 in which the writer alludes to an expedition against Cáceres, and Alfonso IX issued a further document, in favor of the bishop of Astorga, in which he promised that neither he nor his successors would ever force the prelates of Astorga to provide the king with knights (“milites”), as he acknowledged that when the bishop led his forces “in exercitum de Cazeres et in alium quem postea contra sarracenos feci” (“to the army of Cáceres and others that later I formed against the Saracens”), he had done it of his own volition and for the remission of his sins.78 It may be observed that this document mentions Cáceres, and a subsequent expedition: the first allusion could be to the same expedition as the document of 1223 refers to, whereas the second may be the foray undertaken in 1225 over the Aljarafe, in which the Christians defeated the army of Seville which had been sent to face them, around Tejada. According to Lucas de Tuy, in the latter instance Alfonso IX sent Leonese noblemen with Martín Sánchez, illegitimate son of Sancho II of Portugal (1223–48), who was at that time in his service. The Muslim chronicler al-Himyari also notes the defeat of the Sevillans in Tejada, but he attributes the offensive to the Christians of the Algarve.79 Perhaps the lack of concordance among the sources reflects the fact that both Portuguese and Leonese soldiers participated in this expedition. We cannot bypass the question, and we may admit the possibility that Leonese combatants alone were involved; but if so it would hardly constitute an argument against the thesis I have been defending, but rather an exception. Certainly in his most significant campaigns, Alfonso IX either counted on reinforcements from the neighboring Christian kingdoms, or the offensive was executed in a

78 79

J. González, Alfonso IX, 1, pp. 198–99, and 2, doc. 467, pp. 579–80. Lucas de Tuy, Crónica 89, p. 422; al-Himyari, Kitab al-rawd al-mi‘tar fi jabar al-aqtar, ed. and trans. Évariste Lévi-Provençal, La Péninsule Ibérique au moyen-âge d’après le Kitab I […] (Leiden, 1938), p. 156. Francisco García Fitz admits that it is plausible that the reports of Lucas de Tuy and al-Himyari refer to the same event. He remains uncertain, however, as the sources disagree about the origins of the contingents: Relaciones políticas, pp. 171–72. Huici Miranda nonetheless considered that the episode narrated by both texts was the same: Historia política, pp. 454–55.



Treaties between León and the Almohads

183

coordinated manner, or else he made it simultaneously with attacks launched by other Christian kingdoms. Indeed, in 1226 the Leonese king assaulted Badajoz at the same time as Sancho II moved against Elvas, and possibly whilst Fernando III was engaged in conquering Capilla. According to Lucas de Tuy this was a predatory action, but a Portuguese document refers to some fighters entering the moat of the fortress at Elvas. It is possible that the kings were considering taking both fortresses, but after a long-lasting siege both had to give up their hopes. The Portuguese king was abandoned by his people, while Alfonso IX was not willing to endure the summer in that area.80 Once more, the chronicler’s animus against the Leonese king is evident in the explanation he offers of the events: if Alfonso IX lifted the siege because of the heat, it might well have been for sound reasons rather than out of some personal laxness, since summers in that area are usually extremely hot, and if the temperatures were especially high that year a decision to abandon the siege is understandable. It is moreover reasonable to suppose that resistance by the besieged would also affect his decision to halt the attack. Alfonso IX launched his penultimate expedition against Muslim territory in 1229, and this time conquered Cáceres. Lucas de Tuy attributes this initiative to the pontifical legate John of Abbeville, sent to the peninsula by Pope Gregory IX (1227–41) to exhort the Christians to take up arms against the infidel, amongst other things. The same chronicler tells us that on that occasion Castilian troops joined the Leonese forces in the attack. The offensive in question resulted in the conquest of Mérida and Badajoz, in 1230. This time the narrative sources do not detail the composition of the army, but they do offer evidence for the participation of foreign warriors. Thus, Juan de Osma reports that the inhabitants of Elvas fled the city when they heard about Ibn Hud’s defeat in Alange. On their way back from the battle some Portuguese entered the city, as they found its gates open and nobody inside. They sent the good news to the Portuguese king, who immediately sent reinforcements to retain the city under his control.81 This story indicates that Alfonso IX counted on the assistance at least of Portuguese soldiers. III These events patently demonstrate, I would argue, how the kings of León used political/diplomatic relations to support their strategy of expansion against the Andalusi territory: first to preserve their expansion zone through alliance with the Muslims, and then, when Alfonso IX went on the offensive, to gain the military cooperation of forces from other kingdoms in his campaigns. It is true that Alfonso IX did not take any Muslim strongpoint by purely political means, as did Fernando III, the latter king furnishing the paradigmatic example of terri80 81

Lucas de Tuy, Crónica 89, p. 422; José Mattoso, História de Portugal 2:121–22; Crónica latina 50, p. 82. Crónica latina 57, p. 88.

184

María Dolores García Oliva

torial expansion by means of efficient political strategies, as García Fitz has pointed out. He highlights the manner in which Fernando III pursued to its conclusion every political mechanism to destabilize al-Andalus, in the context of the third taifa period – a factor that contributed decisively to the Castilian territorial expansion. García Fitz compares Fernando’s performance on some occasions with that of Alfonso IX. For example, he contrasts the Castilian territorial advance between 1224 and 1226, when thanks to the support he provided to ‘Abd Allah “el Baezano” Fernando obtained from him Martos, Andújar, Salvatierra, Burgalimar, and Capilla (which last he had to take by force), with Leonese passivity during the same period. Such a contrast, in García Fitz’s opinion, “can serve to give an idea of the importance that, from the point of view of expansion, some well-oriented political relations or taking advantage of a favorable juncture might have had.”82 This might be taken to imply that Alfonso IX lacked this capability or political intelligence; but in my opinion that idea does not quite meet the case. Castile was able to develop such a political strategy because it was the hegemonic kingdom, a circumstance that its king unquestionably knew how to exploit. Alfonso IX might not have had the same ability to extend his domain as Fernando III, but he always kept in mind political relations with the other Christian kingdoms within his project for expansion, and he took advantage of Andalusi political decay to push his borders from the Tagus to the Guadiana. This surely proves that he knew how to use more than wisely any means available to him.

82

García Fitz, Relaciones políticas, pp. 159 ff., quotation at p. 172.

Map 1: The Conquest of Estonia

8 A Strategy of Total War? Henry of Livonia and the Conquest of Estonia, 1208–1227 John Gillingham

In 1227 a priest, known today as Henry of Livonia, finished writing a history of events in his part of the world (the modern Baltic states of Latvia and Estonia) during the previous thirty or so years. His chronicle, by far the best of the written sources for the conquest and conversion of the pagan peoples of the region by military and missionary forces largely coordinated by the recently founded Latin church of Riga, is one of the most important of all sources for the conduct of war in the European Middle Ages. What Henry wrote about sieges and battles is undoubtedly valuable; these encounters were the action highlights of war and as such have long been of interest to medieval chroniclers and modern military historians. But rather less attention has hitherto been paid to what he wrote about the more mundane subject of raiding – almost certainly the most common form of warfare throughout the Middle Ages, and probably throughout most of world history too. As the first European author to describe what previous chroniclers had either disregarded as unremarkable routine or had, in twelfth-century England, treated rhetorically as barbaric atrocity, his insider’s account of the early thirteenth-century conquest of Estonia, an unusually dispassionate narrative of raid and counter-raid, offers revealing insights into the obscure subject of the conduct of war in earlier centuries.1 Henry and the Church of Riga Nothing is known about Henry’s earlier life apart from what can be inferred from his chronicle.2 This suggests that, no matter whether his parents were German 1

2

“Most early medieval campaigning was probably small-scale and aimed at the acquisition of booty,” but “the majority of such warfare escapes notice” and “how campaigns were carried out is, in detail, obscure.” Guy Halsall, Warfare and Society in the Barbarian West, 450–900 (London, 2003), pp. 134–62, esp. 145, 156–62. See also Helen Nicholson, Medieval Warfare (Basingstoke, 2004), p. 3. Heinrici Chronicon Livoniae, ed. L. Arbusow and A. Bauer, MGH SRG 31, 2nd ed. (Hanover, 1955). The chronicle [hereafter HCL] is well known in the English-reading world thanks to James Brundage’s translation, The Chronicle of Henry of Livonia (New York, 1961) [hereafter Brundage], reprinted unaltered except for a new introduction in 2003. Unfortunately this translation, “substantially completed” as early as 1951, is based on the 1874 MGH edition by Wilhelm Arndt. In this paper all translations, although deeply indebted to Brundage’s, are my own.

188

John Gillingham

or not, he was educated in north Germany, perhaps in the Magdeburg region and/or the monastery of Segeberg in Holstein, and that he entered the Baltic region in the service of Albert of Buxtehude, bishop of Riga/Livonia, probably in 1205, the year from which his narrative became noticeably more detailed.3 In 1208 he was made a priest and was sent to minister to the newly-baptized Letts living in the valley of the Ümera (Jumar), a tributary of the river Gauja (Aa), to the northeast of Riga.4 Here, right on the frontier of Latin Christendom, he was, in his own words, “exposed to many dangers.” In 1211 Estonian raiders burned his church down; in 1223 they plundered or burned his crops, houses, and belongings and stabled their horses in his (by now stone-built?) church.5 He was passionately involved in the warfare of the Rigan church, especially that engaged in by his own parishioners, the Letts, and was himself present at many of the decisive moments. “We put down the holy chrism and other sacramental articles and took up sword and shield,” is how he described a moment of alarm during a raid into Vironia in 1219.6 He offers a vivid account of the problems faced by a small number of technologically well-equipped troops trying to conquer and hold territory in difficult terrain while simultaneously trying to persuade the natives to adopt a new code of social and moral values (in this case Latin Christianity).7 Since Henry was still alive in 1259, it is conceivable that he composed a continuation dealing with the years after 1227, but if so, it does not survive.8 Indeed the oldest extant manuscript of the chronicle, the Codex Zamoscianus (Z) dating from c. 1300, contains only about two thirds of 3

4 5 6

7

8

Albert, bishop from 1199 to 1229, was a nephew of Hartwig, archbishop of Hamburg-Bremen (1185–1207). On arguments concerning Henry’s descent and loyalties, see Jüri Kivimäe, “Henricus the Ethnographer: Reflections on Ethnicity in the Chronicle of Livonia,” in Crusading and Chronicle Writing on the Medieval Baltic Frontier. A Companion to the Chronicle of Henry of Livonia ed. Marek Tamm, Linda Kaljundi, and Carsten Selch Jensen (Farnham, 2011), pp. 77–106 at 78–80. On the wider context of the struggle between Christian and pagan in northeastern Europe see above all Eric Christiansen’s masterpiece, The Northern Crusades, 2nd ed. (London, 1997). HCL XI, 7, p. 55 (Brundage, p. 75); XIV, 12, p. 86 (Brundage, p. 109); XXVII, 1, p. 194 (Brundage, p. 213). HCL XXIII, 7, pp. 161–62 (Brundage, p. 179). Bauer’s introduction emphasizes not only the extent of Henry’s direct participation in warfare but also his deep understanding of it, HCL, pp. x–xiv, xxiii–iv. Later writers, notably Peter of Dusberg and the author of the Livonian Rhymed Chronicle, share many of Henry’s attitudes to war and their enemies: The Livonian Rhymed Chronicle, trans. Jerry C. Smith and William L. Urban, 2nd ed. (Chicago, 2001). For a valuable study, strong on literary theory, of Henry and Peter of Dusberg, see Claudia Müller’s doctoral dissertation, Information und Transformation an der Grenze. Kriegschroniken im Baltikum – Textanalyse (Greifswald, 2006). I owe my awareness of this to the kindness of Christian Lübke. But within the flourishing field of frontier studies, frontier warfare remains a stepchild. For instance, not one of the ten sections in Grenzräume und Grenzüberschreitungen im Vergleich. Der Osten und der Westen des mittelalterlichen Lateineuropa, ed. Klaus Herbers and Niklaus Jaspert (Berlin 2007), was devoted to warfare. He is mentioned, as the priest of Papendorf (now Rubene, about nine miles southwest of Wolmar), in documents dated to 1234 and 1259: HCL, p. xv.



Henry of Livonia and the Conquest of Estonia

189

the accepted text; the remaining third covering the years from 1219 onwards has had to be supplied from seventeenth- and eighteenth-century transcripts. Unfortunately these later versions lack the rubrics which had been an integral feature of Z, and which, since they appear to have been composed by Henry himself, offer a useful shorthand guide to those topics which he wanted to emphasize.9 It seems likely that he began his chronicle as a guide to the recent past, and to the rights claimed by the bishop of Riga, for William of Modena, the first papal legate sent to Livonia and Estonia.10 The story he chose to tell was one of great triumph. When Albert of Buxtehude was consecrated bishop of Livonia in 1199, he was appointed to a church which, despite the efforts of his two predecessors, Bishops Meinhard (1186–96) and Berthold (1197–98), still did not exist.11 Like Meinhard and Berthold, Bishop Albert owed his appointment to Archbishop Hartwig of Hamburg-Bremen, yet the embryonic new diocese of Livonia lay in the valley of the Lower Dvina about a thousand kilometers east of Bremen. Livonia lay not so much on the edge of Latin Christendom as some 500 km beyond its edge; to the west for about 500 km the lands were held by pagan peoples, chiefly Prussians, Curonians, and Lithuanians. Yet by 1225 Bishop Albert had, in Henry’s words, “planted a vineyard which had grown so much that its branches extended for the length of 10 days journeys, as far as Reval (Tallinn) in one direction, as far as Pskov in another and along the River Dvina as far as Gerzike (Jersika) in a third.”12 The church of Riga now ruled a mini-empire roughly 400 km by 400 km in size. The various peoples who lived within that area, Finno-Ugrian speaking Livs and Estonians, and Balt-speaking Letts, had been baptized as Christians.13 It had not been a peaceful conversion. In addition to the demand that converts pay tithes, the missionaries had intended that baptism to Latin Christianity should bring about a radical change in matters as important as marriage customs and the disposal of the dead. In the war of Estonian independence in 1223 the insurgents not only killed as many Germans as they could, but also “took back their wives who had been sent away during the Christian period. They disinterred the bodies of their dead, who had been buried in cemeteries, and cremated 9 10

11

12

13

See the Appendix for discussion of the authorship of the rubrics. Anti Selart, “Iam tunc… The Political Context of the First Part of the Chronicle of Henry of Livonia,” in The Medieval Chronicle, 5, ed. Erik Kooper (Amsterdam, 2008), pp. 197–209. Henry evidently brought his narrative to a first conclusion with an account of the legate’s activity in 1225–26, but then added another chapter as a result of the legate’s decision to launch a crusade against the Öselians. According to Henry, after Berthold had been killed in a skirmish against pagan Livs and his body torn limb from limb, the clergy were all driven out – though not before the baptized Livs asked for a new bishop to be sent from Germany: HCL II, 6–10, pp. 10–11 (Brundage, pp. 33–34) HCL XXIX, 2, p. 208 (Brundage, p. 230). Both Reval and Pskov were about 280  km from Riga, Gerzika only 150 km, but upstream against the current this may also have been a ten-day journey. Alan V. Murray, “Henry the Interpreter: Language, Orality and Communication in the ThirteenthCentury Livonian Mission,” in Crusading and Chronicle, ed. Tamm et al., pp. 107–34.

