The Journal of Medieval Military History. Volume VI [6] 1843834081, 9781843834083

Latest volume in the leading forum for debate on aspects of medieval warfare. This sixth volume continues the journal�

206 103 1MB

English Pages 170 [174] Year 2008

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
1. Cultural Representation and the Practice of War in the Middle Ages / Richard Abels 1
2. The Brevium Exempla as a Source for Carolingian Warhorses / Carroll Gillmor 32
3. Infantry and Cavalry in Lombardy (11th–12th Centuries) / Aldo A. Settia, translated by Valerie Eads 58
4. Unintended Consumption: The Interruption of the Fourth Crusade at Venice and Its Consequences / Gregory D. Bell 79
5. Light Cavalry, Heavy Cavalry, Horse Archers, Oh My! What Abstract Definitions Don’t Tell Us About 1205 Adrianople / Russell Mitchell 95
6. War Financing in the Late-Medieval Crown of Aragon / Donald J. Kagay 119
7. National Reconciliation in France at the end of the Hundred Years War / Christopher Allmand 149
Recommend Papers

The Journal of Medieval Military History. Volume VI [6]
 1843834081, 9781843834083

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

THE JOURNAL OF

Medieval Military History Volume VI

THE 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 Editions of previously unpublished documents of significance to the field are included.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.

THE JOURNAL OF

Medieval Military History Volume VI Edited by CLIFFORD J. ROGERS KELLY D EVRIES JOHN FRANCE

THE BOYDELL PRESS

© Contributors 2008 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 2008 The Boydell Press, Woodbridge

ISBN 978–1–84383–408–3

The publisher has no responsibility for the persistence 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

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, USA website: www.boydellandbrewer.com

A CIP catalogue record for this book is available from the British Library

This publication is printed on acid-free paper Typeset by Frances Hackeson Freelance Publishing Services, Brinscall, Lancs

Printed in Great Britain by CPI Antony Rowe Ltd, Chippenham, Wiltshire

Contents

ARTICLES 1.

Cultural Representation and the Practice of War in the Middle Ages Richard Abels

1

2.

The Brevium Exempla as a Source for Carolingian Warhorses Carroll Gillmor

32

3.

Infantry and Cavalry in Lombardy (11th–12th Centuries) Aldo A. Settia, translated by Valerie Eads

58

4.

Unintended Consumption: The Interruption of the Fourth Crusade at Venice and Its Consequences Gregory D. Bell

79

5.

Light Cavalry, Heavy Cavalry, Horse Archers, Oh My! What Abstract Definitions Don’t Tell Us About 1205 Adrianople Russell Mitchell

95

6.

War Financing in the Late-Medieval Crown of Aragon Donald J. Kagay

119

7.

National Reconciliation in France at the end of the Hundred Years War Christopher Allmand

149

1 Cultural Representation and the Practice of War in the Middle Ages Richard Abels

Several years ago I was inspired by John Keegan’s provocative book A History of Warfare (1993) to ask my students on their final exam to assess whether medieval warfare was “Clausewitzian.”1 Over the years this evolved into a much broader question that required the students to weigh the merits of two competing approaches to the study of medieval warfare. The version that appeared on last semester’s final exam reads: There are at least two ways of understanding the history of war. The first is a “scientific” model of war that emphasizes unchanging principles of strategic conduct and inherent military probability. According to this model, regardless of the era or society, war is a rational endeavor carried out according to tactical and strategic pragmatic necessities and directed at achieving the goals of a state. This approach also puts a priority on the material factors in war, in particular technological determinacy, and tests what the historical sources claim to have happened against what we know to be physiologically or technically possible, or, in some cases, militarily sensible. If the details recorded in even an authentic primary source fail this test, or stretch credibility, then they are to be rejected and material reality upheld. Others contend that war is a cultural activity: the reasons why societies engage in war and the methods by which they fight them are defined by the particular norms, values, institutions, and mentalities of a society passed on from one generation to the next.

1

This paper is based on my “Journal of Medieval Military History Lecture” sponsored by De Re Militari, presented at the 41st International Congress on Medieval Studies, 5 May 2006. The idea for the talk can be traced to a stimulating and wide-ranging dinner discussion with Bernard Bachrach, John Gillingham, and Stephen Morillo at the 2001 meeting of the Haskins Society at Cornell University. We spent hours over Indian food debating the influence of culture on the practice of war and to what degree, if any, “chivalry” influenced medieval military commanders. My thinking on these subjects has also profited from the insights of my colleagues in the History Department of the US Naval Academy, both military and civilian. I owe particular debts of gratitude to Capt. Timothy Feist, USMC, John France, Ernest Tucker, John Gillingham, Stephen Morillo, Giles Constable, and Clifford Rogers for critiquing drafts of this article. Their questions and criticisms led me to rethink and clarify a number of points and saved me from several errors. I would also like to thank my former colleague Victor Davis Hanson, who as the visiting Shifrin Professor of Military History at the US Naval Academy (2002–03) organized and generously funded a stimulating lecture series and faculty seminar on the question “whither military history?”

2

Richard Abels What we call the “unchanging principles of war” are themselves a cultural construct derived from a particular approach to war and a particular organization of the state characteristic of the West from the late eighteenth century to the present. We may call this the “culturalist” approach. The question: Assess the influence of culture, political policy, technology and practical military necessity upon why and how wars were fought in the Middle Ages. Which of the above interpretations do you think is more correct, and why? (You need not answer this as either/or. In the wars that you analyze you may find elements of each to be true.)2

It probably comes as no shock that my own view underlies the concluding caveat. But if some unreasonable professor forced me to declare an allegiance, I would have to count myself a “culturalist.” I do not doubt that there are certain constants in the experience of war over time, among them the physiological states of fear and excitement of those who kill and risk death, as well as material constraints upon the ways in which war can be conducted. The role played by the introduction and development of new military technologies, especially since the seventeenth century, undoubtedly has also been critical in shaping approaches to warfare although, like Kelly DeVries, I am wary of arguments based upon technological determinism and the ever increasing number of “military revolutions” technology has supposedly spawned.3 Fear of death, the desire for victory, material conditions, technology, economic development, and individual genius have certainly conditioned the historical development of the practice of war. But culture, by which I mean “the system of shared beliefs, values, customs, behaviours, and artifacts that the members of society use to cope with their world and with one another, and that are transmitted from generation to generation through learning,” has played an equally, perhaps even more, important role.4 The causes of 2

3

4

Similar dichotomies are proposed by Jeremy Black, Rethinking Military History (London, 2004), pp. 232–5, and Guy Halsall, Warfare and Society in the Barbarian West, 450–900 (London, 2003), pp. 6–8. The latter distinguishes between “functionalist” and “substantivist” approaches, which correspond closely to what I term “scientific” and “cultural.” See, e.g., Kelly DeVries, “Catapults are Not Atomic Bombs: Towards a Redefinition of ‘Effectiveness’ in Premodern Military Technology,” War in History 4 (1997): 454–70. See also Jeremy Black, “Determinisms and Other Issues,” The Journal of Military History 68 (2004): 1217–1232. This definition of culture is taken from Daniel G. Bates and Fred Plog, Cultural Anthropology, 3rd edn. (New York, 1990), p. 7. The term “culture” is notoriously slippery. By 1952 Alfred Kroeber and Clyde Kluckhohn were able to amass 156 definitions, which they classified under six headings. A. L. Kroeber and C. Kluckhohn, “Culture: a Critical Review of Concepts and Definitions,” Papers of the Peabody Museum of American Archeology and Ethnology 47 (1952): 41–72. The working definition I have adopted from Bates and Plog is, however, what most cultural anthropologists today mean by “culture.” Clifford Geertz offers a related definition that emphasizes the semiotic character of culture: “[culture] denotes an historically transmitted pattern of meaning, embodied in symbols, a system of inherited conceptions expressed in symbolic forms by means of which men communicate, perpetuate, and develop their knowledge about and attitudes toward life.” Geertz, “Religion As a Cultural System,” The Interpretation of Culture (New York, 1973), Chapter 4, p. 89. The danger in all cases is to define culture so broadly that it becomes tautological to say that it influenced conceptions and practice of war. This article emphasizes the normative element of culture (e.g., values, ethos, religious beliefs) rather than its institutional or material aspects.

Cultural Representation and the Practice of War

3

war – what is deemed worth killing and dying for – are defined by culture. Culture also dictates who fights and why, draws distinctions between legitimate and illegitimate conduct in war, and defines what constitutes victory. Culture, moreover, shapes how the physiological and psychological states associated with war – fear, excitement, joy – are understood and experienced as “emotions.”5 Even technological development is conditioned by culture, as Stephen Morillo and John Keegan have shown in their comparisons of the development and use of gunpowder weaponry in sixteenth- and seventeenth-century Europe and Japan.6 What complicates matters is the multiplicity of cultures contained within societies and even professions. In my twenty-five years teaching at the Naval Academy, I have worked with literally hundreds of military officers and have come to appreciate the profoundly different cultures of the various warfare specialties within the United States’ naval service. That the self-conscious esprit de corps of the Marines distinguishes them from naval officers did not come as a surprise, but the profound cultural differences among (and within) the Nuclear Power, Naval Aviation, and Surface Warfare Communities did. The situation for those who fought and those who reported that fighting in twelfth- and thirteenth-century England and France was no less complicated. In the High Middle Ages there was no one chivalric ethos, let alone a single culture shared by those who prayed, those who fought, and those who worked – even among those who fought, there were significant cultural differences between the elite combatants and the nonnobles who composed the mass of the fighters. Scholarship on medieval “chivalry” has emphasized the contested and dynamic character of the construct, as representatives of the Church, royalty, and the knights themselves competed to define a normative ethos for the warrior aristocracy.7 But, as dangerous as it is to rely upon cultural constructs, I think it incontestable that cultural considerations and constraints fundamentally shaped medieval warfare on all levels, from defining casus belli, to strategic and tactical decision-making, to the conduct and experiences of ordinary soldiers.8 In addressing the issue of cultural influences upon the practice of war I am entering into an ongoing debate that until recently has tended to focus on modern Western warfare, in particular attempts to explain the exceptional ferocity of 5 6

7

8

On emotions as cultural and social constructs, see the essays in Barbara H. Rosenwein, ed., Anger’s Past: The Social Uses of an Emotion in the Middle Ages (Ithaca, NY, 1998). S. Morillo, “Guns and Government: A Comparative Study of Europe and Japan,” Journal of World History 6 (1995): 75–106; John Keegan, A History of Warfare (New York, 1993), pp. 42–6. See also Noel Perrin, Giving up the Gun: Japan’s Reversion to the Sword, 1543–1879 (Boston, 1979); but cf. the review by Conrad Totman in The Journal of Asian Studies 39 (1980): 599–601. Cf. also Black, “Determinisms and Other Issues,” pp. 1219–23. See in particular Maurice Keen, Chivalry (New Haven, 1984), Richard Kaeuper, Chivalry and Violence in Medieval Europe (Oxford, 1999), and Constance Bouchard, Strong of Body, Brave & Noble: Chivalry and Society in Medieval France (Ithaca, NY, 1998), pp. 103–16, 129–44. I am using the term “soldier” here to mean all those who fight in war under the direction of commanders. In doing so, I realize that I am blurring the distinction between warriors, those who fight as a condition and expression of their membership in a social class, and soldiers, those who fight under discipline as members of an institutional military.

4

Richard Abels

the American Civil War and the two World Wars.9 The “culturalist” ranks boast a number of prominent military historians, notably John Dower, Craig Cameron, John Keegan, Victor Davis Hanson, John Lynn, Jeremy Black, and Ted Lendon, and among medievalists they include Stephen Morillo, Richard Kaeuper, and Matthew Stickland. The “scientific school,” however, still dominates the field and its practitioners are legion. This is certainly true for traditional operational military history, whether academic or popular. “Scientific” assumptions also underlie much of the (now not-so-new) “new military history” written about the organization, recruitment, financing, and logistics of military forces.10 The most notable proponent of the “scientific” school among today’s medieval historians is undoubtedly Bernard S. Bachrach, arguably the reigning doyen of American medieval military studies. John France, although more reluctant than Bachrach to apply modern military jargon and strategic analysis to medieval warfare, also emphasizes material over cultural factors. For France, cultural constructs take a back seat to geography, climate, topography, technology, and political and social structures in determining how societies wage war.11 (One might question, however, whether one can truly separate political and social structures from culture.12 ) Standing somewhere in between is John Gillingham, whose seminal 1984 article “Richard I and the Science of War in the Middle Ages” argued for a rational medieval “science” of war, based on an inherited late-Roman strategic doctrine that twelfth-century military commanders adapted to the particular castledotted military topography that made battle-avoidance militarily prudent. In more recent work, however, Gillingham treats chivalry (in his words) “not as a mere gloss upon the brutal realities of life, but as an important development in political morality,” one which contributed greatly to the Normans’ impression of the non-chivalrous Welsh, Scots, and Irish as barbarians, a conclusion to which Matthew Strickland independently arrived.13

9 10 11 12

13

John Shy, “The Cultural Approach to the History of War,” The Journal of Military History 57 (1993): 13–26. E.g., Geoffrey Parker, The Military Revolution: Military Innovation and the Rise of the West, 1500–1800 (Cambridge, 1988). John France, Perilous Glory (New Haven, CT, forthcoming). David Kaplan, for example, claims that most American anthropologists view human society as “a sub-system or part of culture” and includes social organization as a constituent element of culture. Kaplan, “The Superorganic: Science or Metaphysics,” American Anthropologist 67 (1965), p. 960, n. 2. See also Edward Sapir, Selected Writings of Edward Sapir in Language, Culture, and Personality, ed. David Mandelbaum (Berkeley, CA, 1949), p. 515. Miguel A. Cabrera’s Postsocial History: An Introduction (Lanham, MD, 2004), argues that all social structures are culturally determined and that social history is merely a subset of cultural history. But cf. the critical review by S. H. Rigby, “History, Discourse, and the Postsocial Paradigm: A Revolution in Historiography?” History and Theory 45 (2006): 110–23. John Gillingham, “Conquering Barbarians: War and Chivalry in the Twelfth Century Britain.” The Haskins Journal 4 (1992): 67–84, in John Gillingham, The English in the Twelfth Century: Imperialism, National Identity and Political Values (Woodbridge, 2000); Matthew Strickland, “Slaughter, Slavery or Ransom? The Impact of the Conquest on Conduct in Warfare,” in England in the Eleventh Century, ed. C. Hicks (Stamford, 1992), pp. 41–60.

Cultural Representation and the Practice of War

5

Victor Davis Hanson and John Keegan have probably done more than anyone to popularize the culturalist approach to military history. Keegan may even be credited with initiating the wider debate with the sustained attack he launched against Clausewitz’s definition of war as the continuation of policy by other means in his A History of Warfare. And yet Hanson and, to a lesser extent, Keegan have also done a disservice to the “culturalist” position by advancing the reductionist idea that there has been over the centuries a particular Western Way of War distinct from and deadlier than the approaches to war found in other parts of the globe.14 For Hanson – and it is his thesis that Keegan advances – this “Western Way of War” is characterized by citizen armies and navies organized tactically in close-order formations designed for a battle-seeking strategy aimed at the annihilation of the enemy. Imbued with Western ideals of individual freedom, democracy, and political equality, these citizen-soldiers and sailors have demonstrated through the ages exceptional discipline in the ranks, while simultaneously appreciating and encouraging individual initiative and innovation. The West’s dominance in military technology, according to Hanson, is also due to its unique culture, “the product of a long-standing Western approach to applied capitalism, science, and rationalism not found in other cultures.”15 The result is a rational, amoral, and lethal approach to war, in which ideas are adopted or rejected upon the basis of their military efficacy. The classicist Hanson, who has authored several superlative books about classical Greek combat, unsurprisingly traces the roots of this style of warfare to the experience of Greek hoplites in battle.16 John France has shown, however, that Hanson’s thesis of Western military exceptionalism does not bear up under scrutiny. As France observes, “Battle was only one aspect of war: in both Europe and the Middle East the focus of fighting was ravaging and destruction, and in both areas strong fortifications were the key to ruling the countryside.” “In both East and West,” he adds, “readiness for battle governed the make-up of armies that had always to campaign with battle in mind, if not in intention. At heart, the differences between Eastern and Western styles of war were the result of climate, geography, and topography.”17 Although France does not discount culture completely, acknowledging that “few cultural constructs are more obvious, enduring, and powerful than chivalry,”18 for him “the emphasis on culture as the setting for developments can mask the underlying realities that shaped it, such as climate and topography, whose impact 14 15 16

17 18

Victor Davis Hanson, Carnage and Culture: Landmark Battles in the Rise of Western Power (New York, 2001); Keegan, History of Warfare, pp. 386–92. Hanson, Carnage and Culture, p. xx. Warfare and Agriculture in Classical Greece (Berkeley, 1983); The Western Way of War (New York, 1989); ed., Hoplites: The Ancient Greek Battle Experience (London and New York, 1991); Wars of the Ancient Greeks (London, 1999); A War Like No Other: How the Athenians and Spartans Fought the Peloponnesian War (New York, 2005). John France, “Close Order and Close Quarter: The Culture of Combat in the West,” The International History Review 27 (2005), p. 502. France, “Close Order,” p. 500.

6

Richard Abels

upon human behaviour is not necessarily mediated by cultural constructs.”19 France’s objections to Hanson’s cultural determinacy apply equally to John Lynn’s more theoretically sophisticated approach to culture and warfare. Lynn’s Battle: A History of Combat and Culture rejects Hanson’s Western essentialism in favor of cultural complexity and cultural change over time and space. Applying the theoretical methodology of the “new cultural history,” Lynn differentiates “between the reality of war and the way in which a culture conceives of war.” The latter he terms “the discourse on war,” and acknowledges that “a single society can harbor several discourses on war that vary by class, gender, and profession.” 20 Lynn posits a dialectical interaction between the discourse and reality of war, in which the former “tries to modify reality to more nearly resemble conceptions of how war should be” and, simultaneously, adjusts to reality, “if for no other reason than survival.”21 In terms of medieval warfare, Lynn identifies “the formalized and elegant discourse of chivalry and the exceptionally brutal reality of warfare during the Late Middle Ages, exemplified in the armed raids, or chevauchées, of the Hundred Years’ War.” The discord between the two led to the development of two distinct “perfected” versions of war, the tournament and the Crusades, which represented, respectively, the discourses on war of the martial aristocracy and the papacy.22 A major flaw in Lynn’s thesis of cultural determinacy is his failure to explain how and why discourses on war and the reality of its practice fall out of synch. Surely, material and technological changes must play a major role in this, and Lynn’s model would be more persuasive – and far more complicated – if it accommodated the dialectical relationship amongst changing material conditions, practice, and discourse instead of being dyadic. Despite Lynn’s refusal to acknowledge the importance of material factors in shaping and altering both the discourse and the practice of warfare, his approach and methodology represent a real contribution to the study of military history. I have serious reservations, however, about his application of them to medieval warfare and the distinction he draws between “the Ideal, the Real, and the Perfect.” “Discourse,” as the term is used in cultural studies, is not an idealization of reality but rather the shared symbolic representation of experience by the members of a group, expressed through language, rituals, and behavior. Discourse is the framework within which members of a social or culturally defined group think about, understand, and express their experiences.23 In this sense, the term “discourse” overlaps ethnographer Clifford Geertz’s influential semiotic definition of culture as “webs of significance” that belong collectively to a particular people and which are manifested in that group’s rituals, public practices, and symbolic expressions, including 19 20

21 22 23

Ibid. John Lynn, Battle: A History of Combat and Culture (Cambridge, MA, 2003), p. xx. A good introduction to the new cultural history is Lynn Hunt, ed., The New Cultural History (Berkeley and Los Angeles, 1989). Lynn, Battle, p. xxi. Lynn, Battle, pp. xxiii, pp. 72–109. Cabrera, Postsocial History, pp. 21–27.

Cultural Representation and the Practice of War

7

language.24 Medieval discourses on war, then, are the various vocabularies and modes of expression employed by the lay and ecclesiastical elites to describe and explain why and how wars were fought and experienced in their time. Lynn’s distinction between a medieval aristocratic discourse on war, in which tournament is “perfected warfare” and chevauchées the grubby reality, misses the point. Battle and chevauchée belonged to the same aristocratic discourse on war; they simply occupied different places in it. Tournaments, on the other hand, occupied a different sphere, that of sport, which was related to but distinct from war. Both war and tournament, moreover, belonged to the same aristocratic (though not clerical) discourse on chivalry. Chivalric writers such as Ramon Lull and Geoffroi de Charny and even the author of the Histoire de Guillaume le Maréchal understood the difference between war and martial sports and accorded the former a higher degree of honor. One sought chivalric perfection, according to de Charny, in war, not tournaments.25 Central to the aristocratic twelfth- and thirteenth-century discourse on war were cultural concepts of honor and shame. When chroniclers and poets wrote about war, they distinguished between actions they deemed honorable, that is, behavior that enhanced reputation, and the normal and unremarkable activities of war, which included ravaging fields, burning villages, and the drudgery of siege warfare. The latter were activities accepted by knights and kings, if not by reforming churchmen, as necessary for achieving victory and economic solvency, but which did not often provide opportunities for accruing honor.26 As Matthew Strickland observed, the consensus among the knights themselves was that, although they accepted ravaging as a legitimate wartime activity fully compatible with their notions of class-based “chivalry,” “there was equally a sense that burning of fields or attacks on peasantry gained the warrior more material profit than martial glory.”27 Finally, there were actions and behavior that brought shame to

24

25 26

27

Clifford Geertz, “Description: Toward an Interpretive Theory of Culture,” The Interpretation of Culture, Chapter 1, pp. 5–13. See also Geertz’s fuller definition of culture in note 10 above. Cf. William H. Sewell, “The Concept(s) of Culure,” in Victoria E. Bonnell and Lynn Hunt, eds., Beyond the Cultural Turns: New Directions in the Study of Society and Culture (Berkeley and Los Angeles, 1999), pp. 31–61. Richard Kaeuper and Elspeth Kennedy, ed. and trans., The Book of Chivalry of Geoffroi de Charny: Text, Context, and Translation (Philadelphia, 1996), pp. 84–91. Recent scholarship emphasizes the compatibility between the aristocratic elite’s conception of chivalry, which had little or nothing to do with the treatment of the lower orders, with the tendency in medieval warfare to target noncombatants in waging war. See Matthew Strickland, War and Chivalry (Cambridge, 1996), pp. 258–90, especially pp. 281–90; Richard Kaeuper, Chivalry and Violence, pp. 176–85; Maurice Keen, Chivalry (New Haven, 1984), pp. 227–33. The attempts of churchmen from the late tenth-century Peace of God movement on to protect non-combatants from the ravages of war and to foster a Christian chivalry in which the knight was the protector of the inermes seem to have had as little practical effect as the ecclesiastics’ repeated condemnations of the tournament. Kaeuper and Maurice Keen both argue that by linking honor to prowess in war, chivalry promoted rather than restrained violence and disorder. Strickland, War and Chivalry, p. 290. Cf. Keen, Chivalry, pp. 227–33.

8

Richard Abels

the perpetrator. The boundaries delineating these three categories were fuzzy. What was honorable, unremarkable, or shameful was determined both by the author who described the actions and by his audience, and we need not assume that author and audience always agreed. Over time discourses on war evolve, and with it the responses of readers to narrative accounts of warfare change. Take, for example, Henry V’s command to kill the prisoners during the battle of Agincourt, an action which for some historians unmasks Shakespeare’s heroic warrior-king for what he truly was, a cold, ruthless, and brutal hypocrite who thirsted for victory at any cost.28 This disapproval would have surprised Henry’s contemporaries. Despite general admonitions in late medieval treatises on the laws of arms against the mistreatment of unarmed prisoners,29 no fifteenth-century chronicler, English or French, condemns Henry for this incident. The chroniclers and their intended audiences understood the act to be a military necessity precipitated by the fog of war. The earliest English account of Agincourt, the Gesta Henrici Quinti, written soon after Henry’s triumphant return to London by a royal chaplain who had witnessed the battle, represents the incident as a spontaneous response by the English soldiers: because of what wrathfulness on God’s part no one knows, a shout went up that the enemy’s mounted rearguard (in incomparable number and still fresh) were re-establishing their position and line of battle in order to launch an attack on us, few and weary as we were. And immediately, regardless of distinction of person, the prisoners, save for the dukes of Orleans and Bourbon, certain other illustrious men who were in the king’s battle, and a very few others, were killed by the swords of either captors or of others

28

29

So Desmond Seward, Henry V as Warlord (London, 2002), pp. 80–1. Cf. Juliet Barker’s judgment: “In humanitarian terms, Henry’s decision was indefensible: to order the killing of wounded and unarmed prisoners in such a cold and calculated way violated every principle of decency and Christian morality. In chivalric terms it was also reprehensible.” Agincourt: Henry V and the Battle that Made England (New York, 2005), p. 289. It is unclear whether this represents Barker’s personal opinion or that of contemporaries. If the latter, she presents little evidence in support of it other than citing Christine de Pizan’s admonition in chapter 17 of The Book of Deeds of Arms and of Chivalry that one not only ought to spare those who surrender but is obliged to defend the prisoner against all those who would harm him. See, e.g., Honoré Bonet (or Bouvet), Tree of Battles (c. 1387), ed. and trans. G. W. Coopland (Liverpool, 1949), p. 134; Christine de Pizan, The Book of Arms and Deeds of Chivalry (c. 1410), ed. Charity Cannon Willard, trans. Sumner Willard (University Park, Pennsylvania, 1999), pp. 169–70. Both Bonet and Christine de Pizan, however, made exceptions for prisoners who would pose a continuing danger to the prince if allowed to live. See Theodor Meron’s discussion of the status of the question at the time Shakespeare wrote: Meron, Henry’s Wars and Shakespeare’s Laws: Perspective on the Law of War in the Later Middle Ages (Oxford, 1993), pp. 154–71; idem, Bloody Constraint: War and Chivalry in Shakespeare (Oxford, 1998), pp. 191–202. On the status of prisoners of prisoners of war in the medieval laws of arms, see Maurice Keen, The Laws of War in the Late Middle Ages (London, 1965), pp. 104–6, 156–85. It is noteworthy that late medieval legal theorists devoted much more attention to the rules of ransom than the question of quarter.

Cultural Representation and the Practice of War

9

following after, lest they should involve us in utter disaster in the fighting that would ensue.30

This is an explanation, not an apology. Nonetheless, Henry’s absence from the account is striking; the explanation is to be sought in the message and agenda of the Gesta. The Gesta is the “official English version” of Henry’s campaign, and in it Agincourt is presented as a trial by combat which proved the validity of Henry’s claim to the French throne. The Gesta’s Henry is so confident of the justice of his cause that he rejoices in the disproportion between his small band and the much larger French forces. In response to Sir Thomas Hungerford’s expressed desire that the king have ten thousand more of England’s best archers to add to his “little company” (paucam familiam), Henry responds: “By the God in Heaven upon Whose grace I have relied and in Whom is my firm hope of victory, I would not, even if I could, have a single man more than I do. For these I have here with me are God’s people, whom He deigns to let me have at this time. Do you not believe,” he asked, “that the Almighty, with these His humble few, is able to overcome the opposing arrogance of the French who boast of their great number and their own strength?”31

As in the case of the Maccabbees, the victory of the few over the many could only be accounted for by divine intervention. Even if militarily understandable, an order to kill the prisoners out of fear of impending disaster would hardly square with the desired image of Henry as a new Judas Maccabaeus relying with faith upon the will of God.32 For contemporary and near-contemporary French chroniclers, the battle’s outcome was not a judgment of God upon Henry’s claim to the French throne but additional evidence of the ineptitude, selfishness, and division amongst the French nobility that cost them victory.33 The French chroniclers agree that Henry, fearing

30

31 32

33

Gesta Henrici Quinti, ed. and trans. Frank Taylor and John S. Roskell (Oxford, 1975), pp. 90– 3, excerpted in Anne Curry, The Battle of Agincourt: Sources and Interpretations (Woodbridge, 2000), p. 37. Cf. Thomas Elmham, Liber Metricus de Henrico Quinto, chap. 37, in Curry, Agincourt, p. 47; Tito Livio Frulovisi, Vita Henrici Quinti, ibid., p. 62; Pseudo Elmham, Vita et Gesta Henrici Quinti, ibid., p. 73 Gesta Henrici Quinti, ed. Taylor and Roskell, pp. 78–9. Gesta Henrici Quinti, ed. Taylor and Roskell, p. 78: “Et non potuit, iudicio meo, ex vera justicia dei, filio tam grandis confidencie infaustam quid accidere, sicuti nec Iude Machabeo accidit usque in diffidenciam cecidit, et inde merito in ruinam.” Accounts of Monk of Saint-Denis, Histoire de Charles VI (c. 1415–22), in Curry, Agincourt, p. 108; Mémoires de Pierre de Fenin (1430s), ibid., p.118; Chronique de Ruisseauville (1420s– 1430s), ibid., p.125; Jean Juvenal des Ursins, Histoire de Charles VI (1430–1440s), ibid., p. 125; Enguerran Monstrelet, Jean Waurin and Jean Le Fèvre (1444–1460s), ibid., pp. 163–4; and Edmond de Dynter, Chronique des ducs de Brabant (early 1440s), ibid., p. 174. Jean Waurin and Jean Le Fèvre report that when the English men-at-arms refused to obey Henry’s order to kill the prisoners they had taken, “for they were all hoping to collect a large ransom,” Henry turned the job over to his archers (ibid., pp.163–4). Edmond de Dynter blames Clignet de Brabant for having misled the English into thinking that he was about to renew the assault when actually he “returned not to fight but to pillage” (ibid., p. 174).

10

Richard Abels

that the French were re-forming for another assault, ordered his men to kill the prisoners. Not one chronicler, however, reproaches him for it. Instead, they blame either feckless French noblemen or opportunistic locals whose looting of the English baggage train gave the false impression of a renewed attack. These French nobles, like the English knights who refused to obey their king’s order lest they lose out on ransom money, acted from self-interest to the detriment of king and country.34 Shakespeare, writing almost two centuries after the event, apparently felt less certain about the propriety of the order. He did not ignore the incident, which he found in Holinshed, but he did the next best thing, devoting only a few lines to it and providing not one but two justifications. Shakespeare has Henry become aware of a possible renewed French attack while trying to control his tears over the news of the deaths of his kinsmen the Duke of York and the Earl of Suffolk. The king responds quickly and decisively, crying out: But hark, what new alarum is this same? The French have reinforced their scattered men. Then every soldier kill his prisoners. Give the word through. (Henry V, iv.vi. 35–8)

Suddenly Shakespeare shifts the scene to the devastated English rear camp, where the professional soldiers Captain Fluellen and Gower survey with anger and horror the bodies of the massacred valets. FLUELLEN: Kill the poys [sic] and the luggage! ’tis expressly against the law of arms: ’tis as arrant a piece of knavery, mark you now, as can be offer’t; in your conscience, now, is it not? GOWER: ’Tis certain there’s not a boy left alive; and the cowardly rascals that ran from the battle ha’ done this slaughter: besides, they have burned and carried away all that was in the king’s tent; wherefore the king, most worthily, hath caused every soldier to cut his prisoner’s throat. O, ’tis a gallant king!35 (iv.vii. 1–11)

34

35

The Monk of Saint-Denis condemns the French to eternal opprobrium for their cowardly flight and explains that the order to kill the prisoners was precipitated by a panicked movement to the rear of French troops trying to flee the battle (Curry, Agincourt, p. 108). Pierre de Fenin and Monstrelet blame Isambart d’Azincourt and Robert de Bournville and report that both men “were later much blamed for this and also were punished by Duke John of Burgundy” (ibid., p. 118). On the fifteenth-century idea that the duty of knights and nobles was to promote the “common good,” see Christopher Allmand, The Hundred Years War: England and France at War, c.1300–c.1450 (Cambridge, 1989), pp. 52–4, 147, 149–50, 154–7, and Malcolm Vale, War and Chivalry: Warfare and Aristocratic Culture in England, France, and Burgundy at the End of the Middle Ages (Athens, Georgia, 1981), pp. 167–70. Meron, Bloody Constraint, p. 192, characterizes Gower’s response as “scathing sarcasm” that “highlights Shakespeare’s substantial doubts about the legality of the order.” See also Meron, Henry’s Wars, p. 160. The discussion between Fluellen and Gower that follows, however, does not support this construction.

11

Cultural Representation and the Practice of War

In Shakespeare’s hands, Henry’s decision to kill the prisoners pales to insignificance when juxtaposed with the atrocity committed by the French in clear violation of the laws of arms. Lest his audience miss the point, Shakespeare has Henry burst out upon viewing the bodies of the dead boys, “I was not angry since I came to France until this instant.” Turning to his herald, he orders him to tell the enemy either to engage him without expectation of quarter or to quit the battlefield. Almost as if he had forgotten the previous scene, Shakespeare now alters the reason for Henry’s order to kill the prisoners. Besides we’ll cut the throats of those we have, And not a man of them that we shall take Shall taste our mercy. Go and tell them so. (iv.vii. 54–64)

No longer is the order to kill the prisoners simply a response to a perceived military threat; it now also becomes a righteous act of anger and reprisal against the “arrant piece of knavery” the French had perpetrated “expressly against the laws of arms.” Theodor Meron tellingly points out that in order to provide Henry with this second justification, Shakespeare altered his historical source. Holinshead had the French killing only “such servants as they found to make any resistance”; the indiscriminate massacre of “the boys” is Shakespeare’s own invention.36 Through artistry and ingenuity Shakespeare thus minimized and excused what the playwright himself apparently saw as a blemish on his hero. Laurence Olivier and Kenneth Branagh in their film versions of the play apparently were even more uncomfortable with the idea that Henry had ordered a massacre of unarmed prisoners. Shakespeare made it easy for them to eliminate this problematic incident entirely. Olivier, who filmed the play in the midst of World War II to raise the patriotic spirits of his British audience, sanitized the play further by cutting Henry’s blood-curdling threat to unleash his troops upon the population of Harfleur if the town’s governor did not capitulate. Shakespeare’s audience understood well what the inhabitants of a town could expect if taken by storm, especially if that town was in rebellion against its lawful lord. But Olivier and his audience could not square the notion that Henry’s “band of brothers” could become “the blind and bloody soldier with foul hand” defiling townsmen’s “shrill-shrieking daughters,” dashing out the brains of their most reverend fathers, and spitting upon pikes their naked infants. Images like that were reserved for the barbarity of the Hun. Branagh, whose 1989 movie version of the play drew upon the grit, mud and confusion of Keegan’s Agincourt in The Face of Battle, restored the speech, but softened its impact by having Henry visibly sigh with relief when the governor of Harfleur surrenders the city, suggesting that Henry was bluffing, or at the very least reluctant to let loose the dogs of war upon the innocents of the town.

36

Meron, Henry’s Wars, pp. 156–7.

12

Richard Abels

Here laid out before us is the changing discourse on war between 1415 and the present. The killing of the prisoners was for contemporary English and French chroniclers a reasonable response to the threat of a renewed attack. Although it did not enhance Henry V’s honor and embarrassingly (for Henry) highlighted his uncertainty about the battle’s outcome, the king’s order to kill the prisoners was, given the circumstances, not shameful. Shakespeare, whose Henry V at times comes close to being a dissertation on the law of arms, was less certain, and by the middle of the twentieth century it had become an atrocity impermissible for a hero. What is deemed to be acceptable behavior in warfare is culturally determined. On 16 March 1968, a platoon led by Lt William Calley crossed the line for most Americans: they massacred between 350–500 unarmed villagers, mostly women, children, and elderly men, in the Vietnamese hamlet of My Lai. More would have been killed if not for the intervention of an American helicopter pilot, Warrant Officer Hugh Thompson, Jr., who ordered his machine-gunners to open fire upon the American troops if they did not cease the slaughter. Thirty years later the Army would recognize Thompson’s intervention with the Soldier’s Medal, the highest military award for heroic action “not involving actual conflict with an enemy.” At the time, however, the Army wasn’t interested in acknowledging either Thompson’s heroism or the massacre that had made it necessary. Thompson’s report was ignored. Instead, the Army trumpeted the action at My Lai as a victory in which 128 insurgents had been killed, and covered up allegations that a massacre occurred. When reporter Seymour Hersh finally broke the story a full year later, the Army responded by indicting Calley for premeditated murder and 25 other soldiers for related offenses. The Army’s inquiry revealed gruesome details. One participant, for example, described how a baby had been used for target practice: “He fired at it with a .45. He missed. We all laughed. He got up three or four feet closer and missed again. We laughed. Then he got up right on top and plugged him.”37 Another soldier testified that he went into shock after shooting a woman and her baby: “The baby’s face was half gone, my mind just went … and I just started killing. Old men, women, children, water buffaloes, everything … I just killed … That day in My Lai, I was personally responsible for killing about 25 people.”38 Many young girls of the village were raped before they were killed, some while their horrified families looked on. In the end, only Calley was convicted. The military court imposed a life sentence at hard labor, which was subsequently reduced on appeal to ten years. After serving three and a half years of house arrest in his quarters at Fort Benning, Georgia, Calley was paroled by the Army Secretary Howard H. Callaway and released.

38

“Into the Dark: The My Lai Massacre,” http://www.crimelibrary.com/notorious_murders/mass/ lai/dark_5.html, accessed 3 May 2007.

Cultural Representation and the Practice of War

13

The Army’s worries may have been justified; a month after Calley’s conviction the Gallup poll registered disapproval for the war at a new high, 61%, and a Harris survey found that for 42% of Americans polled, My Lai served as proof that US involvement in Vietnam had been morally wrong from the start.39 Paradoxically, Calley’s conviction was greeted by the public with an outpouring of popular support that cut across the hawk/dove divide. Most of Calley’s sympathizers viewed him as a scapegoat punished for following the criminal orders of his superiors, but a vocal minority refused even to acknowledge that a crime had been committed.40 For them Calley was a young patriot who was being punished for valuing his troop’s lives more highly than the enemy’s.41 But over time, however, as the details of the massacre became better known and passions over the war cooled, public sentiment gradually shifted to shock, horror, and shame. My Lai brought up unwelcome and often suppressed memories among veterans of World War II, and many of those who participated in the massacre were psychologically scarred for life.42 The question that My Lai raised, and which it continues to raise today, is how could Americans, the good guys, commit such an atrocity? The late twelfth-century knight and troubador Bertrand de Born would have found this national soul-searching very puzzling. What do Americans think happens in war? he might have asked. Rape, pillage, and slaughter, he would have reminded us, are what one does when sacking a city. A number of Calley’s apologists voiced sentiments that on first glance appear similar. Judge Robert Elliot expressed the views of many when he observed upon releasing Calley on parole in 1974, that “War is war, and it’s not unusal for innocent civilians such as the My Lai victims to be killed. It has been so throughout recorded history.”43 There is a profound cultural difference, however, between General Sherman’s observation that war is hell and the sentiments of twelfth-century knights such as Bertrand de Born. For the leadership of today’s professional US military and the American public at large, war is, at best, a regrettable necessity, fought for a righteous cause, and to be engaged in as a last resort. Incidents such as My Lai, although perhaps understandable in the heat of combat, nonetheless are regarded as moral 39

40

41 42

43

Poll released on 6 June 1971, in George H. Gallup, The Gallup Poll: Public Opinion 1935– 1971, 3 vols. (New York, 1972), 3:2309. Cf. poll of 28 June 1970, which showed support at 56%. Ibid., 3:2254. According to a Gallup poll of 7 April 1971, 79% disapproved of Calley’s conviction and sentence, although only 15% thought believed him guilty of no crime. Gallup Poll, 3:2296 (poll of 7 April 1971). For the Harris survey, see Michael R. Belknap, The Vietnam War on Trial: The My Lai Massacre and the Court-Martial of Lieutenant Calley (Lawrence, Kansas, 2002), 211. This was Calley’s own justification at his sentencing hearing. Michael R. Belknap, The Vietnam War on Trial, p. 190. For the response of World War II veterans, see Kendrick Oliver, “Atrocity, Authenticity and American Exceptionalism: (Ir)rationalising the Massacre at My Lai,” Journal of American Studies 37 (2003): 257–8. For the psychological impact, see Kendrick Oliver, The My Lai Massacre in American History and Memory (Manchester, UK, 2006), pp. 255–8. Oliver, “Atrocity, Authenticity and American Exceptionalism,” p. 259.

14

Richard Abels

lapses, and are usually attributed to the influence of a few “bad apples.” Bertrand de Born, on the other hand, gloried in the chaos and slaughter of war. “War is no noble word,” he announced, “when it’s waged without fire and blood.”44 King Henry V of England enthusiastically agreed: “War without fire,” he is reputed to have said, “is as worthless as sausages without mustard.”45 Medieval clergy, indeed, could even rejoice in a massacre if done in a good cause. Consider Abbot Arnold-Almaric’s notorious solution to the problem of distinguishing between the faithful and heretics during the sack of Béziers – “Caedite eos. Novit enim Dominus qui sunt eius,” “Kill them. For God will know His own.”46 In its punchier paraphrase, “Kill ’em all and let God sort them out,” the Cisterican abbot’s Solomonic judgment was adopted as the unofficial motto by Green Berets and Ranger units in Vietnam in what was clearly – I hope – an exercise in military black humor. Those who scribbled this on their fatigues and tee shirts did so with the intention of shocking and offending ‘peaceniks.’ But what is more shocking to modern sensibilities is that we owe the anecdote not to a critic of the Albigensian Crusade but to one of its most ardent supporters, Caesarius of Heisterbach. Caesarius, a fellow Cistercian monk and preacher, reported Arnold-Amalric’s words approvingly in his “Dialogue on Miracles” as an example of righteous Christian zeal. Caesarius’s spokesman in the dialogue, an older monk, explains to a novice about various heresies that threaten the Church. Chief among them is the Albigensian heresy, which, he tells the novice, had spread so widely that “if it had not been cut back by the swords of the faithful … it would have corrupted the whole of Europe.”47 When the Crusaders arrived before the walls of heretic-infested Béziers, the heretics within the city “defiled in an unspeakable manner the book of the sacred gospel; and then cast it from the wall towards the Christians, and sending arrows after it, cried: ‘There is your law, miserable wretches!’ But Christ, the author of the gospel, did not suffer such an insult to be hurled at Him unavenged.”48 And avenge the blasphemy He did through the swords of the Crusaders. Caesarius’s audience certainly would have appreciated this. They may also have known that Bishop Renaud of Montpellier had called upon the Catholics of Béziers either to hand over their heretical neighbors or at least to quit the city. Peter of les Vaux-de-Cernay’s reaction to their refusal was that by setting “themselves up against God and the Church, [Béziers’s

44 45

46

47 48

William D. Paden, T. Sankovitch, and P. H. Stäbelein, eds. and trans., The Poems of the Troubadour Bertran de Born (Berkeley, 1986), pp. 358–9. For the French, see Juvenal des Ursins, Histoire de Charles VI, quoted in John Gillingham, “Richard I and the Science of War in the Middle Ages,” in John Gillingham and J. C. Holt, eds., War and Government in the Middle Ages (Woodbridge, 1984), 85n. Caesarius Heiserbacensis monachi ordinis Cisterciensis, Dialogus miraculorum, Distinctio Quinta, Cap XXI , 2 vols., ed. J. Strange (Cologne, 1851), 1:302. The second clause is a paraphrase of 2 Timothy 2:19. Ibid., 301. Ibid., 302. Cf. Peter of les Vaux-de-Cernay, The History of the Albigensian Crusade, §89, trans. W. A. Sibly and M. D. Sibly (Woodbridge, 1998), p. 50.

Cultural Representation and the Practice of War

15

Catholics] had made a covenant with death” [Isaiah 28:15].49 They would also have understood Arnold-Amalric’s admonition within a broader Christian theological discourse in which the true life of a faithful Christian only began with his physical death. But did Arnold-Amalric actually say the words attributed to him by Caesarius? At best we can say, “perhaps.” The Dialogus Miraculorum was written about forty years after the sack and the story is found in no contemporary source. Caesarius, moreover, did not claim first-hand knowledge of the abbot’s words, which he prefaced with the phrase “fertur dixisse,” “he is reputed to have said.” But we should not dismiss out of hand the report by a Cistercian monk writing about a Cistercian abbot within living memory of the event.50 The Albigensian Crusade was what Stephen Morillo terms “subcultural warfare,” warfare within a shared larger culture in which the foe is understood “as an incarnation of evil, or more practically as maliciously motivated underminers of order, whether sociopolitical, cosmic, or both.” In warfare of this sort, the enemy is constructed as “devils or fallen humans in league with devils” who have willfully rejected Truth. The threat they pose is total, justifying, even necessitating, their extermination. Given this, it is not surprising that a massacre took place in Béziers on 22 July 1209 and that Catholics as well as heretics were put to the sword.51 In a letter to Pope Innocent III written within weeks of the massacre, Arnold-Amalric and his fellow legate Milo reported that “our men spared no one, irrespective of rank, sex or age, and put to the sword almost 20,000 people.”52 This may be an exaggeration, but it confirms what all the contemporary sources report: the slaughter was exceptionally ferocious even by thirteenth-century standards. We can be equally certain that, whether he actually said them or not, the words Caesarius attributed to Arnold-Amalric were not intended to mark him as hypocritical or unfeeling, but to illustrate his pious zeal and trust in the justice of God. Similarly, Richard the Lionheart’s decision to massacre 2600 Saracen prisoners at Acre presents greater difficulties for modern admirers of Richard than it did for contemporary and near-contemporary Christian chroniclers who reported

49 50 51

52

Peter of les Vaux-de-Cernay, History of the Albigensian Crusade, §89, trans. Sibly and Sibly, 50. See discussion and references in Sibly and Sibly, “Appendix: Massacre of Béziers,” in Peter of les Vaux-de-Cernay, History, pp. 289–93. Stephen Morillo, “A General Typology of Transcultural Wars – The Early Middle Ages and Beyond,” in Transcultural Wars from the Middle Ages to the 21st Century, ed. Hans-Henning Kortüm (Berlin, 2006), pp. 36–7. Malcolm Barber concluded that “the Albigensian crusades went far beyond the normal conventions of early thirteenth-century warfare, in the scale of the slaughter, in the execution of high-status opponents, male and female, in the mutilation of prisoners, in the humiliation and shaming of the defeated, and in the quite overt use of terror as a method of achieving one’s goals.” M. Barber, “The Albigensian Crusades: Wars Like Any Other?” in Dei gesta per Francos: Études sur les croisades dédiées à Jean Richard: Crusade Studies in Honour of Jean Richard, ed. M. Balard, B. Z. Kedar, and J. Riley-Smith (Aldershot, Hants., 2001), p. 53. Sibly and Sibly, “Appendix: Massacre of Béziers,” in Peter of les Vaux-de-Cernay, History, pp. 289–90.

16

Richard Abels

the event either without comment or with approval.53 And anyone who has taught the Crusades has probably elicited frissons from a class by reading to them Raymond of Aguilers’ horrific description of the sack of Jerusalem by the Crusaders: Some of our men (and this was more merciful) cut off the heads of their enemies; others shot them with arrows, so that they fell from the towers; others tortured them longer by casting them into the flames. Piles of heads, hands, and feet were to be seen in the streets of the city. It was necessary to pick one’s way over the bodies of men and horses. But these were small matters compared to what happened at the Temple of Solomon, a place where religious services are ordinarily chanted. What happened there? If I tell the truth, it will exceed your powers of belief. So let it suffice to say this much, at least, that in the Temple and porch of Solomon, men rode in blood up to their knees and bridle reins (in sanguine usque ad genua et ad frenos equorum).

My midshipmen are stunned when, after pausing so that they may appreciate fully the horror of the scene, I continue with Raymond’s verdict, “Indeed, it was a just and splendid judgment of God.”54 A My Lai type massacre suddenly becomes an act of Christian piety. That the victors indulged in an orgy of pillage and violence is not difficult for the students to understand. After all, the Crusaders had imbibed a “cocktail of idealism, self-righteousness, indiscipline, alienation and stress [that] created almost copybook conditions for expressions of inhumanity.”55 That a Christian chronicler, however, should have gloried in this inhumanity and even exaggerated it, is more difficult for them to comprehend.56 What my students miss is what Raymond’s intended audience would have immediately grasped, the apocalyptic biblical imagery of Revelation (14:20: “exivit sanguis de lacu usque ad frenos equorum”) and the allusions to the Book of Joshua. (This may also be true of the more “realistic” representations of the massacre by Fulcher of Chartres and the anonymous author of the Gesta

53

54 55 56

John Gillingham, Richard I (New Haven and London, 1999), pp. 169–70. Military historians, following the lead of both Christian and Muslim sources, tend to explain the massacre in terms - ad-Din Ibn Shaddad of military necessity or advantage. Although Saladin’s biographer Baha’ did not know for certain why Richard ordered the prisoners to be killed rather than enslaved, as he had (supposedly) promised Saladin, among the possible explanations he offered was that “the King of England decided to march on Ascalon and take it, and did not want to leave behind him in the city a large number (of enemy soldiers).” Francesco Gabrieli, ed. and trans., Arab Historians of the Crusades (Berkeley, 1984), p. 224. Raymond of Aguilers, Le ‘Liber’ de Raymond d’Aguilers, ed. J. H. Hill and L. L. Hill (Paris, 1969), pp. 150–1; trans. Edward Peters, ed., The First Crusade (Philadelphia, 1971), p. 214. Jonathan Riley-Smith, “The Crusading Movement,” in War, Peace and World Orders in European History, ed. Anja V. Hartmann and Beatrice Heuser (London, 2001), p. 137. John France, Victory in the East: A Military History of the First Crusade (Cambridge, 1994), pp. 355–56, downplays the severity of the massacre, arguing not only that many Jews and Muslims survived but that the sack of the city was “not far beyond [the] common practice of the day.” But cf. Benjamin Z. Kedar, “The Jerusalem Massacre of July 1099,” Crusades 3 (2004): 67–8 for a critique of France’s argument. Kedar himself concludes on the basis of the testimony of a contemporary Muslim writer, Ibn al’Arabi, that about 3,000 of Jerusalem’s 20,000–30,000 inhabitants were killed by the Crusaders. Ibid., pp. 73–4.

Cultural Representation and the Practice of War

17

Francorum.)57 This clerical discourse on war, inspired by topological biblical exegesis, followed a historiographical principle that understood historical writing as a moral endeavor that was concerned less with depicting what literally happened than with highlighting the spiritual dimensions of human actions for the moral edification of the reader.58 In this sense, the discourse was a lens through which reality was distorted into truth. The Christian chroniclers of the First Crusade saw in the massacre at Jerusalem the hand of God. We should not doubt the sincerity of their zeal, nor that of the Muslims of Syria, Palestine, and Egypt who by the middle of the twelfth century had come to recognize the Frankish conquest of Jerusalem as a religious war upon Islam demanding a Countercrusade.59 Although it is out of fashion to portray Saladin as a holy warrior lusting for Christian blood, this is precisely how he appears in the admiring biographies written about him by his household - and ‘Imad - ad-Din al-Isfahani; the same is true officers Baha- ad-Din Ibn Shaddad of the portrait of Saladin that emerges from the Kamil at-Tawarikh (The Epitome - ad-Din relates admiringly of Histories) of the more critical Ibn al-Athir. ‘Imad how two days after Hattin, Saladin sat on his dais and watched on with joy as a whole band of scholars, Sufis, ascetics, and other devout men took turns slashing away at captured Templars and Hospitallers. “How many ills did he cure by the - ad-Din. “I saw how he killed unbeills he brought upon a Templar,” exults ‘Imad lief to give life to Islam, and destroyed polytheism to build monotheism.”60 The Muslim warrior elite shared with the Franks an ethos of reciprocity,61 and 57

58

59

60 61

Fulcher of Chartres, Historia Hierosolymitana, §13, ed. H. Hagenmeyer (Heidelberg, 1913), trans. in Peters, First Crusade, p. 77. Cf. Gesta Francorum et aliorum Hierosolimitanorum, ed. K. Mynors, trans. Rosalind Hill (London, 1962), p. 92, trans. in Peters, First Crusade, p. 209. Cf. Letter of Godfrey, Raymond, and Daimbert to Pope Paschal II, in Peters, First Crusade, p. 236: “in Solomon’s Porch and in his temple our men rode in the blood of Saracens up to the knees of their horses.” William of Tyre and Albert of Aachen are exceptional among the Latin chroniclers in expressing compassion for the victims of the slaughter. For a full discussion of the historiography of the massacre, see Benjamin Z. Kedar, “Jerusalem Massacre,” pp. 15–75. Kedar’s tabulation of the details of the massacre in the various Latin sources (pp. 28–9) is particularly valuable. See, e.g., Roger D. Ray, “Bede’s Vera Lex Historiæ,” Speculum 55 (1980): 19–20. Kaspar Elm argues that the Latin chroniclers adopted Biblical images and prototypes to emphasize (and exaggerate) the bloodletting in 1099 so as to portray the massacre as a cleansing of the Holy City in the blood of the Muslims and Jews who had polluted it. “Die Eroberung Jerusalems im Jahre 1099. Ihre Darstellung, Beurteilung und Deutung in den Quellen zur Geschichte des Ersten Kreuzzugs,” in Jerusalem im Hoch- und Spätmittelalter: Konflikte und Konfliktbewältigung – Vorstellungen und Vergegenwärtigungen, D. Bauer, K. Herbers, and N. Jaspert (eds.) (Frankfurt, 2001), pp. 42–6. But cf. the response of Kedar, “Jerusalem Massacre,” pp. 70–2. Hadia Dajani-Shakeel, “Some Medieval Accounts of Salah al-Din’s Recovery of Jerusalem (Al-Quds),” in Hisham Nashabe (ed.) Studia Palaestina: Studies in honour of Constantine K. Zurayk, Institute for Palestine Studies (Beirut, 1988). Accessed online at http:// www.fordham.edu/halsall/med/salahdin.html Gabrieli, Arab Historian, pp. 138–9. - thought it plausible that Richard ordered the massa- ad-Din Ibn Shaddad Interestingly, Baha’ cre of the prisoners at Acre as a reprisal for Christian prisoners previously killed by Saladin. Gabrieli, Arab Historians, p. 224.

18

Richard Abels

the two together, religious zeal and the drive to avenge injury, brought an enormous pressure to bear upon Saladin in 1187 to take Jerusalem by storm rather than accept surrender, to deal with the Franks just as the Franks had dealt with the population of Jerusalem when they had taken it almost a century earlier, - ad-Din assures “with murder and enslavement and other savageries!”62 ‘Imad his readers that in response to Balian of Ibelin’s plea to spare the inhabitants of Jerusalem, Saladin responded, “Neither amnesty nor mercy for you! Our only desire is to inflict perpetual subjection upon you. … We shall kill and capture you wholesale, spill men’s blood and reduce the poor and women to slavery.”63 - ad-Din explain that the only reason that Saladin Both Ibn al-Athir and ‘Imad failed to carry through on his threats is because Balian countered with his own. As recounted by Ibn al-Athir, Balian, despairing of obtaining the sultan’s mercy, declared: Know, O Sultan, that there are very many of us in this city, God alone knows how many. At the moment we are fighting half-heartedly in the hope of saving our lives, hoping to be spared by you as you have spared others; this is because of our horror of death and our love of life. But if we see that death is inevitable, then by God we shall kill our children and our wives, burn our possessions, so as not to leave you with a dinar or a drachma or a single man or woman to enslave. When this is done, we shall pull down the Sanctuary of the Rock and the Masjid al-Aqsa and the other sacred places, slaughtering the Muslim prisoners we hold – 5,000 of them – and killing every horse and animal we possess. Then we shall come out to fight you like men fighting for their lives, when each man, before he falls dead, kills his equals; we shall die with honour, or win a noble victory!64

Faced with the spectre of the destruction of Islam’s shrines and the slaughter of thousands of Muslims, Saladin called a council of his advisers. “All of them were in favour,” Ibn al-Athir writes, “of granting the assurances requested by the Franks, without forcing them to take extreme measures whose outcome could not be foreseen.”65 Saladin saw the wisdom of this counsel and began negotia- ad-Din tells the same story, tions with Balian for the Christians’ surrender. ‘Imad 66 although in his inimitable inflated style. This apparently was the “official version” of the surrender emanating from Saladin’s camp, as it was the account - and other digincorporated into the letters the sultan sent to the Caliph al-Nasir 67 nitaries reporting the sultan’s recovery of Jerusalem.

62 63 64 65 66 67

Gabrieli, Arab Historians, p. 141. Gabrieli, Arab Historians, p. 156. Gabrieli, Arab Historians, pp. 142, 156–7. Gabrieli, Arab Historians, p. 142. Gabrieli, Arab Historians, pp. 156–8. - ad-Din al-Isfahani, Conquête de la Syrie et de la Palestine par Saladin, trans. Henri Imad Massé (Paris, 1972), 97–8; S. J. Allen and Emilie Amt, eds., The Crusades (Toronto, 2003), no. - ad-Din reports that he wrote seventy such letters. See Malcolm C. Lyons 42 (pp. 162–3). ‘Imad and D. E. P. Jackson, Saladin, The Politics of the Holy War (Cambridge, 1982), p. 276.