190

John Gillingham

them according to their native pagan custom. They washed themselves, their houses and their forts with brooms and water, trying to wash away the sacrament of baptism.”14 It took repeated military defeats before first the Livs and Letts and then the Estonians finally accepted the new rules. This inspired Henry to write with fiery eloquence of the martial triumphs of Mary, the patron saint of the church of Riga.15 Not surprisingly, those who submitted to the mother of God were routinely required to supply hostages as guarantors of their good faith.16 If peace now reigned – in fact a much bigger “if” than Henry appears to have realized in 1227 – it was because, as he put it, “Fear of the Rigans and the Germans had fallen over all neighbouring lands and over all people who lived in the region.”17 In narrowly military terms the imposition of fear was undoubtedly a formidable achievement, the consequence of war against Russians and Lithuanians as well as against Livs, Letts, and Estonians. The Russian Orthodox Christian princes, notably the rulers of Pskov and Polotsk (Polatsk, Belarus), had not willingly relinquished the tributes which they had been accustomed to take from Livs, Letts, and Estonians. As for the Lithuanians, in Henry’s words, “they were such lords over all peoples, both Christian and pagan, that […] all fled like rabbits before hunters, for they were sheep and the Lithuanians wolves.18 The Germans had intruded into the Lithuanians’ hunting grounds, and so they too had not taken kindly to interlopers from the West. True, as Henry saw very clearly, the invaders possessed superior military technology. The pagans did not wear body armor, and did not know how to make crossbows or siege machines for hurling stones, or how to build stone and mortar fortresses.19 The engineering skill needed to build siege machines was the “German technique” (ars Theutonicorum).20 Of course the technology of warfare, particularly on land, could be acquired and turned against the invaders, but there is no evidence of this happening here before the 1220s.21 In any case – as in other frontier regions of Latin Europe – the local peoples did not have the industrial resources to manufacture large quantities of armor and ammunition.22 In this respect, though 14 15 16 17

18 19 20 21

22

HCL XXVI, 8, p. 191 (Brundage, p. 210). HCL XXV, 2, pp. 179–81 (Brundage, pp. 198–200). Here the “mother of God” was as much a war-goddess as she was in the Livländische Reimchronik. See, for instance, the dispute about the handover of hostages in 1212, HCL XVI, 3, pp. 106–08 (Brundage, pp. 125–26) HCL XXIX, 1, p. 207 (Brundage, p. 229). For Henry’s concept of peace, see Simon Gerber, “Heinrich von Lettland – Ein Theologe des Friedens,” Zeitschrift der Kirchengeschichte 115 (2004), 1–18. HCL XIII, 4, p. 69 (Brundage, pp. 90–91). See, for example, HCL I, 6, p. 3 (Brundage, p. 27); X, 12, p. 42 (Brundage, pp. 63–64). HCL XXVII, 3, p. 197 (Brundage, p. 217), when used, in vain, by the Russians in 1223 in their siege of Danish Reval. Cf. more Theutonicorum, HCL X, 12, p. 42 (Brundage, p. 63). In 1222 the Öselians employed siege machines to capture a stone fort built by Danes, HCL XXVI, 3, p. 188 (Brundage, p. 206). On this transference of skills see Robert Bartlett, The Making of Europe (Harmondsworth, 1993), pp. 72–76. On the material culture of this Late Iron Age society and its chiefs with their fortresses of earthwork and timber, see Andris Sne, “The Emergence of Livonia: The Transformations of Social and Political



Henry of Livonia and the Conquest of Estonia

191

much always depended upon the nature of the terrain, the newcomers enjoyed a clear military superiority over both Christian Russians and pagans. This aspect of the wars has been well treated in the literature, and often used as an illustration of medieval asymmetric warfare.23 The Conquest of Livonia, 1199–1208 Bishop Albert’s first action after his consecration in 1199 clearly revealed his strategic realism. Instead of visiting his diocese he went to the island of Gotland, where he recruited crusaders (“pilgrims” in Henry’s terminology). As Henry saw it, the offer of remission of sins to those who “took the cross” was crucial in attracting volunteers to so dangerous an enterprise.24 No doubt it was, but it is significant that Albert looked to Gotland for his first pilgrims. German businessmen from Lübeck and from Visby (their colony on Gotland) had for some time before this been sailing to the mouth of the Dvina, from where they took either the route to Novgorod via Pskov or went upstream as far as Polotsk, where there was a porterage to the river Dnieper and so on to Kiev, the Black Sea, and Constantinople. The trade along these routes offered great profit from furs, honey, beeswax, and slaves.25 The choice of the Lower Dvina as a sphere for missionary activity had clearly been determined by commercial interests.26 At critical moments it would be the merchants of Lübeck and Visby who defended and promoted the church which they had, in effect, called into being.

23

24

25 26

Structures in the Territory of Latvia during the Twelfth and Thirteenth Centuries,” in The Clash of Cultures on the Medieval Baltic Frontier, ed. Alan V. Murray (Farnham, 2009), pp. 53–72 at 58–63. On Europe’s northwest frontier, see John Gillingham, “Conquering the Barbarians: War and Chivalry in Britain and Ireland,” in Gillingham, The English in the Twelfth Century (Woodbridge, 2000), pp. 41–58 at 49; repr. from The Haskins Society Journal 4 (1992), 67–84. Christiansen, Northern Crusades, pp. 38–9, 101; Sven Ekdahl, “Horses and Crossbows: Two Important Warfare Advantages of the Teutonic Order in Prussia,” in The Military Orders 2: Welfare and Warfare, ed. Helen Nicholson (Aldershot, 1998), pp. 119–51; Stephen Turnbull, “Crossbows or Catapults? The Identification of Siege Weaponry and Techniques in the Chronicle of Henry of Livonia,” in Clash of Cultures, ed. Murray, pp. 307–19, esp. pp. 312–18; Kurt Villads Jensen, “Bigger and Better: Arms Race and Change in War Technology in the Baltic in the Early Thirteenth Century,” in Crusading and Chronicle, ed. Tamm et al., pp. 245–64; Ain Mäsalu, “Mechanical Artillery and Warfare in the Chronicle of Henry of Livonia,” in ibid., pp. 265–90; Valter Land and Heiki Valk, “An Archaeological Reading of the Chronicle of Henry of Livonia,” in ibid., pp. 291–316. HCL I, 12, p. 7, and II, 3, p. 9 (Brundage pp. 30, 32). See Christopher Tyerman, “Henry of Livonia and the Ideology of Crusading,” in Crusading and Chronicle, ed. Tamm et al., pp. 23–44. On Bishop Albert’s approaches to the papal curia see Barbara Bombi, “Innocent III and the praedicatio to the Pagans in Livonia,” in Medieval History Writing and Crusading Ideology, ed. Tuomas M. S. Lehtonen and Kurt Villads Jensen (Helsinki, 2005), pp. 232–56 at 232–41. The civil war in Germany may also have made going away in a good cause seem an attractive option. “The route to the Dvina via Gotland was a well-known way to riches, and therefore a not-unattractive path to salvation,” Christiansen, Northern Crusades, p. 98. Mark R. Munzinger, “The Profits of the Cross: Merchant Involvement in the Baltic Crusade (c. 1180–1230),” Journal of Medieval History 32 (2006), 163–85.

192

John Gillingham

Not until the second year of his episcopate (1200) did Albert, having recruited a fleet and a small army, at last feel able to visit his new diocese. The problem he faced here was that Bishop Meinhard had chosen to build his church at Üxküll (Ikskile, Latvia), about 50  km up the Dvina, well placed for preaching to his prospective converts, but not a secure base with easy communication to Gotland and Lübeck.27 Albert moved his cathedral down-river to Riga.28 The site was well-chosen, not right on the coast where it would have been vulnerable to attack, but where there was none the less a harbor deep enough to provide safe anchorage for cogs, the largest seagoing ships of the day.29 In this sphere too superior technology was crucial. Without cogs, the merchants and missionaries sailing into the Gulf of Riga would have been at immense risk from pirates (as they saw them) putting out from the harbors of Curonia and Ösel (Saaremaa). The cog was in effect a floating fortress. In 1216 a cog carrying fifty men with crossbows and arms was stationed at the mouth of the Dvina to guard, like a fort (tamquam castrum), the approach to Riga.30 On one occasion in 1215, Henry was himself on board one of nine cogs which held off, according to his own dramatic estimate, no fewer than two hundred Estonian ships. “At dawn the sea opposite us looked black with their pirate ships.”31 At other times the capacity of these bulk-carrying ships to bring in supplies was vital. Riga suffered a great famine in 1206 and was saved, according to Henry, by the arrival of two cogs filled with grain from Gotland.32 The sea-lane from Lübeck to Riga via Gotland was the strategic lifeline of the Livonian church. Without it, it could not have survived. Yet remarkably it was a lifeline which every year was closed by ice from October/November to March/April. In the perspective of Riga’s church the annual re-opening of navigation in spring marked the departure of the “winter pilgrims” and the arrival of summer ones. This rhythm, together with the impact of weather upon conditions of travel on land, shaped the seasonal pattern of warfare in the Baltic.33 The sea-lane moreover was a lifeline which could be closed in summer too – if the king of Denmark so chose: a fact which, as was to become apparent after 1219 (see below, p. 197), left the new church disturbingly dependent upon good relations with Denmark. 27 28

29 30 31 32 33

Albert’s ships came under attack as they sailed to and from Üxküll; one was captured and most of its crew killed. HCL IV, 2–3, p. 13 (Brundage, p. 36). He and the merchants associated with him promoted his port’s prosperity by taking harsh measures against traders going to Semgallia (south of the Dvina) who tried to by-pass Riga. HCL IV, 5–7, V, 1, VI, 3, pp. 14–17 (Brundage, pp. 37–40). German merchants are first recorded at Riga in 1158. At the river mouth Albert founded a Cistercian abbey, Dünamünde, in 1205. HCL XIX, 11, p. 134 (Brundage, pp. 154–55). HCL XIX, 5–6, pp. 127–30 (Brundage, pp. 148–50). HCL X, 9, p. 39 (Brundage, p. 60). And it came to shape the structure of Henry’s chronicle. Initially he divided his narrative of Albert’s episcopate into sections (termed “chapters” in modern editions of the chronicle), one for each episcopal year and starting at the anniversary of the bishop’s consecration (in late February or early March). But from the eleventh “chapter” (1207–08) onwards each chapter began with the spring re-opening of navigation: HCL, pp. xxviii–xxxiii.



Henry of Livonia and the Conquest of Estonia

193

Despite their technological superiority on land and sea, the incomers faced massive problems. Few were prepared to settle in a dangerous pagan land. The distance from the main recruiting bases in northern Germany meant that their numbers were always small; in the early years they were tiny. Henry was fully aware of the manpower problem and of Albert’s herculean efforts to deal with it. An observation made at the start of his narrative of the sixth year of Albert’s episcopate (1204) expresses the retrospective view of the time of writing: “The bishop, fearing the dangers which the pagan threat posed to the city, still small and weak owing to the small number of the faithful, went again to Germany. He was so committed to the task of converting the pagan which had been laid upon him that he undertook the almost intolerable burden of going to and returning from Germany every year.”34 But the outcome of each exhausting preaching tour was uncertain, and in any case sooner or later pilgrims, having fulfilled their vow, went home.35 A permanent settlement of fighting men was needed. Nobles could be, and were, given fiefs, but in the early stages most of these were speculative grants of places not yet acquired.36 In this difficult situation Albert took an extraordinary strategic step. In 1202/3 he became the first bishop in Christendom to institute his own military order: the Brothers of the Knighthood of Christ in Livonia.37 These “Sword Brethren” rapidly became a permanent and effective fighting force. In the retrospect of the 1220s it seemed to Henry that, whether the bishop was present or absent, it was Folkwin, the Sword Brethren’s Master (1209–36), who “led and commanded the Lord’s army on every expedition, setting out again and again against neighbouring pagans and joyfully fighting the Lord’s battles.”38 In formal terms, the Sword Brethren were the bishop’s servants, bound by oath to obey him, but in practice their indispensability meant that they were able to constrain the strategic choices he could make. Unlike the Templars, they did not enjoy income from manors throughout Europe, and in consequence were dangerously dependent on the profits of aggressive war and territorial expansion. In negotiation with Bishop Albert in 1207 they demanded not only one third of Livonia, but also one third of future conquests. Albert refused the latter, but the claim is a telling sign of their aggressive mindset.39

34 35 36 37

38 39

HCL VIII, 1, p. 23 (Brundage, p. 45). In fact he certainly stayed in Livonia over the winters of 1205–06 and 1207–08: HCL X, 2, p. 34, XI, 5, p. 52 (Brundage, pp. 55, 72). Excessively optimistic on this point is Arnold of Lübeck, “De conversione Livoniae,” Chronica Slavorum, MGH SS 21, pp. 212–17 at 216. The enfeoffment of Conrad of Meiendorf with Üxküll, for example: HCL V, 1, pp. 15–16, IX, 7, p. 29 (Brundage, pp. 39, 50). Henry attributed the foundation to Theodoric, a Cistercian monk (and from 1211 bishop to the Estonians), HCL VI, 4, p. 18 (Brundage, p. 40). But Innocent III’s bull of October 1204 approving the order named Albert as founder and head of the order, and Theoderic must have acted with the bishop’s approval. HCL XIII, 2, p. 68 (Brundage, p. 89) HCL XI, 3, pp. 49–50 (Brundage, p. 69). In quiet times they pressed their subjects all too hard, triggering a rebellion of Livs and Letts in 1212, and of Estonians in 1223. Cf. HCL XXIX, 3, p. 210 (Brundage, p. 231) for the legate’s advice to the Sword Brethren of Fellin in 1225.

194

John Gillingham

During the first few years, however, there can have been few disagreements about strategy. At all costs, even if it meant making alliances with pagan neighbors, the bridgehead in the valleys of the Dvina and the Gauja had to be established. In the first five years, from 1200 to 1204, all the incidents of war reported by Henry took place at Riga, Holme (Salaspils, Latvia), Treiden (Turaida), and Üxküll. 1205 saw the newcomers advance another 50  km up the Dvina, taking and burning Lennewarden and Ascheraden (modern Aizkraukle).40 1206 was a crisis year. In June it took a fierce struggle, in which the Sword Brethren distinguished themselves, to recover the fort of Holme (only 20  km upstream from Riga) from “apostate” Livs.41 When the bishop then returned to Germany to recruit more pilgrims, very few were left behind to defend Riga at a time when “the city walls had not yet been made secure.”42 At this point the Russians of Polotsk moved down the Dvina and laid siege to Holme. For eleven days the fate of Riga appeared to depend upon Holme’s twenty-strong German garrison. But the Russians and their “apostate” allies were unable to take the small fort, in part owing to their ignorance of the ars ballistaria, and this failure Henry saw as a turning point. “After the retreat of the king of the Russians the fear of God took hold of the whole land of Livonia” and all the peoples of the land were baptized.43 Next year (1207), with the help of a new group of “summer pilgrims,” the walls of Riga were completed and – though Riga was to be attacked again – this led to a new confidence. “From now on the attack of the pagans was no longer feared.” In 1208 it was announced that there would be no more alliances with pagans.44 In 1208–09 campaigns directed by Bishop Albert saw major advances up the Dvina as far as Kukenois (Koknese, Lat, 30 km upstream from Ascheraden), Selburg, and Gerzike (80  km further still), places hitherto ruled by Russian Christians.45 As Henry observed, controlling these places had the additional strategic advantage of denying to the Lithuanians the crossing points which they had customarily used in their great raids into the lands north of the Dvina. Now, wrote Henry, “Bishop Albert freed his sheep from the jaws of the [Lithuanian] wolves.”46 What was much less clear was the direction which any further expansion should take. Should the logic of safeguarding the vital sea-lane make the Öselian archipelago at the entrance to the Gulf of Riga the next target? 40 41 42 43 44 45

46

HCL IX, 8–9, p. 30 (Brundage, p. 51). HCL X, 8–9, pp. 37–39 (Brundage, pp. 58–60). HCL X, 11–12, pp. 41–42 (Brundage, pp. 62–63). HCL X, 12–15, XI, 1, pp. 41–47 (Brundage, pp. 62–68). HCL XI, 1, p. 48, XII, 3, p. 60 (Brundage, pp. 68, 81). Not until 1212, however, did the king of Polotsk accept that Bishop Albert ruled Gerzike and all Livonia: HCL XI, 2, p. 48, XIII, 4, pp. 69–71, XVI, 2, pp. 102–04 (Brundage, pp. 68–69, 90–93, 121–23). On the Russians of Gersike and Koknese and Albert’s campaigns see Torben K. Nielsen, “Sterile Monsters? Russians and the Orthodox Church in the Chronicle of Henry of Livonia,” in Clash of Cultures, ed. Murray, pp. 227–52 at 233–41. HCL XI, 6, pp. 53–4, XIII, 4, p. 69 (Brundage, pp. 73–74, 91).



Henry of Livonia and the Conquest of Estonia

195

The Danes had indeed already attempted this in 1206.47 Or should the goal be Curonia, both by the same naval logic and as part of a drive southwestward in an attempt to close the huge “pagan gap” between Livonia and Pomerania?48 In the summer of 1208 a heavily-armed German raiding party moved south and invaded Lithuanian territory. But this ended badly.49 Years later it was to be an advance into the difficult terrain south of the Dvina which, at Saule in 1236, brought catastrophic defeat, the death of Master Folkwin, and the extinction of the Sword Brethren.50 In and after 1208, however, an entirely different theater of war was given priority: Estonia. Henry had begun his chronicle by devoting a book to each of the first three bishops. But at this point, in the tenth year of Albert’s episcopate, he intruded two new rubrics: “Here ends Book Three concerning Livonia,” immediately followed by “Here begins Book Four concerning Estonia.”51 And there were to be no more books. Book Four continues until the end of the narrative. The conquest and conversion of Estonia was to be the great theme of the rest of his chronicle. The Conquest of Estonia, Phase One: 1208–1212 Henry himself was involved in the meetings in 1208 at which the first attack on the Estonians was decided upon. He had arrived among the Letts earlier that year in response to their offer to convert to Christianity. They hoped, he noted, that against their traditional enemies, Lithuanians, Livs, and Estonians, the Germans would be more powerful protectors than the Russians had been.52 His new parishioners wasted no time before advocating an attack on their neighbors to the northeast: the Ugaunian Estonians. In the absence of Bishop Albert recruiting pilgrims in the West, their proposal was backed by Rigan businessmen who blamed the Ugaunians for disrupting their trade with Pskov, and by Berthold, the belligerent commander of the local fort of the Sword Brethren at Wenden (Cesis).53 Once the decision to attack Ugaunia had been taken, a large force of Germans, Letts, and Livs marched through the night, 47

48 49 50 51 52 53

HCL X, 13, p. 43 (Brundage, p. 64). The Öselians posed many threats, e.g. HCL XIV, 12, pp. 86–87, XV, 3, p. 89, XIX, 1–2, pp. 122–24 (Brundage, pp. 108–9, 142–43), but their sea power meant that during the summer months the only military challenge they had to fear came via the Danish navy. Marika Mägi, “Ösel and the Danish Kingdom: Revisiting Henry’s Chronicle and the Archaeological Evidence,” in Crusading and Chronicle, ed. Tamm et al., pp. 317–41 at 322. Not until 1218 was a Rigan assault on the islands seriously considered – only to be aborted when the ice melted. XXI, 5, pp. 144–45 (Brundage, p. 164). On the Curonian threat to pilgrims and Riga in 1210, HCL XIV, 1, 5, pp.72–77 (Brundage, pp. 94–99). Some Livs wanted to invade Curonia in 1215, HCL XVIII, 5, p. 117 (Brundage, p. 137). HCL XII, 2, pp. 59–60 (Brundage, pp. 79–81). Friedrich Benninghoven, Der Orden der Schwertbrüder (Cologne, 1965), pp. 85–91, 98–104, 254–69. HCL XII, 5–6, p. 61. These two rubrics are rightly given great prominence in Brundage’s translation (pp. 82–83). Unfortunately almost all the other rubrics are omitted (See Appendix). HCL XI, 7, p. 55 (Brundage, p. 75). HCL XII, 6, pp. 61–62 (Brundage, pp. 83–84).