Cultural Representation and the Practice of War

19

When we turn to the main contemporary Christian accounts of this event, the anonymous Libellus de expugnatione terrae sanctae per Saladinum and the Chronique d’Ernoul, a far different picture of Saladin and the negotiations emerges. The latter source, apparently the work of Balian’s squire, represents the Ibelin version of events.68 Oddly enough, it portrays Saladin as courteous and generous, and Balian as far less bellicose and heroic. Ernoul’s Saladin is so loathe to take Jerusalem by force that he summons to Ascalon a delegation of burgesses from the city in hopes of negotiating a surrender. “I firmly believe,” Ernoul has Saladin declare to the burgesses, “that Jerusalem is the house of God, both in your religion and in mine. I shall not lay siege to the house of God nor shall I have siege engines bombard it if I can have it by peaceful agreement.”69 He bluntly tells the emissaries that the city will be his, one way or another. But rather than threaten them with the dire consequences of resistance, Saladin offers instead a deal. He will give the Christians of Jerusalem 30,000 bezants to strengthen the city’s defenses; guarantee them an area five miles in any direction of the city in which they could freely move around; and arrange for sufficient foodstuff for the city’s population to last until Pentecost, seven and a half months away. If, however, no Christian army had come to their relief by then, they will pack up their belongings and leave the city; he even guarantees them safe escort along with their possessions to Christian-held territory.70 On its surface, Saladin’s proposal is remarkably generous, especially given the imbalance in power between his forces and what remained of the Latin East. But that imbalance was precisely what Saladin wished to emphasize. The sotto voce message was that the burgesses could expect no help. When the burgesses reject the deal, Saladin swears an oath that he will never accept a negotiated surrender but would take Jerusalem by force. And yet, despite this oath, Ernoul’s Saladin remains flexible and reasonable. Indeed, he repeats the same offer – and threat – to the people of Jerusalem as he stands before the city preparing the assault. Once again the Christians spurn his peace overtures, and once again Saladin swears to take the city by force. The Christians fight bravely on the plains outside the walls, and even force Saladin to relocate his efforts from the western to the northern side of the city. But after

68

69

70

On Ernoul and his chronicle, see Margaret R. Morgan, The Chronicle of Ernoul and the Continuators of William of Tyre (Oxford, 1973), pp. 41–6. The fullest French treatment of the events of 1187 is given in the so-called Lyon Eracles, which contains material not found in surviving manuscript of the Chronique d’Ernoul but which may be closer to Ernoul’s original. See Peter W. Edbury, ed. and trans., The Conquest of Jerusalem and the Third Crusade: Sources in Translation (Aldershot, 1998), pp. 4–7. I have followed Edbury’s translation where the Chronique d’Ernoul and Lyon Heracles agree. Chronique d’Ernoul et de Bernard le Trésorier, ed. L. de Mas-Latrie (Paris, 1871), chap. 16, p. 186: “Je croi bien que Jherusalem est maison Dame Diu, et c’est vo creance; ne je volentiers ne meteroie siege en le maison Dame Diu, ne ne feroie assalir. Se je le pooie avoir par pais et par amour, je vous dirai que je vous ferai;” trans. Edbury, “The Old French Continuation of William of Tyre, 1184–1187,” chap. 49, in Conquest of Jerusalem, pp. 54–5. Chronique d’Ernoul, chap. 16, p. 186, trans. Edbury, ed., Conquest of Jerusalem, p. 55.

20

Richard Abels

more than a week of steady assaults, it becomes clear to Muslim and Christian alike that the city would soon fall. The brave defenders, knights and burgesses alike, argue for a suicidal night attack on the Saracen army, “for they would rather die honourably in battle than be captured shamefully in the city,” but the patriarch persuades them instead to seek terms from Saladin, pointing out that a heroic sally would leave the many women and children in the city to a fate worse than death: “If we are dead, the Saracens will take the women and children. They will not kill them but will make them renounce the faith of Jesus Christ, and they will be lost to God.”71 Balian is sent to make whatever peace he can with Saladin. Saladin, confident that the city is on the verge of falling and conscious of the oath he had sworn, refuses to listen. However, the city does not fall; the Christians manage to clear the walls and ramparts of Saracens. Saladin, disappointed and mortified, sends Balian back. The next morning Balian returns to plead the Christians’ case, again humbly appealing to Saladin’s piety and humanity. “For the sake of God,” Balian implores Saladin, “have mercy on them.” Touched by Balian’s entreaty, the sultan finally gives way: “Sir, out of love for God and for you [Sire, pour l’amor de Diu et de vous] I shall tell you what I will do. I shall have mercy on them in a manner that will save my oath.”72 What this entails is a disguised negotiation, during which Saladin demonstrates extraordinary compassion and courtesy by allowing the Christians to purchase their freedom on reasonable terms, even allowing Balian to bargain him down on the price set for women, children, and the poor. Indeed, Saladin and his brother prove themselves far more charitable than the Templars and the Hospitallers, whom we are told “did not give as much as they should” for the poor. When it becomes clear that despite Balian’s and the patriarch’s best efforts, forty thousand poor people still remain unransomed, Saladin’s brother Saif al-Din makes a noble gesture, asking of his brother a gift of a thousand slaves, whom he then frees “for the sake of God.” Not to be outdone in largesse, Saladin makes gifts of thousands more to Balian, the patriarch, and other Christian notables.73 The anonymous author of the Libellus, a soldier who had fought in defense of Jerusalem, was less impressed with Saladin’s generosity. Unlike Ernoul and ‘Imâd ad-Din, the author of the Libellus was not privy to the details of the peace negotiations; he could only report the “word in the street.” He had heard that Saladin’s initial rejoinder to the offer of surrender was the curt reply that Jerusalem could only be washed clean with Christian blood, but knew nothing of Balian’s supposed threat to destroy the Muslim shrines; he, like Ernoul, represents the surrender as having been accomplished largely through peaceful negotiations and - ad-Din’s appeals to Saladin’s magnanimity. He does confirm, however, ‘Imad 71 72 73

Chronique d’Ernoul, chap. 18, pp. 214–15; Edbury, “Old French Continuation,” chap. 53, in Conquest of Jerusalem, p. 58 Chronique d’Ernoul, chap. 18, p. 217; trans. Edbury, “Old French Continuation,” chap. 54, in Conquest of Jerusalem, p. 59 Chronique d’Ernoul, chap. 19, pp. 227–8; trans. Edbury, “Old French Continuation,” chap. 57, in Conquest of Jerusalem, pp. 62–3.

Cultural Representation and the Practice of War

21

account on one detail. The anonymous author had heard that the negotiators told Saladin that they would comply with his terms for surrender if possible, but if not, would hold out to the death (sin autem, ad interitum sui remaneret).74 The Muslim and Christian narratives of the siege and surrender share the same basic outline. They agree that Jerusalem lacked the manpower and resources to resist a determined siege; that the city’s inhabitants fought fiercely and delayed the inevitable; and that Balian negotiated a surrender which involved the payment of ransom for the city’s Christian population. Nonetheless, the differences among these eyewitness accounts are significant and difficult to reconcile.75 Unlike - ad-Din’s Balian, Ernoul’s hero neither threatens Saladin Ibn al-Athir’s and ‘Imad with a battle to the death,76 nor with mass slaughter of Muslim prisoners and desecration of Islam’s holy places. The tone of Ernoul’s account and his representation of the personalities of Balian and Saladin, moreover, could not be more different. Ibn al-Athir’s proud and threatening Balian contrasts markedly with Ernoul’s humble supplicant, begging for mercy from Saladin. Similarly, ‘Imad ad-Din’s Saladin, the Holy Warrior unyielding in his intention to cleanse the streets of Jerusalem with Christian blood, becomes in Ernoul’s account a wise, courteous, and compassionate prince, nobler than most of the Latin leaders who appear in the Chronique. A modern reader’s expectations are turned topsy-turvy by the Muslim and Christian eyewitness accounts of these events.77 One can only speculate why 74

75

76

77

Libellus de expugnatione terrae sanctae per Saladinum. Rerum britannicarum medii aevi scriptores 66, ed. J. Stevenson, London 1875. Versio digitalis: Brian Smith (Hong Kong): http:/ /www.fh-augsburg.de/~harsch/Chronologia/Lspost12/Libellus/lib_expu.html: “Interim miserunt legatos ad regem Syriae, supplicantes quatenus indignationem animi sui circa eos tempareret, atque sicut ceteras gentes eos etiam haberet foederatos. At ille renuente, tale fertur dedisse responsum: ‘Ego vero a sapientibus nostris Alphachinis frequenter audivi Ierusalem non posse mundari nisi sanguine Christianorum lavetur, atque super hoc eorum volo habere consilium.’ Et sic incerti redierunt. Miserunt et alios, Balisanum et Rainerium Neapolenses et Thomam Patricium, centum millia bisantinorum offerentes; nec voluit eos recipere, et spe frustrati reversi sunt. Remiserunt itaque eos iterum cum aliis flagitantes quatenus ipse Saladinus conventionem quam vellet diceret; et si fieri posset, fieret; sin autem, ad interitum sui remaneret.” Historians have tended to “reconcile” the differences by ignoring them and creating from the Muslim and Christian accounts a single, composite narrative. This is true even for studies of the source evidence for the surrender. See, e.g., Hadia Dajani-Shakeel, “Some Medieval Accounts of Salah al-Din’s Recovery of Jerusalem (Al-Quds),” via http://www.fordham.edu/halsall/ med/salahdin.html The Lyon Eracles (in a passage that is not in Ernoul’s chronicle) has Balian preface his last appeal to Saladin with an implied threat: “Sire, have mercy for the sake of God. For the people of the city are despairing of their lives and would rather kill each other than be taken by force. There will be great loss of life on both sides before you can take the city by force as you intend.” The Eracles’ Balian here implies what Ibn Athîr’s Balian explicitly states: the Franks will kill their wives and children before allowing them to be enslaved, and, if given no alternative, will sell their lives dearly in battle. The Eracles’ Balian, however, reports this less as a threat than as evidence of the desperation of the defenders. Whether Ibn al-Athîr was in fact an eyewitness to the surrender is questionable. All we can say for certain is that he had joined Saladin’s army by the summer of 1188. Lyons and Jackson, Saladin, p. 287.

22

Richard Abels

Balian should appear more bellicose and heroic in the Muslim sources and Saladin more compassionate and reasonable in Ernoul’s chronicle. However, the authors’ purposes in writing their narratives and the very different cultural imperatives facing Saladin and Balian in the aftermath of the fall of Jerusalem may provide some answers. If, as John Gillingham contends, Ernoul’s contribution to the chronicle that bears his name extends only through 1187,78 Ernoul may have been writing within months of the surrender of the city, perhaps during or shortly after Saladin’s unsuccessful siege of Tyre. Ernoul’s presentation of Saladin served as timely reassurance that Saladin was a reasonable and even generous opponent with whom the leaders of the remaining Latin principalities could negotiate. Ernoul’s readers could hardly have missed the not so hidden message that Balian had influence and favor with the conqueror. Saladin, after all, had spared the inhabitants of Jerusalem “out of love for God and for Balian,” a love that he had demonstrated by allowing Balian’s wife and children safe passage to the coast, even while Balian was holding the city against him in violation of a sacred oath that he would stay in the city for no longer than one day and one night.79 The ethos of reciprocity embedded in the value-system of the late twelfth-century French aristocracy demanded not only that injuries be avenged but that acts of good will and friendship be answered in kind. Ernoul’s representation of Saladin as a paragon of piety and chivalry may be seen in this context as partial repayment for the sultan’s exceptional generosity to an enemy. Ernoul wrote with the intent of justifying and celebrating the actions of Balian. His narrative, while that of an eyewitness to the events described, is often tendentious. Neverthless, my suspicion is that for the events leading up to the surrender of Jerusalem the Chronique d’Ernoul is closer to the truth than the Muslim sources. ‘Imâd ad-Din, secretary and chancellor to Saladin, was no less committed to justifying his master’s actions and enhancing his reputation. The recapture of Jerusalem and its holy shrines had been, without doubt, an extraordinary coup, but how he had achieved his goal was problematic. Saladin had over his career assiduously cultivated a reputation for devotion to jihad and justified his campaign to unite all the Muslim territories of Egypt, northern Mesopotamia, Syria, and Palestine under his authority as necessary for the Countercrusade that would follow.80 With Jerusalem clearly incapable of defending itself against a sustained siege, the expectation among Saladin’s followers was that the sultan would fulfill his oath to take the city by force and wash it clean with the blood of

78

79 80

John Gillingham, “Roger of Howden on Crusade” (1982), reprinted in John Gillingham, Richard Coeur de Lion: Kingship, Chivalry and War in the Twelfth Century (London, 1994), pp. 147–8, n. 33. Balian’s virtual disappearance from the narrative after 1187 strongly implies a change in authorship or at least patronage. Chronique d’Ernoul, chap. 15, pp. 186–7; Edbury, “Old French Continuation,” chaps. 45, 49, in Conquest of Jerusalem, pp. 49, 55. See Lyons and Jackson, Saladin, pp. 155–220, 275–6.

Cultural Representation and the Practice of War

23

Christians.81 By taking Jerusalem, instead, peacefully through treaty, he risked the same sort of condemnation that Frederick II was to suffer forty-two years later when he recovered the city through shameful negotiations rather than honorable violence. I suspect that the Muslim versions of the fall of Jerusalem were written with this in mind; they offered their intended audience, other Muslim leaders and Saladin’s own followers, an explanation that they could understand and accept for the sultan’s failure to repay the Franks in kind.82 The aristocratic medieval discourse on war also, at times, shaped the representation of events to accord better with what the authors thought ought to have been the case. Honor and shame are central not only to the aristocratic discourse on war as revealed in the vernacular romances and chansons de geste of the twelfth, thirteenth, and fourteenth centuries but to the less “romantic” literature of the aristocratic elite, chronicles, memoirs, and “lives.” Given their culture, it is not surprising that medieval authors were more concerned with matters of honor when describing military actions than modern historians interested in the “practical realities” of war. Nonetheless, I think we need to take their testimony seriously, if not uncritically, about the motivations and thinking of military leaders and the behavior of men in the field. This is particularly true in the case of Joinville. There are few medieval accounts of warfare more vivid and realistic than Joinville’s narrative of St Louis’s Crusade. These chapters of the Life of St Louis are exceptional in their immediacy and detail, suggesting perhaps that they were based upon a personal memoir that Jean de Joinville wrote soon after the events described. Joinville devotes far greater attention in these chapters to his own experiences and those of his companions than he does to the actions of his putative subject, St Louis. What particularly sets his narrative apart from other medieval descriptions of war and combat is his apparent frankness and lack of vanity. He tells us of the terror of being showered with Greek Fire, and of the limits of conventional piety when facing death. Without apology, Joinville relates how he and all his companions rejected out of hand the advice of his cellarer that they allow themselves to be slain and so achieve paradise rather than surrender to the 81

82

Libellus; Gabrieli, Arab Historians, pp. 142, 156–7. Cf. Lyon Eracles, chap. 53, in which Saladin explains to Balian his reluctance to accept a peaceful surrender: “In any case the fakirs and the hadjiis and the other Muslim men of religion are pressing and urging me not to give you any terms but to take revenge for those Muslims whom Godfrey slew by shedding the blood of those who were in the streets and even at the Temple in Jerusalem.” - who entered Saladin’s service in In this context it is interesting that Baha- al-Din Ibn Shaddad, 1188 and wrote his biography of Saladin between 1198 and 1216, does not mention any threats made by the Franks. His account, in fact, is closer to Ernoul’s than to the other Muslim sources. He reports that the defenders of the city, realizing their position to be indefensible and death inevitable, humbled themselves before Saladin and sought terms from him for their surrender. - position in Saladin’s court, it is likely that he would have been familiar Given Ibn Shaddad’s with the official version of the taking of Jerusalem. His departure from it may therefore be significant, reflecting perhaps changed political circumstances that obviated the need to jus- The Rare and Excellent History of Saladin, tify Saladin’s restraint. See Baha- al-Din Ibn Shaddad, trans. D. S. Richards (Aldershot, 2001), pp. 77–8.

24

Richard Abels

Saracens.83 He also recalls how he and his fellows, certain that they were about to be executed, confessed to one another. When it was his turn, he was unable to recall a single sin and when he heard the confession of a companion, he gave him absolution “as far as God gives me power,” but couldn’t remember a thing that the man had said.84 Joinville’s narrative of the Battle of Mansourah and its aftermath is unrivalled in its depiction of the chaos, confusion, and terror of medieval warfare. He recalls how at the beginning of the battle he and his knights found themselves in a precarious situation. On the advice of one of his knights, Erard de Sivery, Joinville and his retinue retreated to the apparent refuge of a “tumble-down house.” Soon they found themselves surrounded by Turks. As he tells it: There the Turks attacked us on all sides. Part of them got into the ruins, and thrust at us with their spears from above. Then my knights desired me to take hold of their horses’ bridles, which I did, to prevent the horses from stampeding; and they warded off the Turks so vigorously, that they were praised by all the champions of the army, both by those who saw the deed, and those who only heard it told. … There Lord Hugh of Scots was wounded with three spear-wounds in his face, and Lord Ralph too; and Lord Frederick of Loupey was wounded with a spear between his shoulders, and the gash was so wide, that the blood spurted out of his body as through the tap of a cask. Lord Erard of Siverey got such a sword-cut across his face that his nose hung down onto his lip. Then I bethought me of Our Lord Saint James: “Fair Lord Saint James,” I prayed, “Help and save me in this need!” No sooner had I made my prayer, than Lord Erard said to me: “Sir, if you thought it would be no reproach to me and my heirs, I would go and fetch you help from the Count of Anjou, whom I see yonder in the fields.” And I said to him: “Sir Erard, I think it would be greatly to your honour, if you were to fetch us aid to save our lives, for truly your own life is in danger.” (And indeed I spoke the truth, for he died of that wound.) He asked the opinion of all my knights who were there, and they took the same view as I did; and thereupon he asked me to let go his horse whom I was holding by the bridle along with the rest, and I did so. He came to the Count of Anjou, and begged him to come to the assistance of me and my knights. … The Count of Anjou said he should do what my knight asked him; and he turned rein to come and help us, and several of his serjeants spurred on ahead; and when the Saracens saw them coming they let us be.85

83

84

85

The Memoirs of the Lord of Joinville, trans. Ethel Wedgewood (New York, 1906), p. 158 (University of Virginia Library Electronic Text Center, http://etext.virginia.edu/etcbin/toccernew2?id=WedLord.sgm&images=images/modeng&data=/texts/english/modeng/parsed&tag =public&part=14&division=div2) Joinville, trans. Ethel Wedgewood, pp. 176–7 (http://etext.virginia.edu/etcbin/toccer-new2?id =WedLord.sgm&images=images/modeng&data=/texts/english/modeng/parsed&tag= public&part=15&division=div2) Joinville, trans. Ethel Wedgewood , pp. 104–07, (http://etext.virginia.edu/etcbin/toccer-new2?id= WedLord.sgm&images=images/modeng&data=/texts/english/modeng/parsed&tag= public&part=10&division=div2)

Cultural Representation and the Practice of War

25

As Richard Kaeuper observes, Joinville’s narrative “rushes us into the vortex of the world of chivalry: we see bloody hand-to-hand combat, and hear serious talk of honour.”86 Kaeuper’s interest in the story is in how it illustrates the close linkage between prowess and honor in the minds of knights. What intrigues me most is the image of Erard of Sivery, his nose dangling over his lips from a wound from which he would die, refusing to ride off to seek necessary help without first asking for and obtaining assurance, not only from his lord, Joinville, but from everyone present that to do so would bring him honor and not reproach. That Joinville’s role in the tale was simply to hold the horses’ bridles, while his companions fought for their collective lives, makes the story all the more plausible. The ethos of reciprocity governed the moral economy of Joinville and his audience. They would have understood that Joinville’s purpose in relating this story was to repay with lasting honor the actions of a man who had saved his life. Here in its most distilled form is the medieval aristocratic discourse on war and honor. Erard’s decision was made not on the basis of its possible efficacy or need, but on how it would be perceived by others. Even if it meant his death and that of his companions, he would not ride out until he was certain that to do so would not bring reproach upon him and his heirs. It would be a mistake to view this as merely a romantic beau geste. A noble’s honor was something very tangible and practical in the twelfth and thirteenth centuries. “Honour,” as Gerd Althoff reminds us, “determines rank, and rank is something vitally important – not only in the Middle Ages, but here in a special way – because on it depends all possibilities of cooperation, exertion of influence and structuring in the social and political order of the Middle Ages. Honour, the position in such hierarchies, is therefore logically not something static, but is contested, defended and attacked.”87 The same point is made by the poet of Raoul de Cambrai, when he has Bernier explain to his mother that, even though his lord Raoul was “more wicked than Judas,” he “would not fail him for all the wealth of Damascus until such time as all should say [he] acted rightly.”88 Medieval writers like Joinville, Jordan Fantosme, and Froissart played a role similar to that of fourteenth- and fifteenth-century heralds as arbiters of honor. This is a role still played, to some extent, by war correspondents, in particular those embedded in military units. To appreciate this, one needs only to read Evan Wright’s Generation Kill about the experiences of a platoon of First Reconnaissance Battalion Marines that

86 87

88

Kaeuper, Chivalry and Violence, p. 129. G. Althoff, “Compositio, Wiederherstellung verletzter Ehre im Rahmen Gütlicher Konfliktbeendigung,” in K. Schreiner and G. Schwerhoff, eds., Verletze Ehre, Ehrkonflikte in Gesellschaften des Mittelalters und der Frühen Neuzeit (Cologne, 1995), pp. 63–4, quoted and translated in Jan Willem Honig, “Warfare in the Middle Ages,” in War, Peace and World Orders, ed. Hartmann and Heuser, p. 120. Raoul de Camrbai, stanza 67, trans. Jessie Crosland, rev. by Richard Abels (London, 1926; rev. 1995).

26

Richard Abels

spearheaded the American drive into Iraq in 2003.89 There is perhaps no better evidence of how a military commander wished to be remembered and what he thought worth remembering than his memoirs or dispatches. We do not have the memoirs of Richard the Lionheart, but we do have a number of his letters describing his military activities and campaigns. Richard was celebrated as a paragon of chivalric prowess even during his lifetime. John Gillingham performed an impressive feat in recasting the chivalric Richard as the paradigmatic twelfth-century general guided by a well-conceived “scientific” strategy that taught him to avoid battle and focus instead on ravaging the enemy’s territory as a prelude to laying siege to its castles. As Gillingham noted, in Richard’s twenty-five years of campaigning he fought at most two or three pitched battles.90 His real competence as a general, according to Gillingham, is revealed through his methodical planning and close attention to logistics. Gillingham’s argument is persuasive, and one is now tempted to think of Richard as much as “Richard the Quartermaster” as Coeur de Lion.91 One cannot, however, read contemporary accounts of Richard’s campaigns without also noticing that, as cautious as Richard was strategically, he was wont to take extraordinary personal risks by performing his own reconnaissance and by leading cavalry charges on the rare occasions they occurred. This is reported not only by romantic eulogists such as the poet Ambroisie or the author of the Itinerarium Peregrinorum et Gesta Regis Ricardi but by historians as sober as Roger of Howden and, apparently, by Richard himself. In 1198, during the interminable castle warfare in the Vexin that dominated the final years of his reign, Richard defeated Philip Augustus in an engagement fought on the road between the castles of Curcelles, which Richard had just taken, and Gisors, which was still in Philip’s possession. Roger of Howden not only wrote a detailed account of this “open field battle” (de proelio campestri)92 but also included a letter that Richard sent to Philip, Bishop of Durham, reporting his victory:

89

90 91 92

Evan Wright, Generation Kill (New York, 2004). Wright’s book ought to be compared with Nathaniel C. Fick’s One Bullet Away: The Making of a Marine Officer (New York, 2005). Fick commanded the Marine reconnaissance platoon in which Wright was embedded. As Fick told a roomful of midshipmen on 19 January 2006, he had been upset when he first read Wright’s account of events until he realized that Wright was merely expressing the views of the Marine enlisted, who believe that their platoon is the best in the Corps; their company is fundamentally okay, although a bit screwed up; and Battalion is trying to get them killed. On the role of heralds as arbiters of honor, see Maurice Keen, Chivalry (New Haven and London, 1984), pp. 134–42. Gillingham, “Richard I and the Science of War,” in War and Government in the Middle Ages, ed. J. Gillingham and J. C. Holt (Woodbridge, 1984), pp. 78–91. Gillingham represents both sides of Richard in his writings. See, e.g., John Gillingham, Richard I (New Haven, CT, 1999), pp. 16–17, 260–1, 299, 348, and 315–17. Gillingham does not count this engagement as one of Richard’s “two or three” pitched battles. “Richard I and the Science of War,” in Gillingham and Holt, eds., War and Government, pp. 80–1. Roger of Howden, however, labels it a “proelium” and I think an engagement in which Philip had 300 knights and a large number of horse and foot sergeants must qualify as a “battle.”

Cultural Representation and the Practice of War

27

You are to know that on the last Lord’s day before the feast of St Michael we entered the territory of the king of France, at Danju, and made an assault on Curcelles, of which we took the castle with the town. … On hearing of this, … the king of France came from Mante with three hundred knights, and with men-at-arms and townsmen, for the purpose of succouring the castle of Curcelles, as he did not believe that it was taken. On this, as soon as we learned that he was approaching, we went forth with a small number of troops, but sent the main body of our forces to line the bank of the river Epte, as we supposed that he would come upon our people on the opposite bank of the river from the side of Danju. He, however, with his forces made a descent in the direction of Gisors, on which, we put him and his people, after taking to flight, into such consternation on their way to the gate of Gisors, that the bridge broke down beneath them, and the king of France, as we have heard say, had to drink of the river, and several knights, about twenty in number, were drowned. Three knights also we unhorsed with a single lance … and have them as our prisoners. …

Then after explaining that they had captured 100 knights; 200 warhorses, of which 140 were barded, and an unknown number of horse and foot sergeants, Richard closed the letter with: Thus have we defeated the king of France at Gisors; but it is not we who have done the same, but rather God, and our right, by our means; and in so doing, we have put our life in peril, and our kingdom, contrary to the advice of all our people. These things we signify to you that you may share in our joy as to the same.93

Although Bishop Philip had until recently been a clerk of the king’s chamber and Richard’s constant companion, this letter was almost certainly intended as an “official” public pronouncement rather than private tidings, and the fact that Roger of Howden saw a copy of it would seem to confirm this.94 What then did Richard want Bishop Philip and the “political nation” to know about this campaign? Clearly, he took relish in the shame suffered by King Philip. The image of the king of France taking “a drink of the river” Epte while in headlong flight is meant to make his rival look ridiculous. That he won with only a fraction of his forces made the victory even greater. Although Richard boasts of having taken Curcelles, he devotes even more attention to the prisoners captured in battle, and at the head of these he names the three whom he himself unhorsed “with a single lance.” Finally, he wanted his audience to know that he personally risked battle, placing his life and his kingdom in jeopardy, against the advice of all his people (and undoubtedly the ghost of Vegetius), because he knew that God and the right were on his side. In this letter we hear the voice not of Gillingham’s cautious

93

94

Chronica magistri Rogeri de Houedene, ed. William Stubbs, 4 vols.., Rolls Series 51 (London, 1868–71), 4:58–9; The Annals of Roger de Hoveden, trans. Henry T. Riley, 2 vols. (London, 1853; rpt 1968), 2:429–30. Roger of Wendover states that Richard sent copies of the letter to all the prelates and barons in England. See Gillingham, Richard I, p. 316, n. 78. For Richard’s use of such circulars, see Gillingham, “Royal Newsletters and Forgeries,” La Cour Plantagenét (1154–1204): Actes du colloque tenu à Thouars du 30 Avril au 2 Mai 1999, ed. Martin Aurell (Poitiers, 2000), pp. 171–86.

28

Richard Abels

strategist and logician, but of the Richard of popular history and memory. And both versions of Richard are equally true. Finally, I would like to consider briefly how cultural considerations of honor and shame could come into play in devising strategy, and, in particular, in choosing a militarily disastrous course of action. The example I would offer is one well known to this journal’s readers and to millions more since the release of the movie “Kingdom of Heaven”: King Guy de Lusignan’s fatal decision to march out of his camp at Saffuriya into the desert in the height of summer to relieve the siege of Tiberias. This military blunder that resulted in the loss of Jerusalem and the near extinction of the Crusader principalities was foreseeable and avoidable. Raymond of Tripoli and most of Guy’s advisors urged him to stay in Saffuriya and wait Saladin out.95 Guy’s decision made no military sense, but it was eminently understandable nonetheless. As R. C. Smail pointed out, Guy really had no choice.96 He had faced Saladin once before, in 1183, when he was a brash and hated newcomer to the kingdom of Jerusalem, and had disgraced himself (at least in the eyes of some) by choosing to outmaneuver Saladin rather than offer him battle. Now he was king but no more secure politically. If everyone had agreed with Raymond’s advice, Guy could have remained in camp and waited for Saladin’s supplies to run out. But the Master of the Templars, Gerard de Ridefort, had other plans. Burning with desire to avenge what he and his Order had suffered at Cresson and hostile toward his former lord Raymond of Tripoli, Gerard urged Guy to act as a king should and answer a call for aid from a vassal. That the vassal in question was the wife of Guy’s most dangerous political rival, Raymond of Tripoli, and that Raymond himself urged caution, underscored for Guy what was really at stake: his honor. Guy’s fear of being branded a coward; his obligation to do his duty as a lord, even at the risk of disaster; Gerard de Ridefort’s desire to avenge the deaths of his knights: these cultural imperatives are what drove Guy to march the army of the Kingdom of Jerusalem into the desert on the morning of 3 July 1187. One needs to be cautious about accepting Ernoul’s account at face value. His is very much the Ibelin version of events. Hattin was such a disaster that everyone involved in it, including Ernoul’s master Balian, was undoubtedly eager to fix the blame on others, and the triad of Master Gerard, Raynald of Chatillon, both of whom died either in the battle or its aftermath, and the indecisive King Guy, the rival of Balian’s preferred candidate for king, Conrad of Montferrat, made excellent fall guys. Nonetheless, in its general outlines the account rings true. But whether factually accurate or not, it is significant that in Ernoul’s

95 96

The narrative of events are taken from “The Old French Continuation of William of Tyre,” ed. and trans. Edbury, The Conquest of Jerusalem, pp. 36–40. R. C. Smail, Crusading Warfare 1097–1193 (Cambridge, 1956), pp. 153–5, 189–95; R. C. Smail, “The predicament of Guy de Lusignan, 1183–87,” in Joshua Prawer, B. Z. Kedar, H. E, Mayer, and R. C. Smail, Outremer: Studies in the history of the Crusading Kingdom of Jerusalem presented to Joshua Prawer (Jerusalem, 1982), pp. 159–76

Cultural Representation and the Practice of War

29

chronicle Guy’s decision came down to considerations of personal honor and shame. Historians must treat their sources with skepticism, and this is particularly true for the type of narrative sources with which medieval military historians must work. But this does not mean that we ought to dismiss twelfth- and thirteenth-century narrative accounts of warfare out of hand, even those found in imaginative chivalric literature, as hopelessly romantic or unrealistically prescriptive just because we find the motivations and actions described therein militarily irrational or even ridiculous.97 Richard Kaueper justifies his use of the vast corpus of chivalric imaginative literature to recover “the living reality of chivalry” by observing, “Knights, in sum, say that they have read this literature, which itself does not distinguish genres closely; they show that they have read it by using it in their own writings, and they show by their actions that they have read it and are bringing it into their lives.”98 This begs the question whether chivalric biographies, medieval chronicles, and handbooks for knights, even those composed by knights like Geoffroi de Charny, are themselves reliable witnesses to the reality of medieval warfare. I would suggest that they are, if not in every detail, at least in what they report as the motivations and intentions of those who fought. In this my views, admittedly, are shaped by my experience working with military officers. I have interviewed Marine officers of various ranks, including a former Commandant of the Marine Corps, on whether there is such a thing as a “Marine Way of War” that distinguishes the Corps from the Army or Navy apart from differences in mission and weaponry. Unhesitatingly, all of them say yes and provide examples. Unsurprisingly, they attribute the differences to the Marine Corps’ distinctive culture. I think we need to take what they say seriously, if not uncritically.99 Delbrückian Sachkritik may be able to answer factual questions such as could a twelfth-century knight really have split an armored opponent in half with his sword, and can be useful in pointing up the material implications of figures found in medieval sources, but it is dangerous to use it as a substitute for the testimony of those who practiced war and those who wrote with them as the intended audience. To do so is akin to functionalist and materialist anthropologists arguing that all primitive wars are fought for economic and

97

98 99

See, e.g., Bernard S. Bachrach, “Some Observations on Administration and Logistics of the Siege of Nicaea,” War in History 12 (2005): 250: “The Tendenz to focus on ‘knights’ is the result of a romantic bent which gives focus to chivalry as having some deep meaning for the study of medieval military history, which it does not, and is far more consistent with the colorful fiction of Sir Walter Scott than the gritty realities that are inherent in the execution of military operations.” Richard W. Kaeuper, Chivalry and Violence, pp. 32–3. John Dower, War Without Mercy: Race and Power in the Pacific War (New York, 1986), and Craig M. Cameron, American Samurai: Myth, Imagination, and the Conduct of Battle in the First Marine Dvision, 1941–1951 (Cambridge, 1994), provide controversial and critical analyses of the impact of the culture and cultural expectations of the Marine Corps upon its approach to warfighting during World War II and the Korean War.

30

Richard Abels

demographic reasons, even if the participants think otherwise. “When one reads such vain attempts to explain primitive warfare by appeals to its ecological effects or functions,” a critic of this school observes, “one realizes that ‘function’ has frequently the covert significance of ‘What a twentieth-century materialist rationalist intellectual from Europe or America thinks is a sensible allocation of labor and resources’. When such a person encounters primitive societies, he is baffled by their indifference to his criteria of what is sensible, and therefore casts about for some hidden reason which will be the real explanation for their behavior.”100 The same is true of scientific military historians who dismiss as romantic fantasy contemporary accounts in which the eternal principles of the art of war seem to take a back seat to militarily irrational cultural motivations such as honor and shame, or who rationalize apparently irrational behaviors by finding in them hidden military logic. This is not to say that medieval warfare was “chivalric” in the romantic sense of the nineteenth century. Of course it wasn’t. The ordinary activities of war in Western Europe from the eleventh through fourteenth century consisted largely of ravaging, pillaging, and siege. That medieval soldiers ransacked peasant villages should come as no surprise; war in the Middle Ages was meant to pay for itself, and poorly paid soldiers expected to be able to supplement their wages with booty, just as aristocrats looked to profit monetarily from the chivalric restraint they showed in taking other nobles prisoner for ransom. But we must also recognize that such activities were legitimated by the dominant secular aristocratic discourse on war in the Middle Ages, and that there was nothing dishonorable about them. The author of the metrical L’Histoire de Guillaume le Maréchal saw nothing incongruous in praising his hero as “most courteous” (molt corteis) for advising King Henry II to launch a chevauchée into the territory of his enemy Philip Augustus of France, or in having the Marshal tell Henry that, if done “in force, prudently and promptly,” ravaging the enemy’s countryside would be “a better and finer deed.”101 In characteristically blunt language, William Marshal’s bellicose contemporary, the troubadour-knight Bertran de Born, unabashedly summed up his class’s excitement at the prospect of war:

100

101

J. M. G. van der Dennen, “The Origin of War: the Evolution of a Male-coalitional Reproductive Strategy,” Dissertation, University of Groningen, 1995, chap. 4 (“Biological and Ecological Theories of the Origin and Evolution of War”), accessed online at: http:// dissertations.ub.rug.nl/FILES/faculties/jur/1995/j.m.g.van.der.dennen/OW_C4.pdf L’Histoire de Guillaume le Maréchal, ed. P. Meyer, 3 vols. (Paris, 1891–1901), lines 7782– 852, quoted and translated in J. Gillingham, “War and Chivalry in the History of William the Marshal,” in Thirteenth Century England 2, ed. P. R. Cross and S. D. Lloyd (Woodbridge, 1988), p. 12. We must remember that this same poet also praised the seventy-year-old Marshal for plunging into the ranks of the French during the battle of Lincoln with the ferocity of a “ravenous lion” and the alacrity of “a sparrowhawk or an eagle,” “thinning their ranks by main force and breaking up in his path a press which was very tightly formed and crowding in on him.” Histoire Guillaume le Marechal, ed, and trans. A. J. Holden, Stewart Granger, with David Crouch, vol. 1: Text and Translation (Anglo-Norman Text Society Occasional Series, 2002), lines 16608–123, posted at http://www.deremilitari.org/resources/sources/marshal3.htm.

Cultural Representation and the Practice of War

31

Trumpet, drums, flags and pennons, standards and horses white and black – that is what we shall shortly see. And it will be a happy day; for we shall seize the usurers’ goods, … nor shall the burgess journey without fear, nor the merchant on his to France; but the man who is full of courage will be rich.102

The acceptance of war waged upon peasant noncombatants was as much a matter of culture as was the chivalric ideal that enjoined nobles to capture and ransom others of their own social class rather than kill them. That a society as hierarchical as that of France and England in the late twelfth century should have created a code of war in which the lives and welfare of aristocratic warriors were privileged over those of ordinary combatants and even non-combatants should come as no shock. Nor should it be surprising to a student of medieval religion, society, and culture that medieval commanders could engage in military actions as a matter of course that today would be deemed atrocities, or even that commanders who failed to massacre an enemy thought deserving of such treatment might be castigated by Christian clerics for a lack of moral zeal. These are simply reminders that why and how men fought in the Middle Ages were shaped by values peculiar to that culture, and I would dare say that the same could be said for the present.

102

Quoted in Marc Bloch, Feudal Society, Vol. 2: Social Classes and Political Organization, trans. L. A. Manyon (Chicago, 1961), p. 296.

2 The Brevium Exempla as a Source for Carolingian Warhorses Carroll Gillmor

During the Avar campaign in 791, an equine epidemic is reported to have destroyed ninety per cent of the horses in Charlemagne’s army, comprising Charlemagne’s own horses as well as those of his vassals, both cavalry mounts and pack animals. Fortunately, the breeding stock of the Carolingian army survived the epidemic because of their location far removed from Pannonia, the site of the epidemic. Even to begin the process of replenishing the lost animals entailed a lengthy period of eleven months, the gestation time for horses. It would take even longer – some three years – to raise and train newborn colts to develop them into warhorses.1 Two collections of Carolingian estate inventories, the Brevium Exempla ad describendas res ecclesiasticas et fiscales (BE),2 dated to c. 801,3 and the better In addition to the editors, especially Professor Clifford Rogers of the United States Military Academy, the author would like to express gratitude to Professor Richard Abels, United States Naval Academy, Professor Bernard S. Bachrach, University of Minnesota, and Professor Peter von Sivers, University of Utah for their thorough evaluations of the manuscript. Just as important, this article could not have been written without the generous assistance of the Interlibrary Loan Department of the J. Willard Marriott Library, University of Utah. 1

2

3

Carroll Gillmor, “The 791 Equine Epidemic and its Impact on Charlemagne’s Army,” Journal of Medieval Military History 3 (2005), 23–45, and especially pp. 41–2, on the interval of replenishing the supply of warhorses between the epidemic and Charlemagne’s next major campaign in 794 against the Saxons. This designation was assigned by the MGH editor of the capitularies in Capitularia Regum Francorum, ed A. Boretius, MGH, Leges, II, 2 vols (Hanover, 1893), 1:250–56l. Also, G. Baist, “Zur Interpretation der Brevium Exempla et des Capitulare de villis, “Vierteljahrschrif für Sozial-und Wirtschaftsgeschichte 12 (1914), 22–70. Alfons Dopsch, Die Wirtschaftsentwicklung der Karolingerzeit, 2 vols, 3rd ed. (Cologne, repr. 1921; 1962), 1, pp. 75–101. Klaus Verhein, “Studien zu den Quellen zum Reichsgut der Karolingerzeit: Die Brevium Exempla,” Deutsches Archiv 11 (1955), 333–92. Wolfgang Metz, “Zur Entstehung der Brevium Exempla,” Deutsches Archiv 11 (1955), 395–415, and his Das karolingische Reichsgut (Berlin, 1960), pp. 26–53. Charles-Edmond Perrin, Seigneurie rurale en France et en Allemagne du début du IXe à la fin du XII siècle, 3 vols. (Paris, 1966), 1:43–46. Adriaan Verhulst, The Carolingian Economy (Cambridge, UK, 2002), p. 127. Hans J. Hummer, Politics and Power in Early Medieval Europe, Alsace and the Frankish Realm, 600–1000 (Cambridge, UK, 2005), pp. 82–4, discussed the second part of the BE. Verhein, “Brevium Exempla,” p. 388.

The Brevium Exempla

33

known Capitulare de villis (CV), shed valuable light upon the steps taken and the difficulties of royal stewards tasked with restoring Charlemagne’s supply of warhorses. Although the CV has been used by previous scholars, most notably R. H. C. Davis,4 as evidence for Carolingian approaches to horse breeding, it is not until one reads the CV together with the BE that one can fully understand Carolingian horse breeding management. A careful reading of the CV in conjunction with the BE inventories that survive from Charlemagne’s royal estates will show that horse breeding was done as scientifically as had been the case in the Later Roman Empire and as would be in the High Middle Ages, and may reflect a continuity of some Roman practices as evidenced by the presence of manuscripts of the Roman agronomists in Carolingian libraries. The BE contains inventories of ecclesiastical and royal estates in three sections. The first segment describes possessions of an island on Staffelsee in Bavaria in the see of Augsburg; the next section consists of precarial grants from the possessions of the abbey of Wissembourg in Alsace; the third part lists the inventories of five crown estates (fisci), of which one was a vineyard, while the remaining four include information on horses. The BE and the (CV)5 were bound together in the same manuscript.6 Although the BE immediately precedes the CV in the manuscript, the BE actually was compiled afterwards as a supplement to the CV, dated c. 800.7 Several considerations suggested that the CV and part three of the BE were interrelated.8 First of all, both documents share the feature of being administered by the royal

4 5

6

7 8

R. H. C. Davis, “The Warhorses of the Normans,” Anglo-Norman Studies 10 (1988), pp. 71–2, repr. John France, ed. Medieval Warfare 1000–1300 (Ashgate 2006). MGH, Capit. 1:84, c. 13–15, and Dopsch, Die Wirtschaftsentwicklung der Karolingerzeit, 1, pp. 28–74; Verhein, “Studien zu den Quellen zum Reichsgut der Karolingerzeit: Das Capitulare de villis,” Deutsches Archiv 10 (1954), 322–94; Theodor Mayer, “Das Capitulare de villis,” Zeitschrift der Savigny-Stiftung für Rechtsgeschichte, Germanistische Abteilung 79 (1962), 1– 31. François L. Ganshof, Frankish Institutions under Charlemagne, trans. B. and M. Lyon (Providence, R.I., 1968), pp. 35–40, and notes, pp. 134–8, on both the CV and the BE. Verhein, “Capitulare de Villis,” pp. 363–4, explained that the manuscript is presently located in the library at Wolfenbüttel. The manuscript is not an original, but a copy dated by Bernhard Bischoff to c. 825. The origin of the manuscript is unknown, but various theories have been advanced, most notably that of Dopsch, Die Wirtschaftsentwicklung der Karolingerzeit, 1, pp. 60ff., and 93–8, that the CV originated in Aquitaine in 794–5. A monk of Reichenau named Tatto brought the manuscript of the CV to his monastery as part of Benedict of Aniane’s monastic reform movement, a thesis which Verhein refuted in his CV article, which was devoted primarily to disputing Dopsch’s hypotheses on the dating and provenance of the CV. As to the BE, each of these estates is described as a fisc, a part of the royal domain. Verhein, “Brevium Exempla,” pp. 375–6, offered explanations of how documents of the royal fisc were preserved in a private collection: the BE inventories of the fiscs were either copied from exemplars at the chancery at Aachen or more likely were borrowed and copied by a bishopric or monastery. Also, Metz, Karolingische Reichsgut, p. 28, and Mayer, “Das Capitulare de villis,” p. 16. Verhein, “Capitulare de villis,” pp. 379–85. Metz, Das karolingische Reichsgut, pp, 29–45, based on a comparison of plant and animal terminology.

34

Carroll Gillmor

court. The CV concerns only estates of the fisc directly administered by the royal court. Part three of the BE describes royal estates in the north-east of modern France that were also under immediate royal control. Horse breeding was not the exclusive domain of the royally supervised fisc in the Regnum Francorum; other lands in the royal fisc with horse breeding operations were held in benefice by the magnates. Another point needs to be made at the outset. These horse breeding establishments should not be designated as “stud farms” in the modern sense of estates devoted exclusively to horses.9 Charlemagne’s horse breeding operation differed from modern practice in that horse breeding took place on agriculturally diversified estates which included other species of livestock and the cultivation of various crops. The interrelationship of the CV and BE further extended to the maintenance of records. In c. 62 of the CV, each estate manager was required to submit an annual statement of possessions at the Christmas court, listing all valuable items including the number of colts and fillies.10 This requirement of CV, c. 62 presumably inspired the compilation of the BE inventories in part three.11 The inventories of the BE part three are examples of these annual asset statements.12 9

10 11

12

Davis, “Warhorses of the Normans,” p. 73. Bernard S. Bachrach, “Are They Not Like Us? The Carolingian Fisc in Military Perspective,” in Paradigms and Methods in Early Medieval Studies (The New Middle Ages) (New York, 2007), p. 127. MGH, Capit. 1:88–89, c. 62, specifying colts and fillies. R. H. C. Davis, “Domesday Book: Continental Parallels,” in Domesday Studies, papers read at the Novocentenary Conference of the Royal Historical Society and the Institute of British Geographers, Winchester 1986, ed. J. C. Holt (Woodbridge, Suffolk, 1987), p. 19, believed that two capitularies may have initiated Charlemagne’s vast survey of the lands and assets comprising his realm. With reference to MGH, Capit. 1, 136, c. 4, (Capitulare de causis diversis) Davis explained that Charlemagne in 807 “ordered the missi to send him inventories of the benefices and allods of the churches in every county.” In MGH, Capit. 1, 177, c. 5 (Capitulare de iustitiis faciendis) Charlemagne in 811–13 “ordered the missi diligently to hold inquests and describe (diligenter inquirere et describere faciant) what everyone had in the way of a benefice and how many tenants (homines casatos) were in it …” The 807 capitulary is nearer the mark as a precedent for the BE in its specific request for inventories of the material assets contained on these lands. The 811–13 capitulary emphasized the collection of data on land holdings and tenants, not the material assets of these estates. CV c. 62 offers a third possibility as the inspiration for Charlemagne’s survey of estates as a more specific directive than the 807 capitulary. Both CV c. 62 and the 807 capitulary issue a directive for the preparation of these annual income statements, but CV, c. 62 went into greater depth in providing a list of the items to be included. In placing the CV to c. 800, Verhein, “Capitulare de villis,” p. 380, eliminated the 811–13 capitulary as evidence for dating the CV against the argument that this capitulary was the first evidence of Charlemagne’s interest in obtaining an accurate knowledge of his possessions. In citing CV c. 44 (MGH, Capit. 1:87), Verhein pointed out that Charlemagne was reinstating an old obligation of officials that had lapsed. Metz, “Die Entstehung der Brevium Exempla,” pp. 401–3, and Das karolingische Reichsgut, p. 33, derived the format of BE part three from a comparison of the list of animals from the fisc of Annappes with the list of animals in the Hermeneumata Vaticana, a third century Greek– Latin glossary, and the Cassel glossary which was ultimately derived from the Hermeneumata Vaticana via a late eighth-century Old High German inventory of animals from Fulda. This conclusion would not exclude the possibility that the BE inventories were used to comply with the directive of CV c. 62.

The Brevium Exempla

35

The CV set forth directives for the maintenance of the crown estates, while the inventories of four of the five fisci of the BE (as the fifth was a vineyard) reflect the practical application of these standards with enumerated descriptions of each estate’s assets. A significant portion of estate resources was reserved for provisioning the army, as in CV, c. 30 on carts and produce for the army, c. 64 on specially constructed waterproof carts to transport perishable supplies such as flour and grain as well as weapons, and c. 68 on barrels bound with iron for the army or palace.13 These military provisions of the CV were geared towards logistical preparations for campaigns.14 The CV was also concerned with ensuring the supply of horses. In c. 50 of the CV, the steward of each estate was required to determine the number of colts in each stable and the number of grooms the colts would need.15 While this provision does not explicitly connect these colts with future military use, the above measures provide the context of producing war materiel. This provision is specifically concerned exclusively with colts, the source of potential warhorses. These estates were largely preoccupied with the logistics of war material. As part of that military focus, horse breeding on these royal estates was geared to the production of warhorses. The equine segments of the four crown estates in part three of the BE, a microcosmic view of Carolingian horse breeding operations on royal estates, will elucidate how Charlemagne restored the supply of his cavalry horses after the epidemic of 791. The BE provides a wealth of information that has largely been overlooked in discussions about Carolingian horse breeding.16 While these studies have focused exclusively on the equine content of the CV, the approach here will extract the data from the BE on horses to integrate with the content of the CV and enhance our knowledge of Carolingian horse breeding. These sources share the features of chronological proximity and thematic continuity as legal documents on estate management, so that ample justification exists to assign them equal interpretative weight. The CV set down the general guidelines for horse breeding and the BE reflected the application of these principles with the categories of breeding stock and the numbers of each. Taken together, the CV and the BE form a better picture of Carolingian horse breeding than what could be achieved from exclusive reliance on either source. The following chart is derived from data in the BE on the four estates that formed an administrative unit17 located in present

13 14 15 16 17

MGH, Capit. 1:85, c. 30, 1:89, c. 64, and 1:89, c. 68. Bachrach, “Are They Not Like Us?” pp. 124–6. MGH, Capit. 1:87–8, c. 50. Davis, “Warhorses of the Normans,” pp. 71–2. Ann Hyland, The Medieval Warhorse, from Byzantium to the Crusades (Conshohocken, PA, 1994), pp. 62–3. MGH, Capit. 1:128: “This is the total of the above villas,” clearly signifies that the estates formed an administrative unit.

Carroll Gillmor

36

day north-eastern France,18 and lists the categories of horses according to gender with their ages and numbers.

Mares

3-year-olds

51

5 fillies

79

24 fillies

2-year-olds

yearlings

Annappes19 7 7 10 colts 8 fisc II20 12 13 6 colts 12

stallions

3 2 asses

4 stallions and mules 2 asses

fisc III21 44

10 fillies

12 7 colts

15 2 stallions and mules

fisc IV22 “so many”

18

19 20 21 22

“so many” “so many”

10 10

11 5

2 stallions

Of the four estates containing horses only one is named: (Asnapium) Annappes. Philip Grierson, “The Identity of Unnamed fiscs in the Brevium Exempla ad discrebendas res ecclesiasticas et fiscales,” Revue belge de philologie et d’histoire, 18 (1939), 437–61, attempted to identify the three remaining estates based on the transferral of these estates from the royal fisc as a wedding gift to Eberhard of Friuli who married Gisela, daughter of Louis the Pious in 836–7. Grierson explained that the identities of the other three were supplied by the will of Eberhard of Friuli and three charters to the monastery of Cysoing by Gisela in Ignace de Coussemaker, Cartulaire de l’abbaye de Cysoing (Lille, 1886), nos. I–III, pp. 1–8. Eberhard’s will of 867, No. I, p. 1, listed the place names as Anaspio, Cisonium, Summinium and Vitriacum. Grierson, p. 442, argued for a different sequence of place names: that Vitry, Cysoing, and Somain-enOstrevant were identical with fiscs II, III, and IV of the BE. The BE designated Annappes and the three unnamed estates as fisci. Cysoing and Somain-en-Ostrevant were designated as fisci in a diploma of Gisela, No. III, pp. 7–8, while Vitry-en-Artois was called a villa publica in the same document. Verhein, “Brevium Exempla,” pp. 336–7, agreed with Grierson that Annappes and these estates formed an administrative unit, but pointed out that uncertainty arises with superimposing the place names Eberhard’s will and Gisela’s diplomas onto the unnamed BE inventories. Most importantly, the named estates in the will and diplomas do not contain sufficient information, especially on buildings, to match the property descriptions of the individual BE inventories. As a consequence, Verhein retained the numbers II–IV instead of the place names, a conclusion this study will follow. MGH, Capit. 1:87–8, c. 50. MGH, Capit. 1: 128, c. 25: “ … iumenta maiora capita LI, de anno tertio V, de preterito VII, de presenti VII; poledros bimos X, annotinos VIII; emissarios III, … asinos II, … ”. MGH, Capit. 1: 128, c. 31: “Iumenta maiora capita LXXIX, pultrellas trimas XXIV, bimas XII, annotinas XIII, peledros bimos VI, annotinos XII, emissarios vel burdones IV, … asinos II, … ” MGH, Capit. 1: 128, c. 33: “Iumenta maiora capita XXXXIV; putrellas trimas X, bimas XII, anniculos XV, poledros bimos VII, emissarios vel burdones II.” MGH, Capit. 1: 128, c. 35: “Iumenta maiora capita tantum; putrellas trimas tantas; bimas X,

The Brevium Exempla

37

The evaluation of the issues raised by the BE generally will follow the gender and age categories of the horses in these inventories, beginning with mares, younger stock, stallions and mules. Each of these headings presents its unique difficulties of interpretation, specifically the large number of mares in relation to the younger stock, the numerical data on two-year-old colts, the absence of yearling colts on fisc III, the replacement of mares, and finally stallions and mules. Definitions of terminology will be derived so far as possible from internal evidence from the two interrelated texts. Compilers wrote the inventories under royal instructions to provide accurate information about the possessions of these estates, so their use of terminology will have been precise, especially regarding gender differentiation among the age groups. This study will present a chain of argument that the horse breeding operations of the BE estates had the primary objective of producing warhorses. The translation of terminology in the inventories significantly affects the interpretation of the equine segments in the BE, especially the translation of iumentum. Loyn and Perceval translated iumenta as older horses, but the Carolingian definition of the term is mare.23 The BE and the CV reaffirm this definition. Other than the iumenta, the groups of horses in the BE are designated as three-year-old fillies, two-year-old fillies, yearling fillies, two-year-old colts, yearling colts, and stallions. As the categories of horses on these estates clearly represent a breeding population, the iumenta in the BE should be construed as mares, defined as female horses four years of age and older, as three-year-old females are designated as fillies, pultrellae.24 Moreover, the CV, c. 13, leaves no doubt as to the meaning of this term, referring to sending stallions in among the iumenta during the spring breeding season.25 The maintenance of stallions and mares on these properties offers the reasonable assumption that these younger horses (colts and fillies) were their offspring. The practice of segregating the breeding categories by age and gender may have had roots in the treatise of the Roman agronomist, Columella, who wrote in the second century AD.

23

24

25

anniculos XI, peledros de tertio anno tantos; bimos X, anniculos V, emissarios II, … ” This sequence of categories presented two groups of yearling colts. The first group of eleven yearling colts appeared in the female categories. The format of the categories will take precedence over the possibility of a scribal error, so that the anniculos XI should read anniculas XI. On the series of numbers replaced by tantus, see below nn. 58, 87. H. R. Loyn and John Percival, The Reign of Charlemagne, Documents on Carolingian Government and Administration (London, 1975), 102–5. Jan Niermeyer, Mediae Latinitatis lexicon minus (Leiden, 1976), p. 566, translated jumentum as mare, as did Charles du Fresne du Cange, Glossarium mediae et infimae Latinitatis, rev. ed. G. A. L. Henschel, 7 vols. (Paris, 1840–50), 3:923. Three-year-old females were designated as putrellae in fiscs II, III, and IV while the female three-year-olds, two-year-olds, and yearlings were not designated at Annappes. The designation of putrellae for three-year-old females on three of the estates implies that the three-yearold females at Annappes were also putrellae. MGH, Capit. 1:84, c. 13: “antequam tempus veniat ut inter iumenta debeant.”

38

Carroll Gillmor

Columella drew a comparison between ordinary and noble horses: “Stallions and mares of common stock are allowed to be pastured everywhere together, and no fixed seasons are observed for breeding. The stallions of noble stock will be put to the mares at the time of the spring equinox, … ”26 A little later on in his work Columella added, “ … but during the rest of the year valuable stallions should be kept away from the mares.”27 Assuming that Carolingian stewards were following Roman tradition, the separation of the stallions from the mares in the BE further suggests the idea that the breeding stock on these estates was limited to the best available, to the exclusion of common stock. Moreover, the Roman practice of separating stallions from mares is also found in the inventories of these estates with the difference that the Carolingian stewards extended the concept to include the remaining categories, also organized according to age and gender. At the other end of the medieval chronological spectrum, similar age and gender categories appeared in a 1293 inventory of horses at four English and five Welsh studs.28 This practice was likely to have been dictated by the age when colts and fillies reached sexual maturity, both colts and fillies as yearlings or during their second year.29 It is reasonable to infer that the horses of the BE were also pastured according to these categories, a measure that would help to prevent overgrazing, and an administrative convenience with a view to a wellorganized operation. Separation of horses by age and gender combined with the earlier internal evidence from the CV for these estates as centers for the production of warhorses advance the argument that the best horses available were used for breeding on these estates. The absence of three-year-old colts in any of the age and gender categories lends support to the idea of these fiscs as centers for the production of warhorses, their absence being explained by the initiation of their training to be warhorses at that age.30

26

27 28 29

30

Columella, Lucius Junius Moderatus, De re rustica, 6.27.3, trans. Harrison Boyd Ash, 3 vols. (Cambridge, MA, 1941–55), 2: 190–1. Columella’s work may have been accessible to the stewards of the BE and CV and their compilers. See Texts and Transmission: A Survey of the Latin Classics, ed. L. D. Reynolds and P. K. Marshall (Oxford, 1983), p. 146, for the preservation of Columella’s work in the libraries of Fulda and Corbie, and David Ganz, Corbie in the Carolingian Renaissance (Sigmaringen, 1990), p. 61. Columella, De re rustica, 6.27.8; 2: 192–3. Charles Gladitz, Horse Breeding in the Medieval World (Dublin, 1997), p. 166, table 5.2. Angus O. MacKinnon and James L. Voss, Equine Reproduction (Philadelphia, 1993), pp. 678– 80 for colts between fourteen months and two years and p. 116, for fillies aged twelve to fifteen months on their first ovulation. See also Sue M. McDonnell, “Sexual Behavior,” in The Domestic Horse: The Origins, Development and Management of its Behavior, ed. D. S. Mills and S. M. McDonnell (Cambridge, Engl., 2005), p. 118, that fillies are fertile at about age one, while colts achieve fertility as early as eight months to a year. Gladitz, Horse Breeding in the Medieval World, p. 167: “The practice of not gelding and the absence of three and four year old colts was attributed to their involvement in the breeding of destriers and to the fact that three years was considered the age at which a horse could be sent to be broken in for riding.”

The Brevium Exempla

39

A significant observation emerges from the numbers of horses in these breeding categories. The number of mares on three of the estates is well more than triple the number of younger horses in each age group, except for the incomplete data on mares on estate IV. The apparent disproportionately large number of mares as compared with the number of offspring requires an explanation. Information on live foal percentage and neonatal sex distribution ratio will assist in evaluating the data concerning the mares and their offspring on the estates of the BE. Approximately 60% of range-bred mares to 75% of pasture-bred mares will give birth to a live foal.31 Confinement to an enclosure made the mares more accessible to the stallions than the greater mobility under range conditions; this environmental consideration largely explains the difference in these percentages. Of these live foals, the neonatal sex ratio of colts to fillies is roughly 50% each.32 With 75% of mares bred producing a live foal of which 50% were fillies, the number of fillies in each age group ought to average about one-third the number of mares, but the actual figure is much lower than that in two of the three calculable estates. This phenomenon of a high number of iumenta in relation to younger horses conceivably might be explained because a substantial portion of the iumenta were geldings rather than mares, considering that the Roman definition of iumentum was simply “beast of burden.”33 This, however, is unlikely. The presence 31

32

33

The 75% figure was derived from a study of pasture breeding of ponies conducted under the direction of Sue M. McDonnell, founding head of the Havemeyer Equine Behavior Programme, University of Pennsylvania, School of Veterinary Medicine, New Bolton Center, in McDonnell, “Sexual Behavior,” p. 117. The 60% figure appears in the works of Ann Hyland, Equus: The Horse in the Roman World (New Haven, 1990), p. 31, and cited by Bachrach, “Are They Not Like Us?,” p. 128, both without supporting statistical data, but appear to have been derived from horses bred under range conditions. S. R. Spellman, W. M. Dawson and R. W. Phillips, “Some Aspects of the Fertility of Horses raised under Western Range Conditions,” Journal of Animal Science, 3 (1944), 234–5 a fifteen-year study (1926–39) involving 953 matings which resulted in 567 with 568 foals, with one set of twins, producing a fertility level of 59.6 percent. On p. 239, the authors explained that the fertility level of their study was “a little above but quite close to the average reported by most investigators.” Gladitz, Horse Breeding in the Medieval World, pp. 154, 168, was unable to find data on the sex ratio distribution of neonatal foals from his investigations of Teutonic Knights stud records and a 1293 survey of nine royal studs, four in England and five in Wales. These later medieval estates did not keep records on horses less than one year of age. No neonatal rates were reported on the English or Welsh estates because the survey of 1293 was compiled in early March, when the mares were still in foal. An earlier document of 1291 from the estate of Woodstock, dedicated to the production of warhorses, was the single exception in providing information on newborn foals, specifically coat colors, markings, and sexes of the foals. The ratio of colts to fillies was 4:1. The sex ratio of colts to fillies at one year or older could not be determined accurately on the various estates, because these individuals had been moved around from estate to estate. Speelman, Dawson, Phillips, “Some Aspects of Fertility,” p. 238, for the sex ratio of 97.6 males per 100 females in their study. On p. 239, these authors cited Joe Palmer, “Sidelights,” The Blood Horse, 87 (1942), p. 760, which “reports that for the years 1922–39 inclusive, of 67,044 Thoroughbred foals registered in England, 50.56% were males, while of 80,296 in the United States (disregarding late registrations) there were 50.78% females.” Charlton T. Lewis and Charles Short, A Latin Dictionary (Oxford, 1966), p. 1017. Oxford Latin Dictionary, ed. P. G. W. Glare (Oxford, 1982), p. 981

40

Carroll Gillmor

of geldings among the iumenta would doubtless have risked confrontational behavior with the stallions who were likely to have perceived geldings as rivals in competition for the mares. Carolingian stewards, having sound practical knowledge of horse behavior, would not have allowed geldings to be intermingled with mares in the presence of stallions, as this practice had the potential to endanger the lives of prized breeding stock. The absence of terminology for geldings (castratio, eviratio, cantherius, equus castratus) in any of the age and gender categories of the BE thus further advances the argument for these fiscs as centers for the production of warhorses.34 Moreover, the breeding and age categories of the BE were established by the compilers of these inventories, and they were acting under royal order to provide accurate lists of assets. If geldings were not included in these inventories, they were most likely not present on these estates.The definition of iumentum as mare permits us to use the numerical data of the BE estates to resolve some problems in the text. Alternatively, the high proportion of mares in relation to offspring could also reflect the practice of breeding approximately one half of the mares in alternate years. As we shall see, the Roman agronomists highly recommended biennial breeding. A discussion favoring biennial breeding should include the initial step of presenting the disadvantages of annual breeding. A mare comes into her estrous cycle approximately 5–12 days after foaling; this estrous cycle is commonly called the “foal heat.”35 A greater risk (10–20%) of embryonic death exists if the mare conceives during the foal heat.36 Offering an explanation for the decreased pregnancy rate of foal heat breedings, modern veterinarians have hypothesized that the uterus has not recovered sufficiently to grow a developing embryo.37 Moreover, approximately 50% of the new foals will develop diarrhea if their dams conceive during the foal heat,38 causing a weakening of the foal due to dehydration, a condition that may have lessened the foal’s chances of surviving its first winter. From sheer experience, Carolingian stewards would eventually have recognized the disadvantages of breeding mares during the first estrous cycle after foaling. The next estrous cycle occurs approximately 18 days after the foal heat; annual breeding on the second or later estrous cycle after foaling

34

35 36

37

38

Comparable evidence is to be found in a 1293 survey of four royal English studs and five Welsh studs for which statistical data was compiled by Gladitz, Horse Breeding in the Medieval World, p. 167: 75% of the Welsh colt yearlings were gelded, but no gelding was reported for the four English studs, an indication that these studs were engaged in the production of destriers. MacKinnon and Voss, Equine Reproduction, p. 623 MacKinnon and Voss, Equine Reproduction, p. 521, and on p. 623, “Pregnancy rates achieved by breeding during the foal heat are generally reported to be 10–20% lower than those obtained by first breeding at subsequent estrous periods.” MacKinnon and Voss, Equine Reproduction, p. 623: “The decreased pregnancy rate associated with foal heat breedings has been suggested to be caused by the failure of the uterus, particularly the endometrium, to be completely involuted and, therefore, ready to support a developing embryo.” MacKinnon and Voss, Equine Reproduction, p. 1027: “Foal heat diarrhea is a term used for diarrhea that develops in foals 6 to 10 days of age when the mare is in estrus.”

The Brevium Exempla

41

delays the birth of foals by approximately a month every year.39 Breeding mares during a later cycle poses a significant disadvantage to a breeding operation concerned with the winter survival of its foals, progressively shortening the time the foal will have to gain strength for winter. In an age largely without the benefits of veterinary medicine, the production of warhorses depended initially on the ability of these foals to survive their first winter. The disadvantages of breeding mares annually make the practice of biennial breeding worth exploring. Several considerations suggest that not all of the mares were being bred every year. The stewards of the Carolingian estates may have bred their mares biennally, a standard practice of the Roman agronomists.40 According to Columella, “It is usual for common mares to foal every year, but it is proper to restrain one that is well-bred so that she does so every other year; as the offspring receive greater strength from their mother’s milk.”41 Palladius (fl. c. AD 350)42 relied on Columella’s advice that well-bred mares should be covered only every other year, “We ought to breed well-bred mares and those who produce colts in alternate years so that they infuse strength from milk to colts to be large and strong: the others to be bred again at random.”43 Columella and Palladius could be construed to mean that with the annual breeding of mares, the new pregnancy decreased the flow of milk to the live foals, but this interpretation is unsupported by modern veterinary medicine which explains that mares continue to secrete milk “so long as milk is removed regularly from the mammary gland.”44 What the Roman agronomists actually seem to be advocating is that the biennial

39

40

41 42

43

44

MacKinnon and Voss, Equine Reproduction, p. 623: “… failure to establish pregnancy by breeding on the foal heat results in an 18 day drift toward a later conception and a corresponding delay in the foaling date the following year.” Reynolds and Marshall, Texts and Transmission (Oxford, 1983), pp. 40–2, connect the manuscript tradition of the agricultural treatises by Cato and Varro with the Renaissance. These treatises do not appear to have been known in the Carolingian era. Columella, De re rustica, 6.27.13; 2: 196–7. Translation by Gladitz, Horse Breeding in the Medieval World, p. 126. The earliest surviving manuscripts of Palladius’ work, books 1–13, dating from the ninth century, were derived from one lost copy, and had a provenance in north-eastern France. See Reynolds and Marshall, Texts and Transmission, pp. 287–8. Situated in the same general area, the crown estates of the BE would seem to represent a continuity of animal husbandry practices. Rutilius Taurus Aemilianus Palladius, Opus Agriculturae de veterinaria medicina de insitione, ed. Robert H. Rodgers (Leipzig, 1975), IV, xiii, 5, p. 137: “Generosas equas et quae masculos nutriunt, alternis annis summittere debebimus, ut pullis puri et copiosi lactis robur infundant: ceterae passim replendae.” To provide the mares bred biennally with exercise and a useful task during their year of recovery before being bred again, estate managers may have used them as post horses requisitioned for the retinues of the missi in accordance with an obligation specified in the CV (MGH, Capit. 1:85, c. 27). For a definition of paraveredus including a citation to this text, see Niermeyer, Mediae Latinitatis lexicon minus, pp. 762–3. Also Heinrich Dannenbauer, “Paraveredus Pferd,” Zeitschrift der Savigny-Stiftung für Rechtsgeschichte, Germanistische Abteilung 71 (1954), 61–62. McKinnon and Voss, Equine Reproduction, p. 591.

42

Carroll Gillmor

interval of rest and nourishment of the mare will ensure the production of more nutritious milk than if the mare had produced a foal every year. Palladius refined the criteria for breeding in alternate years. Columella had advised the practice for well-bred mares, expecting common mares to be bred every year. Palladius accepted Columella’s recommendation of breeding wellbred mares biennally, introducing the additional qualification that mares who produced colts also should be bred every other year (Generosas equas et quae masculos nutrient). The “et quae” of this passage introduces another category of mares separate from the well-bred mares. The “quae” refers to common mares who were nursing colts and includes them in the group of select mares to be bred in alternate years. The “ceterae” of the passage must refer to the other possible category, the common mares who produced fillies. These mares were to be bred again (replendae) indiscriminately (passim) which is construed as allowing these mares to breed randomly without the intervention of stewards to regulate a specific breeding season. Four categories of mares appear in the text of Palladius: well-bred mares who produced colts, well-bred mares who produced fillies, common mares foaling colts, and common mares giving birth to fillies. In Palladius’ view, all well-bred mares should be bred biennally, regardless of whether they produced colts or fillies, the rationale being that future broodmares should be generated from the best stock. Palladius departed from Columella’s advice on the annual breeding of all common mares, adding ordinary mares who produced colts in the biennial breeding program. The addition of these mares to the biennial breeding pool would not affect the overall sex distribution ratio of approximately 50% colts to 50% fillies, as these percentages are not determined by the perceived quality of the mares. From the text it is clear that Palladius advocated letting an ordinary mare who had just foaled a colt to have a year off breeding, presumably so that her body’s strength could go into milk for the colt rather than supporting a new embryo. Palladius’ advocacy of this practice had two differing yet complementary objectives having to do with quality and quantity. The inclusion of ordinary mares into a biennial breeding program would tend to make larger and stronger colts, not more of them. Logically, this concept would not favor the production of more colts since these horses might (though probably not really) have a somewhat higher than 50% chance of a colt, but perhaps a 60% chance of a colt every other year. Palladius also may have had a dual objective of producing not only larger and stronger colts, but more of them. If Palladius did not want more colts, he probably would not be advocating the introduction of common mares who tended to produce colts (who have produced colts in the past) on a biennial breeding schedule. Operating on the practical level, this practice would mean that if these common mares in the biennial program were to produce fillies, presumably these mares would have been transferred back to the annual breeding group, as ordinary mares thought to produce fillies were excluded from biennial breeding and maintained on an annual schedule. Following this line of reasoning, Palladius would have been willing to compromise on quality in order to have an opportunity to increase the number of colts. Palladius’ recommendation of including

The Brevium Exempla

43

ordinary mares who tended to produce colts into the biennial group of well-bred mares had the twofold purpose of providing these common mares with a year of recuperation from breeding so they could produce larger and stronger colts. Moreover, the addition of these mares to the biennial group of select mares also would increase the number of mares being bred, so that hopefully more colts would be foaled. Palladius’ work was available to Carolingian scholars living in the vicinity of the BE estates. As a consequence, it is worth considering the feasibility of applying Palladius’ idea of introducing common mares who tended to produce colts into the group of Charlemagne’s selectively bred mares, especially since at the end of 791 Charlemagne had to decide on a course of action to replace the warhorses lost on the Avar campaign. Charlemagne would not have succeeded in replenishing his warhorses at a faster rate by 794, the date of his next full-scale campaign,45 had he followed Palladius’ advice on the introduction of ordinary mares who tended to foal colts. The earliest such mares could have been bred after the epidemic in autumn of 791 was the spring of 792. The first foals of these ordinary mares would have arrived in the spring of 793 after an eleven-month gestation period. It would not have been until they reached age three, in 796, before their training as warhorses could have begun, two years after the 794 Saxon campaign. Had Charlemagne chosen the “quick” solution and tried to increase the number of colts by breeding less desirable stock, he would have been living with the negative results for generations of horses. These horses would have lacked the requisite athletic ability and mind, the willingness and intelligence to be good warhorses. Instead of following Palladius’ advice on the introduction of ordinary mares to a biennial breeding group, Charlemagne’s best course of action would have been to continue breeding the best stock and wait until these colts matured sufficiently for cavalry service. In the interest of preserving quality rather than increasing quantity, Palladius’ advice for the inclusion of ordinary mares who produced colts can be safely eliminated from Charlemagne’s breeding program, but the issue of whether or not mares on the estates of the BE were bred annually or in alternate years still needs to be resolved. This problem can be approached in several ways, initially by analyzing the data on the number of mares bred to each stallion. The recommendations of the Roman agronomists will be compared with the statistical data Gladitz compiled from the stud records of the Teutonic Knights and thirteenthcentury royal studs in England and Wales, and then applied to the BE data. Columella advised that a stallion should be capable of covering from fifteen to twenty (average of 17.5) mares per season.46 Palladius was more subjective in his assessment of a stallion’s strength, insisting that each stallion should be evaluated individually, but the number of mares should never exceed twelve to fifteen,47

45 46 47

Gillmor, “The 791 Equine Epidemic and its Impact on Charlemagne’s Army,” pp. 42–3. Columella, De re rustica, 6.27.9–10; 2: 194–5. Palladius, Opus Agriculturae, IV, xiii.1, pp. 135–6.

44

Carroll Gillmor

averaging 13.5 mares per stallion. The advice of Palladius resembles the data in the stud records of the Teutonic Order in Prussia. At these studs in seven districts dated 1374–1409, a stallion covered between eleven and seventeen mares per season with an average of thirteen mares per stallion.48 In a 1293 inventory of nine royal studs, four in England and five in Wales, the number of mares per stallion in England ranged from six to ten (average eight) as compared with twelve to one exceptionally fertile stallion in Wales covering fifty-four mares.49 The number of stallions per mare at Annappes and estates II and III, assuming the stewards followed the practice of biennial breeding, compares more favorably with the data from the studs of the Teutonic Knights in Prussia and the royal estates in England and Wales than the advice of the Roman agronomists. Of the fifty-one mares at Annappes, about twenty-five were bred each year, averaging eight mares per stallion. Seventy-nine mares resided at fisc II of whom around forty mares were bred each year for an average of ten mares for each of the four stallions. The estate of fisc III owned forty-four mares, averaging eleven mares per stallion from approximately twenty-two mares bred. Taking the three fiscs together and assuming that all 174 mares on the fisci of the BE were bred annually by the nine stallions of Annappes and estates II and III, the average number of mares per stallion would be nineteen. While this would, barely, fall within Columella’s acceptable range, it would substantially exceed the maximum ratio recommended by Palladius, and generally observed on the Prussian stud farms. If, however, the mares were bred only biennally (87 each year), the average number of mares per stallion would be approximately ten, closer to being in line with Palladius’ advice and very similar to the practice on the English and Welsh thirteenth-century stud farms. This seems more probable. The Roman agronomists recommended the practice of biennial breeding. Bernard S. Bachrach, whose writings generally favor arguments for continuity with the Roman past,50 did not include a discussion of breeding mares in alternate years in relation to the BE estates. No citations to the works of the Roman agronomists on this point appear in his article which relies on the work of Hyland who simply mentioned biennial breeding.51 Bachrach assumed that mares on the BE estates were bred annually, “If the horse breeding studs at the villae of the fisc each serviced an average of fifty brood mares about eighty-five per cent of the average found in the sample above [51 mares at Annappes, 79 at fisc II, and 44 on fisc III], and about sixty per cent of these produced a live and healthy foal each year, the 600 or so units of the royal fisc would have produced some 18,000

48 49

50 51

Gladitz, Horse Breeding in the Medieval World, pp. 138–9. Gladitz, Horse Breeding in the Medieval World, p. 165, and chart p. 166. These averages were based on total numbers of breeding stock of 7 stallions, 50 mares, and 45 younger horses at the four estates in England; the estates in Wales comprised 5 stallions, 157 mares and 113 younger stock. For example, Bachrach, Early Carolingian Warfare, Prelude to Empire (Philadelphia 2001). See above n. 31

The Brevium Exempla

45

horses on an annual basis for direct royal use.”52 In this computation, the 174 mares divided by the three estates averaged 58 mares per estate. Apparently, Bachrach rounded out the numbers to simplify the calculations, assuming an average of 50 mares (85% of the actual average of 58) were being bred annually. The likely occurrence of deaths and disabling pasture accidents raises doubt as to the assumption that all sixty per cent of these live foals remained healthy and survived to adulthood. In order to get a sense of the scale of Carolingian warhorse production, it would be better to work from the number of two-year-old colts on the BE estates, which averaged eight. If we multiply this by Werner’s figure of 600 fisci, we can estimate an annual production of roughly 4,800 potential warhorses. This data from the BE offers more detailed information about the number of potential cavalry forces available to Charlemagne, and can be used to advance the discussion on the size of Carolingian armies.53 Karl Ferdinand Werner used two methodological approaches to arrive at 30,000–35,000 cavalry in Charlemagne’s armies. In the first scenario Werner posited 700 fiscal units with each contributing a contingent of fifty cavalry. Alternatively, taking the figure of 1,500 direct royal vassals, each mobilizing a contingent of twenty horsemen, Werner arrived at 30,000 cavalry.54 The annual replacement rate for 30,000 cavalry can be derived from the BE data on two-year-old colts. From the four estates, the two-year- old colt production averaged eight. With eight colts annually per fiscal unit, the replacement rate for 30,000 would be 16%. Even allowing for losses of an estimated 300 two-year-old colts overall because of injury or disease, the number of three-year-old colts would then be 4,500 and the replacement rate not significantly lower at 15.5%, corresponding to an average service-life of seven years for a war horse, a very reasonable figure.55 These computations thus offer some support to the large-numbers side of the Carolingian army size dispute. However, it is unlikely, given their location, that the four BE estates examined were merely average in their horse production. Moreover, the average of 8 colts per fiscal estate is based only on the four horse breeding estates of the five BE fisci. If we factor in the vineyard, the average would be only 6.6. Considering the varying quality of the soils from region to region as well as densely forested 52

53

54

55

Bachrach, “Are They Not Like Us?,” p. 181 and n. 78. For the figure of 600 fiscal units, see Karl Ferdinand Werner, “Missus-marchio-comes: Entre l’administration centrale et l’administration locale de l’empire carolingien,” in Histoire comparée de l’administration (IVe– VIIIe siècle), ed. Werner Paravicini and Karl Ferdinand Werner, Beihefte der Francia 9 (Munich 1980), pp. 231–3. Robert-Henri and Anne-Marie Bautier, “Contribution a l’histoire du cheval au moyen age,” Bulletin philologique et historique, 1978, p. 26, also assumed the annual breeding of mares. For the current state of the question and the pertinent literature, see John France, “The Composition and the Raising of the Armies of Charlemagne,” Journal of Medieval Military History 1 (2002), 69 Werner, “Heeresorganisation und Kriegführung im deutschen Königreich des 10. und 11. Jahrhunderts,” Settimane di Studi del Centro Italiano sull’alto Medioevo 15 (Spoleto, 1968), pp. 821–2. On the service life of a warhorse, see Gillmor, “The 791 Equine Epidemic and its Impact on Charlemagne’s Army,” p. 42 and n. 77.

46

Carroll Gillmor

areas56 and insufficient rainfall and irrigation systems57 in the semi-arid lands of the Carolingian Empire, it is likely that away from the north-east of France there would be a smaller proportion of horse breeding estates, and a smaller production of horses on the estates that did maintain stud operations. Still, even after accounting for those variations, the BE and the CV data continue to support the conclusion that these fiscal estates alone would have been able to supply warhorses for a force of well over 10,000 cavalrymen. These computations based on the number of two-year-old colts, the actual potential warhorses, on the estates is preferable methodologically than information derived from the annual breeding of mares. In pursuit of additional depth of understanding of biennial breeding, the mares will be compared with the younger stock to see whether the offspring of these mares approximated the neonatal birth rate of 75% for pasture bred horses. For the foal crops of the four estates, nearly complete figures exist for the two-year-olds and the yearlings. The following computations rest on the assumption that roughly one-half of the mares on each estate were bred every year. From the fifty-one mares at Annappes, about twenty-five were bred in a given year. Of these, the seventeen two-year-olds equal 68 % of the colts and fillies that reached that age; a total of nineteen would equal 75%. The fifteen yearlings suggest that 60% of the mares bred gave birth to a healthy foal which matured to age one. In the case of fisc II, approximately forty of the seventy-nine mares were bred, producing eighteen two-year-olds. According to these numbers, only 45% of the mares bred foaled offspring which lived long enough to reach age two, while the twenty-five yearlings represent 63%. Of the forty-four mares on the fisc III, twenty-two produced nineteen two years olds or 86% healthy foals; the fifteen yearlings indicate that 68% of the mares foaled offspring that survived to age one. The percentage of mares that produced healthy yearlings and two year olds averaged 65. These computations assume the breeding of mares in alternate years and coincide approximately with the 60% average for horses bred under range conditions, but the foals on these estates resulted from pasture breeding with a live foal birth rate, as we have seen, of roughly 75%. This higher figure is more appropriate, as it represents the pasture conditions under which these foals were bred and makes allowance for losses ascribable to disease and accidents. The argument for biennial breeding can be strengthened with a computation of the percentages for foals reaching age two with the annual breeding of all 174 mares at Annappes and estates II and III, assuming the number of mares remained approximately the same. Breeding all fifty-one mares at Annappes would result in only 33% of the colts and fillies reaching age two; 22% of the

56

57

Chris Wickham, “European Forests in the Early Middle Ages: Landscape and Land Clearance,” Settimane di Studi del Centro Italiano sull’alto Medioevo 37 (Spoleto, 1990), 479–548, esp. 500–1. Paolo Squatriti, Water and Society in Early Medeival Italy, AD 400–1000 (Cambridge, Eng., 1998), pp. 79–96, that irrigation was practiced mainly for horticulture, though p. 86 n. 55 mentions irrigated pasture but without reference to species of livestock.

The Brevium Exempla

47

colts and fillies at age two from the seventy-nine mares at estate II; the forty-four mares on estate III produced nineteen two-year-olds for a total of 43%. These low percentages based on annual breeding would not be feasible for a viable livestock operation; the breeding of mares in alternate years more closely corresponds to the live foal birth rate of pasture-bred horses. Of all the approaches to biennial breeding already discussed, including the disadvantages of annual breeding, the treatises of the Roman agronomists, and the number of mares per stallion, this data on survival of offspring in comparison with the number of mares is the most convincing argument for biennial breeding. The proportion of young horses to mares at first glance seems low, considering the potential of a 75% neonatal reproduction rate. Likewise, the number of stallions would be on the low side to cover all of the mares on the estates. If, however, as we have argued, the number of mares on these estates were bred biennially, the number of young horses would match very closely with what one might predict. The number of stallions would then be somewhat on the high side relative to the prescriptions of the Roman agronomists, but in line with the practices of the medieval stud farms. Biennial breeding was in fact recommended by the Roman agronomists. Palladius, however, advised annual breeding for common mares, or at least for common mares not nursing colts, and the breeding in alternate years for the best mares Having established biennial breeding as the more workable choice for preserving Charlemagne’s horse-breeding program on the BE estates, we can now focus more narrowly on a single group of these offspring, the one most important for the production of warhorses, specifically the two-year-old colts for whom the data is most complete. The combined estates of Annappes plus fiscs II and III owned 174 mares excluding fisc IV for which the number of mares was not recorded58 The expected annual production (assuming half the mares were bred and 75% of those produced foals, divided equally between each sex) would then be 32.6 each for colts and fillies. We actually see among the two-year-olds 31 fillies (a nearly perfect match) and 23 colts. This implies that roughly 97% of the fillies born reached age two and 72% of the male foals lived (and were retained on the estate) to age two. The disparity in the number of two-year-old colts at thirty-three, as compared with forty-one fillies on all four estates, in part could be ascribable to annual variation in the production of colts, as the 50/50 ratio of colts to fillies applies to a neonatal average. As these colts were two year olds, the intervening factor of disease,59 though it affects both colts and fillies without preference for colts, may have affected randomly the number of colts on these estates. Alternatively, perhaps the relative shortfall of colts to fillies could be explained because these colts were more likely to be transferred to other estates as yearlings. This stage

58

59

Verhein, “Brevium Exempla,” pp. 348–50: In explaining the use of tantus, Verhein drew a distinction between a model manuscript and a copy. The copyist replaced illegible numbers with tanti (so many), as he was unable to read the numbers from the model he was using. McKinnon and Voss, Equine Reproduction, pp. 993–1094 for neonatal disorders.

48

Carroll Gillmor

of their lives was preoccupied with growing and playing; as we shall see, as yearlings it would have been premature to transfer them to the Christmas court to begin training as warhorses. These factors help to explain the smaller number of two year old colts, but do not explain adequately the absence of yearling colts from estate III. At first sight, this entry (anniculos XV) suggests the fifteen yearlings were colts rather than fillies.60 This ambiguity possibly originated in the manuscript with the penstroke connecting about halfway the terminal “s” with the previous letter which could be read either as “as” or “os.” Emending this “anniculos” to “anniculas” would be consistent with the steward’s previous entry, which clearly distinguished the genders in the two year old category of estate III. Moreover, and this point needs to be re-emphasized, the stewards acted under royal mandate to provide accurate inventories of estate holdings including the number of each sex. As formularies these inventories follow a definite order of categories to facilitate the recording of data: mares, three year old fillies, two year old fillies, yearling fillies; two year old colts, yearling colts, and stallions. The most complete categories are those of Annappes and fisc II. The sequence of categories for fisc III does not follow precisely the order of those at Annappes and fisc II. Fifteen yearling colts are listed in the place where yearling fillies appeared in the first two inventories. Adhering to the sequence of gender categories presented in the first two inventories provides a resolution to this problem, so that the anniculos XV should be read as anniculas XV. On the other hand, however, reading “anniculas” for “anniculos” raises the question of why there would be such a great imbalance between yearling fillies and colts on estate III. The total of fifteen yearlings is about what we would expect from the forty-four mares on this fisc. It could happen by chance that fifteen out of fifteen foals would all be female – but the odds would be essentially the same as the odds of tossing a coin fifteen times and having all fifteen turn up “heads.” Statistically, that would be one in two to the fifteenth power, or 1 in 32,768. It is therefore more likely that this particular estate manager used anniculos just to mean “yearlings” without gender differentiation, since in Latin mixed-gender groups receive the masculine plural ending.61 The textual difficulties arising from the BE tabular data have included the large number of mares in relation to the younger stock and the disparity in the number of colts to fillies. A topic of special significance for the maintenance of a horse-breeding operation, involves establishing the breeding life of broodmares to ensure their continual replacement. The Roman agronomists provided directives on the breeding life of mares. Columella advised that it was acceptable to breed a filly at age two, so that she would be three when bearing her first offspring; the mare was considered to be of no use for breeding after her tenth year, because the offspring of older mares were slow and lazy.62 Palladius followed 60 61 62

MGH, Capit. 1: 128, c. 33. Text above n. 21. The author is indebted to Professor Clifford Rogers for this statistical computation. Columella, De re rustica, 6.xxviii.1, pp. 196–9.

The Brevium Exempla

49

Columella on the breeding of fillies at age two and on the supposed negative qualities of foals of mares older than ten.63 Contrary to the advice of the Roman agronomists, the age categories of the BE suggest that female horses were first bred at age three, producing their first foal at four, as females were classified as mares from age four upwards. Modern practice generally advises against breeding a filly until she reaches age three or four. Fillies can be sexually mature as early as a year,64 but they do not achieve musco-skeletal maturity at least until reaching age three.65 At age two, fillies are still growing. If a filly is bred at that age, a portion of the nutrients that are supposed to ensure her growth will be diverted to the developing embryo. This practice would also incur the risk of stunted growth for the filly and less milk production, which would compromise the production of large and strong colts, the primary objective of Charlemagne’s warhorse breeding program. Breeding fillies at age one or two also incurs an increased potential to result in a difficult pregnancy. Once the pregnancy has come to term, the foal may be simply too large to be delivered by the immature filly, resulting in the death of either or both. The ability to bear weight largely determines the breeding age at three and likewise the age to begin the training of a colt to bear the weight of a rider. At three a colt is asked to bear weight of rider; a three-year-old filly must carry increasing weight of embryo. The weight-bearing capacity is related to the closure of the growth plates, estimates for which range from twenty-seven to forty months for appendicular growth plates and as long as 5.5 years for all 32 vertebrae of the vertebral column.66 Muscular strength, especially along the back, for mares to carry foals to term and for colts to bear the weight of a rider are not sufficiently developed until age three. The age and gender categories of the BE estates draw the distinction in the nomenclature of fillies and mares at the dividing line of ages three and four. This indicates that breeding was delayed until age three, and that is because this practice resulted in larger and stronger foals, a conclusion that would fit with the general trend we see with biennial breeding, specifically, the emphasis on producing higher quality rather than higher numbers of horses.

63 64 65 66

Palladius, Opus Agriculturae, IV.13.6, p. 137. See above n. 29 on the age of sexual maturity for fillies. James M. Giffin and Tom Gore, Horse Owner’s Veterinary Handbook (New York, 1989), p. 295. Eric Strand et al., “Radiographic closure time of appendicular growth plates in the Icelandic Horse,” Acta Veterinaria Scandinavica 49 (2007), 1–3: the growth plates in the regions of the distal radius closed at 27.4 to 32 months and the stifle from 27 to 40 months. “The Icelandic horse appears to have similar radiographic closure times for most of the growth plates of its limbs as reported for large new breeds of horses developed during the past few centuries.” The statistical sample was derived from 64 Icelandic horses, including 38 mares, 15 stallions and 11 geldings. The closure time for the vertebral growth plates is to be found in D. Bennett, “Timing and Rate of Skeletal Maturation in Horses, “Equine Studies Institute (2005) at www.equinestudies.org.

50

Carroll Gillmor

Establishing the age of the first pregnancy at age three is a necessary precondition for determining the breeding life of the mare. For our purposes, the breeding life of a mare will be computed from the conception of the first foal at three to the birth of the last foal at age ten in alternate years as recommended (for “noble” horses) by the Roman agronomists. On this eight-year breeding cycle, a mare would be bred an average of four times, and would produce three foals (at 75% production) during her breeding life. Moreover, since not all mares would live out their eight-year breeding life, the average would be somewhat lower than three .67 From the birth of the first foal in the spring of the mare’s fourth year, the mare will be bred again in the spring of her fifth year, so that she will bear her second foal in the spring of her sixth year. During the remainder of her breeding cycle, she will have her third foal in the spring of her eighth year, and her fourth and probably last foal in her tenth year, a breeding life of eight years. Mares were capable of bearing foals after age ten, but Carolingian stewards may have been influenced by the perception of the Roman agronomists that mares older than ten gave birth to slow and lazy foals. Ten-year-old mares retired from breeding were still healthy, viable horses for other uses, such as serving as pack horses for the army. On an eight-year breeding cycle, one-eighth of the mares required annual replacement. The total of 174 mares divided by eight would mean the number of mares needing replacements would be twenty-two. Excluding the yearlings (due to the uncertainty of how to interpret the anniculos of Estate III) the calculation of the average annual mare production can be based on the two- and three-yearolds only. That would produce an average of 35 (as derived from a total of 70 over two years equals 35 per year), with twenty-two to provide for replacements for mares retiring annually from the broodmare bands and a surplus of thirteen. It is doubtful that there would be much need to allow for growth of the broodmare bands, except rarely, since the estate would probably have already as many mares as the land would support. More likely, the surplus and the retiring broodmares could be used to provide the court and army with riding and pack horses, which they needed just as much as warhorses. The breeding life of mares and provision for their replacement was a vital component of Charlemagne’s horse-breeding operation. Just as important for the success of the breeding program, stewards had to possess a keen sense of how to evaluate the conformation and temperament of young colts as stallion prospects. All four estates produced far more colts than necessary to sustain the number of stallions. The number of colts, yearlings and two year olds combined, easily exceeded the number of stallions, affording stewards the luxury of being able to reserve only the finest colts as breeding stallions. A puzzling terminological difficulty in the tabular data emerges from the use of the conjunction

67

K. D. White, Roman Farming (Ithaca, NY, 1970), p. 296, estimated four to five foals during a mare’s breeding life, but he did not consider the live foal birth rate of pasture-bred horses.

The Brevium Exempla

51

“vel”68 to distinguish mature male equids (stallions) from (burdones) defined as mules (offspring of a male ass and a mare) or hinnies (offspring of a stallion and a female ass)69 on fiscs II and III. The phraseology initially appears to have overlooked a principle of animal genetics. The use of vel suggests that the compiler assumed an equivalency of the two terms, but a Carolingian estate manager with practical experience of livestock would obviously have known that a mule or hinny is not the equivalent of a stallion. A mule is a hybrid, and, as a consequence, is infertile. The presence of a breeding population on these crown estates of mares, two and three year olds, and yearlings necessarily included stallions on these estates, but does not explain the mules or hinnies. When construed to mean “and” rather than “or,” the vel of the BE reconciles the presence of stallions with mules on the two estates. The numerical designation would then denote four stallions at fisc II and two stallions at fisc III with an indeterminate number of mules or hinnies. Burdones are unlikely to have been bred on these warhorse-producing estates, especially as the terms for the younger stock poledri (colts) and pultrellae (fillies) designate the offspring of horses. Moreover, the task of the compilers of these inventories was to list each and every asset of the estate, so it is clear that a mule breeding operation did not exist on these estates. Mules/hinnies and asses could, however, have performed useful tasks, serving as light draft animals to pull carts and hay wagons to feed the livestock on the estates. The foregoing evaluation of the age and gender categories from the tabular data of the BE presents a sequence of argument that these estates were dedicated to the production of warhorses, as explained by the military provisions of the CV and the separation of age and sex categories in the BE, and particularly the practice of biennial breeding of select mares to ensure the supply of large and strong colts. Indispensable to the production of war horses, the resupply of broodmares and the careful selection of stallions was integral to the maintenance of the horsebreeding program on these estates. The tabular data of the BE would be incomplete if left to stand alone. More information about the workings of Charlemagne’s

68

69

Lewis and Short, A Latin Dictionary, p. 1963, described “vel” as a “disjunctive conjunction used to introduce an alternative as a matter of choice or preference.” The Polyptych of Irminon c. 830 enumerated livestock and equipment in phraseology resembling that of the BE. On the similarity of the BE to the polyptychs, see Dopsch, Die Wirtschaftsentwicklung der Karolingerzeit, 1, p. 75, and Verhein, “Brevium Exempla,” pp. 374–5. In examining these passages, Jean Durliat, “Le polyptyque d’Irminon et l’impôt pour l’armée,” Bibliothèque de l’école des chartes, 141 (1984), 191, and documents, pp. 202, 204, 206–7, noticed that the conjunction “aut” actually meant “and.” The application of Durliat’s reasoning to an understanding of “vel” in referring to adult male equids of the BE resolves the problem of equivalency. Baist, “Zur Interpretation der Brevium Exempla und das Capitulare de villis,” p. 26, stated the impossibility of rendering vel in its traditional meaning, and suggested that vel should be construed as “que.” Niermeyer, Mediae Latinitatis lexicon minus, p. 108, burdo, defined as mule or hinny. Du Cange, Glossarium, 1:810, defined burdones as asses, mules, or hinnies. Estate II of the BE resolves this terminological dilemma in clearly distinguishing burdones from asini.

52

Carroll Gillmor

horse-breeding program can be derived from a comparison of the BE data with the equine provisions of the CV. The breeding categories and numerical data of the BE provide a practical reality to the general practices contained in the CV. The provisions of the CV will lay the groundwork for a comparison of the two sources, specifically stallion management and the competence of the stewards, R. H. C. Davis’s interpretation of stallion management based on the CV, and the terminology relating to colts and fillies. Chapters 13–15 of the CV explain the stages in the horse-breeding operation. At the beginning of c. 13, stewards received the directive not to allow stallions to remain for a long period in the same pasture, lest it be destroyed. 70 Chapter 13 further explains the necessity of rotating pastures, and also implies that stallions were kept in the same enclosure, away from mares, during the non-breeding months. As described explicitly at the end of CV c. 13, stewards carefully scrutinized each stallion’s general health to ensure his good condition “before the season comes for sending them into the mares.”71 This passage on pasture breeding is unclear as to whether all of the stallions on an estate were sent in among the mares, or whether just one was sufficient to cover a band of mares in a field. To arrive at a more precise understanding of this passage, it is necessary to examine the implications of each alternative. The introduction of two or more stallions to a group of mares risked the occurrence of outright physical confrontations between the stallions in sorting out the mares.72 Carolingian stewards, knowledgeable of the potential for stallions’ aggressive behavior during the breeding season, presumably would not have endangered a valuable stallion, and would have preferred to rely on a single stallion to cover a group of around ten mares. Moreover, allowing multiple stallions into a band of mares would have resulted in some inbreeding in subsequent generations, compromising the maintenance of genetic diversity. Unfortunately, we do not have any clear evidence of how the preservation of hybrid vigor may have worked. The random effect of pasture breeding with two or more stallions would have prevented the collection of accurate information, specifically the precise number of mares each stallion covered and the exact sire of each foal. Such records would not have been functional without positive identification of the individual horses. This could have been readily achieved with the practice of branding, a symbol designating the offspring of a given stallion. With this information, none of which has survived, stewards could have devised ways to prevent inbreeding. It has been suggested that the stewards rotated each stallion to a different broodmare band every year.73

70

71 72 73

MGH, Capit. 1:84, c. 13: “Ut equos emissarios, id est waraniones, bene praevideant et nullatenus eos in uno loco diu stare permittant, ne forte prohoc pereat.” On waraniones, see Baist, “Zur Interpretation der Brevium Exempla und des Capitulare de villis,” pp. 44–5. MGH, Capit. 1: 84, c. 13. Stephen Budiansky, The Nature of Horses, Exploring Equine Evolution, Intelligence, and Behavior (New York, 1997), pp. 96, 135. Hyland, The Medieval Warhorse, p. 62.

The Brevium Exempla

53

The first interpretation of the CV passage, random pasture breeding, would have resulted in fighting stallions with the potential of disabling injuries and uncontrolled inbreeding. With the second option, individual stallions were bred to bands of mares so as not to endanger their physical well being, a practice which would have allowed for the keeping of pedigree records to prevent inbreeding. Since the latter method was clearly better than the former, the interpretation of this passage ultimately depends on our assessment of the knowledge and experience of the stewards. Internal evidence from the CV contains information on the competence of the stewards. Charlemagne intended for his stewards to be knowledgeable about every aspect of estate management,74 and so the practical knowledge of the stewards acquired from years of experience and the oral tradition passed down from their predecessors favors the more sophisticated approach. The preceding integration of supplementary data from the BE with the provisions of the CV on stallion management modifies several generally accepted notions about Carolingian horse breeding. Segregation of categories by age and sex in the BE appears to reflect a planned breeding operation. Accordingly, each group would have been kept in a separate enclosure or pasture, including the stallions who apparently were not stabled separately from each other, a practice corroborated by CV, c. 13, providing that, to prevent overgrazing, stallions should not be allowed to remain for a long time in the same pasture.75 The analysis undertaken above, based on an integration of data from the BE and the CV, poses a challenge to recent scholarship on the subject of Carolingian horse breeding. R. H. C. Davis, working purely from the CV without reference to the BE, concluded that Charlemagne probably put stallions “in with several groups of mares, one after another.” Davis argued that “in this way each stallion could service more mares than before; and since the number of stallions at stud could be reduced, there would have been no need to use any which were not of

74

75

MGH, Capit. 1: 83, c. 5–10. Bachrach, “Are They Not Like Us?,” p 127. Publius Vegetius Renatus, De mulomedicina, ed. Ernst Lommatzsch (Leipzig, 1903), v–xxvi, for manuscript with the provenance of Corbie. The treatise mainly contains descriptions of diseases and ailments of mature horses with detailed recipes for cures, or drenches, concoctions of various ingredients to be poured down the throats of afflicted animals. References to the duties of stewards on the care of horses are scattered throughout the work. Vegetius in the preface to book I, p. 15, emphasized the daily oversight of the animals. In book I.56, pp. 80–90, Vegetius dedicated this section of his work to the preservation of the health by careful oversight in preference to remedies. Also in c. 56 Vegetius advised the paterfamilias (steward) to entrust the handling of horses to “suitable, restrained, and knowledgeable men.” In the prologue to book II, p. 96, Vegetius recommended that the paterfamilias keep records of the deaths of animals for comparison with the costs of medicine and veterinary fees so as to prevent one undesirable animal from detracting from the health needs of the many. The ethical obligations of the veterinarian appear in the prologue to book III, pp. 244–5, and book III.7, p. 252, on the care of the stables to ensure proper food and attention to the horses. The author would like to acknowledge the unpublished English translation of Vegetius De mulomedicina by the late Professor Margaret Mezzabotta, Department of Classics, University of Cape Town. MGH, Capit. 1:84, c. 13.

54

Carroll Gillmor

the highest quality.” He believed that this change amounted almost to a revolutionary advance in horse breeding, concluding that “it is impossible to overemphasize the importance of this development.”76 But the data derived from the BE – and supported by later medieval records – suggest that this supposedly critical change never took place. The inventories from Annappes and the related estates indicate that stallions covered a relatively small number of mares and did not breed large numbers of mares in successive bands in a given year. Provided the crown estates of the BE are considered typical, Carolingian equine breeding populations should not be described as “herds,” which implies a large group or multitude, but rather as “bands.”77 Moreover, the facilities on which Carolingian horse breeding took place operated on a much smaller scale than Davis’s notion of vast herds of mares being serviced by only a few highly prized stallions. These were agriculturally diversified estates, not comparable to huge modern stud farms devoted exclusively to horses. Chapters 14–15 of the CV described the stages in the horse-breeding process that can be significantly enhanced with additional insights from the BE. Chapter 14 of the CV stipulated that colts (poledri) should be segregated at the proper time. “And if the fillies (pultrellae) increase in number, let them be separated so that they can form a new band by themselves.”78 The listing in the BE of yearling colts and fillies as separate categories of the breeding populations indicates the separation of foals occurred either before or upon reaching one year, the approximate age for reaching sexual maturity for both colts and fillies.79 This contrasts with Davis’s less precise assertion that foals should be separated from the

76 77 78

79

Davis, “Warhorses of the Normans,” p. 72. Davis, “Warhorses of the Normans,” p. 71. MGH, Capit. 1:84, c. 14, described the group of fillies as a grex, which designates either a herd or a band. MGH, Capit. 1:84, c. 14: “Ut iumenta nostra bene custodiant et poledros ad tempus segregent; et si pultrellae multiplicatae fuerint, separatae fiant et gregem per se exinde adunare faciant.” Translation by Loyn and Percival, Reign of Charlemagne, p. 67. Bachrach, “Are They Not Like Us?” p.127, thought these fillies were separated in order to form a “new stud.” More plausibly, some of them served as replacements for deceased mares or mares who were considered too old for breeding. A possible implication of the fillies forming a new band by themselves is that the number of mares on each estate was divided in half according to the practice of biennial breeding. These halves were then divided into separate groups, in keeping with the data on the number of mares per stallion on each estate: at Annappes, six groups of eight plus two to be distributed amongst the groups at the discretion of the steward, estate II with eight groups of ten mares, estate III with four groups of eleven for a total of eighteen bands. From the total of 174 mares on an eight year breeding cycle, twenty-two would retire annually. If one mare were to be retired every year from each breeding band, then eighteen mares would be withdrawn from the breeding program annually, four fewer than required by the breeding cycle. The deficit could be adjusted at the discretion of the stewards. Of the average of thirty-five fillies produced annually, twenty-two would replace retired brood mares, leaving a group of thirteen to form a new band by themselves. On the ages of reaching sexual maturity for colts and fillies, see n. 29 above. See McDonnell, “Sexual Behavior,” p. 114, for “copulatory encounters” between young fillies by gangs of young males, when both sexes are still in their natal bands or in transition from their natal bands.

The Brevium Exempla

55

herd by age two.80 Again, information from the BE clarifies a provision of the CV and raises doubt as to a conclusion reached solely on the basis of the CV. Chapter 15 of the CV ordered the transfer of the poledri to the winter palace before the Feast of St Martin, 11 November.81 The translation of this term significantly affects our understanding of the age when young male horses were sent to the winter palace. If translated as “foals,” horses of either gender less than six months old, the poledri would be identified with the weanling colts foaled the previous spring of the same year who were sent to the winter palace presumably for training as warhorses.82 Hyland construed poledri here to mean foals, and cobbled together CV, chs. 14 and 15, maintaining that weanling colts were sent to the winter palace by 11 November, when they would be about six months old, for training as warhorses, while the fillies were retained to form future bands of broodmares.83 Bringing weanling colts to a central facility when their training as warhorses did not begin until age three,84 was not desirable logistically. Space for the free range growth and play of colts could be achieved more easily in a decentralized regime as represented by the estates of the BE than a central facility such as the winter palace. Leaving the colts on the estates to mature distributed the burden of their maintenance, avoiding the logistical difficulties of centralizing the tremendous quantities of resources needed for their sustenance over a two year period until the beginning of their rigorous training. The integration of the BE with the CV enables us to establish the age of the poledri who were sent to the winter palace so as to determine whether they were foals or colts. The terminology of gender in CV c. 14 distinguishes the poledri from the fillies (pultrellae). The distinct term for the fillies who formed future broodmare bands gives away the meaning of poledri as colts, for the text is mainly concerned with segregating groups of breeding stock. The same gendered pair designations of poledri and pultrellae also are listed in c. 62 of the CV estate assets to be included in the annual income statements.85 The gendered terminology of pultrellae and poledri appears in the BE categories of estates II, III, and IV. A textual problem exists with Annappes on these gender categories. The equine data of the Annappes estate first lists the mares (iumenta maiora) followed by the unnamed categories, “de anno tertio V, de preterito VII, de presenti VII.”86 The text suggests that these unnamed categories of younger horses also should be construed as iumenta. This presents the gendered pairs as iumenta and poledri. Two explanations can be advanced for considering these unnamed categories of

80 81 82 83 84 85 86

Davis, “Warhorses of the Normans,” p. 72. MGH, Capit. 1:84, c. 15: “Ut poledros nostros missa sancti Martini hiemale ad palatium omnimodis habeant.” Niermeyer, Mediae Latinitatis lexicon minus, p. 871 (pulletrus). Ann Hyland, The Medieval Warhorse, from Byzantium to the Crusades, pp. 62–3. Columella, De re rustica, 6.29.4; 2:200–1, explained that horses should not be ridden until age three. For text, see Gillmor, “Equine Epidemic,” p. 40 n. 68. MGH, Capit. 1:88–89, c.62. MGH, Capit. 1:128, c. 25

56

Carroll Gillmor

younger females as pultrellae. The formulaic entries for the remaining three estates (II, III, and IV) supply the term for fillies (pultrellae) to designate these categories. These estates (II, III, and IV) consider females of three years and younger to be pultrellae, and females aged four and older as iumenta. Arguably, to have the Annappes entry conform with the formulaic entries of the remaining estates would require the unnamed categories to be designated as pultrellae. This anomaly is probably the result of a scribal oversight in the entry for the estate of Annappes, and was corrected with the insertion of the designation “pultrellae” in subsequent entries. Moreover, the designation of iumenta for these younger females would be incorrect for the term “iumenta” denotes a female horse that has reached age four. As argued above, the disadvantages of breeding two-yearolds and yearlings would have been counterproductive to a program dedicated to the production of warhorses. The three-year-old females at Annappes would not have become mares until they produced their first foal at age four. Thus, if we construe the unnamed female categories of the Annappes estate as pultrellae, we have gendered pairs in all four estates. Females three years old and younger were enumerated separately as pultrellae as distinct from poledri which clearly means that poledri was not being used in its inclusive meaning. The use of these terms in gendered pairs within c. 14 and c. 62 of the CV and within the age groups of the BE leaves no doubt that poledri should be construed as colts instead of foals. The presence of two-year-old colts on the estates of the BE provides clear evidence that that the colts sent to the winter palace by 11 November were not six-month-old weanlings and that c. 15 of the CV must refer to another group of colts. The real difference among the age groups of the BE appears in the significant gap of information in the data on the three year olds. Only fillies are specified; colts are noticeably absent except on fisc IV where they are represented by tanti (so many), a number that cannot be determined.87 Estate managers retained the fillies to replenish the broodmare band as explained in CV c. 14, but the absence of three year old colts from Annappes and fiscs II and III of the BE needs to be reconciled with the arrival of colts at the winter palace by 11 November in CV c. 15. The colts dispatched to the winter palace should not be identified with those who turned age three the previous spring, otherwise they would have missed nearly an entire year of training. This scenario would have placed three year old colts on the three properties (Annappes, fiscs II and III) of the BE for most of the year, a view that would not reconcile with the data from the BE in the explicit exclusion of three year old colts from residing on these fisci. The colts of CV c. 15 most likely were nearing the end of their second year in November so that they could begin training as warhorses the following spring, when they reached age three and no longer would be residing on crown estates similar

87

Verhein, “Brevium Exempla,” pp. 348–50: The copyist substituted tantus (so many) for illegible numbers of his model manuscript, indicating here an indeterminate number of three-yearold colts.

The Brevium Exempla

57

to those described in the BE. Thus, the colts of CV c. 15 correlate with the absence of three year old colts on the three crown estates of the BE. The equine sections of the BE inventories will have been written before early November, when the colts departed from the estates. At the time of separation in November, estate managers of the fisci would reserve only the outstanding colts to serve as breeding stallions, or only as many as needed to replace deceased or retired stallions or to increase the number of stallions to correspond with an increased herd of mares. These are plausible explanations for the three year old colts still residing on fisc IV. Once the colts arrived at the Christmas court, the king or his advisers would choose the best ones for future training. Not every colt sent to the winter palace possessed the qualities to develop into a warhorse. It is unlikely that the rejects were sent back to the royal estates, because, except at estate IV, no three year old colts or stallions explicitly designated as four year olds appeared on the inventories. Moreover, the colts rejected as prospective warhorses would not be returned to the royal estates, for they had already been deemed undesirable as breeding stallions. Those without the requisite combination of athletic ability and mind, or natural mental aptitude, would have been relegated to other tasks. These colts were bred to be docile and obedient, qualities they would need as warhorses and to be ridden by cavalrymen with average skills. They could remain entire and function as riding horses, especially if they were not to be used for breeding. The production of warhorses was a costly enterprise; their development entailed a unique combination of attributes that made such individuals difficult to find. The numerical data of the BE reflects the difficulties of raising warhorses, while nevertheless tending to support the larger numbers side of the Carolingian army size debate. The BE contains significant information about Carolingian horse breeding that enhances the equine provisions of the CV. The fiscal estate accounts of the BE as integrated with the CV are the primary sources in explaining how Charlemagne replenished his supply of cavalry horses after the 791 epidemic, and more generally for understanding the horse-breeding and horseraising practices of the era.

3 Infantry and Cavalry in Lombardy (11th–12th Centuries) Aldo A. Settia [Originally published as “Fanti e cavalieri in Lombardia (secoli XI–XII)” in Aldo A. Settia, Comuni in Guerra. Armi ed eserciti nell’Italia delle città (Bologna, 1994)]

Translated by Valerie Eads

Milites and Pedites at Milan* The patarene leader Erlembaldo, anticipating an action between two cittadine factions in June 1075, first prepared “cavalry, and infantry assigned to the transport of assault ladders, supplies and diverse kinds of war machines” and then prepared in secret “crossbowmen, slingers, and ladders 20 cubits high fitted with iron at their bases so that they could stand by themselves.” When the moment of battle arrived, the leader, “wearing splendid armor, mounted a bay horse, grasped the banner” and theatrically took out his beard from under the armor “in order to arouse greater fear,” then threw himself first upon the enemy. Too trusting in his recurring good luck, he fell almost immediately, pierced through.1 Leaving aside the ideological reasons for which Erlembaldo fought, we can gather from the preparations he had made some of the essential elements that characterize the methods of warfare in the second half of the eleventh century. The mention of ladders “for taking houses” (ad capiendas domos) and of “various machines” (machinasque diversas) lets us guess at the existence of numerous fortified houses which had to be taken one by one. Siege warfare therefore found application even inside the city, and its importance is such that the infantry as well appear enslaved to it, having become mere auxiliaries assigned to the transport of the machines and supplies.

* The Latin terms miles/milites and pedes/pedites are left untranslated throughout. Miles in particular has multiple levels of meaning in terms of class as well as military function and translation can be impractical in a study comparing milites and pedites. Similarly, the Italian word cittadine/cittadini which can refer to citizens of a city, residents of the region, or urban dwellers as opposed to countryfolk, etc., has been left untranslated. The Italian cavalieri, which can mean either “cavalry” or “knights,” has been translated as the former. 1

The story, very well known, is to be found in Landulphus Senior, Mediolanensis historiae libri quatuor, ed. Alessandro Cutolo. Rerum Italicarum Scriptores [hereafter RIS], n. s., 4.2 (Bologna, 1942), 2.30, p. 123 [hereafter cited as Landulphus].

Infantry and Cavalry in Lombardy

59

There are, however, combatants on foot whom Erlembaldo secretly holds in reserve, together with other, more sophisticated, siege equipment, expecting decisive results from their intervention: they are chosen marksmen (balistas ac fundibularios) among whom those armed with a crossbow stand out because of their rareness and importance. From the height of his destrier, however, the figure of Erlembaldo dominates all, imposing and terrible in his splendid armor (loricam admirabilem). Besides being the leader, he also looks like the personification of the elite warrior – the heavily armed horseman, a veritable icon of military power. The intense internal military activity that convulsed pre-communal Milan is therefore marked by the same technical elements that predominated in the central regions of the West, i.e., the persistent supremacy of armored cavalry, the growing importance of portable projectile weapons and the multiplicity of fortifications which causes the parallel development of siege techniques.2 Although in many aspects representative of a general reality, the cited episode does not do justice to the role of the infantry which, in Italy as elsewhere in Europe, did not in fact stay confined to the humble tasks to which Erlembaldo appointed them.3 In the tenth century, Liutprand of Cremona could completely ignore the pedites, but mention of them becomes more and more frequent in the sources at the turn of the millennium,4 and for Landolfo, a Milanese who wrote at the end of the eleventh century, it is totally normal to consider every army as composed of a minority of milites and a majority of pedites.5 If, referring to his city, he speaks first generically of “citizen warriors” (cives bellatores), he then distinguishes between “worthy milites” (boni milites) and “very vigorous citizens” (strenuissimi cives) and finally, more clearly, between “very vigorous milites” (milites strenuissimi) and “strong pedites” (pedites fortes),6 two classes that, while socially distinguished, are both marked by adjectives intended to flatter their respective military value on an almost equal level. Although the distinction between milites and pedites interests us here above all from a practical point of view, i.e., with regard to those who operate on foot or in the saddle, we cannot ignore its different social connotations at the time. In 2 3

4

5

6

Cf. Robert J. Bartlett, “Technique militaire et pouvoir politique, 900–1300,” Annales ESC 41 (1986), 1135–59 who, however, excludes the Mediterranean zones from his analysis. See, for example, Claude Gaier, Art et organization militaire dans la principauté de Liège et dans la comté de Looz au moyen âge (Brussels, 1968), pp. 49–51, 145–7, as well as the debate between John France, “La guerre dans la France féodale à la fin du IXe siècle,” Revue Belge d’histoire militaire 23 (1979), 177–98, and J. F. Verbruggen, “L’art militaire dans l’empire carolingien (714–1000),” ibid., 289–411. See, for example, the sources cited by Giovanni Tabacco, “Il regno italico nei secoli IX–XI,” in Ordinamenti militari in Occidente nell’alto medioevo 2, Settimane di studi del Centro italiano di studi sull’alto medioevo 15 (Spoleto, 1968), pp. 786–8. The emperor Lambert would have besieged Milan [896] with “militibus et peditibus,” p. 28; Otto I goes to Rome [962] “cum innumerabili atque ineffabili peditum virorum fortium multitudine,” p. 49; Conrad II besieges Milan [1037] while disposing of “peditum multitudine ineffabili,” p. 59. Landulphus, pp. 29, 50, 59.

60

Aldo A. Settia

the first decades of the eleventh century, the milites cittadini still identified themselves in that elite group endowed with public and ecclesiastical possessions that allowed them to equip and train themselves for combat on horseback. The pedites for their part are the heirs of those free people who, although practicing other activities, since the early Middle Ages were required to perform armed service for the “defense of the fatherland” (defensio patriae). Thus, they perhaps possessed traditional military attitudes that in Milan put them in a position to successfully oppose the hegemony of the milites.7 The exchange between the two categories, however, appears very active. Already by the time of Landolfo the equation miles = eques is more apt than the original equation of miles = vassus, whatever the social origin of those who fight on horseback.8 And in the twelfth century, when a large part of the cavalry and of the infantry is no longer socially the same as their predecessors, this equation is all the more valid.9 Next to them a “people without arms” continues to exist in the city10 probably a reservoir from which new pedites can emerge to replace those who move up to the superior category.

Training and Cooperation Given the scarcity of information on the technico-military organization of the city in the pre-communal era, the episodes that Landolfo connects to the resistance put up by the Milanese to Conrad II in 1037 assume a particular value.11 The first item of a general character, not attested in any other source, regards training in arms, which it turns out is already undertaken at the expense of the city itself. At Milan there exist in fact “magistri belli,” i.e., military instructors 7 8

9

10

11

Tabacco, “Il regno italico,” pp. 782–90. As was noted by Georges Duby, “La diffusion du titre chevaleresque sur le versant méditerranéen de la Chrétienté latine,” in La noblesse au moyen âge, XIe–XVe siècles. Essais à la mémoire de Robert Boutruche, ed. Philippe Contamine (Paris, 1976), pp. 64–5. Cf. Cinzio Violante, La società milanese nell’età precomunale (Rome and Bari, 1974), pp. 261–5; Piero Brancoli Busdraghi, La formazione storica del feudo lombardo come diritto reale (Milan, 1965), pp. 154–5; Hagen Keller, Adelsherrschaft und städtische Gesellschaft in Oberitalien (9.–12. Jahrhundert) (Tübingen, 1979), p. 33, n. 73, p. 43, n.113; idem, “Militia. Vasallität und frühes Rittertum im Spiegel Oberitalienischer Miles-Belege des 10. Und 11. Jahrhunderts,” Quellen und Forschungen aus italienischen Archiven und Bibliotheken 62 (1982), p. 97. Gesta Federici I. imperatoris in Lombardia, auctore cive Mediolanensi, ed. Oswald HolderEgger, MHG SS in usum scholarum 27 (Hannover, 1892), p. 48. On 8 August, 1161, in response to a German attack “ad casinum Thome,” “egressus est autem populus sine armis [italics added], quidam de militibus et peditibus cum armis exierunt” [hereafter Gesta Federici]. Landulphus, pp. 61–2. The liveliness and the very confusion of the exposition allows us to think that the account could well reflect the report of persons who had actually seen those moments as does the presence of details that the chronicler had no interest in inventing such as, for example, the “flight” from the Roman arch accomplished by one of the defenders, p. 63. See also the opinion of Giorgio Giulini, Memorie spettanti alla storia, al governo e alla descrizione della città e campagna di Milano nei secoli bassi, 2 (Milan, 1855), p. 234.