196

John Gillingham

captured and burned Odenpäh (Otepää), returning home with their winnings: flocks, captives, and loot.54 In retaliation the Ugaunians called upon the aid of Saccalian Estonians and raided Lettish territory.55 Then it was the Letts’ turn again. They raided Saccala, and returned in triumph to Beverin to distribute gifts to all, including to their priest who was there “with soldiers and some of the bishop’s crossbowmen.” That day, wrote Henry, noting that it was Gaudete Sunday (18 December 1208), all rejoiced that God had taken such vengeance upon the pagan. Not all were happy with the turn of events in 1208. Hermann, whom Bishop Albert had appointed as magistrate to the Livs, was angered “because the Letts again and again renewed the war against the Estonians,” and he put pressure on the parties to agree to a year’s truce.56 When Albert returned to Livonia in 1209, he concentrated on the Dvina valley, not on the Estonian frontier. When the truce with the Ugaunians expired, he agreed to renew it.57 But in 1210 while the bishop was on another recruiting drive in Germany, Berthold of Wenden and the Letts launched more raids into Ugaunia.58 The consequence was – just as Hermann and the bishop had feared – a massive retaliatory raid which culminated in heavy defeat for the Christians in battle at the Ümera in 1210.59 Yet this only led to a thirst for vengeance and an intensification of the struggle, during which the logic of raiding (see below, p. 202–03), led the Rigans to launch attacks further and further north – a northwards thrust which suited the ambitions of both the Sword Brethren and the Letts.60 Such expansionist aggression inevitably alarmed the Russians and so late in 1210, in order to prosecute this war against Estonians without Russian “interference,” the Rigans agreed that the Livs, or the bishop on their behalf, should pay tribute to the ruler of Pskov. Placated, in January 1211 the Russians went so far as to lend support to the first raid into Sontagana.61 54 55

56 57 58 59 60 61

HCL XII, 6, pp. 62–63 (Brundage, p. 84). In the bishop’s absence, his brother Theodoric was among the commanders. Henry tells us that the priest of the Letts – i.e. Henry himself – “mounted the ramparts of Fort Beverin and played a musical instrument in praise of God.” This, he claimed, “made the barbarians stop fighting, for in their country they had no such instrument,” HCL XII, 6, p. 63 (Brundage, p. 85). See Alan V. Murray, “Music and Cultural Conflict in the Christianization of Livonia,” in Clash of Cultures, ed. Murray, pp. 293–305 at 297–99. It was here, and only here, that Henry referred to the pagans as “barbari.” For a favorite phrase of his, “ferocitas gentium,” see Marek Tamm, “A New World into Old Words: The Eastern Baltic Region and the Cultural Geography of Medieval Europe,” in Clash of Cultures, ed. Murray, pp. 11–35 at 25–26. HCL XII, 6, pp. 64–65 (Brundage, p. 87). HCL XIII, pp. 66–72.(Brundage, pp. 87–94). HCL XIII, 5, pp. 71–72, XIV, 4, 6, pp. 74, 77–78 (Brundage, pp. 93–94, 96, 99–100). HCL XIV, 7–8, pp. 78–80 (Brundage, pp. 100–102). The Estonian raid coincided with an attack on Riga by Curonians and others. Their first raid on Sontagana took place in January 1211 and the first on Jerwia in January 1212: HCL XIV, 10, pp. 81–82, XV, 3, 7, pp. 89–91, 97–98 (Brundage, pp. 103–04, 110–13, 117–18). HCL XIV, 9–10, p. 81 (Brundage, pp. 102–03).



Henry of Livonia and the Conquest of Estonia

197

In Henry’s narrative a key moment occurred in the spring of 1211 at what he called “the first siege of Fellin” (Viljandi), the chief Saccalian fortress. Here, for the first time, he recorded the Germans employing a siege machine, the like of which the defenders had never seen. When they surrendered, they acknowledged, wrote Henry, “that your god is greater than our gods” and agreed to be baptized.62 Before this, when Estonian forts were captured, all that happened was they were plundered and burned down.63 After another eleven months of destructively predatory warfare both sides found themselves struggling against disease and famine. In consequence a three-year peace was made in 1212, with Saccala as far as the river Pala (Navesti) being assigned – in theory at least – to the power of the bishop and Germans.64 The Conquest of Estonia, Phase Two: Expansion and Escalation, 1215–1220 During the years of peace with the Estonians, Henry had little to report.65 In 1215 the war was renewed and before the end of the year, after a campaign of ferocious intensity, the Ugaunians submitted.66 Over the next few years the northward tide of Rigan aggression continued unabated. In 1216 Sontagana was invaded and submitted; in 1217 Jerwia was invaded and submitted, while a Saccalian bid for independence was crushed; in 1218 Maritima was invaded and submitted, and a Jerwian bid for independence was crushed; in 1219 Vironia was invaded and submitted. But the submission of the Ugaunians to Riga had precipitated a major strategic shift. The loss of their Ugaunian tribute meant that from 1216 on Russian forces regularly took up arms in support of Estonians. In response Rigan raiding parties extended their field of operations eastwards into the territory of Novgorod and Pskov.67 This in turn led to that worry about the approach of a “great war” against the Russians which persuaded Bishop Albert to ask for Danish military assistance. In response King Valdemar II of Denmark led an expedition to northern Estonia and after hard fighting secured the submission of the province of Harria, including Reval (Tallinn) in 1219–20.68 This marked the completion of the initial conquest of mainland Estonia.

62 63

64 65

66 67 68

HCL XIV, 10–11, pp. 83–85 (Brundage, pp. 105–07). Odenpäh was burned down in 1208 and 1210: HCL XII, 6, p. 63, XIV, 6, p. 78 (Brundage, pp. 84, 100); for other examples in 1211 see XIV, 10, p. 82, XV, 2, p. 88, XV, 7, pp. 95–96 (Brundage, pp. 104, 109, 116). HCL XVI, 1, p. 102 (Brundage, p.121). Whereas Henry’s narratives of 1211 and 1212 take up fifteen pages each in the printed edition, six pages suffice for 1213 and 1214 combined. The renewal of war resulted in sixteen pages on 1215. HCL XIX, 4, 7, p. 126–7, 132 (Brundage, pp. 147, 152). HCL XX, 5, p. 138 (Brundage, pp. 157–58). HCL XXII, 1, p. 147, XXIII, 2, pp. 154–56 (Brundage, pp. 186, 190).

198

John Gillingham

The Conquest of Estonia, Phase Three: Division and Rebellion, 1220–1224 The Öselians continued to resist, even defeating at Leal in Rotalia in August 1220 an attempt by a Swedish force to acquire a share in the spoils of conquest.69 Danes and Rigans fell out when the former laid claim to the whole of Estonia and Livonia. Henry preaching and baptizing in Vironia found himself in competition with Danish priests; the Danes incited “their” Estonians to attack those whom Henry regarded as “our” Estonians. Not surprisingly, neither the bishop of Riga nor the Sword Brethren were willing to accede to Danish demands.70 In 1221 the Sword Brethren strengthened and garrisoned the fortresses from which they governed Saccala and Ugaunia.71 At this point King Valdemar compromised, claiming only the northern provinces of Harria, Vironia, and Jerwia in return for Rigan military assistance in his expedition to Ösel in 1222. After building a fortress there, he returned home, only to be kidnapped in May 1223 by a Saxon enemy – an unforeseeable event which for the Rigan church had the advantage of ensuring that more ambitious Danish claims would not be revived.72 In the meantime, taking advantage of Valdemar’s departure, the Öselians captured and dismantled the Danes’ fort. This success inspired a massive Estonian revolt.73 In 1223 Fellin, Odenpäh, and Dorpat (Tartu), the principal fortresses underpinning the Christian occupation of Saccala and Ugaunia, fell to insurgents. The Russians lent a hand; their king installed garrisons in Dorpat and Odenpäh “ut haberet dominium in Ugaunia et per totam Estoniam.”74 That the Germans and Danes eventually managed to defeat both the Estonians and their Russian allies in 1224 was in part due to a second unforeseeable event, the arrival of the Mongols and their victory over a large Russian army in May/June 1223.75 According to Henry, it was also due to a very deliberate policy of terror. The Rigans, he wrote, decided to storm “the very strong fort of Dorpat” and then slaughter all the defenders, “for in all forts hitherto taken by the church of Livonia, they have always been given life and freedom, and so the rest have not been made afraid.” After the successful assault, some women and all the 69 70

71 72 73 74 75

HCL XXIV, 3, pp. 172–73 (Brundage, pp. 190–91). HCL XXIII, 10–11, pp.167–68 (Brundage, pp. 186–87). On Henry’s anti-Danish sentiments, see Eva Eihmane, “The Baltic Crusades: A Clash of Two Identities,” in Clash of Cultures, ed. Murray, pp. 37–51, 40–41. HCL XXV, 5, p. 185 (Brundage, p. 204). HCL XXVIII, 1, p. 200 (Brundage, p. 220). He was not released until 1227. HCL XXVI, 2–4, pp. 187–89 (Brundage, pp. 205–07). HCL XXVII, 3, p. 196 (Brundage, pp. 216–17). This assumes that Henry’s report of a great Mongol victory in battle bringing the Russians to seek peace, which he placed at the beginning of his account of the bishop’s twenty-fourth year, 1222–23 (HCL XXVI, 1, pp. 186–87; Brundage, p. 205), actually relates to the presence of Russian envoys at Riga at the end of 1223 (XXVII, 6, p. 199; Brundage, p. 229). Continuing uncertainty about the intentions of the Mongols may well explain why the Novgorodian attempt to relieve their garrison in Dorpat in August 1224 came all too late: HCL XXVIII, 6, p. 205 (Brundage, p. 226). On the battle of May–June 1223 and Russian internal politics, see John Fennell, The Crisis of Medieval Russia 1200–1304 (London, 1983), pp. 57–58, 63–68, 71–72.



Henry of Livonia and the Conquest of Estonia

199

Estonian and Russian men were killed – bar one Russian who was given a horse and told to take to the people of Novgorod the news of what had happened to their fellow countrymen. In consequence “fear of the Rigans and Germans fell upon all the neighboring lands and peoples.”76 The Conquest of Estonia, Phase Four: Ösel, 1227 After two years of peace (1225–26), the final phase in Henry’s account of the conquest of Estonia was squeezed into a couple of weeks early in 1227. Following the papal legate’s call for a crusade against the Öselians, a large Rigan army crossed the ice, besieged, and after a fierce struggle (graphically described by Henry) burned down, the stronghold of Maalinn on the island of Mone. This sufficed to persuade those inside the main Öselian stronghold Wolde (Valjala) to capitulate as soon as the crusaders sat outside their walls.77 At this point Henry’s chronicle ends with a paean in praise of the blessed Virgin and Riga’s military power.78 The strategy of conquest As Brundage observed, “The very repetitiveness of the chronicler’s account year after year testifies to the systematic strategy that had by this time been worked out for the Christianization and subjugation of the Baltic peoples.” It is, however, not quite right to continue by saying that “the process began with military expeditions against one after another of the Estonians’ fortified strongholds.”79 Henry’s often very vivid descriptions of attacks on forts have attracted much interest, in part owing to the information he supplies about the use of siege artillery.80 But, as Friedrich Benninghoven long ago showed, the fundamental ingredient in warfare in the East Baltic region throughout the thirteenth and fourteenth centuries was not the attack on a stronghold but the Reise.81 There were, it is true, Reisen which targeted forts and involved sieges, but most Reisen are best characterized as chevauchées – or, in common usage, as raids: military expeditions which set out neither to capture an enemy fortress nor to bring enemy forces to battle. The objective of raids was to do as much damage and take as much plunder as possible, and then return home safely with all loot intact. Some raids clearly also had another purpose: to persuade 76 77 78 79 80 81

HCL XXVIII, 5–6, pp. 203–5, XXIX, 1, p. 207 (Brundage, pp. 224–26, 229). HCL XXX, 1–5, pp. 215–21 (Brundage, pp. 238–450. In this campaign, unlike those on mainland Estonia, Bishop Albert once again took a prominent role. HCL XXX, 6, pp. 221–22 (Brundage, pp. 245–46). Brundage, p. xxii. See, for example, Jim Bradbury, The Medieval Siege (Woodbridge, 1992), pp. 259, 271–72, 274. Friedrich Benninghoven, “Zur Technik spätmittelalterlicher Feldzüge im Ostbaltikum,” Zeitschrift für Ostforschung 19 (1970), 631–51. I am grateful to Carsten Selch Jensen for sending me a copy of this fine analysis.

200

John Gillingham

the victims that their rulers could not protect them and that a change of political allegiance was advisable.82 Compared with the Reise both siege and battle were of secondary importance.83 Henry’s narrative certainly bears out Benninghoven’s analysis. Henry described or mentioned more than fifty expeditions (expeditiones) launched by the Rigans and Danes during the twenty years (1208–27) of the Estonian war, including six which went into Russian territory. He included accounts of seven sieges of Estonian forts by Rigan and Danish forces.84 In addition he mentioned, though without supplying details, seven other occasions when the Rigans targeted Estonian forts.85 In all then, Estonian fortresses appear to have been major targets on fourteen occasions, but were ignored or by-passed during at least thirty-six Rigan and Danish campaigns. In other words, in more than two out of every three expeditions, the Rigans focused on ravaging and plundering the countryside and its people. Henry also described or mentioned thirty-six raids launched against the Rigans and their adherents by the Estonians and their allies (twenty-eight by Estonians, seven by Russians, and one by Lithuanians).86 Out of these thirty-six only five involved sieges.87 All this is similar to the result obtained by Paravicini in his study of the Preussenreisen. He distinguished between “die Belagerungsreise,” in which the main purpose of the campaign was the siege of an important fort, and “der Verheerungsfeldzug,” during the course of which a wood and earthwork fort might occasionally be attacked, but

82 83

84

85

86

87

In Henry’s view, in 1220 the Danes instigated raid after raid against the Jerwanians in order to push them out of the Rigan camp and into their own: HCL XXIV, 2, pp. 171–72 (Brundage, p. 190). In Bennighoven’s view, the Reise’s importance had been badly underestimated in previous scholarly literature, but he cited no general study of medieval warfare later than J. F. Verbruggen, De Krijgskunst in West-Europa in de Middeleeuwen (Brussels, 1954); he evidently did not know R. C. Smail, Crusading Warfare 1097–1193 (Cambridge, 1956), nor H. J. Hewitt, The Black Prince’s Expedition of 1355–1357 (Manchester, 1958), and idem, The Organization of War under Edward III (Manchester, 1966). Smail and Hewitt had already shown that there was much more to medieval warfare than battles and sieges. 1. Siege of Fellin in March 1211; 2. Leole in 1215; 3. Sontagana in Jan 1216; 4. Fellin again, in 1223; 5. Lone/Loal in January 1224; 6. Dorpat in 1224; 7. Mona in 1227 (HCL XIV, 11, p. 83, XVIII, 7, p. 120, XIX, 8, p. 133, XXVII, 2, p. 195, XXVII, 6, p. 198, XXVIII, 5–6, pp. 202–05, XXX, 4, pp. 217–20; Brundage, pp. 105, 139, 153, 215, 218–19, 223–26, 241–43). On the relatively small number of sieges, see Lang and Valk, “An Archaeological Reading,” 305. 1. Odenpäh in 1208; 2. Odenpäh again in 1210; 3. Three unnamed forts in Sontagana in Jan. 1211; 4. Owele and Purke in 1211; 5. Dorpat in 1211; 6. Danes in 1219 at Reval/Tallinn; 7. Dorpat in 1223 (HCL XII, 6, pp. 62–63, XIV, 6, p. 78, XIV, 10, p. 82, XV, 2, p. 88, XV, 7, p. 96, XXIII, 2, p. 154–55, XXVII, 4, p. 197; Brundage, pp. 84, 100, 103, 109, 117, 173, 217). For other estimates of number, see Sne, “Emergence of Livonia,” p. 69. A precise count of those raids which Henry referred to is not possible since he sometimes used phrases such as “they continually invaded,” e.g. HCL XXV, 5, p.185 (Brundage, p. 204). Of Fort Beverin for one day in 1208; Odenpäh by Russians in 1210; Wenden in 1210; Odenpäh again in 1217; Leal in 1220 (HCL XII, 6, p. 63, XIV, 2, pp.73–74, XIV, 7, p. 78, XX, 7, pp.139–40, XXIV, 3, p. 172; Brundage, pp. 85, 95, 100, 159, 191).



Henry of Livonia and the Conquest of Estonia

201

the point of which, as his choice of label indicates, was to ravage. Paravicini counted thirty-five of the former and 127 of the latter.88 From Henry’s narrative it is possible to discern a developing strategy with regard to fortresses. In the early years of the Estonian war captured fortresses were not retained, but merely sacked and burned down. An important stage was reached in 1216 when the Ugaunians asked the Sword Brethren to help them hold Odenpäh against the Russians.89 By the 1220s the presence of Sword Brethren and German traders in the Ugaunian and Saccalian centers of Fellin, Odenpäh, and Dorpat, both symbolizing and upholding Rigan domination of those provinces, meant that these fortresses became the focal points of the Estonian war of independence (1223–24).90 Siege warfare became significantly more prominent in the 1220s than in earlier years. The place of attacks on fortresses in Henry’s understanding of warfare before 1220 is indicated by the frequency with which words for siege (obsidio) and capture (expugnatio) of fortresses occur in his rubrics (subheadings). In the eighty-three rubrics which relate to the conquest of Estonia in the period from 1208 up to December 1219 (which is as far as the only manuscript containing rubrics goes), obsidio and expugnatio occur ten times. By contrast the word for a Reise (expeditio) occurs no less than forty times, in these rubrics far more often than any other noun.91 Battle There are, of course, encounters which are not easy to classify, particularly when, as here, we are dependent upon a single source. Defining “battle” as “a head-to=head encounter between the main forces operating in a theater of war,” I count ten battles in Henry’s narrative of the Estonian war.92 These ten are: 1. The Estonian victory at the Ümera in 1210 (HCL XIV, 8, pp. 79–80, Brundage, pp. 101–02); 2. The Rigan victory near Caupo’s fort in 1211 (HCL XV, 3, pp. 90–91, Brundage, pp. 111–13); 3. The Rigan victory near Fellin in 1217 (HCL 88

89

90

91 92

Werner Paravicini, Die Preussenreisen des europäischen Adels (Sigmaringen, 1995), 2, pp. 56–59. Paravicini also identified a third type of campaign: die Baureise. This seems to have been the preferred Danish strategy in 1206, 1219, and 1222, but to judge from Henry’s narrative, the Rigans adopted it only when Bishop Albert advanced on Kukenois in 1209 (HCL X, 13, p. 43, XIII, 1, p. 66, XXIII, 2, pp. 154–55, XXVI, 2, pp. 187–88; Brundage, pp. 64, 88, 173–74, 205–06). The Germans, however, were soon forced to leave Odenpäh: HCL XX, 5, 8, pp. 137, 140 (Brundage, pp. 157, 160). The Danes garrisoned Tallinn in 1219: XXIII, 2, pp. 154–56 (Brundage, pp. 173–74). After the shock and losses suffered by the Rigans in 1223, when they recovered and rebuilt these fortresses, this time they systematically kept Estonians out (at least in the case of Odenpäh): HCL XXVI, 5–7, pp. 189–90, XXVIII, 8–9, p. 206 (Brundage, pp. 207–09, 227–28). Brundage’s decision to omit the rubrics means that this kind of arithmetic cannot be attempted on the basis of the English translation. For this definition, see John Gillingham, “‘Up with Orthodoxy!’ In Defense of Vegetian Warfare,” Journal of Medieval Military History 2 (2004), pp. 149–58 at 152–53. In the rubrics to Henry’s narrative (i.e. up 1219 only) words for battle (bellum and pugna) occur only eight times.

202

John Gillingham

XXI, 2–4, pp. 142–44, Brundage, pp. 161–63); 4. The Rigan victory in Metsepole in 1218 (HCL XXI, 7, p. 146, Brundage, p. 165); 5. The Rigan victory over the Rus in Puidise in 1218 (HCL XXII, 2–3, pp. 148–50, Brundage, pp. 167–69); 6. The Danish victory at Tallin/Reval in 1219 (HCL XXIII, 2, pp. 155–56, Brundage, pp. 173–74); 7. The Rigan victory near Carethen in 1220 (HCL XXIII, 9, pp. 165–66, Brundage, pp. 184–85); 8. The Estonian victory over the Swedes at Leal in 1220 (HCL XXIV, 3, pp. 172–73, Brundage, p. 191); 9. The German and Danish victory at Reval/Tallin in 1223 (HCL XXVI, 11, p. 192, Brundage, p. 211); 10. The Rigan victory in the second battle at the Ümera in 1223 (HCL XXVII, 1, pp. 193–94, Brundage, p. 214). By one measure battles were frequent: ten battles in sixteen years of war (1208–12, 1215–24, 1227). By another measure they were rather infrequent; it appears that only ten out of at least eighty-six raids involved a battle – despite the fact that all expeditions into enemy territory unavoidably carried a risk of being brought to battle by defenders who wanted either to protect their territory or to punish invaders on the way home. Four battles fall into this category: three times victory went to the defenders, in 1211, in Metsepole in 1218, and at the Ümera in 1223; but on one occasion, in 1210, an over-eager pursuit enabled the raiding force, in this case Estonians, to win. Obviously the greater immobility incurred by engaging in siege warfare or by building a fortress necessarily increased the risk of being brought to battle. Hence five battles occurred during the nineteen campaigns which revolved around fortresses: the Estonian sieges of Wenden in 1210 and of Caupo’s fort in 1211, of Tallin/Reval in 1219, at Leal in 1220, and Reval/Tallin in 1223.93 In the more routine context of the Reise, battles were distinctly more unusual. The Germans among the Rigan forces enjoyed a marked superiority in arms and armor over Estonians and Russians.94 Even so, German-led invading forces very rarely sought to bring their enemies to battle. One battle occurred more or less by accident in 1220 when two raiding armies unexpectedly bumped into each other.95 Only two of the ten battles were actively sought by the aggressor. In 1218 the Rigans were on their way north to ravage the region around Tallin/Reval when they learned from captured Russian and Öselian messengers that plans for a union of Estonian and Russian forces were well advanced and that a Novgorodian army had already set out. At this point the Rigans abruptly changed direction and succeeded in intercepting the Russians before they could join their allies. Only once did invaders seek battle right from the start of the expedition – this time because the Rigans received early warning of a planned union of Estonian and Russian forces. They 93 94 95

I count as battles two occasions when fighting occurred as a result of sorties out of newly acquired strongpoints, by the Swedes at Leal in 1220 and by Danes and Germans at Reval/Tallinn in 1223. As is clear from Henry’s references to the equipment of the two sides in his account of the battle near Caupo’s fort in 1211. In 1220 a Rigan force passing through Jerwan on its way to raid Harrien stumbled across Öselians who were busy raiding Jerwanians. In consequence battle ensued at Carethen, after which the Rigans resumed their raid on Harrien.



Henry of Livonia and the Conquest of Estonia

203

moved rapidly and managed to reach the Estonians and bring them to battle in September 1217, near Fellin, before the Russians arrived. Unquestionably victory in battle brought some advantages, and the fact that Rigan and Danish forces won eight out of the ten battles clearly contributed to the conquest of Estonia. The victory at Fellin in 1217, in which the Saccalian chief Lembit was killed, induced the Saccalians to submit (again), and permitted the Germans to ravage the Maritime Provinces, inducing their leaders to submit to baptism and payment of annual tribute. The victory in 1218, on the other hand, did not prevent the Russians resuming their advance three days later, going on to ravage Liv and Lettish settlements and churches. What it did do, according to Henry, was dissuade the Öselians and Saccalians (evidently still smarting from their defeat in 1217) from joining the Russians.96 More decisive than either of these battles was the encounter in 1211 when the defeat of Estonians besieging Caupo’s fort led to such heavy casualties among their leaders that Henry proclaimed that “in that battle Estonia’s head fell.”97 As will become apparent, the killing of leaders was an important war aim, but it could be achieved without having to take the risk of battle. According to Henry, for example, in a raid in December 1208 his Letts killed no less than three hundred of the leading Saccalians.98 Raiding Since most expeditions involved neither siege nor battle, it follows that it was the ravaging and plundering of raiding which lay at the heart of the strategy of conquest – all the more so since it is clear that ravaging and plundering the countryside were also essential ingredients of those campaigns in which fortresses were attacked. “This must have been,” as Villads Jensen has recently emphasized, “a deliberate strategy to weaken the enemy. And of course this worked both ways, so that the pagans also raided the Christians in a neverending cycle of retribution.”99 In 1211, for example, various Estonian forces invaded the territory of the bishop’s Livs: “one army followed another, when one returned another came, so that day and night they gave the Livs no rest.”100 On the Christian side Henry made the underlying strategy crystal clear in his account of the ravaging, burning, and plundering of Ugaunia by a combination of Letts, Sword Brethren, and episcopal troops in 1215.

Even this gain was limited, however, by the fact that the Harriens did join the Russians, while the Öselians raided along the Dvina. In retaliation the Letts raided Russian territory: HCL XXII, 8, p. 152 (Brundage, p. 171). 97 HCL XV, 3, p. 91 (Brundage, p. 113). 98 This, he noted, persuaded the Saccalians to ask for a truce: HCL XII, 6, pp. 64–65 (Brundage, p. 87). 99 K. V. Jensen, “Bigger and Better,” p. 260. 100 HCL XIV, 12, p. 86 (Brundage, p. 108). 96

204

John Gillingham

Having done as much damage there as they could, they returned to Livonia, arranging for others to go immediately into Ugaunia and do similar damage. And when these returned, then still others were dispatched. The Letts continued in this way, neither giving the Estonians any rest nor taking any rest themselves, until in that one summer nine armies, ravaging one after the other, left the province so desolate and deserted that neither people nor food could be found there. Their aim was to keep on making war until either those who survived came to ask for peace and baptism or they were completely extirpated from the land.101

According to Henry, this was the dilemma which confronted the Estonians from 1208 onwards: either submit and convert to Christianity or face the consequences.102 No earlier medieval chronicle provides a description of raiding warfare to match Henry’s narrative. He was an eyewitness of some raids, observing them from the point of view both of the invader and the invaded, and he did so in a remarkably even-handed way. In its even-handedness his chronicle is very different from the only detailed narratives to have been composed earlier: those twelfth-century English accounts of Scottish and Welsh raids which so strongly adopt the victims’ point of view that they were long discounted as nothing more than propagandistic atrocity stories.103 Henry, by contrast, not only laments the damage which his own people suffered at the hands of their enemies, but also notes the damage which his people inflicted, sometimes exultingly, sometimes in a more somber tone.104 101 “Cogitabant

enim eos tam diu debellare, donec aut pro pace et baptismo venirent, qui residui erant, aut omnino eos exstirpare de terra.” HCL XIX, 3, p. 126. Cf. Brundage’s “or they would be completely wiped out from the earth,” p. 146. For the meaning of ex(s)tirpare see below, p. 00 102 HCL XII, 6, p. 64 (Brundage, p. 85). 103 Because authors such as Ailred of Rievaulx accused the invaders of activities such as drinking children’s blood, they made it easy for modern historians to dismiss their accounts. But for the grim realities behind such all too colorful details see Robert J. Bartlett, “Technique militaire et pouvoir politique, 900–1200,” Annales ESC (1986), 1135–59 at 1147–49, recapitulated in English in Robert Bartlett, The Making of Europe: Conquest, Colonization and Cultural Change 950–1350 (Harmondsworth, 1993), chapter 3; also Matthew Strickland, War and Chivalry: The Conduct and Perception of War in England and Normandy, 1066–1217 (Cambridge, 1996), pp. 313–17; and Strickland, “Rules of War or War without Rules? Some Reflections on Conduct and the Treatment of Non-Combatants in Medieval Transcultural Wars,” in Transcultural Wars from the Middle Ages to the 21st Century, ed. Hans-Henning Kortüm (Berlin, 2006), pp. 107–40 at 118–26. The inclination to treat such narratives simply as rhetorical exercises remains visible in Gabriela Signori, “Frauen, Kinder, Greise und Tyrannen. Geschlecht und Krieg in der Bilderwelt des späten Mittelalters,” in Bilder, Texte, Ritual Wirklichkeitsbezug und Wirklichkeitskonstruktion politisch-rechtlicher Kommunikationsmedien in Stadt– und Adelsgesellschaften des späten Mittelalters, ed. Klaus Schreiner and Gabriela Signori (Berlin, 2000), pp. 139–64 at 145ff. 104 As well as noting that the Livs and Letts returned joyfully after a successful raid on Jerwia in 1212, he added: “Erat autem tunc villa Carethen pulcherrima et magna et populosa, sicut omnes ville in Gerwen et in tota Estonia fuerunt, que postmodum omnes sepius a nostris vastate et incense sunt,” HCL XV, 7, p. 98 (Brundage, p. 118).



Henry of Livonia and the Conquest of Estonia

205

What emerges most clearly from his repetitive phrases is the importance of taking the intended victims by surprise. Time and again a successful raid found its intended victims going about their usual business. In 1222, for example, raiders entering Novgorodian territory “found that land full of people whom no reports had forewarned.”105 It is clear that all were aware of the vital importance of accurate news-gathering.106 His narrative of a raid in 1211 reveals in passing that communities stationed “guards of the roads” (custodes viarum) whose job it was to give timely warning.107 If forewarned, the intended victims had time to take themselves and their animals into fortresses or hiding places in the forests where, with luck, they could wait until the raiders moved on or withdrew.108 Worse still for the raiders, of course, was the situation in which the defenders had time to take active counter-measures. On one occasion in 1208, the invaders’ ravaging parties entered empty settlements. Realizing what this portended, they hastily reunited their forces and retreated, but suffered heavy losses as they did so.109 In order to evade the guards of the roads, invaders commonly marched at night or chose difficult and therefore unexpected routes. In 1211, for example, the Rigans chose a route which took them six days, the last three days “by a very bad road through swamps and forests.” This succeeded in taking the Saccalians by surprise, but cost the lives of almost a hundred horses.110 Not quite so successful was the Rigans’ march into the province of Reval first across the sea-ice and then by night in the teeth of a freezing north wind in February 1219. The raiders reached their target while it was still dark, but some were so cold that to warm themselves up they set fire to a settlement. Alerted by the flames, many Estonians had time to flee to their hiding places.111 On reaching enemy territory, each contingent picked a base of operations and then scattered in order to ravage and plunder – normal practice made all the more necessary by the fact that their intended victims lived in small, dispersed settlements. When the burning and looting was over, the invaders concentrated invenerunt terram illam repletam hominibus et nullis rumoribus premunitam,” HCL XXV, 6, p. 185. The translation “they found that land full of men and forewarned by no rumours” (Brundage, p. 204), is doubly misleading. 106 Translating rumores as “rumours” (for other examples see HCL XI, 5, p. 50, XVIII, 5, p. 119, Brundage, pp 70–7, 138) has probably led to an underestimate of the extent to which the vital importance of news-gathering was recognized. 107 HCL XIV, 10, p. 82 (Brundage, p. 103). 108 Of course, they were not always lucky: see Torben Kjersgaard Nielsen, “Henry of Livonia on Woods and Wilderness,” in Crusading and Chronicle, ed. Tamm et al., pp. 157–78 at 163–67. 109 HCL XII, 2, pp. 59–60 (Brundage, p. 80). According to Henry, when the Germans had news of a Lithuanian raid in 1214 they rejoiced, and were able to organize a successful ambush: HCL XVII, 6, pp. 114–15 (Brundage, p. 134). 110 HCL XV, 7, p. 94 (Brundage, p. 115). Moving at night was routine. For two early examples, see HCL XII, 6, pp. 62, 64 (Brundage, pp. 84, 86). 111 Even so, though many lost toes, fingers and noses to frostbite, the invaders managed to take prisoners and plunder enough to return home in good cheer, HCL XXII, 9, p. 153 (Brundage, pp. 171–72). 105 “Et

206

John Gillingham

their forces again at their agreed assembly point, for which the Estonian word, Henry noted, was maia.112 They then had to get home safely with their plunder and prisoners.113 The return journey was predictably dangerous, and nerves could be on edge. On one occasion in 1216 as they marched back over the sea-ice after a raid on Ösel, panic caused by a cry that they were being pursued led to some falling over and freezing to death.114 Occasionally returning raiders decided to jettison all their gains in an attempt to escape pursuit.115 It is worth noting, however, that raiders did not always return at speed. Henry’s account of the first raid launched against the Saccalians in December 2008 says they travelled back slowly, often halting and waiting, prepared for the eventuality that the surviving Estonians might fall upon them from the rear. Similarly at the end of the 1219 raid on Tallin/Reval, the carefully organized return journey took ten days over the ice.116 The over-hasty pursuit of an invader who conducted a skillful withdrawal could lead to disaster, as the Rigans discovered in 1210 when they lost what Henry called the first battle on the Ümera.117 Killing men, capturing women and children In many respects these raids were similar to chevauchées such as those during the Hundred Years War. But in other very important ways they were different. In these wars ransom is notable by its absence. Only once in the whole of his chronicle did Henry use the word redemptio, when the Lithuanians captured Uldewene, a leader of the Livs of Lennewarden, and then exchanged him for the head of one of their own dead chiefs.118 Only rarely does Henry report that men were taken prisoner, and those who were, were often spared only to be put to death soon afterwards.119 They might be slaughtered within sight of their

the routine of dispersing then re-assembling see, e.g., HCL XVIII, 5, p. 119 (Brundage, p. 138); for the word maia, XV, 7, p. 94 (Brundage, p. 115); XXIII, 7, p. 160 (Brundage, p. 178); XXIII, 9, p. 165 (Brundage, p. 184). 113 There are many examples of this. For a case in which Henry shared, see HCL XII, 6, p. 65 (Brundage, p. 87). On the theme of the joyous return, see Jaan Undusk, “Sacred History, Profane History: Uses of the Bible in the Chronicle of Henry of Livonia,” in Crusading and Chronicle, ed. Tamm et al., pp. 45–75 at 71–73. 114 HCL XIX, 9, p. 134 (Brundage, p. 154). 115 As Estonian raiders did in 1208 and again, this time in vain, in 1223: HCL XII, 6, p. 64, XXVII, 1, p. 194 (Brundage, pp. 86, 214). 116 In 1206 “revertentes paulatim multis diebus per viam moram faciebant, existentes parati si forte Estones residui bellum eis post tergum intulissent.” As it turned out, however, the enemy had collected the bodies of their dead, cremated them, and held funerals with much lamenting and drinking, “according to their custom,” HCL XII, 6, p. 65 (Brundage, pp. 86–87). Cf. “revertentes paulatim in glacie decem diebus propter captivos et predam moram fecimus,” XXII, 9, p. 153 (Brundage, p. 172). 117 HCL XIV, 8, pp. 79–80 (Brundage, pp. 101–02). 118 HCL XVII, 5, p. 114 (Brundage, p. 134). 119 At times there appears to have been a ritual element in these killings: HCL XIV, 8, p. 80, XV, 7, p. 96, XXVI, 6, 13, pp. 190, 193 (Brundage, pp. 102, 116, 209, 212). 112 For