Infantry and Cavalry in Lombardy

61

specifically assigned to preparing men for combat. It would be of great interest to know who they were and where they had learned the ideas they were called upon to teach. In Italy, as elsewhere, the Epitoma of Vegetius, the basic text of military knowledge throughout the Middle Ages, was certainly known even then.12 It is true that the sources for Lombardy in the eleventh and twelfth centuries do not ever make explicit mention of it, but this signifies mainly that the work of Vegetius was unknown to the chroniclers who have handed down the facts to us and does not exclude the possibility that it could have been known to those who held the reins of the military organization.13 This would perhaps explain the considerable and well-attested diffusion that Vegetius would enjoy in the later communal age.14 It appears nonetheless unlikely that the Milanese instructors recorded by Landolfo would turn back to a precise codified tactical doctrine and not instead – as was then done elsewhere – to empirical

12

13

14

A copy of the Epitoma rei militaris made in Italy in the tenth century was offered to Emperor Henry IV; there are in addition numerous other Italian manuscripts of the tenth–eleventh centuries. Flavi Vegeti Renati, Epitoma rei militaris, ed. Carl Lang (Leipzig, 1885; rpt. Stuttgart, 1967), pp. xxviiii, xxxi, xxxvii-xli [hereafter Epitoma]; among the available translations, Flavius Vegetius Renatus, Epitoma rei militaris, ed. and trans. Leo F. Stelten (New York, 1990) gives the Latin and English text on facing pages. On the diffusion of the work of Vegetius and its importance in the Middle Ages in general see also Philippe Contamine, La guerra nel medioevo (Bologna, 1986), pp. 289–91; an English translation, War in the Middle Ages, trans. Michael Jones (Oxford and New York, 1984) is available; Foster Halberg Sherwood, “Studies in the Medieval Use of Vegetius” Epitoma rei militaris (Ph.D. dissertation, University of California at Los Angeles, 1980); Mathias Springer, “Vegetius im Mittelalter,” Philologus 123 (1979), pp. 85–90; Josette A. Wisman, “L’Epitoma rei militaris de Vegèce et sa fortune au moyen âge,” Le moyen âge 85 (1979), pp. 13–31; for an essentially complete list of surviving manuscripts see Charles R. Shrader, “A Handlist of Extant Manuscripts Containing the De Re Militari of Flavius Vegetius Renatus,” Scriptorium 33 (1979), pp. 280–305. One can perhaps gather some indirect traces of the presence of Vegetius in our chroniclers: one could, for example, see reminiscence of Vegetius in Landolfo where he speaks of ladders and machines “ad capiendas domos” (see the text corresponding to note 1, above) to compare with Epitoma 4.30, “ad capiendos muros scalae vel machinae plurimum valent,” p. 149. The clever device put into action by Frederick I, who, before the battle of Carcano in 1060 “viam hostibus obstruxit tantum ut etiam arbores precisas in via sterneret,” Gesta Federici, p. 42, could trace back to a teaching of Vegetius, “rursus post se praecisis arboribus vias claudunt,” Epitoma 3.22, p. 114. Among the thirteenth-century writers, Salimbene asserts that he knows Vegetius, attesting in the meantime to the use that the papal legate Gregorio di Montelongo made of him. Salimbene de Adam, Cronica, ed. Giuseppe Scalia (Bari, 1966), p. 562; English translations include The Chronicle of Salimbene de Adam, trans. Joseph L. Baird, Giuseppe Baglivi, and John Robert Kane, Medieval & Renaissance Texts & Studies 40 (Binghamton NY, 1986). In the same period, the Epitoma was translated into Italian in Florence by Bono Giamboni: Cesare Segre, “Jean de Meun e Bono Giamboni tradutttori di Vegezio” in idem, Lingua, stile e società (Milan, 1976), pp. 271–300. Finally, an acquaintance with Vegetius—even if not directly named— shows through in the account of the siege of Este by Rolandino of Padua: Rolandinus Patavinus, Cronica in factis et circa facta Marchie Trivixane, ed. Antonio Bonardi, Rerum italicarum scriptores, new series, 8.1 (Città di Castello, 1905), p. 91. See also Aldo A. Settia, Castelli e villaggi nell’Italia padana. Popolamento, potere e sicurezza fra IX e XIII secolo (Naples, 1984), pp. 363, 417.

62

Aldo A. Settia

knowledge traditionally transmitted “through experience and oral instruction.”15 Such knowledge could, however, be influenced in a direct or indirect manner by occasional reading of Latin treatise writers or also by appropriate biblical passages.16 We can draw some features of the military instruction imparted to the Milanese from the same text of Landolfo who, in describing the defensive action carried out by the milites and pedites jointly engaged beneath the walls of Milan in 1037, seems to refer directly to certain elementary rules of this training: no combatant should throw himself forward against the enemy without judgment nor, to the contrary, can he abandon his post, at least if he is not gravely wounded; similarly, no one should dare to depart individually from the ranks, not even to strike an adversary who appears within range; and, if enemy pressure increases in a given sector, only the operating unit designated to do so, by means of the proper visual signal from the height of the Roman arch, prepares to rush to reinforce it.17 It is worthwhile to call attention to this exceptional command post dominating the entire battlefield, from which one can discern the movement of the units engaged in combat.18 In a period when the commander usually stood at the head of his troops, such a solution seems completely extraordinary and greatly ahead of its time.19 Under the eyes of the archbishop, Landolfo continues, “cavalry and infantry manfully take up their arms, as they have learned from their instructors, competently defend the positions assigned to them and fight with skill and prudence,”20 following procedures that, although they are elementary and purely 15 16

17

18

19

20

Gaier, Art et organisation militaire, p. 143. As Salimbene, for example, suggests, “in libris Machabeorum similiter multe sagacitates habentur, que valent ad artem pugnandi. Similiter in libri Regum et in libro Iudicum et Iudith,” Cronica, p. 562; see also Aldo A. Settia, “‘De re militari’: cultura bellica nelle corti emiliane prima di Leonardo e Machiavelli,” in Le sedi della cultura nell’Emilia Romagna. L’epoca delle signorie. Le Corti, ed. Giorgio Chittolini (Milan, 1985), especially p. 68. Certain names that Landolfo applies to Milanese military formations could also have biblical roots: “Fortissimis centenis dimissis” and “Eriprandus vicecomes, miles milenarius” are comparable to 2 Samuel 18:4, “ … egrediebanturque populus per turmas suas centeni et milleni” [all italics added]. Landulphus, p. 62. “Itaque inimicorum telis callide exceptis, viriliter sua arma regentes milites ac pedites, ut a magistris belli erant docti, certatim ut erant ordinati suum locum custodientes, competenter et caute certabant. Nemo sino concilio in hoste feriendo irruebat; nemo suum terminum sine gravi hostium concursione aut vulnere deserebat: nemo hostem, etiamsi opportunum ad feriendum conspiceret, solus ex suis exiliens percutere audebat. At ubi bellum in aliquam gravescebat partem, et pondus eiusdem supereminebat, non omnes sed legio ad quam qui super turrim astabat signum faciebat, ordinatis signis statim subveniebant.” Landulphus, p. 61. “Primo iuxta archum triumphalem, quem Heribertus papilione superimposito et viris fortibus superimpositis mirifice armis munitionibus tormentis muniverat.” This certainly corresponds to the “tower” from which the signals came. Gaier, Art et organisation militaire, pp. 195–6; J. F. Verbruggen, The Art of Warfare in Western Europe during the Middle Ages, from the Eighth Century to 1340, trans. Sumner Willard and S. C. M. Southern (Amsterdam and New York, 1977), pp. 197–200; the 2nd ed., rev. and enlarged (Woodbridge, Suffolk, 1997), pp. 217–21, of this translation restores Verbruggen’s original notes. Vegetius also advises this: “Dux, qui praecipuam sustinet potestatem, inter equites at pedites in parte dextra stare consuevit,” Epitoma 3.18, pp.102–3. See note 17, above.

Infantry and Cavalry in Lombardy

63

defensive, certainly imply notable disciplinary cohesion, harmony and capacity for coordination, the fruit, no doubt, of a considerable level of individual and group training. Even if the Milanese army at the time of Archbishop Ariberto (1018–45) would not really have possessed the capability described by the chronicler, his words do demonstrate in an unequivocal way what conception he would have of the level of military organization possible and necessary in his own time.21 The passage shows infantry and cavalry who fight side by side in perfect unity of purpose, a unity that will shortly afterward be broken, hurling the two categories against one another in a long conflict;22 nevertheless, a similar cooperation, much more durable, will be newly attested fifty years later by the anonymous author of the Gesta Federici. He constantly mentions milites and pedites together implying that their continual cooperation and the great harmony between them were completely normal. It is not rare, in fact, that he omits the very distinction between milites and pedites, simply indicating the area of the city, the “gate” (porta), i.e., the organizational units that includes both milites and pedites,23 and among which flourishes a healthy spirit of rivalry.24 We see later cavalry and infantry preparing spiritually for a battle by assisting at religious ceremonies together.25 The names of cowards belonging to one or the other category are indiscriminately registered to transmit their shame to posterity.26 In 1162, five hundred milites and as many pedites, with the carroccio, the standards, and other symbols of both footmen and horsemen, surrendered to Frederick I on a level of equality.27 The solidarity and comradeship seem to reach their peak in certain episodes of 1155. Cavalry and infantry of the Porta Comasina and the Porta Nova sent to reinforce the defenses of Piacenza “feared for their brothers” detached to Tortona

21 22

23 24

25 26

27

It cannot be at all excluded that Landolfo has a didactic intention and by re-evoking an already distant past intends in reality to refer to his Milanese contemporaries. Landulphus, pp. 63–7, 74 ff. On the course of events see Gian Luigi Barni, “Dal governo del vescovo a quello dei cittadini,” in Storia di Milano 3, Dagli albori del comune all’incoronazione di Federico Barbarossa (1002–1152) (Milan, 1954), pp. 79 ff. There are numerous examples in the Gesta Federici, e.g. pp. 17–23, 41–2, 47–8, 53. In May 1122, the Romana and Ticino “gates” easily took the castello of Stabbio “quasi portis aliis tribus invidentibus que paulo ante castellum de Civassio et turres violenter et quasi desperate ceperant.” Gesta Federici, p. 22. Gesta Federici, pp. 40–2. At Tortona in May 1155, “de Mediolanensibus tamen multi in ecclesia fugerunt quorum nomina ad eorum ignominiam scripta fuerunt,” Gesta Federici, p. 20. It would be interesting to know if the names of the fallen recorded by the chronicler on other occasions belonged all to the class of milites or also to that of pedites, ibid., pp. 17, 38. “Miserunt quingentos milites et totidem pedites cum carocero et vexillis et aliis signis equestribus et pedestribus,” Gesta Federici, p. 53, but see also the same episode in Otto Morena and Continuators, Historia Frederici I., ed. Ferdinando Güterbock, MGH SS, n. s., 7 (Berlin, 1930) p. 153, where he speaks of three hundred milites and nearly a thousand pedites; see also Ferdinando Güterbock, “Le lettere del notaio imperiale Burcardo intorno alla politica del Barbarossa nello scisma ed alla distruzione di Milano,” Bullettino dell’Istituto storico italiano per il medio evo e Archivio muratoriano 61 (1949), pp. 62–3.

64

Aldo A. Settia

and rushed to relieve them. A little afterward, the fellow-soldiers of the Ticino and Vercelli gates pair up and are sincerely moved to find their own brothers weak, wounded and sick; the latter, at the sight of reinforcements, rejoice evangelically, “with exceeding great joy.”28 In the cases in which the chronicler mentions discords, we always see individuals among them at adds and never milites against pedites,29 a sign without doubt of the homogeneity and internal solidarity that characterized the army at this moment. All of this seems to testify to a perfect unity of purpose between infantry and cavalry which must translate into a strong capacity to operate jointly on the battlefield. One could suspect that such an idyllic “concord of the ranks” (concordia ordinum) is only a rose-tinted version of the facts that the chronicler is pleased to present to readers for the sake of propaganda. We know, for example, that Milan was not monolithically anti-Swabian,30 but the solidarity between the milites and pedites is not only a Milanese fact, and it would be hard to believe that authors of diverse times and places found themselves always in agreement when presenting a completely false picture. The concord of the ranks, therefore, if not absolute, can be understood as a genuine fact on both the social and military level. It is nonetheless astonishing that such a close cooperation turns out to be possible especially in the latter case where – as Vegetius had already observed in his own time – “by their nature cavalry are not accustomed to getting on well with infantry.”31 Why ever, in fact, did soldiers who were well trained and equipped, quick and resolute in combat, consent since the pre-communal age to burden themselves with an encumbering plethora of men on foot who were so much slower and more vulnerable? There was, in the first place, certainly a practical reason: the cittadina society of the time was not able send forth a very high number of combatants on horseback, and they alone would not be sufficient to conduct whatever actions of war might be required, other than a simple raid. The participation of the much more numerous pedites was therefore indispensable in practice. The presence of external enemies was determinative in eliminating the internal discord, which sprang up precisely – notes Landolofo – “whenever enemies on every side are lacking.”32 The incessant wars conducted by Milan throughout the first half of the twelfth century (resulting, in the opinion of contemporaries,

28 29

30

31 32

Gesta Federici, pp. 18 and 20 respectively. In 1162 in besieged Milan concerning the timeliness of surrendering or not, “disessio maxima orta est inter cives, videlicet inter patrem et filium et virum et uxorem et inter fratrem et fratrem,” ibid., pp. 50–2. As has been brought to light very well by Livia Fasola, “Una famiglia di sostenitori di Federico I,” Quellen und Forschungen aus italienischen Archiven und Bibliotheken 52 (1972), p. 116– 218. “ … cum naturaliter equites a peditibus soleant discrepare.” Epitoma 2.21, p. 55. Landulphus, p. 63. “Quin etiam pacem cum hominibus habentes, cum iam inimici undique deficerent, gladios in semetipsos ferentes, hostes sibimetipsis effecti sunt.”

Infantry and Cavalry in Lombardy

65

in true excesses of bellicosity)33 restored in the city the conditions of solidarity that had existed in the times of the war against Conrad II, and the exaltation of civic patriotism found an effective symbol in the carroccio, which was created precisely to underline the need for unity between infantry and cavalry.34 It remains, nonetheless, to explain how a cooperation imposed by practical and political necessity could develop into something that it is not exaggerated to call affectionate comradeship. The active social mobility that characterized the cittadina society of the time probably comes into play here. Infantry and cavalry were not recruited from closed and rigidly opposed classes, but rather from people who lived together and collaborated daily in other activities. As was seen, already in the second half of the twelfth century the equites were no longer only vassals, but also well-to-do cittadini who were equal to them. So, beside the professional soldiers there were part-time soldiers as well, which the pedites always were. If we can better understand the premises that made the cooperation between infantry and cavalry practicable, the ways in which it was in fact realized still remain uncertain. For the most part the sources limit themselves to briefly noting that the two categories of combatants acted together without adding any more. One encounters this in many episodes of the war between Como and Milan (1118– 27). In the course of an expedition against Mariano, the Comaschi (troops from Como) were surprised by the Milanese and their allies from Cantù so that suddenly – as the poet says – “the streaming horse and foot rise up from all sides” (undique consurgunt equites peditesque fluentes).35 Operating around Dervio in cooperation with the fleet that blockaded the castle from the lake, “pedites and equites” (pedites equitesque) pressed it from the land.36 Against the Monzesi (people of Monza), put in command of a new castle by Milan, “resplendent in arms the cavalry and similarly the infantry all together” of Como proceeded; having caught the enemy by surprise, the Comaschi succeeded in capturing “all the cavalry and likewise all the infantry.”37 Again, defending the region near

33

34

35 36 37

Compare the judgment on the “dominandi libido” of the Milanese expressed in Carmen de gestis Frederici I. imperatoris in Lombardia, ed. Irene Schmale-Ott, MGH SS in ususm scholarum 62 (Hanover, 1965) ll. 21–4, pp. 1–2, with that of Gotifredus Viterbiensis, Gesta Friderici I. et Henrici VI., ed. Georg Heinrich Pertz, MGH SS in usum scholarum 30 (Hanover, 1870) ll. 20–1, 418–19, pp. 1, 16; see also Giorgio Picasso, “Il sermone inedito di Uberto abate milanese del secolo XII,” in Contributi dell Istituto di storia medioevale 1, Raccolta di studi in memoria di Giovanni Soranzo (Milan, 1968), pp. 324–48. See, in general, Hannelore Zug Tucci, “Il carroccio nella vita comunale italiana,” Quellen und Forschungen aus italienischen Archiven und Bibliotheken 65 (1985), 1–104, but consult also John C. Koenig, Il ‘popolo’ dell’Italia del nord nel XIII secolo (Bologna, 1986), pp. 163–4. De bello Mediolanensium adversus Comenses liber Cumanus, ed. Joseph M. Stampa, RIS 5 (Milan, 1724) l. 1284, p. 440 [hereafter Cumanus]. “Insidias circum componunt undique castrum / Clam pedites equitesque ex una parte sedentes / Ast alia de parte rates tectaeque morantes.” Cumanus, ll. 1503–5, p. 445. “Talibus auditis dictis, fulgentibus armis / Procedunt equites, pedites pariter simul omnes … / Et capiunt omnes equites, pedites simul omnes.” Cumanus, ll. 1703 ff., p. 450.

66

Aldo A. Settia

their city, the citizens of Como “sent out cavalry and infantry.” After a silent nocturnal ambush they surprised a Cremonese contingent, “demolished the cavalry” and “cut the infantry to pieces.”38 In a Milanese “task force” operating in 1060 against Lodi we see without differentiation “pedites and equites [acting] together as one.”39 And, one could add the better known cases of the battles of Carcano and Legnano in which, as was noted, cavalry and infantry also worked together next to the carroccio.40 There is an instance, although extra-military, in which the manner of cooperation between the mobilized Milanese infantry and cavalry is shown. In 1155, occupied with the rebuilding of the walls of Tortona, “the milites with horses brought the sand from the bank of the Scrivia while the infantry carried the lime from there to S. Maria in Scultabis.”41 These are tasks that complement each other and were certainly assigned with due regard for the attitudes and the means appropriate to each of the two categories, and are not without allusion to the cooperation usually realized in military actions. Historiography now takes it for granted that in all of western Europe, and therefore in the communal armies of Italy, the units of infantry armed with pikes and drawn up in formations had the fundamental task of furnishing security to the cavalry, allowing them to order themselves shoulder to shoulder in the presence of the enemy before delivering the decisive charge, an essentially static and defensive task as opposed to the mobile and offensive role of the cavalry.42 The procedure is explicitly described to us in the conclusive episode of the battle of Legnano. The Milanese infantry “with shields raised and spears extended” resist the enemy cavalry, allowing their own cavalry to re-form and to go on to the counter-attack. This narrative, while it confirms the generic scheme of interpretation previously indicated, also shows its incompleteness since it goes on to describe the infantry participating in the victorious final movement.43

38

39 40 41 42

43

“Deducunt equites peditesque in valle sedentes / Nocte struunt latebras caute sine murmure tecti … / Sternunt equites et ubique potentes / Detruncant pedites, calcant pedibus fugientes,” Cumanus, ll. 1780 ff., p. 451. “donec Mediolanenses equites simul et pedites unanimiter in Laudenses impetum magnum fecere,” Otto Morena, Historia (note 27, above), p. 104. Zug Tucci, “Il carroccio,” pp. 43–4, 52–3, and the sources cited there. “Siquidem milites cum equis suis duxerunt sabulum de ripa Scrivie, pedites vero calcem ad Sanctam Mariam in Scultabis inde portabant.” Gesta Federici, p. 21. This position is well examined by Zug Tucci, “Il carroccio,” pp. 42–3; see also above all Piero Pieri, “Alcune quistioni sopra la fanteria in Italia nel periodo comunale,” Rivista storia italiana 50 (1933), 564–6; reprinted in idem, Scritti vari (Turin, 1966) with the title “L’evoluzione delle milizie comunali italiane,” pp. 34–6; a similar conclusion had already been reached by Henri Delpech, La tactique au XIII siècle 1 (Paris, 1886), pp. 277–80, an author whom one tends to forget, but who deserves to be remembered. At the first attack, the Lombards “sunt ab imperatore in fugam conversi. Pedites vero Mediolanenses cum paucis militibus, qui circa caruciam erant, fugere non volentes, simul conglomerati stare ceperunt. Imperator autem videns Lombardos milites aufugisse, pedestrem multitudinem, que remanserat, credidit facile superare. Cumque congregata sua militia super eos vellet irrumpere, illi oppositis clipeis et porrectis astis ceperunt eorum furore resistere, et

Infantry and Cavalry in Lombardy

67

The Offensive Capacity of the Infantry It is therefore wrong to believe in a purely static role for the infantry. In Italy at least the accepted scheme was not the only one, and it may not even have been the most usual. We can extract confirmation of this from one of the rare pictorial representations of communal infantry in the twelfth century, probably fairly close to reality, offered by the famous bas-reliefs of the Porta Romana sculpted after 1171.44 All the pedites shown here wear on their heads helms with nasals and at their necks a shield to protect the front part of the body; nonetheless only some of them have pikes at their disposal, indispensable in static defensive actions, while the majority hold short weapons suitable for close-quarters combat. The documentation offered by the reliefs, although not giving a complete picture of the armaments (above all the marksmen are missing), is sufficient to raise questions about the actions generally attributed to infantry for whom a broader range of tactical possibilities is perhaps now recognized. We find an example of it in the war between Como and Milan. Around 1123, it was necessary to open the road barred by the men of Vertemate on foot and on horseback. The pedites of Como “all joined together for battle” preceding the intervention of their equites.45 The initial offensive movement here is therefore undertaken by infantry and only supported in a second action by the cavalry. It is natural that such autonomy of action shows itself especially against enemy infantry. In the text of Landolfo already cited, after the first phase of hostilities, which saw milites and pedites act with close coordination between them, there follows another phase in which “for some days cavalry fought against cavalry and infantry against infantry.”46 This all too concise declaration does not

44

45

46

advenientes animose repellere,” until “Lombardi, qui fugerant, resumptis viribus, et aliis qui de novo venerant, sociali, ad pugnam sunt animose reversi, et simul cum suis peditibus super imperatoris exercitum impetum facientes, ipsum in fugam unanimiter converterunt.” Romualdus Salernitanus, Cronicon, ed. C. A. Garufi, RIS, n.s., 7.1 (Città di Castello, 1935), p. 266. See the examples of them reproduced in Edoardo Arslan, “La scultura romanica” in Storia di Milano 3, Dagli albori del commune all’incoronazione di Federico Barbarossa (1002–1152) (Milan, 1954), pp. 591–2; Gian Luigi Barni, “La lotta contro il Barbarossa,” ibid. 4, Dalle lotte contro il Barbarossa al primo signore (1152–1310), pp. 2, 103; Maria Theresa Binaghi Olivari, “I rilievi di porta Romana e alcune sculture milanesi del XII secolo,” in Atti del’II° Congresso internazionale di studi sull’Alto Medioevo (Milano 26–30 ottobre 1987) 2 (Spoleto, 1989), pp. 783–91. There is some information concerning equipment in Lionello G. Boccia and E. T. Coelho, Armi bianche italiane (Milan, 1975), p. 324, note 6; Lionello G. Boccia, Francesco Rossi and Marco Morin, Armi e armature lombarde (Milan, 1980), p. 23. “Inque via pedites tunc stabant Vertematenses. / Et iaciunt hastas, iaculos, funduntque sagittas. / Sic iter impediunt, nequeunt transire volentes. / Vertematensis miles protectus in armis, / Cominus ense parat duro contendere ferro. / Committunt bello pedites, communiter omnes. / Haec equites cernunt, continuo signa resolvunt, / Voces emittunt, sic ad certamina tendunt, / Obstantes pellunt, vallo pluresque revolvunt.” Cumanus, l. 1218–26, p. 439. Landulphus, p. 62. “Demum, sic facientes, multis inimicorum ferro trucidatis, per aliquot dies equites cum equitibus, pedites cum peditibus proeliati sunt.”

68

Aldo A. Settia

allow us to learn either the reasons or the means of such a different use of opposing forces. There are, however, reliable comparisons in analogous episodes of the successive communal age.47 At Palosco in 1156 the infantry of Bergamo, arrayed in a wedge formation, attacked alone and put to flight the infantry from Brescia who were at once aided by their cavalry. The Bergamaschi were then caught between two forces, and “not wanting to turn their backs,” says the poet, “gathered together in the middle of the field and repelled equally the attacks of both the infantry and the cavalry, the few fighting against so many thousands.”48 Although, because of the overwhelming number of adversaries, the final result of the battle was defeat, the episode illustrates very well the offensive and defensive possibilities of the Lombard communal infantry and its capacity to operate with full autonomy with respect to the cavalry.49 Cavalry and infantry could however operate in indirect coordination, as Landolfo tells, each engaging the corresponding enemy forces. The pedites were able to act in an offensive mode against cavalry, though obviously with more difficulty. Still, this certainly was possible at least in some specific conditions: for example when the opposing cavalry was already in a bad way (as must have happened at Legnano), or had been surprised on the march using the lassoes and traps already known to Vegetius,50 or by using other means which it is worthwhile to dwell on. Landolfo, referring to the actions of 1037, recounts that certain Milanese “killed their enemies after having drawn them in with iron hooks.”51 It was evidently a case of hooks prepared just for harpooning the enemy from a certain distance.

47

48

49

50

51

See for example, Rolandinus, Cronica (note 14, above): in 1202, during an engagement among Veronese “in Braida,” “milites occurrerunt militibus, pedites cum peditibus manualiter pugnaverunt,” p. 22; in August 1256, Ezzelino offers battle, but from Padua there come out only “multi pedites” who “cum inimicis pedestribus, qui ante venerant strenue pugnaverunt” without the cavalry of one side or the other intervening (but this is probably a case of a simple initial skirmish which then has no continuation, as is usual, in a full battle). “At peditum pars magna coit validisque resistit / Viribus et pugnam solita feritate capessit, / Direxitque aciem cuneus violenter unus / Adversum pedites Brixiane gentis et illos / Acriter obtruncans gladiis, dare terga coegit. / Ast equites socios superari marte pedestri / Ut videre suos, laxis properanter habenis / Subveniunt dictisque animos in prelio firmant / … / Talia dicentes equites per gramina currunt / Atque fugam socium cohibent hostesque repellunt. / Undique Pergameos urguet miserabile bellum. / Hinc equites, illinc pedites, … qui vertere terga / Nolentes medio tandem glomerantur in agro / Incursumque equitum pariter peditumque repellunt / Acrius et pauci contra tot milia pugnant.” Carmen de gestis Frederici (note 33, above) ll. 1171–90, pp. 39–40. (See also note 85, below, and the corresponding text.) The Carmen also names one of the attack formations of the infantry, the cuneus (not to be considered—as it could appear at first glance—as a simple reuse of a classical term) which must have been usual in Italy as it was elsewhere. Gaier, Art et organisation militaire, pp. 159– 60. According to Delpech, La tactique 1, p. 273, this would have resulted from reading Vegetius. “Cataphracti equites … propter impedimentum et pondus armorum capi facile et laqueis frequenter obnoxii … ” Epitoma 3.23, p. 115; also, the cavalry of Isola Comacina that besieged Como, “laqueos et ubique paratos / Credit, in adversam se certa evolvere fossam,” Cumanus, ll. 240–1, p. 419. “alius ferreis uncinis hostes quod ad se trahere poterat, trucidabat,” Landulphus, p. 62.

Infantry and Cavalry in Lombardy

69

During the first crusade, at the siege of Marra the attackers turned to similar implements, wielding them from the top of a mobile tower against the defenders on the wall.52 These latter, in their turn, made use of them at Crema in 1159 to demolish the aggressors.53 In both cases it was a case of opponents who faced one another from low to high, and it is credible that the Milanese of whom Landolfo speaks acted in similar conditions. They, however, operated on the ground, and it is natural to think that the infantry used hooks against men on horseback. The point acquires greater interest if it is connected to some iconographic sources from the eleventh century depicting men equipped with bills. These are arms formed with a bent blade, like a pruning knife furnished with a side hook,54 that served precisely – as we know they did later – to allow infantry to unseat cavalry.55 The words of Landolfo can therefore be interpreted as the oldest testimony for the use of such a weapon with which a man on foot could henceforth face, in certain conditions, an armored horseman and have the upper hand. There are no proofs that a similar form of combat would be practiced in a systematic manner, but in the age here examined field battles between troops arrayed in precise formations – on which the traditional military historiography loves to dwell – represent, as is well known, exceptional actions; while the greater part of the very intense activity undertaken by the communal armies unfolds in minor actions that constitute, in short, the true reality of war. In just such an

52

53 54

55

“alii vero tenebant in hastis honorabilia signa et cum lanceis et hamis ferreis putabant eos trahere ad se,” Histoire anonyme de la première croisade, ed. and trans. Louis Bréhier (Paris, 1924), p. 172. For an English translation see The Deeds of the Franks and the Other Pilgrims to Jerusalem, ed. and trans. Rosalind Hill (London and New York, 1962). “Alios etiam tres ipsius ducis milites cum rampinis ferreis de ponte in terram trahentes ipsi Cremenses vivos retinuerunt,” Otto Morena, Historia, p. 91. For illustrations of these weapons refer to Giorgio Dondi, “Del roncone, del pennato e del cosidetto scorpione. Loro origini,” Armi antiche. Bollettino dell’Accademia di San Marciano (1976), pp. 22–5; see also David Nicolle, “Armes et armures dans les épopées de la croisade,” in Les épopées de la croisade, published by Karl-Heinz Bender (Stuttgart, 1987) figs. 2 and 45, reproducing respectively details of a bas-relief from Vézelay (circa 1130) and a sculpture from S. Paolo fuori le mura at Rome (circa 1170). See “Fauchart,” in Eugène-Emmanuel Viollet le Duc, Dictionnaire raisonné du mobilier français de l’époque carlovingienne à la Renaissance 5 (Paris, 1874), pp. 421–3; it appears in the thirteenth century in the southern provinces of France and in Italy as an infantry weapon with two functions, i.e., to cut and to hook the weapons of the horsemen in order to unhorse them (“fournire des coups d’estoc et à accrocher les armes des cavaliers afin de les désarçonner”). A practical example is given by Marino Sanuto Torsello, Istoria del regno di Romania in Chroniques gréco-romanes inédites ou peu connues, ed. Charles Hopf (Berlin, 1873), pp. 104–5. In 1258, the Venetians prevailed against French cavalry, and “the principal cause of the victory was that many Venetians had placed poles with iron hooks in their hands with which they pulled the cavalry down from their horses and pulled them in and easily overcame them.” The statutes of Pernumia (Padua) forbid, among other weapons, the carrying of “runconem cum manico minus duobus pedibus curto” and prohibit the blacksmiths from making “runconem falçonatum cum punta et uncino a manico,” Sante Bortolami, Territorio e società in un comune rurale veneto (sec. XI–XIII). Pernumia e i suoi statuti (Venice, 1978), pp. 202, 204.

70

Aldo A. Settia

environment can we find appropriate use of both less conventional forms of warfare and of other tasks in which infantry and cavalry can operate both on their own and in cooperation. In 1160, for example, the pedites of two Milanese porte were sent to garrison the encampments outside Carcano, replacing the milites who were called to intervene in a different sector. Afterwards milites and pedites acted against the rebels of Seprio at Appiano and Mozzate, each one independent of the other.56 There were later actions that we do not consider here such as laying waste systematically to the enemy territory, which pertains more specifically to the infantry,57 and a large part of the siege operations, both offensive and defensive, which in any case the cavalry could not avoid. Besides direct cooperation there existed, therefore, a whole range of tasks equitably distributed between the forces of the infantry and cavalry. The services offered by each need not have differed greatly from one another, as the rare and concise judgments that it is occasionally possible to gather from the sources confirm. We have already noted the use of adjectives such as such as “milites strenuissimi” and “pedites fortes” attested in Landolfo;58 the pedites from Como are called “powerful” (validi) and “tried,” (probati)59 the Brescians, “worthy” (boni).60 They do not come off at all belittled compared to their more prestigious fellow warriors on horseback. It is the extra-urban militia that furnishes the negative comparison.

“Rustici” and “Cittadini” In 1125, the Comasche, aligning themselves with Arnaldo of Albareto, a local lord who disposed of his own forces, tried to bar the valley of the Lugno to the Milanese. Along the river, where the defense was entrusted to the cittadini, the Milanese, “admit that the Comasche fought ferociously there,” to their cost. Their allies, charged with holding the nearby mount, quickly dispersed, causing the failure of the entire action. Before it was made, this negative judgment was already implicit in the expressions with which the author designates these inefficient

56

57

58 59

60

Gesta Federici, respectively p. 42, “Pedites vero burgorum porte Ticinensis et pusterle S. Eufemie reliquerunt, ut castra a Carcanensibus custodirent. Pedites porte Cumane posuerunt ad Taxariam,” and p. 47, “Post paucos dies consules Mediolani posuerunt milites centum in Crema; in Aplano milites, et pedites in Mozate iam ante posuerant. Postea archiepiscopus intravit Varixium cum centum militibus; et tenuerunt Arsizate et Indunum et Blandronum et hiemaverunt ibi et multum Seprienses oppresserunt.” We limit ourselves to citing Cumanus, ll. 864–5, p. 432, describing the attack by Como at the Isola Comacina, “Descendunt naves, equites, pedites properantes / Iamque parant pedites totum vastare ruentes.” See the text corresponding to note 6, above. The men of Porlezza ask of the consuls of Como “ut sibi statim / Mittant auxilium socios peditesque probatos;” at Vertemate, “Tunc validi pedites sternunt in limine stantes,” Cumanus, ll. 704–5, p. 428, and l. 1229, p. 439, respectively [italics added]. See below note 75 and the corresponding text.

Infantry and Cavalry in Lombardy

71

combatants: “plebeian soldiers” (plebei milites), “a crowd on foot” (turba pedestris) and “a rustic mob” (rustica turba).61 Whatever their military value might be with respect to the cittadini, pedites and milites are at this time certainly also enlisted in the countryside. We know that in 1039 Archbishop Ariberto called all the inhabitants of the diocese of Milan to the defense of the city “from peasant to miles.”62 The same thing must have happened two years earlier against Conrad II. Landolfo mentions in fact “urban pedites” which certainly implies the presence of others who were not.63 The men of the countryside were mobilized all the more in the actions that took place subsequently: in 1066 Erlembaldo, when organizing his expedition to recover the body of Arialdo, called “all those suited for war” to participate, peasants included.64 A few years later at the siege of Castiglione Olona “all the countryfolk of that region” participated.65 The rustici were to be feared, and were not accustomed to distinguishing their opponents by rank – as a “very vigorous miles” slain at Lecco in 1073 by the inhabitants of that place learned to his cost.66 Thanks to the greater availability of the sources we see the active and passive co-involvement of the rural populations become more marked in the first Swabian period, when they were called upon to take part in the battles of both parties in the struggle.67 Together with the Milanese cittadini who in 1158 sought to hold the passage over the Adda river against the imperial army, there are “many peasants” from among whom emerges the figure of a proud “country bishop” armed with a sling.68 Other countryfolk who took refuge in the castle of Trezzo took an

61 62

63 64

65 66

67

68

Cumanus, ll. 1723–48, pp. 450–1. The archbishop “iubet illico convenire ad urbem omnes Ambrosianae parochiae incolas armis instructos, a rustico usque ad militem, ab inope usque ad divitem, ut in tanta cohorte patriam tueretur ab hoste.” Arnulfus, Gesta archiepiscoporum Mediolanensium, ed. Ludwig C. Bethmann and Wilhelm Wattenbach, MGH SS 8 (Hanover, 1848), p. 16. It is they who display as a public laughing stock the corpse of the German warrior killed in battle by Viscount Eriprando. Landulphus, p. 63. “Discurrant itaque nuntii circumquaque velociter, qui omnes ad bella aptos nobiscum cum armis rusticosque hoc venire hortentur cum plaustris, quatinus nobiscum pugnent, isti vero nostra arma et victus deportent.” Andreas abbas Strumensis, Vita S. Arialdi, ed. Friedrich Baethgen, MGH SS 30.2 (Hanover, 1934), p. 1070. “Ad hoc etiam conctos regionis illius instant aggregare ruricolas, diuturnam iuvantes obsidionem.” Arnulfus, Gesta, p. 24. Archbishop Goffredo attempted to occupy the castle of Lecco, but “irruentibus ab urbe militibus violenter eicitur, amisso Marchione illo strenuissimo milite, quem loci incolae rupe praecipitantes ab alta, crudeliter trucidant.” Ibid., p. 26. See the numerous examples from the side of Pavia and Piacenza in Luigi Cesare Bollea, Documenti degli archivi di Pavia relativi alla storia di Voghera (Pinerolo, 1910). From among them we cite only a few examples: “Vidit homines Monticelli et Plebis ad destructionem Mediolani per consules Papie,” doc. 46 (11 November 1184), p. 87; “Item dicit quod vidit missos Placentie in Monticello causa faciendi venire illos de Monticello in expeditione Nuceti cum personis et carriolis,” doc. 55 (11 November 1184), p. 159. Otto Morena, Historia, pp. 48–9; for the episode of the “country bishop” see Vincentius Pragensis, Annales, ed. Wilhelm Wattenbach, MGH SS 17 (Hanover, 1861), pp. 669–70; Carmen de gestis Friderici, ll. 2046 ff, p. 67.

72

Aldo A. Settia

active part in its defense, while in the village of Sale near Tortona it was the peasants (villani), drawn up in formation, who (along with a few milites) victoriously resisted the Milanese.69 Alongside these always more or less effective regular combatants there gathered finally a picturesque assemblage of irregulars operating at the frontier between war waged in bands and outright banditry. Here on one hand the “looters from Pavia” captured by the Milanese while sacking and burning the houses of Siziano and, on the other, the “great company” (magna societas) of the “sons of Arnoldo” who participated in the first phase of the siege of Crema in 1159, a band of poor but dangerous combatants, probably so called in mocking allusion to the poverty preached by Arnoldo of Brescia. Even a local Robin Hood is not lacking, the dreaded Milanese Bagnagatta, “a very audacious predator” who selected the thick woods along the road between Lodi and Pavia as his area of operations and there worked “to capture Germans” and others of the imperial party.70 The inhabitants of Alessandria as a whole enjoyed a reputation as irregulars and marginals, “a multitude of bandits and robbers,” and of fugitive slaves, whom the haughty royal milites adjudged “worthless in combat” (in certamine viles)71 at least before the horrendous and disastrous siege attempted against the city in 1174.

The Cavalry on Foot The cooperation between infantry and cavalry can also be explained in ways other than those examined until now. In 1160, during an umpteenth Milanese effort against Lodi, we know for certain that “pedites as well as milites without horses” participated in the attack.72 It was a case of an engagement that took place near the moat without, however, having scaling the wall as an immediate objective; the same thing had taken place outside Milan in 1037 where the narration of Landolfo already cited gives the very clear impression that the cavalry acted on foot together with the infantry.73 The same impression is obtained from descriptions of other encounters such as those taking place in 1118 between Milan and Como in the fields of Rebbio,74 and in 1191 at Rudiano between the Cremonese and Brescians. Here the singer of the exploit assures us that the ag69 70

71 72 73 74

Otto Morena, Historia, pp. 66 and 25 respectively. Otto Morena, Historia, pp. 28, 73 and 127 respectively. On the “sons of Arnaldo” and the historiographic debate about them see Aldo A. Settia, “‘Kremam Kremona cremabit.’ Esperienze d’Oltremare e suggestioni classiche nell’assedio del 1159,” in Crema 1185. Una contrastata autonomia politica e territoriale (Crema, 1988), p. 78; idem, Comuni in guerra. Armi ed eserciti nell’Italia delle città, pp. 261–76. Respectively, Chronica regia Coloniensis, ed. Georg Waitz, MGH SS 13 (Hanover, 1880), p. 126, and Gotifredus Viterbiensis, Gesta, l. 896, p. 34. Otto Morena, Historia, p. 111. See note 17, above, and the corresponding text. Cumanus, ll. 37–90, pp.414–15; see also note 82, below, and the corresponding text.

Infantry and Cavalry in Lombardy

73

gressors were defeated by the “Brescian milites” while “our worthy pedites were fighting along with them, all together as one.”75 Besides, the already mentioned bas-relief of the Porta Romana shows precisely what we can consider a unit of Milanese pedites on the march preceded by two milites – probably with command functions76 – also on foot. One must, therefore, admit that in northern Italy – as is contemporaneously attested for other regions of Europe77 – the dismounted horsemen sometimes operated in the open field next to the pedites. As is obvious, they must also have come down from their mounts in the course of numerous sieges. It is enough to note the princes and horsemen, “Teutons and Lombards,” who in the winter of 1159 moved “manfully” (viriliter) to the attack of Cremona on the rickety bridge of a mobile tower.78 It is less obvious to observe the same action on other occasions. In 1158, without waiting for orders, the German count Erchempert heads toward Milan on his own and is intercepted near Chiaravalle by the soldiery (militia) of Milan, who are on reconnaissance. Then Erchempert – reports Vincent of Prague – “gets down from his destrier to fight.”79 It is certainly not credible that he would get down and face on foot adversaries who remained in the saddle so one must believe that the Milanese fought in the same way. More than one reason could have suggested such behavior to the cavalry. If one thinks in the first place of the broken landscape of the Lombard plain, then still dominated by a thick natural vegetation, and here and there broken up by areas of cultivation with ditches and rows of grapevines or hedgerows, and in either case therefore unsuitable for the use of mounted men, any contenders on it were constrained (as happened in the following century) to select by common agreement places in which nothing would impede the maneuvers of the cavalry.80

75

76

77 78 79 80

Federico Odorici, “Codice diplomatico bresciano,” in idem, Storie bresciane dai primi tempi sino all’età nostra 6 (Brescia, 1865) doc. 180 (a. 1191), p. 69; see also, below, the text corresponding to notes 87 and 89; see also, idem, “La battaglia di Rudiano detta di malamorte (a. 1191),” Archivio storico italiano, n.s., 3.2 (1856), pp. 3–28; Bortolo Belotti, Storia di Bergamo e dei Bergamaschi 1 (Bergamo, 1959), pp. 388–9. The correct interpretation of the figures depicted in the bas relief as “two of our milites” followed by “simple soldiers” is already in Giulini, Memorie 3, p. 710 (note 11, above); see also Barni, “La lotta,” p. 85 (note 44, above). It is not unusual to see groups of pedites framed by milites. See, for example, Il “Registrum magnum” del comune di Piacenza 1, ed. Ettore Falconi and Roberta Peveri (Milan, 1984) doc. 201 (27 February 1198), p. 427. The commune of Piacenza defended the castle of Bargone, sending sixty infantry “ … armatos de osbergis et panzeriis at alios centum quinquaginta, et eciam usque in duobus centum, et unum militem qui preesset illis” (italics added). Gaier, Art et organisation militaire, pp. 191–4. Otto Morena, “Historia, p. 89–90; see also Settia, “Kremam Kremona cremabit,” pp. 71–2. “Ipse vero nobilis princeps Herkenbertus … cum eis pugnaturus de suo descendit dextrario,” Vincentius Pragensis, Annales, p. 671. See, for example, Rolandinus, Chronica (note 14, above): milites of the two sides in the war agreed to clash “in planam viam, ” p. 118; the work of clearing the ground in expectation of a cavalry engagement, pp. 115, 135.

74

Aldo A. Settia

The ever more sophisticated use of terrain for tactical and even strategic purposes, with a prevalently defensive mentality, has come to be considered a salient characteristic of the wars conducted in Italy in the Po River region in the later communal age.81 To the contrary, this turns out to have been in current use since the first decades of the twelfth century. In the summer of 1118 at Rebbio, where they had been defeated the day before, the Milanese silently occupied “ditches and ruins, fields and ravines, and places for battle” before daylight. For the Comasche who arrived on the field too late there did not remain “any place suitable in which to lie in wait” and to be able adequately to resist their enemy. As a result of the clash “between woods and hills” cavalry and infantry fell on both sides.82 From then on the mounted warriors had to adapt themselves to the systematic exploitation of natural and artificial obstacles by, if necessary, giving up riding. There is a third motive – and not the least important – that could sometimes induce cavalry to fight on foot like the infantry. The horse, a formidable means by which to land deadly charges against the enemy, serves equally well to remove one from danger by a rapid and shameful flight. The horseman, childishly proud of his shining armor and colorful pennoncels, was not in fact always inclined to really risk his life, so that at delicate moments the temptation became strong to unilaterally dissolve the contract that bound him to the foot soldier, leaving the latter to die alone on the battlefield. The documented cases of such behavior are not few. In 1118 the entire Milanese army was put to flight by the Comaschi so that “afterwards the equites flee, leaving the pedites to death.” Many of the latter were captured alive, but more than one thousand actually lost their lives.83 In 1155 the Pavians were surprised by the Milanese, and at once “the milites fled, their foot having been left behind.” One hundred infantry were taken and more than twice that many remained behind on the field.84 The following year the cavalry of Bergamo abandoned their infantry at Palosco; the number of dead and prisoners taken here is also

81

82

83 84

See, for example, Pieri, “Alcune quistioni,” pp. 589–90 (note 42, above); François Menant, “‘Fossata cavare, portas erigere.’ Le rôle des fossées dans les fortifications médiévales de la plaine padane,” in Mélanges d’ archéologie et d’histoire médiévales en l’honneur du doyen Michel de Boüard (Geneva, 1982), pp. 284–85, with the authors there cited; the same article also appears in Aevum 56 (1982). Cumanus, ll. 78–84, p. 415. “sed incauti veniunt ad proelia tardi / Comenses equites fidentes, non dubitantes. / At Mediolanenses taciti multumque timentes / Ante diem fossas ceperunt et ruinosas. / Hesternae pugnae capiunt loca plurima dense, / At cives tardi veniunt ad proelia cauti. / Aptaque nulla vident loca quae sint, unde repugnent / Ut possunt instant illis, pro tempore pugnant. / De scopulis cernunt; silvas montesque tenebant. / Hostibus instabant, et corpora multa iacebant. / Hinc atque inde cadunt equites, peditesque resolvunt.” “Post equites fugiunt, pedites in morte relicti. / Sunt plures tenti, sed sunt plus mille perempti.” Cumanus, ll. 110–11, p. 416. Gesta Federici, p. 21.

Infantry and Cavalry in Lombardy

75

huge.85 Finally, it should be remembered that in the first phase of the battle of Legnano a large part of the cavalry fled as far as Milan, and the day was saved by the capacity for resistance shown by the infantry.86 A very similar if less wellknown event would be in 1191 at Rudiano where in the face of a Cremonese attack in force the Brescian cavalry “fled shamefully, leaving behind their citizens who fought boldly” and only thanks to the commitment of the “worthy pedites” did the fight, in the end, turn into a disaster for the aggressors.87 Given the frequency of such episodes one may wonder if they could always be caused by purely military considerations. In may cases, the behavior of the cavalry was effected by genuine cowardice; but we know of analogous examples of defection from the field both before and after the times with which we are dealing88 which legitimizes the suspicion that sometimes the flight of the cavalry had a political significance. The internal disputes, forcibly quieted in face of external menaces, could in fact express themselves in the form of a desertion, as much ignoble as self-wounding, that allows the quicker to save their own precious lives and at the same time pursue their vendetta.89 In order to escape from the eventuality of being abandoned to themselves just at the moment of greatest danger, the communal Italian infantry, as happened elsewhere,90 could sometimes demand that at least a part of the cavalry would face the enemy on foot with them. 85

86

87 88

89

90

“Pergamei socio disperso robore cedunt / Mox equites, nam fata vident urguere maligna,” but the infantry resist until “Brixia castra petit gaudens bellumque reliquit / Captis quingentis, aliquot quoque marte peremptis.” Carmen de gestis Frederici, ll. 1169–70, 1203–4, pp. 39, 40. On this battle see also Belotti, Storia di Bergamo, 342–3. “Imperator vero milites qui erant ex una parte iuxta carocerum fugavit, ita quod fere omnes Brixienses et de ceteris pars magna fugerunt usque Mediolanum et pars magna de melioribus Mediolanensibus. Ceteri steterunt iuxta carocerum cum peditibus Mediolani et viriliter pugnaverunt,” Gesta Federici, p. 63; see also note 43, above. See note 75, above, and the corresponding text. A similar case is that of the battle on the Brenta in September 899 where a part of the army of Berengar I which faced the Hungarians, “ … Hungariis non solum pugnam non inferebant, sed, ut proximi caderent, anhelebant; atque ad hoc perversi ipsi perverse fecerant, quatinus, dum proximi caderent, soli ipsi quasi liberius regnarent. Qui dum proximorum necessitatibus subvenire neglegunt eorumque necem diligunt, ipsi propriam incurrunt. Fugiunt itaque Christiani, saeviuntque pagani … .” Liudprand of Cremona, Antapodosis in Die Werke Liudprands von Cremona. Opera Omnia, 3rd ed., ed. Joseph Becker, MGH SS in usum scholarum 41 (Hanover and Leipzig, 1915), pp. 44–5. Similarly, at San Procolo in 1275 the Bolognese cavalry abandoned their infantry on the field “some said because of their cowardice, and some said because the people of Bologna treated the nobles badly, and the nobles were content to leave the people to the said danger.” Giovanni Villani, Cronaca dall’anno MCCLXXX al MCCCXI, (Venice, 1833) bk. 7, chap. 48, p. 129. The poet who celebrated the battle of Rudiano had very hard words against those who effectively abandoned the combat. “Qui redere timuerunt—semper debent despici, / Et sint viles et abiecti—sicut tabernarii; / Non honor sublimentur—sed sint semper infimi, / Verecundi atque tristes—homines vilissimi. / O qua fronte te videre—possunt, bona Brixia, / Qui pugnare noluerunt—pro tam dulci patria, / Et se ipsos relinquerunt—suaque omnia; / illos enim manet sola—verecundia.” Odorici, Codice, p. 69 (note 75, above). Gaier, Art et organisation militaire, p. 196. Besides, Narses at Taginae had the Lombards and Heruli dismount from their horses “so that if during the battle they had conducted themselves

76

Aldo A. Settia

On the one hand, the frequently unheroic behavior of the warriors on horseback invites us to resize the halo of exaggerated prestige that an immense literary production wove around their image while, on the other hand, poetry confers on them a greater humanity and, as a result, renders less deep the gap that also separates the combatants on horseback from those on foot in action. There is, however, a final traditional element, regarding this time the skill of the cavalry in the use of certain weapons, which deserves to be at least quickly reviewed.

The Marksmen: a Prejudice If we place ourselves under the walls of Milan in 1037 to be present at the conflict between the cittadini and the royal troops of Conrad II, we see first a defensive phase in which on both sides the use of missile weapons prevails decisively. Our side is now on the attack, “one brandishing a sword,” Landolfo relates, “the other grasping a lance; some hitting with arrows, others are killing enemies drawn to them with iron hooks.”91 The text, as already seen, makes it impossible to distinguish the actions of the infantry from those of the cavalry and allows us to understand that in crisis the same arms would be used indifferently by all the combatants. The importance of the marksmen, the arms used by them, and their effects on the tactical level constitute an important chapter in the history of the techniques of combat which we can only touch on here. Certainly, archers, slingers and crossbowmen, i.e., the specialists in the use of arms that strike from a distance, belong in general to the lower social strata, and there is a deep-rooted conviction that the horseman would always shun them by reason of the chivalric ethos.92 This is, however, an opinion that should be looked at again, at least for the time before the last decades of the twelfth century. Not only does Godfrey of Bouillon handle a crossbow with deadly skill in the Aachen chronicle, which tells of the events of the First Crusade, 93 but closer to this inquiry in time and space, Frederick

91

92 93

badly and had had intentions of throwing down their arms, as could have happened, it would not have been easy for them to give themselves to flight.” Procopius of Caesaria, La guerra gotica, 4.31, in idem, Le guerre persiana, vandalica, gotica, ed. M. Craveri and Filippo M. Pontani (Turin, 1977), p. 750; for Greek and English text see Procopius, A History of the Wars, Books, 5–8. The Gothic Wars, ed. and trans. H. B. Dewing, Loeb Classical Library 431–3 (Cambridge MA, 1919–28; rept. 1962) 8:31.5, p. 371. “Propterea adunati valdeque constricti hostes obnixe iam secure irrumpentes, alius ensem regens, alius lanceam manu tenens, Teutonicos et ceteros hostes feraliter percutiebant; alius sagittis ceterisque tormentis ipsos percutiendo atque minuendo perterrefaciebat, alius ferreis uncinis hostes quos ad se trahere poterat, trucidabat.” Landulphus, p. 62. See in general A. T. Hatto, “Archery and Chivalry: A Noble Prejudice,” Modern Language Review 35 (1940), pp. 40–50. Faced during an action with a Turk who caused heavy injuries in the Christian ranks, Godfrey “arrepto arcu baleari et stans post scuta duorum sociorum, eundem Turcum trans vitalia cordis percuti,” killing him (Albert of Aachen consistently indicates a crossbow with “arcus balearis”).

Infantry and Cavalry in Lombardy

77

I is presented on several occasions as an excellent marksman. In 1158 at the siege of Milan he completed a turn of reconnaissance around the walls and, to check whether or not defenders were installed in an ancient monument, “he braced a crossbow on his arm, placed himself below the monument, and after firing an arrow from a long distance persuaded himself that the enemy was lying in wait at the top.”94 In January 1160, when the walls of Crema were overhung by a gigantic mobile tower, the same emperor, “who knew very well how to shoot,” as Otto Morena reports, killed many of the besieged with his own hand.95 The deed, carried out with complete naturalness, does not betray either prejudice against these weapons that could be considered “unfair,” or crises of conscience for the ecclesiastical interdictions that only a few years before had forbidden their use.96 Here, very likely, we must once again take account of the importance assumed by the work of Vegetius and of the possibility that it suggested models of behavior. The portrait of Frederick seems in fact to have features in common with the anonymous “unconquered emperor” to whom Vegetius dedicated his Epitoma. He, who could teach exercises in arms to his instructors, awoke envy for the “skill and elegance of his riding;” Saracens and Indians were unable to defeat him in a race, and the very Persians admire his “skill in shooting.”97 If at that time a complete warrior education, not yet a chivalrous and courteous one, 98

94

95 96

97

98

Albertus Aquensis, Historia Hierosolymitana, in Recueil des historiens des croisades. Historiens occidentaux, 4 (Paris, 1879), p. 324; for an English translation see Albert of Aachen, Historia ierosolimitana. History of the Journey to Jerusalem, ed. and trans. Susan B. Edgington (Oxford, 2007). “Ipse etiam sumpto ductor subit impiger arcu / Letiferamque procul iaciens per inane sagittum / Percellit superis residentem sedibus hostem,” Carmen de gestis Frederici, ll. 2399–401, p. 80. “Ipsemet quoque imperator, optime noscens sagittare, ut extimo, multos de Cremensibus interfecit,” Otto Morena, Historia, p. 92. Paul Fournier, “La prohibition par le IIe concile Lateran d’armes jugées trop meurtrières (1139),” Revue générale de droit international public 23 (1916), pp. 471–9; see also Settia, “Kremam Kremona cremabit,” p. 75; idem, Comuni in guerra, pp. 261–76. Digesta sunt, imperator invicte, quae nobilissimi auctores diversis probata temporibus per experimentorum fidem memoriae prodiderunt, ut ad peritiam sagittandi, quam in serenitate tua Persa miratur, ad equitandi scientiam vel decorem, quae Hunnorum Alanorumque natio velit imitari, si possit, ad currendi velocitatem, quam Saracenus Indusque non aequat, ad armaturae exercitationem, cuius campidoctores vel pro parte exempla intellexisse gaudent, regula proeliandi, immo vincendi artificium iungeretur, quatenus virtute pariter ac dispositione mirabilis reipublicae tuae et imperatoris officium exhiberes et militis,” (italics added) Epitoma 3.26, pp. 124–5. Nothing certain is known of the education received by Frederick I; see Franco Cardini, Il Barbarossa. Vita, trionfi e illusioni di Federico I imperatore (Milan, 1985), p. 79. That he knew Vegetius is not specifically attested (see, however, note 13, above), but it is at least quite probable. On the later integration of warrior values and courtly values see, e.g., Alessandro Barbero, “La corte dei marchesi di Monferrato allo specchio della poesia trobadorica. Ambizioni signorili e ideologia cavalleresca fra XII e XIII secolo,” Bollettino storico-bibliografico subalpino 81 (1983), 683–9.

78

Aldo A. Settia

comes from sporting skills and military demands, the race and the use of distance weapons were included – i.e., skills suited to a good infantryman – and this points out another factor apt to make the undertaking of actions in common more easy and to stimulate a reciprocal emulation on the field of battle. Above all, when social and ideological motives induce equites and pedites to line up one against the other, it creates fractures that not even charismatic captains like Ariberto and Erlembaldo had succeeded in composing and which had to as a result lead to the definitive crisis of communal civilization.

4 Unintended Consumption: The Interruption of the Fourth Crusade at Venice and Its Consequences Gregory D. Bell

In the fall of 1202, the soldiers of the Fourth Crusade and their Venetian allies sacked the Christian city of Zara instead of sailing directly to fight Islam in the east. The crusaders owed a substantial debt to the Venetians, who had prepared a large fleet to transport the soldiers and horses and organized provisions for the army. Historians have concluded that too few crusaders – somewhere between 10,000 and 12,000 of the anticipated 33,500 soldiers – arrived at Venice, which was the designated rendezvous for the gathering forces. Therefore, the crusaders agreed to a “diversion” to Zara in order to postpone payment of their debt to the Venetians and thereby ensure that the crusade would continue.1 But this interpretation may be wrong on two counts. There may have been more crusaders than has generally been thought, and logistical considerations, particularly the duration of the crusaders’ stopover at Venice, left the army with little choice but to move on from the Italian city at the onset of winter. This article revisits the question of the size of the crusader army at Venice in the summer of 1202 and explores the consequences of feeding the crusaders as they waited to be taken to the Levant. It may be that the crusader army in Venice was too small to pay its debts but too large for the Venetians to sustain.

The comments of the editors of the Journal of Medieval Military History, especially Professor Clifford Rogers, helped focus my estimation of the size of the army, particularly the section on the number of horses. They also guided me to Ramón Muntaner’s description of his company billeting over the winter, discussed below. Joseph Shatzmiller, Alex Roland, and Kenneth S. Friedman also read and commented on versions of this paper. To these scholars I offer my sincerest gratitude. Any remaining errors are mine. 1

Donald E. Queller and Thomas F. Madden, The Fourth Crusade: The Conquest of Constantinople, 2nd ed. (Philadelphia, 1997), pp. 47, 55; Jonathan Riley-Smith, The Crusades: A Short History (New Haven, 1987), p. 125; Jonathan Phillips, The Fourth Crusade and the Sack of Constantinople (London, 2004), p. 106; Edgar H. McNeal and Robert Lee Wolff, “The Fourth Crusade,” in The Later Crusades, 1189–1311, eds. Robert Lee Wolff and Harry W. Hazard, A History of the Crusades, ed. Kenneth M. Setton, vol. 2 (Madison, 1969), pp. 153–86 at pp. 166–7. Queller and Madden said that 10,000 to 12,000 soldiers participated in the expedition, while Riley-Smith maintained that about one-third of the anticipated 33,500 troops joined. Phillips set the figure at 12,000 soldiers, and McNeal and Wolff said between 10,000 and 12,000 likely participated.

80

Gregory D. Bell

Primary Sources and the Size of the Army of the Fourth Crusade The chroniclers of the Fourth Crusade did not clearly specify the size of the army. Historians have relied primarily on Geoffrey of Villehardouin, who participated in the expedition. Villehardouin knew the leaders of the crusade and acted as one of the envoys to Venice when the original contract was made between the crusaders and the Venetians. The envoys commissioned a fleet for 33,500 men, including “4,500 horses and 9,000 squires, and other ships to accommodate 4,500 knights and 20,000 foot sergeants.”2 These figures were clear, but apparently optimistic. When later describing the Venetian fleet, Villehardouin said that “It comprised so great a number of warships, galleys, and transports that it could easily have accommodated three times as many men as were in the whole of the army.”3 As noted above, historians have interpreted Villehardouin’s report as supporting the conclusion that about one-third of the expected 33,500 troops actually reached Venice, or around 11,667 soldiers. A second chronicler, Robert of Clari, also provided information about the size of the army. Like Villehardouin, Robert of Clari participated in the Fourth Crusade.4 Though a knight, he was not among the top tier of noblemen. Crusade historians consider him to be the voice of the everyday soldier of the Fourth Crusade. He offered descriptions of the places visited and expressed the emotions of the crusaders, but, as Donald E. Queller and Thomas F. Madden have noted, Robert of Clari “had no head for numbers or dates, so he should never be relied upon for amounts of money, numbers of men, or chronology without the greatest caution.”5 Fewer than one thousand knights reached Venice, according to Robert of Clari.6 Two fourteenth-century Venetian chronicles gave more precise figures than Robert of Clari or Villehardouin. Andrea Dandolo said that 4,500 horsemen and 8,000 foot soldiers embarked on crusade for a total of 12,500 troops, and the Venetiarum historia vulgo petro iustiniano iustiniani filio adiuidicata, hereafter referred to as the Venetiarum historia, said that Venice transported 5,000 horsemen and 8,000 foot soldiers, or 13,000 troops.7 The overall

2

3 4 5 6 7

Geoffrey of Villehardouin, La Conquête de Constantinople, ed. and trans. M. Natalis de Wailly, 2nd ed., (Paris, 1874, online reprint via www.books.google.com), 5.21, p. 14; English quotation from Villehardouin, Chronicles of the Crusades, trans. M. R. B. Shaw (Baltimore, 1963), p. 33. Villehardouin, La Conquête, 12.56, pp. 32, 34; English quotation from Villehardouin, Chronicles, p. 42. Robert of Clari, The Conquest of Constantinople, trans. Edgar Holmes McNeal (Toronto, 1996), p. 4. Queller and Madden, The Fourth Crusade, pp. 43–4. Robert of Clari, La Conquête de Constantinople, ed. Philippe Lauer (Paris, 1924), p. 10. Andrea Dandolo, “Chronica per extensum descripta,” Rerum italicarum scriptores, 12, ed. Lodovico Antonio Muratori (Mediolani, 1728), p. 320. Venetiarum historia vulgo petro iustiniano iustiniani filio adiuidicata, Monumenti Storici, 18, eds., Roberto Cessi and Fanny Bennato (Venice, 1964), p. 134.

The Interruption of the Fourth Crusade

81

number of soldiers comports with Villehardouin’s estimate of 11,667. These figures make it possible to determine a rough approximation of the number of knights and squires in the army of the Fourth Crusade. The Venetian sources said that 4,500 to 5,000 horsemen were transported. Andrea Dandolo and the Venetiarum historia, however, seem more concerned with the number of individuals and horses that the Venetians actually had to transport than distinguishing between knights and squires. Hence they use the generic terms equites, which suggests horse-soldiers or cavalry, not necessarily knights, and pedites. In comparison, both Villehardouin and Robert of Clari distinguished between knights and squires. Clari, as mentioned above, spoke of fewer than one thousand knights. Since he is unreliable, a slightly higher estimate of 1,500 knights based on Villehardouin’s vague notion of a third might be applied, suggesting an army comprising about 1,500 knights and 3,000 squires.8 That is 4,500 horsemen, as is found in Andrea Dandolo. Of course this is an approximation, but it clarifies the accounts. Robert of Clari might not have been too far off when he said that there were not quite one thousand knights in Venice. In order to determine a more precise estimate of the size and composition of the crusade army, however, the overall debt owed to the Venetians must first be established. The original debt was settled during the negotiations between the crusade envoys, including Villehardouin, and the Venetians. Villehardouin offers specific numbers when describing the initial treaty. As noted above, the Venetians agreed to supply transports to carry 4,500 horses and 33,500 troops.9 In addition, they also agreed to provide nine months’ provisions for the soldiers and horses, although the treaty itself clarifies that the provisions were to last up to one year (ad unum annum).10 The cost was set at four marks per horse and two marks per soldier, totaling 85,000 marks.11 These were the clearest figures that

8 9

10

11

The figure of 3,000 squires is based on the original treaty, which provided for twice as many squires as knights. Villehardouin, La Conquête, 5.21, p. 14. Villehardouin distinguished between transport ships, which would carry the horses and squires, and regular ships, which would carry the knights and foot sergeants. Ibid., 5.21, p. 14; “Pactum domini Balduini comitis Flandrensis, et Theobaldi comitis Trecensis, et Lodovici comitis Blesensis, factum cum prefato domino Henrico Dandulo, Duce Venetie, pro passagio terre sancte,” in Urkunden zur älten Handels- und Staatsgeschichte der Republik Venedig, 1, eds. G. L. Fr. Tafel and F. M. Thomas, no. 92 (Amsterdam, 1964), p. 365. Villehardouin, La Conquête, 5.21–22, p. 14. As the documentation suggests, it was much more expensive to transport and sustain a horse than a soldier. There have been a number of studies on how much food and water individual soldiers, mounts, and even pack horses and mules required for sustenance on the march. For information on the amount of food and water individual soldiers consume on a daily basis, see Donald W. Engels, “Apendix I,” Alexander the Great and the Logistics of the Macedonian Army (Berkeley, 1978), pp. 123–30. For horses and other types of mounts, see John Haldon, “Roads and Communications in the Byzantine Empire: Wagons, Horses, and Supplies,” Logistics of Warfare in the Age of the Crusades, ed. John H. Pryor (Burlington, VT, 2007), pp. 144–6 and Bernard S. Bachrach, “Some Observations on the Military Administration of the Norman Conquest,” Chapter 14, Warfare and Military Organization in Pre-Crusade Europe (Burlington, VT, 2002), pp. 1–25, especially pp. 12, 14–15.

82

Gregory D. Bell

Villehardouin offered. The crusaders may have paid some of this obligation before their arrival at Venice. According to Villehardouin, the envoys borrowed 2,000 marks from the people of Venice “so that the building of the fleet might begin.”12 This payment probably did not alter the overall debt of 85,000 marks, as the crusaders borrowed the 2,000 marks from the Venetians. The crusaders, therefore, owed the Venetians a substantial sum. During the summer of 1202, as too few crusaders arrived, the expedition’s leaders must have worried that they faced a potential financing crisis of major proportions. The leaders’ anxieties pervade Villehardouin’s account, which contains numerous references to crusaders who embarked from other ports, thereby leaving the main army at Venice with too few soldiers. It was at this point, when it appeared that fewer soldiers than anticipated had arrived, that it was decided to levy a fee on each individual and horse as outlined in the original contract. Once the fees were collected, the apprehensive crusade leaders could see where things stood. Villehardouin stated that during this levy “every man in the army was called upon to pay his passage,” and that the barons took “what money they could get” from those who could not pay in full. Rather than specify the total amount that the crusaders actually paid, Villehardouin said that “the money collected did not amount to half – much less the whole – of the sum required,” this despite the fact that everyone in the army contributed.13 So according to Villehardouin, the crusaders were able to raise less than 42,500 marks during this initial levy. Robert of Clari’s figures were more specific. He said that each knight was to pay four marks for himself. Each horse cost four marks as well. Squires, or mounted sergeants, were to pay two marks apiece, and those who could not pay the full amount should pay one mark.14 Of course Robert was often inaccurate when it came to numbers and dates, but in this case perhaps his figures should be taken more seriously. The payment was not some abstract concept for Robert, it was a fee that he was asked to pay himself, and his figures look similar to those that Villehardouin remarked were part of the original agreement. According to Robert, not only were the knights to pay for their horses as anticipated, but they also paid two marks more than was expected of them in the contract while the poorer crusaders were permitted to pay less than the standard fare. If Robert was correct in his recollection, then the wealthier crusaders helped pay for their impoverished counterparts. When all the money was collected and paid, Robert of Clari said the crusaders still owed the Venetians 50,000 marks.15 The crusaders were therefore able to pay 35,000 marks toward the total amount owed, according

12

13 14 15

Villehardouin, La Conquête, 6.32, p. 20; English quotation from Villehardouin, Chronicles, p. 35. There are discrepancies among the manuscripts concerning the amount the crusaders borrowed. For example, Shaw’s translation said that the Venetians loaned 5,000 marks to the crusaders so that they might make initial payments on the fleet. Villehardouin, La Conquête, 12.58 p. 34; English quotation taken from Villehardouin, Chronicles, p. 42. Robert of Clari, La Conquête, p. 10. Ibid., p. 10.

83

The Interruption of the Fourth Crusade

to Robert of Clari. Based on these figures, the crusaders’ initial payment was something less than 42,500 marks, or half of the total debt of 85,000 marks, but apparently at least 35,000 marks. Since the sources suggest that everyone in the army was asked to make a set payment toward the debt, it is possible to come to a general estimate of the size of the army of the Fourth Crusade during the summer of 1202. Using the numbers of four marks per horse and two marks per person from Villehardouin’s contract, an estimated 1,500 horses, and 4,500 knights and squires based on Andrea Dandolo’s reference to 4,500 “horsemen” being transported, the knights and squires would have had to pay 9,000 marks for themselves and 6,000 marks for their horses, or 15,000 marks total. This would leave 20,000–27,500 marks for other participants. If all of the crusaders paid two marks per person, then there were 10,000 to 13,750 foot soldiers and non-combatants.16

Chart 1 Size of Army with 1,500 horses / Total payment of 35,000 marks Based on pay scale Villehardouin described (four marks per horse and two marks per soldier)

Size of Army with 1,500 horses / Total payment of 42,500 marks Based on pay scale Villehardouin described (four marks per horse and two marks per soldier)

Number of Crusaders and Horses

Number of Crusaders and Horses

Cost

4,500 knights and squires at two marks apiece 1,500 horses at four marks apiece 13,750 foot soldiers at two marks apiece

9,000 marks

4,500 knights and squires at two marks apiece 1,500 horses at four marks apiece 10,000 foot soldiers at two marks apiece

Cost 9,000 marks

6,000 marks 20,000 marks

16

27,500 marks

Totals:

Totals: 14,500 crusaders 1,500 horses

6,000 marks

35,000 marks

18,250 crusaders 1,500 horses

42,500 marks

See Chart 1. As mentioned above, the total payment made to the Venetians after the first levy was 35,000 marks – based on Robert of Clari’s assertion that they still owed the Venetians 50,000 marks after making the initial payment (85,000 - 50,000 = 35,000) – up to 42,500 marks, as Villehardouin said that the crusaders were able to pay less than half of the 85,000 marks.

Gregory D. Bell

84

Chart 1A Size of Army with 1,500 horses / Total payment of 35,000 marks Based on pay scale Villehardouin described (five marks per horse and two marks per soldier)

Size of Army with 1,500 horses / Total payment of 42,500 marks Based on pay scale Villehardouin described (five marks per horse and two marks per soldier)

4,500 knights and squires 1,500 horses at five marks apiece 9,250 foot soldiers at two marks apiece

4,500 knights and squires 1,500 horses at five marks apiece 13,000 foot soldiers at two marks apiece

9,000 marks 7,500 marks 18,500 marks

Amended Totals: 13,750 crusaders 1,500 horses

9,000 marks 7,500 marks 26,000 marks

Amended Totals: 35,000 marks

17,500 crusaders 1,500 horses

42,500 marks

Chart 2 Size of Army with 1,500 horses / Total payment of 35,000 marks Based on pay scale Robert of Clari described (Knights pay four marks, others pay two marks, and horses cost four marks)

Size of Army with 1,500 horses / Total payment of 42,500 marks Based on pay scale Robert of Clari described (Knights pay four marks, others pay two marks, and horses cost four marks)

Number of Crusaders and Horses

Number of Crusaders and Horses

1,500 knights at four marks apiece 3,000 squires at two marks apiece 1,500 horses at four marks apiece 8,500 foot soldiers at two marks apiece

Cost 6,000 marks 6,000 marks 6,000 marks 17,000 marks

Totals: 13,000 crusaders 1,500 horses

1,500 knights at four marks apiece 3,000 squires at two marks apiece 1,500 horses at four marks apiece 12,250 foot soldiers at two marks apiece

Cost 6,000 marks 6,000 marks 6,000 marks 24,500 marks

Totals: 35,000 marks

16,750 crusaders 1,500 horses

42,500 marks

The Interruption of the Fourth Crusade

85

Using Villehardouin’s figures, the army would have been between 14,500 and 18,250 crusaders total including knights and squires. Some manuscripts, however, say that the contract called for five marks per horse, not four, and if the crusaders paid an extra mark for every horse, the army would be slightly smaller in size, totaling some 13,750 to 17,500 participants.17 Robert of Clari provided a different set of dues for the crusaders during the first levy. His comment that those who could not pay the full amount should pay one mark might suggest that other than the knights, all crusaders – not just squires – were asked to pay two marks as Villehardouin said. Taking this possibility into account, Robert of Clari’s numbers suggest an army size similar to that found in the Venetian sources and Villehardouin. There would be from 8,500 to 12,250 foot soldiers and non-combatants in the army for a total of 13,000 to almost 17,000 crusaders.18 Based on these estimations, it is entirely possible that the army of the Fourth Crusade comprised more than 12,000 individuals during the summer months. These numbers are at best approximations. It is clear that not everyone paid the full fare. Villehardouin and Robert of Clari both said that some people simply contributed what they could. So the army was probably somewhat larger than the figures given here suggest. For example, if Robert of Clari’s assertion that anyone who was unable to pay the regular fare should contribute a mark is applied to all of the foot soldiers and non-combatants, the size of the army balloons.19 However, none of the sources support the possibility of an army of almost 30,000 crusaders being transported. Such an army would have filled the entire Venetian fleet; there would not have been space for three times as many crusaders, as Villehardouin stated there was. On the other end of the spectrum, there could have been individuals who paid more than their fair share during the initial levy. This seems less likely. The only way for the crusader leaders to determine exactly what resources they could marshal to compensate the Venetians was to have each person pay what was expected, or what they could, and see what was left over. This appears to be exactly what the crusaders did. The number of horses was another variable in determining the size of the army. The crusaders might well have brought more than 1,500 mounts. The original contract only allowed for a single horse per knight, and so bringing more horses was possible after too few crusaders arrived outside Venice. Knights in the early thirteenth century normally brought multiple horses with them. If the animals were not being shipped, then the knight would usually have at least a warhorse, a riding horse, and a pack animal. Further, others might also have been labeled “horsemen” in the Venetian sources. Robert of Clari distinguishes between “knights” and “mounted sergeants.”20 Even priests might have brought 17 18 19 20

See Chart 1A. See Chart 2. See Chart 3. Robert of Clari, La Conquête, p. 10: “si eswarderent entr’aus que cascuns chevaliers donroit quatre mars, et cascuns chevax quatre et cascuns serjans a cheval dues mars … ”

Gregory D. Bell

86

Chart 3 Size of Army with 1,500 horses / Total payment of 35,000 marks Based on pay scale Robert of Clari described (Knights pay four marks, squires pay two marks, horses cost four marks, and others pay one mark)

Size of Army with 1,500 horses / Total payment of 42,500 marks Based on pay scale Robert of Clari described (Knights pay four marks, others pay two marks, horses cost four marks)

Number of Crusaders and Horses

Number of Crusaders and Horses

Cost

Cost

1,500 knights at four marks apiece, 3,000 squires at two marks apiece, and 1,500 horses at four marks apiece

18,000 marks

1,500 knights at four marks apiece, 3,000 squires at two marks apiece, and 1,500 horses at four marks apiece

18,000 marks

17,000 foot soldiers at one mark apiece

17,000 marks

24,500 foot soldiers at one mark apiece

24,500 marks

Totals:

Totals: 21,500 crusaders 1,500 horses

35,000 marks

29,000 crusaders 1,500 horses

42,500 marks

horses or mules with them. The Venetians certainly provided a sufficient number of horse transports if the crusaders decided to bring additional horses.21 Although horses were more expensive to transport than soldiers, it is hard to imagine the leaders of the crusade or other wealthy noblemen and knights taking only one horse if they had the space for more and the money to pay for the transportation of additional mounts. The sources offer modest support for this scenario. Villehardouin and Robert of Clari mentioned that fewer individuals arrived in Venice, but they said nothing of the horses. Both Venetian sources cited above stated that 4,500 or more horsemen were transported. While the squires might have taken care of the knights’ horses rather than bring their own, perhaps the mounted sergeants or priests, mentioned above, account for the discrepancy of 500 horsemen between Andrea 21

See John H. Pryor, “Transportation of Horses by Sea During the Crusades,” Commerce, Shipping, and Naval Warfare in the Medieval Mediterranean, Chapter 5, pp. 21–2, and Pryor, “The Venetian Fleet for the Fourth Crusade and the Diversion of the Crusade to Constantinople,” in The Experience of Crusading, 1, eds. Marcus Bull and Norman Housley (Cambridge, 2003), pp. 103–23, especially pp. 118–19.

The Interruption of the Fourth Crusade

87

Dandolo and the Venetiarum historia. At least one contemporary account spoke of horses in the plural. In a letter to his friend and vassal, Hugh of Saint Pol described how he was so hungry toward the end of the assault on Constantinople that he exchanged his surcoat for bread, but kept his horses and weapons.22 Of course, Hugh of Saint Pol was an important and affluent aristocrat, and, therefore, not typical of the average knight who participated on the Fourth Crusade. However, since the crusaders had sailed to Constantinople and would have filled the horse transports that they did bring with them, Hugh might have purchased additional horses after disembarking at the Byzantine capital or brought multiple horses with him from the start. Or perhaps he did both. Extra horses would alter the calculations of the size of the army. There would be fewer soldiers if the crusaders bought 2,000 horses with them based on the estimate of 1,500 mounts for the knights plus an additional 500 horses for horse sergeants, priests, and knights who brought multiple horses. Using Robert of Clari’s description of the payments made during the first levy (four marks per knight, four marks for each horse, and two marks apiece for the other crusaders), an army with 2,000 horses would number 12,000 to 15,750 rather than 13,000 to 16,750 crusaders.23 In this case, the additional mounts would reduce the size of the army by around 1,000 participants. There was also the matter of Count Thibaut III of Champagne. Count Thibaut, who was one of the original leaders of the Fourth Crusade, died in May of 1201, leaving a large sum of money that was divided into two portions. Some of this money was distributed among his followers. The rest was given to the crusading army. In The Fourth Crusade, Queller and Madden, citing Villehardouin and Robert of Clari, said that the crusaders augmented the sum paid to the Venetians during the levies with Thibaut of Champagne’s money.24 These sources, however, did not say explicitly that Thibaut’s money was used during the first levy. In fact, they suggest the opposite. Thibaut was originally Villehardouin’s lord and trusted the chronicler enough to send him as one of two envoys to negotiate a treaty with the Venetians. Villehardouin, in turn, often lauded Thibaut in his account of the Fourth Crusade. Yet, Thibaut’s name was absent during the description of the initial levy. This was a curious omission. Villehardouin could easily have derided the crusaders who traveled through other ports, as was his want, saying that Thibaut’s generosity was insufficient due to the negligence of others. He might at least have acknowledged the posthumous donation that his lord made. Instead, Villehardouin remained silent. When he did mention Thibaut’s

22

23 24

Count Hugh of Saint Pol, “To R. de Balues” in Rudolph Pokorny, “Zwei unedierte Briefe aus der Frühzeit des lateinischen Kaiserreichs von Konstantinopel,” Byzantion 55 (1985), pp. 180– 209 at p. 203. “Nocte vero precedenti diem, in qua se civitas reddidit nobis, ad tantam inopiam deveneram, quod oportuit me supertunicale meum mittere ad panem, retentis tamen equis et armis.” Italics are mine. See Chart 2. Queller and Madden, The Fourth Crusade, pp. 23, 51, 234–5, n. 83.

88

Gregory D. Bell

money, Villehardouin said that it was split among his followers and companions, and a portion was left “to the army, to be spent there as might seem best.”25 Later, Villehardouin said that the crusaders asked Marquis Boniface of Montferrat to take command of the army in order to replace Count Thibaut and take control of his money and his men.26 The marquis had no obligation to use the money during the first levy. Like Villehardouin, Robert of Clari did not mention Thibaut’s money during the levies. He said that the crusaders gave Boniface “twenty-five thousand livres out of the money which the count of Champagne had left to the crusaders,” and that some of the marquis’s money went toward a 25,000-mark payment to the Venetians, but this assertion remains unsubstantiated.27 Unlike Villehardouin, who was a friend and confidant of Thibaut and Boniface, Robert was far removed from the negotiations between these lords and did not know either one personally. Therefore, it is unlikely that Robert of Clari knew of a 25,000-mark payment while Villehardouin did not. The army, therefore, might well have been larger than historians have thought. Villehardouin and the Venetian accounts of Andrea Dandolo and the Venetiarum historia documented the size of the army at the time of its embarkation in October of 1202. If the Venetian figures and Villehardouin’s impression of the army’s size were close to accurate, then the army contained between 11,667 and 13,000 soldiers. These sources, however, refer to the army in October. Earlier, at the time of the first levy, the army might have been significantly larger. The figures given above suggest that it could have contained more than 18,000 crusaders. The crusade chronicles support the scenario of a diminished army in the fall. Descriptions of the army’s stay clearly describe a shrinking crusader force due to adverse conditions on the Isle of St Nicholas, sometimes called the Lido, where the Venetians had stationed the crusaders. The sources suggest that the living standard for the crusaders gradually deteriorated, and this coincided with a high mortality rate in the camp. Initially, the Venetians established a market for the crusaders, described by Villehardouin “as abundantly supplied as any one could desire with everything necessary for the use of horses and men.”28 Even after the crusaders failed to fulfill their contract with the Venetians after the first levy, the Doge “did not cease from having brought to them [the crusaders] enough to eat and drink,” although he did threaten to stop sending provisions.29 Then the situation changed. The Devastatio Constantinopolitana, an eyewitness account of a crusader, probably from the Rhineland, said that food prices were high, and the poor were trapped and resentful of their leaders. Further, “an unusual mortality

25 26 27 28 29

Villehardouin, La Conquête, 7.36, p. 22; English quotation from Villehardouin, Chronicles, pp. 36–7. Villehardouin, La Conquête, 9.43, p. 26. Robert of Clari, La Conquête, pp. 6, 8; English quotation from Robert of Clari, Conquest, p. 36. Villehardouin, La Conquête 12.56, p.32; English quotation from Villehardouin, Chronicles, p. 42. Robert of Clari, La Conquête, p. 10; English quotation from Robert of Clari, Conquest, p. 40.

The Interruption of the Fourth Crusade

89

rate arose … the dead could barely be buried by the living.”30 Disease might have caused the high mortality rate.31 Queller and Madden suggested that death and illness stemmed from living in foreign, unsanitary conditions.32 Although the Venetians had apparently agreed to supply the expedition with provisions, several sources state the opposite. Villehardouin said that the Venetians provided a “market” that supplied the crusaders, and the Devastatio Constantinopolitana mentioned high food prices. If these sources are correct, then the crusaders were not given provisions outright. They had to trade or even pay for them. Robert of Clari also stated that the crusaders were so poor from paying their debts that they barely had enough resources to survive.33 Perhaps the Venetians only intended to distribute the food and drink after the crusade departed, or they felt justified in charging for food until the crusaders paid the contract. According to Robert of Clari, the Doge warned the crusaders that they would not leave the island of St Nicholas nor receive provisions until the debt was paid.34 Regardless, the crusaders were all asked to pay something toward the overall debt to the Venetians, and, as might be expected, the additional expenditures for food generated hardship and animosity within the crusade army. Desertion also took a toll on the crusade army.35 According to the Devastatio Constantinopolitana, a majority of the poor crusaders staying on the Lido either went home or traveled to other ports.36 Gunther of Pairis, a German chronicler who lauded abbot Martin of Pairis’s role on the expedition, concurred with the Devastatio, saying “many poor persons who had brought little with them [to Venice], and had consumed it” left and went home.37 The poorer crusaders, as well as the sick and women, were enough of a drain on the army to warrant a decision by the papal legate to send them home soon after his arrival at Venice on 22 July 1202.38 A significant percentage of these deaths and desertions probably took place before the crusade leaders collected money towards repaying their debt. Villehardouin was not specific about the exact chronology of the levies. He said that the Venetians, having fulfilled their end of the agreement, were ready to

30

31 32 33 34 35 36 37

38

“Devastatio Constantinopolitana,” Chroniques Gréco-Romanes, ed. Charles Hopf (Berlin, 1873), p. 87; English quotation from “Devastatio Constantinopolitana,” in Contemporary Sources for the Fourth Crusade, ed. and trans. Alfred J. Andrea with contributions by Brett E. Whalen (Boston, 2000), pp. 212–21, at p. 214; Phillips, The Fourth Crusade, pp. 107–8. Phillips, The Fourth Crusade, p. 108. Ibid., p. 108; Queller and Madden, The Fourth Crusade, p. 53. Robert of Clari, La Conquête, p. 11. Ibid., p. 10. Phillips, The Fourth Crusade, p. 107; Queller and Madden, The Fourth Crusade, pp. 53–4. “Devastatio Constantinopolitana,” Chroniques, 87; “Devastatio Constantinopolitana,” Contemporary Sources, pp. 214–15. Gunther of Pairis, Hystoria Constantinopolitana, p. 122; English quotation from The Capture of Constantinople: The “Hystoria Constantinopolitana” of Gunther of Pairis, ed. and trans., Alfred J. Andrea, p. 78. “Devastatio Constantinopolitana,” Chroniques, p. 87.

90

Gregory D. Bell

embark and so approached the counts and barons to pay the debt owed.39 The second levy must have taken place quite late, as Villehardouin mentioned that the Marquis de Montferrat donated with the other lords at this time.40 The marquis did not arrive until 15 August 1202 during the feast of the Assumption of Blessed Mary.41 Queller placed the levies at the end of the summer after most of the crusaders had arrived in Venice.42 His chronology appears to follow Robert of Clari, who said that the Venetians approached the crusaders about payment sometime after the Doge saw that they had all arrived.43 Even though the deaths and desertions among the poorer crusaders appear to have occurred before the crusade leaders called on all participants to contribute to the cause, the repercussions of these departures likely did not have much of an effect on the amount collected during the first levy, as many of these impoverished crusaders had spent all their money earlier in the summer. In August, they would have had little to contribute to the larger cause. If these sources were correct, a number of crusaders, including most of the poor, traveled to Venice and then left without going on to the Holy Land. Thus, by the time the army finally embarked for Zara, it was already diminished, perhaps significantly. It is entirely plausible that during the summer of 1202 the army of the Fourth Crusade was substantially larger than the 13,000 individuals described in the Venetiarum historia.

The Consequences of Sustaining a Large Army at Venice An army of this size would have been an enormous burden on its Venetian hosts, as the crusaders would have to be fed. In “Food and the Fourth Crusade,” Madden points out that the Venetians were required to provide a predetermined quantity of provisions over the course of the year.44 If this is correct, then the Venetians would fulfill their end of the contract once the crusaders consumed the collected provisions and wine. Contractually, the Venetians were to provide the crusade six sextaria of bread, flour, grain, and legumes and a half amphora of wine for each individual, and three modia for each horse using Venetian

39 40 41 42 43

44

Villehardouin, La Conquête, 12.57, p. 34. Ibid., 12.61, p. 36. “Devastatio Constantinopolitana,” Chroniques, p. 87. Donald E. Queller, The Fourth Crusade: The Conquest of Constantinople, 1201–1204 (Philadelphia, 1977), pp. 46–7. Robert of Clari, La Conquête, pp. 9–10: “Quant li dux de Venice vit que tot li pelerin furent venu,” he initially organized the Venetians who were going on crusade, and then sought payment from the crusaders. Italics are mine. Thomas F. Madden, “Food and the Fourth Crusade: A New Approach to the ‘Diversion Question,’” in Logistics of Warfare in the Age of the Crusades, ed. John H. Pryor (Aldershot, 2006), pp. 209–28, at p. 211.

The Interruption of the Fourth Crusade

91

measures.45 Madden estimated that the Venetians would have been responsible for about 15,422 tons of provisions for the men and horses, and, using his figures, around 3,009,247 gallons of wine.46 The provisions would have lasted the crusaders approximately 300 days, or about ten months. Since these estimates were based on the soldiers consuming a minimum amount of calories, or the number of calories needed for a soldier at a low activity level, it is possible that the food was eaten more quickly.47 Perhaps this was why Villehardouin stated that the Venetians agreed to nine-month’s worth of provisions.48 The problem was the duration of the crusaders’ stay. Although the crusade army was smaller than anticipated, it remained on the Lido for much longer than expected. The original plan had been for the crusaders to arrive in Venice in the summer of 1202 and immediately set out for the Holy Land. Instead, the crusaders spent summer negotiating with the Venetians about how to pay their debt. Rather than send the soldiers on their way at once, as expected, the Venetians had to feed them for the entire summer. The sources clearly state that the delay was an issue. The Doge considered the postponement of embarkation a problem and blamed the crusaders for it, saying that he “would have had you [the crusaders] make the [sea] crossing long ago.”49 The chronicler Gunther of Pairis remarked that “a good deal of time passed” as the crusaders resisted the Venetian proposal that they aid in an attack on Zara.50 This probably would not have been as much of a concern if the crusade was only delayed three months. However, winter loomed when the Doge offered Zara as an alternative. The Mediterranean was dangerous during the cold months, and maritime powers rarely transported large crusade armies over the winter. Without embarking, the army would be stationary another five months before setting out for the east. The crusaders would have been consuming for eight months by the spring. Venice was not in a position to feed the army should the collected supplies run out. The Venetians had already taken a loss over the past year as they prepared

45

46

47 48 49 50

“Pactum domini Balduini,” no. 92, p. 365; “Pactum initum inter dominum Henricum, Ducem Venecie, et Bonifacium, Marchionem Montisferrati, et Baduinum comitem Flandriensem, et Ludovicum, comitem Blesensem, in captione urbis Constantinopolitane,” in Urkunden zur älten Handels- und Staatsgeschichte der Republik Venedig, 1, eds. Tafel and Thomas, no. 93, p. 371. Both versions of the treaty were almost identical. See Madden, “Food,” p. 211, n. 1 for a short discussion of the discrepancies. Madden, “Food,” pp. 211–14. Madden’s figures were in metric tonnes, which were converted to US tons. For the purposes of this article, one tonne equals 1.1023 tonnes. Based on the size of a fourteenth-century Venetian amphora, which measured about 680 liters, Madden estimated that each crusader would have had 340 liters of wine. Liters were converted to US gallons using one gallon equals 3.785 liters. Ibid., p. 212. Villehardouin, La Conquête, 5.21, p. 14. Robert of Clari, La Conquête, p. 12; English quotation from Robert of Clari, Conquest, p. 42. Gunther of Pairis, Hystoria Constantinopolitana, p. 122; English quotation from Capture of Constantinople, p. 77.

92

Gregory D. Bell

the fleet. The food itself was probably obtained from the region of the Po Valley between Cremona and Venice, as the crusaders were asked not to purchase food “at Cremona and beyond towards Venice” without Venetian consent.51 The regions around Venice were, therefore, depleted. Provisions would be more difficult to locate. Further, it would not make sense for the Venetians to bring all of the contracted foodstuffs to Venice. Doubtless they stored the provisions in caches along the route rather than ship them west to Venice only to pack them on ships and send them back east with the army. In other words, the Venetians probably used up local food supplies to feed the army, and it is unlikely that they had all of the contractual provisions for the crusaders at Venice. The crusaders might also have consumed more resources than expected. Although the army did not contain 33,500 men, it might have comprised upwards of 18,000 participants before the deaths, desertions, and the papal legate’s mandate to the elderly, women and sick to go home. It is not hard to imagine the additional mouths, the deserters and those who died before the army departed, consuming prodigiously during those hot summer months. This problem was not unique to the Venetians in 1202. For example, the chronicler Ramón Muntaner described members of his Catalan Company, who were billeting for the winter, taking two, three, and even four times what they should have by season’s end.52 Further, the food that Gunther of Pairis said the poorer crusaders brought with them was probably consumed quickly.53 And the provisions that Venetians sold to the soldiers on the Lido would deplete the city’s overall food supply. Adding to the possibility that the crusaders consumed a significant portion of the gathered resources, contemporary accounts suggest that the crusaders lacked provisions even on the short trip to their agreed upon destination of Zara. The Anonymous of Soissons, a clerical account likely based on the experiences of a chief prelate of the army named Nivelon de Chérisy, stated that from the Lido, “with their supplies almost consumed, our men [the crusaders] were compelled by the Venetians to sail across the sea against the inhabitants of Zara.”54 As the main army sailed for Zara, the Venetians left the fleet to visit their allies along the Istrian coast, stopping at the ports of Trieste and Muggia to demand that these cities provide rowers, marines, and supplies. This was purportedly done to demonstrate Venetian authority and force tribute, but the outcome was to accumulate provisions.55 The main crusade army also had to stop and pay for provi51 52

53 54

55

Madden, “Food,” p. 215; “Pactum domini Balduini,” no. 92, p. 367; “Pactum initum inter dominum Henricum,” no. 93, p. 371. Ramón Muntaner, The Chronicle of Muntaner, trans. Lady Henrietta Margaret Goodenough, online reprint (Cambridge, Ontario, 2000), 411–13, especially p. 413. This source is available online at http://www.yorku.ca/inpar/muntaner_goodenough.pdf. The events Ramón described took place roughly a century after the Fourth Crusade; however, the root of his concern is the same as the Venetians. See above. Anonymous of Soissons, “Concerning the Land of Jerusalem and the Means by Which Relics Were Carried to this Church from the City of Constantinople,” in Contemporary Sources, pp. 230–8, at p. 233. “Devastatio Constantinopolitana,” Chronique, p. 87.

The Interruption of the Fourth Crusade

93

sions at Pola, again highlighting the lack of readily available supplies that the Venetians had promised the crusaders.56 And the poor in the army continued to starve. At Zara, many poor soldiers left the army because the wealthier crusaders would not share with their poorer counterparts.57 The prospect of feeding the crusaders for an additional five months confronted the Doge in October. The Venetians could not risk running out of provisions for an army, which, though smaller, still remained sizeable. The immediate concern of the Doge, therefore, was to find a place other than Venice that was nearby and wealthy enough to sustain the crusade army over the winter should the stockpiled provisions run out. In order to meet these needs, the Doge suggested that the crusaders help him take the city of Zara on the Dalmatian coast. His purpose was clear: find a place to winter the fleet and the crusaders. “There is a city near here, Zara is its name … we will go there and stay there this winter until toward Easter, and then we will make ready our fleet and go oversea to the service of God.” The Doge also mentioned that Zara was “a very fine city and plenteous in all good things.”58 If the crusaders helped conquer Zara, they could postpone payment of their debts to the Venetians and continue on to the Holy Land the following spring. For the Doge, taking Zara not only bolstered Venetian hegemony in the region, but also distanced the crusaders from Venice. He clearly had in mind, at least in part, the necessity of protecting his city should supplies run out. And perhaps Zara, which had not had to provision an army over the summer, would have food supplies should the reserved quantity that the Venetians had collected be consumed, something that Venice would probably not be able to offer. The size of the army of the Fourth Crusade was a concern from its inception through the summer of 1202. It was too small to generate the fees needed to pay the debt owed to the Venetians, and yet it may have been too large to provision after a long summer of stagnation on the Isle of St Nicholas. The army that sat on the Lido may have been a little larger than historians have allowed, comprising somewhat more than the 12,500 to 13,000 soldiers that embarked for Zara according to the Venetian sources. Rather than setting out as armed pilgrims seeking to reconquer the Holy Land at the behest of the pope, the crusaders instead found themselves following the far less spiritually minded dictates of the Venetians. Even though many crusaders resisted the notion of attacking Zara (in fact, some did not participate in the assault), without Venetian support the crusade would be left with insufficient provisions and transportation, the two basic concerns of any logistical operation. In other words, from a logistical perspective the crusaders discovered that they had little choice but to do what the Venetians wanted.

56 57 58

Robert of Clari, La Conquête, p. 13. “Devastatio Constantinopolitana,” Chronique, p. 88, stated that 2,000 poorer crusaders left the army, 1,000 with leave and another 1,000 without. Robert of Clari, La Conquête, p. 12; English quotation from Robert of Clari, Conquest, p. 42.

94

Gregory D. Bell

As was common among twelfth-century chroniclers, Villehardouin focused on the actions of the crusade leaders and soldiers rather than the role of the noncombatants, yet both groups appear to have influenced the outcome of the crusade. While soldiers organized the expedition and paid a majority of the debt, all of the sources make it clear that there were non-combatants, too. Although the opinions of these less affluent non-combatants did not influence the major decisions of the crusade leaders, they consumed resources while contributing little toward the overall levy. The Venetians, too, experienced the consequences of a significant logistical error. The Doge was upset that the information about the size of the crusade army been incorrect. Further, he almost led Venice into a serious economic quandary and could have lost the city entirely to the crusaders. He seems to have been aware that Venice would have a difficult time sustaining the army through the winter. He probably did not want to make another logistical error and leave Venice to the whims of an unsatisfied and potentially underfed army in the coming spring. The prolonged consumption of provisions, therefore, appears to have put significant pressure on the crusaders and Venetians and was probably a key factor in their decision to move forward. The Venetians, in particular, were quite concerned with the delays over the summer and the onset of winter. The crusade army might well have consumed more than expected as it sat on the Lido, and the threat of an additional five months provisions being used up over the winter loomed. With reduced provisions and questionable reserves, the Doge convinced the crusade leaders to relocate to Zara. Having the army remain near Venice was probably no longer a viable option.

5 Light Cavalry, Heavy Cavalry, Horse Archers, Oh My! What Abstract Definitions Don’t Tell Us About 1205 Adrianople Russell Mitchell

It is somewhat odd, given the amount of attention that has been paid to the military history of the Crusades, and the degree to which one is currently wary of using the words “decisive” and “battle” in the same sentence without seeming hopelessly antiquated, that the battle outside of Adrianople in 1205 does not seem to have attracted much attention. After all, most of the other “decisive battles” of the Crusades have certainly received ample consideration. One would think that it would receive more attention, given that it is also one of our relatively few clear battle descriptions of the Cumans, a people who were as important in the thirteenth century as they are almost completely unknown today. And perhaps that, in conjunction with the brutal clarity of Geoffroi de Villehardouin’s chronicle, helps to explain why the topic doesn’t seem to generate a lot of discussion. That is a pity, because while the Cumans themselves seem to be of fairly marginal interest except to Russian and Hungarian scholars, the events outside of Adrianople have quite a bit to teach us. To begin with, by the time of the Fourth Crusade, one cannot posit that the Crusaders had no experience fighting light cavalry and horse archers pursuing nomadic tactics. At the same time, however, as much as one has to draw parallels, the phrase “nomadic tactics” itself is a definition that is almost meaninglessly abstract, and if one uses it as an explanatory bucket into which to simply pour all of the horsemen of the Eurasian steppe, we miss an opportunity to explain what happened at one of the few medieval battles where one can seriously suggest that the future of Europe would have looked very different had it had a different outcome. First and foremost, we can summarize the main points of the Crusader-Cuman encounter as follows, using Villehardouin’s account:1

1

Geoffrey de Villehardouin Chronicle of The Fourth Crusade and The Conquest of Constantinople, trans. Frank T. Marzials (London, 1908). Available via Internet Medieval Source Book at: http://www.fordham.edu/halsall/basis/villehardouin.html.

96

Russell Mitchell 1 The day before the main battle, the Cumans trick the Crusaders into chasing them, and wound many of their horses with arrows when the Crusaders disengage. And he sent his Comans running before our camp, and a cry was raised throughout the camp, and our men issued therefrom helterskelter, and pursued the Comans for a full league very foolishly; for when they wished to return, the Comans began to shoot at them in grievous wise, and wounded a good many of their horses.2

2 The encounter is immediately followed by a meeting in which the Crusaders all vow not to repeat the same mistake. 3 The following morning, the Cumans attack the Crusader camp. 4 The Crusaders assemble and pursue the Cumans 5 The Emperor is wounded during heavy fighting. 6 The Cumans rout the Crusaders at close quarters 7 Villehardouin assembles his contingent to shelter the survivors as they flee back to the camp. 8 The Cumans choose not to engage Villehardouin’s fresh troops, and instead gall them with arrow fire until withdrawing at nightfall. Given the beating that the Latins took during their first encounter with the Cumans on the previous evening, when all agreed that giving chase was foolish on account of the enemy’s lighter equipment,3 one has to ask the question: since we can assume that the Crusaders, having recently conquered Constantinople, weren’t idiots, why did they again chase the Cumans, after publicly committing themselves not to do so? As already noted, by the Fourth Crusade, one can reasonably make the argument that there was little to no culture shock regarding how to deal with light cavalry and horse archers. Arsuf stands as a clear and ready example of how Crusader heavy cavalry were able to maintain their discipline prior to catching their lighter-equipped opponents at closer quarters.4 The trick, of course, was to bring about the circumstances in which heavy cavalry was able to perform this task. The Turcoman horse archer consistently gives way before the heavily armed cavalryman because he’s faster and because he has a bow, which acts as a “30foot spear” – he can use his weapon at relatively close quarters, but he’s much more effective if he can “control the distance,” to use a fencing metaphor, and hold his opponent just close enough that it’s easy to shoot at him, but far enough away that the damage is all one-sided. Poorer Turcomen may not even have

2 3

4

Ibid. Villehardoin’s account is sufficiently vivid that one can almost visualize the post-mortem with the Crusaders in a football huddle and the Emperor saying “okay, repeat after me, we’re not going to chase the Cumans.” The fact that the Crusaders played directly into Cuman tactics twice in a row suggests that while the Crusaders had learned how to handle horse archers in general, they were not prepared for the differences between tactics between different Turkic groups.

Light Cavalry, Heavy Cavalry, Horse Archers

97

possessed a hand-to-hand weapon other than a knife. The absolute last thing they wanted was to be within a lance thrust of their opponents. An excellent example of this is the almost comical prelude to Lord Walter of Châtillon’s death as depicted by Joinville, fifty years later, during the retreat to Damietta from Al Mansourah: A street ran straight through the village, so that one could see the fields on either side. In this street was my Lord Walter of Châtillon with his naked sword in his hand. As often as he saw the Turks entering this street, he charged upon them, sword in hand, and hustled them out of the place; and whilst the Turks were fleeing before him, they (who shoot as well backwards as forwards) would cover him with darts. When he had driven them out of the village, he would pick out the darts that were sticking all over him; and put on his coat-of-arms again; stand up in his stirrups, and brandishing his sword at arm’s length cry, “Châtillon! knights! where are my paladins?” Then, turning round, and seeing that the Turks had come in at the other end of the street, he would charge them again, sword in hand, and drive them out. And this he did about three times in the manner I have described.5

On the other hand, the Byzantine writer, Choniates, describes the Cumans both generally: Their weapons consisted of a quiver slung along the waist, curved bows and arrows; in battle, they wheel about with spears.6

and more specifically at Beroe in 1187: They fought in their ancestral customary way: they let fly with their darts and attacked with their lances. After a short while, they turned their onslaught into flight, and, enticing their adversaries to follow hard behind them as though they were in retreat, then cleaved the air sharper than do birds and wheeled about to face their pursuers, whom they fought with even greater bravery. Repeating this tactic time and time again, they so prevailed over the Romans that they no longer bothered to turn around, but with naked swords and terrifying shouts fell upon the Romans almost faster than thought; overtaking both him who gave battle and the coward, they mowed them down.7

In comparing the first passage with the latter two, it is immediately obvious that, even though both groups of warriors depicted are “nomad horsemen,” their fighting methods are very different.8 Although Choniates may be classicizing when he

5

6 7 8

Joinville, The Memoirs of the Lord of Joinville: A New English Version, trans. Ethel Wedgwood (New York, 1906), pp. 197–8. Available online at: http://etext.lib.virginia.edu/toc/modeng/public/ WedLord.html. Niketas Choniates, O City of Byzantium: Annals of Niketas Choniates, trans. Harry J. Magoulias (Detroit, 1984), p. 54. Choniates, Annals, p. 218. This is not to say that no individuals in the Turkic world fought at close quarters. On the contrary, we know that there was much heavier cavalry. However, when looking at two groups instantly recognizable as “light horsemen,” who are horse archers, we see distinct differences. The description given of the Cumans suggests that their tactical response to Châtillon would have been much different than the Turcomans actually described in the chronicle.

98

Russell Mitchell

describes the Cumans fighting with the lasso,9 it’s very clear that although the Cumans use the bow, their sense of tactics is clearly not defined by it. And this is important, because Villehardouin’s account of the battle is in absolute agreement with Choniates’ assessment of how the Cumans fought, and these in turn are in agreement with what we know of Cuman military equipment from archaeological sources.10 So while the Cumans may have been horse archers, they also clearly seem to have been equipped for close combat.11 In this, they probably were not at all unusual for the region, as such an arrangement was described clearly by Maurikios in the Strategikon.12 Which begs another question: how did they manage to do this? Typically, when heavy cavalry comes into contact with lighter cavalry, the latter is simply destroyed for lack of protection – light horsemen rely on speed, not heavy armor, and at close quarters speed no longer applies. Thus, the Crusaders were able to rout the light cavalry and horse archers at Arsuf completely, and Byzantine strategy against the Pechenegs was almost always an attempt to bring them to battle where they could be crushed with combined arms. A closer look at the equipment and tactics of both the Crusaders under Baldwin and the Cumans under Qoja reveal that the differences between the two were significant, but not primarily because one group wore a heavier armor than another. To begin with, by the standards of later medieval Europe, the Crusaders at Adrianople were hardly what would qualify as heavy cavalry. Forgive the obviousness of the statement: it is there for a purpose. Obviously cavalry equipped in mail, without benefit of armored surcoats, cuirie, and coats of plate, let alone full white harness, do not qualify as heavy cavalry when set beside the juggernauts of the fifteenth and sixteenth centuries. However, when one looks at the standard of protection enjoyed by cavalrymen in this period, there is a group that might qualify – the same Imperial Byzantine cavalry that the Normans, and then the Crusaders, had consistently routed. 13 As counter-intuitive as it may sound, if one takes an 9

10

11

12

13

Choniates, Annals, p. 337. Perhaps classicizing in a reference to the Huns, perhaps not. Like the bow, the lasso is everyday equipment for the herdsman. Its use differing only in the nature of the target, it is plausible to suggest both that they would have been expertly wielded and that it would likely have been the gear of choice for those unable to afford “true” weapons like the lance, axe, mace, or much more expensive sabre. András Pálóczi-Horváth, Pechenegs, Cumans, Iasians: Steppe Peoples in Medieval Hungary (Budapest, 1989), pp. 68–85. See also illustrations of Cuman “babas,” or funerary statues, ´ ¸ ´ referred to in Witold Swietoslawski, Arms and Armour of the Nomads of the Great Steppe in the Times of the Mongol Expansion (12th–14th centuries) (Lodz, 1999), p. 116. But, notably, without the general use of the shield, which otherwise defines close-combat troops. And in this we have a useful example of diversity in the Turkic world, since the wellknown Middle Eastern technique of “shooting (from) under the shield,” was simply irrelevant to the Cumans. Surviving Cuman lances appear to be distinctly military in origin. See Œwiêtos³awski, Arms, p. 53. For Byzantine hybrid archer/lancer cavalry, see George T. Dennis (ed.), Maurice’s Strategikon: Handbook of Byzantine Military Strategy (Philadelphia, 1984), p. 11. Much more heavily equipped than other Pecheneg, Alanian, or Seljuk cavalry in Byzantine service (but said lighter contingents would also have been employed whenever possible). John Birkenmeier, The Development of the Komnenian Army 1081–1180 (Boston, 2002), pp. 67–8, 82.

Light Cavalry, Heavy Cavalry, Horse Archers

99

unprejudiced look at the armor worn by horsemen in the eleventh to thirteenth centuries, the so-called “heavy western knight” wears armor that is significantly lighter than the banded scale armor breastplate worn by his heavy-cavalry counterparts among the Byzantines, and vastly lighter than the lamellar harnesses taken for granted among the heavy cavalry of Eastern Europe for hundreds of years before that. Byzantine and Eastern European heavy cavalrymen were wearing armor that, for sheer durability and toughness, wouldn’t be matched by anything in Western Europe until at least the fourteenth century.14 However, the difference between asking how protective a particular kind of armor is, and looking at the same harness while asking “how is it protective,” is both subtle and profound. On the one hand, understanding the first question lets us frame an abstract response, which is perfectly sufficient for some historical examples but which leaves us in the dark regarding others – debates regarding the so-called “evolution” of European armor being only one example. On the other hand, investigating precisely what an armor does in order to protect its wearer allows us to get an insight as to why one armor might have been preferred over another, and to gain the sort of concrete understanding that can shed light on the subject matter at hand. As one moves out of the Carolingian era, in which the scale brunia was widespread, and into the “age of mail,” we find that short thrusting swords generally go by the wayside in favor of swords with large, cleaving blades, and this is for a very good reason. Mail can be defeated in two ways: by shearing the links directly, or by putting so much pressure on a link that it tears apart from the inside out.15 The latter is quite common when faced with a narrow penetrator like an arrowhead, and so the nature of any supplemental protection provided to the mail by a pourpoint or the like is vitally important for gauging mail’s protectiveness, because the extra material helps to support the mail and reduce the amount of “pull” on each individual link, thus effectively

14

15

A Byzantine banded scale harness bears a distinct functional similarity to a mid-fourteenthcentury coat-of-plates. For excellent recent work on the topic, see Timothy Dawson, “Suntagma Hoplôn: The Equipment of Regular Byzantine Troops, c. 950 to c. 1204,” in A Companion to Medieval Arms and Armor, ed. David Nicolle (Woodbridge, 2002), pp. 82–90 and particularly Figure VII–12. Note that many Byzantines employed westerners with their equipment, and at this point the only consideration of “heaviness” being considered is in terms of the equipment used, rather than training and tactics, which will be described in turn. For example, one can readily see a clear difference in the equipment of the warriors depicted in the Carolingian Golden Psalter and those within the eleventh-century Bayeux Tapestry. By this time, the sax has vanished from the battlefield, and (using Oakeshott’s typology) the type X sword common to both eras, with its rounded point (perfectly adequate for thrusting into an unarmored body as well as lopping off an unprotected forearm) have begun to change, getting longer, heavier, and possessing a more distinct tip, necessary for finding the much smaller gaps provided by a heavily mailed opponent. And of course, all of these are clearly cutting blades when placed besides the type XV and type XVI thrusting blades of the fourteenth century and beyond, when early armored surcoats have given way to the widespread adoption of plate armors.

100

Russell Mitchell

increasing the metal’s tensile strength.16 Mail is much less effective at distributing impacts from mass weapons and resisting hard thrusts than scale, banded scale or hooped harnesses, and especially lamellar, the last of which, if constructed of iron and laced rigidly, moves as a single piece, and provides protection loosely comparable to late-medieval plate defenses.17 Although lamellar that is loosely laced would not distribute force away from its wearer as well, a rigidly laced suit of lamellar will present at least two sheets of metal which must be cut through or penetrated in order to harm the wearer – in the case of Byzantine banded scale, this would involve both layers of metal and at least one layer of leather sandwiched between the overlapping plates. It clearly requires much more energy to punch through overlapping sheets of plate iron than to tear a link apart with a penetration weapon (such as Cuman lances and arrows). First of all, the weapon must not glance off the hardened plate or scale on impact. Second, although the thickness of lamellar plates is typically not mentioned, data presented by Alan Williams in the Knight and the Blast Furnace suggests that the range of energy required for an arrow to defeat two (rather thin) 0.75 mm plates would be a minimum of 110J. 18 A more likely 1 mm plate19 moves the impact energy required for an arrow penetration to at least 175J, and much, much greater for a lance or sword penetration. (These figures assume a perpendicular hit).20 The energy required to penetrate with a lance would be notably higher due to its broader surface area. This compares very favorably with Williams’ tests on mail, indicating that a bodkin arrowhead strike providing 120J is sufficient to completely defeat both the replica and authentic mail tested, when the mail is supplemented with a 26-layer linen jack.21 The presence of backing material dramatically enhances the mail’s effective tear strength.22 Even here, in this apples-to-oranges comparison, 140J with a lance

16

17

18 19

20 21 22

Russell Mitchell, “Archery vs. Mail: Experimental Archaeology and the Value of Historical Context,” Journal of Medieval Military History 4 (2006), pp. 18–28. Mail’s protectiveness against archery in the experiments performed literally spans a range from “worthless” to “invulnerable.” This is a classic case of “asserting something which ought to be proven.” Nevertheless, experimental practice has shown that lamellar, when tightly laced, is very rigid, and has much more in common with wearing a breastplate than with wearing scale armor. Slag within the metal is obviously an issue here. However, note that it would also be an equally important issue for mail, particularly before late-medieval improvements in metallurgy. Alan R. Williams, The Knight and the Blast Furnace: A History of the Metallurgy of Armour in the Middle Ages & The Early Modern Period (Leiden-Boston, 2003), p. 928. This will obviously remain conjectural until plate thicknesses are published. Note that most reconstructions of lamellar plates are somewhere from 1–1.5 mm thick, and thus the data given above is an attempt to portray a conservative estimate of these armors’ protectiveness. Williams, Knight, p. 928. Williams, Knight, pp. 942–3. See footnote 16, above. In repeated tests, the presence and type of backing material turned out to be the single most significant determinant for whether arrows defeated mail.

Light Cavalry, Heavy Cavalry, Horse Archers

101

strike (inherently requiring much greater energies to penetrate than a bodkin) was sufficient to completely defeat both the authentic mail and the jack. Also, note that this is an extremely stout jack, sufficient to be considered a substantially protective primary armor in and of itself. This may be reasonable for Williams’ historical context, but is an inappropriate match as an early thirteenthcentury underpadding.23 There is a further difficulty here: Williams’ figures are controversial. Clifford Rogers discounts them while citing previous tests, in which 70-lb. bows providing roughly 55J penetrated 1.5 mm of mild steel at a 45-degree angle. He also argues, at least partially based on my own previous research, that 45J would be sufficient to defeat both mail and light underpadding, while penetrating sufficiently to provide a lethal wound.24 Adjusting the penetration energy of a successful a 45-degree impact to one that hits squarely would suggest that mail and 1.5 mm of mild steel are roughly equivalent. This would mean that mail has a protective value against bodkin-type penetrators roughly equivalent to that of scale or brigandine – so long as one is not able to bypass the link structure with a very narrow penetrator (in which case the scale or brigandine is more protective).25 I do not believe that this can be the case, and although I do not disagree with Rogers’ conclusions vis-à-vis the battle of Agincourt, there is another interpretation of the data that allows us to integrate all of the previous studies, rather than having to pick and choose amongst them. This is particularly important, given that Rogers and Williams both draw from Peter Jones’ work.26 I am choosing to use Williams’ figures, on the grounds that he cites the performance characteristics of the mild steel in question and chooses to use a steel with a 0.15% carbon content, directly in the middle of the “negligible-to–0.3%” that characterizes a steel being “mild.” Otherwise, the term “mild steel” itself is too vague a term to allow an assumption that one is engaged in an apples-to-apples comparison, as some “mild steels” tested may in fact be performing no better than samples of soft iron would. I believe that this is likely to have been the case in tests that feature unusually successful penetrations of low-carbon steel of unknown

23

24 25

26

The 16-layer “padding” tested (Williams, Knight, p. 943) is a more likely analog for an earlier pourpoint, and provides notably less protection. Even this may be so heavy as to be inappropriate: the pourpoint provided in my own tests was a mere 7 layers, but still sufficiently stout that its bulk was visibly obvious underneath the mail, in stark contrast to the much more closefitting depictions of mail universal to this period. See Clifford Rogers, “The Battle of Agincourt,” in L. J. Andrew Villalon and Donald J. Kagay (eds.), The Hundred Years War, vol. 2:Different Vistas (Leiden, forthcoming 2008), Appendix I. These armors are essentially equivalent, differing primarily in whether the plates are placed on the outside of the backing material, as in scale, or the inside, as with brigandine. Both are less rigid than a tightly laced lamellar cuirass, but provide somewhat similar protection due to the presence of overlapping metal plates. Peter Jones, “The Target”, in Longbow: A Social and Military History, ed. Robert Hardy (Cambridge, 1992) pp. 206–8.