Henry of Livonia and the Conquest of Estonia

207

friends, as in 1211 when the Letts, in an attempt to terrorize the defenders of Fellin into surrender, first threatened to kill prisoners and then carried out the threat, throwing the corpses into the moat.120 Other killings were carried out in a spirit of vengeance. When Thalibald of Beverin, one of the Lettish chiefs most prominent in Henry’s narrative, was captured by Estonians and roasted to death, his sons, accompanied by Sword Brethren and other Germans, raided Ugaunia and burned alive all the men they could capture, before carrying off their women and children.121 A few men were taken prisoner by Estonians and kept alive in order to be exchanged – a practice presumably stimulated by the Rigan custom of requiring hostages from those Estonian peoples who submitted to baptism. In 1223, for example, the Saccalians liberated their hostages in return for some merchants and Sword Brethren captured at Fellin and Dorpat.122 It may have been for this purpose in this same year that a Saccalian and Ugaunian raiding army kept male prisoners alive long enough for them to be rescued in 1223.123 Just twice Henry made a point of the fact that his own people had taken male (by implication adult male) as well as female prisoners, and both times it was in 1221 in the context of raids into Russian, that is Christian, territory.124 By contrast it soon becomes all too clear that his concluding words on his own parishioners’ first raid into Saccala are entirely typical. “Finally exhausted by so much killing, they collected all their plunder and returned home, bringing back with them herds of animals and a great many girls, the only people whom the armies of this region normally spare.”125 The prisoners being carried off by a pagan Öselian raiding army who, seizing the opportunity presented by the accidental encounter referred to above (p. 201), attacked their captors from behind and won back their freedom, were apparently all female.126 Russian raiders acted in the same way. When they attacked the Ugaunians in 1216 (so after the latter had submitted to Riga), they “killed many males and carried women and children into captivity.”127 Despite the lack of direct evidence for the conduct of war in this region in earlier periods, it seems highly likely that people had acted in this way in the years before Henry arrived on the scene to record their military exploits.128 But the practice was not confined to pagans,

120 HCL

XIV, 11, p. 84 (Brundage, p.105). XIX, 3–4, pp. 124–26 (Brundage, pp. 145–46). 122 HCL XXVI, 5, 7, 9, pp. 189–91 (Brundage, pp. 208–10). 123 HCL XXVII, 1, p. 194 (Brundage, p. 214). 124 HCL XXV, 5–6, p. 185 (Brundage, p. 204). 125 “Villisque omnibus sanguine multo paganorum coloratis sequenti die redierunt et per omnes villas spolia multa colligentes iumenta et pecora multa et puellas quam plurimas, quibus solis parcere solent exercitus in terris istis, secum abduxerunt,” HCL XII, 6. pp. 64–65 (Brundage, p. 86). 126 HCL XXIII, 9, p. 166 (Brundage, p.185). 127 “multos viros interfecerunt mulieresque et parvulos captivos deduxerunt,” HCL XX, 3, p. 136 (Brundage, pp. 156–57). 128 Hence I am sceptical of the view that prior to the arrival of Germans there had been an “eternal, relatively bloodless petty warfare among neighbours,” William Urban, The Baltic Crusade of The Thirteenth Century (London, 1979), p. 38. 121 HCL

208

John Gillingham

Orthodox Christians, and recent converts such as the Letts. The Germans, the bishop’s military household, and the Sword Brothers regularly pursued the same policy. This, for example, is how Henry described a Rigan raid into Estonia in 1216: “When we arrived there we burned and devastated everything, killed all the males, captured women and children, and drove off their horses, cattle and sheep.”129 Laconic phrases stating explicitly that males were killed and females captured are repeated over thirty times in twenty passages.130 Moreover this was Latin Christian warfare not only as practiced against pagans, but also, after initial hesitation, from 1217 on against Orthodox Christians.131 It was war as a hunt for women and children with men being killed in pursuit of this end.132 Consider Henry’s narrative of the attack on Ugaunia in 1215: “They seized those who came out of the forest to get food in the fields or in their settlements. Burning some and cutting the throats of others, they tortured them until they […] led them to the secret refuges in the forest, and delivered the women and children into their hands.”133 Just occasionally Henry explicitly noted that women and children were killed too, for instance during the first raid on Saccala in December 1208. When he recorded this happening again in 1215 he was moved to remark that the Livs and Letts were crueler than other peoples.134 There may also have been other occasions when women were among the “people” who were killed, though whenever the word homines has been translated as “men,” this possibility has been

129 “Quo

cum pervenimus, exercitum nostrum per omnes vias ac villas nec non et provincias illius terre divisimus, incendentes omnia et vastantes, quidquid masculi sexus interficientes, mulieres et parvulos capientes, pecora multa nec non et equos eorum minantes,” HCL XX, 2, p. 135 (Brundage, p. 156). 130 HCL XIV, 6, p. 78 (1210); XIV, 12, p. 86 (1211); XV, 2, p. 88 (1211); XV, 7, pp. 94–96 (1211); XVI, 8, p. 112 (1213); XVIII, 5, p. 119 (1215); XIX, 3, pp. 125–26 (1215); XIX, 8, 9, 11, XX, 2–3, pp. 133–36 (1216); XX, 6, p. 138 (1217); XXI, 5, 7, pp. 145–46 (1218); XXII, 4, p. 150 (1218); XXIII, 5, p. 158 (1219); XXIII, 6, p. 159 (1219); XXIII, 7, p. 160 (1219); XXIII, 9, p. 167 (1220); XXVI, 13, p. 193 (1223); XXVII, 1, p. 193 (1223); XXVIII, 6, p. 205 (1224). 131 “Ruthenos propter nomen christianitatis non audent interficere,” HCL XI, 8, p. 56 (Brundage, p. 76) in 1208; “Theutonici […] pre reverencia christiani nominis paucos occidentes,” XIII, 4, p. 70 (Brundage, p. 91) in 1209; “Interfecerunt populum multum et mulieres quam plurimas captivas duxerunt et equos et pecora multa depellentes spolia multa tulerunt,” XX, 5, p. 138 (Brundage, p. 157–58) in 1217; “viros interficientes et mulieres captivantes” (twice), XXIII, 5, p. 158–59 (Brundage, p. 177) in 1219. For discussion of similar formulae in Christian– Muslim warfare in Outremer see Yvonne Friedmann, Encounters between Enemies. Captivity and Ransom in the Latin Kingdom of Jerusalem (Leiden, 2002), pp. 163–66. 132 The extent to which females were taken has sometimes been underestimated as a consequence of the “natural” translations of parvulos and pueros as “boys” rather than as “children.” 133 “Venientes de silvis ad agros aut ad villas pro cibariis comprehendunt et alios igne cremantes, aliis gladiis iugulantes diversis tormentis afficiunt, donec omnes pecunias suas eis aperiunt, donec ad omnia nemorum suorum latibula eos deducunt et mulieres et parvulos in manus eorum tradunt. Sed nec sic quidem mitigatur animus Lettorum […] pertranseuntes omnes provincias […] nemini parcunt, sed quidquid masculine sexus est, interficiunt, mulieres et parvulos captivos trahunt,” HCL XIX, 3, p. 126 (Brundage, p. 145). 134 HCL XII, 6, p. 64 (Brundage, p. 86); XVIII, 5, p. 119 (Brundage, p. 138).



Henry of Livonia and the Conquest of Estonia

209

rendered invisible to anyone dependent upon the English translation.135 More clearly than in Henry’s laconic phrases, Richard of Hexham’s narrative of Scottish raids in 1138, for all its propagandistic language, makes plain why raiders aiming to capture women and children in practice found it necessary to kill some of them too: babies and nursing mothers as well as the elderly of both sexes.136 A question arises: when the raiders shared out their booty at the end of a successful expedition, what happened to the women and children they had carried off? Whether the raiders were pagan or Christian, this is a subject on which Henry was remarkably unforthcoming.137 Only once did he claim to know what his pagan enemies did with their captives. Writing about what had provoked the papal legate to preach a crusade against the Öselians in 1227, he asserted that they treated their captives, young women and virgins, very badly, violating them or taking them as their wives, each man having two or three wives or more, while selling others to Curs and other pagans.138 Only once did he say what the Rigan forces did. This was in 1209 when Bishop Albert gave back the captives (the queen, women, and children) taken during his attack on Gerzike in return for the king’s submission and agreement to align his policies to suit the bishop.139 But this was before relations between Rigans and Russians had deteriorated, and while both sides were still more conscious of their common Christianity. By contrast not once did Henry say anything at all about the fate of the pagan women and children whom the Rigans captured in raid after raid. His silence here has allowed historians to tiptoe around the subject. Some follow Henry in mentioning enslavement by pagans, but either saying nothing about Christian practice or referring only to them using POWs as hostages.140 It seems much more likely, however, that, just as in later wars against Prussians and Lithuanians, the Christians enslaved their pagan captives.141 Chris135 HCL

XII, 6, p. 62 (Brundage, p. 84) in 1208. The fact that here the adjective took the masculine plural form (“homines paganos”) made the published translation seemed to confirm the natural one. But masculine plurals are deceptive. For other instances where “homines” has been rendered as “men” instead of as “people” see HCL XXII, 7, 9 (Brundage, pp. 171–72), XXV, 3, 6 (Brundage, pp. 201, 204), XXVII, 6, (Brundage, p. 219). 136 John Gillingham, “Conquering the Barbarians: War and Chivalry in Twelfth-Century Britain and Ireland,” Haskins Society Journal 4 (1992/93), repr. in Gillingham, The English, pp. 44–46. 137 “Es ist schwer, Aussagen über das Schicksal der im Laufe der baltischen Missionskriege gefangen genommenen und nach Auskunft der Chronisten auch deportierten Menschen zu treffen,” Müller, Information und Transformation, p. 312. 138 HCL XXX, 1, p. 216 (Brundage, p. 238). Mägi, however, only mentions the Öselians holding hostages: “Ösel and the Danish Kingdom,” p. 325. 139 HCL XIII, 4, pp. 69–71 (Brundage, pp. 91–93). On this episode see Torben K. Nielsen, “Sterile Monsters? Russians and the Orthodox Church in the Chronicle of Henry of Livonia,” in Clash of Cultures, ed. Murray, pp. 227–52 at 238–41. 140 Turnbull, “Crossbows or Catapults,” p. 311; Nielsen, “Sterile Monsters,” p. 248. See also William Urban, “The Frontier Thesis and the Baltic Crusade,” in Crusade and Conversion on the Baltic Frontier 1150–1500, ed. Alan V. Murray (Aldershot, 2001), pp. 45–71 at 61–62, 65. 141 Paravicini, Die Preussenreisen, 2, pp. 101–10; Sven Ekdahl, “The Treatment of Prisoners of War during the Fighting between the Teutonic Order and Lithuania,” in The Military Orders: Fighting for the Faith and Caring for the Sick, ed. Malcolm Barber (Aldershot, 1994), pp. 263–69 at 265–67; idem, “Christianisierung – Siedlung – Litaureise. Die Christianisierung Litauens als

210

John Gillingham

tians in the Baltic region may well have behaved much like their fellow Christians in late medieval Spain and Italy, in that they owned slaves of both sexes, but chiefly female domestics: they were there to do the housework, look after children – and to provide sex.142 That Henry’s Christians kept pagan captives as slaves is implied by a letter from Pope Gregory IX to his legate in Livonia, instructing him on the proper consequences of baptism for those in servitude. He is to give them “the freedom of confessing their sins, going to church, and of hearing the divine office.” The instructions were no doubt carefully drawn; the freedom to be accorded to baptized slaves was very circumscribed.143 Probably most captives were eventually assimilated into the society which took them.144 But whatever improvements to the lives of such captives may have occurred over time, the fact remains that they had been violently carried off and their families had suffered the slaughter that went hand in hand with enslavement. From Henry’s narrative it seems clear that the strategy of making relentless and destructive war until the survivors either submitted to conversion or were “completely extirpated from the earth” did not involve the slaughter of every single man, woman, and child, as a translation of extirpare as “wipe out” might suggest. It seems rather that it involved the systematic capture and removal of women and children. Given the human preference for making a profit out of war, is it not likely that this was also what was meant in those very much briefer and more opaque early medieval narratives in which Latin words have often been given genocidal translations?145 Take, for example, Bede’s description of Dilemma des Deutschen Ordens,” in Die Christianisierung Litauens im mitteleuropäischen Kontext, ed. Vydas Dolinskas et al. (Vilnius, 2005), pp. 189–205 at 192–95; Alvydas Nikžentaitis, “Prisoners of War in Lithuania and the Teutonic Order State (1283–1409),” in Der Deutsche Orden in der Zeit der Kalmarer Union 1397–1521, ed. Zenon Hubert Novak and Roman Czaja (Toruń, 1999), pp. 193–208. I owe these last two items to the kindness of Sven Ekdahl. See also Christiansen, Northern Crusade, pp. 45–46; Jensen, “Bigger and Better,” p. 260. 142 For a study with references to earlier literature, see Jeffrey Fynn-Paul, “Tartars in Spain: Renaissance Slavery in the Catalan city of Manresa, c. 1408,” Journal of Medieval History 34 (2008), 348–59. 143 “[…] si quos de servili condictione seu alios alterius ditione subiectos ad baptismi gratiam […] de onere servitutis facias aliquid relaxari, et dari eis liberam facultatem confitendi peccata, adeundi ecclesiam et divina officia audiendi,” in Liv-, Esth- und Curländisches Urkundenbuch, ed. F. G. von Bunge et al. (Reval, 1853), 1, no. 158. For commentary see Benjamin Z. Kedar, Crusade and Mission: European Approaches toward the Muslims (Princeton, 1988), pp. 147–49; Tyerman, “Henry of Livonia,” p. 42. It is noteworthy that when Henry referred to what the Öselians did to their female captives as “illicita” it was on the grounds that it was not right for a pagan to be coupled with a Christian (HCL XXX, 1, p. 216, Brundage p. 238). This might be thought to imply that he had to accept that in this part of the world Christian lords might keep and have sex with their female slaves so long as they baptized them. 144 On “assimilation slavery” in Eastern Europe, see Henryk Lowmianski, “Die Sklaverei und ihre Rolle in der Entstehung der mittelalterlichen Gesellschaft bei den Slawen,” Studia Historiae Oeconomicae 1 (1966/67), 43–59. 145 “Images of ethnic purge and replacement were written into the very DNA of medieval religious and political assumption.” Len Scales, “Bread, Cheese and Genocide: Imagining the Destruction of Peoples in Medieval Western Europe,” History 92 (2007), 284–300 at 300.



Henry of Livonia and the Conquest of Estonia

211

Cædwalla’s conquest of the Isle of Wight: “stragica caede omnes indigenas exterminare,” translated by Bertram Colgrave as “to wipe out all the natives by merciless slaughter.”146 Or the Frankish annalist on Charlemagne’s determination to wage war on the Saxons until they were either subjected to the Christian religion or “entirely exterminated (‘aut omnino tollerentur’).”147 Or St. Bernard’s view that the Saxons in 1147 were to fight the enemies of Christ living beyond the Elbe “until either the people or their form of worship shall be destroyed.”148 Henry of Livonia’s repetitive phrases indicate that when a people was “extirpated” it was rather that their women and children were forcibly taken over by a different people, with the implication that any children these captives subsequently had would be children of the victors, not of the defeated.149 In this way a defeated people could be made to disappear – “vanish like the Avars,” as a proverb put it.150 In the medieval period the earliest fairly long accounts from the victims’ point of view of raids in which human beings were among the most desirable objects of plunder were composed in twelfth-century England, but they were long discounted as being unreliable anti-Scottish or anti-Welsh atrocity tales written by authors who were victims. Henry of Livonia offers us something entirely different. His narrative, I suggest, allows us to see something of the reality of raids, and from the points of view of both invader and victim.151 It gives us valuable insights into far more than just warfare in the East Baltic of his own day. It provides clues, perhaps the best clues we have, into what was going on behind the laconic phrases of early medieval authors such as Gregory of Tours, Bede, and Widukind of Corvey, and in this way sheds a harsh light on

Historia ecclesiastica gentis Anglorum 4.16, ed. and trans. Bertram Colgrave and R. A. B. Mynors (Oxford, 1969) pp. 382–83. Cf. Bede’s use of the verbs exterminare and eradere when describing the ambitions of the Northumbrian Æthelfrith and the Briton Cædwalla, ibid. 1.34, pp. 116–17 and 2.20, pp. 204–05. For discussion of these terms see James Fraser, “Early Medieval Europe” and Len Scales, “Central and Late Medieval Europe” in The Oxford Handbook of Genocide Studies, ed. Donald Bloxham and A. Dirk Moses (Oxford, 2010). 147 Annales Regni Francorum, ed. F. Kurze, MGH SRG 6 (Hanover, 1895), p. 41; trans. B. W. Scholz as Carolingian Chronicles (Ann Arbor, 1972), p. 51. 148 “Donec, auxiliante Deo, aut ritus ipse aut natio deleatur.” This, Bernard wrote, in a phrase to be echoed by Henry, was “ad faciendam vindictam in nationibus et exstirpandos de terra christiani nominis inimicos.” The text is printed and discussed by Hans-Dietrich Kahl, “Crusade Eschatology as seen by St Bernard in the Years 1146 to 1148,” in The Second Crusade and the Cistercians, ed. Michael Gervers (New York, 1992), pp. 35–47. See also Christiansen, Northern Crusades, pp. 53–54. 149 Thus putting “an end to existing pagan dynasties” – but not just ruling dynasties, as seems to be implied by Torben K. Nielsen, “Mission and Submission. Societal Change in the Baltic in the Thirteenth Century,” in Medieval History Writing, ed. T. M. S. Lehtonen and K. V. Jensen, pp. 216–31 at 222. 150 Walter Pohl, Die Awaren: ein Steppenvolk in Mitteleuropa, 567–822 n. Chr. (Munich, 1988), pp. 315–28. 151 As Paravicini observed of the Preussenreisen, “War was waged with extreme violence, yet with no qualms of conscience,” Die Preussenreisen, 2, p. 56. 146 Bede,

212

John Gillingham

other places and times when wars were fought according to values very different from those which came to prevail in the “Age of Chivalry.”