102

Russell Mitchell

characteristics.27 This has been directly confirmed in conversation with a steelworker, who mentions that stress tests he has conducted have shown as great as a 10,000 psi difference in tensile strength between lower-carbon and higher-carbon “mild steels” that are otherwise indistinguishable.28 Therefore, I believe that we are still able to keep his 45J figure for mail,29 since bodkins are able to bypass much of mail’s link structure, and literally achieve the 40 mm penetration that is Williams’ standard, without actually tearing any mail links. It is clear that the majority of the energy being used to penetrate armor in this circumstance is being expended in defeating the padding, not the mail itself. The consequences of this are quite clear. On the other hand, mail does not restrict the warrior’s movement. Most importantly, it leaves no gaps through which one might stab, particularly if the garment was tailored for its owner. The more rigid armors, precisely because they are rigid, cannot articulate around the joints, and thus there are sections left vulnerable unless the rigid armors are supplemented with an underlying defense. A highly trained warrior would be adept at exploiting those gaps. Mail, on the other hand, does not suffer from this problem. This is especially the case for those warriors at Adrianople who had the more expensive, up-to-date long-sleeved hauberks with attached mittens. Add a helmet (which may or may not possess a rigid visor), ventail, and chausses, and when writers describe mail-clad warriors as being armed “cap a pie” (“head to foot”), they mean it. Mail is an unusual and innovative armor that doesn’t seek to redirect a blow or stop it outright due to a hard surface, but rather to catch an incoming blow while protecting the wearer from a sharp edge and impact energy. Mail was already an expensive armor. Although its construction requires relatively little in the way of specialized tools and expertise in comparison to making plate armor and some leather armors, and thus can be produced as a “cottage industry,” it still involves the painstaking process of riveting literally thousands of rings together. It is quite clear that the protection afforded by mail was well worth this expense. Mail was typically made of iron that was unhardened. Thus, it has a tendency to give under a blow, and to bend, rather than break. In this way the energy of the blow is kept away from its wearer, because it is taken up inside the mail as the incoming weapon is forced to move the weight of the garment. A warrior in

27

28

29

“Generally, commercial ‘mild stee[l]’ can be expected to be readily weldable and have reasonable cold bending properties but to specify ‘mild steel’ is technically inappropriate and should not be used as a term in engineering.” Viewable at http://www.azom.com/ details.asp?ArticleID=2537#_Mild_Steel,_Normal. Private conversation with David Rylak, 19 February 2007. Note that although I therefore disagree with how Rogers interprets data from previous archery tests, my interpretation of the data actually supports, rather than refutes, the chronicle evidence that is the basis of Rogers’ hypothesis. For bodkin-type penetrators, which in this case includes Cuman lance heads, but discounts wider penetrators with cutting edges. Arrows fitted with broadheads are more likely to perform along the lines of the “lance” category provided by Williams.

Light Cavalry, Heavy Cavalry, Horse Archers

103

battle wearing only mail without padding would expect to receive bruises through his armor, but the kind of cuts that would penetrate such armor when fighting on foot would generally be short and shallow.30 This is because one cannot cut one’s opponent merely by severing a single link: one must split several links and still have enough energy left in the blow to cut into his body. There is a limit to the damage that happens, simply because mail, not being solid, is not subject to crack propagation or material failure: each link is a separate piece of armor all by itself. Even if a cut managed to sever links and damage the warrior’s body beneath, there is a limit to how far the sword can cut into the body of one’s opponent without coming into contact with yet another, perfectly undamaged, link, at precisely the moment when his weapon has lost all of the momentum and kinetic energy required to sever it. Thus, one reads passages such as the following, in Chrétien de Troyes’ The Knight of the Cart: With their swords they dented their shields, helmets, and hauberks; they split the wood and broke the chain-mail, and each was wounded several times.31

The degree to which literature mirrors reality is an obvious and ongoing controversy. However, the text clearly matches what the intended audience for these tales would have expected to see. As the two heroic figures fight, it is quite clear that the wounds are being inflicted through mail, each of which would likely have killed or maimed an unarmored man. Their heroic stature is revealed by their ability to consistently wound through armor, as well as simply denting the mail, rather than merely bruising each other. And, note that, on foot, even these heroic figures are not expected to cleave through the solid metal of their opponent’s helms. Mail is relatively less effective against cleaving weapons such as axes and halberds, because of the weapons’ concentration of mass directly behind the cutting edge, which allows it to cut like a chisel, inflicting a deep wound, even if relatively few links are broken. (One can understand the difference between the two by visualizing a man splitting wood with an axe, versus carving a ham with a knife.) Still, in order to cut, rather than simply bruise or bludgeon, the axe head must hit squarely on the armor. There is some evidence that this is actually how Western Europeans of the High Middle Ages attacked with their swords, as well – the long, extended position regularly depicted in strikes, with the wrist deeply bent, is completely unsuitable for any kind of long cut, and entirely consistent with trying to focus the energy of one’s blow into a fairly small area for a short, cleaving impact.32 30 31 32

Fighting from horseback, where the weight of the horse can be put behind a blow, is potentially a very different story. Chrétien de Troyes, The Knight of the Cart, in Arthurian Romances, trans. William Kibler (London, 1991), ll. 2776–8. If any “history of fencing” were to attempt to include the entire medieval period, it would have to account for the notable change in the depiction of fighting postures and actions (and the resulting changes in movement that these imply) that occurs between the 13th and 14th centuries.

104

Russell Mitchell

Mail’s great value for protecting against cuts without hindering mobility comes at a price. The armor’s lack of rigidity means the wearer is vulnerable to being battered spectacularly and suffering from blunt trauma, even if the armor itself is not defeated. Therefore, in some respects, mail itself may have encouraged the practice of prisoner-taking and ransom, because it was perfectly possible to beat somebody into submission without doing much damage to their armor at all. The following is an example involving archery, but the implications for close combat are quite clear: … among these was a fiery young fighter named Benkin, expert and swift in shooting arrows. He kept going around the walls in the fighting, running here and there, and though he was only one he seemed like more because from inside the walls he inflicted so many wounds and never stopped. And when he was aiming at the besiegers, his drawing on the bow was identified by everyone because he would either cause grave injury to the unarmed or put to flight those who were armed, whom his shots stupefied and stunned, even if they did not wound.33

It is for this reason that mail was frequently paired with padded garments. Against archery itself, this combination was valuable, but not invulnerable. There are many examples where mail was proof against arrows and equally there are other examples where it was not. The lighter leather armor worn by the Cumans, on the other hand, is still a matter of some conjecture. It is generally accepted that they wore armor made of multiple layers of leather, probably soft leather. Matteo Villani describes the Hungarians, who in this case may or may not have included Cuman contingents, as wearing “farsetti di cordovano,” made up of four layers of leather put together: They generally wear cordovan [leather] doublets, that they wear over their clothes, and to that they add a new (layer), and to this another, and attach another, and in this way they are made strong and sufficiently defensive.34

This accords well with images of Cumans, both those produced by themselves (the well-known funereal statues) and by others (images within Russian chronicles). His description of layers being added sequentially, rather than simply put together and then stitched, implies that they may have been glued as well as sewn together, and he describes the resulting armor as being “sufficiently defensive” for close-quarters combat; it should be noted that this account comes from an Italian writer of the mid-fourteenth century. 35

33 34

35

Galbert of Bruges, The Murder of Charles the Good, trans. J. B. Ross (Toronto, 1982), p. 165. Cronica di Matteo Villani: A Miglior Lezione Ridotta, coll’aiuto De’testi a Penna, ed. Giammaria Mazzuchelli (Florence, 1825), Book 3/7, Chapter 56, “come si reggeano gli Ungheri in oste.” Author’s own translation. Villani doesn’t specify the nature of the combat, but the context is clear: just prior, he describes them as being armed, in addition to bows, with long swords for defending themselves.

Light Cavalry, Heavy Cavalry, Horse Archers

105

We know that such armor would have to be sufficiently flexible to be worn in the saddle, and also fairly thick, given the popular use of “breast phalerae” (or the “Varangian bra”) to secure the upper section to the chest and prevent gapping.36 These are round metal disks that are connected with straps across the shoulders and back (thus the “bra” nickname). It is not precisely certain what type of leather Villani meant by “cordovano” in this instance, as that term had not yet settled into its very specific modern use. It could refer to an alum-tawed leather, which would be pliable, though not particularly soft, with a dense grain (this is commonly used by taxidermists to constrict the surface of a hide in order to keep the fur from dropping out), or else it could have been a relatively stiff and bulky reddish bark-tanned leather, perhaps with a thin rawhide core (due to the technical considerations involved in a hard, fast leather-tanning using a very strong tanning liquor), or else it could simply refer to a fine or soft leather, such as an oil-tanned leather or a sueded bark-tan. The last possibility is probably the most accurate, given tawed leather’s theoretical vulnerability to moisture combined with the Cumans’ habit of campaigning during the winter and spring, and the serious problems that a fast-tanned leather would present in terms of the upper-body flexibility that horseback archers require. What we know that it was not, was a cuir-bouilli-type breastplate, for two reasons. First, a doublet-type garment could not be made of cuir bouilli if intended for a horseman, because the garment would effectively act as a straightjacket – the rider would have had difficulty getting into the saddle in the first place, and would certainly not be able to engage in any sort of fighting. The second reason is Villani’s use of “farsetto” (meaning “doublet” or “jacket”) as a term to describe the armor, and the time he spent to describe its construction, rather than simply saying “thick leather breastplate.” Furthermore, there are many parallels – too many to list – for such a military garment throughout the Kipchak world. Therefore, one can reasonably posit that the Cumans were wearing a sort of exceptionally thick multiple-layer “proto-buff coat,” less difficult to cut through than equally thick cuir bouilli, but without cuir bouilli’s and lamellar’s primary disadvantage – gaps in the armor through which the warrior is completely vulnerable.37 36

37

For illustrations of the “Baba” statues in which these are depicted, in which exceptionally heavy center seams are clearly represented, see Œwiêtos³awski, Arms, p. 116. The “chest phalerae” disappear after the first half of the 13th century, but technological arguments are dangerous here: changes are just as likely to be due to the vast cultural crisis of the Cumans after their defeat by the Mongols and the ensuing (frequently violent) process of cultural assimilation in Hungary. There is an additional option, were the hides taken specifically intended for use as armor. Softening a hide is a specific step in the production of leather, and this phase may have been done more extensively for areas that would require high levels of flexion, and less for areas across the chest or back that were acceptable when more stiff. This would tend to accommodate the natural variations in stiffness and thickness inherent to various regions of a given hide, and would handily account for the necessity of the “breast phalerae” on a garment otherwise sufficiently flexible to allow for horse archery. While perfectly obvious from a tanner’s perspective, there is, of course, no way that these sorts of construction questions can be addressed from current archaeological evidence.

106

Russell Mitchell

Thus, when we stack up the relative protection afforded by the armor of the average Crusader and the average Cuman warrior, we find that they are surprisingly similar.38 Both wear protective gear that is flexible and allows for excellent freedom of motion and both lack the “exoskeletal” protection provided by stronger and more rigid armors. The armors compensate for this by protecting nearly the entirety of the warrior’s body. The Cumans’ armor was probably a little lighter and therefore easier on their horses, and the Crusaders’ a little heavier and possibly a bit more protective – and definitely requiring less repair after battle. 39 The numbers here indicate that 120J is sufficient for an arrowhead to defeat both mail and jack.40 For layered leather, we must extrapolate from the data that Williams has provided us, wherein he notes that “the energy needed by points and edges to penetrate plate increases very approximately with the square of its thickness.”41 Williams describes a “buff leather” sample being defeated at 70J by a blade, and 30J by an arrowhead. If we assume that sheets of leather and metal will react similarly by thickness, then, using the formula he provides, that a quadrupled thickness requires 9.2 times as much energy to defeat, four layers of the buff leather that he tested would provide a staggeringly protective garment, failing at 644J and 276J, respectively.42 This supports Villani’s commentary, and we know that these garments were thick, but this degree of protection is, to put it bluntly, distinctly counter-intuitive, even if one factors in the degree to which leather armor is vastly more ablative than its metal counterparts.43 If, in the absence of

38 39

40 41 42

43

At this stage we are purposefully neglecting the (immense) role of the shield. My own experience as a leatherworker suggests that such a garment might have weighed as little as half as much as the comparable garment in mail. This is supported by the fact that the Crusaders themselves regarded it as lighter equipment, but there is no way to make estimates without engaging in dangerous speculation regarding the armor’s thickness. For instance, a typical modern “suede” split might be “4 ounce” leather, meaning that one square foot weighs a quarter pound. Four layers of something like this with short sleeves could in theory result in a garment weighing as little as eight pounds. But make even a minor modification to the leather in question (such as preserving the “grain layer” of the leather), and the weight estimates will swing wildly. Williams, Knight, pp. 942–3. Williams, Knight, p. 928. Thickness to the power of 1.6. Note that although thickness is a major factor for protectiveness (for example, walrus skin is sufficiently thick that it will resist a bear’s bite, at the same time as pigs have been known to bite cleanly through articles of iron or steel), I do not believe that the multiple-hundreds figure applies so much as that they need to be mentioned. As much as these numbers seem an embarrassing “reach” for the purposes of argument, note Williams’ figures provided for another defense made of organic materials. When one compares the 16-layer “padding” Williams describes, versus the 26-layer “jack” made out of the same material, we see penetration energies of 80J and 200J, respectively. In other words, a jack roughly 1.6 times as thick requires a penetration energy 2.5 times greater: just like sheet steel, this is not quite a quadratic increase. Further experimental data is clearly required. After all, a hard cut on mail that bends a few links has done almost no appreciable damage to the armor. A similar cut that slashes through three layers of leather, but not the fourth, has definitely damaged the garment.

Light Cavalry, Heavy Cavalry, Horse Archers

107

experimental data supporting such an unlikely figure, one assumes a linear relationship and instead simply quadruples the numbers given (70J and 30J, respectively), one still arrives at a garment either out-performing or matching the performance of mail, even if a layer of the leather is removed.44 Dr Williams’ data aside, the widespread use of armor-penetrating weapons by the Cumans meant that any superiority of mail that did exist was not sufficiently vast as to provide the Crusaders an overwhelming advantage in mounted combat.45 In period artwork, both armors are defeated by lances on horseback. For example, mail is inadequate protection for Haraldr III (“Hardrade”) in the Life of Edward the Confessor, where he is shown being speared on horseback. “The Cuman” of the Legend of St. Ladislas depicted in a wall painting at Szepesmindszent (present Bijacovce in Slovakia) is similarly run through by a lance.46 In wall paintings showing this story, King Ladislas and “The Cuman” are almost always painted wrestling on foot.47 This is a typical response to an opponent who is wearing armor through which one cannot cut.48 Both armors are therefore vulnerable to the primary weapon wielded by their opponents on horseback, and both are potentially both protected from and vulnerable to secondary hand weapons such as the Crusaders’ swords and the Cumans’ maces. By 1205 it is almost certain that some of the Crusaders would have possessed doubled mail, which would have been dramatically more protective than the

44

45

46

47

48

This since 120J was sufficient to completely defeat the (stronger) replica mail and jack, and 80J sufficient to defeat the mail and padding such that “the jack was holed completely.” Williams, Knight, p. 942. We are unfortunately left somewhat in the dark by Williams’ apples-to-oranges comparison, but clearly the mail itself is failing at energies far below 80J. See footnote 16, above. There is no way to create an “average” mail shirt for this comparison without devising some means by which one could assess the amount and distribution of slag within the metal. Williams’ modern replica thus constitutes a “best case” notably superior to most original pieces with slag inclusions. Cambridge University Library MS. Ee.3.59, f. 32v. Latest comprehensive discussion in: L. Gyula, László-legenda közékori falképei [The Medieval Wall paintings of the Legend of St Ladislas], (Budapest, 1993). Depictions of said wrestling are never the friendly “collar and belt” wrestling, but rather the “unfriendly” postures attempting to control or destroy the opponent’s elbows and shoulders, similar in numerous respects to wrestling as depicted in Fiore dei Liberi’s Flos Duellatorum: Il Fior di battaglia di maestro Fiore dei Liberi da Premariacco, ed. F. Novati (Bergamo, 1902) The Pissani-Dossi version is available online via the Academy of European Medieval Martial Arts: http://www.aemma.org/onlineResources/liberi/liberiHome.htm. Students of this manuscript will recognize the heavy use of the “posta frontale” within various versions of these paintings. This is the consensus view of scholars working to understand medieval martial arts manuscripts, and is based primarily on private conversation with several researchers working on the Flos Duellatorum. Almost all surviving manuscripts of this sort demonstrate a specialized, separately trained approach to personal combat when faced with armored opponents. Note that if the energies theoretically suggested above are accurate, they would indeed be far above the 60–130J that Williams posits available from using a sword or axe (Williams, Knight, p. 945).

108

Russell Mitchell

normal variety.49 It is unfortunate that there is still no consensus on how such garments would have been constructed. There is an embarrassment of possibilities, some of which, such as adding additional layers to parts of the garments where flexibility is not an issue, clearly contradict the logic of further thirteenthcentury armor developments, such as the use of the cuirie and cuir-bouilli hard armor pieces such as vambraces and rerebraces.50 It is theoretically possible that a few might have possessed very early forms of cuirie or armored surcoats that do not appear within surviving artwork until much later. Cuman warriors also would have possessed numerous options for “upgrades” to supplement the minimum armor assumptions: lamellar harness, scale vests, or breastplates, all of hardened leather, and for the wealthier combatants, suits of iron lamellar. When one combines the stouter of these options with their tactics, the Cumans begin to show an eerie resemblance to a typical sixteenth-century reiter. One area in which the two differ starkly is in the Crusaders’ use of the shield, and it is a significant issue for understanding what made western cavalry so effective. We have almost no evidence for widespread use of the shield amongst the Cumans. This is in stark contrast to the Crusaders, amongst whom one can reasonably make the argument that the shield was the principal form of protection, which defended its user not merely by stopping incoming blows, but by literally closing off significant portions of his body as a viable target. Unlike the round or lenticular shield common among Byzantine cavalrymen, the Crusaders used a form of shield that was tailored specifically for use on horseback and often simply hung as a passive defense from the neck and shoulder. By the beginning of the thirteenth century, it had begun to grow shorter, probably because the couched lance did not present as great a threat to the leg. Unlike the older overhand technique, the angles involved with the couched lance technique did not favor the possibility of simply dropping the point of one’s lance into the opponent’s leg as one rode by, and letting the lance head tear a huge hole in his thigh on its way out. On an individual basis, such a use of the shield balanced out the flexibility of mail by putting a huge emphasis on defense, at the expense of mobility and offensive options: For a Frank on horseback is invincible, and would even make a hole in the walls of Babylon, but directly he gets off his horse, anyone who likes may make sport of him.

49

50

And, theoretically, even tripled. Such provision was not simply restricted to the wealthiest warriors, either: Usama’s memoirs note a captured Crusader in doubled mail, but is clear from his text that said warrior is merely a sergeant, and neither a particularly wealthy, nor wellconnected one at that. The circumstances of this sergeant’s capture demonstrate his utter immunity to sword strikes while wearing this armor. Ibn-Munqidh, Usama, An Arab-Syrian Gentlemen & Warrior in the Period of the Crusades, trans. Philip K. Hitti (New York, 2000), p. 104. This is important: since we can directly trace changes in knightly equipment, the “standard” technological-determinist argument of “armor evolution” would suggest that doubled or tripled mail would be less protective than a combination of mail plus cuir-bouilli. In this case, the “standard argument” may have merit: cuir-bouilli, particularly if the leather has a rawhide core, is both hard and extremely tough, and, unlike metal, not subject to failure due to slag inclusions.

Light Cavalry, Heavy Cavalry, Horse Archers

109

Anna Comnena’s contrast between the effectiveness of the individual Norman on foot and on horseback makes sense in this context, and deserves to be considered seriously.51 The advantage that such an arrangement provides, on the other hand, is significant, and, if carried out correctly, well worth its costs. By restricting both the movement of the shield and of the lance, the western cavalrymen were able to operate in a tighter formation that threatened a denser combination of lance points and horseflesh-in-motion than one can achieve wielding the lance overhand. The latter method likely struck with just as much force, since the lance rest, which relied upon a solid breastplate for its foundation and made fifteenth- and sixteenth-century cavalry nigh-unstoppable without blocks of pike, hadn’t yet been invented. The overhead method also allows vastly more freedom of movement with the lance, which means that a group of horsemen using the lance this way requires more space than a group using the couched technique. Otherwise, a bad bounce at impact places the horse and rider in the front rank at serious risk of being “clotheslined” by one of his companions’ lances. This is not to say that earlier cavalry did not charge in good order – there is much evidence that this was practiced obsessively amongst the Carolingians, for example – but that the definition of “good order” changes dramatically according to the techniques the warriors are using. There is a good reason that historians regard the use of the lance held this way as an attribute of light cavalry. Similarly, compared to the more fencing-oriented training for the lance among Byzantine and Mamluk horsemen (as seen throughout the furusiyya literature and as written about by David Nicolle), a European cavalry unit uses the lance as a unit, in order to turn the unit into a “wrecking ball” – a very specialized tool with overwhelming power if used correctly.52 If the Norman or western cavalry unit can bring itself to bear properly upon an enemy formation, the resulting advantage in both reach and unit density means that there is a very good chance that the opposing array will be dispersed and its order completely disrupted, at which point it can be dealt with easily, as the remaining enemy lacks both the density and numbers to withstand successive charges. Needless to say, the strength of the enemy’s armor is a moot point if he is violently unhorsed into a crowded mass of men, horses, and weapons. If western cavalry focused on destroying the enemy’s cohesion through the use of an obsessively trained dense fighting order, the Cuman cavalry attempted to achieve similar disruption, but using very different tools. To begin with, the Cumans were accustomed to fighting against well-armored cavalry, and this is reflected in their weapons. Choniates may be entirely literal, rather than

51

52

Anna Comnena, Alexiad, trans. Elizabeth A. Dawes (London, 1928), Book 13, Chapter 8. Cited from Medieval Sourcebook online: http://www.fordham.edu/halsall/basis/AnnaComnenaAlexiad00.html#BOOK%20XIII. David Nicolle, “The Impact of the European Couched Lance on Muslim Military Tradition,” in his Warriors and their Weapons around the Time of the Crusades (Burlington, 2002), pp. 6–40.

110

Russell Mitchell

classicizing, about the Cuman use of the lasso – it would have possessed the virtue of allowing a poorer and presumably less-well-equipped horseman the ability to fight while keeping safely out of sword’s reach. In the process, he could target either the man or his horse, and the effects of either one would be significant. We have significant archaeological evidence, on the other hand, that even the less exotic Cuman weapons were designed with an armored opponent in mind. The Cuman mace was light compared to those used in Western Europe, but was mounted on a light, long shaft presenting a greater lever arm, and probably hit with similar force. Similarly, Cuman lance heads were not the bladed variety seen in Western Europe, but instead were built on a narrow, pyramidsectioned armor-piercing design – essentially a bodkin arrowhead scaled up to fit a lance shaft.53 Until heavier armors such as coats of plate came into use, there was no need for such a lance head in Western Europe – the bladed, multi-purpose lance was a better deal than the specialty can-opener. Therefore, the Cuman engages cavalrymen differently than the western cavalryman does. The aim is the same, but the method is completely different – and one with which the Crusaders, unlike their Byzantine, Russian, or Hungarian counterparts, were obviously unfamiliar. The Cuman cavalryman, using his lance in two hands, did not have the reach of the Crusaders as their unit bore down, but the Cumans had no intention of charging into a dense formation of horsemen – just as they did not attempt to overwhelm Villehardouin’s reserve at the end phase of the battle. Instead, they turned, and fought as they were being pursued. One thing that Nicolle ought to have emphasized in his article on the use of the lance is that, just as the horse archer could fire backwards while moving, he could also fight with the lance the same way.54 This is directly depicted in the furusiyya literature, 55 and the exact technique depicted is described in the Munyatu’l Ghuzât: When you are pursued and are in front of your opponent, defend (with your lance) to your left and put your lance on your left wrist.56

Unlike the sabre, the lance has sufficient reach to easily defend the horse’s crupper, particularly given that the cavalryman with a shield was limited in how he could use the weapon on the charge – the Byzantine cavalryman somewhat limited, the Crusader extremely so.57 The Cuman warrior was relatively unlikely to 53

54 55

56 57

´ ¸ ´ Swietoslawski, Arms, p. 53. This is not an exaggeration: the lance heads described are universally narrower than 3 cm at the socket, and frequently narrower than 2 cm, with a long taper from the point. Nicolle, Plate V, figures E and G; Plate VI, figure H; Plate XII, figures D and E. British Library, Add 18866, fol. 97a, “Illustration of two horsemen whose lance-heads are between each other’s shoulder blades,” in Medieval Muslim Horsemanship: A FourteenthCentury Arabic Cavalry Manual. (London, 1979), p. 5. - A 14th-Century Mamluk-Kipchak Military Treatise, trans. Kurtulus¸ Öztapçu Munyatu’l Ghuzat: (Harvard, 1989), p. 65. Lenticular shields, still commonly represented in Byzantine artwork, would have given those using them more freedom with which to handle the lance: at a price, compared to the much greater protection afforded by the Crusader’s larger, significantly more restrictive shield.

Light Cavalry, Heavy Cavalry, Horse Archers

111

do much real damage this way unless he managed to poke the Crusader’s horse in the nose, but that wasn’t really the issue. All he has to do is bat his opponents’ lances away from a roughly three-foot-wide area for long enough to draw out the charging force until it either tired itself out, or else began to lose its cohesion, at which point it could be defeated in detail. If the enemy refused to remain engaged, as in the first day’s encounter at Adrianople, the Cumans would use the bow to either punish the enemy and their horses as they retreated (the opponent’s shield now a non-issue, because it would be facing the wrong way), or else to force the opponent to remain engaged. Once the opponent showed signs of disorder, the Cumans then engaged in earnest. We can follow this directly in Villehardouin’s text, which describes the encounter differently than Choniates does. Choniates posits that the Cumans were able to hide their numbers due to the terrain, and so lured the Crusaders into an ambush.58 Numbers likely played some role, given that the Crusader force was now split into two groups that were unable to cooperate.59 But Villehardouin, to his great credit, does not simply describe the Crusaders simply being mobbed by overwhelming numbers of Cumans in ambush, as does Choniates. In fact, Villehardouin says nothing at all about an ambush or about numbers, except earlier to lament that the Emperor had not brought all of his troops with him to Adrianople.60 Rather, forced to rely upon the hearsay of those who fled, the Marshal’s account clearly shows him struggling to come to grips with precisely what happened, with the count gravely injured and groups of Crusaders suddenly and somewhat inexplicably61 wavering once the Cumans began to attack: And then the Cumans turned back upon them, and began to cry out and to shoot. On our side there were battalions made up of other people than knights, people having too little knowledge of arms, and they began to wax afraid and be discomfited. 62 And Count 58 59

60

61

62

Choniates, Annals, p. 337. Though the guard at the camp was clearly a reserve, rather than an equal force, as Villehardouin recalls his post-battle conversation with the Doge, indicating that they had “lost the Emperor Baldwin and Count Louis, and the larger part of our people, and of the best.” Villehardouin, Chronicle, p. 96. Even this is a bit sketchy. Villehardouin describes a force of twenty-five knights fleeing the disaster at Adrianople to Constantinople. If his force had originally possessed only slightly more than a hundred knights (the lowest possible interpretation of Baldwin’s forces possible from the chronicle), where did these men come from? A rearguard consisting of such a tiny force would not have been militarily significant unless augmented with a very large force of sergeants – otherwise there would have been nothing to prevent the inhabitants of Adrianople from sallying. Unless one is ready to believe that Baldwin and Louis were suicidally rash (not likely given their success at Constantinople), I believe that the Doge’s statement, combined with the Crusaders’ readiness to bring battle, suggests that the forces were not so small. After all, the Crusaders had recently made short work of Byzantine cavalry and archers in similar circumstances. Villehardouin clearly expected that the Crusaders should have been able to withstand the Cuman archery fire. One would presume that these are sergeants, as they clearly are intended to fight on horseback, and Villehardouin’s chronicle nowhere discusses any unusual contingents, such as Turcopoles, amongst the Crusaders. But there is not enough data within the chronicle to draw more meaningful conclusions. It is entirely possible that Villehardouin is merely drawing a social distinction in order to contrast the evident bravery of Baldwin and Louis with those who fled the battle.

112

Russell Mitchell

Louis, who had been the first to attack, was wounded in two places full sorely; and the Cumans and Wallachians began to invade our ranks; and the count had fallen, and one of his knights, whose name was John of Fraise, dismounted, and set him on his horse.63

Since one cannot rationally posit the Count of Blois having taken the field in sub-standard equipment, the Cumans’ archery was obviously effective – much more effective than the chronicler is comfortable admitting. Villehardouin does not describe how the Cumans shot, some of which was undoubtedly direct, pointblank shooting. He describes the severe damage that was done to the Crusaders’ horses the first night, and the enemy “invading the ranks,” which implies that Louis and Baldwin’s formations were already disordered prior to the actual Cuman assault. This is supported by the fact that there was sufficient space that the Count fell when wounded, rather than being caught and supported, and enough space for other knights to be able to dismount to get him back up on horseback, all within the portion of their formation most dependent on maintaining close order.64 This is sufficient information for us to fill in the gaps and explain what happened. One of the reasons that the Cumans’ archery likely was effective was for precisely the same reasons that the western cavalry tended to be effective – the inherent density of the Crusaders’ formation. Unlike foot archers occupying a prepared position or able to make small adjustments in order to obtain a shooting window, when the mounted Cuman units turned to shoot prior to engaging the Crusaders, only those in the front rank would have been able to shoot directly. The rest, lacking a clear shot, would have had to shoot at a high angle up into the air. This is an aspect of mounted archery that is alien to Western Europe, and so deserves some attention. The technique involves firing at a very steep angle – 70 or 80 degrees – and allowing the arrow’s forward momentum to be extinguished while the arrow gains height, at which point the arrow falls directly (vertically) to the earth under constant acceleration. The technique requires consistent practice in order to gain accuracy, and cannot be used effectively by an amateur except perhaps against a large, densely packed formation. The skill involved is as different from “typical” archery as pitching a baseball is from pitching a horseshoe, and involves being able to gauge and time the fall of the arrow with the intended target. My two attempts to recreate the technique, for instance, have never resulted in a shot grouping of less than five meters, and this with a very light bow at ranges no greater than fifty meters, and I have yet to encounter a modern bowhunter willing to admit the possibility of so much as hitting a barrel this way.65 63 64

65

Villehardouin, Chronicle, p. 94. Note the similarity to Choniates’ description of Cuman tactics. It is conceivable that the count had left the formation. This might have some plausibility insofar as it might make him more vulnerable to arrows (one’s shield only being able to cover one “line” of the body). However, doing so would have been both foolhardy and contrary to Crusader “doctrine.” This is, admittedly, similar to sitting down to one’s second piano lesson and pronouncing Mozart unplayable. On the other hand, I have heard stories of Turcoman shepherds hunting goats from horseback using this technique, and it is considered an important practice amongst ´´ ´´ Gabor of the Agrisome “traditional archers” in Hungary (private conversation with Szollosy cultural Museum of Budapest, 22 December 2006).

Light Cavalry, Heavy Cavalry, Horse Archers

113

It is entirely possible that these Crusaders had never encountered this technique in the field before, because the horse archers who engaged them during the Third Crusade are commonly described as shooting short arrows with the use of an arrow guide. First, these arrows, while carrying for very long ranges, are limited insofar as they can only mount a bodkin-type arrowhead – anything larger would not fit within the guide. The short “dart,” therefore, tends to lack sufficient mass to stabilize the arrowhead and keep it from yawing (in which case penetration is extremely unlikely).66 Like an English “Clout Shoot,” or traditional Korean archery practice, the arrow could certainly be used to hit a specific target or piece of ground at long range, but would involve a much lower shooting angle.67 The Cumans, on the other hand, used a variety of arrows, and the extremely heavy kite- and chisel-shaped arrowheads, which have no parallel in Western Europe, would have been well suited either for direct shooting at close ranges, or else for stabilizing the arrow when put to use as plunging shot. 68 The second reason that it is fairly unlikely that this technique was used except during sieges is that the stand-off horse archers fighting in the Levant were also engaged in return by Crusader archers and crossbowmen. The latter, shooting powerful projectiles, were able to hold the horse archer off at a distance. The plunging fire technique, on the other hand, is not a technique that allows for long-range archery – it sacrifices anywhere from half to two-thirds of its theoretical maximum range in order to obtain an angle which is difficult to defend against. An arrow shot in this way is extremely difficult to spot, even for the archer, unless one has tracked its entire flight. Significantly, the arrow also hits at something close to its initial power at point-blank range, because it’s accelerating all the way down. If the elevation difference is sufficient, it actually hits harder than if shot directly at point-blank range, and while that may seem obvious, it is something that needs to be kept in mind given Adrianople’s hilly terrain.69 It is relatively unlikely that this technique was ever used amongst western archers. We can deduce this for several reasons: first, infantry archers almost never need to shoot at this kind of angle – they do not have the mobility that allows them to sacrifice range, and, not being dependent on their horse for

66

67 68 69

A practical demonstration of this principle involves simply throwing an arrow with a very light arrowhead, versus a similar one mounting a much heavier arrowhead, and observing how much more stable the latter arrowhead is in its descent. A hit with the arrow shaft, after all, is meaningless in combat. Which also includes the use of an arrow guide. ´ ¸ ´ For a (very) small sample, see Swietoslawski, Arms, pp 128–30. This also demonstrates why the “Scythians” depicted in the Strategikon have no problem with fighting in very deep formations of cavalry. It may also explain the Strategikon’s depiction of an extremely deep infantry-archer formation, with men ranged up to 16 deep! (Strategikon, pp. 149–50), which is in stark contrast to English longbow deployment … but this depends on the extent to which one buys the argument that infantry archers couldn’t obtain a good shooting angle by simply angling around their fellows. That question is unlikely to be settled in the absence of an experiment with a couple hundred Boy Scouts or the like with their archery equipment.

114

Russell Mitchell

positioning, can easily get an angle at which to safely shoot past their fellows in formation. Second, the technique is generally unsuitable for siege because of the range one must sacrifice in order to pursue the technique.70 When fighting in a heavily wooded environment, shorter sight lines and the interference of the forest canopy make it a loser’s game compared to direct shooting. Third, when longbowmen are deployed by the English in great numbers during the Later Middle Ages, they do not seem to have used the technique – had they been aware of it or practiced with it, they surely would have put it to use at Poitiers, when the French heavy cavalry moved to screen their light infantry. In addition to being able to simply shoot the infantry and cavalry to pieces from above, the longbowmen would not have had to move to an alternate position in order to get a shot at the horses’ unarmored flanks – and those flanks, like those of steppe heavy cavalry, likely would have been armored had plunging fire been a threat.71 It is no great mystery, therefore, why Villehardouin paints less-experienced contingents amongst the Crusaders as “waxing afraid and becoming discomfited.” Like a game of “scissors, paper, rock,” the same close-order formation that gave the Normans and Crusaders an advantage over both light cavalry and their significantly more heavily armored Byzantine opponents turns out to create the almost ideal target for the sort of archery the Cumans would have had to use to engage the Crusaders at short-to-medium range. What would be a very difficult shot against an individual rider or against a loosely arrayed formation of riders would instead have been reasonably likely to have done significant damage – and we know, direct from Villehardouin, that it did. They would have shot at a relatively solid carpet of horse and rider, where it would have been much less difficult to hit, and where the arrows, rather than coming in at an angle against the best-equipped Crusaders’ shields (much like Poitiers later), instead would have ripped holes at random throughout the ranks, as stricken horses caused increasing disruption throughout the Crusaders’ formations.72 The difficulty here

70

71

72

This is certainly not to say that archers didn’t use indirect fire during sieges, but rather that circumstances that would allow an archer to stand both safely and sufficiently close to engage in the technique would be relatively rare. Since they were, after all, shooting at formations. Clifford Rogers, War Cruel and Sharp (Woodbridge, 2000), pp. 378–9, and maps, pp. 374–5. Note that the first shooting against the mounted men-at-arms is mid-to-long-range shooting, whereas the shorter-range shooting from the flank that disperses the men-at-arms and destroys the vanguard is done from the side. Having “crossed the T” on the vanguard, the archers were in an ideal situation in which to shoot, but would have been unable to see most of the individuals within the formation itself, whether mounted or not. Or difficult to miss, depending on the extent to which one accepts the Crusaders’ use of dense formations. An exceptionally dense formation on the lines of the “an apple could not fall to the ground” topos would carry its own obvious ramifications. Note the emphasis in the Strategikon on attacking archers on open terrain with even ranks, thus diminishing the density of fire that would otherwise skyrocket as targets get bunched up behind a stricken horse. Strategikon, p. 115: “When lancers attack archers, as we have said, unless they maintain an even, unbroken front, they sustain serious damage from the arrows and fail to come to close quarters. Because of this they require more even ground for such fighting.”

Light Cavalry, Heavy Cavalry, Horse Archers

115

lies in accepting that horsemen who are engaged in indirect plunging fire could hit other horsemen on the move.73 However, one does not need to posit overwhelming damage from the Cumans’ shooting for members of the Crusading contingents to have “waxed afraid and become discomfited.” In particular, since we know that the Cumans destroyed the Crusaders at close quarters without the benefit of an extended archery barrage, Villehardouin’s battle narrative does not allow for such a possibility. There are two primary ways in which this could be interpreted: 1 Cuman armor was lighter, and the Crusaders possessed a definite advantage in hand-to-hand combat. Therefore, the Cuman archery barrage must have been sufficiently effective to overcome this disadvantage. 2 The Cumans, while lacking the protection of a shield, were armored in a way that did not put them at a serious disadvantage at close quarters, and therefore an archery barrage that was relatively less effective, but still sufficient to disrupt unit cohesion, would have been sufficient to allow the Cumans a close-quarters victory. In either interpretation, with the enemy in disorder, they finished the fight at close quarters, causing a rout. Villehardouin’s depiction of the count and emperor’s heroic behavior may have been a direct reflection of personal virtue or simply told out of a desire to praise his superiors, but it was also the correct tactical move for the situation. Turning one’s back on faster opponents en masse in order to flee while they shot and gave chase at close quarters would have been suicidal, and one can reasonably posit that the close fighting in which the emperor was captured bought the lives of those who fled by providing them the time they needed in order to return to the refuge of the camp, two leagues away. Significantly, they did not engage Villehardouin’s relief force as it moved into position to shelter the retreating Crusaders, but simply galled them with arrows until sundown.74 This may be attributable to the Cumans simply being fatigued, but it also squares very well with how we understand their concept of warfare – like western heavy cavalry, they sought to disrupt their opponent’s formations. Unlike said cavalry, their strategy for doing so involved pulling the formation apart, rather than smashing it into pieces, and that was much easier to achieve if one

73

74

An exploration of the practices of Turkic horse-archers (and by this I mean their actual day-today drills, which are well-documented) demonstrates a clear ability to time the arrows’ fall – such a skill is actually required not to accidentally injure oneself during training. But unless a really large elevation difference occurred, many of the Cumans would not have had a clear field of fire. I believe that we must accept a general fire by the Cumans unless we are prepared to argue for an exceptional effectiveness of such weapons against not simply the mail, but also the large shields, which would be positioned to counter direct fire. We have no data that might illustrate how the Cumans shot during this period, as, unlike the constant motion involved to avoid a charge, the Cumans were clearly able to shoot at their leisure, and possibly pick their shots.

116

Russell Mitchell

could separate horsemen from their supporting infantry. The reserve guard at Adrianople, on the other hand, having literally just raised their siege in order to form a defensive curtain for those in flight, was not about to go anywhere. By paying careful attention to the equipment and tactics of the forces in play at 1205 Adrianople, rather than simply speaking about light and heavy cavalry as abstract concepts, we are able to learn useful things about both the Cuman “light cavalry” and the Crusader “heavy cavalry.” Unlike fights between western heavy cavalry and stand-off horse archers such as those encountered in the Holy Land, this was not a contest between two competing specialists. As we have seen from metallurgical data, against certain types of attack, metallic lamellar, banded scale, and similar armors were potentially much more protective than the mail and underpadding that the Crusaders wore, and assorted leather armors could, theoretically at least, match mail’s protectiveness.75 The so-called “heavy cavalry” of Western Europe was right to distinguish between heavily armored and lightly armored foes. But any notion of an armor evolution in these circumstances needs to be looked at skeptically. One could advance an argument that at the beginning of the First Crusade, most Western European cavalrymen were unfamiliar with lamellar defenses. However, one cannot begin to make this argument by the time of the Fourth Crusade, let alone the Seventh. Comparative arguments based on technological diffusion fall apart: by the time Constantinople fell to Baldwin and the Doge, these armors had been encountered on a regular basis by those on Crusade, and rejected, in favor of combining an exceptionally large and protective shield with gear that is theoretically less protective, at least against certain types of attack, but which has the advantage of protecting the warrior’s entire body. This leads us to a couple of conclusions. The first is that the notion of “heavily armored” in the High Middle Ages needs to be understood drastically differently than in the Late Middle Ages and Renaissance, with an emphasis on the degree to which the individual’s body was covered, rather than the extent to which the individual was equipped with various progressively more-advanced plate armors. The second conclusion, since the combination of mail and shield is a perfectly adequate defense for the purposes of close combat, particularly when supplemented with the somewhat earlier pourpoint and the somewhat later cuirie, is that rigid armors were adopted fairly grudgingly, and primarily as a defense against the increasing employment of missile troops on the battlefield.76 Similarly, the shield’s length diminishes almost directly in relation to the adoption of the couched lance technique, and begins to disappear in Western Europe as soon

75

76

Though these probably came nowhere near the protection provided by doubled mail. But note that it was the Cumans, not the Western Europeans, who sacrificed cutting edges on their lance heads in order to improve their ability to defeat heavy armor. After all, in this period, rigid and semi-rigid armors appear to have dominated in precisely those regions of Europe where battlefield archery was a serious threat. The lamellar and scale armors routinely depicted amongst the Cumans’ Byzantine, Russian, and Hungarian enemies make perfect sense in this respect.

Light Cavalry, Heavy Cavalry, Horse Archers

117

as the underlying armor worn is sufficiently heavy that it is no longer needed. Western heavy cavalry may have pursued a specialized – and widely successful – tactic, but their gear remained relatively general purpose, and one can argue that it was kept as light as the combination of battlefield threats and the shocktroop role would allow. The Cumans who crushed the Crusaders at Adrianople were by no means specialists, either. If anything, they were even more general purpose than the Crusading cavalry, able to fight both at close quarters and at a distance. They were horse archers, but not like those encountered on previous Crusades. Unlike other “nomad cavalry,” the Cumans built their tactics not only around the bow and their mounts’ famous endurance, but also around their ability to engage in whatever kind of fighting best fit the circumstances. The Cumans did not use the bow as their primary weapon, as did many peoples on the central Eurasian steppe, but as one tool among many, used to either harass an enemy, disrupt a formation by shooting the opponent’s horses out from underneath him, or else to force the opponent to engage them at close quarters under advantageous circumstances.77 The Cumans did not engage heavy cavalry on terms favorable to the latter.78 According to Choniates, the battle of Beroe was salvaged by the Emperor’s personal cavalry moving to attack the Cumans while they were engaged. The Cumans withdrew rather than put themselves into a “meat grinder” on the opponents’ terms. The same applies to the battle at Adrianople: even though the Cumans, as at Beroe, held the possibility of completely eliminating their enemy, they did not engage Villehardouin’s contingent. Indeed, they felt no need to do so, and Villehardouin’s chronicle itself makes it perfectly clear that the Crusaders’ nighttime flight is all that saved them. The Crusaders lost at Adrianople over the course of two days, and lost so badly – with the count and other notable nobles slain, and Emperor Baldwin captured – that Villehardouin records seven thousand Crusaders simply giving up and going home at the first opportunity, in spite of Villehardouin’s and other survivors’ best attempts to talk them into remaining.79 Kaloyan was able to completely dominate the Balkans so long as he had Cuman mercenaries available to him, and the resulting lack of manpower doomed the Latin Empire, allowed the Byzantine Empire a chance to regain its feet, and set in motion the events that would give rise to the Ottoman Empire’s foothold in Europe, as Ottoman mercenaries were called in to make up for the primary antagonists’ manpower shortages. The stage for all of this was set because on 14 April 1205, the Cumans

77

78 79

This is generally true of all Turkic peoples. However, the stereotypical image of a Turkic or steppe warrior is hardly that of a number of horsemen “wheeling about with lances.” While there may be a “Steppe norm,” it is equally likely that the steppe possessed numerous distinct variations. As this article describes, I believe those variations to be critically important once one is looking at specific engagements. See footnote 8, above: it is clear that in the same circumstances, Cumans would have simply mobbed Châtillon from all sides. Choniates, Annals, p. 218. Villehardouin, Chronicle, pp. 98–9.

118

Russell Mitchell

under Qoja’s command were able to force Emperor Baldwin’s Crusaders to give chase in spite of their public resolution not to do so. They did not do so because they were incompetent – by the Fourth Crusade, crusading cavalry knew better than to try to chase down light cavalry, and had proven themselves more than capable of keeping discipline in the face of archery. They did so because, unlike the horse archers that previous Crusaders had encountered in the Holy Land, the Cumans did not simply stand off at a distance and hit them with showers of arrows – instead, Qoja’s troops forced them to engage by opening the battle with a close-quarters attack, and forced them to remain engaged by more or less continually threatening immediate attack even when being chased. The Cumans, who were in regular combat with the Byzantines, Russians, and Hungarians, knew how to fight heavy cavalry, and how to take advantage of heavier cavalry’s dependence on the horse for striking power. Unlike these three groups, however, Baldwin’s forces were not so varied. At no point does Villehardouin ever depict the Emperor’s army fielding an equivalent to the Turcopoles that could operate as an effective screen against horse archers, or which could catch the Cumans and engage them in holding attacks that would allow the crusading heavy cavalry to hit them at close quarters and in good order.80 Why did they chase the Cumans? As we have seen, they chased them because they had to do so. But what put them into that position is all a matter of details. The Crusaders’ actions make no sense if we simply reduce each party’s capabilities to an abstract definition. The abstract, bird’s-eye view that we need in order to make meaningful historical comparisons fails us in the specific case. It’s only when we take the worm’s-eye view – and look beyond the abstract definitions to the fine print – that 1205 Adrianople begins to make sense.

80

Again, see footnote 62, above. The only variations that Villehardouin describes at all are individuals who are clearly expected to fight according to standard western doctrine, and are mentioned solely due to their perceived cowardice.

6 War Financing in the Late-Medieval Crown of Aragon1 Donald J. Kagay

Medieval soldiers would agree wholeheartedly with the political maxim which had grown hackneyed by the time of the Renaissance in its assertion that “money constitutes the sinews of war.”2 Medieval sovereigns often came to the painful conclusion that warfare required only three things: “money, more money, and yet more money;”3 they would also agree that its predictable end was the inexorable “increase in taxes.”4 To explore in very specific terms how these realities were managed by and simultaneously affected medieval sovereigns, this paper will focus on the family of realms of eastern Spain known as the Crown of Aragon during one of the longest and most expensive conflicts of its history known by later historians as the War of the Two Pedros (1356–66).5

1

2 3

4 5

I would like to thank Dr. Clifford Rogers and the other readers of this article for their insightful comments, most of which I have incorporated in this final version. Any remaining errors, of course, I claim as my own. J. R. Hale, War and Society in Renaissance Europe, 1450–1620 (Baltimore, 1985), p. 232. Ibid., 232; I. A. A. Thompson, “‘Money, Money, and Yet More Money’ Finance, the FiscalState, and the Military Revolution: Spain 1500–1650,” in The Military Revolution Debate: Readings on the Military Transformation of Early Modern Europe, ed. Clifford J. Rogers (Boulder, Colo., 1995), pp. 273–4. Hale, War, 232; Bruce D. Porter, War and the Rise of the State: The Military Foundations of Modern Politics (New York, 1994), p. 14. For the Crown of Aragon, see, T. N. Bisson, The Medieval Crown of Aragon. A Short History (Oxford, 1986); Donald J. Kagay, “The Institutional Blue Print of a Crusader Land: The Case of the Medieval Crown of Aragon,” The Journal of the Georgia Association of Historians 24 (2003): 25–60; Joan Reglá, Introducció a la historia de la Corona de Aragón (Palma, 1973). For War of the Two Pedros, see Antonio Gutiérrez de Velasco, “La conquista de Tarazona en la Guerra de los Dos Pedros (Año 1357),” Cuadernos de la Historia Jeronimo Zurita [hereafter cited as CHJZ] 10–11 (1960): 69–98; idem, “Las fortalezas aragonesas ante la gran ofensiva castellana en la Guerra de los Dos Pedros,” CHJZ 12–13 (1961): 7–39; idem, “La contraofensiva aragonesa in la Guerra de los Dos Pedros: Actitud militar y diplomática de Pedro IV el Ceremonoso (años 1358 a 1362),” CHJZ, 14–15 (1962): 7–30; María Teresa Ferrer i Mallol, “La frontera meridional valenciana durant la guerra amb Castella dita dos Peres,” in Pere el Cerimoniós I la seva època, ed. María Teresa Ferrer i Mallol (Barcelona, 1989), pp. 245–357.

120

Donald J. Kagay

I As early as the twelfth century, the fiscal exigencies of war acted as a powerful catalyst for the development of treasuries across Europe. In the Low Countries, England, and France, body servants to the sovereign inexorably took on a wider public role as collectors of war taxes and finally as members of an emerging treasury.6 As conflicts became more extended and costly in the late thirteenth and fourteenth centuries, treasury officials in both England and France grew expert in squeezing taxes and subsidies from increasingly restive populations – all in the name of the “evident necessity” and “common good” of the realm.7 After several decades of growing royal insolvency and desperate bureaucratic efforts to address it, all sources of war funding began to dry up in a firestorm of opposition from many different groups, caused by lengthening episodes of warfare. Far from totally rejecting the expensive martial policies of their sovereigns, the English Parliament and French Estates General, distrusting royal officialdom, began to assert their rights to control the fiscal course of such conflicts.8 The global cost of the decades of fighting associated with the Hundred Years War has been estimated at over £8,000,000 for England and at a much higher level for the French.9 Despite the fiscal creativity and ruthlessness royal officials in both states utilized to rake money into the overburdened treasuries of their sovereigns. The bottom line was more favorable to the English, however, because of strong monetary policies which monarchs down to Richard II (1377– 99) found difficult to overturn, despite a growing deficit.10 In France, Philip VI 6 7

8

9

10

Bryce Lyon and A. E. Verhulst, Medieval Finance: A Comparison of Financial Institutions in Northwestern Europe (Providence, RI, 1967), pp. 36, 47, 56. For these terms, see M. S. Kempshall, The Common Good in Late Medieval Political Thought (Oxford, 1999), pp. 179, 187, 249–61; Kenneth Pennington, The Prince and the Law, 1200– 1600: Sovereignty and Rights in the Western Legal Tradition (Berkeley, 1993), pp. 232–7; Joseph Canning, A History of Medieval Political Thought, 300–1450 (London, 1996), pp. 112–13, 128–30, 166. Ibid., 59; John Bell Hennemann, Royal Taxation in Fourteenth Century France: The Development of War Financing 1322–1356 (Princeton, NJ, 1971), pp. 5–10, 126, 155–6, 250, 285–6, 292, 322–3; idem, “Financing the Hundred Years’ War: Royal Taxation in France in 1340,” Speculum 42 (1967): 15–43; G. L. Harriss, King, Parliament, and Public Finance in Medieval England to 1369 (Oxford, 1975), pp. 231–69; Chris Given-Wilson, The Royal Household and the King’s Affinity: Service, Politics, and Finance in England 1360–1413 (New Haven, Conn., 1986), pp. 121–30. K. B. McFarlane, “War, the Economy and Social Change: England and the Hundred Years War,” Past and Present 22 (Jul., 1962): 6; M. M. Postan, “The Costs of the Hundred Years’ War,” Past and Present 27 (Apr., 1964): 40; Richard Bean, “War and the Birth of the Nation State,” Journal of Economic History 33 (1973): 215–17. The estimate for French expenses can not be as accurate as that of the English because of the many breaks in French accounts of the era and the great proliferation of tax collection areas. Peter Spufford, Money and its Use in Medieval Europe (Cambridge, 1988), p. 295; GivenWilson, Royal Household, 113. One of the greatest complaints again Richard was “the great and excessive costs of the court.”

War Financing in the Late Medieval Crown of Aragon

121

(1328–50) and his immediate successors soon ran short of operating funds and turned to the intermittent devaluation of their principal coinage. This deleterious practice proved a mainstay for the French war effort despite the bitter complaints of churchmen, nobles, and merchants that they were being “reduced to poverty” from the wildly fluctuating prices that closely followed the debasement.11 The same forces of increasing administrative efficiency and professionalism pitted against mounting costs and unwise monetary policies would hold sway far from the heartlands of the Hundred Years War in the Iberian Peninsula.

II In the 150 years before the great Castilian war of the mid-fourteenth century, both the central and fiscal administration of the Crown of Aragon steadily grew more complex and separate from the royal household. To protect and merge his scattered political and financial interests, sovereigns, such as the great warrior king, Jaume I (1213–76), utilized members of the royal family as “lieutenants” (locumtenentes) and also endowed several skilled, Jewish servitors with the office of “bailiff” (baiulus, batlle).12 Associated with the Templars from his earliest days as king, Jaume relied on the order for long-term revenue management and short-term loans when martial emergencies loomed.13 Much of the king’s fiscal planning was underpinned by a kind of “crisis financing,” consisting of a stream of loans from the king’s barons and prelates who were paid back from future tax revenues. These monies were converted into pensions (violaris) and annuities (censals mort) in the newly conquered lands of Valencia and the Balearics as well as among the Aragonese and Valencian urban populations.14

11

12

13

14

Spufford, Money, pp. 276, 300–1, 305–6; Hennemann, Royal Taxation, 355–7; Henry Miskimin, Money, Prices, and Foreign Exchange in Fourteenth Century France (New Haven, Conn., 1963). Donald J. Kagay, “Administration, Central, Aragon-Catalonia,” in Medieval Iberia: An Encylopedia, ed. E. Michael Gerli (New York, 2003), p. 214; Robert I. Burns, S.J., Medieval Colonialism: Postcrusade Explotation of Islamic Valencia (Princeton, NJ, 1975), pp. 274–75; J. Lee Shneidman, The Rise of the Aragonese-Catalan Empire 1200–1350, 2 vols. (New York, 1970), 2:471–5; Jesús Lalinde Abadía, La gobernación general en la Corona de Aragón (Madrid/Zaragoza, 1963), pp. 24–40; J. N. Hillgarth, “The Royal Accounts of the Crown of Aragon,” in Spain and the Mediterranean in the Later Middle Ages: Studies in Political and Intellectual History (Aldershot, Hampshire, 2003), study III, p. 5. Thomas N. Bisson, “Las finanzas del joven Jaume I (1213–1228),” in X Congreso de historia de la corona de Aragón [hereafter XCHCA] (Jaume X y su época) (Zaragoza, 1975), Comunicaciones 1–2, pp. 170–5. For Templar banking, see Alan Forey, The Military Orders from the Twelfth to the Early Fourteenth Centuries (Toronto, 1991), pp. 114–16; The Templars, ed. Malcolm Barber and Keith Bate (Manchester, 2002), pp. 14–15. Burns, Medieval Colonialism, pp. 221, 237–40; Max Turull, “Finances i fiscalitat municipals a Catalunya durant la baixa edat mitjana,” L’Avenç 133–43 (1990): 62; Manuel Sánchez Martínez, Pere Orti Gost, Max Turull Rubinat, “La genesis de la fiscalidad municipal en Cataluña,” Revista d’Història Medieval 7 (1996): 115–34.

122

Donald J. Kagay

The next stage of eastern Spanish, fiscal development was ushered in with the reign of Pere II (1276–85) and his immediate successors. Adapting the efficient accounting methods of his new Sicilian realm, Pere also imported into the Crown of Aragon such new fiscal offices as the “scribe of accounts” (escrivá de ració) and the “master of accounts” (maestre racional).15 Bringing new accounting and banking methods to eastern Spain, these officials also incurred such discontent that Alfons II (1285–92) found it expedient to discontinue the Sicilian innovations. He then restored the power of the “treasurer” (tresorer).16 With the increased volume of royal revenues and the accession of Jaume II (1291–1327), a new sovereign fresh from the rule of the politically sophisticated realm of Sicily, the maestre racional reappeared in royal government, again sharing power with the treasurer.17 To attain the necessary modernization of the rather haphazard fiscal administration he had inherited, Jaume was thus forced to marry a “new financial superstructure” to the older forms.18 With the reign of the cautious Aragonese monarch, Pere III (1336–87), officials found their duties laid out in minute detail in the Ordenacions (1344), a simplified version of the Majorcan household guide, the Leges Palatinae.19 In this careful expression of governmental protocol, the maestre racional emerges as a kind of fiscal manager who audited the books of bailiffs and other curial officials, established the assessment rate for all imposts, and received the vouchers, receipts, and other fiscal paper work of subordinate officials, eventually incorporating all these records into a set of general accounts.20 A commission of

15

16 17

18 19

20

Arxiu de la corona d’Aragó [hereafter ACA], Cancillería real, R. 1529, ff. 4v–5; Kagay, “Administration,” p. 15; Marta Vanlandingham, Transforming the State: King, Court, and Political Culture in the Realms of Aragon (1213–1387) (Leiden, 2002), pp. 208–9; Tomás de Montagut i Estragués, El maestre racional a la Corona d’Aragó (1283–1419), 2 vols. (Barcelona, 1987), 1:44–6, 62–7. ACA, Cancillería real, R. 74, ff. 98, 102; Vanlandingham, Transforming, pp. 216–19. Luis G. de Valdeavellano, Curso de Historia de las Instituciones españoles. De los origenes al final de la Edad Media (Madrid, 1982), pp. 594–5; J.E. Martínez Ferrando, “Jaume II,” in Els descendents de Pere el Gran (1954; reprint, Barcelona, 1980), pp. 61–2; Denis Mack Smith, Medieval Sicily 800–1713 (New York, 1968), pp. 77–8; Enrique Cruselles, El maestre racional: función política administrativo del oficio público (Valencia, 1989). Hillgarth, “Royal Accounts,” p. 5. James III, King of Majorca, Leges Palatinae (facsimile edition), ed. José J. de Olañeta (Palma de Mallorca, 1994); Thomas N. Bisson, “Statebuilding in the Medieval Crown of Aragon,” in XV congrés d’história de la corona d’Aragó (El poder real en le Corona de Aragó, Siglos XIV–XVI) [hereafter XV CHCA], 5 vols. (Zaragoza, 1993), 1:148; Hillgarth, “Royal Accounts,” p. 6; Montagut i Estragués, Maestre racional, 1:79–80, 120–4. “Ordenacions fetes per lo molt alt senyor En Pere Perç rey d’Aragó sobra lo regiment de tots los officials de la sua cort,” in Colección de documentos inéditos del archivo general de la corona de Aragón [hereafter CDACA], ed. Próspero de Bofarull y Moscaró, 41 vols. (Barcelona, 1847–1910), 5:150–5, 161–2. For full discussion of late medieval, Iberian courts, see Rita Costa Gomes, The Making of a Court Society: Kings and Nobles in Late Medieval Portugal, trans. Alison Aiken (Cambridge, 2003), pp. 27–30; J. Jordán de Urríz y Azara, “Las Ordenaciones de la Corte Aragonesa en los siglos XIII y XIV,” Boletín de la Real Academic de Buenas Letras de Barcelona 13 (1913): 220–9, 284–92.

War Financing in the Late Medieval Crown of Aragon

123

1348 gives a clear idea of the maestre racional’s duties. It enjoined the official to “demand and receive” all money owed the crown and to pay all sums of money which he was directed to by the sovereign.21 The treasurer, who maintained the bulk of royal funds, made disbursements for both subjects and officials whom the Crown owed money.22 The scribe of accounts, the treasury’s agent in the field, was responsible for tax collection and the settlement of royal debts on the local level.23 Pere lavished a great deal of care on spelling out the duties of his fiscal officials so his household and, by extension, his government would be “wisely arranged in all things.”24

III In some ways, the curial and governmental officials of the medieval Crown of Aragon were organized along disciplinary lines not unlike armies of the time. At the commencement of their service, fiscal bureaucrats swore solemn oaths to reject favoritism and extortion while advancing the royal patrimony and maintaining the secrecy of court matters.25 These same promises bound officials even more when their master marched off to war and “lived under tents and pavilions.”26 During these dangerous periods, the king’s officials were crucial not only as tax collectors and paymasters, but also as spies and trusted messengers to field commanders.27 Royal government and its “financing on the fly” was often well adapted to the exigencies of intermittent border warfare so common in most phases of the Spanish reconquista.28 As Jaume I made war on Islam almost continually between 1229 and 1244, however, fiscal officials were forced to one height of creativity and determination after another in making royal means fit military demands.29 Even though Jaume the Conqueror could rely on large feudal hosts from the clergy, nobility, and townsmen of his lands, these troops had to be supplied and,

21 22 23 24 25 26 27 28

29

ACA, Patrimonia real, Maestre Reacional, R. 642, ff. 317v–8; Hillgarth, “Royal Accounts,” p. 8, n. 29. “Ordenacions,” pp. 158–9. Ibid., 161–2. Ibid., 149–50. Ibid., 156, 159–60, 186–7. Ibid., 84. Ibid., 163. Derek W. Lomax, The Reconquest of Spain (London, 1978), pp. 99–101; James F. Powers, A Society Organized for War: The Iberian Municipal militias in the Central Middle Ages 1000– 1284 (Berkeley, 1988), pp. 102–4; Antonio Palomeque Torres, “Contribución al estudio del ejercito en los estados de la reconquista,” Anuario de Historia del Derecho Español 15 (1944): 319–41. For Jaume’s own account of these conquests, see The Book of Deeds of James I of Aragon [hereafter BD], trans. Damian Smith and Helena Buffery (Aldershot, 2003), pp. 69–282 (chaps. 47–337).

124

Donald J. Kagay

if kept in the field long enough, had to be paid a daily wage.30 The fuel of government operations, the traditional revenues associated with clerical and urban institutions, were difficult to convert into the ready cash that lengthy military campaigns could not do without. To gain this readily spendable income, Jaume occasionally auctioned off future revenues for a wholesale price.31 As the reality of military service drew ever closer, royal officials could harvest from a windfall of funds made in lieu of military service or as a fine for non-compliance with the king’s military summons. Though providing ready money, these expedients shackled the king with the promise that he would not collect them again for a specified time.32 To fill the fiscal void for his army in the meantime, Jaume relied on longterm loans from foreign bankers and repeated “exactions” (questiae) from the churchmen and townsmen as well as from Jewish and Muslim “communities” (aljamas) in his own realms.33 For larger subsidies such as the Catalan bovatge and Aragonese monedaje, the king had to grudgingly go before his parliaments (cortes, corts).34 For a man of action like Jaume, this proved exceedingly unpleasant since the members of the cortes proved very penurious paymasters who larded every extraordinary grant with galling conditions, which often attempted to hem in royal power.35 In spite of the steady escalation of military costs which had almost doubled over Jaume’s lifetime,36 the promise of rich Muslim plunder and the advancement 30

31

32 33

34

35

36

Ibid., 154 (chap. 151); Donald J. Kagay, “Structures of Baronial Dissent and Revolt under James I (1213–76),” in War, Government, and Society in the Medieval Crown of Aragon (Aldershot, Hampshire, 2007), study VII, p. 64; idem, “Army Mobilization, Royal Administration, and the Realm in the Thirteenth-Century Crown of Aragon,” in Iberia and the Mediterranean World of the Middle Ages, ed. P. E. Chevedden, D. J. Kagay, and P. G. Padilla, L. Simon, 2 vols. (Leiden, 1996), 2:100–1. For traditional revenues of the Crown of Aragon, see Rentas de la antigua Corona de Aragón, ed. Manuel de Bofarull y de Sartorio in CDACA, vol. 39. For sale of revenues, see Christian Guilleré, “Les finances de la Corona d’Aragó,” L’Avenç 133–43 (1990): 54–5; Bisson, “Finanzas,” pp. 186–8. ACA, Cancillería real, R. 23, ff. 8v–9, 70v; Cartas reales [Jaime I] (extra series), no. 64; Pergaminos de Jaime I, no. 1916; Kagay, “Army Mobilization,” pp. 102. ACA, Cartas reales [Jaime I], no. 61; Luis J. Fortun Perez de Ciriza, “Relaciones financieras entre Sancho el Fuerte y los monarchos de la Corona de Aragón,” in X CHCA, Comunicaciones, 3, 4, y 5: 171–2; A. J. Forey, The Templars in Corona de Aragon (London, 1964), p. 350; Yitzak Baer, A History of the Jews in Spain, trans. Louis Schoffman, 2 vols. (Philadelphia, 1961), 1:84–6; Kagay, “Army Mobilization,” pp. 102–3. Jaume’s parliaments dealing with war funding and other military matters were: Tortosa (1225), Barcelona (1228), Monzón (1232, 1236), Barcelona (1264), Zaragoza (1264), and Lerida (1272, 1275). Jaume and his people commonly accepted the idea that “in such important matters as … [war], it is necessary to celebrate the Corts.” The king also knew full well that “the members of the Corts are generally divided whenever we ask for advice on a great matter because they cannot agree well” [BD, 286–87 (chaps 381–2)]. ACA, Cancillería real, R. 20, ff. 342, 345; Kagay, “Army Mobilization,” pp. 104–5. For contemporary military costs, see Philippe Contamine, War in the Middle Ages, trans. Michael Jones (Oxford, 1990), pp. 94–7; Michael Prestwich, Armies and Warfare in the Middle Ages: The English Experience (New Haven, Conn., 1996), pp. 84–8.

War Financing in the Late Medieval Crown of Aragon

125

of status made royal service appealing and somewhat lowered the expense of Jaume’s armies which could number up to 15,000 men.37 With the conquest and distribution of the riches of the Balearics and Valencia, plunder as an inducement for military support and service largely disappeared. To defray military costs and assure that his soldiers in the field were paid, Jaume increasingly had to admit, as he did in 1264, that without the cortes “the business of war could not be arranged.”38 Such cash wrangling from the recalcitrant members of the parliament was hardly Jaume’s long suit. He complained that these assemblies were “generally divided in opinion … and could never be made to agree.”39 Even with a desperate need for money to support his armies, Jaume’s temper occasionally forced him to leave a cortes in high dudgeon “as much displeased as any lord ever was with his people.”40

IV The difficulties Jaume the Conqueror would face in extracting war funds from his people presaged quite clearly the fiscal travails of his great-great-grandson, Pere III (1336–87), when caught in the vice of a bitter, decade-long border war with his Castilian arch-enemy, Pedro I (1350–66/69). In the two decades since his accession to the throne in 1336, Pere the Ceremonious had engaged in a number of costly conflicts including the lingering duel with Genoa over the control of the island of Sardinia and other central Mediterranean territories,41 as well as the naval war against his cousin, King Jaume III of Mallorca (1324–49) which brought about the Aragonese conquest of the Balearics (1343), and an Mallorcan invasion of Catalonia (1344–49) which eventually cost Jaume his life.42 On the domestic front, Pere spent much of the late 1340s in suppressing the baronial and urban insurrection of the Unión, which, under the leadership of the king’s half-brother, Prince Ferran (d. 1363), brought great

37

38 39 40 41

42

For distribution of plunder, see Repartamiento de Valencia, ed. Próspero de Bofarull y Mascaró in CDACA, 11:143–656; Repartamiento de Mallorca, ed. Próspero de Bofarull y Mascaró in CDACA, 13:1–141. For estimates of Jaume’s armies, see Alvaro Santamaria, “La expansion politico-militar de la Corona de Aragón bajo la dirección de Jaime I: Baleares,” in XCHCA, Ponencias, 122–3. BD, 286 (chap. 381). Ibid., 287 (chap. 382). Ibid., 288–9 (chap. 384). Antonio Arribas Palau, La conquista de Cerdeña por Jaime II (Barcelona, 1952); David Abulafia, The Western Mediterranean Kingdoms 1200–1500: The Struggle for Dominion (London, 1997), pp. 123–8, 179–80; Steven A. Epstein, Genoa and the Genoese, 958–1528 (Chapel Hill, NC, 1996), pp. 222–3. Pere III of Catalonia (Pedro IV of Aragon), Chronicle [Pere III], trans. Mary Hillgarth, ed. J. N. Hillgarth, 2 vols. (Toronto, 1980), 1:253–78 (III:21–41); J. E. Martínez Ferrando, La tràgica história dels reis de Mallorca (1960; reprint, Barcelona, 1979), pp. 118, 113–35.

126

Donald J. Kagay

portions of Aragon and Valencia into rebellion against the crown.43 In all of these campaigns, the fiscal weaknesses of the Crown of Aragon and the constant financial and bureaucratic adaptation Pere engaged in to pay for his wars long and short, is clearly evident. No matter which enemy he faced, the Aragonese ruler encountered similar economic barriers created by the directives of royal and municipal law in his realms. The initial problem Pere and many other medieval kings at war faced was the rather limited extent of the royal revenue base. The sovereigns of the Crown of Aragon were financially supported by a hotchpotch of regular and extraordinary dues that were somewhat different in each of his realms. The perennial revenues emerged from such feudal obligations as “hospitality” (cena), from such military dues as “horse support” (cabelleria), and from such economic imposts as the “excise tax” (lezda, leuda) and the “toll” (peaje, peatge).44 The extraordinary sources of Pere’s income came from imposts such as the bovatge and mondatge which were assessed on the entire realm but only once in a king’s lifetime as well as from such military dues as the redemptio exercitus and the defectus servitii which were paid in lieu of or for avoidance of royal military service.45 Even with this steady income, however, the king, especially when forced to wage a lengthy war, was prone to live beyond his means. To make up such shortfalls, Pere and many of the urban centers he relied on for extraordinary subsidies sold “government bonds” (censales, censals) and auctioned off “future tax revenues” (violarios, violaris). When in desperate fiscal straits, the Aragonese monarch fell back on the dangerous expedient of selling off segments of the royal patrimony. Despite all of these measures, Pere, from the increasing financial pressure brought about by year after year of conflict, had no choice but to follow the dangerous course of deficit spending.46 Though Pere was supported by feudal and national defense levees who had to fight at their own expense when their lord and king was in danger, these

43

44

45 46

Pere III, 2:433–5 (IV:45); Vincente Boix y Ricarte, Crónica de la Provincia de Valencia, 3 vols. (Madrid, 1867), 3:56. For Prince Ferran and the Unión, see María Teresa Ferrer i Mallol, “The Southern Valencian Frontier during the War of the Two Pedros,” in The Hundred Years War: A Wider Focus, ed. L. J. Andrew Villalon and Donald J. Kagay (Leiden, 2005), pp. 80–1, 106–7; Shneidman, Rise, 2:500–2; Manuel Dualde Serrano, “Tres episodios de la lucha entre ‘Pere el de Punyalet’ y la Unión aragonesa, relatados por el monarca a su tió Pedro, conde de Ribagorza,” Estudios de la Edad Media de Aragón 2 (1946): 350–1; Donald J. Kagay, “The Captaincy System in the Late-Medieval Crown of Aragon: An Administrative Scaffolding of Defense and Imperialism,” Studies in Medieval and Renaissance History, forthcoming. Jaime Vicens Vives, Manual de historia economica de España, 2 vols. (Barcelona, 1959), 1:215–16; Valdeavellano, Curso, p. 607; Pierre Bonnassie, La Catalogne du milieu du Xe à la fin du Xe siècle: Croissance et mutation d’une société (Toulouse, 1975–76), 1:751–2. Kagay, “Army Mobilization,” pp. 102–3. Vicens Vives, Manual, 1:216–19.; Hillgarth, “Royal Accounts,” p. 10. For the sale of bond and future tax revenues, see notes 13 and 31 above.

War Financing in the Late Medieval Crown of Aragon

127

uncompensated troops were hardly free and seldom proved militarily effective.47 To get better value for his money, the king generally relied on paid soldiers who received a daily wage of between 3 and 8 sous depending on the armor they wore, the weapons they wielded, and the horses they rode.48 While the campaigns against Sardinia of 1323–24 and 1354–55 required between 10,000 and 11,000 troops, and that against Jaume III in 1343 involved some 6,000 men, the great expenses associated with such sizeable conflicts were fairly short-lived.49 By contrast, the first phases of the Castilian war seldom involved more than 4,000 Aragonese, Catalan, and Valencian troops as well as Castilian mercenaries per year.50 The difference from the earlier wars was clearly the length of the conflict with Pere’s “principal adversary,” Pedro the Cruel. The many years of combat, even though carried out at a fairly low level, made the payment of sizeable military units a regular item on Pere’s annual expense sheet. In spite of the great fiscal pressure this placed on his realms, the king could hardly skimp on military spending, given the vast demographic advantage of Castile over the Crown of Aragon.51

V Ironically, Jaume the Conqueror’s “war machine,” marked by continual fiscal, political, and military accommodations to gain its rock-solid military objectives, was largely recreated in Pere the Ceremonious’s conflict against Castile. Even the theater of the struggle–the far western Aragonese frontier bounded by such cities as Calatayud, Teruel, and Tarazona and the southern Valencian borderland opposite Murcia–had long stood as zones of defense against Islam and witnessed some of Jaume’s great victories against his Muslim

47

48

49 50 51

For a good example of the defective nature and great expense associated with such troops, see Manuel Sánchez Martínez, “The Invocation of Princeps namque in 1368 and its Repercussions for the City of Barcelona,” in The Hundred Years War: A Wider Focus, ed. L. J. Andrew Villalon and Donald J. Kagay (Leiden, 2005), pp. 297–332. Maria Teresa Ferrer i Mallol, “La organización militar en Cataluña en la Edad Media,” Revista de historia militar, extra number 45 (2001), p. 178; Donald J. Kagay, “A Government Besieged by Conflict: The Parliament of Monzón (1362–1363) as Military Financier,” in Hundred Years War, pp. 140–1. Ferrer i Mallol, “Organización,” pp. 171–3; Arribas Palau, Conquista, p. 159. Ferrer i Mallol, “Organización,” pp. 174; idem, “Frontera meridional,” pp. 252–3, 266–70. The estimates drawn from tax and clerical records put the population of Castile in the midfourteenth century at approximately 4,500,000 and that of the contemporary Crown of Aragon at a little over 1,100,000. This demographic advantage is also reflected in city populations of the time. Castile thus had thirteen cities with more than 20,000 residents while the Crown of Aragon had only two. N. J. G. Pounds, An Historical Geography of Europe (Cambridge, 1990), p. 151; Josiah Cox Russell, Medieval Regions and their Cities (Newton Abbot, 1972), pp. 169, 175, 190, 195.

128

Donald J. Kagay

neighbors.52 The same land fought over by Cross and Crescent would now form the tilting ground for two men of mediocre Christianity. These areas, which endured war for a full ten years, have been described by R. I. Burns as “a series of irrigated huertas and lively commercial cities set like a necklace of emeralds and pearls”53 around swaths of treeless badlands characterized by the unflappable, nineteenth-century traveler, Richard Ford, as some of “the bleakest and most inhospitable land in the whole of Spain.”54 The origins of the Castilian war were as muddled and petty as its two major protagonists. After a century of sniping at each other over borders between Castile, Aragon, and Valencia, the encouragement of domestic rebellion across these fluid borderlands, and dynastic disagreements between the Aragonese and Castilian royal families which were often bound by marriage ties,55 Castile and the Crown of Aragon commenced hostilities in 1356 for ostensibly unimportant reasons.56 From a remarkable series of letters which Pere and Pedro exchanged in the first months of the war, it is obvious that the monarchs were motivated by deepseated enmity and frustration. The self-images projected in these missives are crucial for understanding the psychological parameters in which the Castilian war was fought. Portraying himself as a friend of Aragon whose reputation had been cruelly maligned by Pere, Pedro swore to avenge his honor no matter how long it took. Pere, on the other hand, comes across as an innocent bystander befuddled by the madness of the Castilian king’s brutal over-reaction, but one equally determined to defend his scattered lands from every Castilian threat.57

52

53 54 55

56

57

Bonifacio Palacios Martin, “La frontera de Aragón con Castilla en la época de Jaime I,” in XCHCA, Comunicaciones 1–2, 490–5; Miguel Angel Ladero Quesada, “Sobre la evolución de las fronteras medievales hispánicas (siglos XI a XIV),” in Identidad y representación en a España medieval (siglos XI–XIV) (Madrid, 2001), pp. 17, 21–44; J. M. Lacarra de Miguel, “Acerca de las fronteras del Ebro (siglos VIII–XII),” En la España Medieval 1 (1980): 181–91. Burns, Medieval Colonialism, 14. Richard Ford, A Hand-book for Travellers in Spain and Readers at Home, ed. Ian Robertson, 3 vols. (1845; reprint, Carbondale, Ill., 1966), 2:606–7. M. T. Ferrer i Mallol, “Causes i antecedents de la guerra dels dos Pedros,” Boletín de la Sociedad Castellonense de Cultura 43 (1987): 446–50; Joseph F. O’Callaghan, “Kings and Lords in Conflict in Late Thirteenth-Century Castile and Aragon,” Iberia, 2:117–38; Angeles Masiá de Ros, Relación castellano-aragonesa desde Jaime II a Pedro el Ceremoniosos, 2 vols. (Barcelona, 1994) 1:56–72, 127–8. The war commenced in late summer of 1356 when Francesç Perrellós, a Catalan privateer, captured two Italian merchantmen at Cadiz. The ships were under the protection of Genoa and their ally, Castile, Pedro I viewed the attack as a clear provocation. He demanded restitution from Perellós’s overlord, Pere III, and when it was not immediately forthcoming, he broke feudal ties with Pere III which served as a declaration of war. Pere III, 2:495–6 (VI:3); Pedro López de Ayala, Crónica de Pedro I de Castilla, (Biblioteca de Autores Españoles, 66), ed. Llaguno y Amírola, 3 vols. (Madrid, 1953), 1:473–4 (1357:7–8); Jerónimo Zurita y Castro, Anales de la Corona de Aragon, ed. Angel López Canellas (Zaragoza, 1967–85), 4:292–3 (IX:i); María Teresa Ferrer i Mallol, “The Southern Valencian Frontier during the War of the Two Pedros,” in The Hundred Years War, pp. 76–7. Pere III, 2:498–503 (VI:3–4); ACA, Cancillería real, R. 1379, ff. 12v–15v; Masiá de Ros, Relación, 1:249–53.

War Financing in the Late Medieval Crown of Aragon

129

The tone of these letters–offensive for Pedro and defensive for Pere–would determine the overall trend of the conflict for the next eight years. Pere’s role as defender, though approved as intelligent by Papal and royal theorists of the time,58 was, in reality, dictated by a tightening fiscal bind that forced the king to engage in both time-worn financial practices and numerous administrative innovations. The surprising speed of Castilian attack in the first few months of the war induced Pere to name in great haste “frontier captains” (frontalers) who would rely on both feudal and paid levies.59 The wages were divided into two classes– “heavy cavalry” (caballo armado, cavall armat) and “light cavalry” (cavall alforrat, caballo desarmado)–with the first rank drawing the highest pay.60 Because of the difficulty of keeping these paid forces on the frontier indefinitely, Pere (at least in Catalonia) could call out troops under the auspices of Princeps namque, the national defense clause of the Usatges of Barcelona.61 Though this expedient allowed the sovereign to mount a sizeable force which served for national duty rather than for pay, Pere soon found that such national defense contingents proved better consumers of supplies than they were warriors. He thus called them out sparingly and only when more professional troops were not available.62 Largely mistrusting his uncompensated national defense levees, the beleaguered Aragonese king had no choice but to rely on paid troops. To pay them, he squeezed money from traditional, extraordinary, and hitherto unheard of sources. His war aims were thus twofold: (1) to put men in the field who would resist the ever-present threat of enemy incursion along his frontiers and (2) to find sufficient money to pay and supply these defenders.

58

59

60 61

62

Joseph R. Strayer, Medieval Statecraft and the Perspectives of History (Princeton, NJ, 1971), pp. 295–8; Donald J. Kagay, “The Defense of the Crown of Aragon during the War of the Two Pedros (1356–1366),” The Journal of Military History, 71 (2007): 11–34. ACA, Cancillería real, R. 1381, ff. 41v–42, 48v, 49v, 183v; R. 1383, f. 173; Donald J. Kagay, “A Shattered Circle: Eastern Spanish Fortifications and their Repair during the Calamitous Fourteenth Century,” in War, study III, p. 130. ACA, Cancillería real, R. 20, ff. 342, 345; Kagay, “Army Mobilization,” p. 105; idem, “Government,” pp. 140–1. The Usatges of Barcelona: The Fundamental Law of Catalonia, trans. Donald J. Kagay (Philadelphia, 1994), p. 80 (art. 64); Donald J. Kagay, “The National Defense Clause and the Emergence of the Catalan State: Princeps namque Revisited,” in Crusaders, Condottieri and Cannon: Medieval Warfare in Societies around the Mediterranean, ed. L. J. Andrew Villalon and Donald J. Kagay (Leiden, 2003), pp. 57–100. For a list of the occasions on which Pere invoked Princeps namque, see ACA, Cancillería real, R. 1519, ff. 1–71, 72–73v; Kagay, “National Defense Clause,” 95–97.

130

Donald J. Kagay

VI The rapidity of the Castilian attacks in the fall of 1356 left the Aragonese king gasping for fiscal breath.63 Desperate to post garrisons in his exposed castles on the Aragonese and Valencian frontiers, Pere could not long rely on the slowmoving revenue system that supported the Crown and its retainers who controlled castles.64 Since the birth of both Aragon and Catalonia in the tenth and eleventh centuries, great and middling nobles had held castles as fiefs and were supposed to support their garrisons from the “castellany” (castellania, castlania)– a complex of rents and taxes paid by the surrounding populace to the castellan.65 Such fief holders, however, often made use of the castle revenues without carrying out the duties of castle and garrison maintenance.66 Because of this lapse of feudal defense, especially in areas where a Muslim threat had long before been eradicated, Pere had few options but to pour ready cash into the protection of his exposed frontiers from periodic enemy attacks. While the Castilian king ultimately faced the same fiscal exigencies as his Aragonese counterpart, he was initially able to minimize some of his military costs with the conquests his troops made in Aragon and Valencia.67 Pere, on the other hand, had few such options and the financial demands of defense thus periodically left him inundated by “great and immense expenses.”68

63

64 65

66

67

68

For this confused period down to Pedro I’s conquest of the Aragonese town of Tarazona in 1357, see Gutiérrez de Velasco, “Conquista de Tarazona,” 83–9; Zurita, Anales, 4:309–17, 321–7 (IX:v–vii, ix–x). For an early fourteenth-century list of these castle revenues, see Rentas in CDACA, 39:315– 74. Pierre Bonnassie, From Slavery to Feudalism in South-Western Europe, trans. Jean Birrell (Cambridge, 1991), pp. 166, 178–9; Flocel Sabaté, “Les castlanies i la comissió reial de 1328,” in Estudios sobre renta, fiscalidad y finanzas en la Cataluña bajomedieval, ed. Manuel Sánchez Martínez (Barcelona, 1993), pp. 177–241. ACA, Cancillería real, R. 231, ff. 52–3; Cartas reales [Jaime II], no. 9300; María Teresa Ferrer i Mallol, “La Tinença a custum d’Espanya en els castells de la frontera meridional Valenciana (Segle XIV),” in Miscellània de textos medievals (Barcelona, 1988), pp. 50–1, 58–9 (doc. 4, 15); The Customs of Catalonia between Lords and Vassals by the Barcelona Canon, Pere Albert: A Practical Guide to Castle Feudalism in Medieval Spain, trans. Donald J. Kagay (Tempe, Ariz., 2002), pp. 13, 16–17 (arts. 15, 18–19). Gutiérrez de Velasco, “Conquista de Tarazona,” 83–7; idem, “La financiación aragonesa de la ‘Guerra de los Dos Pedros’,” Hispania 19 (1959): 25–6. Though Pedro possessed armies ranging from some 9,000 up to 42,000 men in the major campaigns against Alicante, Tarazona, and Teruel, the raids he unleashed on the Aragonese and Valencian frontiers normally involved small units of a few hundred men. M. T. Ferrer i Mallol, “Organización,” pp. 174–5. ACA, Cancillería real, R. 331, f. 46; Documenta Selecta Mutuas Civitatis Arago-Cathalaunicae et Ecclesiae Relationes Illustrantia [hereafter DS], ed. Joannes Vincke (Barcelona, 1936), pp. 30–1 (doc. 58).

War Financing in the Late Medieval Crown of Aragon

131

To pay off his mounting war bills, Pere time and again had to turn to the various classes or estates of eastern Spanish society for several different types of monetary grants. In the first weeks of the Castilian war, he managed to pressure individual churchmen, monasteries, military orders, and members of his Jewish and Muslim communities to make sizeable contributions “for the good of the commonwealth of our lands.”69 In the same period, a number of knights made emergency loans to their cash-strapped sovereign; there is little indication that these were ever repaid.70 The widening military crisis of 1356–57 soon consumed these small fonts of money and led the king to even more drastic measures. In February 1357, he began pawning royal lands, a process that would continue throughout the war.71 In April of the same year, Pere wrote to his “dear companion,” Queen Elionor, instructing her to pawn her “banner and as many of her jewels” as it took to raise at least 50,000 sous.72 Though not often forced to resort to such humiliating measures in his war with Castile, the Aragonese king had few compunctions about engaging in them when absolutely necessary. Thus, in 1364, when Pedro held Valencia under siege,73 Pere ordered that all the silver ornaments in the churches of the beleaguered city be pawned to pay the salary of the troops he had brought into the southern capital. “For the release of our conscience,” Pere insisted that his treasurer generate a thorough description of each ornament, presumably so they could be redeemed at some time in the future.74 Though these examples represented immediate fiscal emergencies for the Crown, Pere’s overall war effort put him under constant financial pressure that he could only assuage by selling royal properties (including towns) and the revenues they generated.75 This action which reduced royal rents and other operating funds by half during the Castilian war intensified the king’s personal financial embarrassment.76 Despite these emergency measures, the Aragonese war expenses showed no signs of decreasing until the late spring of 1357 when the efforts of a Papal

69 70 71 72

73 74 75 76

ACA, Cancillería real, R. 1149, f. 74v; R. 1327, f. 201; R. 1379, ff. 93r–v, 95v, 134.150v–1; R. 1380, ff. 109v–110,134; DS, 424–5, 431 (docs. 560, 566). ACA, Cancillería real, R.. 1380, f. 44. ACA, Cancillería real, R. 1379, f. 134. ACA, Cancillería real, R. 1152, f. 172; Documents Historichs Catalans del Segle XIV : Colecció de cartas familars corresponents al regnats de Pere del Punyalet y Johan I [hereafter DHC] (Barcelona, 1889), p. 59. For this crucial Valencian campaign, see Pere, III, 2:562–7 (VI:51–2); Zurita, Anales, 4:485– 7, 499–504 (IX:li,liv). ACA, Cancillería real, R. 1197, f. 184; DHC, 59–60. There is little indication that such a tally was compiled or that the ornaments were indeed returned to their original owners. Gutiérrez Velasco, “Financiación,” pp. 16–17. Manuel Sánchez Martínez, El naixement de la fiscalitat d’Estat a Catalunya (segles XII–XIV) (Gerona, 1995), pp. 85–7, 117; Hillgarth, “Royal Accounts,” p. 12. For similar fiscal strategies in Habsburg Spain, see Helen Nader, Liberty in Absolutist Spain: The Hapsburg Sale of Towns, 1516–1700 (Baltimore, 1990).

132

Donald J. Kagay

nuncio bore fruit with the arrangement of a truce between the warring parties.77 When Pedro rapidly violated this instrument, the Aragonese king, suffering from growing insolvency, turned to more formal and general means of gaining operating funds for his armies.78 As the first phases of the Castilian war passed from one disaster to another for the Aragonese king, he came to the obvious conclusions that “his soldiers [could not] do their duty without money”79 and that he himself would not be able “to sustain the burden of the war without the assistance of his subjects.”80 To accomplish this aim in the short term, Pere resurrected a lucrative impost utilized by Jaume I, the redemptio exercitus, a one-time payment that allowed individuals or corporate groups to buy exemption from military service for a specified period.81 He also began to privatize future royal revenues and taxes to his subjects by permitting clerical and urban authorities to sell them for a base price and for a specified period to individuals with the proviso that the money be used for paying military salaries or repairing fortifications.82 The king also allowed Enrique de Trastámara and other captains who controlled a great number of castles to supply the fortifications from the population that surrounded them.83 Though their paid protectors, people living in the jurisdictional shadow of castles found Pere’s captains and their fiscal agents as dangerous as any Castilian raider and often met them with “injurious and threatening words of death.”84 With the mounting financial pressure the war placed on the urban, clerical, and infidel communities, tax redistribution schemes proved ineffective for the tax middlemen and the Crown. Instead, both parties increasingly relied on fiscal negotiation on an ever wider scale. To accomplish this complex task, Pere appointed “commissioners” (commissarii, comissaris) from his court officials who would work out with the leaders of an entire group the parameters of a general “subsidy” (subsidium) or a “grant in aid of war” (auxilium, auida, ayuda, proferta). The commissioners would then oversee the collection of these funds and turn over the proceeds to treasury officials who would begin the process of distributing them to forces in the field. This process proved much more effective for the royal administration and was also preferred by the various groups since, in theory,

77 78

79 80 81 82 83

84

Zurita, Anales, 4:324–32 (IX:x–xi); Ferrer i Mallol, Southern Valencian Frontier, pp. 84–7. For the question of the royal government’s management of deficits before and during the Castilian war, see Christian Guilleré, “Les finances de la Couronne d’Aragon au début du XIVe siècle (1300–1310),” in Estudios, p. 493. Pere III, 2:554 (VI:45). ACA, Cancillería real, R. 1382, f. 104v. ACA, Cancillería real, R. 1380, ff. 18v, 55v–56v, 57v, 97. ACA, Cancillería real, R. 1379, f. 25v; R. 1380, ff. 18, 136–7v; R. 1381, ff. 60–1; R. 1383, f. 208v. ACA, Cancillería real, R. 1387, ff. 171, 195v. For the same practice in England and France of the same period, see Prestwich, Armies, pp. 259–62; Helen Nicholson, Medieval Warfare (New York, 2004), pp. 122–4. ACA, Cancillería real, R. 1543, f. 37.

War Financing in the Late Medieval Crown of Aragon

133

the king could only claim such grants every few years.85

VII With the almost constant collection of war taxes across the Crown of Aragon in the first years of the Castilian war, a methodology, drawn from the operating principles of the offices of treasurer and maestre racional, was adapted to the pressing matter of finding funds for Pere’s expansive war effort. In the case of a contribution of an individual or a single community to the war fund, Pere normally dispatched two or more of his court officials to levy the specified amount, properly docket the collected sum, and then transport the money back to the royal court where treasury officials or other extraordinary fiscal delegates would take charge of it.86 Since such operations were often conducted with town or village councils or clerical groups, these communal authorities were occasionally given some autonomy in laying out how the collection would be conducted and the installments in which taxes would be paid off.87 This simple model of court dominance of the process was firmly turned on its head when town councils, such as Perpignan, rejected limited fiscal autonomy and completely took over the raising of war funds. To meet royal quotas in 1359, the consuls of the royal village took it on themselves to sell off imposts and annuities for several years into the future. When the purchaser of these revenues moved to reimburse himself by collecting them, these traditional revenues applied to all property holders in Perpignan, including members of the royal family and great nobility. Even more shocking to the concept of royal order was the city council’s insistence that all aspects of this tax collection operation would be under the supervision of the municipal authorities. What is more, royal officials would have nothing to do with the process until collected monies were turned over to treasury officials.88 In reality, Perpignan rode a wave of fiscal autonomy that had already swept across the representative assemblies of the Crown of Aragon’s major states. As the years of war passed and the sources of money in eastern Spain shrank, Pere could do little else but to respect, at least partially, the claims of some clerical and urban institutions.89 By and large, however, he persistently attempted to maintain the system that emerged from the fiscal methods of his own government.

85 86 87

88 89

ACA, Cancillería real, R. 1380, ff. 29, 76r–v, 96r–v, 100v, 178–79; R. 1382, f. 104v; R. 1383, ff. 209v–210v. ACA, Cancillería real, R. 1379, ff. 27v–28; R. 1381, f. 31; R. 1382, ff. 141v–42. ACA, Cancillería real, R. 1380, ff. 46v–48, 112r–v; María Teresa Ainaga Andrés, “El fogaje aragonés de 1362: Aportación a la demografía de Zaragoza en el siglo XIV,” Aragón en la Edad Media 8 (1989), pp. 48–9 (doc. 1). ACA, Cancillería real, R. 1383, ff. 58v–60. ACA, Cancillería real, R. 1380, ff. 123r–v.

134

Donald J. Kagay

To collect imposts from large groups of ratepayers, the Crown utilized the same method that had proved effective with individuals: the appointment of commissioners with extensive power to collect tax at the level negotiated with the group authorities and then to transfer it to the royal treasury. These special agents of the king’s war effort had “full power … to seek, demand, and receive” the imposts and could designate “suitable persons” to do the same.90 While the tax rolls themselves were often woefully incomplete, the process by which taxes were gathered was a fairly efficient one and even provided an example for the fiscal activities of the Aragonese, Catalan, and Valencian parliaments for the next century and a half.91 The procedure used was as follows: the tax burden of each city, town, village, or hamlet, publicized in writing or by crier, was assessed on each “household” (foch). The resulting “household tax” (fogatge, fogaje) was an impost that held both families and communities responsible for fulfilling the tax levels assigned by the Crown.92 It was also surprisingly flexible and allowed the tax commissioners to renegotiate the household tax, even after the collection process had begun.93 When “money … [was] very necessary” for the war effort, collection of the fogatge could be turned over to the urban community or clerical institution in return for a small lump-sum payment in cash to be made immediately.94 Not all of Pere’s tax-gathering ventures always went according to the written orders his commissioners used as both credentials and authorization. Depending on the ebb and flow of campaigns, tax collectors might be commanded to postpone their operations or to completely cancel them.95 The curtailment of these operations was understandable when small hamlets and villages had been so destroyed by the constant Castilian raiding that their traditional populace had deserted them. Commissioners, however, must have found royal orders to break off collections with the arrangement of a truce in 1357 and a peace treaty in 1361 particularly troubling in light of Pedro’s penchant for violating these instruments.96 One of the most unpredictable aspects of the tax collection process faced by Pere’s commissioners was the level of general compliance or opposition to the extraordinary subsidies expressed by the taxed population. The king warned his agents against antagonizing the communities that fell under extraordinary tax

90 91

92 93 94 95 96

ACA, Cancillería real, R. 1380, ff. 40v–41v. J. M. Pons Guri, “Un fogatjament desconegut de l’any 1358,” Boletín de Real Academia de Buenas Letras de Barcelona 30 (1963–4): 322–498; Ainaga Andrés, “Fogaje aragonés,” pp. 33–59. ACA, Cancillería real, R. 1380, f. 65; R. 1381, ff. 167r–v. For the similar French impost, the fouage, see Henneman, Royal Taxation, 4–5, 211–15, 255–9, 281–3. ACA, Cancillería real, R. 1381, ff. 176r–v. ACA, Cancillería real, R. 1381, f. 152v. ACA, Cancillería real, R. 1381, f. 39v. ACA, Cancillería real, R. 1381, ff. 35–6, 128, 131. The truce was accepted in May 1357 and violated by Pedro in the spring of 1358; the peace of Terrer was concluded in May 1361 and violated by Castilian attacks in summer 1361.

War Financing in the Late Medieval Crown of Aragon

135

burdens. He specifically ordered them to avoid violating the privileges of clerical and noble lords who exercised suzerainty over many of the urban sites that were repeatedly called on to contribute to the war effort.97 Most of Pere’s Christian subjects were given some time to pay their tax bills; many were even allowed to leave a small monetary “surety” (securitas) with the commissioners as a guaranty that they would pay their share into the war chest at some time in the near future.98 Royal officials could confiscate such bonds if the persons did not contribute to the war subsidy within a grace period of a few weeks.99 After some months had passed, however, the names of those who had not completed their fiscal responsibilities had to be shared with the commissioners and with the entire community.100 From their credentials, Pere’s agents were allowed to take legal action against recalcitrant taxpayers and were well within their rights in “compelling and distraining the persons and goods” of those who had avoided their patriotic duty of contributing to the war against Pedro.101 The king even warned his subjects that failure to contribute to the subsidy was tantamount to treason since such actions evaded each citizen’s duty “to resist the iniquitous purpose of our enemy, the Castilian king” and, as a result, might “overturn the defense of the kingdom.”102 Though the tax collectors could use legal methods to extort promised tax payments, the ever-cautious Pere was nervous about such legitimate initiatives and often warned his taxmen: “refrain from acting until you have something ordered from us.”103 Though based on the procedures of the royal treasury, the collection of war subsidies took on a life of its own that occasionally seemed to be transferring the entire process into near-anarchy. Commissioners and treasury officials were clearly required to issue “receipts” (apochas, apocas) to all ratepayers and then to enter these transactions in a special “account” (compte).104 These records would prove crucial if the communities contributing to the subsidy decided to contest legally the actions of the Crown’s tax collectors.105 Royal personnel involved in the war financing effort were also to keep careful records of all their operating expenses and salaries.106 The collected funds were then transferred to the royal treasury by the royal “messenger” (porter) who swore homage and fealty “through mouth and hand” to faithfully carry out this crucial duty.107 At times, the king designated

97 98 99 100 101 102 103 104 105 106 107

ACA, Cancillería real, R. 1380, ff. 59v–60, 104r–v. ACA, Cancillería real, R. 1380, f. 84. ACA, Cancillería real, R. 1383, f. 235v. ACA, Cancillería real, R. 1382, f. 147. ACA, Cancillería real, R. 1382, ff. 49v, 146; R. 1383, f. 44. ACA, Cancillería real, R. 1382, ff. 153–54; R. 1383, ff. 216v–7. ACA, Cancillería real, R. 1384, f. 3v. ACA, Cancillería real, R. 1381, f. 192. ACA, Cancillería real, R. 1379, f. 95. ACA, Cancillería real, R. 1380, f. 84v; R. 1381, ff. 70r–v. ACA, Cancillería real, R. 1381, ff. 174r–v.

136

Donald J. Kagay

a special agent, usually a great “banker” or “money changer” (cambiador) to maintain “the great sums of money” generated from the subsidy and then make disbursements from this fund.108 In both cases, the trail of tax money into royal coffers and from there to the frontier battle zones was supposed to be meticulously recorded. Pay vouchers (albarani, albaras) and “letters of credit” (cartas de creença) directed officials to pay the salaries of military units once their commanders produced a proper “muster list” (mostra).109 The fiscal operation that took place within the precincts of Pere’s government because of the “great expenses and costs” that harried the Aragonese king for the entire decade of the Castilian war was both an adaptative and inefficient process.110 From the collection of the extraordinary subsidy to the transport and storage of these funds to their ultimate distribution among the troops in the field, Pere’s system of war-financing was a monetary pipeline that leaked at every administrative connecting point of its journey. The clerical and urban authorities as well as the king’s own fiscal servitors all had ample opportunity to engage in kickbacks, double billing, false accounting, and other forms of malversation that eventually put most of the government of Pere’s successor, Joan I (1387–96), on trial for malfeasance.111 Even military commanders had ample opportunity to enrich themselves from frontier service by maintaining the troop strength of their units below that specified in the muster list and happily pocketing salaries for troops that did not exist. They also fell into the habit of enriching themselves by holding back a portion of money intended for the salaries of the troops they did have.112 The passage of war finances through zones dominated by urban, clerical, and royal administrations necessarily made for an uncoordinated and unwieldy arrangement, not unlike the governance structure of many an American university. Several examples of this jealous and overlapping ineptitude will be sufficient. In 1349, long before the War of the Two Pedros broke out, Pere had sent a number of congratulatory embassies to the Castilian king, King Alfonso XI (1312–50), the conqueror of the important Muslim port of Algeciras in 1344.113 In the last of these in the late spring of 1349, the Aragonese sent a squadron of four ships commanded by his admiral, Ramón de Vilanova, and his principal adviser, Bernat de Cabrera, to aid Alfonso in defending the straits of Gibraltar against the

108 109 110 111 112 113

ACA, Cancillería real, R. 1380, ff. 118v–9. ACA, Cancillería real, R. 1381, ff. 123r–v, 181; R. 1383, ff. 221v–22v. ACA, Cancillería real, R. 1383, ff. 245r–v. Donald J. Kagay, “Poetry in the Dock: The Court Culture of Joan I on Trial (1396–1398),” in War, study XI, pp. 66–7; Hillgarth, “Royal Accounts,” p. 15. ACA, Cancillería real, R. 1381, ff. 173r–v; R. 1382, f. 149v; R. 1383, ff. 221r–v, 243. Nicholás Agrait, “The Experience of War in Fourteenth-Century Spain: Alfonso XI and the Capture of Algeciras (1342–1344),” in Crusaders, pp. 213–38; idem, “The Reconquest during the Reign of Alfonso XI (1312–1350),” in On the Social Origins of Medieval Institutions: Essays in Honor of Joseph F. O’Callaghan, ed. Donald J. Kagay and Theresa M. Vann (Leiden, 1998), pp. 157–62.

War Financing in the Late Medieval Crown of Aragon

137

threatened attack of the Muslim dynasty of Morocco, the Merinids.114 To cover the expenses of this small naval force, the admiral was given a set fee by the royal treasurer. As often happened with military and naval expeditions, real costs exceeded administrative estimates. As a result, Vilanova had to pay out of his own pocket the extra 6,000 sous which the final campaigning bill came to and then wait to be reimbursed from the Crown. Despite a number of attempts to recover this money, the admiral was unsuccessful until 1366, when, ironically, his royal master was at war with the son of the Castilian monarch he had set out to help so long before.115 In 1359, at the height of the Castilian war, Miguel Pérez de Gotor, referred to as the “shieldbearer, esquire” (scrutifer, escudero), a scion of an old Aragonese baronial family with holdings along the Murcian frontier,116 served on the Aragonese frontier with a company of horsemen led by his father, Eximen Pérez de Gotor. During this tour of duty, Miguel lost a “roan” (caballo de pelo runo) from his string of warhorses.117 The scribe of accounts set the worth of the animal, which may have been killed in action or stolen, at 800 Jacan sous and directed the representatives of the hamlets of Teruel to pay this amount since the animal died during the defense of these Aragonese frontier settlements.118 Carrying the pay voucher that documented his loss through his entire war service, Miguel spent much of the next year going through channels – all to no avail. After appealing to Aragon’s highest judicial authority, the Justicia Major, Juan López de Sesse, he “complained bitterly” to the king that he had not been reimbursed for his loss. Incensed, Pere demanded that the city fathers of Teruel’s hamlets accept his “remedy of justice” (remedio justicie) by paying Miguel the money owed him. Otherwise, Pere darkly promised them his “ire and indignation.” Pere issued this stern warning in June 1360 (over a year after the loss of the horse), but its ultimate effect is not known.119 What must have become painfully

114 115 116 117

118

119

ACA, Cancillería real, R. 1134, f. 167; Masiá de Ros, Relación 1:239; 2:364–7 (docs. 196/35; 197/37). ACA, Patrimonio real, Maestre Racional, R. 644, ff. 255–6; Hillgarth, “Royal Accounts,” 9, n. 33. Pedro Garcés de Cariñena, Nobilario de Aragón, ed. María Isabel Ubieto Artur (Zaragoza, 1983), pp. 103–5. For a list of common horse types from the period of the War of the Two Pedros; i.e. the “pack horse” (rocio) or “saddle horse” (palafrey), see Manuel Alvar, Textos hispanicos dialectales: Antología historica, 2 vols. (Madrid, 1960), 1:331 (doc. 160). As with Gotor’s mount, horses were also described by color and pattern: “chestnut” (castaynno), “dark brown” (moreno), “gray” (gris), “piebald” (cardeno),“sorrel-colored” (alazan), and “white” (blanco). They were also differentiated by body characteristics and types: “small” (chico), “with balanced feet” (balçan de los dos pies), “with face marking” (frontino), “with no markings” (sin senyal), “with Moorish head” (cabeça de moro), “with swelling rump” (lobado de tras). For such unitary actions of Aragonese hamlets, see Donald J. Kagay, “Two Towns Where There Was Once One: The Aldea in Medieval Aragon,” Journal of the Rocky Mountain Medieval and Renaissance Association 14 (1993): 40–1. ACA, Cancillería real, R. 1382, ff. 138v–39.

138

Donald J. Kagay

obvious to both Ramón de Vilanova and Miguel Pérez de Gotor, however, was the difficulty of getting of getting reimbursed from Pere’s administration, whether functioning in war or peace.

VIII As the Aragonese sovereign utilized his fiscal institutions old and new to squeeze every possible sou out of his realms, the representative assemblies of the Crown of Aragon and the estates of its various lands grew increasingly restive since they had been supporting Pere’s often unsuccessful military endeavors for years before the conflict with Pedro I.120 While Pere had managed to extract subsidies from clergy, nobles, and townsmen across eastern Spain in the first few months of the Castilian war, the pressing need for money forced the king to rely ever more heavily on the curia generalis [corts (Catalan); cortes (Aragonese, Castilian)], an institution he would grow to detest. In 1357, when “the Castilian king had inflicted great damage on the kingdom of Aragon in capturing some castles and other fortresses,”121 Pere frantically called together a corts at Lérida to seek advice and help in regard to the worsening military situation he commanded. Though hoping that the Catalans would supply funds to staff his embattled frontier posts in Aragon and Valencia, he was quickly disappointed.122 In August of the next year, the mounting expenses of a new campaigning season forced Pere to call out another Catalan assembly–this time at Barcelona. He again harbored the hope that he might convince the Catalans to grant a general aid. After months of pointless wrangling, the members of the corts and the frustrated sovereign came to an agreement that obliged the estates of Catalonia to raise a subsidy that would pay military salaries for the next two years. Pere appointed four special tax commissioners to collect this money “in the manner of a fogatge” on all “castles, square towers and closed fortified places” held by Catalan clergy and laymen.123 Despite the parliamentary pledge of monetary support for the king in 1358, many Catalans felt that Pere had committed “a great prejudice and grievance” against their homeland by extracting large sums of money from them to support military ventures “in remote lands”; that is, across the Catalan border in Aragon and Valencia.124 120 121 122 123

124

Manuel Sánchez Martínez, “Corts, parliaments y fiscalidad en Cataluña: Las ‘profertes’ para las guerras mediterraneas (1350–1356),” XV CHCA, Comunicaciones, 253–72. José-Luis Martin, “Las cortes catalanas en la Guerra Castellano-Aragonesa (1356–1365),” VIII CHCA (La corona de Aragón en el siglo XIV), 2 vols (Valencia, 1970), 2:81 (n.7). José-Luis Martin, “Las cortes de Pedro el Ceremonioso,” in Pere el Cerimoniós i la seva època, ed. María Teresa Ferrer i Mallol (Barcelona, 1989), pp. 102–4. Ibid., 105; idem, “Les Corts catalanes del 1358,” Estudis d’história medieval 4 (1971): 71–5; Colección de los cortes de los antiguos reinos de Aragón y de Valencia y del principado de Cataluña [hereafter CAVC], ed. Fidel Fita y Colomé and Bienvenido Oliver y Estreller, 27 vols. (Madrid, 1896–1922), 1, pt. 2:701–2. Usatges, trans. Kagay, p. 72 (arts. 30, 32); Customs of Catalonia, pp. 36–7 (art. 37).

War Financing in the Late Medieval Crown of Aragon

139

In Pere’s realms where the Castilian threat was the greatest, assemblies took a much harder line concerning war financing. Following a course that many town councils would choose, the Aragonese and Valencian cortes demanded an ever increasing share of autonomy in the collection and disbursement of war funds. These meetings appointed executive committees (eventually known as the Diputació del General in Aragon and the Diputació de la Generalidad in Valencia) whose principal functions were to collect the war subsidies pledged in a full meeting of the cortes and then rapidly to pay off the frontier troops whose duty was to ward off Castilian attacks.125 A good example of how these standing committees operated during the War of the Two Pedros is provided from records emanating from the Aragonese assembly of Cariñena in 1357 that pledged a subsidy to pay the salaries of 700 horsemen for the next two years.126 During the period of the grant, the committee of “deputies, dividers, and distributers,” chosen from each Aragonese estate (churchmen, barons, knights, townsmen), assessed the taxes promised by their fellows and gathered the required money by utilizing a tax similar to the Catalan fogatge. The parliamentary agents appointed in 1357 followed the contemporary fiscal methodology of the Crown by issuing receipts for monies expended, by keeping lists of those whose goods had to be distrained to force their contributions to the war fund, and by issuing vouchers for payment of military salaries. Out of a desperate desire to have his soldiers paid for the next two years, Pere allowed these executive committees a great deal of leeway by warning his own officials not to interfere in the process, but instead to give “counsel, favor, and aid” to parliamentary agents whenever required.127 At the end of this two-year period, Cariñena stood as dangerous precedent of parliamentary autonomy that would soon engulf the entirety of the Crown of Aragon. With the failure of the treaty of Terrer of 1361 to establish peace between Castile and the Crown of Aragon,128 Pere’s financial pessimism deepened to such a point that he could only see one way to rapidly retool for war–the gathering of representatives from all his realms in one place. The site of this “parliament” (parliamentum) was Monzón, a town perched on the AragoneseCatalan border which would become the customary center for such general

125

126

127 128

J. A. Sesma Muñoz and J. A. Armillas, La Diputación de Aragón: El gobierno aragonés del Reino a la Comunidad Autónoma (Zaragoza, 1991), pp. 16–19; Josè María Font y Rius, “The Institutions of the Crown of Aragon in the First Half of the Fifteenth Century,” in Spain in the Fifteenth Century 1369–1516: Essays and Extracts by Historians of Spain, ed. Roger Highfield, ed. Frances M. López-Morillas (New York, 1972), pp. 175–7; Bisson, Medieval Crown, p. 118. J. A. Sesma Muñoz and Esteban Sarasa Sánchez, Cortes del reino de Aragón, 1357–1451. Extractos y fragmentos de procesos desaperacidos (Valencia, 1976), pp. 17–34; Esteban Sarasa Sánchez, Las cortes de Aragón en la Edad Media (Zaragoza, 1979), pp. 44–5. ACA, Cancillería real, R. 1381, ff, 157v–68, 203v–4. For negotiations leading to and content of the treaty, see ACA, Cancillería real, R. 1163, f. 83v; R. 1176, ff. 11, 90; Zurita, Anales, 4:412–18 (IX:xxxiii); Masiá de Ros, Relación, 1:280–5.

140

Donald J. Kagay

gatherings. 129 Convening the assembly in November 1362, Pere begged the Aragonese, Valencian, Catalan, and Majorcan representatives to aid him “with subsidies, provisions, and other preparations for the defense of the republic of all our kingdoms and lands.”130 The king then spent the next six months in an attempt to force his badly divided subjects to act in unison against a real and present danger that could topple the “commonwealth” (cosa publica) that bound them as one.131 At long last, in March 1363, the Castilian danger coupled with their sovereign’s insistence finally led the members of the Monzón assembly to give in and donate to the war effort. The Crown’s fiscal victory, however, was clearly a limited one. The churchmen, nobles, and townsmen who stood before Pere in the cathedral of Santa Maria in Monzón did agree to pay the huge sums demanded of them, but did so on their own and not on the king’s terms. Each of the realm’s estates conferred separately and assigned the requisite share of the subsidy to its fellows on the basis of rank and wealth.132 Executive committees were chosen in each realm to handle all phases of the war effort from tax collection to the storage of these funds to the payment and deployment of troops along the Crown of Aragon’s many threatened frontiers.133 Though Pere was theoretically able to divert the Monzón subsidy in any way he thought best for “the use of the war,” the weakened and outmaneuvered ruler had no option but to follow the lead of the “shadow government” that the Monzón assembly and its executive committees briefly exercised.134 The Aragonese ruler, canny realist that he was, had no intention of giving up even a small portion of his sovereignty and so set out to use his own officials to “shadow” the work of Monzón deputies. While Pere could not formally reject a parliamentary tax-collection scheme he had sworn to uphold, he had no intention of folding up his own fiscal tents. Even after the parliamentary deputies of Monzón had begun their work, he continued to appoint tax commissioners and charge them with the responsibility of collecting special war imposts and receiving subsidy installments from the parliamentary agents.135 Though careful to instruct his men to do “nothing which is [not] pleasing to the Generalitat,”136 Pere repeatedly put them in situations that did just that. In one instance, he might overturn the tax collection activities of

129

130 131

132 133 134 135 136

Valdeavellano, Curso, 383; Luis González Antón, Las Cortes de Aragón (Zaragoza, 1978), pp. 68–9. For configuration of Monzón, see Spanish Cities of the Golden Age: The Views of Anton van den Wyngaerde, ed. Richard L. Kagan (Berkeley, 1989), pp. 150–4. CDACA, 48:13; Kagay, “Monzón” in 126–7. For Pere’s appeals to the parliament for unified action, see CDACA, 48:54, 63; Parlaments a les corts catalanes, ed. Ricard Albert and Joan Gassiot (Barcelona, 1928), pp. 10, 25; Suzanne F. Cawsey, Kingship and Propaganda: Royal Eloquence and the Crown of Aragon, c.1200– 1450 (Oxford, 2002), p. 137; Kagay, “Monzón,” pp. 128–30. Kagay, “Monzón,” pp. 135–6. Ibid., 136–7. Ibid., 133–4; CDACA, 48:108. ACA, Cancillería real, R. 1382, ff. 146r–v; R.1383, ff. 209v–10v. ACA, Cancillería real, R. 1383, f. 208v.

War Financing in the Late Medieval Crown of Aragon

141

royal officials as prejudicial to the fiscal operations of the parliament’s executive committees,137 but, before the ink was dry on the first dispatch, would order the same functionaries to confiscate sureties which pledged persons or groups to pay the parliamentary subsidy in the future and to distrain the goods and persons of all those who had not yet paid into the war subsidy. All of these provocative actions enraged the parliamentary leaders who saw them as blatant attacks on their authority.138 Even the point of contact between the parliamentary and royal tax collection efforts that both sides accepted–the final royal audit of subsidy revenues by the maestre racional–could prove inflammatory.139 The audit, after all, might reveal malfeasance or incompetence on the part of the deputies, ultimately throwing into question the accounts of the entire parliamentary subsidy collection.140 Direct royal interference in the parliament’s collection process could also cast doubt on these same financial records. On more than one occasion, Pere, moved by the “great poverty” of many of his over-taxed villagers, declared them exempt for a time from the general subsidy–a move that increased the complexities of an already complex procedure.141 In the end, the many-sided relationship of the parliament’s executive committees to the royal administration reflects the open and covert actions of a sovereign who desperately needed his national assemblies, but was thoroughly galled by this dependence.

IX Though the violence unleashed on the Crown of Aragon during the War of the Two Pedros was interspersed with periods of truce or military inactivity, the mere threat of “heavy blows from the king of Castile”142 continually weighed down the Aragonese sovereign with one “urgent crisis and emergency”143 after another, imposing on him at the same time “an intolerable lack of funds.”144 This interminable financial pressure forced Pere to ignore certain aspects of the “commonwealth” and “republic” he repeatedly pledged to defend in exchange for the ready cash necessary to pay troops to secure his frontiers. The irrationality of this position placed Pere in a bind that surely contributed to the many “sleepless nights” he complained of in the Monzón assembly.

137 138 139

140 141 142 143 144

ACA, Cancillería real, R. 1382, ff. 90r–v. ACA, Cancillería real, R. 1381, f. 241; R. 1383, f. 221. CDACA, 48:97, 140 (arts. 34, 40); Kagay, “Monzón,” p. 140. The maestre racional would issue a final audit of the two-year subsidy collection after six years. If the parliamentary agents were proved guilty of any “crimes and excesses” by this instrument, the Crown, and not the cortes, would punish them ACA, Cancillería real, R. 1381, ff. 243r–v; R. 1383, f. 209. ACA, Cancillería real, R. 1381, ff. 220–1; R. 1383, f. 245v. ACA, Cartas Reales [Pedro IV], caja 52, no. 6241. CAVC, 2:8 ACA, Cancillería real, R. 1327, f. 201; DS, 424 (doc.560).