Appendix: Authorship of the Rubrics On manuscript grounds alone it is possible that the rubrics were composed c. 1300 by a later reader interpreting what Henry had written. In fact, however, the editors of both the 1874 (Wilhelm Arndt) and 1955 (Albert Bauer) editions regarded the rubrics as in some sense original: “wohl vom Verfasser der Chronik herrührend,” in Bauer’s words; while Arndt drew attention to two rubrics written in the first person plural and concluded that “eas in archetypo iam existisse patet.”152 Despite this, Arndt chose to relegate almost all the rubrics to footnotes, and then Brundage, in translating Arndt’s text, decided to omit all those so relegated. The rubrics indicate a text divided into thirty sections, the first two devoted to the missionary careers of the first two bishops, Meinhard and Berthold, and then each of the next twenty-eight to a year in the episcopate of Bishop Albert.153 This seems fairly straightforward, and on this basis Hansen in 1853 divided the text into thirty chapters.154 He also divided the chapters into subsections as seemed to him to make sense, and numbered them. The rubricator had also created subsections. Unfortunately Z was unknown to Hansen, and his subsections do not coincide with those of the rubricator. For example, Hansen divided chapter eleven into nine subsections instead of the rubricator’s sixteen, and chapter twelve into eight subsections instead of the rubricator’s nine. It might have made sense to replace Hansen’s divisions with those of the manuscript, but unfortunately Z is incomplete, missing about one third of the whole text. In consequence, since it is impossible to “follow” the rubricator throughout, it has seemed convenient to all subsequent editors to retain Hansen’s divisions and their numbering. A further complication: the rubricator referred to the first three of Hansen’s “chapters” as books, and inserted into the middle of Hansen’s chapter twelve the words: “Amen. Explicit liber IIIus de Lyvonia. Incipit liber quartus de Estonia.”155 Curi-

152 HCL,

p. xliv; MGH SS 23, p. 234, referring to two first-person rubrics in XIX, 5: “Qualiter ab ignibus liberati sumus per Dei misericordiam” and “Qualiter conclusionem portus evasimus.” For the story of Arndt’s discovery of Z in the 1860s, ibid., pp. 231–32. 153 The first is headed “Liber primus. De Lyvonia. De primo episcopo Meynardo”; the second “Liber secundus. De episcopo Bertoldo”; and the third “Liber tercius. De episcopo Alberto.” The fourth is headed “De secundo anno consecratione episcope Alberti,” the fifth “De anno tercio”, the sixth, “De anno quarto,” and so on until the thirtieth which begins with a reference to the bishop’s twenty-eighth year. 154 On Hansen’s edition see Tiina Kala, “Henry’s Chronicle in the Service of Historical Thought: Editors and Editions,” in Crusading and Chronicle, pp. 385–407 at 395–96. 155 HCL XII, 5–6, p. 61.



Henry of Livonia and the Conquest of Estonia

213

ously Arndt did not consign this to the footnotes, and in the English translation it is highlighted by being printed in bold and large type.156 Up to Bishop Albert’s eighth year, each rubric marking the beginning of the year retains that simple form – e.g., “De anno octavo.”157 The rubric at the start of the next year is “De baptismo Lettorum. Annus nonus.” Yet there were to be thirteen more rubrics dealing with many different matters in seven pages of printed text before the author reached the subject of the baptism of the Letts who lived in the Ümera valley. Evidently for the rubricator this event was the highlight of the whole year.158 On many different occasions groups of newly converted people were killed in fighting. But just once, in the rubric “De martyrio Lettorum ad Ymeram,” the fate of such a group is termed “martyrdom.”159 The Letts of the Ümera valley were Henry’s parishioners. It appears that the composer of the rubrics was as attached to the Letts as Henry was, and the most economical explanation for this is that the rubricator was, as Arndt and Bauer thought, Henry himself.

156 Brundage,

pp. 82–83. X, 1, p. 32. 158 HCL XI, 1, 7, pp. 47, 55. 159 HCL XIV, 8, p. 80. And see ibid., n. 1, pointing out that according to Henry only Livs (not Letts) fled from the battle and, though it later transpires that Livs too were killed in the pursuit, only Letts (not Livs) were martyred. 157 HCL

9 The English Longbow, War, and Administration John France

In England when medieval warfare is mentioned almost everybody immediately thinks of the longbow. It is the iconic weapon in English military experience and the symbol of victory. To discuss it evokes memories of Crécy and Agincourt. Its status was already long established when it was fixed in people’s minds by Laurence Olivier’s famous film of Shakespeare’s Henry V with the thrilling scenes of Agincourt. Significantly, this was essentially a propaganda film, made during the Second World War. A well-known account of Agincourt speaks of the volleys of arrows making “a weird clanking and banging” as they struck the armor of French men-at-arms.1 By Tudor times the romance of the longbow was well established. Latimer, bishop of Worcester, preaching in 1549 proclaimed, “The art of shooting has been in times past much esteemed in this realm. It is a gift of God […] whereby he hath given us many victories against our enemies.”2 It was probably conservatism and the prestige of the weapon which account for its continuing use in England to the end of the sixteenth century. In 1596 Sir John Smythe was still drilling companies of archers and pikemen in Essex. He was admittedly an eccentric, and his fondness for the bow owed much to its association with the decisive campaigns of the past which he contrasted with the long and indecisive conflicts of the late Elizabethan period.3 Bows still had limited uses, moreover, notably at long range against cavalry. But by this time England was relatively backward, and continental armies, notably the Spanish Army of the Netherlands, seem almost to have forgotten the weapon. Underlying the nostalgia of Smythe and others, however, was a reality. There is no doubt that in the fourteenth and fifteenth centuries the longbow was highly important on the battlefield, and archers were a vital part of English armies. Modern academic studies have stressed the role of the heavily-armored men-atarms of the English armies. It was the shrewd combination of these arms and 1 2 3

John Keegan, The Face of Battle. A Study of Agincourt, Waterloo and the Somme (Harmondsworth, 1976), p. 93. Quoted in Matthew Strickland and Robert Hardy, The Great Warbow: From Hastings to the Mary Rose (Stroud, 2005), p. 30. John Smythe, Certain Discourses Military, ed. J. R. Hales (Ithaca, 1964), 85–87. On Smythe’s place in English military development, see Mark C. Fissel, English Warfare 1511–1642 (London, 2001).

216

John France

new battle tactics which maximized possibilities of late medieval armies and brought victories.4 As Clifford Rogers explains of dismounted knights in the tight close-order English formations, “What made these horsemen on foot so effective was their integration with the archers. Because the English formations had missile superiority, their dismounted men-at-arms could not be dispersed by enemy archers.”5 Although this is a serious modification of the simple view of the longbow as the queen of the battlefield, it nonetheless remains the case that longbows quite suddenly achieved great military importance. How can we explain such a remarkable phenomenon? The bow, after all, is one of man’s oldest weapons. One of the earliest representations we have of a battle is a relief carved in stone in the tomb of Inty, dating from the later years of the third millennium B.C. This shows Egyptians firing arrows at their Canaanite enemies before charging to close quarters with copper axes.6 Many of the Canaanites, despite being stuck by arrows, actually fight on in the relief, which reveals something of the limitations of the bow as a weapon. In 1191, at the battle of Arsuf, Richard I of England surrounded his cavalry with footsoldiers, spearmen, crossbowmen, and archers to hold the Turkish mounted archers out of range of their vulnerable horses, and one of Saladin’s staff commented, The enemy army was already in formation with the infantry surrounding it like a wall, wearing solid iron corselets and full-length well-made chain-mail, so that arrows were falling on them with no effect […] I saw various individuals amongst the Franks with ten arrows fixed in their backs, pressing on in this fashion quite unconcerned.7

This warns us that the effectiveness of archery was limited by many factors. Range, weather, the quality of the bow, and the nature of the arrow are obvious variables. Above all, of course, are the quality and abilities of the archer himself. It is for this reason that before the development of really effective firearms, battle was decided at close quarters by troops, cavalry or infantry, fighting in close-order with edged weapons. William the Conqueror used bowmen at Hastings in 1066, but only to attrite the Anglo-Saxon battle line; to break it, he turned to his heavy infantry and cavalry. At Bourgethéroulde in 1124, really a large-scale skirmish, the Anglo-Norman forces put their archers forward and they cut down a charge by the Norman rebels. At the Battle of the Standard in 1138 English archers destroyed a charge by

4 5 6 7

Clifford J. Rogers, War Cruel and Sharp: English Strategy under Edward III, 1327–1360 (Woodbridge, 2001). Clifford J. Rogers, “The Military Revolutions of the Hundred Years War,” in The Military Revolution Debate, ed. Clifford J. Rogers (Boulder, CO, 1995), p. 58. W. J. Hamblin, Warfare in the Ancient and Near East to 1600 BC (London, 2006), p. 359; A. Kuhrt, The Ancient Near East, c. 3000–330 BC, 2 vols. (London, 1995). Baha‘ al-Din Ibn Shaddad, The Rare and Excellent History of Saladin, trans. D. S. Richards (Aldershot, 2002), p. 170.



The English Longbow, War, and Administration

217

lightly-armed Scots.8 But in many battles archery played no part at all, or was so limited that it is not mentioned. At Benevento in 1266 Manfred of Sicily seems to have had a very large force of archers and foot, perhaps ten thousand, to five thousand cavalry, but the battle was decided by the knights; while at Tagliacozzo in 1268 neither side seems to have deployed any infantry.9 At Courtrai in 1302 the French had about a thousand crossbowmen against nine hundred on the Flemish side, but they played a negligible role and there were apparently no ordinary archers.10 Very shortly after, Edward II of England mustered very large numbers of archers for his disastrous battle at Bannockburn in 1314.11 This is not an exhaustive survey, but the point I am trying to make is that archers were not invariably present in battle, and often were not very effective. Now, it has to be admitted that our sources for battles were usually aristocratic, and therefore focused on the knights and leaders. Even so, enough is known to make it clear that the effectiveness of archery fluctuated enormously. And yet, in the fourteenth and fifteenth centuries archers proved to be immensely effective against knights wearing highly effective and very sophisticated armor, much of which was specifically designed with oblique surfaces to deflect arrows. Before we can start to discuss why this was, we need to distinguish and describe the weapon under discussion. Other types, the composite bow, a short bow made of a core of wood glued to bone and cartilage or sinew, was little used in northern Europe. It was the weapon of the steppe horse-archers, providing enormous poundage in a short bow which could be wielded on horseback. It was complex to make, however, and the glue came unstuck in damp climates. It had a distinctive D-shape, with curved horns, and it might be thought that this identifies it in contemporary pictures; but artists were influenced by a whole series of factors in their portrayals, and could have copied the type from classical reliefs. Nor here are we discussing the crossbow, which was a mechanical bow of extraordinary power. This was very slow to load, so its chief value was in siege and naval warfare, which is why the French hired Genoese crossbowmen and used them at Courtrai in 1302 and Crécy in 1346. The weapon we are discussing here is rather the simple stave-bow (or self-bow): a length of pliable wood whose ends are “nocked” so that a string can be tied from one to the other, thereby creating a pull or poundage which can be drawn to launch the arrow. 8

9 10

11

Jim Bradbury, “Battles in England and Normandy, 1066–1154,” Anglo-Norman Studies 6 (1983), 1–12 at 8; Matthew Bennett has drawn attention to these earlier battles: “The Development of Battle Tactics in the Hundred Years War,” in Arms, Armies and Fortifications in the Hundred Years War, ed. Anne Curry and Michael Hughes (Woodbridge, 1994), pp. 1–20. John France, Western Warfare in the Age of the Crusades, 1000–1300 (London, 1999), pp. 178–83. J. F. Verbruggen, The Battle of the Golden Spurs (Courtrai, 11 July 1302): A Contribution to the History of Flanders’ War of Liberation 1297–1305, ed. Kelly DeVries, trans. David Richard Ferguson (Woodbridge, 2002), p. 194. G. W. S. Barrow, Robert Bruce and the Community of the Realm of Scotland (Edinburgh, 1965), pp. 323–24.

218

John France

We have learned a great deal about military bows from the wreck of a Tudor ship, the Mary Rose, which sank on 19 July 1545, and was recovered on 11 October 1982. Careful examination of its contents revealed numerous bowstaves. This showed that the bows were much heavier, thicker, and stronger than had previously been assumed. They were carefully cut from the point on the yew-tree where the sapwood and the heartwood met, and required an enormous draw – the pull required to fire them varied from 110 lbs to an incredible 185 lbs. Moreover, the arrows found with them were often much heavier than had been anticipated, and the combination of power and weight gave them a penetrative power which was very dangerous to even the best late-medieval plate armor.12 This bow, then, is a very lethal weapon, and experiments have borne out the impression of its lethality derived from accounts of battles like Crécy and Agincourt. It is fairly natural to ask how the longbow, which had hitherto such a patchy history, suddenly leapt to prominence in the first half of the fourteenth century, eclipsing even the crossbow. The reason most usually cited is essentially technological: that the longbow of the fourteenth century was different in kind from the bows that had been used before. This is a perfectly natural explanation, which comes readily to mind. Moreover, it has long been accepted that the early fourteenth century saw improvements in tactics, so the overall sense has been that better bows came to be used more intelligently and, importantly, in large numbers, with volley-fire creating those sleets of arrows which made “a weird clanking and banging” as they struck the armor worn by their targets. Moreover, illustrations of bows are very common in medieval manuscripts, as are carvings of them in churches, in stone and wood. Thus some historians have taken a technological view, that the longbow simply evolved from earlier bows: the shortbow, a limited weapon, was superseded by the longbow. Charles Oman has been immensely influential in shaping ideas about medieval warfare; he thought that Edward I came across Welsh bowmen with their rough elm bows during his conquest of Wales, and their adapted weapon subsequently became the longbow, whose first victory was at Falkirk in 1298. John E. Morris has since pointed to the plentiful evidence for English bowmen much earlier, however, and argued that Edward I developed in particular the use of the longbow in conjunction with heavily armed infantry, in confrontations such as Maes Moydog in 1294. This tactical development, he suggested, was the real foundation of the greatness of archery. But it is the opinion of both Oman and Morris that in the twelfth century only the inferior and less powerful shortbow had been known.13 They justify this stance by reference to a large number of illustrations which appear to show that the bows of the eleventh and twelfth centuries were rather weak-looking affairs, about four to five feet feet (1.2 to 1.5 m) in length, which 12 13

Strickland and Hardy, Great Warbow, pp. 3–33. Charles Oman, The Art of War in the Middle Ages, 2 vols. (1924, repr. London, 1998), 1:43–44, 59–60; John E. Morris, The Welsh Wars of Edward I (Oxford, 1901, 2nd ed. Stroud, 1996), pp. 26, 32–33.

1. Marginalia drawing of a Welsh archer from the Littere Wallie, c. 1282–92 (Kew, The National Archives, E 36/274, f. 356r). Likely an illustration of the short bow.

220

John France

were apparently drawn back only to the chest, thus suggesting a very limited pull and consequently feeble performance. This view of an evolution from the shortbow has long been accepted by historians, not least because one set of the representations on which it is based, which appears to show small weapons, forms part of the famous Bayeux Tapestry.14 A well-known manuscript picture, dating from 1267, of a Welsh archer wielding a short and very thick, knobbly bow, perhaps suggests that even such a short weapon was capable of highvelocity fire; but it is possible that the illustration is a caricature.15 The working hypothesis which emerges from such studies is that in the course of the thirteenth century the bow evolved, perhaps most critically by the selection of wood from good yew cut where heart and sapwood meet, and that leaders, especially Edward I, recognized that this new weapon could be used in conjunction with other arms. At Maes Moydog, the earl of Warwick with two hundred cavalry trapped the rebel Moydog in the open: the Welsh formed tight circles bristling with spears, but these were broken up by Warwick’s archers, to the point where the cavalry could charge them down.16 Clearly the tactics used at Maes Moydog were not precisely those employed at Crécy, but represented a step in that direction. This comfortable hypothesis, of a short bow evolving into something stronger and more formidable, was systematically attacked in Matthew Strickland and Robert Hardy’s authoritative study Great Warbow, following the view expressed by J. C. Holt that “the shortbow, as a category of weapon was invented by the military historian, Sir Charles Oman in the nineteenth century.”17 They pointed out that there is evidence of longbows six feet or more in length with a thickness comparable to those of the Mary Rose long before the Norman conquest of England. Further, they follow Holt in his skepticism as to the value of the many pictures which show apparently short bows or bows drawn only to the chest or chin rather than the ear, arguing that those who accept them as evidence do so from “a too literal and too unwary interpretation of the iconography of the bow in medieval art […] what changed from the thirteenth century was not the bow itself but rather the manner in which many artists depicted bows.”18 From this they argue that there never was a short bow, only distortions of the longbow. There is much to be said for such caution about illustrations. The bowmen in the Bayeux Tapestry are set in the lower margin and their figures are bent to fit in. The whole picture is distorted. There is one archer – an Anglo-Saxon – shown in the main part of the embroidery, but he is depicted as a small figure in contrast to his comrades, which may signify a desire to show how unimportant the bowmen were in Harold’s army.19 In fact pictures of bowmen show an astonishingly diverse range of weapons of various 14 15 16 17 18 19

The Bayeux Tapestry, ed. D. M. Wilson, (London, 1985), pp. 68–70. A marginal drawing from the Treaty of Montgomery, also shown in Strickland and Hardy, Great Warbow, p. 35. France, Western Warfare, p. 173. J. C. Holt, Robin Hood, rev. ed. (London, 1989), pp. 79–80. Strickland and Hardy, Great Warbow, p. 38. Bayeux Tapestry, ed. Wilson, p. 62.



The English Longbow, War, and Administration

221

lengths, some pulled back to the cheek, others drawn in quite different ways. At the same time, I rather agree with Clifford Rogers that small bows are pictured so often that they cannot all simply be distortions. In addition, children and women used smaller bows, as Holt recognized. The weakness or strength of a bow is not, moreover, related purely to its length: Strickland and Hardy cite Gaston Phébus, count of Foix, who in the fourteenth century recommended a weak bow for hunting because it was easier to bend when concealment from the prey was important.20 More importantly, perhaps, we should recognize that the bow was used by a very wide range of people for a number of purposes, and that the expense and trouble of creating weapons like those found in the Mary Rose would have been beyond the means or the needs of most of them. The earliest Assize of Arms issued by Henry II in 1181 for England did not mention bows, and this is sometimes taken as meaning that the weapon was despised; but then again, neither does it mention swords (though that for the continental lands does).21 This Assize was less concerned with defense against external enemies than with internal security, and the king wanted to ensure that substantial citizens were equipped to defend the peace. By contrast, Clause 2 of the Assize of the Forest issued in 1184 prohibits anyone from entering the forest with bows and arrows except with royal permission. The Ordinance for the Preservation of the Peace of 1242 required the poorest to equip themselves with bows, while the Assize of Arms of 1252 required all “citizens, burgesses, free tenants, villeins and others from 15 to 60 years of age” to be armed, and the poorest of them were expected to have at least a bow. 22 The bow was clearly the weapon of the poor. It was made from easily available materials and, unlike other arms, required very little expensive metal. But it seems unlikely that many of these bows were the monsters, of over 100 lbs pull, found in the Mary Rose. These would simply not have been needed to shoot pigeons or other small game, and would have been unwieldy if used for such purposes. Moreover, relatively few men would have had the musculature required. When Richard II of England wanted to recruit an elite group of archers as a personal bodyguard a contemporary recorded that he “sent a certain squire through Essex, Cambridge, Norfolk and Suffolk who […] made any of the stronger and more powerful men of those counties swear to him that, putting aside all other lords, they would hold with the king.”23 Anyone who has ever tried to draw a longbow will understand the instructions given to the squire. Men of sufficient strength must have been relatively rare.

20 21 22

23

Strickland and Hardy, Great Warbow, p. 38; Holt, Robin Hood, p. 80. Roger of Howden, Chronica magistri Rogeri de Houedene, ed. William Stubbs, 4 vols. (London, 1868–71), 2:253, 260–63. Sources of English Constitutional History: A Selection of Documents from AD 600 to the Present, ed. Carl Stephenson and Frederick George Marcham (New York, 1937), pp. 85–87, 139; France, Western Warfare, pp. 32–33. Monk of Westminster, quoted in J. L. Gillespie, “Richard II’s Archers of the Crown,” Journal of British Studies 18 (1979), 14–29 at 21.

222

John France

Edward I needed infantry for his attacks on Wales and Scotland, and for this purpose issued “Commissions of Array,” ordering his household knights to recruit soldiers from the levy of the shires. In practice, local royal officers negotiated with communities to provide young men. The Commissions of Array persisted as a means of recruiting soldiers, and clearly did raise some, but the system was not a notable success. Large numbers of archers were recruited, but their equipment was often limited, and there was mass desertion.24 Moreover, the bodies of archers found in the wreck of the Mary Rose showed skeletal deformations which could be associated with pulling the heavy longbows, and this points to self-selection through a lifetime of training. It is unlikely that many of the infantry masses recruited under Edward I were made of such mettle. On the other hand, some may have been. In the light of this the controversy over long and short bows seems a little irrelevant. Given the widespread use of the bow and its diffusion amongst the population, the likelihood is, as our visual evidence suggests, that there were many sizes and shapes, and many qualities of bow. Even in 1542 a statute distinguished between “bows of the best sort” and those “of the second sort.”25 Though elm was sometimes used, the finest bows seem always to have been made of yew, which is slow-growing and produces a very dense wood. But it does not grow everywhere in the British Isles, and is often stunted in its habit.26 This is perhaps why it was imported from Spain as early as the fourteenth century. A murder case tried in 1315 refers to a bow of Spanish yew.27 Bows were, then, diverse, and while it seems entirely possible that many were of high quality and great strength, it is simply not possible to dismiss the notion that there were other bows of varying strengths and lengths. The question is, therefore: How did effective and skilled archery arise from the circumstances of the end of the thirteenth century? This is clearly not merely an issue of technology, for I am sure that Strickland and Hardy are right, that powerful longbows were known before 1300. Their caution over illustrations is quite understandable. A fifteenth-century (c.  1450) mural in Pickering church shows men wielding what are meant to be powerful bows, but drawing them only to the chest, an extent of draw associated with weaker weapons. The length of these weapons is quite difficult to guess, but they do clearly show the distinction between heartwood and sapwood. Another illustration of c. 1483 shows the same distinction in the wood, and clearly the bows are very long.28 But these pictures date from a time when a degree of perspective and “realism” was becoming generally evident in illustration. I suspect that the appearance of really effective archery was at least partly a matter of military administration. England in the thirteenth century 24 25 26 27 28

The limitations of the Commissions of Array are discussed by Michael Prestwich, Armies and Warfare in the Middle Ages: The English Experience (New Haven, 1996), pp. 123–37. Strickland and Hardy, Great Warbow, pp. 30–31, 38. John Lowe, Yew Trees of Great Britain and Ireland (London, 1897), p. 24. Prestwich, Armies and Warfare, p. 131. Strickland and Hardy, Great Warbow, pp. 382, 52.

2. Archer depicted in a thirteenth-century English bestiary, c. 1225–50 (London, The British Library, Harley MS 4751, f. 14v). Note the different shades of wood picked out in the bow.

224

John France

was a comparatively peaceful country until Edward I’s imperial ambitions demanded the creation of armies. While the upper class rallied, or could be made to rally by distraint of knighthood, the need for large infantry forces was novel.29 In these broken and often wooded lands footsoldiers were essential, and they were vital in creating and attacking fortifications. Amongst the poor who were now recruited, archers had always been numerous; this must have led commanders to a recognition of their skills in such wars. Commissions of Array, which were the Crown’s chosen method of recruiting, were often inefficient and open to abuse, but they tapped reserves of manpower not previously touched. This was important: all ancient and medieval armies depended upon native skills in the population, for they lacked the means and infrastructure to train soldiers. The old general obligation to defend the realm did nothing to discover the best men, but as applied in an increasingly selective way by Edward I, it began to do so. Edward recruited fewer infantry, but tried to ensure that they were of high quality. Warfare in the rough lands of the British periphery demonstrated the need for good infantry, and the utility of archery. And the sheer numbers recruited demonstrated that volley fire provided a means of overcoming the limitations of individuals in a body where there were bound to be inconsistencies. Compulsion, applied via the system of Commissions of Array, seems always to have played its role in filling the ranks of English armies; but there were other processes at work too. It has been noted that some (but not all) of the archers who served in the Hundred Years War were people of substance, yeomen and petty gentry.30 At the end of the thirteenth century and the start of the fourteenth English agriculture entered a period of difficulty. This coincided with the end of the wars in Wales, and may account for the prominence of Cheshire archers: the successes of the war in France and the accompanying loot made such service attractive. Ultimately, however, the Crown fell back upon an older system – recruiting the followers of the great – but via a new instrument, the indenture, which suited the emerging royal bureaucracy. In the twelfth and thirteenth centuries magnates had recruited elements of their retinues for royal service, from men endowed on their lands, but for the most part they hired paid men, professionals of war: hence the growth in references to “men-at-arms.”31 In one way or another, the Crown paid these men indirectly. In the circumstances of war on the British periphery at the end of the thirteenth century, such men continued to be essential; but infantry were also needed in large numbers, and had to be drawn from all over the realm, imposing a heavy burden on local communities 29 30 31

Peter Coss, The Knight in Medieval England 1000–1400 (Stroud, 1993), pp. 60–71. See especially Gillespie, “Richard II’s Archers,” pp. 14–29. Undoubtedly they were an elite group, but the very fact that such substantial men took up archery is important. Nicholas Vincent, Peter des Roches: An Alien in English Politics, 1205–38 (Cambridge, 1996), p. 63; David Crouch, William Marshal. Court, Career and Chivalry in the Angevin Empire (London, 1990), pp. 195–205.



The English Longbow, War, and Administration

225

who paid for their equipment and maintenance and lost their labor.32 By the end of the thirteenth century English wealth was growing and the Crown had developed a much more efficient taxation structure, hence it enjoyed a new capacity to pay. It is this process which underlay the emergence of the contractual indenture system. Local leaders knew their areas and could seek out willing and able soldiers, amongst whom archers were very important.33 Ultimately, even local lords like Sir Hugh Hastings in 1380 turned to men-at-arms to find archers for him, and also to military subcontractors like Nowell.34 Recruitment in this way established a relationship between the armored men, whose close-order formations were so essential to the new tactics, and the archers and others, laying the foundation for close cooperation. If many of our examples come from a period when the indenture was firmly established, there is little reason to doubt its evolution from the earlier system. The indenture was the product of a world where liquid wealth and trade were making contracts more necessary, but the actual process of recruitment which it reveals must have been very ancient. The whole process must have been given impetus by the Scottish war, with its disaster in 1314 at Bannockburn and its triumph at Halidon Hill in 1333, which owed much to the victory the year before at Dupplin Moor. At Halidon Hill the Scots attacked a strongly positioned English army: They (the Scots) marched towards the town with great display, in order of battle, and recklessly, stupidly and inadvisedly chose a battle ground at Halidon Hill, where there was a marshy hollow between the two armies, and where a great downward slope, with some precipices, and then again a rise lay in front of the Scots, before they could reach the field where the English were posted.35

Their advance was then weakened by archery: the Scots who marched in the front were so wounded in the face and blinded by the multitude of English arrows that they could not help themselves, and soon began to turn their faces away from the blows of the arrows and fall.36

32 33

34

35 36

Morris, Welsh Wars, pp. 35–110; Prestwich, Armies and Warfare, pp. 115–46. Adrian R. Bell, War and the Soldier in the Fourteenth Century (Woodbridge, 2004), p. 116; N. Jamieson, “The Recruitment of Northerners for Service in English Armies in France, 1415–50,” in Trade, Devotion and Government in Later Medieval History, ed. D. L. Clayton, R. G. Davies, and P. McNiven (Stroud, 1994), pp. 102–15. A. E. Prince, “The Indenture System under Edward III,” in Historical Essays in Honour of James Tait, ed. J. G. Edwards, V. H. Galbraith and E. F. Jacob (Manchester, 1933), pp. 283–97; Anthony Goodman, “The Military Subcontracts of Sir Hugh Hastings,” English Historical Review 95 (1980), 114–20 at 116. Book of Pluscarden, ed. Felix J. H. Skene, 2 vols. (Edinburgh, 1877–80), 1:202. Herbert Maxwell, trans., The Chronicle of Lanercost, 1272–1346, trans. Herbert Maxwell, new ed. with introduction by James Wilson, 2 vols. (Glasgow, 1913, repr. Cribyn, 2001), 1:279.

226

John France

As a result when they finally clashed with the dismounted English men-at-arms they were thrown back in confusion and their retreat turned into a rout as the victors mounted their horses and massacred their retreating enemy. This is the system foreshadowed by Maes Moydog, and, it could be said, is rooted in the tactics of earlier ages. Applying the system and understanding the tactical context depended on competent leadership, which, of course, arises largely through chance. Military talent contributes to a commander’s rise, but rank remained important in all armies, and loyalty was (and is) always a serious consideration; thus the ability to negotiate the treacherous currents of court life was essential, though a skill little related to military capacity. It so happened that Edward III was a very able commander, and I think Clifford Rogers has rightly distinguished his importance in the development of the tactics which were to become known in the First World War as “bite and hold.” But the roots of those tactics lay in long experience which interacted with administrative development and social change, largely as a result of the demands of conquest in the British periphery. We have given so far a great deal of attention to English developments, because English sources are so strong. However, we perhaps ought to consider too why the French proved so vulnerable to these tactics, and what factors gave French military development its particular direction. Chance factors of course, like leadership, had a hand in this, but what the main influences were upon French development in this period is something that has been little explored, even in Contamine’s history of the French army.37 And how do other regions compare? Austria and Hungary spring to mind as possessing armies about whose formation we do not seem to be well informed, for example. Whatever the outcome of enquiry into such topics, however, I am quite certain that the popular emphasis on technology as the agency of military change is mistaken, and that we need to look beyond this, to the broad context of war and the development of the state.

37

Anne Blanchard, Philippe Contamine, André Corvisier, Jean Meyer, and Michel Mollat du Jourdain, Histoire militaire de la France. 1: Des origines à 1715, ed. Philippe Contamine (Paris, 1992).

Contributors

Bernard S. Bachrach is currently Professor of History at the University of Minnesota, Minneapolis, and has an extensive publication record on military and political history from Late Antiquity to the High Middle Ages, including Fulk Nerra, the Neo-Roman Consul 987-1040: A Political Biography of the Angevin Count (Berkely, 1993) and Early Carolingian Warfare: Prelude to Empire (Philadelphia, 2001). His latest book, Medieval Warfare (2017) was co-authored with David S. Bachrach and his latest article “Is the Song of Roland’s Roncevalles a Military Satire?” also appeared in 2017. Leif Inge Ree Petersen is Associate Professor of Late Antique and Early Medieval History at the Norwegian University of Science and Technology, Trondheim, and his major publication to date is Siege Warfare and Military Organization in the Successor States, 400-800 AD: Byzantium, the West and Islam (Brill, 2013). Richard Abels is Professor Emeritus of History at the United States Naval Academy. He is the author of two books, Alfred the Great, War, Kingship and Culture in Anglo-Saxon England (1998; revised edition forthcoming), and Lordship and Military Obligation in Anglo-Saxon England (1988), and a number of articles on warfare, politics and culture in early medieval England. Dr Matthew Bennett taught at The Royal Military Academy Sandhurst (19842014) and is currently a Visiting Lecturer at Winchester University. The focus of his research is the ethos and practice of warfare in the High Middle Ages, especially chivalry, largely through the medium of Old French literature. Medieval Hostageship: hostage, captive, prisoner of war, guarantee, peacemaker, jointly edited with Dr Katherine Weikert, University of Winchester (Routledge Research in Medieval Studies, vol. 9, 2017) is his most recent publication. Luis García-Guijarro is Reader in Medieval History at the University of Zaragoza. He has published books and articles dealing with the Reconquista, the crusades, the military orders and the Latin Church in the Central Middle Ages, such as Papado, cruzadas y órdenes militares, siglos XI-XIII (1995). An additional line of his research is centered on the birth and development of political units in Eastern Iberia between the ninth and eleventh centuries.

228 Contributors

Manuel Rojas Gabriel is Professor of Medieval History at the University of Extremadura. He has published extensively on medieval warfare, frontiers and castellology, and he has directed several national and international research projects on these subjects. He is currently engaged in a general study with the title La Guerra en el Occidente Medieval (c. 450-c- 1500). María Dolores García Oliva is Profesora Titular of Medieval History at the University of Extremadura. She has published widely on several aspects of medieval Iberian history, mainly rural enviroment and town councils. An additional line of her research is centred on the political and military development of the kingdom of León in the twelfth and thirtheen centuries. John Gillingham is a Fellow of the British Academy and Emeritus Professor of History at the LSE, where he taught from 1965 to 1998. His books include Richard I (Yale UP, 1999), The English in the Twelfth Century (Boydell, 2000), The Angevin Empire, 2nd ed. (Hodder, 2001), Conquests, Catastrophe and Recovery. Britain and Ireland c. 1066-c.1485 (Random House, 2014) and William II. The Red King (Penguin, 2015). John France is Professor Emeritus at Swansea University, specializing in the study of crusading and medieval warfare represented by works such as Victory in the East: A Military History of The First Crusade (Cambridge, 1994) and Western Warfare in the Age of the Crusades 1000-1300 (London, 1999). His latest book is Hattin (Oxford, 2015).