142

Donald J. Kagay

The “continuous zeal and anxiety”145 of fighting his “principal adversary” on one hand while squeezing money from his own realms on the other occasionally led Pere to lash out in a violent or even quixotic fashion. In 1363, after months of bargaining with the tight-fisted deputies of the Monzón assembly, the frustrated monarch challenged his startled subjects “on horseback, on foot, or only with the shirts on their backs” to follow him against the Castilian enemy.146 Seven years before the impasse that caused the king’s outburst, Pere, an unlikely chivalric warrior, shocked his own people and surely amused his Castilian rival by challenging Pedro to settle their differences on “the field of honor.”147 No matter how real the emotions underpinning such events might have seemed to Pere or his subjects, the king was essentially driven not to vindicate his honor, but to gain victory over his arch-enemy and this would be accomplished not by courage, but through patience and guile.148 This meant that the king would spend whatever it took to defeat Pedro, even if this meant all his “Christian, Muslim, and Jewish [subjects]” would be driven into bankruptcy.149 This win-at-all-cost attitude would eventually undermine the psychological and demographic structures that had for so long supported Pere’s war effort. The concomitant stress that a decade of war had on the Aragonese king was reflected in similar fashion among the rate-paying public of the Crown of Aragon. After three general parliamentary subsidies, at least twice as many impost commissions of the Crown, and an almost continual royal extortion of cash and matériel from individuals and groups in all quarters of eastern Spain, a palpable “tax weariness” gripped Pere’s subjects and forced the king to adapt to it. The first sign of this exhaustion was the purposeful delay or total avoidance of paying war taxes at every level of the Crown of Aragon’s several societies. Because all of the estates had sworn to contribute to general aids in specified increments and their members had put up bonds to guarantee their future compliance, failure to pay into the war fund was tantamount to both perjury and treason–at least in the mind of the Aragonese king.150 Though Pere and his tax collectors had the right to distrain those whose duty it was to support the war financially, the very number of these slackers among the clergy, nobility, and townsmen, whose tax-evasion strategies (including litigation, jurisdictional casuistry, and simple non-compliance) caused untold

145 146 147

148 149 150

CDACA, 48:51. Ibid., 64. ACA, Cancillería real, R. 1380, f. 12; Pere III, 2:512–13 (VI:11), 547 (VI:40); Donald J. Kagay, “The Theory and Practice of Just War in the Late-Medieval Crown of Aragon,” Catholic Historical Review 91 (2005): 596–7. For assessment of Pere’s relationship to war, see David A. Cohen, “Secular Pragmatism and Thinking about War in Some Court Writings of Pere III el Cerimoniós,” in Crusades, 50–1. ACA, Cancillería real, R. 1379, f. 115v. Donald J. Kagay, “Defending the Western and Southern Frontiers in the War of the Two Pedros: An Experiment in Nation-Building,” Journal of the Georgia Association of Historians 23 (2002), p. 89.

War Financing in the Late Medieval Crown of Aragon

143

“damage to the defense of the commonwealth,” forced new solutions on the Crown.151 A past master of both the carrot and the stick, Pere turned first to threats and then to fines for those who set out to avoid purposefully paying their tax bill. For the worst of these offenders, he promised “large punishments” and instructed his officials to document their delinquency and draw up a list of these shirkers of national duty.152 He occasionally used this information to turn up the pressure even further. In 1360, he turned over the list of lesser Aragonese nobles who had not paid their taxes to the Justicia Mayor of Aragon, the highest legal authority in Aragon, instructing him to investigate the delinquent aristocrats and, if he could prove incontrovertibly their guilt, to strip them of their noble status.153 Despite the firm line that Pere communicated to local communities through his officials, the sovereign realized that the prosecution of his tax paying would ultimately prove detrimental to the general effectiveness of war financing. Behind the royal bluster, then, stood the king’s ultimate intention to compromise. He allowed communities to set their own schedules for the collection and delivery of imposts and did not himself neglect the postponement or cancellation of taxes if the situation warranted it.154 To put troops on his endangered frontiers, Pere was not above making deals with individuals willing to exchange military service for taxes owed.155 He made this practice even more attractive by assuring these substitute soldiers that all their legal and fiscal obligations would be held in abeyance during their entire tour of duty.156 In a set of lands as litigious as the Crown of Aragon, it was inevitable that Pere’s extraordinary efforts at war financing would spark one court challenge after another. In 1360, after two full years of non-support of the Cariñena subsidy, Count Cecilie of Urgel and her nobles ended a long legal battle with the Crown. Cecilie’s case was based on the contention that payment into an extraordinary subsidy was prejudicial to the “fueros, privileges, usages, and liberties” of the Countess and her subjects. After weeks of behind-the-scenes negotiations, a deal was brokered that allowed the countess and the residents of her lands to pay a reduced lump-sum payment into the aid and absolved them from full military service.157 In addition to these “zero-sum” suits which sought to question a litigant’s very obligation to pay war taxes, the disputes of Catalan taxpayers centered on how their hard-earned money would be used and, in particular, who would profit

151 152 153 154 155 156 157

ACA, Cancillería real, R. 1379, ff. 120, 121v–22,142; R. 1381, ff. 124r–v; R. 1383, f. 221; Ainaga Andrés, “Fogaje,” 49–52. ACA, Cancillería real, R. 1381, ff. 163, 231r–v; R. 1382, f.f. 91v, 156v–57; R. 1383, f. 216. ACA, Cancillería real, ff. 145r–v. ACA, Cancillería real, R. 1379, ff. 65r–v; R. 1380, f. 62v; R. 1381, ff. 128, 239v–40. ACA, Cancillería real, R. 1382, ff. 133r–v. ACA, Cancillería real, R. 1381, f. 165. ACA, Cancillería real, R. 1383, ff. 199v–201.

144

Donald J. Kagay

from it. In the first months of the war, the ecclesiastical estate of Catalonia filed suit against the government’s collection of war taxes, saying that the process violated national law that mandated the use of Catalan tax money only for the defense of the Principate. The royal judges reinforced the king’s right to use Catalan tax receipts even outside the borders of Catalonia by pointing to the precedent of the expeditions against Sardinia (1323–24, 1354–55) that were broadly supported by Catalan tax money. The obvious lesson for the Catalans was that they could not selectively choose what royal projects their tax receipts were used for.158 A great many lawsuits during the Castilian war involved litigants who did not oppose the subsidy itself, but instead attempted to reduce their tax bill by denying lordship and fiscal liability for territory that the Crown was attempting to tax. Ironically, the basis for most of the suits was contested boundaries–the very casus belli for the Castilian conflict itself. When an Aragonese subsidy was approved at Cariñena in 1358, ecclesiastical and lay lords with holdings straddling the Catalan–Aragonese border made the case that they had no obligation to contribute into the Aragonese war chest since their lands were Catalan. Playing no favorites, they presumably claimed that their lands were Aragonese when pressed by Catalan tax gatherers In 1359, Pere’s cousin, Count Alfonso of Ribagorza and Denia, refused to pay into the Aragonese subsidy for the entire Pyrenean region of Montanya which he claimed had settled under “the custom of Catalonia” and thus had to be considered Catalan territory.159 In the next year, the royal judiciary was called on to render a settlement for a much more complicated case between the ecclesiastical estate of Aragon led by the archbishop of Zaragoza on one side and the principal official of the Hospital in the Crown of Aragon, the castellan of Amposta, on the other. The castellan, whose order had tax responsibilities in both Aragon and Catalonia, attempted to lessen, at least temporarily, his organization’s final obligation to the 1358 subsidy by determining if four villagers were Aragonese or Catalan.160 In 1361, the king’s uncle, Ramon Berenguer, count of Amposta, made the same claim of no liability for the Cariñena subsidy in regard to the border village of Oya which, he asserted, had long been considered Catalan.161 In none of these cases was Pere’s judiciary in the mood to engage in cartography; instead, his judges acted as arbitrators and hammered out compromises that required

158

159 160 161

ACA, Cancillería real, R. 1380, ff. 44r–v. For the Sardinian expedition, see David Abulafia, The Western Mediterranean Kingdoms 1200–1500: The Struggle for Dominion (London, 1997), pp. 123–7, 179–80; idem, A Mediterranean Emporium: The Catalan Kingdom of Majorca (Cambridge, 1994), pp. 249–52; Antonio Arribas Palau, La conquista de Cerdeña per Jaime II de Aragón (Barcelona, 1952); Vincente Salvert y Roca, “La isla de Cerdeña y la política internacional de Jaime II de Aragón,” Hispania 10 (1950): 211–65. ACA, Cancilllería real, R. 1381, f. 242v. ACA, Cancillería real, R. 1383, f. 242v. The four border villages litigated over were Orta, Gandesa, Miravet, and Alquezar (?). ACA, Cancillería real, R. 1383, ff. 197r–v.

War Financing in the Late Medieval Crown of Aragon

145

the litigants to pay into the 1358 subsidy, if only at a reduced rate. Even more serious for the financing of Pere’s war effort than outright opposition or legal foot-dragging was the demographic and social breakdown that ten years of war brought in its wake. Great swaths of Aragon and Valencia were “at the point of depopulation and loss”;162 even the Castilian borderland was in jeopardy of becoming a “wasteland … caused by the great slaughter.”163 Many hamlets were ruined, their dwellings demolished and open to the “frigid weather.”164 A great number of villages “which had been settled from olden times were so destroyed that hardly any vestiges of [them] remained.”165 While much of the destruction could rightly be attributed to his Castilian enemies, the king could not ignore the fact that his own troops were also guilty of such mayhem.166 No matter who was responsible for the ruin of Aragonese and Valencian frontiers, most of eastern Spain, like contemporary France, had grown used to being held hostage by war.167 The effect of the long border war with Castile on the social order of the Crown of Aragon was equally devastating in both the threatened borderlands and in the large cities of the region. In the first months of the war, Pere received a disturbing report concerning “sons of the devil jealous of peace and quiet … unleash[ing] banditry” on his Catalan capital. He instructed the Procurator of Catalonia, Ramon de Montcada, to “block the evil attempts … of the iniquitous population” in any way his “true power” saw fit.168 At the same time, long-dormant feuds between important Valencian families, encouraged by the violent instability of the period, burst again into existence.169 Across the Crown of Aragon, clerics found themselves at the mercy of royal officials who took the unprecedented step of imprisoning “priests, tonsured men, and those bound by ecclesiastical judicial authority,” often for the non-payment of the latest round of war taxes.170 Churchmen were also at the mercy of increasingly violent parishioners; not even bishops were safe from being “daily … abused, oppressed, and damaged by multiple lawsuits, deputies, dangers, and discord.”171

162 163 164 165 166

167 168 169 170 171

ACA, Cancillería real, R. 1383, ff. 194v–95. Cortes de los antiguos reinos de León y de Castilla, ed. Real Academia de Historia, 5 vols. (Madrid, 1882–1903), 2:149 ACA, Cancillería real, R. 1381, ff. 33r–v. María Desamparados Cabanes Pecourt, “Cartas de población en le dominio verolense,” Aragón en la Edad Media 6 (1984): 102. ACA, Cancillería real, R. 1379, ff. 117, 123r–v. Since one of Pere’s largest contingents was led by the Castilian royal contender, Enrique de Trastámara, Aragonese settlements repeatedly had to be reminded that Pedro’s half-brother was an Aragonese ally and, as such, would be treated with respect. Marc Bloch, French Royal History: An Essay on its Basic Characteristics, trans. Janet Sondheimer (Berkeley, 1966), p. 113. ACA, Cancillería real, R. 1380, ff. 49r–v. ACA, Cancillería real, R. 1184, f. 1; DHC, 19–20. ACA, Cancillería real, R. 722, f. 34; DS, 463–4 (doc. 612). DS, 466–67 (doc. 616).

146

Donald J. Kagay

The Crown of Aragon had clearly seen first-hand what Eustache Deschamps had described in France: “War is nothing but damnation … The walls are falling. Things go ill and it’s dangerous.”172 Pere could hardly avoid the “sterility of the times” and the “indescribable poverty” he saw with his own eyes.173 As year after year of war passed, he could not escape the fact that much of the instability and psychological distress suffered by his people was not brought on solely by an odious Castilian enemy, but was also due, at least partially, to the many “different sums of money … [his people] always … [had] to give and pay.” In one letter, he assumed the shame of allowing “the places of our frontiers … [to become] depopulated.”174 Despite this regret (whether real or feigned), the war had to be paid for and this meant that taxes continued to be levied and collected with deadening regularity. No group bore a heavier fiscal burden for the continuance of the war than the infidel populations of eastern Spain. Subject to all the same general subsidies as their Christian counterparts, Jewish and Muslim aljamas also had to provide money and supplies whenever Pere’s military effort needed such support.175 Other royal directives attempted to criminalize the trading between regions of grain and other foodstuffs earmarked for troops, thus closing off a profitable black market that might have buoyed the sagging economies of Aragonese and Valencian infidel populations.176 As it was, the combination of Castilian raiding which left frontier settlements (many of them with large aljamas) “very often devastated” and the steady exactions of the Aragonese government left many infidel communities as “paupers who … [could] not bear customary tax burdens.”177 The extent to which Pere’s “royal treasure” had been systematically emptied by the ravenous demands of the Castilain war became more obvious as each year passed and the Crown had no choice but to grant one “tax extension” (elongamentum) after another to financially drained aljamas.178 Despite the obvious danger signs that Pere formally recognized, the king, like his Christian, Muslim, and Jewish subjects, was enthralled by the war and the financial machinery devised to serve it.

172 173 174 175

176 177 178

H. J. Hewitt, The Organization of War under Edward III, 1338–62 (Manchester, 1966), p. 134. ACA, Cancillería real, R. 1381, ff. 28v–30. ACA, Cancillería real, R. 1381, f. 245v. ACA, Cancillería real, R. 1379, ff. 144r–v; R. 1380, f. 129v; R. 1381, ff. 66v–67; R. 1384, f. 2. For the general economic situation of the Jewish and Muslim communities during the period of the Castilian war, see Baer, History, 2:28–34; John Boswell, The Royal Treasure: Muslim Communities under the Crown of Aragon in the Fourteenth Century (New Haven, Conn., 1977), pp. 196–226; Jonathan Ray, The Sephardic Frontier: The Reconquista and the Jewish Community in Medieval Iberia (Ithaca, NY, 2006), pp. 96–7. ACA, Cancillería real, R. 1382, ff. 131r–v. ACA, Cancillería real, R. 1381, f. 62; Cartas reales [Pedro IV], nos. 6100, 6107. Boswell, Royal Treasure, pp. 227–8.

War Financing in the Late Medieval Crown of Aragon

147

X In 1366, the bitter struggle between the rulers and realms of Castile and the Crown of Aragon had ended in a whimper in the midst of a Castilian civil war which, by 1369, cost Pedro his life and given his crown to his arch-enemy and half-brother, Enrique de Trastámara.179 The War of the Two Pedros had set the stage for these dramatic events, however, and presaged the course of Iberian events for decades to come. The connection of Pere’s role as war financier to these momentous events is a direct one. Disdaining his enemy’s policy of coinage debasement to make up the shortfalls brought on by the long war, Pere, like contemporary English sovereigns, did not dare to stray from the road of strong money.180 With the avenue of devaluation closed to him, the Aragonese king had to fall back on the time-worn methods of customary taxation and feudal military service. As one campaigning season melded into another, however, Pere realized that traditional means of supplying and paying for war would suffice no longer. The maintenance of standing armies, however, was not really the problem. Jaume I, Pere II, Jaume II, and even Pere himself had put large armies in the field for fairly long periods. These campaigns, however, were wars of conquest and were not fought to defend the frontiers of the Crown of Aragon. The standing war that engulfed much of eastern Spain was a different thing and required novel solutions that attempted to group into a military union peoples with different languages who considered each other “foreigners.”181 To survive, the Aragonese king had increasingly to rely on his parliaments and grudgingly assent to their increased power and autonomy. Whether or not Pere’s innovations stand as one of the headwaters of what Thompson calls the “fiscal-state” that grew up in Spain from the time of the Catholic Kings,182 Pere’s actions in the Castilian war does not signify–as Bisson suggests–a complete “loss of initiative in policy and finance.”183 Instead, it points to a sovereign “of a poor state in a rich country”184 who quickly realized that the battlefield that faced him could

179

180 181 182 183 184

Clara Estow, Pedro the Cruel of Castile, 1350–1369 (Leiden, 1995), pp. 232–69; L. J. Andrew Villalon, “Pedro the Cruel: Portrait of a Royal Failure,” in Medieval Iberia: Essays on the History and Literature of Medieval Spain, ed. Donald J. Kagay and Joseph T. Snow (New York, 1997), pp. 208–11; P. E. Russell, The Spanish Intervention in Spain and Portugal in the Time of Edward III and Richard II (Oxford, 1955), pp. 108–48; Kenneth Fowler, The Great Companies, vol. 1 of Medieval Mercenaries, 1 vol. to date (Oxford, 2001), pp. 163–222. Spufford, Money, pp. 314–15; Octavio Gil Farres, Historia de la moneda española (Madrid, 1959), pp. 210–13. Sesma Muñoz, “Fijación,” pp. 155–61. Thompson, “‘Money,” pp. 288–9. Bisson, Medieval Crown, p. 118. C.-E. Dufourcq and J. Gautier-Dalché, Histoire économique et sociale de la Espagne chrétienne au Moyen Âge (Paris, 1976), p. 205; Hillgarth, “Royal Accounts,” p. 14.

148

Donald J. Kagay

not be understood through the military truisms of the past. Victory could not be won with courage and advanced strategy alone; it now could never be attained without money. This realization had its limits, however, for (to paraphrase R. I. Burns), money talks, but has a limited vocabulary.185 For Pere III in his epic struggle against Pedro I, the key words in this vocabulary were “determination, cunning, and survival.”

185

Burns, Medieval Colonialism, p. 344.

7 National Reconciliation in France at the end of the Hundred Years War* Christopher Allmand

Among the problems most feared by any ruler in the medieval world – and later – was division among his subjects. A divided society, as Christ had proclaimed, could not long survive. Conflict was a threat to peace which every prince hoped to avoid.1 It was for this reason that steps had long been taken, with mixed success, to ban and prevent private war in the lands over which French kings ruled; over centuries men came to accept that only the ruler might wage war with proper authority.2 And, as if in conscious search of ways of resolving disputes when these arose (as they inevitably did), the late Middle Ages witnessed the development of legal procedures leading to compromises and negotiated settlements between those who sought to resolve their differences through the law.3 The responsibilities of rulers to work actively for peace were causing them to take a firm lead and to show by example that society need not, indeed should not, live in conflict, but in harmony and justice. In the mid-fifteenth century the kingdom of France was emerging from a lengthy period of war against England whose kings had pretensions to the crown of France. Since the 1340s, in their attempts to translate claims into reality, the

*

I am grateful to the editors for their comments and suggestions regarding the content and presentation of this article. Its faults are mine alone.

1

Many of the royal letters cited in the pages which follow began with a “scene-setting” description of the rifts and divisions in French society which it was the king’s wish and duty to heal. The measures being taken by him were intended to bring unity and justice (“bonne concorde et justice”) to all under his rule. As late as 1439 Charles VII was obliged to emphasise that “no one can or ought to assemble men and take to arms, nor maintain men-at-arms … without our authority … and if anyone does the contrary, he becomes a traitor … and an enemy of the public weal.” (Cited by S. H. Cuttler, The Law of Treason and Treason Trials in Later Medieval France [Cambridge, UK, 1981], p. 198). “Above all, and this from necessity as much as from choice, the king preferred private arrangements that served his political interests to general methods.” (R. Boutruche, “The Devastation of Rural Areas during the Hundred Years War and the Agricultural Recovery of France,” in The Recovery of France in the Fifteenth Century, ed. P. S. Lewis [New York and London, 1972], p. 39).

2

3

150

Christopher Allmand

English had invaded the country several times. Most recently, under their kings, Henry V (1413–1422) and his son, Henry VI (1422–1461), they had conquered and occupied Normandy and much of northern France. The years of war and occupation had ended in 1450 when Charles VII (1422–1461) had finally regained control of the duchy through military force. Nonetheless, in the early fifteenth century, the serious failings of the monarchy had allowed, even encouraged, civil war between the followers of the dukes of Burgundy on the one hand and those who favored the duke of Orléans and the count of Armagnac on the other (hence the names “Burgundians” and “Armagnacs” given to those who supported the rival groups). At the very core of the troubles (in addition to strong personality conflicts) lay a struggle for control of the monarchy, weakened by the poor mental health of Charles VI (1380– 1422), as well as for the control of the country’s towns, above all, of Paris. The plight of the capital, so well reflected in the pages of the so-called Journal d’un Bourgeois de Paris,4 was at times desperate, as it changed hands between the parties, frequently suffered virtual blockade by surrounding garrisons, and witnessed the horrors of the massacre of Armagnac supporters by their Burgundian rivals in the spring of 1418. There can be no doubt about the rivalries and divisions created in Paris and elsewhere by the internecine civil conflicts of these years. It was these which the English to a certain extent came to control while they occupied the capital between 1420 and 1436, but to which they, in their turn, were also to contribute. The two conflicts, one among Frenchmen, the other between Frenchmen and Englishmen, including Gascons, shared an important characteristic. In each, property lying in territory controlled by one side but whose owners had fled to live in that of the other was confiscated as punishment and was often given as reward to supporters of the party in power. This practice had already been in evidence in Paris when the Armagnacs seized the capital in 1413; the roles would be reversed in 1418. In Normandy, the English followed suit. Lands and estates belonging to Normans and others who refused to accept English rule and fled their homes, only to be declared rebels, were confiscated by the English and granted either to Frenchmen ready to accept the English presence and live under their rule, or to Englishmen whose settlement in Normandy was encouraged by the authorities. It was such a policy that the English followed in Paris, too, where many confiscated houses were granted to Englishmen or to Frenchmen of “Burgundian” sympathies who shared the government and administration of the capital.5 4

5

Journal d’un Bourgeois de Paris 1405–1449, ed. A. Tuetey, Société de l’Histoire de Paris (Paris, 1881), or, more recently, Journal d’un Bourgeois de Paris de 1405 à 1449, ed. Colette Beaune (Paris, 1990). See also the English translation, A Parisian Journal, 1405–1449, trans. Janet Shirley (Oxford, 1968), to which reference will be made in what follows. For Normandy see C. T. Allmand, Lancastrian Normandy, 1415–1450. The History of a Medieval Occupation (Oxford, 1983), chaps. 3 and 4; Robert Massey, “The Land Settlement in Lancastrian Normandy,” in Property and Politics: Essays in Later Medieval English History, ed. Tony Pollard (Gloucester and New York, 1984), pp. 76–96; Massey, “Lancastrian Rouen:

National Reconciliation in France

151

What, then, had these conflicts done to France, and what problems had they created for the future? In the first half of the fifteenth century France was a country divided by loyalties and allegiances, as well as by the presence of an English “occupant.” Charles VII, whose authority extended only over the center and the south of the divided country, tried to reassert the power of the crown, and to insist on the obligation of all loyal Frenchmen to give him their allegiance as their true lord. Weakened by the civil conflict, relying upon the support of the Armagnac faction during the first years of his reign (a fact which underlined that he was not king of all Frenchmen), bitterly at odds with the Burgundians who had established an alliance with the English, the king sought to regain the areas over which he had lost control and to re-establish his authority as king recognised by all Frenchmen. The first years had been difficult, but hope had been raised in 1429 when Joan of Arc, having helped to defeat the English at the siege of Orléans, had led Charles to his coronation at Rheims. Crowned and anointed, Charles was now undoubtedly stronger, and in the ensuing years the military conflict began to turn slowly in his favor. At Arras, in 1435, he managed to be reconciled formally with Philip, duke of Burgundy, who, abandoning his alliance with England, acknowledged Charles as king of France. Some months later, a French army appeared outside Paris; within a short time the demoralised English had surrendered, and Paris became, once again, the capital of a legitimate France. It was at this stage that the problems associated with what would prove to be the long process of ending and healing the “times of division” (“temps des divisions”), and of winding down the war, began. In fact, the need to find solutions to a particular problem created by war and occupation had already been acted upon by Charles VII at the time of his coronation in 1429, when talks had taken place with Burgundian representatives regarding the return of land, under the terms of a general pardon, to those who had held them before hostilities had begun. However, a vital factor had been ignored or forgotten. If this simple plan were acted upon, would not those loyal to the king risk incurring financial loss if their estates were returned to them in their current, often run-down condition? Anxious to give momentum to his campaign, Charles issued what came to be known as the Edict of Compiègne, which decreed the return of confiscated lands in the condition in which they had been left in the past. All current holders of such lands, even those who had come to them quite legally in the intervening years, were to be ousted.6 This was more than

6

military service and property holding, 1419–49,” in England and Normandy in the Middle Ages, ed. David Bates and Anne Curry (London and Rio Grande, 1994), pp. 269–86; Philippe Cailleux, “La présence anglaise dans la capitale normande: quelques aspects des relations entre Anglais et Rouennais,” in La Normandie et l’Angleterre au Moyen Age, eds. Pierre Bouet and Véronique Gazeau (Caen, 2003), pp. 265–76. For Paris, see Guy Llewelyn Thompson, Paris and Its People under English Rule. The Anglo-Burgundian Regime, 1420–1436 (Oxford, 1991), chap. 5. A. Bossuat, “The Re-establishment of Peace in Society during the Reign of Charles VII,” in The Recovery of France, ed. Lewis [see above, n. 3], pp. 67–8.

152

Christopher Allmand

simply a proclamation that measures to re-establish peace would have to be taken, and that the crown was accepting responsibility for doing this. In 1429 Charles’s aim was the publication of an overtly political manifesto, at once a claim to sovereignty in France and a partisan move aimed at rewarding those who had remained faithful to him and to his cause. But its approach had been unsubtle, its proposed terms not fully considered. What would Burgundian beneficiaries of grants think of it all, for it was largely at their expense that the king’s followers stood to gain? Yet, imperfect as it certainly was, it signalled that consideration had been given to the future when war might end and France would once again be a united country. A start, at least, had been made. However, at the great international congress held at Arras in the summer of 1435, while Charles and Philip of Burgundy were outwardly reconciled, the difficulty of making lasting peace between them was underlined by one clause in the treaty which concerns us here. The general terms agreed between France and Burgundy, which were to the advantage of the latter, and would be the cause of future disputes between both Philip and Charles and their respective heirs, Charles “the Bold” (1467–1477) and Louis XI (1461–1483), included a clause which stated that while “each one of either side should return to his own,” all lands and ecclesiastical benefices were to be accepted in their current state and condition, not, as set out in the Edict of Compiègne, in the state in which they had been in former, often better, times. Furthermore, no compensation for losses could be demanded. Not being in a political or military position to impose their view which, in any event, years of economic loss resulting from constant war would have made it impossible to put into effect, the French had allowed this clause, in spite of its direct contravention of the conditions set out in the Edict of 1429, to be included. They could hardly do otherwise.7 It was in April 1436 that the first major attempt to achieve conciliation within a large community was attempted. The scene was Paris, the occasion the capture of the nation’s capital from the English by forces fighting for Charles VII, now aided by the Burgundians and their sympathisers living there. The king’s “modération,”8 referred to by a recent historian of Paris, may have been influenced by that factor. The vocabulary of official documents, later reflected in the writings of the chroniclers, specifically emphasized the king’s wish and intention not to wreak vengeance. All acts of treason were to be forgotten, and a general pardon (“abolicion généralle”) was published. The language emphasised the need to forget past events and to encourage all to be united as loyal subjects in the king’s “grace and good will” (“bonne grace et bienveillance”).9 Behind this lay the presumption that the Parisians had come under enemy rule in spite of themselves, rather than out of choice. They were not to be held responsible for the ambiguous, not to say treasonable position in which they had lived for the

7 8 9

E. Cosneau, Les grands traités de la guerre de Cent Ans (Paris, 1889), p.146 (cl.35). Jean Favier, Nouvelle histoire de Paris. Paris au XVe siècle 1380–1500 (Paris, 1974), p. 237. Histoire de Charles VII, roy de France, ed. D. Godefroy (Paris, 1661), p. 796.

National Reconciliation in France

153

past eighteen years, and were certainly not to be treated as “collaborateurs” in the modern sense of the word. The king, furthermore, undertook to be “generous” (“liberal”) towards those who had sustained losses in his service. The Parisian solution reflected Charles’s growing confidence that he was now on the road to victory.10 Military considerations influenced what would otherwise be regarded as political decisions. There was no purpose in provoking resistance or outright opposition which might hinder the momentum of the military tide now running in his favor. The king had to be ready to make concessions. Those who preferred to continue serving the English were allowed to leave. When the Valois Parlement was restored to Paris early in 1437, several of its members were chosen from appointees of the Anglo-Burgundian régime of the previous years, a good example of the “honest peace, accord, and reconciliation” which made it possible that this institution – and others – “should be restored, re-established, and allowed to function in our town of Paris and its neighborhood in the way it had done before the divisions and for a long time before.”11 For example, Jean de Longueil, appointed lieutenant civil of the Châtelet by the English in 1431, was still in possession of the same office nearly thirty years later.12 Others, too, continued to serve in the administration. Such were the signs of that “moderation” which inspired the king as he reasserted his authority over the capital. This pragmatic approach to the employment of personnel from both sides of the political divide was to be followed by guidelines given by the crown on the validity of legal decisions made in the Parlement of Paris during the years 1418 to 1436. Writing from Poitiers, where the rival Valois Parlement had sat for the past fourteen years, on 15 March 1436 (a date which anticipated the capture of Paris by some weeks, and which therefore strongly suggests a policy which was premeditated rather than spontaneous), Charles VII admitted that in the preceding years many had obeyed the English king “both in the exercise and administration of justice as well as in other ways, as if he had been their sovereign lord.” This was true of Burgundian supporters who were now “reconciled, returned and reunited with us and to our obedience.”13 Having been asked to declare valid those judgements made in the part of France outside his control since 1418, the king set out the reasons which encouraged him and his council to accede to the request. Scarcely in a position to act otherwise, Charles none the less made the best use of necessity. Wishing, he wrote, to live “in harmony and justice” yet without implying recognition of the jurisdiction and authority of the English,

10 11

12 13

Recueil général des anciennes lois françaises, eds. F.-A. Isambert et al., 29 vols. (Paris, 1822– 33), 8:832–4. “ … bonne paix, accord & reconciliation … soient remises, restablies, tenues & exercées en nostredicte ville de Paris, ès lieux, ainsi & par la maniere qui avant lesdictes divisions avoient acoustumé estre tenues & exercées d’ancienneté.” (Ordonnances des roys de France de la troisième race, 22 vols, [Paris, 1723–1849], 13: 229). [Hereafter Ordonnances] Favier, Paris au XVe siècle, p. 238. “ … reconciliez envers nous & reduiz & réunis à nous et à nostre obéissance.” (Ordonnances, 13: 216).

154

Christopher Allmand

and certainly without prejudice (real or implied) to himself or to any supporters, he agreed that judgements made “by those who called themselves judges”14 outside his effective jurisdiction between persons living in those parts should be regarded as valid, although this was not to apply to cases involving persons residing within his jurisdiction. From this decision it was logical to allow those responsible for these judgements to remain in their positions. The compiler of the register of the Parlement criminel marvelled at the ease with which the transition took place, in which few incidents occurred, and for which all should thank “the sweet Jesus … for the goodness and kindness which, in his mercy, he had shown today in this town” which had enabled the events to take place “without the shedding of blood, or any violence or incidents … or at least very few … which all creatures must regard as the work of divine not human action.”15 The continuity of the proceedings of the Parlement were an indication of the fact that all sides saw the provision of justice as a necessary function of state which could not be halted or delayed even in these most unusual times. The future was to bring further evidence of the continuation, one might say the development, of such a policy. At an assembly held with certain princes of the blood at Nevers in March 1442 the king was asked to appoint good counsellors, and to be reconciled to some of the highest nobility who, it was claimed, had sustained huge losses of land to the English while supporting their king against them. The terms used to present these petitions are instructive: “that he should appoint to his council notable God-fearing persons, who have not adopted extreme positions with regard to the divisions of the past,” to which the king replied that in choosing counsellors “he has had no regard to these divisions, which he has chosen to forget.”16 To the request that he be fully reconciled with several dukes, among them Orléans, Alençon, Vendôme, and Nevers, the reply given was that the king had been as “generous” as he could be to those lords of the blood and others who had suffered losses in his service. A reference later in the document to “peace and reunion” (“paix et reunion”) supports the view that, outwardly at least, it was official royal policy to encourage reconciliation between the parties in what still remained a disunited and war-torn country. It was this ideal which appears to have governed and directed the positive and seemingly consistent policy of Charles VII as he strove to reconcile his people to one another and towards the crown of which he was the living embodiment. In this matter the king gave a strong lead, presenting himself as a figure ready to

14 15 16

Ibid., 13: 217. Journal de Clément de Fauquembergue, greffier du Parlement de Paris, 1417–1435, ed. A. Tuetey, 3 vols, Société de l’Histoire de France (Paris, 1903–15), 3:194–5. “ … qu’il luy plaise eslire en son grant conseil gens notables, crémans Dieu, et non extremes ou passionés ès divisions passées … Le roy n’a eu regard aux divisions passées, et les a et tient pour oubliées.” (Chronique d’Enguerran de Monstrelet, ed. L. Douët-d’Arcq, 6 vols, Société de l’Histoire de France [Paris, 1857–62], 6:40). Variant texts in Recueil général, ed. Isambert, 9:109; and in Chronique de Mathieu d’Escouchy, ed. G. du Fresne de Beaucourt, 3 vols, Société de l’Histoire de France (Paris, 1863–64), 3:77.

National Reconciliation in France

155

forgive and to forget events of the past, preferring to look to the future. This last point is important, as the king and his counsellors had before them the daunting task of rebuilding France economically and socially, and giving it a future under the authority and leadership of the monarchy. There were limits to what could be done about events and decisions of the past generation or two, and there should be as few recriminations as possible. If blame for what had occurred had to be apportioned, then it could be laid firmly at the door of the English who, having brought the war to France, had become the cause of that country’s divisions. Frenchmen, if loyal to the Valois dynasty, could be praised for their fidelity; on the other hand, Frenchmen who had temporarily “gone English” should be pardoned for having acted under the coercion of military force. In the final analysis, no distinction should be made between them. France should try to achieve a new beginning under a leader who showed virtues which, if inspired by Christian teaching, were also strongly political in character. Christian influence upon the forming of political attitudes was best expressed, in terms of the needs of mid-fifteenth-century France, in the writings and exhortations of Jean Juvénal des Ursins, once bishop of Beauvais, later archbishop of Rheims and counsellor to the king. Tireless in his efforts to make Charles into a “most Christian king,” Jean Juvénal reiterated time and again that, for the sake of the common good, a king, while always striving for justice, should always be ready to exercise the royal attribute of mercy. “Sire,” he pleaded about 1440, “do not have regard to our faults committed in times past, nor to those of our relatives and ancestors, nor to our failure to obey your orders, nor to have recognized early enough the obedience which we owed you … Alas, sire, you must be ready to show us mercy.”17 The law should not always be applied to the letter; rather, the king should have some latitude in its application. Mercy should be shown to those who erred; this would bring credit to the ruler himself. Jean Juvénal was anxious that his king should rule with moderation. “Justice without mercy,” he wrote, “is simply harshness;” unlike others, a king is able to write “we remit and pardon” rather than simply “we have punished.”18 His image was of a caring monarchy, readier to pardon than to exercise its rights and exact punishment to the full. The terms granted by the king’s commanders to the towns of Normandy as these were recaptured from the English mainly in the course of the year 1449 constitute an interesting series of texts which enable us to observe royal policy in action as the war drew to an end. On 17 July, on the eve of the final “push” towards reconquest (“recouvrement”) the king authorised commissioners to grant

17

18

“N’ayez aucun regart a noz faultes du temps passé, ne a celles de noz parens et predecesseurs, car ce que n’avons pas obey a voz commandmens, et que nous n’avons pas assez tost congneu la vraye obeissance que nous vous devions … Helas, sire, il est de present temps que tu nous faces misericorde.” (Ecrits politiques de Jean Juvénal des Ursins, ed. P. S. Lewis, 3 vols, Société de l’Histoire de France [Paris, 1978–92], 1:373–4). “ … justice sans misericorde c’est severité … ’; … ‘remettons et pardonnons’ mais seulement ‘avons pugny.’” (Ibid., 2: 415).

156

Christopher Allmand

favorable terms to the citizens of towns returning into his obedience.19 The military campaign was going so well that the king’s commissioners felt obliged, as their predecessors had felt at the time of the capture of Paris in 1436, not to propose unacceptable conditions which, by provoking resistance, could delay the military advance. Time, or rather the lack of it, was on the side of the defenders who were thus able to gain favorable terms from the king who needed to make the maximum possible progress in the remaining summer months available to him before the campaign would have to be halted. Since it was already getting dangerously near to August, the last thing Charles wanted was to be compelled to spend time besieging towns and castles. The form and content of the letters actually granted suggest that, when negotiations for surrender were taking place, the citizens’ representatives expressed concern about two matters in particular: that they might expect to pay a heavy penalty for their allegiance to the English over the past thirty years, and that their rights as property owners might be open to question or challenge. The views of those seeking readmission into the French obedience gave the king the opportunity to emphasize that, even in the moment of success, the crown would not act in a triumphalist way nor exert the maximum penalty from those who had lived in the enemy obedience. On the contrary, the people of Neufchâtel were told that the king, “wishing to draw our vassals and subjects to us, and to treat them with love and kindness” so as to encourage their future loyalty, was willing to “put behind him, pardon, suppress all memory of, and put aside any wrongdoing” in order to receive the people of the town into his good grace.20 The texts of these letters are all different, but their contents are very similar in tone. At both a moral and a political level the king was trying to make the kingdom whole, to stabilise it, and to welcome back all who wished to live under his authority. Consequently, all who had turned their backs on the crown and had encouraged the enemy might now benefit from the king’s magnanimity; their treason (the term was not used, but the concept was clearly there) and all other crimes against whomsoever committed would now be “regarded as never having happened,”21 and they would receive the royal pardon. The granting of a general pardon (“abolicion généralle”) along with an undertaking not to permit royal prosecutors to pursue miscreants, and a promise to seek out and punish those who might try to avenge acts committed during the past years, constitutes further evidence of a forward-looking approach which must have been an important factor in making a conscious policy of conciliation succeed. What was lacking in all this was a clear sense of legal consistency. We have already observed that the terms agreed at Arras in 1435 relating to the condition

19 20

21

Ordonnances, 14:59–61. “ … voulans nos vassaulx & subgects recueillir & retraire à nous, & iceulx traicter en toute amour & débonnaireté … mectre en oubly & icelles [choses] leur abolyr, pardonner, et remectre.” (Ibid., 14: 65) “ … repputez comme non faicts et non avenus.” (Ibid., 14: 66, 72).

National Reconciliation in France

157

in which properties should be returned to their original owners contradicted those set out in the Edict of Compiègne of 1429. The terms granted to the towns of Normandy added further to the confusion. Probably to secure or, in places which had offered no resistance to the French king’s army, to reward a rapid surrender, the interests of those in possession at the time appear to have been recognised once more at the expense of those loyal to the king. Confirmed in possession of their rights and revenues, the “possessioners” were the beneficiaries of the policy of granting general amnesties even though the Edict of Compiègne would not have judged in their favor. To add to the confusion, the Edict was to be confirmed by the king at Montbazon on 28 October 1450. The text of the accompanying letter emphasised his desire to “ensure and maintain good peace and union among our subjects, so that they will no longer recall their conflicts, nor the evils and difficulties associated with time of war and division which our kingdom has experienced.”22 All judgements made in the courts, including the Parlement, contrary to the terms of the Edict were declared void; the law was to be observed, a decision which suggests that the courts knew or suspected that non-legal considerations, such as military necessity, had influenced earlier decisions. The text of the Edict, favoring those loyal to the king, was registered by the Parlement for the first time only on 15 February 1451; a royal letter hinted strongly that the court had hitherto been reluctant to take this step. In Rouen, the news of the confirmation caused confusion, as it appeared “to be greatly to the prejudice of several notable persons from this region of Normandy and beyond, and even several from this city, as it seemed to be in complete contravention of certain things granted by the said lord in the terms of surrender and pardon granted by the said lord to this city.”23 Attempts were to be made to secure letters from the king which would ensure that the terms of surrender of November 1449 were in no way affected by this recent development. On 8 April the delegation reported to the people of the city assembled “in large numbers.” Its members were thanked for their work; but the record does not tell us whether they achieved what they had set out to do. However, this was not the end of the story. The Parlement, reluctant to judge solely on legislation without being able to consider each case separately, and aware that the war had created complex problems which were beyond the

22

23

“ … tenir & garder bonne paix & union entre noz subgetz, sans ce qu’ilz aient cause de rémembrer les ungs contre les autres, les maulx & inconvéniens faiz & perpetrez durant les guerres & divisions qui ont esté en nostredit royaume.” (Ibid., 14:105). “ … grandement preiudicier pluseurs notables personnes de ce pays de Normendie et d’ailleurs, et mesme pluseurs de cette ville, et qu’ilz estoient directement contre aucunes choses accordées par ledit seigneur par les traictié, composicion [et] abolicion de ceste ville de Rouen puis naguere fait et donné par icellui seigneur.” (Rouen, Bibliothèque Municipale, MS.A. 7 [Délibérations de la ville, 1448–53], fol. 90v. See also fols. 91 and 114; C. T. Allmand, “Local Reaction to the French Reconquest of Normandy: The Case of Rouen,” in The Crown and Local Communities in England and France in the Fifteenth Century, eds. J. R. L. Highfield and R. Jeffs [Gloucester, 1981], pp. 146–61).

158

Christopher Allmand

capability of any single legislative text to resolve, preferred to consider its judgements in accordance with its age-old practices, which did not necessarily lead to what might appear as consistency of judgement. Moreover, in July 1454 it published and registered the terms of amnesty granted by the king to the citizens of Rouen in November 1449, declaring that it did so without prejudice to the declaration on the Edict of Compiègne made by the king at Montbazon on 28 October 1450. It is clear that the Parlement was intent upon securing maximum freedom for itself in the matter of resolving the complex legal implications of royal ordinances such as the Edict of Compiègne.24 At the local level, too, the Edict was not always the basis of arrangements made between those returning to reclaim their inheritances and those in possession whom they could challenge by virtue of its terms. The evidence of local legal records suggests that many compromise agreements, or accords, were reached between parties, often leaving in possession the persons occupying the property concerned, although they may have made some kind of financial arrangement in order to do so. Such compromises, rather than the rigid application of the law expressed in the Edict of Compiègne, were encouraged by a number of factors. One was the long absence in the Valois obedience of those upon whom rights had devolved, persons who may have settled elsewhere in France and had little inclination to return to a society from which they been absent for, in many cases, more than thirty years. Nor were they always in a position to provide deeds or other legal proof of the validity of their claims. Economic factors, too, discouraged supporters of the king from pursuing these claims: the costs of litigation could be high; regular references to the need for repairs suggest that the physical condition of the estates to which they laid claim was not always sound, and the profits to be made might not be high; there were often debts to be paid, and these, too, might be considerable. In brief, the encouragement to return given by the terms of the Edict of Compiègne, while a factor likely to lead to a greater degree of legal stability and so to a greater determination to exploit the land by those who saw themselves as its proper owners, did not always prove to be advantageous to those concerned. Cheaper than going to law, arbitration was a procedure often more likely to find practical solutions than decisions handed down by the courts. It was a device much used in these times as France sought to come to terms with the difficult legacy of the divisions created by war.25

24

25

Bossuat, “Re-establishment of peace,” pp. 78–80, 366, n. 45. On the part played by accords and by the judges of the Parlement in the judicial processes of this period, see Claude Gauvard, “Justification and Theory of the Death Penalty at the Parlement of Paris in the Late Middle Ages”, in War, Government and Power in Late Medieval France, ed. Christopher Allmand (Liverpool, 2000), pp. 190–208. On this see Bossuat, “Re-establishment of peace”; C. T. Allmand, “The Aftermath of War in Fifteenth-century France,” History, 61 (1976), pp. 344–57; Allmand, Lancastrian Normandy, chap. 11.

National Reconciliation in France

159

We have already observed that Charles VII’s relations with his great nobility constituted one of the major political problems faced by the king in the middle years of his reign. One whose case proved particularly difficult was Jean II, duke of Alençon, many of whose lands had been overrun by the English and who, taken prisoner at the battle of Verneuil in 1424, had been obliged to alienate others to pay the ransom needed to secure his freedom. Out of favor since 1440, he had unsuccessfully sought financial help from the king. Later, out of malice (so a royal clerk was to claim),26 he began to plot with the English with the intention of helping them return to Normandy, hoping to regain for himself the places he had lost. Such was the core of the accusations levelled against him when, after enquiries lasting some two years, he was finally brought to trial at Vendôme in September 1458. The occasion may best be contextualized by considering the king’s reaction to other treasonable acts perpetrated against him in these years. Generally speaking, Charles VII “demonstrated a firmness of purpose, tempered by leniency towards the higher nobility, in meeting the challenge to his authority posed by the princes” at this time; “clemency was a distinct characteristic of his rule.”27 Those who opposed him were largely reprieved and pardoned, in some cases even returned to favor. So Jean IV, count of Armagnac who, among other crimes against the royal authority, had flirted with the English, was nonetheless pardoned.28 Alençon’s crime, however, was more serious; in his case, many appear to have favored the death penalty. By contrast, what proved significant was the king’s lack of vindictiveness towards the accused. At his trial, the representative of the duke of Burgundy was allowed to plead for clemency;29 another plea was entered by Alençon’s father-in-law Charles, duke of Orléans who, taken at Agincourt in 1415, had spent the next twenty-five years in English captivity;30 while Jean Juvénal des Ursins, also present, composed his Exortation faicte au Roy as a plea for mercy to be shown to Alençon. The king, he claimed, had pardoned many others, individually or through general pardons; would he not do the same for a prince of the blood? He would gain greater credit from an act of mercy than from seeing the law take its course.31 Although Alençon was sentenced to die, Charles VII reduced the sentence to imprisonment during his pleasure. In a very public case, he had preferred mercy to the weight of the law. By the time that Alençon’s trial had taken place, further events had occurred by which we may judge the degree of reconciliation being achieved in France at the end of the Hundred Years War. The first was the conquest of Guyenne, following upon the recovery of Normandy in 1450. By the summer of 1451 superior military force had won Charles VII the city of Bordeaux and, with it, control

26 27 28 29 30 31

M. G. A. Vale, Charles VII (London, 1974), p. 157. Cuttler, Law of Treason, p. 196. Ibid., pp. 202–3. Ibid., pp. 103–4. Ibid., p. 105; P. Champion, Vie de Charles d’Orléans, 1394–1465 (Paris, 1911), pp. 542–8. Ecrits politiques, 2:421–2.

160

Christopher Allmand

of most of Guyenne. Historians are generally agreed that, since Guyenne had never taken the oath of allegiance to the king or to his predecessors, it could not be accused of treason and treated accordingly. As such, the terms granted to its urban populations were favorable: a general pardon32 for those who chose to enter the French king’s obedience and the confirmation of privileges were the most important and the most normal. The agreement made in June 1451 between the Three Estates of the Bordelais and the king’s representatives was likewise moderate in tone and content. “The terms imposed by Charles VII were generous and Gascon liberties and immunities were to be respected.” People were free to leave the king of France’s jurisdiction if they wished. However, if they chose to stay and took the oath of allegiance, they would benefit from a general pardon, fiscal concessions, the continuation of trade (vital for the economic prosperity of the region), access to justice,33 and a new currency to be minted in Bordeaux, the king undertaking to pay any profits from this operation into the local economy. Charles VII was doing his best to forget the past and to win over new subjects. However, the return of the English to Bordeaux in October 1452, which necessitated a second French attempt to regain the city, which was successfully completed in October 1453, was to test severely previous French policy of conciliation. Giving “support” (“attrait et confort”) to the enemy was rebellion, since it involved the breaking of oaths and undertakings given only months earlier. The tone of the pardon eventually granted to the city revealed the king’s anger. Admitting that only a few individuals had actively supported the rebellion, and accepting that the majority were afraid of being punished for something for which they had not been responsible, the king agreed to receive the people of Bordeaux into his grace and grant them a pardon for acts committed against him. Nevertheless, he withheld their earlier privileges, and exempted certain persons from the general pardon. Worse was to come. Heavy taxes were imposed on both buyers and sellers of wine as well as on other forms of trade on the river Garonne. There followed measures to restrict the wine trade on the river, and to limit the role

32

33

“ … mettre en oubly & tout pardonner et abolir.” (Ordonnances, 14: 176–7). Jean de Wavrin maintained that the places taken by agreement could equally well have capitulated to assault, had the king not had other ideas: “ … le roy de sa benignité voulloit que len les prensist par composition adfin de obvier à leffusion du sang humain et destruction de son pays et du peuple quy estoit enclos es dites villes et forteresses.” Note how, according to this witness, the king intended to spare the sufferings “de son pays et du peuple.” (Recueil des croniques et anchiennes istories de la Grant Bretaigne, à present nommé Engleterre, par Jehan de Waurin, seigneur de Forestel (1447–1471), eds. W. and E. L. C. P. Hardy, 5 vols, Rolls Series [London, 1891], 5:165) Malcolm Vale, “France at the End of the Hundred Years War (c.1420–1461),” in New Cambridge Medieval History VII, ed. Christopher Allmand (Cambridge, UK, 1998), 7:402. A “cour souveraine” was functioning by 1452, although thereafter it met only twice, in 1456 and 1459 (G. Hubrecht, “Jurisdictions and Competences in Guyenne after Its Recovery by France,” in The Recovery of France, ed. Lewis, pp. 88–9). A sovereign parlement would be established in Bordeaux in 1462.

National Reconciliation in France

161

played in it by the city of Bordeaux, while further curbs on the trade of other commodities were also imposed. With regard to its legal status, Bordeaux was maintained within the jurisdiction of the Parlement of Paris, to be visited every year or two by judges and others who, with local assistance, would judge appeals locally. A fine of 100,000 écus, imposed in 1451, was reduced to 30,000 écus only in response to a petition from the citizens of Bordeaux that the destruction caused by the English on their return to the city had made the collection of such a sum impossible.34 The contrast between the documents of June 1451 and of April 1454 could scarcely have been greater. Most significant were the measures clearly intended to reduce the value of the trade, in particular the wine trade, for which Bordeaux was the region’s centre and main beneficiary. This was a body blow, a punishment which, seemingly uncharacteristic, has tarnished a little the king’s reputation. What was important for Bordeaux (and Charles must have known it and intended the result of his punitive decision) was the decline in trade which followed. The de facto absence of English traders, which followed not unnaturally at the end of the war and which a king with good intentions would have taken into account, only exacerbated the commercial situation. This was allowed to continue until it was resolved, eleven years later, with the decision of Louis XI to permit the port to be opened up again to wider commercial activity.35 Both the city and the royal coffers benefited from the decision. Like Normandy, Guyenne had a problem with the Edict of Compiègne. Charles VII had entered the territory relatively “peacefully” in 1451; but after the shortlived return of the English to Bordeaux in the following year, the area had been reconquered by force, and an altogether tougher attitude taken towards its people. This gave some the opportunity of seeking to apply the terms of the Edict of Compiègne, attempts in which they were opposed by those who claimed that, issued a generation earlier (in 1429) for application in a France which did not, at that time, include Guyenne, it could not now be applied to that territory. Its significance for our purpose is twofold. First, legal records underline that the problems of the recovery of landed interests by individuals citing the Edict were not limited to the more northern parts of France, and that the future social peace was also threatened in Guyenne by conflicts arising out of the problems involving land.36 Secondly, the introduction into Guyenne, in the years following 1453, of French legislation and decrees (the Edict of Compiègne was one of these) was an unpopular measure, underlining the fact that Guyenne was now part of that much larger area where the law of France would be applied. While it did little to retard that eventual process, the introduction of “alien” laws into a territory proud of its

34 35 36

Ordonnances, 14:270–5. C. T. Allmand (ed), Society at War. The experience of England and France during the Hundred Years War (Edinburgh, 1973: new edn Woodbridge, 1998), pp. 182–4. M. G. A. Vale, English Gascony, 1399–1453 (Oxford, 1970), pp. 221–5; Allmand, “Aftermath of War,” pp. 253–6; R. Harris, Valois Guyenne (Woodbridge and Rochester, N.Y., 1994), pp. 101–2.

162

Christopher Allmand

ancient customs and legal procedures, just at the time when Normandy had been brought, once again, within the French kingdom, was not always conducive to harmony and good relations between “conquerors” and “conquered.”37 A final example of the practical problems faced by the crown may be seen in the way in which the role of Joan of Arc in securing victory in the years 1428– 1430 was treated a generation later. The apparent lack of any attempt to vindicate her in the period of almost twenty years after her execution at Rouen in May 1431 has long been noted by historians. Even when her case was raised in 1450 in order “to know the truth” about her trial and condemnation, it led nowhere. Although a brief enquiry was held in 1452, and was reopened in 1455 as an attempt to vindicate the honor of the Arc family, the process of rehabilitation, as it came to be known, closed with only the annulment of the sentence pronounced against her. Why, by contrast, was there no condemnation of the surviving judges and assessors? A number of reasons have been advanced: the delicate state at the time of France’s relations with the papacy; the reluctance to suggest that Charles VII’s victory against the English around 1450 had owed much to events involving Joan twenty years earlier. But the most convincing reason offered is the king’s unwillingness to raise the matter of “collaboration” by a large group of leading clergy, mainly alumni of the university of Paris, “our very dear and much loved first-born daughter,”38 whose privileges the king had confirmed in May 1436, into an issue involving men who, since that day, had mostly demonstrated their allegiance to the crown of France. In 1456 the process of rehabilitation pronounced on the verdict delivered by the court in Rouen in 1431. Significantly, because it could not afford to do so, it did not deal with the individuals concerned, particularly those who were still alive. That task would be left to history. In 1456, resurrecting certain aspects of the past would not help to bring about the restoration of harmony in society. Nevertheless there was irony in the fact that Thomas de Courcelles, chosen to preach the sermon at Charles VII’s funeral in August 1461, had been involved in the decision to submit to torture the young girl who, many thought, had given impetus to the movement which had restored the king to the full exercise of his royal authority.39 What is the broad context of these attempts, largely successful, to achieve reconciliation in mid-fifteenth-century France? First, they may be seen as an essential part of restoring peace and order in a much divided kingdom which had experienced both civil conflict and occupation of part of its territory by the English. If there was to be the peace which all longed for, the past must be cast into oblivion, crimes pardoned, purges and witch hunts forbidden, the signs and symbols of factions and divisions outlawed, and the continuity of justice maintained. The policy of granting general pardons was both sensible and realistic; reassuring to those entering – or re-entering – the royal obedience was the promise that 37 38 39

Harris, Valois Guyenne, chap. 5. “ … nostre très-chere & très-amée Fille premiere née, l’Université de l’Estude de Paris.” (Ordonnances, 13: 219). On this matter see Vale, Charles VII, chap. 3.

National Reconciliation in France

163

none would be allowed to rake up the ashes of the past. Further, the outlawing of the old party titles40 and party badges (the bende or distinguishing sash worn by the Armagnacs,41 the cross of St Andrew worn by the Burgundians) was a further step in the direction of banishing provocative and emotive political symbols from public life.42 The moderation of the language of official documents and, following them, that of the chroniclers, corresponded to the king’s conciliatory approach. Persons recognising his sovereignty must be received “peacefully,” and none was in any way to be hounded or libelled.43 The king, at least, knew that, for the sake of the future, the past should be forgotten. Secondly, we may see this way forward as part of the reassertion of royal authority in France, an authority reflected in the growing effectiveness of the crown which, by 1453, had been re-emerging for a decade or more. In a country so recently divided, the monarchy was the only institution which could unite and give it a lead. It needed peace both to succeed and to bring it the benefits which would follow from that peace. Only the monarchy could successfully bring about the reconstruction of France, keep the nobility in its place, encourage (where necessary) the repopulation of town and country, and stimulate trade and economic regeneration. In order to achieve all – or any – of these, the king had to demand that there be peace within his kingdom. What may surprise is the apparent inconsistency of the law as a means of achieving that peace. The Edict of Compiègne, for example, represented a “hardline” and partisan solution to the problems raised by returning populations expecting to be rewarded for their loyalty to the Valois cause. In practice, politics (in particular when it came to the reality of the bargaining which took place at city gates)44 dictated that concessions should be made to further the process of reconquest; the talking and bargaining between individuals who had legal cases to pursue could take place later. Generally, a pragmatic approach to practical problems was often more telling than a rigid application of a legal text. However, two factors certainly helped to resolve a potentially divisive situation. First, not all who had a claim to press wished to make much of it in reality, in particular if

40

41 42

43 44

“Ever since France first heard the names of ‘Burgundian’ and ‘Armagnac,’ every crime that can be thought of or spoken of has been done in the kingdom of France, so that innocent blood cries for vengeance before God.” (Parisian Journal, p.146). Ibid., p. 54, n. 1. Bossuat, “Re-establishment of peace,” pp. 78, 365, nn. 42–3; P. S. Lewis, Later Medieval France. The Polity (London and New York, 1968), p. 75, for an example of the name “Armagnac” being used as a term of abuse. On the other hand, it is striking how the anonymous Parisian, normally no lover of the Armagnacs or of their treatment of his city and its people (see Parisian Journal, p. 54) refers no more to them after 1436, significantly returning to the term “Français” (“Francoys”) which he had used in the period up to 1417, before the major divisions which rent the French nation became too apparent. The use of “Français” was now intended to indicate supporters of Charles VII as king of France. “ … en quelque manière calumpniez.” (Ordonnances, 14:77). The terms for the surrender of Lisieux were agreed “devant la porte de ladicte ville de Lisieux.” (Ibid., 14:64).

164

Christopher Allmand

difficulties arose, or economic reasons did not make it worth their while spending time and money in the courts. Secondly, the spirit of compromise, preached by (among others) Jean Juvénal des Ursins, and formalised in out-of-court settlements, or accords, negotiated by persons acting as arbitrators, must often have taken the bitterness out of what might have been acrimonious conflicts at law. Society, as well as individuals, benefited from such procedures. In 1444 Rolando Talenti, the Milanese humanist living in Bayeux in the service of Zenon da Castiglione, the town’s Italian bishop, wrote that the peacemaking of the time must not be held up by the “hatred and injuries of past times,” nor by the “long-standing harshness of mind” of “obstinate and rebellious spirits” of certain individuals.45 Old hatreds, he was suggesting, existed; yet, over all, our sources tell us little about them.46 Sectional interests do not appear to have delayed the advances made towards unity by the kingdom of France in the second half of the fifteenth century. It was for the country as a whole that Charles VII ordered Masses of thanksgiving to be said annually on 12 August (the day of the capture of Cherbourg from the English).47 As Charles of Orléans expressed it in the oration delivered in favor of the duke of Alençon in 1458, “for a long time past none of your predecessors has held the kingdom so united in his grasp as you hold it now.”48 Spoken as part of a plea that mercy be shown to a man on trial for his life, these words were a recognition that reconciliation, although it had met challenges, had largely been successful.

45 46

47 48

“odia et veteres iniurias … inveteratam animorum duriciam … obduratos et efferatos animos.” (Bayeux, Bibliothèque Capitulaire, MS. 5, fol. 12). See Bossuat, “Re-establishment of peace,” pp. 70, 364, n. 29: Lewis, Later Medieval France, pp. 75–6; Thompson, Paris and its People, pp. 236–9; Hubrecht, “Jurisdictions and Competences,” p. 91. Allmand, Lancastrian Normandy, pp. 304–5. “ … car, passé long temps, nulz de vos prédecesseurs n’ont eu le royaume si entier en leurs mains comme vous l’avez.” (Champion, Charles d’Orléans, p. 543).