Journal of Medieval Military History 1477–545X

Volume I 1 The Vegetian “Science of Warfare” in the Middle Ages – Clifford J. Rogers 2 Battle Seeking: The Contexts and Limits of Vegetian Strategy – Stephen F. Morillo 3 Italia – Bavaria – Avaria: The Grand Strategy behind Charlemagne’s Renovatio Imperii in the West – Charles R. Bowlus 4 The Composition and Raising of the Armies of Charlemagne – John France 5 Some Observations on the Role of the Byzantine Navy in the Success of the First Crusade – Bernard S. Bachrach 6 Besieging Bedford: Military Logistics in 1224 – Emilie M. Amt 7 “To aid the Custodian and Council”: Edmund of Langley and the Defense of the Realm, June–July 1399 – Douglas Biggs 8 Flemish Urban Militias Against the French Cavalry Armies in the Fourteenth and Fifteenth Centuries – J. F. Verbruggen (translated by Kelly DeVries)

Volume II 1 The Use of Chronicles in Recreating Medieval Military History – Kelly DeVries 2 Military Service in the County of Flanders – J. F. Verbruggen (translated by Kelly DeVries) 3 Prince into Mercenary: Count Armengol VI of Urgel, 1102–1154 – Bernard F. Reilly 4 Henry II’s Military Campaigns in Wales, 1157–1165 – John Hosler 5 Origins of the Crossbow Industry in England – David S. Bachrach 6 The Bergerac Campaign (1345) and the Generalship of Henry of Lancaster – Clifford J. Rogers 7 A Shattered Circle: Eastern Spanish Fortifications and Their Repair during the “Calamitous Fourteenth Century” – Donald Kagay 8 The Militia of Malta – Theresa M. Vann 9 “Up with Orthodoxy”: In Defense of Vegetian Warfare – John B. Gillingham 10 100,000 Crossbow Bolts for the Crusading King of Aragon – Robert Burns

Volume III 1 A Lying Legacy? A Preliminary Discussion of Images of Antiquity and Altered Reality in Medieval Military History – Richard Abels and Stephen F. Morillo 2 War and Sanctity: Saints’ Lives as Sources for Early Medieval Warfare – John France

The 791 Equine Epidemic and its Impact on Charlemagne’s Army – Carroll Gillmor The Role of the Cavalry in Medieval Warfare – J. F. Verbruggen Sichelgaita of Salerno: Amazon or Trophy Wife? – Valerie Eads Castilian Military Reform under the Reign of Alfonso XI (1312–1350) – Nicolas Agrait 7 Sir Thomas Dagworth in Brittany, 1346–1347: Restellou and La Roche Derrien – Clifford J. Rogers 8 Ferrante d’Este’s Letters as a Source for Military History – Sergio Mantovani 9 Provisions for the Ostend Militia on the Defense, August 1436 – Kelly DeVries 3 4 5 6

Volume IV 1 The Sword of Justice: War and State Formation in Comparative Perspective – Stephen F. Morillo 2 Archery versus Mail: Experimental Archaeology and the Value of Historical Context – Russ Mitchell 3 “Cowardice” and Duty in Anglo-Saxon England – Richard Abels 4 Cowardice and Fear Management: The 1173–1174 Conflict as a Case Study – Steven Isaac 5 Expecting Cowardice: Medieval Battle Tactics Reconsidered – Stephen F. Morillo 6 Naval Tactics at the Battle of Zierikzee (1304) in the Light of Mediterranean Praxis – William Sayers 7 The Military Role of the Magistrates in Holland during the Guelders War – James P. Ward 8 Women in Medieval Armies – J. F. Verbruggen 9 Verbruggen’s “Cavalry” and the Lyon-Thesis – Bernard S. Bachrach 10 Dogs of War in Thirteenth-Century Valencian Garrisons – Robert I. Burns

Volume V 1 2 3 4 5 6 7 8

Literature as Essential Evidence for Understanding Chivalry – Richard W. Kaeuper The Battle of Hattin: A Chronicle of a Defeat Foretold? – Michael Ehrlich Hybrid or Counterpoise? A Study of Transitional Trebuchets – Michael Basista The Struggle between the Nicaean Empire and the Bulgarian State (1254–1256): towards a Revival of Byzantine Military Tactics under Theodore II Laskaris – Nicholas S. Kanellopoulos and Joanne K. Lekea A “Clock-and-Bow” Story: Late Medieval Technology from Monastic Evidence – Mark Dupuy The Strength of Lancastrian Loyalism during the Readeption: Gentry Participation at the Battle of Tewkesbury – Malcolm Mercer Soldiers and Gentlemen: The Rise of the Duel in Renaissance Italy – Stephen Hughes “A Lying Legacy” Revisited: The Abels-Morillo Defense of Discontinuity – Bernard S. Bachrach

Volume VI 1 Cultural Representation and the Practice of War in the Middle Ages – Richard Abels 2 The Brevium Exempla as a Source for Carolingian Warhorses – Carroll Gillmor 3 Infantry and Cavalry in Lombardy (11th–12th Centuries) – Aldo Settia 4 Unintended Consumption: The Interruption of the Fourth Crusade at Venice and its Consequences – Greg Bell 5 Light Cavalry, Heavy Cavalry, Horse Archers, Oh My! What Abstract Definitions Don’t Tell Us About 1205 Adrianople – Russ Mitchell 6 War, Financing in the Late-Medieval Crown of Aragon – Donald Kagay 7 National Reconciliation in France at the end of the Hundred Years War – Christopher Allmand

Volume VII 1 The Military Role of the Order of the Garter – Richard W. Barber 2 The Itineraries of the Black Prince’s Chevauchées of 1355 and 1356: Observations and Interpretations – Peter Hoskins 3 The Chevauchée of John Chandos and Robert Knolles: Early March to Early June, 1369 – Nicolas Savy 4 “A Voyage, or Rather an Expedition, to Portugal”: Edmund of Langley’s Journey to Iberia, June/July 1381 – Douglas Biggs 5 The Battle of Aljubarrota (1385): A Reassessment – João Gouveia Monteiro 6 “Military” Knighthood in the Lancastrian Era: The Case of Sir John Montgomery – Gilbert Bogner 7 Medieval Romances and Military History: Marching Orders in Jean de Bueil’s Le Jouvencel introduit aux armes – Matthieu Chan Tsin 8 Arms and the Art of War: The Ghentenaar and Brugeois Militia in 1477–1479 – J. F. Verbruggen 9 Accounting for Service at War: the Case of Sir James Audley of Heighley – Nicholas Gribit 10 The Black Prince in Gascony and France (1355–1356), According to MS 78 of Corpus Christi College, Oxford – Clifford J. Rogers

Volume VIII 1 People against Mercenaries: The Capuchins in Southern Gaul – John France 2 The Last Italian Expedition of Henry IV: Re-reading the Vita Mathildis of Donizone of Canossa – Valerie Eads 3 Jaime I of Aragon: Child and Master of the Spanish Reconquest – Donald Kagay 4 Numbers in Mongol Warfare – Carl Sverdrup 5 Battlefield Medicine in Wolfram’s Parzival – Jolyon T. Hughes 6 Battle-Seeking, Battle-Avoiding or Perhaps Just Battle-Willing? Applying the Gillingham Paradigm to Enrique II of Castile – Andrew Villalon 7 Outrance and Plaisance – Will McLean

8 Guns and Goddams: Was there a Military Revolution in Lancastrian Normandy 1415–1450? – Anne Curry 9 The Name of the Siege Engine Trebuchet: Etymology and History in Medieval France and Britain – William Sayers

Volume IX 1 The French Offensives of 1404–1407 against Anglo-Gascon Aquitaine – Guilhem Pépin 2 The King’s Welshmen: Welsh Involvement in the Expeditionary Army of 1415 – Adam Chapman 3 Gunners, Aides and Archers: The Personnel of the English Ordnance Companies in Normandy in the Fifteenth Century – Andy King 4 Defense, Honor, and Community: the Military and Social Bonds of the Dukes of Burgundy and the Flemish Shooting Guilds – Laura Crombie 5 The Battle of Edgecote or Banbury (1469) through the Eyes of Contemporary Welsh Poets – Barry Lewis 6 Descriptions of Battles in Fifteenth-Century Urban Chronicles: A Comparison of the Siege of London in May 1471 and the Battle of Grandson, 2 March 1476 – Andreas Remy 7 Urban Espionage and Counterespionage during the Burgundian Wars (1468–1477) – Bastian Walter 8 Urban Militias, Nobles, and Mercenaries: The Organization of the Antwerp Army in the Flemish-Brabantine Revolt of the 1480s – Frederik Buylaert, Jan Van Camp and Bert Verwerft 9 Military Equipment in the Town of Southampton during the Fourteenth and Fifteenth Centuries – Randall Moffett

Volume X 1 The Careers of Justinian’s Generals – David Parnell 2 Early Saxon Frontier Warfare: Henry I, Otto I, and Carolingian Military Institutions – David S. Bachrach and Bernard S. Bachrach 3 War in The Lay of the Cid – Francisco Garcia Fitz 4 The Battle of Salado (1340) Revisited – Nicolas Agrait 5 Chivalry and Military Biography in the Later Middle Ages: The Chronicle of the Good Duke Louis of Bourbon – Steven Muhlberger 6 The Ottoman–Hungarian Campaigns of 1442 – John Jefferson 7 Security and Insecurity, Spies, and Informers in Holland during the Guelders War (1506–1515) – James P. Ward 8 Edward I’s Wars in the Chronicle of Hagnaby Priory – Michael Prestwich

Volume XI 1 Military Games and the Training of the Infantry – Aldo A. Settia 2 The Battle of Civitate: A Plausible Account – Charles D. Stanton

3 The Square “Fighting March” of the Crusaders at the Battle of Ascalon (1099) – Georgios Theotokis 4 How the Crusades Could Have Been Won: King Baldwin II of Jerusalem’s Campaigns against Aleppo (1124–1125) and Damascus (1129) – T. S. Asbridge 5 Saint Catherine’s Day Miracle: The Battle of Montgisard – Michael Ehrlich 6 The Military Effectiveness of Alan Mercenaries in Byzantium, 1301–1306 – Scott Jesse and Anatoly Isaenko 7 Winning and Recalling Honor in Spain: Pro-English Poetry in Celebration of the Battle of Nájera (1367) – Donald J. Kagay and L. J. Andrew Villalon 8 The Wars and the Army of the Duke of Cephalonia Carlo I Tocco (c. 1375–1429) – Savvas Kyriakidis 9 Sir John Radcliffe, K. G. (d. 1441): Miles famossissimus – A. Compton Reeves 10 Defense Schemes of Southampton in the Late Medieval Period, 1300–1500 – Randall Moffett 11 French and English Acceptance of Medieval Gunpowder Weaponry – Kelly DeVries

Volume XII 1 Some Observations Regarding Barbarian Military Demography: Geiseric’s Census of 429 and Its Implications – Bernard S. Bachrach 2 War Words and Battle Spears: The kesja and kesjulag in Old Norse Literature – K. James McMullen 3 The Political and Military Agency of Ecclesiastical Leaders in Anglo-Norman England: 1066–1154 – Craig M. Nakashian 4 Couched Lance and Mounted Shock Combat in the East: The Georgian Experience – Mamuka Tsurtsumia 5 The Battle of Arsur: A Short-Lived Victory – Michael Ehrlich 6 Prelude to Kephissos (1311): An Analysis of the Battle of Apros (1305) – Nikolaos S. Kanellopoulos and Ioanna K. Lekea 7 Horse Restoration (Restaurum equorum) in the Army of Henry of Grosmont, 1345: A Benefit of Military Service in the Hundred Years’ War – Nicholas A. Gribit 8 The Indenture between Edward III and the Black Prince for the Prince’s Expedition to Gascony, 10 July 1355 – Mollie M. Madden 9 Investigating the Socio-Economic Origins of English Archers in the Second Half of the Fourteenth Century – Gary Baker 10 War and the Great Schism: Military Factors Determining Allegiances in Iberia – L. J. Andrew Villalon

Volume XIII 1 Feudalism, Romanticism, and Source Criticism: Writing the Military History of Salian Germany – David S. Bachrach 2 When the Lamb Attacked the Lion: A Danish Attack on England in 1138? – Thomas K. Heebøll-Holm 3 Development of Prefabricated Artillery during the Crusades – Michael S. Fulton 4 Some Notes on Ayyubid and Mamluk Military Terms – Rabei G. Khamisy

5 Helgastaðir, 1220: A Battle of No Significance? – Oren Falk 6 Por la guarda de la mar: Castile and the Struggle for the Sea in the Thirteenth and Fourteenth Centuries – Nicolás Agrait 7 The Battle of Hyddgen, 1401: Owain Glyndwr’s Victory Reconsidered – Michael Livingston 8 The Provision of Artillery for the 1428 Expedition to France – Dan Spencer 9 1471: The Year of Three Battles and English Gunpowder Artillery – Devin Fields 10 “Cardinal Sins” and “Cardinal Virtues” of “El Tercer Rey,” Pedro González de Mendoza: The Many Faces of a Warrior Churchman in Late Medieval Europe – L. J. Andrew Villalon 11 Late Medieval Divergences: Comparative Perspectives on Early Gunpowder Warfare in Europe and China – Tonio Andrade

Volume XIV 1 Anglo-Norman Artillery in Narrative Histories, from the Reign of William I to the Minority of Henry III – Michael S. Fulton 2 Imperial Policy and Military Practice in the Plantagenet Dominions, c. 1337–c. 1453 – David Green 3 The Parliament of the Crown of Aragon as Military Financier in the War of the Two Pedros – Donald J. Kagay 4 Chasing the Chimera in Spain: Edmund of Langley in Iberia, 1381/82 – Douglas Biggs 5 Note: A Medieval City under Threat Turns Its Coat, While Hedging Its Bets – Burgos Faces an Invasion in Spring, 1366 – L. J. Andrew Villalon 6 Medieval European Mercenaries in North Africa: The Value of Difference – Michael Lower 7 Medieval Irregular Warfare, c. 1000–1300 – John France 8 Muslim Responses to Western Intervention: A Comparative Study of the Crusades and Post-2003 Iraq – Alex Mallett 9 “New Wars” and Medieval Warfare: Some Terminological Considerations – Jochen G. Schenk 10 Friend or Foe? The Catalan Company as Proxy Actors in the Aegean and Asia Minor Vacuum – Mike Carr

De Re Militari and the Journal of Medieval Military History De re militari: Society for the Study of Medieval Military History (DRM) was founded in 1992 during the annual International Congress for Medieval Studies meeting at Kalamazoo, Michigan. Our organization takes its name from the military treatise by Flavius Renatus Vegetius, the famous Roman work. From its beginning, the Society has had its most significant presence at the Congress, where each year it holds its annual meetings and sponsors a number of sessions, as well as a special lecture delivered by a distinguished medieval military historian. In addition, the society has sponsored sessions at other conferences, including the International Medieval Congress in Leeds, UK, and the annual meeting of the Society for Military History (SMH). DRM maintains a website which has become a principal online source for scholarly information about warfare in the Middle Ages, easily available for use by scholars and students in the United States and abroad. On this website, the organization hosts many primary sources, articles, and dissertations, as well as an extensive book review section. The Society’s Journal of Medieval Military History (JMMH), founded in 2003, is published annually by Boydell & Brewer. It is very much in the interests of both the field of medieval military history, and scholars within that field, to support DRM and the JMMH. Consequently, we encourage scholars to take out membership. At the time of writing, the annual dues for DRM membership are $35 (if paid by check or money order); $38 (if paid by credit card). Payment should be made to the treasurer (or his/ her delegate), if at all possible by check or money order: please see the website for further details, www.deremilitari.org. A DRM membership includes: one hardcover copy of the annual JMMH the right to serve as an officer of DRM the right to vote in all elections of officers and on all issues relevant to the society preference in participating on panels sponsored by DRM the right to compete for the Gillingham Prize annually awarded to the best article in the JMMH the right to buy earlier copies of the JMMH at a reduced price. Those renewing their membership or joining for the first time can obtain an application form via the De re militari site at the following URL: www://deremilitari.org/society/drm-membership-form/ L. J. Andrew Villalon President, De Re Militari (2014–2017)

CONTENTS BERNARD S. BACHRACH

Later Roman Grand Strategy: The Fortifcation of the Urbes of Gaul

LEIF INGE REE PETERSEN

In Search of Equilibrium: Byzantium and the Northern Barbarians, 400–800

RICHARD ABELS MATTHEW BENNETT LUIS GARCÍA-GUIJARRO RAMOS MANUEL ROJAS GABRIEL

Evolving English Strategies during the Viking Wars Norman Conquests: A Strategy for World Domination? The Papacy and the Political Consolidation of the Catalan Counties, c.1060–1100: A Case Study in Political Strategy Alfonso VII of León-Castile in Face of the Reformulation of Power in al-Andalus, 1145–1157: An Essay in Strategic Logic

MARIA DOLORES GARCÍA OLIVA

The Treaties between the Kings of León and the Almohads within the Leonese Expansion Strategy, 1157–1230

JOHN GILLINGHAM

A Strategy of Total War? Henry of Livonia and the Conquest of Estonia, 1208–1227

JOHN FRANCE

The English Longbow, War, and Administration

Cover: Coll. Bibliothèques municipales de Chambéry, MS13, fo. 105r (circa 1170–1190). The Italian artist who illuminated this copy of Gratian’s Decretum beautifully captured the deadly seriousness of medieval warfare – for townsmen as well as knights. Image used by generous permission of the Bibliothèques municipales de Chambéry.

JOURNAL OF MEDIEVAL MILITARY HISTORY

The Journal’s hallmark is broad chronological, geographic and thematic coverage of the subject, bearing witness to ‘the range and richness of scholarship on medieval warfare, military institutions, and cultures of conflict that characterises the field’.  HISTORY

XV

J ournal of

MEDIEVAL MILITARY HISTORY

PETERSEN, ROJAS (editors)

·   XV · Edited by Leif Inge Ree Petersen and Manuel Rojas Gabriel